Handbook on Sentencing Policies and Practices in the 21st Century 2019011041, 2019012916, 9780429027765, 9780367136499

Citation preview

HANDBOOK ON SENTENCING POLICIES AND PRACTICES IN THE 21ST CENTURY

The Handbook on Sentencing Policies and Practices in the 21st Century focuses on the evolution and consequences of sentencing policies and practices, with sentencing broadly defined to include plea bargaining, judicial and juror decision-​making, and alternatives to incarceration, including participation in problem-​solving courts. This collection of essays and reports of original research explores how sentencing policies and practices, both in the United States and internationally, have evolved, explores important issues raised by guideline and non-​guideline sentencing, and provides an overview of recent research on plea bargaining in the United States, Australia, and the United Kingdom. Other topics include the role of criminal history in sentencing, the past and future of capital punishment, strategies for reducing mass incarceration, problem-​solving courts, and restorative justice practices. Each chapter summarizes what is known, identifies the gaps in the research, and discusses the theoretical, empirical, and policy implications of the research findings. The volume is grounded in current knowledge about the specific topics, but also presents new material that reflects the thinking of the leading minds in the field and that outlines a research agenda for the future. This is Volume 4 of the American Society of Criminology’s Division on Corrections and Sentencing handbook series. Previous volumes focused on risk assessment, disparities in punishment, and the consequences of punishment decisions. The handbooks provide a comprehensive overview of these topics for scholars, students, practitioners, and policymakers. Cassia Spohn is a Foundation Professor and Director of the School of Criminology and Criminal Justice at Arizona State University. She is a Fellow of the American Society of Criminology, a Fellow of the Academy of Criminal Justice Sciences, and a Fellow of the Western Society of Criminology. Her research interests include prosecutorial and judicial decision-​making, the intersections of race, ethnicity, gender, crime and justice, and sexual assault case-​processing decisions. Pauline K. Brennan is a Professor and the Ph.D. Program Director for the School of Criminology and Justice Studies at the University of Massachusetts Lowell. Her research focuses on racial and gender inequities in court processing, corrections policy, and issues related to female offenders and victims. Her work has appeared in the top journals in the field, including Justice Quarterly, The Journal of Quantitative Criminology, and Criminal Justice and Behavior.

THE ASC DIVISION ON CORRECTIONS & SENTENCING HANDBOOK SERIES Edited by Pamela K. Lattimore and John R. Hepburn.

The American Society of Criminology’s Division on Corrections & Sentencing sponsors a series of volumes published by Routledge on seminal and topical issues that span the fields of sentencing and corrections. The critical essays, reviews, and original research in each volume provide a comprehensive assessment of the current state of knowledge, contribute to public policy discussions, and identify future research directions. Each thematic volume focuses on a single topical issue that intersects with corrections and sentencing research. The contents are eclectic in regard to disciplinary foci, theoretical frameworks and perspectives, and research methodologies. EDITORIAL BOARD Gaylene Armstrong, University of Nebraska Omaha Todd Clear, Rutgers University Francis T. Cullen, University of Cincinnati Jodi Lane, University of Florida Dan Mears, Florida State University Joan Petersilia, Stanford University Cassia Spohn, Arizona State University Jeffery Ulmer, Pennsylvania State University Steve Van Dine, Ohio Department of Rehabilitation and Correction Christy Visher, University of Delaware HANDBOOK ON RISK AND NEED ASSESSMENT Theory and Practice Taxman HANDBOOK ON PUNISHMENT DECISIONS Locations of Disparity Ulmer and Bradley HANDBOOK ON THE CONSEQUENCES OF SENTENCING AND PUNISHMENT DECISIONS Huebner and Frost HANDBOOK ON SENTENCING POLICIES AND PRACTICES IN THE 21ST CENTURY Spohn and Brennan

HANDBOOK ON SENTENCING POLICIES AND PRACTICES IN THE 21ST CENTURY

Edited by Cassia Spohn and Pauline K. Brennan

First published 2020 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2020 Taylor & Francis The right of Cassia Spohn and Pauline K. Brennan to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-​in-​Publication Data Names: Spohn, Cassia, editor. | Brennan, Pauline Katherine, 1967– editor. Title: Handbook on sentencing policies and practices in the 21st century / Cassia Spohn & Pauline K. Brennan. Description: Abingdon, Oxon ; New York, NY : Routledge, 2019. | Series: American Society of Criminology§s Division on Corrections and Sentencing handbook series ; Volume 4 Identifiers: LCCN 2019011041 (print) | LCCN 2019012916 (ebook) | ISBN 9780429027765 (Ebook) | ISBN 9780367136499 (hardback) Subjects: LCSH: Sentences (Criminal procedure)–English-speaking countries. Classification: LCC K5121 (ebook) | LCC K5121 .H36 2019 (print) | DDC 345/.0772–dc23 LC record available at https://lccn.loc.gov/2019011041 ISBN: 978-​0-​367-​13649-​9  (hbk) ISBN: 978-​0-​429-​02776-​5  (ebk) Typeset in Bembo by Newgen Publishing UK

We dedicate this book to Chet Britt, Marie Griffin, and Ben Steiner, exemplary scholars within the Division on Corrections and Sentencing. You are deeply missed by your friends and colleagues. Chester L. Britt, III (1962–​2016) Professor, Iowa State University Marie Griffin (1967–​2016) Professor, Arizona State University Benjamin M. Steiner (1975–​2019) Professor, University of Nebraska Omaha

CONTENTS

Contributors 

x



1

Sentencing Policies and Practices in the 21st Century: An Introduction  Cassia Spohn and Pauline K. Brennan

PART I

The Evolution of Sentencing Policies and Practices 

17

1 The Transformation of Sentencing in the 21st Century  Megan C. Kurlychek and John H. Kramer

19

2 Sentencing Guidelines in the United States  Richard S. Frase and Kelly Lyn Mitchell

43

3 Sentencing Guidelines Outside the United States  Julian V. Roberts and Lyndon Harris

68

PART II

Issues in Guideline and Non-​Guideline Sentencing  4 Inter-​district Differences and Extra-​legal Disparity in Federal Sentencing: The Trees Are “Substantially” More Important Than the Forest  Richard D. Hartley

vii

87

89

Contents

 5 Minimum Sentencing for Serious Offenses: Lessons from Australia  Kate Fitz-​Gibbon and James Roffee  6 Discretionary Release Practices for Juveniles Facing Life: A Review of State Parole and Resentencing Procedures  Stuti S. Kokkalera and Simon I. Singer

114

129

PART III

Plea Bargaining 

151

 7 Plea Negotiations: An Australian Perspective  Arie Freiberg and Asher Flynn

153

 8 Plea Bargaining in the Shadow of the Trial  Amy Dezember and Allison D. Redlich

168

 9 Estimating the Size of Plea Discounts: Why Does It Matter?  Shi Yan

188

10 To Plead or Not to Plead? “Guilt” is the Question: Rethinking Sentencing and Plea Decision-​Making in Anglo-​American Countries  Jay Gormley and Cyrus Tata

208

PART IV

Capital Punishment 

235

11 Evolving Attitudes toward Capital Punishment  Amy L. Anderson,Weng-​Fong Chao, and Philip Schwadel

237

12 Disparities in Death Penalty Prosecution and Punishment: A Review of Recent Research and an Expanded Agenda  Jeffery T. Ulmer and Lily Hanrath

254

PART V

Current Controversies 

281

13 Rethinking the Role of Criminal History in Sentencing  Rhys Hester, Richard S. Frase, Julia Laskorunsky, and Kelly Lyn Mitchell

283

14 AB109 in California: Realignment, Decarceration and Crime in Los Angeles County  Katharine Tellis and Cassia Spohn

viii

312

Contents

15 The Problem with Problem-​solving Courts: The Black Box Remains Unopened after Thirty Years  Eileen M. Ahlin and Anne S. Douds

339

16 Restorative Justice Practices and Challenges in the United States  Jennifer L. Lanterman

360

Index 

378

ix

CONTRIBUTORS

Dr. Eileen M. Ahlin is Assistant Professor of Criminal Justice in the School of Public Affairs at Penn State Harrisburg. Her research focuses primarily on the use of formal and informal social controls across the ecological framework to prevent and reduce criminal behavior. Her research appears in the journals such as Youth Violence and Juvenile Justice, Journal of Interpersonal Violence, Journal of Youth and Adolescence, Aggression and Violent Behavior, Journal of Criminal Justice, and American Journal of Public Health. She is a co-​recipient of the 2016 W. E. B. Du Bois Fellowship from the National Institute of Justice, for research examining the role of immigrant generation status and race/​ethnicity on the effects of inner-​city mobility on youth violence and exposure to violence. Amy L. Anderson is Professor of Criminology and Criminal Justice at the University of Nebraska Omaha. Her research examines the relationship between social contexts and delinquency and public perceptions of criminal justice policies. Professor Anderson’s research has been published in Criminology, Justice Quarterly, Journal of Research in Crime and Delinquency, Journal of Developmental and Life-​Course Criminology, and other peer-​reviewed journals. Pauline K. Brennan is Professor and the Ph.D. Program Director for the School of Criminology and Justice Studies at the University of Massachusetts Lowell. Her research focuses on racial and gender inequities in court processing, corrections policy, and issues related to female offenders and victims. She has been the principal investigator on projects funded by the National Institute of Corrections and the Foundation for the Carolinas, and has worked closely on policy-​relevant research with community practitioners in North Carolina, Nebraska, and Iowa. Her work has appeared in the top journals in the field, including Justice Quarterly, The Journal of Quantitative Criminology, and Criminal Justice and Behavior. Weng-​Fong Chao is currently pursuing his Ph.D.  in Criminology and Criminal Justice at the University of Nebraska Omaha. His current research interests include criminological theories, public opinion about crime and justice, and research methods and statistics. Amy Dezember is a doctoral candidate in Criminology, Law and Society at George Mason University. Her research interests include plea bargaining, Alford pleas, legal decision-​making, and

x

Contributors

the role of evidence in the decision-​making process. Her dissertation research examines Alford pleas and the presumption of strong evidence, which includes examining court records and conducting interviews with prosecutors, judges, and defense attorneys to learn more about the process for offering, negotiating, and accepting these unique pleas. In addition to her dissertation research, she currently works as a Graduate Research Assistant on a National Science Foundation and National Institute of Justice funded-​project looking at youth and adult engagement in the guilty plea process. She has been published in Psychology, Public Policy, and Law and has written numerous book chapters about misdemeanor guilty pleas and the use of risk assessments in the criminal justice system. Anne S. Douds, JD, Ph.D. is Assistant Professor of Public Policy at Gettysburg College and the faculty lead for the Eisenhower Institute’s Women and Leadership Program. A retired trial attorney with more than 20  years of litigation experience, Dr.  Douds’ research centers on court program evaluations, judicial and legal policy analysis, and social justice issues. Dr. Kate Fitz-​Gibbon is Senior Lecturer in Criminology at Monash University,Victoria, Australia, and an Honorary Research Fellow in the School of Law and Social Justice at University of Liverpool (UK). She is a lead researcher in the Monash Centre for Gender and Family Violence Prevention. Dr. Fitz-Gibbon conducts research in the area of family violence, legal responses to lethal violence, youth justice and the effects of homicide law and sentencing reform in Australian and international jurisdictions. Dr. Fitz-Gibbon has served as an advisor on homicide reform law, family violence, youth justice, and sentencing law reviews in several Australian and international jurisdictions. Dr.  Asher Flynn is Associate Professor of Criminology, and Director of the Social and Political Sciences Graduate Research Program at Monash University, Australia. Dr Flynn’s research utilizes a sociolegal framework to understand, critique and transform legal policy and practice, with a particular focus on gendered and technology-​facilitated violence. Informed by national and international context, her research examines experiences of accessing and negotiating justice. She has published widely on technology-​facilitated violence and harassment, with a specific focus on image-​based sexual abuse, as well as sexual and gender-​based violence, plea negotiations, and access to justice. Dr. Flynn is Chief Investigator on an Australian Research Council Discovery Project and recently completed two Australian Criminology Research Council-​funded projects. Richard S. Frase is the Benjamin Berger Professor of Criminal Law at the University of Minnesota Law School and Co-​Director of the Law School’s Robina Institute of Criminal Law and Criminal Justice. His principal research interests pertain to Minnesota’s and other state sentencing guidelines, purposes and limits of punishment, comparative criminal procedure, and comparative sentencing within the US and between the US and other Western nations. He teaches criminal law, criminal procedure, and the federal defense clinic, and has previously taught prosecution clinics operating in state and federal courts. His seminars include sentencing guidelines, sentencing policy, and comparative criminal procedure. After earning a B.A. in Psychology from Haverford College and a J.D. from the University of Chicago Law School, he practiced law in Chicago before joining the Minnesota faculty. He is the author or co-​editor of ten books and over 95 articles, chapters, and essays on criminal justice topics. Arie Freiberg is Emeritus Professor of Law, Monash University, Victoria, Australia. He is the chair of the Victorian and Tasmanian Sentencing Advisory Councils. He has written extensively on sentencing, non-​adversarial justice and regulatory theory.

xi

Contributors

Jay Gormley, LLB Hons (First Class), MPhil (Law) has submitted his Ph.D. thesis at the Strathclyde University Law School. His thesis interrogated the nature and extent of sentence discounting for guilty pleas in Scotland and involved empirical research with defense lawyers, prosecutors, judges, and defendants. The Strathclyde Excellence Scholarship provided funding for Dr. Gormley’s doctoral education. In 2017 he was also awarded a doctoral internship with the Scottish Prison Service sponsored by the UK Economic & Social Research Council’s Scottish Graduate School of Social Sciences. He has a keen interest in criminal justice, processual legal justice, and is a member of the ESC Working Group on Sentencing and Penal-​Decision Making. He teaches on the LLB modules in Legal Process and in Law and Society. Lily Hanrath is a Criminology Ph.D. Candidate at Penn State University. Her work focuses on courts and sentencing, and morality and deviance. Lyndon Harris is currently reading for a DPhil in criminology at the University of Oxford. He is a Lecturer in Criminal Law at Keble College, University of Oxford, and is a criminal barrister. He is the author of Thomas’ Sentencing Referencer, editor of Current Sentencing Practice, and a contributing editor to Archbold, Criminal Pleading, Evidence and Practice. Richard D. Hartley, Ph.D., is Professor in the Department of Criminal Justice at the University of Texas, San Antonio. His research examines decision-​ making practices surrounding criminal court processes and has been funded by the Bureau of Justice Statistics and the American Statistical Association, as well as the National Institute of Justice (NIJ). His most recent NIJ funded projects include a multi-​site evaluation of veterans treatment courts, and a study examining the correlates of federal prosecutorial charging decisions. His currently published research appears in Justice Quarterly, Journal of Criminal Justice, and Criminal Justice Policy Review. Rhys Hester is Assistant Professor in the Department of Sociology, Anthropology & Criminal Justice at Clemson University, South Carolina. His research has been published in Criminology, Journal of Quantitative Criminology, and Crime and Justice: A Review of Research. He holds a J.D. and a Ph.D. and previously served as a research fellow at the University of Minnesota Law School’s Robina Institute of Criminal Law and Criminal Justice, and was also Deputy Director of the Pennsylvania Commission on Sentencing. Stuti S. Kokkalera is a doctoral candidate in the School of Criminology and Criminal Justice at Northeastern University. Her current research focuses on state discretionary release practices for juveniles sentenced to life. Her articles have been published in Societies, Journal of Youth and Adolescence, and The Criminologist. John H.  Kramer is Professor Emeritus of Sociology and Criminology at the Pennsylvania State University where recently the John Kramer Professorship in Criminology was established. He taught at Mankato State University from 1970–​1973 and at Penn State from 1973 until retirement in 2015. His career included serving as Executive Director of the Pennsylvania Commission on Sentencing from 1979–​1998 and as Staff Director of the United States Sentencing Commission from 1996–​1998. Dr.  Kramer’s research focuses on sentencing, sentencing disparity, and sentencing reform and his research has been published in major journals in the field including Criminology. His book, Sentencing Guidelines: Lessons from Pennsylvania, co-​authored with Dr. Jeff Ulmer, in 2009, examined the history and impact of the Pennsylvania sentencing guidelines. He recently completed a study of the death penalty in Pennsylvania with Drs. Jeff Ulmer and Gary Zajak. Megan C. Kurlychek is Associate Professor in the School of Criminal Justice at the University at Albany. She also serves as Executive Director of the New York State Youth Justice Institute and Editor xii

Contributors

of Justice Quarterly. Her primary research interests encompass the courts and sentencing, the juvenile justice system and juvenile delinquency, and the lasting impact of a juvenile or criminal record on later life outcomes. Her work has been published in the field’s top journals, including Criminology, Criminology and Public Policy, Justice Quarterly, and Punishment and Society. Jennifer L. Lanterman is Assistant Professor in the Department of Criminal Justice at the University of Nevada, Reno. Her research focuses on the management and treatment of high-​risk and high-​ need offenders, cultural aspects of firearm violence, and restorative practices. Her recent work has appeared in Criminal Justice and Behavior, Criminology and Public Policy, Dialogues in Social Justice, Feminist Criminology, Journal of Qualitative Criminal Justice and Criminology, Justice Research and Policy, Studies in Social Justice, and The Prison Journal. Julia Laskorunsky is Research Fellow at the Robina Institute of Criminal Law and Criminal Justice. She is working on the Criminal History Enhancements Project and the Parole Release and Revocation Project. She received her Ph.D. in Criminology from Pennsylvania State University. Her dissertation focused on the use of actuarial risk assessments at sentencing and was funded by the National Institute of Justice Dissertation Fellowship. Dr. Laskorunsky’s research focuses on the ways sentencing structures and correctional practices affect incarceration rates and racial disparity in the criminal justice system. Her work can be seen in the Journal of Crime and Justice, Advancing Criminology and Criminal Justice Policy, and Oxford Handbooks Online. She was formerly a Deputy Project Director and Research Assistant for Development Services Group, Inc., where she worked on multiple projects for the Office of Juvenile Justice and Delinquency Prevention and the National Institute of Justice. Kelly Lyn Mitchell is the Executive Director of the Robina Institute of Criminal Law and Criminal Justice, at the University of Minnesota Law School. She was the Executive Director of the Minnesota Sentencing Guidelines Commission from 2011 to 2014, and served as the President of the National Association of Sentencing Commissions from 2014–​2017. Prior to joining the Robina Institute, she worked at the Minnesota Judicial Branch from 2001–​2011 as a staff attorney and manager, where she served as the Branch’s liaison to other criminal justice agencies and was responsible for several statewide programs and services such as drug courts, the court interpreter program, and examiner services for sex offender civil commitment exams. She also provided legal support to trial court judges and court administrators on issues ranging from criminal and juvenile delinquency law to court records access and fines and fees in the criminal justice system. She also provided legal support for several Minnesota Supreme Court rules and policy committees, and in this role led efforts to fully revise the Minnesota Rules of Criminal Procedure and the Minnesota Juvenile Delinquency Rules of Procedure. Over the course of her career, Ms. Mitchell has held numerous appointments on committees and task forces on issues such as prison population control, probation supervision, sex offender management, and collateral consequences. She earned her J.D. from the University of North Dakota Law School, and has a Master’s in Public Policy from the University of Minnesota’s Humphrey School of Public Affairs. Allison D. Redlich is Professor of Criminology, Law and Society at George Mason University, and serves as the department’s Associate Chair and Graduate Director. She was trained as an experimental psychologist but uses multiple methods to conduct her research. To a large degree, her research centers on whether legal decision-​making is knowing, intelligent, and voluntary. She examines such decision-​making in vulnerable (juveniles and persons with mental health problems) and non-​ vulnerable defendants, and across several different contexts—​in the interrogation room, during the guilty plea process, and in mental health courts. Professor Redlich also studies wrongful convictions, with a particular focus on false confessions and false guilty pleas. In addition to publishing numerous xiii

Contributors

articles on these and related topics, she has co-​authored/​edited five books. To pursue her research, Professor Redlich has received funding from the National Science Foundation, the National Institute of Justice, the Brain and Behavior Research Association, and the Federal Bureau of Investigation, among others. Julian Roberts is Professor of Criminology in the Faculty of Law, University of Oxford. He was a member of the Sentencing Council of England and Wales from 2009–​2018. Dr. James Roffee is Associate Professor of Law, and Associate Dean International in the Faculty of Business and Law at Swinburne University of Technology. James leads the Faculty’s international portfolio; this includes strategic and operational responsibility for delivery of undergraduate and postgraduate transnational education, student mobility and exchange, bespoke programs for industry partners, and sponsored students. Prior to joining Swinburne in 2018, James received the 2017 Vice-​ Chancellor’s Award for Teaching Excellence from Monash University where he held positions in the Faculty of Business and Economics and the Faculty of Arts. He fosters an inclusive approach to teaching and is recognized as a leader and innovator within the field of inclusive higher education. James’s research in criminal justice and criminology concerns victimization, hate-​and bias-​motivated crime, and serious and violent offenses. James was appointed to the Victorian Government’s LGBTI Taskforce Justice Working Group in 2017. Philip Schwadel is Professor of Sociology at the University of Nebraska-​Lincoln. His research examines the intersections of religion, politics, social status, and social change. Recent research focuses on how social contexts influence the effects of religion and politics on public opinion. Professor Schwadel’s research has been published in Social Forces, Social Science Research, Journal for the Scientific Study of Religion, Criminology, and other peer-​reviewed journals. Simon I.  Singer is Professor of Criminology and Criminal Justice at Northeastern University. His scholarship focuses on youth crime and justice. He has received awards from the American Sociological Association (Albert Reiss Book Award, 1999) for his book, Recriminalizing Delinquency, and from the American Society of Criminology (Hindelang Book Award, 2014) for his most recent book, America’s Safest City. His current research focuses on juveniles subject to long-​term adult maximum sentences. Cassia Spohn is Foundation Professor and Director of the School of Criminology and Criminal Justice at Arizona State University. She is a Fellow of the American Society of Criminology, a Fellow of the Academy of Criminal Justice Sciences, and a Fellow of the Western Society of Criminology. She is the recipient of numerous academic awards, including the University of Nebraska Outstanding Research and Creative Activity Award, the W. E. B. DuBois Award for Contributions to Research on Crime and Race/​Ethnicity, and Arizona State University’s Faculty Achievement Award for Defining-​ Edge Research in the Social Sciences. She is the author or co-​author of seven books, including The Color of Justice: Race, Ethnicity and Crime in America, Policing and Prosecuting Sexual Assault: Inside the Criminal Justice System, How Do Judges Decide? The Search for Fairness and Equity in Sentencing and Drugs, Crime and Justice. Her research interests include prosecutorial and judicial decision-​making, the intersections of race, ethnicity, gender, crime and justice, and sexual assault case-​processing decisions. She currently is a member of the Defense Advisory Committee on the Investigation, Prosecution and Defense of Sexual Assault in the Military. She also serves as a consultant to the MacArthur Foundation’s project on advancing prosecutorial effectiveness and fairness through data and innovation. In 2020 she will be the president of the Academy of Criminal Justice Sciences.

xiv

newgenprepdf

Contributors

Professor Cyrus Tata, Ph.D., FRSA is Professor of Law and Criminal Justice at Strathclyde University Law School, where he is Director of the Centre for Law, Crime and Justice, and Programme Director of the LLM/​MSc in Criminal Justice & Penal Change. For over 25 years he has conducted and published research into various aspects of criminal justice in Scotland and abroad, including: the impact of legal aid reforms; plea decision-​making and plea negotiation; lawyer-​client relations; the role of pre-​sentence reports, mitigation and individualization, sentencing; and the use of information technology (including, for example, the introduction of a Sentencing Information System for the High Court). He is founder and chair of the European Group on Sentencing and Penal Decision-​Making (a network of academic, policy, and practice members in over 25 countries). Regularly invited to speak to policy and practice audiences around the world, he has also served as adviser to governments in several countries and recently to the senior judiciary and court service of the Irish Republic and is currently assisting Northern Ireland’s Review of Sentencing Policy. He is currently collaborating in research into public attitudes to sentencing, including in specific areas such as causing death by driving offenses and sexual offenses. He is currently leading a small exploratory study examining the communication and interpretation of sentencing in the same cases from the perspectives of: the judicial sentencer, the defense lawyer, and the sentenced person. He is also currently working (with partners in Italy, Sweden, and the USA), on a European Research Council-​funded research study into cognition-​emotion processes in the criminal process. His new book, Sentencing as a Social Process: Re-​ Thinking Research & Policy, will be published in 2019. In 2017 he was invited and elected as a Fellow of the Royal Society of Arts. Dr.  Katharine Tellis is Associate Professor and the Director of the School of Criminal Justice and Criminalistics at California State University, Los Angeles. A social worker and criminologist by training, her areas of interest center on intersectionality, criminal justice policy, and violence prevention. She received a B.A. in Gender Studies and a Master’s in Social Welfare from UCLA, and a Ph.D. in Criminology and Criminal Justice from the University of Nebraska Omaha. Jeffery T. Ulmer is Professor of Sociology and Criminology at Penn State University, and also serves as Associate Department Head. His work spans topics such as courts and sentencing, criminological theory and symbolic interactionism, religion and crime, and violent crime. He received the 2001 Distinguished New Scholar Award and the 2012 Distinguished Scholar Award from the American Society of Criminology’s Division on Corrections and Sentencing. He and his coauthors won the American Society of Criminology’s 2012 Outstanding Article Award. He and Darrell Steffensmeier were also awarded the ASC’s 2006 Hindelang Award for Confessions of a Dying Thief: Understanding Criminal Careers and Illegal Enterprise (2005, Transaction). Shi Yan is Assistant Professor in the School of Criminology and Criminal Justice, Watts College of Public Service and Community Solutions, Arizona State University. His research interests include sentencing, plea bargaining, and criminal careers. His research has been published in Journal of Quantitative Criminology, Justice Quarterly, Crime & Delinquency, among others.

xv

SENTENCING POLICIES AND PRACTICES IN THE 21ST CENTURY An Introduction Cassia Spohn and Pauline K. Brennan

This is the fourth volume of the American Society of Criminology’s Division on Corrections & Sentencing (DCS) handbook series, which was established by the DCS to disseminate research to scholars, students, practitioners, and policymakers. Each thematic volume features state-​of-​the-​art work on seminal and emerging issues at the intersection of corrections and sentencing. The goals of each volume are to summarize the overarching themes revealed by research and scholarship, identify theoretical and methodological limitations of the extant research, and provide suggestions about directions for future research. Volume 1, edited by Faye Taxman, addressed the important and controversial issue of risk assessment for correctional populations, and Volume 2, edited by Jeffery Ulmer and Mindy Bradley, focused on disparities based on race, ethnicity, sex, and class in pretrial processes and sentencing decisions. In Volume 3, edited by Beth Huebner and Natasha Frost, contributors turned their attention to the consequences of punishment and sentencing decisions in an era of mass incarceration. The essays and empirical studies contained in these three volumes highlight punishment issues that evoke controversy and spark debate, illustrate the ways in which decision-​maker discretion produces disparities and differential treatment over the “life course” of a criminal case (Spohn, 2015: 76), and explore “the intended and unintended consequences of punishment and sentencing decisions that extend far beyond the confines of the justice system” (Huebner & Frost, 2018). In this volume our focus shifts to the evolution of sentencing policies and practices, both in the United States and elsewhere. Concerns about disparity, discrimination, and unfairness in sentencing led to a “remarkable burst of reform” (Walker, 1993: 112) that began in the mid-​1970s and continues today.The initial focus of reform efforts was the indeterminate sentence, in which the judge imposed a minimum and maximum sentence and the parole board determined the date of release. Both liberal and conservative reformers challenged the principles underlying the indeterminate sentence and called for changes designed to curb discretion, reduce disparity and discrimination, and achieve proportionality and parsimony in sentencing. This process of reconsideration and reform led many jurisdictions to adopt structured sentencing proposals designed to control the discretion of sentencing judges. A  number of jurisdictions adopted determinate sentencing policies that offered judges a limited number of sentencing options and included enhancements for use of a weapon, presence of a prior criminal record, or infliction of serious injury. Other jurisdictions adopted presumptive or voluntary sentence guidelines that incorporated crime seriousness and prior criminal record into a sentencing protocol that judges were to use in determining the appropriate sentence. Other reforms enacted in the United States and 1

Cassia Spohn and Pauline K. Brennan

(to a lesser extent) elsewhere included mandatory minimum penalties for certain types of offenses (especially drug and weapons offenses), habitual offender and “three-​strikes-​and-​you’re out” laws that mandated long prison sentences for repeat offenders, and truth-​in-​sentencing statutes that required offenders to serve a specified portion of the sentence before being released. This process of experimentation and reform revolutionized the sentencing process. Thirty years ago, many jurisdictions had an indeterminate sentencing system and “the word ‘sentencing’ generally signified a slightly mysterious process which … involved individualized decisions that judges were uniquely qualified to make” (Tonry, 1997: 3). The situation today is much more complex. Sentencing policies and practices vary enormously on a number of dimensions, and there is no longer anything that can be described as the modal approach. Volume 4 of the Handbook series addresses this evolution of the sentencing process. It includes contemporary essays, legal analyses, and empirical research on sentencing policies and practices in the 21st century, with sentencing broadly defined to include plea bargaining, judicial and juror decision-​ making, and alternatives to incarceration.The contributors to the volume tackle a diverse set of issues, including the development and refinement of sentencing guidelines and mandatory penalties in the United States, the United Kingdom, and Australia; the consequences of life sentences for juvenile offenders; plea bargaining practices and the factors that entice defendants to plead guilty; attitudes toward and disparities in capital punishment; the role that criminal history should play in calculations of sentence severity; and the development and implementation of alternatives to traditional adjudication and to incarceration. The insights and conclusions of the contributors to this volume—​ who represent both established and emerging scholars—​will shape the ongoing debate regarding the future evolution of the punishment process.

Chapters The first part of the handbook, “The Evolution of Sentencing Policies and Practices,” focuses on how sentencing practices have transformed over time and the reasons for changes in sentencing policy. In Chapter 1, Megan C. Kurlychek and John H. Kramer discuss the relationship between punishment ideology and sentencing policy, and highlight shifts in ideological thinking and policy over time. As they point out, a belief in the values of deterrence and proportionality dominated sentencing policy for the first 100 years of US history. A major shift in sentencing policy occurred by the end of the 1800s wherein the goal of rehabilitation via indeterminate prison sentences was emphasized. But, by the mid-​1970s sentencing policy predicated on a belief in rehabilitation came under attack, due (partly) to both a lack of empirical evidence on the utility of treatment as a crime control strategy and rising crime rates. The crime problem also became politicized, as candidates for political office increasingly declared “wars on crime,” and victims increasingly became more involved in the sentencing process (via the Victim Rights Movement and laws mandating victim restitution and participation), which served to inflate support for retribution.Therefore, sentencing policy based on notions of retribution, incapacitation, and deterrence became the underlying justifications for punishment by the 1980s, and by the 1990s all states and the federal government implemented more punitive laws (such as mandatory minimum sentencing laws, habitual offender sentence enhancements, three-​strikes legislation, truth-​in-​sentencing policies, determinate prison sentences, and sentencing guidelines). Such punitive policies resulted in surging US prison populations with unprecedented cost. But, recent winds of change suggest that modifications to sentencing policy in the 21st century have occurred and will continue. In particular, both liberals and conservatives have argued that mass incarceration is problematic, and that the resources allocated to corrections have not been wisely or effectively utilized. Thus, mandatory sentencing laws have been repealed or modified in several jurisdictions, correctional interventions are increasingly based on evidence of effectiveness (e.g., use of needs and responsivity assessment tools for probation supervision), the Supreme Court has restricted use of the death penalty and incarceration (for juveniles, in particular), and problem-​solving 2

Introduction

courts have been implemented across the country. Although Kurlycheck and Kramer believe that such changes provide some evidence of a return to a belief in the value of rehabilitation in sentencing policy, they caution that progress made in sentencing reform may be derailed by the current Trump administration, due to increased politicization of the crime problem and recent highly conservative appointments to the federal bench. In Chapter  2, Richard S.  Frase and Kelly Lyn Mitchell add to the discussion of sentencing guidelines systems in the United States, and note within the first paragraph of their chapter that “since the late 1970s, no competing sentencing reform model of comparable scope has been implemented, or even seriously proposed.” Their chapter begins with an overview of the purposes of guidelines reforms, including reduced sentencing disparity (via limitations on the discretion of judges and parole boards), improved sentencing policy (via sentencing commissions that develop, monitor, and evaluate guidelines), improved management of correctional resources (via projections of prison populations), encouraged use of intermediate sanctions (in some jurisdictions), and enhanced transparency in sentencing (via stipulated sentences that closely resemble the amount of time an offender will actually serve in prison). Although the stated goals of sentencing guidelines are largely consistent across jurisdictions, Frase and Mitchell stress that no two jurisdictions utilize the same guidelines system. Across jurisdictions, guidelines differ in terms of their scope (i.e., the types of cases/​ offenses included within the guidelines), format (e.g., a two-​dimensional grid specifying a recommended punishment versus a worksheet that computes a recommended sentence), number and/​or weight of important variables (including the types of variables used to ascertain offense severity and offender prior record), and the degree to which judges are required to follow stipulated punishments. Such variations are important to acknowledge and consider in studies of sentencing guidelines, because differences will limit the generalizability of research findings. In terms of research findings, Frase and Mitchell stress that we do not yet have a clear empirical understanding of whether sentencing guidelines provide a superior (or inferior) policy choice relative to other options such as indeterminate sentencing. Our theoretical understanding is also limited in that much of the work on sentencing outcomes is dominated by the focal concerns theory, wherein disparities in sentencing outcomes are believed to arise from differences in judicial assessments of offender blameworthiness, the extent to which society may be threatened if an offender were allowed to remain in the community, and the costs associated with imposing a given sanction on a given offender (e.g., added prison overcrowding, potential disruption to family ties). With regard to our limited empirical understanding of outcomes in jurisdictions with guidelines, the majority of research on sentencing guidelines is based on assessments of outcomes in federal courts (i.e., under US sentencing guidelines), but findings cannot necessarily be generalized to state guidelines systems due to significant differences between state and federal systems. Moreover, since state guidelines systems differ widely, one should also exercise caution in attempting to generalize findings across states. Concerns about generalizability aside, Frase and Mitchell then highlight areas in need of more research attention. The vast majority of the research on sentencing guidelines has modeled the variables that drive sentencing outcomes, with considerable attention placed on the effects of offender race/​ethnicity and gender after guidelines were implemented in a single jurisdiction. By comparison, there are very few studies of variation in sentencing under pre-​and post-​guidelines regimes. Studies of outcomes such as probation or intermediate sanctions are also rare, although guidelines-​based structures in many states allow these punishment options; most of the existing research examines the in-​out (i.e., whether offenders are sentenced to prison or not) or prison-​length outcomes. And, much of the research on state-​level outcomes has focused on what occurs in Pennsylvania or Florida; research in other states with sentencing guidelines is either lacking or non-​existent. Overall, Frase and Mitchell urge future researchers to conduct cross-​jurisdictional analyses involving multiple distinct guidelines systems and of non-​prison outcomes. More would also be learned if researchers examined opinions about guidelines held by the public and by criminal justice administrators. 3

Cassia Spohn and Pauline K. Brennan

With regard to opinions about guidelines, in Chapter  3, Julian V.  Roberts and Lyndon Harris acknowledge that sentencing guidelines have not been a “popular penal export” to Canada, Australia, Scotland, South Africa, Ireland, India, and other countries, despite the reality that disparity or discrimination in sentencing is not a uniquely American problem. While many countries have rejected sentencing guidelines altogether, a guidelines-​based system has been adopted in England and Wales. Much of Chapter 3 provides an explanation of the evolution of the English system and important lessons we may extract from the process of successful implementation outside the United States. In particular, the English judiciary came to accept sentencing guidelines because judges played a primary role in their creation and because the guidelines system evolved slowly. Initial efforts began in the 1970s, but a true guidelines model did not emerge until 2003. Interestingly, it will take until 2020 for the English Sentencing Council to issue guidelines for all principal offense categories. As a point of contrast, Roberts and Lyndon argued that sentencing guidelines failed in Canada because judges were not directly consulted; judges resisted guidelines they perceived as the work of bureaucrats and academics. The English guidelines would not have been adopted if Parliament had created a sentencing authority modeled after the Minnesota Sentencing Commission. Because the English judiciary was largely responsible for crafting sentencing guidelines, judges across England and Wales were confident that the guidelines were not intended to restrict judicial discretion because they were developed by those who understood the task of sentencing. In addition, any modification to the guidelines would take time. The English Sentencing Council is not required or authorized to consider the size of the prison population in amending its guidelines. The Sentencing Commission in Minnesota, by comparison, may amend prescribed sentences relatively quickly in an effort to reduce prison admissions. English sentencing guidelines also differ from Minnesota’s in that they provide separate and individualized guidelines for different offenses (versus grids that include all offenses). Sentencing judges in England and Wales must follow nine steps in determining an appropriate punishment, with focus placed on assessments of harm and culpability (rather than on crime seriousness and criminal history). With regard to prescribed sentences, English judges are also required to consider whether aggravating or mitigating circumstances ought to “cross the custody threshold in one direction or the other.” This point is particularly interesting because it indicates that both community and custodial alternatives are allowable by law for many offenses. English guidelines further specify that defendants who enter guilty pleas earlier in case processing should receive greater sentencing discounts than those who plead guilty later. In the US, by contrast, the extent to which defendants will benefit from pleading guilty is left to the discretion of the sentencing judge. Overall, Roberts and Lyndon maintain that “English guidelines are more demanding of judges, less prescriptive, and less restrictive than their Minnesota equivalent.” While the English guidelines system differs in notable ways from Minnesota’s, is the English system better or worse? The answer is unknown because sentencing guidelines in England and Wales have not yet been comprehensively evaluated. This is an area of research worth pursuing. Part II of the handbook examines “Issues in Guideline and Non-​Guideline Sentencing.” In Chapter  4, Richard D.  Hartley addresses the question of whether federal sentencing guidelines are uniformly and consistently applied in federal sentencing via an examination of the extent to which prosecutors’ decisions to grant substantial assistance departures (i.e., 5K1.1 departures) to drug offenders across federal courts may result in inter-​district and extra-​legal disparity in federal courts. His focus on drug offenders is justified in that approximately 60% of the substantial assistance departures granted over a recent five-​year period were for defendants with drug charges, and the use of such departures is often intended to mitigate sentences for those facing mandatory minimum prison sentences. However, the frequency with which 5K1.1 departures are used varies widely across federal district courts. Hartley, therefore, conducted aggregated analyses (where data from 45 federal districts were combined) and disaggregated analyses (where data were partitioned by district) of whether narcotics offenders received substantial assistance departures. At the aggregate level, substantial assistance 4

Introduction

departures were affected by both legally relevant case characteristics and legally irrelevant offender characteristics. With regard to the extra-​legal variables, departures were less likely for minority offenders (black and Hispanic), males, and those who had not graduated from high school. Hartley then disaggregated the data by district and found that his analyses of aggregated sentencing data may have somewhat overstated the significance of extra-​legal variables across federal districts; an offender’s gender, race, ethnicity, and education influenced the decision in some districts but not in others. Females were more likely to be granted substantial assistance departures in 12 districts, but offender sex did not matter in 33 districts. Relative to whites, blacks were less likely to get a substantial assistance departure in seven districts and these departures also were less likely for Hispanics. But, offender race and ethnicity did not predict the likelihood of a departure in 37 districts. Hartley stressed that males and minorities in some federal districts had vastly different probabilities of having their sentences reduced via a substantial assistance departure and, consequently, ended up with sentences that differed notably from those given to white and female drug offenders. Overall, even under sentencing guidelines that exist in the 21st century, localized sentencing practices (which may arise from decisions made by prosecutors) may result in considerable sentencing variation for legally similar offenders. Sentencing policies that mandate at least a minimum amount of incarceration also exist in the United States and in other countries, but approaches and support for these policies are not universal. In Chapter 5, Kate Fitz-​Gibbon and James Roffee provide a cautionary tale of mandatory minimum sentencing laws adopted for serious offenders during the first two decades of the 21st century in two Australian states. In 2003 New South Wales (NSW) became the first Australian jurisdiction to introduce “a presumptive minimum sentencing scheme,” called the Standard Non-​Parole Period Sentencing Scheme (SNPP). It was designed to enhance punishment severity for offenders convicted of 35 serious and violent offenses (such as murder, serious drug offenses, person offenses, sexual offenses, and firearms offenses).The legislation required judges, for example, to sentence those convicted of killing a child to 25 years in prison and those convicted of manufacturing prohibited drugs to 10  years in prison. Critics immediately argued that the law was arbitrary, punitive, and complicated, and researchers found that prison sentences lengthened soon after the law went into effect. By 2011, a Law Reform Commission reviewed the SNNP law at the behest of the NSW Attorney General and suggested that complete repeal of the law be considered. Legal challenges soon followed and culminated with the High Court of Australia ruling that the SNPP law should never have included language that “implied a mandatory application”; the law should never have limited what judges were otherwise required or permitted to take into account when imposing punishment. Another highly controversial law was passed inVictoria in 2014—​the baseline sentencing scheme—​ after being hastily conceived by politicians who used high-​profile incidents of serious crime to raise the alarm over inadequate sentencing practices. This law increased punishment severity by creating baseline minimum sentences for a range of offenses, and abolished suspended sentences and home detention. The law stipulated that baseline medians for prison sentences be greatly increased. For example, the law provided for prison sentences that were 7½ years longer for drug traffickers and six years longer for those convicted of incest. Judges were instructed to sentence half the offenders above the baseline minimum and half below (based on consideration of mitigating and aggravating circumstances) and to provide their written reasons for the sentences they imposed. Criticism of the law was widespread and came in the form of newspaper editorials, media releases from the Victorian Bar and the Law Institute of Victoria, and negative commentary from judges and prosecutors. Critics argued the law would unduly limit judicial discretion, overcomplicate the sentencing process, increase pressure on the criminal justice system because prison sentences would lengthen, do little to deter or rehabilitate offenders, and overburden the courts with more trials and appeals. About a year after the law went into effect, the Victoria Court of Appeal ruled that the baseline sentencing scheme was “inoperative” and “incurable” and, thereby, nullified the law. 5

Cassia Spohn and Pauline K. Brennan

The overarching lesson from Chapter 5 is that any legislative attempt in Australia to reduce judicial discretion, such as via a law directing the judiciary to impose denoted minimum terms of imprisonment for serious offenders, will be met with strong resistance by the legal community. Therefore, in Australia and elsewhere, in moving forward legislators must work in tandem with officials from the court and criminal justice system to devise effective sentencing reform (echoing Roberts and Lyndon’s argument in Chapter 3). In Chapter 6, Stuti S. Kokkalera and Simon I. Singer provide a review of discretionary release practices for juveniles facing life sentences.They begin their chapter by discussing the rise and decline of the rehabilitative model for juveniles, and the harsh legislation that followed (see also Chapter 1). In particular, the automatic waiver or transfer to adult court of juveniles who committed serious crimes became increasingly common from the 1980s onward. Upon being transferred to the adult court, a juvenile could receive a sentence of life without any possibility of parole (LWOP). Prior to 1980, when laws requiring automatic waivers were uncommon, judges rarely sentenced youth to LWOP; but, by 2017, there were over 3,000 juveniles serving LWOP sentences across the country. And, such sentences were more often imposed on black and Hispanic youth than white youth, which raises the issue of equal protection under the law. Over time, the US Supreme Court heard cases related to the sentencing of juvenile offenders and made several important rulings. In particular, the Court referred to developmental literature when issuing its ruling in a death penalty case and stressed that children differ from adults, which mitigates their culpability and, therefore, affects the type of sentence that may be imposed.The ruling prohibited the execution of anyone who had committed their offense under the age of 18 years old. Seven years later, in 2012, the Court ruled that all legislated LWOP sentences for juveniles were unconstitutional, although states are still allowed to mandate lengthy prison sentences and judges are still allowed to impose consecutive sentences that may result in de facto life sentences. As a result of the 2012 ruling, all juveniles serving mandatory life without parole sentences became entitled to an opportunity to gain early release from prison via a review process. Review occurs via a discretionary parole process in some states or a resentencing process in other states. Regardless of whether a parole board or a judge is involved, the Supreme Court stipulated that a review should take into account the fact that juvenile lifers were adolescents when they committed their offenses, and release decisions should consider demonstrated growth and maturity in a determination of future risk of reoffending.The Court went further and suggested several factors that ought to be considered at the time of review, including a juvenile’s role in the crime, home environment, peer pressure, and rehabilitation in prison. Although these considerations may seem straightforward, there are several issues that become apparent upon closer inspection. For example, the Court left it to states to determine when a juvenile sentenced to LWOP would become eligible for a parole or resentencing hearing. Second, for those who may gain early release, the Court did not specify the duration or type of required community supervision. Third, the Court did not establish procedural standards for a parole board or judge to use when determining whether release should be granted. This is concerning in that researchers find that parole boards use inconsistent and generally unknown criteria when making their decisions. And, nearly all states consider victim input when making decisions, but victims may give little thought to an offender’s adolescence, growth, or maturity when recalling the victimization of a loved one. Kokkalera and Singer conclude their chapter by recommending that adolescence needs to be formally recognized by the criminal justice system, and calling for transparency in decision-​making and evaluation of the reentry needs of juvenile lifers. Part III of the handbook examines sentencing within the context of plea bargaining. In their introduction to Chapter 7, Arie Freiberg and Asher Flynn acknowledge that plea bargaining “is an accepted element of the criminal justice process in many common law jurisdictions and Australia is no exception.” Negotiated guilty pleas are purported to be advantageous for defendants in that they often reduce sentence severity. Case resolution via guilty pleas is also believed to benefit the public in that it saves the criminal justice system time and money, and relieves victims and witnesses of the 6

Introduction

burden of coming to court. But many argue that the plea negotiation process coerces defendants to plead guilty by threatening harsher punishment than would be meted out following a guilty verdict at trial. Defendants who lack attorneys to help them navigate the legal process and negotiate deals with prosecutors may be especially vulnerable. This last point must be emphasized because access to legal aid services in Victoria is extremely restrictive; in recent years, the number of unrepresented defendants has increased dramatically. Despite concerns about the plea negotiation process, regulation of the process is limited and generally consists of guidelines about when case resolution may occur, if and when approval from senior staff must be obtained, and record keeping. There are three primary forms of plea bargaining in Australia: charge bargaining (dropping or reducing charges), fact bargaining (watering down case facts), and sentence bargaining (wherein a prosecutor submits a recommendation for a less severe punishment to a judge). Other less common forms of plea negotiations were also noted, such as agreements where prosecutors decide not to proceed with the prosecution of a spouse or family member, promise financial support for children of the defendant, and agree not to pursue other charges or investigations against the defendants. By law, Australian defendants entering their guilty pleas earlier in the process are expected to received greater sentencing discounts (i.e., shorter terms of imprisonment, suspended sentences, or non-​custodial punishments) than those wait longer to plead guilty (see also Chapter 3). Interviews with prosecutors, defense attorneys, and judges indicated that between 90 to 100% of cases in Victoria were handled through plea negotiations. An empirical review of 50 case files indicated that charge reductions were most common, and the majority of cases resulted in the most serious charge being dropped. Drug and theft offenses were more likely to be negotiated than murder, sex offenses (especially familial sex offenses involving children), and family violence offenses. Although we now know a bit more about plea negotiations in Australia, Frieberg and Flynn conclude that “they ultimately remain a pragmatic form of somewhat mysterious justice in a largely unknown and imperfect criminal justice system.” In Chapter 8, Amy Dezember and Allison D. Redlich take a closer look at why prosecutors offer plea bargains and why defendants accept them. Consistent with figures reported in the proceeding chapter, between 97 to 99% of juvenile and adult convictions in the United States are the result of guilty pleas. And, based on research recently conducted in New  York City, adult and juvenile offenders who plead guilty received large sentencing discounts. Although the benefit of a guilty plea may seem obvious, there is limited research on why defendants decide to forgo trials and why prosecutors offer plea bargains. To provide more understanding of the pervasiveness of plea bargaining, Chapter 8 begins with a review of institutional explanations, with emphasis on a courtroom workgroup’s development of going rates for punishments (especially for defendants willing to plead guilty), desire to alleviate pressing caseloads, and critical focal concerns in the imposition of punishment (i.e., defendant culpability, community protection, and immediate and collateral costs linked to specific punishments). Under this theoretical approach, individuals who are convicted at trial (as opposed to those who plead guilty) get longer sentences. The “going to trial penalty” materializes because defendants who insist on trials add to caseload pressures, reduce system efficiency, and add to justice uncertainty. In contrast to an institutional explanation, which indicates that defendants are rewarded for pleading guilty by the courtroom workgroup, the shadow of the trial model starts from the perspective of the defendant. Defendants must decide whether they want a near-​certain outcome that comes from pleading guilty or an uncertain trial outcome. The decision is based on an assessment of the likelihood of conviction, which prosecutors also carefully consider. Assessments of conviction likelihood are largely predicated on evidentiary considerations. In cases where there is strong evidence and therefore a high likelihood of conviction, the plea discount will be smaller; larger discounts will arise in cases with greater conviction uncertainty. Few researchers, however, have tested the shadow model, and critics maintain that the model is too simplistic because plea negotiations are at least partially 7

Cassia Spohn and Pauline K. Brennan

influenced by factors other than the strength of evidence (such differences in attorney skill level, lack of understanding on the part of the defendant, and greater distrust in the system among minority defendants). Nonetheless, a notable body of research finds that the evidence against a defendant plays a role in the plea offers extended to defendants, although it is important to acknowledge that not all evidence is of the same quality or perceived utility. To complicate the issue further, tests of evidentiary strength are difficult because datasets often exclude measures or contain overly simplistic and subjective measures. Given limited research on the shadow model and concerns over the measurement of evidentiary strength in previous studies, Dezember and Redlich conducted a study wherein they analyzed how the strength of evidence influenced the sentencing discounts defendants received and whether defendants considered the evidence against them when pleading guilty. Chapter 8 presents findings from their preliminary analyses, which are based on data derived from observations of court hearings, examinations of court records, and interviews with defendants who pled guilty in a medium-​sized suburban county between February 2017 and March 2018. Results from a multivariate analysis indicated that the presence of strong evidence in a case did not influence the sentence discount received by a defendant who pled guilty. In addition, many defendants reported pleading guilty even when they thought the evidence against them was weak; among those who pled guilty, only 40% thought the evidence against them was strong. Although the findings of this preliminary study point to something other than evidentiary strength in plea outcomes, Dezember and Redlich caution against drawing definitive conclusions. They argue, instead, that future research is needed with improved evidentiary measures (including measures that assess the quality of individual types of evidence), on negotiations that result in non-​custodial sanctions, and on how courtroom actors work separately and together. Shi Yan advances research on pleas negotiations in Chapter 9 through his discussion of the measurement of plea discounts. He argues that the disparity between plea and trial sentences raises alarm and, therefore, necessitates valid estimates of the size of plea discounts in order to more fully understand the trial penalty. The considerable variation in the estimated magnitude of discounts in the empirical literature is related to the variety of theoretical frameworks used to examine plea bargaining, including theories about the importance of shared courtroom workgroup norms (regarding the quick disposition of cases, group cohesion, and uncertainty reduction) and the shadow model (where plea offers and decisions to accept those offers are rationally made based on evidentiary considerations related to conviction likelihood) (see also Chapter 8). Shi argues that findings from prior research on sentence discounts resulting from guilty pleas should be carefully evaluated in regard to employed research methodologies. A  common research technique is to estimate a multivariate regression model where the mode of disposition (guilty plea versus trial) is assessed for its effect on the sentencing outcome. A second, more complicated approach entails estimating what would have happened to defendants who pled guilty if they had opted to go to trial. Such an approach produces an observed plea sentence and an estimated trial sentence from which the size of a plea discount can be derived. Researchers find that the average size of plea discounts varies widely across jurisdictions—​from 13% in North Carolina to 50% in Florida. There is also some evidence to suggest that case-​related variables, such as the type of offense, affect plea bargaining outcomes; plea discounts are higher for defendants facing conviction on drug charges. The evidence related to the effects of defendant sociodemographic characteristics (such as race and ethnicity), however, is mixed.And, some researchers have found that the magnitude of a potential plea discount affects the decision of whether or not to plead guilty. More work needs to be done, and Shi specifies steps for moving forward at the end of Chapter 9. First, there is a need for research on sentencing alternatives to incarceration, given the extent to which community sanctions are used and because prosecutors often encourage guilty pleas via promises to recommend probation or an intermediate sanction instead of incarceration. Future researchers should explore ways that alternative sanctions may be integrated into empirical models of plea bargaining. 8

Introduction

Second, researchers should pay more attention to omitted variable bias because the exclusion of known predictors may distort estimates of the magnitude of plea discounts. However, administrative databases often exclude variables of interest to scholars. Therefore, qualitative studies (observations and interviews) and experiments may be necessary.Third, the length of time spent on plea bargaining should be analyzed, including how defendant perceptions about the process may alter over time. In Chapter 10, Jay Gormley and Cyrus Tata further emphasize that more research on plea bargaining is desperately needed, along with more critical thought about the assumptions that underlie the practice. From a justice system perspective, there are three main reasons offered to explain why sentencing differentials that materialize from plea bargaining are encouraged and tolerated—​the efficiency rationale, the victim rationale, and the remorse rationale. With regard to efficiency arguments, many believe the court saves time and money when defendants plead guilty (especially when they do so at earlier junctures in the process); guilty pleas are necessary to minimize court backlog. But plea bargains are used widely in both high-​and low-​volume courts, which indicates that concerns over caseload pressure do not drive the practice. Moreover, purported cost savings are countered by a “culture of delay” wherein wasteful actions, lack of preparation, and the need for adjournments are common. Prosecutors overcharge in order to have bargaining chips. Defense attorneys advise their clients against pleading guilty to exaggerated initial charges. Motions for continuances are filed in order to allow back-​and-​forth dialogue between parties. All the while, however, court actors and defendants are required to go to court. In short, there are problems with the assumption that guilty pleas enhance system efficiency. A second assumption is that guilty pleas spare victims from the ordeal of a trial, which must be carefully reconsidered given that some cases either have no victim or have victims who would prefer contested trials for cathartic reasons. The remorse rationale must also be scrutinized, despite its enduring appeal. The overall argument is that sentencing discounts are accepted and tolerated for defendants who plead guilty because their admittance of guilt in open court demonstrates remorse, warranting less severe punishment. Related to this, those who plead guilty also affirm the “righteousness” and legitimacy of the criminal justice system. However, the argument that a guilty plea provides evidence of defendant’s remorse and acceptance of the legal process is weak, because both guilty and innocent defendants may plead guilty for any number of reasons. Many assume that defendants freely decide whether to plead guilty or opt to go to trial; many believe that innocent defendants are not somehow coerced into admitting guilt. But, as Gormley and Tata point out, there are multiple reasons why innocent defendants would enter guilty pleas. The stress and anxiety that come from criminal prosecution, unfamiliar vernacular, complicated legal procedures, limited control, and a general lack of understanding can quickly overwhelm any defendant, which could result in hasty and uninformed decisions to plead guilty. Other defendants may plead guilty due to “passive resignation” rather than any true acceptance of guilt; some may feel unequipped to challenge a prosecutor’s version of events because they were under the influence of drugs or alcohol at the time of the alleged offense and have only a hazy recollection of what happened. Still other defendants may plead guilty because doing so amounts to immediate liberation from pretrial detention and/​or the legal process as a whole. Gromley and Tata provide evidence of such realities in Chapter 10 via their informative summary of findings from a recent study of 12 Scottish defendants. In-​depth interviews with these defendants revealed that their guilty pleas resulted from pressures about wasting the court’s time and only having minutes to decide how to plea. Most felt their cases and time were trivialized. Interviews further revealed that defendants, even those with extensive prior records, experienced profound “pains of waiting” and associated costs, which significantly factored into their decisions to plead guilty. Some unemployed defendants had to travel notable distances to court, incurring cost. Others had to miss work and/​or arrange child care. Many learned their court hearings had been rescheduled only after they arrived in court, but were nonetheless reminded of the requirement to make themselves available to court at short notice. Others had to wait all day for their cases to be heard. In short, the criminal 9

Cassia Spohn and Pauline K. Brennan

process became like a “part-​time prison” from which many defendants sought relief by pleading guilty. And, even among the guilty, the threat of a much more punitive sanction following trial forced guilty pleas. Interviewed Scottish defendants, however, did not consider their possible chances of non-​conviction at trial when making their decisions, a finding that runs counter to the shadow model discussed in Chapters 8 and 9. In addition, defendants seemed to give little thought to the legal distinctions between bargains related to sentence discounts, charge reduction, or fact modification. Moreover, defendants did not think that less severe punishments amounted to “some kind of discount.” Other defendants stated they were unclear about the nature of their charges, although they accepted wrongdoing. It is also possible that some defendants plead guilty because they distrust the system and feel that resistance is futile. Many have speculated that levels of distrust in the system are high among minority defendants, which leads to their disinclination to be cooperative. Indeed, as Gormley and Tata point out in Chapter 10, previous research indicates that minorities are less likely to plead guilty and, if they do, they do so later.Therefore, minority defendants are at a greater risk than whites of being adversely affected by plea-​dependent sentencing differentials, which amounts to more punitive sentencing outcomes for the former. Such a reality challenges the principles of equal protection under the law and the presumption of innocence. Our fourth part, “Capital Punishment,” focuses on evolving attitudes toward capital punishment and disparity in the application of this sanction. In Chapter 11, Amy L. Anderson, Weng-​Fong Chao,  and Philip Schwadel review support for capital punishment over time, subgroup levels of support, and the effects of crime rates, racial animus, and the media in shaping the public’s attitudes about the death penalty.Their examination is important because it considers the US Supreme Court’s stipulation in Furman that the constitutionality of the death penalty under the 8th Amendment be linked to “evolving standards of decency,” which could be derived from periodic assessments of an informed public. The question, however, is the extent to which the opinions derived from a biased public should be considered. US levels of support for capital punishment have varied over time; levels of support were the lowest in the mid-​1960s (at about 40%) and the highest in the early 1990s (at roughly 80%). Fifty-​five% of Americans currently support the death penalty, although many soften their attitudes if presented with options such as life without parole. In addition, levels of support will depend on differences in respondent social and demographic characteristics. In general, Southerners, whites, males, religious fundamentalists (especially evangelical Christians), and conservatives are more likely to favor capital punishment, but the relationship between respondent age and support for capital punishment is more nebulous. Across all sociodemographic subgroups, Republicans are most supportive of the death penalty, with support close to 80%. In contrast, only 35% of Democrats support capital punishment. What may be perhaps more interesting is that the political gap in attitudes about the death penalty has widened over time, with “striking divergence” identified at the close of the 20th century. More research is needed on how views about capital punishment have contributed to partisan sorting and vice versa. Research on what shapes death penalty opinion is evolving in its complexity. Although many assume that crime rates affect support for the death penalty, empirical evidence is mixed. However, Anderson and her colleagues point out in Chapter 11 that most studies used cross-​sectional designs or examined only a few points in time. By contrast, more sophisticated age-​period-​cohort modeling techniques may more aptly tease out the effects of crime rates on death penalty attitudes. One recent study employing such a technique, for example, found that the Uniform Crime Reports overall crime rate significantly predicted period-​level variation in death penalty attitudes. Future researchers interested in life-​course effects on attitudes about capital punishment are encouraged to use methods to adjust for collinearity between age, time period, and cohort effects. An examination of period-​ specific prevalence of racist discourse, for example, could meaningfully explain differences in support for capital punishment among blacks and whites. 10

Introduction

This last point may be especially meaningful in that findings from many empirical studies demonstrate that levels of racism significantly affect support for capital punishment among whites; racist whites are more likely to support the death penalty than non-​racist whites. In fact, many argue that racial prejudice, more than anything else, influences death penalty support. Racism arises from processes connected to the development of racial stereotypes, including how societal events are interpreted and explained by political leaders and by the media. Racially charged accounts, especially when such accounts link racial images to violence, affect societal levels of racial prejudice. Stereotypes of minorities as crime-​prone, violent, and dangerous are indicative of prejudiced thinking, which contributes to ideas about the use of capital punishment. More research is needed on how attitudes about capital punishment are shaped by politics and by media accounts of crime and criminals. Data on social media sources such as Facebook and Twitter, in particular, are needed. Researchers are also encouraged to examine whether news stories that stoke racial animus and/​or whether partisan news sources differentially affect death penalty attitudes. To that end, Anderson and her colleagues conducted a preliminary examination of death penalty support and partisan news viewership (operationalized as the extent to which individuals watched Fox News or MSNBC shows), with data from the 2016 Times Series Study of the American National Election Studies. Their findings, which are reported at the end of Chapter 11, suggest that individuals who watched more Fox News programs more strongly supported the death penalty, while support for the death penalty was weaker among more frequent viewers of MSNBC programs. Given that Fox News and other conservative news outlets are inclined to showcase President Trump’s punitive and racially charged rhetoric, support for capital punishment is expected to increase among partisan viewers. Overall, due to social and political forces, which have produced striking political divisions among the US electorate, one must question whether public opinion should be considered in assessments of evolving standards of decency in debates on capital punishment or other sanctions. Although it is clear that there is support for the use of capital punishment, very few people are sentenced to death and even fewer are executed. The infrequent use of the death penalty is likely an artifact of the many moral, legal, and empirical questions it raises regarding fairness in application and possible irreversible wrongful conviction. Questions of fairness, for example, materialize when one considers that blacks and Hispanics are disproportionately given the death penalty, with recent figures indicating that racial and ethnic minorities constitute 54% of death row inmates nationwide. In Chapter 12, Jeffrey Ulmer and Lily Hanrath closely examine the issue of racial disparity in the imposition of the death penalty. They begin with a discussion of US Supreme Court decisions regarding disparity in the application of the death penalty, including majority rulings in key cases (in Furman v.  Georgia, Gregg v.  Georgia, and McKleskey v.  Kemp) and notable dissenting opinions from more recent cases (i.e., Callins v. Collins and Glossip v. Gross). This collective body of case law has shaped policies and ideas about the process by which the death penalty may be imposed. Nonetheless, prosecutors exercise vast discretion in deciding whether to file a capital charge or subsequently reduce a death-​eligible charge to a lesser murder charge. If a defendant is convicted of first-​degree murder, a prosecutor must then file a motion to seek the death penalty (i.e., in jurisdictions that allow the punishment). A judge or jury must then decide whether the defendant deserves the death penalty, based on the existence of statutorily defined aggravating and mitigating circumstances. An appellate process may then follow, which could result in resentencing. The point of all of this is that case attrition occurs at each stage of the process, which narrows the universe of cases that may ultimately receive the death penalty. Moreover, each point in the process may lead to disparate decision-​making based on defendant characteristics such as race and ethnicity. Chapter 12 continues with a comprehensive review of findings from 27 studies published between 1995 and 2017 on the effects of the race of the defendant and victim in prosecutorial decisions and in court sentencing decisions for death penalty cases. In general, most researchers find that prosecutors are more likely to seek the death penalty in cases involving a white victim or in cases where the 11

Cassia Spohn and Pauline K. Brennan

defendant is black and the victim is white. This particular offender-​victim dyad also matters at the time of sentencing; a large body of research finds that black defendants convicted of killing white victims are the most likely to be sentenced to death. Fewer studies have examined the effects of defendant or victim gender, although there is some research to suggest that the death penalty is less likely for female defendants and more likely in cases involving female victims and especially white female victims. Racial disparities in case processing may, of course, be an artifact of defense lawyer competence (or incompetence), but, as Ulmer and Hanrath note, most death penalty research does not include a measure of the type of legal representation in analyses of charging or sentencing decisions. Findings from the few studies with such a measure indicate that plea bargains are more likely for defendants with private attorneys; those with public defenders are more likely to be charged with capital murder and to have the death penalty sought. Defendants with public defenders are also more likely to receive the death penalty, but the effects of defense attorney differences are likely to vary greatly across counties in the same state. Ulmer and Hanrath then provide an overview of the highly localized nature of death penalty sentences and executions; 31% of death penalty sentences in 2017, for example, were imposed in three US counties. Further evidence of substantial variability between local court jurisdictions in decisions regarding the death penalty comes from a recent study of first-​degree murder convictions in 18 Pennsylvania counties. As noted in Chapter 12, prosecutors in Pittsburgh were much less likely to seek the death penalty and prosecutors in Philadelphia were much more likely to retract the death penalty than prosecutors in other counties. Pittsburgh and Philadelphia prosecutors were also less likely to seek the death penalty for defendants represented by public defenders than prosecutors elsewhere. In moving forward with research on the death penalty, Ulmer and Hanrath advise that scholars utilize a combination of the focal concerns and court communities’ perspectives to guide their research (for earlier discussions of these perspectives see Chapters 8 and 9). Prosecutors, judges, and juries are the products of their socio-​political environments, which means they must be sensitive to their broader communities when making decisions. Perceptions about what constitutes a heinous murder will differ from place to place, and it’s possible that such perceptions would be shaped by a victim’s or a defendant’s race in different ways. Resources to pursue the death penalty may also vary by jurisdiction, so prosecutors in some places may sparingly file capital charges. It is also possible that more politically and religiously conservative jurisdictions may foster more punitive postures by prosecutors, judges, and juries in capital murder cases (for a similar discussion see Chapter 11). And, given that opinions about the death penalty vary among blacks and whites (and especially among racist whites), the local racial or ethnic composition of a community may condition punitive court responses and, thereby, the effects of defendant and victim race. Such possibilities should be explored both quantitatively and qualitatively in future studies, along with further examinations of the effects of legal counsel and whether disparate outcomes accumulate as cases moved through the court process. The final part of the handbook, “Current Controversies,” examines the role that criminal history plays in sentencing, responses in California to mass incarceration, and whether problem-​solving courts and restorative justice programs may offer viable options for sentencing in the 21st century. In Chapter  13, Rhys Hester, Richard S.  Frase, Julia Laskorunsky and Kelly Lyn Mitchell tackle the important issue of the role of criminal history in US sentencing guidelines systems. Arguing that “laws and policies that impose hefty increases in punishment on the basis of past crimes need careful rethinking,” the authors discuss the adverse consequences of basing sentence severity on criminal history, document jurisdictional variations in policies regarding the use of criminal history scores to enhance sentences, and summarize inherent difficulties in attempting to justify prior record enhancements using retributive or utilitarian theories of punishment. Their chapter concludes with a review of empirical research designed to evaluate the risk-​predictive accuracy of criminal history scores used in federal and state courts. 12

Introduction

An important contribution of this chapter is the fact that it raises both normative and empirical questions about the legitimacy of criminal history-​based sentencing enhancements, especially in the context of sentencing guidelines. The authors document that there is little evidence demonstrating that the various prior record formulas and resulting sentence enhancements serve either retributive or utilitarian goals. Moreover, they note that even if a particular prior record formula can be justified based on the greater culpability of repeat offenders and/​or by incapacitation or deterrence of higher risk offenders, these positive effects must be balanced against the adverse consequences that result from using criminal history to enhance sentence severity—​undercutting policies that prioritize use of prison for offenders who commit the most serious crimes; increasing the prison population by imprisoning more low-​level and non-​violent offenders and by lengthening sentences for all offenders; reducing proportionality in sentences by imposing more punitive sentences on offenders convicted of low-​level offenses than on those convicted of higher-​severity offenses as a result of prior record enhancements; and increasing the racial disproportionality of the prison population. As the authors note, the lack of research on either the positive or negative effects of prior record enhancements, coupled with serious questions about the degree to which such enhancements can be justified on normative grounds, suggests that it is indeed time to rethink these policies and practices. In Chapter 14, the focus turns to the impact of a California law—​AB109 or the Public Safety Realignment Act—​that was designed to decrease the California prison population while ensuring public safety. Passed in response to a US Supreme Court decision which ordered the state to reduce its prison population, AB109 provided that offenders sentenced after October 1, 2011, on non-​serious, non-​violent, and non-​sex offenses (the so-​called “N3s”), were, with certain limited exceptions, no longer eligible for state prison sentences. It also transferred responsibility for supervising N3 offenders from the state to the county. In their mixed-​methods study, Katharine Tellis and Cassia Spohn evaluate the impact of AB109 in Los Angeles County; they determine whether AB109 affected crime rates in LA County, document thepercentage of offenders who successfully completed the county post-​ release supervision program, evaluate the recidivism rates of offenders who participated in the post-​ release supervision program, and assess the attitudes of relevant LA County stakeholders (i.e., police chiefs and probation officers) toward AB 109. Tellis and Spohn used Uniform Crime Report (UCR) data on crimes known to law enforcement and on drug arrest rates from 2009 through 2013 in Los Angeles County to examine the impact of AB109 (i.e., Realignment) on crime. Using interrupted time series analysis, they found that Realignment had no effect on the overall crime rate, the violent crime rate, the property crime rate, or the rate of arrests for drug sales and manufacturing. Their evaluation of Realignment also examined program completion and recidivism for offenders who participated in the post-​release community supervision program and were supervised by the Los Angeles County Probation Department. They found that two-​thirds of the offenders successfully completed probation, that most supervised offenders were not convicted of a new felony while on supervision, and that the likelihood of recidivism was affected by the offender’s age, number of prior felony convictions, and risk score. An important contribution of Tellis and Spohn’s chapter is its extensive discussion of how the criminal justice system in Los Angeles County adjusted to the changes wrought by Realignment. Their interviews with Los Angeles County Probation Department personnel and law enforcement executives from the Los Angeles County Police Chiefs Association revealed that although AB109 encouraged inter-​agency collaboration and led to a greater emphasis on reentry and rehabilitation (as opposed to enforcement) strategies, it also raised concerns about the safety of unarmed probation officers, the degree to which AB109 shifted the burden to local jails, and the challenges created by homelessness and mental health issues among probationers. The results of the Tellis and Spohn study suggest that Realignment, which they characterize as “a major shift in the philosophy and practice of criminal justice in California,” may be part of the solution to the problem of mass incarceration. Their research, which was conducted in the largest county in the State of California, demonstrates that diversion of non-​violent, non-​serious, and 13

Cassia Spohn and Pauline K. Brennan

non-​sex offenders from prison can be accomplished without an increase in crime or recidivism rates. However, they end their chapter with a cautionary note, stating that successful implementation of a major policy change like Realignment requires “transparency, evidenced-​based practice, and evaluation research.” Problem-​solving courts offer another solution to the problem of mass incarceration. Such courts are purported to provide a cost-​effective community-​based option for low-​level offenders for whom incarceration may be unnecessary. In Chapter  15, Eileen M.  Ahlin and Anne S.  Douds provide a broad overview of problem-​solving courts and organize their discussion around the following four areas: the evolution of these courts, post-​hoc theoretical justifications for the existence of such courts, the need for theoretically grounded empirical research, and directions for future research. Problem-​solving courts developed as an alternative to traditional court sanctions, which were seen as ineffective in reducing crime. Evidence of the ineffectiveness of incarceration as a crime prevention strategy for drug offenders, combined with increases in drug cases arising from drug-​war policies, called for solutions that would address underlying drug abuse problems, and the first drug court was created in Miami-​Dade, Florida in 1989. As positive anecdotal accounts from judges began to emerge, political support materialized and abundant federal funding for nationwide drug court expansion efforts followed. The number of drug courts quickly grew, and problem-​solving strategies expanded to include mental health courts, veterans’ treatment courts, domestic violence courts, family treatment courts, and reentry courts. An estimated 3,000 problem-​solving courts currently exist across the nation. Although various names may be used, all problem-​solving courts seek to address unique situations for targeted populations of offenders via concentrated expertise and close judicial oversight. Individualized team approaches to supervision are typical in these courts, where members of the courtroom workgroup (prosecutors, defense attorneys, and judges), corrections personnel, peer mentors, and social service providers within a community work together to devise focused strategies intended to address an offender’s underlying criminogenic needs, in order to prevent crime, enhance social well-​being, and reduce levels of incarceration. While many different types of problem-​solving courts now exist across the country and while some studies show that such courts reduce recidivism, Ahlin and Douds argue that the extant empirical research is disjointed and lacks a clear theoretical basis. Evaluations of problem-​solving courts should be informed by theory, and an integrated theoretical framework to guide empirical research is offered in Chapter 15. The framework begins with the notion that problem-​solving courts operate from a perspective of applied therapeutic jurisprudence, wherein judicial oversight of comprehensive approaches to solve or mitigate offender problems occurs. But, varying oversight strategies, treatments, social services, and methods to induce participant compliance will affect whether rehabilitation and deterrence materialize.These key punishment objectives are more likely to be achieved if participants have trust in a fair process wherein they are given a voice, feel as if they have some control, believe they matter, and think the justice system’s response is impartial, justified, and ethical. In other words, perceptions of procedural justice should play a critical moderating role in program success. Also, if rehabilitation and deterrence are achieved, then restorative justice is possible, which some believe should be the ultimate goal of any intervention. In moving forward with empirical research, Ahlin and Douds recommend that evaluations should be theoretically driven and employ longitudinal designs with measures of treatment quality and perceptions of procedural justice. Overall, further examination of why problem-​solving courts may work is required in order to determine the best course of action for sentencing options in the 21st century. The therapeutic jurisprudence model, which provides the foundation for problem-​solving courts, connects well to ideas about restorative justice. Both therapeutic jurisprudence and restorative justice focus on attempts to bring about positive change in offenders, rather than on the retributive necessity of punishment. In Chapter 16, Jennifer L. Lanterman traces the development and use of restorative justice practices, reviews research findings, discusses the challenges that lie ahead, and offers 14

Introduction

recommendations in moving forward. Contemporary restorative justice practices are derived from methods used by ancient societies, indigenous cultures, and religious groups to resolve conflicts and address harms. Modern-​day restorative justice practices seek resolutions to crime-​related harm with involved victims, offenders, and communities. Key components include voluntary participation, victim-​centered approaches, and offender accountability. Although voluntary participation is essential, many believe an element of coercion exists for offenders due to the threat of formal criminal justice processing looming in the background. In addition, while the victim’s healing should be the ultimate goal, critics assert that the restorative justice process could be “co-​opted for other purposes,” including a desire for expedited case resolution. Moreover, Lanterman argues that the coercive nature of the process “diminishes the offender’s autonomy and detracts from his or her ability to be actively responsible to the victim.” The overall criticism is the restorative justice practices may be no different than methods employed within the traditional criminal justice system. Nevertheless, the notion of restorative justice has gained support across the United States, and 35 states have adopted legislation encouraging the use of restorative programs for juvenile and adult offenders. Lanterman questions the extent to which such programs may provide an “an antidote” to what ails the conventional system. Restorative justice is most often used as a diversionary option for juvenile and first-​time offenders who have committed low-​level offenses. Diversionary restorative justice is less likely in cases of violent crime. Instead, post-​sentencing programs are more likely, where encounters between victims and offenders generally occur years later within a confinement facility under the direction of very experienced facilitators who have taken considerable time to prepare participants. But, post-​sentencing programs with inmates will do little (if anything) to reduce prison populations. The four primary forms of restorative justice for criminal matters include victim-​offender mediation (or victim-​offender dialogues), restorative conferences, family group conferences, and circle processes. Of the four, victim-​offender mediation is most often used in the United States. Several researchers have found that victims who participate in restorative justice become less fearful of future victimization, have fewer emotional symptoms, and experience less post-​traumatic stress associated with their victimization. However, others have found that victims may have negative experiences if facilitators are insensitive or unprepared, if the need for case resolution dominates, and if victims do not know what to expect of the process. Research on whether restorative justice reduces reoffending has produced inconclusive findings. Some argue that restorative justice used in conjunction with conventional prosecution is more effective at lowering recidivism, when compared with diversionary alternatives. Others argue that a lack of success is to be expected if offenders believe the restorative option provides a way to skirt the system or if they think the experience is counterproductive. When findings of victim and offender outcomes are taken together, it appears that implementation by well-​trained facilitators is critical. The challenge, however, lies in finding the resources necessary to properly train facilitators and to evaluate their performance. Related to this, stated requirements for facilitators are generally lacking, so concerted efforts must be made to create basic regulatory standards. And, the sustainability and expansion of restorative justice programs will be an issue without government funding. In moving forward, Lanterman suggests that exit surveys be employed to ascertain participant perceptions of the process (echoing the importance procedural justice from Chapter 15). A research strategy such as this would help shed light on the mechanisms by which beneficial outcomes may be achieved. In addition, a process of case review should be used to evaluate the strengths and weaknesses of facilitators, given the seminal role they play. Finally, it will be crucial for restorative justice programs to obtain statutory funding required by law, although non-​traditional funding through social impact bonds may provide a viable alternative funding mechanism. Lanterman concludes that as we advance into the 21st century, “instead of thinking of restorative justice as a discrete criminal justice system reform that will independently reduce the issues of victim marginalization and high rates of offender 15

Cassia Spohn and Pauline K. Brennan

recidivism … thinking of restorative justice as one of several reforms to collectively address these issues may result in more realistic expectations.”

Conclusion We believe that the papers included in this volume make a significant contribution to our understanding of the punishment process and extend research in a number of important ways. The contributors highlight not only the ways in which sentencing policies and practices have evolved over the past several decades, but also the consequences of these changes for fairness, equity, and proportionality in punishment. The papers emphasize that sentencing is a collaborative process that involves legislators, prosecutors, and corrections officials, as well as judges and jurors, and that plea bargaining plays a central role in determinations of punishment severity. They also illustrate that policy changes may have unintended collateral consequences and raise questions about our underlying normative assumptions regarding the factors that should determine the punitiveness of the sentence. We hope that this collection of thought-​provoking and policy-​relevant papers will stimulate future research on the important issues that the authors address.

References Spohn, C. (2015). Race, crime and punishment in the twentieth and twenty-​first centuries. Crime & Justice: A Review of Research, 44: 49–​98. Walker, S. (1993). Taming the System: The Control of Discretion in Criminal Justice, 1950–​1990. Oxford: Oxford University Press.

16

PART I

The Evolution of Sentencing Policies and Practices

1 THE TRANSFORMATION OF SENTENCING IN THE 21ST CENTURY Megan C. Kurlychek and John H. Kramer

Scholars of criminal justice and criminology often debate the philosophical purpose and procedures of our nation’s criminal justice system as they relate to current sentencing practice, but seldom do they reflect upon its origins, the twists and turns it has undertaken over time, and how these pathways might inform and even predict future directions. Originally modeled from English practice rooted in the seminal work of Cesare Beccaria’s “On Crimes and Punishment” (1764), the system was designed to deter crime by providing penalties that were proportionately severe to the offense committed. Thus, the determinate sentence handed down by a judge was to be just severe enough to outweigh any benefits reaped from the commission of the offense. Viewing man as a rational and utilitarian being, this philosophy would work because the pain caused by the punishment would outweigh any possible benefit. Of equal import for our discussion is the central tenet of proportionality in that the punishment “… should fairly reflect the degree of reprehensibleness (that is, the harmfulness and culpability) of the actor’s conduct” (von Hirsch & Ashworth, 2005). If the punishment did not seem to fit the crime, that is, if it was too severe, this philosophy suggests and research supports that the punishment could actually undermine the legitimacy of the system and lead to increased crime (Tyler, 2006; Baker, 2017). For the first 100  years of our nation’s history, deterrence and proportionality reigned supreme over sentencing policy. However, a significant re-​direction of our philosophy of punishment came in 1870 when the Prison Congress proposed treatment rather than punishment as the purpose of incarceration and encouraged the adoption of indeterminate sentencing practices to replace the use of “proportionately severe” determinate sentences. Put simply, instead of the sentence being prescribed as a fixed amount of time proportional to the severity of the crime, the sentence length was open (e.g., indeterminate), to end only when the individual was deemed rehabilitated.1 By the mid-​1940s, the federal government and all states had switched to a primarily indeterminate sentencing model that adopted this new philosophy (Tonry, 1999). But, after another 100 years from the original shift, this new model was to come under question and a return to the deterrence philosophy would begin. Only this time, as we will extrapolate below, the new philosophy would expand to include notions of retribution and incapacitation as well, which would in turn result in an increased emphasis on severity, rather than proportionality, of sentence. In this chapter we explore not only the original shift from deterrence to treatment, but the more recent shift from treatment to punishment that has been the query of several scholars—​framing the question in both the past and future. We finish

19

Megan C. Kurlychek and John H. Kramer

with a discussion of recent winds of change we see on the horizon and the complex task of making predictions about what sentencing in the 21st century will become.

Implementing the Rehabilitation Model The rehabilitative model of criminal justice (sentencing), although seemingly benevolent and therefore desirable, holds the somewhat questionable assumption that offenders suffer from some physical, mental or social “sickness” that can be diagnosed and treated. Individuals do not therefore act fully from a point of rationality and choice, but rather may suffer from an ailment (ranging from a diagnosable mental condition to a drug or alcohol addiction) or suffer from extreme negative social circumstances such as poverty and exposure to violence and victimization that place the individual at risk for criminal behaviors. Such diagnoses are typically the function of intake officers, probation workers, and pre-​sentence investigations that often rely on risk (and in more modern time, risk, needs and responsivity) assessments (Andrews, Bonta, & Hoge, 1990). Once the ailment is diagnosed, the system can then prescribe the proper treatment to cure, or rather, rehabilitate, the offender into a functional and productive citizen. While such a model seems progressive and desirable, having been first introduced in the late 1880s, it wasn’t until decades later that social scientists began to question the lack of empirical evidence on its effectiveness in crime control (Bailey, 1966; MacNamara, 1977; Martinson, 1974). In fact, still high on the rehabilitative ideal, in 1966 New York Governor Rockefeller formed the Governor’s Special Committee on Criminal Offenders. This committee was charged with devising imaginative new approaches to rehabilitate offenders for “any way we can rehabilitate more of these criminal offenders and reduce the number of repeaters” was proposed as the best method to significantly reduce crime in New  York. To fulfill this end the Committee and the New  York Office of Crime Control Policy (later to become the Division of Criminal Justice Services) contracted with two criminologists to conduct a survey regarding what was known about the effectiveness of rehabilitative efforts. The final report, which was not published until 1974 for mostly political reasons, reviewed 231 evaluations of a variety of correctional treatment programs including educational training, vocational and skills training, counseling/​psychotherapy, surgical treatments, and intensive community supervision. Although the report did find that about half the programs reported success, at least with certain types of offenders, the final results of this “What Works?” report were instead interpreted as “Nothing Works” (Sanchez, 1990, p.  131). Although Robert Martinson himself (1979) would later acknowledge that his conclusions were overstated, until this time the benefits of the treatment philosophy had just been assumed (Clark, 1970). Thus, Martinson’s work exposed a critical gap in scientific knowledge and created a crisis in the field. With crime on the rise and conservatives blaming the increase on lenient liberal policies, this lack of evidence for the medical model delivered a nearly fatal blow to the rehabilitative philosophy of corrections. Historian David Rothman (1980) in his book, Conscience and Convenience, further unveils the imperfections of correctional developments during the latter part of the 19th century through the mid-​20th century. As the title of his book suggests, the progressives’ (1870–​1970) agenda of reforming offenders was undermined as it ran into institutional “convenience” in its application. In Rothman’s (1980, p. 10) view, “… when conscience and convenience met, convenience won. When treatment and coercion met, coercion won.” That is, what may appear to be treatment on the surface was not always so carefully implemented or followed when one dives below the surface. In this sense, Rothman saw the essential ingredients of the anti-​progressive movement even as it seemingly operationalized the progressives’ treatment model. Given this backdrop, perhaps Martinson’s conclusions were not as startling as they first appeared, as much treatment was never fully implemented. Or, perhaps, the goals of treatment were subservient to the interests and goals of those in the organization who had minimal interest or training in the treatment agenda. 20

Sentencing in the 21st Century

From this view, one might ask why reform a system that already operated with a coercive punitive orientation despite its shiny, rehabilitative “cover?” Yet, in the late 1960s and 1970s the attacks on the progressive treatment model were broad-​based, coming from both liberals and conservatives, and they were strident and vicious. Liberals (American Friends Service Committee, 1971; Frankel, 1972; von Hirsch, 1973) argued that treatment didn’t work, caused unwarranted disparity, and was overly punitive. Conservatives believed treatment did not work and that the policies of the time were overly lenient on the offender without paying needed attention to protection of victims and society (Carrington, 1983; Cary, 1993; Wilson, 1975). Thus, the liberals wanted to replace the indeterminate sentencing system with determinate sentences built on a “just desert” framework that would correct the unfairness in the indeterminate model while maintaining or reducing prison populations. On the other hand, conservatives viewed the fix to increase the certainty and severity of sanctions. In the context of a rising crime rate, however, the balance was certain to tip in the conservative direction. The conservative attacks came from the view that the individualized treatment model had resulted in lenient penalties that undermined the rule of law and lowered the deterrent and incapacitative effect of sentences. These attacks were grounded in two key developments. First, beginning in the mid-​1960s, crime rates had dramatically increased. Second, reviews of correctional treatment by Bailey (1966) and Martinson (1974) found little support for the effectiveness of treatment. Indeed, Robert Martinson’s comprehensive review, as noted above, was broadly interpreted as concluding that “nothing works.” The conservative view pounced upon the perspective that treatment did not work and that its leniency had caused a breakdown in social control, resulting in a crisis and the need to reestablish law and order (Carrington, 1975; Cary, 1993). It is from the more conservative stance that mandatory minimum sentences, three-​strikes legislation, and truth in sentencing became the discretionary control mechanism that resonated across the country, such that by the 1990s all states had passed some flavor of these reforms (MacKenzie, 2001; Stemen, Rengifo, & Wilson, 2006). Sentencing reform, however, did not just rest with conservative efforts; liberals were also concerned with sentencing practices for a different reason.The liberal attacks were first resoundingly expressed by the American Friends Service Committee (1971), where in its report, Struggle for Justice, it concluded “… the individualized treatment model, the ideal toward which reformers have been urging us for at least a century, is theoretically faulty, systematically discriminatory in administration, and inconsistent with some of our most basic concepts of justice” (American Friends Service Committee, 1971, p. 12). But this was just the beginning. Struggle for Justice was followed in the next decade by a series of commentaries and proposals (see, for example, von Hirsch, 1976; Twentieth-​Century Fund, 1976; Singer, 1979) that indicted the treatment model for its unbridled discretion and harshness. Perhaps the most compelling attack on sentencing came from inside the judiciary. Federal Judge, Marvin Frankel, gave a series of lectures in the early 1970s that were published under the title Lawlessness in Sentencing (1972) and later as a book entitled Criminal Sentences: Law without Order (1973). Frankel argued that judicial sentencing discretion was “lawless,” as judges did not have to justify their sentences and their broad discretion was unreviewable. The result of this broad discretion was, to him, unacceptable levels of sentencing disparity. Consequently, Frankel argued that we needed an administrative agency (sentencing commission) that would establish sentencing guidelines to structure judicial sentencing discretion. Frankel’s proposal stimulated a major series of sentencing reforms across the country at both the federal and state level.The commission promulgated guidelines were designed to help judges determine the appropriate sentences to be served for a crime, to reduce judicial discretion and, thus, reduce disparities in sentencing. However, hidden, perhaps not so carefully, within these policies, were definitions of crimes and their proposed punishments that would adversely impact minority populations—​particularly, young black males (Tonry, 1997). Moreover, in many cases these new policies would actually introduce, rather than reduce, disparities in sentencing outcomes. Probably the most visible of these was the USSC cocaine guidelines that extended mandatory minimums and disproportionately applied penalties to crack versus powder cocaine (Tonry, 1997; 2016). 21

Megan C. Kurlychek and John H. Kramer

Catalyst for Change As suggested above, the period of the 1970s signaled the demise of rehabilitation as the driving purpose of sentencing in both academic and political circles and began a thirty-​year revolution in sentencing policy and practice. These changes, particularly conservative initiatives to implement mandatory and truth-​in-​sentencing laws, set the stage for what often has been referred to as America’s great experiment with mass incarceration. In Garland’s (2001) work, The Culture of Control, he details multiple societal factors that he believed changed the criminal justice landscape in the period 1970–​2000. In our discussion we particularly hone in on three of these areas: 1) the decline of rehabilitation as the prime purpose at sentencing (already discussed above), 2) the return of victims to the sentencing process, and 3) the politicization of crime.We further delve deep into these areas as we see them overlapping with other proposed influences, such as the re-​emergence of punitiveness and an over-​emphasis on emotion and fear—​particularly fear of suspect populations—​as it relates to the politicization of crime, criminal justice policy, and the creation of a perpetual sense of crisis.

The Victims’ Rights Movement With the rehabilitative model on its knees after assaults from the left, the right, and academia, what would rise to take its place was to be a more punitive model that, as noted above, included not just a return to principles of deterrence, but also a focus on retribution. Garland (2001) partially credits this change to the return of the victim to the sentencing process and makes a strong case for its importance in increasing emphasis upon the severity of the offense. Indeed the Victims’ Rights Movement became an important force, perhaps in response to the increase in defendants’ rights that was witnessed under the Warren Court of the 1960s (Carrington & Nicholson, 1989; Green, 2002). President Reagan appointed a President’s Task Force on Victims of Crime to explore the ways for victims to be informed of, and included in, court decision-​making processes. Even before the Task Force issued its report, Congress passed the Victim Witness Protection Act (96 STAT. 1248 Public Law 87–​291, 1982) that authorized the payment of restitution to victims and the use of victim impact statements at sentencing in federal cases. Shortly thereafter, Congress would revise the Federal Rules of Criminal Procedure to require pre-​sentence reports to include information on “any harm done to or loss suffered by any victim of the offense” (CT 3, 96 stat at 1249, amending Fed R Crim P 32.) The Victim Rights and Restitution Act, (34 U.S.C. § 20141, 1990) extended to crime victims in federal cases the right to notification of court proceedings and the right to attend them, the right to notice of changes in a defendant’s detention status, the right to consult with prosecutors, and the right to protection against offender aggression. Subsequently, in 1994 the omnibus Violent Crime Control and Law Enforcement Act gave victims in federal cases the right to speak at sentencing hearings, made restitution mandatory in sexual assault cases, and expanded funding for local victim services (Davis & Mulford, 2008). Most recently, the Crime Victims’ Rights Act of 2004 (18 U.S.C. section 3771(a)(8)) expanded victims’ rights in federal cases to include the right to be present and heard at all public court proceedings involving plea bargains, sentencing, and parole. In addition to federal action, numerous states also went on to explore the issue and to pass similar legislation. Nearly two-​thirds of the states have amended their state constitutions to include victims’ rights to date and Pennsylvania has a proposed victims’ rights constitutional amendment that will be on the 2019 statewide ballot (Dees, 1999). A  study by researchers at the Vera Institute (Dees, 1999) analyzed victim rights legislation from every state and found that all states allow for victim compensation, the court appearance of the victim, and the submission of a victim impact statement before sentencing.2 A majority of states also gave victims the right to restitution, to attend sentencing hearings, and the right to consult with officials before offers of pleas or release of defendants from custody (Dees, 1999). 22

Sentencing in the 21st Century

As important as it is for a victim of a crime to have a voice and to be treated with respect and dignity, the reason we believe these laws increased the severity of sentences is that they inflate the need for retribution at a personal level to a status that outweighs the confines of justice.Victim input affected both individual sentences as well as legislatively established offense grading, mandatory minimums, and sentencing guidelines.3 Remember that the original deterrence model called for the punishment to be “proportionately severe” and warned that sanctions that were overly harsh could actually increase crime by delegitimizing the system. However, when working from a retributive model with vocal victim input, rather than a deterrence model, it is much more difficult to limit sanctions to those proportionately fair. That is, what might satisfy a victim may not be what necessarily fits the offense or the offender. For Garland (2001), the victim movement was a particularly powerful component of the “get tough on crime” movement, as it established the victim as a politically powerful ingredient in the processing and punishment of offenders. In Garland’s words: “The new political imperative is that victims must be protected, their voices must be heard, their memory honored, their anger expressed, their fears addressed” (2001, p. 11). Moreover, victims’ voices resounded in the halls of legislatures as they testified to the need to increase sanctions for offenses such as DUI where offender culpability was often viewed as a serious mitigating circumstance, but the effects on victims were often death or serious injuries (Green, 2002).

Politicization of Crime Movement The importance of the victim’s movement in state and federal legislatures leads us naturally to Garland’s (2001) view that the politicalization of crime and criminal justice became the most influential source of criminal justice reform. At the same time that some were holding on to the hopes of rehabilitation, others were preparing for war. In 1965 (prior to Governor Rockefeller’s declaration above), the federal government was already warning of an upsurge in crime and preparing for a “war on crime,” as declared by then-​President Lyndon B. Johnson. Indeed, the figures were shocking. During the decade of the 1960s, property crime increased 147% and violent crime 126%, according to the FBI Uniform Crime Reports. Although this upsurge in crime may have had much less to do with any lenient policies than it did with population composition,4 civil unrest in society, and great divides in income, opportunity, and liberty, that is not how it would be interpreted. Moreover, the catchy rhetoric of getting tough on crime was to continue through subsequent administrations with President Nixon attacking the growing problem of crime through enhanced law enforcement and penalties and President Reagan declaring his own war on drugs in 1982. Continuing through the 1990s, even when crime was dropping, “get tough” on crime policies provided an easy and actionable platform for candidates at the state and federal level, which fed upon the fear of the American public. What is so unique and concerning about this politicization of crime is that it takes the emphasis away from science, out of the hands of practitioners who know the system, and out of the realm of philosophical ideas of right and wrong, humanity, and civilization. Instead, criminal justice policy is reduced to what we refer to as its lowest common denominator—​fear. The average citizen is taught to fear for his/​her life and in many cases to fear a specific crime or a specific “type” of criminal (Jiwani, 2002; Kohm, 2009). Thus, the politicization of crime involves and accentuates many of the above factors such as the emphasis on punishment/​retribution over rehabilitation and the involvement of the victim. This first part was, and is, often accomplished through, or at least aided by, media representation of the most horrific cases (Anderson, 1995; Cavender, 2004). We then feel the need to protect society from imminent harm. This intertwines with the selling of an inevitable crisis as well as the culprit of the crisis (Toffler & Alvin, 1980; King, 1987). We will return to these notions in the last part of this chapter as we emphasize current influences and future predictions. 23

Megan C. Kurlychek and John H. Kramer

From Rehabilitation to Retribution: The New Face of Sentencing Combined, these influences led to shifts in policy from policing through sentencing and parole. In this section we will focus only on the main policies we believe changed the face of sentencing in America.

Mandatory Sentencing and Three Strikes Laws: The Conservative Influence Our purpose here is not to review in any detail the passage of mandatory sentencing laws across the United States during the latter part of the 20th century, but rather to acknowledge the resounding political approval of mandatory minimum sentencing policies despite considerable criticism of such penalties. In effect, mandatory penalties epitomize the ascendance of emotion to the public policy arena driven by the politicization of crime, despite concerns and criticisms raised by academic and policy and practitioner organizations, including the American Bar Association (1968, standard 2.3; 1994, standard 18–​3.21 as noted in Tonry, 1997, p. 134).Therefore, as we later reflect on what the 21st century holds for the courts and sentencing, we must consider what has and is changing in the realm of mandatory sentencing laws. Historically, America’s use of mandatory sentences began in 1790 after debate regarding the then-​ disproportionately harsh penalties attached to many offenses. At the time, capital punishment was widely used and the Federal Sentencing Act of 1790 actually de-​mandatorized the penalty of death for seven offenses and set periods of incarceration for seven offenses as well as mandatory periods of incarceration of one to seven years for 11 other offenses mainly dealing with crimes of treason, piracy, and other interference with governance.5 Thus, the original use of mandatory sentences was to decrease not increase the severity of punishment. Indeed the 1790 debate in many ways marked the movement in the United States to establish penalties that were fair relative to the seriousness of the offense and to reduce the use of overly harsh sanctions.6 This purposeful and restrictive use of mandatory sentences would not change drastically until centuries later during the conservative political climate of the 1950s. It was during this time that mandatory sentences were expanded to a new variety of crimes, particularly drug offenses.7 Interestingly, many of these same policies would be repealed under the Nixon administration—​a conservative regime—​but also a time of great social uprising in our nation. However, this was not a wholesale rejection of mandatory minimums, but rather a shift in direction. For example, in 1970 Congress repealed many mandatory minimum penalties for drug crimes, while enacting new mandatory minimums for several weapon-​related crimes.8 The expansion of mandatory minimums to weapon and firearm violations would continue into the 1980s, with Congress amending Section 24 of Title 18, Crimes and Criminal Procedure (18 U.S.C. §924) to apply a mandatory minimum sentence of five years to anyone brandishing a firearm during the commission of a crime of violence and later to the carrying of a firearm during certain drug-​related offenses (Public Law 99 No. 308 §104 100 Stat 449,456, 1986). Another major historical development was the United State Congress’ differentiation of crack cocaine from powder cocaine, with the former being viewed as a much more lethal and dangerous form of the same drug. This view materialized into policy in the now infamous 100-​to-​1 differential in which an offender was subject to a five-​year mandatory minimum sentence for possession of only 5 grams of crack cocaine while the same penalty was only invoked for possession of 500 grams of powder cocaine (The Anti-​Drug Abuse Act of 1986 Pub L. No. 99–​570, 100 Stat. 3207, 1986). Proponents focused on the drug’s dangerousness to the users and the community within which it was trafficked. Subsequently the distinction, and in particular the magnitude of the distinction, has been widely criticized for producing institutionalized racism (Tonry, 1997). The suggestion that mandatory minimums are racially charged has found much support in the empirical literature. For example, a study by Schwarzer (1992) found that given equal offenses that 24

Sentencing in the 21st Century

would qualify an individual for a mandatory minimum sentence, the mandatory minimum was invoked in reality 20% more often on non-​white than white defendants. This study also found that the differences in the length of federal sentences imposed on black offenders versus white offenders increased from 28% to 49% from the time period of 1984 (pre Anti-​Drug Control Act) to 1990. These disparities in outcomes will be discussed in more detail below as they relate to both mandatory minimums and the birth of sentencing commissions. It is to the latter we now turn our attention.

The Evolution of Sentencing Reforms and Sentencing Commissions: The Liberal Influence While mandatory minimums were seen as simple “fixes” to concerns with leniency at sentencing and were passed in all 50 states, the attacks from critics such as the American Friends Service Committee and Judge Marvin Frankel required comprehensive reform of sentencing for all crimes to create a fair system for all defendants. The three basic components of the sentencing reforms that were to be adopted included: 1) a flat determinate sentencing structure without parole release (often referred to as “truth in sentencing” and related to the Victims’ Rights Movement noted above; 2) the establishment of guidelines to structure judicial discretion; and 3) the establishment of sentencing commissions or legislative committees empowered to write sentencing guidelines. The first major sentencing reform proposal came in 1975 from Senator Ted Kennedy. Senator Kennedy’s legislation proposed creating a federal sentencing commission to write sentencing guidelines, but this legislation did not pass until 1984. In the meantime, several states took several different pathways to reform sentencing. First, Maine in 1976 abolished the indeterminate sentence and, while not restricting judicial discretion, did establish a relatively clear “truth in sentencing” model where offenders would be sentenced to a flat sentence and then they would be able to earn up to 10 days a month good time. California took a somewhat similar approach in that it abolished the indeterminate sentence but took an additional step by establishing legislatively created guidelines that provided for four categories of offenses with a presumptive sentence attached to each category. Judges were provided a very limited range to aggravate or mitigate the sentence. Under California’s initial statute the judge was given broad discretion on whether to incarcerate, but very limited discretion as to the length of incarceration. Illinois and Indiana adopted flat sentencing legislation that abolished parole release while retaining parole supervision. (Hewitt & Clear, 1983; Kramer et al., 1978). While indeterminate sentences and parole release were targeted in these initial reforms, only California established serious restriction on judicial discretion. The most significant sentencing movement then started with Kennedy’s guideline commission model and it soon became the model for major sentencing reforms across the country. In 1978, Minnesota became the first jurisdiction to pass legislation creating a sentencing commission to write sentencing guidelines and establish determinate sentencing (flat sentences without release decisions made by a parole authority). As the first sentencing commission bill to be created, it served as model legislation for other states to follow. There were several important ingredients contained in the bill that became the foundation for other states to emulate. First, it created a sentencing commission that contained broad representation. Second, it directed the commission to construct sentencing guidelines that eliminated the indeterminate sentence and replaced it with a flat determinate sentence that could be reduced only by earned good time set at the rate of one day of good time for two days of violation-​free behavior. Third, it restricted the width of the guideline ranges. Fourth, the commission had to consider the correctional resources in the guideline recommendations (Parent, 1988). The Minnesota guideline legislation served as model legislation for other jurisdictions and its commission an archetype for thoughtful and cautious sentencing reform. Following the Minnesota experience, the guideline movement seemed to be taking off, with Pennsylvania creating a sentencing commission in 1978, followed by New York in 1979, Washington in 1981 and the US Congress in 1984. The country seemed poised for a major shift in sentencing. 25

Megan C. Kurlychek and John H. Kramer

However, the guideline movement slowed after the passage of federal legislation, with another five years passing before several other states adopted the commission model. Since the early 1990s several more states have joined the commission framework, but the strength of the movement seems to have dissipated with only three states adopting the commission-​based guideline model (Alabama, 2000; Arkansas, 2011; and District of Columbia, 2006) since the turn of the century. Another interesting issue is that even states adopting sentencing commissions and developing guidelines have generally kept the indeterminate sentence paradigm and have made the guidelines voluntary or advisory rather than presumptive.

Sentencing at the Turn of the Century: The Aftermath of the Get Tough Movement What we were left with at the turn of the century was a sentencing “system” that primarily abandoned rehabilitation as a focus and replaced it instead with a hodgepodge of policies using both determinate and indeterminate sentencing to reflect elements of deterrence, incapacitation, and retribution philosophies of justice. While the goals of increased severity were realized, the movement towards increased predictability and decreased bias had stalled in its tracks. The result was not only increased use of incarceration (Blumstein, 1983; Greenberg & West, 2001), but increased disparities in sentencing and a new crisis in the legitimacy of the sentencing policy (Blumstein, 1983; Duster, 1995; Tonry,  1997).

Mass Incarceration According to the American Civil Liberties Union (2018), between 1970 and 2015 America’s prison population increased by over 700%. This means that more than 2  million Americans are behind bars on any given day. To put this in global perspective, America is home to about 5% of the world’s population but 25% of the world’s incarcerated population. Figure 1.1 shows incarceration rates for the US and the other 9 countries rounding out the “top ten” for incarceration rates. As is evident, not only are we far above even our closest competitors, but the other nations closest to the US in incarceration rates are not those thought of as having strong human rights records. As is vividly clear, the result of the wars on crime and drugs has been an unparalleled expansion of imprisonment, depriving millions of Americans of civil liberties and costing American taxpayers an estimated $80 billion dollars per year just in actual incapacitation expenses (ACLU, 2018). While this

Countries with Highest Incarcera‚on Rates Scotland England/Wales Brazil Belize Poland Ukraine South Africa Russia USA 0

100

200

300

400

500

Rate per 100,000

Figure 1.1  Incarceration Rates

26

600

700

800

Sentencing in the 21st Century

mere fact is alarming, it is even more concerning, though not surprising, that these effects are felt the greatest by the weakest and poorest members of our society (Western and Pettit, 2010).

The Unanticipated International Crime Decline As the frenzy of policy to scare Americans into believing there was an inevitable crime wave continued, unbeknownst to most, crime was actually on the decrease. In fact, crime has declined steadily and somewhat dramatically since the early 1990s. According to the Uniform Crime Report, both the violent crime and property crime rates fell 48% between 1993 and 2016. As Zimring points out (2007) this “Great American Crime Decline” was not limited to one area or city, to one demographic or even to America. That is, it was replicated in Canada, our neighbor, which enacted none of the get tough policies implemented in America. Also, it was not linked with any sudden change in the fabric of American society, as many sociologists would hope, such as a sudden down swing in poverty, upswing in employment or decrease in prejudice (Zimring, 2007). Rather, it continued without explanation.What this decline does tell us, though, is that there are forces at large, outside the criminal justice system and its direct policies that cause, perpetuate and potentially ameliorate crime. While we do not offer here an explanation of the decline above and beyond what has been offered in prior studies, we note that it has opened the door for returned thought to “what works” and a look at empirical science (Sherman et al., 1997; MacKenzie, 2006).

Disparities in Outcomes As previously discussed, while conservatives were interested in becoming tougher on crime, liberals were initially concerned that the prior system of indeterminate sentencing and vast judicial discretion had led to disparate outcomes, particularly for minority members of society. Combined, these concerns led to the myriad of practices noted above, including mandatory sentencing laws, truth in sentencing practices, and sentencing guidelines, all of which, in some way, reduce judicial discretion. For liberals, the focus of early attacks on sentencing was on both reducing disparity and reducing prison populations (see, for example, American Friends Service Committee, 1971; von Hirsch, 1974). However, the liberal cause transitioned toward getting on board the get tough on crime movement as voters’ fears grew with increasing crime rates and policies such as sentencing guidelines, that were initially envisioned to reduce differences at sentencing, working in many instances instead to increase overall punishment severity. While little research exists on the impact of truth in sentencing laws on increasing or decreasing disparity, much research implicates both mandatory sentencing, and certain aspects of either federal or state sentencing guidelines in either increasing, or at least, failing to reduce disparities. Concerns were raised about the disproportionate representation of minorities (particularly African-​ Americans) in America’s prison system, and the extent to which minority overrepresentation resulted from disparate decision-​making by system actors such as police (Kochel, Wilson, & Mastrofski, 2011), prosecutors, judges, and parole boards (Huebner & Bynum, 2008). For example, one early study on the disproportional representation of minorities in America’s prison population found that while a large proportion of the disparity (80%) could be traced to arrests, there were still significant biases at sentencing that led to African-​Americans being over-​represented in the nation’s prison population (Blumstein, 1983). A follow-​up report produced just a decade later, but after the Anti-​Drug Abuse Act of 1986, found that not only did this over-​representation still exist, but that it was greatly magnified in one specific area—​drug offenses (Blumstein, 1993). According to the study’s author, Alfred Blumstein: The striking new issue in 1991 is the saliency of drug offenders in prison. Blacks comprise 57.7% of the prisoners for drug offenses, but they are only 40.4% of the arrestees for drug offenses, so that they are over-​represented in prison by forty-​three% compared to arrest (p. XX). 27

Megan C. Kurlychek and John H. Kramer

The fact that this disparity still existed and was increasing, and that it was particularly tied to drug offenses, led many to blame this disparity on America’s war on drugs and its mandatory and enhanced sentences for drug offenders (Mauer, 2006; Lynch & Sabol, 2000). Indeed Tonry (1995) went as far as to proclaim that conservative officials from the Reagan and Bush administration knew that these policies would disproportionately impact poor and minority communities and, thus, should be held accountable for the damage caused by these policies and others declared these new laws “The New Jim Crow” (Alexander, 2010). In this last statement, Alexander is primarily referring to mandatory sentencing laws when speaking of the new Jim Crow as these laws mainly targeted drug offenders. For example, Myers (1989) examined a sentencing reform law in Georgia that increased penalties for drug trafficking and reduced judicial discretion in the application of these sentences. While the author anticipated that the reduced judicial discretion would reduce disparity, the opposite was found with blacks being more likely than whites to be incarcerated for drug trafficking (25-​percentage point difference), drug distribution (19-​percentage point difference), and even more minor drug use (12-​percentage point difference). Moreover, some studies suggest that even the application of mandatory minimums can be disproportionately felt by minorities. That is, in most jurisdictions a “mandatory” is only so if requested by the prosecutor. In a study of the application of mandatory sentences in Pennsylvania, Ulmer, Kurlychek, and Kramer (2007) found that the decision to apply the mandatory was complex and was related not only to legal factors such as prior record and offense seriousness, but also to extra-​legal factors such as gender and race. The largest negative impact was suffered by young Hispanic males subject to a drug offense mandatory sentence. In addition to mandatory sentencing, sentencing disparities have been reported even within jurisdictions operating sentencing guideline systems. Perhaps, these disparities have been most greatly noted in the US sentencing guidelines system. One early study of federal sentencing guidelines (Albonetti, 1997) found that not only did racial differences still exist in the sentencing of drug offenders, but that it was tied to prosecution decisions and the magnitude of plea reductions offered. This latter point resonates with the role of prosecutors noted in the Ulmer study above and is a theme to which we will later return. Later studies continued to report ongoing disparities. Mustard (2001) found that blacks, Hispanics, and males were less likely to receive downward departures from the federal guidelines and smaller discounts. Steffensmeier and Demuth (2000) found Hispanics to be particularly disadvantaged compared to black or white drug offenders in receiving substantial assistance departures. Since this time many studies have reported similar findings (Everett & Wojtkiewicz, 2002; Mustard, 2001; Johnson, Ulmer, & Kramer, 2008) and find that disparities exist not only for black defendants but also, and perhaps, even more significantly for Hispanic drug offenders (Steffensmeier & Demuth, 2000). Most recently the United States Sentencing Commission (USSC) itself issued a report that reveals that these disparities may even be increasing post-​Booker. In 2006, the USSC reported that black male offenders received sentences that were 15.2% longer than those imposed on white males, however, in more recent years this difference increased to 23.3% longer. Hispanic males were also found to receive longer sentences than white males in recent years, with a difference of 6.8%. Similar to the earlier reports, non-​citizens also received sentences that were longer than those imposed on citizens by 11.5% (USSC, 2012). Thus this quasi-​experiment with the federal guidelines changing from highly presumptive (often referred to as “mandatory”) to “advisory” appears to have increased federal sentencing disparity (see, Ulmer, Light, & Kramer, 2011 for an alternative view and there are others). The findings at the state level regarding sentencing guidelines and disparity are more mixed. Although 16 states and the federal courts have sentencing guidelines, most of the studies have been done either on the federal guidelines or in the states of Minnesota and Pennsylvania. Early 28

Sentencing in the 21st Century

examinations of the Minnesota guidelines found that while at first they did serve to reduce disparities (Miethe & Moore, 1986; Moore & Miethe, 1986), this impact later greatly diminished. In fact, one study suggested that the guidelines exacerbated rather than resolved any problems of disparity, concluding that: The imposition of the Minnesota Sentencing Guidelines on the judiciary has not succeeded in carrying out many of its stated objectives. It seems to have exacerbated the importance of many impermissible criteria in sentencing decisions in such a way as to work in favor of the socially and politically advantaged while harming the socially and politically disadvantaged. (Nelson, 1991, p. 251) Studies of Pennsylvania’s state guidelines that provide for a greater level of judicial discretion have also found extra-​legal factors such as race, age, and gender to have direct and indirect impacts on sentencing that disadvantage, particularly, young black males (Johnson, 2003; Kramer & Ulmer, 1996; Steffensmeier, Ulmer & Kramer, 1998). However, there are two important points here to be considered. First, these studies do not indicate, per se, that the guidelines are the problem as no counterfactual is available to determine what level of disparity might exist if the guidelines were not in play. Rather, they suggest that guidelines are perhaps a less than perfect mechanism for eliminating disparity. Indeed, although disparities have been found, much has been attributed to departures from recommended guideline sentences (Engen et al., 2003; Johnson, 2003, 2005). Another controversy surrounding guidelines is the difference between presumptive and voluntary guidelines, with the first assuming the judge should or must follow the guideline recommendation and the latter providing the guidelines as advisory only to help judge’s determine appropriate sentences. Wang et al. (2013) compared voluntary and presumptive guideline systems and found no racial disparity in the imprisonment decision under either guideline system, but there was racial disparity in non-​guideline jurisdictions. However, in their examination of the imprisonment decision by ethnicity,Wang et al. (2013) did find a difference between voluntary and presumptive systems with the presumptive system reducing disparity and the voluntary system producing similar levels of disparity to non-​guideline states. Thus, neither mandatory sentences nor sentencing guidelines proved a panacea for reducing disparity. One final point we would like to return to here is the role of the prosecutor as noted in the discussions of mandatory sentencing and guidelines. While our focus here is, and will remain, on sentencing, one cannot address this without acknowledging that the judicial sentence is only the last step in a much longer process. The prosecutor plays a particularly looming role in this process, deciding upon charges, requesting pretrial detention, and negotiating pleas—​all of which have significant impact on the final sentence (Hagan, 1975; Kutateladze et al., 2014; Stolzenberg et al., 2013; Schlesinger, 2005; Spohn, 2008; Heaton, Mayson, & Stevenson, 2017; Wooldredge et al., 2015). Moreover, guidelines arm the prosecutor with charging power that can pressure negotiated pleas (Frase, 2005; Reitz, 1998; Stith & Cabranes, 1998; Wright, 2005). Thus, as we continue to advance our discussion of directions in sentencing, it will include, as appropriate, references to prosecutorial discretion as well.

Four Signs of an Emerging Policy Shift With the sentencing reforms of the past not achieving their desired goals of reduced disparities and the rising costs (financial and human) of mass incarceration becoming more and more apparent, there are signs that the reliance on “getting tough on crime” as the most effective and efficient approach is changing. Conservatives and liberals are both indicating that we have over spent on locking people up and that we have created a system that uses our resources neither wisely nor effectively. What follows is a discussion of what we perceive as the most salient signs of change. 29

Megan C. Kurlychek and John H. Kramer

Repealing Mandatory Sentencing Laws Perhaps less related to the crime drop than to a realization that mass incarceration is too costly to sustain, we have witnessed a recent effort to repeal and redirect mandatory sentencing policy. Based on estimates of the actual financial cost to keep many low-​level offenders in prison (Chris & Sbramanian, 2017), and growing outcries about the human suffering and social costs of imprisonment (Western & Pettit, 2010), many states have been revisiting their mandatory minimum laws. According to the National Conference of State Legislatures (2016), this trend included New York’s repeal of its original “Rockefeller Drug Laws,”9 some of the strictest in the nation, as well as legislation in Minnesota (2016) that also removed mandatory minimums for low-​level drug offenders, legislation in Georgia that allowed for departures from mandatory minimums for drug trafficking offenses, and in Maryland (2015) that authorized departure from mandatory minimums for repeat offenders if not necessary for public protection. In 2015, North Dakota and Oklahoma also passed legislation that would allow for departures from mandatory minimum sentences when they did not serve a public protection need and when the application of the mandatory sentence would result in “manifest” (North Dakota) or “substantial” (Oklahoma) injustice to the defendant, thereby seemingly outwardly admitting to the inherent bias in these policies (National Conference of State Legislatures, 2016). Perhaps one of the key accomplishments of this counter movement was the 2010 passing of the Fair Sentencing Act, in which the crack-​to-​powder cocaine ratio was reduced from 100:1 to 18:1 by establishing the five-​year mandatory minimum for crack cocaine to be triggered by 28 grams and increasing the amount to trigger a 10-​year mandatory minimum from 50 grams to 280 grams (Public Law 111, 220). The Fair Sentencing Act was enacted more than 15 years after the first proposal made by the United States Sentencing Commission to eliminate the penalty disparity between crack and powder cocaine, on the grounds that there were very small differences between the two forms of the drug. This first proposal was quickly rejected by Congress. In 1997, the USSC provided Congress with a range of options that would have reduced the ratio; the various proposed ratios ranged from 2:1 to 15:1. But, again, Congress rejected any reduction in the penalty disparity. There were several other attempts to reduce the disparity. Senator Jeff Sessions in 2001 introduced The Drug Sentencing Reform Act that would have raised the five-​year mandatory minimum for crack to take effect at 20 grams rather than 5 grams and lowered the five-​year mandatory minimum for powder to 400 grams from 500 grams. This would have reduced the 100:1 ratio to 20:1. Senator Orin Hatch sponsored the Fairness in Drug Sentencing Act of 2007 (S1685), which was intended to reduce the disparity by setting the five-​year mandatory minimum for crack at 25 grams and leaving the five-​year mandatory minimum for powder at 500 grams. Both Senators Sessions and Hatch were conservatives who had fought earlier reductions in the 100:1 ratio, but by the early part of this century they took significant steps to recognize the unfairness of the cocaine mandatory minimums. The importance of the Fair Sentencing Act for sentencing in the 21st century should not be overlooked. One of its key sponsors was Senator, and recent Attorney General, Jeff Sessions. The fact that conservatives such as Senators Hatch and Sessions proposed significant reductions to the mandatory minimums for cocaine signals a significant transformation from their strong rejections of the USSC proposals for change in 1995 and 1997. While it is not clear whether their support came from concerns about the costs of confinement, unwarranted disparity, or some other issue, it is important that such leading conservatives on crime and sentencing would change course so dramatically.

Support from the Supreme Court The Supreme Court also began tackling issues of mandatory sentencing including sentencing guidelines that mandated consideration of factors other than those used in computing the standard-​ range guideline sentence in a series of state-​specific cases in the early 2000s. The Supreme Court 30

Sentencing in the 21st Century

considered the “mandatory” (presumptive) nature of some sentencing guideline systems such as the Washington Guideline’s in Blakely v. Washington (542 US 296, 2004). The Blakely decision essentially established the upper guideline range in Washington as basically a statutory maximum sentence, thus limiting judges to facts pled to by the defendant or found in a jury verdict. The implication of the Blakely decision was that presumptive guideline systems such as US sentencing guidelines were in constitutional jeopardy. In United States v. Booker (543 US 220, 2005) the Court extended the Blakely decision to the federal guidelines. In Booker the Court held in a 5–​4 decision that any fact, other than a prior conviction, that is required to support a judge handing down a sentence that exceeds the otherwise maximum sentence for an offense must be either admitted by the defendant or proven to a jury beyond a reasonable doubt. A similar decision was reached in Cunningham v. California (549 US 270, 2007), in which the Court struck down California’s Determinate Sentencing Law on the grounds that it violated the defendant’s Sixth Amendment right to a jury trial. This decision was based on the clause in the law that allowed the judge to impose enhanced sentences based on facts concluded by the judge, but not found by the jury. Following the passage of the Fair Sentencing Act (2010), the Court also ruled that Congress intended this reduction in disparities between crack and powder cocaine at sentencing to be applied retroactively. In Freeman v. United States (131 S.  Ct. 2685, 2011)  the court ruled by plurality that Freeman who had been sentenced under the guidelines three years prior to a retroactive guideline amendment that reduced Freeman’s applicable guideline from 46 to 57 to 37 to 46  months was eligible for a sentence reduction as a result of a retroactively applicable guideline amendment. The District Court and the Sixth Circuit had ruled that the district judge in this case did not have to reconsider the sentence when the sentence was pursuant to a plea agreement (Rule11©(1)©). However, the US Supreme Court found that the sentencing judge must consider the applicable guideline that had been revised and had been made retroactive to this case. In the following year the Court in Dorsey v. United States (132 S.Ct. 2321, 2012) further ruled that the reduced sentencing disparity provided by the Fair Sentencing Act applied to offenders who committed their offense prior to the change, but were sentenced after the change. In addition to mandatory sentencing, the Court also began to tackle issues of rehabilitation and individualized justice. For example, in Pepper v. United States (131 S. Ct. 1229, 2011), the Court ruled 7–​1 that when a sentence was set aside on appeal, at resentencing the court may consider evidence of the defendant’s rehabilitation and that evidence of rehabilitation may be used to support a downward variance from the federal sentencing guidelines. Another sign of the changing times was witnessed by Supreme Court decisions that began to place some limits on sentencing and in particular on the death penalty. The Court first limited the use of the death penalty to only offenses that involve murder or crimes against the state (Coker v. Georgia 433 US 584, 1977). However, one of the more recent developments that we note as perhaps not just a sign of a reduction in punitiveness but also a return to the idea of rehabilitation was the 2005 decision in Roper v. Simmons (543 S. CT. .551, 2005). This decision specifically declared the use of the death penalty on anyone who was under the age of 18 at the time of their offense to be cruel and unusual punishment and, therefore, unconstitutional. Interestingly, this decision cited heavily social science findings regarding both adolescent psychological development, peer influence, and decision-​making, and the ability of a youthful offender to change as he or she matured. Indeed, in the end the justices ruled that because the individual was not yet done developing, he or she could not be determined to be “the worst of the worst,” which is the population for whom this penalty is reserved. While this decision did not impact the use of the death penalty on adults, we believe it clearly marks a return to the consideration of individual factors at sentencing. In this case the factor to be considered was the age of the individual as it directly impacts determinations of culpability and blameworthiness. It also has a hint of rehabilitation in its voice as it suggests that just because the individual commits a horrid act as a teenager he or she is not doomed to a life of crime, but rather, because he or she is not done developing a more positive and law-​abiding adult 31

Megan C. Kurlychek and John H. Kramer

could emerge. This decision was followed closely by the Supreme Court rulings Graham v. Florida and Miller v. Alabama that further restricted the sentencing of youth in adult courts. This time the topic was life without parole (LWOP). Graham v. Florida (560 S. CT. 48, 2010) used similar logic as the Simmons’ decision to ban life without parole for all non-​homicide offenses committed by juveniles which was closely followed by the Miller v. Alabama decision (567 S. Ct. 460, 2012) that banned the use of “mandatory” LWOP even for homicide offenses and required individualized assessments of culpability. Overall, these decisions indicate that the Supreme Court is able and willing to tackle key issues of disparity as they relate to mandatory sentences and sentencing guidelines, to address key disparities that violate citizen’s rights, and to rely on empirical science in such determinations. Also, and of key importance for our following arguments, the latter series of decisions also notes a possible return to individualized justice and a recognition of an individual’s ability to change—​or dare we say—​be rehabilitated.

The Rebirth of Rehabilitation While some declared rehabilitation “dead” after the Martinson report, others have been working to resuscitate it for the past 30 years. Even as support for treatment was waning, some were fighting to keep it alive. First, perhaps, was Ted Palmer’s work that systematically addressed and rebutted the original findings of Martinson reviewing Martinson’s own data and finding that about half of the programs did indeed work (Palmer, 1975, 1978). That is, is the cup half empty or half full? Cullen and Gilbert shortly thereafter joined the fray with their work Reaffirming Rehabilitation (Cullen & Gilbert, 1982). In this piece the authors duly note that the abandonment of the indeterminate sentence and rehabilitation model was driven by a fear of state power and misuse of discretion. However, ironically, it left the state with an even greater power of punishment against which the authors warned of dire consequences. Other major pushes for rehabilitation came from scholars such as Andrews, Bonta and Hoge (Andrews, Bonta, & Hoge, 1990; Bonta, 1996) and Latessa (2004). Andrews and colleagues would argue that the true reality was not that nothing worked, but that some things worked for some people. This notion lead to the evolution of more standardized risk, needs, and responsivity (RNR) assessment tools in probation and correctional programming. Latessa not only worked to assess the validity of such tools, but to evaluate their use in practice and to promote the notion of fidelity to the treatment model (Latessa, 2004). Thus, it was not rehabilitation per se that did not work, but the lack of funding and true implementation that may have led to its demise. Since this time, Latessa has devoted much of his career not only to proving the possibility of effective treatment programming, but working directly with practitioners to bring science into practice. The importance of these and other scholars is more fully spelled out by Frank Cullen in his address to the American Criminology Association in 2004 (Cullen, 2005).10 In addition to these scholars we would like to add Larry Sherman to the list and bring attention to the important report that he and a team of colleagues produced for the National Institute of Justice—​“Preventing Crime”: What Works, What Doesn’t and What’s Promising (Sherman et al., 1998). What is so important about this report is that while recognizing that not everything works and some programs might be better off if abandoned (e.g., scared straight), it is not necessary to throw the baby out with the bathwater! That is, many programs do work and show potential for even greater success. Our cup is indeed at least half full. While much of this discussion revolves around prevention and correctional programming, it is important to realize that our faith in indeterminate sentencing inherently rests upon faith in the system to do more than just punish. Second, some of the programs noted as promising in the works noted above directly involve court practices. In fact, we believe the Treatment Court movement to be a potential key force in the return of rehabilitation to the corrections and sentencing philosophies. 32

Sentencing in the 21st Century

The Promise of the Treatment Court Problem-​solving courts began as drug courts in 1989 when Dade County, Florida initiated the first such court (Chriss, 2002). Problem-​solving court is the label currently being used to describe such courts but other terms such as treatment courts and specialty courts have also been used to identify them. Problem-​solving courts (PSCs) are an important movement because they focus on local communities identifying particular crime problems facing the community and because they marshal local community resources to address the problem. Miller and Johnson (2009) identify five key components to PSC: 1) individualizing treatment such that each offender’s needs are assessed and resources to address those needs are provided; 2) merging and coordinating both public and non-​profit organizations to service the client; 3) sharing information to enhance decision-​making; 4) holding clients and service providers accountable through judicial monitoring; and 5) focusing on outcomes to monitor success of the PSC. Thus, PSCs incorporate the RNR components we described earlier which have been proven, when properly implemented, to be successful in reducing recidivism and increasing public safety (Gottfredson & Exum, 2002; MacKenzie, 2006; Taxman, 2002). Further, the problem-​solving court philosophy requires significant planning and a team or workgroup approach to case management. This is crucial as not only are guidelines set forth as to how the court will function and who will be eligible to participate but, perhaps more importantly, it brings together a multi-​disciplinary team made up of the judge, treatment specialists, probation officers, defense attorneys, and district attorneys—​e.g., the courtroom workgroup (Eisenstein & Jacobs, 1977). By agreeing to participate and participating in the problem-​solving court process, the district attorney is implicitly shifting his/​her focus from retribution and/​or deterrence to interceding and protecting the community through treatment/​rehabilitation. Participation involves working with and listening to input from not just criminal justice members, but from treatment needs assessment experts and treatment providers. Most importantly, problem-​solving courts identify a serious problem to the community that has not been addressed by the traditional punishment system and provide an alternative approach to protecting the community by client needs assessment, treatment focused on client needs, and closer supervision with more certain and timely sanctions. We use the term “problem-​solving court movement” to reflect the length of time that such courts have been on the criminal justice landscape, but mostly to capture the expansion in both numbers and types of problem-​solving courts. Problem-​solving courts are entering their fourth decade of development and there is growing support for them at all levels of government. What began as a drug court has expanded to focus also on special offenses including DUI, domestic violence, as well as special types of offenders, including homelessness, mental health, and veterans. The near future promises further expansion with continuing support from the Bureau of Justice Assistance. From one court in 1989, the most recent estimate we can find is that this movement has grown to 3,142 drug courts as of 2015 with 1,558 being adult drug courts (National Institute of Justice, 2018). Moreover, with the devastating effects of the opioid crisis in the past several years, we anticipate that there will be additional courts to assist in addressing the tremendous human costs created by opioids. With 72,306 overdose deaths in the United States in 2017, which is almost 10,000 more than reported in 2016 (National Institute on Drug Abuse, 2018), the drug crisis has expanded to every corner of the country. Further, problem-​solving courts link the community into the treatment and reentry of the offender; thereby, they can serve to educate the public and the local criminal justice community that crime is a local problem and the response to it is most effectively addressed by the community. If problem-​solving courts serve to invigorate local justice systems to reduce their reliance on incarceration, we think the problem-​solving court movement will serve as a foundation for the rediscovery of rehabilitation and individualized justice that is stronger and more evidence based than the model that served this nation from the late 1800s to the late 1900s. 33

Megan C. Kurlychek and John H. Kramer

Justice Reinvestment A final initiative we would like to briefly highlight is justice reinvestment as we believe this initiative is most directly related to the backlash against the nation’s mass incarceration rate and its cost to taxpayers. Justice reinvestment is a term that was coined in 2003 by Tucker and Cadora relating to “redirect some portion of the $54 billion America now spends on prisons to rebuilding the human resources and physical infrastructure—​the schools, healthcare facilities, parks, and public spaces—​of neighborhoods devastated by high levels of incarceration” (Tucker & Cadora, 2003, p. 3). It has since been interpreted by many states however as a “data-​driven” approach to corrections policy that seeks to reduce the excessive costs and destructive forces of mass incarceration by using proven community-​ based strategies to reduce crime and then “reinvesting” those savings back into the system to help support additional community-​based programs (Austin et  al., 2013). While some of these policies may be rehabilitative, such as the treatment courts noted above, they may also be designed more specifically to hold offenders accountable through restorative justice practices (James & Agha, 2013). Overall, justice reinvestment activities fall into three primary categories: alternatives to pretrial jail/​ detention, alternative-​/​community-​based sentences, and the sustainability of programs (hence, the reinvestment). The expansion of such alternative sentencing practices increases the options available to judges, thus increasing the ability to help offenders and keep society safe without the use of mass incarceration. While federal legislation has been introduced, this initiative has mainly taken shape at the state level with 35 states implementing some form of justice reinvestment policy since 2007 (Pew, 2018). Perhaps the most remarkable example of the potential impact of the justice reinvestment movement took place this summer in Louisiana. Louisiana has the highest incarceration rate in the country and during the summer of 2018 Louisiana passed legislation to reduce incarceration and use the savings to reinvest in other programs, including those for victims. Specifically, the legislation that is designed to reduce the prison population by 10% in ten years and to save the state $262 million will reduce drug and theft penalties and eliminate some mandatory minimum sentences and reduce habitual offender penalties. To date, however, we must admit that the future of justice reinvestment is questionable as it is newly implemented in most areas and seems to take a variety of shapes and forms which make empirical assessments difficult. Critics are currently concerned that these policies might lead to increased disparities and that any saving may not be redirected back into either the system or communities as intended (Austin et al., 2013). However, we chose to include it here as a “sign” that the nation is looking for ways to move away from an over-​reliance on get tough and mass incarceration policies.

Cause for Pause: Countervailing Forces and the Current Administration As previously mentioned, the politicization of crime is of grave concern because it has the potential to steer policy away from empirical science and towards rhetoric and politically savvy sound bites. For example, the recent killing of college student Molly Tibbetts by an individual who was in the country illegally plays neatly into the hands of the Trump Administration’s campaign of fear regarding illegal aliens. The representation of Mexicans as rapists and murderers creates false stereotypes based on isolated cases. By doing so, the administration becomes moral entrepreneurs labeling an entire population as “outsiders” and threats to law and order. To the degree that such stereotypes become a part of the community cultures in the United States, they can initiate targeting and unwarranted disparity in our handling of members of the minority. One only has to briefly review some of the propaganda from the Trump campaign referring to Mexicans as rapists, or comparing immigrants to “animals” to have fear—​not of the criminal—​but of those in charge of crime control.We believe three key features of the current administration that go hand-​in-​hand with the past politicization of crime also hold the potential to derail the recent progress in sentencing reform. 34

Sentencing in the 21st Century

First is the movement away from science and towards further politicization of the judiciary to serve executive priorities. President Trump has already begun to blame the small upswing in crime in the past two years on the “lenient” policies of the Obama administration, with a special focus on a decline in drug prosecutions. In particular, President Trump tweeted that federal drug prosecutions had decreased from 2011 to 2016 and claimed that this decline was responsible for the opioid epidemic. However, while it is a fact that drug prosecutions decreased, there is no evidence that it was in the area of opioids, or particularly the prescription opioids responsible for much of the current overdose and addiction crisis. Indeed, many more informed policy makers would argue that the decrease in prosecutions was a good thing. Similar to the fashion in which Reagan was able to use the crack cocaine crisis to invoke fear, it appears the current administration hopes to use the opioid epidemic. In addition, with terrorism on the rise internationally, the current administration is aiming to create a new population of suspect others who, being of certain religions and national origins, provide an easy scapegoat for the crime problem. For example, the scientific literature shows that immigrant populations are less criminal than citizens (Burrington, 2018) and that cities with a large influx of immigrants do not suffer from escalating crime rates (Martínez-​Schuldt & Martínez, 2017). However, one might not know this fact if relying only on media reports and political rhetoric rather than empirical fact (Ousey & Kubrin, 2018). This leads us to our second concern with the current administration, which builds from one of the most troubling aspects of the prior mandatory sentencing and war on drugs policies. In addition to blatantly targeting certain suspect crimes and classes of criminals, we have shown how these policies have institutionalized forms of racism into the system. For example, the war on drugs that resulted in many of the mandatory minimums focused on relatively low-​level dealers rather than the high-​ end traffickers (Mauer, 2006; Wilson, 1978). As an example of this rhetoric, President Trump faults loose border patrol for Mexican drug dealers and “mules” being able to transport opioids into the country while little to no national attention is being directed at American M.D.s and pharmaceutical representatives despite the fact that research suggests the medical overprescribing of opioids is a large portion of the current crisis (Stuart & Joshi, 2018). Finally, and perhaps of greatest consequence for future sentencing policy, is the President’s power to appoint officials. For example, President Trump appointed Jeff Sessions as Attorney General of the United States, positioning him in charge of all law enforcement for the nation. Although this post was short lived, it should be remembered that Sessions was an avid promoter of mandatory sentencing and during his short stay in this post overturned many of the policies of the prior administration. In specific he overturned a memo delivered by Eric Holder that sought to curb mass incarceration via avoiding mandatory sentencing, by instead ordering federal prosecutors to begin seeking the most serious charges possible and to invoke mandatory sentencing. We must also note the President’s ability to appoint members to the nation’s Supreme Court who serve, unlike Sessions, not only while the current President is in office, but until their own retirement or death. Almost all of the previous Supreme Court decisions noted that turned this nation towards rehabilitation and fair sentencing practices hinged on a small majority, often with Justice Kennedy representing the swing vote.With Justice Kennedy announcing his retirement in the summer of 2018, the new appointee, Judge Kavanaugh, may therefore hold great power in directing the nation’s courts if any of these previous narrow decisions are challenged and revisited by the Court.

Sentencing in the New Era We began this chapter with a brief reminder that major shifts in sentencing are not new to the last 50 years but have been an integral part of our history. In fact, during the relatively brief existence of the United States we have experienced significant changes in our sentencing philosophy and procedures about every 100 years. The most recent shift, starting in the 1970s, has led us into a system reliant upon mass incarceration to the point that our nation accounts for upwards of 25% of the world’s prisoners despite being only 5% of its population (Garland, 2001). With the plethora of 35

Megan C. Kurlychek and John H. Kramer

literature dispelling the myth of mass incarceration as a solution to our problem and the reliance on legislative policies and sentencing commissions as a remedy for disparity, the question we are left to answer is “Where do we go from here?” We see some positive signs that some of draconian measures adopted primarily during the 1970s to mid-​1990s (Tonry, 2016) may be losing favor with policymakers. Michael Tonry refers to the period since the mid-​1990s as an “equilibriums period” that might lead one to believe that not much has changed. However, as we point out above, much has been changing in the 21st century and we see significant signs of a potential turning point with the repeal of mandatory minimum sentencing statutes, the return of rehabilitation through treatment courts and the support of the Supreme Court for a more restricted use of incarceration. Yet on the heels of this change, we also must acknowledge the re-​emergence of punitive trends under the current administration. These countervailing forces promote stereotypes to reinforce fear of offenders and victimization—​particularly those not of white Anglo-​Saxon background—​and, even more specifically, immigrants. President Trump has characterized illegal immigrants as “rapists and murderers,” drawing examples from extraordinary events to characterize the population consistent with Garland’s elaboration of the methods for the politicization of crime.These stereotypes have served as the foundation to develop policies to remove and exclude these people from the United States and to exclude scientists, criminologists, and, perhaps, common sense, from the table. Thus, we face a quandary as to the likely path sentencing will take in the future. Will it be the path more traveled—​back to harsh sentencing and ill-​informed policy—​or might it be the path less traveled? In the short term we see hope in movement away from mandatory sentencing practices, particularly for low-​level drug offenders, as well as in the growth of treatment or specialty courts and moves by the Council of State Governments, the Pew Center on the States, and the Department of Justice promoting justice reinvestment (Gottschalk, 2015). We do not see any of these initiatives as changing the face of sentencing in and of themselves, but combined, they provide positive signs that we may indeed be on a move away from mass incarceration. We find optimism in state legislative changes and local court community initiatives, particularly the treatment court movement, as a significant reframing of the primary purpose of sentencing for vulnerable offenders. The fact that these courts bring a “treatment” team to the sentencing table may, and we think will, set the stage for greater realization in the court community that imprisonment is an ineffective short-​term result and an endangering outcome for returning offenders. Thus, as long as drug abuse and mental illness are left to the local criminal justice system to address, we believe rehabilitation will grow as a mechanism to help the offender and protect the community. However, we would be naïve to assume such rehabilitation and care will be afforded to all. That is to say we predict “justice” for some and the continuation of disparities at all points of the criminal justice system, including sentencing. Just as the original juvenile court excluded African-​Americans from its care and rehabilitation, we fear the current rehabilitation and reform efforts in the adult court will be colored by fear and misinformation. Moreover, some will certainly claim that we are amiss to have overlooked the importance of the prosecutor as a resisting factor to shifts in sentencing (see, for example, Pfaff, 2017). As discussed above, there is no question that the prosecutor has been empowered by mandatory minimums and sentencing guidelines (Ulmer, Kurlychek, & Kramer, 2007). Further, it is primarily in prosecutors’ offices that the role of the victim has been strengthened. However, there are significant indications that the prosecution side of the adversary system may also be recognizing the failure of imprisonment as the primary protector of the community. Important changes are occurring that suggest that the traditional prosecutor model as case processor with success measured by conviction rates is moving toward a “community prosecutor” (Worrall & Wheeler, 2018). Under the community prosecution approach, the prosecutor’s role shifts to focusing on making communities safer and receiving input from the community they are elected or appointed to protect (Nugent-​Borakove & Fanflik, 2008; Worrall & Wheeler, 2018).This role of the prosecutor is evidenced by their participation in treatment courts. 36

Sentencing in the 21st Century

Thus, rather than looking to Congress and the current executive administration to guide sentencing practice, we propose here that the future may indeed be a grassroots, or bottom-​up, approach.We think that the change will likely come from the local court communities and filter up to the legislature (Tonry, 2016; Gottschalk, 2015; Pfaff, 2017). In many respects justice is a local phenomenon and even when there are mandatory minimums, they are often ignored or avoided in order to maintain local sentencing standards (Ulmer, Kurlychek, & Kramer, 2007). It is these court communities that set the standards of justice in the 1900s until the sentencing revolution in the 1970s took place and we believe it is these same court communities that will play a major role in the future of sentencing.

Conclusions and Directions It seemed in the new millennium, with crime on the decline and the nation feeling the aftermath of its exercise in mass incarceration, we were perhaps on the brink of a new era in which empirical scientific evidence would be used more readily to inform sentencing, and criminal justice practices in general.Yet, despite empirical evidence that mandatory sentencing is not the fix to our problems, it is still a popular crime control strategy and one that is supported by many members of the current administration. As discussed in this chapter, there are many mandatory minimums still in place. In fact, more than one-​fifth of federal offenders sentenced in 2016 (21.9%) were convicted of an offense carrying a mandatory minimum penalty and overall, 55.7% of federal inmates currently serving their sentence were convicted of an offense that carried a mandatory minimum penalty. In addition, as discussed, the President has the power to appoint many other officials that directly or indirectly impact criminal justice and sentencing, not the least of which include Supreme Court Justices, the nation’s Attorney General and members of the United States Sentencing Commission. With the continued and even increased politicization of criminal justice and sentencing, we predict that these appointments will be tied to a more conservative political regime than what we have experienced in the recent past and that harken back to the era of the 1980s. But we have learned much since the 1980s, most as a result of the failed experiment in mass incarceration. While many are skeptical that social science research can truly influence policy, it is important to note that such research was influential at the highest level in the recent Supreme Court decisions and that local court communities strive to find alternatives to incarceration, particularly through treatment, and to seek scientific evidence of the effectiveness of such treatment as evidenced by the work of scholars such as Sherman and Latessa. Thus, unlike the early rehabilitation efforts of the 1900s, today’s programming can learn from and build upon its own successes and failures. Our overarching conclusion is then that we must not get caught into short-​term politics as any significant change requires steps forward and backward. Research and information are tough and recalcitrant and while set aside for a period, we do not believe they can be suppressed in the long run. Further, when the dishonesty and failures created by those who ignore the truth becomes visible then, as it has in the past, the country will marshal itself to be and do better. Scholars, as the conveyers of science, can and should seek to shape legislative and judicial policy not only at the highest level, but directly through work with our local court communities where the rubber ultimately meets the road.

Notes 1 The indeterminate sentence viewed defendants’ treatment needs at sentencing as generally unknown and beyond the capability of the judge to adequately assess because treatment assessment was not a skill set of the judiciary and because the treatment is a process to which offenders respond differently and the length of time that treatment is needed must be assessed continuously to determine when the offender is release ready. 2 In some states this is awarded to all victims, in others only to victims of violent crimes or serious felonies. 3 Seven of the 16 states operating sentencing commissions require a victim representative to sit on the commission.

37

Megan C. Kurlychek and John H. Kramer 4 In 1961 the “baby boomers” would enter their teenage crime-​prone years thus altering the age composition of American Society. Some studies have attributed much of the crime boom of the 1960s to the “baby boom” of the 1940s (Wellford, 1973). 5 See 1790 Crimes Act §§ 2, 5, 6, 7, 11, 12, 13, 15, 18, 23, 26, and 28. 6 13 Documentary History of the First Federal Congress of the United States of America, Debates in the House of Representatives 968–​74 (Helen E. Veit et al. eds, 1994). 7 See Pub Law No. 82–​255 §1,65 Stat 767 (1951) and Narcotics Control Act of 1956 § 103,105,107 & 108. 8 See Omnibus Crime Control Act of 1970 Pub L. No. 91–​644 Title II §13, 84 Stat. 1880, 1889–​90 (1971). 9 On December 14, 2004, New York Governor George Pataki signed into law the Drug Law Reform Act (DLRA) (2004 N.Y. Laws Ch. 738 (effective January 13, 2005)), which replaced the indeterminate sentencing scheme of the Rockefeller Drug Laws with a determinate system and under Governor Paterson In April 2009, the New York Penal Law and the New York Criminal Procedure Law were revised to remove the mandatory minimum sentences for drug offenses all together. 10 Scholars attributed with the rebirth of rehabilitation include Don Andrews, James Bonta, Frank Cullen, Paul Gendreau, Karen Gilbert, Scott Henggeler, Edward Latessa, Mark Lipsey, Doris, McKenzie, Ted Palmer, Joan Peterseilia and Patricia Van Voorhis.

References Albonetti, C. A. (1997). Sentencing under the federal sentencing guidelines: Effects of defendant characteristics, guilty pleas, and departures on sentence outcomes for drug offenses, 1991–​1992. Law and Society Review, 31(4), 789–​822. Alexander, Michelle (2010). The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New  York: The New Press. American Civil Liberties Union. (2018). Mass Incarceration. New York: Author. American Friends Service Committee. (1971). Struggle for Justice. New York: Hill & Wang. Anderson, D. C. (1995). Crime and the Politics of Hysteria: How the Willie Horton Story Changed American Justice. New York: Random House. Andrews, D. A., Bonta, J., & R. D. Hoge. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17(1),  19–​52. Austin, J., Cadora, E., Clear,T. R., Dansky, K., Greene, J., Gupta,V., & Young, M. C. (2013). Ending Mass Incarceration: Charting a New Justice Reinvestment. Washington, DC: Sentencing Project. Bailey,W. C. (1966). Correctional outcome: An evaluation of 100 reports. Journal of Criminal Law, Criminology, and Police Science, 57(2), 153–​160. Baker, T. (2017). Exploring the relationship of shared race/​ethnicity with court actors, perceptions of court procedural justice, and obligation to obey among male offenders. Race and Criminal Justice, 7(1), 87–​102. Beccaria, C. (1764). On Crimes and Punishments. Oxford, UK: Clarendon Press. Bersani, B. E. (2014). An examination of first and second-​generation immigrant offending trajectories. Justice Quarterly, 31(2), 315–​343. Blumstein, A. (1983). The impact of changes in sentencing policy on prison populations. In A. Blumstein, J. Cohen, S. E. Martin, & M. Tonry (eds.), Research on Sentencing: The Search for Reform (Vol. 2, pp. 460–​489). Washington, DC: National Academy Press. Bobo, L. D., & Thompson,V. (2006) Unfair by design: The war on drugs, race, and the legitimacy of the criminal justice system. Social Research: An International Quarterly, 73(2), 445–​472. Bonta, J. (1996). Risk-​needs assessment and treatment. In A. T. Harland (ed.), Choosing Correctional Options that Work: Defining the Demand and Evaluating the Supply (pp. 18–​32). Thousand Oaks, CA: Sage Publications. Burrington, L. A. (2018). Neighborhood advantage, relative status, and violence among foreign-​born adolescents. Justice Quarterly, 35(4), 699–​725. Bushway, S. D., & Piehl, A.M. (2007). Social science research and the legal threat to presumptive sentencing guidelines. Criminology & Public Policy, 6(3), 461–​482. Clark, R. (1970). Crime in America. New York: Simon and Schuster. Carrington, F. (1975). The Victims. New Rochelle, NY: Arlington House. —​—​. (1983). Crime and Justice: A Conservative Strategy. Washington, DC: The Heritage Foundation. Carrington, F., and Nicholson, G. (1989). Victims’ rights: An idea whose time has come—​five years later: The maturing of an idea. Pepp. L. Rev., 17,  1–​19. Cary, M. K. (1993). How States Can Fight Violent Crime: Two Dozen Steps to a Safer America. Washington, DC: The Heritage Foundation. Cavender, G. (2004). Media and crime policy: A reconsideration of David Garland’s The Culture of Control. Punishment & Society, 6(3), 335–​348.

38

Sentencing in the 21st Century Chiricos, T. G., & Bales, W. D. (1991). Unemployment and punishment: An empirical assessment. Criminology, 29(4), 701–​724. Chiricos,T. G., & Crawford, C. (1995) Race and imprisonment: A contextual assessment of the evidence. Ethnicity, Race, and Crime: Perspectives across Time and Place, 13, 281–​309. Chriss, J. J. (2002).The drug court movement: An analysis of tacit assumptions. In J. L. Nolan Jr. (ed.), Drug Courts in Theory and in Practice (pp. 189–​214). New York: Aldine De Gruyter. Cullen, F. T. (2005). The twelve people who saved rehabilitation: How the science of criminology made a difference: The American Society of Criminology 2004 presidential address. Criminology, 43(1),  1–​42. Cullen, F. T., & Gilbert, K. E. (1982). Reaffirming Rehabilitation. Cincinnati, OH: Anderson. Cullen F. T., & Gendraeu, P. (2001). From nothing works to what works: Changing professional ideology in the 21st century. The Prison Journal, 81(3), 313–​338. Davis, R. C., & Mulford, C. (2008).Victim’s rights and new remedies: Finally getting victims their dues. Journal of Contemporary Criminal Justice, 24(2), 198–​208. Dees, P. (1999). Victims’ Rights: Notification, Consultation, Participation, Services, Compensation, and Remedies in the Criminal Justice Process. New York:Vera Institute of Justice. Demuth, S. (2003). Racial and ethnic differences in pretrial release decisions and outcomes: A comparison of Hispanic, Black, and White felony arrestees. Criminology, 41(3), 873–​908. Demuth, S., & Steffensmeier, D. (2004). Ethnicity effects on sentence outcomes in large urban courts: Comparisons among White, Black, and Hispanic defendants. Social Science Quarterly, 85(4), 994–​1011. Duster, T. (1995). The new crisis of legitimacy in controls, prisons, and legal structures. The American Sociologist, 26(1),  20–​29. Eisenstein, J., & Jacobs, H. (1977). Felony Justice: An Organizational Analysis of Criminal Courts. New York: Little, Brown and Company. Engen, R. L., Gainey, R. R., Crutchfield, R. D., & Weis, J. G. (2003). Discretion and disparity under sentencing guidelines: The role of departures and structured sentencing alternatives. Criminology, 41(1), 99–​130. Everett, R. S., & Wojtkiewicz, R. A. (2002). Difference, disparity, and race/​ethnic bias in federal sentencing. Journal of Quantitative Criminology, 18(2), 189–​211. Fishman, J. B., & Schanzenbach, M. M. (2012). Racial disparities under the federal sentencing guidelines:The role of judicial discretion and mandatory minimums. Journal of Empirical Legal Studies, 9(4), 729–​764. Frankel, M. E. (1972). Lawlessness in sentencing. U. Cin. L. Rev., 41,  1–​54. —​—​ . (1973). Criminal Sentences: Law without Order. New York: Hill and Wang Inc. Frase, R. S. (2005). Sentencing guidelines in Minnesota, 1978–​2003. In Tonry, M. (ed.), Crime and Justice: A Review of Research (pp. 543–​562). Chicago, IL: University of Chicago Press. Garland, D. (2001). The Culture of Control: Crime and Social Order in Contemporary Society. Chicago, IL: University of Chicago Press. Gelman, A., Fagan, J., & Kiss, A. (2007). An analysis of the New  York City police department’s “stop-​and-​ frisk” policy in the context of claims of racial bias. Journal of the American Statistical Association, 102(479), 813–​823. Goddu, C. R. (1993).Victim rights or a fair trial wronged? Buff. L. Rev, 41, 245–​272. Goel, S., Rao, J. M., & Shroff, R. (2016). Precinct of prejudice? Understanding racial disparities in New York City’s stop and frisk policy. The Annals of Applied Statistics, 10(1), 365–​394. Gottfredson, D. C., & Exum, L. M. (2002). The Baltimore City Drug Treatment Court: One-​year results from a randomized study. Journal of Research in Crime and Delinquency, 39(3), 337–​356. Gottschalk, M. (2015). Caught: The Prison State and the Lockdown of American Politics. Princeton, NJ: Princeton University Press. Greene, J. (2002). Getting tough on crime: The history and political context of sentencing reform development leading up to the passage of the 1994 crime act. In C. Tata & N. Hutton (eds.), Sentencing and Society: International Perspectives (pp. 43–​64). Hampshire: Ashgate. Greenberg, D. F., & West, V. (2001). State prison populations and their growth, 1971–​1991. Criminology, 39(3), 615–​654. Hagan, J. (1975). Law, order and sentencing: A study of attitude in action. Sociometry, 38(3), 374–​384. Hewitt, J. D., & Clear, T. R. (1983). The Impact of Sentencing Reform: From Indeterminate to Determinate Sentencing. Lanham, MD: University Press of America. Huebner, B. M., & Bynum, T. S. (2008). The role of race and ethnicity in parole decisions. Criminology, 46(4), 907–​937. James, J., & Agha, S. (2013). Justice Reinvestment in Action: The Delaware Model. New  York: Vera Institute of Justice. Jiwani, Y. (2002). The criminalization of “race”, the racialization of crime. In W. Chan & K. Mirchandani (eds.), Crimes of Colour: Racialization and the Criminal Justice System in Canada (pp. 67–​86). Peterborough: Broadview Press.

39

Megan C. Kurlychek and John H. Kramer Johnson, B. D. (2003). Racial and ethnic disparities in sentencing departures across modes of conviction. Criminology, 41(2), 449–​490. Johnson, B. D. (2005). Contextual disparities in guidelines departures: Courtroom social contexts, guidelines compliance and extralegal disparities in criminal sentencing. Criminology, 43(3), 761–​796. Johnson, B. D., King, R. D., & Spohn, C. (2016). Socio-​legal approaches to the study of prosecutorial discretion and plea bargaining. Annual Review of Law and Social Science, 12, 479–​495. Johnson, B. D., Ulmer, J. T., & Kramer, J. H. (2008). The social context of guideline circumvention: The case of the US district courts. Criminology, 46(3), 737–​783. Kim, B., Cano, M.V., Kim, K., & Spohn, C. (2016). The impact of United States v. Booker and Gall/​Kimbrough v. United States on sentence severity: Assessing social context and judicial discretion. Crime & Delinquency, 62(8), 1072–​1094. King, P. (1987). Newspaper reporting, prosecution practice and perceptions of urban crime: The Colchester crime wave of 1765. Continuity and Change, 2(3), 423–​454. Kochel, T. R., Wilson, D. B., and Mastrofski, D. D. (2011). Effect of suspect race on officers’ arrest decisions. Criminology, 49(2), 473–​512. Kohm, S. A. (2009). Naming, shaming and criminal justice: Mass-​mediated humiliation as entertainment and punishment. Crime, Media, Culture, 5(2), 188–​205. Kramer, J. H. (2016). Centre County DUI-​Court: An Evaluation of the Processes and Outcomes. Centre County: Centre County DUI Court. Kramer, J. H., Hussey, F. A., Lagoy, S. P., Katkin, D., & McLaughlin, C.V. (1978). Assessing the Impact of Determinate Sentencing and Parole Abolition in Maine. Washington, DC: Law Enforcement Assistance Administration. Kramer, J. H., & Ulmer, J. T. (1996). Sentencing disparity and departures from guidelines. Justice Quarterly, 13(1), 81–​106. Kutateladze, B. L., Andiloro, N. R., Johnson B. D., & Spohn, C. (2014). Cumulative disadvantage: Examining racial and ethnic disparity in prosecution and sentencing. Criminology, 52(3), 514–​551. Latessa, E. J. (2004). The challenge of change: Correctional programs and evidence based practices. Criminology and Public Policy, 3(4), 547–​559. Lin, J., Grattet, R., & Petersilia J. (2010). “Back-​end sentencing” and reimprisonment: Individual, organizational, and community predictors of parole sanctioning decisions. Criminology, 48(3), 759–​795. Lynch, J. P., & Sabol, W. J. (2000). Prison Use and Social Control. In Policies, Processes, and Decisions of the Criminal Justice System (Vol. 3, pp. 7–​44). Washington, DC: US Department of Justice. MacNamara, D. E. (1977). The medical model in corrections requiescat in pace. Criminology, 14(4), 439–​448. MacKenzie, D. L. (2001). Sentencing and Corrections in the 21st Century: Setting the Stage for the Future. Washington, DC: National Institute of Justice. —​—​ . (2006). What Works in Corrections? Reducing the Criminal Activity of Offenders and Delinquents. New York: Cambridge University Press. Mai, C. & Sbramanian, R. (2017).The price of prisons: Examining state spending trends, 2010–​2015.Vera Institute of Justice. Retrieved from: www.Vera.org/​publications/​price-​of-​prisons-​what-​incarceration-​costs-​taxpayers Martínez-​Schuldt, R. D., & Martínez, D. E. (2017). Sanctuary policies and city-​level incidents of violence, 1990 to 2010. Justice Quarterly, 1–​27. https://​doi.org/​10.1080/​07418825.2017.1400577 Martinson, R. (1974). What works? Questions and answers about prison reform. The Public Interest, 35,  22–​54. —​—​. (1979). New findings, new views: A note of caution regarding sentencing reform. Hofstra Law Review, 7(2), 243–​258. Mauer, M. (2006). Race to Incarcerate. New York: New Press. Miethe, T., & Moore, C. (1986). Racial disparities in criminal justice processing: The consequences of model selection on conclusions about differential treatment. The Sociological Quarterly, 27, 217–​237. Miller, J., & Johnson, D. G. (2009). Problem Solving Courts: New Approaches to Criminal Justice. Lanham, MD: Rowman & Littlefield Publishers, Inc. Moore, C., & Miethe, T. (1986). Regulated and non-​regulated sentencing practices under Minnesota Felony Sentencing Guidelines. Law & Society Review, 20(2), 253–​265. Mustard, D. B. (2001). Racial, ethnic, and gender disparities in sentencing: Evidence from the US federal courts. The Journal of Law and Economics, 44(1), 285–​314. Myers, M. A. (1989). Symbolic policy and the sentencing of drug offenders. Law and Society Review, 23(2), 295–​315. National Conference of State Legislatures. (2016). Drug Sentencing Trends. Washington, DC: Author. Retrieved from: www.ncsl.org/​research/​civil-​and-​criminal-​justice/​drug-​sentencing-​trends.aspx National Institute of Justice. (2018). Drug Courts. Washington, DC: Author. National Institute on Drug Abuse. (2018). Overdose Death Rates. Bethesda, MD: Author. Najaka, K. (2003). Effectiveness of drug treatment courts: Evidence from a randomized trial. Criminology and Public Policy, 2(2), 171–​196.

40

Sentencing in the 21st Century Nelson, B. (1991). The Minnesota sentencing guidelines: The effects of determinate sentencing on disparities in sentencing decisions. Law & Inequality: A Journal of Theory & Practice, 10(2), 217–​251. Nugent-​ Borakove, M. E., & Fanflik, P. (2008). The Changing Role of the American Prosecutor. Albany, NY: SUNY Press. Ousey, G., & Kubrin, C. (2018). Immigration and crime: Assessing a contentious issue. Annual Review of Criminology, 1,  63–​84. Palmer, T. (1975). Martinson revisited. Journal of Research in Crime and Delinquency, 12(2), 133–​152. —​—​ . (1992). The Re-​emergence of Correctional Intervention. Newbury Park, CA: Sage Publications. Parent, D. (1988). Structuring Criminal Sentences:The Evolution of Minnesota’s Sentencing Guidelines. Stoneham, MA: Butterworth Legal Publishers. Pew Charitable Trusts. (2018). Fact Sheet: 35 States Reform Criminal Justice Policies through Justice Reinvestment. Washington, DC: Author. Pfaff, J. (2017). Locked In: The True Causes of Mass Incarceration—​And How to Achieve Real Reform. New  York: Basic Books. Reitz, K. R. (1998). Modeling discretion in American sentencing systems. Law & Policy, 20(4), 389–​428. Ridgeway, G., & MacDonald, J. M. (2009). Doubly robust internal benchmarking and false discovery rates for detecting racial bias in police stops. Journal of the American Statistical Association, 104(486), 661–​668. Rothman, D. J. (1980). Conscience and Convenience: The Asylum and its Alternatives in Progressive America. Boston, MA: Little Brown. Sanchez, J. E. (1990). The uses of Robert Martinson’s writings on correctional treatment: An essay on the justification of correctional policy. Journal of Contemporary Criminal Justice, 6(3), 127–​138. Schlesinger, T. (2005). Racial and ethnic disparity in pretrial criminal processing. Justice Quarterly, 22(2), 170–​192. —​—​. (2013). Racial disparities in pretrial diversion: An analysis of outcomes among men charged with felonies and processed in state courts. Race and Justice, 3(3), 210–​238. Schwarzer, W. (1992). Sentencing guidelines and mandatory minimums: Mixing apples and oranges. S. Cal. L. Rev. 66, 405–​411. Sherman, L. W., Gottfredson, D. C., MacKenzie, D. L., Eck, J., Reuter, P., & Bushway, S. (1998). Preventing Crime: What Works,What Doesn’t,What’s Promising. Washington, DC: US Department of Justice. Singer, R. G. (1979). Just Deserts: Sentencing Based on Equality and Desert. Cambridge, MA: Ballinger Publishing. Spohn, C. (2000).Thirty years of sentencing reform:The quest for a racially neutral sentencing process. In Policies, Processes and Decisions of the Criminal Justice System (Vol. 3, pp. 427–​501). Washington, DC: US Department of Justice. —​—​. (2008). Race, sex, and pretrial detention in federal court: Indirect effects and cumulative disadvantage. U. Kan. L. Rev., 57, 879–​902. Spohn, C., & Brennan, P. K. (2011). The joint effects of offender race/​ethnicity and gender on substantial assistance departures in federal courts. Race and Justice, 1(1),  49–​78. Stith, K., & Cabranes J. A. (1998). Fear of Judging: Sentencing Guidelines in the Federal Courts. Chicago, IL: University of Chicago Press. Steffensmeier, D., & Demuth, S. (2000). Ethnicity and sentencing outcomes in US federal courts:Who is punished more harshly? American Sociological Review, 65(5), 705–​729. Steffensmeier, D., Ulmer, J., & Kramer, J. (1998). The interaction of race, gender, and age in criminal sentencing: The punishment cost of being young, black, and male. Criminology, 36(4), 763–​798. Stemen, D., Rengifo, A., & Wilson, J. (2006). Of Fragmentation and Ferment: The Impact of State Sentencing Policies on Incarceration Rates, 1975–​2002—​Final Report to the National Institute of Justice. Washington, DC: National Institute of Justice. Stith K., & Cabranes J. A. (1998). Fear of Judging: Sentencing Guidelines in the Federal Courts. Chicago, IL: University of Chicago Press. Stolzenberg, L., D’Alessio, S. J., & Eitle, D. (2013). Race and cumulative discrimination in the prosecution of criminal defendants. Race and Justice, 3(4), 275–​299. Stuart, R., & Joshi, N. (2018). Dreamland: The True Tale of America’s Opiate Epidemic. New  York: Bloomsbury Publishing. Taxman, F. S. (2002). Supervision: Exploring the dimensions of effectiveness. Federal Probation, 66(2),  14–​27. Toffler, A., & Alvin, T. (1980) The Third Wave (Vol. 484). New York: Bantam Books. Tonry, M. (1995). Malign Neglect: Race, Crime, and Punishment in America. New York: Oxford University Press. —​—​ . (1997). Sentencing Matters. New York: Oxford University Press. —​—​ . (1999). Reconsidering Indeterminate and Structured Sentencing (NCJ 175722). Washington, DC: National Institute of Justice. —​—​ . (2016). Sentencing Fragments: Penal Reform in America, 1975–​2015. New York: Oxford University Press.

41

Megan C. Kurlychek and John H. Kramer —​—​. (2017). Making American sentencing just, humane and effective. In M. Tonry (ed.), Crime and Justice: Reinventing American Criminal Justice (Vol. 46, pp. 441–​504). Chicago, IL: University of Chicago Press. Tonry, M., & Melewski. M. (2008). The malign effects of drug and crime control on Black Americans. In M. Tonry (ed.), Crime and Justice:A Review of Research (Vol. 37, pp. 1–​44). Chicago, IL: University of Chicago Press. Tucker, S., & Cadora, E. (2003). Justice reinvestment: To invest in public safety by reallocating justice dollars to refinance education, housing, healthcare, and jobs. Ideas for an Open Society. Occasional Papers Series, 3(3). New York: Open Society Institute. Twentieth-​Century Fund. (1976). Fair and Certain Punishment: Report of the Twentieth Century Fund Task Force on Criminal Sentencing. New York: McGraw-​Hill. Tyler, T. (2006). Why People Obey the Law. Princeton, NJ: Princeton University Press. Ulmer, J. T., Light, M. T., & Kramer, J. H. (2011). Racial disparity in the wake of the Booker/​Fanfan decision: An alternative analysis to the USSC’s 2010 report. Criminology and Public Policy, 10(4), 1077–​1118. Ulmer, J. T., Kurlychek, M. C., & Kramer, J. H. (2007). Prosecutorial discretion and the imposition of mandatory minimum sentences. Journal of Research in Crime and Delinquency, 44(4), 427–​458. United States Sentencing Commission. (2006). Final Report on the Impact of United States v.  Booker on Federal Sentencing. Washington, DC: Author. United States Sentencing Commission. (2012). 2012 Report to the Congress: Continuing Impact of United States v. Booker on Federal Sentencing. Washington, DC: Author. von Hirsch, A. (1976). Doing Justice:The Choice of Punishments. New York: Hill and Wang. von Hirsch, A., & Ashworth, A. (2005). Proportionate Sentencing: Exploring the Principles. New  York: Oxford University Press. Wang, E., Mears, D. P., Spohn, C., & Dario, L. (2013). Assessing the differential effects of race and ethnicity on sentence outcomes under different sentencing systems. Crime & Delinquency, 59(1), 87–​114. Weisburd, D., Telep, C. W., & Lawton, B. A. (2014). Could innovations in policing have contributed to the New York City crime drop even in a period of declining police strength? The case of stop, question and frisk as a hot spots policing strategy. Justice Quarterly, 31(1), 129–​153. Weisburd, D., Wooditch, A., Weisburd, S., & Yang, S. M. (2015). Do stop, question, and frisk practices deter crime? Criminology & Public Policy, 15(1),  31–​56. Wellford, C. F. (1973). Age composition and the increase in recorded crime. Criminology, 11(1),  61–​70. Western, B., & Pettit, B. (2010). Incarceration & social inequality. Daedalus, 139(3),  8–​19. Wilson, J. Q. (1975). Thinking about Crime. New York: Basic Books. Wilson, J. Q. (1978). The Investigators: Managing FBI and Narcotic Agents. New York: Basic Books. Wooldredge, J., Frank, J., Goulette, N., & Travis L. (2015). Is the impact of cumulative disadvantage on sentencing greater for black defendants? Criminology & Public Policy, 14(2), 187–​223. Worrall, J. L., & Wheeler, A. P. (2018). Evaluating community prosecution code enforcement in Dallas, Texas. Justice Quarterly,  1–​30. Zimring, F. E. (2007). Protect individual punishment decisions from mandatory penalties. Criminology & Public Policy, 6(4), 881–​886.

42

2 SENTENCING GUIDELINES IN THE UNITED STATES Richard S. Frase and Kelly Lyn Mitchell

Introduction Sentencing guidelines reforms, developed by an independent sentencing commission, have been in effect in some states for almost 40  years, and 19 state and federal jurisdictions currently sentence under a guidelines system.1 This is also the sentencing structure recommended by the American Law Institute (2017), in the recently revised sentencing and corrections provisions of the Institute’s influential Model Penal Code (MPC); commission-​drafted sentencing guidelines have likewise been strongly endorsed by the American Bar Association (1979, 1994). And since the late 1970s, no competing sentencing reform model of comparable scope has been implemented, or even seriously proposed. But guidelines systems are very diverse; no two systems are the same, and in some respects they differ dramatically. This rich diversity, and the wealth of sentencing data available in many guidelines jurisdictions, provides many opportunities for research on the ways in which sentencing reforms are implemented, and the effects that different guidelines structures and rules have on sentencing practices and outcomes. In a few systems (notably Minnesota, Pennsylvania, and the federal system) there is a considerable body of research by commission staff and outside researchers, but much more work needs to be done. And for many guidelines jurisdictions there is almost no published research. Until recently, it was difficult for researchers to even identify which jurisdictions have guidelines, and to learn about the details of each system, but this critical background information is now available online at the University of Minnesota’s Sentencing Guidelines Resource Center, of which we are the co-​directors (Robina Institute, 2018). The goals of this chapter are to encourage and facilitate research on American guidelines systems by explaining how these systems work, documenting the major variations across the systems, providing an overview of recent research, and suggesting the issues most in need of further research.

What Are Sentencing Guidelines, and Where Have They Been Adopted? The second half of the 20th century witnessed a number of sentencing reforms designed to reduce or eliminate the broad sentencing and parole release discretions that had existed in all jurisdictions for many decades (and still exist in the majority of American states today) (Tonry, 1996, 2016; Kurlychek & Kramer, this volume). Sentencing guidelines are only one of these “structured” sentencing 43

Richard S. Frase and Kelly Lyn Mitchell Table 2.1  American Sentencing Guidelines Systems, as of June 2018 Jurisdiction

Guidelines Commissions—​years in operation

Years when each commission’s guidelines first went into effect

Minnesota Pennsylvania Maryland Florida Washington Delaware Federal courts Oregon Tennessee Kansas Arkansas No. Carolina Virginia Massachusetts Ohio Utah Michigan Washington, DC Alabama

1978–​ 1978–​ 1996–​ 1982–​1997 1981–​ 1984–​ 1984–​ 1985–​ 1986–​1995 1989–​ 1993–​ 1990–​ 1994–​ 1994–​ 1990–​ 1983–​ 1994–​2002, 2015–​ 1998–​ 2006–​

1980 1982 1983 1983 1984 1987 1987 1989 1989 1993 1994 1994 1995 1996 1996 1998 1999 2004 2006

Source: Robina Institute of Criminal Law and Criminal Justice, 2018

reforms (US Dept. of Justice, 1996, 1998); moreover, not all true guidelines systems are referred to as “guidelines”—​some jurisdictions deliberately avoided the use of that term because it was associated with the widely criticized federal guidelines, and the reformers wanted to signal that they were taking a very different approach (Tonry, 1996). For purposes of this chapter we define a “guidelines” system as one in which: 1) judges are given recommended sentences for most felonies or other serious crimes; 2) that apply to typical cases of that type (i.e., cases without aggravating or mitigating factors justifying departure from the recommendation); and 3) that were developed by a legislatively created sentencing commission (regardless of whether the rules were then embodied in statutes, and even if the commission was abolished after the guidelines went into effect).2 This definition excludes presumptive sentences like those that legislatures in California and several other states adopted in the mid-​to late 1970s, without a sentencing commission.3 The existence of a commission, especially one that remains in existence after the guidelines are implemented, is important when examining any issue of sentencing policy—​such a body can (and most do) collect and analyze data on sentences imposed under the guidelines and on the ways in which sentencing decisions translate into higher or lower prison populations. The 19 state and federal jurisdictions that currently have guidelines meeting the above definition are listed in Table 2.1, showing the initial effective dates of each jurisdiction’s guidelines and the years when that sentencing commission was in operation.4 As discussed more fully below, these 19 systems vary in a number of important respects, including their primary reform and punishment goals; the crimes and sentencing decisions they cover; and the ways in which those decisions are regulated.

Purposes of Guidelines Reforms When comparing guidelines provisions and impacts across jurisdictions, it is important to take into account each system’s reform and sentencing policy goals.While some goals have been endorsed in all 44

Sentencing Guidelines in the US

guidelines jurisdictions, systems often differ in the ways in which they define and prioritize the same goal (e.g., disparity reduction; resource management) (for further discussion, see Frase, 2015b). Other goals have been recognized in some systems but not others. These differences mean that, in some respects, guidelines systems may not be truly comparable; it is not fair to critique a guidelines system for failing to achieve a goal that the system chose not to recognize (although it is fair to question that decision itself).

Disparity Reduction In response to the widespread concerns that began to be expressed, in the late 1960s, about sentencing disparities under traditional sentencing regimes, the primary policy goal of the earliest sentencing guidelines reforms was to reduce sentencing disparity by limiting and structuring the discretion of judges and parole boards (von Hirsch, 1987). However, the meaning of sentencing disparity in any given jurisdiction depends on what sentencing purposes that jurisdiction considers to be most important; those purposes determine when two cases are “similar” and should receive similar penalties (the goal of sentencing “uniformity”), and when they are sufficiently different that the penalties should be different (the goal of sentencing “proportionality”). In the pre-​guidelines period the most important punishment goals were rehabilitation and public protection, but in the 1970s these punishment rationales began to fall out of favor and more attention began to be given to retribution (just deserts) and general deterrence (Blumstein et al., 1983; Reitz, 2001; Tonry, 1996). Both of the latter goals require sentence severity to be proportionate to the seriousness of the offender’s current offense, rather than to the offender’s need for rehabilitation or restraint (incapacitation) (Frase, 2005a). But probably because an offender’s prior record of convictions has always been deemed an important sentencing factor, and one that is arguably relevant to all traditional punishment purposes especially those related to promoting public safety, all guidelines reforms have chosen to also base recommended sentences on prior record, as well as on conviction offense severity.Thus, under sentencing guidelines “disparity” exists when offenders with similar current offenses and prior records do not receive similar penalties, or when offenders committing more serious current offenses, and/​or with more serious prior conviction records, do not receive more severe penalties. The emphasis that guidelines reforms have given to offense severity and prior conviction record also largely explains the widespread use of a two-​dimensional grid format to display recommended sentences.

Improved Sentencing Policy Development and Implementation The earliest guidelines reforms were also strongly influenced by the writings of federal judge Marvin Frankel (1973). Frankel proposed both the concept of guidelines that would provide recommended sentences for typical offenses of each type, and the idea that guidelines should be developed and monitored by a specialized agency akin to the rule-​drafting and enforcing administrative agencies legislatures have often created to address complex areas of law and public policy. Accordingly, the first guidelines enabling statutes in Minnesota and Pennsylvania established independent and permanent sentencing commissions, with the goal of promoting more rational and informed sentencing policy. This goal has two facets. First, the use of an independent, appointed commission to develop and monitor guidelines serves to partially insulate sentencing policy decisions from short-​term political pressures to appear “tough on crime.” Second, the commission, like other administrative agencies, is directed to collect data and develop expertise which will facilitate evidence-​based policy development and improvement. The commission’s database and expertise (and in many states, the goal of avoiding prison overcrowding, discussed below) also encourage a more comprehensive, long-​term, and fiscally responsible perspective: sentencing policy is determined on a statewide basis and for all crimes (or at least, all felonies), thus avoiding piecemeal reforms in response to the news media’s “crime of the week” and helping to more effectively manage correctional resources. 45

Richard S. Frase and Kelly Lyn Mitchell

Improved Management of Correctional Resources Another early goal, recognized in Minnesota but not initially in Pennsylvania, was to use the more consistent and predictable nature of guidelines sentencing, combined with the improved sentencing data and analytic capabilities of a sentencing commission, to achieve better management of limited and expensive correctional resources, set priorities for the use of those resources, and avoid prison overcrowding (Frase, 2005b). A  commission’s data on past and current sentences imposed allows the commission to forecast the impact that particular recommended sentences will have on future inmate and supervision populations, and on the public costs of increased correctional populations. Advocates of more severe penalties can then be asked to show how the increased penalty will be paid for (by raising taxes, and/​or by lowering penalties for other offenders). And if additional prison beds will be required, the legislature can appropriate and expend the funds in time to avoid any serious overcrowding. Almost all systems with guidelines eventually recognized the goal of improved correctional resource management, although they have defined and applied it in different ways (some have only used prison bed projections to warn the legislature of the need for more prison beds, not to set priorities, Frase, 1995, 2005b).

Effecting Changes in Particular Sentencing Policies Minnesota and some other guidelines states have also used guidelines to make prescriptive changes in sentencing policy, and in particular, to send more violent and fewer property offenders to prison (Frase, 2005b). By contrast, guidelines reforms in other jurisdictions (e.g., the District of Columbia) have only sought to encourage judges to more consistently apply existing sentencing norms (“descriptive” or “historical” guidelines).

Encouraging More Frequent Use of Intermediate Sanctions In response to the substantial costs and burdens of rising prison populations, some guidelines states began to use their guidelines to encourage more frequent, effective, and consistent use of intermediate sanctions that are less restrictive than prison but more restrictive than traditional probation (Frase, 2005b, 2013; Morris & Tonry, 1990). However, as further discussed below, guidelines provisions regulating intermediate sanctions are usually much less restrictive than that jurisdiction’s guidelines regulating the use and duration of prison sentences.

Truth in Sentencing Another goal that has been recognized in a number of guidelines systems is to promote transparency and “truth in sentencing” by reducing the difference between the prison term imposed at sentencing and the actual amount of time the offender spends in prison (Frase, 1999, 2005b). Advocates for crime victims and their families increasingly objected to parole release practices under which an offender might be sentenced to ten years in prison but released after one or two years. Even in states that had already abolished parole release discretion, there was pressure to lower the proportion of prison terms that could be reduced by good-​conduct credits. A 1994 federal statute strongly encouraged truth in sentencing goals by providing substantial funds for prison construction to states that required inmates convicted of serious crimes to serve at least 85% of their sentences.

Summary The importance of the reform goals listed above (when recognized) varies from minor to substantial. Achievement of disparity reduction depends on the degree of binding force in a system’s guidelines 46

Sentencing Guidelines in the US

(further discussed below). Improved sentencing policy requires a commission that is both independent and adequately funded. The goal of better correctional resource management can be mostly ignored, or it can greatly constrain sentencing policy.The basis for guidelines recommendations varies from entirely historical to highly prescriptive. Greater use of intermediate sanctions can be weakly or strongly encouraged (see further discussion below, under Scope). The degree of truth in sentencing depends mostly on whether parole release discretion has been abolished for some or all crimes, but abolition systems also vary greatly in the magnitude of good-​conduct credits they allow (also discussed under Scope).

Varieties of Guidelines The structure and content of state and federal guidelines systems differ substantially on a number of dimensions. The most important of those dimensions, discussed more fully below, relate to the scope of the guidelines (cases and sentencing issues covered), grids and other formats used to display recommended typical-​case sentences, the nature and weighting of factors that determine the recommended sentence, and the degree of binding force of guidelines recommendations. Except as otherwise noted, the source for the summary below is the University of Minnesota’s Sentencing Guidelines Resource Center (Robina Institute, 2018).

Scope: Cases and Sentencing Issues Covered by the Guidelines Guidelines in 12 of the 19 jurisdictions only cover felony-​level crimes (Alabama, Arkansas, District of Columbia, Florida, Kansas, Michigan, Minnesota, Ohio, Oregon,Tennessee,Virginia, and Washington). The federal guidelines cover class A misdemeanors (the most serious class) as well as felonies. And in six states (Delaware, Maryland, Massachusetts, North Carolina, Pennsylvania, and Utah) guidelines coverage extends to all non-​petty misdemeanors. In many systems certain crimes are specifically excluded from the guidelines, either because the crimes involve extreme or indeterminate sentences (e.g., Minnesota excludes first-​degree murder and other crimes punishable by life imprisonment), or because local politics or other considerations made it difficult to incorporate all crimes into the guidelines when they were enacted (e.g., North Carolina excludes driving under the influence and certain drug trafficking offenses). All guidelines provide recommendations as to whether the offender should be sentenced to prison or placed on probation (this is often referred to as the “disposition” decision), and also recommendations as to the duration of prison sentences. Most guidelines specify a recommended prison duration or range even if the prison sentence will be suspended (or “stayed”). As further discussed below, recommended guidelines prison durations have much more practical significance in guidelines systems that have abolished parole release discretion than in systems that retain parole discretion. Some guidelines also contain recommendations related to non-​prison sanctions. A few systems (e.g., North Carolina and Pennsylvania) define two or more broad levels of sanction intensity that may be imposed for different groups of offenders, for example, differentiating between more severe intermediate sanctions such as jail or inpatient treatment and less severe sanctions such as community service or outpatient treatment (Frase, 2013, chapter 3). Where probation is the recommended sanction, a few provide guidance as to the length of probation terms and the use of revocation and other sanctions for violations of probation conditions. Some provide a maximum or a permitted range of local jail custody for each grid cell or other offender group, with rules translating non-​jail community penalties into jail days (e.g., Oregon and Washington). But in most systems, judges have complete discretion to fashion a non-​prison sentence and select probation conditions which may include onerous custody requirements (for example, in Minnesota up to a year in jail may be imposed as a condition of felony probation). 47

Richard S. Frase and Kelly Lyn Mitchell

Guidelines systems generally use three kinds of backup sanctions, alone, or in combination, to enforce probation conditions (Frase, forthcoming b): 1) full or partial execution of a prison sentence that had previously been suspended; 2) revocation of a prior decision suspending imposition of sentencing (in which case the court holds a deferred sentencing hearing and decides what prison term to impose and whether to execute it); and/​or 3) application of jail or other non-​prison sanctions for violation of release conditions. Some guidelines reforms simply continued pre-​guidelines backup-​ sanction options, while other jurisdictions sought to restrict or restructure these options. Oregon and Washington abolished suspended sentences for most offenders, substituting shorter jail terms as backup sanctions; in these states, most offenders thus receive probation as a free-​standing penalty, not as a condition of a suspended-​execution or suspended-​imposition sentence, and prison is essentially “off the table” once the court decides to grant probation. In contrast, North Carolina, Utah, and the federal system retain the use of suspended-​execution or suspended-​imposition sentences as backup sanctions, but encourage and in some cases require judges to use custody terms shorter than the recommended guidelines prison term applicable to that offense and offender, when sanctioning initial and less serious violations of probation conditions. However, in all five of the systems named above, judges retain considerable discretion when selecting from the available sanctions for probation violations, and they retain almost complete discretion in the remaining guidelines jurisdictions that made no change in their system of backup sanctions. The presence or absence of parole release discretion is one of the most important differences among guidelines systems, since this determines the extent to which the guidelines regulate the duration of prison terms—​ in parole-​ retention jurisdictions, recommended and imposed prison terms only determine the minimum and/​or the maximum length of imprisonment, which usually leaves a substantial range within which the parole board can select the actual amount of time to be served before release. Eleven of the 19 guidelines systems have abolished discretionary parole release for all or most offenders; inmates serve their entire executed prison term minus any earned good-​conduct and/​or program-​participation credits. The other eight guidelines systems—​Alabama, Arkansas, Maryland, Massachusetts, Michigan, Pennsylvania, Tennessee, and Utah—​retain parole discretion for most offenders serving felony sentences; five of these states use or are developing parole release guidelines (Watts, 2018b). In many parole-​abolition guidelines systems, good-​conduct and participation credits are about 33% of the maximum imposed prison term, but such credits can be as low as 15% in some systems (e.g., the federal system and the District of Columbia) and as high as 50% for some offenders in some states (e.g., Washington). In parole-​abolition systems most released inmates have a period of post-​prison supervision; the duration of supervision can be equal to the offender’s remaining unserved prison term (e.g., Minnesota) or can be a fixed period that usually depends on the seriousness of the conviction offense. Under the latter approach, supervision periods are sometimes shorter than typical traditional parole terms. For example, the supervision period in North Carolina is either 9 or 12 months, depending on the offense severity level. In Washington, some non-​violent offenders have no required supervision period, and for most other offenders the period is one year, 18 months, or three years. In all systems, with or without abolition of parole release discretion, there are few rules limiting the conditions of post-​prison supervision or regulating revocation and other sanctions for violations of release conditions.

Recommended Sentences for the “Typical” Case Sentencing guidelines generally recommend a sentence for the “typical” case. Thus, the sentence recommended by the guidelines is deemed to be appropriate for all similar offenses committed in the typical manner, and for all offenders with similar criminal histories. But at the same time, all guidelines recognize that there are atypical cases or situations that may warrant a departure (i.e., a sentence other than that recommended by the guidelines) such as when the manner in which the crime was committed was more serious than average (e.g., committed with particular cruelty). 48

Sentencing Guidelines in the US

Grids and Other Guidelines Formats Most state and federal guidelines use a two-​dimensional grid to display recommended typical-​case sentences, but three states—​Alabama, Florida, and Virginia—​compute recommended sentences on one or more worksheets (Frase et al., 2015; Watts, 2018a). Ohio and Delaware use neither grids nor worksheets. Ohio statutes contain standards to help judges decide between prison and probation and select a prison duration; in Delaware a sentence range is provided for each offense class and sub-​class (violent; non-​violent), with specified adjustments for listed aggravating and mitigating circumstances. On guidelines grids, offense severity is almost always represented on the vertical axis (grid rows), while prior record (“criminal history”) forms the horizontal axis (grid columns) (in Utah these two axes are reversed). The grid cell at the intersection of these two points determines the recommended sentence. Four states and the federal system have a single grid for all offenses; nine guidelines jurisdictions have multiple grids (usually a “main” or “standard” grid, with additional grids for sex, drug, or other offense types). The use of multiple grids means that the scoring of offense severity and criminal history may depend on the conviction offense type, and such grids may also reflect particular sentencing policy choices; in Minnesota, for example, the creation of a separate sex-​crimes grid in 2006 was designed to increase sentence severity for these offenses, while the addition of a drug crimes grid in 2016 was part of a set of reforms designed to lower sentence severity for some drug offenses. Grids differ in a number of other respects, including: the number of severity levels and criminal history categories (see further discussion, below); the size of the sentence ranges contained within the cells and the degree of overlap, if any, between cells or rows; use of a “disposition” line separating grid cells with recommended executed prison sentences from cells recommending probation;5 and the order of offense severity levels and criminal history categories.6 In most systems the prison terms shown on the grid (and any higher or lower terms chosen by judges, in cases of departure) represent the maximum terms offenders will serve; that maximum is then usually reduced by good-​conduct and program-​participation credits (and/​or discretionary early release on parole, if that has been retained). But in three guidelines states, Michigan, North Carolina, and Pennsylvania, the prison terms shown on the grid and chosen by judges are the minimum term the offender will serve if he or she receives all available credits and/​or is paroled at the earliest allowable time. In these systems the maximum prison term is then determined by other rules (for example, except for the most serious felonies the maximum in North Carolina is automatically set at 120% of the minimum).

Nature and Weighting of Factors Determining Recommended Sentences In almost all guidelines systems, including those using worksheets rather than a grid format, typical-​ case recommended sentences are based almost entirely on the nature and severity of the offender’s conviction offense(s)7 and his or her prior conviction record. Under the federal guidelines, however, sentence enhancements and upward departures are often based on crimes that did not result in conviction (because they were dismissed, acquitted, or never charged).8 Grid systems usually have about 10 to 15 offense severity levels on their main grids, but Tennessee has only five and the federal grid has 43; these differences are only partly explained by the use of single versus multiple grids, and the inclusion or exclusion of misdemeanor-​level crimes. The use of more offense severity levels (or a more detailed worksheet) means that a larger number of aggravating and mitigating circumstances can be specified in advance, with narrower ranges within the grid cells and fewer determinations left to the sentencing judge. Some systems (e.g., Tennessee) assign offenses to severity levels according to statutory categories (e.g., felony classes), while in other systems (e.g., Minnesota) the sentencing commission makes an independent assessment of the relative seriousness of each offense (Parent, 1988). Although no systematic assessment has been made, the severity ranking of common offenses appears to follow a similar pattern across guidelines systems (homicide offenses 49

Richard S. Frase and Kelly Lyn Mitchell

are ranked highest followed by other serious violent crimes, serious drug crimes, serious property crimes, and less serious violent, drug, and property crimes). As for criminal history, most systems have five to seven categories, but in Kansas, Oregon, and Washington there are nine or ten.9 Again, the use of more categories permits narrower ranges within the grid cells, with fewer determinations left to the sentencing judge. All guidelines systems using a grid format have a separate criminal history score that forms one axis of the grid or grids; in systems using worksheets rather than grids, prior record factors add points that determine the form (prison versus probation) of the recommended sentence and the duration of imprisonment. Some systems use prior offense-​type categories rather than points. For example, in Kansas offenders with three or more prior violent felonies are placed in the highest criminal history category (A), while offenders with only misdemeanors or no prior convictions are placed in the lowest category (I). Criminal history enhancements are sometimes said to be justified under a retributive punishment theory (i.e., that repeat offenders are more culpable for their current offenses), but that rationale is highly contested (Frase & Roberts, forthcoming, ­chapter 1; Roberts & von Hirsch, 2010). And although prior record is generally viewed as a proxy for the offender’s risk of recidivism, very few guidelines systems have attempted to validate the risk-​predictive accuracy of their criminal history scores and score components (Frase & Roberts, forthcoming, c­ hapter 2). Guidelines criminal history scoring rules vary in terms of which convictions and other prior record factors are counted, their weighting, and how much a high score increases the recommended sentence (Frase et al., 2015; Hester et al., 2018). All guidelines systems count prior adult felony convictions, and almost all systems include at least some prior misdemeanors and juvenile court adjudications Most systems also add a point or points for “custody status” if the offender was under criminal justice supervision or was in jail or prison, at the time he committed the offense being sentenced. Almost all point-​based systems count prior felonies more heavily than misdemeanors and juvenile adjudications, and weight felonies according to their seriousness. About half of the systems have rules limiting eligibility for the highest criminal history categories to offenders with at least one prior violent or high-​severity felony. Many guidelines systems stop counting very old priors at some point, but in more than half of the systems all or at least the most serious prior felony convictions are counted, no matter how long ago they were entered. And even when a lookback limit (typically 10 or 15 years) does apply, the “clock” often does not begin to run until the offender is discharged from probation, released from prison, or discharged from post-​prison supervision. Thus, a 15-​year lookback limit can easily count convictions that are 20 or 25 years old, and a 50-​year-​old offender can have his sentence enhanced based on crimes committed when he was a young man. Despite the well-​known age-​crime-​curve of declining criminality, almost no guidelines system considers the offender’s current age as a formal sentencing factor. Although no studies have specifically addressed this question, it seems likely that offender age and other factors unrelated to current offense and prior record are given less weight under guidelines sentencing than in systems without guidelines. A high criminal history score, category, or point total can substantially increase the severity of an offender’s recommended sentence (Frase et al., 2015; Frase & Roberts, forthcoming, ­chapter 5). High scores cause many offenders convicted of medium-​to low-​severity offenses to be recommended for and receive a prison sentence rather than probation. High scores also greatly increase the duration of recommended custody sentences—​across guidelines systems the average recommended custody sentence for offenders in the highest criminal history category is six times longer than for offenders in the lowest category, and in some systems the ratio of highest-​history to lowest-​history recommended custody is over 10.These major sentence enhancements have a number of problematic consequences: they greatly increase the size and expense of prison populations; make sentences less proportional to conviction offense seriousness; imprison many non-​violent offenders; increase the numbers of aging, low-​r isk, high-​cost inmates; and add to the racial disproportionality of prison populations. It seems likely that at least in some guidelines systems, prior record receives more weight at sentencing than 50

Sentencing Guidelines in the US

it did before guidelines, or does in comparable systems without guidelines (these questions have not been studied, however).

How Guidelines Sentence Recommendations Relate to Statutory Sentencing Ranges In most guidelines systems the longest recommended prison term in each grid cell or worksheet group, even for the highest criminal history offenders, is lower than the applicable statutory maximum (although prison terms above the guidelines range can be imposed by means of upward durational departure). In a few systems (e.g., the District of Columbia) the top of the highest-​history guidelines range was set equal to the preexisting statutory maximum. And in some systems the tops of guidelines range effectively replaced previous statutory maximums. For example, the top of the “aggravated” range in each cell of the North Carolina grid became the maximum prison term allowed in that cell; and in Kansas, upward durational departures cannot exceed twice the top of the applicable guidelines grid-​cell range. With regard to statutory minimums, state systems sometimes take them into account when setting recommended guidelines sentences but, in most states, the recommended prison term is determined independently and then is over-​r idden if a longer statutory minimum applies. By contrast, under the more punitive approach of the federal guidelines all recommended sentences are equal to or greater than any applicable or related mandatory minimum prison term (Tonry, 1996, pp. 96–​98).

Degree of Binding Force of Guidelines Recommendations Although some guidelines are considered “mandatory,” it is important to stress that no guidelines rules are truly mandatory in the way that mandatory or mandatory minimum statutes are (i.e., requiring the judge to impose a particular sentence or minimum sentence, with no power to disregard the rule). One of the distinctive features of so-​called mandatory guidelines is that the recommended sentences provided for each group of cases are only presumed to be correct, based on the characteristics of typical cases of that type, and the presumption may be rebutted—​judges may depart10 from the recommended sentence if the case is sufficiently atypical to meet that system’s standard for departure, and the judge states reasons that are consistent with general guidelines policies. Similarly, just because a system refers to its guidelines as “advisory” does not mean that the guidelines rules lack any binding effect; for example, sentences under the advisory guidelines in Pennsylvania and the federal courts are subject to appellate review and may be reversed if found “unreasonable.” Thus it is best to view the binding effect of guidelines as forming a continuum rather than a “mandatory” versus “advisory” dichotomy (Frase, 2015a; Mitchell, 2017). A given system’s location on this continuum depends on the degree to which that system’s typical-​case sentence recommendations and departure standards are formally or in practice binding on judges, as measured by factors such as: the extent to which they are enforced by active appellate review and occasional reversals; whether a departure standard is provided, and if so, how strict it is; procedural requirements to state reasons for departure; and other factors (e.g., publication of judge-​specific departure rates) that tend to encourage compliance. Legally binding guidelines systems with active appellate review are found in Kansas, Minnesota, Oregon, and Washington. But even in these systems trial courts retain considerable discretion as to both the type and the severity of sanctions, and appellate review does not appear to have unduly limited trial court discretion (Reitz, 1997). At the other end of the binding-​force continuum are guidelines systems in which recommendations and departure standards are expressly described as “voluntary” or “advisory”—​sentencing judges may ignore guidelines recommendations and impose any sentence within the broader statutory range for that offense, and departures are not subject to appeal. Jurisdictions following this model include Arkansas, Delaware, the District of Columbia, Maryland, Massachusetts, Utah, and Virginia. Systems falling in the middle of the binding-​force continuum display numerous variations. Some “advisory” guidelines systems have features that strongly 51

Richard S. Frase and Kelly Lyn Mitchell

encourage judicial compliance (for example: the requirement to state reasons for departure; publication of judge-​specific departure rates; strong collegial bonds in a small or geographically concentrated jurisdiction). Conversely, in some legally binding guidelines jurisdictions judges retain broad discretion when making certain decisions, so in those contexts the system functions more like an advisory system. Some systems allow for appellate review of all sentences, whereas others only permit review of departures. There are also major differences in the standard of review and degree of deference to the trial court where appeal is allowed. For example, in North Carolina the standard of appellate review of departures is so deferential that the typical-​case presumptive sentence range is, as a practical matter, only advisory; sentences in the “aggravated” or “mitigated” ranges are rarely reversed on appeal.11 With or without strict or any appellate review, the extent to which judges feel bound by guidelines recommendations will depend on the applicable departure standard. Six states (e.g., Pennsylvania and Virginia) have no stated standard that judges must meet before departing from recommended guidelines sentences (Mitchell, 2017). Some stated departure standards only require that the case be “atypical” (Arkansas Sentencing Commission, 2018, p. 1) or that there are “aggravating” or “mitigating” circumstances (Mass. Gen. Laws ch. 211E, § 3(a)(2) (2018)). In other states (e.g., Delaware and Minnesota) a stricter standard applies: the judge must find that there are “substantial and compelling” circumstances justifying departure. Under the federal guidelines, departure requires a finding that the case involves aggravating or mitigating circumstances “of a kind, or to a degree, not adequately taken into consideration by the sentencing commission in formulating the guidelines” (USSC, 2016, § 5K2.0). Because of the extremely detailed nature of the federal guidelines, this standard strongly discouraged departures in the period prior to the Booker decision (discussed below) when the guidelines were legally binding. In two respects, the pre-​Booker guidelines also strongly discouraged downward departure based on offender cooperation. Under the guidelines (USSC, 2016, §3E), offenders who show “acceptance of responsibility” (by pleading guilty) may receive no more than a 3-​level reduction in their offense severity level (the federal grid contains 43 levels). And offenders who provide “substantial assistance” to law enforcement may receive sentence mitigation only upon motion by the government (USSC, 2016, §5K1). The degree to which any given set of guidelines are legally binding took on federal constitutional significance with the US Supreme Court’s decision in Blakely v.Washington (2004). Blakely held that, in legally binding guidelines systems like the one in Washington state, any fact (other than the fact of a prior conviction) that is used to enhance a sentence above the top of the applicable guidelines range must be either admitted by the defendant or found by a jury beyond a reasonable doubt. The same rule was applied to the federal guidelines in United States v. Booker (2005); but the latter case also made clear that the Blakely requirements do not apply in an “advisory” guidelines system. The Court, in Booker, further held that the federal guidelines would thereafter be deemed “advisory” unless and until Congress specified that it wanted the guidelines to once more be “legally binding” and subject to Blakely proof standards (which Congress has not done, as of this writing); yet the Booker court also held that defense and prosecution appeals of sentences under the federal guidelines would continue to be available, subject to an overall “reasonableness” standard of review—​that degree of “binding force” does not trigger Blakely proof requirements. In a later case, Alleyne v.  United States (2013), the Supreme Court extended the Blakely proof requirements to facts that trigger the application of a mandatory minimum statute, and lower courts have further extended this ruling to facts that increase the lower end of the applicable guidelines range (see, e.g., People v. Lockridge (2015), applying Blakely to the Michigan guidelines). Several states (Kansas, Minnesota, Oregon, North Carolina, and Washington) have chosen to retain their legally binding guidelines and comply with Blakely proof requirements; two states (Ohio and Tennessee) chose to avoid Blakely by making their guidelines completely advisory (Frase, 2007). Alabama is the only state to move from advisory to mandatory guidelines since the Supreme Court’s decision in Blakely. In 2006, Alabama enacted presumptive

52

Sentencing Guidelines in the US

guidelines for property and drug offenses (but its guidelines for person and violent offenses remain advisory). Another way to assess the binding effect of a jurisdiction’s guidelines is to examine rates of departure from or compliance with recommended guidelines rules. Most sentencing guidelines commissions collect data on compliance rates; these can be used to encourage judges to comply (especially if judge-​specific rates are published), and this data can also be studied by the commission to see if some guidelines rules are seriously out of synch with the views of judges and practitioners. Published compliance data shows a wide range across guidelines systems, but these figures are very difficult to compare across jurisdictions (Frase, forthcoming a). As previously noted, there is substantial variation as to which crimes are covered by the guidelines; moreover, systems differ in what qualifies as a “departure” (Tonry, 1996). For example, in the District of Columbia departure rates do not include plea-​bargained sentences or sentences by judges who simply state that they choose not to apply the guidelines (Mitchell, 2017); and in Washington state, “first-​offender waivers” and several other options for mitigation are not treated as departures (Frase, 2013). In addition, guidelines grids and worksheets vary a lot in how wide a range is provided, for each grid cell or other offense/​offender group, within which judges may sentence without departing.12 Also, some guidelines grids have “border boxes” or grid zones in which judges have multiple dispositional options (prison, intensive probation, regular probation), and any of these can be chosen without the sentence being considered a departure.

Differences Between Federal and State Guidelines As further discussed below, the majority of criminological research on guidelines sentencing has focused on the federal system.Yet there are significant differences between state and federal systems; these differences suggest that findings based on research at the federal level cannot necessarily be generalized to state guidelines systems. Here is a summary of the most important state-​federal differences, some of which have already been noted above. The federal guidelines are more complex and more punitive than in most state guidelines systems. In particular: 1) The federal grid has by far the largest number of offense severity levels (43), and this translates into narrower sentencing ranges from level to level. 2) Unlike most state systems in which the severity level of offenses is fixed, each offense under the federal system is assigned a base offense level and then the severity level is increased or decreased by specified amounts based on listed Specific Offense Characteristics that apply to each offense, along with generally-​applicable adjustments for, e.g., the defendant’s substantial role in the offense, or the presence of multiple current convictions. 3) Uniquely under the federal guidelines, recommended sentences are often further increased by, and upward departures may be based on, “relevant conduct”—​aggravating facts that are found by the judge using the preponderance of evidence standard and that are often not related to any conviction offense. Instead, they reflect offenses that were dismissed, acquitted, or never charged. 4) In computing criminal history, the weighting of each prior conviction depends on the severity of the sentence that was imposed (not, as in state systems, the legal classification or severity ranking of the offense leading to that conviction),13 and a prior sentence of more than 13 months (even if fully suspended) is sufficient to trigger the highest weight (3 points). 5) Resource-​impact and potential prison overcrowding are not seriously considered when drafting or modifying federal guidelines rules.14 6) Many federal cases are subject to statutory mandatory minimum prison terms, and guidelines sentences build up from those terms (rather than, as in state systems, being set independently and superseded when the mandatory is higher).

53

Richard S. Frase and Kelly Lyn Mitchell

7) Federal plea bargaining concessions are limited by several features not found in state guidelines systems: a) the relevant conduct provisions, noted above, mean that the recommended guidelines sentence is often not reduced even if the prosecutor agrees to not file or to drop additional charges; b) the guidelines rules limiting formal plea concessions to a 3-​level reduction in offense severity level (for “acceptance of responsibility”) mean that judges may not grant any greater sentence mitigation in return for a guilty plea; c) the requirement of a motion by the government in order to reward offenders who provide “substantial assistance” to law enforcement prevents the court from granting such mitigation on its own motion or solely at the defendant’s request. The combined effect of the punitive rules listed above is that federal prison populations have grown much faster than in most states, with or without guidelines (Frase, forthcoming a).

Sentencing Guidelines Research Sentencing guidelines have been the subject of research in economics, law, sociology, and criminology literature since they were first enacted nearly 40 years ago. But the research questions have changed over time. This section explains some of the research implications of varying sentencing guidelines structures and rules, details recent trends in empirical guidelines research and primary theoretical perspectives, and suggests priority issues and areas for further research. This section does not attempt to establish overarching findings, however, because few studies tackle larger questions about the success or failure of sentencing guidelines relative to other forms of sentencing.

Research Implications of the Various Guidelines Structures and Rules The primary goal of most sentencing research is to determine whether variations in sentences imposed are the result of bias or of legally relevant factors. The multiple structures and rules that comprise sentencing guidelines, described earlier, mean that the researcher must have a keen understanding of the system(s) being studied in order to determine whether the variation is due to the structure of the guidelines themselves or the exercise of discretion by actors within the system. As noted above, sentencing guidelines often do not cover all types of offenses in every jurisdiction. In some systems, the guidelines cover both felonies and misdemeanors; in others, they cover only felonies, but not necessarily all felonies. For example, the North Carolina guidelines do not encompass felony driving under the influence. And in Arkansas, the guidelines do not apply to defendants found guilty in a jury trial (the guidelines only apply to defendants who have entered a plea of guilty or no contest, or who are tried by the bench after being permitted by the prosecutor to waive a jury trial). Empirical models may therefore need to account for the applicability (and non-​applicability) of the guidelines. Several important issues are encompassed in the sentencing guidelines grid structure. Though all sentencing grids recommend increased punishment based upon a combination of offense severity and criminal history, that relationship is not necessarily linear (Engen & Gainey, 2000). Punishments often increase by smaller increments for less serious offenses such as property crimes, and by much larger increments for more serious offenses such as sexual assault. Along the criminal history axis, sentence ranges may be narrower at lower criminal history scores and wider at higher criminal history scores, or the magnitude of punishment may increase more dramatically as criminal history scores increase. In jurisdictions that use a categorical versus a point-​based criminal history score, movement across the grid is sporadic rather than incremental because prior person offenses push offenders into higher criminal history categories instantly, while a record consisting solely of prior property crimes will prevent an offender from reaching the highest criminal history categories regardless of the number of prior offenses. Additionally, some jurisdictions have multiple grids, and each may establish 54

Sentencing Guidelines in the US

different mathematical relationships between offense severity and criminal history. Thus, more complex functions may be needed to account for the relationships between offense severity and criminal history and sentencing outcomes. When modeling the factors that drive the sentencing outcome, in most jurisdictions, it is important to control for the guidelines recommended sentence. This is usually the starting point, and represents the punishment established by policy. If unwarranted disparity appears to exist even when all defendants receive the recommended sentence, then other factors besides discretion may be causing it. Departures are the main exercise of discretion available to judges under guidelines systems. But it is important to understand what is and is not considered a departure. For example, in DC, a sentence that results from a plea agreement is not a departure even if the agreed upon sentence is wildly different from the guidelines presumptive sentence. In that case, plea agreement practices could be driving variation that cannot be attributable to the guidelines or the court. But to understand the importance of the recommended guidelines sentence, it is also imperative to understand how binding the guidelines are in a particular jurisdiction. In Minnesota, for example, the recommended sentence must be imposed unless the court can articulate substantial and compelling circumstances that justify a departure, and even then, an appellate court might overrule the decision. By contrast, in DC, the sentencing judge can simply choose not to apply the guidelines, and the resulting sentence will not be considered a departure.Thus, in some jurisdictions the guidelines serve as an anchor against which the actual sentence can be measured while in others the guidelines are merely advisory and may in fact mask the exercise of judicial or prosecutorial discretion. Finally, there may be outside factors such as mandatory minimums that override the guidelines and judicial discretion. In such cases, the root cause of disparity may be the legislative policy or the exercise of prosecutorial discretion in charging, or both.

Examples of Recent Sentencing Guidelines Research In order to get a sense of recent sentencing guidelines research within the field of criminology, the authors reviewed articles published within the last decade (2008–​2018) in six leading journals: Crime and Delinquency, Criminal Justice and Behavior, Criminology, Criminology & Public Policy, Journal of Quantitative Criminology, and Justice Quarterly.15 Within each journal, the authors ran a targeted search for the phrase “sentencing guidelines.” Articles were excluded if they mentioned, but did not analyze sentencing guidelines, fell outside of the ten-​year window, addressed non-​US sentencing guidelines, or took the form of a commentary, response, editorial introduction, book review, or policy essay; articles that focused solely on theory development were also excluded.This approach yielded 53 articles (see Appendix), the majority of which were found in two journals: Crime and Delinquency (17) and Justice Quarterly (16); the rest were fairly evenly distributed among the other four journals. As shown in Table 2.1, sentencing guidelines currently exist in 19 jurisdictions, comprising 17 states, the US Federal Courts, and the District of Columbia. But only ten of the 19 jurisdictions were represented in the sample of articles included in this literature review (Figure 2.1). By far, the US Sentencing Guidelines were the most studied, appearing as the target jurisdiction in 29 of the 53 articles. In the vast majority of articles (44), the authors focused on determining the effects of a single set of sentencing guidelines within one jurisdiction. This is likely due to the fact that the design and implementation of guidelines vary considerably from state to state (Frase, 2005b), complicating the determination of causality in sentencing outcomes. Nevertheless, in four articles, the authors conducted a comparative analysis across two or more sentencing guidelines systems, and in five articles, the authors broadened to a cross-​jurisdictional analysis across both guidelines and non-​ guidelines states by using alternative data sources such as the State Court Processing Statistics. These findings suggest that the vast majority of our current knowledge about the operation of sentencing guidelines is based on the federal system, which handles just a fraction of all criminal convictions sentenced in the United States each year. 55

Richard S. Frase and Kelly Lyn Mitchell

Figure 2.1  Sentencing Guidelines Jurisdictions Researched in Sample of Articles 2008–18

Because sentencing guidelines were designed to reduce sentencing discretion and disparities, it should come as no surprise that explaining disparity (or variation) in sentencing was the goal in all but four of the 53 articles; the remaining four articles focused on the effect of various sentences or punishment types on recidivism. When guidelines were first enacted, sentencing research focused on whether sentencing guidelines reduced disparity by comparing sentences imposed under the pre-​ and post-​guidelines regimes (e.g., Moore & Miethe, 1986). Within the 53-​article sample of more recent work, such comparisons were rarely made. Instead, current research takes it as a given that there will still be major variation in sentencing under guidelines and attempts to discern how much of that variation is attributable to extra-​legal factors such as race, gender, and socioeconomic status versus legal factors such as offense severity and criminal history. It should be noted that, although the term “disparity” is typically used, it is the identification of discrimination that is the goal of these studies. As described by Spohn (2000), Disparity refers to a difference in treatment or outcome, but one that does not necessarily involve discrimination … Disparity exists when “like cases” with respect to case attributes—​regardless of their legitimacy—​are sentenced differently … Discrimination, on the other hand, is a difference that results from differential treatment based on illegitimate criteria, such as race, gender, social class, or sexual orientation. With respect to sentencing, discrimination “exists when some case attribute that is objectionable (typically on moral or legal grounds) can be shown to be associated with sentence outcomes after all other relevant variables are adequately controlled.” (pp. 432–​433) The majority of the articles in this sample focused on examining the relationship between sentencing and particular contexts or offender characteristics, suggesting that disparity research may be moving toward the micro rather than macro (system-​wide) perspective. Important contexts that were studied within the sample were the impact of the Booker case on the use of judicial discretion, sentencing patterns for specific crimes such as immigration or sex offenses, and the impact on the sentencing outcome of pretrial decisions such as detention or whether to enter a guilty plea or go to trial. With regard to extra-​legal factors, several articles continued the traditional focus on race or ethnicity, but researchers also explored relationships between sentencing and other individual offender characteristics such as educational attainment, a history of hard drug use, and whether the offender was a juvenile tried as an adult. Some articles delved deeper, exploring interaction effects between multiple characteristics. For example, in a study of federal sentences pronounced in 2001, Doerner and Demuth (2010) examined interaction effects between age, race, and gender (all extra-​legal 56

Sentencing Guidelines in the US

factors), and found that young Hispanic male defendants had the highest odds of incarceration while young black male defendants received the longest sentences. And in examining sentences for federal narcotics cases, Tillyer, Hartley, and Ward (2015) found an interaction effect between criminal history (a legal factor) and gender (an extra-​legal factor) such that female defendants with low criminal history scores received more lenient treatment than males but female defendants with higher criminal history scores were sentenced more harshly than males. A few articles took a different approach and focused on discerning how much variation in sentencing is attributable to the policy structures themselves versus the exercise of individual judicial discretion. Bushway and Forst (2013) describe this as the difference between Type A and Type B discretion: The choice by legislators to impose sentencing guidelines is an act of Type B discretion—​ the legally allowable choice by judges of sentences within the sentencing guideline ranges is an act of “weak” or Type A discretion. Type B discretion, the ability to create rules and policies, can be used to limit and shape Type A discretion, and there is often a tension between the rules set by actors (Type B) and the discretion available to lower level actors within those rules (Type A). (p. 201) Bushway and Forst (2013) note that individual or Type A discretion is the more studied of the two, but make the case for greater attention being paid to Type B discretion in order to discern whether racial disparities have been codified into the legal structure that governs sentencing. Though many different outcome measures were utilized in this sample of articles, the two most common measures were type of disposition, which refers to the decision to sentence an individual to a prison or non-​prison sanction; and sentence length, which refers to the length of the prison sentence in months. By contrast, it was rare to find studies that examined other dispositions types such as probation, jail, or intermediate sanctions as outcome measures, or that measured differences in jail terms or probation length. The third most common outcome measure involved departures, in some cases merely testing for their presence or absence, in others, measuring the likelihood and magnitude of departure.

Theoretical Perspectives Although there were many different theories utilized in the 53 articles in this review, one theory dominated the work: focal concerns. As will be explained further below, this theory is built upon a constellation of theories from different disciplines, including uncertainty avoidance and causal attribution, bounded rationality, and perceptual shorthand. Additional theories that stood out were the court community perspective, racial or minority threat, and the liberation hypothesis.16

I  Focal Concerns

The focal concerns framework was utilized for all types of sentence disparity research. The theory posits that judges are influenced by three focal concerns: the offender’s blameworthiness, protection of the community, and practical implications of sentencing decisions. The first focal concern—​ blameworthiness—​is associated with the “just deserts” or retributive philosophy of punishment. The seriousness of the offense is often the most significant factor informing this concern, but offender factors are also relevant, including “criminal history (which increases perceptions of blameworthiness and risk) or prior victimization at the hands of others (which tends to mitigate perceived blameworthiness), and the offender’s role in the offense, such as whether the offender was a leader, organizer, or a follower” (Steffensmeier, Ulmer, & Kramer, 1998, pp. 766–​767). The second focal concern—​protection of the community—​focuses on the offender’s danger to the community or likelihood of reoffense, and relates more closely to the need to incapacitate the offender or deter future criminal conduct (Steffensmeier, Ulmer, & Kramer, 1998). The third focal concern—​the 57

Richard S. Frase and Kelly Lyn Mitchell

practical implications of sentencing decisions—​relates to a mix of organizational and individual considerations. Organizational considerations include ensuring that cases are moving through the system, and being sensitive to the capacities of state and local correctional resources. Individual considerations may relate to the offender, and include factors such as the health of the offender and potential disruption of ties to children and other family members. Individual considerations may also relate to the judge, including factors such as the judge’s political standing within the community and his or her perception of the effect of the defendant’s recidivism on that standing (Steffensmeier, Ulmer, & Kramer, 1998). These focal concerns are undergirded by a constellation of theories fusing together to explain judicial sentencing decisions. The first is Albonetti’s (1991) uncertainty avoidance and causal attribution. As Albonetti explains, judges strive to make rational decisions, but to do so, they need complete information, and judges rarely have this. Instead, uncertainty surrounding the sentencing decision arises from an inability to predict accurately future criminal behavior. Using defendant characteristics, circumstances of the crime, and case processing outcomes, judges assess the defendant’s disposition toward future criminal activity … These attributions provide a basis for arriving at rational decision in a domain of responsibility characterized by uncertainty. (Albonetti, 1991, p. 250) This process results in “bounded rationality,” in which judges are forced to make sentencing decisions based on “too little information to accurately assess the focal concerns” (Franklin & Henry, 2018, p. 119). “[A]‌s court actors attempt to efficiently process and sentence offenders with imperfect information, often under considerable caseload pressure, they may draw upon prevalent criminal stereotypes to aid in their judgments about offender blameworthiness and dangerousness” (Franklin, 2015, p. 656). These “patterned responses” or “perceptual shorthands” are based not only on legal factors, “but also on a variety of stereotypes, including prevalent race-​based stereotypes that identify different groups as more or less crime prone and dangerous” (Franklin & Fearn, 2015, p. 98). Thus, discrimination in sentencing decisions may be the result of judges attempting “to achieve a ‘bounded rationality’ in sentencing by relying on stereotypical images of which defendant is most likely to recidivate” (Albonetti, 1991, 250).

II  Court Community Perspective

The court community perspective is often utilized when analyzing inter-​court or racial disparity. Developed by Eisenstein, Flemming, and Nardulli (1988, p. 14), the court community perspective posits that there is wide variation in “the informal structure of relations among the people who work in courts, the values and informal norms that guide their behavior, and shared understandings about how they treat one another and dispose of cases.” Like other theories mentioned here, the court community perspective may also be used in conjunction with focal concerns to explain the development of perceptual shorthand. Under this theory, “[c]‌ourt communities are said to exhibit different, locally varying substantive legal rationalities … that influence sentencing outcomes and processes at least as much as formal policies and legal structures” (Ulmer, Light, & Kramer, 2011, p. 806). This perspective emphasizes the importance of court size and posits that variations in court size directly produce several crucial court community differences … Larger court communities are said to have reduced media visibility in routine case processing, greater bureaucratization of sponsoring agencies (e.g., U.S.  attorneys’ offices and federal public defenders), greater ease of plea bargaining, and a normative tolerance or desensitization of deviant/​criminal behavior. (Johnson, Ulmer, & Kramer, 2008, p. 746)

58

Sentencing Guidelines in the US

In short, this theory posits that sentencing decisions are as much a product of the court in which they are imposed as the legal structures within which sentencing occurs.

III  Racial or Minority Threat

Racial or minority threat is a critical theory asserting that those who occupy positions of power will attempt to maintain their status through subordination of minority groups (Franklin, 2013). Within the criminal justice system, this means that incarceration and other forms of criminal punishment are used as mechanisms of control (Rengifo & Stemen, 2015). As a particular minority group increases in size, the dominant group begins to feel uneasy and fears that crime will increase as a direct result of the increased minority presence. Studies have demonstrated that this results in increased spending on policing and corrections, which correlates with increased incarceration rates for the minority group (Johnson, Ulmer, & Kramer, 2008). This racial threat perspective has been utilized in research about sentencing discretion both as a stand-​alone theory and as an augmentation to the focal concerns theory. When used in conjunction with focal concerns, racial threat theory explains the mechanism by which negative racial imagery influences the assessment of key focal concerns (Ulmer & Johnson, 2004).

IV  Liberation Hypothesis

The liberation hypothesis is a theoretical perspective suggesting that extra-​legal influences on legal decisions increase as the opportunity for discretion increases (Ulmer, Light, & Kramer, 2011). It was first developed in the context of jury decisions (Kalven & Zeisel, 1966), where researchers found that “certain case conditions (e.g., contradictory or weak evidence) … ‘liberated’ jurors to consider extra-​legal factors such as their own personal values or their beliefs about the victim when determining case outcomes” (Hauser & Peck, 2017, pp. 167–​168).When applied to sentencing decisions, the liberation hypothesis starts from the premise that “decision-​making will be strongly bound to legally relevant factors by the seriousness of the offense” (Hauser & Peck, 2017, p. 168). Decisions will be based more on legal criteria and less on personal feelings when the crime is serious but may be based on extra-​legal factors and personal typescripts when the crime is less serious because in such cases, it is harder to discern the appropriate punishment from the nature of the offense. A  version of the liberation hypothesis has served as the underlying theoretical perspective to examine situations in which guidelines shift from mandatory to advisory as the federal sentencing guidelines did after Booker (Hauser & Peck, 2017). Here, the theory posits that “[i]‌f judges become ‘liberated’ in their discretion relative to past restrictions, they might exhibit a greater tendency to consider extra-​legal factors, resulting in greater social status disparities” (Ulmer, Light, & Kramer, 2011, p. 805).

Priority Issues and Areas for Further Research Although numerous articles examining the impact of sentencing guidelines have been written, the majority of these articles involve the examination of a single guidelines system, or at best, two or three systems at once. Because of this single-​jurisdiction approach, existing research is unable to clearly discern sentencing patterns resulting from different guidelines structures, whether any particular guidelines system functions better than another, or whether particular features are more desirable than others. Thus, the greatest research need is cross-​jurisdictional analysis involving multiple distinct sentencing guidelines systems. Within the general heading of cross-​jurisdictional analysis there are three potential lines of research. The first is research that attempts to discern how well different guidelines structures function in order to determine whether certain structural features uphold or hinder the goals for which guidelines were created, such as fairness, uniformity, proportionality, and correctional resource management.

59

Richard S. Frase and Kelly Lyn Mitchell

While there has been research that has attempted to address these concepts within jurisdictions (see, e.g., Miethe & Moore, 1985; Tonry, 1996), what is needed is comparative research across jurisdictions and differing types of guidelines to support sentencing commissions in making better policy choices. Within this line of research, relevant questions might include: • Is it possible to devise comparable measures of guidelines departure or compliance rates across jurisdictions?17 • What potential sources of bias exist in the varying guidelines structures? Which policy choices best reduce these sources of bias? • How do different kinds of guidelines relate to levels of prison racial disproportionality? Are some kinds of guidelines better able to curb disparity? Is it possible to devise comparable measures of disparity reduction? • Are some guidelines more susceptible to being undermined by prosecutorial charging and plea bargaining decisions? • Do offenders’ prior records tend to receive more weight at sentencing in jurisdictions with guidelines than in those without guidelines?18 • To what extent are sentencing guidelines undermined by other policy choices such as mandatory minimums and “off-​grid” sentencing? A second line of cross-​ jurisdiction research might provide more detail about non-​ prison sentences. By focusing so heavily on the disposition decision (prison or non-​prison) and the length of the prison sentence as outcome measures, researchers are missing other areas of potential disparity such as jail sentences, length of probation, and the imposition of intermediate punishments. Sentencing guidelines do not often control these decisions, and that may be the reason for their lack of study, but some guidelines systems do control these decisions, and should be studied. In others, it may be fair to hypothesize whether adding the formal structure of sentencing guidelines has the spillover effect of making other aspects of sentencing more uniform and proportional as well. A third line of cross-​jurisdictional research should be aimed at determining the extent to which sentencing guidelines have contributed to or tempered prison growth in the age of mass incarceration. Every guidelines system has a stated goal to manage prison and correctional resources to avoid prison overcrowding. But did the predictability of sentencing guidelines merely serve to provide advance warning when additional prison resources would be needed? Or did it serve to slow prison growth by promoting a more informed and deliberate decision-​making process? Or did all of the above occur in tandem with varying guidelines structures? Though some analyses suggest that guidelines states may perform better on this front (American Law Institute 2011, pp. 154–​155; Frase, forthcoming a), the differences in prison growth have never been subject to rigorous empirical analysis. Thus, relevant questions within this line of research would seek to determine: • Can we agree on how to measure prison population growth, to permit assessment of the effects of different guidelines structures?19 How do these measures compare to prison population growth in non-​guidelines systems? What about growth in supervision populations? • How do the variations in system context and guidelines structure affect rates of prison growth? For example, researchers could examine how the following factors affect growth rates: imprisonment scale (low vs high per capita rate at the outset); determinacy (retention vs abolition of parole discretion);20 guidance re: probation conditions; different ways of structuring sanctions for violation of the conditions of non-​prison sentences (i.e.., abolishing or placing limits on revocation to prison); location on the mandatory to advisory continuum; form of guidance (grid vs non-​g rid); variations in criminal history scoring and in the magnitude of sentence enhancements based on 60

Sentencing Guidelines in the US

criminal history; extent to which resource limitations and impact are considered when drafting guidelines rules or assessing proposed legislation). • How do guidelines jurisdictions compare to non-​guidelines jurisdictions in prison racial disproportionality, after controlling for other relevant differences between the jurisdictions? Additionally, as greater awareness is raised about the policy choices that led to mass incarceration, it will be important to gauge professional and public opinion about potential policy changes to reverse this trend. Thus, research will be needed to determine: • What do the major criminal justice actors (judges, prosecutors, defense counsel, offenders, probation officers, other correctional officers) think about guidelines rules, and how do their attitudes support or undermine guidelines reforms? • What does the public think about guidelines rules? How well do such attitudes reflect the actual content of the rules? In what ways does public opinion support or undermine guidelines reforms?

Conclusion It has been 40 years since the first independent sentencing guidelines commission was established, and just under 40 years since the first set of guidelines was enacted.Yet we still do not have a clear empirical understanding of whether and in what ways sentencing guidelines are a superior policy choice to other options such as statutory structured sentencing or indeterminate sentencing with parole release. We have also not yet subjected the varying guidelines structures and rules to rigorous empirical analysis to determine which best achieve reform goals, and the ways in which success depends on the particular social, economic, and political context. By focusing on cross-​jurisdictional analysis, the research agenda outlined above should begin to provide empirical answers to these questions, but only if researchers understand and take into account the many important variations in guidelines structures and rules.

Notes 1 Federal judge Marvin Frankel proposed the idea of commission-​drafted sentencing guidelines in the early 1970s, and in 1978 the Minnesota and Pennsylvania legislatures each created a commission with a mandate to develop and implement statewide sentencing guidelines; Minnesota’s guidelines went into effect in 1980, and Pennsylvania’s guidelines became effective two years later. For further details on how this article defines a “guidelines” system, and on the location and timing of American guidelines reforms, see infra, text at notes  2–​3. 2 Sentencing guidelines commissions were abolished in Florida and Tennessee, but the commission-​drafted guidelines or successor laws are still in effect in those states. In Louisiana and Missouri guidelines were repealed but non-​guidelines sentencing commissions remain. The state of Wisconsin implemented guidelines at two points in time, but in each case both the guidelines and the commission were later abolished (Frase, 2013; Watts,  2016). 3 Florida’s commission-​ drafted guidelines were replaced in 1998 with statutory presumptive minimum sentences; however, those statutes carry over elements of the former commission-​drafted guidelines system. We also count Massachusetts as a guidelines state, even though its commission-​drafted guidelines have not received legislative approval, because judges generally choose to sentence within the guidelines ranges (in effect, this is an advisory guidelines system). Alaska is sometimes classified as a “guidelines” system (see, e.g., Kauder & Ostrom, 2008). However, that state’s statutory presumptive sentences were drafted by the legislature, not a commission (additional presumptive sentences have been recognized in appellate case law, see Carnes, 1993). 4 For prior reviews of American guidelines systems, as of 2004 and 2012, see Frase (2005, 2013). See also Tonry (1988, 1996); Frase (forthcoming a); and Frase (forthcoming b). For surveys and critiques of one very important aspect of guidelines sentencing—​criminal history enhancements—​see Frase et al. (2015), Hester et al. (forthcoming), and Frase and Roberts (forthcoming).

61

Richard S. Frase and Kelly Lyn Mitchell 5 Most grids have some kind of disposition zone structure. A few grids (e.g., Oregon’s sole grid) have “border boxes”—​cells in which either prison or probation is permitted without departure. Some grids have large zones in which judges have a choice of two or more disposition types (one-​third of the cells on North Carolina’s felony grid are of that nature). 6 On most grids the highest-​severity offenses are on the top row of the grid and the highest criminal history category is the far-​r ight column, but on some grids one or both of these orders is reversed. 7 Most guidelines systems retain the judge’s traditional discretion to choose between concurrent and consecutive sentencing of multiple current offenses, but a few systems have presumptions in favor of concurrent and/​ or consecutive sentencing; some of them also use intermediate options such as adding current offenses to the offender’s criminal history score. For further discussion see Frase et al. 2015 and Hester 2018a. 8 See USSC 2016, Sec. 1B1.3 (defining “relevant conduct” that courts may consider when calculating the recommended guidelines sentence, or when departing from that sentence). Most other guidelines systems allow upward departures to be based on aggravating facts of the conviction offense(s) that were not proven or admitted by the defendant (e.g., committing that offense with unusual cruelty), but such facts only occasionally increase the ordinary-​case recommended sentence (for example, when the conviction offense involves weapon use or victim injury). Except in the federal system, guidelines rarely permit the alleged commission of non-​conviction offenses to serve as grounds for departure, and almost never allow such allegations to increase the ordinary recommended sentence. 9 For further discussion of this aspect of guidelines sentencing, see the separate chapter on Criminal History Enhancements in this volume (Hester et al., forthcoming). See also Frase et al. (2015); Hester et al. (2018); and Frase and Roberts, forthcoming. 10 A departure is a sentence other than that recommended in the sentencing guidelines. Every guidelines system permits judges to pronounce sentences that are harsher than the recommended guidelines sentence (aggravated departure) or less severe (mitigated departure) (Mitchell 2017). 11 (Wright, 2002, p. 94). However, the North Carolina guidelines should still be classified as (relatively) “mandatory” because they completely prohibit extreme durational departures (sentences above or below the aggravated and mitigated sentencing ranges); they also prohibit upward dispositional departures (e.g., imposition of a prison sentence rather than the recommended probation), while providing very narrow grounds for downward dispositional departure. See Wright (2002, p. 79), noting that judges may not impose a custody sentence above the aggravated range or below the mitigated range. See also Frase (2013, p. 160), noting that upward dispositional departures are prohibited in North Carolina, and that mitigated dispositional departures are allowed but seem to occur only rarely. 12 For example, in Florida, sentencing ranges extend all the way up to the statutory maximum, so there is no such thing as an upward durational departure (Florida Departments, 2014, p. 13). 13 The decision to use imposed sentences rather than severity ranking to weight prior convictions was most likely due to the high proportion of non-​federal priors, and the reality that state law offense definitions often vary dramatically across states and between state and federal courts. 14 The lack of concern for resource-​impact is most likely because, unlike state governments which must balance their budgets each cycle, the federal government can run a deficit. In addition, even rapidly escalating prison costs remain a tiny fraction of the huge federal budget. 15 The journals were identified by reviewing the 2017 Web of Science rankings for journals in the criminology & penology category. Journals were not included in the review if they were too narrowly focused (e.g., Trauma Violence & Abuse) or contained primarily theoretical works (e.g., Theoretical Criminology). 16 Multiple discrete theoretical concepts were utilized in the sample of articles. This part focuses only on the most commonly used theories appearing in more than one article. 17 As noted earlier, guidelines systems differ substantially in ways that make compliance rates difficult to compare, such as in the width of allowed sentencing ranges and what is defined as a “departure.” Another major source of potential non-​comparability across systems is the hidden effect of charging and plea bargaining decisions. Most research on state guidelines systems has controlled for offense type and severity by using measures of conviction offenses, but this can yield apples-​oranges comparisons if some systems make much greater use of charge reductions but sentence, at least in part, according to what is perceived to be the “real offense.” In this regard, research on sentencing under the federal guidelines is less problematic; due to that system’s unique Relevant Conduct provisions, discussed above, federal sentencing data partially controls for “real-​ offense” severity, factoring out at least some of the effects of charging and bargaining differences across federal districts. For further discussion of comparability issues in cross-​jurisdictional guidelines research, see Frase (forthcoming a). 18 For further discussion of this and other research topics related to prior record enhancements, see Hester et al. (forthcoming). 19 See Reitz (2018), arguing that prison growth should be measured in terms of the absolute number of per capita units added to a given jurisdiction’s incarceration rate, not the relative growth (added per capita units as a percentage of the starting-​year per capita incarceration rate).

62

Sentencing Guidelines in the US 20 See, e.g., Stemen and Rengifro (2011) examining incarceration rates for all states from 1978 to 2004 and finding that determinate sentencing policies constraining prison release discretion were more effective at controlling incarceration rates than structured sentencing policies constraining sentencing discretion. See also Frase (forthcoming a), examining the evidence that parole-​abolition guidelines have had slower rates of prison growth than guidelines systems that retained parole release discretion.

References Albonetti, C. A. (1991). An integration of theories to explain judicial discretion. Social Problems, 38(2), 247–​266. American Bar Association. (1979). Standards Relating to the Administration of Criminal Justice: Sentencing Alternatives and Procedures. Washington, DC: Author. —​—​. (1994). ABA Standards for Criminal Justice: Sentencing. Washington, DC: Author. American Law Institute. (2011). Model Penal Code: Sentencing. Philadelphia, PA: Author. —​—​. (2017). Model Penal Code: Sentencing. Philadelphia, PA: Author. Arkansas Sentencing Commission. (2018). Sentencing Standards Grid Offense Seriousness Rankings & Related Material. Little Rock, AR: Author. Blumstein, A., Cohen, J., Roth, J., & Tonry, M. (eds.). (1983). Research on Sentencing: The Search for Reform. Washington, DC: National Academy Press. Bushway, S. D., & Forst, B. (2013). Studying discretion in the processes that generate criminal justice sanctions. Justice Quarterly, 30(2), 199–​222. Carnes, T. (1993). Sentencing reform in Alaska. Federal Sentencing Reporter, 6, 134–​137. Doerner, J. K., & Demuth, S. (2010).The independent and joint effects of race/​ethnicity, gender, and age on sentencing outcomes in US federal courts. Justice Quarterly, 27(1),  1–​27. Eisenstein, J., Flemming, R. B., & Nardulli, P. F. (1988). The Contours of Justice: Communities and Their Courts. Boston, MA: Little, Brown. Engen, R. L., & Gainey, R. R. (2000). Modeling the effects of legally relevant and extralegal factors under sentencing guidelines: The rules have changed. Criminology, 38(4), 1207–​1230. Florida Department of Corrections and Office of the State Courts Administrator (“Florida Departments”) (2014). Florida Criminal Punishment Code Scoresheet Preparation Manual. Tallahassee, FL: Authors. Frankel, M. (1973). Criminal Sentences: Law without Order. New York: Hill & Wang. Franklin,T.W. (2015). Race and ethnicity effects in federal sentencing: A propensity score analysis. Justice Quarterly, 32(4), 653–​679. —​—​. (2013). Sentencing Native Americans in US federal courts: An examination of disparity. Justice Quarterly, 30(2), 310–​339. Franklin, T. W., & Fearn, N. E. (2015). Sentencing Asian offenders in state courts: The influence of a prevalent stereotype. Crime & Delinquency, 61(1), 96–​120. Franklin, T. W., & Henry, T. K. S. (2018). One day makes all the difference: Denying federal offenders access to “good time” through sentencing. Crime & Delinquency, 64(1), 115–​140. Frase, R. (1995). Lessons of state guideline reforms. Federal Sentencing Reporter, 8,  39–​41. —​—​. (1999). Sentencing guidelines in Minnesota, other states, and the federal courts: A twenty-​year retrospective. Federal Sentencing Reporter, 12,  69–​82. —​—​. (2005a). Punishment purposes. Stanford Law Review, 58, 67–​83. —​—​. (2005b). State sentencing guidelines: Diversity, consensus, and unresolved policy issues. Columbia Law Review, 105, 1190–​1232. —​—​ . (2007).The Apprendi-​Blakely cases: Sentencing reform or counter-​revolution? Criminology and Public Policy, 6(3), 403–​432. —​—​ . (2013). Just Sentencing: Principles and Procedures for a Workable System. New  York: Oxford University Press. —​—​ . (2015a). Varying Binding Effects of Guidelines—​The Mandatory-​to-​Advisory Continuum. Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. —​—​. (2015b). Why have U.S. state and federal jurisdictions enacted sentencing guidelines? Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. —​—​. (forthcoming a). Forty Years of American Sentencing Guidelines: What Have We Learned?. Crime and Justice: A Review of Research. Chicago, IL: University of Chicago Press. —​—​. (forthcoming b). Suspended sentences and free-​standing probation orders in U.S. guidelines systems: A survey and assessment. Law and Contemporary Problems. Frase, R., & Roberts, J. (forthcoming). Paying For the Past: The Case Against Prior Record Sentence Enhancements. New York: Oxford University Press.

63

Richard S. Frase and Kelly Lyn Mitchell Frase, R., Roberts, J., Hester, R., & Mitchell, K. (2015). Criminal History Enhancements Sourcebook. Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. Hauser,W., & Peck, J. H. (2017).The intersection of crime seriousness, discretion, and race: A test of the liberation hypothesis. Justice Quarterly, 34(1), 166–​192. Hester, R., Frase, R., Roberts, J., & Mitchell, K. (2018). Prior record enhancements at sentencing: Unsettled justifications and unsettling consequences. In Tonry, M. (ed.), Crime and Justice: A Review of Research (pp. 209–​ 354). Chicago, IL: University of Chicago Press. Hester, R., Frase, R., Laskorunsky, J., & Mitchell, K. (forthcoming). Rethinking the Role of Criminal History in Sentencing. In C. Spohn & P. K. Brennan (eds.), Handbook on Sentencing Policies and Practices in the 21st Century (­chapter 13). Johnson, B. D., Ulmer, J. T., & Kramer, J. H. (2008). The social context of guidelines circumvention: The case of federal district courts. Criminology, 46(3), 737–​783. Kalven, H., Jr., & Zeisel, H. (1966). The American Jury. Boston, MA: Little, Brown. Kauder, N., & Ostrom, B. (2008). State Sentencing Guidelines: Profiles and Continuum. Williamsburg, VA: National Center for State Courts. Miethe, T. D., & Moore, C. A. (1985). Socioeconomic disparities under determinate sentencing systems: A comparison of preguideline and postguideline practices in Minnesota. Criminology, 23(2), 337–​363. Mitchell, K. (2017). State sentencing guidelines: A garden full of variety. Federal Probation, 81(2),  28–​36. Moore, C., & Miethe,T. (1986). Regulated and unregulated sentencing decisions: An analysis of first-​year practices under Minnesota’s felony sentencing guidelines. Law & Society Review, 20(2), 253–​277. Morris, N., & Tonry, M. (1990). Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System. New York: Oxford University Press. Parent, D. (1988). Structuring Criminal Sentences: The Evolution of Minnesota’s Sentencing Guidelines. Stoneham, MA: Butterworth Legal Publishers. Reitz, K. (1997). Sentencing guidelines systems and sentence appeals: A comparison of federal and state experiences. Northwestern University Law Review, 91, 1441–​1506. Reitz, K. (2001). The disassembly and reassembly of U.S. sentencing practices. In M. Tonry & R. S. Frase (eds.), Sentencing and Sanctions in Western Countries (pp. 222–​258). New York: Oxford University Press. Reitz, K. (2018). Measuring change in incarceration scale. Berkeley Journal of Criminal Law, 23(2),  1–​35. Rengifo, A. F., & Stemen, D. (2015). The unintended effects of penal reform: African American presence, incarceration, and the abolition of discretionary parole in the United States. Crime & Delinquency, 61(5), 719–​741. Roberts, J., & von Hirsch, A. (eds.). (2010). Previous Convictions at Sentencing: Theoretical and Applied Perspectives. Oxford, UK: Hart Publishing. Robina Institute of Criminal Law and Criminal Justice. (2018). Sentencing Guidelines Resource Center. Minneapolis, MN: Author. Spohn, C. (2000).Thirty years of sentencing reform:The quest for a racially neutral sentencing process. In Policies, Processes and Decisions of the Criminal Justice System (Vol. 3, pp. 427–​501). Washington, DC: US Department of Justice. Steffensmeier, D., Ulmer, J., & Kramer, J. (1998). The interaction of race, gender, and age in criminal sentencing: The punishment cost of being young, black, and male. Criminology, 36(4), 763–​798. Stemen, D., & Rengifo, A. F. (2011). Policies and imprisonment: The impact of structured sentencing and determinate sentencing on state incarceration rates, 1978–​2004. Justice Quarterly, 28(1), 174–​201. Tillyer, R., Hartley, R. D., & Ward, J. T. (2015). Differential treatment of female defendants: Does criminal history moderate the effect of gender on sentence length in federal narcotics cases? Criminal Justice and Behavior, 42(7), 703–​721. Tonry, M. (1988). Structuring sentencing. In M. Tonry (ed.), Crime and Justice: A Review of Research (Vol. 10, pp. 267–​337). Chicago, IL: University of Chicago Press. —​—​ . (1996). Sentencing Matters. New York: Oxford University Press. —​—​ . (2016). Sentencing Fragments: Penal Reform in America, 1975–​2025. New York: Oxford University Press. Ulmer, J. T., & Johnson, B. (2004). Sentencing in context: A multilevel analysis. Criminology, 42(1), 137–​178. Ulmer, J., Light, M. T., & Kramer, J. (2011). The “liberation” of federal judges’ discretion in the wake of the Booker/​Fanfan decision: Is there increased disparity and divergence between courts? Justice Quarterly, 28(6), 799–​837. US Department of Justice. (1996). National Assessment of Structured Sentencing. Washington, DC: Author. —​—​. (1998). National Survey of State Sentencing Structures. Washington, DC: Author. United States Sentencing Commission. (2016). Federal Sentencing Guidelines Manual. Washington, DC: Author. von Hirsch, A. (1987).The sentencing commission’s functions. In A. von Hirsch, K. Knapp & M.Tonry (eds.), The Sentencing Commission and its Guidelines (pp. 3–​15). Boston, MA: Northeastern University Press.

64

Sentencing Guidelines in the US Watts, A. (2016). Timelines of Sentencing Commissions and Sentencing Guidelines Enactments: 1978 to the Present. Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. —​—​ . (2018a). Sentencing Guideline Grids. Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. —​—​ . (2018b). Sentencing Guidelines and Discretionary Parole Release. Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. Wright, R. (2002). Counting the cost of sentencing in North Carolina, 1980–​2000. In M. Tonry (ed.), Crime and Justice: A Review of Research (Vol. 29, pp. 39–​112). Chicago, IL: University of Chicago Press.

Cases Cited Alleyne v. United States, 570 U.S. 99 (2013). Blakely v. Washington, 542 U.S. 296 (2004). People v. Lockridge, 870 N.W. 2d 502 (Mich. 2015). United States v. Booker, 543 U.S. 220 (2005).

Appendix List of Authorities Included in Review of Recent Sentencing Guidelines Research Anderson, A. L., and Spohn, C. (2010). Lawlessness in the federal sentencing process: A test for uniformity and consistency in sentence outcomes. Justice Quarterly, 27(3), 362–​393. Bales, W. D., and Piquero, A. R. (2012). Racial/​ethnic differentials in sentencing to incarceration. Justice Quarterly, 29(5), 742–​773. Bradley-​Engen, M. S., Engen, R. L., Shields, C., Damphousse, K. R., and Smith, B. L. (2012). The time penalty: Examining the relationship between time to conviction and trial vs. plea disparities in sentencing. Justice Quarterly, 29(6), 829–​857. Britt, C. L. (2009). Modeling the distribution of sentence length decisions under a guidelines system: An application of quantile regression models. Journal of Quantitative Criminology, 25(4), 341–​370. Bushway, S. D., and Forst, B. (2013). Studying discretion in the processes that generate criminal justice sanctions. Justice Quarterly, 30(2), 199–​222. Bushway, S. D., and Piehl, A. M. (2007). Social science research and the legal threat to presumptive sentencing guidelines. Criminology & Public Policy, 6(3), 461–​482. Cano, M. V., and Spohn, C. (2012). Circumventing the penalty for offenders facing mandatory minimums: Revisiting the dynamics of “sympathetic” and “salvageable” offenders. Criminal Justice and Behavior, 39(3), 308–​332. Caravelis, C., Chiricos, T., and Bales, W. (2013). Race, ethnicity, threat, and the designation of career offenders. Justice Quarterly, 30(5), 869–​894. Cassidy, M., and Rydberg, J. (2018). Analyzing variation in prior record penalties across conviction offenses. Crime & Delinquency, 64(7), 831–​855. Cochran, J. C., Mears, D. P., and Bales, W. D. (2014). Assessing the effectiveness of correctional sanctions. Journal of Quantitative Criminology, 30(2), 317–​347. Doerner, J. K., and Demuth, S. (2010). The independent and joint effects of race/​ethnicity, gender, and age on sentencing outcomes in US federal courts. Justice Quarterly, 27(1),  1–​27. Franklin, T. W. (2017). Sentencing outcomes in US district courts: Can offenders’ educational attainment guard against prevalent criminal stereotypes?. Crime & Delinquency, 63(2), 137–​165. Franklin,T.W. (2015). Race and ethnicity effects in federal sentencing: A propensity score analysis. Justice Quarterly, 32(4), 653–​679. Franklin, T. W. (2013). Sentencing Native Americans in US federal courts: An examination of disparity. Justice Quarterly, 30(2), 310–​339. Franklin, T. W., Dittmann, L., and Henry, T. K. S. (2017). Extralegal disparity in the application of intermediate sanctions: An analysis of US district courts. Crime & Delinquency, 63(7), 839–​874. Franklin, T. W., and Fearn, N. E. (2015). Sentencing Asian offenders in state courts: The influence of a prevalent stereotype. Crime & Delinquency, 61(1), 96–​120. Franklin, T. W., and Henry, T. K. S. (2018). One day makes all the difference: Denying federal offenders access to “good time” through sentencing. Crime & Delinquency, 64(1), 115–​140. Freiburger,T. L., and Hilinski, C. M. (2013). An examination of the interactions of race and gender on sentencing decisions using a trichotomous dependent variable. Crime & Delinquency, 59(1),  59–​86.

65

Richard S. Frase and Kelly Lyn Mitchell Harmon, M. G. (2013). “Fixed” sentencing: The effect on imprisonment rates over time. Journal of Quantitative Criminology, 29(3), 369–​397. Hartley, R. D., and Tillyer, R. (2012). Defending the homeland: Judicial sentencing practices for federal immigration offenses. Justice Quarterly, 29(1), 76–​104. Hauser, W., and Peck, J. H. (2017). The intersection of crime seriousness, discretion, and race: a test of the liberation hypothesis. Justice Quarterly, 34(1), 166–​192. Haynes, S. H., Ruback, B., and Cusick, G. R. (2010). Courtroom workgroups and sentencing:The effects of similarity, proximity, and stability. Crime & Delinquency, 56(1), 126–​161. Hofer, P. J. (2007). United States v. Booker as a natural experiment: Using empirical research to inform the federal sentencing policy debate. Criminology & Public Policy, 6(3), 433–​460. Johnson, B. D., and Kurlychek, M. C. (2012).Transferred juveniles in the era of sentencing guidelines: Examining judicial departures for juvenile offenders in adult criminal court. Criminology, 50(2), 525–​564. Johnson, B. D., and DiPietro, S. M. (2012). The power of diversion: Intermediate sanctions and sentencing disparity under presumptive guidelines. Criminology, 50(3), 811–​850. Johnson, B. D., Ulmer, J. T., and Kramer, J. H. (2008). The social context of guidelines circumvention: The case of federal district courts. Criminology, 46(3), 737–​783. Kaiser, K. A., and Spohn, C. (2014). “Fundamentally flawed?” Exploring the use of policy disagreements in judicial downward departures for child pornography sentences. Criminology & Public Policy, 13(2), 241–​270. Kim, B., Cano, M.V., Kim, K., and Spohn, C. (2016).The impact of United States v. Booker and Gall/​Kimbrough v. United States on sentence severity: Assessing social context and judicial discretion. Crime & Delinquency, 62(8), 1072–​1094. Lehmann, P. S., Chiricos, T., and Bales, W. D. (2018). Juveniles on trial: Mode of conviction and the adult court sentencing of transferred juveniles. Crime & Delinquency, 64(5), 563–​586. Metcalfe, C., and Chiricos, T. (2018). Race, plea, and charge reduction: An assessment of racial disparities in the plea process. Justice Quarterly, 35(2), 223–​253. Nagin, D. S., and Snodgrass, G. M. (2013). The effect of incarceration on re-​offending: Evidence from a natural experiment in Pennsylvania. Journal of Quantitative Criminology, 29(4), 601–​642. Nowacki, J. S. (2015). Race, ethnicity, and judicial discretion: The influence of the United States v. Booker decision. Crime & Delinquency, 61(10), 1360–​1385. Oleson, J. C., Lowenkamp, C. T., Wooldredge, J., VanNostrand, M., and Cadigan, T. P. (2017). The sentencing consequences of federal pretrial supervision. Crime & Delinquency, 63(3), 313–​333. Rengifo, A. F., and Stemen, D. (2015). The unintended effects of penal reform: African American presence, incarceration, and the abolition of discretionary parole in the United States. Crime & Delinquency, 61(5), 719–​741. Rydberg, J., Cassidy, M., and Socia, K. M. (2018). Punishing the wicked: Examining the correlates of sentence severity for convicted sex offenders. Journal of Quantitative Criminology, 34(4),  1–​28. Sevigny, E. L. (2009). Excessive uniformity in federal drug sentencing. Journal of Quantitative Criminology, 25(2), 155–​180. Shermer, L. O. N., and Johnson, B. D. (2010). Criminal prosecutions: Examining prosecutorial discretion and charge reductions in US federal district courts. Justice Quarterly, 27(3), 394–​430. Spohn, C., and Belenko, S. (2013). Do the drugs, do the time? The effect of drug abuse on sentences imposed on drug offenders in three US district courts. Criminal Justice and Behavior, 40(6), 646–​670. Spohn, C., and Fornango, R. (2009). US attorneys and substantial assistance departures:Testing for interprosecutor disparity. Criminology, 47(3), 813–​846. Spohn, C. C., Kim, B., Belenko, S., and Brennan, P. K. (2014). The direct and indirect effects of offender drug use on federal sentencing outcomes. Journal of Quantitative Criminology, 30(3), 549–​576. Spohn, C., and Sample, L. L. (2013).The dangerous drug offender in federal court: Intersections of race, ethnicity, and culpability. Crime & Delinquency, 59(1),  3–​31. Tillyer, R., and Hartley, R. (2016).The use and impact of fast-​track departures: Exploring prosecutorial and judicial discretion in federal immigration cases. Crime & Delinquency, 62(12), 1624–​1647. Tillyer, R., Hartley, R. D., and Ward, J. T. (2015). Differential treatment of female defendants: Does criminal history moderate the effect of gender on sentence length in federal narcotics cases?. Criminal Justice and Behavior, 42(7), 703–​721. Ulmer, J. T., Eisenstein, J., and Johnson, B. D. (2010). Trial penalties in federal sentencing: extra-​guidelines factors and district variation. Justice Quarterly, 27(4), 560–​592. Ulmer, J., Light, M. T., and Kramer, J. (2011). The “liberation” of federal judges’ discretion in the wake of the Booker/​Fanfan decision: Is there increased disparity and divergence between courts?. Justice Quarterly, 28(6), 799–​837. Ulmer, J. T., Light, M. T., and Kramer, J. H. (2011). Racial disparity in the wake of the Booker/​Fanfan decision: An alternative analysis to the USSC’s 2010 report. Criminology & Public Policy, 10(4), 1077–​1118.

66

Sentencing Guidelines in the US Ulmer, J., Painter-​Davis, N., and Tinik, L. (2016). Disproportional imprisonment of Black and Hispanic males: Sentencing discretion, processing outcomes, and policy structures. Justice Quarterly, 33(4), 642–​681. Wang, X., Mears, D. P., Spohn, C., and Dario, L. (2013). Assessing the differential effects of race and ethnicity on sentence outcomes under different sentencing systems. Crime & Delinquency, 59(1), 87–​114. Warner, T. D., and Kramer, J. H. (2009). Closing the revolving door? Substance abuse treatment as an alternative to traditional sentencing for drug-​dependent offenders. Criminal justice and behavior, 36(1), 89–​109. Wooldredge, J. (2012). Distinguishing race effects on pre-​trial release and sentencing decisions. Justice Quarterly, 29(1),  41–​75. Wooldredge, J. (2009). Short-​versus long-​term effects of Ohio’s switch to more structured sentencing on extralegal disparities in prison sentences in an urban court. Criminology & Public Policy, 8(2), 285–​312. Wu, J., and Spohn, C. (2010). Interdistrict disparity in sentencing in three US district courts. Crime & Delinquency, 56(2), 290–​322. Zhang,Y., Zhang, L., and Vaughn, M. S. (2014). Indeterminate and determinate sentencing models: a state-​specific analysis of their effects on recidivism. Crime & Delinquency, 60(5), 693–​715.

67

3 SENTENCING GUIDELINES OUTSIDE THE UNITED STATES Julian V. Roberts and Lyndon Harris

Most readers of this handbook will be familiar with the sentencing guidelines currently functioning across the United States. These have been operating for 40 years now, following their introduction in Minnesota in 1979 (Frase, 2005a; Kauder & Ostrom, 2008). Far fewer readers, however, will be aware of the guidelines found in other jurisdictions. The purpose of this chapter is to introduce these alternative schemes and to make some limited comparisons across legal boundaries. The US guidelines were created as a response to concerns about disparity at sentencing (Frankel, 1973). Prior to the creation of guidelines courts had exercised wide discretion at sentencing, with very few restrictions on that discretion and little guidance as to how to exercise discretion at sentencing. The inevitable consequence of this state of affairs was some degree of unwarranted sentencing disparity (see discussion in Spohn, 2002, c­ hapter 4). Yet disparity and discrimination at sentencing are not exclusively American problems, and guidelines are not a uniquely American solution. Other countries facing this common challenge have responded with different kinds of guidelines. Most US states have a formal sentencing guidelines scheme to assist judges at sentencing. The best-​known guidelines model involves a two-​dimensional sentencing grid—​much like a mileage chart which shows the distance between two cities. Under a sentencing grid, the two dimensions are crime seriousness and criminal history. In order to determine the sentence that should be imposed, a court selects the appropriate level of seriousness, and the appropriate criminal history category. Where the crime seriousness row and the criminal history column intersect, there is a grid cell containing a relatively narrow range of sentence length. Sentencing grids of this kind are found in a number of states including Minnesota (e.g., Minnesota Sentencing Guidelines Commission, 2018; Frase, 2005b; 2009) and are used in the federal district courts in the US. Structuring judicial discretion at sentencing is one of the most significant challenges for a legislature. If they prescribe specific sentences—​such as mandatory terms of custody—​courts may be prevented from doing justice by reflecting the individual circumstances of specific offenders. For example, legislating a mandatory sentence of imprisonment of a specified length for all convictions of robbery means that offenders of different levels of culpability who have committed offenses involving differing levels of harm will receive the same sentence—​a clear injustice, as it violates the key principle of proportionality at sentencing. This principle requires that the severity of punishments be proportionate to the seriousness of the crime and the culpability of the offender. In many countries the principle has been codified (see Roberts & Baker, 2008). For example, in 1996 as part of a sentencing reform law, the Canadian Parliament legislated the following provision in 68

Sentencing Guidelines Outside the US

that country’s Criminal Code: “A sentence must be proportionate to the seriousness of the crime and the offender’s degree of responsibility” (s. 718.1; see Roberts & Cole, 1999). If legislatures leave the courts to regulate themselves at sentencing, outcomes may be too variable, leading to sentencing disparity (Palys & Divorksi, 1986). In short, there is a fine balance between offering too much and too little structure.

Chapter Overview All common law jurisdictions confront the challenge of guiding courts in the critical stage of sentencing, and disparity is an inevitable consequence of wide discretion across all systems. Many countries have risen to this challenge and created sentencing guidelines. This chapter reviews these guidelines schemes in the following way. The first part provides some general commentary on common law sentencing, followed by introductory comments about sentencing guidelines and sentencing commissions. The second part discusses schemes of sentencing guidance which do not include specific sentence length recommendations or ranges of sentence length. Such schemes are found in many Scandinavian countries as well as Israel. The third part focuses on the jurisdiction outside the US with the most well-​developed guidelines: England and Wales. The English and Welsh sentencing guidelines have been evolving for the past 20 years (Ashworth & Roberts, 2013a; Roberts & Ashworth, 2016) and represent the principal alternative to the grid-​based approach found in several US jurisdictions.This section summarizes the research into the effects of the guidelines on sentencing practices. The fourth part briefly summarizes developments in other parts of the world. We exclude Australia from our survey of foreign jurisdictions because it is the subject of a separate chapter in this handbook (see Freiberg & Flynn, Chapter 12). The fifth part draws some limited conclusions about the future of sentencing guidelines around the world.

Sentencing in Common Law Countries Sentencing in common law jurisdictions has traditionally been grounded in wide judicial discretion and a resistance to structuring of the decision-​making process. A combination of significant stock placed in the experience of the sentencing judge and a lack of concern as to the existence or extent of disparity in sentencing made for a sentencing landscape which paid little regard to consistency at sentencing. Emphasis was typically placed on the need to reflect the individualities of each case and therefore recourse to general common law principles at sentencing was, for a considerable period of time, as far advanced as matters became.This inevitably produced subjective interpretation and application of the proper approach to sentencing, which would result in disparities. While there was likely to be intra-​judge consistency, with judges sentencing (within the law but) according to their own penal philosophies, inter-​judge consistency was unlikely. For instance, in the 1830s in England, the Criminal Law Commissioners sought to bring a sense of order to the discretionary sentencing practices which they considered to be leading to arbitrary sentencing (Radzinowicz & Hood, 1979). Concerns regarding disparity led to corrective measures. In England and Wales, this led to the creation of a criminal appeal court and a right to apply for leave to appeal in the early 20th century. The court had the ability to review sentencing decisions and thereby provide some “light touch” guidance and structure by way of appellate review.There was increasing interest in the wide discretion afforded to sentencing judges in common law jurisdictions and greater involvement from appellate courts in the structuring of that discretion by the issuing of guideline judgments (Pattenden, 1996). Greater political interest and involvement followed. This led to legislation providing increasing levels of structure and constraint placed on judicial discretion, alongside calls for the creation of guidelines bodies.The claims for guidelines bodies were, in some jurisdictions, resisted by reference to the “need” for wide judicial discretion to do justice to difference, yet the direction of travel was towards more constrained judicial freedom at sentencing. That is not to say judicial 69

Julian V. Roberts and Lyndon Harris

involvement must be limited, but merely that structured discretionary decisions limit the scope of the range of penalties which may be imposed in a particular case. While there are distinct differences between the approaches to sentence, the role of the prosecutor and what is required of a sentencing judge in common law jurisdictions, there are overwhelming similarities. Most common law jurisdictions operate a sentencing system based around the principle of proportionality: Canada, England, USA, Australia, and other countries. This requires the advocates for the prosecution and defense to address matters of offense seriousness at the sentencing hearing. The role of the attorneys differs from jurisdiction to jurisdiction, however. For instance, in Australia, a prosecutor may not make a submission as to sentencing ranges (Barbaro v. The Queen [2014] HCA 2). In England, Wales, and Canada, prosecutors play a more active role, making submissions on relevant factors and setting out the court’s sentencing powers (Roberts, 2012b). In England, this represents a shift from a more traditional view that the prosecution had no role at sentencing. England and Wales have traditionally been an outlier in this regard. Although prosecutors in Canada, the US, and Australia make robust submissions at sentencing, this has typically not been the case in England. English prosecutors normally identified the most important aggravating factors for the court but stopped short of making specific recommendations as to the appropriate sentence to impose. As noted, that is now changing as a result of the English sentencing guidelines (discussed below). The story of sentencing in common law jurisdictions therefore appears to be one of increasing restraint placed on the discretion afforded to sentencing judges in the pursuit of consistency. From a position of limited structure provided by a combination of sparse statutory provisions and appellate oversight, towards an interest in the practice of sentencing and the desirability of greater consistency, there has been a seismic shift in the sentencing landscape in common law jurisdictions over the past 150 years.

Sentencing Structures: Commissions and Councils Assuming a legislature has been convinced of the need for greater structure at sentencing, how do countries go about creating a sentencing guidelines scheme? The first step in any move towards structuring judicial discretion involves the creation of an independent authority to develop and issue sentencing guidelines.The most common approach to this is the creation of such an authority as a statutory body. It is this step which, as is discussed below, can stall the move towards a guidelines system, in circumstances where the legislature does not enact the required legislation. All US guidelines schemes emerge from a sentencing commission, such as the Minnesota Sentencing Guidelines Commission or the US Sentencing Commission at the federal level. In other countries these bodies are usually called “Sentencing Councils,” and there is significant variation in their structures and functions. The Sentencing Council of England and Wales is headed by the Lord Chief Justice and is tasked with devising and disseminating guidelines as well as a range of other functions (see Roberts, 2012a). By contrast, Sentencing Councils in Australia such as the Sentencing Advisory Council in New South Wales are, as the name implies, purely advisory in nature. These councils do not issue sentencing guidelines per se, but rather provide advice and conduct research upon a wide range of sentencing matters. All Sentencing Councils and panels are involved in public legal education of one kind or another. This may mean publishing reports to help the public understand the sentencing process better, or it may mean releasing comprehensive sentencing statistics. For example, some guidelines authorities publish periodic Sentencing Bulletins which summarize sentencing trends for specific offenses (see http://​sentencingcouncil.vic.gov.au/​page/​about-​us/​council). The public typically rely on news media accounts of sentencing decisions, and these generally focus on unusual or exceptionally lenient sentences—​those which are newsworthy in some respect (Roberts & Hough, 2005). Often media reporting of sentencing decisions is inaccurate or misleading; non-​immediate custodial sentences are described as the offender having “walked free from court” or having been “let off ” and errors are made in relation to the period of time to be spent in custody, and the period of time spent 70

Sentencing Guidelines Outside the US

on license subject to recall to custody. Overwhelmingly in England and Wales, for example, the media reporting suggests that sentencing judges are too lenient. It is important therefore for a guidelines authority to dispel public misperceptions of sentencing.

Guideline Structures If appropriately constructed, and not subject to political interference, sentencing guidelines represent the best way of constraining prison populations and achieving principled sentencing (Stemen & Rengifo, 2011; von Hirsch, Ashworth, & Roberts, 2009), but the question remains: What form of guidelines are appropriate for any given jurisdiction? The guidelines movement remains strong across the US, but despite its high profile, the model employed in states such as Minnesota has not proven a popular penal export. Canada was the first country to reject this approach to structured sentencing. In 1984, Canada created a term-​limited sentencing commission which visited several American states (including Minnesota and Pennsylvania) and concluded that two-​dimensional grids held no promise for sentencing in Canada (see Canadian Sentencing Commission, 1987). A generation later the Sentencing Commission Working Group in England and Wales visited the home of numerical guidelines and drew the same conclusion. Western Australia considered adopting a two-​dimensional sentencing grid in 1999, but also ultimately abandoned the idea. The grid-​based approach has therefore found no support in foreign jurisdictions. The proliferation of two-​dimensional sentencing grids across the US since the 1970s may paradoxically have undermined the appeal of all presumptively binding guidelines. Sentencing guidelines of any kind are often regarded by judges as harbingers of grids and as being antithetical to sentencing as a “human process” (Hogarth, 1971). Calls for the introduction of any kind of sentencing guidelines system are perceived as an attempt to move towards the ultimate goal of a grid and a reduction in judicial discretion. Indeed, opposition in Canada (and England and Wales) to sentencing guideline schemes of all stripes was fueled by predictions that any move towards structuring judicial discretion would culminate in the imposition of a rigid two-​dimensional grid. In England and Wales, despite considerable judicial and professional resistance to the concept of guidance derived from a source other than the Court of Appeal, guidelines have slowly emerged over the past 15 years. Definitive guidelines now exist for most high frequency offenses and enjoy widespread support from the judiciary, practitioners, politicians, and the public.

Sentencing in Other Jurisdictions: Guidance by “Words Alone” When most people think about sentencing guidelines, formal structures usually come to mind, involving guideline sheets or grids, sentencing tables and manuals and so forth. However, guidance does not have to be numerical in nature, providing a specific range of sentence for each crime. A number of Scandinavian countries have developed what may be termed “guidance by words” (see Ashworth, 2009). This approach to structured sentencing involves the legislature placing relatively detailed guidance in a sentencing law. For example, the Swedish Penal Code identifies proportionality as the primary rationale for sentencing and requires courts to assess the seriousness of the crime in order to determine sentence. A number of mitigating and aggravating factors are also specified in the Swedish sentencing law, in order to guide judges in the determination of sentence. Finally, the law also contains guidance for courts with respect to the choices they should make between different sentencing options (for further information, see von Hirsch & Jareborg, 2009). In theory, an advantage of the “guidance by words” approach is that it leaves courts with considerable flexibility to determine an appropriate and proportionate sentence, thereby doing justice to difference. On the other hand, this may result in much greater disparity than would be the case in 71

Julian V. Roberts and Lyndon Harris

a jurisdiction such as Minnesota where judges have to follow detailed and prescriptive sentencing guidelines. As is discussed in more detail below, the resolution of that issue rather depends on one’s conception of “consistency.” The best example of sentencing guidance by words alone can be found in the state of Israel. In 2012 the Israeli Parliament (Knesset) approved a sentencing law. This law adopted parts of a Bill which provide for “guidance by words” but without establishing the guidelines authority which would have been empowered to develop and issue guidelines scheme involving “starting point sentences” (Gazal-​Ayal & Kannai, 2010; Roberts & Gazal-​Ayal, 2013). Under the legislation, courts are required to devise their own proportionate sentence range for the case being sentenced, and to provide reasons if they impose a sentence outside this range. This novel approach is very different from the US grids. For example, in Minnesota, the grid will determine the sentence length range on the basis of the offense level and the offender’s criminal history score. This range has therefore been decided, a priori, by the Minnesota Sentencing Guidelines Commission. Courts in Israel determine their own proportionate range. The consequence of this approach is likely that consistency across sentences by the same judge is likely to be high but sentencing between judges or courts will be less consistent, as each judge will presumably devise his or her own proportionate range. This said, the Sentencing Law in Israel provides a great deal more guidance than any other sentencing statute, including direction about mitigating and aggravating factors; sentencing procedure; reasons for a court to impose a sentence outside the proportionate sentence range and much else besides. Unfortunately, to date there has been no published evaluation of the new law, so it is unclear whether sentencing in Israel has become more consistent or principled since the reform was introduced.

Sentencing Guidelines in England and Wales The English guidelines have been operating for almost 20 years now. Although most publications trace the origins of sentencing guidelines to the proposals made by Judge Frankel in 1972, in fact a number of writers in Victorian England first proposed creation of sentencing commission and guidelines (see Roberts and Ashworth, 2016). These proposals were never actually adopted by the United Kingdom Parliament, and it was not until years after the US guidelines had been operating that a sentencing commission with authority to issue guidelines was created. The evolution of the English sentencing guidelines has been documented in earlier publications (e.g., Ashworth, 2015; Roberts & Ashworth 2016). However, a brief summary may help to contextualize the discussion. Until 1998, English courts enjoyed widespread discretion at sentencing, guided only by limited appellate review. Guideline judgments from the Court of Appeal were rare and only expressly acknowledged in the 1970s (Pattenden, 1996, p. 271); this state of affairs changed with the creation of the Sentencing Advisory Panel in 1998. Created to provide advice to the Court of Appeal, this statutory body was subsequently joined by a second statutory body, the Sentencing Guidelines Council (SGC). The SGC issued the first formal guidelines in 2004. Both bodies remained in existence until 2010 when they were replaced by the Sentencing Council of England and Wales (SCEW).

Structure and Functions of the Sentencing Council of England and Wales The English judiciary has historically opposed the imposition of more structured sentencing.Why then, did the English judiciary ultimately come to accept the creation of definitive sentencing guidelines? The explanation lies in the origins of the Council and the nature of its guidelines. Members of the judiciary constitute a majority of the Council’s 14 members yet representatives of key stakeholders are also included. There are no members of the general public on the SCEW, and the perspective of victims is represented not by an individual crime victim (as is the case in several Australian Sentencing Councils) but rather a professional working in a victim-​related organization. Finally, the SCEW is 72

Sentencing Guidelines Outside the US

apolitical in the sense that no members are appointed to represent political parties or to provide “political experience and connections” (Frase, 1993a, p. 369).The US Commissions are undoubtedly more political in nature than their European counterparts. The SCEW was created in close conjunction with the senior judiciary. The Lord Chief Justice serves as President and a senior Court of Appeal (Criminal Division) judge sits as the Chair. The influence of the judiciary can also be found in the nature of the enabling statute which created the Council and which specifies the kinds of guidelines the Council should be developing. This significant degree of judicial engagement was critical to ensuring acceptance by sentencers. Judicial dominance of the Council has been questioned by some scholars, but the English guidelines would not have emerged if Parliament had created a Council modeled on the Minnesota Commission. An additional benefit of the judicial majority is that in practice it confers more independence upon the Council and provides confidence to sentencers in the courts that the guidelines developed are developed by those who understand the task of a sentencing judge and are not designed to undermine or restrict judicial discretion. Judicial membership additionally adds to the development of the methodology employed by the guidelines, ensuring that the theoretically rigorous process is also practically workable. The Minnesota Governor makes appointments to the MSGC and while the UK Lord Chancellor and Secretary of State for Justice, also an elected politician, appoints members to the English Council, decisions are taken in conjunction with the Lord Chief Justice. To date, neither the government nor Parliament has intruded into the Council’s activities save for requiring the Council to produce guidelines on sentencing when the offender has multiple convictions and on the reduction in sentence for a guilty plea. The judicial majority may explain this uncharacteristic reticence on the part of legislators. By contrast, the Minnesota Commission appears to have been under almost constant political pressure since its creation. This pressure has resulted in an escalation in sentence severity over time. Frase noted in 1991: “the pressure for increased sentence severity and legislative control” (p. 732). The SCEW has been spared this pressure, and the guidelines have not been amended in response to any external political influences. The judicial dominance comes with a cost, however. One consequence of this judicial presence will be seen in the more discretionary nature of the English guidelines (discussed later in the chapter).

Descriptive or Prescriptive Guidelines? An important policy decision for any legislature contemplating creating a Sentencing Council to develop guidelines concerns its role in the sentencing environment. Guidelines can be descriptive in nature, simply reproducing current judicial practice, or prescriptive, with a mandate to change current practice. Von Hirsch (1987) argued that “The enabling statute [of any commission] should make clear that the commission’s role is a policy-​making one” (p. 62, emphasis in original). The Minnesota guidelines are much closer to the prescriptive model (Frase, 1993; Tonry, 1987). As the MSGC noted, “In developing guidelines we have been informed by, but not bound to, current practice” (Minnesota Sentencing Guidelines Commission, 1980, p. 30). A related question is whether the guidelines should be sensitive to prison capacity. Most US guidelines are sensitive to the prison population, and therefore have the potential to prevent serious overcrowding. Chapter 23 of the Minnesota Laws 1978, ch. 244 et seq. directs the Commission to “take into substantial consideration current sentencing and release practices and correctional resources, including but not limited to the capacities of local and state correctional facilities.” The Minnesota Commission noted that it interpreted its enabling legislation, “to mean that the guidelines should produce prison populations which do not exceed the current capacity of state correctional institutions” (1980, p.  2). US academics share the opinion firmly expressed by Frase that “an assumption of limited prison capacity is an essential component of guidelines development” (1990, p. 734). 73

Julian V. Roberts and Lyndon Harris

Unlike Minnesota, the English Council is not required to consider the size of the prison population. For this reason, the English guidelines are founded upon judicial practice prevailing at the time and designed simply to promote a more consistent approach to current sentencing. This approach has been criticized by a number of UK scholars, who have argued that the Council should address the high (relative to other Western European nations) prison population in England and Wales by amending its guidelines (e.g., Allen, 2016). Even if the Council wished to change sentencing trends, in contrast to Minnesota, there are practical impediments to reducing the use of custody, or the size of the prison estate. The principled objection is that the Council has no legal authority to engineer changes in the volume or duration of custodial sentences. The practical objection is that it is hard to see how the Council could go about the task, even if it had the authority. The Minnesota Commission can lower the volume of prison admissions very expeditiously by reducing the grid sentence ranges. For example, it could reduce sentence lengths by 10% across the entire grid. This step would result in a reduction in the sentence lengths (and ultimately a smaller prison population) without disturbing ordinal proportionality—​all offenses would be affected to the same proportionate degree. The English Council has sought to make a small alteration to sentencing levels in one case—​drug mules. It considered that current levels (prior to the guideline being issued) were too high and ought to be reduced. This was expressly consulted upon and endorsed by consultees.The change was introduced to correct a tendency towards disproportionate sentencing, rather than to achieve a reduction in the volume of admissions to custody. The US Sentencing Commissions (including the federal Commission) can achieve expeditious reductions in the prison population (see Adelman, 2013). However, if the SCEW wished, for example, to promote a greater use of non-​custodial sentences for, say, drug offenders, this would require launching a professional and public consultation and then ultimately issuing an amended guideline. The typical duration for creation, consultation, and release of a guideline is approximately 12 months. Moreover, if sentences for drug offenders changed it would also be necessary to review all other offense-​specific guidelines to ensure that proportionality between offenses was not undermined.This limitation is one of the drawbacks of an offense-​specific approach to issuing guidelines and represents one of the advantages of the US approach, where all offenses are assigned to a single grid, or a small number of grids.

Consistency of Outcome versus Consistency of Approach The US and English guidelines have adopted different conceptions of consistency. Should a guidelines scheme prioritize, or exclusively pursue, consistency of approach—​a conception of consistency concerned with process over substance whereby the desire is to influence sentencers’ behavior so as to achieve greater likeness between the methodological approach to the discretionary sentencing decision—​or consistency of outcome—​a conception of consistency more concerned with substance so as to achieve a greater likeness between the sentences imposed in similar cases. An increase in the consistency of sentencing outcomes as an aim of a guideline scheme places more emphasis upon the sentence imposed for the offense; it appeals more obviously to our innate sense of justice and may tend towards a more prescriptive guideline scheme as outcomes are more measurable than adherence to process. By contrast, a guideline scheme which aims to achieve more consistency in the approach to the determination of sentence favors a more nuanced sense of justice grounded in procedural fairness. It would however, be wrong to view the two conceptions as antithetical; on the contrary, they will frequently go hand-​in-​hand: a more consistent process is, a priori, more likely to lead to a more consistent outcome, than an unregulated process, and a restriction to bring about more consistent outcomes will likely be enforced by the same methodology across different cases.

74

Sentencing Guidelines Outside the US

Whether a preference for consistency of outcome ought to result in more prescriptive guidelines with narrower ranges and greater limitation on departures, or looser guidelines requiring more subjective assessment on the part of the sentencer is up for debate. It is clear, however, that whichever way one construes the concept of consistency, a desired result—​direct or latent—​will be more consistent outcomes. The Minnesota guidelines (and indeed all US presumptive schemes) emerged in response to recognition that disparity was the inevitable consequence of highly discretionary sentencing. Judge Frankel’s seminal volume inspired the creation of the Minnesota Sentencing Guidelines Commission (see Frase, 2005). Disparity was the problem, guidelines the remedy. In light of this, it is unsurprising that the Minnesota guidelines assumed the form of a relatively restrictive grid which would ensure predictable and consistent outcomes. More recently, a number of authors have questioned this emphasis on reducing disparities, linking such a goal to the high use of imprisonment. For example, Berkow (2012) wrote that “the [guidelines] movement’s reaction against the prior regime often placed too much emphasis on uniformity and not enough on individualization” (p.1620) while Stith (2000) and others have drawn a similar conclusion about the federal guidelines.The grid-based approach has also been criticized by European scholars such as Wandall (2006), who concluded that “Working from the ideal that sentencing frameworks should accommodate a demand for individualistic equality in sentencing, the Danish [more discretionary] model is to be preferred to the Minnesota model” (p.40). There was no such recent trigger for the English guidelines, with predictable consequences for their structure. Although research in this country also documented a lack of consistency, there was far less pressure to achieve greater uniformity. To date, no study has directly compared levels of disparity in the two jurisdictions. Empirical research on sentencing variation in England and Wales suggests that disparity may be less striking in this jurisdiction (e.g., Pina-​Sanchez & Grech, 2018). This more relaxed approach to the problem of disparity is reflected in the more flexible format of the English guidelines.

Development of Guidelines: Concurrent or Consecutive? Unlike the US Commissions, the English Council issues its guidelines incrementally, offense by offense. The incremental approach confers benefits but also creates challenges. The more protracted approach is a consequence of the decision to issue offense-​specific guidelines. Guidelines are more easily and expeditiously developed if all offenses are assigned to a single grid. Constructing a separate guideline with different sentence recommendations, starting point sentences, and mitigating and aggravating factors takes much longer. The Council could well have taken several years to develop all of its offense-​specific guidelines in preparation for a mass release. Instead, it chose to identify key offenses and to issue guidelines one by one. By 2021, a full decade after its creation, the Council will have issued guidelines for all principal offense categories. In short, the English guidelines have taken much longer than those in Minnesota to cover all offenses, though this is primarily due to the consultation and drafting process. Other Councils have taken even longer. The Scottish Sentencing Council began in 2016 and is also issuing its guidelines sequentially. In 2019, the Scottish Council had issued just one guideline.

Structure of English Guidelines: Offense Specific vs a Single Grid The principal structural difference between the regimes in Minnesota and England/​Wales is that the Minnesota guidelines use three grids to accommodate all offenses, whereas the English guidelines provide guidance in separate (and individualized) guidelines for different categories of offending. Despite some structural variation reflecting the nature of the offense, all the offense-​specific guidelines

75

Julian V. Roberts and Lyndon Harris

contain a number of common elements. Broadly speaking, they require courts to follow a step-​by-​ step methodology when determining sentence. The first two steps of all the English guidelines are the most important as they begin the process of providing guidance and have the greatest impact upon the eventual sentence. The Minnesota guidelines reflect a modified just deserts rationale, although the enabling statute articulates no single rationale (see Minnesota Sentencing Guidelines Commission, 2018). The English guidelines incorporate two primary dimensions, harm and culpability, the principal components of a retributively proportionate sentence. It is noteworthy that the guidelines developed in other jurisdictions have adopted the harm-​culpability combination rather than the US crime seriousness-​criminal history alternative. These include South Korea (Park, 2010), New Zealand (Young & Browning, 2013) and China (Chen, 2010; Roberts & Wei, 2016).

Guideline Example: Street Robbery The street robbery sentencing guideline contains nine separate steps for courts to follow (Appendix A contains an extract from the guideline). At Step 1, in assessing the seriousness of the offense, the court must assign the offender to one of three levels of culpability (high, medium, and lesser) and the case to one of three levels of harm (1, 2, and 3). Consistency at this crucial first step of the guidelines methodology is encouraged by requiring all courts to apply a common set of factors to determine which category of harm and culpability is appropriate. The list of factors at Step 1 which determine category sentence range is exclusive; courts may consider other factors only later, at Step 2. The exclusive nature of this list is one of the most restrictive elements of the English guidelines and ensures that the primary determinant of the sentence—​the initial assessment of culpability and harm—​is approached in as consistent a way as possible. Step 2 uses these two dimensions to create a matrix which contains starting point sentences and sentence ranges. For example, if the court assigns the case to lesser culpability and intermediate harm, the guideline provides a starting point sentence of two years and range of 1–​4  years. Step 2 also provides a non-​exhaustive list of mitigating and aggravating factors to be considered by the court (along with any other factors proposed by advocates) in order to determine a provisional sentence within the range. Thereafter, the court works through a series of additional steps, including awarding credit for any assistance to the police or prosecution or for entering a guilty plea. These two considerations are external to considerations of harm or culpability and for this reason they are considered at a separate step. The guideline thus creates categories of factors: those of primary relevance (located at Step 1); those of more limited relevance (Step 2); and those relevant to the sentencing decision but unrelated to harm or culpability (state assistance; plea, found at Steps 3 and 4).The guidelines note that the Step 1 factors “comprise the principal factual elements of the offence” (see Sentencing Council, 2011, p. 4). In all, this guideline contains nine steps, with Steps 5 to 9 concerning other relevant considerations, such as totality, the assessment of dangerousness, and whether an offender should be subject to a sentence for public protection and a reduction for time spent on remand. Box 3.1 summarizes the operation of the robbery sentencing guideline, using a specific case. The advent of the English guideline has significantly affected sentencing practices at the trial court level. As a direct result of the increased certainty introduced by the guidelines, advocates are able to provide more detailed and concrete advice to defendants prior to entering a plea. At sentencing hearings, submissions now revolve around the interpretation of the guideline rather than discussion of appellate decisions which are said to inform the approach to sentence.There are probably no fewer appeals against sentence as the guidelines have merely changed the nature of the appeal (House of Commons Justice Select Committee, 2018).

76

Sentencing Guidelines Outside the US

BOX 3.1  Example of Applying the Street Robbery Guideline An offender aged 30, wearing a hooded sweater and a scarf covering half of his face, approaches a lone male walking in a poorly lit alley way at night in darkness. He stops the man and reveals a knife stored in the waist band of his trousers. He instructs the man to hand over his mobile telephone and wallet. He then leaves the scene.The victim was scared but otherwise unscathed.The offender has no previous convictions and has shown some remorse and insight into his offending. In applying the guideline, the court must first assess culpability. The offense appears to fit into Category B, Medium Culpability, by reference to “Production of a weapon other than a bladed article or firearm or imitation firearm to threaten violence” or “Threat of violence by any weapon (but which is not produced)”.The court must then assess harm. It appears to fit into Category 3, “No/​minimal physical or psychological harm caused to the victim”. That produces a category range of 1–​4 years’ custody and a starting point of 2 years’ custody. The court must then assess the aggravating and mitigating factors. There appear to be three aggravating factors present: the timing of the offense (at night after dark), the location of the offense (a poorly lit alley way), and the offender has attempted to conceal their identity (by the use of a hooded sweater and face-​covering). This warrants an increase from the starting point, perhaps in the region of 6–​9 months. There appears to be one mitigating factor present: good character. That warrants a reduction in sentence, perhaps in the region of 6 months. That produces a sentence in the region of 27–​ 30 months. The court must then consider Steps 3–​9, dealing with, inter alia, any reduction for a guilty plea and assistance to the prosecution.

Greater Guidance, More Onerous Requirements Compared to the Minnesota grid, the English guidelines provide guidance on a wider range of issues. This can be demonstrated by considering a typical case of robbery in which the offender has no prior convictions and has pleaded guilty to the offense. Under the Minnesota guidelines a court must impose a sentence of imprisonment between 41 and 57 months or find “substantial and compelling circumstances” to justify a downward or upward departure. Judicial decision-​making therefore focuses on whether such circumstances exist, and most of the time the sentence falls within the recommended range. By contrast, an English court would first apply the robbery guideline, proceeding through the nine steps enumerated above. At Step 1 the court considers an exhaustive list of factors needed to assign the case to high, medium, or lesser culpability, and one of three levels of harm. This determination produces a starting point sentence as a point of departure and a sentence range (at Step 2). The judge then takes all other relevant mitigating and aggravating factors into account to arrive at a provisional sentence.This sentence would then be modified to reflect other considerations such as plea (found in Steps 3 through 9 of the guideline). The offense-​specific robbery guideline is not the only guidance that the court will have to follow. Since the guideline sentence ranges encompass both custodial and non-​custodial dispositions, the court will need to consult the separate guideline on the use of the principal sanctions. If the defendant has pleaded guilty, the court will also have to apply the guideline regulating plea-​based sentence reductions. Unlike other common law jurisdictions, including the US, courts in England and Wales follow a clear guideline when determining the level of reduction that a defendant should receive if he or she enters a guilty plea. The level of reduction in all other countries is left to the individual court to determine. However, in England and Wales there is a guideline which contains clear recommendations. If the defendant enters his or her plea at the first opportunity to do so, the

77

Julian V. Roberts and Lyndon Harris

savings in terms of case preparation and court time are greatest.Victims and witnesses are best off if they know early in the proceedings that their testimony will not be needed (because the defendant has decided to plead guilty). For this reason, the guideline recommends the greatest reduction—​one-​ third off the sentence—​to defendants who enter their plea early. If they choose to wait, and perhaps plead guilty only on the day of trial, the savings are minimal and accordingly the recommended reduction is modest—​only 10% off the sentence (see Appendix B for an extract from the guideline). Having a guideline of this kind makes the consequences of pleading guilty far more predictable, and attorneys are better able to advise their clients about the anticipated benefits of entering a plea. Research has shown that the courts in England generally follow the recommendations of the guideline (Roberts & Bradford, 2015). Finally, there is an additional relevant guideline addressing factors affecting seriousness.1 As with other Commissions and Councils,2 the SCEW also publishes a range of statistical and analytic information relevant to sentencing in general and with respect to specific offenses. It is unclear how often this material is accessed by sentencers or whether it affects judicial decision-​making. The custody threshold is another example of the deeper judicial processing required by the English guidelines. Under the Minnesota guidelines the offender can know in advance of sentencing whether he is in the custody zone of the grid, based upon his offense of conviction and criminal history score.The offender can arrive at the sentencing hearing knowing their fate is largely determined by the decisions of the Minnesota Commission. The dispositional departure statistics suggest that in most cases, this a priori classification will accurately predict whether he is incarcerated. In 2016, mitigated dispositional departures occurred in approximately one-​third of cases (MSGC, 2017, p. 27) and aggravated dispositional departures were extremely rare. As the Commission notes in a recent data release, the aggravated dispositional departure rate is very low (MSGC, 2017, p.  25). Judicial reflection is required only to determine whether there are substantial and compelling circumstances to justify overturning the presumptive disposition. In applying the English guideline, courts must consider whether an offense ought to be mitigated or aggravated to the extent that the sentence ought to cross the custody threshold in one direction or another. It is noteworthy that the statutory duty to “follow” the English guidelines requires only that the sentencer imposes a sentence between the bottom of the lowest category range and the top of the highest category range. Using the street robbery example relied upon above, this would require a sentence between a high-​level community penalty (a sentence lasting up to three years served in the community consisting of requirements which must be complied with) to an immediate custodial sentence of 12 years. The English guidelines are therefore more demanding of judges, less prescriptive, and less restrictive than their Minnesota equivalent. Does the former lead to a more reasoned sentencing decision? It is unclear what kind of critical test could be devised to establish the superiority of one model. Perhaps the best we can say is that sentencing involves a more in-​depth judicial processing of all relevant variables, albeit at the cost of greater variability in outcomes. The sentence ranges for many offenses encompass custodial and non-​custodial options and many offense ranges span the custody threshold. For example, the category 3 sentence range for unlawful wounding (maximum penalty of five years) runs from a low-​level community order to 51 weeks’ custody.3 A court must therefore grapple with the question of whether the custodial threshold has been passed in a high volume of cases. In resolving this issue, the court will be assisted by submissions from the advocates; compared to offenders in Minnesota, advocates representing offenders sentenced under the English guidelines have more to play for in terms of mitigating the effects of prior offending.

78

Sentencing Guidelines Outside the US

Compliance Requirement A key question about any guidelines scheme is the degree of constraint that is imposed upon courts. In Minnesota, and several other US states, the guidelines are presumptively binding. This means that the defendant is presumed to receive a sentence within the sentence length range recommended by the grid (if the offense falls with the custodial zone). A court may impose a sentence above or below the range, but in order to do so it must find “substantial and compelling reasons” to justify this “departure” sentence (see MSGC, 2018). This requirement ensures that courts apply the guideline recommendations in most cases and allows attorneys and defendants to have a clear idea of the sentence that will likely be imposed. Some states operate guidelines that are purely advisory in nature; the court is not compelled to follow any specific recommendation. In England, courts are also required by law to “follow any relevant guideline” at sentencing. But comparisons of the two sets of guidelines (Minnesota; England and Wales) makes it clear that the English guidelines allow more discretion within the guideline. In addition, as with Minnesota, a court in England can impose a different sentence if it finds that it would be “contrary to the interests of justice” to follow the guideline.

Research on the Impact of the English Guidelines Since the English guidelines have been existence for a much shorter period than those found in Minnesota, Pennsylvania or other States, there is far less research into their effectiveness. Nevertheless, a number of studies have demonstrated positive impacts in terms of consistency. Academic analyses published to date suggest a positive impact on consistency across courts and the application of the offense-​specific guidelines. Pina-​Sanchez and Linacre (2013) demonstrated that for a number of high-​ volume offenses, the sentencing factors provided in the guideline were being applied in a consistent way across courts. Pina-​Sanchez (2015) evaluated the impact of the new assault guideline on variability in sentencing, conducting a pre-​post analysis using the Council’s dataset and concluded that “consistency improved in all the offenses studied after the new guideline came into force” (p. 87). Irwin Rogers and Perry (2015) focused on sentencing for domestic burglary and found that “the courts were sentencing in a manner that was consistent with the domestic burglary guideline” (p. 210).These studies (see also essays in Roberts, 2015), while limited in scope, suggest the English guidelines have had a positive effect on promoting consistency. More research is clearly needed, however. The introduction of guidelines may have contributed to improving public confidence in sentencing. One empirical exploration of public opinion suggested that greater public awareness of the guidelines may promote public confidence in sentencing and possibly mitigate criticism of sentencers. Members of the public in the UK were strongly supportive of the concept of guidelines. Over 90% of a representative sample of the public endorsed the view that guidelines were definitely or probably a good idea (Roberts et al. 2012). In addition, respondents in the same survey who had been informed of the guidelines were less likely to rate specific sentences as being too lenient than were people who reacted to the same cases without having been made aware of the guidelines (Roberts et al. 2012).

Guidelines in Other Jurisdictions Progress towards developing sentencing guidelines in other counties has been slower. However, as of 2019, a number of other countries have created guidelines for courts at sentencing. The Sentencing Commission for Scotland recommended creation of an Advisory Panel on Sentencing to assist with the introduction of sentencing guidelines, and a Scottish Sentencing Council now exists (see

79

Julian V. Roberts and Lyndon Harris

Hutton & Tata, 2010). The Scottish Council has issued one overarching guideline on the principles of sentencing, and is likely to follow the offense-​specific rather than the grid-​based approach to guidance. South Korea launched a guidelines scheme over a decade ago (see Park, 2010). The Korean guidelines are issued by the Supreme Court. Although they are modeled loosely on the English format and are offense-​specific in nature (Sentencing Commission of the Supreme Court of Korea, 2014), the Korean guidelines are less detailed and allow greater discretion for courts. For most offenses they prescribe a standard sentencing range, applicable in most cases, as well as a mitigated and an aggravated sentence range. An example illustrates. The standard sentencing range for a robbery conviction is from two to four years, while the mitigated range is 18 months to three years and the aggravated range is three years to six years (Sentencing Commission of the Supreme Court of Korea, 2014, p. 76). Unlike the English guidelines, the Korean guidelines do not contain a series of steps to follow. None of the existing foreign guidelines have been subject to systematic research, so no conclusions may be drawn about their relative effectiveness in reducing disparity of outcome or achieving other goals usually set for guideline structures. Aside from the countries that have implemented formal sentencing guidelines, a number of others have expressed an intention to do so or have created prototypes. The Law Commission of New Zealand developed a comprehensive and principled set of guidelines but the legislature in that jurisdiction has yet to enact the necessary legislation to permit implementation (see Young & Browning, 2008). The New Zealand scheme involved a comprehensive guideline for each offense; the guideline contained categories of crime seriousness, each with an associated range of sentence. A sentencing court would match the case appearing for sentencing to the guideline category using information in the guideline.The system was more flexible than the US-​based schemes.The New Zealand guidelines also included “generic” advice—​guidelines which apply to more than a single offense. For example, the guidelines provide guidance on considering the impact of the crime upon the victim and also the way in which courts should approach the sentencing of multiple crimes on the same occasion. This adopts a similar approach to that taken by the SCEW, which has issued a number of overarching guidelines on topics such as totality of sentence, assessing seriousness, domestic abuse as a factor in other offenses, and the imposition of custodial and community sentences. In 2017, the government of the Australian state of Victoria announced it would introduce legislation to establish a Sentencing Guidelines Council.This body would be granted the powers to issue sentencing guidelines for courts in that state (Sentencing Advisory Council, 2018). Other jurisdictions—​ including Western Australia and Northern Ireland—​ have explored the use of guidelines for sentencers, but so far have not actually adopted a formal scheme. Following recommendations from a Sentencing Working Group in 2010, Northern Ireland held a consultation on the possible options for a form of sentencing guidelines (Criminal Policy Unit, 2010), but unlike Scotland, no Council has been created. Several jurisdictions (including New South Wales and the state of Victoria in Australia) have created advisory bodies which disseminate information about sentencing but which do not issue guidelines (see Freiberg, this volume; and more generally, Freiberg & Gelb, 2008). Finally, other countries—​Canada, South Africa, Ireland, and India for example—​have resisted all appeals for greater structure at sentencing (e.g., O’Malley, 2003; Roberts, Azmeh, &Tripathi, 2011). Although scholars and practitioners in those countries have long advocated creation of some kind of guidelines scheme, legislatures in these countries have so far rejected calls to introduce sentencing guidelines. The consequence is that judges in these jurisdictions continue to impose sentence much as they have for decades, with the only guidance coming from the appellate courts. This approach to sentencing may be termed “judicial self-​regulation” (see Ashworth, 2009). The limitation of this approach is that higher courts hear only a small proportion of cases on appeal, which means that the opportunities for guidance are limited. When a court of appeal does hear a sentence appeal it does not always give

80

Sentencing Guidelines Outside the US

general guidance, other than by way of general comment as to the level of sentence imposed when dismissing or allowing the appeal.

Conclusions What have we learned about the experience with guidelines outside the United States? A number of lessons can be drawn. First, the Scandinavian model suggests that numerical guidelines are not necessarily the only model to follow. It is possible to offer guidance to courts without prescribing specific sentencing ranges in terms of numbers of months or years.Whether the “guidance by words” approach is sufficient to achieve adequate levels of consistency is debatable, however. Most sentencing scholars appear to agree that some form of guidelines is necessary in order to achieve an acceptable degree of consistency in sentencing. Second, the English guidelines demonstrate that there is a middle ground lying between the relatively tight sentencing guidelines grids found across the United States and the looser systems of “guidance by words” found in countries like Sweden and Finland (see von Hirsch et al., 2009, Chapter 6). The English guidelines offer a system which is numerical (in the sense that it contains specific sentence recommendations), prescriptive, and yet quite flexible in application. Both systems (the US and the English) represent an improvement upon the highly discretionary sentencing arrangements found in countries like Canada, South Africa, and India. Third, judicial acceptance of greater structure (and reduced discretion) is more likely when the judiciary are heavily implicated in the development and evolution of the guidelines. The statutory bodies responsible for the guidelines in England and Wales have generally been dominated by the judiciary.The Canadian Sentencing Commission proposals failed, in part because judges perceived the guideline scheme to be a bureaucratic scheme created by academics. Similarly, in the Australian states, judicial resistance to the introduction of guidelines has been an obstacle to adoption of a guidelines scheme. Fourth, there may be an advantage to the gradual evolution of the guidelines. The English guidelines have been criticized for being slow to cover all offenses. The English guidelines for specific offenses have been issued serially over the years rather than in one step as was the case in the United States. In retrospect, this potential weakness of the guidelines may paradoxically have ensured their survival and development. More structured sentencing, by way of guidance, has evolved very slowly, beginning with the first guideline judgments from the Court of Appeal (Criminal Division) in the mid-​1970s (Pattenden, 1996). These increased in frequency over the following decades and eventually led to the creation of the Sentencing Advisory Panel in 1998, which provided guidance to the Court of Appeal (Criminal Division). From these modest origins in 1999 (see Ashworth & Wasik, 2010) came a true guidelines model in 2003 providing increased structure. This incremental development allowed the creation of the much more comprehensive and detailed guidelines which have been issued since 2011 (Ashworth & Roberts, 2016). Judges who are traditionally resistant to any attempts to curb their discretion may be more likely to accept guidance when it comes in this format. The ultimate question, however, is the following: Are the guidelines proposed or implemented in other countries better or worse, more or less effective than those developed across the United States? Does the experience in other countries carry any lessons for US guideline commissions? Unfortunately, the absence of truly comparative research makes it impossible to resolve the issue one way or another. In addition, the non-​US-​based guidelines including those in operating in England have yet to be comprehensively evaluated. At the very least, however, the experience in that country demonstrates that is possible to introduce detailed and prescriptive sentencing guidelines even in a common law jurisdiction which, in the 1980s and 1990s, was committed to the traditional model of privileging judicial discretion (see Ashworth, 2015).

81

Julian V. Roberts and Lyndon Harris

Appendix A: Example of Offense-​Specific Sentencing Guideline Outside the US: Sentencing Robbery in England and Wales STEP ONE Determining the offense category The court should determine the offense category with reference only to the factors listed in the tales below. In order to determine the category, the court should assess culpability and harm. The court should weigh all the factors set out below in determining the offender’s culpability. Where there are characteristics present which fall under different levels of culpability, the court should balance these characteristics to reach a fair assessment of the offender’s culpability. Culpability demonstrated by one or more of the following: A—​High culpability

• Use of a weapon to inflict violence • Production of a bladed article or firearm or imitation firearm to threaten violence • Use of very significant force in the commission of the offense • Offense motivated by, or demonstrating hostility based on any of the following characteristics or presumed characteristics of the victim: religion, race, disability, sexual orientation or transgender identity

B—​Medium culpability

• Production of a weapon other than a bladed article or firearm or imitation firearm to threaten violence • Threat of violence by any weapon (but which is not produced) • Other cases where characteristics for categories A or C are not present

C—​Lesser culpability

• Involved through coercion, intimidation or exploitation • Threat or use of minimal force • Mental disability or learning disability where linked to the commission of the offense

Harm The court should consider the factors set out below to determine the level of harm that has been caused or was intended to be caused to the victim. Category 1

• Serious physical and/​or psychological harm caused to the victim • Serious detrimental effect on the business

Category 2

• Other cases where characteristics for categories 1 or 3 are not present

Category 3

• No/​minimal physical or psychological harm caused to the victim • No/​minimal detrimental effect on the business

82

Sentencing Guidelines Outside the US

STEP TWO Starting Point and category range Having determined the category at step one, the court should use the corresponding starting point to reach a sentence within the category range below. The starting point applies to all offenders irrespective of plea or previous convictions. A case of particular gravity, reflected by multiple features of culpability or harm in step one, could merit upward adjustment from the starting point before further adjustment for aggravating or mitigating features, set out on the next page. Consecutive sentences for multiple offenses may be appropriate—​please refer to the Offences Taken into Consideration and Totality guideline. Culpability Harm

A

B

C

Category 1

Starting point 8 years’ custody

Starting point 5 years’ custody

Starting point 4 years’ custody

Category range 7–​12 years’ custody

Category range 4–​8 years’ custody

Category range 3–​6 years’ custody

Starting point 5 years’ custody

Starting point 4 years’ custody

Starting point 2 years’ custody

Category range 4–​8 years’ custody

Category range 3–​6 years’ custody

Category range 1–​4 years’ custody

Starting point 4 years’ custody

Starting point 2 years’ custody

Starting point 1 years’ custody

Category range 3–​6 years’ custody

Category range 1–​4 years’ custody

Category range High-​level community order 3 years’ custody

Category 2

Category 2

The table on the next page contains a non-​exhaustive list of additional factual elements providing the context of the offense and factors relating to the offender. Identify whether any combination of these, or other relevant factors, should result in an upward or downward adjustment from the sentence arrived at so far. In particular, relevant recent convictions are likely to result in an upward adjustment. In some cases, having considered these factors, it may be appropriate to move outside the identified category range. Source: Sentencing Council of England and Wales.

Appendix B Example of a ‘Generic’ Sentencing Guideline, Applicable to All Crimes (Extract from Plea-​based Sentence Reduction Guideline, England and Wales) In each category, there is a presumption that the recommended reduction will be given unless there are good reasons for a lower amount. After a trial date is set

First reasonable opportunity

Door of the court/ after trial has begun

======= | ==================== | =================== | recommended 1/3

recommended 1/4

Source: Sentencing Council of England and Wales.

83

recommended 1/10

Julian V. Roberts and Lyndon Harris

Notes 1 All the English Council’s guidelines are available on the Council’s website (see www.sentencingcouncil.org. uk/​). 2 Particularly the Sentencing Councils in Australia (see for example www.sentencingcouncil.vic.gov.au/​). 3 www.sentencingcouncil.org.uk/​ w p-​ c ontent/​ u ploads/​ A ssault_​ d efinitive_​ g uideline_​ - ​ _ ​ C rown_​ C ourt. pdf, p. 8.

References Adelman, L. (2013). What the sentencing commission ought to be doing reducing mass incarceration. Mich. J. Race & L., 18: 295. Allen, R. (2016). The Sentencing Council for England and Wales: Brake or Accelerator on the Use of Prison. London: Transform Justice. Ashworth, A. (2009). Techniques for reducing sentence disparity. In A. von Hirsch, A. Ashworth & J. V. Roberts (eds.), Principled Sentencing: Readings on Theory and Policy (pp. 243–​257). Oxford: Hart Publishing. —​—​ . (2015). Sentencing and Criminal Justice (6th ed.). Cambridge: Cambridge University Press. Ashworth, A., & Roberts, J.V. (2013a). The origins and nature of the sentencing guidelines in England and Wales. In A. Ashworth & J. V. Roberts (eds.) Sentencing Guidelines: Exploring the English Model (pp. 1–​12). Oxford: Oxford University Press. Ashworth, A., & Roberts, J. V. (eds.) (2013b). Sentencing Guidelines: Exploring the English Model. Oxford: Oxford University Press. Ashworth, A., & Wasik, M. (2010). Ten Years of the Sentencing Advisory Panel: The Sentencing Advisory Panel Annual Report. United Kingdom and Wales: The Sentencing Council for England and Wales. Retrieved from: www. sentencingcouncil.org.uk/​ Canadian Sentencing Commission. (1987). Sentencing Reform. A Canadian Approach. Ottawa, Canada: Supply and Services Canada. Chen, X. (2010). The Chinese sentencing guideline: A preliminary analysis. Federal Sentencing Reporter, 22(4), 213–​216. Criminal Policy Unit. (2010). Consultation on a Sentencing Guidelines Mechanism. Belfast: Northern Ireland Department of Justice. Frankel, M. (1973). Criminal Sentences: Law without Order. New York: Hill and Wang. Frase, R. (1990). Sentencing reform in Minnesota, ten years after: Reflections on Dale G. parent’s structuring criminal sentences: The evolution of Minnesota’s sentencing guidelines. Minnesota. Law Review, 75, 727–​754. —​—​. (1993). The role of the legislature, the sentencing commission, and other offcials under the Minnesota sentencing guidelines. Wake Forest L. Rev., 28: 345. —​—​. (2005a). State sentencing guidelines: Diversity, consensus, and unresolved policy issues. Columbia Law Review, 105, 1190–​1232. —​—​. (2005b). Sentencing guidelines in Minnesota, 1978–​2003. In M. Tonry (ed.), Crime and Justice. A Review of Research (pp. 131–​219). Chicago, IL: University of Chicago Press. Freiberg, A., & Gelb, K. (eds.) (2008). Penal Populism, Sentencing Councils and Sentencing Policy. Cullompton:Willan Publishing. Gazal-​Ayal, O., & Kannai, R. (2010). Determination of starting sentences in Israel—​System and application. Federal Sentencing Reporter, 22, 232–​242. Hogarth, J. (1971). Sentencing as a Human Process. Toronto: University of Toronto Press. House of Commons Justice Select Committee. (2018). Oral Evidence: Prison Population 2022: Planning for the Future. London: House of Commons. Hutton, N. and Tata, C. (2010). A sentencing exception? Changing sentencing policy in Scotland. Federal Sentencing Reporter, 22, 272–​278. Irwin Rogers, K., & Perry, T. (2015). Exploring the impact of sentencing factors on sentencing domestic burglary. In J. V. Roberts (ed.), Exploring Sentencing Practice in England and Wales (pp. 194–​220). London: Palgrave Macmillan. Kauder, N., & Ostrom, B. (2008). State Sentencing Guidelines: Profiles and Continuum. Washington, DC: National Center for State Courts. Minnesota Sentencing Guidelines Commission. (1980). Report to the Legislature. St. Paul, MN: Minnesota Sentencing Guidelines Commission. —​—​ . (2017) Report to the Legislature. St. Paul, MN: Minnesota Sentencing Guidelines Commission. —​—​. (2018). Minnesota Sentencing Guidelines and Commentary. St. Paul, MN: Minnesota Sentencing Guidelines Commission.

84

Sentencing Guidelines Outside the US Northern Ireland Assembly. (2011). Comparative Research into Sentencing Guidelines Mechanisms. Belfast: Northern Ireland Assembly. O’Malley, T. (2013). Living without Guidelines. In A. Ashworth & J. V. Roberts (eds.), Sentencing Guidelines: Exploring the English Model. Oxford: Oxford University Press. —​—​. (2016). Sentencing Law and Practice (3rd ed.). Dublin: Thomson Reuters. Palys, T. S., & Divorski, S. (1986). Explaining Sentence Disparity. Canadian Journal of Criminology, 28: 347. Park, H. (2010). The basic features of the first Korean sentencing guidelines. Federal Sentencing Reporter, 22, 262–​271. Pattenden, R. (1996). English Criminal Appeals, 1844–​1994: Appeals against Conviction and Sentence in England and Wales. Oxford: Clarendon Press. Pina-​Sanchez, J. (2015). Defining and measuring consistency in sentencing. In J. V. Roberts (ed.), Exploring Sentencing Practice in England and Wales (pp. 76–​92). London: Palgrave Macmillan. Pina-​Sanchez, J., & Grech, D. (2018). Location and sentencing: To what extent do contextual factors explain between court disparities? The British Journal of Criminology, 58(3), 529–​549. Pina-​Sanchez, J., & Linacre, R. (2013). Sentence consistency in England and Wales. The British Journal of Criminology, 53(6), 1118–​1135. —​—​. (2014). Enhancing consistency in sentencing: Exploring the effects of guidelines in England and Wales. Journal of Quantitative Criminology, 30(4), 731–​748. Radzinowicz, L., & Hood, R. (1979). Judicial discretion and sentencing standards: Victorian attempts to solve a perennial problem. University of Pennsylvania Law Review, 127, 1288–​1349. Roberts, J. V. (2012a). Structured sentencing: Exploring recent developments in England and Wales. Punishment and Society, 14(3), 267–​288. —​—​. (2012b) Structuring sentencing in Canada, England and Wales: A tale of two jurisdictions. Criminal Law Forum, 23, 319–​345. —​—​. (ed.) (2015). Exploring Sentencing Practice in England and Wales. London: Palgrave Macmillan. Roberts, J. V., & Ashworth, A. (2016). The evolution of sentencing policy and practice in England and Wales, 1996–​2015. In M. Tonry (ed.), Sentencing Policies and Practices in Western Countries: Comparative and Cross-​ National Perspectives (pp. 307–​356). New York: Oxford University Press. Roberts, J. V., Azmeh, U., & Tripathi, K. (2011). Structured sentencing in England and Wales: Lessons for India? National Law School of India Review, 23,  27–​46. Roberts, J. V., & Baker, E. (2008). Sentencing structure and reform in common law jurisdictions. In S. Shoham, O. Beck, & M. Kett (eds.), International Handbook of Penology and Criminal Justice (pp. 551–​586). New York: Taylor and Francis. Roberts, J.V., & Bradford, B. (2015). Sentence reductions for a guilty plea: New empirical evidence from England and Wales. Journal of Empirical Legal Studies, 12(2), 187–​210. Roberts, J.V., & Cole, D. (eds.) (1999). Making Sense of Sentencing. Toronto: University of Toronto Press. Roberts, J. V., & Gazal-​Ayal, O. (2013). Sentencing reform in Israel: An analysis of th statutory reforms of 2012. Israel Law Review, 46, 455–​479. Roberts, J. V., & Hough, M. (2005). Understanding Public Attitudes to Criminal Justice. Maidenhead: Open University Press. Roberts, J. V., Hough, M., Jackson, J., & Gerber, M. (2012). Public attitudes toward the lay magistracy and the sentencing council guidelines:The effects of information on opinion. The British Journal of Criminology, 52(6), 1072–​1091. Roberts, J.V., & Pei,W. (2016). Structuring judicial discretion in China: Exploring the 2014 sentencing guidelines. Criminal Law Forum, 27(1),  3–​33. Sentencing Advisory Council. (2018). A Sentencing Advisory Council for Victoria. Melbourne,Victoria: Sentencing Advisory Council. Sentencing Commission of the Supreme Court of Korea. (2014). Sentencing Guidelines. Seoul: Supreme Court of Korea. Sentencing Council of England and Wales. (2011). Assault: Definitive Guideline. England: Crown Court. Retrieved from: www.sentencingcouncil.org.uk/​wp-​content/​uploads/​Assault_​definitive_​guideline_​-​_​Crown_​Court. pdf —​—​. (2016). Robbery: Definitive Guideline. England: Crown Court. Retrieved from: www.sentencingcouncil.org. uk/​wp-​content/​uploads/​Robbery-​definitive-​guideline-​web2.pdf —​—​. (2017). Sentence Reductions for a Guilty Plea: Definitive Guideline. England: Crown Court. Retrieved from: www.sentencingcouncil.org.uk/​wp-​content/​uploads/​Reduction-​in-​Sentence-​for-​Guilty-​plea-​Definitive-​ Guide_​FINAL_​WEB.pdf Sentencing Working Group of Northern Ireland. (2010). Monitoring and Developing Sentencing Guidance in Northern Ireland. Belfast: Northern Ireland Department of Justice.

85

Julian V. Roberts and Lyndon Harris Spohn, C. (2002). How Do Judges Decide? The Search for Fairness and Justice in Punishment. London: Sage Publications. Stemen, D., & Rengifo, A. (2011). Policies of imprisonment:The impact of structure sentencing and determinate sentencing on state incarceration rates, 1978–​2004. Justice Quarterly, 28(1), 174–​201. Stith, K. (2000). Sentencing guidelines: Lessons for the states. Saint Louis University Law Journal, 44, 387. Terblanche, S. (2003). Sentencing guidelines for South Africa: Lessons from elsewhere. The South African Law Journal, 120, 858–​882. Tonry, M. (ed.) (2016). Sentencing Policies and Practices in Western Countries: Comparative and Cross-​ National Perspectives. New York: Oxford University Press. Tonry, M., & Frase, R. (2001). Sentencing and Sanctions in Western Countries. New York: Oxford University Press. von Hirsch, A., Ashworth, A., & Roberts, J.V. (eds.) (2009). Principled Sentencing: Readings on Theory and Policy (3rd ed.). Oxford: Hart Publishing. von Hirsch, A., & Jareborg, N. (2009).The Swedish sentencing law. In A. von Hirsch, A. Ashworth, & J.V. Roberts (eds.), Principled Sentencing: Readings on Theory and Policy (pp. 258–​269). (3rd ed.). Oxford: Hart Publishing. Wandall, R. (2006). Equality by numbers or words: A comparative study of sentencing structures in Minnesota and in Denmark. Criminal Law Forum, 17, 1–​41. Young,W., & Browning, C. (2008).The origins and evolution of sentencing guidelines: A comparison of England and Wales and New Zealand. In A. Ashworth & J.V. Roberts (eds.), Sentencing Guidelines: Exploring the English Model (pp. 202–​217). Oxford: Oxford University Press. Young, W., & Browning, C. (2013). New Zealand's sentencing council. Criminal Law Review, 287.

86

PART II

Issues in Guideline and Non-​Guideline Sentencing

4 INTER-​DISTRICT DIFFERENCES AND EXTRA-​LEGAL DISPARITY IN FEDERAL SENTENCING The Trees Are “Substantially” More Important Than the Forest1 Richard D. Hartley

The intent of promulgating guideline-​based sentencing at the federal level was to provide uniformity in sentencing and to ameliorate unwarranted disparity concerning legally irrelevant factors such as race, ethnicity, and gender. The United States federal prison population since guideline implementation, however, has become increasingly made up of racial and ethnic minorities; recent statistics show that roughly 70% of federal inmates are racial or ethnic minorities (USSC, 2015). This increased disparity in sentencing and incarceration statistics has raised questions about whether guideline-​based sentencing policy has achieved the goals of uniformity and consistency in federal sentencing (for more thorough reviews, see Altshuler, 2005; Bushway & Piehl, 2007; Cole, 1997; Doob, 1995). Research and commentary on federal sentencing has also noted the complexity of applying the guidelines across the 94 federal districts. Due to varied types of caseloads, organizational practices, and localized discretion, individual courts may develop their own modes of implementing federal policy under the guidelines (Ulmer, 2005; Ulmer & Johnson, 2004). Roughly 25 years ago, and about six years after federal sentencing guideline implementation, Weisselberg and Dunworth (1993) observed that the “guidelines do not impact all cases and all districts equally” (pp. 26–​27). Their research led them to conclude that “the trees may be more significant than the forest.” In other words, they contended that district caseloads were not homogenous and thus any system-​wide analyses drawing general conclusions would not accurately reflect the effect of the guidelines on sentencing for each district. In the more than two decades since the publication of this paper, district-​level statistics validate that caseloads, departure rates, and average sentence length vary substantially across the 94 federal districts (Kautt, 2002; Johnson, Ulmer, & Kramer, 2008). Weisselberg and Dunworth’s comments would portend that localized and informal case processing norms (Nardulli, Eisenstein, & Fleming, 1988) would function as obstacles to uniformity and the amelioration of unwarranted disparity under the federal sentencing guidelines. The idea that a presumptive sentencing guideline structure would lead to more uniform and fairer sentence outcomes was championed by politicians and some in the judiciary. The upshot of guideline implementation, however, is a rather complex federal sentencing system; sentences under the guidelines are calculated using offense seriousness scores and criminal history scores, with level additions and subtractions for various aggravating and mitigating factors. District-​level variations in 89

Richard D. Hartley

the use of mandatory minimums, government-​sponsored departures (both substantial assistance and fast track), and safety valve reductions, as well as differences in dismissal, plea, and trial rates, differential application of prior history counts, and judicial departures further complicate achievement of uniformity in sentencing. Empirical research attempting to uncover the correlates of unwarranted disparity are further complicated by the fact that this process includes multiple decision points across many criminal justice system actors (Cassell & Luna, 2011; Mauer, 2010; Weisselberg & Dunworth, 1993). The practical result of guideline implementation was a more punitive sentencing structure; more offenders receiving prison sentences and for lengthier time periods.Whether fairer and more effective sentences have been produced as a result, however, is still today a subject for debate. The current study provides commentary on the nature of inter-​district differences and extra-​legal disparity in federal sentencing in general, and specifically utilizes the example of prosecutorial decision to grant substantial assistance departures, to exemplify how inter-district variation in the application of these departures can lead to extra-​legal disparities in federal sentencing outcomes.

Discretion Under the Federal Sentencing Guidelines Under the Federal Sentencing Guidelines, hereinafter referred to as the FSG, judges are prohibited from considering any extra-​legal factors in determination of an appropriate sentence. A large body of empirical research, however, has revealed that extra-​legal disparity persists post-​guidelines in federal sentencing (for extensive reviews, see Baumer, 2013; Spohn, 2000; Ulmer, 2012; Zatz, 2000). Other studies conclude that presumptive and determinate sentencing structures such as the FSG cannot practically remove extra-​legal disparity from sentencing because of the numerous methods that decision makers can use to manipulate the guidelines, or to circumvent them altogether (Hartley, Maddan, & Spohn, 2007; Nagel & Schulhofer, 1992). In the last decade, Supreme Court decisions (i.e., United States v. Booker, Gall v. United States, and Kimbrough v. United States) making the federal sentencing guidelines advisory in nature have resulted in further criticisms that judicial philosophies, in addition to district-​specific organizational priorities, might also drive inter-​district variation in sentencing. Subsequent to these SCOTUS decisions, judges need only consult the guidelines to obtain a “starting point” for an appropriate sentence but can then sentence above or below the guideline range for various reasons, including disagreement with sentencing commission policy. Post-​Booker, a majority of sentences continue to be within the guideline-​specified ranges (United States Sentencing Commission [USSC], 2009; USSC, 2012), despite a rise in government-​sponsored departures (Blackwell & Baisinger, 2016; Tillyer & Hartley, 2013), (i.e., substantial assistance departures and early disposition or fast-​track departures). A survey of federal district judges by the United States Sentencing Commission in 2010 revealed that only 32% of judges strongly agreed that the FSG have reduced unwarranted disparity for similarly situated defendants (USSC, 2010). Likewise, 30% strongly agreed that the FSG has increased certainty in sentencing, and only 22% strongly agreed that the FSG have increased fairness in sentencing (46%, 46%, and 45% of judges surveyed somewhat agreed with the above statements). In the same survey, however, 75% of judges responded that the current advisory guidelines system achieves the purposes of sentencing; only 8% of judges stated that no guidelines (i.e., the pre-​1987 era) would best achieve the purposes of sentencing, 3% preferred the mandatory guidelines (i.e., the system in effect prior to the SCOTUS decision in Booker), and 14% favored a revised guidelines system with broader sentencing ranges and fewer statutory mandatory minimums. The takeaway from this survey is that the majority of federal judges prefer the current advisory guidelines to the previous mandatory version, and believe that the purposes of sentencing are being fulfilled under them. Many of the judges who responded to the survey, however, also believe that the guidelines have not fully achieved the expected reduction in unwarranted disparity in sentencing practices. 90

Inter-district Differences

Sentencing Outcomes under the Federal Sentencing Guidelines Despite that most sentences handed out today (~60%) are within-​range sentences, critics argue that judicial departures and variances, and prosecutorial decisions related to charging, plea bargaining, and mandatory minimums still inhibit uniformity in sentencing for similarly situated offenders (USSC, 2012). Empirical research also reveals that certain constellations of extra-​legal factors influence the above decisions, thereby producing sentencing outcomes that are fraught with unwarranted disparity (Baumer, 2013; Mauer, 2010; Spohn, 2006). To date there is a large corpus of research examining extra-​legal disparity, and inter-​district differences in sentencing outcomes. This literature finds that the chief predictors of federal sentencing decisions continue to be legally relevant determinants (i.e., offense seriousness and criminal history); however, non-​legal (i.e., demographic) factors also influence outcomes. Collectively, the majority of these studies find that male and minority defendants receive the harshest court outcomes, net of controls for other legal, case processing, and extra-​legal, factors (for contemporary reviews, see Baumer, 2013; The Sentencing Project, 2005; Spohn, 2000; Ulmer, 2012; Zatz, 2000). Research has also revealed that extra-​legal variables interact with legally relevant factors to differentially influence sentencing outcomes across different combinations of offender race, ethnicity, and gender (Tillyer, Hartley, & Ward, 2015). Recent research has also attempted to understand how earlier decisions (i.e., the exercise of prosecutorial discretion in charging and plea bargaining) play a role in determining final outcomes (see for example, Hartley, Maddan, & Spohn, 2007; Johnson, Ulmer, & Kramer, 2008; Kutateladze, Lynn, & Liang, 2012; Maxfield & Kramer, 1998, O’Neill-​Shermer & Johnson, 2010; Spohn & Fornango, 2009; Tillyer & Hartley, 2013; Ulmer & Miller, 2002; Wilmot & Spohn, 2004). Finally, a limited number of studies have begun assessment of how district-​specific practices might also impact disparity, net of case-​specific factors (Anderson & Spohn, 2010; Johnson, 2014,, 2006; Johnson, Ulmer, & Kramer, 2008; Kim, Spohn, & Hedberg, 2015; Ulmer, 2005; Ulmer & Johnson, 2004). Indeed, many factors other than the FSG also influence sentencing outcomes. For example, mandatory minimum penalties greatly affect sentencing outcomes under the FSG, especially sentence length decisions. According to the USSC (2018a) in fiscal year 2017, approximately 22% of all offenders were convicted on offenses subject to mandatory minimums; approximately 70% of these involved drug trafficking offenses. Almost 40% of the offenders subject to mandatory minimums, however, received relief from the minimum penalty at sentencing; of these, about 20% received relief from the penalty due to provision of substantial assistance to the prosecutor, 13% received relief under the safety valve departure provision, and roughly 5½% received relief under both of the above. Being convicted of an offense carrying a mandatory minimum, or being relieved of that minimum, both of which are predicated on decisions by federal prosecutors, are important case-​ processing decisions because mandatory minimum penalties trump the presumptive guideline sentence, which drastically affects the sentence length that the presiding judge must impose. In 2017, for example, offenders who were sentenced under a mandatory minimum received an average sentence of 138  months, whereas offenders who got relief from the mandatory minimum received average sentences of 72 months and those sentenced for an offense carrying no mandatory minimum received average sentences of 28 months (USSC, 2018a). Statistics also show that being convicted of an offense carrying a mandatory minimum also varies by race/​ethnicity: 48% of those subject to a mandatory minimum were Hispanic, 30% were black, and 27% were white. Relief granted from the mandatory minimum penalty also varied by race/​­ethnicity: 51% of offenders who received relief from the mandatory minimum were Hispanic, followed by 30% who were white, and 24% who were black (USSC, 2018a). These figures reveal that prosecutorial decisions under the FSG are important considerations when examining unwarranted disparity in sentencing (Nagel & Schulhofer, 1992). Other research on prosecutorial discretion in the federal system reveals that females more often received charge reductions (O’Neill-​Shermer & Johnson, 91

Richard D. Hartley

2010), and males and minorities were generally less likely to receive a departure (Hartley, Maddan, & Spohn, 2007; Spohn & Fornango, 2009; Tillyer & Hartley, 2013; Ulmer, Light, & Kramer, 2011). Regarding prosecutorial influences on sentence lengths specifically, a USSC (2011) report revealed that the statute that doubles the mandatory minimum for offenders who have a prior conviction was used regularly by prosecutors in some federal districts, and not at all in others. Finally, Wilmot and Spohn (2004) found prosecutorial discretion related to “relevant conduct” resulted in drug offenders convicted of the same offense receiving dissimilar sentences. Sentencing under the FSG has been characterized as complicated at best, and at worst, as a “tragic mess” (Hofer, 2016, p.  649). Legislators believed judges’ decisions were the source of the lack of uniformity and unwarranted disparity in federal sentencing, but others argued that prosecutorial charging and bargaining decisions under the FSG would result in increased disparities in federal sentencing (Schulhofer & Nagel, 1989). Under the FSG, AUSAs have wide-​ranging discretion in the filing of charges, the level of charges, and the dismissal of charges (Albonetti, 1987). They also have independent discretion of information that sets the guideline range and control most opportunities for leniency and sentence reduction via substantial assistance and other government-​sponsored departures (Hartley, Maddan, & Spohn, 2007; Hofer, 2016; Tillyer & Hartley, 2013). A recent survey of federal judges (USSC, 2010) reported that judges believe that prosecutorial decision-​making contributes greatly to disparities in sentencing; when asked to rank the factors that they believed contributed to disparities in sentencing, 45% ranked statutory mandatory minimums and 57% ranked prosecutorial charging decisions as the number one or two causal factor. In the same survey, 76% of judges responded that the guidelines manual doesn’t contain provisions for departure that adequately reflect the reasons for which they want to impose sentences outside the guideline range, and 65% responded that the departure policy statements in the guidelines manual were too restrictive. Clearly, judges want more discretion to fashion appropriate sentences under the guidelines.

The FSG and Federal Narcotics Offenders Nowhere has criticism regarding fair and effective sentencing practices under the Federal Sentencing guidelines been debated more than in the case of narcotics offenses. The United States war on drugs was heightened during the time the guidelines were being implemented due to a moral panic surrounding crack cocaine that surfaced in the mid-​1980s (Chiricos, 2004). As such, the war on drugs, coupled with the Federal Sentencing Guidelines, produced large increases in the number of federal prosecutions and convictions for narcotics offenses (Saris, 2015). Although the overall proportion of federal inmates serving time for narcotics convictions has been declining over the last 15 years—​from 63% in 1997, and 55% in 2004 (Mumola & Karberg, 2006) to 48% in 2011 (Carson & Sabol, 2012)—​drug prosecutions in the federal system exploded from the early 1980s to the early 1990s; drug convictions comprised 18% of the federal criminal caseload in 1981, but comprised 34% in 1990 and represented 44% of convictions in 1992 (Weisselberg & Dunworth, 1993). Furthermore, the lengthy sentences handed out to narcotics offenders under the federal sentencing guidelines was the largest contributor to federal prison population increases from 1998 to 2010 (Mallik-​Kane, Parthasarathy, & Adams, 2012; Saris, 2015). In 2017, however, federal sentencing statistics revealed that roughly 40% of convicted drug traffickers were sentenced below the guideline range as a result of a government-​sponsored departure; another roughly 24% received a nongovernmental below-​range sentence, either via a judicial downward departure or variance (USSC, 2018b). Further, almost half (45.3%) of offenders convicted of drug trafficking offenses were convicted of offenses that triggered a mandatory minimum penalty, but half of those offenders received some form of relief from the minimum at sentencing. About 21% received relief from the mandatory minimum as a result of a government-​sponsored substantial assistance departure; another 20% were eligible for relief under safety valve departure provisions 92

Inter-district Differences

and roughly 8% received relief as a result of both a substantial assistance departure and a safety valve departure.

Substantial Assistance Departures A motion for a departure for substantial assistance, which is commonly referred to as a §5K1.1 departure, reflects the statutory authority set forth in Title 18 U.S.C. §3553(e), and can only be filed by a US attorney. Defendants who provide “substantial assistance”—​that is, information that leads to the prosecution and conviction of another offender—​can be sentenced outside the applicable guideline range.2 This type of departure is used frequently in narcotics cases. For example, from 2009–​2014, there were 57,485 substantial assistance departures granted; 59.2% of these were in drug cases, and another 5.5% were in firearm and drug cases (Blackwell & Baisinger, 2016). A substantial assistance motion made by the prosecutor and granted by the court removes the mandatory minimum sentence that otherwise would be binding at sentencing. Empirical research has shown that substantial assistance departures have been increasingly used to mitigate the sentences of convicted drug offenders, often in cases where a mandatory minimum sentence is applicable (Hartley 2008; Hartley, Maddan, & Spohn, 2007). Federal prosecutors unilaterally decide for whom motions for substantial assistance departures will be filed. Similarly, when a departure for substantial assistance is granted, a federal judge has unfettered discretion to determine the amount of sentence discount from the guideline sentence. Decision-​making regarding the granting of substantial assistance departure motions therefore has important implications on the types of sentences that narcotics defendants in the federal system receive. Moreover, according to the USSC (2017), the use of §5K1.1 departures varies widely across the 94 district courts. For example, departure rates across all offense types vary from highs of 76.2%,, 53.2%,, and 49.8% in the Southern District of California, the Western District of Washington, and the Middle District of Tennessee, respectively, to lows of 6.7%, 6.8%, and 10.3% in the Eastern District of Virginia, the Southern District of Mississippi, and the Western District of Oklahoma, respectively. These statistics reveal that exploration of inter-​district variation concerning these decisions may therefore have important implications for unwarranted disparity under the Federal Sentencing Guidelines.

Empirical Research on §5K1.1 Departures Research on substantial assistance departures has been relatively sparse. Nagel and Schulhofer (1992) studied §5K1.1 departures in three federal districts and found that the standard for determining whether the offender had provided substantial assistance varied among the districts. In two districts, prosecutors were allowed to file the motion even if the offender did not provide information that led to additional convictions; in these districts, substantial assistance motions frequently were used to mitigate the sentences of sympathetic defendants facing harsh punishment. In the other district, there was a stricter standard; defendants were typically required to provide information leading to an arrest and also be willing to testify in court. The researchers concluded that, “while the sentencing guidelines have reduced judicially created disparity, prosecutorial behavior may reproduce unwarranted disparity or, worse, discrimination based on race, sex, and social class, thereby compromising the goals of the Sentencing Reform Act” (Nagel & Schulhofer, 1992, p. 560). Ulmer and Miller (2002) studied four US District Courts and found that the odds of a substantial assistance departure varied by district, and that receiving this departure had a significant, but variable, effect on sentence severity in each district. Like Nagel and Schulhofer (1992), they also discovered that the size of the sentence discounts for §5K1.1 departures and the meaning of “substantial assistance” varied among the districts. In two of the districts, “providing information deemed useful was generally sufficient,” while in one of the other districts substantial assistance was defined 93

Richard D. Hartley

“as putting oneself in some position of risk in order to assist in the prosecution of another” (Ulmer & Miller, 2002, p. 14). Hartley, Maddan, & Spohn (2007) utilized federal data on crack and powder cocaine cases to examine prosecutorial discretion to grant substantial assistance departures. Their results show that for powder cocaine offenders, the odds of a substantial assistance departure are lowest for black and Hispanic males, while both white males and females had the largest odds of a departure. Similar results emerged for crack offenders with the exception that Hispanic females enjoyed the greatest odds of departures for substantial assistance. Other research has also supported a localized justice perspective where local politics and priorities at the district level predict differing adherence to the guidelines (Ulmer, 2005; Johnson, Ulmer, & Kramer, 2008). Ulmer (2005) found that some districts granted “soft 5K1s” in cases where the defendant provided questionable information; his qualitative interview data revealed that the origination of these so-​called soft departures is from a want to mitigate overly harsh guideline sentences for certain offenders (p. 263). Other studies have found inter-​district gaps in the use of government-​sponsored departures (Stith & Cabranes, 1998; Hartley, 2008; Hartley & Tillyer, 2012; Tillyer & Hartley, 2013), as well as interprosecutor variation within and across district courts (Spohn & Fornango, 2009; Kim, Spohn, & Hedberg, 2015). There also is evidence of racial disparities in the substantial assistance process. Research has found that among those who received a substantial assistance departure, the magnitude of the sentence discount was smaller for blacks than for whites (Maxfield & Kramer 1998; see also Steffensmeier & Demuth, 2000). Finally, some research reveals that jurisdictional variation exists regarding the case-​level factors that affect substantial assistance motions (Kautt, 2002; Spohn & Fornango, 2009; Ulmer & Johnson, 2004). These studies support criticisms that the federal sentencing guidelines did not ameliorate disparities but rather simply displaced discretion from judges to prosecutors (Miethe, 1987). In practice, prosecutors have unfettered discretion in many decision-​making areas under the federal sentencing guidelines, such as the filing of charges, the level of charges, and the dismissal of charges (Albonetti, 1987). Under the FSG, AUSAs also enjoy discretion in the decision to mitigate sentences by granting government-​sponsored departures in certain cases. These earlier discretionary decisions have become an important area of research as relatively little is known regarding the determinants of prosecutorial decision-​making at the federal level (Baumer, 2013; Bushway & Forst, 2012). These earlier decision-​making stages are also important as empirical research has concluded that they are significant determinants of final sentencing outcomes (Ulmer, 2012).

The Current Study The current research is an attempt to add to the extant literature regarding extra-​legal disparity in sentencing outcomes by exploring decision-​making practices of prosecutors to grant a departure for substantial assistance. More specifically, the current study reports findings from a larger project seeking to explore variation in the application and determinants of substantial assistance departures for federal narcotics offenders examining data disaggregated to the district level. Scholars of sentencing have recently emphasized that other process-​related and jurisdictional factors need to be considered to add to the modal model of sentencing research (Bushway & Forst, 2012; Baumer, 2013; Ulmer, 2012). Process-​related variables in each jurisdiction, from the type of courtroom workgroup and its organization, to practical application of the law, are as varied as its defendants. Therefore, exploring data disaggregated to the district level may uncover interesting findings regarding contexts for extra-​legal disparity in the system. This study therefore examines inter-​district variation in sentencing practices by analyzing data disaggregated to the district level. More explicitly, this study explores the predictors of prosecutorial decisions to grant a substantial assistance departure across federal district courts.Variation in the factors that affect this decision point across districts would indicate that the unwarranted disparities that the federal sentencing guidelines 94

Inter-district Differences

were designed to ameliorate may be creeping back into the system at these important decision-​ making stages.

Methodology The data utilized in this study come from the Bureau of Justice Statistics and reflect information on narcotics cases sentenced in United States Federal District Courts in 2004. The data were obtained from the Federal Justice Statistics Resource Center (file name: SC04OUT) online at http://​fjsrc. urban.org/​download/​dtsheet.cfm?year=2004, and then transferred into SPSS. This project focused solely on narcotics offenses and examined data for 20,5423 offenders convicted in federal courts in 2004. The reason for the focus on drug cases was two-​fold: offenders charged with drug offenses are more likely than other types of offenders to receive substantial assistance departures, and since guideline implementation, drug offenders have comprised a majority of inmates admitted to federal prison. The study’s analyses include variables measuring a number of case characteristics that, according to the United States Sentencing Commission, are legally relevant predictors of sentence severity. The main legal variable included in this analysis is the presumptive sentence.4 The presumptive sentence is the minimum guideline sentence that applies to an offender and is itself calculated by an offender’s final offense level, final criminal history category, and any upward or downward adjustments to an offender’s sentence. An offender’s presumptive sentence is also affected by any mandatory minimums that may be applicable. Mandatory drug minimums trump the guideline minimum and the offender’s presumptive sentence becomes whichever of these is longer. Factors that are not legally relevant to sentencing under the federal sentencing guidelines were also included as controls. They include the offender’s race/​ethnicity, sex (female = 1; male = 0), education, pretrial detention status (detained = 1; released = 0), and citizenship status (US Citizen = 1; Non-​citizen = 0). The offender’s race/​ethnicity was measured using three dummy variables (white, black, and Hispanic; other races were removed from the analyses due to small numbers in some of the districts); white offenders are the reference category. Education is measured by three dummy variables—​some high school, high school graduates, and some colleges which also included college graduates; some high school is the reference category.Variables were also created to control for type of drug; separate dummy variables were created for marijuana, powder cocaine, crack cocaine, heroin and methamphetamine; marijuana is the reference category. Finally, because this study analyzes data from 2004, a variable was included that controls for whether the case was before or after the Blakely v. Washington decision.5 Table 4.1 displays the frequencies or means for the variables included in the study. As shown in Table 4.1, roughly 25% of narcotics offenders in federal district court in 2004 received a departure for substantial assistance, and the mean magnitude of sentence discount for these departure recipients was almost 50%.The types of drug charges were varied, and included marijuana (30.5%), cocaine (25.2%), crack (22.5%), heroin (8.2%), and meth (13.7%). Approximately 30% of the offenders were black, 24% were white, and 44% were Hispanic; the majority of offenders were United States citizens (70.9%). Most were male (87.1%), almost half (49.8%) did not have a high school degree, and the majority were detained prior to sentencing (76.6%).

Analytic Strategy The dependent variable for the study is the receipt of a substantial assistance departure and is binary; coded 1 if the offender received a §5K1.1 departure and 0 if the defendant did not receive a departure. Of the 20, 542 narcotics offenders in this data set, roughly 25% (5,029) received a departure for substantial assistance. The initial step in this analysis was to examine the narcotics cases utilizing logistic regression on the decision to grant a departure for substantial assistance. 95

Richard D. Hartley Table 4.1  Frequencies and Means for Narcotics Defendants in Federal District Court 2004 Variable

N = 20,542 N

Presumptive Sentence (mean) Substantial Assistance Departure Granted Percentage Sentence Discount (mean) Mandatory Minimum Applicable Post-​Blakely Case Type of Drug Marijuana Cocaine Crack Heroin Meth Offender Race/​Ethnicity White Black Hispanic Offender is Female Offender’s Education Some High School HS Graduate Some College Offender was Detained Pretrial Offender is US Citizen

93.31 5,029 48.83 13,314 5,216

% 24.9 64.8 25.4

6,263 5,172 4,620 1,679 2,808

30.5 25.2 22.5 8.2 13.7

4,901 6,203 9,071 2,652

23.9 30.2 44.2 12.9

10,234 6,588 3,258 15,584 14,468

49.8 32.1 15.9 76.6 70.9

Results Results from the logistic regression model using the aggregated data (not shown here due to the focus on district-​specific analyses and results) suggests that prosecutors’ decisions regarding substantial assistance departures are based on both legally relevant case characteristics and legally irrelevant offender characteristics. An offender’s presumptive sentence is significant, and so is the variable for mandatory minimum. Interestingly the coefficients for these two variables are positive, meaning that as an offender’s presumptive sentence increases, and if the offender has a minimum apply, the odds of receiving a substantial assistance departure increase. This finding is interesting because those with longer presumptive sentences and those facing charges that carry a mandatory minimum are presumably more serious offenders, or at least are being charged with trafficking larger amounts of drugs. Despite this, these offenders are more likely to receive a §5K1.1 departure. An argument could be made that offenders with higher presumptive sentences and who qualify for a drug minimum should be receiving the lengthiest sentences. However, it could also be that those who are convicted on more serious narcotics offenses and have lengthier prior records play more important roles in the drug distribution network, or are key players in the drug trade; as a result, they may have more information to provide by way of assistance in another defendant’s case. An alternative explanation is that these findings might suggest that prosecutors are using substantial assistance departures to mitigate the sentences of offenders who are facing severe punishments or harsh mandatory minimum sentences. Prosecutors, in other words, may be using substantial assistance departure motions to reduce the sentences of certain offenders to a level they feel is appropriate. 96

Inter-district Differences

The likelihood of receiving a substantial assistance departure also varied by the type of drug. Results show that compared with marijuana, the likelihood of a substantial assistance departure was increased for cocaine (odds ratio = 1.42), crack (odds ratio = 1.37), and heroin (odds ratio = 1.38). Extra-​legal factors also played a significant role in the receipt of a substantial assistance departure. An offender’s educational attainment had significant effects on the likelihood of receiving a §5K1.1 departure. Those who were either high school graduates (odds ratio = 1.107), or had some college (odd ratio = 1.321) had increased odds of departure compared with those offenders who had not graduated high school. Females (odds ratio = 1.450), and US citizens (odds ratio = 1.221), also had increased odds of receiving a departure for substantial assistance, compared to males and non-​US citizens. Those who were detained (odd ratio = 0.554), however, were less likely to receive a substantial assistance departure than those released. The variables measuring race/​ethnicity had significant effects. Both black (odds ratio = 0.610) and Hispanic (odds ratio = 0.581) offenders had lower odds of receiving a substantial assistance departure than white offenders. These race and ethnicity effects, along with those for gender, citizenship, and educational attainment, suggest that extra-​legal factors are still influential under the Federal Sentencing Guidelines in this case in the context of substantial assistance departure decisions. Unwarranted disparity, therefore, still exists in the federal sentencing process.6 Discretionary decisions that occur prior to the sentencing stage are important in examining any extra-​legal disparity that may be influencing final sentencing outcomes (Ward, Hartley, & Tillyer, 2016). Because the contexts surrounding the filing of substantial assistance motions vary by district, the next step in the analysis was to explore this decision-​making stage for each district separately. This required partitioning the data by district and running logistic regression analyses on the disaggregated data.

Descriptive Statistics for Data Partitioned by District The federal districts cannot be uniformly characterized, and therefore it is difficult to predict the factors that affect decision-​making in each district with any consistency. Previous research has shown that analyzing aggregated data may mask inconsistencies that are present across jurisdictions (Spohn, 2006). When the data were disaggregated by district, however, the sample sizes for some of the districts were small. A power analysis estimating the sample size needed in order to obtain medium (desired) effect sizes was calculated and results determined that a minimum of 100 cases would be necessary in order to perform meaningful analyses at the district level. The number of cases in some of the districts fell below this 100-​case threshold and, therefore, these districts were excluded from the subsequent analyses. After removing districts where the number of narcotics cases was less than 100, 59 districts remained. In some of these districts, however, there were still problems related to cell sizes for some of the variables, such that the logistic regression models could not be estimated; there were also some districts where none of the variables used in the analyses reached statistically significant levels. The district-​level logistic regression analyses for receipt of a substantial assistance departure, therefore, were performed for 45 districts. An examination of the frequencies for variables of interest across the 45 districts (not shown) revealed that the percentage of narcotics offenders receiving a substantial assistance departure varied considerably from a low of 3.1% in the Eastern District of Virginia to a high of 56.6% in the Eastern District of Pennsylvania; 27 of the 45 districts granted substantial assistance departures in over 25% of cases. In 82% of the districts, over half of the narcotics offenders were subject to a mandatory drug minimum. Further, half of the districts had mandatory drug minimums apply in over 75% of cases. Average presumptive sentences ranged from a low of 33 months in the Southern District of California to 181 months in the Northern District of Florida, with the mean presumptive sentence across all districts being roughly 106 months. 97

Richard D. Hartley

The average age of offenders across districts was early thirties, and a majority of the offenders in all districts were male; there were only four districts (the Eastern District of New York, the Southern District of Alabama, and the Districts of Vermont and New Jersey) where females made up more than 20% of the cases. Racial and ethnic minorities comprised the majority of offenders in all but five districts (the District of Vermont, the Eastern District of Tennessee, the Central District of Illinois, the Northern District of Iowa, and the Western District of Washington). As for drug type, marijuana was the most prevalent drug in 20% of the districts, while powder cocaine and crack cocaine comprised the largest caseload in 29 and 51% of the districts, respectively. A majority of offenders in all of the districts, with one exception, were detained; the exception was in the District of Vermont, where the percentage of offenders released was equal to the percentage who were detained. Finally, in most districts (84.5%) the majority of offenders were United States citizens.

Disaggregated District Logistic Regression Results Logistic regression analyses were performed for each of the 45 districts; however, because a discussion of each of these regression models separately would be both a cumbersome and tedious task, Table 4.2 summarizes the results of these 45 models to display the collective results of the district-​ level analyses.7 As shown in Table 4.2, the only variables that affect the receipt of a substantial assistance departure with any degree of consistency are the presumptive sentence, the presence of a mandatory minimum sentence, whether the offender was female, and whether the offender was detained. However, the effect of the presumptive sentence on the likelihood of a substantial assistance departure varies among the districts. In over half of the districts (25), the effect is positive, meaning that the higher an Table 4.2  Summary of Logistic Regression Results across Districts on the Likelihood of Receiving a Substantial Assistance Departure Variable

Presumptive Sentence Mandatory Minimum Post-​Blakely Type of Drug (Marijuana = Reference) Cocaine Crack Heroin Meth Age Offender is Female Offender Race/​Ethnicity (White = Reference) Black Hispanic Offender’s Education (Some High School = Reference) HS Graduate Some College Offender was detained Offender is US Citizen

Coefficient Positive & Significant

Coefficient Negative & Significant

Coefficient Nonsignificant

N

%

N

N

%

25 16 4

55.5 35 8.9

0 0 4

–​ –​ 8.9

20 29 37

45.5 65 82.2

4 2 1 3 1 12

9.1 4.7 4.3 8.3 2.2 26.7

3 5 2 2 4 0

6.8 11.6 8.7 5.6 8.9 –​

37 36 20 31 40 33

84.1 83.7 87 86.1 88.9 73.3

0 0

–​ –​

7 4

15.9 8.9

37 41

84.1 91.1

1 7 1 5

2.2 15.6 2.2 11.9

1 0 16 0

2.2 –​ 35.6 –​

43 38 28 37

95.6 84.4 62.2 88.1

98

%

Inter-district Differences

offender’s presumptive sentence was, the greater the likelihood of receiving a departure for substantial assistance, but in the other just under half of the districts (20) the effect of the presumptive sentence is nonsignificant; there were no districts where presumptive sentence had a negative effect on receiving a substantial assistance departure. The presence of a mandatory minimum also had significant effects in some districts. Although for most of the districts (29), the presence of a minimum has no significant effect on the filing of substantial assistance motions, in 16 districts, those caught with enough narcotics to trigger mandatory minimum sentences actually have increased chances for §5K1.1 departures. Again, a possible explanation for this finding may be that these offenders are higher up in the narcotics trade and may have more to offer by way of substantial assistance. Alternatively, prosecutors may believe that sentence mitigation for these offenders is appropriate because the mandatory minimum sentence they are facing is overly harsh. The findings in Table  4.2 also reveal that analyses of aggregated federal sentencing data may overstate the significance of certain variables which includes the significance of extra-​legal factors. Nonetheless, extra-​ legal factors (gender, race, ethnicity, education) significantly influence these decision-​making points in some districts. Table 4.2, for example, reveals that the sex of the offender significantly affects the likelihood of a substantial assistance departure in just over one-​third of the district studied here. In 12 of the 45 districts, females have a greater likelihood of §5K1.1 departure than males; in 33 of the districts, there is no statistically significant gender differences. Explanations proffered in past research for this finding point to the fact that prosecutors may be sympathetic to female offenders who are facing lengthy prison sentences, or that prosecutors may charge certain females with narcotics offenses and then use substantial assistance motions as leverage to get them to turn evidence against their boyfriends or spouses who play a more major role in the offense (Hartley, Maddan, & Spohn, 2007). Likewise, in the disaggregated analyses performed here, the race and ethnicity of the offender was significantly related to receiving a substantial assistance departure in only 11 of the 45 districts. In seven districts, black offenders were significantly less likely to receive a departure for substantial assistance than white offenders whereas in four of the districts Hispanic offenders had significantly lower odds of a departure than whites. In the majority of the districts, 37 and 41 respectively, the race and ethnicity of the offender had no statistically significant relationship with the odds of receiving a departure for substantial assistance. Finally, in one district, being a high school graduate increased the odds of a departure, and in another seven districts, offenders with some college credits had larger odds of a departure than those without a high school degree. Detention status also has significant effects in some districts. For most of the districts, it does not affect the receipt of a substantial assistance departure but in 16 districts it negatively affected the likelihood of a departure; in only one district was being detained correlated with increased chances for departure. Explanations for this finding might be that prosecutors in some districts do not believe that those who are detained are deserving of departures, or it may also be that those detained are less likely or available to cooperate with prosecutors in other cases. The findings from the district-​level analysis reveal that there is inter-​district variation in the correlates of substantial assistance departures, and that extra-​legal factors play a role in these prosecutorial decisions in some districts. However, because the log of the odds is not a practical or logically understandable way to interpret the effects of x on y (Roncek, 1991, 1993), the interpretation of results for particular independent variables from logistic regression analyses (odds ratios) can be a complex and misunderstood concept. As such, calculating predicted probabilities is sometimes a preferred method for interpretation of logistic regression results; this is especially the case for the district-​level comparisons attempted here. Interpretation of logit coefficients therefore should be related back to probabilities; referring back to the equation of ln(P/​1-​P) = a + bx, then exponentiating the right side of the equation and carrying out the simple algebra yields that the probability is P = ea+bx /​1 + ea+bx. 99

Richard D. Hartley

This gives the predicted probability for an individual case from the logistic regression model, and tends to be more easily interpreted than the log of the odds (Roncek, 1991). Predicted probabilities are commonly calculated for meaningful values of x.  In this way, the predicted probability of a substantial assistance departure can be calculated for typical offenders in each district. Likewise, these probabilities can also be calculated for the different offender race/​ethnicity and gender combinations utilizing the coefficients in the models. Recall that race and ethnicity had negative effects on receipt of a departure in seven and four districts respectively. Gender effects, however, were present in just over a quarter of the districts (12). Therefore, to further explore inter-​district differences in the likelihood of receiving a substantial assistance departure, predicted probabilities of receipt of a departure were calculated for offenders with six different demographic combinations of race/​ethnicity and gender (white males and females, black males and females, and Hispanic males and females).

Disaggregated District Predicted Probability Results For the purposes of revealing the extent of inter-​district disparity in receipt of 5K1.1 departures, the figures that follow below display the predicted probability of a substantial assistance departure for the typical offender for the data utilized in the current study. For comparative purposes, the predicted probabilities were calculated where the typical offender was a US citizen, had a high school degree, was sentenced under a mandatory minimum, was charged with a powder cocaine offense, and was detained prior to sentencing.The average age of offenders (32), and median presumptive sentence for all defendants (101) was also used in the calculation for these predicted probabilities. The figures that follow (4.1 through 4.11) show the differences in predicted probabilities by district for white, black, and Hispanic, males and females. For simplicity and comparative purposes, the district probabilities are nested within circuits. Figure  4.1 displays the results for the Districts of Massachusetts, Connecticut and Vermont. These districts were combined for comparative purposes even though they are in different circuits; Massachusetts is in circuit 1, while Connecticut and Vermont are in circuit 2. Circuits 1 & 2

Predicted Probability of Substantial Assistance Departure

90

84

89

80

88

81

70 60 50 40

36

30

0

38

25

20 10

39 36

10 6 3

9

WM

BM

2

13

13

12

7 HM WF Race/sex

Massachusetts

BF

Connecticut

Figure 4.1  Predicted Probability of Substantial Assistance Departure: Circuit 1

100

HF Vermont

Inter-district Differences

Predicted Probability of Substantial Assistance Departure

The most obvious result from Figure 4.1 is that predicted probabilities of departure are highest in Vermont, but not for all offenders. White males have the highest probability of receiving a departure for substantial assistance at 89%, followed by Hispanic females at 88% and then Hispanic males and white females at 84 and 81%, respectively. Black males and females have the lowest probability of departure in Vermont at 9 and 13%, respectively. This means that in Vermont, white males have an 80-​percentage point greater probability of being granted a departure for substantial assistance than black males. The difference between White and Black females is 68 percentage points. These results show an obvious racial disparity in the filing of substantial assistance motions in Vermont. Conversely, there is only a 5-​percentage point difference in probabilities between white and Hispanic males, and Hispanic females actually have a 7-​percentage point greater probability of receiving a departure than white females. The probability differences are not as pronounced for Massachusetts and Connecticut. In Connecticut, black males and females have a greater probability of departure than their white and Hispanic counterparts, and in Massachusetts, Hispanics (both male and female) have the highest probability of departure. In all three districts, females of any race/​ethnicity have greater probabilities of departure than males (except for whites in Vermont). Figure 4.2 displays the results for the four federal districts in New York. The results in Figure 4.2 show that offenders in the Northern and Western Districts of New York have much greater probabilities of departure for substantial assistance than offenders in the Eastern and Southern Districts. Univariate statistics show that the Eastern and Southern Districts of New York granted substantial assistance departures in roughly 20% of cases, whereas the Western and Northern Districts granted them in more than 40% of cases. The Eastern District of New York has the least racial/​ethnic and gender disparity in probabilities of departure, followed by the Southern District. In the Northern District of New York, white males have the lowest probability of departure (32%); the probability for black males is twice as great, and for Hispanic males it is 25 percentage points greater. In the Western District of New York, however, the opposite is true; white males have higher probabilities than both black and Hispanic males. Regarding gender, in three of the four New York Districts, females, regardless of race/​ethnicity, have greater probabilities of departure than males. Further, within the state of New York itself, probabilities for substantial assistance for powder cocaine offenders facing mandatory minimums differ by as much as 84 percentage points (black females in New York North versus New York East).

100 90 80 70 60 50 40 30 20 10 0

Circuit 2 86 71

88

69

68

82 78

67

66

57

55 32 17 5 WM

17

15

9 BM

7 4

2

4

9

HM WF Race/sex

BF

NY North

NY East

NY South

NY West

Figure 4.2  Predicted Probability of Substantial Assistance Departure: Circuit 2

101

12 7 HF

Predicted Probability of Substantial Assistance Departure

Richard D. Hartley Circuit 3

100 90 80 70

88

70

88

60

58

50 40 30

42

42 22

43 40

27

20 15

10 0

88

WM

BM

HM WF Race/sex New Jersey

BF

HF

Maryland

Figure 4.3  Predicted Probability of Substantial Assistance Departure: Circuit 3

Figure 4.3 displays the results for the third circuit, which includes the Districts of New Jersey and Maryland. Again, in both districts female offenders of any race/​ethnicity have greater probabilities than their male counterparts of receiving a departure. In Maryland, white, black and Hispanic females have equal probabilities of departure. In New Jersey, Hispanics males and females have greater probabilities of departure than all other offenders. Finally, in Maryland, white males have an almost 30-​percentage point greater probability of receiving a departure than either black or Hispanic males. Both of these districts granted departures in roughly 36% of narcotics cases. Figure 4.4 displays the results for the districts in the fourth circuit, which include the Eastern and Western districts of North Carolina, the District of South Carolina, and the Eastern District of Virginia. As can be seen from the figure, offenders in the Eastern District of Virginia have a very low probability of departure compared with the other districts, and especially the Eastern District of North Carolina. Females, and especially Hispanic females, have the greatest probability of departure regardless of race/​ethnicity and regardless of district. What is also worth noting from these results is that despite the fact that these districts are all in the same circuit, the probability of substantial assistance motions being filed varies widely across districts. Even within the same state (North Carolina), probabilities of a 5K1.1 departure differ across districts by as much as 38 percentage points for Hispanic males, 35 percentage points for black males, and 32 percentage points for Hispanic females. Figure 4.5 displays the results of the probability of substantial assistance for districts within the fifth circuit. Again, the probabilities of receiving a substantial assistance departure differ substantially across the districts; most notably, there is a 91-​percentage point difference in the probability of departure for white males in Southern Mississippi versus white males in Eastern Louisiana. The Southern District of Mississippi has the largest disparities between white, black, and Hispanic defendants, and the Eastern District of Louisiana has the smallest. In Texas, among males, whites have the greatest probability in the Northern District whereas Hispanics have the greatest probability of departure in the Southern and Western Districts. Among females convicted in Texas, the probabilities are greatest for whites in the Northern and Western Districts but in the Southern District black females have the greatest probability for departure. Again (except for Whites in the Southern District of Mississippi), females, regardless of race, ethnicity, or district, have higher probabilities of departure for substantial assistance than their male counterparts. 102

Inter-district Differences

Predicted Probability of Substantial Assistance Departure

Circuit 4 80

80

75

78

60

59 51 43 37

40

50

78

58

59

48

47

39

34

32

24

20 3

1

WM

BM

0

48

7

3

HM WF Race/sex

BF

3

6 HF

Virginia East

N Carolina East

N Carolina West

S Carolina

Figure 4.4  Predicted Probability of Substantial Assistance Departure: Circuit 4

Predicted Probability of Substantial Assistance Departure

100 90

Circuit 5 94 87

80 70 60

52

50

44

40 30 20 10 0

29 21 20 13 3 WM

20 17 2 2 BM

25 24 21 6 5

39 20 8

HM WF Race/sex

Miss South Louisiana East Texas North

37 35 29

29 24 4

11 8

4

BF

HF

Texas South Texas West

Figure 4.5  Predicted Probability of Substantial Assistance Departure: Circuit 5

Figure  4.6 displays the probabilities of substantial assistance departure for three districts in the Sixth Circuit. The Eastern District of Tennessee shows similar results to the other districts shown thus far in that females have higher probabilities of departure than males for all of the race/​ethnicity categories. In the Eastern District of Michigan and the Northern District of Ohio, however, males, regardless of race or ethnicity, have higher probabilities of departure than females. Male and female whites have the highest probabilities for departure in the Eastern District of Michigan, whereas in the Northern District of Ohio and Eastern District of Tennessee, black males 103

Predicted Probability of Substantial Assistance Departure

Richard D. Hartley

70 60 50

Circuit 6 64 57 55

55

55 48

50

49

67

65

43 41

40 36

30

31

40 34 30

30

20 10

WM

BM

HM WF Race/sex

Mich East

BF

Ohio North

HF

Tenn East

Predicted Probability of Substantial Assistance Departure

Figure 4.6  Predicted Probability of Substantial Assistance Departure: Circuit 6

Circuit 7

70

58

60

54

50

53

50

40 30

24

22

20 10 0

24 23

5

8 1.5

1

WM

BM

Illinois Cent

0.5

29 20 1.4

HM WF Race/sex Illinois North

1.9 BF

0.7 HF

Wisconsin East

Figure 4.7  Predicted Probability of Substantial Assistance Departure: Circuit 7

and females have higher probabilities of receiving a departure for substantial assistance than other offenders. Again, inter-​district disparity in the probability of departure is present in the Sixth Circuit. Figure  4.7 shows the departure probabilities for three districts within the seventh circuit. Interestingly, the probabilities of receiving a substantial assistance departure in the Eastern District of Wisconsin, regardless of the offender’s race/​ethnicity and gender, are almost non-​existent. For the two districts in Illinois, however, departure probabilities are higher, especially in the Northern District of Illinois. Female offenders in these districts have the greatest probability of departure, moreover, Hispanic females in the Central District of Illinois have the highest probability of departure of any offenders in these districts; their probability of departure is almost 30 percentage points greater than the probability for black females and 38  percentage points greater than the probability for white females. Interestingly, within the Northern District of Illinois, there appears to be some degree of parity between white, black, and Hispanic males, and white, black, and Hispanic females; there is only a 104

Inter-district Differences Circuit 8 65 Predicted Probability of Substantial Assistance Departure

60

61 46 42 41

40 34 32 28 22 20

0

18 17 12 7 0.4 WM

12 10 1.4 BM

38

33

27 23

21 17

19 15

15 3.2

4.4

HM WF Race/sex

BF

2

0.6

45 42 37 36

5 1.8 HF

Iowa North

Iowa South

Missouri East

Missouri West

Nebraska

Minnesota

Figure 4.8  Predicted Probability of Substantial Assistance Departure: Circuit 8

2-​percentage point difference for males and a 1 to 4 percentage point difference for females in the probability of departure. For this district there is almost no racial or ethnic disparity but a large gender disparity regardless of race. Figure 4.8 displays the probabilities for a substantial assistance departure for districts within the eighth circuit. Similar to Eastern Wisconsin, offenders in the District of Nebraska, regardless of race, ethnicity or gender have a very small probability of receiving substantial assistance departures. In all six districts, however, females, regardless of race or ethnicity, have a greater probability than males of receiving a departure for substantial assistance. Black male and female offenders have the greatest probability of departure in the Western District of Missouri and in Minnesota. Hispanic males and females have the greatest departure probabilities in the Northern District of Iowa but the lowest probabilities in the Southern District of Iowa. In the Eastern District of Missouri, there is more racial/​ethnic parity regarding departure probability for both males and females than in the other districts shown here. Again, Figure 4.8 illustrates that disparity in departure probabilities across districts is considerable in the eighth circuit and even across districts within the same state, as in Iowa and Missouri. Figure 4.9 displays the departure probabilities for four districts within the ninth circuit; note there were too few black offenders in the Western District of Washington who received a departure for substantial assistance to be included in this analysis; therefore, there are no predicted probabilities listed in the figure for these offenders. In the Eastern District of Washington, both white males and females have higher probabilities of receiving a departure for substantial assistance than their Hispanic counterparts. In both Districts in California, females, regardless of race/​ethnicity, have higher departure probabilities than males. By contrast, in Arizona both black and Hispanic males (but not white males) are more likely than their female counterparts to receive departures. Also, in the Eastern District of California, black offenders, both male and female, have the highest probability of departure, followed by Hispanics. In the Southern District of California, for males, Hispanics have the greatest probability and for females, 105

Richard D. Hartley

Predicted Probability of Substantial Assistance Departure

Circuit 9 80

78

60

40

57

49 42

47

39 41

30

31

26 9

7 WM

58 51 40

38 35

24 14

20

0

59

57

BM

HM WF Race/sex

BF

HF

Wash West

Arizona

Cali East

Cali South

Figure 4.9  Predicted Probability of Substantial Assistance Departure: Circuit 9

Circuit 10

100 Predicted Probability of Substantial Assistance

85 80

77 62

60

60

46

52 48

40 38

41

39

20

15

9 0 WM

BM

HM WF Race/sex Kansas

BF

HF

New Mexico

Figure 4.10  Predicted Probability of Substantial Assistance Departure: Circuit 10

whites enjoy higher probabilities. Again, inter-​district disparity in the probability of departure exists, even between the two districts within the State of California. Figure 4.10 displays the probability for substantial assistance departure for two states in the tenth circuit, Kansas and New Mexico. Hispanic males and females in both districts have the highest probability of departure, followed by white males and females and then black males and females. In New Mexico, the difference in predicted probability of departure between Hispanic males and black males is 68percentage points, and between Hispanic females and black females, the difference is 70 percentage points. In Kansas, the difference is not as pronounced; there is a 21-​percentage point difference between both Hispanic and black males and females. Figure 4.11 displays the predicted probabilities for six districts in the eleventh circuit. Probabilities of a substantial assistance departure are highest in the Southern District of Alabama, and lowest in 106

Inter-district Differences Circuit 11

Predicted Probability of Substantial Assistance Departure

100

93

95

40

0

69 66

70 63

55 49 31

20

94

78

80

60

96

19 5 WM

62 43 39 36

38 35

41 29 27

18 8 BM

17 7

63

37 35 26 19

13

HM WF Race/sex

BF

AL South

FLA Mid

FLA North GA North

FLA South GA Mid

28 24 18

HF

Figure 4.11  Predicted Probability of Substantial Assistance Departure: Circuit 11

the Middle District of Georgia. Hispanics, both male and female, have the lowest probability for departure in the Southern District of Alabama, and the Northern and Southern Districts of Florida, but the highest probabilities for departure in the Middle District of Georgia. Whites, both male and female, have the highest probabilities of departure in the Middle and Northern Districts of Florida, as well as the Southern Districts of Florida and Alabama, but the difference between blacks and whites is not as great in these latter two districts. Black offenders have the highest probability for departure in the Northern District of Georgia; in this district, however, probabilities for offenders overall are low.Again, there is a great deal of inter-​district disparity regarding the probability for departure across these six districts and even between districts within the states of Florida and Georgia.There is an especially wide gap between the probability for departure between the Northern and Middle Districts of Georgia, particularly for Hispanic males and females. This finding is especially surprising given that in the Middle District of Georgia roughly 10% of offenders were Hispanic compared with the over 46% of Hispanic offenders in Georgia’s Northern District.This finding, therefore, probably reflects local district-​level philosophies about which offenders are deserving of a substantial assistance departure and not reluctance on the part of Hispanic defendants to provide substantial assistance.

Discussion Presumptive sentencing guidelines were championed as an apparatus for the achievement of more uniform and less disparate sentencing outcomes. Sentencing under the FSG, however, is a rather complex process wherein multiple courtroom actors across several decision-​making stages undertake decisions which increase or decrease the final presumptive sentence. District-​level variations in the use of legal mechanisms (i.e., mandatory minimums, departures, safety valve reductions, and plea and dismissal rates) also affect final court outcomes which further support arguments that unwarranted disparity is still present under the current guideline structure. The current study was an attempt to provide commentary on the nature of inter-​district differences and extra-​legal disparity in federal sentencing. This study also sought to identify the correlates of 107

Richard D. Hartley

prosecutorial decisions to grant substantial assistance departures and to examine the consistency in the effects of these correlates across federal districts. The findings highlight how racial/​ethnic and gender disparities can be produced at earlier stages of the federal sentencing process via prosecutorial decisions. These types of contextual disparities are not modeled in most post-​conviction sentencing studies, and overall the results here reveal how the varying application and influences of substantial assistance departures at the district level can lead to extra-​legal disparities in federal sentencing outcomes. More specifically, the results of this study reveal two consistent findings about the predicted probability of a substantial assistance departure for the six racial/​ethnic and gender groups examined here. The first finding is that in most districts, female offenders, regardless of race or ethnicity, have a higher probability than male offenders of receiving a §5K1.1 departure. Again, whether this is evidence that prosecutors in these districts are sympathetic towards female narcotics offenders, who in their minds are not as culpable as their male counterparts, yet are facing lengthy mandatory minimum prison sentences, is not known. Another explanation might be that prosecutors are using substantial assistance departures to get females to testify against their husbands or boyfriends who are likely more involved in the narcotics trafficking offense. Unfortunately, with these data it is not known what criteria prosecutors consider in the granting of a substantial assistance departure. In other words, some districts may require testifying against other defendants, or providing information leading to the arrest of other narcotics offenders, whereas in other districts the criteria may simply constitute providing names of other offenders (Ulmer, 2005). What is known from these findings, however, is that regardless of which of the above explanations is true, gender disparities are present in the granting of substantial assistance departures, and therefore, a source of continued gender disparity in the sentencing of narcotics offenders under the federal guidelines sentencing system. There is also some racial and ethnic disparity within districts (and circuits) regarding the probability of receiving a departure for substantial assistance. This disparity, however, is not always in favor of white defendants and to the disadvantage of black and Hispanic defendants. Likewise, these racial and ethnic disparities in 5K1.1 decisions have implications for the racially and ethnically neutral sentencing practices that the United States Sentencing Commission has attempted to realize through promulgation of federal sentencing guidelines. The second finding that emerges from these analyses is that there is also substantial disparity across districts regarding the probabilities that prosecutors will file 5K1.1 motions for any defendant. This inter-​district disparity in the filing of substantial assistance motions was present across districts within the same circuit, and even between districts within the same state. The findings from the predicted probability analyses revealed differences across race, ethnicity and gender in receipt of these departures. From these findings, unwarranted disparity (that due to extra-​legal factors) is persistent under the sentencing guidelines system that was implemented to ameliorate it. Considering that average sentence discounts for offenders who received these departures was 50% across the districts studied here, and in some districts it was upwards of 80%, substantial assistance departures could be a continued source of unwarranted disparity in final sentencing outcomes.

Conclusions Those advocating for federal sentencing guidelines, including members of Congress, believed that the result would be a more punitive, more effective, and more just federal sentencing system. Although research on the results of reforms has been mixed, sentences are more punitive today than they were in the pre-​guideline era. The observable result has been a significant increase in the nation’s federal prison population. Questions about the effectiveness and fairness of guidelines, however, have been difficult to answer because courtroom actors have found ways to circumvent or manipulate the guidelines. In the post-​guideline era, sentences are more uniform and less disparate because there is less inter-​judge disparity, and because sentences are now attached to the seriousness of the 108

Inter-district Differences

offense and an offender’s prior record. Extra-​legal factors, however, still play a role at some decision-​ making stages. The findings from this study provided evidence of this; racial and ethnic minorities, and females, in some districts have vastly different probabilities of having their sentences reduced via a substantial assistance departure and therefore arrive at final sentencing outcomes for drug offenses differently than whites and males. Given these findings, it is clear that extra-​legal variables are still affecting sentencing outcomes under guideline-​based sentencing, but here these effects are via prosecutorial discretion rather than the more oft studied judicial discretion. The findings from this study also highlight inter-​district variation in federal sentencing outcomes for narcotics offenders; this suggests that discretionary decisions regarding substantial assistance departures may be a continuing source of unwarranted disparity in the federal sentencing process. Even though the motivation behind the decisions of prosecutors in each of these districts is not known, it is clear from the findings of this study that their decisions are influenced by both legally relevant and irrelevant variables, and perhaps by a desire to reduce the sentences of certain offenders. Moreover, inter-​district variation regarding the filing of substantial assistance motions indicates that the application of the FSG may not be as uniform as it appears in the aggregate. Even for districts within the same circuit, and in some instances, within the same state, the decisions of prosecutors and judges may vary, and localized sentencing practices may result in considerable sentencing variation for similarly situated offenders. A recent investigation of roughly 1.8  million federal criminal investigations from 2005–​2015 revealed wide disparities in prosecution rates for the same crimes for districts within the same state and districts in bordering states (Conte & Bowling, 2016). Using United States Department of Justice Data, the authors compared declination rates of certain offenses, citing, for example, that attorneys in the Southern District of Florida declined roughly 25% of healthcare fraud cases whereas 35 miles up the road in the Middle District of Florida, attorneys declined almost 60% of the complaints of healthcare fraud. Likewise, attorneys in South Carolina prosecuted 89% of the counterfeiting crimes referred to them, whereas across state lines in the Eastern District of North Carolina, attorneys prosecuted only 35.7% of such cases. Arrests for drug trafficking were declined in over 75% of cases in Northern Ohio encompassing the city of Toledo, but the US attorney across city and state lines on Interstate 75 in the Eastern District of Michigan declined to prosecute in only 16.6% of federal drug trafficking cases. These statistics reveal that US Attorney’s offices in the same region might have differing priorities and decide to spend their prosecution resources in very divergent ways, with the result that the same federal crimes fully prosecuted in one district may be declined or referred to state or local prosecutors in another district (Conte & Bowling, 2016). In the article, Justice Department officials explain that these statistics do not tell the entire story and that prosecutors base their decisions on priorities in the district and cooperation with other local agencies. Interviews with one US Attorney cited that prosecutorial decisions take into account the seriousness of the offense as well as the offender’s criminal history and federal priorities. Others stated that the lead US Attorney for each district determines what cases to prosecute, which might be based on thresholds for drug amounts or money embezzled. Conte and Bowling (2016) state the reasons most often cited by AUSAs for declining cases include lack of evidence or weak evidence, and priorities related to federal interests and resources. The above statistics also highlight the problem with not having publicly available data on the decision-​making practices of federal prosecutors. Because the Department of Justice does not make these data available to the public, we have no information on what other factors, in addition to evidence, or lack thereof, and district priorities, might also be influencing prosecutorial declination decisions.This inhibits the ability of the general public to evaluate whether certain types of offenders are over-​represented in declination decisions, and the ability of researchers to analyze whether extra-​ legal factors are influencing these decisions. The current study was unable to determine whether the offenders in some of the districts studied here were more willing to provide substantial assistance or had more to offer by way of substantial 109

Richard D. Hartley

assistance, or whether the priorities and /​or beliefs about the appropriate use of 5K1.1 departures of prosecutors across these districts are different. What is clear, however, is that attempts to make the federal sentencing system uniform and proportional by promulgating guidelines may not have fully achieved the desired effects, at least in the case of narcotics offenses adjudicated in the districts studied here. Despite a uniform federal law, visible differences in sentencing practices surface at the district level. Future research should continue to examine decision-​making across jurisdictions, especially where a single set of laws and/​or guidelines exist. This research demonstrated that the odds of a substantial assistance departure varied enormously across districts and that the demographic factors examined (i.e., race/​ethnicity and gender) also had widely divergent effects across districts. It would be faulty therefore to make general conclusions about sentencing practices across the federal district courts using aggregate analytic strategies. Merely controlling for district-​level influence through multilevel models likewise doesn’t allow researchers to ferret out the many contexts within which extra-​legal disparities might persist. Departure decisions for substantial assistance in the federal narcotics cases examined here were shown to be a context in certain districts where disparity, and possibly discrimination, persists under the federal sentencing guidelines system. As such, those wishing to explore this line of inquiry in the future would be wise to disaggregate data to the district level, as inter-​district differences in sentencing practices can be masked in the aggregate. The time is also ripe for the Department of Justice to make data on all decision-​making stages in the federal criminal justice system publicly available, from initial law enforcement arrest, through prosecutorial screening and charging, and judicial disposition. If we espouse the importance of uniformity under the law, it is necessary to have a system that is not only characterized as uniform and without unwarranted disparity, but also one that is transparent and considered by the populace to be just. This determination cannot be made without access to data which allow for public observation and scholarly research to expand its examination of decision-​making practices at all stages of the federal criminal process from arrest to final outcome. The United States Federal Sentencing Guidelines have made sentencing practices more mechanical, and in the process, less individualized, and less discretionary (Spohn, 2002). Attempts to restrain judicial discretion by promulgating guidelines, however, have not fully removed unwarranted disparity from the system. Legislators were mistaken in their belief that one single and prescriptive sentencing system would result in more uniform and fairer sentences across all districts, especially since they knew that the FSG constrained judicial discretion yet relinquished control of the mechanisms to aggravate or mitigate sentences to prosecutors (Hofer, 2016). Inter-​district differences in sentencing practices have almost certainly hindered the guidelines’ ability to ameliorate unwarranted disparity in the system. From the results of this study, it is clear that when examining uniformity in federal sentencing practices, the trees may be more important than the forest.

Notes 1 This manuscript reports some of the results from a larger study that was funded by the American Statistical Association (ASA) and the Bureau of Justice Statistics (BJS) through the Statistical Methodological Research Program. Special thanks to Dr.  Cassia Spohn, School of Criminology & Criminal Justice, Arizona State University and Dr.  Alan Dabney, Department of Statistics, Texas A&M University who were consultants on the grant. The viewpoints and conclusions herein are strictly those of the author and do not necessarily represent those of the ASA or BJS. 2 These departures can also be granted for up to one year after sentencing if the offender provides the government assistance in another case under Federal rule of Criminal Procedure 35(b); the court would then resentence the offender reducing the original sentence. 3 These are cases for narcotics trafficking offenses, the cases where the charge was simple possession or a charge related to a communication facility were left out of the analyses because of the small percentage of overall cases that they represented.

110

Inter-district Differences 4 For the purposes of this study, the presumptive sentence equals the drug minimum or gun minimum if it was longer than the sentence under the guidelines, and if the safety valve was not applied. The safety valve mechanism allows low-​level and first-​time offenders to avoid the mandatory minimum sentence. If a defendant was facing both a drug minimum and a gun minimum, the presumptive sentence equals the drug minimum plus the gun minimum.The presumptive sentence equals the guideline minimum if: (1) the offender was not facing a mandatory minimum sentence, (2) the offender was facing a mandatory minimum sentence but the mandatory minimum sentence was shorter than the guideline minimum sentence, or (3) the offender was facing a mandatory drug minimum sentence that exceeded the guideline minimum but the safety valve was applied. Whether the offender had a minimum applicable was also included as a separate variable in the analysis. 5 Although Blakely v. Washington only affected cases in Washington State, and not the Federal Sentencing guidelines, several federal district judges meted out sentences that were below the guideline minimum after this case. This variable was included in the analyses to empirically test for differences before and after the Blakely decision. 6 One caveat to this finding, however, is that it is not known whether U.S. citizens, females, whites, and those with more education were better able to provide assistance in the prosecution of other cases. There is some research that argues that differences exist in beliefs about working with law enforcement officials according to one’s class, race, culture and subculture (see for instance, Mirande, 1987; Zatz, 2000). 7 Separate results for each district are available from the author upon request.

References Albonetti, C. (1987). Prosecutorial discretion: The effects of uncertainty. Law and Society Review, 21(2), 291–​313. Alshuler, A.W. (2005). Disparity: The normative and empirical failure of the federal guidelines. Stanford Law Review, 58, 85–​117. Anderson, A.L., & Spohn, C. (2010). Lawlessness in the federal sentencing process: A test for uniformity and consistency in sentencing outcomes. Justice Quarterly, 27(3), 362–​393. Baumer, E. (2013). Reassessing and redirecting research on race and sentencing. Justice Quarterly, 30(2), 231–​261. Blackwell, K., & Baisinger, J. (2016). The use of Federal Rule of Criminal Procedure 35(b). Washington, DC: United States Sentencing Commission. Bushway, S.D., & Forst, B. (2012). Studying discretion in the processes that generate criminal sanctions. Justice Quarterly, 30(2), 199–​222. Bushway, S.D., & Piehl, A.M. (2007). Social science research and the legal threat to presumptive sentencing guidelines. Criminology and Public Policy, 6(3), 461–​482. Carson, A., & Sabol, B. (2012). Prisoners in 2011 (NCJ 239808). Washington, DC: Bureau of Justice Statistics. Cassell, P.G., & Luna, E. (2011). Sense and sensibility in mandatory minimum sentencing. Federal Sentencing Reporter, 23(3), 219–​227. Chiricos,T. (2004). Media, Moral Panics and the Politics of Crime Control. In G. Cole, M.G. Gertz, & A. Bunger (eds.), The criminal justice system: Politics and policies (pp. 41–​61). Belmont, CA: Wadsworth. Cole, K. (1997). The empty idea of sentencing disparity. Northwestern University Law Review, 91, 1336–​1337. Conte, A., & Bowling, B. (2016, April 23). U.S. Attorneys’ priorities differ depending on region. Trib Live. Retrieved from: https://​triblive.com/​usworld/​nation/​9819990-​74/​federal-​cases-​prosecutors Doob, A. (1995). The United States sentencing commission guidelines: If you don’t know where you are going, you might not get there. In C. Clarkson and R. Morgan (eds.), The Politics of Sentencing Reform (pp. 199–​250). Oxford: Clarendon-​Oxford University Press. Hartley, R.D. (2008). Sentencing reforms and the war on drugs: An analysis of sentence outcomes for narcotics offenders adjudicated in U.S. District Courts on the southwest border. Journal of Contemporary Criminal Justice, 24(4), 437–​461. Hartley, R.D., Maddan, S., & Spohn, C. (2007). Prosecutorial discretion: An examination of substantial assistance departures in federal crack-​cocaine and powder-​cocaine cases. Justice Quarterly, 24(3), 383–​407. Hartley, R.D., & Tillyer, R. (2012). Defending the homeland: Judicial sentencing practices for federal immigration offenses. Justice Quarterly, 29(1), 76–​104. Hofer, P.J. (2016). After ten years of advisory guidelines, and thirty years of mandatory minimums, federal sentencing still needs reform. University of Toledo Law Review, 47, 649–​693. Johnson, B.D. (2014). The Missing Link: Examining Prosecutorial Decision-​Making Across Federal District Courts—​ Final Report. Washington, DC: National Institute of Justice. Johnson, B.D. (2006). The multilevel context of criminal sentencing: Integrating judge-​ and county-​ level influences. Criminology, 44(2), 259–​298. Johnson, B.D., Ulmer, J.T., & Kramer, J.H. (2008). The social context of guideline circumvention: The case of federal district courts. Criminology, 46(3), 711–​783.

111

Richard D. Hartley Kautt, P. (2002) Location, location, location: Interdistrict and intercircuit variation in sentencing outcomes for federal drug-​trafficking offenses. Justice Quarterly, 19(4), 633–​671. Kim, B., Spohn, C., & Hedberg, E. (2015). Federal sentencing as a collaborative process: Prosecutors, judges, prosecutor-​judge dyads and disparity in sentencing. Criminology, 53(4), 597–​623. Kutateladze, B., Lynn, V., & Liang, E. (2012). Do Race and Ethnicity Matter in Prosecution? A Review of Empirical Studies. New York:Vera Institute of Justice. Retrieved from: https://​storage.googleapis.com/​vera-​web-​assets/​ downloads/​Publications/​do-​race-​and-​ethnicity-​matter-​in-​prosecution-​a-​review-​of-​empirical-​studies/​ legacy_​downloads/​race-​and-​ethnicity-​in-​prosecution-​first-​edition.pdf Miethe, T. (1987). Charging and plea bargaining practices under determinate sentencing: An investigation of the hydraulic displacement of discretion. Journal of Criminal Law and Criminology, 78(1), 155–​176. Mallik-​Kane, K., Parthasarathy, B., & Adams, W. (2012). Examining Growth in the Federal Prison Population, 1998–​ 2010. Washington, DC: Urban Institute Justice Policy Center. Retrieved from: www.urban.org/​research/​ publication/​examining-​growth-​federal-​prison-​population-​1998–​2010/​view/​full_​report Mauer, M. (2010). The impact of mandatory minimum penalties in federal sentencing. Judicature, 94(1), 6–​8, 40. Maxfield, L.D., & Kramer, J.H. (1998). Substantial Assistance: An EmpiricalYardstick Gauging Equity in Current Federal Policy and Practice. Washington, DC: United States Sentencing Commission. Mirande, A. (1987). Gringo Justice. Notre Dame, IN: University of Notre Dame Press. Mumola, C.J., & Karberg, J.C. (2006). Drug Use and Dependence, State and Federal Prisoners, 2004.Washington, DC: Bureau of Justice Statistics. Nagel, I., & Schulhofer, S.J. (1992). A tale of three cities: An empirical study of charging and bargaining practices under the federal sentencing guidelines. Southern California Law Review, 66, 501–​566. Nardulli, P.F., Eisenstein, J., & Flemming, R.B. (1988). The Tenor of Justice: Criminal Courts and the Guilty Plea Process. Chicago, IL: University of Illinois Press. O’Neill-​Shermer, L., & Johnson, B.D. (2010). Criminal prosecutions: Examining prosecutorial discretion and charging decisions in U.S. federal district courts. Justice Quarterly, 27(3), 394–​430. Roncek, D. (1993).When will they ever learn that first derivatives identify the effects of continuous independent variables or “Officer, you can’t give me a ticket, I wasn’t speeding for an entire hour”. Social Forces, 71(4), 1067–​1078. —​—​. (1991). Using logit coefficients to obtain the effects of independent variables on changes in probabilities. Social Forces, 70(2), 509–​518. Saris, P.B. (2015). A generational shift for federal drug sentences. American Criminal Law Review, 52(1),  1–​24. Schulhofer, S.J., & Nagel, I. (1989). Negotiated pleas under the federal sentencing guidelines: The first fifteen months. American Criminal Law Review, 27(2), 239–​288. Spohn, C. (2006). Sentencing decisions in three U.S. District Courts:Testing the assumption of uniformity in the federal sentencing process. Justice Research & Policy, 7(2),  1–​27. —​—​ . (2002). How Do Judges Decide:The Search for Fairness and Justice in Punishment. Thousand Oaks, CA: Sage. —​—​. (2000). Thirty years of sentencing reform: The quest for a racially neutral sentencing process. In Policies, Processes and Decisions of the Criminal Justice System (Vol. 3, pp. 427–​501). Washington, DC: U.S. Department of Justice. Spohn, C., & Fornango, R. (2009). U.S. Attorneys and substantial assistance departures:Testing for interprosecutor disparity. Criminology, 47(3), 813–​846. Steffensmeier, D., & Demuth, S. (2000). Ethnicity and sentencing outcomes in U.S.  federal courts: Who is punished more harshly? American Sociological Review, 65(5), 705–​729. Stith, K., & Cabranes, J.A. (1998). Fear of Judging: Sentencing Guidelines in the Federal Courts. Chicago: University of Chicago Press. The Sentencing Project. (2005). Racial Disparity in Sentencing: A Review of the Literature. Washington, DC: Author. Tillyer, R., & Hartley, R.D. (2013).The use and impact of fast-​track departures: Exploring prosecutorial and judicial discretion in Federal immigration cases. Crime & Delinquency, 62(12), 1624–​1647. Tillyer, R., Hartley, R.D., & Ward, J.T. (2015). Differential treatment of female defendants: Does criminal history moderate the effect of gender on sentence length in federal narcotics cases? Criminal Justice and Behavior, 42(7), 703–​721. Ulmer, J.T. (2012). Recent developments and new directions in sentencing research. Justice Quarterly, 29(1),  1–​40. —​—​. (2005). The localized use of federal sentencing guidelines in four U.S. District Courts. Symbolic Interaction, 28(2), 255–​279. Ulmer, J.T., & Johnson, B. (2004). Sentencing in context: A multi-​level analysis. Criminology, 42(1), 137–​177. Ulmer, J.T., Light, M., & Kramer, J. (2011). Racial disparity in the wake of the Booker/​Fanfan decision: An alternative analysis to the USSC’s 2010 report. Criminology and Public Policy, 10(4), 1077–​1118. Ulmer, J.T., & Miller, L.L. (2002, November). Plea Agendas in Federal Sentencing: Quantitative and Qualitative Evidence from Four District Courts. Paper presented at the meeting of the American Society of Criminology, Chicago, IL.

112

Inter-district Differences United States Sentencing Commission. (2018a). Quick Facts: Mandatory Minimum Penalties. Washington, DC: Author. —​—​. (2018b). Quick Facts: Drug Trafficking Offenses. Washington, DC: Author. —​—​. (2017). 2017 Sourcebook of Federal Sentencing Statistics. Washington, DC: Author. —​—​. (2015). Overview of Federal Criminal Cases Fiscal Year 2014. Washington, DC: Author. —​—​. (2012). Continuing Impact of United States v. Booker on Federal Sentencing. Washington, DC: Author. —​—​. (2011). 2011 Report to the Congress: Mandatory Minimum Penalties in the Federal Criminal Justice System. Washington, DC: Author. —​—​. (2010). Results of Survey of United States District Judges January 2010 through March 2010. Washington, DC: Author. —​—​. (2009). U.S. Sentencing Commission Post-​Kimbrough/​Gall Data Report. Washington, DC: Author. Ward, J.T., Hartley, R.D., & Tillyer, R. (2016). Unpacking gender and race/​ethnicity biases in the federal sentencing of drug offenders: A casual mediation approach. Journal of Criminal Justice, 46, 196–​206. Weisselberg, C.D., & Dunworth, T. (1993). Inter-​district variation under the guidelines: The trees may be more significant than the forest. Federal Sentencing Reporter, 6(1),  25–​28. Wilmot, K., & Spohn, C. (2004). Prosecutorial discretion and real offense sentencing: An analysis of relevant conduct under the federal sentencing guidelines. Criminal Justice Policy Review, 15(3), 324–​343. Zatz, M. (2000). The convergence of race, ethnicity, gender, and class on court decision making: Looking toward the 21st Century. In Policies, Processes and Decisions of the Criminal Justice System (Vol. 3, pp. 503–​552). Washington, DC: US Department of Justice.

Cases Cited Gall v. United States. (2007). Supreme Court of the United States 552 US 38. Kimbrough v. United States. (2007). Supreme Court of the United States 552 U.S. 85. United States v. Booker. (2005). Supreme Court of the United States 543 US 220.

113

5 MINIMUM SENTENCING FOR SERIOUS OFFENSES Lessons from Australia Kate Fitz-​Gibbon and James Roffee

Introduction There is currently no nationally agreed upon approach to the minimum sentencing of offenders convicted of a serious offense in Australia. While this is to be somewhat expected given that sentencing practices for criminal offenses are implemented at a state rather than national level (Fox & Freiberg, 2014), the last decade has seen a plethora of divergent sentencing reform activity across Australian state and territory jurisdictions. Interestingly, despite the motivation for reform often being connected with penal populist tendencies, the approaches introduced by individual Australian state jurisdictions concerning the minimum sentencing of serious offenses have differed significantly. This divergence affords an important opportunity to examine the merits and impact of minimum sentencing law, practice and its reform across the different Australian jurisdictions. This chapter will provide an overview and critical analysis of recent sentencing law reform activity across Australia with a focus on recently introduced laws for the minimum sentencing of serious offenses, including murder, manslaughter and sexual offenses. In doing so, the chapter will document the different approaches that have been introduced, overview the impacts of sentencing reform in practice, and consider the extent to which reforms introduced have impacted sentencing practice. Enhancements sought through the introduction of recent sentencing legislation have included achieving consistency in approach and outcome, more just outcomes for serious offenses, and improving public confidence in the sentencing process. As will be explored, it remains questionable if the reforms introduced have in each case achieved their intended goals in practice.The chapter details the operation of, and interrogates how well, a minimum sentencing scheme could provide for appropriately proportionate sanctions and investigates the relationship with presumptive sentencing. The evolution of sentencing practice for serious offenses across Australia provides a worthy reference point and, in this case, a cautionary tale, for those seeking to reform sentencing laws and practice in other jurisdictions.

The Turn Towards Minimum Sentence Schemes for Serious Offenses in Australia Australia has one federal, six state and two territory criminal justice systems, of which each has their own sentencing legislation. At the state and territory level, each of the eight jurisdictions adopt 114

Minimum Sentencing for Serious Offences

their own approach to sentencing whereby sentencing is governed by a combination of legislation made by Parliament and common law (Fox & Freiberg, 2014). As such, for all offenses not governed by federal law, the relevant state or territory parliament creates the legislation, which defines the details of how the offense is committed along with the maximum penalty that can be imposed on a person after they are convicted of committing the specified offense. Courts, while independent of Parliament, must therefore interpret and apply the laws within the framework created by Parliament. In New South Wales (NSW) and Victoria, the two Australian state jurisdictions under focus here, sentencing has traditionally been governed by a process undertaken by judges with little guidance from any institution outside of the judiciary, known as the “sentencing synthesis” (Markarian v. The Queen (2005) 228 CLR 357, as per McHugh J, at 378) or the “instinctive synthesis” (Nguyen v. The Queen [2010] VSCA 284, as per Ashley JA, at 21). The sentencing synthesis requires the court to take into consideration all relevant factors of the case before reaching a conclusion concerning the appropriate sentence. The applicability of this approach, though still widely used for many summary and non-​violent offenses, has been curtailed in recent decades through the introduction of presumptive minimum sentencing schemes in both jurisdictions. In contrast to the discretionary sentencing models favored in comparable jurisdictions, presumptive minimum sentencing schemes impose reference points for members of the judiciary to take into account when determining the minimum term of imprisonment in cases involving serious offenses. Whilst presumptive minimum sentencing schemes vary in form from jurisdiction to jurisdiction, Hoel and Gelb (2008, p. 2) describe them as entailing an approach to sentencing “in which Parliament prescribes both a sanction type and a minimum level of severity for a given offense which the court must impose unless there is a demonstrable reason—​which may be broadly or narrowly defined.” This approach to sentencing, predominately favored for serious offenses, has been implemented by several comparable Western jurisdictions over the past ten years and is explored here with reference to the 2003 introduction of standard non-​parole periods in NSW and the 2014 introduction of a baseline sentencing scheme in Victoria. Both schemes were novel in their creation in that they have not previously been introduced in any Australian jurisdiction, and for the purposes of this chapter both are important as they provide an opportunity for a comparative analysis of the merits, impacts and failures of such schemes in practice. This chapter is structured into three main parts.The chapter first examines the 2003 introduction, operation of, and the High Court of Australia’s ruling on, the standard non-​parole periods scheme (SNPPs) in New South Wales. Following, the chapter sets out a detailed examination of another approach to minimum sentencing, that of the failed baseline sentencing scheme as introduced in Victoria in 2014. The analysis first details the law and order climate that gave rise to the introduction and drafting of the scheme, its short operation and the impact of its repeal on future sentencing practice for minimum sentences in the state of Victoria. In the final part we consider what lessons can be gleaned from the NSW and Victorian experiences and provide a cautionary note on the introduction of presumptive (and mandatory) minimum sentencing schemes in any context. In doing so, we seek to build on the body of criminological and legal scholarship that has previously debated the merits of minimum sentencing schemes (see, inter alia, Ashworth & Player, 2005; Fitz-​Gibbon, 2016; Hutton & Tata, 2010; Mitchell, 2013).

Standard Non-​Parole Periods in New South Wales In 2003 NSW became one of the first Australian jurisdictions to introduce a presumptive minimum sentencing scheme targeted at the sentencing of serious and violent offenders across a range of offenses. The scheme itself had not been previously introduced elsewhere, and as such its drafting and subsequent application in NSW represented a first. In examining the impact of this approach in practice, the following parts examine the introduction and operation of the standard non-​parole sentencing scheme (SNPP) for serious offenses in NSW. In so doing, we first provide an overview of 115

Kate Fitz-Gibbon and James Roffee

the introduction of the scheme and initial responses from the legal community followed by an examination of its impact and operation to date, including an intervention (and arguably a correction) by the High Court of Australia altering its use in practice.

The Introduction of the Standard Non-​Parole Period Sentencing Scheme While NSW retains what could be classified as a discretionary sentencing system in comparison to many other Western jurisdictions, the 2003 implementation of SNPPs for serious offenses arguably threatened this approach. Devised by the Labor Government during 2002, before the election of March 2003, the SNPP scheme entered into force on February 1, 2003 through the Crimes (Sentencing Procedure) Amendment (Standard Minimum Sentencing) Act 2002 (NSW).1 Through Part 4, Div 1A of the Crimes (Sentencing Procedures Act) 1999 (NSW), specific non-​parole periods are recommended for a variety of serious offenses. These periods apply to the sentencing of all offenders over the age of 18 who are convicted following trial of an offense committed after February 1, 2003 (Sentencing Council, 2011). In cases where an offender pleads guilty prior to trial, the scheme allows for the SNPP to be used as a “guidepost” by the sentencing judge (Sentencing Council, 2011, p. 10). SNPPs are currently prescribed for a total of 35 serious offenses including murder, serious drug offenses, sexual offenses, offenses against the person, and property offenses. Examples of the SNPP prescribed for such offenses are set out in the table below. As shown in the table below, for the offense of murder, for example, the legislation requires that judges adopt a SNPP of 20 years except in cases where the victim is under 18 years old or is a specific member of the community2 in which case the standard non-​parole is increased to 25 years. In all other cases, the SNPP must be implemented except in circumstances where a sentencing judge can justify setting a minimum term that is shorter or longer than the SNPP by reference to the established mitigating and aggravating factors included within section 21A of the Crimes (Sentencing Procedures Act) 1999 (NSW). Table 5.1  Standard non-​parole periods prescribed for some serious offenses in NSW1 Item no (as prescribed in the Act)

Offense

SNPP

1B 1 2 4B 5 7 9 9A 10 11 16

Murder—​where the victim was a child under 18 years of age Murder—​in other cases Conspiracy to murder Reckless causing of grievous bodily harm Assault of police officer occasioning bodily harm Sexual assault Aggravated sexual assault in company Aggravated indecent assault Sexual intercourse with a child under 10 Robbery with arms etc. and wounding Manufacture or production of commercial quantity of prohibited drug, being an offense that: (a)  does not relate to cannabis leaf; and (b) if a large commercial quantity is specified for the prohibited drug concerned under that Act, involves less than the large commercial quantity of that prohibited drug Unauthorised sale of prohibited firearm or pistol

25 years 20 years 10 years 4 years 3 years 7 years 15 years 5 years 15 years 7 years 10 years

21

10 years

Note: This is a selective, non-​exhaustive list of the SNPPs prescribed. For the full list of SNPPs prescribed see Part 4 Division 1A Crimes (Sentencing Procedure) Act 1999 (NSW).

116

Minimum Sentencing for Serious Offences

Initial Responses and the Operation of the Scheme in NSW At the time of its introduction the SNPP scheme was criticized as being arbitrary, unnecessarily complicated, and punitive (Johnson, 2003; Warner, 2003). Loukas (2002: p.  54) suggested that the guideline judgment system, supplemented by legislation detailing, for example, relevant factors to be taken into consideration on sentencing should be preferred over the SNPP system that “must unavoidably represent a concession to the most insupportable misconceptions as to the reality of the justice system.” Less than ten years after its 2003 introduction, in 2010 a study conducted by the Judicial Commission of NSW revealed that the scheme had led to an increase in the length of sentences imposed for offenses covered under the legislation. The Judicial Commission found that since February 2003 the standard non-​parole period scheme has led to an increase in the severity of penalties imposed and the duration of sentences of full-​time imprisonment (for discussion, see also Jensen, 2010). Specifically, the study observed that sentences for murder, particularly in cases where an offender pled not guilty and was convicted following a contested trial, had “significantly increased since the commencement of the statutory scheme” (Judicial Commission of NSW, 2010, p.  25). Beyond increasing sentences imposed for the relevant offenses, the Judicial Commission observed that a clear impact of the SNPP scheme had been to increase consistency in sentences imposed in both the NSW District and Supreme Courts (Judicial Commission of NSW, 2010). However, and as later noted by the NSW Sentencing Council (2011, p. 19), it is unclear whether this represents a consistency resulting from “like cases being treated consistently” or “dissimilar cases being treated uniformly in order to comply with the scheme.” This consequential increase in sentences-​imposed post-​February 2003 has also been recognized by members of the NSWSC judiciary, with NSW Supreme Court Justice Howie, in Regina v. LR ([2010] NSWSC 22 (at 30)), noting that “there is little doubt that since the introduction of the standard non-​parole period sentencing for murder have increased.” The increase in length of sentences by way of setting prescribed terms in legislation represents a key impact of the scheme, which may have the consequential effect of failing to allow for proportionality in sentencing. This is particularly concerning given research has found that members of the public are more concerned with the notion of proportionality in sentencing, than the need for deterrence and denunciation (see Roberts, 2003). As explained by Roberts (2003, p. 504), “By limiting a court’s ability to impose a proportional sentence, mandatory minima can violate the principle of proportionality, and this is likely to undermine, rather than enhance, public confidence in the courts.” As such, while intended to deliver on a promise to implement tougher penalties, these policies can actually serve to undermine the operation of sentencing for serious offenses. As explained by Palmer (2005, p. 29), a “tough on crime” approach to punishment leads to the implementation of sentences that are “proportional to the anger and resentment the public feels toward criminals, rather than proportional to the crime.” This increase in the length of sentences imposed is also particularly concerning given that past research has often identified that the implementation of restrictive sentencing practices, such as presumptive minimum sentencing as well as mandatory sentencing schemes, rarely leads to a reduction in offending (see, inter alia, Brown, 2001; Hoel & Gelb, 2008;Tonry, 2009), raising the question of what benefits such policies actually have in practice. Another intended effect of the scheme may have been to force defendants to plead guilty in the hope of receiving a sentencing discount, with research indicating the number pleading guilty increased from 78.2% to 86.1% (see Poletti & Donnelly, 2010). Following on from the work of the Judicial Commission, in September 2011 the NSW Attorney General requested that the NSW Law Reform Commission (NSWLRC) conduct a review of the Crimes (Sentencing Procedures Act) 1999 (NSW), with a specific focus in the terms of reference on the operation of the SNPP scheme.The review sought to build upon a background review of SNPPs and guideline judgments already conducted by the NSW Sentencing Council (2011). Initial stages of the review were published in a Preliminary Outline of the Review (NSWLRC, 2011), and an Interim Report 117

Kate Fitz-Gibbon and James Roffee

on Standard Minimum Non-​Parole Periods (NSWLRC, 2012). Specifically, the Interim Report detailed several options for reform including the retention of the SNPP scheme with various amendments as well as a complete repeal of the scheme.3 Questions as to the workability of the legislation, including its drafting, have also emerged in case law. Specifically, the legislation’s failure to define or provide a sufficient guide as to what constitutes a “mid-​range” of objective seriousness offense has been noted in case law (see, for example, R v. Way [2004] NSWCCA 131). In 2006 the NSW Court of Criminal Appeal provided a description of the process for determining whether a case should be considered a mid-​range offense (Vu v. R [2006] NSWCCA 188, at para. 30–​31). However, even with this clarification the NSW Sentencing Council (2011, p. 16) has since noted that the determination of a mid-​range offense could still “give rise to error,” highlighting the continuing problematic nature of this complicated legislation.

The High Court of Australia’s Intervention In 2011, the High Court of Australia’s decision in Muldrock sought to alleviate some of the concerns raised in the initial operation of the SNPP Scheme (Muldrock v. The Queen [2011] HCA 39).4 The High Court held in Muldrock that the approach to sentencing as outlined in the 2004 decision of R v. Way was wrong.The Muldrock case involved a “mentally retarded sex offender” who was sentenced to 96 days in the District Court of NSW following a guilty plea to the offense of sexual intercourse with a child under ten years old (see further case summary statement, High Court of Australia, 2011). On appeal, the offender’s sentence was increased, with reference given to the need to apply the SNPP, to a term of six years and eight months. Subsequently, Muldrock appealed to the High Court of Appeal by special leave. In deciding the Muldrock appeal, the High Court ruled that the assessment of standard non-​ parole periods should be made in the context of the sentencing methodology previously set out in Markarian v. The Queen ([2005] HCA 25, hereinafter Markarian), and most importantly, ruled that “it was an error” of previous case law to frame the scheme “in mandatory terms” (Muldrock, at 25; this referred to earlier decisions of the NSW courts in R v. Way [2004] NSWCCA 131; (2004) 60 NSWLR 183). As a result of this framing, the High Court found that the NSW Court of Appeal had erred in giving “determinative significance” to the legislation (Muldrock, at 32). The judgment also clarified that the application of the SNPP scheme should not be limited to cases that fall only within the middle of the range of objective seriousness and thus the SNPP has a role in sentencing for an offense found in the high or low range of objective seriousness. In Markarian it was held that: the judge identifies all the factors that are relevant to the sentence, discusses their significance and then makes a value judgment as to what is the appropriate sentence given all the factors of the case. Only at the end of the process does the judge determine the sentence. (Markarian v. The Queen [2005] HCA 25, at 51) The Muldrock judgment also somewhat addressed ambiguity in the definitions of the legislation in that it sought to define how the objective seriousness of an offense should be assessed. The judgment set out that: The objective seriousness of an offence is to be assessed without reference to matters personal to a particular offender or class of offenders. It is to be determined wholly by reference to the nature of the offending. (Muldrock, at 27) However, as noted by the NSWLRC (2012, at 22) in the following year, “this [judgment] does not resolve all the difficulties concerning the factors that must be taken into account when considering 118

Minimum Sentencing for Serious Offences

objective seriousness.” Despite this, the judgment did have significant impact in that it gave rise to a series of legislated sentencing amendments within NSW, introduced as a first step in bringing the law in line with the decision of the High Court. Specifically, following the decision in Muldrock, the Crimes (Sentencing Procedure) Amendment (Standard Non-​parole Periods) Act 2013 (NSW) was introduced. The Amendment Act, amongst other reforms (see further Donnelly 2013), sought to clarify the process through which a sentencing judge is required to apply a standard non-​parole period in an individual case. Donnelly (2013) indicated that the “most notable” change in the Amendment Act relates to the repeal of s 54B(2) of the previous Act: 54B(2) When determining the sentence for the offence (not being an aggregate sentence), the court is to set the standard non-​parole period as the non-​parole period for the offence unless the court determines that there are reasons for setting a non-​parole period that is longer or shorter than the standard non-​parole period. (emphasis added by Donnelly, 2013) By replacing this section of the Act, the Amendments bring to the fore the ruling of the High Court that the SNPPs should never have been legislative form using language that implied a mandatory application of the SNPP. The amended s 54B(2) provides: The standard non-​parole period for an offence is a matter to be taken into account by a court in determining the appropriate sentence for an offender, without limiting the matters that are otherwise required or permitted to be taken into account in determining the appropriate sentence for an offender. The approach as it now stands in legislation and as it has been applied subsequently in the NSW courts represents an attempt, some ten years after the SNPPs were first introduced, to clarify the appropriate meaning and application of the scheme in a way that brings together Parliament’s intent, the prescribed minimum terms of imprisonment and case law.

Baseline Sentencing in Victoria Proffered to ensure that sentencing practices in Victoria met the perceived expectations of the Victorian community, the 2014 introduction of the baseline sentencing system through the Sentencing Amendment (Baseline Sentences) Act 2014 (Vic) was a highly controversial change to the state’s sentencing regime. At the time of its introduction, it represented one of a number of recent changes to the Victorian sentencing landscape, including the abolition of suspended sentences and home detention. The baseline sentencing scheme, while sharing some similarities with the NSW SNPPs, represents another interesting point of analysis in understanding divergent approaches taken to the minimum sentencing of serious and violent offenders in Australia.The following parts will review the origins of baseline sentencing, before placing the scheme within the wider Victorian sentencing landscape. We explain what baseline was and how it was intended to operate.We then review many of the concerns that were raised in its initial operation, before detailing its nullification and eventual replacement with a new standard sentencing scheme.

The Origins of Baseline Sentencing The baseline sentencing scheme was devised hastily before the 2010 Victorian State election as one of a range of measures to combat what was presented by the opposition party as a spiraling law and order crisis. Specifically, the scheme was announced on November 23, 2010, as an election pledge by the Honourable Ted Baillieu MP, leader of the then-​opposition party the Liberal-​National Coalition. 119

Kate Fitz-Gibbon and James Roffee

The election, held only four days later on November 27, 2010, saw the Coalition elected, and thus placed an arguably ill-​thought out election pledge as a central tenet of a new-​government’s law and order regime. The baseline scheme was, in essence, a rather blunt attempt to force the judiciary to increase sentences for a range of offenses. While the scheme remained on the statute book for three years, its use was quickly nullified by the Courts, in the first sentencing appeals concerning the new scheme, where it was declared “incapable” of application by judicial members of the Court of Appeal. The offenses for which the baseline scheme was determined to apply, detailed below, are of a type and range to cause significant public concern and contain serious violent and sexual offenses, including against children, as well as commercial drug trafficking. The criticism of the sentencing regime incumbent at the time of the election not only come from politicians, but also from a number of individual Judges of Appeal, who had commented on the adequacy of the then-​sentencing practices (see inter alia Director of Public Prosecutions v. CPD (2009) 22 VR 533; Leeder v. The Queen [2010] VSCA 98; Nguyen v. The Queen (2010) 208 A Crim R 464). In its 2016 report on Sentencing Guidance in Victoria, the Victorian Sentencing Advisory Council (SAC) identified a number of cases prior to the 2010 election where criticism of the adequacy of current sentencing practice had been made (SAC, 2016). A number of the cases cited within the SAC report share a common member of the judiciary, and it thus remains open to question the extent to which the cited “judicial concern with the adequacy of sentencing practices” was widespread or rather concentrated among individual members of the judiciary who happened to be sitting on high-​profile cases. Nonetheless, the concern with inadequate sentencing, alongside a number of high-​profile incidents of serious crime, created the perfect law and order climate within Victoria for the introduction of the baseline scheme.

The Victorian Sentencing Advisory Council As of 2004, Victoria has a Sentencing Advisory Council, which was established by the Victorian Parliament as an independent statutory body. Given the significant advisory role that the Council has played in the formation of approaches to minimum sentencing in the state it is worthy of a short diversion here to explain the introduction and scope of its work. The Council’s establishment followed a review into Victorian sentencing laws resulting from earlier concerns raised over the need for reform in the sentencing process more generally. The review was commissioned by the Victorian Government and led to the Pathways to Justice: Sentencing Review 2002 report, which recommended a number of improvements to the sentencing system including the establishment of a Sentencing Advisory Council and the introduction of guideline judgments (Freiberg, 2002). The Pathways to Justice report stated: this final report is written in the context of community and media calls to increase the use and severity of imprisonment in Victoria. The Review has not found evidence to support a significant shift away from Victoria’s long-​standing use of imprisonment only as a last resort. (Freiberg, 2002, p. 1) The Sentencing Advisory Council created after the review has similar functions to the councils found in England, Scotland, and other Australian jurisdictions including New South Wales and Tasmania (for more on Sentencing Councils generally, see Freiberg & Gelb, 2008). Like the statutory bodies found in these other jurisdictions, one of the aims of the Victorian SAC is to bring together community, courts, and government in a process to inform, educate, and advise on issues relevant to sentencing and the impact of sentencing. The legislation governing the SAC details its functions:

120

Minimum Sentencing for Serious Offences

• provide statistical information on sentencing, including information on current sentencing practices to members of the judiciary and other interested persons. • conduct research and disseminate information to members of the judiciary and other interested persons on sentencing matters. • gauge public opinion on sentencing. • consult on sentencing matters with government departments and other interested persons and bodies as well as the general public. • advise the Attorney General on sentencing issues. • provide the Court of Appeal with the Council’s written views on the giving, or review, of a guideline judgment. (Sentencing Act 1991 s.108C)

The Creation of a Baseline Sentencing Scheme In April 2011, the Attorney General wrote to the SAC asking it to advise him on the introduction of a baseline sentencing scheme.The Council was not required to advise on the merits of the scheme in and of itself, but rather to provide information regarding: a. the levels of baseline sentence to be legislated for offenses to which the baseline sentencing regime will apply. b. whether offenses additional to those committed to by the government should be included, either in the initial introduction of baseline sentences or subsequently. c. whether the legislation should provide for different levels of baseline sentence for a particular offense to be applicable in different circumstances, or for certain offenses to be redefined to achieve better applicability of baseline sentences in different circumstances. d. how baseline sentences should best apply in cases involving multiple offenses. e. the amendments required to relevant legislation to introduce baseline sentencing. f. the likely effects of recommendations and options put forward by the Council on sentencing levels for the relevant offenses and on the numbers of persons serving custodial and non-​custodial sentences; and g. any other matters the Council considers relevant. (SAC, 2012) The government made clear, when tasking the SAC, that they wanted baseline sentences to apply to serious offenses as defined in the Sentencing Act 1991 (Vic) and to additional offenses such as arson, recklessly causing serious injury, aggravated burglary and major drug trafficking (SAC, 2012, para 1.4). In seeking the Sentencing Advisory Council’s advice, the then-​Attorney General stated that baseline sentences were to provide the starting point for the court in determining the minimum sentence (the non-​parole period) to be imposed where a baseline sentence applies, and the baseline sentence is to act as an indicator of the sentence that Parliament expects to be the median or midpoint of minimum sentences imposed for cases involving that offense (SAC, 2012, para 1.4). The scheme, while unique, was noted as being similar to the NSW SNPPs (as discussed above). In introducing the scheme, the government intended that sentencing practices in Victoria for the relevant offenses would adjust over a period of time (upwards) so that half of the sentences sit above and half sit below the baseline sentence. Importantly, the Bill that introduced the baseline sentencing scheme departed from the advice given by the SAC and applied to the entire sentence rather than just the non-​ parole period. In their 2012 Report, the SAC suggested that they “[did] not consider that the terms of reference require a court to apply the Baseline Sentencing Scheme in every case involving a baseline offence” (SAC, 2012, p. 21). However, the Act was not drafted in line with the SAC’s recommendations. Instead the baseline scheme was to apply to all sentences given for the specified offenses. This is a significant difference in terms of the later application and deemed workability of the scheme.

121

Kate Fitz-Gibbon and James Roffee

In 2014, the SAC produced a report—​Calculating the Baseline Offence Median Report—​to assist the courts in understanding how the baseline system works. The median is explained to be the “numerical value separating the lower half of a distribution form the higher half ” (SAC, 2014, p. 13). As such, it is the value that is in the middle of a set of values ordered sequentially. The median sentence is calculated taking into account both the custodial and non-​custodial sentences. The SAC noted that the Act is silent about the duration over which the baseline median is calculated, as such in its 2014 Report it selected a five-​year period for analysis (SAC, 2014). During the passage of the Act the then-​Attorney General, Robert Clark, indicated a purpose of the Act was to increase the length of sentences in order to meet community expectations (Vic Parliament Sentencing Amendment (Baseline Sentences) Bill 2014, April 3, 2014, 1275). The Sentencing Guidance Report indicated that the baseline reforms were intended to serve as a guidepost for judges. As such, Parliament had the expressed intention that judges use the baseline to increase sentences upwards over time to ensure that the proscribed sentence became the median. In line with this, the Bill set the baseline sentences at a level above the existing median sentences in order to ensure that over time those offenses that would otherwise have been given a sentence at the current median, would instead be given the legislated new upwards trending baseline. At the time of its introduction, Professor Arie Freiberg, Chair of the SAC, called the changes introducing baseline a “major change in sentencing methodology” (Spooner & Bucci, 2014). The Sentencing Amendment (Baseline Sentences) Act 2014 (Vic) received Royal Assent on 12 August 2014. As incorporated into the Sentencing Act 1991 the Scheme stipulates that when sentencing an offender, the court must therefore consider the prescribed baseline sentence for the offense and in so doing must sentence half the offenses above and half below the baseline. In practice, Freiberg described this new approach as a two-​step process, whereby judges would be required to start with the baseline prescribed for the offense and then go up or down depending on the aggravating or mitigating circumstances of the individual case (Malik, 2012). In deciding upon a sentence, the Court is also required to state the reasons why it is imposing the sentence in relation to the baseline. The baseline sentences stipulated in the Act, as compared to the median sentence imposed prior to the Act, were as set out in Table 5.2 (below). The Act also stipulated a baseline sentence of 30 years for the murder of an emergency worker while on duty under the Crimes Act 1958 (Vic) s3(2)(a). The Act provides that baseline sentencing does not apply to cases that are determined summarily (a summary offense is an offense that can be heard by a magistrate sitting alone, rather than a judge and jury), nor was the Scheme applied to offenses committed by a person who was under 18 at the time of the commission of the offense.

Table 5.2 The difference between baseline medians prior and the median established under the Act Offense

2008–​09 to 2012–​13 baseline median

Baseline sentence under the Act

Difference between 2008–​09 to 2012–​13 baseline median and baseline sentence

Culpable driving causing death Incest Persistent sexual abuse of a child under 16 Sexual penetration of a child under 12 Trafficking in a large commercial quantity of a drug of dependence Murder

5 years and 6 months 4 years 6 years

9 years 10 years 10 years

3 years and 6 months 6 years 4 years

3 years

10 years

7 years

6 years and 6 months

14 years

7 years and 6 months

20 years

25 years

5 years

122

Minimum Sentencing for Serious Offences

How Was It Applied in Practice? The Act commenced on November 2, 2014 and applied to all offenses committed on or after that date. Providing an early signal of the disquiet that would follow, at the time of its introduction, an editorial in the widely respected state newspaper The Age ran with the headline “Baseline sentencing is rank populism,” mounting an argument that the baseline system was fundamentally flawed and reduced the sentencing exercise to maths (Editorial, 2014). The scheme’s introduction was similarly met with significant criticism from those within the legal profession and the judiciary.5 The Victorian Bar, for example, issued a media release indicating its concern with the scheme reducing the independence of the Courts (Bar, 2014), while the Law Institute of Victoria (LIV), the professional association of solicitors, issued an even more strongly worded critique of the proposed system (LIV, 2014). The LIV indicated that it was disturbed by the decision to introduce a baseline scheme that would unduly limit judicial discretion, overcomplicate the sentencing process and fail to act as a deterrent. The LIV was concerned that the scheme would increase pressure on the criminal justice system with the resulting increase in duration of prison sentences. The LIV (2014) saw the scheme as one of a “raft of similarly punitive measures … that do little to reduce crime, rehabilitate offenders or benefit the community.” Among the judiciary, County Court Chief Judge Michael Rozen warned in early 2015 after the scheme had been introduced, that it would lengthen trials and the number of appeals as concern was raised over the “substantial complexity” that the scheme will bring to the sentencing process (Lee, 2015). The then-​head of the County Court criminal jurisdiction, Judge Mark Taft, also indicated that the median acted as a constraint and raised similar concerns over the additional delays, cost and heightened prospects of appeal that would arise as a result of the scheme’s operation (Lee, 2015). In perhaps the strongest critique of the Scheme, in May 2015 Chief Crown Prosecutor Gavin Silbert indicated that the scheme was causing effects contrary to what the Attorney General and government had anticipated, and suggested that prosecutors would be accepting plea bargains on lesser charges to stop the delays being caused by sentencing under the baseline scheme which would take longer to dispose of and thus clog the courts (White, 2015). The strongly worded criticism from the Chief Crown Prosecutor suggested the legislation was the “worst drafting he has ever seen” written by “someone [lawyers] on some substance [drugs]” (White, 2015). Writing somewhat in defense of the scheme and in contrast to the mounting view of the Victorian legal community, then-​Attorney General Robert Clark made clear in 2014 that he believed that the Courts should be listening to Parliament as a body elected by the people (Clark, 2014).The Attorney General intimated that he saw the role of judges as being required to follow Parliament’s directions as to the sentencing levels, with the suggestion being that judges were too lenient (Clark, 2014). Thus, the introduction of baseline became part of a larger struggle in the battle of supremacy between the judiciary and the then-​executive acting in Parliament. On the background of such debate significant interest arose in the first case sentenced in line with the scheme. R v. IRT ([2015] VSC 372 (30 July 2015)) involved two charges of indecent act with a child under 16 (s.47(1) Crimes Act 1958 (Vic)) and four counts of incest with a child/​step-​child under (s.44(1) Crimes Act 1958 (Vic)). The court found in IRT that the offending was a level of seriousness below that which would have been found for the median sentence. This meant that as per standard sentencing practice, the sentencing range below the median remained open to the judge to apply as seen fit. Importantly, and fatally for the baseline scheme, the court found that it was therefore not required to scale offenses. It held: It must be accepted as a matter of mathematics, that so long as half of all sentences are at or above, and half of all sentences are at or below a particular figure, the distribution of sentences above and below that point do not affect the median. Therefore, sentencing in a 123

Kate Fitz-Gibbon and James Roffee

manner compatible with Parliament’s intention does not require that sentences below the median are scaled in proportion to the baseline sentence. (R v.  IRT [2015] VSC 372, at para 141) Less than six months following the decision in IRT, the Victorian Director of Public Prosecution appealed against the sentence imposed to the Victorian Court of Appeal on the basis that the total sentence and non-​parole period were “manifestly inadequate.” The November 2015 decision of the Court of Appeal in DPP v. Walters ((a pseudonym) ((2015) 49 VR 356)) became critical in sealing the fate of the Victorian short-​lived baseline sentencing scheme. In Walters, the Victorian Court of Appeal ruled that the Baseline Sentencing Scheme in Victoria was inoperative. Specifically, the Court held that amendments to the Sentencing Act 1991 to introduce the baseline scheme had insufficient detail as to how a judge was to sentence in a manner compatible with the Act in order to achieve the baseline becoming the median sentence. Further it held: the defect in the legislation is incurable. Parliament did not provide any mechanism for the achievement of the intended future median, and the Court has not authority to create one … [and] To do so would be to legislate, not to interpret. (Walters, at para. 8) As such, the court ruled that “baseline sentencing provisions are incapable of being given any practical operation” (at para. 9). The concern was the difficulty inherent in determining what constitutes the distinctive attributes or features of a median sentence, such that it can be compared with an historical case or a future hypothetical case in order to move the sentencing upwards. The Court explained: there remains a lack of precision arising from the lack of any specified time in the future, or specified time period, by which the intended median is to be achieved. Crucially, however, the legislation is wholly silent as to the means by which a judge imposing sentence for incest is to do so “in a manner compatible with” the intention to achieve the intended median sentence in the future. (Walters, at para. 6) This decision represented the courts’ nullification of the Baseline Sentencing Scheme in Victoria.

Sentencing in Victoria after the Baseline Scheme Following a change in state government, the new Attorney General, following the decision in DPP v. Walters (a pseudonym), asked the Victorian SAC on November 24, 2015 for advice on legislative mechanisms for sentencing guidance in Victoria. In providing guidance to the Attorney General, the SAC in its 2016 Report recommended the baseline sentencing provisions be repealed in their entirety (SAC, 2016, p. xvii). The government responded and alongside the repeal in the Sentencing Amendment (Sentencing Standards) Act 2017 the government introduced a standard sentence scheme. The standard sentence scheme commenced on February 1, 2018 for all offenses committed on or after this date. Unlike the baseline scheme, it is legislated that the standard sentence is not intended to affect the instinctive synthesis approach to sentencing as per section 5B(3). While the impact of the newly prescribed standard sentencing scheme in Victoria is yet to emerge in case law, there are some important differences to note between the heavily critiqued baseline scheme and the newly formulated standard sentencing scheme. While a standard sentence is prescribed  for  12  named offenses (including all those that were within the baseline scheme), the approach by which members of the judiciary are required to apply the scheme differs. The standard sentence for an offense is the sentence that falls in the middle of the range of seriousness for that specified offense, taking into account only objective factors. The relevant objective factors are determined solely by reference to the nature of the offending, with no reference being made to a 124

Minimum Sentencing for Serious Offences

particular offender or class of offender (Sentencing Act 1991 ss5A(1)(b), 5A(3)). Factors include current sentencing practices, any aggravating or mitigating factors and the maximum penalty for the offense. The new standard sentences are 40% of the maximum for the specified offenses except two named offenses; murder and trafficking a large commercial quantity of drugs. This new scheme is designed to be a legislative yardstick for judges, acting in conjunction with the maximum penalty found in legislation. In addition, the Sentencing Amendment (Sentencing Standards) Act 2017 introduced non-​parole periods for the offenses subject to the standard sentence scheme. The non-​parole periods are set at a defined percentage of the “relevant term.” The relevant term, for a single standard sentence offense is the sentence imposed for that offense, and where there are two or more offenses including at least one single standard sentence offense, the relevant term is the total effective sentence.The non-​parole periods are set, unless it is in the interests of justice not to do so, at: • 30 years, if the “relevant term” is life imprisonment • 70% of the “relevant term”, if that term is for 20 years or more • 60% of the “relevant term”, if that term is less than 20 years. (Sentencing Act 1991 s11A(4)) The scheme has been devised to avoid some the pitfalls of the previous baseline scheme. The legislation provides that when sentencing for a standard sentence offense the court may only consider sentences previously imposed for the relevant offense where the offense was subject to the standard sentence scheme (s5B(2)(b)). Courts therefore cannot have regard to the existing sentencing practices for offenses that were committed before the standard sentence provisions entered into force. Thus, the standard sentence scheme provides a level of direction to judges that signals the intention of Parliament to increase sentences in line with public expectations for the sentencing of serious offenses. However, and importantly, it leaves the judiciary with sufficient discretion. Even the introduction of the non-​parole period is contingent on the judiciary’s discretion that it is just to apply to the case. Further, judges are required to state reasons for imposing the sentence, and if they wish to set a non-​parole period shorter than the fixed duration in s.11A(4) they must state the reasons for doing this, as set out in s.5B(4). The new legislation therefore guides judges to achieve a standard sentence for middle of the range of seriousness offenses, while still allowing the necessary flexibility to ensure sentencing is done in the interests of justice.

Where to Next for Minimum Sentences in Australian Jurisdictions? While the impact of the new standard sentencing scheme in Victoria remains to be seen in practice, and the NSW scheme of standard non-​parole periods continues albeit in a form heavily restricted in comparison to its application prior to the intervention of the High Court of Australia, these two approaches to minimum sentencing provide valuable lessons for other Australian and comparable jurisdictions. There sits an uncomfortable similarity between the attempts to enact such legislation in both states, that being the proximity of the decision to pursue presumptive minimum sentences to impending state parliamentary elections. In both instances the legislation was criticized for being hastily devised, unnecessarily complex, and resulting in excessive costs to the judicial system. Notably, in both cases, the respective Parliament’s attempts to direct to the judiciary the terms at which minimum non-​parole periods for serious offenses should be set—​driven either by a pre-​election law and order punitive maneuver or, as is often the expressly stated case, the commendable intention of seeking consistency in sentencing practice and the alignment of community expectations with the delivery of justice, were heavily resisted. In both NSW and Victoria, the legal community, including legal practitioners and members of the judiciary, expressly and quite openly stated their disbelief in such an approach to sentencing—​be it presumptive or mandatory—​acting to further curtail judicial discretion. The primacy of judicial discretion has been noted at length elsewhere, and while not seeking to rehash that well-​trodden ground (see, inter alia, Fitz-​Gibbon, 2013; Hoel & Gelb, 2008; 125

Kate Fitz-Gibbon and James Roffee

Tonry, 1992) we believe that these recent Australian experiences further evidence the push back that curtailment of judicial discretion is receiving across Western jurisdictions. This analysis demonstrates that debates surrounding the restriction and fettering of judicial discretion not only occur in relation to maximum penalties imposed—​including mandatory terms of life imprisonment and other extreme sentences—​but that the experiment towards presumptive minimum terms gives rise to similar concerns. In both instances, the introduction of these sentencing reforms was designed to increase the terms of imprisonment imposed upon offenders and to deplete the degree to which the circumstances of the offender could be taken into account in sentencing. As evidence continues to mount as to the complex needs and disadvantages often faced by those who come into conflict with the law, particularly in the context of Aboriginal and Torres Strait Islander communities in Australia (Australian Law Reform Commission, 2017), any approach to sentencing that impedes the ability to achieve individualized justice, including allowing judges sufficient latitude to impose a just sentence, should be heavily scrutinized and avoided wherever possible. Finally, in NSW and Victoria the ambiguities following the introduction of the respective minimum sentencing schemes reveal uniform concerns surrounding the unintended consequences of reform that further complicate the sentencing exercise for serious offenses. Similar concerns have been noted in England and Wales, where the 2003 introduction of Schedule 21 and within it the minimum starting points for serious offenses (Fitz-​Gibbon, 2016, p. 16) was noted by English legal practitioners and judicial officers to be an unnecessary complication, which left sentencing legislation open “to individual misinterpretation,” an interpretation which undermined the stated goal of achieving consistency in sentencing. In both Australian states under focus here, the drafting of the amended sentencing legislation to prescribe the respected minimum sentencing schemes were heavily criticized in later court decisions as unworkable in practice and lacking the necessary definitions to allow for consistent application to individual cases. This in itself highlights the need moving forward, both within Australia and elsewhere, for sentencing reform to occur in consultation with those working within the courts and criminal justice system and that which is evidence based. On both fronts, the recent Australian approaches towards presumptive minimum sentencing schemes arguably fall short.

Notes 1 Interestingly, the NSW Sentencing Council (2011) has acknowledged that at the time of its implementation the scheme was “presented as an alternative to the approach of mandatory sentencing that was being advocated by the party then in Opposition.” 2 This is a selective, non-​exhaustive list of the SNPPs prescribed. For the full list of SNPPs prescribed see Part 4 Division 1A Crimes (Sentencing Procedure) Act 1999 (NSW). 3 This includes where the victim is a police officer, emergency services worker, correctional officer, judicial officer, council law enforcement officer, health worker, teacher, community worker, or other public official, exercising public or community functions, and the offense arose because of the victim’s occupation or voluntary work. 4 For a detailed description of each of the options for reform given by the NSWLRC, see NSWLRC, 2012, pp. 32–​44. 5 For a detailed discussion of the implications of the High Court of Australia’s decision in Muldrock, see Donnelly (2012) and Hulme (2012).

References Ashworth, A., & Player, E. (2005). Criminal Justice Act 2003: The sentencing provisions. Modern Law Review, 68(5), 822–​838. Australian Law Reform Commission. (2017). Pathways to Justice—​An Inquiry into the Incarceration Rate of Aboriginal and Torres Strait Islander Peoples: Final Report. ALRC Repotr 133. Sydney, NSW: Australian Law Reform Commission.

126

Minimum Sentencing for Serious Offences Bar, V. (2014). Victorian Bar has some concerns over introduction of baseline sentences. Media Release, April 15, 2014. Brown, D. (2001). Mandatory sentencing: A criminological perspective. Australian Journal of Human Rights, 7(2),  31–​49. Clark, R. (2014, 8 April). No apology for being tough on serious crime. Herald Sun, p 21. Donnelly, H. (2012). The diminished role of standard non-​parole periods. Judicial Officers’ Bulletin 24(1), 1. —​—​. (2013). Crimes (Sentencing Procedure) Amendment (Standard Non-​parole periods) Act 2013. CLE talk Aboriginal Legal Service. Redfern: 5 December. Editorial. (2014, 5 April). Baseline sentencing is rank populism. The Age, p. 34. Fitz-​Gibbon, K. (2013). The mandatory life sentence for murder: An argument for judicial discretion in sentencing. Criminology and Criminal Justice, 13(5), 506–​525. Fitz-​Gibbon, K. (2016). Minimum sentencing for murder in England and Wales: A critical examination 10 years after the Criminal Justice Act 2003. Punishment and Society, 18(1), 47–​67. Fox, R.G. & Freiberg, A. (2014). Sentencing: State and Federal Law in Victoria. 3rd ed. Pyrmont, NSW: Lawbook Co. Freiberg, A. (2002). Pathways to Justice: Sentencing Review. Melbourne,Victoria: Department of Justice. Freiberg, A., & Gelb, K. (2008) Penal Populism, Sentencing Councils and Sentencing Policy (eds.), Annandale, New South Wales: Hawkins Press. Hoel, A., & Gelb, K. (2008). Sentencing Matters: Mandatory Sentencing Research Paper. Melbourne, Victoria: Sentencing Advisory Council. High Court of Australia. (2011, 5 October) Statement: Muldrock v. The Queen [2011] HCA 39. Canberra, Australia: High Court of Australia. Hulme, R.A. (2012). After Muldrock—​sentencing for standard non-​parole period offences in NSW. Judicial Officers’ Bulletin, 24, 81–​85. Hutton, N., & Tata, C. (2010). A sentencing exception? Changing sentencing policy in Scotland. Federal Sentencing Reporter, 22(4), 272–​278. Jensen, E. (2010, 25 May). Jail time increases up to 60%, reports says. Sydney Morning Herald, p. 4. Johnson, P. (2003). Reforms to New South Wales Sentencing Law: The Crimes (Sentencing Procedure) Amendment (Standard Non-​parole Periods) Act 2002. Seminar Paper for Judicial Commission of New South Wales presented March 12, 2003. Judicial Commission of New South Wales. (2010). The Impact of the Standard Non-​parole Period Sentencing Scheme on Sentencing Patterns in New South Wales, Research Monograph 33. Sydney, New South Wales: Judicial Commission of New South Wales. Lee, J. (2015, 27 February). Judges warn on baseline sentences: County Court judges warn on baseline sentences. The Age, p. 9. Loukas, C. (2002) Crimes (Sentencing Procedure Amendment) Standard Minimum Sentencing Bill 2002. Bar News:The Journal of the New South Wales Bar Association, Summer, 51–​54. Malik, S. (2012, 3 May). Tougher jail terms possible for Victoria. The Australian, unpaginated. Mitchell, B. (2013). Sentencing guidelines for murder: From political schedule to principled guidelines. In A. Ashworth & J. Roberts (eds.), Sentencing Guidelines: Exploring the English Model. Oxford: Oxford University Press. New South Wales Law Reform Commission. (2011). Sentencing: Preliminary Outline of the Review. Sydney, New South Wales: New South Wales Law Reform Commission. —​—​. (2012). Interim Report on Standard Minimum Non-​Parole Periods, Report 134. Sydney, New South Wales: New South Wales Law Reform Commission. Palmer, S. (2005). Law and order sentencing: Political practice in search of a rationale. Journal of Public Affairs, XVII, 24. Patterson, A. (2014, 27 October). Sentencing changes come with mandatory debate. Melbourne Law School News. Poletti, P., & Donnelly, H. (2010). The Impact of the Standard Non-​Parole Period Sentencing Scheme on Sentencing Patterns in New South Wales. Research monograph 233, NSW: Judicial Commission of NSW. Roberts, J. (2003). Public opinion and mandatory Sentencing: A review of international findings. Criminal Justice and Behaviour, 30(4), 483–​508. Sentencing Advisory Council. (2012). Baseline Sentencing Report. Melbourne, Victoria: Sentencing Advisory Council. —​—​. (2014). Calculating the Baseline Offence Median: Report. Melbourne,Victoria: Sentencing Advisory Council. —​—​. (2016). Sentencing Guidance in Victoria: Report. Melbourne,Victoria: Sentencing Advisory Council. Sentencing Amendment (Baseline Sentences) Act 2014 (Vic). Sentencing Amendment (Baseline Sentences) Bill 2014 (Vic). Sentencing Amendment (Sentencing Standards) Act 2017 (Vic). Sentencing Council. (2011). Standard Non-​Parole Periods: A Background Report by the NSW Sentencing Council. New South Wales: Sentencing Council.

127

Kate Fitz-Gibbon and James Roffee Spooner, R., & Bucci, N. (2014, 18 September). Jail terms to jump under new laws: Sentences to more than double for some offences, under new medians. The Age, p. 2. Tonry, M. (1992). Mandatory penalties. Crime and Justice: A Review of Research, 16, 243–​274. —​—​. (2009). The mostly unintended effects of mandatory penalties. Crime and Justice, 38, 65–​114. Warner, K. (2003).The role of guideline judgments in the law and order debate in Australia. Criminal Law Journal, 27,  8–​22. White, A. (2015, 31 March). Killers may beat jail: Baseline sentencing may help worst criminals avoid jail, warns Chief Crown Prosecutor Gavin Silbert. Herald Sun.

Statutes Cited Crimes Act 1958 (Vic). Crimes (Sentencing Procedure) Amendment (Standard Minimum Sentencing) Act 2002 (NSW). Crimes (Sentencing Procedure) Amendment (Standard Non-​parole Periods) Act 2013 (NSW). Crimes (Sentencing Procedures Act) 1999 (NSW). Sentencing Act 1991 (Vic). Sentencing Amendment (Baseline Sentences) Act 2014 (Vic). Sentencing Amendment (Baseline Sentences) Bill 2014 (Vic). Sentencing Amendment (Sentencing Standards) Act 2017 (Vic).

128

6 DISCRETIONARY RELEASE PRACTICES FOR JUVENILES FACING LIFE A Review of State Parole and Resentencing Procedures Stuti S. Kokkalera and Simon I. Singer

We know little about states of discretionary release practices for juvenile offenders, and by states we mean actual US states and their discretionary release practices. The term discretionary release is quite general, and we use it here to refer to parole board decision-​making.Whether to grant or deny parole is a bottom-​line parole board decision, but it is not the only decision that is routinely made to lead juveniles to a life sentence. Legislators, prosecutors, judges, correctional officials (just to name a few in the criminal justice system) have created America’s population of juvenile lifers. The status of America’s juvenile lifer population has always been initially directed by state laws. State laws could require juveniles convicted in criminal court to face the exact same sentences as adults, such as life without the possibility of parole. Or they could set relatively low and high minimum sentences with a maximum of life. Although life without the possibility of parole has been judged as unconstitutional based on the 8th Amendment’s cruel and unusual clause, high minimum sentences could legislate virtual life. In order to sentence juveniles to adult maximums of life, state legislators had to buy into the retributive, vengeful pursuit of popular punitiveness (Bottoms, 1995). Juvenile lifers emerged in increasing numbers because legislators were unwilling to recognize their adolescence as a reason to mitigate culpability, especially for juveniles charged with violence. A political mantra of adult time for adult crime enabled the sort of waiver legislation that would create our current population of juvenile lifers, and a subset of juveniles sentenced to the adult maximum of life without the possibility of parole (juvenile LWOPs). The penal state consists of a complex nexus that not only determines law but also the processes that lead to frontline decision-​making (Garland, 2013). The front line depends on the stage in the system. For our purpose, the critical end-​point in determining adult sentencing for juveniles is the parole board, or a state’s resentencing procedures. To learn more about today’s state of incarceration that has impacted juveniles, in the first part we discuss in the reasons for our population of juvenile lifers. The reasons for automatic waiver legislation are discussed, along with the Supreme Court’s 21st-​century recognition of adolescence. The Court’s recognition of adolescence is further related to the context of discretionary release decisions as they impact the possibility of parole for juvenile lifers. We conclude the first part with a review of the parole board decision-​making literature on instrumental and attributional assessments. In the second part we provide a review of the numbers; 129

Stuti S. Kokkalera and Simon I. Singer

the number of juvenile lifers with a focus on juvenile LWOPs. Our focus is on how states have produced their populations of juvenile lifers and their efforts at providing a meaningful opportunity of release.We conclude with a recommendation for formally recognizing adolescence, and evaluating the reentry needs of juvenile lifers.We finally make a plea for opening the doors of parole boards and their decision-​making processes.

The Status of Juvenile Lifers In an earlier era, only a few states were willing to subject youth to adult determinate sentences of life, most often for older adolescents who had a history of violent crimes (Bishop & Feld, 2014; Brown, 2012). The charging process was usually via the juvenile court. Judicial waiver procedures in juvenile court meant that juvenile justice officials would determine who would be eligible for criminal court. There was more care in terms of considering the reasons for juvenile offending by juvenile justice officials who were more familiar with adolescent development. In other words, there was usually no direct route to criminal court; adult sentences were less likely via judicial waiver from juvenile court (Feld, 2017). The rise and decline of the juvenile court for adolescents corresponds to the rise and decline of the rehabilitative ideal (Bishop & Feld, 2014).The decline of the rehabilitative led to automatic waiver legislation, and adult maximum sentences of life. The 1960s were a period of social anxiety about juvenile violence. Some states were earlier than others in making criminal court the first court of jurisdiction for juveniles. Automatic offense-​based waiver became the preferred techniques (Bishop & Frazier, 2000). States expanded or created waiver legislations that brought a larger proportion of the delinquent population directly into the criminal court (Singer, 1996). Even the most liberal of states were quick to produce automatic waiver legislation. New York became one of the first to go from a state without any possibility of waiver for juveniles (younger than 16) to a state with automatic waiver. New York’s Juvenile Offender Act of 1978 required juveniles as young as 13 to be charged directly in criminal court where they could receive a maximum sentence of life with the possibility of parole. Similarly, amendments to Massachusetts’ transfer laws in 1990 included lowering the age in which a juvenile could be automatically placed in the adult criminal court to 14; the maximum sentence would be life without any possibility of parole for first-​degree murder (Monahan & Kaban, 2009). It is important to note that a perfect storm led to the expansion and creation of waiver legislation for juveniles, and today’s populations of juvenile lifers. First, the criminalization of juvenile justice can be attributed to the due process revolution of the 1960s, as represented in the Supreme Court’s decisions in Kent (Kent v.  U.S. [1966]), Gault (In Re Gault [1967]), and Winship (In Re Winship [1970]). The Court’s decision in Kent (1966) is based on a case of judicial waiver. The Court mandated an adversarial hearing whereby the judicial waiver process would include legal representation. In Kent (1966), the Court viewed the power of juvenile court judges to make an administrative decision as too arbitrary. Gault (1967) and Winship (1970) moved the court into more of a due process-​oriented juvenile court, leading Feld (1999) to refer to the contemporary juvenile court as a criminalized juvenile court. The second major factor to expand waiver legislation was a perceived rise in violent juvenile crime, particularly in America’s urban centers (Zimring, 2005). Middle-​class white flight from the city to the greener pastures of suburbia left the impoverished inner city even more impoverished. Many of its youth were too much on their own, and a more due process-​oriented juvenile court seemed incapable of producing the needed security and control that many citizens desired. The most extreme form of waiver and the maximum adult sentences for juveniles began to sell in one state after another. Third, rehabilitation was perceived as ineffective, and less just than retribution (Allen, 1978). The juvenile court was based on the rehabilitative ideal, and it soon was associated with general 130

Discretionary Release Practices

critiques of rehabilitation in the sense that nothing was believed to work, as is often cited based on Robert Martinson’s 1974 evaluation of treatment programs. Not only was the juvenile court viewed as ineffective, but rehabilitation in general became an idea that passed its time. The decline of the rehabilitative ideal not only impacted juvenile justice, but also the length and likelihood of juvenile incarceration for those sentenced as adults (Bishop & Feld, 2014). Prisons began to operate fewer rehabilitative programs, and to merely warehouse a significantly larger proportion of its inmate population. Parole was impacted by determinate sentencing legislation. A few states abolished parole, and established commutation boards instead. A shift from actually meeting with parolees to merely monitoring and assessing risk contributed to a decline in aftercare programs (Clear & Frost, 2013). The end result is the most extreme form of adult maximum punishment—​life without any possibility of parole (LWOP) became America’s new death penalty (Ogletree & Sarat, 2012). For the adolescent population, the shift has been devastating. For example, prior to the Massachusetts 1996 Youthful Offender Act, a 14-​year-​old could receive a maximum of life with the possibility of parole, but that same juvenile now faced a maximum of life without the possibility of parole. Barry Feld (2017) summarized the numbers: “prior to 1980, judges rarely sentenced children to LWOP; now they impose LWOPs on youths three times as often as they did in 1990” (p. 210). Race is a factor too; black youths who killed a white victim were twice as likely to receive LWOP than white youths who killed a black victim (Feld, 2017).

The Supreme Court Recognizes Adolescence Now enter a series of US Supreme Court decisions recognizing the adolescence of juveniles sentenced for capital crimes.The first that explicitly referred to the developmental literature is Roper v. Simmons (2005). Roper (2005) raised the minimum age of capital punishment to 18, but just as important was the Court’s acknowledgement that children are not adults, and their culpability should be mitigated for the sole reason of their age.The dissenting justices disagreed, essentially arguing that judges should be free to decide on a case-​by-​case basis. Roper paved the way for the Court’s decisions in Graham (Graham v. Florida [2010]) and Miller (Miller v.  Alabama [2012]). Graham (2010) raised the bar for adult LWOP sentences to only nonhomicide. Up until Graham, states could impose a LWOP sentence on a juvenile who committed a non-​homicidal offense. Miller (2012) ruled that all legislated juvenile LWOP sentences are unconstitutional. However, Miller (2012) did not prevent states from retaining or creating minimums, and judges from imposing consecutive sentences that produced virtual life. Several aspects of Graham and Miller are worth highlighting for the purpose of this review. First, the Court has made clear that juvenile offenders are entitled to a “meaningful opportunity to obtain release based on demonstrated growth and maturity” (Graham, [2010] p. 75). Juvenile lifers should expect a meaningful review that recognizes their capacity to mature into law-​abiding adults. Second, the adolescence of juvenile lifers should be taken into account. An adolescent’s culpability should be mitigated by his/​her decreased aversion to risk, likelihood of being influenced by peers, familial dependency, and heightened impulsivity. Taking into account a constellation of developmental factors assumes that parole boards are capable of recognizing the adolescence of their juvenile lifers. Parole board members are in the habit of dealing with adult lifers. The juvenile lifers that appear before them are adults, most often in their middle-​aged years. A juvenile lifer may appear no different than an adult lifer. Adolescence is specifically recognized in a juvenile court and its systems of juvenile justice; it is not once a juvenile enters the criminal justice system. Although this review is limited to juvenile lifers and in particular juvenile LWOPs, state discretionary release practices impact a larger population of juveniles who have been sentenced in criminal court to adult maximums. The exact number is difficult to determine. Even if a very small segment of the approximately 250,000 juveniles initially brought into the adult criminal justice 131

Stuti S. Kokkalera and Simon I. Singer

system received long-​term sentences, over time that number would represent a significant part of the incarcerated population (Arya, 2011).The discretionary release practices of state parole boards are critical to developing sensible sentencing policies, and to reversing America’s high incarceration rate. As mentioned, state waiver legislation only partially accounts for the numbers of juveniles that are serving life sentences. Graham (2010) and Miller (2012) do not end the possibility that a segment of juveniles will spend the rest of their life in prison. The meaning of a meaningful review is left open to debate. To juvenile justice advocates it means recognizing the adolescence of juveniles, and in only the most extreme cases allowing a juvenile to serve more than a reasonable minimum sentence. But the reason why juvenile lifers might be denied parole may not have anything to do with their minimum sentence, their growth and maturity, their low reoffending risk score, and having positively responded to correctional programs. Rather the reason may be related purely to the seriousness of the offense. In other words, recognizing the juvenile lifer’s adolescence may not be enough to avoid repeated denials. Indeed, parole boards have a statutory right to selectively refer to the sentencing offense, the stated concerns of victims (if they are alive) and their family members, prosecutors, and other law enforcement officials, as well as interested members of the public. The possibilities are many, including how offense severity may be considered indicative of reoffending risk (Latessa & Lovins, 2010). Appeals based on the Supreme Court’s recognition of adolescence are emerging in response to the administrative procedures of parole boards. States are creating legislation that would require parole boards to modify their administrative procedures. Parole boards that previously made decisions without mentioning age are now acknowledging the offender’s adolescence. How that recognition occurs requires detailing the discretionary release practices in each state.

Parole in Context Parole has more than one set of meanings. It initially was envisioned as a sentencing institution based on the theory that rehabilitation was the dominant approach. Indeterminate sentencing made parole boards institutions in determining the amount of time that an offender would eventually serve. Retribution and even deterrence were secondary to the main objective of parole as a method for observing the offender and saving his labor to meet the demands of late 19th-​century industrialization (Simon, 1990). Of late it has come to mean much more. The punitive turn in American penology has shifted the focus of parole to include risk-​management (Ball, 2011). The dangerousness of the offender is also defined in terms of the seriousness of the offense, and the degree to which members of the community object to the offender’s release. In other words, parole has developed into ways that draw on a multitude of normative concerns –​ones that fit a post-​modern, decentralized society where the mechanisms of state control are diverse. At one time, parole could be easily identified in terms of its rehabilitative mission. That is no longer the case. Parole fell out of favor as determinate sentencing systems replaced less determinate ones, and the prison would take on a warehousing look. Rehabilitation is there, but more as an indicator of reoffending risk. And reoffending risk has become closely associated with offense seriousness and dangerousness (Feeley & Simon, 1993). An understanding of the parole board’s discretionary release practices is important for not only reasons of fairness, but also for providing juvenile lifers with the hope of treatment and potential aftercare. States are grappling with how to fulfill the promise of a meaningful review. The words are there in that juveniles are entitled to a “meaningful opportunity to obtain release based on demonstrated growth and maturity” (Graham, 2010, p. 75). Juvenile lifers might have expected a meaningful review which, among other things, recognized the fact that at the time of their offense they were adolescents. Less understandable are reasons that ignore the offender’s adolescence, as mentioned. The ease in which to ignore a juvenile offender’s adolescence is there in legislation, sentencing, and then subsequently through the possibility of parole. The juvenile lifer’s adolescence at time of offense and good 132

Discretionary Release Practices

conduct may have been acknowledged and might not matter, given a range of parole board concerns. According to one state’s parole statute, the board cannot grant parole “merely as a reward for good conduct” (Massachusetts General Laws, Chapter  127, Section 130). As noted, the possibilities for denial are endless.

Instrumental Assessments Parole boards are in the business of assessing reoffending risk and appropriate sentencing time, given the severity of the offense (Latessa & Lovins, 2010). They are also required to take into account the concerns of victims (if alive), the victim’s family members, prosecutors, and citizens who might wish to speak for or against the offender’s parole. Posted guidelines emphasize the responsibility of the parole board to review the nature of the crime and entire criminal history (Ball, 2011). They also emphasize that an offender’s eligibility for parole is contingent on the reasonable probability that an individual will not violate the law if released. In regard to reoffending risk, a recent survey of parole boards showed that parole boards are giving increasing weight to risk assessment scales. According to Ruhland and colleagues (2016), 36 out of 40 responding states reported that they use some form of risk assessment in determining release. Although parole boards routinely draw on standardized risk assessment scales, they are not automatically applied. Tidmarsh (1997) concluded that objective risk assessment scales were of limited value in the locale he observed. Standardized risk scales were seen as having limited value because parole board members believed them to contain questionable actuarial data. Instead, parole board members weighed the risks of release based on their personal feelings about incarceration. Some parole members noted “the possibility that keeping someone inside may not only postpone risk but increase it” (Tidmarsh, 1997, p. 274). Tidmarsh further noted that the parole board members are socialized agents, with newly appointed members going through a process of “organizational learning.” Hawkins (1986) interpreted the instrumental concerns of parole boards operating in several states. He documented how decisions to grant parole in one state are more contingent on an inmate’s prison record than on the parole board hearing itself. In the state where the inmate’s record matters more, Hawkins concluded that discretionary release is used more as an “internally directed management tool.” He observed that “the board’s general approach was to depend more on the assessments of institutional authorities” (1986, p. 1199), making decisions highly related to institutional behavior rather than sentencing offense. Hawkins further observed that in another state, the parole board placed less weight on institutional adjustment and more on assessments made at time of the hearing, such as an offender’s display of remorse. Parole boards also make judgments based on how they see the offender’s culpability and offense severity in terms of deserved time served. Going rates are established, and parole boards may adjust the actual amount of time served based on how its members feel about the offender and the offense. For example, in their examination of 421 cases heard in California, Holland and colleagues (1978) noted that correctional caseworker evaluations were of lesser importance than the severity of the sentencing offense. Parole boards made decisions according to their expectations for time served/​ deserved for certain kinds of offenses, thus relying mainly on descriptions of the sentencing offense. Similarly, Morgan and Smith (2005) analyzed Alabama parole board decisions for 762 cases between 1993 and 1994 and found that the strongest predictor of parole was the amount of time served on the sentence. However, controlling for offense severity, they found that total number of past felonies and the “senior officer’s” (i.e., corrections official) recommendation were also influential factors related to release. The age of the inmate at time of hearing was not significant. Notably, however, their multivariate analysis did not include a measure of age at time of offense. How the factor of age at time of the offense or characteristics associated with adolescence plays into expectations for time served is relatively unexplored. Caldwell (2016) analyzed 107 Youth Offender parole hearings in California and found that age at the time of the offense was a significant 133

Stuti S. Kokkalera and Simon I. Singer

predictor of release decisions. Specifically, a younger offender was more likely to be found suitable for release than an older offender. Caldwell’s study also examined the reasons given by the parole board to deny release for offenders who committed their crimes as adolescents. She found that “the reasons Commissioners provided for denying parole in many cases focused on the nature of the commitment offense, or to closely related issues such as an individual’s lack of insight or honesty about the offense” (2016, p. 295). Parole boards are also required to take into account community concerns. Other variables that may play an influential role in the parole decision-​making process are victim, prosecutorial, and solicited community impact statements (Proctor, 1999). As victims’ groups and prosecutors have gained increasing influence in discussions of justice policy and court processes, they have also become more integral to the parole process. According to a national survey conducted by the Association of Paroling Authorities International (Kinnevy & Caplan, 2008), nearly all states consider victim input when making release decisions. Whether victim, prosecutorial, and solicited community impact statements influence on parole decisions in the case of adolescent offenders, however, is undocumented.

Attributional Assessments Assessments of who deserves to be paroled are not isolated individual decisions. Emerson (1991) describes how legal concepts like culpability and reoffending risk are socially constructed. A holistic approach to case-​processing decisions suggests that attributes of responsibility are relative, emerging from a stream of cases (Emerson, 1983). Singer (1996) showed how in jurisdictions where violent juvenile offending is relatively rare, there are higher levels of punishment. As noted, parole boards are in the business of attributing criminal responsibility, and they may do so based on a limited set of factors.These factors can be broadly grouped into two categories: external factors, including the environment or situation in which the crime was committed (and youth itself), and internal factors, including supposedly “deep” facts about a person’s moral orientation or underlying disposition. When responsibility for the crime is attributed primarily to external factors, parole board members are judging a person as less morally deserving of blame (reduced “blameworthiness”), at lower future risk of future criminal activity, and of better moral character. By contrast, attributing responsibility for the crime to internal factors is associated with judgments of increased blameworthiness, higher risk of reoffending, and “bad character” (Bridges & Steen, 1998; Emerson, 1969). Attributions as they relate to the sentencing offense further relate to reoffending risk. For instance, Carroll (1978) found that “the higher the stability of internal attributions regarding the cause of the parole applicant’s offense, the higher the board member’s prediction of offender risk and the more reluctant he or she is to grant parole” (pp.  1509–​1510). Similarly, Hawkins indicates that parole board members’ attributional choice of words is impacted by the offense, so that a horrific crime is considered indicative of bad character. Internally and externally directed attributions of juvenile lifers may be related to their age.The fact that the juvenile offender committed a violent offense at a relatively young age may be considered indicative of an intractably aggressive person, or one that was only temporarily corrupted by a highly dysfunctional family. A juvenile lifer described as particularly dangerous based on a series of documented antisocial behaviors extending from childhood through adolescence may be viewed as indicating high offending risk. Alternatively, the dependent status of adolescents may be considered as a way to include externally driven attributions, such as the reason for the offense being an abusive parent or involvement in a violent gang for the purpose of protection. If the juvenile lifer’s offense is perceived to have been the product of a group action, and therefore subject to situational factors that precipitated the sentencing offense, then the offender’s blameworthiness and reoffending risk may be viewed as relatively low (Bridges & Steen, 1998). As in research on sentencing, parole decision-​making has been related to personal characteristics. Gobeil and Serin (2009) presented vignettes to a sample of parole board members in Canada and 134

Discretionary Release Practices

New Zealand, finding that, even after controlling for level of risk, the likelihood of granting parole varied by gender and race of the offender. The decision to release was more common for female offenders, and less common for minority (Aboriginal) offenders. Although hampered by a small sample, an important point of the study was to show that parole board members operated with a limited set of information, drawing on a few (stereotypical) characteristics to make assessments about offender risk. Huebner and Bynum (2008) also examined the effects of race and ethnicity by examining parole decision-​making for 423 young adult inmates who became eligible for parole during an eight-​year period in one state. Their dependent variable was the amount of time to parole after the minimum sentence was served. Controlling for a range of variables, race and ethnicity were found to have significant effects. Black offenders, more than white and Hispanic offenders, spent significantly more time awaiting release, independent of any effect of neighborhood disadvantage. Although not examining juvenile offenders, this work demonstrates that “parole actors may perceive blacks as more threatening and deserving of longer terms of imprisonment” (Huebner & Bynum, 2008, p.  925). Other research has focused on race disparities in the decision-​making that leads to parole revocation hearings or pre-​sentencing reports conducted by probation officers. Though the focus is not on parole but pre-​sentence reports, Bridges and Steen (1998) found that the crimes of black juveniles were more often attributed to their bad character, while the attributions of white juveniles were focused on external factors that could be remedied, like peer group influences. The extent to which adolescence matters and the dynamics of its effects on the decision-​ making process are open questions. The only studies we know of that have touched upon adolescence and parole decision-​making are recent studies by Caldwell (2016) and by Bell (2018), who have each independently examined data on California’s specialized youth offender parole hearing board, which was created in the wake of Miller. Although the rate of release at youth offender hearings was slightly higher than that of adult hearings in 2014, it has since fallen slightly below that of adult hearings. Our review leads us to conclude that discretionary release practices of parole boards are impacted not only by statutorily, normative concerns, but also by an organizational one that relates to the meanings attached to culpability and reoffending risk. As noted, the new concept for parole is “the meaning of a meaningful review.” For our purpose it is how states can better recognize the adolescence of their juvenile offender populations, including those eligible for life. In the next part, we detail the populations of juvenile lifers with a particular focus on juvenile LWOPs.

Populations of Juvenile Lifers Reports conflict with one another on the number of juvenile LWOPs. According to Nellis and King, “6,807 juveniles are serving life sentences, of which 1,755 or 25.8% are serving sentences of life without parole” (2009, p. 3). More recently, Rovner (2017, 2014) reported that about 2,500 individuals are directly impacted by the Miller decision whereas the Associated Press (2017) estimated that there are 3,150 juvenile LWOPs across the country. According to Figure  6.1, which presents data on the distribution of juvenile lifers across the country, in one-​fifth of the states and the District of Columbia, no juvenile lifers were sentenced to life without parole. A further look at these states’ statutes on sentencing reveals that some states lack the option to sentence juveniles to life with or without parole. For example, Alaska does not have life without parole for juvenile homicide offenders. Instead, judges can sentence an offender convicted of first-​degree murder to a term of 99 years in prison (Alaska Code of Criminal Procedure, Chapter 55, Section 125). However, they may consider mitigating factors, such as a defendant’s age and the group context of the offense in reducing the term of sentence (Alaska Code of Criminal Procedure, Chapter 55, Section 155(d)(5); Alaska Code of Criminal Procedure, Chapter 55, Section 165). In other states that do not have any juvenile LWOPs, judges did not impose life without parole 135

Stuti S. Kokkalera and Simon I. Singer

Number of States and the District of Columbia*

25 21 20

15 10 10

8 5

5

5 1 0 0

1 to 25

26 to 50

51 to 100

101–200

More than 200

Number of Juvenile Lifers serving LWOP

Figure 6.1  State-​wise distribution of incarcerated individuals sentenced to life without parole as juveniles1 1Number of juvenile lifers serving life without parole sentences (mandatory or discretionary) and currently incarcerated (not including those who have been resentenced) were collected by AP and presented as graph here. *AP was unable to get an accurate count from New Jersey and is not included in this figure.

because of provisions allowing judges to take age at the time of the offense into account. For example, criminal court judges in Kansas can hold hearings to take the age of an offender into account to lower terms of incarceration (Crimes and Punishments, Chapter 21, Section 6625). Figure 6.1 also shows that 21 states have between 1 and 25 juvenile lifers sentenced to life without parole. Of these 21 states, 9 states have juvenile lifers as a result of legal provisions that provide judges the discretion to impose such a sentence (i.e., discretionary life without parole). The remaining 12 states with juvenile lifers had mandatory provisions of life without parole for certain offenses, such as first-​degree murder. Among the other states in Figure  6.1, five states have between 26 to 50 juvenile lifers; eight states have between 51 and 100 juvenile lifers and six states have more than 100 juvenile lifers incarcerated due to provisions of either mandatory or discretionary life without parole. Of the states with especially large numbers, Arkansas has 114 juvenile LWOPs, whereas California, Florida, Louisiana, Michigan and Pennsylvania have more than 200 juvenile LWOPs. This may be directly related to the size and the population of these of states. For instance, Philadelphia County in Pennsylvania is responsible for 9% of all life without parole sentences for juvenile offenders nationwide (Mills, Dorn, & Hritz, 2015).

Juvenile LWOPs and State Waiver Policies As mentioned, the numbers of juvenile lifers are largely driven by state policies on prosecution and sentencing of juvenile offenders. While the transfer of juveniles to criminal court is not new, changes across states in the late 1980s and early 1990s expanded the scope of transfer policies (Griffin et al., 2011). Several states reduced the use of judicial waiver and introduced statutory exclusion laws and/​ or increased prosecutorial discretion in filing cases in criminal court (Brown, 2012; Feld & Bishop, 2012; Zimring, 2010). States with statutory exclusion laws jumped from 20 to 38 in the 1990s and those with prosecutorial discretion laws rose from 7 to 15. These changes had the effect of increasing the prosecution of juveniles charged with serious crimes in criminal court (Griffin et al., 2011). Once waived into and convicted in criminal court, juvenile offenders are eligible for adult maximum sentences, such as life. The number of cases transferred to the adult system nearly doubled between 1985 and 1994 and contributed to more juvenile offenders receiving life sentences (Nellis 136

Discretionary Release Practices

& King, 2009). Before Miller (2012), 28 states and the federal government had mandatory provisions of parole-​ineligible life sentences for juvenile homicide offenders (Rovner, 2014). For example, Pennsylvania had one of the highest number of juvenile lifers (more than 500) because judges were required to sentence individuals to life without parole and were not allowed to consider a juvenile’s age or related circumstances when youth were tried in criminal court (Nellis, 2012). When judges impose mandatory life sentences on juvenile offenders, they fail to take into account factors that mitigate culpability. For instance, Monahan and Kaban’s study (2009) of juvenile LWOPs in Massachusetts found that 41% of cases involved youth who were first-​time offenders; 40% of cases involved youth who offended with an adult codefendant and 20% of cases were those in which the juvenile offender was not the principal actor. Although Miller (2012) banned an automatic life without parole sentence for juveniles, most states have retained discretionary life without parole. Only 20 states and the District of Columbia have banned a discretionary sentence of life without parole for juvenile offenders, but three-​fourth of these states had no juvenile LWOPs (Rovner, 2017). For states that retained discretionary life without parole, a judge must find that a juvenile’s crime reflected “irreparable corruption” to sentence them to life without parole (Montgomery v. Louisiana, [2016]). Alternatively, state provisions may have the option of life without parole but can permit judges to impose parole-​eligible sentences for juvenile offenders. For example, certain crimes in Arizona require judges to impose a “natural life sentence” (Arizona Criminal Code, §13–​751(A)¶2) without any parole eligibility, although judges may impose a parole-​eligible life sentence based on the age of an offender (Arizona Criminal Code, §13–​751(G)  ¶5). No national data set currently tracks juveniles prosecuted and sentenced in criminal court (Griffin et al., 2011; Nellis & King, 2009). Studies that examined the numbers of juveniles serving life without parole sentences requested and received data from individual states (AP, 2017; Mills, Dorn, & Hritz, 2015; Nellis, 2012). Even with limited data, there is evidence of the correlation between the expansion of waiver policies and increasing number of juvenile LWOPs (Mills, Dorn, & Hritz, 2015). In recent years, states have started amending their transfer provisions by restricting the criminal court’s jurisdiction to only the most serious crimes (Brown, 2012). The number of cases transferred into criminal court has seemingly declined, indicating that the number of life sentences has dropped over time (Nellis & King, 2009).Yet, in light of Graham and Miller, states have continued to focus on amending sentencing provisions without further reexamining their transfer policies as they apply to juveniles charged with serious offenses.

“Virtual” lifers Few studies have examined the number of juveniles whose incarceration is the result of a parole-​ eligible life sentence or terms that resemble a life sentence (Ford, 2017; Nellis & King, 2009). The Office of Juvenile Justice and Delinquency Prevention (OJJDP) releases national data on the processing of juvenile court cases, including those that are waived annually (Easy Access to Juvenile Court Statistics [EZAJCS], 2015).These data are not disaggregated by type of waiver or by disposition once waived into criminal court.We cannot tell the range of prison sentences, including life sentences and cumulative aggregate terms that resemble life. A few state appellate courts have examined the constitutionality of a juvenile’s life sentence. The Nebraska Supreme Court held that a life sentence for a 15-​year-​old is the functional equivalent of a life without parole sentence and violates the Eighth Amendment (State v. Mantich [2014]). Similarly, the Mississippi Supreme Court ruled that a life sentence with eligibility for parole when a juvenile offender is 65 years old is prohibited by Miller (Parker v. State [2013]). By contrast, the Tennessee Court of Criminal Appeals held that a sentence of life with possibility of parole at 51 years for juvenile homicide offenders is not unconstitutional under the Eighth Amendment (Perry v. State [2014]).This court ruled that a minimum mandatory term of 51 years can be characterized as a virtual life sentence, 137

Stuti S. Kokkalera and Simon I. Singer

but it is not legally inconsistent with Graham or Miller. In Massachusetts, the Supreme Judicial Court (SJC) held that a parole-​eligible life sentence may be applied to a juvenile offender and is not inconsistent with the Eighth Amendment (Commonwealth v. Okoro [2015]). However, such parole-​eligible life sentences can be unconstitutional if the state’s parole practices do not provide juvenile lifers with a meaningful review (Russell, 2016). Providing a meaningful review allows for a thoughtful consideration of whether a juvenile lifer is ready to transition into the community (Nellis & King, 2009). In Miller (2012), the Court ruled that mandatory life without parole is cruel and unusual punishment because legislators and sentencing courts did not recognize the adolescence of juvenile lifers. The same rationale should apply for juveniles sentenced in criminal court to long-​term mandatory terms of less than life (Hilliard, 2016). Since the Court did not clarify a “meaningful opportunity” (Graham, 2010, p. 75) states are left to establish release mechanisms, and to determine if juveniles serving de facto life sentences are eligible for discretionary release (Moriarty, 2017b). Thousands of juveniles are serving de facto life sentences as the result of long-​term mandatory sentences. For example, the District of Columbia alone has around 100 juvenile offenders serving lengthy mandatory sentences that are not sentences of life without parole (AP, 2017; Mills, Dorn, & Hritz, 2015). Two state appellate courts have categorically recognized such sentences to be unconstitutional under the Eighth Amendment’s proportionality clause (Kirby, 2015). The Iowa Supreme Court held that a mandatory term of 60 years without parole for a juvenile offender was the functional equivalent of life without parole under Miller (State v. Ragland [2013]). Similarly, the Connecticut Supreme Court held that sentencing a juvenile to 100 years in prison resembled a life sentence without parole, and Miller remedies of resentencing or determining parole eligibility would apply (State v. Riley [2015]). Aggregate or consecutive sentences are another way to reproduce LWOP sentences. For example, the Massachusetts SJC declined to extend the Eighth Amendment to a juvenile sentenced to aggregate sentences for non-​homicide offenses, where the aggregate sentences exceeded the parole eligibility for a murder conviction (Commonwealth v. Perez [2017]). In Perez, the SJC remanded the matter to the lower court for resentencing because the sentence violated the state’s constitution; however, the SJC was reluctant to rely on the Eighth Amendment since the Supreme Court has not extended the Amendment’s protection to aggregate sentences for juveniles. In 2018, the Supreme Court had the opportunity to clarify whether Graham or Miller remedies applied to juvenile offenders who received aggregate sentences for their crimes. The Court declined to hear the case (Bostic v. Dunbar [2018]) where a juvenile was sentenced to 241 years in prison at the age of 16 for multiple counts of robbery, making him eligible for parole when he would be 112 years old (Wheeler, 2018). Therefore, without a Supreme Court mandate, state appellate courts may be reluctant to extend the Eighth Amendment’s protection against cruel and unusual punishment to virtual lifers or the protections afforded under Graham, Miller and Montgomery.

Determining Eligibility for Early Release Thirteen states have responded to Graham (2010) and Miller (2012) by either abolishing or restricting the use of life without parole for juvenile offenders (Mills, Dorn, & Hritz, 2015; Rovner, 2014). Some of these states introduced new terms of parole eligibility on a life sentence (e.g., 30 years in Alabama) or reduced the term of eligibility (e.g., Delaware reduced the term of parole eligibility from 30 years to 25 years for first-​degree murder). Four of these states clarified that the new laws applied retroactively to juvenile LWOPs (see Rovner, 2014 for list). Connecticut joined this list in 2015, when it passed a new law that made juvenile lifers eligible for parole if they had served at least 30 years of their life sentence1 (EJI, 2015).Three states (Rhode Island,Tennessee and Texas) introduced legislative bills that would lower the term of eligibility of parole for juveniles convicted of homicide offenses, but all three bills failed in the state legislatures. Among these three states,Tennessee currently provides the longest minimum term to serve before becoming eligible for a parole hearing at 51 years (Kirby, 2015), whereas juvenile offenders in Texas must serve a term of 40 years of their life sentence before 138

Discretionary Release Practices

becoming eligible for parole (Rovner, 2014). On the other hand, Rhode Island has no juvenile lifers serving non-​parole-​eligible life sentences, and a bill to lower the term of parole eligibility for juvenile offenders may have failed because it did not seem warranted at the time. Laws attempting to amend sentencing schemes also raise a separation of powers issue between the legislature and the judiciary. The failure of the three state bills may be due to a lack of support from lawmakers who do not want judges to determine parole eligibility. In Montgomery (2016), the Supreme Court suggested that a state can remedy a life without parole sentence by making juvenile LWOPs eligible for a parole hearing if they have served a certain amount of time. However, with the onus on states to determine eligibility, several states continue to provide resentencing hearings. As shown in Figure 6.2, three states (Arizona, Florida and Maine) provide resentencing hearings to determine a fixed sentence (anywhere between 25 to 100 years), primarily because none of these three states have a parole system. Maine was the first state to eliminate parole altogether (AP, 2017) and sentencing judges there retained the discretion to impose life without parole (NCSL, 2010). Arizona and Florida were among the 13 states that passed legislation or issued judicial rulings to provide resentencing hearings before the Court’s decision in Montgomery (2016). In these states, juvenile lifers must file a petition for a resentencing hearing and the court follows criminal sentencing procedures. Figure 6.2 also shows that 27 states provide resentencing hearings to determine parole eligibility, even if they have a process of discretionary release for violent offenders. A closer look at these states’ procedures reveals two reasons for providing resentencing hearings instead of making juvenile lifers immediately eligible for parole. First, though states have passed laws that ban life without parole for juvenile lifers, these laws apply only prospectively. For instance, Pennsylvania amended its legal provisions to create more options of sentences with reduced terms of parole eligibility2 (Kirby, 2015). Pennsylvania’s new sentencing options of life with parole are only available to juveniles convicted after 2012 and do not allow for retroactive application to those serving life without parole. Following Montgomery (2016), Pennsylvania has not made any other legislative changes and continues to provide resentencing hearings to juvenile LWOPs. Second, some states have eliminated discretionary parole

30

Number of States and the District of Columbia

27 25 21 20

15

10

5

3

0 Resentencing before determining parole eligibility

Resentencing to determine fixed sentence

No resentencing

Figure 6.2  State-​wise distribution of determining parole eligibility for juvenile lifers previously sentenced to life without parole1 1Data for Figures 2 to 6 were collected by the authors.

139

Stuti S. Kokkalera and Simon I. Singer

altogether and, hence, resentencing hearings serve as the appropriate remedy. On the other hand, Figure 6.2 also shows that 21 states provide parole hearings without resentencing through legislation, following Miller (2012) or through judicial rulings prior to Montgomery (2016). Four states passed laws that apply retroactively to juvenile lifers serving life without parole (Rovner, 2014), whereas nine state Supreme Courts have held that Miller (2012) applies retroactively. Prior to the Court’s decision in Montgomery (2016), these nine states began conducting parole hearings for juvenile LWOPs as early as 2014 (e.g., Massachusetts). However, neither Miller (2012) nor Montgomery (2016) address the duration of community supervision for juvenile lifers. Generally, the term of supervision corresponds to the sentence received, meaning that individuals sentenced to life will be supervised for the rest of their lives. Most states do not allow individuals serving life sentences to petition for early discharge and only two states (Idaho and Georgia) allow individuals convicted of violent offenses to petition for an early discharge from parole after five years. There are no studies that have examined the impact of early discharge from parole for juvenile lifers;3 presumably because few states provide an opportunity for early discharge for individuals serving life sentences. We also know little about recidivism rates or the rates and types of parole revocations for juvenile lifers released on parole. Therefore, the next frontier of parole research and policymaking is examining whether lifetime supervision is necessary for juvenile lifers, and whether such supervision constitutes a disproportionate punishment under the Eighth Amendment (Russell, 2014).

Resentencing versus Parole It is too early to tell the impact of resentencing as opposed to the opportunity for immediate parole eligibility (Drinan, 2017). Resentencing allows states to comply with Miller and Montgomery (Rovner, 2017). It retains the authority of the sentencing court, especially in states that provide for a discretionary sentence of life without parole for juvenile homicide offenders (e.g., Alabama). Resentencing also maintains the separation of powers between an executive agency (the parole board) and the courts. However, resentencing hearings focus on the sentence needed for a crime committed many years ago. While they consider mitigating circumstances at the time, they do not necessarily consider the juvenile lifer’s conduct while incarcerated (Boone, 2015). By contrast, parole boards typically discuss the facts of the crime as well as a juvenile lifer’s institutional conduct and preparedness for release. Yet, states can circumvent the requirement of a meaningful opportunity of release by delaying parole eligibility, since the Supreme Court provided no minimum standard for release (Russell, 2014). States are also cautious about determining eligibility for parole for those sentenced to multiple life sentences since a juvenile lifer’s eligibility is usually established by the courts (Ruhland et al., 2016). For instance, the Supreme Court of Minnesota ruled that Miller and Montgomery did not apply to a juvenile offender who received three consecutive life sentences (Minnesota v. Mahdi Hassan [2017]). Some scholars have argued that parole eligibility should be determined by a juvenile lifer’s demonstrated rehabilitation as opposed to the circumstances of the crime (Caldwell, 2016; Russell, 2016), but this requires serious legislative changes which states seem reluctant to do. The Supreme Court has also erred on the side of caution in its decisions in Graham, Miller and Montgomery by not clarifying the aspects of a meaningful review for juvenile lifers; there is no minimum threshold for release, or established procedural standards (Moriarty, 2017a). As a result, the procedures used to provide juvenile lifers with a meaningful opportunity for release vary significantly across states.

Providing a Meaningful Review Determining when a juvenile lifer is eligible for parole is only the first step (Russell, 2014). Miller (2012) suggested several factors that a court at the time of sentencing a juvenile offender or a resentencing 140

Discretionary Release Practices

court or state parole board should consider when deciding release for juvenile lifers. These include juvenile lifers’ understanding of their lack of maturity at the time of the crime; their role in the crime; their home environment; peer pressure; their understanding of legal proceedings and their rehabilitation in prison (Levick & Schwartz, 2013). Twenty-​five states and the District of Columbia have incorporated Miller factors through amendments to their criminal provisions, including recognizing age at the time of the crime as a mitigating factor in the culpability of juvenile offenders. For example, Hawaii introduced a new law that requires sentencing judges to consider 15 Miller factors before imposing a life sentence (Moriarty, 2017a). Some states (e.g., Massachusetts, New York) issued parole guidelines that required board members to recognize Miller factors when determining release for juvenile lifers. Arkansas introduced a new law for juvenile lifers who could seek parole after 25 to 30 years and required the parole board to recognize factors that mitigate a juvenile lifer’s culpability (The Fair Sentencing of Minors Act, Section 13 (b)(2)). However, this law was struck down by a state circuit (trial) court (State v. Hardman [2017]). Though the decision was appealed, the Arkansas Supreme Court (the state’s highest appellate court) at this time has not yet decided on the constitutionality of the Act (Lasley v. State [2017]). Twenty-​five states have not incorporated Miller factors in their statutes or parole schemes. Of these, Iowa’s Supreme Court (State v.  Zarate [2018]; State v. Lyle [2014]) and Oklahoma’s Criminal Court of Appeals (Luna v. State [2016]) have ruled that a resentencing court must recognize Miller factors when deciding release for juvenile lifers. The recognition of Miller factors in statutorily legislated rules and in the discretionary release practices of parole boards is critical to a meaningful review. Most juvenile lifers have experienced difficult childhoods, including instances of violence and abuse (Drinan, 2017; Rovner, 2014). Ideally, resentencing courts and parole boards should recognize criminological reasons for juvenile violence which can range from childhood trauma to neighborhood violence (Singer, 2017). Additionally, when parole boards evaluate release, they should consider a juvenile lifers’ prison behavior in the context of adolescent development. Risk assessments should incorporate recent research on adolescent behaviors (Caldwell, 2016). If adolescent developmental factors remain unacknowledged, then the possibility of parole is more symbolic than real in providing juvenile lifers a meaningful review (Caldwell, 2016; Cohen, 2014; Russell, 2014).

Discretionary State Release Practices The type of parole board hearing determines how juvenile lifers previously ineligible for parole may receive a meaningful review. As noted in Graham (2010) and Miller (2012), a meaningful review should be based on demonstrated maturity and rehabilitation for all juveniles serving long-​term adult sentences. Figure  6.3 shows that 32 states and the District of Columbia have no discretionary release procedures for individuals convicted of violent offenses. Once resentenced to the possibility of parole, the challenge for these states is to determine how to provide hearings for juvenile lifers. One solution is to offer “hybrid hearings” instead of resentencing where the trial court looks at the mitigating circumstances at the time of the crime and examines a juvenile lifer’s rehabilitative efforts (Boone, 2015). The hybrid hearing combines elements of a resentencing hearing and parole hearing while providing some due process protections. Four out of these 33 jurisdictions (New Mexico, Ohio, Utah and Wisconsin) have created a separate parole eligibility mechanism (without resentencing) for juvenile lifers. Figure 6.3 also shows that three jurisdictions that lack discretionary parole rely exclusively on resentencing hearings to determine release of juvenile LWOPs. For the 15 states that have discretionary release procedures, the possibility of providing a meaningful review is less challenging. For example, in 2014, the Massachusetts Parole Board issued a policy with criteria that must be considered by board members when it conducts hearings for juvenile lifers, including evaluating their adolescence at the time of the crime. It remains to be seen how parole boards have applied their revised guidelines in arriving at decisions to parole juvenile lifers. 141

Stuti S. Kokkalera and Simon I. Singer

Number of States and the District of Columbia

35

33

30 25 20 15 15 10 5

3

0 No Parole Authority

Parole Board with determinate sentencing for violent offenses

Parole Board with indeterminate sentencing for violent offenses

Figure 6.3  State-​wise distribution of type of parole board

States also vary in their discretionary release practices because of previous Supreme Court rulings on parole.The Court found a specific liberty interest for parolees who face returning to prison when their parole is revoked (Morrissey v. Brewer [1972]). According to the Court, terminating a parolee’s liberty “calls for some orderly process, however informal” (Morrissey, 1972, p. 482) to determine the facts that support the revocation of parole. Five years later, the Court established that a state’s parole statute can create a “protectible expectation of parole”4 in which a minimal standard of due process applies (Greenholtz v.  Inmates of the Nebraska Penal and Correctional Complex [1979]). While Miller (2012) and Montgomery (2016) do not address the elements of a meaningful review for juvenile lifers, both rulings put the onus on states to determine some form of procedural due process (Moriarty, 2017b; Russell, 2016). Importantly, Miller (2012) did not address the paradox of considering the adolescence of a juvenile lifer who is no longer a juvenile, making it necessary to clarify the due process requirements in a parole hearing (Boone, 2015). Procedural due process begins with an opportunity to be heard (Russell, 2014). In the context of juvenile lifers appearing for a parole hearing, other due process considerations include the right to counsel (Moriarty, 2017a; Russell, 2014); the burden of proof to support a decision and whether an individual is entitled to a written decision from the parole board (Russell, 2014).

An Opportunity to Be Heard According to Figure 6.4, three state parole boards (Illinois, Indiana and Minnesota) lack hearings for violent offenders and instead, they provide resentencing hearings. In 23 states, parole boards must hold a public hearing for lifers. Open hearings are conducted in a specific location like the parole board office (e.g., Massachusetts), prison (e.g., California), or via video-​conferencing (e.g., South Dakota) where several interested parties may attend, such as the victim’s family members. Figure 6.4 also shows that 22 state parole boards conduct closed hearings either at the offender’s prison or via video-​conferencing and this would preclude anyone, except the offender, from attending the hearing. Only one state (North Carolina) conducts closed hearings without any in-​person testimony from the offender. Most parole boards seek victim input either through the victim’s family or a representative or through the submission of a victim impact statement. States may also allow representatives 142

Discretionary Release Practices 25 23

Number of Parole Authorities

22 20

15

10

5 3

0 Open Parole Hearings

Closed Parole Hearings

No Parole Hearings (with a Parole Authority)

Figure 6.4 Type of parole hearing for juvenile lifers

of law enforcement to attend, such as from the district attorney’s office (e.g., Connecticut), the police department where the crime occurred (e.g., New Hampshire), or a sentencing judge (e.g., Ohio). Officials and citizens in open parole hearings have the right to make statements for or against the offender’s release. Ultimately, providing a public hearing allows juvenile lifers to prove their preparedness for release, comporting with a meaningful review (Russell, 2014).

Right to Counsel States vary as to the availability of legal representation during the process to secure release on parole. The US Supreme Court has been less than clear as to a parole-​eligible offender’s right to legal representation. State parole boards usually hold three types of hearings: initial, review, and parole revocation. A juvenile lifer appears for an initial hearing after serving a minimum amount of time, ranging from 15 (e.g., Massachusetts) to 51 years (e.g., Tennessee). If a juvenile lifer is denied parole at the initial hearing, she/​he may receive a review hearing before the board, with the time ranging from one (e.g., Nebraska) to ten years (e.g., California). A revocation hearing occurs when parole is revoked due to a violation of parole conditions. Usually, a parolee must return to custody before they are provided a review hearing to determine if the parole revocation is valid (e.g., Massachusetts) but revocation hearings can have a two-​step process involving a preliminary hearing followed by a hearing to consider mitigating circumstances (e.g., Texas). Generally, states are reluctant to provide counsel for any of these types of hearings (Cohen, 2014). The Supreme Court has previously held that an individual facing the possibility of parole, including after revocation, can be represented by counsel; this would be determined on a “case-​by-​case basis” (Gagnon v. Scarpelli [1973]). Essentially, Scarpelli allowed states to decide whether attorney representation should be provided for or even allowed at a parole hearing. Figure 6.5 shows that 16 state parole boards do not allow any type of legal representation during any type of parole hearing. Some states allow legal representation either at the initial hearing (13) or during the parole revocation hearing (4) but the costs are borne by the inmate. Eleven states guarantee the appointment of counsel for indigent juvenile lifers at their initial hearing. Seventeen states provide legal representation to indigent parolees whose parole is revoked, presumably to comply with 143

Stuti S. Kokkalera and Simon I. Singer 17

18 Number of Parole Authorities

16 16 14

13 11

12 10 8 6 4 4 2 0

No legal State provides Counsel appointed Counsel appointed State provides representation at cost of inmate legal at cost of inmate legal allowed during initial during parole representation at representation at parole revocation hearing revocation hearing initial parole hearing hearings

Figure 6.5  Access to legal representation for juvenile lifers during parole hearings

Morrissey (1972), Scarpelli (1973), and Greenholtz (1973). Additionally, revocation proceedings closely resemble a trial where there are complicated issues of fact and law and parolees are expected to defend their conduct on parole supervision. However, with the Supreme Court’s meaningful review mandate as expressed in Graham (2010) and Miller (2012), parole hearings may be characterized as trials where juvenile lifers must present their case for release. Therefore, providing legal counsel assists juvenile lifers to appropriately relate their circumstances to the parole board (Russell, 2014).

Burden of Proof and Right to a Written Decision State parole boards also vary in terms of the standard of persuasiveness or the burden of proof required for an offender to show his/​her readiness to reenter society. In a national survey, most parole authorities reported that there is no burden of proof requirement, since the parole hearing is administrative (Ruhland et al., 2016). Thirteen state parole boards reported that the burden of proof is preponderance of evidence and two reported the use of clear and convincing evidence for contested facts. No state based its discretionary release practices on the standard of proof beyond a reasonable doubt (Ruhland et al., 2016). State parole boards differ on voting procedures. Thirty-​six states decide based on majority vote, or two-​thirds majority. Seven states require a unanimous vote to grant parole for violent offenders and juvenile lifers. A few parole boards require the state Governor (e.g., California) or sentencing judge (e.g., Michigan) to affirm parole or use a system in which the board recommends and the Governor grants parole (e.g., Oklahoma). Some states require those granted parole to serve time in a lower security setting before being released on community supervision (e.g., Kentucky; Massachusetts). Parole boards also diverge on whether offenders are entitled to a written decision; 24 state parole boards provide written decisions to inmates, and seven states have no such rule (Ruhland et al., 2016). The upshot is that states vary enormously in their procedures for providing juvenile lifers with a meaningful review (Moriarty, 2017b).

Appeals Depending on a state’s parole statute and guidelines, juvenile lifers may or may not have the right to appeal a denial of parole. 144

Discretionary Release Practices 10 9 Number of Parole Authorities

9 8 7 6 5 5 4 3 2

2

2 1 0 Administrative Review Administrative Review Judicial Review of a Judicial Review of only of any Parole Board of only Parole Parole Board Decision Parole Revocation Decision Revocation Decision Decision

Figure 6.6  Review of decisions made by the parole board

Figure 6.6 shows that nine states allow administrative review of any parole decision (typically a denial of parole), while two states restrict an administrative review to the denial of parole after revocation. Seven states allow the judicial review of a parole board decision (either in a trial court or an appellate court) of which two states restrict judicial review to denials of parole following revocation hearings. Only Michigan allows the prosecutor to appeal the decision to grant parole. The US Supreme Court has not yet clarified if juvenile lifers have an automatic right to a judicial review of a parole denial. New York is among a handful of states that allow a judicial review of a parole board decision. Dempsey Hawkins’ case illustrates the necessity for judicial review of parole decisions that are inconsistent with Graham and Miller. In 2014, Hawkins challenged his latest denial for parole. At the age of 16, Hawkins strangled his 14-​year-​old-​g irlfriend to death. He was convicted of second-​degree murder and sentenced to 22 years to life. After his ninth parole hearing in 2014, the New York Board of Parole denied his request for release based on his crime. He appealed this decision to the appeals unit of the Board which failed to reconsider parole within the mandated period of four months. Subsequently, Hawkins filed a petition with the Supreme Court of Sullivan County (the trial court with jurisdiction), arguing that the parole board’s decision was arbitrary and capricious because it did not consider his age at the time of the crime. The court agreed with him and granted an order for a new parole hearing, but this was appealed by the parole board. The appellate division of the Supreme Court of New York (Third Judicial Division) agreed with the lower court’s decision to grant Hawkins a new hearing. The court concluded that the parole board did not provide Hawkins with a fair hearing because they did not evaluate his “transient immaturity” at the time of the crime (Hawkins v. New York Department of Corrections and Community Supervision [2016]). The appellate court further clarified that juvenile lifers, like Hawkins, should receive a parole hearing that is equivalent to a resentencing hearing, where a juvenile lifer’s youth and associated circumstances are fully considered by the Board. The right to appeal is critical to ensuring that juvenile lifers have the opportunity to address procedural and jurisdictional hurdles to a meaningful review (Russell, 2014). In 2016, the Court clarified that it could intervene in cases where resentencing courts overemphasize the severity of the offense without meaningfully considering the age of a juvenile lifer at the time of the crime (Tatum v. Arizona [2016]). Consequently, the next constitutional challenge for the Court could be the case of a juvenile lifer whose parole is denied due to the seriousness of the sentencing offense. Parole 145

Stuti S. Kokkalera and Simon I. Singer

guidelines that give the same or more weight to the facts of the offense ostensibly violate a meaningful review for juvenile lifers.

Conclusion The population of juvenile offenders serving life sentences is a product of state laws that have expanded the population of juvenile offenders eligible for these sentences. Policies that have created a population of juvenile lifers fall within the “get tough” legislative era, beginning in the late 1970s. The expansion is most notable with legislative wavier—​which places the decision to charge as an adult in the hands of prosecutors. Juvenile justice advocates would like to see no juvenile enter the criminal court. For those who recognize the need to punish in criminal court, waiver policies would take a middle course by being much more selective; only older juveniles who have a repeated history of violence would be subject to adult criminal prosecutions (Scott & Steinberg, 2008). Still the juvenile’s diminished culpability would preclude the possibility of an adult life sentence. But there is more that we could do in terms of recognizing the adolescence of juveniles in the adult criminal justice system. First, we need to recognize that there is a mishmash of concerns that are impacting the sentencing of juveniles to life. These concerns include the seriousness of the offense, and the desire to see the offender punish, no matter how young. The politics of popular punitiveness means that the victim, the victim’s family, and prosecutors have an impact on a juvenile lifer’s possibility of parole. Statutory guidelines repeatedly emphasize the responsibility of the parole board to examine the nature of the crime and an individual’s criminal history (Ball, 2011). We also need to bear in mind that the US Supreme Court has yet to decide whether imposing long mandatory sentences or cumulative sentences on juvenile offenders violates the Eighth Amendment’s prohibition against cruel and unusual punishment. Without a national consensus on whether parole-​ eligible life sentences or aggregate sentences for juveniles are constitutional, states can continue to impose these sentences without legally violating the mandates in Graham (2010), Miller (2012), and Montgomery (2016). State long-​term minimum sentencing policies are reproducing LWOP sentences, increasing the number of individuals serving “virtual life sentences.” We are also confronted with the issue of whether lifetime supervision on parole is necessary for juvenile lifers, especially when we know little about their risk to reoffend. States will continue to diverge on how they conduct a meaningful review. States are facing and will continue to face numerous appeals because their resentencing or parole board hearings neglect to recognize a juvenile lifer’s adolescence. We suggest the following as a way for policy makers to proceed in their state’s best interests. First, we need greater transparency in the decision-​making of parole boards. As mentioned, parole is administrative, an extension of the correctional system, and in many states the hearings are closed. We need to know more about how parole board members are deciding the fate of juveniles sentenced as adults. Comparative data that are more accurate than that which we could gather are required to assess discretionary release practices as they impact juvenile lifers. Second, we need to study and recommend policies that relate to discretionary release practices of parole boards and resentencing hearings of juvenile lifers. States might even consider a hybrid system in which elements of resentencing and parole board hearings are merged (Boone, 2015). The advantage of resentencing is that it involves a full due process court hearing where the adolescence of the juvenile lifer at the time of the offense is addressed explicitly.This explicit recognition is important to understanding the reasons why an adolescent would commit a serious and violent offense. The disadvantage of resentencing is that there is no consideration of the offender’s rehabilitative efforts. The advantage of the parole board hearing is that it is explicit in taking into account rehabilitation, and in turn reoffending risk.The parole board hearing also includes conditions and recommendations for 146

Discretionary Release Practices

reentry. However, due to its administrative nature, parole hearings are conducted with minimal due process protections. Finally, we know virtually nothing about the reentry process for juvenile lifers. We need to know more not only about the juvenile lifers who are able to be granted parole, but also those who are denied. The reasons for denial based on their inability to meet the rehabilitative standards of the parole board should be considered.The failure of lifers to make parole may be related to the failure of prisons to provide the treatment-​oriented services that would allow for the kind growth and maturity that would lead to law abiding conduct in prison and then beyond. When conditions of parole are established by a parole board, they may not reflect the specific needs of juvenile lifers (Mulvey & Schubert, 2012). Difficulties in reentry are often the result of conditions that are unrealistic and challenging to comply with (Solomon et al., 2008), and these conditions are more pronounced for juvenile lifers. This is a growing population of individuals returning to communities after many years of confinement. Examining their reentry needs should be a priority for state policymakers, and this should not be restricted to those who received life, but also to all juveniles who have been sentenced to adult maximums.

Notes 1 Connecticut legislature passed the Public Acts 15–​84 of 2015 titled “An Act Concerning Lengthy Sentences for Crimes Committed by a Child or Youth and the Sentencing of a child or Youth Convicted of Certain Felony Offenses”. 2 SB 850 (PA) in 2012 provides for two parole-​eligible options: For a conviction of 1st degree murder, (LWOP is a discretionary option) a sentence of life with a minimum of 35 years to life for youth ages 15 to 17 and a minimum of 25 years to life for youth under age 15. For a conviction of 2nd degree murder, (LWOP is a discretionary option) a sentence of life with a minimum of 30 years to life for youth ages 15 to 17 and a minimum of 20 years to life for youth under age 15 (Kirby, 2015). 3 We conducted a search on Google Scholar and Northeastern University’s ScholarOne database with the key terms “early discharge” AND “parole” AND “juvenile lifer”. 4 The wording of the Nebraska statute created an expectation of parole rather than a hope of release by using the words “shall grant parole” (D.B.H, 1980).

References Allen, F. (1978). Decline of the Rehabilitative Ideal—​ Penal Policy and Social Purpose. New Haven, CT: Yale University Press. Arya, N. (2011). State Trends: Legislative Victories from 2005 to 2010 Removing Youth from the Adult Criminal Justice System. Washington, DC: Campaign for Youth Justice. The Associated Press. (2017, July 31). 50-​state examination. The Associated Press. Retrieved from: www.ap.org/​ explore/​locked-​up-​for-​life/​50-​states Ball, D.W. (2011). Normative elements of parole risk. Stanford Law & Policy Review, 22(2), 395–​411. Bell, K. (2018). A Stone of Hope: Legal and Empirical Analysis of California Juvenile Lifer Parole Decisions. Forthcoming in Harvard Civil Rights-Civil Liberties Law Review. Available at http://dx.doi.org/10.2139/ ssrn.3228681 Bishop, D., & Feld, B.C. (2014). Juvenile justice in the get tough era. In G. Bruinsma & D. Weisburd (eds.), Encyclopedia of Criminology and Criminal Justice (pp. 2766–​2773). New York: Springer. Bishop, D., & Frazier C. (2000). The consequences of transfer. In J. Fagan & F. Zimring (eds.), The Changing Borders of Juvenile Justice: Transfer of Adolescents to the Criminal Court (pp. 227–​276). Chicago, IL: University of Chicago Press. Boone, B.H. (2015). Treating adults like children: Resentencing adult juvenile lifers after Miller v. Alabama. Minnesota Law Review, 99, 1159–​1194. Bottoms, A.E. (1995). The philosophy and politics of punishment and sentencing. In C. M.  V Clarkson & R. Morgan (eds.), The Politics of Sentencing Reform (pp. 17–​49). Oxford, UK: Oxford University Press. Bridges, G.S., & Steen, S. (1998). Racial disparities in official assessments of juvenile offenders: Attributional stereotypes as mediating mechanisms. American Sociological Review, 63, 554–​570. Brown, S.A. (2012). Trends in Juvenile Justice Legislation: 2001–​2011. Washington, DC: National Conference of State Legislatures.

147

Stuti S. Kokkalera and Simon I. Singer Caldwell, B. (2016). Creating meaningful opportunities for release: Graham, Miller and California’s youth offender parole hearings. N.Y.U Review of Law and Social Change, 40, 245–​304. Caplan, J. (2010). Parole release decisions: Impact of victim input on a representative sample of inmates. Journal of Criminal Justice, 38(3), 291–​300. —​—​. (2012). Protecting parole board legitimacy in the twenty-​first century: The role of victims’ rights and influences. Victims and Offenders, 7(1),  53–​76. Carroll, J. (1978). Causal attributions in expert parole decisions. Journal of Personality and Social Psychology, 36(12), 1501–​1511. Clear, T.R., & Frost, N.A. (2013). The Punishment Imperative: The Rise and Failure of Mass Incarceration in America. New York: NYU Press. Cohen, L. (2014). Freedom’s road: Youth, parole, and the promise of Miller v. Alabama and Graham v. Florida. Cardozo Law Review, 35, 1031–​1089. D.B.H. (1980). Procedural due process in parole rescission hearings. Virginia Law Review, 66(4), 779–​795. Drinan, C.H. (2017). The War on Kids: How American Juvenile Justice Lost its Way. New York: Oxford University Press. Emerson, R.M. (1969). Judging Delinquents; Context and Process in Juvenile Court. Chicago, IL: Aldine Pub. Co. —​—​. (1983). Holistic effects in social control decision-​making. Law & Society Review, 17(3), 425–​455. —​—​. (1991). Case processing and interorganizational knowledge: Detecting the ‘real reasons’ for referrals. Social Problems, 38(2), 198–​212. EJI. (2015, July 2). Connecticut abolishes juvenile life-​without-​parole sentences. Equal Justice Initiative. Retrieved from: https://​eji.org/​news/​connecticut-​abolishes-​juvenile-​life-​without-​parole Feld, B.C. (1993). Criminalizing the American juvenile court. In M. Tonry (ed.), Crime and Justice: An Annual Review of Research (pp. 197–​280). Chicago, IL: The University of Chicago Press. —​—​ . (1999). Bad Kids: Race and the Transformation of the Juvenile Court. New York: Oxford University Press. Feld, B.C. (2017). The Evolution of the Juvenile Court: Race, Politics and the Criminalizing of the Juvenile Justice. New York: NYU Press. Feld, B.C., & Bishop, D.M. (2012). Transfer of juveniles to criminal court. In B.C. Feld & D.M. Bishop (eds.), The Oxford Handbook of Juvenile Crime and Juvenile Justice (pp. 801–​842). New York: Oxford University Press. Feeley, M.M., & Simon, J. (1992). The new penology: Notes on the emerging strategy of corrections and its implications. Criminology, 30(4), 449–​499. Ford, M. (2017, July 14). The reckoning over young prisoners serving life without parole. The Atlantic. Retrieved from: www.theatlantic.com/​politics/​archive/​2017/​07/​juvenile-​life-​without-​parole/​533157/​ Garland, D. (2013). The 2012 Sutherland address: Penalty and the penal state. Criminology, 51(3), 475–​517. Gobeil, R., & Serin, R.C. (2009). Preliminary evidence of adaptive decision-​making techniques used by parole board members. International Journal of Forensic Mental Health, 8(2), 97–​104. Griffin, P., Addie, S., Adams, B., & Firestine, K. (2011). Trying juveniles as adults: An analysis of state transfer laws and reporting. Juvenile offenders and victims: National report series bulletin. Washington, DC: OJJDP. Hawkins, K. (1986). Discretion in making legal decisions. Washington & Lee Law Review, 43, 1161–​1195. Hilliard, E. (2016). A life without: Juveniles spending their lives in Oregon prisons and the need for change following Miller and Graham. Lewis and Clark Law Review, 20(1), 331–​379. Holland, T., Holt, N., & Brewer, D. (1978). Social roles and information utilization in parole decision-​making. Journal of Social Psychology, 106(1), 111–​120. Huebner, B.M., & Bynum, T.S. (2008). The role of race and ethnicity in parole decisions. Criminology, 46(4), 907–​938. Kinnevy, S.C., & Caplan, J.M. (2008). Findings from the APAI International Survey of Releasing Authorities. Philadelphia, PA: Center for Research on Youth and Social Policy, University of Pennsylvania. Kirby, M. (2015). Juvenile Sentencing Laws and Court Decisions after Miller v. Alabama. Hartford, CT: Connecticut Legislative Assembly Office of Legislative Research. Latessa, E.J., & Lovins, B. (2010).The role of offender risk assessment: A policy maker guide. Victims and Offenders, 5(3), 203–​219. Levick, M.L., & Schwartz, R.G. (2013). Practical implications of Miller v. Jackson: Obtaining relief in court and before the parole board. Law & Inequality: A Journal of Theory and Practice, 31(2), 369–​409. Lynch, M. (2016). Hard Bargains: The Coercive Power of Drug Laws in Federal Court. New  York: Russell Sage Foundation. Maruna, S., & Toch, H. (2005). The impact of imprisonment on the desistance process. In J. Travis & C. Visher (eds.), Prisoner Reentry and Crime in America (pp. 139–​179). New York: Cambridge University Press. Martinson, R. (1974). What works? Questions and answers about prison reform. The Public Interest, 35,  22–​54. Mills, J.R., Dorn, A.M., & Hritz, A.C. (2015). Juvenile life without parole in law and practice: Chronicling the rapid change underway. American University Law Review, 65, 535–​605. Monahan, L., & Kaban, B. (2009). Until They Die a Natural Death: Youth Sentenced to Life without Parole in Massachusetts. Boston, MA: Children’s Law Center of Massachusetts.

148

Discretionary Release Practices Morgan, K.D., & Smith, B. (2005). Parole release decisions revisited: An analysis of parole release decisions for violent inmates in a southeastern state. Journal of Criminal Justice, 33(3), 277–​287. Moriarty, P.L. (2017a). The trilogy and beyond. South Dakota Law Review, 62, 539–​558. —​—​. (2017b). Implementing proportionality. U.C. Davis Law Review, 50, 961–​1028. Mulvey, E.P., & Schubert, C.A. (2012).Youth in prison and beyond. In B.C. Feld & D.M Bishop (eds.), The Oxford Handbook of Juvenile Crime and Juvenile Justice (pp. 843–​870). New York: Oxford University Press. National Conference of State Legislatures (NCSL). (2010). Juvenile life without parole [Pamphlet]. Retrieved from: www.ncsl.org/​documents/​cj/​jlwopchart.pdf Nellis, A. (2012). The Lives of Juvenile Lifers: Findings from a National Survey. Washington, DC: The Sentencing Project. Nellis, A., & King, R.S. (2009). No Exit: The Expanding Use of Life Sentences in America. Washington, DC: The Sentencing Project. Office of Juvenile Justice and Delinquency Prevention [OJJDP]. (2015). Easy Access to Juvenile Court Statistics [EZAJCS]. Retrieved from: www.ojjdp.gov/​ojstatbb/​ezajcs/​ Ogletree, C.J., & Sarat, A. (2012). Life without Parole: America’s New Death Penalty? New York: NYU Press. Proctor, J.L. (1999). The “new parole”: An analysis of parole decision-​making as a function of eligibility. Journal of Crime and Justice, 22(2), 193–​217. Rovner, J. (2014). Slow to Act: State Responses to 2012 Court Mandate on Life without Parole. Washington, DC: The Sentencing Project. Rovner, J. (2017). Juvenile Life without Parole: An Overview. Washington, DC: The Sentencing Project. Ruhland, E.L., Rhine, E.E., Robey, J.P., & Mitchell, K.L. (2016). The Continuing Leverage of Release Authorities: Findings from a National Survey. Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice, University of Minnesota. Russell, S.F. (2014). Review for release: Juvenile offenders, state parole practices and the Eighth Amendment. Indiana Law Journal, 89, 373–​440. —​—​. (2016). The role of crime at juvenile parole hearings: A response to Beth Caldwell’s Creating Meaningful Opportunities for Release. The Harbinger, 41, 227–​232. Scott, E.S., & Steinberg, L. (2008). Rethinking Juvenile Justice. Cambridge MA: Harvard University Press. Scott, E.S. (2000). Criminal responsibility in adolescence: Lessons from developmental psychology. In T. Grisso & R.G. Schwartz (eds.), Youth on Trial: A Developmental Perspective on Juvenile Justice (pp. 291–​324). Chicago, IL: The University of Chicago Press. Scott, E.S., Thomas, G., Levick, M., & Steinberg, L. (2016). Juvenile sentencing reform in constitutional framework. Temple Law Review, 88, 675–​716. Simon, J. (1993). Poor Discipline: Parole and the Social Control of the Underclass, 1890–​1990. Chicago, IL: The University of Chicago Press. Singer, S.I. (1996). Recriminalizing Delinquency: Violent Juvenile Crime and Juvenile Justice Reform. New  York: Cambridge University Press. —​—​. (2003). Incarcerating juveniles into adulthood: Organizational fields of knowledge and the back end of waiver. Youth Violence and Juvenile Justice, 1(2), 115–​127. —​—​. (2017, October 16). Parole boards treat adolescents who grow up in prisons like adults—​and that’s wrong. Juvenile Justice Information Exchange. Retrieved from: http://​jjie.org/​2017/​10/​16/​parole-​boards-​ treat-​adolescents-​who-​grow-​up-​in-​prison-​like-​adults-​and-​thats-​wrong/​ Solomon, A.L., Osborne, J.W.L., Winterfield L., Elderbroom, B., Burke, P., Stroker, R.P., Rhine, E.E., & Burrell, W.D. (2008). Putting Public Safety First: 13 Parole Supervision Strategies to Enhance Reentry Outcomes.Washington, DC: The Urban Institute. Tidmarsh, D. (1997). Risk assessment among prisoners: A view from a parole board member. International Review of Psychiatry, 9, 273–​282. Wheeler, L. (2018, April 23). Supreme Court refuses to hear appeal of teen sentenced to 241  years in prison. The Hill. Retrieved from: http://​thehill.com/​regulation/​court-​battles/​384401-​supreme-​court​refuses-​to-​hear-​appeal-​of-​teen-​sentenced-​to-​241-​years Zimring, F.E. (2010). The power politics of juvenile court transfer: A mildly revisionist history of the 1990s. Louisiana Law Review, 71,  1–​15. Zimring, F.E. (2005). American Juvenile Justice. New York: Oxford University Press.

Cases Cited Alaska Code of Criminal Procedure, 12 ALASKA STAT. §55.125 (2008) Alaska Code of Criminal Procedure, 12 ALASKA STAT. §55.155, 165 (2008) Arizona Criminal Code, ARIZ. REV. STAT. ANN. §13–​751 (2015)

149

Stuti S. Kokkalera and Simon I. Singer Commonwealth v. Okoro, 471 Mass. 51 (2015) Commonwealth v. Perez, 477 Mass. 677 (2017) Diatchenko v. District Attorney for Suffolk District (Diatchenko I), 466 Mass. 655 (2013) Gagnon v. Scarpelli, 411 U.S. 778 (1973) Graham v. Florida, 560 U.S. 48 (2010) Greenholtz v. Inmates of the Nebraska Penal and Correctional Complex, 442 U.S. 1 (1979) Hawkins v. New York Department of Corrections and Community Supervision, 140 A.D.3d 34 (N.Y., 2016) In Re Gault, 387 U.S. 1 (1967) In Re Winship, 397 U.S. 358 (1970) Crimes and Punishments, KAN. STAT. ANN. § 21–​6625 [2015] Kent v. U.S., 383 US 541 (1966) Lasley v. State, 2017 Ark. 311 Luna v. State, 387 P.3d 956 (Oklahoma 2016) Miller v. Alabama, 567 U.S. 460 (2012) Minnesota v. Mahdi Hassan, 895 N.W.2d 237 (Minn., 2017) Montgomery v. Louisiana, 136 S. Ct. 718 (2016) Morrissey v. Brewer, 408 U.S. 471 (1972) Parker v. State, 119 So.3d 987 (Miss. 2013) Perry v. State, 2014 Tenn. Crim. App. LEXIS 327 (Tenn. Crim. App. Apr. 7, 2014) Prisons, Imprisonment, Paroles and Pardon, MASS. GEN. LAWS. c. 127, §130 [2015] Roper v. Simmons, 543 U.S. 551 (2005) State v. Hardman, No. 60CR-​00-​1457, Ark. 2017 State v. Lyle, 854 N.W.2d 378 (Iowa 2014) State v. Mantich, 287 Neb. 320 (2014) State v. Ragland, 836 N.W.2d. 107 (2013) State v. Riley, 315 Conn. 637 (2015) State v. Zarate, 908 N.W.2d 831 (Iowa 2018) Tatum v. Arizona, 137 S.Ct. 11(2016)

Statutes Cited The Fair Sentencing of Minors Act, 2015 Ark. Acts. 539

150

PART III

Plea Bargaining

7 PLEA NEGOTIATIONS An Australian Perspective1 Arie Freiberg and Asher Flynn

Introduction That the criminal trial has long been a vanishing phenomenon is an open secret. That most criminal proceedings culminate in a plea of guilty that is the result of negotiations between the prosecution and defense is also well known, at least amongst those closely involved in the criminal justice system. Plea negotiations, plea bargaining, charge bargaining or negotiated resolutions are an everyday form of justice for most accused persons, an accepted element of the criminal justice process in many common law jurisdictions and Australia is no exception. Plea negotiation, which can be understood as a process whereby a defendant will plead guilty in exchange for agreed concessions from the prosecution, has long been the subject of trenchant critiques (Johns, 2002; McConville & Mirsky, 2005; McConville & Marsh, 2014; Woolard, Henning, & Fountain, 2016). It has been argued that negotiated guilty pleas have the potential to undermine fundamental criminal justice principles including the presumption of innocence and the defendant’s right to a contested trial. As McConville and Marsh (2014) maintain, “because the process of State-​ induced guilty pleas is intended to replace in whole or part the promise of adversary justice, it challenges the foundations on which the claim of legitimacy of criminal courts has traditionally rested” (p. 216). Concerns relating to plea negotiations have heightened in light of recent, ongoing austerity measures that have seen a reduction or shift in funding for government and state-​funded legal services, and a greater focus on increased efficiency and timely finalizations in the courts more generally (Flynn & Hodgson, 2017a, 2017b). In this context, the criminal justice system has become focused “on the immediate case facts and outcomes, with less time and attention reserved for discussions about rights, or long-​term collateral consequences” (Fountain & Woolard, 2017, p. 1). It has also been argued that guilty pleas resulting from a negotiation process are problematic because they trade the contested trial, with its strict rules of procedure, for an informal method of case resolution, thereby removing the public’s access to a transparent process and creating uncertainty and misunderstanding about “justice” outcomes (Buckle & Buckle, 1977; Euvard & Leclerc, 2017; Flynn, 2016; Westling, 1976). Plea negotiations are said to be unfair because they create power imbalances between the prosecutor and the defendant, particularly where the possible benefits of the plea negotiation may, and generally will, outweigh the risks of contesting the case at trial. In this regard, it has been argued that defendants may plead guilty because they feel coerced to do so, with pressures arising from the threat of a more severe sentence, from the need to have the matter resolved 153

Arie Freiberg and Asher Flynn

speedily, and from the promise of a more certain outcome, compared with that of a trial. In other words, a plea of guilty is a more expedient way to finalize one’s case, enabling the defendant to not only resolve the process, but to save money and get back to their “normal” life (Redlich, Wilford, & Bushway, 2017, p. 464). Yet the persistence and pervasiveness of plea negotiations points to an underlying utility and logic to the practice; to a system that may be better described as “managerial” rather than “adversarial” (Freiberg, 2005). Considerations such as cost savings, reductions in court backlogs, and greater certainty of outcomes for both prosecution and defense have substantial force. Indeed, it might be argued that most defendants plead guilty because they are, in fact, guilty, meaning that a plea negotiation that results in a reduced sentence may be regarded as a better, or “the best”, outcome. It has also been argued that plea negotiation can free up court resources for trials where there is a real contest, possibly improving the quality of justice in those cases (Gazal-​Ayal & Tor, 2012, p. 389). In some jurisdictions, the development of non-​adversarial processes and problem-​oriented courts such as drug courts and mental health courts may mean that a guilty plea produces a more expeditious and therapeutically oriented result to the benefit of both the offender and ultimately the community. Extensively studied in the United States (US), Canada and the United Kingdom (UK), less is known of the nature and extent of plea negotiations in Australia. While there are many commonalities between these jurisdictions, each is unique in terms of its legal and jurisprudential contexts and cultural practices. In Australia, plea negotiations are not of the kind depicted in fictionalized American television dramas in which plea deals are done and presented as fait accompli to the court, but are part of everyday legal life in what is essentially becoming a semi-​adversarial criminal justice system. This chapter broadly describes plea negotiations in Australia, but with particular emphasis on the state of Victoria—​the second-​most populated state in Australia, comprising approximately 6.3 million people. It reports on a study of plea negotiations undertaken between 2014 and 2017 which sought to shed light on the process by providing an empirical account of plea negotiation practices, including documenting the frequency of plea negotiations, their different forms, the common outcomes and the processes involved in the parties reaching an outcome. It involved obtaining a dataset of 50 de-​identified cases from Victoria Legal Aid, the primary provider of state-​funded legal services in the state, who handle approximately 80 per cent of all criminal cases in Victoria (Flynn, 2010a; PricewaterhouseCoopers, 2008), as well as undertaking 48 qualitative interviews with police prosecutors (who handle the less serious cases) and the Office of Public Prosecutions’ (OPP) solicitors and prosecutors (who handle indictable offenses) (n = 10), defense practitioners from state-​funded and private practice (n = 25) and judicial officers (n = 13). A two-​hour consultation workshop with key legal and policy stakeholders was also held (n = 15).2

The Legal Context In Victoria, the terms “negotiations,” “early resolutions,” or “settlements” are preferred to “plea bargaining” due to the latter’s perceived pejorative undertones, its “American connotations” and its implications of inappropriate secretiveness, unfair deals, and undue pressure being placed on defendants (Flynn & Freiberg, 2018). Fox and Deltondo (2015) suggest that “the phrase ‘plea bargaining’ fails to adequately represent the complexity of the processes that may lead representatives of the prosecution and defense … to engage in discussions aimed at the early resolution of a criminal case”. For most participants in the current study, the term “plea bargaining” was similarly considered as failing to capture the many nuances inherent in the plea negotiation process. Both the prosecutor and defense have a role in shaping plea negotiations in Australia. It is for the prosecutor to decide the charges upon which to proceed and, although his or her discretion may not be absolute, it is large. The defense must take into consideration the best interests of the defendant, and work towards a resolution that reflects this usually in the form of reduced charges, fewer charges 154

Plea Negotiations

and a “watered down” (Flynn & Freiberg, 2018) summary of facts. Courts have a limited ability to ensure that the appropriate charges have been laid. In a resource-​constrained criminal justice system, a judge’s ability to determine the “true” facts will be very limited. Where courts face a situation where both parties have agreed upon an outcome, but where that outcome may not represent the interests of all those concerned, such as the victim, a court must take special care to ensure that it is being given “accurate, reliable and complete information” with regard to relevant facts in order to ensure that public confidence in the administration of criminal justice is not diminished.3 Plea negotiation has been justified by the Victorian prosecution authority, the Director of Public Prosecutions (DPP), on the basis that it relieves victims and witnesses of the burden of having to give evidence and may help victims put their experience behind them. It provides certainty of outcome and saves the community the cost of trials. (DPP n.d., p. 21) The DPP’s policy, which deals specifically with the “resolution” of cases, states that resolution is “necessary for the effective and efficient conduct of prosecutions” (p. 21). Among the factors identified as being relevant to achieving resolution of a case, the “public interest” is the overarching consideration. In determining whether a proposed resolution is in the public interest, regard must be had to: (a) whether there is a reasonable prospect of a conviction on each charge. If there is no reasonable prospect of a conviction on a charge, then that charge must not proceed. It is improper for such a charge to proceed to committal with a view to the prospects of conviction being reassessed after the committal. (b) the strength of the evidence on each charge. (c) any defenses. (d) the likelihood of an acquittal on any of the charges. (e) whether the charge or charges to which the accused will plead guilty: i. adequately reflect the accused’s criminality ii. allow for the imposition of an appropriate sentence iii. allow for the making of all appropriate ancillary orders. (f) the views of the victims and the informant. (g) the need to minimize inconvenience and stress to witnesses, particularly those who may find it onerous to give evidence; and (h) the likely length of a trial. Prosecutors must also ensure that the charge, or charges, to which the accused will plead guilty as part of a resolution4 will adequately reflect the accused’s criminality, allow for the imposition of an appropriate sentence and allow for the making of all appropriate ancillary orders (DPP n.d., p. 21). In Australia, regardless of whether the outcome involves a guilty finding by judge or jury, or a guilty plea (with or without plea negotiations), the sentence imposed remains the responsibility of the sentencing judge alone (Freiberg 2014, p. 107). No agreement between the prosecution and defense as to sentence can bind the sentencing judge, nor any appeal court, although the prosecution may limit the sentencing options available to the judge by determining the number and seriousness of the charges presented.5 A firm reminder of this occurred in R v. Williams.6 In this somewhat unusual case, a plea negotiation had been arranged between the prosecutor and the defendant’s son, a notorious underworld figure, in an unrelated case, which included an agreement that the prosecutor in Williams would recommend a wholly suspended sentence for the defendant.This agreement was made in light of the defendant’s ill health and as recognition of the resource savings for the community, the courts and the prosecutor’s office from both the guilty plea to drug offenses and his son’s guilty pleas to 155

Arie Freiberg and Asher Flynn

several murders and conspiracy to commit murder charges. In reflecting on this agreement, the sentencing judge, Justice Betty King (King J) of the Victorian Supreme Court observed: There is no doubt that it is important that the court uphold, if possible, any concessions or so called “deals” that are made between the Crown and the defence, if they are proper and appropriate concessions. Here, there is no doubt that the Crown were extremely anxious to conclude the potentially lengthy and costly series of trials that were being conducted in respect of these underworld murders, and there is no doubt that those involved in the prosecution of those matters determined that it was appropriate to make the concession.7 However, the Justice went on to state that: [I]‌t is ultimately for the court to determine what sentence is appropriate. The Court is not bound, in any way, by the negotiations between counsel as to the appropriateness or otherwise of a sentence. That is clearly the role and function of the court …8 For reasons that I shall set out shortly, I do not agree with the Crown and defence submission that a wholly suspended term of imprisonment would be appropriate in this case.9 King J sentenced the defendant to four and a half years’ imprisonment, with a minimum non-​parole period of 20 months, a sentence significantly longer in duration and more severe in nature than that “agreed” to by the negotiating parties. Additionally, no agreement made between parties can prevent the Director of Public Prosecutions (DPP) (acting in the public interest) from appealing against a sentence imposed.10 In Malvaso, the High Court of Australia, the country’s ultimate appellate authority, stated that a plea negotiation did not affect: the duty of either of the sentencing judge or of the Court of Criminal Appeal … to impose the sentence which appeared appropriate to the Court in the circumstances. The Court’s sentencing discretion is to be exercised in the public interest; it cannot be fettered by a plea-​ bargaining agreement.11 Indeed, in Australia, following a brief four year period where prosecutors were required to submit what they considered to be an appropriate sentencing range if “(a) the court requests such assistance; or (b) even though no such request has been made, the prosecutor perceives a significant risk that the court will fall into error regarding the applicable range unless such a submission is made,”12 the High Court ruled that the prosecution may not make submissions to the court regarding the bounds of the available range of sentences, holding that such practice is wrong in principle.13 Consequently, the negotiated guilty plea can, at best, be an influential factor in sentencing, but never a determining one, with the power ultimately lying with the sentencing judge. In most of the larger Australian jurisdictions, including Victoria, there are three levels of court that hear criminal cases. The Magistrates’ Court, or court of summary jurisdiction, is comprised of a single magistrate sitting without a jury.The magistrate hears summary, or relatively minor offenses, or indictable offenses which are triable summarily. The length of prison sentence that can be imposed is relatively limited—​two years for a single offense and five years for multiple offenses (in Victoria). These courts determine over 90 per cent of criminal cases in Australia. At the next level, the County or District Courts (intermediate courts) hear indictable offenses, and finally, the Supreme Court hears the most serious indictable cases, such as murder and serious sexual and drug offenses. These courts sit with juries but determine only around five per cent of all criminal cases.14 In Victoria, there is a single police force for the whole state. The different levels of court have very different processes of prosecution, but prosecution process across all three court levels is strongly 156

Plea Negotiations

oriented towards early resolution. Summary prosecutions (in the Magistrates’ Court) are conducted by police prosecutors. Some, but not all, have a law degree.The prosecution process begins with what is known as a summary case conference and a contest mention. The summary case conference was introduced with the aim of encouraging out-​of-​court communication between the prosecution and the defendant (or their representative) in order to manage the progression of the case (for example, to avoid unnecessary adjournments or resolve matters at an early stage). The conference is intended to help the defendant understand the evidence available to the prosecution, enable the parties to identify any issues in dispute and have the parties identify the steps required to advance the case, which includes the possibility of an early resolution (Flynn, 2010a). The conference does not need to occur in the court setting and instead takes place by email, telephone or face-​to-​face meetings. There are summary case conference police prosecutors whose role is only to engage in these discussions. The contest mention involves a magistrate, defense practitioner and police prosecutor meeting in a relatively informal manner, in court, to identify any matters or issues that could be resolved, without having to proceed to a contested hearing. The key aims of the contest mention are to refine the case issues and, where relevant, to estimate the likely length of a contested hearing (Flynn, 2010a). The magistrate can also provide a sentence indication at this hearing if requested by the defense (Flynn, 2009), or as the current study found, at the discretion of the magistrate with or without a specific request from the defense (see Flynn & Freiberg, 2018). In the higher courts, once the police charge a defendant, carriage of the file is transferred to the OPP who are responsible for prosecuting matters in the indictable stream. The OPP is established under statute and is an entity separate and independent of government. Crown prosecutors, who appear in court, are barristers appointed to work for the DPP by the government of the day. Although they are co-​located with the OPP, they are not its employees. Unlike many American states, prosecutors (like judicial officers) are not elected officers and are required to act independently, impartially and fairly to the accused. There are a number of steps in the indictable prosecution process which encourage or facilitate early resolution. Some—​for example, the committal mention, the case conference, the initial directions hearing, among others—​are presided over by judicial officers who are encouraged to discuss and resolve key case issues, with a view towards early resolution (where possible) (Flynn & Freiberg, 2018). Another (rarely used) procedure in the indictable stream is the sentence indication hearing, in which a defendant (subject to prosecutorial approval) can request an indication of whether they would receive a custodial or non-​custodial sentence were they to enter a guilty plea, although an indication may not be given about the quantum of any sentence likely to be imposed (Flynn, 2009; 2010b; Flynn & Freiberg, 2018). The adversarial paradigm that is supposed to characterize the criminal justice process is therefore some way from the reality of every day practice. In Victoria, a culture of resolution has slowly developed in the OPP over recent years. This may partly be a response to increased caseloads and decreasing resources, but it may also be partly due to a recognition of the fact that most cases can, and should, be resolved before trial. In Victoria, this culture was institutionalized in the mid-​2000s through the establishment of an “early resolution unit” whose function was to proactively identify and finalize matters that could be resolved prior to trial. Early Resolution Advocates were responsible for managing most negotiations with defense practitioners in the first instance.The unit was disbanded in 2016, with the DPP being of the view that seeking early resolution was the function of every person involved in the prosecution process. Early resolution is now regarded not as a specialist function, but as part of the culture and every day processes of the OPP, at least for the majority of OPP employees.

Sentencing and Plea Negotiations There is a statutory incentive to plead guilty and to plead guilty early in Australia. Most jurisdictions have provisions that state that a court must have regard to whether a defendant pleaded guilty and 157

Arie Freiberg and Asher Flynn

the stage of proceedings at which the offender did so, or intended to do so (Tasmanian Sentencing Advisory Council, 2018, p. 40). Earlier pleas “should” receive a greater discount than later pleas. The discount may be provided on utilitarian grounds (that is, regardless of whether the offender shows remorse), although a greater discount may be awarded if true remorse is shown.The discount may be in the form of a lesser sentence of imprisonment or a different form of sentence, such as a suspended sentence rather than an executed one, or a non-​custodial sentence rather than a custodial one. In some jurisdictions, the discount is not prescribed, while in others, a sliding scale model has been adopted which links the extent of the discount to a specified time in the pretrial process (Tasmanian Sentencing Advisory Council, 2018, p. 40). There is evidence that overall, in Victoria, the discount amounts to around a 25–​30% reduction in imprisonment terms and in around one-​third of cases, the nature of the sentence changes (Victorian Sentencing Advisory Council, 2015). A live question in Australia is whether a defendant should receive what is effectively a “double discount” for a plea of guilty that has been the product of a plea negotiation. The argument is that if the plea negotiation has resulted in less severe, or fewer charges, being pleaded to, which are likely to result in a lower sentence, should there then be a further reduction on account of the guilty plea (Tasmanian Sentencing Advisory Council, 2018, p. 76)? It is estimated that between 70 and 80% of cases in the intermediate courts are resolved by the defendant entering a guilty plea (Victorian Sentencing Advisory Council, 2015, p. 11), a rate that has grown from the low 60s over the last decade or so. In the state of Tasmania, for example, between 2011 and 2016, 78.6% of defendants in the Supreme Court, in cases where the defendant was proven guilty were finalized by a guilty plea, while in the Magistrates Court, 86.2% were finalized by guilty plea (Tasmanian Sentencing Advisory Council, 2018, p.  8). Only five% of all cases in the Tasmanian Supreme Court were finalized by an acquittal. While it is possible to monitor the rate of guilty pleas, there remains very limited information as to how or why such pleas are entered, and it appears in most jurisdictions that there are no official data available documenting the role of plea negotiations in facilitating guilty pleas. However, based on the analysis of de-​identified files provided by Victorian legal aid, the current study found that approximately 87% of cases for which there was sufficient information to determine the charges pre and post the entering of a guilty plea involved some form of negotiation.The interview data shed further light on the frequency of plea negotiations in Victoria, extending beyond this initial estimate to suggest it occurs in between 90 and 100% of cases (see Flynn & Freiberg, 2018).Thus, the question of whether a “double discount” should apply would be a relevant issue in most, if not all, cases resolved by a guilty plea in Victoria (and most likely, many Western jurisdictions). In only one Australian jurisdiction, the Australian Capital Territory, the legislation provides that in sentencing following a guilty plea, the court may take into account “whether the guilty plea was related to negotiations between the prosecution and defense about the charge to which the offender pleaded guilty” (Crimes (Sentencing) Act 2005 (ACT), s 35(2)(c)). This provision is unusual in that it recognizes the existence of plea negotiations but the language in which it is framed does not indicate specifically whether its application should inevitably result in a moderation of the amount of the discount. Sentencing guidelines of the kind found in the US federal and some state systems that have the effect of limiting judicial discretion do not operate in Australia. General guidance can be found in comprehensive sentencing statutes that set out, in broad terms, the aims and purposes of sentencing, relevant factors that may or must be taken into account, and the sanctions that may be imposed (e.g., Sentencing Act 1991 (Vic)).The prevailing sentencing methodology is known as an instinctive or intuitive synthesis which requires a judge to identify: all the factors that are relevant to the sentence, discuss … their significance and then make … a value judgment as to what is the appropriate sentence given all the factors in the case. Only at the end of the process does the judge determine the sentence.15 158

Plea Negotiations

This method makes it difficult to predict the outcome of either a trial or plea of guilty. However, over recent years, various Australian legislatures have increasingly introduced mandatory or presumptive sentences, the levels of which have usually been much more severe than existing sentencing practices. For example, during the data collection period for the current study, there were two main presumptive sentencing policies in effect in Victoria. The first was baseline sentences, which set a median sentence for seven nominated offenses (Flynn & Freiberg, 2018), and the second was the Crimes Amendment (Gross Violence Offences) Act 2013 (Vic), which introduced two new offenses into s 15A and s 15B of the Crimes Act 1958 (Vic)—​intentionally causing serious injury when committed with gross violence, and recklessly causing serious injury when committed with gross violence—​and amended the Sentencing Act 1991 (Vic) to require that a statutory minimum non-​parole period of four years’ imprisonment be applied to adult defendants and a two-​year minimum detention sentence be applied to juvenile defendants (aged 16 or 17 years) found guilty (by plea or trial) of either of these offenses.16 Faced with such provisions and the constrictions on judicial discretion that they entail, together with the prospect of lengthy terms of imprisonment, defendants are likely to be less inclined to seek an agreement with the prosecution in favor of going to trial and taking their chances for an acquittal. Evidence from the interviews conducted for the current study also found that in the face of mandatory sentences, defendants were tempted, or pressured, to plead guilty to a lesser charge that did not carry a mandatory sentence. However, there was little evidence to indicate that defendants pleaded guilty, when they were not, as a result of such pressures, although such dangers are ever present. The interview data also indicated that for prosecutors, the greater reluctance of defendants to plead to the higher charge meant that they either needed to run more trials, or accept guilty pleas to lesser alternative offenses. As the Law Council of Australia (2014, p. 35) has argued, because “mandatory sentencing offenses may result in more contested cases, prosecutors and police may feel additional pressure to negotiate with the offender and/​or the defense and agree to pursue lesser charges to prevent court delay and a backlog of cases”. These potential outcomes raise serious questions about the effects of mandatory sentencing regimes, and how they may affect the quality of justice provided—​in terms of both pressuring defendants to plead guilty to lesser charges, and influencing prosecutors to accept guilty pleas to lesser charges to reduce court backlogs.

Forms of Negotiation To date, discussions of plea negotiations have generally focused on three main forms: charge bargaining, which incorporates withdrawing and substituting charges; fact bargaining, which incorporates discussions on which facts will form the basis of the agreed summary of facts presented to the court from which the defendant will be sentenced; and sentence bargaining, which incorporates discussions on the prosecutor’s sentencing submission, agreements as to what stage in the proceedings the defendant indicated an intention to plead guilty and in which jurisdiction the offense should be heard (see e.g., Flynn, 2016; Fox & Deltondo, 2015; Johns, 2002; Mackenzie,Vincent, & Zeleznikow, 2015; Manikis, 2012; Wren & Bartels, 2014;Yang, 2013). Although these are the most common forms of negotiation, they are not the only ones. Indeed, the current study identified 14 forms of plea negotiation within the case files and interview datasets: • Withdrawing and substituting charges: these are the two most common forms of plea negotiation. They involve the prosecution withdrawing charges that may have been laid in the alternative17 or were duplicitous or contained lesser included charges, and substituting these with less serious or fewer offenses. Negotiation in these cases involves identifying which charge or charges the defendant will plead guilty to and which will be withdrawn. This has the effect of simplifying the case without necessarily affecting the overall criminality of the offending conduct, although where an accused pleads guilty to a substituted charge, this is likely to reduce the severity and 159

Arie Freiberg and Asher Flynn



•



•

•

aggravation of the original charge or charges, thereby reducing the overall likely sentence and seriousness of the conviction recorded. There have been persistent allegations that police “overcharge” offenders; a practice known colloquially as “the hamburger with the lot.” Flynn and Fitz-​Gibbon (2011, p. 916) define overcharging as “the situation in which a prosecuting agency charges an accused with a more serious offense or an accumulation of offenses, which creates a strong incentive for the accused to enter into plea bargaining.” Although this is a matter of concern, the evidence in the current study appears to indicate that this may by the product of inexperience on the part of police, and that overcharging is a defensive approach aimed at ensuring that a conviction for something is secured. Charging numerous offenses to cover the same course of conduct may also be a product of the existence of many alternative offenses in relation to some forms of conduct, such as assault,18 and time limits that apply to the charging process, particularly in summary matters, which restrict the power of prosecutors to amend the charges. Rolled up and representative charges: offenders often commit multiple offenses of a similar kind against the same statutory provision on different occasions. An example would be multiple thefts or frauds against the same victim. Under Victorian law, these offenses can be “rolled up” under one charge, although the summary of facts presented to the court will explain that the charge encompasses a number of distinct offenses. For the court, this has the benefit of simplifying the proceedings; for the prosecution, it protects them from allegations that they are overloading the indictment or overcharging, and for the accused, it means that they will have fewer convictions against their name. For the public and the victim, it will likely result in a sentence that is similar to that which may have been imposed had the accused been found guilty of every charge, owing to the operation of the sentencing principles of totality and concurrency. A representative charge is similar in that it will reduce the number of charges to which an offender will plead guilty where there have been multiple offenses—​for example, historical sexual offenses where it is difficult for the victim to identify exact dates and times. The court will be aware that the offenses to which the offender pleaded guilty are not singular, unusual, or out of character. Therefore, as in the case of rolled-​up charges, the fact that the offense is representative should be specified in the summary of facts presented to the court for sentencing to ensure that the penalty acknowledges that this is a course of conduct, not simply a one-​off. However, unlike rolled-​up counts, the interview data suggests that representative counts both reduce the number of charges on the indictment and generate a less severe sentence, because the judge cannot sentence the defendant for charges not listed on the indictment. Taking offenses into account: provisions similar to rolled up and representative charges allow a court to take into account other pending charges that have been filed with the court and that the defendant admits to having committed. This procedure allows offenses in relation to which the offender has not been formally convicted to be taken into account in sentencing while barring further proceedings in respect of those offenses. Additionally, the admission is not regarded as a conviction, but the offenses are formally accounted for which obviates the need for further investigation. Offenses taken into account will generally augment the sentence for the principal offense, but the increase is likely to be less than would have been the case had the charges been separately prosecuted. Negotiating the agreed summary of facts: negotiating a settled summary of facts is a very common occurrence and comprises a very major factor in determining whether a case will resolve. As the summary of facts will often form the basis upon which a court will decide what sentence to impose in the absence of a trial, and as judicial officers will often be pressed for time, the description of the offense or offenses is a crucial part of the sentencing process. However, there is a fine balance between providing a generous description and providing a misleading account that fails to reflect the true nature of the crime (see Flynn, 2012).

160

Plea Negotiations

• Agreements on the prosecutor’s sentencing submission: in contrast to some international practice, in most Australian jurisdictions prosecutors cannot agree to specific punishments, either quantum or type, in exchange for a guilty plea.The High Court of Australia has made it clear that: It is for the prosecution, alone, to decide what charges are to be preferred against an accused person. Second, it is for the accused person, alone, to decide whether to plead guilty to the charges preferred. … Third … it is for the sentencing judge, alone, to decide what sentence will be imposed. … [The defendant’s] decision cannot be made with any foreknowledge of what sentence will be imposed. Neither the prosecution, nor the offender’s advisers, can do anything more than proffer an opinion as to what might reasonably be expected to happen.19 In Victoria, prosecutors can only make sentencing submissions in relation to the type of sentence that should be imposed, but not in relation to the numerical range within which the sentence should fall or the duration of the sentence. They may comment on current sentencing practices, comparable cases and a number of other matters [DPP Prosecution Policy nd, pp. 11–​ 12]. However, while specific agreements between the parties as to sentencing submissions may be prohibited, a prosecutor may indicate in discussions that they may not object to the imposition of a non-​custodial sentence or that they are unlikely to appeal if such a sentence were imposed. Accordingly, sentencing discussions remain an important part of plea negotiations in Victoria (Flynn & Freiberg, 2018). • Agreement to change jurisdiction: many serious cases in Australia are classified as “indictable offenses triable summarily”. This means that in certain circumstances an indictable offense may be determined in the summary jurisdiction, before a single magistrate with lower penalty ranges, rather than before a judge and jury. There are a number of advantages to this course of action. The first is that it may reduce the severity of the sentence that may be imposed. Second, cases in the summary jurisdiction are dealt with more expeditiously and at less expense. Third, appeals from sentences imposed in the summary jurisdiction are available de novo [a form of rehearing] rather than requiring leave to appeal from the Supreme Court on the basis that the sentencing judge has made an error in law. For these reasons, this can be an attractive resolution outcome for a defendant. • Agreement not to pursue charges against another person known to the defendant: this is a less common form of negotiation which involves the prosecution assessing the strength of the evidence against a number of parties, usually related, and deciding, after negotiation, not to proceed against one party, often a spouse, partner or family member. • Agreement not to pursue other charges or investigations into the defendant: similar to the previous type of negotiation, this may occur occasionally where there may be a number of minor charges or those that might be difficult to pursue and the prosecution may agree not to pursue these charges. However, prosecuting authorities do not have any power over police investigations and cannot direct them to discontinue a line of inquiry. • Providing the defendant and/​or their family with protection and/​or financial support: on some occasions, support may be provided to high-​level, high-​r isk informers or their families in the form of witness protection or, in very rare cases, some financial support to spouses or children of the defendant. • Agreement not to oppose bail: As of July 2018, some 35% of the male prison population and 43% of the female prison population in Victoria is unsentenced and awaiting trial. The average time served on remand by prisoners tried in the higher courts is around nine months and for those tried in the Magistrates’ Court, it is just over two months (Victorian Sentencing Advisory Council, 2016, pp. 48–​49). Being held in custody awaiting trial may involve loss of employment, housing, and/​or contact with family. An agreement to plead guilty in exchange for the prosecution not to oppose an application for bail is a form of negotiation that does not relate to the length or nature of the sentence, but is a concession nonetheless. Increasingly in Australia, the discretion of courts and other bail decision makers in relation to the granting of bail is being restricted following

161

Arie Freiberg and Asher Flynn

amendments to the law which emerged in direct response to a number of egregious incidents involving alleged offenders on bail committing serious offenses (Bartels et al., 2018). The mandatory or presumptive nature of these provisions is likely to diminish the scope for this form of negotiation. • Agreement to recommend a diversion program: in relation to less serious offenses,Victoria has a scheme whereby offenses heard in the court of summary jurisdiction may be adjourned for a period of up to 12 months to enable the defendant to participate in and complete a diversion program. This program may require an offender to apologize to the victim, make a donation or compensation, undertake voluntary work, an anger management program, drug and alcohol awareness, counseling or treatment programs, or other conditions relating to the offender’s behavior. The offender must acknowledge responsibility for the offense, be a first-​time offender, and the offense must not be considered to be too serious, which may, in turn, depend upon the agreed summary of facts. The police informant or prosecutor must agree to this option, as must the magistrate. If the program is completed satisfactorily, the defendant is discharged without a finding of guilt or criminal record. • Agreement not to seek costs: this is an unusual form of negotiation in that it does not involve a defendant agreeing to plead guilty, but rather agreeing not to pursue costs in the event that a police prosecutor decides to withdraw charges against them. Under Victorian law, a defendant is entitled to seek costs in the summary jurisdiction if charges are dismissed or not proceeded with. The advantage to a defendant is that they may avoid acquiring a criminal record had the case proceeded, but on the other hand, they must absorb the cost of defending themselves. For the prosecution, it may mean that they avoid running a relatively weak case and save paying the costs in the event of a loss. As can be seen, the forms of plea negotiations in Victoria are varied, complex and situationally specific. They also involve many aspects of the criminal justice process, not just the charges being proceeded with or the sentence outcome.

Negotiation Outcomes The empirical study of plea negotiations in Victoria described above revealed that across the 50 case files, there was a reduction of almost 50% in the number of charges ultimately proceeded with—​from 325 charges to 162 charges. The average number of charges per case prior to resolution was 6.43 and the mean number of charges withdrawn was 3.24. The majority of cases involved the most serious charge being withdrawn in exchange for less serious alternatives, but in some cases, the defendant pleaded guilty to the most serious offense in exchange for other charges being withdrawn. The most common offenses negotiated were theft, injury, or serious injury offenses and drug use or possession. These offenses tended to be negotiable either because there were numerous offenses, the reduction of which was not likely to affect the overall outcome, or injury offenses where there is some doubt regarding the mental state of the offender (intention versus recklessness) or the degree of injury (serious/​less serious), each aspect of which amounted to a separate or different offense carrying different maximum penalties. In relation to drug offenses, plea negotiation was common due to the potential length and expense of a trial as a result of issues relating to the testing of drugs, their amount and the relative roles of the parties. Incentives to inform on other members of a syndicate in return for sentence discounts were also a factor in drug offense negotiation processes. The more difficult cases to negotiate were serious cases such as murder, particularly where there was a high level of public interest, and sex offenses involving adults, where there were issues of consent, lack of corroborating evidence and the absence of alternative offenses. Familial sexual offenses involving children were also difficult to resolve due to the reluctance of offenders to acknowledge their behavior. In other words, a guilty finding at trial could still allow the defendant to deny the 162

Plea Negotiations

abuse occurred. Another impediment to negotiation has been the introduction of the sex offender registration legislation that requires offenders to be placed on a register for up to 15 years, depending upon the seriousness of the offense. The consequences of being placed on such a register are considerable and judicial officers in Australia generally have little or no discretion whether to place a person on the register. Although there is little research on how sex offender registers affect plea negotiations, the interview data generated by the current study suggested that potentially being placed on a register is a major contributing factor for the lack of guilty pleas in sex offense cases. A final group of offenses that have been difficult to negotiate is that involving family violence. Victoria, and Australia more generally, has identified family violence as a major public policy issue. A lengthy and important inquiry in the form of a Royal Commission into Family Violence (RCFV, 2016) produced an extensive report that highlighted the prevalence and seriousness of family violence. It recommended extensive changes to police, prosecutorial and court practices. The growing public interest in family violence has meant that such offenses are to be treated with utmost seriousness and this has resulted in changes in police charging practices and approaches to prosecuting family violence offenses. This study found that fears by police and prosecutors that they might be regarded as not taking these offenses seriously enough and that failure to respond appropriately might result in family violence escalating, even to the extent of homicide, has meant that negotiations are far less likely. This was even the case where the decision to pursue the case was contrary to the victim’s preferences.

The Rule of Law or of Men and Women? The negotiation process tends to take place either in private, between prosecutors and defense counsel, or before magistrates and judges in relatively informal proceedings. Although there exist some broad legal rules regarding the process and substance of negotiation, this study found evidence that the outcome of some negotiations was dependent upon the personalities of the parties. Some prosecutors were more amenable to negotiation and some judicial officers were more encouraging of settlement. There was a consistent theme that some judicial officers were allocated to the contest mention jurisdiction because they were known to be more lenient in their sentencing, therefore providing an incentive to defendants to settle the case at that point in the proceedings rather than facing the uncertainty of a trial and sentence before a different judicial officer. This suggests that personality may play a key role in determining justice outcomes when it comes to negotiated guilty pleas.

Unrepresented Defendants The pressures defendants face in plea negotiations and in the criminal justice system more generally are even more prevalent in light of continuing constraints on legal aid services which limit access to lawyers throughout the legal process, and leave defendants to navigate complex legal processes without adequate support (Flynn & Hodgson, 2017a; Flynn et al., 2016). In this regard, a recognized safeguard for defendants in combatting pleading pressures is legal representation. This view is based on the idea that a lawyer will provide quality advice about a guilty plea which helps the defendant make a principled, informed, and appropriate pleading decision. In practice, however, this safeguard is threatened given the very strict eligibility criteria for legally aided assistance, including means, merits, and case assessments which have been a consequence of increasing budget restrictions and rising demand for legal services. In Victoria, the eligibility criteria are extremely restrictive, which means that only those on very low incomes will obtain representation. Over recent years there has been a dramatic increase in unrepresented defendants attempting to engage in plea negotiations, particularly in the lower courts, raising a number of concerns around power and justice. This outcome also creates additional resourcing pressures within the system. As Flynn and Hodgson (2017a, p. 11) observe, an increase in 163

Arie Freiberg and Asher Flynn

unrepresented defendants is likely to increase “the use of court resources and costs, and decrease services and access to justice for already vulnerable and marginalized groups”. The lack of representation often results in offenders pleading guilty rather than contesting the charges or entering into negotiations with only limited knowledge of the law and the consequences of their actions. This, in turn, casts a considerable burden upon judicial officers and prosecutors, who are becoming quasi-​defense practitioners in order to assist unrepresented defendants. The potential for miscarriages of justice in these circumstances is real and present.

Constraints on Plea Negotiations and Reform Regulation of the process of plea negotiation is limited. Prosecutorial procedures and practices are regulated by internal guidelines that are public documents but do not have the force of the law.These are extensive documents that cover such matters as the role of the prosecutor, the exercise of prosecutorial discretion, the role of victims, undertakings or agreements to assist law enforcement authorities and indemnities from prosecution, appeals, juries, family violence and the like. In respect of plea negotiations, the guidelines cover matters including when resolution may occur, when the views of the victim and informant must be sought, when approval is required from more senior staff, relations with opposing legal representatives, record keeping, and unrepresented accused. In the state of New South Wales, the Crimes (Sentencing Procedure) Act 1999 (NSW), s 35A(2) provides that a court may not take into account offenses other than the principal offense,20 that were the subject of charge negotiations,21 unless the prosecutor has filed a certificate with the court verifying that: (a) The requisite consultation [with the victim] has taken place, or, if consultation has not taken place, the reasons why this has not occurred; and (b) Any statement of agreed facts arising from the negotiations tendered to the court constitutes a fair and accurate account of the objective criminality of the offender having regard to the relevant and provable facts or has otherwise been settled in accordance with the applicable prosecution guidelines. In Victoria, victims do not have the right to review decisions made by prosecutors. In late 2016, the Victorian Law Reform Commission (2016, p. xxii, recommendation 9) recommended that the DPP be required to provide victims with written reasons for any decision to modify charges, not to proceed with some or all charges, or to accept a guilty plea to a lesser charge (in other words, when a plea negotiation occurs), unless the victim has expressed a wish not to be so informed. They also recommended (VLRC, 2016, p. xxii, recommendation 10) that the Victims’ Charter Act 2006 (Vic) should be amended to: (a) establish a right for victims to seek internal review of a decision by the DPP to discontinue a prosecution or to proceed with a guilty plea to lesser charges; (b) require the DPP when informing the victim of these decisions and the reasons for these decisions to notify the victim of their right to seek internal review and the procedure for doing so.22 Similarly, the Royal Commission into Institutional Responses to Child Sexual Abuse (2017) proposed that each Australian Department of Public Prosecutions or Office of the Director of Public Prosecutions should adopt a formalized internal complaints mechanism which would allow victims to seek an internal merits review of key decisions, particularly decisions that would result in a prosecution not being brought or being discontinued in relation to charges for alleged offending against that victim. It is yet to be seen whether or how such reforms would be implemented in practice within a system that claims to be adversarial. 164

Plea Negotiations

Conclusion Despite the criticisms made of the plea negotiation process over the years, it appears that in Victoria, at least, the plea negotiation system is operated by dedicated professionals working within the context of significant financial constraint, which places them under considerable pressure to deal with large workloads quickly and efficiently. All parties—​courts, prosecutors, defense and police—​work within legal, administrative and ethical guidelines. There are benefits and limitations of this method of case finalization, all of which are made much more complex by the limited data, research, and evidence of the process to those outside the legal community. Plea negotiations do not operate in a lawless silo. They play a valuable role in the legal realm. However, they ultimately remain a pragmatic form of somewhat mysterious justice in a largely unknown and imperfect criminal justice system.

Notes 1 This chapter is drawn from Negotiated Guilty Pleas—​Pragmatic Justice in an Imperfect World, London: Palgrave Macmillan (2018), parts of which are reproduced with permission. The research upon which this book was based was funded by the Australian Criminology Research Council (Grant No. 53/​13–​14). 2 For a detailed description of the research methodology see Flynn and Freiberg (2018, Appendix A). 3 Gallagher (1991) 23 NSWLR 220, 232. 4 Unlike many US jurisdictions, an Australian accused can only plead guilty or not guilty (other than a plea of not guilty on the grounds of insanity). Australian law does not recognize a nolo contendere plea (no contest) or what is termed an Alford plea whereby a guilty plea may be entered but the accused is able to maintain his or her innocence (Redlich, Wilford, & Bushway, 2017, p. 459). 5 JO [2009] NTCCA 129 at [105]. 6 [2007] VSC 490 (15 November 2007). 7 Ibid. [at 4]. 8 Ibid. [at 6]. 9 Ibid. [at 8]. 10 Gallagher (1991) NSWLR 220. 11 [1986] HCA 58 at [14]. 12 [2008] VSCA 190 [at 3]. 13 Barbaro v. The Queen; Zirilli v. The Queen [2014] HCA 2. 14 The Children’s Court hears the remainder of the cases. 15 Markarian [2005] HCA 25 at [51]. 16 A court is not required to impose the mandatory sentence if “special reasons” exist, namely, that the defendant has assisted or has given an undertaking to assist law enforcement authorities, the defendant is between the ages of 18 and 21 at the time of the commission of the offense and proves that he or she has a particular psychosocial immaturity or impaired mental functioning, the court proposes to make a secure treatment order or residential treatment order in respect of the defendant, or that there are substantial and compelling circumstances that justify doing so, Sentencing Act 1991 (Vic) s 10A; see, generally, Sentencing Advisory Council (2012) (baseline sentencing report) and Sentencing Advisory Council (2011) (Statutory Minimum sentences for Gross Violence Offences). Since the introduction of these pieces of legislation, the Victorian Parliament has added a number of other offenses and circumstances in which presumptive sentences will apply, including certain offenses against emergency workers and custodial officers on duty—​Sentencing Act 1991 (Vic) s 10AA; offenses of breaching supervision orders under the Serious Sex Offenders (Detention and Supervision) Act 2009 (Vic), Sentencing Act 1991 (Vic) s 10AB; the offense of aggravated home invasion, Sentencing Act 1991 (Vic) s 10AC; and the offense of aggravated carjacking, Sentencing Act 1991 (Vic) s 10AD. 17 An alternative charge is one where several offenses may cover the same or similar course of conduct and once the facts have been clarified or agreed to, one or more charges may be withdrawn. For example, an accused person may face multiple charges arising from the one incident such as armed robbery or robbery, or intentionally causing serious injury or recklessly causing serious injury. 18 For example, the possible offenses in Victoria include common assault, aggravated assault, assault in company, assault by kicking, assault with a weapon, assault or threatening to assault a person with intent to commit an indictable offense, resisting, obstructing assault or threatening to assault an emergency worker on duty, assaulting or threatening to assault a person with intent to resist or prevent arrest, causing serious injury intentionally in circumstances of gross violence, causing injury intentionally, causing serious injury recklessly and causing injury recklessly or intentionally.

165

Arie Freiberg and Asher Flynn 1 9 Barbaro v. The Queen; Zirilli v. The Queen [2014] HCA 2 [at 47]. 20 See discussion above with regard to “offenses taken into account”. 21 Defined as “negotiations between the prosecution and an offender with respect to a plea of guilty in relation to an offense other than the principal offense”. 22 See also Iliadis and Flynn (2018) for a discussion on victims’ right to review prosecutorial decisions in the UK context. Most of these recommendations were adopted in the Victims and Other Legislation Amendment Act 2018 (Vic).

References Bartels, L., Gelb, K., Spiranovic, C., Sarre, R., & Dodd, S. (2018). Bail, risk and law reform: A review of bail legislation in Australia. Criminal Law Journal, 42(2), 91–​107. Buckle, S., & Buckle, L. 1977. Bargaining for Justice: Case Disposition and Reform in the Criminal Courts. New York: Praeger. Euvrard, E., & Leclerc, C. (2017). Pre-​trial detention and guilty pleas: Inducement or coercion. Punishment and Society, 19(5), 525–​542. Flynn, A. (2009). Sentence indications for indictable offences: Increasing court efficiency at the expense of justice—​a response to the Victorian legislation. Australian and New Zealand Journal of Criminology, 42(2), 244–​268. —​—​. (2010a). Victoria’s legal aid funding structure: Hindering the ideals inherent to the pre-​trial process. Criminal Law Journal, 34(1),  48–​63. —​—​. (2010b). An indication of injustice: An analysis of the problems inherent to maintaining the sentence indication scheme in Victoria’s Higher Courts. Flinders Law Journal, 12(2),  41–​78. —​—​. (2012). Bargaining with justice:Victims, plea bargaining and the Victims’ Charter Act 2006 (Vic). Monash University Law Review, 37(3),  73–​96. —​ —​ . (2016). Plea negotiations, prosecutors and discretion: An argument for legal reform. Australian and New Zealand Journal of Criminology, 49(4), 564–​582. Flynn, A., & Freiberg, A. (2018). Negotiated Guilty Pleas—​Pragmatic Justice in an Imperfect World. London: Palgrave Macmillan. Flynn, A., & Fitz-​Gibbon, K. (2011). Bargaining with defensive homicide: Examining Victoria’s secretive plea bargaining system post law reform. Melbourne University Law Review, 35(3), 905–​932. Flynn, A., & Hodgson, J. (2017a). Access to justice and legal aid cuts: A mismatch of concepts in the contemporary Australian and British legal landscapes. In A. Flynn & J. Hodgson (eds.), Access to Justice and Legal Aid: Comparative Perspectives on Unmet Legal Needs (pp. 1–​22). Oxford: Hart Publishing. Flynn, A., & Hodgson, J. (eds.). (2017b). Access to Justice and Legal Aid: Comparative Perspectives on Unmet Legal Needs. Oxford: Hart Publishing. Flynn, A., Hodgson, J., McCulloch, J., & Naylor, B. (2016). Legal aid and access to legal representation: Redefining the right to a fair trial. Melbourne University Law Review, 40(1), 207–​239. Fountain, E., & Woolard, J. L. (2017).The capacity for effective relationships among attorneys, juvenile clients, and parents. Ohio State Criminal Law Review, 14, 492–​519. Fox, R., & Deltondo, N. (2015). Victorian Criminal Procedure. Sydney: The Federation Press. Freiberg, A. (2005). Managerialism in Australian criminal justice: RIP for KPIs? Monash University Law Review, 31(1),  12–​36. Freiberg, A. (2014). Fox and Freiberg’s Sentencing: State and Federal Law in Victoria (3rd ed.). Sydney: Thomson Reuters. Gazal-​Ayal, O., & Tor, A. (2012). The innocence effect. Duke Law Journal, 62, 339–​401. Iliadis, M., & Flynn, A. (2018). Providing a check on prosecutorial decision-​making: An analysis of the victims’ right to review reform. The British Journal of Criminology, 58(3), 550–​568. Johns, R. (2002). Victims of Crime: Plea Bargains, Compensation,Victim Impact Statements and Support Services (Briefing Paper No. 10/​ 02). Retrieved from: www.parliament.nsw.gov.au/​researchpapers/​Documents/​victims-​of-​ crime-​plea-​bargains-​compensation-​vict/​10-​02.pdf Law Council of Australia. (2014). Policy Discussion Paper on Mandatory Sentencing. Canberra: Author. Mackenzie, G.,Vincent, A., & Zeleznikow, J. (2015). Negotiating about charges and pleas: Balancing interests and justice. Group Decision and Negotiation, 24(4), 577–​594. Manikis, M. (2012). Recognizing victims’ role and rights during plea bargaining: A fair deal for victims of crime. Criminal Law Quarterly, 58(3–​4), 411–​432. McConville, M., & Marsh, L. (2014). Criminal Judges: Legitimacy, Courts and State-​Induced Guilty Pleas in Britain. Cheltenham: Edward Elgar Publishing. McConville, M., & Mirsky, C. (2005). Jury Trials and Plea Bargaining: A True History. Oxford: Hart Publishing.

166

Plea Negotiations PricewaterhouseCoopers. (2008). Review of Fees Paid by Victoria Legal Aid Barristers in Criminal Cases. Melbourne, Australia: Author. Redlich, A.D.,Wilford, M.M., & Bushway S. (2017). Understanding guilty pleas through the lens of social science. Psychology, Public Policy and Law, 23(4), 458–​471. Royal Commission into Family Violence. (2016). Summary and Recommendations (Parliament Paper No. 132). Retrieved from: http://​files.rcfv.com.au/​Reports/​Final/​RCFV-​All-​Volumes.pdf Tasmanian Sentencing Advisory Council. (2018). Statutory Sentencing Discounts for Pleas of Guilty: Consultation Paper. Hobart, Australia: Author. Retrieved from: www.sentencingcouncil.tas.gov.au/​_​_​data/​assets/​pdf_​file/​ 0020/​412292/​Final-​SAC-​Consultation-​paper-​Statutory-​Sentencing-​Discounts-​for-​Pleas-​of-​Guilty.pdf Victoria, Director of Public Prosecutions. (n.d.). Policy of the Director of Public Prosecutions for Victoria. Retrieved from: www.opp.vic.gov.au/​getattachment/​b5d48af4-​3bef-​4650-​84fa-​6b9befc776e0/​DPP-​Policy.aspx Victorian Law Reform Commission. (2016). Victims of Crime in the Criminal Trial Process: Final Report. Melbourne: Author. Victorian Sentencing Advisory Council. (2011). Statutory Minimum Sentences for Gross Violence Offences—​Report. Melbourne: Author. —​—​. (2012). Baseline Sentencing—​Report. Melbourne: Author. —​—​. (2015). Guilty Pleas in the Higher Courts: Rates,Timing, and Discounts. Melbourne: Author. —​—​. (2016). Victoria’s Prison Population 2005–​2016. Melbourne: Author. Westling, W. (1976). Plea bargaining: A forecast for the future. Sydney Law Review, 7, 424–​432. Wren, E., & Bartels, L. (2014). “Guilty, Your Honour”: Recent legislative developments on the guilty plea discount and an Australian Capital Territory case study on its operation. Adelaide Law Review, 35, 361–​384. Woolard, J.L., Henning, K., & Fountain, E. (2016). Power, process, and protection: juveniles as defendants in the justice system. Advances in Child Development and Behaviour, 51, 171–​201. Yang, K. (2013). Public accountability of public prosecutions, Murdoch University Law Review, 20(1),  28–​75.

Cases Cited Barbaro v. The Queen; Zirilli v. The Queen [2014] HCA 2. Gallagher (1991) 23 NSWLR 220. JO [2009] NTCCA 129. Malvaso [1986] HCA 58. Markarian [2005] HCA 25. MacNeil-​Brown [2008] VSCA 190. Williams [2007] VSC 490.

Statutes Cited Crimes Act 1958 (Vic) Crimes Amendment (Gross Violence Offences) Act 2013 (Vic). Crimes (Sentencing) Act 2005 (ACT). Crimes (Sentencing Procedure) Act 1999 (NSW). Sentencing Act 1991 (Vic). Serious Sex Offenders (Detention and Supervision) Act 2009 (Vic). Victims’ Charter Act 2006 (Vic).

167

8 PLEA BARGAINING IN THE SHADOW OF THE TRIAL Amy Dezember and Allison D. Redlich

Imagine you are a defendant charged with the three counts of grand larceny. You have been in the county jail for the past three months because you could not afford bail.You know the prosecution’s evidence against you is weak and you have a realistic chance of acquittal if you go to trial. But there is now a plea bargain on the table.The prosecutor is offering to dismiss two of the counts if you plead guilty to one felony count of grand larceny, for which you would receive “time served” and three years of probation. Your defense attorney recommends that you take the deal. You could go home today if you accept the plea offer, or you could continue to wait in jail for your day in court, and if found guilty, you face the maximum punishment of 10 years in prison. What would you do? Although details of the charges, negotiated offers, and punishments vary, the vast majority of defendants in the United States in the above situation choose to accept the plea bargain. In total, 97 to 99% of juvenile and adult convictions are the result of guilty pleas (Bureau of Justice Statistics, 2010; Kaban & Quinlan, 2004). And it is well established that defendants receive a discount for pleading guilty, or alternatively, a penalty for going to trial; the size of this discount varies across individual cases (Bradley-​Engen et al., 2012; Bushway & Redlich, 2012; Smith, 1986; Ulmer & Bradley, 2006; Ulmer, Eisenstein, & Johnson, 2010). While criminologists have tended to use perspectives such as focal concerns theory to explain variation in plea discounts, legal scholars have focused more on the “bargaining in the shadow of the trial” model to explain such variation (Bibas, 2004; Landes, 1971; Mnookin & Kornhauser, 1979). Theorists have argued that the various legal actors involved in the plea process make such plea decisions “in the shadow of trial,” thus the name of the model (Nagel & Neef, 1979). The “shadow of the trial” model argues that decisions to offer, accept, or reject plea bargains derive from the perceived probable outcome of a trial, which in turn is driven by the strength of the evidence. This model was only recently introduced to a criminological audience (e.g., see Bushway & Redlich, 2012) and social scientists are working to empirically test this theory as a way to better understand variations in the size of plea discounts. The “shadow of the trial” model predicts that a defendant would plead guilty if the offered sentence is less than or equal to the expected value of the trial (Bibas, 2004). For example, if the expected sentence for a conviction at trial is 20 years and the defendant believes he or she has an 80% probability of conviction at trial, then a plea to a sentence of no more than 16 years (80% of 20)  represents a rational choice for a risk-​neutral defendant (Bushway, Redlich, & Norris, 2014). Similarly, prosecutors forecast the expected sentence after trial, discount it by the probability of acquittal, and offer some proportional discount to the defendant in the form of a plea bargain (Bibas, 168

Plea Bargaining

2004). The shadow theory also argues that prosecutors, defense attorneys, and defendants base their plea decisions on the strength of the evidence they anticipate being presented at trial, suggesting that evidence plays a key role in the plea decision process (Bushway, Redlich, & Norris, 2014). This indicates that evidence strength should drive plea decisions insofar as strong evidentiary cases lead to trial convictions. This chapter begins with a review of the legal and empirical literature on the shadow of the trial model, the role of evidence in the plea decision-​making process, and research on other factors that could also drive plea bargaining, including prosecutorial discretion. Then, we partially explore these issues using new data obtained from systematic observations of plea hearings, examination of court records, and interviews with defendants who recently pled guilty. These data are analyzed to evaluate how the strength of evidence impacts plea decisions and the value of plea discounts (e.g., distance between trial and plea sentences). We conclude with discussions on how to improve the shadow of the trial paradigm, an examination of how strength of evidence influences plea bargaining, and ideas for future research designed to further our understanding of how the “shadow” model operates in our system of pleas.

Macro-​Level Approaches to Understanding Plea Decision-​Making Plea bargaining is a powerful tool used to dispose of cases and has become an integral part of our criminal justice system. With the vast majority of criminal convictions in the United States resulting in a guilty plea, scholars continue to debate the merits of plea bargaining, the incentives behind guilty pleas, and the influences on the outcomes of guilty pleas (Alschuler, 1981; Wright & Miller, 2002; Johnson, King, & Spohn, 2014). Defendants typically plead guilty as part of an arrangement with the district attorney, wherein the state ensures a conviction and the defendant receives some sort of deal or bargain (Bibas, 2004). These deals commonly include charge and/​or sentence discounts (Redlich et al., 2017) and, depending upon the severity of the crime, can allow for immediate release from jail, which can be a strong incentive to accept the plea offer (Kellough & Wortley, 2002). The constitutionality of plea bargaining was firmly established in Brady v. United States (1970), when the Court held that states can offer benefits to a defendant who is willing to admit guilt, so long as the plea is made knowingly, voluntarily, and without coercion. Before and since this decision, courts have relied on plea bargaining as a way to effectively and efficiently dispose of cases (Fisher, 2000).

Institutional Approaches in Plea Bargaining Research In criminology, research on plea bargaining and plea discounts frequently focuses on the institutional or organizational influences on pleas, such as the role of judges, prosecutors, and defense attorneys working together towards the shared goals of disposing cases efficiently and minimizing uncertainty (Eisenstein & Jacobs, 1977). Within this framework, the courtroom workgroup will establish norms, or going rates, that are less than the punishment at trial for defendants who are willing to plead guilty. In other words, prosecutors must decide what plea offers to make, defense attorneys must decide how to advise clients on these issues, and judges must decide whether to accept or reject the negotiated plea deal. Many studies looking at the effect of courtroom workgroups on plea discounts suggest that the size of any plea-​trial sentencing differences likely varies by jurisdiction (Brereton & Casper, 1982; King et al., 2005). One study looking at courtroom workgroups found that similarity among the court actors, specifically gender and experience-​level similarities, and familiarity between the prosecutor and judge increased the odds of a plea disposition and reduced the number of days to disposition, demonstrating that the level of interaction among court actors is an important factor to consider in case processing (Metcalfe, 2016). One commonly used approach in plea research is the focal concerns theory, which builds on the framework of courtroom work groups and institutional factors that impact pleas. This framework 169

Amy Dezember and Allison D. Redlich

involves the recognition of a court community or organization that comes together around a shared set of norms and expectations (Eisenstein & Jacob, 1977; Eisenstein, Fleming, & Nardulli, 1987; Ulmer, 1997). Focal concerns theory argues that sentencing decisions reflect the judge’s assessment of the offender’s blameworthiness or culpability, the judge’s desire to protect the community through incapacitation of dangerous offenders, and the judge’s concerns about the social costs of a sentencing decision (Hartley, Maddan, & Sophn, 2007). Similarly, institutional factors like caseload size vary by jurisdiction and can affect the probability of conviction, and thus the plea discount will also vary systematically across courts (Ulmer, Eisenstein, & Johnson, 2010). In other words, jurisdictions with higher caseloads or fewer trials are more likely to have lower probabilities of conviction and therefore offer plea discounts that are larger to encourage pleas.

Empirical Tests of the Focal Concerns Model The focal concerns approach to plea bargaining has been used to understand the sentencing discounts that defendants receive when pleading guilty. It is well established that defendants who plead guilty receive comparatively shorter sentences than individuals who are convicted of similar crimes at trial, referred to as the “trial penalty” (Ulmer & Bradley, 2006). Traditionally, sentencing guidelines provide a presumptive sentence, and any departure from this presumptive sentence can be viewed as an exercise of discretion that can be placed directly in the hands of the actor ultimately responsible for assigning sentence lengths (Bushway & Piehl, 2001; Engen & Steen, 2000). Research using the focal concerns approach has found variation in the size of the trial penalty across jurisdictions, though factors such as caseload, criminal history, and jurisdictional characteristics have only a modest ability to explain this variation (Ulmer & Bradley, 2006; Ulmer, Eisenstein, & Johnson, 2010). For example, case processing factors, such as caseloads, likely create practical constraints on decisions to charge and dispose of cases (Ulmer & Bradley, 2006). Additionally, when focal concern is applied to plea bargaining, it accentuates the issues that prosecutors face when negotiating pleas, such as doing justice, trial-​worthiness, convictability of cases, and even the prosecutors’ own long-​term career goals (Johnson, King, & Spohn, 2016). A problem with this approach has been that most of the datasets used for analysis do not contain information on cases that go to trial and end in acquittal, which are rare occurrences. Thus, authors cannot calculate the expected value of the trial but rather must compare similar cases that are convicted at trial to those that are resolved through guilty pleas (Abrams, 2011). Some scholars have begun new attempts to address these weaknesses and provide more fully specified models to explain the sentencing process. For example, Kutateladze and his colleagues (2014) assessed racial and ethnic disparity across multiple discretionary points of prosecution and sentencing using data from the New York County District Attorney’s Office. They found that black and Latino defendants were more likely than white defendants to be detained, to receive a custodial plea offer, and to be incarcerated (Kutateladze, Andiloro, & Johnson, 2014). These findings indicate that racial and ethnic disparities occur across multiple points in the criminal justice system, creating a cumulative disadvantage along the criminal justice continuum. Similarly, other researchers have conducted experiments using hypothetical cases to better understand plea decision-​making, which has opened the door to more in-​depth testing of models, including the shadow model, to explore the influence of legal and non-​legal variables (Bushway, Redlich, & Norris, 2014; Kramer, Wolbransky, & Heilburn, 2007; McAllister & Bregman, 1986).

The Shadow of the Trial Model In contrast to institutional theories like focal concerns, the shadow model is a formal theoretical model that starts from the perspective of the defendant, but ultimately can be extended to predict the behaviors of all court actors who interact with the defendant (Bushway, Redlich, & Norris, 2014). 170

Plea Bargaining

As previously mentioned, the shadow model predicts that plea decision-​making is premised on the perceived probable outcome of a trial (Bibas, 2004; Landes, 1971; Mnookin & Kornhauser, 1979). Under this model, defendants face a choice between two outcomes: (1) the near-​certain outcome that goes along with pleading guilty, and (2) the uncertain trial outcome. According to the model, a rational, risk-​neutral defendant would not accept a plea deal that is less than or equal to the discounted value of the trial outcome (Bushway & Redlich, 2012). Additionally, a prosecutor would not offer a defendant a plea deal that was less than or equal to the discounted trial sentence, since, in theory, the defendant would not accept that offer. This indicates that the sentence as a result of a plea will be equal to the sentence at trial discounted by the probability of conviction at trial (Bushway & Redlich, 2012). For example, the theory posits that the plea discount will be large when the probability of conviction is low, and small when the probability of conviction is high. There is some evidence to support this: prosecutors have been found to offer substantial plea bargains to defendants when they do not have adequate evidence to convict at trial (Kramer & Ulmer, 2002; see also, Alschuler, 1968). More recently, plea bargaining research has focused on the discount that the defendant receives in sentencing. Under the shadow model, the plea sentence is influenced by the probability of conviction if the defendant goes to trial and the sentence the defendant would receive if convicted at trial (Bushway & Redlich, 2012). Research indicates that defendants who plead guilty receive comparatively shorter sentences than individuals who are convicted of similar crimes at trial (Ulmer & Bradley, 2006). For example, Zottoli and her colleagues (2016) reported that juvenile and adult defendants who pled guilty in New York City received, on average, 98% and 80% sentencing discounts, respectively. These discounts on sentencing, particularly when large, provide attractive incentives for defendants to accept the prosecutor’s plea offer rather than exercise their right to trial. Furthermore, studies have shown that higher district court caseload pressure is associated with greater trial penalties (Ulmer et al., 2010). Studies looking at defendants’ plea decision-​making have also found that defendants who plead guilty reported doing so because they felt that if they were to reject a plea deal, they would receive a higher penalty (Daftary-​Kapur & Zottoli, 2014). This indicates that defendants consider the probable outcome at trial when deciding whether to plead guilty, providing some foundation of the shadow model operating in practice. However, legal scholars debate whether the shadow of the trial model provides a truly accurate description of the plea bargain decision-​making process (e.g., Bibas, 2004); thus, it is important to test this paradigm in order to identify if this model provides a better understanding for plea discounts and plea decision-​making.

Empirical Tests of the Shadow Model Despite the prevalence of the shadow theory in legal literature, the shadow model was only recently put to the empirical test (Bushway & Redlich, 2012). In order to conduct an adequate test of this model, ideally researchers need data on the probability of conviction at trial, the sentence at trial, and the value of the plea bargain. The expected value of the plea deal is often represented as X ≤ P(Y) (Bushway & Redlich, 2012). In this equation, the expected value of the plea deal (represented as X) is less than or equal to the probability of conviction at trial (P) multiplied by the sentence at trial (Y). In other words, as mentioned, the value of the plea should be less than or equal to the discounted value of the trial outcome (Bushway & Redlich, 2012). However, these pieces of information are not available for each person because each defendant either pleads guilty or goes to trial (if their case is not dismissed). Researchers have overcome this problem by using statistical models to create predicted counterfactuals for those who pled guilty based on the remaining few who actually went to trial. At the aggregate level, the results of the first-​known empirical test provided support for the shadow model and demonstrated that the average plea sentence for the sample was equivalent to the average sentence at trial discounted by the probability of conviction for the sample (Bushway & Redlich, 2012). More specifically, they found that, on average, those who pled guilty received a sentence that was 77% of the trial sentence, or alternatively, that the sentence at trial was 29.6% higher than the plea 171

Amy Dezember and Allison D. Redlich

sentence (Bushway & Redlich, 2012). In other words, the plea discount appears to be explainable, on average, by the shadow of the trial model. However, when they examined the process at the individual defendant level, Bushway and Redlich (2012) did not find support for the shadow model; they were not able to explain variation in the estimated probability of conviction for defendants who pled guilty compared to those who went to trial. Moreover, the evidence did not explain variation in estimates of probability of conviction for those who pled guilty. For example, a case with a confession led to a nearly 12 percentage point reduction in the probability of conviction at trial for those who pled guilty, which is in opposition to the predicted effect. In many cases, the defendant’s actual plea value was not at all similar to the estimate of the discounted probability of a sentence at trial (Bushway & Redlich, 2012). This indicates that the individual estimates were either uncorrelated with key pieces of evidence known to increase the probability of conviction, such as confessions, or correlated in the opposite direction than predicted by the shadow theory.This leaves open the possibility that plea negotiations are at least partially based on factors that have little to do with evidence, including, as suggested by Bibas (2004), caseload, financial considerations, courtroom culture, and individual differences in risk aversion. As a next step in testing the shadow model, Bushway, Redlich, and Norris (2014) conducted an experimental survey of defense attorneys, prosecuting attorneys, and judges (see also, Redlich, Bushway, & Norris, 2016). In this survey, which was a partial replication of work done 40 years earlier (i.e., Miller, McDonald, & Cramer, 1978), Bushway and colleagues presented respondents with a hypothetical vignette and a set of 31 fact files associated with the crime, each containing a different piece of information pertaining to the hypothetical armed robbery case (e.g., the defendant’s race and alibi).They then randomized the presence and absence of three specific types of evidence (eyewitness identification, confession, and DNA match) and the length of the defendant’s criminal history (short or long). Respondents were asked to assess the probability of conviction, to estimate the average sentence for a trial conviction, to indicate the least severe sentence that would be acceptable for a plea deal, and to describe what their likely course of action would be in the case. The goal was to assess whether the variation in the plea sentence, probability of conviction, and the expected trial outcome induced in the experiment conformed to the shadow model. Overall, they found that prosecutors and defense attorneys behaved in a manner consistent with the basic shadow model (Bushway, Redlich, & Norris, 2014). However, this was not true for the judges in the sample, who appeared to make decisions based on fixed discounts. This study provided a formal mathematical explanation for plea decision-​making within the shadow model, but also called for additional testing of the shadow model to further specify and improve this theory.

Criticisms of the Shadow of the Trial Model A main criticism of the shadow model is that the paradigm is too simplistic (Bibas, 2004; Redlich & Edkins, 2019). Specifically, scholars have argued that structural (e.g., poor lawyering, agency costs) and psychological considerations (e.g., use of heuristics, differences in risk aversion) in plea decisions have sizeable influences, and thus also need to be considered in the plea decision-​making process (Bibas, 2004; Stuntz, 2004). For example, the “shadow of the trial” model argues that the strength of the evidence moderates decisions to offer and accept pleas. However, it has been debated whether strength of evidence is in fact the sole driver of convictions; other factors such as prosecutorial discretion, defendant characteristics, and offense type have been found to impact the plea bargain a defendant receives (King et al., 2005; Rhodes, 1979; Ulmer & Bradley, 2006). Research also indicates that extra-​legal factors, such as race and gender, influence the offering and acceptance of pleas (Albonetti, 1990; Kellough & Wortley, 2002; Kutateladze, Andiloro, & Johnson, 2014). Specifically, black defendants, especially black male defendants, have a significantly lower probability of a charge reduction when they plead guilty, which means they are also expected to receive a lower value for their plea (Metcalfe & Chiricos, 2018). Similarly, prior criminal history is typically not allowed during 172

Plea Bargaining

the trial phase; thus, it is considered an extra-​legal factor that juries, with some exceptions, do not hear (People v. Falsetta, 1999). However, criminal history is one of the two main factors used for sentencing guidelines and an important factor in determining sentencing outcomes, which implies that criminal history is likely to impact trial outcomes and pleas differently (Spohn, 2000; Tonry, 1987). Another critique of the shadow theory is that it is based on a key assumption: that criminal justice actors act rationally. However, many studies have shown that individuals generally, and professionals in the criminal justice system specifically, do no act in strictly rational ways (Guthrie, Rachlinski, & Wistrich, 2001; Plous, 1993; Tversky & Kahneman, 1974). For example, Bjerk (2005) argued that the structure of plea bargaining and the use of harsh penalties will “force” rational people to plead guilty even in cases where they are innocent. Additionally, one study using data from Cook County found that, on average, the plea sentences were actually higher than the expected sentence at trial, suggesting that defendants are acting irrationally more often than not (Abrams, 2011). Finally, there are many structural factors that influence the plea bargaining process, such as the skills of the individual attorneys involved, sentencing laws, and a lack of understanding on the part of the defendants (Bibas, 2004). Thus, there are many other factors coming into play that are not considered in the current, simplistic version of the shadow model. With concerns raised about what variables are most relevant to the shadow of the trial model, there is a clear need for future empirical testing of this model to further specify this theory and identify how the shadow model can provide an alternate explanation of plea decision making.

Additional Factors Related to Plea Decision-​Making Regardless of the macro-​level approach (e.g., focal concerns or shadow of trial model) taken to understand plea decision-​making, there are myriad factors that affect defendants’ and attorneys’ decisions to accept, offer, and advise pleas. Such factors may or may not be subsumed within the larger models. In this part we focus on two factors; prosecutorial discretion in plea bargaining and the role of evidence in the shadow model.

Prosecutorial Discretion in Plea Bargaining In the plea bargaining process, prosecutors have largely unchecked discretion to decide the number and severity of charges to file, to offer plea deals, and to recommend the sentencing outcomes following the acceptance of a guilty plea (Bibas, 2004; Pfaff, 2017). While the shadow model is not concerned explicitly with prosecutors’ behavior, the model theorizes that prosecutors, defense attorneys, and defendants base their plea decisions on the evidence that they anticipate jurors will hear and weigh. Additionally, studies of the shadow model have found evidence that prosecutors do often bargain in the shadow of trial (see Bushway, Redlich, & Norris, 2014; Pezdek & O’Brien, 2014; Redlich, Bushway, & Norris, 2016). For example, in an early study, prosecuting attorneys claimed that “evidence strength” moderated their decision to offer or to accept plea deals (Horney, 1980). Additionally, prosecutors may lower the plea discount below the break-​even point to pressure defendants into pleading guilty as a way to minimize the number of trials (Alschuler, 1979, 2002; Bushway, 2019). Little is known about this highly discretionary aspect of the criminal justice process, and research is needed to identify the decision rules that prosecutors use in making plea bargaining decisions and to pinpoint the legal and extra-​legal factors that affect those decisions (Spohn, 2018). In addition to their charging abilities, prosecutors also have the power to control the evidence to a degree, in that it is the responsibility of the prosecutor to disclose material relevant to a case (Alschuler, 1968). In some cases, the defendant and defense attorney may lack personal knowledge of the evidence against the defendant, and thus may rely on the prosecutor to provide the evidence needed to make a plea decision, giving them more power in the plea bargaining process (Bibas, 2004). One study found that prosecutors offer more punitive charge bargains when certain types of 173

Amy Dezember and Allison D. Redlich

evidence are present, such as eyewitness identification and video or photo evidence, though the evidence may be more important to the decision to charge rather than the decision to offer a plea deal (Kutateladze, Lawson, & Andiloro, 2015; Frederick & Stemen, 2012; Redlich, Bushway, & Norris, 2016). Similarly, in line with the shadow model, one study found that prosecutors’ willingness to offer pleas (in a hypothetical case) significantly increased when there was eyewitness evidence, indicating that prosecutors are likely aware that jurors favor direct evidence (Redlich, Bushway, & Norris, 2016). Despite their importance to the plea bargaining process, there is still an overall lack of research on prosecutors’ charging and plea bargaining decisions (Spohn, 2018).

The Role of Evidence in Guilty Pleas As previously mentioned, the shadow model argues that decisions to offer, accept, or reject pleas derive from the perceived probable outcome of a trial, which in turn is driven by the strength of the evidence (Bushway & Redlich, 2012). In fact, the terms “probability of conviction” and “strength of evidence” have been used interchangeably (Kramer, Wolbransky, & Heilburn, 2007; McAllister & Bregman, 1986). Under the shadow model, researchers have evaluated the impact various types of evidence can have on discretion at various stages of the plea bargaining process. This is a notoriously difficult area of research because access to the plea negotiation process is very limited since negotiations are handled outside the courtroom and are not typically formally recorded into court records (Frederick & Stemen, 2012). One way that researchers have attempted to overcome this challenge is by looking at the role of evidence in pleas, which often translates into focusing on the existence of a plea bargain (i.e., the decision to plead guilty or go to trial, for example), rather than its value (Emmelman, 1998). Strength of evidence has been found to drive convictions both at trial and through plea bargaining (Devine et al., 2001; Emmelman, 1998). For example, prosecuting and defense attorneys claim that evidence strength moderates their decision to offer or to accept plea deals (Emmelman, 1998; Kramer, Wolbransky, & Heilburn, 2007). Evidence appears to have a particularly strong impact at the ­initial case screening and in the prosecutor’s decision to bring charges against a defendant (Albonetti, 1997). This makes sense given that the early stages of a case require prosecutors to evaluate the case based on their perceived probability of achieving a conviction, meaning they need to examine the existing evidence against the defendant and the likelihood of getting a conviction based on that evidence.This also indicates that prosecutors will consider the evidence during the plea bargaining process, and that the plea offers they extend will be based on their assessment of the likelihood of conviction at trial (Kutateladze, Lawson, & Andiloro, 2015). Further, a study of defense attorneys responding to a hypothetical scenario found the “likelihood of conviction based on evidence” to be the most important factor in their decisions to recommend plea bargains (Edkins, 2011). Based on these findings from studies using a variety of different methods, it is highly probable that evidence plays a key role at various points throughout the plea decision-​making process. But not all evidence is the same in quality or perceived utility in securing a conviction, and thus distinctions by evidence type are necessary.

Direct Versus Circumstantial Evidence Research has also looked at how specific types of evidence influence legal-​decision making (e.g., Heller, 2006; Kassin & Neumann, 1997; Thompson, 2006). There are indications that the type of evidence may matter more than the amount of evidence when it comes to legal decision-​making (O’Neill, 2007).There is a distinction between direct evidence, which proves a fact without inference or presumption, and circumstantial evidence, which can be used to infer whether the facts in dispute existed or did not exist (e.g., Heller, 2006). Confessions and eyewitness identification are the quintessential examples of direct evidence, while all types of forensic evidence (e.g., DNA, fingerprints, 174

Plea Bargaining

etc.) are considered to be circumstantial. DNA evidence has been considered the “gold standard” of evidence (Thompson, 2006). Although circumstantial evidence is probabilistically more likely to be reliable than direct evidence, jurors and judges consistently overvalue direct evidence (Heller, 2006; Niedermeier, Kerr, & Messe, 1999; Wells, 1992). This indicates that it is important to consider how different types might influence case outcomes not only at trials but also in guilty pleas. Direct evidence, such as confessions, has been considered by many to be the most potent form of evidence, and one study found that jurors appeared to weigh confessions more heavily than eyewitness testimony (Kassin, Bogart, & Kerner, 2012; Thompson, 2006). Eyewitness evidence, which is also a form of direct evidence, has also been found to be highly valued, especially when contrasted with circumstantial evidence. In experimental studies that have manipulated corroborating eyewitness testimony with mock juries, strength of evidence has been shown to have a strong positive association with jury verdicts of guilt, with a conviction rates ranging from 24 to 70% across various studies (Devine et  al., 2001; Greene, 1988). However, DNA evidence is a form of circumstantial evidence (Heller, 2006). Studies examining the weight of DNA evidence indicate that mock jurors do sometimes undervalue DNA, and do not understand or incorrectly apply statistical probabilities (Koehler, 2001; Schklar & Diamond, 1999; but see Lieberman et al., 2008). There is also some evidence from real cases that confession evidence is valued higher than DNA evidence (see Kassin, 2012). This indicates that despite direct evidence being more problematic, this type of evidence can play a weightier role in a defendant’s perceived outcome at trial and potential decision to plead guilty.

Concerns with Measurement of Evidentiary Factors If plea bargaining does in fact occur in the “shadow of the trial,” then evidentiary factors should also affect the value of the plea discount. However, testing the role of evidence in the plea decision can be challenging and plagued with measurement issues. For example, when attempts to measure strength of evidence have been made, they tend to be overly simplistic and/​or subjective. First, most of the available data on evidence as a factor in convictions comes from self-​reports of legal actors. Studies have found that prosecuting and defense attorneys claim that evidence strength moderates their decision to offer and accept plea deals (Emmelman, 1998; Horney, 1980). Measuring strength of evidence is often criticized for being overly subjective, and therefore susceptible to discretion or bias (Bushway & Redlich, 2012). Second, previous measures of strength of evidence often aggregate evidence, often without knowing anything about the evidence per se (e.g., whether it is reliable or inculpatory). For example, one study used a summary measure for all pieces of evidence, including the presence of physical evidence, confession, positive eyewitness identification, weapon, number of indictment charges, degree of injury to victim, and the number of witnesses (Taylor & Hosch, 2004). Because these pieces of evidence were not examined separately, it was not possible to determine whether any one factor was sufficient to predict convictions. Additionally, strength of evidence may be determined by the quantity and quality of evidence, and thus it is important to look at individual pieces of evidence as well as the overall amount. Recent research has demonstrated that individual pieces of evidence are not viewed independently but rather are more interdependent than previously thought (Kassin, 2012; Kassin, Bogart, & Kerner, 2012). Despite these challenges, Bushway and Redlich (2012) found that, while certain types of evidence (i.e., eyewitness, confession, and physical evidence) were able to explain the probability of conviction for defendants who go to trial, it did not similarly explain variation in the estimates of the probability of conviction for those who pled guilty.This indicates that there is a need for additional research to further specify the role of evidence in plea bargaining, as well as determine what factors can help explain variation in plea decision-​making. The shadow of the trial model provides an alternative way to examine plea bargaining by studying how the expected outcomes at trial may impact the decision to plead guilty. However, the few empirical tests of this model have resulted in mixed findings on support for this model, leaving many 175

Amy Dezember and Allison D. Redlich

questions remaining about the validity of this model. For example, there are concerns regarding the measurement of the strength of the evidence, a key factor in the shadow model, which need to be addressed in order to improve our testing of this model. Additionally, one main criticism of this model is that it is too simplistic and does not include structural, jurisdictional, and psychological considerations that have been shown to impact the plea decision-​making process. Critics of this model have argued that additional research is needed to identify what variables are most relevant to the shadow model in order to develop a more robust model.

New Data Exploring Shadow Model Factors In order to further explore the role of evidence in the shadow model, we present some new, preliminary data to examine various factors included in the shadow model, specifically the prosecutor’s proffer of evidence during plea hearings; that is, evidence against the defendant that would have been used at trial with an aim towards conviction. In this manner, we can begin to address whether strength of the evidence impacts plea discounts and influences the value of these discounts in a way that previous studies have not yet explored. While the prosecutor is not required to proffer all pieces of evidence against the defendant at the plea hearing, most judges request that they present the key elements that would be presented as evidence at trial to ensure that there are sufficient facts to demonstrate the defendant’s guilt. For this reason, we will focus on the evidence that was proffered during the plea hearing in order to examine how these factors, which are key components of the shadow model, may influence the discounts that defendants receive as a result of a plea. The aims of this analysis are to understand how the strength of evidence influences the sentencing discounts defendants receive, and provide insight into whether defendants consider the evidence against them when pleading guilty.

Data and Methods Data were obtained from systematic observations of plea hearings, examination of court records, and interviews with defendants within a month of their plea hearing in a medium-​sized suburban county with a population of roughly 450,000. This county has approximately 8,000 adult arrests each year, with just over 6,000 felony-​level crimes occurring in 2016 (Federal Bureau of Investigation, 2016). First, we conducted observations of court proceedings that occurred at the county’s Circuit Court, indicating that defendants had been charged with at least one felony. A team of trained undergraduate and graduate research assistants observed plea hearings and systematically coded what occurred during the hearings. Each week, 20 to 30 plea hearings are scheduled in Circuit Court (although some are continued to a later date), with the plea docket occurring weekly at the courthouse. Second, after plea hearings were observed, official court records for the cases were gathered from an online database or using the public computers at the county courthouse. Third, interviews with defendants whose plea hearings we observed occurred either at the county jail for defendants in custody or in public spaces (e.g., a library) for defendants in the community. Defendants whose pleas we observed that were between 18 to 35 years of age and spoke English fluently (i.e., without the assistance of an interpreter) were considered eligible and were approached to participate in the study. Data collection is ongoing but the data used here were collected between February 2017 and March 2018; during this time, researchers observed 377 criminal court plea hearings. Of these observed pleas, the court accepted 370 pleas. Five pleas were not accepted by the judge, due to issues or complications that arose during the plea hearing (e.g., attorney did not have the correct plea forms, judge had concerns about the sentencing statutes that needed to be sorted out before plea was accepted). In the other two cases, the defendants withdrew their pleas during the hearing before it could or could not be accepted by the court because the state still had not received lab results for 176

Plea Bargaining

the evidence. Of the 370 pleas that were accepted by the Court, 307 defendants had been sentenced by the time this chapter was written. During the plea hearings, researchers coded for a variety of elements that may or may not be present at each plea hearing. For example, researchers coded demographics of the defendant and court actors, the charges to which the defendant was pleading, the type of plea entered, and elements of the plea colloquy. Additionally, the official court records provided more detailed information about the initial charges filed against the defendant, the charges to which the defendant pled guilty, and the final sentence given to the defendant. These records also provide official information about the maximum incarceration sentences for each charge, based on state law, which were used to calculate plea-​sentencing discounts. Researchers also coded the proffer of evidence by the prosecutor, including the types of evidence presented, and evaluated the strength of the evidence proffered (i.e., strong, medium, or weak). For example, strong evidence was operationally defined as: “it appears to be very clear that the defendant committed the crime for which s/​he is charged. For example, if a police officer witnessed the suspect committing a crime (e.g., driving recklessly) or if there was some kind of “hard” evidence such as DNA, photographs of the defendant committing the crime, etc.” An example of a case with strong evidence would be a defendant who stole merchandise from a store and was stopped by an officer on the way out of the store with the goods and there was surveillance video of the defendant concealing stolen goods. In contrast, weak evidence was defined as, “it does not seem like the evidence strongly links the defendant to the commission of the crime. This could either mean that there is a lack of evidence or that the evidence that exists is not very strong/​reliable.” An example of weak evidence would be if an officer located a gun in a backpack in the trunk of the car that the defendant was in, but the defendant claimed the backpack and gun were not his. In this case, there was nothing linking the defendant explicitly to the possession of the gun, other than it being in the trunk of his car. To our knowledge, this is the first study that has used this measure of strength of the evidence—​one that is based on the prosecutor’s proffer during the plea hearing. Although admittedly an imperfect measure, it may alleviate some of the problems noted above that are associated with measuring this amorphous and hard-​to-​ operationalize concept. The last form of data analyzed here resulted from interviews with a subset of individuals (n = 72) within a month of their observed plea to learn more about their decision-​making processes, including their own judgments of the strength of the evidence against them and their reasons for taking a plea. The interview included questions that required participants to consider the evidence against them and evaluate its strength, their reasons for pleading guilty, and other factors that might have impacted their decision (i.e., pretrial detention, advantages of pleading guilty over going to trial, effects of pleading guilty on their constitutional rights). These questions provided a fuller and more qualitative picture of the factors that may have played a role in the defendant’s decision to plead guilty.

Analysis and Findings Table 8.1 presents descriptive statistics for defendant characteristics in the study sample. Of the pleas observed, 96.4% included a traditional guilty plea, 2.6% included a no contest plea, and 2.3% included an Alford plea (a plea which allows defendants to assert innocence but still plead guilty).1 As previously mentioned, all of the defendants in our sample were initially charged with a felony; however, only 83.4% pled to at least one felony, while 29.6% pled to misdemeanor-​only level offenses. Almost 80% of the sample were men and 39% were white. The average age of defendants in the sample was 32 years old and the average education was just below a 12th grade or a GED-​equivalent education. Additionally, approximately 47% of defendants were in custody immediately prior to their guilty plea and 62% of the defendants received a custodial sentence as part of their actual pled-​to sentence. Based on the evidence proffered during the plea hearing, 84.2% of cases were coded as having strong 177

Amy Dezember and Allison D. Redlich

evidence against the defendant, 13.9% were coded as having medium evidence, and 1.9% were coded as having weak evidence. Since the medium and weak categories had much lower percentages (as compared to strong), these two categories were collapsed together so that the final coding was strong versus medium/​weak evidence. Table 8.2 shows the most serious charges on which defendants were indicted and to which they pled guilty. Person crimes included assault, robbery, sex offenses, murder, kidnapping, and other crimes against a person. Property crimes included burglary, larceny, fraud, auto theft, forgery, and embezzlement. Drug charges included possession, distribution, and other charges involving illegal possession and selling of drugs. Finally, minor crimes or misdemeanor-​only level crimes included DUIs, disorderly conduct, probation violations, and court-​related citations, though these charges had originated as felony offenses. There are several possible ways to quantify a defendant’s plea discount. First, there is the maximum incarceration sentence that a defendant faces based on the indicted charges. Table 8.1 shows that defendants in our sample were indicted on 3.0 charges (SD = 2.57), on average, and faced a maximum incarceration sentence length of 462.7 months (SD = 465.25), or 38.6 years, based on their indicted charges. Second, since charges can be (and most often were) dropped or amended to lesser charges in exchange for a guilty plea, there is the maximum incarceration sentence based on the charges to which they pled guilty. Defendants in our sample pled to an average of 1.7 charges (SD = 1.12) that carried an average maximum incarceration sentence of 215.6 months, or almost 18 years (Table 8.1). Finally, there is also the actual sentence that the defendant received as a result of pleading guilty, which for the purposes of this analysis is defined as the number of months of incarceration the defendant is sentenced to serve after taking the plea. The defendants in our sample were sentenced to an average length of active incarceration (i.e., length of time in jail or prison with no

Table 8.1  Descriptive Statistics for Sample (n = 307) Variable

Mean /​%

Defendant Characteristics Male 76.9% White 39.7% In Custody Prior to Plea 46.9% Age 32.0 Education 11.7 Type of Plea Guilty Plea 96.4% No Contest Plea 2.6% Alford Plea 2.3% Severity of Charges Pled to At Least One Felony 83.4% Pled to Misdemeanor Only 29.6% Number of Indicted Charges 3.0 Number of Pled-​to Charges 1.7 Dependent Variable Strength of Evidence, Strong 84.2% Sentencing Measures Max. Sentence, Indicted Charges (in months) 462.7 Max. Sentence, Pled-​to Charges (in months) 215.6 Average Actual Sentence Length (in months) 11.4 Plea Discount (proportion of discount from indicted 0.96 sentence max to actual sentence) Received a Custodial Sentence 62.2%

178

SD

Min.

Max.

0.42 0.49 0.50 10.94 2.20

0 0 0 17 3

1 1 1 65 24

0.19 0.16 0.15

0 0 0

1 1 1

0.83 0.30 2.57 1.12

0 0 1 1

1 1 24 8

0.37

0

1

465.25 238.46 26.83 0.10

60 12 0 0

4128 1620 252 1

0.49

0

1

Plea Bargaining

time suspended) of 11.4 months, or less than 1 year (see Table 8.1). This indicates that the defendants received, on average, a 96% discount (SD = 0.10) from the maximum incarceration sentence based on the indicted charges (i.e., 11.4 months versus 46.2 months). To exemplify, one defendant in our sample was indicted on six separate charges, which together carried a maximum incarceration sentence of 1,440 months. This defendant pled guilty to one charge that carried a maximum incarceration sentence of 60 months and was sentenced to 20 months’ incarceration (a discount of more than 98% from the maximum possible sentence at indictment). At one extreme, only two defendants did not receive a discount (i.e., the sentence imposed was the same as the maximum sentence based on the charges at indictment). At the other extreme, even though every defendant in our sample was initially charged with at least one felony that carried potential prison time, 141 defendants received no carceral sentence (a 100% discount).

Strength of the Evidence To address whether strength of the evidence influences the plea discounts received, three OLS regressions were conducted. Discount measures were computed as proportions. For example, to calculate the maximum -​indicted-​to-​actual-​sentence discount, we subtracted the actual sentence from the maximum-​indicted sentence and divided it by the maximum-​indicted sentence. In other words, if a defendant faced a maximum-​indicted sentence of 1,200 months and received an actual sentence of 24 months, their indicted to actual sentence discount would be calculated as follows: (1200–​24) /​ 1200 = .98, or a 98% discount. This method was used to calculate all of the discount measures (i.e., max-​indictment to max-​plea discount, max-​indicted to actual-​sentence discount, and max-​plea to actual-​sentence discount). We calculated a separate OLS regression model for each of these outcome measures and used strength of evidence, defendant sex, race, age, whether the defendant was in custody prior to plea, and the most serious indicted charge as predictors. The strength of evidence was initially coded as strong (1), medium (2), and weak (3) based on the prosecutor’s proffer during the hearing. However, as previously mentioned, the medium and weak categories were collapsed for analysis (i.e., 1 = strong, 0 = medium/​weak). Defendant’s sex (1 = male, 0 = female), race (1 = white, 0 = non-​white), and age were coded during observations.The defendant’s custody status was coded as in custody (1) or not in custody prior to plea (0). Finally, the most serious indicted charge was coded into dummy variables: person crime (1 = yes, 0 = no), property crime (1 = yes, 0 = no), drug crime (1 = yes, 0 = no), and minor crime (1 = yes, 0 = no). Table 8.3 displays the results of all three OLS regression models.The first model uses the maximum-​ indicted to maximum-​plea sentence discount measure as the dependent variable. This model was significant, F(8, 256) = 2.102, p = .036, with an R2 of .062. However, the strength of evidence variable was only significant at the .10 level (β = -​.097, p = .083). Of the other variables included in the model, the defendant’s age (β = -​.006, p ≤ .01) and if the defendant was in custody prior to the plea (β = -​.083, p ≤ .05) significantly reduced the maximum-​indicted to maximum-​plea discount a defendant received. Older defendants and defendants who were in custody received smaller plea

Table 8.2  Most Serious Charge (n = 307) Crime Type

Person Crime Property Crime Drug Crime Minor Crime

Indicted Charge

Pled-​To Charge

N

%

N

%

76 103 105 23

24.8 33.6 34.2 7.5

70 83 85 69

22.8 27.0 27.7 22.5

179

Amy Dezember and Allison D. Redlich Table 8.3  OLS Regressions of Various Sentencing Discount Measures Variable

Max-​Indicted to Max-​ Plea Sentence Discount (S.E.)

Max-​Indicted to Actual-​ Sentence Discount (S.E.)

Max-​Plea to Actual-​ Sentence Discount (S.E.)

Strength of Evidence, Strong

-​.097† (.056) -​.007 (.051) .009 (.044) -​.006** (.002) -​.083* (.042) .050 (.056) -​.008 (.051) .114 (.087) .636 (.101) .062

-​.015 (.017) .014 (.015) -​.009 (.013) .001 (.001) -​.029* (.013) -​.052** (.017) -​.014 (.015) -​.071** (.026) .983 (.031) .084

-​.009 (.022) .021 (.020) -​.005 (.017) .001 (.001) -​.034* (.017) -​.094*** (.022) -​.027 (.020) -​.102** (.035) .942 (.040) .116

Defendant Male Defendant White Defendant Age Defendant in Custody Prior to Plea Hearing Most Serious Indicted Charge, Person Crime Most Serious Indicted Charge, Property Crime Most Serious Indicted Charge, Minor Crime Constant R2

p ≤ .10; *p ≤ .05; ** p ≤ .01; *** p ≤ .001 Note: “Most Serious Indicted Charge, Drug Crime” has been excluded from all three models as the reference category. †

discounts than their counterparts. However, this model only explains a small percentage (6.2%) of variation, which indicates that other important variables are not being captured. For example, one variable that is not included is the defendant’s criminal history, which is known to influence sentencing (Ulmer & Bradley, 2006). Although we have not yet coded criminal histories of our sample, we do have access to this information and plan to include these data in future analyses. The second model uses the maximum-​ indicted to actual-​ sentence discount measure as the dependent variable. The regression equation was significant, F(8, 256) = 2.946, p = .004, with an R2 of .084. In this model, strength of evidence was not significantly related to the outcome measure. However, there were three variables that significantly influenced the discount measure: the plea discount was significantly reduced if the defendant was in custody prior to their plea (β = -​.029, p ≤ .05), if the most serious indicted charge was for a person crime rather than a drug offense (β = -​.052, p ≤ .01), and if the most serious indicted charge was a minor crime rather than a drug offense (β = -​ .071, p ≤ .01). This model does a slightly better job at explaining the overall variation in the outcome measure, but still only explains a small percentage of the variation in the maximum indicated sentence to actual sentence discount measure (8.4%). The last model, using the maximum-​plea sentence to actual-​sentence discount measure as the dependent variable, was also significant, F(8, 256)  =  4.191, p ≤ .000, with an R2 of .116. Again, strength of evidence was not significantly related to the outcome measure. Consistent with the results presented in the second model, whether the defendant was in custody prior to the plea (β = -​.034, p ≤ .05), if the most serious indicted charge was a person crime (β = -​.094, p ≤ .001), and if the most 180

Plea Bargaining

serious indicted charge was a minor crime (β = -​.102, p ≤ .01) all reduced the plea discount measure used in this model. Finally, while this model does the best job of accounting for the most variance, it still accounts for a small percentage (11.6%) of the variance in this sentence discount. This, again, indicates that there are other factors impacting this sentence discount not accounted for in this regression model.

Defendant Interview Responses Our last set of analyses concern the defendant interviews. Relevant to our purposes here are defendants’ self-​evaluations of the evidence against them. Of 72 defendants, only 40.3% believed that the evidence against them was strong, while 26.4% responded that they believed the evidence was somewhere between weak and strong and another 29.6% responded that they believed the evidence was weak.2 Perhaps unsurprisingly, defendants’ perceptions of the strength of evidence differed from the observers’ coding of the prosecutor’s evidence. As noted above, coders rated 84% of the proffered evidence as strong. When asked why they chose to take the plea, numerous defendants mentioned that the evidence influenced their decision to accept the plea. For example, one defendant said,“It was a rock-​solid case. The evidence led to me.” Similarly, another defendant said, “Since they had the lab evidence, I felt I had to [take the plea].” Some defendants also noted that they felt they had to accept the plea based on conversations with the prosecutor. For example, one defendant stated, “I had two cases [going on] at the same time. If I didn’t plea to one them, they’d smoke me on the other. Prosecutors said that.” Another defendant stated, “The prosecutor said I was late to the hearing and it was already continued once so I needed to make the decision that day.”These interview responses provide some context into how the evidence may be factoring into defendants’ decision-​making. In summary, these new data, although preliminary and not without limitations, provide some insight into how and whether strength of the evidence impacts pleas or whether other factors might be better predictors of plea discounts. In its simplest form, the shadow of the trial model relies solely on the strength of the evidence, yet critics and early empirical tests of this model indicate that there are other factors that need to be incorporated into the model to more fully account for variation in plea discounts. For example, looking at court actors, such as prosecutors, to understand if they have any impact on the value of plea discounts (e.g., distance between trial and plea sentences) may provide for a fuller picture of what occurs when plea bargains are negotiated. Since there are many factors that could influence the bargaining process, it is important to consider ways in which the shadow model can be strengthened through future research and empirical testing.

Strengthening the Shadow Model Paradigm Since the shadow model was only recently introduced to criminology, it is important to consider what this model adds to the conversation surrounding plea bargaining. One thing that is clear is that there is limited research and empirical testing of this paradigm. As previously mentioned, one of the major criticisms of the shadow model is that it is too simplistic, which calls into question the face validity of the model (Bibas, 2004; Redlich & Edkins, 2019). This can also be seen in the relatively low R2 values in our analysis, which indicate that there are variables not included in the models that affect sentence discounts, such as the defendant’s criminal history, the length of time from indictment to plea, and specific types of evidence. These are factors that we are currently collecting and plan to include in future models. However, we acknowledge the need for future research on the shadow model in order to test more sophisticated versions of the theories by exploring additional variables to include in the model. Since the model is often criticized for being too simplistic, research should aim to strengthen the measurements of variables already included in the model, such as evidentiary variables, and also integrate other variables that may impact the defendant’s decision to plead guilty, 181

Amy Dezember and Allison D. Redlich

such as structural features of the jurisdiction (Bibas, 2004; Ulmer, Eisenstein, & Johnson, 2010). By strengthening and improving the shadow model, we will gain a better understanding of how this model works in practice and how it can help explain outcomes in plea bargaining.

Improving Methodologies to Test the Shadow Model One of the main ways in which the utility of the shadow model can be improved is to conduct additional tests of the shadow model using real court records and case data to ensure the realistic nature of the outcomes. However, one of the biggest challenges to this approach is that the expected outcome at trial or the probability of conviction are not readily available in case files. One way to address this challenge is for researchers to attempt to capture valid estimates of the expected value of the trial sentence during actual plea negotiations as they occur (Bushway, Redlich, & Norris, 2014). Of course, this method would have its own limitations as court actors may face legal and logistical barriers to providing full transparency to this type of information for ongoing cases. As case data and court records become more systematic and available to researchers, this approach may become more feasible for scholars and may provide future research to generate more direct insight into how cases are resolved. Experimental designs may also provide a more realistic and controlled testing of the shadow model as a way to simulate negotiations between actors, which could resemble a research equivalence to a mock trial for the plea bargaining context (see Bushway, Redlich, & Norris, 2014; Devine et al., 2001). Experimental designs are beneficial in that they allow the researcher to isolate and control particular variables in the shadow model to identify and explain variation observed in the plea outcomes. This can be especially useful in trying to further specify the shadow model and identify which factors are most valuable and useful in better understanding how plea bargains work. However, one limitation to this approach may be that the behavior of actors may differ substantially from their responses in a survey versus when confronting the real constraints of the criminal justice system. Future research using hypotheticals or experimental designs should explore alternative ways to solicit information from actors about the expected trial outcomes to better inform how to improve the shadow model.

Better Understanding of the Impact of Evidence The strength of the evidence is a major factor in the shadow model and thus requires ample research and testing to understand the impact it has on the plea process and plea outcomes. Based on the new data presented in this chapter, we were not able to find a statistically significant relationship between the strength of the evidence and the sentence discount a defendant receives for pleading guilty. However, this could be due to limitations in the data, such as the lack of variation in the strength of evidence variable (i.e., that most cases were coded as “strong”) and the fact that our models did not include specific types of evidence proffered (e.g., DNA, video/​photo evidence, etc.), though we are collecting this information and plan to include it in future analyses. For this reason, we caution against drawing strong conclusions about the overall lack of a connection between evidence and plea outcomes. Instead, we argue that further research is needed to more accurately measure evidence and improve our understanding of how evidence impacts pleas. Although Bushway and Redlich (2012) were not able to explain variation in estimates of probability of conviction for those who pled guilty using evidence at the individual level in the first empirical test of the shadow model, they also cautioned against reaching definitive conclusions based on their study and encouraged replications and further testing of the shadow model. Thus, it is important to continue exploring the role of evidence in guilty pleas, as well as the strengths and limitations of measuring evidence, in order to better understand how the strength of the evidence may impact outcomes in the plea bargaining process. One element of the shadow model that needs further development is learning more about how individual actors think about the evidence and how the evidence factors into their decision-​making 182

Plea Bargaining

process. With the data presented in this chapter, it is notable that the defendants interviewed often had a very different evaluation of the strength of evidence than the coding that was done during the plea hearing. Based on the proffer of evidence during the hearing, 84% of cases were coded as having strong evidence, whereas only 40% of defendants interviewed self-​evaluated the evidence against them as strong.This provides empirical support for a disconnect between what the prosecutor proffers and how the defendant views the evidence. This disconnect raises questions about whether the strength of evidence should be considered the driving factor in the shadow model. This shows there is a clear need for more research about how the different individuals involved in the case think about the evidence and evaluate its strength, which will also have implications for the shadow model and the forecasting of trial outcomes.

Developing a Fuller Version of the Shadow Model In order for the shadow model to be a useful tool in understanding and explaining pleas, researchers should work to improve the validity of this model by considering bargaining around issues other than sentence length. Only 62% of the defendants in our sample received incarceration as a sentence, which means there are many other factors that are being negotiated (e.g., restitution, community service hours, length of probation, driver’s license suspension length, etc.). The model, as it is, has had limited success in explaining plea outcomes in practice, which may cause some to doubt the usefulness of this model. However, before we discount this model, scholars should work to develop this model into its fullest form by including other variables that have been shown to influence plea decision-​making beyond the simple predicted outcome at trial. As previously mentioned, jurisdictional characteristics, such as caseload, may be just as important to the plea process and can provide added explanatory value to the shadow model (Bibas, 2004; Bushway, Redlich, & Norris, 2014). Additionally, more research is needed to better understand how all courtroom actors working separately and together (i.e., judges, prosecutors, and defense attorneys) behave within this model. It has been argued that defense attorneys are disadvantaged under the shadow model because they are ill-​equipped to estimate the value of the plea or likelihood of a conviction due to informational deficits (Henderson, 2019). Additionally, research looking at judges found that they do not seem to act in a manner that is consistent with the shadow model and that the judges’ tenure can also impact their rates of plea bargains and the discounts they hand out to defendants (Abrams, 2011; Bushway, Redlich, & Norris, 2014). This is by no means an exhaustive list of other factors that influence the plea process, but we will only learn the true utility of this model by testing the inclusion of these various factors to identify if the shadow model really can help explain plea bargaining more effectively than previous paradigms.

Conclusion In this chapter, we presented an overview of how the bargaining in the shadow of trial model developed in legal scholarship and only recently adapted to the field of criminology as a way to better understand plea bargaining and the variation among plea outcomes. However, the results of the few studies testing the shadow model are mixed and it is clear that more research in this area is needed to draw strong conclusions about the validity and utility of this model. There is limited published research aimed at exploring the ways in which bargaining in the shadow of the trial affects the value of the plea bargain and the “distance traveled” during the bargaining process (Spohn, 2018). Furthermore, the shadow paradigm appears to be the leading model to explain outcomes, indicating that the shadow model needs to be thoroughly researched (Bushway & Redlich, 2012). While early research findings might be mixed on the usefulness of this model, additional work to improve how we measure the strength of evidence and consider the quality of individual types of evidence may help strengthen our understanding of how evidence influences plea decisions within the shadow model. Additionally, we can improve the utility of this model by addressing concerns of the validity 183

Amy Dezember and Allison D. Redlich

of this model through further testing and specification to ensure that all relevant factors are included in the shadow model. Using improved methodologies, such as experimental designs or actual case data, can also help to strengthen and improve this model. Continuing to improve the validity of the shadow model can help continue to shed light on how the “shadow of the trial” model operates in our system of pleas.

Notes 1 The pleas are based on the charge, not the defendant. Thus, defendants charged with multiple charges can take different pleas for separate charges heard during the same plea hearing. For example, one defendant pled guilty to a felony charge and no contest to a misdemeanor charge during the same plea hearing. Due to this, the percentages do not add up to 100 percent. 2 Three interviewed defendants did not provide a response to this question, thus the percentages do not add up to 100 percent.

References Abrams, D.S. (2011). Is pleading really a bargain? Journal of Empirical Legal Studies, 8(1), 200–​221. Albonetti, C.A. (1986). Criminality, prosecutorial screening, and uncertainty: Toward a theory of discretionary decision making in felony case processings. Criminology, 24(4), 623–​644. —​—​. (1990). Race and the probability of pleading guilty. Journal of Quantitative Criminology, 6(3), 315–​334. —​—​. (1997). Sentencing under the federal sentencing guidelines: Effects of defendant characteristics, guilty pleas, and departures on sentence outcomes for drug offenses, 1991–​1992. Law and Society Review, 31(4), 789–​822. Alschuler, A.W. (1968). The prosecutor’s role in plea bargaining. The University of Chicago Law Review, 36(1), 50–​112. —​—​. (1979). Plea bargaining and its history. Columbia Law Review, 79(1),  1–​43. —​—​. (1981). The changing plea bargaining debate. California Law Review, 69(3), 652–​730. —​—​. (2002). Straining at gnats and swallowing camels: The selective morality of Professor Bibas. Cornell Law Review, 88, 1412–​1424. Bibas, S. (2004). Plea bargaining outside the shadow of trial. Harvard Law Review, 117, 2463–​2547. Bjerk, D. (2005). Making the crime fit the penalty: The role of prosecutorial discretion under mandatory minimum sentencing. The Journal of Law and Economics, 48(2), 591–​625. Bradley-​Engen, M.S., Engen, R.L., Shields, C., Damphousse, K.R, & Smith, B.L. (2012). The time penalty: Examining the relationship between time to conviction and trial vs. plea disparities in sentencing. Justice Quarterly, 29(6), 829–​857. Brereton, D., & Casper, J.D. (1982). Does it pay to plead guilty? Differential sentencing and the functioning of criminal courts. Law & Society Review, 16(1),  45–​70. Bureau of Justice Statistics. (2010). Felony Defendants in Large Urban Counties, 2006 (Bulletin NCJ 228944). Washington, DC: U. S. Department of Justice. Bushway, S.D. (2019). Defendant decision making in plea bargains. In V. Edkins & A.D. Redlich (eds.), A System of Pleas: Social Science’s Contribution to the Real Legal System. New York: Oxford University Press. Bushway, S.D., & Piehl, A.M. (2001). Judging judicial discretion: Legal factors and racial discrimination in sentencing. Law and Society Review, 35(4), 733–​764. Bushway, S.D., & Redlich, A.D. (2012). Is plea bargaining in the “shadow of trial” a mirage? Journal of Quantitative Criminology, 28(3), 437–​454. Bushway, S.D., Redlich, A.D., & Norris, R.J. (2014). An explicit test of plea bargaining in the “shadow of the trial”. Criminology, 52(4), 723–​754. Daftary-​Kapur, T., & Zottoli, T.M. (2014). A first look at the plea deal experiences of juveniles tried in adult court. International Journal of Forensic Mental Health, 13(4), 323–​336. Devine, D.J., Clayton, L.D., Dunford, B.D., Seying, R., & Pryce, J. (2001). Jury decision-​making: 45 years of empirical research. Psychology, Public Policy, and Law, 7(3), 622–​727. Edkins,V. A. (2011). Defense attorney plea recommendations and client race: Does zealous representation apply equally to all? Law and Human Behavior, 35, 413–​425. Eisenstein, J., Fleming, R., & Nardulli, P. (1987). The Contours of Justice: Communities and Their Courts. Boston, MA: Little, Brown.

184

Plea Bargaining Eisenstein, J., & Jacob, H. (1977). Felony Justice: An Organizational Analysis of Criminal Courts. Boston, MA: Little, Brown. Emmelman, D.S. (1998). Gauging the strength of evidence prior to plea bargaining: The interpretive procedures of court-​appointed defense attorneys. Law and Social Inquiry, 22(4), 927–​955. Engen, R.L., & Steen, S. (2000). The power to punish: Discretion and sentencing reform in the war on drugs. American Journal of Sociology, 105(5), 1357–​1395. Federal Bureau of Investigation (2016). Uniform Crime Report Statistics, 2016. Washington, DC: US Department of Justice. Fisher, G. (2000). Plea bargaining’s triumph. The Yale Law Journal, 109(5), 857–​1086. Frederick, B., & Stemen, D. (2012). The Anatomy of Discretion: An Analysis of Prosecutorial Decision Making. New York:Vera Institute of Justice. Greene, E. (1988). Judge’s instruction on eyewitness testimony: Evaluation and revision. Journal of Applied Social Psychology, 18(3), 252–​276. Guthrie, C., Rachlinski, J.J., & Wistrich, A.J. (2001). Inside the judicial mind. Cornell Law Review, 86, 777–​830. Johnson, B.D., King, R.D., & Spohn, C. (2016). Sociolegal approaches to the study of guilty pleas and prosecution. Annual Review of Law and Social Science, 12, 479–​495. Hartley, R., Maddan, S., & Spohn, C. (2007). Concerning conceptualization and operationalization: Sentencing data and the focal concerns perspective—​A research note. Southwest Journal of Criminal Justice, 4(1),  58–​78. Heller, K.J. (2006). The cognitive psychology of circumstantial evidence. Michigan Law Review, 105, 241–​305. Henderson, K.S. (2019). Defense Attorneys and Plea Bargains. In V. Edkins & A.D. Redlich (eds.), A System of Pleas: Social Science’s Contribution to the Real Legal System. New York: Oxford University Press. Horney, J. (1980). Plea bargaining decision factors. In A. Cohn & B.Ward (eds.), Improving Management in Criminal Justice (Vol. 17, pp. 73–​83). Beverly Hills, CA: Sage. Kaban, B., & Quinlan, J.C. (2004). Rethinking a “knowing, intelligent and voluntary waiver” in Massachusetts’ juvenile courts. Journal of the Center for Families, Children, and the Courts, 5,  35–​55. Kalven, H., & Zeisel, H. (1966). The American Jury. Chicago, IL: University of Chicago Press. Kassin, S.M. (2012). Why confessions trump innocence. American Psychologist, 67(6), 431–​445. Kassin, S.M., Bogart, D., & Kerner, J. (2012). Confessions that corrupt: Evidence from the DNA exoneration case files. Psychological Science, 23(1),  41–​45. Kassin, S.M., & Neumann, K. (1997). On the power of confession evidence: An experimental test of the fundamental difference hypothesis. Law and Human Behavior, 21(5), 469–​484. Kellough, G., & Wortley, S. (2002). Remand for plea: Bail decisions and plea bargaining as commensurate decisions. British Journal of Criminology, 42(1), 186–​210. King, N., Soule, D., Steen, S., & Weidner, R. (2005). When process affects punishment: Differences in sentences after guilty plea, bench trial, and jury trial in five guideline states. Columbia Law Review, 105, 960–​1009. Koehler, J.J. (2001). When are people persuaded by DNA match statistics? Law and Human Behavior, 25(5), 493–​513. Kramer, J.H., & Ulmer, J.T. (2002). Downward departures for serious violent offenders: Local court “corrections” to Pennsylvania’s sentencing guidelines. Criminology, 40(4), 897–​932. Kramer, G.M., Wolbransky, M.M., & Heilbrun, K. (2007). Plea bargaining recommendations by criminal defense attorneys: Evidence strength, potential sentence, and defendant preference. Behavioral Sciences and the Law, 25(4), 573–​585. Kutateladze, B.L., Andiloro, N.R., & Johnson, B.D. (2016). Opening Pandora’s box: How does defendant race influence plea bargaining? Justice Quarterly, 33(3), 398–​426. Kutateladze, B.L., Andiloro, N.R., Johnson, B.D., & Spohn, C.C. (2014). Cumulative disadvantage: Examining racial and ethnic disparity in prosecution and sentencing. Criminology, 52(3), 514–​551. Kutateladze, B.L., & Lawson, V.Z. (2017). Is a plea really a bargain? An analysis of plea and trial dispositions in New York City. Crime & Delinquency, 64(7), 856–​887. Kutateladze, B.L., Lawson, V.Z., & Andiloro, N.R. (2015). Does evidence really matter? An exploratory analysis of the role of evidence in plea bargaining in felony drug cases. Law and Human Behavior, 39(5), 431–​442. Landes, W. (1971). An economic analysis of the courts. Journal of Labor Economics, 14(1), 61–​107. Lieberman, J.D., Carrell, C.A., Miethe, T.D., & Krauss, D.A. (2008). Gold versus platinum: Do jurors recognize the superiority and limitations of DNA evidence compared to other types of forensic evidence? Psychology, Public Policy, and Law, 14(1),  27–​62. McAllister, H.A., & Bregman, N.J. (1986). Plea bargaining by prosecutors and defense attorneys: A decision theory approach. Journal of Applied Psychology, 71(4), 686–​690. McCannon, B.C. (2019). Prosecutors and plea bargains. In V. Edkins & A.D. Redlich (eds.), A System of Pleas: Social Science’s Contribution to the Real Legal System. New York: Oxford University Press.

185

Amy Dezember and Allison D. Redlich Metcalfe, C. (2016). The role of courtroom workgroups in felony case dispositions: An analysis of workgroup familiarity and similarity. Law & Society Review, 50(3), 637–​673. Metcalfe, C., & Chiricos,T. (2018). Race, plea, and charge reduction: An assessment of racial disparities in the plea process. Justice Quarterly, 35(2), 223–​253. Miethe, T.D. (1987). Charging and plea bargaining practices under determinate sentencing: An investigation of the hydraulic displacement of discretion. Journal of Criminal Law & Criminology, 78(1), 155–​176. Miller, H.S., McDonald, W., & Cramer, J.A. (1978). Plea Bargaining in the United States. Washington, DC: National Institute of Justice. Mnookin, R.H., & Kornhauser, L. (1979). Bargaining in the shadow of the law: The case of divorce. Yale Law Journal, 88, 950–​997. Nagel, S.S., & Neef, M.G. (1979). Decision Theory and the Legal Process. Lexington, MA: Lexington. Niedermeier, K.E., Kerr, N.L., & Messe, L.A. (1999). Jurors’ use of naked statistical evidence: Exploring bases and implications of the Wells Effect. Journal of Personality and Social Psychology, 76, 533–​542. O’Neill, T. (2007). Type of Evidence May Outweigh the Strength. Chicago Daily Law Bulletin. Available from www. chicagolawbulletin.com/​home Pezdek, K., & O’Brien, M. (2014). Plea bargaining and appraisals of eyewitness evidence by prosecutors and defense attorneys. Psychology, Crime, and Law, 20(3), 222–​241. Pfaff, J. (2017). Locked In: The True Causes of Mass Incarceration—​And How to Achieve Real Reform. New  York: Basic Books. Plous, S. (1993). The Psychology of Judgment and Decision-​making. New York: McGraw-​Hill. Redlich, A.D., Bibas, S., Edkins, V.A., & Madon, S. (2017). The psychology of defendant plea decision making. American Psychologist, 72(4), 339–​352. Redlich, A.D., Bushway, S., & Norris, R. (2016). Plea decision-​making by attorneys and judges: The impact of legal and extra-​legal factors. Journal of Experimental Criminology, 12(4), 537–​561. Redlich, A.D. & Edkins, V. (2019). Conclusion: Moving forward in a system of pleas. In V. Edkins & A.D. Redlich (eds.), A System of Pleas: Social Science’s Contribution to the Real Legal System. New York: Oxford University Press. Redlich, A.D., Wilford, M.M., & Bushway, S. (2017). Understanding guilty pleas through the lens of social science. Psychology, Public Policy, and Law, 23(4), 458–​471. Rhodes, W. (1979). Plea bargaining: Its effect on sentencing and convictions in the District of Columbia. Journal of Criminal Law and Criminology, 70(3), 360–​375. Schklar, J., & Diamond, S.S. (1999). Juror reactions to DNA evidence: Errors and expectancies. Law and Human Behavior, 23(2), 159–​184. Smith, D.A. (1986). The plea bargain controversy. Journal of Criminal Law and Criminology, 77(3), 949–​968. Spohn, C. (2000). Thirty years of sentencing reform: The quest for a racially neutral sentencing process. In J. Horney (ed.), Policies, Processes, and Decisions of the Criminal Justice System (Vol. 3, pp. 427–​501). Washington, DC: National Institute of Justice. —​—​. (2018). Reflections on the exercise of prosecutorial discretion 50 years after publication of the challenge of crime in a free society. Criminology & Public Policy, 17(2),  1–​20. Spohn, C., & Holleran, D. (2000). The imprisonment penalty paid by young, unemployed black and Hispanic male offenders. Criminology, 38(1), 281–​306. Stuntz, W.J. (2004). Plea bargaining and criminal law’s disappearing shadow. Harvard Law Review, 117, 2548–​2569. Taylor, T.S., & Hosch, H. (2004). An examination of jury verdicts for evidence of a similarity-​leniency effect, an out-​g roup punitiveness effect or a black sheep effect. Law and Human Behavior, 28(5), 587–​598. Thomspon, W.C. (January–​February, 2006). Tarnish on the gold standard: Recent problems in forensic DNA testing. Champion Magazine, 10. Tonry, M. (1987). Sentencing guidelines and their effects. In A. von Hirsch, K.A. Knapp, & M. Tonry (eds.), The Sentencing Commission and its Guidelines (pp. 16–​46). Boston, MA: Northeastern University Press. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–​1131. Ulmer, J.T., (1997). Social Worlds of Sentencing: Court Communities under Sentencing Guidelines. Albany, NY: State University of New York Press. Ulmer, J.T., & Bradley, M.S. (2006).Variation in trial penalties among serious violent offenses. Criminology, 44(3), 631–​670. Ulmer, J.T., Kurlychek, M.C., & Kramer, J.H. (2007). Prosecutorial discretion and the imposition of mandatory minimum sentences. Journal of Research in Crime and Delinquency, 44(4), 427–​458. Ulmer, J.T., Eisenstein, J., & Johnson, B.D. (2010). Trial penalties in federal sentencing: Extra-​guidelines factors and district variation. Justice Quarterly, 27(4), 560–​592.

186

Plea Bargaining Wells, G.L. (1992). Naked statistical evidence of liability: Is subjective probability enough? Journal of Personality and Social Psychology, 62(5), 739–​752. Wright, R.F., & Miller, M.L. (2002). The screening/​bargaining tradeoff. Stanford Law Review, 55(1), 29–​119. Zottoli, T.M., Daftary-​Kapur, T., Winters, G.M., & Hogan, C. (2016). Plea discounts, time pressures, and false-​ guilty pleas in youth and adults who pleaded guilty to felonies in New York City. Psychology, Public Policy, and Law, 22(3), 250–​259.

Cases Cited Brady v. United States. 397 US 742 (1970). People v. Falsetta. 21 CAL.4TH 903. (1999).

187

9 ESTIMATING THE SIZE OF PLEA DISCOUNTS Why Does It Matter? Shi Yan1

Plea bargaining has become the norm, rather than the exception, for the disposition of criminal cases, as over 95% of convictions are the result of guilty pleas (Cohen & Kyckelhahn, 2010; Reaves, 2013). To reach a plea deal with the defense, the prosecutor usually offers a sentence that is less harsh than the sentence the defendant would receive if convicted at trial. The sentence reduction can take several forms. The prosecutor may either recommend a sentence towards the lower bound of the prescribed range, or reduce the severity or counts of the initial charges. The exact form of sentence reduction depends on state laws as well as local practices (Heumann, 1978; Turner, 2017). Some scholars call the disparities “trial penalties” (e.g., McCoy, 2005; Spohn, 1992; Ulmer & Bradley, 2006; Ulmer, Eisenstein, & Johnson, 2010), whereas others call them “plea discounts” (e.g., Abrams, 2011; Bibas, 2004; Redlich et al., 2018; Zottoli et al., 2016).2 Despite the disagreement on how to characterize the disparities, researchers have detected plea discounts or trial penalties in the vast majority of empirical studies on courts (e.g., Brereton & Casper, 1982; Bushway, Redlich, & Norris, 2014; Engen & Gainey, 2000; Johnson, 2003; N. J. King et al., 2005; Spohn, 1992; Ulmer, 2012; Ulmer & Bradley, 2006; Ulmer, Eisentstein, & Johnson, 2010).3 The dominance of plea bargaining in case processing comes with considerable controversy over the past decades (Alschuler, 1979; Fisher, 2000; Smith, 1986). Many scholars and practitioners viewed plea bargaining as a necessary means to manage the caseload and to incentivize remorse (Heumann, 1978; Scott & Stuntz, 1992; Ulmer, 1997). Meanwhile, other scholars have challenged the legitimacy of plea bargaining, on the grounds that it impinges on the constitutional rights of defendants (Alschuler, 1981; Easterbrook, 1992; Langbein, 1979b).These scholars contended that large disparities between trial and plea sentences penalize defendants for exercising their Sixth Amendment right to trial. Under the threat of “trial penalties,” the voluntariness of plea decisions can be questionable (Brunk, 1979; Langbein, 1978; McCoy, 2005). As a result, some commentators have argued for the complete abolition of plea bargaining (e.g., Palmer, 1999; Schulhofer, 1984; Schulhofer, 1992), and many others called for regulations on the bargaining practice (Bibas, 2011; Covey, 2008; Gazal-​Ayal, 2006; Guidorizzi, 1998). Regardless of the standpoint and the policy recommendations for plea bargaining, researchers agree that valid proposals for reform need to build on the correct identification of the problem (Bibas, 2012; Langer, 2006; Scott & Stuntz, 1992; Turner, 2017). Since the controversy over plea bargaining largely stems from the sheer disparity between plea and trial sentences (Covey, 2008; McCoy, 2005; Rakoff, 2014), valid estimates of the size of plea discounts is crucial in understanding the problem, 188

Estimating the Size of Plea Discounts

and in the design of reforms. Despite the near-​consensus on the existence of plea discounts, there is considerable variation among the estimated magnitude of the discounts in the literature.While different laws and local practices may explain some of the variation, a set of methodological issues can also bias the empirical estimates and lead to imprecise conclusions on plea bargaining. Researchers have realized the significance of pre-​conviction decisions in shaping the sentence (e.g., Heaton, Mayson, & Stevenson, 2017; Kutateladze et al., 2014; Rehavi & Starr, 2014). Several recent and influential literature reviews have identified studies on the earlier stages as the crucial next step of research on courts and sentencing (Baumer, 2013; Johnson, King, & Spohn, 2016; Spohn, 2015a; Tata & Gormley, 2016; Ulmer, 2012). The present chapter heeds the call, and reviews both recent progress and remaining challenges in the study of plea discounts. I begin this chapter by summarizing the theories explaining plea bargaining, with particular attention on the significance of the size of plea discounts. I then move on to review the empirical approaches and findings of recent studies. I conclude the chapter by summarizing some general challenges researchers today are facing, as well as potential ways to move forward.

Theories Explaining Plea Bargaining Historically, both legal scholars and practitioners viewed trials, and particularly jury trials, as the default option for the disposition of criminal cases (e.g., Blumstein et al., 1983; Kagan, 2003; Kalven & Zeisel, 1966; Langbein, Lerner, & Smith, 2009; Thayer, 1890;Vidmar & Hans, 2007). Even though plea bargaining has existed for centuries, its rise and dominance is a relatively recent matter (Fisher, 2000; Langbein, 1979b;Vogel, 2007). The trend is similar in research on courts, as most early studies on courts focused on appellate courts and the US Supreme Court (Flango, Roper, & Elsner, 1983). Studies on trial courts—​courtrooms where both trials and plea bargaining take place—​gained popularity in the second half of the 20th century (e.g., Flemming, Nardulli, & Eisenstein, 1992; Heumann, 1978; Nardulli, Eisenstein, & Flemming, 1988). While legal scholars continued to contribute to plea bargaining research over this period (e.g., Alschuler, 1968; Bibas, 2004; Hogarth, 1971; N.  J. King et al., 2005), they were joined by social scientists from a variety of disciplines, most notably political science (Eisenstein & Jacob, 1977), sociology (Ulmer, 2012), psychology (Redlich et al., 2017), and economics (Bushway & Reuter, 2008). The diversity in academic roots has resulted in a variety of extant theoretical frameworks on plea bargaining. In this part, I present the two most influential theories currently used by researchers—​the norm-​based theories on plea bargaining, and the rational choice model “bargaining in the shadow of trial.”4

Norm-​based Theories on Pleas: From the Courtroom Workgroup to Focal Concerns In the 1970s and 1980s, political scientists conducted several influential studies on courtroom dynamics (Eisenstein, Flemming, & Nardulli, 1988; Eisenstein & Jacob, 1977; Flemming et al., 1992; Heumann, 1978; Nardulli, 1978; Nardulli, Eisenstein, & Fleming, 1988; Nardulli, Flemming, & Eisenstein, 1985). These studies built on the organizational paradigm from political science—​namely, how organizational setups and workgroup dynamics impact decisions—​and often adopted a mix of both ethnographic and statistical analysis. The key claim was that “all courtroom workgroups share values and goals,” despite “the apparent conflicts generated by the formal roles of workgroup members—​the prosecutors’ push towards convictions, the defense attorneys’ quest for acquittals, and judges’ inclination toward neutrality” (Eisenstein & Jacob, 1977, pp. 24–​25). These shared values and goals thus generate rules and norms within courtrooms, which may or may not be visible in criminal statutes. Courtroom norms encompass both external goals such as “doing justice” and disposing of the caseload, and internal goals such as maintaining group cohesion and reducing uncertainty (Eisenstein & Jacob, 1977). Compared with trials, plea bargaining generally 189

Shi Yan

better serves the goals of quickly disposing of cases, maintaining group cohesion, and reducing uncertainty.5 Although there has been major controversy on whether plea bargaining serves justice (see later in the chapter for an in-​depth discussion), studies suggest that plea bargaining usually guarantees less harsh sentences for defendants who demonstrate remorse and who admit guilt (e.g., Di Luca, 2005; Etienne & Robbennolt, 2007; Kaiser & Spohn, 2018; Worden, McLean, & Kennedy, 2013). Because of the separation of these norms from statutes and casebooks, new prosecutors and defense attorneys had to learn them as they transition into their professional role. For example, Heumann (1978) found that fresh defense attorneys often underestimated the proportion of defendants who were factually guilty, and were therefore more likely than their senior colleagues to request trials. Experienced prosecutors, on the other hand, rewarded attorneys who were cooperative (i.e., those who preferred pleas) and “penalized” attorneys who insisted on trials (Heumann, 1978, pp. 62–​63): Q:  How about new attorneys? How do you establish … [omission here original] A:  [W]‌hat I try to tell every new attorney who comes in … I try to “steer them straight.” I’ll call them in on the first case we’re dealing with and say: “You can file all the motions you want, harass me any way you want, but you … in the long run are not getting anything. Or, do you want to come in, I’ll tell you what my file has, I’ll show you what my file has, and we can talk about the case. It’s your choice.” Q:  Let’s say he … hassles you, and then he later comes in and wants to negotiate the case? A:  [I]‌f it was a case you would have settled for three to five [years] eventually, you may say “All right, five to ten.” If he doesn’t want it, you will go on trial, and you know this attorney is scared to death … “Look, you’ve harassed me, now you’re going to be harassed … Number one, you can’t win. Number two, you’re really not prepared to try.” The past two decades saw an infusion of sociological insights into the organizational goals and norms perspectives. Researchers regarded courts as “social worlds” in addition to its organizational role in the bureaucratic system (Dixon, 1995; Ulmer, 1997). As a part of the transition, research on sentencing adopted the focal concerns perspective to explain court decisions (Steffensmeier, Ulmer, & Kramer, 1998). This perspective views the decision-​making of courtroom actors as the holistic product of three focal concerns: the defendant’s blameworthiness, the need to protect the community from dangerous offenders, and the practical constraints on and consequences of decisions. Numerous studies have since used this perspective to explain extra-​legal disparities, particularly the less favorable sentencing outcomes received by defendants who were young, black, and male, suggesting that these defendants are more likely to be stereotyped as having higher levels of blameworthiness and risk (e.g., Demuth & Steffensmeier, 2004; Hartley, Maddan, & Spohn, 2007; Huebner & Bynum, 2006; Kurlychek & Johnson, 2004). More recently, scholars have also adopted the focal concerns framework to explain the disparity between sentences imposed at trial and at plea (Ulmer & Bradley, 2006; Ulmer, Eisenstein, & Johnson, 2010), or the extra-​legal disparities among the size of plea discounts (e.g., whether defendants of certain races received larger “rewards” than others when pleading guilty, see Kutateladze, 2018; Metcalfe & Chiricos, 2018; Shermer & Johnson, 2010; Sutton, 2013). The key argument is that a trial can reveal more “bad facts” associated with the crime and the defendant, and the rejection of plea itself can be regarded as the lack of remorse (Ulmer, 1997; Ulmer & Kramer, 1996). Both factors can eventually lead to higher assessed levels of blameworthiness and risk for defendants who opt for trials, which further result in harsher sentences if guilty verdicts are returned. Moreover, the courtroom rules and norms against trials—​regardless of the caseload—​serve as a practical constraint that favors defendants and defense attorneys who plead. While these points explain why defendants generally receive more favorable outcomes at pleas, the association between stereotyping and the focal concerns can further explain why some defendants receive more favorable plea deals than others (see paragraph above and “The Empirical Estimates” section below). 190

Estimating the Size of Plea Discounts

These norm-​based theories on the courtroom workgroup and focal concerns have strong face validity, as a number of works over the decades have confirmed the role of norms in the processing of criminal cases (e.g., Bogira, 2005; Casper & Brereton, 1984; Church, 1985; Clair & Winter, 2016; Hester, 2017; Lynch & Omori, 2014; Young, 2013). These theories predict that local courtroom norms, or sometimes known as “going rates,” are the ultimate determinant of the size of plea discounts. Although this argument makes sense, the term “local norms” can be difficult to evaluate empirically. Furthermore, some ambiguity remains within this theoretical framework—​for example, how much difference between the trial and plea sentences is necessary to fulfill the three focal concerns, and how much is “too much” or “too little?” These questions require an alternative theory to answer, which I present in detail below.

The Rational Choice Model: “The Shadow of Trial” In the similar time frame as the acceptance of the courtroom norms framework, a different group of legal scholars adopted the rational choice framework from economics, and proposed the theory “bargaining under the shadow of trial” (hereinafter referred to as “the shadow model” for simplicity, see Cooter, Marks, & Mnookin, 1982; Landes, 1971; Mnookin & Kornhauser, 1979; Nagel & Neef, 1979). Researchers have applied this framework to both civil and criminal litigations, and it has since remained an influential perspective in legal research (e.g., Galanter, 1990; Lacasse & Payne, 1999; Stuntz, 2004;Yoon & Baker, 2006). The shadow model, in its simplest form, assumes that prosecutors aim at maximizing the sentence, and defense attorneys and defendants attempt to minimize the sentence. Therefore, the plea bargaining process resembles the negotiation over the price of goods in a perfect market.6 The theory predicts that the parties—​the prosecution and the defense—​will eventually reach an equilibrium (i.e., the balanced state between the parties) at the following point, (1)

sP = p * sT , 0 < p < 1, sP ≥ 0, sT ≥ 0

or, (2)

sT − sP = 1 − p sT

In the equations above, sP stands for the sentence at plea, and sT stands for the sentence at trial. The parameter p stands for the probability of conviction at trial, which is typically associated with the strength of evidence (Bushway et al., 2014; Tor, Gazal-​Ayal, & Garcia, 2010). The right-​hand side of Equation 1 is the mathematical expectation at trial—​the gain (for the prosecution, and the loss for the defense) associated with a conviction at trial, multiplied by the probability of conviction. In this model, rational prosecutors would not offer sentences lower than p * sT , and rational defense attorneys would not accept sentences harsher than p * sT . As a result, the plea would occur at p * sT . For example, suppose a defendant faces a prison sentence of 10 years if convicted at trial, and the strength of evidence indicates a 70% (0.7) probability of conviction, then the shadow model predicts a plea sentence at seven years. Compared with the norm-​based theories, it is easier to test the shadow model empirically, as the strength of case is more concrete and measurable than the concept of local norms.7 An advantage of the shadow model is that it gives an unambiguous prediction of the size of plea discounts. For a given case and a given plea offer, Equation 1 can also predict whether the parties are likely to strike a plea deal.8 However, it is important to realize that the shadow model is not perfect. Equation 1 rests on the assumption of a perfect market and rational market participants. While there can be challenges to the rational choice assumptions in social sciences in general (Hastie & Dawes, 2010; Shepsle, 1989; Tversky & Kahneman, 1986), this issue stands out particularly in plea bargaining research (for an 191

Shi Yan

excellent review on the challenges to the shadow model, see Bibas, 2004). Here it assumes that both parties are fully aware of the expected sentence, as well as the probability of conviction at trial. It also assumes that the severity of the sentence is the only variable of interest to the parties. However, studies have suggested that prosecutors may pursue goals other than maximizing the total sentence (Bandyopadhyay & McCannon, 2014, see also Endnote 8; Rasmusen, Raghav, & Ramseyer, 2009). Moreover, the parties may not be able to predict the case outcomes accurately, due to a set of structural and psychological factors (e.g., Covey, 2007; Gross, 2011; Gross & O’Brien, 2008; Heumann, 1978; Hollander-​Blumoff, 2007; Howe, 2005; Redlich et al., 2017; Scott & Stuntz, 1992;Tversky & Kahneman, 1974). All these factors call the validity of the shadow model into question. Despite these theoretical challenges, a number of empirical studies have found support for the shadow model at the aggregate level (Bushway & Redlich, 2012; Bushway, Redlich, & Norris, 2014; Elder, 1989; LaFree, 1985; Smith, 1986). However, there is less support for the model at the individual case level, as analyses of both empirical sentencing data (Bushway & Redlich, 2012) and experimental data (Bushway, Redlich, & Norris, 2014) found considerable variation in the sizes of plea discounts among similar cases. This pattern is consistent with our knowledge of courts and sentencing. On the one hand, studies have found that learning is crucial to all courtroom actors (Alschuler, 1975, 1983; Heumann, 1978; Nardulli , Eisenstein, & Fleming, 1988; Ulmer, 1997). Experienced courtroom actors, familiar with courtroom norms and going rates, can often predict case outcomes reasonably well. On the other hand, idiosyncratic factors, such as the background of judges and the information presented in courtrooms, can also affect the sentence (e.g., Englich, Mussweiler, & Strack, 2005, 2006; Myers, 1988; Rachlinski et al., 2009). Statistical models on sentencing, particularly those estimated from administrative datasets, tend to capture the systematic variables (such as crime severity and criminal records) much better than the idiosyncratic factors (such as remorse and harm to victims). Therefore, it is safer for us to draw conclusions on the overall pattern (e.g., the average size of plea discounts for the sample or a defendant group) than on the individual-​case level specifics (e.g., the exact size of plea discount for a specific defendant). Even though the shadow model was intended at the individual level, given the theoretical and statistical issues presented above, the rest of the chapter only utilizes the framework at the aggregated level.

The Significance of the Size of Plea Discounts A closer look at the two theories suggests that plea discounts that are either too small or too large would result in problems in courtrooms. For prosecutors, consistently recommending small plea discounts results in insufficient rewards to remorse and cooperation (Bibas & Bierschbach, 2004; Etienne & Robbennolt, 2007; Nardulli, Eisenstein, & Fleming, 1988). Moreover, both the norm-​ based theories and the shadow model suggest that in those cases, a much larger number of defendants would opt for trials (Engen & Steen, 2000; Uhlman & Walker, 1980; Ulmer & Kramer, 1996). Even in courtrooms that do not suffer from serious caseload pressure, this can violate the norms of maintaining group cohesion (Eisenstein & Jacob, 1977; Heumann, 1975, 1978). Therefore, despite the attempts at certain times and places to restrict or even prohibit plea offers (e.g., Acevedo, 1995; Berger, 1976; Langbein, 1979a; Rubinstein & White, 1979; Weninger, 1987), plea bargaining remains the dominant mode of case disposition. A more problematic situation arises when plea discounts are too large. At first glance, it may appear that defendants who plead are getting large rewards. However, according to Equation 2 above, a rational defendant would plead guilty as long as the ratio of the plea and trial sentences is lower than the probability of conviction at trial (in other words, the plea discount in percentage form being larger than the probability of acquittal at trial). As long as this condition holds, the defendant should plead guilty regardless how low the probability of conviction is (i.e., how weak the case is). Suppose a defendant faces a plea offer of one year in prison, and the sentence at trial is 10 years, then the defendant’s rational choice is to plead guilty as long as the probability of conviction is 10% or higher. 192

Estimating the Size of Plea Discounts

However, the question is—​should someone plead guilty with such a weak case against him? Or more formally, is this disparity consistent with the safeguards of defendants’ rights listed in the Fifth and Sixth Amendments? It is noteworthy that neither the norm-​based theories nor the shadow model has been able to set boundaries for the “appropriate” range of plea discounts. Although studies on the courtroom workgroup did highlight “doing justice” as a part of the norms (Eisenstein & Jacob, 1977; Nardulli, Eisenstein, & Fleming, 1988; Ulmer, 1997), there is no guarantee that the workgroup will always prioritize justice when there is a tension with other goals (e.g., the caseload pressure, Alschuler, 1979; Heumann, 1978; Johnson, 2005). The shadow model is able to predict the behavior of courtroom actors (i.e., to accept or to reject the plea given the circumstances), yet the model itself does not contain the moral values to differentiate “good” (rewarding) and “bad” (coercive) plea offers (Bibas, 2004). When the disparity between the trial and plea sentences is large, researchers tend to view it as a penalty applied to those who are convicted at trial, rather than a reward for those who plead (e.g., Burke, 2007; Covey, 2007; Kim, 2015; Lynch, 2016; McCoy, 2005; Rakoff, 2014; Scott & Stuntz, 1992; Ulmer, Eisenstein, & Johnson, 2010). Researchers have observed large trial penalties in a number of cases, as the examples below illustrate. Although Alfred did not actively participate in the event [shoving a boy down and taking 40 dollars from his backpack], the prosecutor charged him under the accomplice theory of criminal liability, for which he could face up to sixteen years in prison. Days later … Alfred’s counsel relayed to him the prosecutor’s so-​called one-​time offer: “If you plead to the robbery charge, the street terrorism charge will be dropped, and you will serve two years in state prison” (Caldwell, 2011, pp. 63–64, clarification added). [T]‌his is a serious offense in terms of what it’s done to the community … in terms of the resources that it took to convict you. We had the police and the prosecutor sitting down here for three days to convict you and impanel this jury and try it, and for that you’re going to be punished … And, ladies and gentlemen … had he been willing to enter a plea on this case, he would have been afforded probation. He wouldn’t have gone to the institution (“State v. Scalf,” 1998, emphasis added). If you’d have come in here, as you should have done in the first instance, to save the State the trouble of calling a jury, I would probably have sentenced you, as I indicated to you I would have sentenced you, to one to life in the penitentiary. It will cost you nine years additional, because the sentence now is ten to life in the penitentiary. (“The People v. Moriarty,” 1962, emphasis added) The Moriarty case here suggests that the 90% plea discount mentioned above is not merely imaginary. The sheer contrast here motivated legal researchers’ criticism of plea bargaining as being coercive—​as Langbein (1978, p.  12) has claimed (see also Alschuler, 1968, 2015; Brunk, 1979; Caldwell, 2011; Lieb, 2014; Lynch, 2016; McCoy, 2005; Smith, 1986), In twentieth-​century America we have duplicated the central experience of medieval European criminal procedure … [O]‌ur means are much politer; we use no rack, no thumbscrew, no Spanish boot to mash his legs. But like the Europeans of distant centuries who did employ those machines, we make it terribly costly for an accused to claim his right to the constitutional safeguard of trial. We threaten him with a materially increased sanction if he avails himself of his right and is thereafter convicted. This sentencing differential is what makes plea bargaining coercive. 193

Shi Yan

Scholars have claimed that in the worst case, the optimal choice for factually innocent defendants is to plead guilty, since the expected sentence at trial is often worse than the plea offer (e.g., Bowers, 2008; Dervan & Edkins, 2013; Hessick & Saujan, 2002; Rakoff, 2014; Redlich et al., 2017; Redlich & Shteynberg, 2016; Wright, 2005).9 Some researchers recommended restriction on the maximum size of plea discounts (e.g., Bar-​Gill & Gazal-​Ayal, 2006; Covey, 2008; Gazal-​Ayal, 2006). While this argument can appear as if it prevents excessive rewards to defendants who cooperate, it also prevents excessive “trial penalties” imposed on those who are convicted at trials. From a social scientist’s perspective, there is still one missing piece between the claims and the policy recommendations. It is true that many individual cases support the claim on coercive plea bargaining. However, we know relatively less about the big picture, that is, whether the 90% discount in Moriarty is an outlier or the mode. Much larger amount of data, as well as much more rigorous statistical analysis, is necessary to answer this question (Tata & Gormley, 2016). Fortunately, a group of studies has already examined the average size of plea discounts at the individual county and the state level. The following part summarizes what we have known from these studies.

The Empirical Estimates: What We Have Known So Far In this part, I first summarize the empirical findings on the size of plea discounts. I then move on to a discussion of the estimation strategies. Today, almost every study that utilizes administrative sentencing data controls for the mode of conviction (i.e., at trial or by plea, Ulmer, 2012), and a complete review of all the findings is beyond the capability of the current chapter. This part will instead focus on studies that specifically examined issues related to plea bargaining—​the size of plea discounts, the variation in plea discounts, the decision to plead, and other related topics.

The Variation in Plea Discounts Across Jurisdictions Criminal law and criminal procedural law are the primary determinants of the plea bargaining practice. Some jurisdictions have presumptive and restrictive rules on plea bargaining that limit the discretion of courtroom actors, whereas other jurisdictions leave the practice mostly at the discretion of the courtroom actors. Moreover, the theories on courtroom norms suggest that even within the same jurisdiction (i.e., under the same set of laws), local practices can lead to variation among counties and courthouses (Heumann, 1978; Nardulli, Eisenstein, & Fleming, 1988; Ulmer, 1997). All these factors suggest that we would expect considerable variation in the size of plea discounts at the jurisdiction level. Using data collected from five sites across the United States by Miller, McDonald, and Cramer (1978), both Smith (1986) and Bushway and Redlich (2012) found average plea discounts at around 25%.10 Using more recent data from Pennsylvania, Ulmer and Bradley (2006) found an average plea discount of 36% for defendants charged with serious violent crimes.11 Additionally, researchers found average plea discounts at around 13 percent using data from both the federal jurisdiction (Ulmer, Eisenstein, & Johnson, 2010) and North Carolina (Abrams & Fackler, 2016), and an average plea discount at 50 percent using data from a large county in Florida (Metcalfe & Chiricos, 2018). Studies that examined statewide data also found variation in the size of plea discounts across counties, as well as an association between the plea discounts and county-​level factors (such as caseload, demographic composition, and economic condition, see Abrams & Fackler, 2016; Ulmer & Bradley, 2006; Ulmer, Eisenstein, & Johnson, 2010). One notable exception to the near-​consensus on the presence of plea discounts is a study conducted by Abrams (2011), which found negative average plea discounts (i.e., “plea penalties”) using data from Cook County, Illinois (see also Abrams, 2013). His explanation was that by pleading guilty, defendants were able to avoid the time and monetary costs associated with trials (see also Feeley, 1979; Kohler-​ Hausmann, 2013, and the above part on the shadow model). Therefore, defendants were willing to

194

Estimating the Size of Plea Discounts

plead guilty even if the offered sentences were slightly harsher than the expectation at trials. While this argument has strong face validity, the finding of “plea penalties” has not emerged from other studies or jurisdictions.

Sentencing Guidelines and Plea Bargaining Some jurisdictions in the United States have adopted sentencing guidelines to promote consistency in sentencing outcomes (Breyer, 1988). While the overall effectiveness of sentencing guidelines is beyond the scope of this chapter, research suggests that the structure of guidelines can affect the plea bargaining practice. King et  al. (2005) analyzed data from five states that had guidelines, and found considerable variation in the size of plea discounts among those states. Piehl and Bushway (2007) further found that presumptive and restrictive guidelines reduced the unexplained variation in the sentences (i.e., increased the consistency at the sentencing stage), but meanwhile increased the impact of charge bargaining in shaping the sentence (see also Bjerk, 2005; Rehavi & Starr, 2014). Furthermore, using data from North Carolina, Wright and Engen (2006, 2007) found that charge bargaining was more likely to occur for crime types with “deeper” charging options in the guidelines (i.e., a higher number of possible charges associated with the same crime event, such as the different degrees of assault or robbery).

Plea Discounts and Case-​level Factors Empirical research on sentencing has long focused on how case-​ level legal and extra-​ legal characteristics explain the variation in sentencing outcomes. This tradition remains in studies on plea bargaining. Unlike the near-​consensus on racial disparities in the observed sentences (Mitchell, 2005; Spohn, 2015b), studies have revealed mixed findings. For instance, King et  al. (2005) found substantial variation in the size of plea discounts across crime types. In the majority of jurisdictions they examined, defendants who faced drug crime charges received larger plea discounts than those charged with other crime types. Meanwhile, Ulmer, Eisenstein, & Johnson (2010) found that for cases processed in the federal jurisdiction, the size of plea discounts increased with the severity of charges. The inconsistency in findings is also present among studies on racial disparities in plea discounts. For example, Ulmer and Bradley (2006) did not find statistically significant differences in plea discounts across different racial and ethnic groups. However, Metcalfe and Chiricos (2018) found significant disparities among various race-​gender combinations, with black female defendants receiving the largest average plea discount and black males receiving the smallest.These patterns suggest that there are multiple factors affecting plea bargaining, and additional studies are needed to disentangle the role of these legal and extra-​legal variables. Both the norm-​based theories and the shadow model emphasized the role of evidence in shaping prosecutors’ plea offers, as well as defense’s plea decisions (Bushway, Redlich, & Norris, 2014; Heumann, 1978; Ulmer & Kramer, 1996). Bushway and colleagues (2014) conducted an experiment to test the relationship between the strength of evidence and plea offers (see also Redlich, Bushway, & Norris, 2016). They recruited real-​world prosecutors, defense attorneys, and judges (as opposed to student or layperson subjects) to review a vignette of a robbery case. They randomly assigned cases with different pieces of evidence to the subjects, and asked the subjects to estimate the trial and plea sentences respectively. Bushway et al. found a positive relationship between the strength of evidence and the size of plea discounts. Depending on the evidence presented, the average size of plea discounts ranged between 31% (when confession, eyewitness identification, and DNA match were all present) and 63% (when none of those evidence types was present). To my awareness, no other published study has adopted a similar design, and additional work is necessary to reach a more solid conclusion on the relationship between evidence and plea discounts.

195

Shi Yan

Plea Offers and the Decision to Plead Critics of plea bargaining have claimed that large disparities between plea and trial sentences force defendants to plead guilty (Lynch, 2016; McCoy, 2005). Moreover, psychologists state that on and above the harshness of plea offer, the labeling and framing of the offer also matters (Bibas, 2004; Redlich et al., 2017). A group of experimental studies tested these claims empirically by examining the relationship between the size of plea discounts (in the format of plea offers) and the defense’s plea decision. Using a vignette of academic integrity violations, Dervan and Edkins (2013) found a positive relationship between the size of plea offers and the probability for subjects to plead guilty. This pattern held for both factually innocent and factually guilty subjects, albeit for any given plea offer, factually innocent subjects were consistently less likely to plead than factually guilty ones. Tor and colleagues (2010) found that plea decisions depended not only on the sentence, but also the perceived fairness of the offer. Using multiple vignettes, they found that for a fixed plea sentence, subjects were more likely to plead guilty if they were informed that the offer was more lenient than the average offer in similar cases. To the contrary, Redlich and Shteynberg (2016) found that the label of the plea offer (i.e., calling the same sentence recommendation a “really good offer” versus “a fair but not great offer”) did not have an effect on the decision to plead guilty.

Overall Summary of Findings A few takeaways can be summarized from the findings presented above. First, there is considerable variation in the size of plea discounts across jurisdictions. Among the studies reviewed, the average plea discount ranged from 13% to 50% (see above). This pattern is expected given the various state and federal laws, as well as the variation in local practices. Unfortunately, the decentralized nature of the court system makes it virtually impossible to compare and cross-​validate the size of plea discounts across jurisdictions. Second, estimates of the average plea discount are much smaller than the extreme values found in individual cases (such as the 90% in Moriarty). However, given the difficulty of setting the theoretical range for “appropriate” plea discounts, I refrain from concluding whether or not the amounts presented here imply coercion. Third, there is evidence suggesting that the size of plea discounts varies with legal and extra-​legal characteristics. Nevertheless, researchers are still far from reaching a consensus on the direction and magnitude of the correlates. Fourth, consistent with the theories, empirical studies have found that the magnitude of plea discounts directly affects the defendant’s decision of whether or not to plead guilty. This finding highlights the significance of research and possible policy reforms (or regulations) on the size of plea discounts.

Regression-​based Strategies of Estimating the Plea Discounts Researchers adopted a variety of research designs in the studies reviewed above. When using surveys, interviews, and, experiments, researchers can estimate the size of plea discounts by comparing plea offers and the trial outcomes for the same cases (or vignettes). However, the sample sizes of these studies are likely to be small. Large, administrative datasets are often necessary for studies of the big picture (Connelly et al., 2016; Ulmer, 2012). In larger-​scale studies, most researchers utilize regression techniques to draw conclusions on the size of plea discounts.Typically, two different regression-​based approaches are used in the studies presented above, but the approaches are built on slightly different assumptions, and may lead to different estimates of the plea discounts. I present the issue here as a methodological note.

The Regression Coefficient Approach The first approach measures plea discount as the difference between the average plea and trial sentences, conditional on observed legal and extra-​legal variables (e.g., Brereton & Casper, 1982; N. J. 196

Estimating the Size of Plea Discounts

King et al., 2005; Spohn, 1992; Ulmer & Bradley, 2006; Ulmer, Eisenstein, & Johnson, 2010). These studies typically estimate a regression model on the sentence received by a sample of defendants, and use the variable plea (or trial) as the key regressor (Baumer, 2013). Or in equation form,  (3) Sentence = β0 + β1Plea + β2Other legal factors + β3Extralegal factors + ε The plea discount is the coefficient for the plea variable.

The Counterfactual Approach The second approach is to estimate what would have happened to the plea defendants had they gone to trial—​the sentence that would have been imposed if the defendants were convicted at trial. This approach, introduced by Smith (1986), was recognized as “the best previous empirical work comparing sentences after trial and after plea bargain” (Abrams, 2011, p. 203). Piehl and Bushway (2007, p. 108) summarized the basics of this approach as follows, In [Smith’s] case, he needed to know the sentencing outcome at trial for someone who pled guilty. To create an estimate of this unobservable value, he first estimated regression models for conviction and incarceration for those who went to trial. He then used the coefficient estimates from these models to predict both the probability of conviction at trial and the probability of incarceration for those who plead guilty. Or in equation form, for defendants convicted at trial, (4)

Sentence |trial , conviction = β0 + β1 * current crime + β2 * criminal history + β3 * extralegal factors + w

Then for defendants who pled guilty, (5) Predicted trial entence | plea = β0 + β1 * current crime + β2 * criminal history + β3 * extralegal factors This procedure results in an observed plea sentence and an estimated counterfactual trial sentence for each defendant who pled. Then, the plea discount can be calculated from the two sentences.

Comparison of the Two Approaches The two approaches presented above are largely similar—​both are regression-​based, and are therefore subject to the same general assumptions of the regression model used. That being said, a crucial difference between the approaches is that the regression coefficient approach assumes that one unified sentence generating process exists at trial and at plea, whereas the counterfactual approach assumes that there are two separate sentence generating processes. In other words, the key difference is whether the regression coefficients for all legal and extra-​legal variables (except for the plea/​trial variable) remain the same or vary at trial and at plea. Theoretically, the counterfactual approach is more consistent with the theories and findings of existing studies on plea bargaining. This is because prosecutors and defense attorneys utilize every detail of the case—​from the harm to the victim to the attitude of the defendants—​as leverages in plea bargaining (Heumann, 1978; Nardulli, Eisenstein, & Fleming, 1988). The courtroom actors—​both the prosecution and the defense—​are aware that certain details of the case can often have a considerable impact on the case outcome if presented at trial (i.e., the “bad facts” under the focal concerns perspective, Ulmer & Bradley, 2006). Pleading guilty can prevent these details from being used against the defense. As Ulmer and Kramer (1996, p. 396) have noted, You might learn details during the trial that you would not learn at a guilty plea. Hearing that an eight-​year-​old may be traumatized is not the same as seeing her get up on the 197

Shi Yan

stand and sob and cry and fall apart. I had a case of an elderly couple that was robbed. Her husband got up to testify and his wife just collapsed in the courtroom. We had to call an ambulance. Knowing that this robbery had so traumatized these two had an effect on me. We can reasonably expect that the information presented at trials and at pleas is systematically different. Therefore, it is also reasonable to assume that a given case variable (such as criminal records and charge level) would play different roles in shaping the sentence at pleas versus at trials. In the regression context, it is likely that the coefficients for the variables will be different between trials and pleas. This additional assumption of the single regression approach may not stand in reality (see also Klepper, Nagin, & Tierney, 1983; Oaxaca, 1973). The degree of the problem depends on the exact differences between the coefficients, which further depends on the jurisdiction and courtroom-​level factors. Unfortunately, studies reviewed above used one approach or the other, and there has been no direct comparison of the performances of the two model specification strategies. Yet this is not to claim that the counterfactual approach is free of methodological concerns. A crucial limitation that both approaches have in common is the issue of unobservable heterogeneity (Bibas, 2004). Specifically, on average, defendants who opt for trial may be systematically different from those who plead guilty, on a set of variables that are typically missing from administrative datasets (and sometimes not even measurable). These variables include risk preference (Bar-​Gill & Gazal-​ Ayal, 2006; Tor, Gazal-​Ayal, & Garcia, 2010), legal representation (Chin & Holmes, 2001; Roberts, 2013), and perceptions of the severity of sentences (Crouch, 1993; May & Wood, 2010). If defendants select themselves into either trials or pleas due to one or more of these unobserved factors, then the coefficients from the model explaining the trial sentence may no longer be applicable to defendants who pled (and vice versa). Regardless of the approach, the estimated size of plea discounts in these cases can be biased. Recently, researchers have attempted to address this issue of unobservable variables with more sophisticated modeling techniques. One attempt by Abrams and colleagues is the use of the tenure of judges as an instrumental variable (Abrams, 2011; Abrams & Fackler, 2016). The rationale was that prosecutors and defense attorneys would be more likely to reach plea agreements when the case outcome at trial was certain, which was more likely to happen if the judge had spent a longer time in the office.The claim was that the length of tenure of the judges was related to the decision to plea, but not the sentence.Yet the instrument can be weak (Abrams, 2011, p. 217). Moreover, instrument variables are often not available in large administrative datasets (Baumer, 2013; Johnson, King, & Spohn, 2016; Ulmer, 2012). To date, there has not been an established way to address this issue in the literature, and it is not very likely for the problem to be resolved unless datasets of a different nature become widely accessible. Nevertheless, I suggest readers keeping this issue in mind when going through the empirical findings of current studies, as well as when designing studies in the future.

Moving Forward In his review of sentencing research in the 2000s, Ulmer (2012, p.  25) listed the study of pre-​ conviction decisions as the top priority for researchers to move forward. The largest gap is in understanding prosecutorial discretion, the importance of which sentencing researchers have agreed on and desired to fill since the 1970s. The largest obstacle in the way of such knowledge is the general pre-​conviction data, especially data on initial or indictment/​information charges, conviction/​acquittal outcomes, and characteristics of guilty pleas (e.g. pleas to charge reductions, sentence recommendations, or both). The statement above highlights the significance of plea bargaining and plea discounts to the research of courts and sentencing (see also Baumer, 2013; Johnson, King, & Spohn, 2016). However, as 198

Estimating the Size of Plea Discounts

summarized in the parts above, estimating the size of plea discounts is not merely an intellectual game. The disparities between sentences imposed at trial and those at plea are directly related to the fundamental rights of criminal defendants (Lynch, 2016; Rakoff, 2014).Therefore, inaccurate estimates can result in misidentifying the problems, as well as the viable ways for policy reform. While researchers have made considerable progress in measuring the plea discounts in the past few decades, there are still issues that remain unresolved. The last part will summarize the theoretical, methodological, and data issues, and suggest some possible ways for research to move forward.

Alternative Sanctions in Plea Bargaining The first place for future studies to investigate is the measurement of the sentence, since it can directly affect the size of plea discount estimates. The vast majority of studies on sentencing examine the decision to incarcerate, or the length of incarceration (Baumer, 2013; Ulmer, 2012), and this focus also characterizes research on plea bargaining. However, alternative sanctions can be crucial in plea decisions. The number of individuals under alternative sanctions is much larger than the number of prisoners (Kaeble, 2018), and prior research suggests that prosecutors often recommend alternative sanctions in lieu of incarceration terms to encourage pleas (Heumann, 1978; Klein, 2006; Nardulli, Eisenstein, & Fleming, 1988; Ulmer, 1997). Still, the attention given to these decisions is less than adequate in current research. The conventional approach of examining solely the incarceration sentences treats all alternative sanctions as having a zero in the dependent variable, regardless of it being the in/​out decision or the length of incarceration (Baumer, 2013). The norm-​based theories suggest that prosecutors and judges do not assign alternative sanctions randomly, but assign specific types of sanctions designed to reflect the assessment of blameworthiness and risk of the defendant (Alschuler, 1968; Kahan, 1996; Ulmer, 1997). Treating all alternative sanctions as if they were the same masks the distinction among different assessments of the defendants. For the shadow model, the equilibrium laid out in Equation 1 rests on the consideration of the sentence severity, and not necessarily the sentence length. Prior research suggests that even though most alternative sanction types—​such as probation, fines, and community service—​do not involve incarceration, they can still bear considerable harshness (Crouch, 1993; Kohler-​Hausmann, 2013; Martin et al., 2018; May & Wood, 2010). If researchers treat the severity of alternative sanctions as zero, then any sentence reduction from an incarceration term to an alternative sanction becomes a 100% plea discount. This can lead to severe overestimation of the average plea discount when the use of alternative sanctions is frequent. Within the voluminous body of research on sentencing, few studies have addressed the issue of alternative sanctions. A group of studies examined the qualitative difference among the alternative sanctions type by using categorical dependent variables to code different sanction types (e.g., a dependent variable where prison = 1, jail = 2, probation = 3, etc. See Engen & Steen, 2000; Holleran & Spohn, 2004; Johnson & DiPietro, 2012; Yan, 2017). While this approach preserves the qualitative differences among the sanction types, it does not establish the “exchange rate” between sanction types. Another group of studies attempted to unify the severity measure for different sanction types by creating severity scales (e.g., May & Wood, 2010; Schiff, 1997; Spelman, 1995). Although this approach is compatible to the shadow model (Equation 1), the scales are subject to external validity issues (Ostrom, Ostrom, & Kleiman, 2004). To date, there is no consensus in the literature on the appropriate strategy for modeling alternative sanctions. An avenue for future research might be to explore ways to integrate alternative sanctions into plea bargaining models.

Personal Traits and Other Understudied Characteristics The issue of unobserved and unobservable variables is present in most studies using administrative datasets (Baumer, 2013; Connelly et al., 2016; Ulmer, 2012). Examples of unobserved variables 199

Shi Yan

include the type of legal counsel, the socioeconomic status of the defendant and the victim, and the defendant’s level of remorse. Unobservable variables include risk preference, level of confidence, and factual guilt. All these variables can be reasonably related to the plea outcome (Bibas, 2004; Bjerk, 2007; Ulmer & Bradley, 2006). Therefore, when regression models fail to include these variables, omitted variable bias may arise and distort the plea discount estimates (Johnson, King, & Spohn, 2016). It is often impossible for administrative datasets to capture personal traits, such as risk preference, confidence, and passion, for either the prosecution or the defense. Therefore, alternative research designs, such as experiments, observations, and interviews, would be necessary and more appropriate. A group of studies in economics (McAdams & Ulen, 2009), psychology (Covey, 2007; Redlich et al., 2017), and sociology and political science (Nardulli, Eisenstein, & Fleming, 1988; Ulmer, 1997) have each contributed to the exploration of those variables, and future studies need to continue this multi-​ disciplinary approach. This, of course, is not to argue that administrative datasets are no longer necessary.To better model the variables that are visible to the parties in plea bargaining, one of the possibilities to move forward is to discover administrative datasets that contain richer information. Some successful examples of this type include recent studies that investigated the role of the facial appearance of defendants during sentencing (Johnson & King, 2017; R. D. King & Johnson, 2016). Another possibility is to search for larger administrative datasets that have unique features, such as judge identifiers (Abrams, 2011) or more comprehensive information on defendants’ criminal careers (Yan, 2018). These variables can potentially increase the fit of models explaining the plea outcomes.

The “Bargaining” Component in Plea Bargaining The term “plea bargaining” suggests that it can often be an iterative process—​the prosecutor makes the initial offer, the defense provides a counter, etc. However, most studies have only modeled the outcome at the end (i.e., the conviction charges).While studies have begun to explore the role of the length of time spent on plea bargaining, many variables that are related to time remain understudied. Out of these, the most important ones are the time-​variant factors (Johnson, King, & Spohn, 2016), such as the “bargaining” process in individual cases, the learning process throughout the career of prosecutors and defense attorneys (Heumann, 1978), or the change in defendants’ perceptions of the severity (Spelman, 1995). Addressing these issues would also promote our understanding of the variation in the size of plea discounts.

Conclusion This chapter has reviewed the literature on recent efforts to estimate the size of plea discounts. Overall, the quality of research on sentencing has improved considerably over the past decades (Baumer, 2013; Blumstein et al., 1983; Ulmer, 2012), and this is true for the study on plea bargaining as well (Johnson, King, & Spohn, 2016). Nevertheless, as summarized in this part, there are still methodological and theoretical issues that limit our further understanding of the plea bargaining process. The mission of future research includes the discovery of new data (both primary collection and second-​hand data), the adoption of innovative analytic techniques (both quantitative and qualitative), and many more. In this process, it is crucial to maintain openness to this multi-​disciplinary approach (Newman, 1993). What I have listed in this chapter is, of course, only a subset of possibilities for future studies, and I hope this chapter will serve as a starting point for a fruitful academic dialogue.

Notes 1 I thank Cassia Spohn for the helpful guidance, and Jason Walker for his help in copyediting.

200

Estimating the Size of Plea Discounts 2 Some researchers use both terms interchangeably in their work (e.g., Kim, 2015; N. J. King, Soule, Steen, & Weidner, 2005). For consistency and simplicity, I generally use the term “plea discounts” in this chapter. 3 But see Abrams (2011) and Abrams (2013) for notable exceptions, and see below in this chapter for a detailed discussion of these studies. 4 While some researchers further differentiated the courtroom workgroup and focal concerns perspectives (e.g., Johnson, King, & Spohn, 2016), I combine them into one broad category in this chapter due to the similarity in their core arguments and hypotheses. 5 Despite the claims that caseload pressure was not the main driving force for plea bargaining to gain the dominance (Heumann, 1975; Nardulli, 1979), there was little challenge to the argument that plea bargaining expedited the movement of the docket. 6 In economics, a market is a “perfect market” when it meets a number of conditions. Some of the conditions include the presence of a larger number of buyers and sellers, homogenous products, and the absence of transaction costs (i.e., neither party pays for anything other than the price of the goods). The equations hold only if these conditions are met. However, as will be presented later in the chapter, there are disagreements on whether these conditions stand in the context of plea bargaining. 7 This, however, does not necessarily mean that the information on evidence is easy to access. 8 Moreover, the shadow model also has some inherent flexibility to allow for the inclusion of other factors, such as the time costs associated with trials (Abrams, 2011, 2013; Feeley, 1979; Kohler-​Hausmann, 2013), the factual guilt/​innocence of defendants (Bjerk, 2007), and the goals of prosecutors (such as the pressure of re-​ election, see Bandyopadhyay & McCannon, 2014; McCannon, 2013). The rest of the chapter will still build on the classic shadow model (i.e., Equation 1), since there is not a consensus on a modified version of the shadow yet. However, it is important to point out that the shadow model does have the potential to formally include other useful variables in plea bargaining. 9 “Expected sentence” here means the sentence if the defendant gets convicted at trial, multiplied by the probability of conviction (Kim, 2015). When the sentence at trial is harsh, even after the discount, this mathematical expectation can still bear considerable harshness. 10 The original dataset contained cases collected from New Orleans, Norfolk, Seattle, El Paso, Tucson, and Delaware County, Pennsylvania. However, both Smith (1986) and Bushway and Redlich (2012) excluded cases from El Paso from their analyses due to missing data. 11 Two further notes. First, some studies cited here further differentiated bench trials and jury trials (e.g., N. J. King et al., 2005; Ulmer & Bradley, 2006). Although the differentiation is important, I only report the findings on jury trials here for simplicity. Second, some studies cited here reported the disparities in sentences in the form of “trial penalties” (e.g., Ulmer & Bradley, 2006). For consistency, I convert all findings to “plea discounts” here (i.e., using trial sentences in lieu of plea sentences as the comparison base). For a trial penalty of x . the corresponding plea discount is 1 −

1 x or  . 1+ x 1+ x

References Abrams, D.S. (2011). Is pleading really a bargain? Journal of Empirical Legal Studies, 8(S1), 200–​221. https://​doi. org/​10.1111/​j.1740-​1461.2011.01234.x —​—​. (2013). Putting the trial penalty on trial. Duquesne Law Review, 51(3), 777–​786. Abrams, D.S., & Fackler, R. (2016). To Plea or Not to Plea: Evidence from North Carolina. Working  paper. Acevedo, R. (1995). Is a ban on plea bargaining an ethical abuse of discretion—​A Bronx County, New York case study. Fordham Law Review, 64(3), 987–​1013. Alschuler, A.W. (1968). The prosecutor’s role in plea bargaining The University of Chicago Law Review, 36(1), 50–​112. —​—​. (1975). The defense attorney’s role in plea bargaining. The Yale Law Journal, 84(6), 1179–​1315. —​—​. (1979). Plea bargaining and its history. Columbia Law Review, 79(1),  1–​43. —​—​. (1981). The changing plea bargaining debate. California Law Review, 69(3), 652–​730. —​—​. (1983). Implementing the criminal defendant’s right to trial: Alternatives to the plea bargaining system. The University of Chicago Law Review, 50(3), 931–​1050. —​—​. (2015). A nearly perfect system for convicting the innocent. Albany Law Review, 79(3), 919–​940. Bandyopadhyay, S., & McCannon, B.C. (2014). The effect of the election of prosecutors on criminal trials. Public Choice, 161(1–​2), 141–​156. https://​doi.org/​10.1007/​s11127-​013-​0144-​0 Bar-​Gill, O., & Gazal-​Ayal, O. (2006). Plea bargains only for the guilty. The Journal of Law & Economics, 49(1), 353–​364. https://​doi.org/​10.1086/​501084 Baumer, E.P. (2013). Reassessing and redirecting research on race and sentencing. Justice Quarterly, 30(2), 231–​ 261. https://​doi.org/​10.1080/​07418825.2012.682602

201

Shi Yan Berger, M. (1976). The case against plea bargaining. American Bar Association Journal, 62(5), 621–​624. Bibas, S. (2004). Plea bargaining outside the shadow of trial. Harvard Law Review, 117(8), 2463–​2547. —​—​. (2011). Regulating the plea-​bargaining market: From caveat emptor to consumer protection. California Law Review, 99(4), 1117–​1162. —​—​. (2012). Incompetent plea bargaining and extrajudicial reforms. Harvard Law Review, 126(1), 150–​174. Bibas, S., & Bierschbach, R.A. (2004). Integrating remorse and apology into criminal procedure. The Yale Law Journal, 114(1), 85–​148. Bjerk, D. (2005). Making the crime fit the penalty: The role of prosecutorial discretion under mandatory minimum sentencing. The Journal of Law & Economics, 48(2), 591–​625. https://​doi.org/​10.1086/​429980 —​—​. (2007). Guilt shall not escape or innocence suffer? The limits of plea bargaining when defendant guilt is uncertain. American Law and Economics Review, 9(2), 305–​329. https://​doi.org/​10.1093/​aler/​ahm010 Blumstein, A., Cohen, J., Martin, S.E., & Tonry, M.H. (eds.). (1983). Research on Sentencing: The Search for Reform. Washington, DC: National Academy Press. Bogira, S. (2005). Courtroom 302: A Year Behind the Scenes in an American Criminal Courthouse. New  York: Vintage Books. Bowers, J. (2008). Punishing the innocent. University of Pennsylvania Law Review, 156(5), 1117–​1180. Brereton, D., & Casper, J.D. (1982). Does it pay to plead guilty? Differential sentencing and the functioning of criminal courts. Law & Society Review, 16(1),  45–​70. Breyer, S. (1988). The federal sentencing guidelines and the key compromises upon which they rest. Hofstra Law Review, 17(1),  1–​50. Brunk, C.G. (1979). The problem of voluntariness and coercion in the negotiated plea. Law & Society Review, 13(2), 527–​553. Burke, A.S. (2007). Prosecutorial passion, cognitive bias, and plea bargaining. Marquette Law Review, 91(1), 183–​212. Bushway, S.D., & Redlich, A.D. (2012). Is plea bargaining in the “shadow of the trial” a mirage? Journal of Quantitative Criminology, 28(3), 437–​454. https://​doi.org/​10.1007/​s10940-​011-​9147-​5 Bushway, S.D., Redlich, A.D., & Norris, R.J. (2014). An explicit test of plea bargaining in the “shadow of the trial”. Criminology, 52(4), 723–​754. https://​doi.org/​10.1111/​1745–​9125.12054 Bushway, S.D., & Reuter, P. (2008). Economists’ contribution to the study of crime and the criminal justice system. Crime and Justice, 37(1), 389–​451. https://​doi.org/​10.1086/​524283 Caldwell, H.M. (2011). Coercive plea bargaining: The unrecognized scourge of the justice system. Catholic University Law Review, 61(1),  63–​96. Casper, J.D., & Brereton, D. (1984). Evaluating criminal justice reforms. Law & Society Review, 18(1), 121–​144. Chin, G.J., & Holmes, R.W., Jr. (2001). Effective assistance of counsel and the consequences of guilty pleas. Cornell Law Review, 87(3), 697–​742. Church, T.W., Jr. (1985). Examining local legal culture. American Bar Foundation Research Journal, 10(3), 449–​518. https://​doi.org/​10.1111/​j.1747–​4469.1985.tb00508.x Clair, M., & Winter, A.S. (2016). How judges think about racial disparities: Situational decision-​making in the criminal justice system. Criminology, 54(2), 332–​359. https://​doi.org/​10.1111/​1745–​9125.12106 Cohen, T.H., & Kyckelhahn, T. (2010). Felony Defendants in Large Urban Counties, 2006. Washington, DC: US Department of Justice. Connelly, R., Playford, C.J., Gayle,V., & Dibben, C. (2016).The role of administrative data in the big data revolution in social science research. Social Science Research, 59, 1–​12. https://​doi.org/​10.1016/​j.ssresearch.2016.04.015 Cooter, R., Marks, S., & Mnookin, R.H. (1982). Bargaining in the shadow of the law: A testable model of strategic behavior. The Journal of Legal Studies, 11(2), 225–​251. Covey, R.D. (2007). Reconsidering the relationship between cognitive psychology and plea bargaining. Marquette Law Review, 91(1), 213–​247. —​—​. (2008). Fixed justice: Reforming plea bargaining with plea-​based ceilings. Tulane Law Review, 82(4), 1237–​1290. Crouch, B.M. (1993). Is incarceration really worse? Analysis of offenders’ preferences for prison over probation. Justice Quarterly, 10(1), 67–​88. https://​doi.org/​10.1080/​07418829300091711 Demuth, S., & Steffensmeier, D. (2004). Ethnicity effects on sentence outcomes in large urban courts: Comparisons among White, Black, and Hispanic defendants. Social Science Quarterly, 85(4), 994–​1011. https://​doi.org/​ 10.1111/​j.0038-​4941.2004.00255.x Dervan, L.E., & Edkins, V.A. (2013). The innocent defendant’s dilemma: An innovative empirical study of plea bargaining’s innocence problem. The Journal of Criminal Law and Criminology, 103(1),  1–​48. Di Luca, J. (2005). Expedient McJustice or principled alternative dispute resolution? A review of plea bargaining in Canada. Criminal Law Quarterly, 50(2),  14–​66. Dixon, J. (1995). The organizational context of criminal sentencing. American Journal of Sociology, 100(5), 1157–​ 1198. https://​doi.org/​10.1086/​230635

202

Estimating the Size of Plea Discounts Easterbrook, F.H. (1992). Plea bargaining as compromise. The Yale Law Journal, 101(8), 1969–​1978. Eisenstein, J., Flemming, R.B., & Nardulli, P.F. (1988). The Contours of Justice: Communities and Their Courts. Boston, MA: Little, Brown. Eisenstein, J., & Jacob, H. (1977). Felony Justice: An Organizational Analysis of Criminal Courts. Boston, MA: Little, Brown. Elder, H.W. (1989).Trials and settlements in the criminal courts: An empirical analysis of dispositions and sentencing. The Journal of Legal Studies, 18(1), 191–​208. Engen, R.L., & Gainey, R.R. (2000). Modeling the effects of legally relevant and extralegal factors under sentencing guidelines: The rules have changed. Criminology, 38(4), 1207–​1230. https://​doi.org/​10.1111/​j.1745–​ 9125.2000.tb01419.x Engen, R.L., & Steen, S. (2000). The power to punish: Discretion and sentencing reform in the war on drugs. American Journal of Sociology, 105(5), 1357–​1395. https://​doi.org/​10.1086/​210433 Englich, B., Mussweiler, T., & Strack, F. (2005). The last word in court—​A hidden disadvantage for the defense. Law and Human Behavior, 29(6), 705–​722. https://​doi.org/​10.1007/​s10979-​005-​8380-​7 Englich, B., Mussweiler, T., & Strack, F. (2006). Playing dice with criminal sentences: The influence of irrelevant anchors on experts’ judicial decision making. Personality and Social Psychology Bulletin, 32(2), 188–​200. https://​ doi.org/​10.1177/​0146167205282152 Etienne, M., & Robbennolt, J.K. (2007). Apologies and plea bargaining. Marquette Law Review, 91(1), 295–​322. Feeley, M.M. (1979). The Process is the Punishment: Handling Cases in a Lower Criminal Court. New York: Russell Sage Foundation. Fisher, G. (2000). Plea bargaining’s triumph. The Yale Law Journal, 109(5), 857–​1086. Flango, V.E., Roper, R.T., & Elsner, M.E. (1983). The Business of State Trial Courts. Williamsburg, VA: National Center for State Courts. Flemming, R.B., Nardulli, P.F., & Eisenstein, J. (1992). The Craft of Justice: Politics and Work in Criminal Court Communities. Philadelphia, PA: University of Pennsylvania Press. Galanter, M. (1990). Civil jury as a regulator of the litigation process. University of Chicago Legal Forum, 1990(1), 201–​271. Gazal-​Ayal, O. (2006). Partial ban on plea bargains. Cardozo Law Review, 27(5), 2295–​2349. Gross, S.R. (2011). Pretrial incentives, post-​conviction review, and sorting criminal prosecutions by guilt or innocence. New York Law School Law Review, 56(3), 1009–​1032. Gross, S.R., & O’Brien, B. (2008). Frequency and predictors of false conviction: Why we know so little, and new data on capital cases. Journal of Empirical Legal Studies, 5(4), 927–​962. https://​doi.org/​10.1111/​ j.1740-​1461.2008.00146.x Guidorizzi, D.D. (1998). Should we really ban plea bargaining: The core concerns of plea bargaining critics. Emory Law Journal, 47(2), 753–​784. Hartley, R.D., Maddan, S., & Spohn, C.C. (2007). Concerning conceptualization and operationalization: Sentencing data and the focal concerns perspective—​A research note. Southwest Journal of Criminal Justice, 4(1),  58–​78. Hastie, R. & Dawes, R.M. (2010). Rational Choice in an Uncertain World: The Psychology of Judgment and Decision Making (2nd ed.). Thousand Oaks, CA: SAGE. Heaton, P., Mayson, S., & Stevenson, M. (2017). The downstream consequences of misdemeanor pretrial detention. Stanford Law Review, 69(3), 711–​794. Hessick, F.A., III, & Saujan, R.M. (2002). Plea bargaining and convicting the innocent:The role of the prosecutor, the defense counsel, and the judge. Brigham Young University Journal of Public Law, 16(2), 189–​242. Hester, R. (2017). Judicial rotation as centripetal force: Sentencing in the court communities of South Carolina. Criminology, 55(1), 205–​235. https://​doi.org/​10.1111/​1745–​9125.12132 Heumann, M. (1975). A note on plea bargaining and case pressure. Law & Society Review, 9(3), 515–​528. —​—​ . (1978). Plea Bargaining: The Experiences of Prosecutors, Judges, and Defense Attorneys. Chicago, IL: The University of Chicago Press. Hogarth, J. (1971). Sentencing as a Human Process. Toronto, ON: University of Toronto Press. Hollander-​Blumoff, R. (2007). Social psychology, information processing, and plea bargaining. Marquette Law Review, 91(1), 163–​182. Holleran, D., & Spohn, C.C. (2004). On the use of the total incarceration variable in sentencing research. Criminology, 42(1), 211–​240. https://​doi.org/​10.1111/​j.1745–​9125.2004.tb00518.x Howe, S.W. (2005). The value of plea bargaining. Oklahoma Law Review, 58(4), 599–​636. Huebner, B.M., & Bynum, T.S. (2006). An analysis of parole decision making using a sample of sex offenders: A focal concerns perspective. Criminology, 44(4), 961–​991. https://​doi.org/​10.1111/​j.1745-​9125.2006.00069.x Johnson, B.D. (2003). Racial and ethnic disparities in sentencing departures across modes of conviction. Criminology, 41(2), 449–​490. https://​doi.org/​10.1111/​j.1745–​9125.2003.tb00994.x

203

Shi Yan —​—​. (2005). Contextual disparities in guidelines departures: Courtroom social contexts, guidelines compliance, and extralegal disparities in criminal sentencing. Criminology, 43(3), 761–​796. https://​doi.org/​10.1111/​ j.0011-​1348.2005.00023.x Johnson, B.D., & DiPietro, S.M. (2012). The power of diversion: Intermediate sanctions and sentencing disparity under presumptive guidelines. Criminology, 50(3), 811–​850. https://​doi.org/​10.1111/​ j.1745-​9125.2012.00279.x Johnson, B.D., & King, R.D. (2017). Facial profiling: Race, physical appearance, and punishment. Criminology, 55(3), 520–​547. https://​doi.org/​10.1111/​1745–​9125.12143 Johnson, B.D., King, R.D., & Spohn, C.C. (2016). Sociolegal approaches to the study of guilty pleas and prosecution. Annual Review of Law and Social Science, 12(1), 479–​495. Kaeble, D. (2018). Probation and Parole in the United States, 2016. Washington, DC: Bureau of Justice Statistics. Kagan, R.A. (2003). Adversarial Legalism:The American Way of Law. Cambridge, MA: Harvard University Press. Kahan, D.M. (1996). What do alternative sanctions mean? The University of Chicago Law Review, 63(2), 591–​653. Kaiser, K., & Spohn, C. (2018). Why do judges depart? A review of reasons for judicial departures in federal sentencing. Criminology, Criminal Justice, Law & Society, 19(2),  44–​62. Kalven, H., & Zeisel, H. (1966). The American Jury. Boston, MA: Little, Brown. Kim, A.C. (2015). Underestimating the trial penalty: An empirical analysis of the federal trial penalty and critique of the Abrams study. Mississippi Law Journal, 84(5), 1195–​1256. King, N.J., Soulé, D.A., Steen, S., & Weidner, R.R. (2005). When process affects punishment: Differences in sentences after guilty plea, bench trial, and jury trial in five guidelines states. Columbia Law Review, 105(4), 959–​1009. King, R.D., & Johnson, B.D. (2016). A punishing look: Skin tone and Afrocentric features in the halls of justice. American Journal of Sociology, 122(1), 90–​124. https://​doi.org/​10.1086/​686941 Klein, R. (2006). Judicial misconduct in criminal cases: It’s not just the counsel who may be ineffective and unprofessional. Ohio State Journal of Criminal Law, 4(1), 195–​222. Klepper, S., Nagin, D.S., & Tierney, L.-​J. (1983). Discrimination in the criminal justice system: A critical appraisal of the literature. In A. Blumstein, J. Cohen, S.E. Martin & M.H.Tonry (eds.), Research on Sentencing:The Search for Reform (Vol. 2, pp. 55–​128). Washington, DC: National Academy Press. Kohler-​Hausmann, I. (2013). Misdemeanor justice: Control without conviction. American Journal of Sociology, 119(2), 351–​393. https://​doi.org/​10.1086/​674743 Kurlychek, M.C., & Johnson, B.D. (2004). The juvenile penalty: A comparison of juvenile and young adult sentencing outcomes in criminal court. Criminology, 42(2), 485–​515. https://​doi.org/​10.1111/​j.1745–​ 9125.2004.tb00527.x Kutateladze, B.L. (2018). Tracing charge trajectories: A study of the influence of race in charge changes at case screening, arraignment, and disposition. Criminology, 56(1), 123–​153. https://​doi.org/​10.1111/​ 1745–​9125.12166 Kutateladze, B.L., Andiloro, N.R., Johnson, B.D., & Spohn, C.C. (2014). Cumulative disadvantage: Examining racial and ethnic disparity in prosecution and sentencing. Criminology, 52(3), 514–​551. https://​doi.org/​ 10.1111/​1745–​9125.12047 Lacasse, C., & Payne, A.A. (1999). Federal sentencing guidelines and mandatory minimum sentences: Do defendants bargain in the shadow of the judge? The Journal of Law and Economics, 42(S1), 245–​270. LaFree, G.D. (1985). Adversarial and nonadversarial justice: A comparison of guilty pleas and trials. Criminology, 23(2), 289–​312. Landes, W.M. (1971). An economic analysis of the courts. The Journal of Law & Economics, 14(1), 61–​107. Langbein, J.H. (1978). Torture and plea bargaining. The University of Chicago Law Review, 46(1),  3–​22. —​—​. (1979a). Land without plea bargaining: How the Germans do it. Michigan Law Review, 78(2), 204–​225. —​—​. (1979b). Understanding the short history of plea bargaining. Law & Society Review, 13(2), 261–​272. Langbein, J.H., Lerner, R.L., & Smith, B.P. (2009). History of the Common Law:The Development of Anglo-​American Legal Institutions. New York: Wolters Kluwer Law and Business. Langer, M. (2006). Rethinking plea bargaining: The practice and reform of prosecutorial adjudication in American criminal procedure. American Journal of Criminal Law, 33(3), 223–​300. Lieb, D. (2014).Vindicating vindictiveness: Prosecutorial discretion and plea bargaining, past and future. The Yale Law Journal, 123(4), 1014–​1069. Lynch, M. (2016). Hard Bargains: The Coercive Power of Drug Laws in Federal Court. New  York: Russell Sage Foundation. Lynch, M., & Omori, M. (2014). Legal change and sentencing norms in the wake of Booker:The impact of time and place on drug trafficking cases in federal court. Law & Society Review, 48(2), 411–​445. https://​doi.org/​ 10.1111/​lasr.12070

204

Estimating the Size of Plea Discounts Martin, K.D., Sykes, B.L., Shannon, S., Edwards, F., & Harris, A. (2018). Monetary sanctions: Legal financial obligations in US systems of justice. Annual Review of Criminology, 1, 471–​495. https://​doi.org/​10.1146/​ annurev-​criminol-​032317-​091915 May, D.C., & Wood, P.B. (2010). Ranking Correctional Punishments: Views from Offenders, Practitioners, and the Public. Durham, NC: Carolina Academic Press. McAdams, R.H., & Ulen,T.S. (2009). Behavioral criminal law and economics. In N. Garoupa (ed.), Criminal Law and Economics (2nd ed.,Vol. 3). Cheltenham: Edward Elgar Publishing. McCannon, B.C. (2013). Prosecutor elections, mistakes, and appeals. Journal of Empirical Legal Studies, 10(4), 696–​714. McCoy, C. (2005). Plea bargaining as coercion: The trial penalty and plea bargaining reform. Criminal Law Quarterly, 50(2), 67–​107. Metcalfe, C., & Chiricos,T. (2018). Race, plea, and charge reduction: An assessment of racial disparities in the plea process. Justice Quarterly, 35(2), 223–​253. https://​doi.org/​10.1080/​07418825.2017.1304564 Miller, H.S., McDonald, W., & Cramer, J.A. (1978). Plea Bargaining in the United States, 1978. Ann Arbor, MI: Inter-​university Consortium for Political and Social Research. https://​doi.org/​10.3886/​ICPSR07775.v1 Mitchell, O. (2005). A meta-​analysis of race and sentencing research: Explaining the inconsistencies. Journal of Quantitative Criminology, 21(4), 439–​466. https://​doi.org/​10.1007/​s10940-​005-​7362-​7 Mnookin, R.H., & Kornhauser, L. (1979). Bargaining in the shadow of the law:The case of divorce. The Yale Law Journal, 88(5), 950–​997. Myers, M.A. (1988). Social background and the sentencing behavior of judges. Criminology, 26(4), 649–​675. Nagel, S.S., & Neef, M. (1979). Decision Theory and the Legal Process. Lexington, MA: Lexington Books. Nardulli, P.F. (1978). The Courtroom Elite: Organizational Perspective on Criminal Justice. Cambridge, MA: Ballinger Publishing Co. —​—​. (1979). The caseload controversy and the study of criminal courts. The Journal of Criminal Law and Criminology, 70(1), 89–​101. Nardulli, P.F., Eisenstein, J., & Flemming, R.B. (1988). The Tenor of Justice: Criminal Courts and the Guilty Plea Process. Urbana, IL: University of Illinois Press. Nardulli, P.F., Flemming, R.B., & Eisenstein, J. (1985). Criminal courts and bureaucratic justice: Concessions and consensus in the guilty plea process. The Journal of Criminal Law and Criminology, 76(4), 1103–​1131. Newman, D.J. (1993).The American Bar Foundation Survey and the development of criminal justice higher education. In L.E. Ohlin & F.J. Remington (eds.), Discretion in Criminal Justice:The Tension between Individualization and Uniformity. Albany, NY: SUNY Press. Oaxaca, R. (1973). Male-​female wage differentials in urban labor markets. International Economic Review, 14(3), 693–​709. Ostrom, C.W., Ostrom, B.J., & Kleiman, M. (2004). Judges and Discrimination: Assessing the Theory and Practice of Criminal Sentencing. Lansing, MI.: Michigan Department of Corrections. Palmer, J. (1999). Abolishing plea bargaining: An end to the same old song and dance. American Journal of Criminal Law, 26(3), 505–​536. The People v. Moriarty, 185 N.E.2d 688 (Illinois Supreme Court 1962). Piehl, A.M., & Bushway, S.D. (2007). Measuring and explaining charge bargaining. Journal of Quantitative Criminology, 23(2), 105–​125. Rachlinski, J.J., Johnson, S.L.,Wistrich, A.J., & Guthrie, C. (2009). Does unconscious racial bias affect trial judges? Notre Dame Law Review, 84(2), 1195–​1246. Rakoff, J.S. (2014, November 20). Why innocent people plead guilty. The New York Review of Books. Retrieved from: www.nybooks.com/​articles/​2014/​11/​20/​why-​innocent-​people-​plead-​guilty/​ Rasmusen, E., Raghav, M., & Ramseyer, M. (2009). Convictions versus conviction rates:The prosecutor’s choice. American Law and Economics Review, 11(1),  47–​78. Reaves, B.A. (2013). Felony Defendants in Large Urban Counties, 2009—​Statistical Tables. Washington, DC: Bureau of Justice Statistics. Redlich, A.D., Bibas, S., Edkins, V.A., & Madon, S. (2017). The psychology of defendant plea decision making. American Psychologist, 72(4), 339–​352. Redlich, A.D., Bushway, S.D., & Norris, R.J. (2016). Plea decision-​making by attorneys and judges. Journal of Experimental Criminology, 12(4), 537–​561. Redlich, A.D., & Shteynberg, R.V. (2016). To plead or not to plead: A comparison of juvenile and adult true and false plea decisions. Law and Human Behavior, 40(6), 611–​625. Redlich, A.D.,Yan, S., Norris, R.J., & Bushway, S.D. (2018). The influence of confessions on guilty pleas and plea discounts. Psychology, Public Policy, and Law, 24(2), 147–​157. Rehavi, M.M., & Starr, S.B. (2014). Racial disparity in federal criminal sentences. Journal of Political Economy, 122(6), 1320–​1354.

205

Shi Yan Roberts, J. (2013). Effective plea bargaining counsel. The Yale Law Journal, 122(8), 2650–​2674. Rubinstein, M.L., & White, T.J. (1979). Alaska’s ban on plea bargaining. Law & Society Review, 13(2), 367–​383. Schiff, M.F. (1997). Gauging the intensity of criminal sanctions: Developing the Criminal Punishment Severity Scale (CPSS). Criminal Justice Review, 22(2), 175–​206. Schulhofer, S.J. (1984). Is plea bargaining inevitable? Harvard Law Review, 97(5), 1037–​1107. —​—​. (1992). Plea bargaining as disaster. The Yale Law Journal, 101(8), 1979–​2009. Scott, R.E., & Stuntz, W.J. (1992). Plea bargaining as contract. The Yale Law Journal, 101(8), 1909–​1968. Shepsle, K.A. (1989). Studying institutions: Some lessons from the rational choice approach. Journal of Theoretical Politics, 1(2), 131–​147. Shermer, L.O.N., & Johnson, B.D. (2010). Criminal prosecutions: Examining prosecutorial discretion and charge reductions in U.S. federal district courts. Justice Quarterly, 27(3), 394–​430. https://​doi.org/​10.1080/​ 07418820902856972 Smith, D.A. (1986).The plea bargaining controversy. The Journal of Criminal Law and Criminology, 77(3), 949–​968. Spelman, W. (1995). The severity of intermediate sanctions. Journal of Research in Crime and Delinquency, 32(2), 107–​135. Spohn, C. (1992). An analysis of the “jury trial penalty” and its effect on Black and White offenders. The Justice Professional, 7(1), 93–​112. —​—​. (2015a). Evolution of sentencing research. Criminology & Public Policy, 14(2), 225–​232. —​—​. (2015b). Race, crime, and punishment in the twentieth and twenty-​first centuries. Crime and Justice, 44,  49–​97. State v. Scalf, 126 Ohio App. 614 (Court of Appeals of Ohio, Eighth District, 1998). Steffensmeier, D., Ulmer, J.T., & Kramer, J.H. (1998). The interaction of race, gender, and age in criminal sentencing: The punishment cost of being young, black, and male. Criminology, 36(4), 763–​798. Stuntz, W.J. (2004). Plea bargaining and criminal law’s disappearing shadow. Harvard Law Review, 117(8), 2548–​2569. Sutton, J.R. (2013). Structural bias in the sentencing of felony defendants. Social Science Research, 42(5), 1207–​1221. Tata, C., & Gormley, J.M. (2016). Sentencing and plea bargaining: Guilty pleas versus trial verdicts. In Oxford Handbooks Online. Oxford: Oxford University Press. Thayer, J.B. (1890). “Law and fact” in jury trials. Harvard Law Review, 4(4), 147–​175. Tor, A., Gazal-​Ayal, O., & Garcia, S.M. (2010). Fairness and the willingness to accept plea bargain offers. Journal of Empirical Legal Studies, 7(1), 97–​116. Turner, J.I. (2017). Plea bargaining. In E. Luna (ed.), Reforming Criminal Justice (Vol. 3, pp. 73–​99). Phoenix, AZ: Arizona State University. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–​1131. —​—​. (1986). Rational choice and the framing of decisions. The Journal of Business, 59(4), S251–​S278. Uhlman, T.M., & Walker, N.D. (1980). “He takes some of my time; I take some of his”: An analysis of judicial sentencing patterns in jury cases. Law & Society Review, 14(2), 323–​341. Ulmer, J.T. (1997). Social Worlds of Sentencing: Court Communities under Sentencing Guidelines. Albany, NY: SUNY Press. —​—​. (2012). Recent developments and new directions in sentencing research. Justice Quarterly, 29(1),  1–​40. Ulmer, J.T., & Bradley, M.S. (2006).Variation in trial penalties among serious violent offenses. Criminology, 44(3), 631–​670. Ulmer, J.T., Eisenstein, J., & Johnson, B.D. (2010). Trial penalties in federal sentencing: Extra-​guidelines factors and district variation. Justice Quarterly, 27(4), 560–​592. Ulmer, J.T., & Kramer, J.H. (1996). Court communities under sentencing guidelines: Dilemmas of formal rationality and sentencing disparity Criminology, 34(3), 383–​408. Vidmar, N., & Hans,V.P. (2007). American Juries:The Verdict. Amherst, NY: Prometheus Books. Vogel, M.E. (2007). Coercion to Compromise: Plea Bargaining, the Courts, and the Making of Political Authority. New York: Oxford University Press. Weninger, R.A. (1987). The abolition of plea bargaining: A case study of El Paso County, Texas. UCLA Law Review, 35(2), 265–​314. Worden, A.P., McLean, S.J., & Kennedy, M. (2013). Sidestepping justice: adjournments in contemplation of dismissal in misdemeanor court. Albany Law Review, 76(3), 1713–​1750. Wright, R.F. (2005).Trial distortion and the end of innocence in federal criminal justice University of Pennsylvania Law Review, 154(1), 79–​156. Wright, R.F., & Engen, R.L. (2006).The effects of depth and distance in a criminal code on charging, sentencing, and prosecutor power. North Carolina Law Review, 84(6), 1935–​1982. —​—​. (2007). Charge movement and theories of prosecutors. Marquette Law Review, 91(1),  9–​38.

206

Estimating the Size of Plea Discounts Yan, S. (2017). Search for the hidden punishments: An alternative approach to studying alternative sanctions. Journal of Quantitative Criminology, 33(1), 21–​44. https://​doi.org/​10.1007/​s10940-​015-​9275-​4 —​—​. (2018). Does criminal specialization predict case processing? Crime & Delinquency. https://​doi.org/​ 10.1177/​0011128718765123 Yoon, A., & Baker, T. (2006). Offer-​of-​judgment rules and civil litigation: An empirical study of automobile insurance litigation in the east. Vanderbilt Law Review, 59(1), 153–​196. Young, R. (2013). Exploring the boundaries of the criminal courtroom workgroup. Common Law World Review, 42(3), 203–​239. Zottoli, T.M., Daftary-​Kapur, T., Winters, G.M., & Hogan, C. (2016). Plea discounts, time pressures, and false-​ guilty pleas in youth and adults who pleaded guilty to felonies in New York City. Psychology, Public Policy, and Law, 22(3), 250–​259.

207

10 TO PLEAD OR NOT TO PLEAD? “GUILT” IS THE QUESTION1 Rethinking Sentencing and Plea Decision-​Making in Anglo-​American Countries2 Jay Gormley and Cyrus Tata

Anglo-​ American justice systems are said to be adversarial. Adversarialism has important implications for conceptions of how truth ought to be determined and the presumption of innocence protected. The Adversarial trial is a cornerstone of the legitimacy of Anglo-​American justice systems based upon the liberal rule of law. Although Anglo-​American justice systems have never perfectly operationalized the ideal of the adversarial trial, the ideal remains important. The presumption of innocence, embodied in the adversarial trial, is perhaps the most important idea binding the process together.3 However, in practice cases are disposed of primarily via guilty pleas and increasingly alternatives to prosecution through court, such as police and prosecutor fines. The dominance of case disposal via guilty pleas means that Anglo-​American justice systems’ key claim to legitimacy, the full adversarial trial, rarely occurs. Depending on the jurisdiction and the court, Anglo-​American justice systems dispose of up to 95% by way of a guilty plea. Nonetheless, defendants still have the right to a trial, even if they choose not to exercise it. Why defendants choose not to go to trial, and whether they do so freely, is one of the great controversies of contemporary criminal justice. It is widely believed that defendants who plead guilty are likely to receive a reduced sentence compared to the sentence if they are convicted of the very same charges following an evidentially contested trial. Various terms are used to describe this practice such as “sentence discount,” “guilty plea discount,” “sentence reduction,” “allowance in respect of a guilty plea,” “trial tax,” and “trial penalty.” While all these terms refer to the difference a plea can make to a sentence, they are all value-​laden, implying different normative positions. Those who object to the idea of altering sentences because of the plea contend that it undermines core rule of law principles, most notably the presumption of innocence. Those who object typically refer to altering sentences based on pleas as a penalty for exercising one’s right to trial (“Trial Tax” is a widely used term). Those who consider altering sentences based on pleas unproblematic for the presumption of innocence use terms such as “sentence discount.” As these different terms for altering sentences based on how the defendant pleads relate to the presumption of innocence, the differences are more than semantics. At stake is the credibility of the criminal law’s claims to be fair and legitimate as well as claim to moral authority in society. After all: [Law] is the most public of the many institutions that regulate social conduct as well as the one that claims to speak with authority for the moral community as a whole. (Weisman, 2009, p. 49) 208

To Plead or Not to Plead?

The underlying claims implicit in these distinct terms are discussed shortly, but for now, in the interests of neutrality, this chapter will primarily use the term “plea-​dependent sentence differential” (or “sentence differential” for short) to refer to the difference to a sentence that a not guilty as opposed to a guilty plea can make. The formal foundation for the plea-​dependent sentence differential varies depending on the jurisdiction. In some jurisdictions, statute or case law may permit, or even require, a sentence differential. In other jurisdictions, the basis of the sentence differential is found in guideline schemes. In certain jurisdictions, the sentence differential has little formal basis and is rooted in the informal practices of legal practitioners. However, regardless of its basis, in all Anglo-​American criminal justice systems of which we are aware, there is a widely held perception that how one pleads affects one’s sentence. While there are procedural nuances that vary between jurisdictions, in general, there are two broad ways through which a sentence differential may result. The first way is direct. A direct sentence differential results from a judge altering a sentence on the basis of the plea entered. For example, a judge may (depending on one’s view) ‘reduce’ a sentence because the defendant pled guilty or ‘increase’ the sentence because the defendant was found guilty after an evidentially contested trial. Alternatively, a sentence differential can occur indirectly. In this instance, the judge is not altering the sentence because of the plea itself. Instead, the judge is only considering the offense that the offender is convicted of.The sentence differential in this instance arises because plea bargaining can significantly alter the charges and the facts of the offense of which a defendant is convicted. For example, there are plea bargaining practices such as “charge bargaining,” where defense lawyers and prosecutors agree on a guilty plea to amended charges. These practices are indirect because they do not directly concern the sentence. However, often defendants may believe that they will result in a different sentence. “Sentence discounting,” “charge bargaining,” and, “fact bargaining” are all theoretically distinct forms of “plea bargaining” but are related in practice. Practitioners (especially defense lawyers and their clients) may consider the net effects of plea bargaining instead of each factor in isolation. Considering the sum of various forms of plea bargaining can create complex dynamics whereby maximizing one form of plea bargaining reduces another. A typical example is that an early guilty plea maximizes the judicially given sentence discount. However, often defense lawyers can extract better charge bargains and fact bargains when their clients continue to contest the charges. Accordingly, in the daily work of the criminal justice process, the relationship between plea decision-​making and sentencing is one of the most pressing questions. It is thought to influence how defendants plead, affect caseload pressures, and potentially have implications for the presumption of innocence as well as the experience of victims. This chapter evaluates the arguments for and against the practice of altering a sentence as a consequence of a plea of “not guilty” as opposed to “guilty” (the sentence differential). It also appraises the state of international knowledge about the practice and proposes agendas for future research. The chapter is organized as follows.The first part investigates whether a sentence differential violates the cherished values of the presumption of innocence and the notion of equality before the law. It examines the criticism that the sentence differential operates to penalize those who plead not guilty by imposing (or threatening to impose) a higher sentence than if they plead guilty. It also considers the criticism that the sentence differential may have disparate impacts on different groups (specifically minorities and those who are socially and economically disadvantaged).The second part reflects on why, in light of the dangers to principled sentencing and liberal rule of law values, justice systems continue to persist with guilty plea discounts. Finally, by reporting recent empirical research, the third part investigates the underresearched question of how the experiences of defendants may be affected by the sentence differential.

Does the Plea-​Dependent Sentence Differential Undermine Rule of Law Values? This debate over whether or not the sentence differential is a discount is one of the great controversies of contemporary criminal justice systems. In Anglo-​American justice systems, legal practitioners 209

Jay Gormley and Cyrus Tata

tend to refer to the sentence differential as something like a “sentence discount” or “reduction” that “rewards” a guilty plea. Articulating the sentence differential as a discount is how legal systems formally argue that the sentence differential does not undermine the presumption of innocence. However, several commentators regard the articulation of the sentence differential as a discount to be little more than legal sophistry to assuage the embarrassment that results from routinely and systemically violating the presumption of innocence. For instance, Darbyshire (2000) argues that the sentence differential: Undeniably punishes those who exercise their right to trial then are found guilty, however much the Court of Appeal tries to disguise a sentence discount as a reward for remorse.This is stunning hypocrisy in the Anglo-​American legal systems, whose rhetoric trumpets the right to trial, especially jury trial, the burden of proof and the presumption of innocence as the hallmarks of the world’s finest democracies. (p. 901) Thus, critics argue that by seeking to encourage guilty pleas (or at least earlier guilty pleas), the sentence differential risks undermining the presumption of innocence, and bypassing the safeguards a trial is argued to provide. For example, McCoy (2005, p. 90) insists that we must ask “whether it amounts to institutionalized coercion.” She argues that: Implicit plea bargains occur for two reasons: (1) the defendant knows the “going rate” of punishment and can accept it, and (2)  for some defendants who believe they have valid legal defenses, the threat of trial penalty will cause them to plead guilty regardless. (McCoy, 2005, p. 94) Any infringement on the presumption of innocence would run contrary to rule of law principles. Policymakers often attempt to rationalize guilty pleas and the sentence differential on the basis that the defendant is usually, and obviously, guilty. However, guilty pleas and sentence differentials challenge the presumption of innocence precisely because the power of the state looms so large over the defendant’s decision-​making. This power the state exercises over defendants’ decision-​making makes it impossible to know for sure whether a defendant’s admission of guilt is genuine or motivated by undue factors, such as fear of a higher sentence. So, how should the sentence differential be characterized? Is it a Trial Tax or a Discount? Does it represent state beneficence or a threat to exercising one’s right to trial? Much depends on what one sees as the “normal,” “default” or “baseline sentence.”Terms such as “reduction” and “discount” imply that the point of reference is the hypothetical sentence a judge would give if the defendant pled not guilty and was convicted following a trial. From this perspective, the sentence differential is a discount because the judge is giving a lesser sentence than she would have if the defendant pled not guilty.This perspective has its logic and is in line with legal notions of “discounts.” However, not all agree that the post-​trial-​conviction sentence should be considered the baseline or starting point. The counter-​argument is that the post-​trial-​conviction sentence is a largely fictitious baseline because, in reality, nearly all sentences follow a guilty plea rather than a conviction after an evidentially contested trial. For example, Lynch argues: In a system where ninety percent or more of cases end in a negotiated disposition, it is unclear why the ‘discounted’ punishment imposed in that ninety percent of cases should not be considered the norm. (Lynch, 2003, p. 1401) In England and Wales, the idea that the baseline sentence should be that following a guilty plea has been considered. In one instance sentencing guidelines were locally modified to assume a timely guilty plea, with the sentence in effect increasing where there was a late guilty plea or a 210

To Plead or Not to Plead?

not guilty plea (Henham, 2000, p. 439). Using guidelines in this way was normatively troubling, as it came close to formally recognizing that the sentence differential is a penalty for going to trial. However, the advantage of this local modification was that it bore more relevance to the majority of cases. Taking the baseline sentence as the lower sentence given in most cases could be considered more straightforward and practical. The usual alternative is a guideline or norm based on a first-​time offender who is convicted following a trial. Yet in most jurisdictions, this is atypical. For example, Ashworth and Roberts note that in England and Wales only 12% of offenders plead not guilty, “and only 10% appeared for sentencing without any prior convictions” (Ashworth & Roberts, 2013, p. 7). Consequently, thepercentage of those sentenced who meet both these conditions is “obviously much smaller than 10%” (Ashworth & Roberts, 2013, p. 7). Thus, guidelines and official norms based on a first-​time offender could be said to give a misleading impression as they do not match the everyday reality of most cases. The argument that the normal reality of sentencing following guilty pleas should form the assumed starting point or baseline challenges the notion that the sentence differential is a discount. However, ultimately the debate is not possible to resolve objectively: it depends on whether or not one sees the likely sentence following an adversarial trial without plea bargaining as the correct starting point. “Discounts” and “penalties,” like notions of “gain” and “loss,” “are malleable concepts” (Bibas, 2004, p. 2512). This malleability means that room for different views is inevitable. Consequently, what may matter more is what the defendant perceives the sentence differential to be. While there is a growing interest in “user perspectives” of the justice system, so far there is only limited knowledge about defendant perspectives—​a deficit which we argue in section three Anglo-​American justice systems need to address.

Does the Sentence Differential Encourage Innocent Defendants to Plead Guilty? The central argument in defense of the sentence differential is that the defendant has a free choice to contest her case and go to trial if she genuinely believes she is not guilty. The assumption is that any innocent defendant will choose to maintain her “not guilty” plea and leave it to the prosecution to try to prove its case if it can. Further protection against wrongful conviction is provided by the defendant’s lawyer who will ensure that she is fully informed. However, is there a genuine danger that the sentence differential will induce innocent defendants to plead guilty and thus result in wrongful convictions? To some, it might seem unlikely that an innocent person would plead guilty when they are innocent, and that this risk is merely hypothetical. Surely, the supposition runs, no person would plead guilty to something they have not done; defendants choose rationally and freely how to plead. Moreover, some might expect that this would not happen because skilled lawyers hold steadfastly to cherished values such as the presumption of innocence and protect the best interests of their clients. However, while this may be true of some defendants, the sentence differential may still contribute to other innocent defendants pleading guilty. There are at least six reasons to doubt the supposition that it is implausible that defendants do not plead guilty to charges of which they may not be guilty. First, the literature on the relationship between criminal defense lawyers and their clients clearly shows that although in formal terms defendants “instruct” their lawyers, the reality is much more complex. Empirical research “has consistently highlighted the relative passivity of most clients” (Tata & Stephen, 2007, p. 732, see also Carlen, 1976; McConville et al., 1994; Baranek & Ericson, 1982). In most contexts, there is an inequality of power in the professional–​client relationship. However, this inequality of power is especially acute in the context of most criminal cases. The weak educational, social and personal resources of many defendants can severely limit their ability to instruct their lawyers and control their cases. 211

Jay Gormley and Cyrus Tata

Additionally, the stress and anxiety of being subject to criminal charges can hinder the ability of defendants to control their cases (especially while held in pretrial detention); the unfamiliar vernacular; and the requirements of criminal and court procedures (which can vary by the court in some jurisdictions). As a result, it can be difficult for many clients to take firm command of their cases. Indeed, it is not inconceivable that some defendants may not always fully understand the charges against them. Even where a defendant knows what she has done, she may or may not be fully aware of what crime in law it may or may not constitute.While research is urgently needed to explore more fully what defendants do and do not understand, for now, there is a reason to question the assumption that guilty pleas are always made freely and are always fully informed. Second, research has shown that the criminal process in the lower and intermediate courts tends to be characterized by a presumption that guilty pleas will settle cases. There is a widespread expectation among court personnel that an admission of guilt by the defendant is inevitable (e.g., McBarnet, 1981; Carlen, 1976; Feeley, 1982; Heumann, 1978; Tata, 2007, 2010). This expectation can become self-​perpetuating in very practical ways. For example, court schedules depend on a high volume of guilty pleas to assist in the expeditious processing of cases. In many jurisdictions, the court, the prosecution, and the defense will often overbook in the expectation that most cases will plead guilty and that trials will be unnecessary. This decision to overbook (to be more “efficient”) makes the criminal process dependent on guilty pleas. If the anticipated guilty pleas are not forthcoming, then the typical working practices of the court cannot function. Thus, a belief that guilty pleas will be forthcoming creates a routine where guilty pleas are necessary, which results in a self-​perpetuating circle of dependence on guilty pleas that come to be sought by pragmatic means. Third, while lawyers may endeavor to look after the best interests of their clients, it is now well established that lawyers also have to be cognizant of a range of potentially competing imperatives, including longer-​term inter-professional​and intra-​professional relationships (e.g., Eisenstein & Jacob, 1977; Heumann, 1978; Tata, 2007) as well as the structuring of financial interests (Tata, 2007). Lawyers not only owe obligations to their clients, as officers of the court they also owe duties to the court. Lawyers must reconcile these potentially conflicting duties alongside their interests: such as the need to remain profitable and secure working relationships with others in court. That lawyers must reconcile their interests does not mean that dedicated professionals discard cherished values for career or financial gain: professional behavior is more complicated than that. It is, however, fair to say that there is a range of competing dynamics which they have to find some way to resolve and which mesh with advice on plea decision-​ making (Tata, 2007). Fourth, some defendants, even where they do not accept the prosecution case, feel unable to challenge it effectively. For example, defendants may be charged with an allegation when they were under the influence of alcohol or drugs and may have only a hazy memory of events. A  defendant with a poor recollection of events relating to a charge may be at a disadvantage, and this disadvantage may make them less willing to contest the charges (Tata, 2010). Other defendants may not trust the system, or defendants may feel that the court will not believe them and not find them credible. In this case, they may plead guilty due to passive resignation rather than any true acceptance of guilt. How often such issues occur is difficult to quantify, and research on this question is crucial. Fifth, defendants may make a deliberate choice to plead guilty to charges of which they are not guilty. Initially, this might seem to conflict with theories of decision-​making that posit the defendant as a rational actor who would not wish to incriminate him/​herself. However, some defendants may feel obliged to plead guilty so that another family member, friend, codefendant or gang member avoids conviction. For example, the prosecution may sometimes offer to drop charges against family members of the defendant to extract a guilty plea. Other defendants may struggle with the demands of the court process and so conclude that a plea of guilty is the least bad option. Indeed, pleading 212

To Plead or Not to Plead?

guilty, even when innocent, can be considered a better option due to the costs which the criminal process inflicts (Feeley, 1982). For example, Albert Alschuler has noted that: A misdemeanor defendant, even if innocent, usually is well-​advised to … plead guilty at the earliest opportunity … to minimize the painful consequences of criminal proceedings. (Alschuler, 1983, p. 953) These painful consequences are especially true where the defendant is held in pretrial/​sentence detention (remand) and the sentence could be backdated to cover, wholly or in part, time served. A  guilty plea, when held in remand, is also beneficial where the final sentence is non-​custodial (a surprisingly common occurrence). In both cases, the effect of a guilty plea can be immediate liberation. Sixth, and finally, the difference in the potential punishment if a defendant goes to trial can be extortionate. Although some argue that innocent defendants still have the choice to go to trial, some commentators have suggested that large sentence differentials can be incompatible with the presumption of innocence or even coercive (Caldwell, 2011; Ashworth, 2006, pp. 256–​257). As McCoy (2005, p. 90) puts it: There is nothing wrong with pleading guilty in the expectation of receiving a ‘going rate’ of punishment. But … there is a lot wrong with pleading guilty if that going rate after trial is so huge as to be the reason a defendant will make a pre-​emptive guilty plea. A defendant may run the risk of a much higher sentence by not pleading guilty. Some of the most infamous examples of this are where going to trial has prompted the prosecution to charge under “three-​strikes-​and-​you’re-​out” laws. In these circumstances, where there is a perception of a large trial tax, an individual may rationally choose to plead guilty even if they are innocent: paying a small but certain cost to avoid the risk of a contingent but larger cost if convicted following a trial. From the perspective of such defendants, this is a meaningful choice. Indeed, statistical theory may also bear this out to be rational in some circumstances. How large the sentence differential must be before it becomes “coercive” is a matter of ongoing debate. Like Appeal Courts, reformers have wrestled with the question of “how large a percentage difference is too large?” Although the question of magnitude is critical, it is defendants’ perceptions that determine their pleading decisions. Defendants form their perceptions of the likely sentence differential by making evaluations through the lens of a broader range of other criminal justice process costs (not least, for example, being held on remand) and a range of personal (e.g., health and family) circumstances. For all these reasons, and more, it is unsafe to assume that innocent persons never plead guilty. The sentence differential may add to this issue by making it more likely that an innocent person will plead guilty. Exacerbating this problem is that legal guilt is often a nebulous concept. Thus, despite legal rhetoric to the contrary, it seems that the sentence differential does pose challenges to the presumption of innocence.

Do Plea-​Dependent Sentence Differentials Violate Legal Equality? One criticism of the plea-​dependent sentence differential is that it has a disproportionately adverse effect on minority groups, thereby making the practice indirectly discriminatory and undermining the principle of legal equality (e.g., Savitsky, 2009). However, identifying indirectly discriminatory practices is challenging, as “a wide variety of practices that are facially neutral … have racially disparate effects” (Tonry, 2012, p. 87). One infamous example of this was the USA’s drug laws penalizing the use of crack cocaine more severely than powdered cocaine. While this seemed racially 213

Jay Gormley and Cyrus Tata

neutral, it disproportionately affected poorer minority groups more likely to use the cheaper crack cocaine (Provine, 2002; Tonry, 2011, p. 1). The sentence differential may indirectly discriminate against minorities if minority defendants: Are in a worse bargaining position than white ones, and that this differential bargaining power makes [minority] defendants more likely to make worse bargains than similarly situated white defendants. (Savitsky, 2012, p. 135) It seems that minorities may both start from a disadvantaged position, be treated differently, and receive poorer bargains from prosecutors. According to Sah, Robertson, and Baughman (2015), Although the U.S. Constitution theoretically limits the discretion of prosecutors (to target a particular race prejudicially, for instance), such protections are exceedingly difficult to invoke, especially if a prosecutor’s unconscious rather than intentional bias is in play. (p. 70) Given that most sentences follow a guilty plea, the receipt of worse bargains may lead to a more severe charge on conviction and a higher sentence, even if judges’ sentencing is not affected by any implicit racial bias. The effect of this can be significant and Berdejó (2018) has argued that: White defendants are twenty-​five percent more likely than black defendants to have their principal initial charge dropped or reduced to a lesser crime. As a result, white defendants who face initial felony charges are less likely than black defendants to be convicted of a felony. Similarly, white defendants initially charged with misdemeanors are more likely than black defendants to be convicted for crimes carrying no possible incarceration or not being convicted at all. (p. 1188) It is also possible that inequality could be the result of minority groups being disposed to plead differently from majority groups. Indeed, Henham (2017) has noted that: Research suggests that the fact ethnic minority offenders were more likely to contest the charges against them inadvertently subjected them to a form of indirect discrimination since the system encouraged sentence discounts for guilty pleas. (p. 4) One factor that may influence whether a defendant pleads guilty or contests the charges is their level of trust in the justice system and legal actors. The effect of trust on plea decision-​ making is problematic as there is evidence that certain minority groups have less faith in the criminal justice system than majority groups (e.g., Shute, Hood, & Seemungal, 2005; Weitzer & Tuch, 2002). This mistrust may be demonstrated by the higher proportion of black defendants in England who plead not guilty and choose to contest their case before a jury of their peers, rather than to trust a judge sitting alone (Hood, 1992, p. 196; Thomas, 2010, p. 21). Furthermore, Michael Tonry has noted that ‘black defendants less often plead guilty and when they do, they do it later’ (Tonry, 2012, p. 75). These differences in inter-​group pleading behaviors are significant in justice systems where pleading decisions may dramatically affect sentences due to a sentence differential. For example, if a ‘guilty plea discount’ is one-​third (e.g., 120  days down to 80)  then by not pleading guilty a defendant’s sentence will be 50% higher than a comparable defendant who pleads guilty (i.e., a sentence of 120 days is 50% more than one of 80 days). Indeed, Hood’s study suggested that pleading not guilty accounted for 13% of the total difference ‘in the black male general population and their proportion among those serving [custodial] sentences’ (Hood, 1992, p. 203).4 Thus, lower levels of 214

To Plead or Not to Plead?

trust may mean that the sentence differential indirectly contributes to racial disparities in criminal justice systems. Why some minorities may have lower levels of trust cannot be thoroughly explored within the confines of this chapter. However, as Bennett (2016) implies, such lower levels of trust should hardly be surprising: That implicit racial bias may affect sentencing should come as no surprise. There “is a rich and overlapping literature” documenting implicit racial bias by white Americans favoring whites over blacks, and commentators “almost universally agree” that racial disparities are pervasive. (p. 396) However, one factor may be that some groups are more likely to feel victimized by agents of the criminal justice system. For example, minority groups may feel targeted in various ways, such as through “racial profiling” (Tator & Henry, 2006; Welch, 2007). The result of this, according to Tonry (2012) is that: Put into a single sentence: young black men [and others] who believe themselves unfairly treated by the police understandably become angry and uncooperative, and are punished more severely as a result. (p. 74) However, understanding and empirically testing the potentially discriminatory effects of the sentence differential is not easy. There are many variables to consider (e.g., intersectionality between race, class, and gender) and the information is limited, particularly regarding defendants’ subjective perceptions and how these influence pleading decisions.Though, for now, it seems safe to suggest that, while ostensibly neutral, the sentence differential may contribute to wider racial disparities (Tonry, 2012, p. 87).

How is the Sentence Differential Justified? There are potentially serious problems with the sentence differential. It may be a threat to the presumption of innocence and notions of equality before the law. Given the significance of these potential problems, why is the sentence differential tolerated and even encouraged by policymakers around the world? Three main rationales are advanced to justify the sentence differential (deemed to be a discount) at a policy level. The section explores these rationales. The first rationale is the remorse rationale. The remorse rationale is the argument that guilty pleas demonstrate remorse and that this warrants the reduction. The second rationale is the victim rationale. The victim rationale claims that reductions are justified on the basis that the guilty pleas spare victims from the further ordeal of a trial. The third rationale is the efficiency rationale. The efficiency rationale supposes that reductions are legitimate as they induce justified guilty pleas (and earlier guilty pleas) by people who are guilty, thereby saving resources such as court time and money. However, all three of these rationales have serious limitations, which begs the question of whether there is some other reason that the sentence differential exists. Let us first scrutinize these three justifications before asking whether there are any other reasons for the sentence differential.

The Remorse Rationale Remorse, empathy, and compassion are difficult questions for judges when sentencing and, more generally, pose challenges for the rule of law itself (Bandes, 2017). Using remorse to justify sentence discounting relies on the notion that a remorseful offender is less deserving of punishment, or that the remorseful offender is worthier of mercy. It also relies on the assumption that the guilty plea is a 215

Jay Gormley and Cyrus Tata

sign of remorse. However, there are problems with this rationale in both practice and theory.The first problem lies in identifying genuine remorse. The second problem is explaining why remorse justifies a lesser sentence. These are dealt with in turn. Identifying genuine remorse is difficult (e.g., Bandes, 2016; Leverick, 2004, pp.  370–​ 372; Rossmanith, 2018; Weisman, 2014). It might be assumed that the guilty plea itself is evidence of remorse. However, the perception of a sentence differential means that this argument is weak. As long as defendants perceive that they can, in relative terms, benefit from a guilty plea it cannot be known whether a guilty plea indicates remorse, a tactical decision to try to benefit from a discount, or both. Additionally, even if the sentence differential were not a factor, there are many other reasons why a defendant might plead guilty, such as to avoid the stress and uncertainty of a trial or to be liberated from pretrial/​sentence detention. Despite these problems, many may feel that they can identify genuine remorse through their ability to empathize with the defendant. However: The problems arise from selective empathy and from empathic inaccuracy (Bandes, 2009). For judges, these are acute problems, because judges are encouraged to believe in their own omniscience. (Bandes, 2017, p. 192) A judge may empathize differently with those who are similar to themselves than those who are different, which may contribute to inequality. For example, emotions may be expressed differently by different cultures and things that add credibility in one culture (e.g., norms regarding eye contact) may not be present in another. As well as adding to inequality, the ability to empathize is imperfect. There is also a risk that some defendants tendering guilty pleas for tactical reasons will be thought remorseful and therefore will benefit from a discount. At the same time, others who are pleading guilty because of genuine remorse may be thought disingenuous and therefore will not benefit. The second issue with justifying the sentence differential based on remorse is that it is unclear why remorse should justify a reduction in sentence. Maslen and Roberts argue that “desert theory provides the primary theoretical basis for sentencing guidance” (Maslen & Roberts, 2013, p. 125). If this is correct, the remorse rationale is questionable as it does not affect culpability or harm. Remorse occurs after the fact of the offense (Leverick, 2004, p. 370).5 Thus, remorse does not affect what the defendant did, nor how blameworthy they were at the time. There is also little evidence that displays of remorse reduce the risk of reoffending and, if remorse is thought to be intrinsically painful for the person, it is questionable whether sentencing should consider this extrajudicial punishment. These limitations of the remorse rationale mean that it is hard to pinpoint a logical reason why a remorseful defendant should receive a reduced sentence based on penal principles. Indeed, on this view, the sentence differential may undermine the logic of principled sentencing.

The Enduring Appeal of Remorse: Affective and Legitimacy Explanations Despite the normative problems that rewarding defendants for expressing remorse create, there is widespread support for the notion that remorse is important and that it can provide an acceptable reason for a reduced sentence (Maslen & Roberts, 2013, p. 124). Remorse can also be an important factor for policymakers as well. For example, in Scandinavia reductions for admissions have been “rationalized in terms of contrition or acceptance of responsibility” (Tonry & Lappi-​Seppälä, 2011, pp. 16–​17). In the US discounts for “acceptance of responsibility” tend to be premised on a remorse rationale. Perhaps the reason for remorse’s enduring appeal is linked to the expressive, affective, and normative elements of the criminal law. Throughout its history, it seems law has never been solely an enterprise concerning the objective application of rules in a formalistic fashion. Indeed, as the classic critiques of formalism powerfully demonstrate (cf. Leiter, 2010), this is not possible. Law is always indeterminate to some extent and judges are human and fallible. 216

To Plead or Not to Plead?

Moreover, imposing punishment can present moral challenges for judges. Judges facing moral dilemmas is not a new phenomenon, and judges have long struggled to reconcile the punishment they impose with notions such as compassion. For example, at one point “police court missionaries were welcomed in the courts because they provided justifications for lenience to magistrates who were increasingly concerned about the harsh effects of the cumulative principle in sentencing” (Vanstone, 2004, p.  36). Indeed, regardless of how it is articulated (whether it is called “leniency,” “mercy,” or “acceptance of responsibility”), these expressive, affective, and normative elements are intrinsic to the criminal law. Indeed, Vogel (2007) traces the origins of plea bargaining to older discretionary concepts of leniency: One notable legal innovation developed as men and women drew on the language and forms of the time-​honored tradition of episodic leniency and reforged it, along with indigenous elements of the Puritan religious practice of admonition, to produce a new discretionary form—​namely, that of plea bargaining. (p. 133) Thus, while a sentence differential (especially a large differential), because an offender is remorseful, is hard to justify in terms of penal principle, remorse does have some enduring appeal. The legacy of formalism creates an inescapable bind which seeks to deny the affective, expressive, and normative role of law. It supposes that punishments are the inevitable consequences of the application of formal law to self-​evident case facts, rather than cathartic retributions, empathetic reactions, etc. Ultimately, this formalistic view obscures the fact that “in law, we are always making choices about which emotions advance legal goals” (Bandes, 2017, p. 185) and in making this choice remorse may play an important part. Thus, a central and inescapable paradox afflicts the courts in its search for genuine remorse. On the one hand, there is the pervasive professional desire for speed, control, and predictability by settling cases as far as possible by way of guilty pleas. On the other hand, there is also a professional requirement to know whether or not the person fully and freely accepts responsibility for the offense she has formally admitted to. Ambiguous, insincere, or confused formal admissions of guilt are highly problematic for court professionals. Formal admissions which appear to be contradicted by the person’s account (for example in a pre-​sentence report or plea in mitigation or allocution by the defendant to the court in which she may explain her position in terms of the formal guilty plea) highlight difficult questions about the fairness and voluntariness of the process through which such admissions are generated (Tata, 2019). In a sense, court communities have to live with the perpetual doubt which they generate: has she pled guilty because she accepts that she really is guilty? Or is she pleading guilty for fear of the sentence differential; or, to get out of remand; or, to take the rap for someone else; or, because she doesn’t expect that she would receive a fair trial, etc.? Operating in the shadow of law’s threats and inducements means court professionals are condemned to keep guessing as to whether or not the person’s formal guilty plea is really free and genuine.6 This requirement to feel that the defendant fully and freely accepts responsibility may be where expressions of remorse play a key role. As Weisman (2014) so well explains, remorse must not simply be experienced by the person but be shown and displayed to the court. The court must look for signs of remorse. It is not enough to hear that a defendant feels contrite. That contrition and regret must be shown to the court. Such a showing should be convincing, authentic, and heartfelt. Court professionals talk of knowing remorse when they are able to sense and feel genuine remorse in front of them (Rossmanith, 2015). Experiencing the defendant’s signs of remorse shows the court (and the wider public) that the person fully and freely accepts her responsibility for her offending (Tata, 2019). She can be understood to be within the court’s (and wider community’s) moral community (Weisman, 2009). There is no equivocation or resistance, no ifs, buts, or excuses in “genuine remorse.” The remorseful defendant is seen by the court (and the wider community) to accept responsibility completely: so much so that even before sentencing she is almost punishing herself. 217

Jay Gormley and Cyrus Tata

The genuinely remorseful defendant fully, freely, and sincerely accepts responsibility and the fairness of the impending punishment. In this sense, then, signs of genuine remorse are sought by the court (at least in part) because they are seen as the ultimate signs of the legitimacy of the criminal-​penal process. Convincing signs of authentic remorse reassure court professionals of the fairness of their penal work (Tata, 2019), and assure the wider community that the defendant is not a fundamental threat to the moral order (Weisman, 2014).

The Victim Rationale Another rationale for sentence differentials is that there are benefits for victims of crime.This rationale assumes that a guilty plea will best serve a victim as this means that she or he will not be subjected to the trial process. Indeed, it is true that a trial can be an arduous ordeal for a victim that, among other things, can require the victim to be cross-​examined and have his integrity questioned—​an experience that some victims may prefer to avoid (Dawes et al., 2011, para 3.2). However, while sparing victims the further ordeal may appear to be compassionate, there are reasons to question this rationale. One reason to object is that in many cases where there is a sentence differential the only witnesses are police officers (Page et al., 2010, para 7.1.2). In these cases, the victim rationale does not apply. Second, it is wrong to stereotype victims. Just as defendants’ circumstances vary, so do those of victims.These variations are significant as there are victims who are not well served by sentence differentials (Darbyshire, 2000, p. 905). For example, some victims may prefer a contested trial if this means they will learn more about the offense and the offender—​which may help them obtain emotional closure. Indeed, there is an ongoing debate about whether victim impact statements may be beneficial for victims (Bandes, 2016). Additionally, some victims may resent being “spared” for the simple reason that the cost of this is that the defendant who wronged them receives a lesser sentence: though this risk must be weighed against the increased odds that an offender will not be convicted without a guilty plea. Thus, while “sparing” victims is a noble endeavor, these good intentions do not always have positive results. While the victim rationale seems plausible, it does not apply to many cases where there is a sentence differential, as there is either no victim or the victim is not one who would benefit from being “spared.” Additionally, something is troubling about the criminal process when it becomes so burdensome on victims that they need to be protected from it, potentially at the expense of sentencing principles. Indeed, Leverick (2004) has criticized the victim rationale for focusing attention on avoiding flawed criminal processes rather than improving them. For example, in some instances, an issue with the trial process has been that alleged attackers may cross-​examine victims. However, the solution to the flaws in the trial process should be to rectify the trial process, rather than avoid the trial altogether for the good of the victim (Padfield, 2012, pp. 361–​377). Another drawback of minimizing the victim’s participation is that it is detrimental to the function of the criminal trial as a public communicative forum playing out the nature and limits of social mores. In this sense, the trial is not something that is only for the disposal of cases concerning the interests of its immediate participants, but also serves as a forum for wider public discussion and debate. Through public displays of emotion, ritual, and drama, the trial may demonstrate the polity’s commitment to justice. It plays an essential cathartic role for victims and the public, which is increasingly encouraged to identify with “the victim” (Sparks, 2011). In so doing, in daily practice and especially in high-​profile cases, the trial enables a moment of collective moral expiation, which in turn helps to constitute a sense of community. In this way, the phenomenon of the vanishing trial not only silences the victims and defendants, but it also denies the ability to hear their stories publicly and for the law to show that it has listened to victims and defendants before coming to a decision. The decreased incidence of trials and the participation of victims entails the loss of public displays of emotion, drama, and ritual, in turn perpetuating a feeling that criminal justice has become a sterile

218

To Plead or Not to Plead?

automated process devoid of moral drama and meaning. An effect of this is that it tends to undermine public confidence in the legitimacy of sentencing and the broader criminal justice system (Tait, 2002).

The Efficiency Rationale Those promoting an efficiency rationale for the sentence differential propose that, while trials may be the ideal way to safeguard the presumption of innocence, resource constraints require that this ideal cannot be achieved in most cases. Consequently, it is deemed necessary to incentivize early guilty pleas with a sentence differential.Thus, the efficiency rationale is based on a claim of pragmatic necessity: inferring that without guilty pleas the system would collapse under the increased workload. In many ways this claim is persuasive. It is uncertain how Anglo-​American justice systems, so accustomed to guilty pleas and plea bargaining, would respond to many contested trials. Additionally, it is important to note that the efficiency argument, properly made, is not necessarily amoral. The justice system runs on public money, and there is a moral duty to ensure it is spent wisely.Thus, the efficiency rationale can be made in such a way to provide it with a moral (utilitarian) foundation. However, is this claim that plea-​dependent sentence differentials save money factually accurate? While removing the sentence differential may result in fewer (and later) guilty pleas, there are other reasons why defendants plead guilty (Schulhofer, 1984, p. 1040; Feeley, 1979). These other reasons why defendants plead guilty mean that it should not be assumed that all guilty pleas are the result of the perceived sentence differential. Indeed, research on why defendants do and do not plead guilty is desperately needed (a point this chapter will stress later). Moreover, several commentators have suggested that the caseload necessity thesis is not as certain as may be assumed. For example, Weigend (2006) notes that the caseload thesis: Has little evidence to support it—​the time and location of system changes from trial to non-​trial adjudication are not related to significant increases in case input. This is especially true for countries where the introduction of bargained case dispositions did not result from overburdened courts’ search for an outlet but was the product of comprehensive legislative reform. (p. 213) Indeed, from both international and historical perspectives, the empirical link between the rise of the sentence differential (and plea bargaining more generally) and the level of workload is tenuous (e.g., Eisenstein & Jacob, 1977; Feeley, 1979, 1982; Heumann, 1975, 1978; Mather, 1979;Vogel, 2007).7 By comparing low and high court volumes across time, it has been found that remarkably similar proportions of cases result in a guilty plea. McCoy (1983) sums up the point neatly: The finding that a high percentage of cases conclude with guilty pleas even when there is very little caseload pressure undermines the conventional wisdom that explains plea bargaining in terms of efficiency. (p. 59) Thus, it cannot be assumed that the sentence differential is necessary and there is a good reason to be skeptical of the efficiency argument. There is also reason to be skeptical of the assumption that the sentence differential promotes the expedient disposal of cases and that guilty pleas save resources. Ironically, the expectation that cases will invariably settle by way of guilty pleas or be dropped can itself lead to a wasteful delay in court proceedings. Prosecutors may tend to be more improvident in how they file charges than they would if they expected a trial, in the partial expectation that those charges will later become useful bargaining chips (Caldwell, 2011, p.  65). Defense lawyers may respond to this by not pleading guilty to the charges initially filed out of a belief that these are

219

Jay Gormley and Cyrus Tata

exaggerated and can be reduced. Lawyers may also adapt by delaying settlement until trial, which may be a tactic to extract a better deal from a time-​pressured prosecutor (Tata, 2007, p. 512). This expectation of a guilty plea means that prosecutors and defense lawyers tend to avoid a thorough examination of the evidence at the earliest opportunity in the expectation of settlement later in the process, or of withdrawal of the case (Tata & Stephen, 2007; Bradshaw et al., 2012, para 6.14).This expectation is self-​perpetuating. The expectation of guilty pleas also leads to a lack of preparation and an ‘adjournments culture’ (Kemp, 2008), which from the perspective of individual practitioners is rational and prudent, but from a system-​wide perspective is inefficient.Thus, the sentence differential may work to normalize an inefficient culture and contribute to ‘phenomena aimed at delaying the progress of cases’ (McInnes, 2004, Chapter 28; see also Bradshaw et al., 2012, para 6.14 for a discussion of reasons to delay a guilty plea). Additionally, there can be a tendency to assume that the processes encouraging guilty pleas cost nothing. However, while the costs may be hard to quantify, they do exist and it is incorrect to assume that they are ‘trivial’ (Schulhofer, 1984, p.  1040). Lawyers may have to devote time to persuade defendants to plead guilty, and prosecutors may have to devote time to overcharging and haggling with the defense. Furthermore, even seemingly minor delays related to guilty pleas have consequences. Schulhofer (1984) has noted that in his sample various delays meant that the average case involved 35 minutes of waiting, and the “average guilty plea proceeding in fact required 55 minutes of courtroom time for the conviction stage alone” (pp. 1056–​1057). While these figures do not take account of the entirety of the potential culture of inefficiency that results from practices designed to promote guilty pleas, they do suggest that the costs might be more significant than is generally assumed. Moreover, it is worth noting that even if the removal of the sentence differential did increase the workload of the justice system, it might adapt rather than grind to a halt. Some have suggested that a simplified trial procedure could be used, while others have argued that a total ban on guilty plea discounts may be unsustainable, but a partial ban could work (Schulhofer, 1984; Alschuler, 1983; Gazal-​Ayal, 2005). Thus, while the necessity argument cannot be dispelled completely, it should not automatically be taken as true. Finally, the widely cited concept of “efficiency” bears scrutiny. Efficiency is, of course, a laudable goal with which no one can reasonably disagree. However, implicit in the debate about the sentencing differential is profound disagreement about what counts as “efficient.” Measuring “efficiency” in the justice system is not a value-​free exercise. It implies normative claims about what is “just,” what is “necessary,” and what is “wasteful.” These claims are based on different and competing perspectives about the “correct” role of prosecution, defense, judge, and indeed the justice of the criminal process as a whole (Tata, 2007). It is easy to imagine that “efficiency” is about doing things more productively. The prevailing image is the factory assembly line where production of the desired output from no more input (e.g., labor, outlay) equals greater productivity.When the output is shoes or widgets, it is relatively clear and easy to measure. Yet, what should be the desired output of criminal justice? Is it the sheer number of cases processed? Surely, the output of a justice system must be justice. Therefore, an increase in case disposals which primarily results in injustice is, by definition, self-​defeating and therefore inefficient. In other words, when even the cheapest justice system produces injustice, it fails to be efficient. In the same way, injustice can only ever be a mark of inefficiency: the system has failed to produce what it should. In other words, logically, it cannot be enough to justify the sentence differential on the grounds of cost alone. Noting the tenuous link between justice and efficiency is not to deny that the sentence differential may dispose of cases in a way that also dispenses justice, but the link between the two cannot be assumed: it is tenuous and contingent.

Is There Another Reason for the Sentence Differential? Notions of the sentence differential may be propagated at a policy level by a belief that the sentence differential results in more cases pleading guilty and pleading guilty at an earlier stage. However, 220

To Plead or Not to Plead?

policies do not explain the existence of sentence discounting. In almost all jurisdictions perceptions of a sentence differential (whether through sentence discounting, plea bargaining, or other means), pre-​date formal policies of the same effect. The perception of a sentence differential emanates from the bottom up (in the reality of daily practice) and eventually finds itself expressed in formal law and policy, with at least one the three rationales noted above ascribed to it. Vogel (2007) offers a nuanced account of the origin of the sentence differential, attributing its origin to wider social and political changes: An urban political elite, seeking to maintain its position of power, played a key role in its establishment. This privileged group, responding to political challenge in a specific social and temporal context, shaped much of the imaginative construction of American legal ideas during this formative era. It was this elite’s perception of crisis and threat, along with its effort to preserve social order, the legitimacy of self-​rule, and its own dominance, which shaped the practice of plea bargaining in a single locale that would then become a national and, eventually, an international phenomenon. (p. 5) These grassroots origins of the sentence differential do not appear to be limited to Anglo-​American systems. For example, Rauxloh (2010) notes the emergence of informal mechanisms for case disposal via a sentence differential in Germany and how the formal law was changed to match practice, rather than the other way around. But why does daily practice generate this perception of a sentence differential? In the case of Anglo-​American justice systems, their adversarial nature is thought of as a cornerstone of the legitimacy of the process. However, this notion of adversarialism does not accord with the reality of what actually happens in practice. Several commentators have argued that plea bargaining has brought about the demise of adversarialism: Blumberg is not alone in ascribing the demise of the adversary system to the rise of plea bargaining. In what have quickly become classics, University of Colorado law professor Albert Alschuler has examined plea bargaining from the perspectives of the prosecutor, judge, and defense attorney … He concludes that from each of these views the prevailing incentive is one of institutional convenience and organizational maintenance rather than the interests of the accused and the concern with justice. (Feeley, 1982, p. 339) In practice, especially in the lower courts, legal actors work together in many complex but collaborative ways (Eisenstein & Jacob, 1977). Legal practitioners form relationships with others in court, case disposals are generated through collaborative means, etc. Even though there is no perfectly adversarial system (it is an ideal (or pure) type) the reality is so far removed from this ideal that it poses uncomfortable questions about the validity of adversarialism as a touchstone of legitimacy. In fact, the manner in which most cases are disposed of in Anglo-​American justice system has about as much in common with Inquisitorial ideals as it does Adversarial ones.8 Regardless of how one characterizes the operation of Anglo-​American systems, it is clear that in many ways the lack of adversarialism and the predominance of collaboration is congenial for legal practitioners. For example, guilty pleas serve various ends. Structures of publicly funded remuneration for defense lawyers may not pay significantly more for the extra work of a trial, and prosecutors and judges may like to clear their cases without the uncertainty and stress of going to trial. However, this collaboration is problematic in terms of the perceived legitimacy of the process since legitimacy is partly drawn from notions of adversarialism and the selflessness of ethical legal practitioners.The sentence differential, in daily practice, works to mitigate this issue. Let us explain. If it appears the guilty plea benefits the defendant, then it also appears that the choice to plead guilty has been made by a defendant freely. The guilty plea also serves a symbolic role in having the 221

Jay Gormley and Cyrus Tata

defendant affirm the righteousness of the criminal process. When defendants admit their guilt to the court (and ideally show remorse), they affirm the legitimacy of the court. This cements the court’s position as a higher form of authority. In the midst of this, the benefits to legal practitioners can be glossed over, and the benefits to a defendant can be presented center stage. This protects the professional self-​image of the honorable, diligent, selfless professional. Part of the social capital (the prestige) of being a legal professional is that they serve a higher calling beyond their self-​interest. For legal practitioners, they serve the court and the wider interests of justice. Indeed, as Flexner (2001) notes, professionals may be expected to self-​ sacrifice to serve this higher calling: Devotion to well-​doing is thus increasingly likely to become an accepted mark of professional activity; and as this development proceeds, the pecuniary interest of the individual practitioner of a given profession is apt to yield gradually before an increasing realization of responsibility to a larger end. (p. 156) This is often drawn upon to justify (sometimes) high salaries and ground claims of the honorable duty of their work (Sommerland, 2015). It is also used as a reason why professions ought to have more autonomy to regulate themselves (e.g., Marshall, 1939; Parsons, 1939): Society’s granting of power and privilege to the professions is premised on their willingness and ability to contribute to social well-​being and to conduct their affairs in a manner consistent with broader social values. (Frankel, 1989, p. 110) However, this argument is tautologous. The system creates its own rules. The legal profession, in particular, has a high level of self-​regulation and an ability to set its own standards. In doing so it has generated the perception of a sentence differential and a permissibility for plea bargaining. It then uses this to explain away high levels of guilty pleas and treats the sentence differential as an inevitable fact over which it has no control. This is akin to a casino claiming it merely provides the games for customers to play and washes its hands of the odds which mean that overall the house always tends to win. Another explanation of the persistence of the sentence differential is that, if defendants plead guilty, it makes the limitations to adversarial ideals of the justice system less notable. There is less scrutiny of the process. Indeed, it is remarkable how prevalent plea bargaining in England and Wales became before it was formally acknowledged (Baldwin & McConville, 1978). Even today, when there is recognition of widespread plea bargaining, there is little extra scrutiny. When a defendant pleads guilty in the public court, the details of the plea bargain rarely emerge. Thus, for practical purposes, plea bargaining remains very much in the shadows. This invisibility of plea bargaining makes it appealing to legal professionals who strive for autonomy, certainty, and are habituated into this way of working.

Defendants’ Views and Perspectives A vital area which requires much deeper research is defendants’ views and perspectives about the sentence differential. Most of the research which has been conducted examines the views and practices of defense lawyers, prosecutors, and, to a lesser extent, judges. Yet, we know remarkably little about the perspectives of defendants. Much of what we think we know is, in fact, derived from the views and assumptions of practitioners. The views of defendants may be overlooked on the basis that they are thought to be less informed than those of legal experts. Defendants’ views may also be overlooked because legal experts often claim to speak on behalf of defendants. Even research on issues faced by defendants tends not to interview the defendants whose experiences the research was seeking. To take just one example, Gibbs 222

To Plead or Not to Plead?

(2016) very usefully explores the challenges faced by unrepresented defendants in England and Wales. However, this did not involve interviewing defendants. Instead, the research interviewed prosecutors and other practitioners, who spoke of what they took to be the experience of defendants. This limitation is increasingly notable in light of the growing volume of research on user perceptions of law, legal consciousness, user perspectives in desistance studies of how people move away from offending (e.g., Maruna, 2001; McNeill, 2015; McNeill & Weaver, 2010; Schinkel, 2014; Weaver, 2016). This growing volume of research highlights that the views of defendants cannot be assumed. As Casper (1978) notes: It is the defendant who must most directly live with the consequences of the administration of criminal justice; moreover, given the current concern with crime, it is the defendant’s past and future behavior that is of concern not only to him but also to society at large. Thus, to examine what the defendant thinks is happening to him, the roots of his behavior, and the lessons he learns from his encounter with criminal justice is of importance in understanding the operation and impact of one set of institutions of … government. (p. XI) Whether the aim is to understand potential normative issues or explore new ways to promote efficiency, research directly asking defendants about their experience has significant potential. However, there is some work which has explored defendants’ perspectives. Notably, Casper (1978) set out to investigate the perspectives of defendants in the American criminal process. More recently, Jacobson, Hunter, and Kirby (2016) examined the experiences of those attending court as witnesses, victims, and defendants in the Crown Court of England and Wales. Rather like Casper (1978), they noted a degree of cynicism about what was perceived as professional “game-​playing,” as well as resigned or “passive acceptance.” Swaner et al. (2018) surveyed a sample of those with experience of being proceeded against and interviewed 102 persons “who had significant experience with the police, the courts, and corrections” (Swaner et al., 2018). Although four-​fifths of respondents felt respected by court officials, they were less satisfied with their ability to participate (e.g., ask questions and tell their side of the story and with waiting times). However, far more needs to be understood about defendants’ experiences and data is needed from the defendants themselves to explore the extent to which their perspectives are shared or influenced by practitioners. Most of the research conducted to date examines the views and practices of defense lawyers, prosecutors, and, to a lesser extent, judges.Yet, we know remarkably little about the perspectives of defendants. Much of what we think we know derives from the views and assumptions of practitioners. How is plea decision-​making approached and interpreted by defendants? How does it connect with their life contexts and previous encounters with the justice system? Do defendants consider the sentence differential to be a discount or a penalty? How do they understand the meaning of guilt in the particular context of their case? How do their perceptions about their treatment (including as to how to plead) play into broader views of the justice system and authority? How might a sense of processual fairness assist (or exacerbate) the desistance journey away from offending? For example, it was already suggested that a guilty plea does not necessarily indicate remorse. However, even if this is true, it may be that the defendant nevertheless comes to internalize some aspect of the guilty plea.9

Defendant Perceptions of the Sentence Differential and the Presumption of Innocence Whether the sentence differential violates the presumption of innocence has been the subject of intense debate.Yet, astonishingly little research has been devoted to understanding the experiences and interpretations of those who have to make plea decisions. A recent small research study in Scotland has explored the perceptions of the criminal process of 12 people who had been accused of a criminal offense (Gormley, 2018).10 For these defendants, the baseline sentence (from which they evaluated increases or decreases) was taken as the likely sentence following a guilty plea. From this perspective, pleading not guilty, and being convicted at trial, ran the risk of a higher sentence. 223

Jay Gormley and Cyrus Tata

That defendants considered there to be a risk of a higher sentence is significant as there are different views. For example, Abrams (2011) argued that the correct comparison should be between the sentence if pleading guilty and the sentence if pleading not guilty. This would factor in the realistic chance of a non-​conviction (whether by way of an acquittal, the case being dropped by the prosecution, or the case being dismissed by the court). While those such as Abrams make an argument for this approach, this was not how defendants evaluated the sentence differential in the research in Scotland. That defendants perceived there to be a risk of a higher post-​trial sentence raises key questions about the nature of the sentence differential and the presumption of innocence. Consequently, defendants viewed going to trial as posing a risk of a higher sentence. This perception of going to trial as a risk challenges the claim that the sentence differential does not undermine the presumption of innocence.While policy architects may genuinely hold the view that the sentence differential is a reward or discount, this view was not shared by the defendants themselves. The practical reality of the presumption of innocence is significantly diminished if defendants perceive (even incorrectly) that it does not operate in practice. For example, Jack11 reported that he pled guilty only to avoid a higher post-​trial sentence, rather than seeing it as a way of benefiting from a sentence discount: I just pled guilty because they had my [social media evidence].They would’ve thought I was lying. So, I just pled guilty otherwise I would’ve got more of a sentence. Of course, it is not possible to test the veracity of Jack’s claim of innocence. Yet, it is significant that even in this small sample of defendants there is one who claims he pled guilty, even though innocent, to avoid “more of a sentence.” This raises questions about whether sentence discounting, and other elements of Anglo-American justice systems, might lead those who are innocent of the charges against them (or at least those who genuinely believe themselves to be innocent or are simply uncertain) to plead guilty. Frank felt the sentence after a trial was worse than the sentence following a guilty plea: Because you’re wasting the court’s time, wasting the court’s funds, and fucking them about. “We [the court] are too busy.” And I agree they are too busy and that [the sentence differential] is the shit they have to go to—​the underhand tactics. Indeed, for Frank, plea bargaining was an underhanded tactic. Other interviewees expressed disdain for this apparent gamesmanship, which was felt to undermine the legitimacy of the law and trivialize their case. Another defendant interviewee, while accepting wrongdoing, was unclear as to what he was being charged with and had only a vague understanding of the nature of the charges. Others noted that by the time formal charges were presented they were already in court and had to decide how to plead within minutes. Defendants felt that having only a short time to decide added pressure. Jack also noted that he felt pressured: I would’ve got locked up if I didn’t plead guilty. That’s why I pled guilty. [Interviewer: “Did the Judge tell you that because you pled guilty you didn’t get the jail?”] No, my solicitor. He said that to me. That if I didn’t plead guilty, that if I said, “not guilty,” they’ve got [evidence] … so I would’ve got fucked if I plead not guilty … It felt like I had pressure put on me and that. [Interviewer: “From the court, or the lawyer, or just everything?”] The court. Fucking bullshit like. Interestingly, while interviewees spoke of pleading not guilty as a risk, they also regularly spoke of the “one-​third sentence discount” as a matter of fact. For example, Alan noted the “discount of up to 224

To Plead or Not to Plead?

one-​third” as a reason why it might have been beneficial to plead guilty. That defendants mentioned the sentence differential was significant because the interviews were designed to avoid mention of the sentence differential to see if defendants would identify it themselves without any prompting and, if so, how they would describe it. Defendants did identify the sentence differential, though this was often conflated with other perceived effects of pleading (namely charge bargaining and fact bargaining). Indeed, defendants seemed to care little for the legal distinction between sentence discounting and other overall effects of a guilty plea. It might be thought that where defendants spoke of the sentence differential as a “discount”, they did not tend to feel that the sentence differential infringed on their right to a trial. At first blush, this is a puzzling contradiction. However, upon further questioning, it transpired that defendants were repeating what their lawyers had told them. Interviews with defense lawyers revealed that they advised clients of “sentence discounts” of “up to one-​third” in more straightforward and more certain terms than the formal law suggests.12 Thus, this apparent incongruity between the perception of a longer post-​trial sentence and the use of the term “discount” was less significant than it appeared. Consequently, defendants used the term “sentence discount” unreflectively and repeated what their lawyers had relayed to them. Several defendant interviewees then asked the interviewer if they recalled the law correctly. But, when reflecting on the plea decision-​making process, they did not, in fact, consider it to be a discount. The overall effect of the sentence differential was seen by them to work against the presumption of innocence. Where it had the most significant impact was in cases like Frank’s, where the plea was thought to be the difference between a custodial sentence and a non-​custodial sentence.

The Pains of Being a Defendant13 The pains of being a defendant may be easily overlooked by some in the belief that those with criminal histories become more or less inured to the process costs of the criminal system. This belief that defendants are more or less immune to process costs can result in a more relaxed attitude to the risks of the sentence differential. For example, there seems to be a perception that those who are in and out of prison and are socio-economically disadvantaged have little to lose compared to those who seem to enjoy the prospect of brighter futures. For example, a judge observed: You sometimes think that the whole process of having to stand in the dock, and plead guilty, and be found guilty. It is a big thing to a lot of people … the court process can be quite a deterrent for a lot of these people in itself … I think, if you are a person that leads an otherwise respectable life and you’re not familiar with the courts, the fact of having been charged and going to court can be a significant deterrent. While there may be truth in this, it is potentially problematic from the perspective of equality before the law. As Bandes (2017) has argued, compassion may lead to a “break” for those a legal actor: Instinctively understands, sympathises with, [and] identifies with. In the US and elsewhere, this kind of compassion may be selectively doled out based on racial bias and class bias, unconscious or otherwise. (p. 190) This selectivity in how judges give “breaks” might lead to the “the danger of cultivating selective empathy for those … who live ‘respectable lives’ ” (Bandes, 2016). It is also important to note that how defendants experience the criminal process is very poorly understood, but it does seem that defendants experience events in significantly different ways. In the research in Scotland, those defendants who were less familiar with the criminal process found it unnerving in a way that those 225

Jay Gormley and Cyrus Tata

with more experience did not. However, the interviews did show that even defendants with extensive experience of incarceration and a notable criminal record can experience profound emotions from going through the process and facing the prospect of custody.Thus, while defendants may subjectively experience the process differently, the dynamics are complex and even experienced defendants are far from inured to the process costs of the criminal justice system. While the sentence differential may appear minimal to legal professionals, when combined with other aspects of the process, it can have a significant impact on defendants’. These burdens of the law underline Sarat’s (1990) claim that for those caught up in the system, “the law is all over” as a form of power and domination. The recognition that “… the law is all over” expresses, in spatial terms, the experience of power and domination; resistance involves efforts to avoid further “spatialization” or establish unreachable spaces of personal identity and integrity. (pp. 347–​348)

The Pains of Waiting We typically conceive of plea bargaining as involving intentional efforts to encourage guilty pleas. For example, the State may purposefully implement some systematic “discount” for those who plead guilty. Alternately, judges, prosecutors, and defense lawyers may consciously implement plea bargaining practices themselves from the bottom up. However, Scottish defendants noted that waiting and court process costs were a significant factor that made guilty pleas tempting. Thus, the State offering defendants a way to end the wait with a guilty plea is a powerful dynamic in the reality of plea bargaining. While waiting may seem neutral (it is an absence of decision) even trivial, it posed significant problems for defendants. Some unemployed defendants had to travel far to attend court and they incurred a significant cost. Others had to arrange childcare or miss work. Following this, many were told to attend another day as the court was unable to hear their case for one reason or another. As one defendant lamented: You get told to be there at quarter to ten. And then to come back at two o’clock. You know they are never going to get to you. Look how many people they have to get through. I know it, and they know it. But they still make you come, and you have to wait until they tell you, “we are putting it back.” Why can’t they tell you that in the morning? They know it’s not going to happen today, but you still have to wait for them to tell you to go away. Another commented that one of his biggest criticisms was: Repetition! You’re going to court, and like the police aren’t there?! They should be there! Come on to fuck; they’re the ones that brought it there. All this, “A trial can’t go ahead because PC [Police Constable] Shiny Buttons is in Marbella for a week.” PC Shiny Buttons knew court was coming so he shouldn’t have booked his holiday. He chose that profession. Or they could do it by video link. Indeed, waiting can be especially stressful because of its attendant uncertainty: The wait is devastating because it is associated with uncertainty, doubt, inability to control, constant questioning and confronting one’s fears. It is associated with constantly thinking about what has happened—​magnifying every detail and reaction, every piece of information—​in an attempt to find spaces of control … (Sales 2016, p. 57)

226

To Plead or Not to Plead?

Court observations revealed that judges told defendants who had traveled significant distances that, as part of bail, “you will have to make yourself available to the court at short notice.” In another observed instance, a defendant had been waiting for his case to call since 9.45 am. A court officer informed the defendant that he was to come back after lunch to see if his case would be called at that time. A friend of the defendant (many attended courts alone) pleaded with a court clerk that this was not possible: You don’t get it. He’s an alcoholic. If he goes for lunch, he is not coming back. That’s just not something he can do. After lunch, the case did in fact call and, as the friend had warned, the defendant was not present. Possibly (per the friend’s warning) the defendant, having stayed sober all morning, had missed court due to his dependency issues. Unfortunately, information about the difficulties of this defendant did not (at least at that time) reach the judge. Instead, the judge was informed (in open court by a different clerk who had not heard the friend’s warning) that a phone caller had left a message that the defendant had to go to hospital for an unspecified reason. The caller left a clearly fictitious and humorous name which elicited laughter in the public gallery.14 Regardless, the difficulties of this accused did not reach the judge at this time. Instead, the impression provided to the judge was that the defendant was mocking the court. The judge appeared singularly unimpressed and noted that the court would seek evidence of this hospital visit, which the defendant had better be able to provide. These difficulties that defendants experience are similar to those of other groups, such as those dependent on welfare: Power and domination are … represented in the legal consciousness of the welfare poor in temporal as well as spatial terms; thus, the people I studied often spoke of an interminable waiting that they said marks the welfare experience. In that waiting they are frozen in time as if time itself were frozen; power defines whose time is valued and whose time is valueless. (Sarat, 1990, pp. 347–​348) In the criminal process, professional actors seek to control time. It is the demands of legal actors and the legal system that primarily determine when cases call, not the defendants. In this way, the criminal process almost becomes like a part-​time prison in various respects. Defendants are physically deprived of their freedom in that they must attend court on multiple occasions, and each occasion may consume the better part of a day. More than this, mentally defendants can struggle to leave the court process behind. These deleterious properties of waiting and uncertainty mean that criminal justice systems can be highly taxing on defendants only through their manner of operation (e.g., Cheng et al., 2018). Even simple matters of routine can have consequences for defendants that are hard to predict. Indeed, research on defendants’ perspectives in the US shows: Delays that filing of motions or demanding trial can produce—​place strong pressures on the defendant to get it over with, to cop out and ‘escape’ to prison. (Casper, 1978, p. 16) Even aside from the pains of being held in custody on remand awaiting trial, in themselves these pressures stemming from waiting support Feeley’s (1979) observation the process is the punishment (see also Cheng et  al., 2018). These waiting-​based process costs mean that it is little wonder that some defendants claimed they might plead guilty to end the pains of being a defendant. The effect of waiting in encouraging guilty pleas means that, the lamentable process costs of Anglo-​American justice systems have become a powerful component of plea bargaining.

227

Jay Gormley and Cyrus Tata

Further Research into Defendants’ Views and Perspectives There is a pressing need for research to devote itself to a much more thorough investigation of defendants’ experiences and perceptions about the sentence differential. Why has there been so little research into defendants’ perspectives and experiences? One reason may simply be convenience for researchers and research-​funders. In general, it is much easier to secure interviews with legal professionals than it is with defendants. Some defendants may also live chaotic lives and may be difficult to follow-​up. This chaos can make it difficult for researchers to plan interviews and court case observations. Second, ethical standards of research clearance have to be higher when speaking to defendants who are in a much more vulnerable position than, for example, lawyers.Third, policymakers, professionals and academic researchers can easily tend to be dismissive of defendant perspectives, which are so often punctuated by loss or gaps in memory, marked by addition and poor health, and confusion about the case. It is commonly assumed that there is little point in seeking the views and experiences of defendants who can offer little certain or factual knowledge. Much better, it is assumed, to stick to seeking the views of informed professionals. However, this assumption is problematic given the “very obvious demographic chasm” between typical defendants and legal actors (Tata, 2008, p. 31). Research should not assume that legal actors will fully understand the perspectives of defendants whose backgrounds are usually significantly different to their own. Moreover, for all its limitations, perceptions help to determine decision-​making as well as the sense of the legitimacy and authority of the court. A fourth reason that research on defendants’ perspectives is sparse is that some defense lawyers may be disinclined to allow their clients to speak to researchers out of concern about what might be said, unfairly implying some failing on the part of the lawyer. Despite these challenges, research on defendants should seek to combine defendant interviews with court observations and interviews with defense lawyers and judges. Furthermore, an understanding of how defendants are invited to consider the future consequences of their plea decision-​making should be central to future work. Although academic work tends to divide up the criminal process into discrete phases of “adjudication” (trial), “mitigation,” “sentencing,” and “corrections,” this is not necessarily the experience of defendants. Defendants may have to consider the consequences of being found guilty following a trial and the prospect of being seen as either “a chancer” (i.e., someone who is trying to exploit the system) or “in denial” at sentencing and missing out on possible mitigation (Tata, 2019). Such a defendant may also face being deemed less suitable for community-​based sentences, and if sentenced to prison may find that her earlier denial makes it more difficult to access prison programmes—​often a crucial way to show that she is progressing in the prison system to gain earlier release (e.g., Schinkel, 2014). So, while academic and professional work divides the criminal process into seemingly autonomous individual decision points, the defendant’s experiences of plea decision-​making may be more inter-​linked and holistic. Finally, research on defendants’ perspectives can be challenging to relate to the formal legal process. Defendants may not speak of their case in the same way that legal actors do. Indeed, they may not always fully grasp the charges against them and may have only a vague understanding of the process. This difference between how defendants and legal practitioners communicate means that defendant interviews are challenging to conduct without court observations or interviews with legal actors involved in the case. While these difficulties may pose a challenge, they do not negate the benefit of research on defendant perspectives. One particularly useful area worthy of investigation is that of defendants’ perspectives on the presumption of innocence and the plea decision making process. A better and deeper understanding of defendants’ perspectives could enable more informed and effective policymaking which may encourage the expeditious disposal of cases while not undermining (or mitigating) the risks to cherished rule of law values.

228

To Plead or Not to Plead?

Conclusions This chapter has investigated issues and questions surrounding the sentence differential that is perceived to exist in most Western jurisdictions. The first part examined the philosophical nature of what constitutes a reward for pleading guilty and what constitutes a penalty for going to trial. While there are valid arguments on both sides, and the matter is open to interpretation, it was highlighted that since most defendants plead guilty, it would make sense to consider the post-​guilty plea sentence as the baseline. This raises the question of whether the sentence differential deters defendants from going to trial and thus compromises the presumption of innocence. Part One also scrutinized the criticism that the sentence differential may have a disparate impact on certain groups, such as racial and ethnic minorities. It was argued that a sentence differential that favors guilty pleas (or disadvantages not guilty pleas) means that those pleading not guilty will fare worse. This is problematic since some groups appear less likely to plead guilty. This section concluded that it was likely that the sentence differential contributes to disparate sentencing outcomes. Part Two of the chapter scrutinized why, considering all the criticism and controversy, some justice systems allow the sentence differential to continue and others seek to encourage and bolster it.To this end, three main rationales are discussed: the remorse rationale, the victim rationale, and the efficiency rationale. All three were found to be limited, but each has its appeal. Notably, courts may be impelled to seek signs of remorse, at least in part, to affirm the validity and authority of its own work. Of the three, the efficiency rationale is generally considered dominant in legal discourse. However, this chapter argues that for the sentence differential to be efficient it must produce justice; otherwise, it cannot, no matter how many trials it avoids, be considered efficient. Part Two also questioned whether the sentence differential, and other practices related to securing guilty pleas, actually save resources, or whether there is, in fact, adaptive behavior and a consequential culture of delay. Part Three scrutinized crucial questions regarding defendants’ perspectives of the sentence differential.This is the most critical set of questions facing future research. Defendants have a complex view of the criminal process and their place within the criminal system and society more generally. The two key points of Part Three were that defendants may view the sentence differential as violating the presumption of innocence, but that they may speak of it by repeating the terminology of legal actors who refer to the sentence differential as a reduction/discount. The first point has serious normative implications. The second point has important implications for how future research should explore the perspectives of defendants.

Notes 1 With apologies to The Bard. 2 This chapter very substantially revises and develops parts of an earlier essay by C. Tata & J. Gormley (2016) “Sentencing and Plea Bargaining: Guilty Pleas Versus Trial Verdicts,” in M.  Tonry (ed.), Criminology and Criminal Justice: Criminal Courts and Prosecutors, Oxford Handbooks Online, Oxford University Press. We would like to thank Cassia Spohn for her invaluable comments on an earlier draft of this chapter, as well as Candace McCoy, Mike Nellis, and Julian Roberts. 3 Woolmington v. DPP (1935) famously identifies the principle of the presumption of innocence as “the golden thread” of the entire criminal process. This image of “the golden thread” articulates the sense of its work in holding the legitimacy of the process together (without that thread it falls apart); its golden preciousness; yet also its fragility. 4 The plea-​dependent sentencing differential may have increased since the time of Hood’s study (Sentencing Council of England and Wales 2014 Table 4.2). 5 Some have argued that remorse can make an offender less blameworthy and therefore justify a (slightly but not significantly) lower sentence (see McCoy, 2005, pp. 79–​82). 6 The controversial American phenomenon of “Alford Pleas” (sometimes called a best interest plea) permits defendants to plead guilty while protesting innocence. Bibas (2012: 61–​64) reports that judges dislike Alford Pleas, and indeed equivocal guilty pleas more generally, because of:

229

Jay Gormley and Cyrus Tata the “message which they send. Pleas without confessions leave victims unvindicated and defendants defiant and resistant to treatment … Moreover, some judges and prosecutors worry that equivocal guilty pleas undermine public confidence, leading defendants’ family, friends, and the public to suspect injustice … Guilty-​but-​not-​guilty pleas send mixed messages, breeding public doubt, uncertainty, and lack of respect for the criminal justice system.” Equivocal guilty pleas fail to show to the court that the admission of guilt is sincere and accepted by the defendant (Tata, 2019). 7 Though it has been suggested that, even if changes in criminal workloads were not related, changing civil workloads may have placed pressure on the courts (McCoy, 2005, p. 77). 8 This is not to say Anglo-​American systems have become Inquisitorial as there are still significant divergences from this ideal. Perhaps, this means Anglo-​American systems are somewhere in between and the challenge is to pinpoint where. Alternatively, it may be that these ideal types are losing their relevance to practice all together. If so new conceptual frameworks may be necessary, along with new yardsticks for what is normatively desirable and new touchstones to ground claims of legitimacy. 9 In the case of white-​collar offenders, it has been suggested that “acceptance or rejection of guilt has an impact upon the process of desistance at various stages” (Hunter, 2015 p. 182). 10 The research observed non-​jury court cases in two neighboring intermediate-​level courts and interviewed 17 legal practitioners (sentencing judges, defense lawyers, prosecutors), and 12 defendants about the relationship between guilty pleas and sentencing. 11 Pseudonyms are used to protect the anonymity of interviewees. 12 Gormley (2018). 13 This term alludes to the work of Gresham Sykes (1958), who argued that the mental “pains of imprisonment” may be less visible than physical affliction, and so require our attention: Attacks on the psychological level are less easily seen than a sadistic beating, a pair of shackles on the floor, or the caged man on a treadmill, but the destruction of the psyche is no less fearful than bodily affliction. (p. 64) 14 Why the phone caller (whether the accused or an acquaintance) chose this route instead of explaining the difficulties is unknown.

References Abrams, D.S. (2011). Is pleading really a bargain? Journal of Empirical Legal Studies, 8, no. s1: 200–​221. Alschuler, A.W. (1979). Plea bargaining and its history. Columbia Law Review, 79(1): 1–​43. —​—​. (1983). Implementing the criminal defendant’s right to trial: Alternatives to the plea bargaining system. The University of Chicago Law Review, 50(3): 931–​1050. Anleu, S.L.R. (2009). Law and Social Change. London: Sage. Ashworth, A. (2006). Four threats to the presumption of innocence. The International Journal of Evidence & Proof, 10(4): 241–​279. Ashworth, A., & Roberts, J.V. (2013). The origins and structure of sentencing guidelines in England and Wales.’ In A. Ashworth & J.V. Roberts (eds.), Sentencing Guidelines: Exploring the English Model (pp. 1–​12). Oxford: Oxford University Press. Bagaric, M., & Brebner, J. (2002). The solution to the dilemma presented by the guilty plea discount: the qualified guilty plea—​“I’m pleading guilty only because of the discount …” International Journal of the Sociology of Law, 30(1),  51–​74. Baldwin, J., & McConville, M. (1978). Plea bargaining and plea negotiation in England. Law & Soc’y Rev., 13, 287. Baldwin, R., & Hawkins, K. (1984). Discretionary Justice: Davis Reconsidered. Public Law, Winter: 570–​599. Baranek, P.M., & Ericson, R.V. (1982). The Ordering of Justice: A Study of Accused Persons as Dependants in the Criminal Process. Toronto: University of Toronto Press. Bandes, S.A. (2017). Compassion and the Rule of Law. International Journal of Law in Context, 13(2), 184–​196. —​—​. (2016) “Remorse and criminal justice.” Emotion Review 8(1),  14–​19. Bennett, M.W. (2016). The implicit racial bias in sentencing: The next frontier. Yale LJF, 126, 391. Berdejó, C. (2018). Criminalizing race: Racial disparities in plea-​bargaining. BCL Rev., 59, 1187. Bibas, S. (2004). Plea bargaining outside the shadow of trial. Harvard Law Review, 2463–​2547. Bradshaw, P., Sharp, C., Duff, P., Tata, C., Barry, M., Munro, M., & McCrone, P. (2012). Evaluation of the Reforms to Summary Criminal Legal Assistance and Disclosure. Scottish Government. Casper, J.D. (1978). American Criminal Justice:The Defendant’s Perspective. Englewood Cliffs, NJ: Prentice Hall.

230

To Plead or Not to Plead? Carlen, P., Campbell, C.M., & Wiles, P. (1976). Magistrates’ Justice (pp. 17–​38). London: Martin Robertson & Co. Ltd. Caldwell, H.M. (2011). Coercive plea bargaining:The unrecognized scourge of the justice system. Cath. UL Rev., 61, 63. Chalmers, J., Duff, P.R., Leverick, F., & Melvin,Y. (2007). An Evaluation of the High Court Reforms Arising from the Criminal Procedure (Amendment) (Scotland) Act 2004. Scottish Executive. Cheng, K.K.Y. (2012). Pressures to plead guilty: Factors affecting plea decisions in Hong Kong’s magistrates’ courts. British Journal of Criminology, 53(2), 257–​275. —​—​. (2014). The practice and justifications of plea bargaining by Hong Kong criminal defence lawyers. Asian Journal of Law and Society, 1(2), 395–​412. Cheng, K.K.Y., Chui, W.H., Young, S.N., & Rebecca, O.N.G. (2018). Why do criminal trials “crack”? An  empirical investigation of late guilty pleas in Hong Kong. Asian Journal of Comparative Law, 13(1), 1–​25. Darbyshire, P. (2000). The mischief of plea bargaining and sentencing rewards. Criminal Law Review, 895–​910. Dawes,W., Harvey, P., McIntosh, B., Nunney, F., & Phillips, A. (2011). Attitudes to guilty plea sentence reductions. London: Sentencing Council of England and Wales. Retrieved from: www.sentencingcouncil.org.uk/​wp-​ content/​uploads/​Attitudes_​to_​Guilty_​Plea_​Sentence_​Reductions_​web1.pdf Eisenstein, J., & Jacob, H. (1977). Felony Justice: An Organizational Analysis of Criminal Courts (p. 21). Boston, MA: Little, Brown. Feeley, M. (1979).The process is the punishment. In Crime, Law and Society (pp. 139–​188). Routledge. New York: Russell Sage Foundation. Retrieved from: www.taylorfrancis.com/​books/​e/​9781351570640/​chapters/​ 10.4324/​9781315095288-​8 —​—​. (1982). Plea bargaining and the structure of the criminal courts. Justice System Journal 7(3): 338–​354. Flexner, A. (2001). Is social work a profession? Research on Social Work Practice, 11(2), 152–​165. Frankel, M.S. (1989). Professional codes: why, how, and with what impact? Journal of Business Ethics, 8(2–​3), 109–​115. Frase, R.S. (2012). Theories of proportionality and desert. In The Oxford Handbook of Sentencing and Corrections (p. 131). Oxford: Oxford University Press. Gazal-​Ayal, O. (2005). Partial ban on plea bargains. Cardozo L. Rev., 27, 2295. Gibbs, P. (2016). Justice denied? The experience of unrepresented defendants in the criminal courts. Retrieved from: www.transformjustice.org.uk/​wp-​content/​uploads/​2016/​04/​TJ-​APRIL_​Singles.pdf Gormley, J. (2018). The Nature and Extent of Sentence Discounting. Paper presented to the Annual Symposium of the European Sentencing and Penal Decision-​Making Working Group: Individualisation in Sentencing & Penal Decision-​Making. Leiden University, April 2018. Henham, R. (2000). Reconciling process and policy: sentence discounts in the magistrates’ courts. Criminal Law Review, June, 436–​451. —​—​ . (2017). Sentence Discounts and the Criminal Process. London: Routledge. Heumann, M. (1975). A note on plea bargaining and case pressure. Law & Society Review, 9(3), 515–​528. —​—​ . (1978). Plea Bargaining: The Experiences of Prosecutors, Judges, and Defense Attorneys. Chicago, IL: University of Chicago Press. Hodgson, J., & Roberts, A. (2010). Criminal process and prosecution. In P. Cane & H. Kritzer (eds.), The Oxford Handbook of Empirical Legal Research. Oxford: OUP. Hood, R.G. (1992). Race and Sentencing. Oxford: Clarendon Press. Hood, R.G., & Cordovil, G. (1992). A Question of Judgement: Race and Sentencing: Summary of a Report for the Commission for Racial Equality Race and Sentencing: A Study in the Crown Court. Commission for Racial Equality. Roberts, J., Hough, M., Jacobson, J., Moon, N., & Steel, N., 2009. Public Attitudes to the Principles of Sentencing. Research Report-​6, Sentencing Advisory Panel. [pdf] Retrieved from: http://​eprints.bbk.ac.uk/​3796/​1/​ 3796.pdf Howe, S.W. (2005). The Value of Plea Bargaining. Okla. L. Rev., 58, 599. Hunter, B. (2015). White-​Collar Offenders and Desistance from Crime: Future Selves and the Constancy of Change. London: Routledge. Hutton, N. (2013). The definitive guideline on assault offences: The performance of justice. In Ashworth, A. & Roberts, J.V. (eds.). (2013). Sentencing Guidelines: Exploring the English Model. Oxford: OUP. Jacobson, J., Hunter, G., & Kirby. A (2016). Inside Crown Court: Personal Experiences and Questions of Legitimacy. Bristol: Policy Press. Kemp,V. (2008).Youth Court Research Project: a scoping study adopting a ‘whole-​systems’ approach to the processing of cases in the Youth Courts. Legal Services Research Centre, Legal Services Commission Research Findings 24.

231

Jay Gormley and Cyrus Tata Kutateladze, B.L., & Lawson, V.Z. (2018). Is a plea really a bargain? An analysis of plea and trial dispositions in New York City. Crime & Delinquency, 64(7), 856–​887. Leiter, B. (2010). Legal formalism and legal realism: What is the issue? Legal Theory, 16(2), 111–​133. Leverick, F. (2004). Tensions and balances, costs and rewards: The sentence discount in Scotland. Edinburgh Law Review, 8(3), 360–​388. Lynch, G.E. (2003). Screening versus plea bargaining: Exactly what are we trading off. Stan. L. Rev., 55, 1399. Marshall,T.H. (1939).The recent history of professionalism in relation to social structure and social policy. Canadian Journal of Economics and Political Science/​Revue Canadienne de economiques et science politique, 5(3), 325–​340. Maruna, S. (2001). Making Good: How Ex-​convicts Reform and Rebuild Their Lives (p. 86).Washington, DC: American Psychological Association. Maslen, H., & Roberts, J.V. (2013). Remorse and sentencing: An analysis of sentencing guidelines and sentencing practice. In Ashworth, A. & Roberts, J.V. (eds.). Sentencing Guidelines: Exploring the English Model. Oxford: OUP. Mather, L.M. (1979). Plea Bargaining Or Trial?: The Process of Criminal-​ case Disposition. Lexington, MA: Lexington Books. McBarnet, D. (1981). Magistrates’ courts and the ideology of justice. British Journal of Law and Society, 8(2), 181–​197. McConville, M., Hodgson, J., Bridges, L., & Pavlovic, A. (1994). Standing Accused: The Organisation and Practices of Criminal Defence Lawyers in Britain. Oxford: Clarendon Press. McCoy, C. (1983). Politics and Plea Bargaining: Victims’ Rights in California. Philadelphia: University of Pennsylvania Press. McCoy, C. (2005). Plea bargaining as coercion: The trial penalty and plea bargaining reform. Crim. LQ, 50, 67. McInnes, J. (2004). The Summary Justice Review Committee: Report to Ministers. Edinburgh: Scottish Executive. McNeil, F., & Weaver, B. (2010). Changing Lives? Desistance Research and Offender Management. Scottish Centre for Crime and Justice Research (SCCJR) Project Report No. 03/​2010. Retrieved from: www.sccjr.ac.uk/​wp-​ content/​uploads/​2012/​11/​Report_​2010_​03_​-​_​Changing_​Lives.pdf McNeill, F. (2015). Desistance and criminal justice in Scotland. In Croall, H., Mooney, G., & Munro, M. (eds.), Crime, Justice and Society in Scotland. London: Routledge. Padfield, N. (2012). The right to self-​representation in English criminal law. Revue internationale de droit pénal, 83(3), 357–​375. Page, L., MacLeod, P., Rogers, J., Kniver, A., Iliasov, A., & Tseloni, A. (2010). 2008/​09 Scottish Crime and Justice Survey:Technical Report. Edinburgh: Scottish Government Social Research. Parsons, T. (1939). The professions and social structure. Social forces, 17(4), 457–​467. Pleasence, P., & Quirk, H. (2001). The Criminal Case Profiling Study Final Report. London: Legal Services Research Centre. Retrieved from: hwww.lccsa.org.uk/​assets/​documents/​misc/​criminalcaseprofilingstudy.pdf Provine, D.M. (2002). Sentencing policy and racial justice. In C. Tata & N. Hutton (eds.), Sentencing and Society: International Perspectives (pp. 483–​503). London: Routledge. Rauxloh, R.E. (2010). Formalization of plea bargaining in Germany: Will the new legislation be able to square the circle. Fordham Int’l LJ, 34, 296. Roberts, J.V. (2012). Crime victims, sentencing, and release from prison. In Petersilia, J. & Reitz, K.R. (eds.), The Oxford Handbook of Sentencing and Corrections. (pp. 104–​131). Oxford: Oxford University Press. Rossmanith, K. (2018). Small Wrongs: How We Sorry in Love, Life and Law Melbourne: Hardie & Grant. —​—​. (2015). Affect and the judicial assessment of offenders: Feeling and judging remorse. Body & Society, 21(2), 167–​193. Sah, S., Robertson, C.T., & Baughman, S.B. (2015). Blinding prosecutors to defendants’ race: A policy proposal to reduce unconscious bias in the criminal justice system. Behavioral Science & Policy, 1(2),  69–​76. Sales, P.P. (2016). Psychological Torture: Definition, Evaluation and Measurement. London: Routledge. Sarat, A. (1990). The law is all over: Power, resistance and the legal consciousness of the welfare poor. Yale JL & Human, 2, 343. Savitsky, D. (2009). The Problem with Plea Bargaining: Differential Subjective Decision Making as an Engine of Racial Disparity In The United States Prison System. ProQuest Dissertations and Theses database. (UMI No. AAI3376664). Available from Social Science Premium Collection. (743061981; 201048407). Retrieved from: https://​search.proquest.com/​docview/​743061981?accountid=14116. —​—​. (2012). Is plea bargaining a rational choice? Plea bargaining as an engine of racial stratification and overcrowding in the United States prison system. Rationality and Society, 24(2): 131–​167. Schinkel, M. (2014). Being Imprisoned: Punishment, Adaptation and Desistance. Houndmills: Palgrave MacMillan. Schulhofer, S.J. (1984). Is plea bargaining inevitable? Harvard Law Review, 97(5), 1037–​1107. Sentencing Council. (2015). Crown Court Sentencing Survey Annual Results 2014. England and Wales: Sentencing Council of England and Wales.

232

To Plead or Not to Plead? Shute, S., Hood, R., & Seemungal, F. (2005). A Fair Hearing? Ethnic Minorities in the Criminal Courts. Cullompton: Willan Publishing. Sommerland, H. (2015). The social magic of merit: Diversity, equity, and inclusion in the English and Welsh legal profession. Fordham L. Rev., 83, 2325. Sparks, R. (2011). Divided sympathies: David Hume and contemporary criminology. In Karstedt, S., Loader, I., & Strang, H. (eds.), Emotions, Crime and Justice (317–​330). London: Bloomsbury Publishing. Spohn, C. (2008). Race, sex, and pretrial detention in federal court: Indirect effects and cumulative disadvantage. U. Kan. L. Rev., 57, 879. Swaner, R., Ramdath, C., Martinez, A., Hahn, J., & Walker, S. (2018). What Do Defendants Really Think?: Procedural Justice and Legitimacy in the Criminal Justice System. Center for Court Innovation. Retrieved from: www.courtinnovation.org/​sites/​default/​files/​media/​documents/​2018-​09/​what_​do_​defendants_​really_​ think.pdf Swenson, T. (1995). The German Plea Bargaining Debate. Pace Int’l L. Rev. 7, 373: 400–​404. Sykes, G.M. (1958). The Society of Captives: A Study of a Maximum Security Prison. Princeton, NJ: Princeton University Press, pp. 63–​78. Tait, D. (2002). Sentencing as performance: restoring drama to the courtroom. In C. C. Tata & N. Hutton (eds.), Sentencing and society: International Perspectives. Hampshire: Ashgate, pp. 469–​480. Tata, C. (1997). Conceptions and representations of the sentencing decision process. The Journal of Law and Society, 24(3), 395–​420. —​—​. (2007). In the interests of commerce or clients? Supply, demand and ‘ethical indeterminacy’ The Journal of Law & Society 38, 489–​519. —​—​. (2010). A sense of justice: The role of pre-​sentence reports in the production (and disruption) of guilt and guilty pleas. Punishment & Society, 12(3), 239–​261. —​—​ . (2019). ‘Ritual Individualisation’: Creative Genius at Sentencing, Mitigation and Conviction. The Journal of Law & Society, 46(1), 112–140. Tata, C., & Stephen, F. (2007).When paying the piper gets the wrong tune:The impact of fixed payments on case management, case trajectories and quality in criminal defence work. In P. Pleasence, A., Buck, & N.J. Balmer (eds.),. Transforming Lives: Law and Social Process. London: The Stationery Office. Tata, C., Goriely, T., McCrone, P., Duff, P., Henry, A., Knapp, M., & Lancaster, B. (2004). Does mode of delivery make a difference to criminal case outcomes and clients’ satisfaction? The public defence solicitor experiment. The Criminal Law Review, 50: 120–​136. Tator, C., & Henry, F. (2006). Racial Profiling in Canada: Challenging the Myth of “A Few Bad Apples”. Toronto: University of Toronto Press. Thaman, S. (ed.). (2010). World Plea Bargaining: Consensual Procedures and the Avoidance of the Full Criminal Trial. Durham: Carolina Academic Press. Thomas, C. (2010). Are juries fair? (pp. 1–​65). London: Ministry of Justice. Retrieved from: www.ohrn.nhs.uk/​ resource/​policy/​arejuriesfair.pdf Tonry, M. (2012). Punishment Politics. London: Routledge. —​—​ . (1996). Sentencing Matters. Oxford: Oxford University Press. —​—​ . (2011). Punishing Race: A Continuing American Dilemma. Oxford: Oxford University Press. Tonry, M.H., & Lappi-​Seppälä, T. (2011). Crime and Justice in Scandinavia. Vol. 40, Chicago, IL: University of Chicago Press. Vanstone, M. (2004). Mission control: The origins of a humanitarian service. Probation Journal, 51(1),  34–​47. Vogel, M.E. (2007). Coercion to Compromise: Plea Bargaining, the Courts, and the Making of Political Authority. Oxford: Oxford University Press. von Hirsch, A., & Roberts, J.V. (2004). Legislating sentencing principles: The provisions of the Criminal Justice Act 2003 relating to sentencing purposes and the role of previous convictions. The Criminal Law Review, 50, 639–​652. Ware, J., & Mann, R.E. (2012). How should “acceptance of responsibility” be addressed in sexual offending treatment programs? Aggression and Violent Behavior, 17(4), 279–​288. Weaver, B. (2016). Offending and Desistance:The Importance of Social Relations. London: Routledge. Weaver, S. (2011). Liquid racism and the ambiguity of Ali G. European Journal of Cultural Studies, 14(3), 249–​264. Weigend,T. (2006).Why have a trial when you can have a bargain. In A. Duff., L. Farmer., S. Marshall & V.Tadros (eds.), The Trial on Trial:Volume 2 Judgment and Calling to Account. Portland, OR: Hart. Weisman, R. (2009). Being and doing: The judicial use of remorse to construct character and community. Social & Legal Studies, 18(1),  47–​69. —​—​ . (2014). Showing Remorse: Law and the Social Control of Emotion (Law, Justice and Power). Farnham: Ashgate Publishing Group. —​—​ . (2016). Showing Remorse: Law and the Social Control of Emotion. London: Routledge.

233

Jay Gormley and Cyrus Tata Weitzer, R., & Tuch, S.A. (2002). Perceptions of racial profiling: Race, class, and personal experience. Criminology, 40(2), 435–​456. Welch, K. (2007). Black criminal stereotypes and racial profiling. Journal of Contemporary Criminal Justice, 23(3), 276–​288. Willis, J.E. (1985). The sentencing discount for guilty pleas: Are we paying too much for efficiency? Australian & New Zealand Journal of Criminology, 18(3), 131–​146. Wren, E., & Bartels, L. (2014). Guilty, your honour: Recent legislative developments on the guilty plea discount and an Australian capital territory case study on its operation. Adel. L. Rev., 35, 361.

234

PART IV

Capital Punishment

11 EVOLVING ATTITUDES TOWARD CAPITAL PUNISHMENT Amy L. Anderson, Weng-​Fong Chao, and Philip Schwadel

Amnesty International recently ranked the United States as eighth in executions worldwide (Amnesty International, 2018). While a majority of Americans have approved of capital punishment across decades of public polling research on the issue, a closer examination of trends and subgroups reveals some interesting patterns and avenues for future research. We begin this chapter with a broad review of public opinion polling on support for the death penalty followed by a breakdown of subgroup differences. These patterns lead to a discussion in the following section regarding the effects of crime, racial animus, and the media on death penalty support.

Trends in Support for Capital Punishment Multiple organizations have polled Americans for nearly a century about their attitudes towards capital punishment. As can be seen from Figure 11.1 below, Gallup reported in 1937 that 60% of respondents favored use of the death penalty and 33% did not (Gallup, n.d.). These data show that the only time that the percentage who favored use of the death penalty was smaller than the percentage who did not favor its use was in 1966. Generally, support declined across the 1950s and 1960s until it reached its low in the mid-​1960s. Support then increased over time until it began to decline again beginning in the mid-​to-​late 1990s. Current polling shows that a majority of Americans still support rather than oppose the death penalty, 55% and 41% respectively (Gallup, n.d.), however, these percentages have been fluctuating in the last few years. For example, data from 2016 showed that 49% of individuals favored use of the death penalty and 42% opposed it (Oliphant, 2016), but two years later support had increased to 54% with 39% opposed (Oliphant, 2018), a 5% increase in support in two years. Support for capital punishment is not distributed equally across the United States as there are regional and state-​level differences in attitudes. In particular, support for the death penalty is higher in the South. This is often attributed to a “southern subculture of violence” that would elevate support for the death penalty among southerners relative to individuals from other regions of the country (see Borg, 1997). An analysis by Shirley and Gelman (2015) found that support increased over time among southerners and Midwesterners, while support decreased among individuals living in the Northeast and West. Their preliminary analyses also demonstrated that there was state-​level variation in both mean levels of support and changes in levels of support over time (Shirley & Gelman, 2015; see also

237

Anderson et al. 90% 80% 70% 60% 50% 40% 30% 20% 10%

Favor

Not in Favor

2018

2015

2012

2009

2006

2003

2000

1997

1994

1988

1991

1985

1982

1979

1976

1973

1967

1970

1964

1961

1958

1955

1952

1949

1946

1943

1940

1937

0%

No Opinion

Figure 11.1  Death Penalty Attitudes, Gallup 1937–​2018

Baumgartner et al., 2018). In terms of individuals who live in the south, Borg (1997) found higher support among Southern individuals who were born and raised in the South and who were identified as politically conservative, a religious fundamentalist, and racially intolerant. The level of support for the death penalty is a function of the question being asked. For example, respondent answers will differ if they are asked whether other criminal justice alternatives such as life without parole (LWOP) should be used or if given additional information regarding either the victim or the offender (Baumgartner et al., 2018; Bowers,Vandiver, & Dugan, 1994; Unnever, Cullen, & Roberts, 2005). Baumgartner and colleagues (2018) compiled data from different sources and examined responses to an array of questions about attitudes towards capital punishment. Their examination showed that support for the death penalty if LWOP was an option was about 62% before the year 2000 but then dropped to about 54% after the year 2000 (pp. 272–​273, Table 13.3 column C). Wozniak (2017), however, found that individuals will support the death penalty over alternatives like LWOP if they do not perceive prison to be harsh enough. This finding is consistent with results from other studies that show support for capital punishment is higher among individuals who distrust the government to handle crime problems (Messner, Baumer, & Rosenfeld, 2006; Soss, Langbein, & Metelko, 2003; Unnever, Cullen, & Bartkowski, 2006). Aside from alternative options, there are questions about the magnitude of support for capital punishment. Unnever and his colleagues (2005) sought to examine the strength of support for and against the death penalty using two sources of data from the year 2000. First, using data from the National Election Study that asked respondents if their death penalty support was weak or strong, they found that 54.6% reported their support for capital punishment was strong although a total of 72.8% expressed support rather than opposition. Similarly, they used data from the Gallup survey and found that 64% of respondents supported the death penalty, but 54% of those individuals had reservations at its use; only 34.5% of all respondents supported the death penalty without any reservations (Unnever, Cullen, & Roberts, 2005). This evidence suggests that the majority of Americans support the use of capital punishment when the question asks about first-​degree murder and the response categories are in a yes, no, unsure or do not know format, but further assessment indicates that support is soft and nuanced among those who report being in favor of capital punishment. 238

Evolving Attitudes

Overall, national trends indicate that while support for capital punishment has declined since its high in the mid-​1990s, it has increased slightly in recent years. Additionally, when alternative options to the death penalty are considered, support for capital punishment appears to be trending towards a minority position in the United States. National-​and state-​level attitudes regarding capital punishment, however, mask social and demographic differences as well as subgroup differences, such as whether support for the death penalty differs for Republicans relative to those with other political party affiliations. We discuss group differences further in the next part.

Subgroup Attitudes towards Capital Punishment Generally, researchers find that support for the death penalty is highest among whites, males, religious fundamentalists, and conservatives (Baumer, Messner, & Rosenfeld, 2003). These groups also tend to hold more punitive attitudes than other groups (Ramirez, 2013a) and death penalty support is associated with broader indicators of punitiveness (Baumgartner et al., 2018; Enns, 2014; 2016). We discuss each of these associations below and then connect them to explanations for support. Those explanations provide avenues for future research, which we present later in this chapter. We begin, however, with a discussion about the mixed findings regarding the relationship between age and attitudes towards capital punishment as more research is needed in this area. There is some evidence that younger individuals are less likely to support the death penalty than middle-​aged individuals (Anderson et al., 2017; Oliphant, 2018; Miller & Nakamura, 1997; Young, 2004). Additionally, while there may be an expectation that elderly individuals are more likely to support the death penalty than other age groups, such as proposed by the aging-​conservatism hypothesis (Danigelis, & Cutler, 1991), most researchers have found no relationship between age and support for the death penalty (Baumer et al., 2003; Cochran, Boots, & Heide, 2003; Unnever & Cullen, 2006, 2010). The lack of a relationship between age and death penalty support was also reflected in recent data from Pew Research Center, with 56% of respondents aged 30–​49, 55% of 50–​64 year olds, and 57% of individuals aged 65 and over in favor of the death penalty (Oliphant, 2018). Individuals aged 18–​29 were evenly split, with 47% favoring and 46% opposing the death penalty. The finding of no relationship between age and death penalty support may be a result of the collinearity between the respondents’ age, the time period or year of the survey, and the birth cohort of the respondent (i.e., birth cohort = time period—​age, discussed in more detail later in this chapter). Researchers who accounted for this collinearity using hierarchical cross-​classified models found that young adults and older adults were less likely to support the death penalty, with support peaking among middle-​ aged individuals (Anderson, et al., 2017; Shirley & Gelman, 2015). As such, more research is needed regarding the relationship between age and capital punishment attitudes. While the relationship between respondent age and support for capital punishment varies across studies, researchers generally find that respondent sex, religious and political affiliation, and race are strongly related to attitudes towards capital punishment. Current figures indicate that approximately 61% of men and 46% of women favor the death penalty for persons convicted of murder while 34% and 45% respectively oppose it (Oliphant, 2018). Researchers consistently find a significant relationship between sex and death penalty support (Applegate, Cullen, & Fisher, 2002; Baumer et al., 2003; Boots & Cochran, 2011; Cochran & Sanders, 2009; Ramirez, 2013b). Typically, cross-​sectional survey data shows that men favor the death penalty over women by about 10 percentage points on average (Ramirez, 2013b). Additionally, research findings and polling show that religious fundamentalists, especially evangelical Christians, are significantly more likely to support the death penalty (Anderson et al., 2017; Borg, 1997) than other types of religious and non-​religious people. Current polling data indicates that 73% of white evangelical Protestants favor the death penalty and 19% oppose it. This compares to 61% of white mainline Protestants who favor and 30% who oppose use of the death penalty, and 53% of Catholics (57% white Catholics) who favor it and 42% (35% of white Catholics) oppose it (Oliphant, 239

Anderson et al.

2018). The relatively lower support among Catholics (20  percentage points) is due in part to the Catholic Church’s formal denunciation of the death penalty, which has translated to its members becoming less supportive over time. Interestingly, Catholic support for capital punishment was higher than evangelical support in the 1970s and it was not until the 1980s that evangelical support for capital punishment was higher than Catholic support (Anderson et al., 2017). While more religiously unaffiliated individuals favor use of the death penalty (48%) than oppose it (45%), the difference is not notable (Oliphant, 2018). Since at least the late 1960s, Republicans have considered themselves the party of “law and order” and “tough on crime” (Beckett, 1997; Beckett & Sasson, 2003; Ramirez, 2013a; Warr, 1995; Welch, 2007). As a result, it is not surprising that macro-​and micro-​level indicators of a conservative political ideology, such as a Republican political affiliation or a context with a conservative political climate, are associated with higher levels of support for capital punishment (Baumer et al., 2003; Brown & Socia, 2017; Ramirez, 2013b; Unnever & Cullen, 2006). The importance of political affiliation is underscored by recent Pew Research Center data, which showed that among all the subgroups related to sex, race, age, education, and political and religious affiliation, it was Republicans who had the highest level of support (77% favor, 17% oppose; Oliphant, 2018). A person’s race is another strong predictor of support for capital punishment. Whites are significantly more likely to support the death penalty than individuals from other racial and ethnic groups (Barkan, & Cohn, 1994; Shirley & Gelman, 2015; Ramirez, 2013b; Rankin, 1979; Unnever & Cullen, 2010). Current polling data show that while 59% of whites favor the use of the death penalty, only 47% of Hispanics and 36% of blacks favor its use (Oliphant, 2018). Long-​term trend data show that blacks have had relatively lower levels of support across time than other racial groups (Anderson et al., 2017; Shirley & Gelman, 2015; Ramirez, 2013b). It generally has been the case that when support for capital punishment increased or decreased, it tended to do so for all subgroups within a given attribute. Recent research, however, identified divergent trends among the subgroups of several demographic characteristics. Two examples from Anderson, Lytle, and Schwadel (2017) are that the level of support for capital punishment began to diverge between men and women for individuals who comprise the post-​1970s birth cohorts and began to diverge since about the year 2000 for Republicans compared to non-​Republicans. We return to recent subgroup trends in attitudes towards capital punishment later in this chapter.

Explanations for Death Penalty Support As the data presented above demonstrate, nationwide trends in capital punishment attitudes as well as subgroup differences are not static over time. In this part, we focus our discussion on three inter-​ related factors that are shown to affect attitudes towards capital punishment: crime, racial animus, and media consumption. We focus on these three because they have strong relationships with attitudes towards capital punishment, are related to the demographic trends described above, and also lead to directions for future research regarding attitudes towards capital punishment.

Crime Perhaps the most obvious reason to support the use of capital punishment is crime. Indeed, support for the death penalty increased as the crime rate increased in the 1960s (Gross & Ellsworth, 1994; Warr, 1995). Crime rates may affect attitudes towards capital punishment by increasing an individual’s support of punitive approaches to handle offenders, perhaps through a desire to deter future offenders but generally to punish already adjudicated offenders. Vollum, Mallicoat, and Buffington-​Vollum (2009), for instance, argued that when people are distressed about crime, they will prefer more punitive policies out of a desire for justice or retribution. In this way, death penalty support is positively associated with concern that formal responses to crime are not adequate (Garland, 2001; Simon, 2007). 240

Evolving Attitudes

Scholars have found mixed evidence for a relationship between crime rates and support for the death penalty (Baranauska & Drakulich, 2018; Baumer et al., 2003; Carll, 2017; Soss, Langbein, & Metelko, 2003). In one study, Baranauskas and Drakulich (2018) found that county-​level violent crime rates were related to death penalty support. In another study, Baumgartner and his colleagues found long-​term but not short-​term effects of homicide rates on death penalty support (Baumgartner, De Boef, & Boydstun, 2008). On the other hand, Messner and his colleagues (2006) did not find a relationship between homicide rates and death penalty support. Similarly, Kleck and Jackson (2017) did not find a significant relationship between rates of murder, robbery, assault and burglary, and death penalty support. Separate from the effect of crime rates on support for the death penalty, there is evidence that the perception that crime is going up increases support for the death penalty (Kleck, & Jackson, 2017; Kort-​Butler & Ray, 2018). Researchers also have found little evidence that personal victimization increases support for the death penalty (Applegate et al., 2000; Borg, 1998; Kort-​Butler & Hartshorn, 2011; Tyler & Weber, 1982; Unnever, Cullen, & Fisher, 2007; Wu, Sun, & Wu, 2011). Kleck and Jackson (2017) examined vicarious victimization along with several different personal victimization and fear of crime variables and found that none of them predicted support for the death penalty. In fact, the only variable that reached significance was the perception that the crime rate was increasing. Most studies of the effects of macro-​or micro-​measures of crime on capital punishment attitudes are cross-​sectional or involve only a few time points. A dynamic examination of whether the crime rate affects an individual’s support for the death penalty, however, necessitates the use of multilevel longitudinal data (e.g., Miller & Nakamura, 1997; Shirley & Gelman, 2015). A recent study used hierarchical age-​period-​cohort modeling techniques, about four decades of General Social Survey data, and included common individual-​level variables as well as five period-​level predictors, one of which was crime rates (Anderson et al., 2017). There are two findings worth highlighting here. First, there was evidence of variation across periods and cohorts in death penalty support, although the cohort-​ level variation was small compared to the period-​level variation. Cohort effects tend to be developed during formative years and are somewhat stable over the life-​course, while period effects are related to short-​term, temporal effects. This distinction is important for the second relevant finding, which was that more than 70% of the significant period-​level variation in death penalty attitudes was explained by the Uniform Crime Reports (UCR) overall violent crime rate. The interesting thing about the effect of the UCR violent crime rate was that this measure explained more period-​level variance than either the UCR homicide rate alone or the National Crime Victimization Survey violent crime rate (Anderson, Lytle, & Schwadel, 2017). This pattern of findings suggests that the UCR was a better predictor of death penalty attitudes when the trends in the UCR and NCVS data have diverged. The above-​stated argument that increasing crime rates, real or perceptual, lead to increased support for capital punishment may reflect some underlying hostilities related to racial animus. For example, researchers argue that overt racism changed during the civil rights movement such that language and policy positions around crime espoused especially by Republican politicians, such as “law and order,” became the racially coded substitute (Beckett and Sasson, 2003; Baumer, Messner and Rosenfeld, 2003; Chiricos, Welch, & Gertz, 2004; Ramirez, 2013a). Consequently, researchers have argued that support for capital punishment by whites represents a modern form of racism manifested as racial resentment (Kinder & Sanders, 1996; Unnever & Cullen, 2007a). In support of this position, Unnever and Cullen (2007a) found that the relationship between crime and death penalty support became insignificant after controlling for measures of racism.We discuss racial animus further in the next part.

Racial Animus Racial stratification in the United States can create racial biases among whites, which may translate into support for punitive policies like capital punishment (Soss et al., 2003; Unnever & Cullen, 241

Anderson et al.

2007a). A  consistent body of literature suggests that white racism is important for explaining the significant differences in support for the death penalty among white and black individuals (Bobo & Johnson, 2004; Peffley & Hurwitz, 2010; Peffley, Hurwitz, & Mondak, 2017; Unnever & Cullen, 2010, 2012; Unnever et al., 2008). Three ways of describing racial animus that are associated with death penalty support include the racial threat hypothesis, racial stereotypes, and racial resentment or symbolic racism (see Unnever & Cullen, 2007). We discuss these three forms further below. A seminal discussion by Blumer (1958) titled “Race Prejudice as a Sense of Group Position” serves as the foundation for some of the research on the relationship between race and support for the death penalty. At the time, racial prejudice was considered to be a micro-​level or individual characteristic. His paper laid out a macro-​level, sociological explanation for how prejudice can be embedded in racial or ethnic groups. Specifically, he advanced the idea that the “dominant” racial group is able to portray though definitions the “subordinate” racial group. This happens through communication and interaction that creates a sense of group position (e.g., us vs. “others”) and also through the use of an abstract image of the subordinate racial group, which creates a collective racial image. He argued that prejudice can become entrenched when the “social order is rooted in the sense of social position” (Blumer, 1958: 7) and pointed to the South as an example of this sort of social order and race relations. Finally, he suggested that the level of prejudice among the dominant group could be affected, for instance, by the positive or negative way societal events are defined and treated by leaders in society. Blumer’s arguments are relevant not only for researchers interested in racial prejudice but also in the context of current events, as we discuss later. Related to Blumer’s theory, researchers interested in racial threat have proposed that perceived threats to the dominant group by the minority group would increase prejudice, such as when the size of a minority group increases relative to the dominate group (Blalock, 1967; Quillian, 1995). This is because of a perception rooted in conflict theory that an increasing minority group poses a threat to the finite social resources available to the dominant group (King & Wheelock, 2007). If more prejudice is associated with a desire for formal control and punishment, then a growing minority population would be associated with higher levels of support for capital punishment among whites. In support of this hypothesis, Soss and his colleagues (2003) found that the effect of racial prejudice on support for the death penalty among whites increased as the percentage of country residents who were black increased. King and Wheelock (2007) found a similar effect; punitive attitudes among white respondents increased as the size of the black population increased, which was mediated by the perception of blacks as an economic threat. Similarly, Lehmann and Pickett (2017) found that the relationship between economic insecurity and support for the death penalty was strongest for white men. Second, aside from the dominate group feeling threatened by an increasing minority population, researchers have also examined the effect of racial stereotypes on attitudes towards capital punishment. Stereotypes that link racial minorities to crime such as through the identification of criminal others or the racial typification of crime serve to validate the racial animus of prejudiced and racist individuals (Chiricos, Welch, & Gertz, 2004; Stewart et al., 2018; Unnever & Cullen, 2010). Perceptions may be influenced by criminal justice practices, such as through disproportionate executions of black offenders with white victims than of white offenders with black victims (Baumgartner et al., 2018; Baumgartner, Grigg, & Mastro, 2015). These types of disparate criminal justice system practices can affect racial stereotypes, which then influence attitudes toward the death penalty (Bobo & Johnson, 2004; Matsueda & Drakulich, 2009; Peffley & Hurwitz, 2007, 2010; Peffley et  al., 2017). Among already prejudiced individuals who desire punishment, the perception of criminality among minority group members would lead to support for capital punishment (Barkan & Cohn, 1994; Dambrun, 2007; Unnever & Cullen, 2007a). While evidence suggests there is a positive relationship between the racial typification of crime and support for the death penalty (Chiricos, Welch, & Gertz, 2004; Unnever & Cullen, 2012), there is also evidence that this effect was more pronounced around 1990 and but was smaller by the year 2000 (Brown & Socia, 2017; Soss et al., 2003; Unnever & Cullen, 242

Evolving Attitudes

2012). Based on this finding, Unnever and Cullen (2012) posited that support for capital punishment may only increase when crime committed by blacks and Hispanics is perceived as being intolerable or abnormal. According to Blumer (1958), the dominant group can contribute to and reinforce these perceptions. Finally, racial resentment arises in part from processes that facilitate the racial stereotypes, specifically the language of Conservative politicians that re-​framed the crime problem in a way that invoked racial images and violence and law with order language as the solution. This language, in combination with racially-​charged language linked to Democrats and welfare policy, has become a so-​called “dog whistle” that affects societal levels of racial prejudice (Beckett & Sasson, 2003), and serves to increase support for capital punishment and other punitive crime control measures (Baumer, Messner, & Rosenfeld, 2003; Unnever & Cullen, 2010). Racial resentment or symbolic racism has become a new generation of racism, as claims of racial inferiority have given way to more subtle claims of problematic communities and minority group members not willing to help themselves (Kinder & Sanders, 1996; Unnever & Cullen, 2007a; Unnever et al., 2008). As a result, white racists are resentful towards groups that they perceive as crime-​prone, violent, and dangerous, while at the same time being given special advantages by the government and not doing enough for themselves (Unnever & Cullen, 2007a); or stated differently, “the undeserving nature of an irresponsible group” (Drakulich, 2015, p. 411). Overall, there is evidence that racial resentment has a significant effect on white support for the death penalty (Brown & Socia, 2017; Soss et  al., 2003; Unnever & Cullen, 2007a, 2007b, 2012). Drakulich (2015) found that both explicit and implicit racial biases affected three distinct measures reflecting respondents’ framing of the crime problem for whites but not for African-​Americans or Hispanics. Unnever and Cullen (2007a) constructed racial stereotype and racism scales and found that white racism accounted for one-​third of the difference between whites and African-​Americans in death penalty support. Among whites, they found that 63% of racist whites supported the death penalty compared with 43% of non-​racist whites, which highlights the role of racial resentment on attitudes towards capital punishment. In sum, although white racism may be conceptualized in different ways, the research suggests that racism significantly affects white support for capital punishment. Racial prejudice among whites increases when the size of minority populations increases, particularly the size of the black population (Soss et al., 2003). A belief in the stereotype of dangerous minorities also increases support for capital punishment (Chiricos, Welch, & Gertz, 2004). Finally, consistent with ideas about racial resentment and symbolic racism, white individuals are more likely to support the death penalty when they perceive black individuals to be violent criminals despite the special societal accommodations they have received (Unnever & Cullen, 2007a).

Media Ideas about both crime and race are socially constructed and amplified through media accounts (Beckett & Sasson, 2003; Chiricos,Welch, & Gertz, 2004), such as through news and crime shows (see Baranauskas & Drakulich, 2018). Displays of racial imagery that increase prejudice and also portrayals of crime or criminals that increase the desire or acceptance of punitive crime control policies may serve to increase support for capital punishment. Media content can also affect viewers by influencing perceptions of the criminal justice systems’ ability to effectively handle crime, such that individuals who do not have faith in the system would be more likely to support capital punishment (Kort-​ Butler & Hartshorn, 2011; Kort-​Butler & Ray, 2018). Media stories serve as a main source of an individual’s ideas about crime, particularly when individuals do not have first-​hand exposure to crime (Pickett et al., 2015). This can be problematic if, as researchers have found, the typical story is consistent with stereotypes (Gruenewald, Chermak, & Pizarro, 2013), is a distorted view of offenders and victims (Chermak, 1994; Pollak & Kubrin, 2007), or puts forward consistently supportive views of 243

Anderson et al.

punitive crime control policies (Kort-​Butler & Habecker, 2018; Pickett et al., 2015; see also Gerbner & Gross, 1976). There is some evidence to suggest that death penalty attitudes may be affected by media consumption. Britto and Noga-​Styron (2014), for example, examined the relationship between the consumption of different media sources, broadly defined, and support for the death penalty. They found that television and radio consumption were significantly related to death penalty support while newspaper and Internet consumption were not. Television viewing is often sorted into categories intended to measure a respondent’s exposure to shows pertaining to crime or the criminal justice system in addition to the news. Britto and Noga-​Styron (2014) examined the specific categories of watching police, news, and crime drama television, and found a positive and significant relationship between each of the three and support for the death penalty. Kort-​Butler and Hartshorn (2011) found that the number of days spent watching TV crime dramas significantly increased support for the death penalty but the number of days spent watching the local news, national news, and nonfiction crime dramas did not affect attitudes toward capital punishment. On the other hand, in an examination of the strength of attitudes towards capital punishment, Unnever, Cullen and Roberts (2005) found that individuals who watched or paid attention to local or national news were more likely to show strong support for the death penalty. Baranauskas and Drakulich (2018) examined the association between a respondent’s primary source for news and support for the death penalty and found that the number of days spent watching television news was positively associated with death penalty support while the number of days of listening to radio news or reading newspaper or online news was not significantly related to death penalty support. Overall, evidence suggests that watching television, especially news shows and crime dramas, increases perceptions of rising crime and support for punitive crime control policies such as the death penalty. There are two recent articles with findings worth highlighting here regarding the effects of news viewership. First, the aforementioned Baranauskas and Drakulich (2018) study found that local TV news viewership was significantly associated with multiple measures reflecting perceptions of crime as well as support for the death penalty, but that national news viewership was not. Further, and consistent with the above discussion pertaining to racial animus and support for the death penalty, they found that racial context mattered. In particular, and consistent with the minority threat hypothesis, they found that the effect of watching crime depictions on TV increased perceptions of crime among whites who lived near greater numbers of African-​Americans. These depictions mattered less, however, for respondents who were not near areas with a greater concentration of African-​Americans. Second, Kort-​Butler and Ray (2018) examined support for the death penalty using data from a yearly omnibus survey conducted in Nebraska.They used two cross-​sectional surveys, 2015 and 2016, the years in which the state was debating the legality of the death penalty and resulted in a vote to reinstate it in November of 2016. This time period also covered a number of high-​profile incidents related to Nebraska’s criminal justice system, including two inmates who died during a prison riot (see Kort-​Butler & Ray, 2018 for further discussion). They found that death penalty supporters were the most likely group to believe the US crime rate was increasing and were the angriest about crime. They also examined individuals who responded that they were unsure of their death penalty attitudes. Interestingly, they found that this group was more engaged with the local TV news in 2016 (when the discussion around the death penalty was ongoing) than those who favored or opposed it. Based on the pattern of their results, they argued that the local criminal justice problems may have persuaded some of the previously unsure respondents that were also more likely to distrust the government to support the death penalty in the 2016 Nebraska election.These findings, along with those of Baranauskas and Drakulich (2018) and others (Chiricos, Padgett, & Gertz, 2000; Roche et  al., 2016), highlight the importance of the local news context on perceptions of crime and attitudes towards capital punishment. A few studies have examined the relationship between media consumption and support for the death penalty by examining how support changes when respondents are exposed to additional 244

Evolving Attitudes

criminal justice options such as LWOP. Niven (2002) assigned subjects to three conditions that all included reading articles prior to them responding to a set of questions about the death penalty. One group represented “business as usual” and read an article that portrayed strong majority support for the death penalty. A  second group was presented with a portrayal that included a mix of options including LWOP+R (life without parole plus restitution). The third group was a control group that read an article about airport renovations. He found that when subjects were exposed to the mixed set of preferences beyond capital punishment that they were significantly less likely to support that death penalty than the control group. Further, there was no significant effect for the typical death penalty portrayal (i.e., business as usual). Similarly, Britto and Noga-​ Styron (2014) found that television and talk radio consumption predicted support for capital punishment, however, when respondents were asked if they would not support the death penalty if the offender would stay in prison for life, only television consumption continued to predict support for capital punishment. The types of television shows that predicted support for capital punishment included crime dramas, news programs and police-​reality programs. Interestingly, when the LWOP option was included, news programs were the only type of television consumption that predicted support for capital punishment. We return to the importance of news programs in the next part.

Directions for Future Research To this point, we have focused our discussion of attitudes towards capital punishment on particular sociodemographic characteristics and explanations because they lead to avenues for future research. We also refer readers to a recent book by Baumgartner and his colleagues (2018) that provides a great deal of detailed, statistical information regarding many aspects of the death penalty. Additionally, a review by Pickett (2019) provides a broad discussion regarding public opinion and criminal justice policies, including capital punishment. We believe it is important to reiterate the oft-​noted methodological point that more research needs to be conducted using longitudinal data and dynamic modeling techniques. Much of the research that examines attitudes towards capital punishment takes a cross-​sectional approach (see Pickett, 2019). Researchers interested in using repeated cross-​sectional surveys to model behavior such as evolving attitudes towards capital punishment should employ methods to adjust for the collinearity between age, time period, and birth cohort (see Yang & Land, 2016; Schwadel, 2010).This is especially important for researchers interested in life-​course effects on punitive criminal justice attitudes. The usefulness of hierarchical age-​period-​cohort (HAPC) modeling to the study of death penalty attitudes can be illustrated by recent findings of significant cohort and period-​level variation in death penalty attitudes (Anderson et al., 2017). Also relevant to our interests in this chapter, they found significant variation in death penalty support across time periods for Republicans and blacks, and that over 70% of the variation in individual support death penalty support across time periods was attributable to the overall UCR violent crime rate. These multilevel, dynamic findings were produced using HAPC modeling techniques, and more research is needed to further examine potential cohort and period effects, such as the punitive tone of presidential statements on punishment (see Ramirez, 2013a). Further, more research is needed to explain subgroup variation in attitudes towards capital punishment. For example, researchers could determine why differences between whites and blacks vary across time periods. The political and cultural context, such as the party of the President or the period-​specific prevalence of and discourse around racism, could play a role in black–​white differences in support for the death penalty. These are but a few examples of the many research questions that now can be addressed due to advances in the modeling of repeated cross-​sectional survey data. Additionally, long-​term data need to be collected on attitudes toward capital punishment with alternative criminal justice responses to murder also presented. This is challenging for researchers 245

Anderson et al.

as we often rely on the data collected by national polling companies. Gallup, for example, does ask about alternative criminal justice options, however techniques like HAPC require many decades of repeated, cross-​sectional surveys. There either needs to be a long-​term data collection effort on behalf of researchers interested in public opinion concerning capital punishment or work needs done to convince polling companies to add at least one follow-​up question that measures the strength or weakness of support or opposition. The public is exposed to favor and oppose percentages without having any sense of strength, and societal attitudes could shift if people were exposed to a wider range of both criminal justice and survey response options. Additionally, repeated cross-​sectional data that spans an extended period of time with respondent answers to questions with criminal justice alternatives like LWOP and LWOP + R are necessary in order to examine whether support for alternatives significantly varies over time. For instance, support for alternative sentences could be affected by levels of trust in the government or perceptions of confinement to prison. Another avenue that researchers should pursue is whether recent trends in death penalty attitudes reflect broader trends regarding political polarization.The most robust finding from recent research on the demographic correlates of support for capital punishment is that Republicans and Democrats are growing farther apart on the issue (Anderson et al., 2017, supplemental analyses not shown but available upon request). Consistent with this, Pew Research Center data show that 87% of Republicans and 71% of Democrats favored the death penalty for persons convicted of murder in 1996, however, by 2018, 77% of Republicans and 35% of Democrats supported capital punishment (Oliphant, 2018). In other words, Republican support dropped by 10  percentage points while Democratic support dropped by 36 percentage points over the course of 22 years. This widening political gap in support for the death penalty comports with a larger body of research on political polarization in the United States. Near the close of the twentieth century, DiMaggio and his colleagues (1996: 738) identified a “striking divergence of attitudes between Democrats and Republicans.” These party differences are most apparent when it comes to issues of sexuality and abortion (Bafumi & Shapiro, 2009; Ura & Ellis, 2012) but are also evident when it comes to views of the death penalty and of punitiveness and punishment for crime more broadly (Davis, 2018; Ramirez, 2013a). From the 1950s to the 1970s, the trend was toward party dealignment. Notable numbers of Republicans expressed liberal viewpoints on some issues and many Democrats voiced conservative perspectives on some issues. By the close of the twentieth century, and now in the first two decades of the twenty-​first century, the trend has reversed itself. Democrats and Republicans have grown father apart on a variety of issues (Abramowitz, 2010). The contemporary American voter is highly ideological, and there is now a far more robust connection between partisanship and ideology (Bafumi & Shapiro, 2009). Consequently, while notable numbers of Americans claim Independent political status, the remaining members of the population have largely sorted themselves into the Democratic and Republican parties over the last few decades (Levendusky, 2009). Empirical research suggests that key factors motivating this “partisan sort” include gendered issue preferences, the southern realignment—​the transition from the South being disproportionately Democrat to disproportionately Republican—​and, relatedly, the growing connection between religious and political affiliations (Gillion, Ladd, & Meredith, forthcoming; Lang & Pearson-​Merkowitz, 2015; Schwadel, 2017). Given the high degree of partisanship in contemporary opinions on capital punishment, it is possible that views of crime and punishment also helped contribute to the partisan sort over the last few decades. Not only have Americans sorted themselves into political parties based on issue preferences, but we are now far more likely to be biased against and feel anger toward people from the other major party (Mason, 2015). As noted above, the Republican Party has cast itself as the party of “law and order” and being “tough on crime” (Beckett, 1997; Enns, 2016). Conversely, the Democratic Party pitches itself as the party that cares about the interests of minority Americans (Grossmann & Hopkins, 2015), and thus attracts people who raise questions about the equal application of the law (Young, 1991), especially in interracial offender-​victim cases (Baumgartner, Gigg, & Mastro 2015; 246

Evolving Attitudes

Keil & Vito 1989; Paternoster 1984; Sorensen & Wallace 1999). Given the strongly stated positions by the major political parties, we expect that along with views of sex and gender, abortion, and the environment, views of punitiveness and of the death penalty in particular may have contributed to the substantial partisan sorting over the last few decades. Still, empirical research is needed to address how views of capital punishment have contributed to partisan sorting; and how partisan sorting may have contributed to views of capital punishment since party affiliation can influence political perspectives (Hart & Nisbet 2012). Longitudinal data are required to differentiate the effects of party affiliation on views of capital punishment from the effects of views of capital punishment on party identification. Future research should also address the social and demographic mechanisms that lead Republicans to be relatively favorable towards the death penalty. Republicans and Democrats differ in their regional concentration (Black & Black, 20013), who they interact with (Huckfeldt & Sprague, 1995), and where (and if) they go to church (Schwadel, 2017). These or other factors may explain the robust connection between party affiliation and views of capital punishment in the United States. Finally, the intersection of political polarization, media accounts and portrayals of crime, and attitudes towards capital punishment warrants further research. Generally, more research is needed into the types and content of media that affect death penalty attitudes. We discussed earlier that research consistently finds television to be the media source related to support for the death penalty, but current data are needed that include social media sources like Facebook and Twitter. Even many recently published studies that examined Internet effects (e.g., Baranauskas & Drakulich, 2018; Roche et al., 2016) used data that are dated relative to the rise of social media platforms as a source of news. Current data collected by the Pew Research Center showed that only 32% of adults reported never getting news from social media (Matsa, & Shearer, 2018). Although the same question was not asked in 2012, this compares with only 49% who reported seeing any news on social media just six years earlier (Gottfried & Shearer, 2016). The importance and potential effects of people getting their news from social media is underscored by the 2016 election for President of the United States, when the average American viewed at least one “fake news” story via social media (see Allcott & Gentzkow, 2017). Generally, the content of national and local news stories as well as the context of the viewers watching those news programs also warrants more attention. Researchers need to examine whether specific news sources affect support for the death penalty through stories and commentary that touch on factors shown to affect death penalty support. For example, experimental findings that used manipulated newspaper articles to portray death penalty attitudes indicate that levels of support for the death penalty decline when it is presented as a minority position and in a context with alternative sentencing options (Niven, 2002). Unnever, Cullen, and Roberts (2005) found that Americans who paid attention to and watched the local and national news were more likely to strongly support the death penalty rather than weakly support it. These findings and others (e.g., Weitzer & Kubrin, 2004) suggest that more research is needed to determine the degree to which local or national news is stoking racial animus through stories about “others” or engaging in the racial typification of crime, thereby affecting death penalty attitudes. Relatedly, researchers should examine the influence of partisan news media on evolving attitudes towards capital punishment. There is evidence that political party affiliation may interact with news type: as Roche and his colleagues (2016) found, Internet news was negatively associated with death penalty support but only among respondents who were not conservative. To the degree that Republicans, evangelical Christians, Southerners, males, and whites disproportionately watch conservative news outlets like Fox News or read conservative news sources like Breitbart News Network, and to the degree that those outlets promote pro-​punitive stances or affect perceptions about crime or risk of victimization (e.g., a focus on “others” who are seeking to do some segment of the population harm), then this content could explain the consistently high support among these groups. It would also be useful to examine whether support is lower among groups that typically do not engage with conservative news outlets or who engage with liberal news outlets. Further, longitudinal 247

Anderson et al.

examinations are necessary as Levendusky (2013) found that extreme viewers became more extreme (i.e., polarization) when they regularly watched like-​minded, partisan media rather than “cross-​ cutting” media and that these effects remained for several days. To quickly examine the relationship between partisan news and the death penalty, we used cross-​sectional data from the 2016 Time Series Study of the American National Election Studies. Respondents were asked whether they had watched a list of a variety of television shows that included primetime news shows from Fox News (e.g., The O’Reilly Factor) and MSNBC (e.g., The Rachel Maddow Show).We conducted ANOVAs on attitudes towards the death penalty (a 1–​4 scale, strongly oppose to strongly favor) and number of shows watched on Fox News and MSNBC (none, one, two, or more) and found a significant relationship for each (F = 39.967, 2 df, p < .01 and F = 48.440, 2 df, p < .01, respectively). As demonstrated in Figure 11.2, individuals who watched more Fox News programs had stronger support for the death penalty, while those who watched more MSNBC shows had weaker support for the death penalty. These findings point to the need for more research about the cumulative effect of partisan news media while also accounting for selection bias. This research could be expanded to broadcast companies such as Sinclair Broadcast Group (founded in 1986).The Sinclair website states they are “the leading local news provider in the country” (http://​ sbgi.net/​). Sinclair is owned by a conservative family and sparked outrage in 2018 when a video was

Fox News Programs and Death Penalty Atudes 100.00% 80.00% 60.00% 40.00% 20.00% 0.00% Strongly oppose

Not strongly oppose None

Watch 1

Not strongly favor

Strongly favor

Watch 2 or More

MSNBC News Programs and Death Penalty Atudes 100.00% 80.00% 60.00% 40.00% 20.00% 0.00% Strongly oppose

Not strongly oppose None

Watch 1

Not strongly favor

Strongly favor

Watch 2 or More

Figure 11.2  Fox News and MSNBC News Viewers and Attitudes toward the Death Penalty

248

Evolving Attitudes

constructed that showed multiple news anchors across the United States giving roughly the same script regarding “dishonesty in the mainstream media,” in compliance with a “must-​run segments” policy (Edevane, 2018). This is important because individuals may be watching what they believe is a local, non-​partisan channel.These local news broadcasts may produce similar effects in different places if a consistent message is mandated. Overall, researchers need to determine whether the trends for Republicans, whites, evangelical Christians, and men in the 21st century can be explained by the way that news outlets typify crime, victims, and racial and ethnic minorities. As a final note, if one takes the position that because the United States is the only Western Democratic country to still use the death penalty and therefore should be on a trajectory towards abolishment (Zimring, 2004), then particularly problematic is the punitive tone used by President Trump.Years ago he publicly voiced his support for the death penalty in the case of the Central Park Jogger via full page advertisements in New York City newspapers (Greene, 2018). As President, he argued that some drug dealers should receive the death penalty (Merica, 2018) and has also stated that people who commit mass murder in religious establishments should be executed (Rogers, & Mays, 2018). Importantly, he has high support among many of the same demographic groups that have maintained high support for the death penalty—​Republicans, white evangelicals and men (Vandermaas-​Peeler et  al., 2018). Given prior research on the relationship between the punitive tone of the President and support for punitive policies (Ramirez, 2013a), we expect that support will increase for capital punishment among those same groups, and possibly for other groups whose support currently is lower. Presidential statements that include the racial typification of crime (e.g., drawing attention to violent crimes committed by minority group members) that are then amplified on partisan news networks can be expected to contribute to any increase in support for the death penalty if the level of prejudice among the dominant group can be increased by the way societal events are defined and treated by leaders in society (Blumer, 1958).

Conclusion US Supreme Court Justice Thurgood Marshall argued in Furman v. Georgia (1972) that the constitutionality of the death penalty under the Eighth Amendment should be linked to public opinion.That is because an informed public’s opinion would represent the current “standards of decency,” which always evolve. Further, he thought that a majority of an informed public would reject the death penalty as unconstitutional. The finding that there is less support for capital punishment once alternative options such as life without the possibility of parole are put forward provides some evidence that attitudes towards capital punishment may be softening. On the other hand, there is no evidence that exonerations affect punitive attitudes such as support for capital punishment (Anderson et al., 2017; Ramirez, 2013a). Importantly, Justice Marshall’s hypothesis is dependent on an informed public for whom retribution is not the primary reason to support the death penalty.Whites who hold racist attitudes may also hold highly retributive beliefs about the purpose of punishment. The effect of the crime rate, which can rise due to factors external to changes in offending behavior (e.g., police patrols; O’Brien, 1996), on death penalty attitudes, along with partisan sorting, news media that covers uncommon crimes, pro-​punitive stances from political leaders and in our television shows and movies, and the like may be working at cross-​purposes with having an informed public. In the end, it may be the inability of states to carry out the death penalty due to problems obtaining the drugs for lethal injection that eventually ends capital punishment in the United States.

References Abramowitz, A. (2010). The Disappearing Center: Engaged Citizens, Polarization, and American Democracy. New Haven, CT:Yale University Press.

249

Anderson et al. Allcott, H., & Gentzkow, M. (2017). Social media and fake news in the 2016 election. Journal of Economic Perspectives, 31(2), 211–​236. Amnesty International. (2018). The death penalty in 2017: Facts and figures. Retrieved from: www.amnesty.org/​ en/​latest/​news/​2018/​04/​death-​penalty-​facts-​and-​figures-​2017/​ Anderson, A.L., Lytle, R., & Schwadel, P. (2017). Age, period, and cohort effects of death penalty attitudes in the U.S., 1974–​2014. Criminology, 55(4), 833–​868. Applegate, B.K., Cullen, F.T., Fisher, B.S., & Ven,T.V. (2000). Forgiveness and fundamentalism: Reconsidering the relationship between correctional attitudes and religion. Criminology, 38(3), 719–​754. Applegate, B.K., Cullen, F.T., & Fisher, B.S. (2002). Public views toward crime and correctional policies: Is there a gender gap? Journal of Criminal Justice, 30(2), 89–​100. Bafumi, J., & Shapiro, R.Y. (2009). A new partisan voter. The Journal of Politics, 71(1),  1–​24. Baranauskas, A.J., & Drakulich, K.M. (2018). Media construction of crime revisited: Media types, consumer contexts, and frames of crime and justice. Criminology. https://​doi.org/​10.1111/​1745–​9125.12189 Barkan, S.E., & Cohn, S.F. (1994). Racial prejudice and support for the death penalty by whites. Journal of Research in Crime and Delinquency, 31(2), 202–​209. Baumer, E.P., Messner, S.F., & Rosenfeld, R. (2003). Explaining spatial variation in support for capital punishment: A multilevel analysis. American Journal of Sociology, 108(4), 844–​875. Baumgartner, F.R., Grigg, A.J., & Mastro, A. (2015). #BlackLivesDon’tMatter: Race-​of-​victim effects in US executions, 1976–​2013. Politics, Groups, and Identities 3(2), 209–​221. Baumgartner, F.R., Davidson, M., Johnson, K.R., Krishnamurthy, A., & Wilson, C.P. (2018). Deadly Justice: A Statistical Portrait of the Death Penalty. New York: Oxford University Press. Baumgartner, F.R., De Boef, S.L., & Boydstun, A.E. (2008). The Decline of the Death Penalty and the Discovery of Innocence. New York: Cambridge University Press. Beckett, K. (1997). Making Crime Pay: Law and Order in Contemporary American Politics. New  York: Oxford University Press. Beckett, K., & Sasson, T. (2003). The Politics of Injustice: Crime and Punishment in America. Thousand Oaks, CA: Sage Publications. Black, E., Black, M., & Black, E. (2009). The Rise of Southern Republicans. Cambridge, MA: Harvard University Press. Blalock, H.M. (1967). Toward a theory of minority-​group relations. The Pacific Sociological Review, 1(1),  3–​7. Blumer, H. (1958). Race prejudice as a sense of group position. The Pacific Sociological Review, 1(1),  3–​7. Bobo, L.D., & Johnson, D. (2004). A taste for punishment: Black and White Americans’ views on the death penalty and the war on drugs. Du Bois Review: Social Science Research on Race, 1(1), 151–​180. Boots, D.P. & Cochran, J.K. (2011).The gender gap in support for capital punishment: A test of attribution theory. Women & Criminal Justice, 21(3), 171–​197. Borg, M.J. (1997). The southern subculture of punitiveness? Regional variation in support for capital punishment. Journal of Research in Crime and Delinquency, 34(1),  25–​45. Bowers, W.J.,Vandiver, M., & Dugan, P.H. (1994). New look at public opinion on capital punishment: What citizens and legislators prefer. American Journal of Criminal Law, 22, 77–​150. Britto, S., & Noga-​Styron, K.E. (2014). Media consumption and support for capital punishment. Criminal Justice Review, 39(1), 81–​100. Brown, E.K., & Socia, K.M. (2017).Twenty-​first century punitiveness: Social sources of punitive American views reconsidered. Journal of Quantitative Criminology, 33(4), 935–​959. Carll, E. (2017). Disparate vantage points: Race, gender, county context, and attitudes about harsh punishments in the US. Social Science Research, 64, 137–​153. Chermak, S.M. (1994). Body count news: How crime is presented in the news media. Justice Quarterly, 11(4), 561–​582. Chiricos,T., Padgett, K., & Gertz, M. (2000). Fear,TV news, and the reality of crime. Criminology, 38(3), 755–​786. Chiricos, T., Welch, K., & Gertz, M. (2004). Racial typification of crime and support for punitive measures. Criminology, 42(2), 358–​390. Cochran, J.K., Boots, D.P., & Heide, K.M. (2003). Attribution styles and attitudes toward capital punishment for juveniles, the mentally incompetent, and the mentally retarded. Justice Quarterly, 20, 65–​93. Cochran, J.K., & Sanders, B.A. (2009). The gender gap in death penalty support: An exploratory study. Journal of Criminal Justice, 37(6), 525–​533. Dambrun, M. (2007). Understanding the relationship between racial prejudice and support for the death penalty: The racist punitive bias hypothesis. Social Justice Research, 20(2), 228–​249. Danigelis, N.L., & Cutler, S.J. (1991). Cohort trends in attitudes about law and order: Who’s leading the conservative wave? Public Opinion Quarterly, 55(1),  24–​49. Davis, J. (2018). Enforcing Christian Nationalism: Examining the Link Between Group Identity and Punitive Attitudes in the United States. Journal for the Scientific Study of Religion, 57(2), 300–​317.

250

Evolving Attitudes DiMaggio, P., Evans, J., & Bryson, B. (1996). Have American’s social attitudes become more polarized? American journal of Sociology, 102(3), 690–​755. Drakulich, K.M. (2015). The hidden role of racial bias in support for policies related to inequality and crime. Punishment & Society, 17(5), 541–​574. Edevane, G. (2018, April 2). Is your local station owned by Sinclair? What you need to know about efforts to #boycottsinclair.Newsweek.Retrieved from:www.newsweek.com/m ​ y-s​ tation-o ​ wned-s​ inclair-b​ oycott-8​ 68305 Enns, P.K. (2014). The public’s increasing punitiveness and its influence on mass incarceration in the United States. American Journal of Political Science, 58(4), 857–​872. —​—​ . (2016). Incarceration Nation: How the United States Became the Most Punitive Democracy in the World. New York: Cambridge University Press. Gallup. (n.d.). Death penalty. Retrieved from: https://​news.gallup.com/​poll/​1606/​death-​penalty.aspx Garland, D. (2001). The Culture of Control: Crime and Social Order in Contemporary Society. Chicago, IL: University of Chicago Press. Gerbner, G., & Gross, L. (1976). The scary world of TV’s heavy viewer. Psychology today, 9(11),  41–​45. Gillion, D.Q., Ladd, J.M., & Meredith, M. (Forthcoming). Party polarization, ideological sorting and the emergence of the US partisan gender gap. British Journal of Political Science (online). Gottfried, J., & Shearer, E. (2016). News Use Across Social Media Platforms 2016. Pew Research Center. Retrieved from: www.journalism.org/​2016/​05/​26/​news-​use-​across-​social-​media-​platforms-​2016/​ Greene, L. (2018, July 19). Central Park jogger attack, and still has no sympathy for accused despite convictions overturned. New  York Daily News, Retrieved from: www.nydailynews.com/​new-​york/​ny-​news-​trump-​ death-​penalty-​central-​park-​five-​20180713-​story.html Ellsworth, P.C., & Gross, S.R. (1994). Hardening of the attitudes: Americans views on the death penalty. Journal of Social Issues, 50(2),  19–​52. Grossmann, M., & Hopkins, D.A. (2015). Ideological Republicans and group interest Democrats:The asymmetry of American party politics. Perspectives on Politics, 13(1), 119–​139. Gruenewald, J., Chermak, S.M., & Pizarro, J.M. (2013). Covering victims in the news: what makes minority homicides newsworthy? Justice Quarterly, 30(5), 755–​783. Hart, P.S., & Nisbet, E.C. (2012). Boomerang effects in science communication: How motivated reasoning and identity cues amplify opinion polarization about climate mitigation policies. Communication Research, 39(6), 701–​723. Huckfeldt, R.R. & Sprague, J. (1995). Citizens, Politics and Social Communication: Information and Influence in an Election Campaign. New York: Cambridge University Press. Keil,T.J. & Gennaro F.Vito. (1989). Race, homicide severity, and application of the death penalty: A consideration of the Barnett scale. Criminology, 27(3), 511–​535. Kelly, C., Sanchez, R., & Stark, L. (2018, October 27). Trump says Pittsburg synagogue should have had armed guards. CNN. Retrieved from: www.cnn.com/​2018/​10/​27/​politics/​trump-​jba-​death-​penalty-​pittsburgh/​ index.html Kinder, D.R., & Sanders, L.M. (1996). Divided by Color: Racial Politics and Democratic Ideals. Chicago, IL: University of Chicago Press. King, R.D., & Wheelock, D. (2007). Group threat and social control: Race, perceptions of minorities and the desire to punish. Social Forces, 85(3), 1255–​1280. Kleck, G., & Jackson, D.B. (2017). Does crime cause punitiveness? Crime & Delinquency, 63(12), 1572–​1599. Kort-​Butler, L.A., & Habecker, P. (2018). Framing and cultivating the story of crime: The effects of media use, victimization, and social networks on attitudes about crime. Criminal Justice Review, 43(2), 127–​146. Kort-​Butler, L.A., & Hartshorn, K.J.S. (2011). Watching the detectives: Crime programming, fear of crime, and attitudes about the criminal justice system. The Sociological Quarterly, 52(1),  36–​55. Kort-​Butler, L.A., & Ray, C.M. (2018). Public support for the death penalty in a red state: The distrustful, the angry, and the unsure. Punishment & Society, Retrieved from: https://​doi.org/​10.1177/​1462474518795896. Lang, C., & Pearson-​Merkowitz, S. (2015). Partisan sorting in the United States, 1972–​2012: New evidence from a dynamic analysis. Political Geography, 48, 119–​129. Lehmann, P.S., & Pickett, J.T. (2017). Experience versus expectation: Economic insecurity, the great recession, and support for the death penalty. Justice Quarterly, 34(5), 873–​902. Levendusky, M. (2009). The Partisan Sort: How Liberals Became Democrats and Conservatives Became Republicans. Chicago, IL: University of Chicago Press. —​—​. (2013). Why do partisan media polarize viewers? American Journal of Political Science, 57(3), 611–​623. Mason, L. (2015). “I disrespectfully agree”:The differential effects of partisan sorting on social and issue polarization. American Journal of Political Science, 59(1), 128–​145. Matsa, K.E., & Shearer, E. (2018). News Use across Social Media Platforms 2018. Pew Research Center. Retrieved from www.journalism.org/​2018/​09/​10/​news-​use-​across-​social-​media-​platforms-​2018

251

Anderson et al. Matsueda, R.L., & Drakulich, K. (2009). Perceptions of criminal injustice, symbolic racism, and racial politics. The Annals of the American Academy of Political and Social Science, 623(1), 163–​178. Merica, D. (2018, March 19). Trump pushes death penalty for some drug dealers. CNN, Retrieved from: www. cnn.com/​2018/​03/​19/​politics/​opioid-​policy-​trump-​new-​hampshire/​index.html Messner, S.F., Baumer, E.P., & Rosenfeld, R. (2006). Distrust of government, the vigilante tradition, and support for capital punishment. Law & Society Review, 40(3), 559–​590. Miller, A.S., & Nakamura, T. (1997). Trends in American public opinion: A cohort analysis of shifting attitudes from 1972–​1990. Behaviormetrika, 24(2), 179–​191. Niven, D. (2002). Bolstering an illusory majority: The effects of the media’s portrayal of death penalty support. Social Science Quarterly, 83(3), 671–​689. O’Brien, R.M. (1996). Police productivity and crime rates: 1973-​1992. Criminology, 34(2), 183–​207. Oliphant, B. (2016). Support for death penalty lowest in more than four decades. Retrieved from: www. pewresearch.org/​fact-​tank/​2016/​09/​29/​support-​for-​death-​penalty-​lowest-​in-​more-​than-​four-​decades/​ —​—​. (2018). Public support for the death penalty ticks up. Retrieved from: www.pewresearch.org/​fact-​tank/​ 2018/​06/​11/​us-​support-​for-​death-​penalty-​ticks-​up-​2018/​ Paternoster, R. (1984). Prosecutorial discretion in requesting the death penalty: A case of victim-​based racial discrimination. Law and Society Review, 18, 437–​478. Peffley, M., & Hurwitz, J. (2007). Persuasion and resistance: Race and the death penalty in America. American Journal of Political Science, 51(4), 996–​1012. Peffley, M., & Hurwitz, J. (2010). Justice in America:The Separate Realities of Blacks and Whites. New York: Cambridge University Press. Peffley, M., Hurwitz, J., & Mondak, J. (2017). Racial attributions in the justice system and support for punitive crime policies. American Politics Research, 45(6), 1032–​1058. Pickett, J.T. (2019, forthcoming). Public opinion and criminal justice policy: Theory and research. Annual Review of Criminology. Retrieved from: www.researchgate.net/​profile/​Justin_​Pickett/​publication/​325575112_​ Public_​Opinion_​and_​Criminal_​Justice_​Policy_​Theory_​and_​Research_​Annual_​Review_​of_​Criminology/​ links/​5b1beb8aaca272021cf46727/​Public-​Opinion-​and-​Criminal-​Justice-​Policy-​Theory-​and-​Research-​ Annual-​Review-​of-​Criminology.pdf Pickett, J.T., Mancini, C., Mears, D.P., & Gertz, M. (2015). Public (mis)understanding of crime policy. Criminal Justice Policy Review, 26(5), 500–​522. Pollak, J.M., & Kubrin, C.E. (2007). Crime in the news: How crimes, offenders and victims are portrayed in the media. Journal of Criminal Justice and Popular Culture, 14(1),  59–​83. Quillian, L. (1995). Prejudice as a response to perceived group threat: Population composition and anti-​immigrant and racial prejudice in Europe. American Sociological Review, 60, 586–​611. Ramirez, M.D. (2013a). Punitive sentiment. Criminology, 51(2), 329–​364. —​—​. (2013b). Americans’ changing views on crime and punishment. Public Opinion Quarterly, 77(4), 1006–​1031. Rankin, J.H. (1979). Changing attitudes toward capital punishment. Social Forces, 58(1), 194–​211. Roche, S.P., Pickett, J.T., & Gertz, M. (2016).The scary world of online news? Internet news exposure and public attitudes toward crime and justice. Journal of Quantitative Criminology, 32, 215–​236. Rogers, K., & Mays, J.C. (2018). Trump calls for unity after synagogue shooting, then swiftly denounces Democrats. The New  York Times. Retrieved from: www.nytimes.com/​2018/​10/​27/​us/​politics/​trump-​ pittsburgh-​synagogue-​shooting.html Schwadel, P. (2010). Age, period, and cohort effects on US religious service attendance: The declining impact of sex, southern residence, and Catholic affiliation. Sociology of Religion, 71(1),  2–​24. —​—​. (2017). The Republicanization of evangelical Protestants in the United States: An examination of the sources of political realignment. Social Science Research, 62, 238–​254. Shirley, K.E., & Gelman, A. (2015). Hierarchical models for estimating state and demographic trends in US death penalty public opinion. Journal of the Royal Statistical Society, Series A, 178, 1–​28. Simon, J. (2007). Governing through Crime: How the War on Crime Transformed American Democracy and Created a Culture of Fear. New York: Oxford University Press. Sorensen, J., & Wallace, D.H. (1999). Prosecutorial discretion in seeking death: An analysis of racial disparity in the pretrial stages of case processing in a Midwestern county. Justice Quarterly, 16(3), 559–​578. Soss, J., Langbein, L., & Metelko, A.R. (2003). Why do white Americans support the death penalty? Journal of Politics, 65(2), 397–​421. Stewart, E.A., Mears, D.P., Warren, P.Y., Baumer, E.P., & Arnio, A.N. (2018). Lynchings, racial threat, and Whites’ punitive views toward Blacks. Criminology, 56(3), 455–​480. Tyler,T., & Weber R. (1982). Support for the death penalty: Instrumental response to crime, or symbolic attitude? Law & Society Review, 17(1),  21–​46. Unnever, J.D. & Cullen, F.T. (2007a). The racial divide in support for the death penalty: Does White racism matter? Social Forces, 85(3), 1281–​1301.

252

Evolving Attitudes —​—​. (2007b). Reassessing the racial divide in support for capital punishment: The continuing significance of race. Journal of Research in Crime and Delinquency, 44(1), 124–​158. —​—​. (2010).The social sources of Americans’ punitiveness: A test of three competing models. Criminology, 48(1), 99–​129. —​—​. (2012).White perceptions of whether African Americans and Hispanics are prone to violence and support for the death penalty. Journal of Research in Crime and Delinquency, 49(4), 519–​544. Unnever, J.D., Cullen, F.T., & Bartkowski, J.P. (2006). Images of God and public support for capital punishment: Does a close relationship with a loving god matter? Criminology, 44(4), 835–​866. Unnever, J.D., Cullen, F.T., & Fisher, B.S. (2007). “A liberal is someone who has not been mugged”: Criminal victimization and political beliefs. Justice Quarterly, 24(2). 309–​334. Unnever, J.D., Cullen, F.T., & Jonson, C.L. (2008). Race, racism, and support for capital punishment. In M. Tonry (ed.), Crime and Justice: A Review of the Research (Vol. 37, pp. 45–​96). Chicago, IL: Chicago University Press. Unnever, J.D., Cullen, F.T., & Roberts, J.V. (2005). Not everyone strongly supports the death penalty: Assessing weakly-​held attitudes about capital punishment. American Journal of Criminal Justice, 29(5), 187–​216. Ura, J.D., & Ellis, C.R. (2012). Partisan moods: Polarization and the dynamics of mass party preferences. The Journal of Politics, 74(1), 277–​291. Vandermaas-​Peeler, A., Cox, D., Majle, M., Fisch-​Friedman, M., Griffin, R., & Jones, R.P. (2018, October 3). Partisanship trumps gender: Sexual harassment, women candidates, access to contraception, and key issues in 2018 midterms. PRRI. Retrieved from: www.prri.org/​research/​abortion-​reproductive​health-​midterms-​trump-​kavanaugh/​ Vollum, S., Mallicoat, S., & Buffington-​Vollum, J. (2009). Death penalty attitudes in an increasing critical climate: Value-​expressive support and attitude mutability. Southwest Journal of Criminal Justice, 5(3), 221–​242. Warr, M. (1995). Poll trends: Public opinion on crime and punishment. The Public Opinion Quarterly, 59(2), 296–​310. Weitzer, R., & Kubrin, C.E. (2004). Breaking news: How local TV news and real-​world conditions affect fear of crime. Justice Quarterly, 21(3), 497–​520. Welch, K. (2007). Black criminal stereotypes and racial profiling. Journal of Contemporary Criminal Justice, 23(3), 276–​288. Wu, Y., Sun, I.Y., & Wu, Z. (2011). Support for the death penalty: Chinese and American college students compared. Punishment & Society, 13(3), 354–​376. Yang, Y., & Land, K.C. (2016). Age-​period-​cohort Analysis: New Models, Methods, and Empirical Applications. Boca Raton, FL: Chapman and Hall/​CRC. Young, R.L. (1991). Race, conceptions of crime and justice, and support for the death penalty. Social Psychology Quarterly: 67–​75. Zimring, F.E. (2004). The Contradictions of American Capital Punishment. New York: Oxford University Press.

253

12 DISPARITIES IN DEATH PENALTY PROSECUTION AND PUNISHMENT A Review of Recent Research and an Expanded Agenda Jeffery T. Ulmer and Lily Hanrath

Capital murder is the most serious criminal offense, and capital sentencing entails the most consequential punishment decisions in society. The death penalty is also an issue that invokes foundational legal, moral, and empirical questions surrounding fairness, proportionality, and wrongful convictions. Partly because of controversy surrounding these issues, the US has seen a long-​term decline in death sentences and executions since 2000. There were 39 death sentences handed down nationwide in 2017, and 23 executions carried out, both the lowest since 1973 (Death Penalty Information Center, 2017). These numbers were down from a peak of 315 death sentences in 1996 and 98 executions in 1999 (Death Penalty Information Center, 2017). In addition, there has been a less pronounced but steady decline in public support for the death penalty in general. In 2017, 55% of nationwide Gallup poll respondents expressed support for the death penalty, the lowest level since 1972 (Death Penalty Information Center, 2017). Nonetheless, the death penalty remains an important contemporary issue, in part because of racial disproportionality among those under sentence death, and the potential for disparity in ongoing death sentences associated with race/​ethnicity, locality, and defense representation. There has been perennial legal commentary and social science research on disparity in these most consequential decisions. Most prominently, empirical research has focused on disparity related to the race/​ethnicity of defendants and victims. According to Spohn (2015), Reducing the racial disproportionality in prison and eliminating racial bias in non-​capital and capital sentencing should be high-​priority goals of policy makers and politicians. Evidence that race infects the sentencing process undermines respect for the law and casts doubt on the ability of the criminal justice system to ensure due process and equal protection for all. Indeed, racial and ethnic disproportionality has been embedded in capital punishment outcomes throughout US history (Banner, 2002; Garland, 2010). This disproportionality continues in contemporary times: 39% of US death sentences in 2017 were given to African-​American defendants, and 18% to Hispanic Latino defendants. On death rows throughout the US, 41% of inmates are black and 13% are Hispanic. Since 1976, 288 black defendant/​white victim capital murder cases have 254

Death Penalty Prosecution and Punishment

resulted in execution, while 20 white defendant/​black victim cases have (Death Penalty Information Center, 2018). In addition, differences in capital case outcomes related to legal representation raise issues of due process and adequate, fair representation for indigent defendants. As Garland (2010) describes, “Capital cases are extraordinarily difficult for defense counsel. Funds are scarce, workloads heavy, and stakes high” (p. 290). The large majority of defendants in capital cases also cannot afford to pay for their own representation. Not surprisingly, the quality of defense counsel has a major effect on death penalty case outcomes (Paternoster, 2011). Moreover, a major contemporary direction of criminal justice research is the study of how discretion in the criminal justice system is shaped by larger socio-​political environments. Criminal justice is largely a local process in the United States and is largely administered at the state and especially the county level, as noted by Garland (2010): “From its beginning, America’s state has been less centralized and more localized, less bureaucratized and more personalistic … than the states of other Western nations” (p. 152).This stands out markedly in death penalty sentencing, where Garland (2010) notes: “the most important decisions about capital charging and sentencing remain matters for county-​level decision makers” (p. 117). This review will focus on contemporary empirical research on disparity in death penalty case processing and sentencing decisions, focusing on the race/​ethnicity of defendants and victims, gender, defense representation, and differences between local courts. We first briefly recall key US Supreme Court decisions involving disparity and provide a short overview of the case processing structure and key discretionary decision points of capital murder cases. We then review research on disparity conducted over the past three decades, delineated by prosecutorial discretion as compared to judicial or jury discretion. We conclude by framing research on the death penalty in terms of prominent theories of sentencing in general, and outlining new directions for research. Such research will hopefully fill important gaps in scholarly understanding of punishment and provide guidance for law and policy efforts to come to gain traction on issues of fairness, arbitrariness, and equal protection surrounding the death penalty.

Key US Supreme Court Cases Regarding Disparity in the Death Penalty There have been several notable Supreme Court cases regarding the death penalty in the last 50 years. The status of the death penalty has gone through multiple phases that have differentially limited or increased discretion of judges and prosecutors in death-​eligible cases. Issues of disparity are front and center in many of the cases taken on by the Supreme Court. In Furman v. Georgia, 408 U.S. 238, decided in 1972, the United States Supreme Court articulated its concerns with unwarranted disparity in the administration of capital punishment. Furman struck down the death penalty in the 40 death penalty jurisdictions, finding that Furman had been deprived of his constitutional rights and that the death penalty was administered in an arbitrary and capricious manner. Specifically, the majority of the Justices ruled that the sentence of death was not unconstitutional per se, but the procedures used in applying the death penalty across the states were unconstitutional in allowing for bias in its application against the poor, uneducated, mentally disabled, and minorities. The message to the states was that they needed to develop and implement procedures to ensure that the application of the death penalty would not be discriminatory against offenders because of their status. Ultimately, in 1976, in Gregg v. Georgia, 428 U.S. 153 (1976), the United States Supreme Court upheld the constitutionality of Georgia’s death penalty statute, and also struck down mandatory death penalty statutes. In response to Furman, Georgia’s new death penalty statute bifurcated the trial in death penalty cases to include separate proceedings to determine guilt and to determine the sentence after consideration of mitigating and aggravating circumstances. Georgia required the jury to find beyond a reasonable doubt that the offender’s case involved at least one of the specified aggravating 255

Jeffery T. Ulmer and Lily Hanrath

circumstances. The Court found that this system of administration of the death penalty contained protections against unfair applications that were at the root of the Furman v.  Georgia decision in 1972. As will be seen later, however, empirical research shows that the procedures approved in Gregg v. Georgia have not eliminated unwarranted disparities. By 1987, scholars had produced several sophisticated, multivariate analyses of disparity in the application of the death penalty in several states; most notably two studies in Georgia done by David Baldus and associates (see Baldus et al., 1983, 1990).The Baldus studies informed the plaintiff ’s case in McKleskey v. Kemp, 481 U.S. 279 (1987), wherein a black death row inmate pointed out that similar crimes committed by white offenders all resulted in life sentences. The court considered empirical research on racial disparity at length and concluded that the Baldus study was “valid statistically.” Ultimately, however, the court posited that overall patterns of discrimination could not be considered because decisions are individual, and it was not proven that the judge and prosecutor in McKleskey’s case “acted with discriminatory purpose.”1 A dissenting opinion by Justice Harry Blackmun in Callins v. Collins 510 U.S. 1141 (1994) provides another articulation of the complexity of both the death penalty and arguments for and against its use. Justice Blackmun chose to issue a dissent despite the Supreme Court electing not to issue a written opinion. He noted that he previously dissented in Furman and had concurred in Gregg v. Georgia, but had come to change his mind about these decisions. Blackmun reached the conclusion that no amount of regulation or narrowing of death penalty eligibility can completely eliminate arbitrary capital sentencing. In particular, his dissent argues that “even under the most sophisticated death penalty statutes, race continues to play a major role in determining who shall live and who shall die.” As in McKleskey, Blackmun cites Baldus’ study as support that racial disparity is both evident and significant in death penalty decision-​making. Justice Blackmun’s dissent stands as an eloquent statement of grave concern about racial and other types of unwarranted disparity in death sentencing. More recently, Justice Stephen Breyer, with a dissent in Glossip v. Gross (576 U.S. 2015), decried that the death penalty, as currently administered, suffered from arbitrary application, in part due to wide geographic disparity in death sentences and their concentration in a comparatively small number of counties nationwide. In a dissent from the Supreme Court’s decision not to grant certiorari in Jordan v. Mississippi and Evans v. Mississippi (585 US 2018), Justice Breyer further expanded on his argument that wide geographic disparity in the application of the death penalty constituted the kind of arbitrariness rejected by Furman.

An Overview of Death Penalty Discretion Points The death penalty decision-​making system consists of a linked series of prosecutorial, defense, and judicial decision-​making processes made largely at the local level. All of these decision points are potential sites of disparity. The system contains numerous points at which discretion is exercised by prosecutors, judges, and juries to exclude individual death-​eligible cases from the risk of a death sentence. This can result from the plea agreement process, in which prosecutors agree to reduce a first-​degree murder indictment to a lesser murder charge, or to waive the death penalty as part of a plea agreement. It can also result from the court acquitting the defendant on the first-​degree charge during the trial phase, or the judge or jury sentencing the defendant to life imprisonment in the penalty phase. The general, typical case processing sequence can be divided into several steps: 1) 2) 3) 4)

A homicide occurs. A homicide suspect is identified and arrested. The prosecutor decides whether to charge first-​degree murder. The prosecutor, or in some states a grand jury, decides whether to indict for first-​, second-​, or third-​degree murder. 256

Death Penalty Prosecution and Punishment

5) If the defendant is indicted for first-​degree murder, the prosecutor decides whether to seek death penalty. 6) The prosecution and defense may reach a plea agreement, perhaps entailing a reduction to a lesser murder charge, or an agreement for the prosecutor to not seek the death penalty, or to retract a death penalty filing already made; or 7) The defendant stands for bench or jury trial for first-​degree murder. 8) If convicted of first-​degree murder, and the prosecution has filed a motion to seek the death penalty, the sentencing authority (either the jury or judge) must decide on whether the defendant deserves the death penalty on the basis of finding either (a) existence of aggravating circumstances and no mitigating circumstances or (b) aggravating circumstances outweigh mitigating circumstances. 9) An appellate process, which may involve appeals in both state and federal courts, and which may or may not result in a new sentence hearing. At each stage in the process, attrition occurs and the universe of cases that could result in a death sentence narrows. By the end of the process, the number of individuals actually sentenced to death is an extremely small fraction of the number of people who are charged with capital homicide, and the number actually executed is even smaller (254 people have been executed from 2010 to 2018; Death Penalty Information Center, 2018). As an example of this winnowing process, Figure 12.1 shows the case flow of a sample of 880 first-​degree murder convictions from a recent study of the death penalty in Pennsylvania (Kramer, Ulmer, & Zajac, 2017). Most states’ homicide statutes provide for first, second, and third degrees of murder. To comply with the logic of Furman v.  Georgia and Gregg v.  Georgia, aggravating circumstances that can lead to the death penalty must be defined in statute, and for a case to be death-​eligible, one or more aggravating circumstances must be found and proved beyond a reasonable doubt. For example, in Pennsylvania these 18 statutory aggravating circumstances are:

D.A Sought and Later Retracted Death Penalty Filing (17%)

D.A. Sought Death Penalty (36%) 880 1st Degree Murder Convicons

Defendant Received Death Sentence (6%)

Cases Ulmately Exposed to Death Penalty at Sentencing (19%) Defendant Received Life Sentence (94%)

D.A. Did Not Seek Death Penalty (64%)

Figure 12.1  Death Penalty Exposure and Sentences for 880 First Degree Murder Convictions in 18 counties, Pennsylvania, 2000–​2010

257

Jeffery T. Ulmer and Lily Hanrath

1. The victim was a firefighter, peace officer, prison guard, judge of any court, the Attorney General of Pennsylvania; a deputy attorney general; district attorney; assistant district attorney; member of the General Assembly; Governor; Lieutenant Governor; Auditor General; State Treasurer; State, Local, or Federal law enforcement official or a person employed to assist any law enforcement official, who was killed in the performance of his duties or as a result of his official position. 2. The defendant was contracted to kill the victim. 3. The victim was being held by the defendant for ransom, or as a shield or hostage. 4. The victim was killed while the defendant was hijacking an aircraft. 5. The victim was a witness for a separate felony case against the defendant and was killed to prevent the victim’s testimony. 6. The defendant committed a killing while in the perpetration of a felony. 7. The defendant knowingly created a grave risk of death to another person in addition to the victim of the offense. 8. The offense was committed by means of torture. 9. The defendant has a significant history of felony convictions involving the use or threat of violence to the person. 10. The defendant has been convicted of another federal or state offense for which a sentence of life imprisonment or death was imposable, or the defendant was undergoing a sentence of life imprisonment at the time of the commission of the offense. 11. The defendant has been convicted of another murder before or during the offense. 12. The defendant has been convicted of voluntary manslaughter, or a substantially equivalent crime before or during the offense. 13. The defendant committed the killing or was an accomplice in the killing, while in perpetration of felony drug trafficking. 14. The defendant committed the killing or was an accomplice to the killing because the victim was or had been involved, associated, or in competition with the defendant in drug dealing. 15. The victim was a nongovernmental informant for law enforcement and the defendant killed them in retaliation for these activities. 16. The victim was a child under 12 years of age. 17. The victim was in her third trimester of pregnancy or the defendant had knowledge of the victim’s pregnancy. 18. The defendant was subject to a court order restricting the defendant in order to protect the victim from them. The following are mitigating circumstances that may be presented by the defense during the sentencing phase of the first-​degree murder trial: 1 . The defendant has no significant history of prior criminal convictions. 2. The defendant was under the influence of extreme mental or emotional disturbance at the time of the commission of the murder. 3. The capacity of the defendant to appreciate the criminality of his conduct or to conform his conduct to the requirements of law was substantially impaired. 4. The age of the defendant at the time of the crime. 5. The defendant acted under extreme duress or acted under the substantial domination of another person. 6. The victim was a participant in the defendant’s homicidal conduct or consented to the homicidal acts. 7. The defendant’s participation in the homicidal act was relatively minor. 8. Any other evidence of mitigation concerning the character and record of the defendant and the circumstances of his offense. 258

Death Penalty Prosecution and Punishment

Pennsylvania’s list is very similar to other typical aggravating and mitigating circumstances listed in other death penalty states’ statutes.2 It is notable that the aggravating circumstances vary a great deal in their specificity and breadth. That is, some of the aggravating circumstances, such as victimizing a firefighter or police officer or victimizing someone under 12 years old, are very clear-​cut and narrow. Others potentially apply to a much broader segment of cases, and/​or leave room for interpretation, such as the offender having a significant history of violent felonies, the murder being committed in perpetration of a felony or drug trafficking, or the defendant “knowingly creating grave risk of death” to others besides the victim. The breadth and interpretive nature of aggravating circumstances such as these latter ones allow prosecutors discretion and leeway in seeking the death penalty, and in the breadth of cases which could be eligible for the death penalty.3 In addition, courts, and/​or juries have substantial discretion at sentencing in terms of finding whether the aggravating circumstances are proven to a reasonable doubt standard, and deciding whether these are outweighed by any mitigating circumstances. Table 12.1 provides an example from the recent death penalty study of Pennsylvania (Kramer, Ulmer, & Zajac, 2017) of the scope of how different aggravating circumstances are filed by prosecutors and found at trial by judges or juries. As shown in Table  12.1, many aggravating circumstances are quite rarely presented and even more rarely found. However, the aggravating circumstances, “Committed while in perpetration of a felony,” “Defendant knowingly created grave risk of death to another,” “Defendant has significant history of violent felony convictions,” and “Defendant has been convicted of another murder” are presented more frequently than others. This is not surprising, as at least one of these aggravating circumstances potentially applies to many first-​degree murder cases. Conversely, there were no aggravating circumstances filed for the highly specific and narrow circumstance of “hijacking an aircraft.” In addition, all of the aggravating circumstances are found by the court much less often than they are presented. The case-​processing decision points involved in capital murder cases all present points of discretion by prosecutors, judges, and juries, as do the consideration of various aggravating and mitigating Table 12.1  Statutory Aggravating Factors Filed by Prosecutors and Found at Trial by Jury/​Judge: 18 counties of Pennsylvania 2000–​2010 Aggravating Circumstances

Prosecutor Filed

Jury/​Judge Found

Victim was firefighter, peace officer Defendant was paid for killing Victim held for ransom, reward, or shield Hijacking aircraft Victim was prosecution witness Committed in perpetration of felony Knowingly created grave risk of death Offense committed by means of torture Significant history of violent felonies Defendant convicted of offense carrying life/​death Defendant convicted of another murder Defendant convicted of voluntary manslaughter Defendant committed killing during drug felony Victim was associated with defendant in drug trafficking Victim was a nongovernment informant Victim was under 12 Victim was in third trimester or def. knew of pregnancy Defendant was under PFA from victim

12 (1.4) 3 (.3) 3 (.3) 0 29 (3.3) 134 (15.2) 136 (15.5) 48 (5.5) 82 (9.3) 39 (4.4) 85 (9.7) 11 (1.3) 13 (1.5) 23 (2.6) 4 (.5) 27 (3.1) 8 (.9) 11 (1.3)

4 (.5) 1 (.1) 1 (.1) 0 5 (.6) 23 (2.6) 21 (2.4) 11 (1.2) 16 (1.8) 8 (.9) 23 (2.6) 0 1 (.1) 2 (.2) 3 (.3) 6 (.7) 4 (.5) 1 (.1)

Frequency (percent of 880 first-​degree murder convictions) (Note: cases can have multiple aggravating factors).

259

Jeffery T. Ulmer and Lily Hanrath

circumstances that may apply, and that shape exposure to the death penalty. Each of these discretionary decisions and the consideration of aggravating and mitigating factors that shape these decisions are potential sites of disparity based on defendant characteristics. In addition, these decision processes potentially vary by jurisdiction and by local contexts. Research has heavily focused on some of these types of disparity, but hardly attended to others.

Prior Research on Disparity in Death Penalty Case Processing Most research on disparity in death penalty decision-​making has focused on the race of the defendant and the race of the victim, both singly and in combination. Comparatively less research has examined disparity connected to victim or defendant gender, defendant socioeconomic status, variations in death penalty decisions and outcomes between local courts, and differences connected to legal representation. Below, we briefly review research on these bases of disparity and variation. The scope of this review emphasizes studies conducted since the 1990 review of disparity in the death penalty by the US General Accountability Office. Table 12.2 presents a listing and brief sketch of key studies since 1995.

Race of Defendant and Victim A major review of research on disparity in administration of the death penalty in the post-​Furman era was conducted by the US General Accountability Office (1990), and this review provides a starting point for our consideration of contemporary empirical research on disparity in the death penalty.The General Accountability Office review concluded that the evidence for the influence of the race of the defendant on death penalty outcomes was mixed. Although more than half of the studies found that the race of defendant influenced the likelihood of being charged with a capital crime or receiving the death penalty, the relationship between race of defendant and outcome varied across studies (US General Accountability Office, 1990). The General Accountability Office review, however, found that the race of the victim had a much stronger influence on capital sentencing outcomes. In 82% of the studies, the race of the victim was found to influence the likelihood of being charged with capital murder or receiving a death sentence; that is, those who murdered whites were found to be more likely to be charged with a capital crime and sentenced to death than those who murdered blacks. This finding was remarkably consistent across data sets, states, data collection methods, and analytic techniques. In the parts that follow, we discuss the results of research on the role of the race of the defendant and victim in prosecutorial decisions and in court sentencing decisions.

Prosecutorial Decisions Prosecutors have broad discretion in their decision to seek the death penalty in a given case (Mangino, & Champion, 2013). Many studies have found that prosecutors were more likely to seek the death penalty for capital murder cases involving a white victim (Bowers & Pierce, 1980; Paternoster, 1983, 1984; Radelet & Pierce, 1985; Keil & Vito, 1995; Brock et al., 1999; Paternoster et al., 2004; Songer & Unah, 2006; Hindson et al., 2006; Williams et al., 2007) than for cases involving a black victim. Other studies have found prosecutors are specifically more likely to seek the death penalty when the defendant was black and the victim was white (e.g., Keil & Vito, 1995; Brock et al., 1999; Lenza et al., 2005; Donohue, 2014). For example, Lenza et al. (2005), in their study of death penalty charging decisions in Missouri from 1978 to 1996, found that blacks who killed whites were five times more likely to be charged with capital murder than blacks who kill blacks. A particularly rigorous recent study by Donohue (2014) of 205 death-​eligible murders in Connecticut from 1973–​2007 found that black defendant/​white victim cases were substantially more likely than other types of cases to 260

newgenrtpdf

Table 12.2  Select Studies on Disparity in Death Penalty Decisions 1995–​2017 Author/​date

Data

Method

Key Disparity Analyses

Findings

Keil, T. J. & Vito, G. F. (1995)

KY 1976–​1986 864 cases Cook County, IL No years given 1,925 jurors

LR O

-​prior  record -​R interactions of D & V

-​past violent offenses did not significantly impact DP decisions -​BD/​WV more to likely to be charged and convicted with the DP

Q

-​Attorney  type -​effect of expert witnesses -​effects of cross-​examination

Baldus, et al. (1997–​1998)

PA 1983–​1993 425 cases

LR

Brock, D. E., Sorensen, J., & Marquart, J. W. (1999)

TX 1991–​1993 179 cases

LR

-​R of D & V -​SES of D & V -​murder severity index -​salient factors -​R of D & V

-​Jury perception of attorneys does not relate to final verdicts -​beliefs and values conditioned the effect of evidence and presentation -​juries pay more attention to substantive aspects than personal qualities of lawyers -​cross-​examination did not change the% favoring DP verdict BD & WV most likely to be convicted

Blankenship, M. B. & Blevins, K. R. (2001)

TN 1977–​1998 152 cases

O

-​R of D & V -​Age of D & V -​G of V

LR

-​R of D & V -​G of D & V -​Age of D & V

Diamond, S. S., Casper, J. D., Heiert, C. L., & Marshall, A. (1996)

Williams, M. R. & Holcomb, J. E. OH (2001) 1981–​1994 5,319 cases

Racial discrimination based on V R in prosecutorial stage of case processing -​strengthened if BD -​More disparity in lower levels of case culpability (because of more discretion) -​Greatest disparity among convicted capital cases V & D relationship & Juror Rs significantly affected number of aggravators found (more aggravators when the V was a stranger & when juror R differed from D) -​no significant differences were found by R of the D,V, or combination V R increases the likelihood of a DP but not as high as previously reported. The R of D was not an independent predictor of a DP WV cases more likely to get DP (continued)

newgenrtpdf

Table 12.2  (Cont.) Author/​date

Data

Method

Key Disparity Analyses

Findings

Brooks, R. & Raphael, S. (2002)

Chicago, IL 1870–​1930 2,631 cases

O

Baldus et al. (2002–​2003)

NE 1973–​1999 184 cases

LR

-​between judge variation -​election  year -​R of D & V -​Age  of  V -​G of V -​county -​R of D & V -​R interactions of D & V -​SES of D & V

Paternoster, R. & Brame, R. (2003)

MD 1978–​1999 1,311 OH 1981–​1997 5,976 cases MO 1978–​1996 574 cases

LR

-​R of D & V -​county

-​significant variation between judges in DP rates -​no significant difference in guilty verdict by election year -​significant difference in cases resulting in a DP by election year (more DPs during election years) -​D & V characteristics not significant predictors -​no evidence that prosecutors or judges treat cases differently based on R of D or V -​State law creates more discrimination against minority Ds -​B Ds more likely to actually be executed -​High SES V led to more charging and sentences with DP -​Large disparities in practices by county Cases with B DF & W V more likely to get capital charge -​large geographical effects on capital charge

LR

-​R & G Interactions of V -​R of D & V -​G of D & V -​R of D & V -​Age of D -​G of D & V -​public/​private defender -​prior convictions

CO 1972–​2005 110 cases

O

Holcomb, J. E., Williams, M. R., & Demuth, S. (2004) Lenza, M., Keys, D., & Guess, T. (2005)

Hindson, S., Potter, H., & Radelet, M. L. (2006)

LR

-​R of D & V -​G of D & V

-​DP more likely when V is W & F -​WFVs were the only statistically distinct V dyad to have an effect on DP -​BD/​WV are 5X as likely to be charged with capital murder than  BD/​BV -​WD/​BV are half as likely to be charged with capital murder than  WD/​WV. -​no evidence found that D/​V R effected jury decisions -​MDs more likely to be taken to trial & get DP -​FV more likely to result in DP -​D with public defender more likely to get DP -​Decision to seek DP influenced by V R &G -​Decision to seek DP influenced by geography

newgenrtpdf

Author/​date

Data

Method

Key Disparity Analyses

Findings

Poveda, T. G. (2006)

VA 1978–​2001 Aggregate analysis of 85 jurisdictions

O

-​Population size and percentage of B residents were significant predictors of DP -​homicide case characteristics and contemporary conditions had a greater impact than historical, economic, and political factors on DP -​conservatism, slavery, lynchings, and economic marginality were not significant predictors of DP

Williams, M. R., Demuth, S., & Holcomb, J. E. (2007)

GA 1973–​1979 1,066 cases

LR

-​Historical variables (slaves, lynchings, pre-​Furman executions) -​Size -​SES -​Political leanings -​Region -​R of D & V -​G of D

Paternoster, R. & Brame, R. (2008)

MD 1978–​1999 1,311 cases

LR O

Phillips, S. (2009a)

Harris County, TX 1992–​1999 504 cases

LR

Phillips, S. (2009b)

Harris County, TX 1992–​1999 504 cases

LR

-​R of V & D -​BD & WV vs. not -​93 background characteristics -​V  SES -​V  Status -​Heinousness (aggravating and mitigating circumstances) -​form of legal counsel (appointed, hired, mixed) -​D SES

-​V G matters for juries but not for prosecutors -​V R influences prosecutors’ decisions more than juries’ decisions -​DP more likely with F V -​cases involving WFVs are treated more harshly -​cases involving BMVs are treated the most leniently -​cases involving BMVs are the most divergent BD/​WV cases more likely to have DP sought BD/​WV are significantly more likely to get DP

DP more likely when the V was of a high-​status and less likely when the V was a low-​status

-​No apparent relationship between SES & form of counsel -​Hired legal counsel increases chance of acquittal -​no Ds who hired counsel were given a DP -​hired counsel were more likely to negotiate a plea bargain (continued)

newgenrtpdf

Table 12.2  (Cont.) Author/​date

Data

Method

Key Disparity Analyses

Findings

Behnken, M. P., et al. (2011)

AR, FL, GA, NC, NJ, OH, OK, TX No year given 654 cases

O

-​Prior offenses -​R of D

Unah, I. (2011)

NC 1993–​1997 520 cases DE 1977–​2007 146 cases

LR

R of D & V D & V R interactions

-​juvenile arrest & prior violent offenses were significantly associated with DP -​prior prison sentences predicted DP -​the number of Vs significantly predicted DPs -​R of D was not found to be a predictor of DP -​juries more likely to give DP when there’s a BV -​prosecutorial decisions not influenced by V or D R

LR

-​R interactions of D & V -​County of case -​Judge vs Jury decision on punishment -​G of V N/​A

Johnson, S. L. et al. (2012)

Peterson, N. & Lynch, M. (2012) Los Angeles, CA 1996–​2008 7,258 cases

LR O

Donohue, J. J. III. (2014)

LR

CT 1973–​2007 205 cases

Richards, T. N., Smith, M. D., NC Jennings, W. G., Bjerregaard, B., 1977–​2009 & Fogel, S. J. (2014) 1,113 cases

PS O

-​Heinousness  Scales -​R of D & V -​RD/​RV Interactions -​County -​G of D

-​Judges gave more DPs than Juries -​BD/​WV more likely to get DP by both Judge and Jury

-​special circumstance committees seek to maximize the odds of a conviction by selectively filing death notices in cases involving multiple special circumstance allegations -​capital cases are more costly and take longer -​filing a motion to seek the DP increases cost and time even when a plea bargain is reached -​Heinousness did not predict DP -​BD/​WV significantly more likely to get DP -​County of case significant influence on DP -​F D less likely to lead to the DP being sought or given -​M D are more likely to have aggravating factors and mitigating factors in their cases -​F D are more likely to have a DP trial if their crime doesn’t fit the “profile” of a F murder

newgenrtpdf

Author/​date

Data

Method

Key Disparity Analyses

Findings

Jennings, W. G. et al. (2014)

NC 1977–​2009 1,113 cases

LR PS

-​R of D & V -​attorney  type -​geography (three regions)

Vito, G. F., Higgens, G. E., & Vito, KY A. G. (2014) 2000–​2010 359 cases

LR

-​R of D & V -​G of D & V

Hans,V. P. et al. (2015)

DE 1977–​2007 146 cases

LR

Kramer, J. H., Ulmer, J., & Zajac, G. (2017)

PA 2000–​2010 880 cases

PS LR

-​Heinousness -​Judge vs. Jury decision on punishment -​R interactions of D & V -​County -​G of V -​R of D & V -​attorney  type -​county

-​WV more likely to result in DP based on LR -​WV not significant influence on DP based on PS -​“R of V effect” may be better described as a “case effect” where R plays a part Plea deals are influenced by D/​V R interaction -​D R did not significantly affect DP seeking decisions -​prosecutors were more likely to seek the DP and less likely to offer a plea with a FV -​prosecutors sought the DP more when judges decided on punishment -​Judges gave the DP more than juries -​DP given more with BD/​WV -​County of case influenced DP decision -​FV more likely to result in the DP -​DR did not significantly affect whether DP sought, retracted, or imposed -​DP 8% more likely imposed in cases with WVs, 6% less likely imposed in cases with BVs -​cases with public defenders were less likely to have DP sought, but more likely to have DP imposed. Private attorney cases much less likely to have DP imposed -​large differences in likelihood of seeking, retracting, and imposing DP by county; differences in effects of VR and defense attorney by county

Jeffery T. Ulmer and Lily Hanrath

be charged as capital crimes and thus exposed to the death penalty. This pattern persisted even when controlling for statutory aggravating circumstances and independent coders’ race-​blinded ratings of the “egregiousness” of the murder cases. Other research, however, presents a more mixed picture. For example, David Baldus and associates’ (1997–​1998) well-​known study of death penalty decisions in Philadelphia from 1983–​1993 did not find a race of defendant or a race-​of-​victim effect in the prosecution’s decision to seek the death penalty but did find that prosecutors were significantly less likely to retract a filing to seek the death penalty when the murder victim was black. By contrast, a recent statewide study of death penalty decisions in Pennsylvania by Kramer, Ulmer, and Zajac (2017) found no pattern of disparity to the disadvantage of black or Hispanic defendants, including those who killed white victims, in prosecutors’ decisions to seek and, if sought, to retract the death penalty. Similarly,Vito et al.’s (2014) study of Kentucky from 2000–​2010 found no significant disparity connected to black defendant/​ white victim cases in prosecutors’ decisions to seek the death penalty. However, they did find that “black offenders charged with killing a white victim were much less likely to benefit from a plea in a capital case” (Vito et al., 2014, p. 763). In other words, black offenders were less likely to obtain plea bargains where prosecutors did not seek the death penalty in exchange for a guilty plea. Complicating the picture further, Unah (2011) found that North Carolina prosecutors were less likely to seek the death penalty in cases with non-​white defendants and white victims than when the defendant and victim were white.

Death Sentences Compared to research on race and prosecutorial decisions to seek the death penalty, considerably more research has investigated racial disparity in death penalty sentencing. A large number of studies have found that black defendants who are convicted of killing white victims are the most likely to receive the death penalty (e.g., Bowers & Pierce, 1980; Gross & Mauro, 1984; Keil & Vito, 1995; Baldus et al., 1990, 1997–​1998; Holcomb, Williams, & Demuth, 2004; Lenza, et al., 2005; Williams, Demuth, & Holcomb, 2007; Paternoster & Brame, 2008; Unah, 2011; Johnson et al., 2012; Donohue 2014; Ohio Joint Task Force, 2014).This includes the famous study of the death penalty in Georgia by Baldus et al. (1990), which figured centrally in the McKleskey v. Kemp (1987) US Supreme Court case. Baldus and colleagues (1990) examined over 2,000 cases involving defendants charged with murder in Georgia and found that a defendant convicted of killing a white victim was four times more likely to receive a death sentence than a defendant convicted of killing a non-​white victim. A re-​analysis by Williams et al. (2007) of the 1970s Georgia data compiled by Baldus and colleagues (1990) also found that cases involving black male offenders with white victims were treated most severely, while black offenders with black victims were treated most leniently. Similar to the Georgia study, the Baldus et al. (1997–​1998, p. 1714) study of Philadelphia also found that the race of the defendant is “a substantial influence in the Philadelphia capital charging and sentencing system, particularly in jury penalty trials.” Specifically, black defendants, and especially those who killed white victims, were more likely to receive the death penalty, and such racial disparity was more prominent in jury sentences than sentences handed down by judges. More recently, Unah’s (2011) study in North Carolina also found that cases with non-​white defendants and white victims were 8% more likely to receive the death penalty. In their comprehensive study of the death penalty in Maryland, Paternoster and Brame (2008; see also Paternoster et al., 2004) found defendants with white victims were six times more likely to receive the death penalty than defendants with non-​white victims. Similarly, the recent study of the death penalty in Pennsylvania by Kramer, Ulmer, and Zajac (2017), using propensity score methods, found no disparities in death sentences connected to race of defendant, but found notable race-​of-​ victim effects. Cases with white victims were 8% more likely to receive the death penalty, while black

266

Death Penalty Prosecution and Punishment

victim cases were 6% less likely to receive the death penalty. These effects were present regardless of defendant race. By contrast, some other studies have failed to find race-​of-​defendant and race-​of-​victim effects, either singly or in combination. Specifically, studies by Blankenship and Blevins (2001) and Jennings et al. (2014) in North Carolina, and by Baldus et al. (2002–​2003) in Nebraska did not find black defendant/​white victim disparities. The Jennings et al. (2014) study analyzed North Carolina capital murder trials between 1977 and 2009 using propensity score matching and found results that conflicted with the findings of Unah (2011), who found black defendant/​white victim disparities using somewhat different data and relying on logistic regression for analysis. That the Baldus et  al. (1997–​1998) Philadelphia study found that race-​of-​victim and race-​of-​ defendant effects were particularly strong in jury decisions raises the question of what lay behind these jury sentences. To shed light on this question, Foglia (2003) interviewed 74 jurors who participated in 27 death penalty trials in Philadelphia. Forty-​three of those interviewed were jurors in cases in which the defendants were sentenced to death, and 31 were jurors in cases where the defendants were sentenced to life without parole. Foglia (2003) found that most of the jurors whom she interviewed misunderstood the law of capital sentencing. They often based their decisions on the erroneous assumption that the defendant would be released after a term of years if given a life sentence. They also failed to understand jury instructions regarding mitigation in their deliberations. Foglia (2003) found that jurors who assumed that defendants given a life sentence would serve 15 years or less in prison were much more likely to vote for the death penalty. While these findings were based on different cases than those included in the Baldus (1997–​1998) study, they reinforce the notion that jurors’ ignorance of the law could result in their reliance on perceptions of risk posed by the defendant and the defendant’s culpability. Finally, some recent research seems to indicate that the race of the defendant and victim can influence judgments about the severity or aggravated nature of murders and might actually outweigh objective indicators of aggravation. Donohue (2014) had 18 coders independently evaluate 205 death-​eligible cases in Connecticut with race and outcome information removed, using three different murder “egregiousness” scales based on Connecticut’s statutory aggravating circumstances. On average, white defendant/​white victim crimes were rated by the coders as objectively most egregious, while minority defendant/​white victim crimes were rated least egregious. Yet, in the real-​life case outcomes, the latter were most likely to result in a death sentence. Overall, Donohue’s (2014) analyses found that the race of the defendant and victim were stronger predictors of death sentences than the egregiousness scales or specific aggravating factors. In fact, only one of the nine cases that ultimately resulted in a death sentence was rated as among the 15 most egregious by coders.

Gender of Defendants and Victims Compared to research on the race of defendants and victims, fewer studies have specifically focused on the gender of the defendant or victim. One reason for this may be the relative rarity of female capital murder defendants compared to males. Two% of death row inmates are female (Death Penalty Information Center, 2018), and women account for less than two% of the US murder rate. Nonetheless, a handful of studies have examined defendant and victim gender in relation to prosecutors’ decisions to seek the death penalty. Studies generally conclude that female defendants are less likely to be prosecuted for capital crimes (Richards et al., 2014). By contrast, studies have found that cases involving female victims are more likely to result in prosecutions for capital crimes (Vito et al., 2014; Lenza et al., 2005; Williams et al., 2007; Royer et al., 2014; Songer & Unah, 2006). For example, a recent study by Vito et al. (2014) of death-​eligible offenders in Kentucky from 2000–​2010 found that prosecutors were 3.17 times more likely to seek the death penalty when the victim was female.

267

Jeffery T. Ulmer and Lily Hanrath

Some studies have examined defendant and victim gender in relation to death sentencing decisions. Jennings et al. (2014), in their propensity score matching analysis of the North Carolina capital murder trials data between 1977 and 2009, found that female defendants were much less likely to receive the death penalty. Other studies have supported the Jennings study findings (Lenza et al., 2005; Holcomb et al., 2004). Additionally, in Unah’s (2011) study of North Carolina death penalty decisions, defendants in cases involving female victims were significantly more likely to receive a death sentence. Hans et al. (2015) examined 146 capital cases in Delaware from 1977–​2007, finding that cases with female victims were more likely to result in death sentences, especially among jury-​ sentenced (compared to judge-​sentenced) cases. Using data from Georgia collected by Baldus and colleagues (1983), Holcomb et al. (2004) and Williams et al. (2007) found that defendants convicted of killing white females were 14.5 times more likely to receive the death penalty than similarly situated offenders accused of killing black males. Holcomb et al. (2004), Williams et al. (2007) and Royer et al. (2014) found that the pronounced likelihood of accused killers of white females being sentenced to death was explained by the sexualized nature of the victimization surrounding such homicides. By contrast, a supplemental analysis in Donohue’s (2014) analysis of Connecticut capital cases failed to uncover significant case outcome differences by gender of defendant or victim.

Type of Legal Representation There have long been concerns with the quality of legal representation in capital cases. Justice Thurgood Marshall (1986) remarked that, due to the unique demands of death penalty cases, “even the most well-​intentioned attorneys often are unable to recognize, preserve, and defend their client’s rights” (p. 1). Some legal experts go as far as to argue that defense lawyer competency is a better predictor of capital case outcomes than the facts of the case (Bright, 1994; Doyle, 2013). Unfortunately, most death penalty research does not include type of defense representation in analyses of charging and sentencing decisions. Exceptions include Phillips (2009b), Unah (2011), Donohue (2014), Lenza et al. (2005), and Anderson and Heaton (2012). Phillips’ (2009b) study of Harris County, TX (one of the counties with the greatest number of death sentences in the US) found that defendants with private attorneys had a greater probability of obtaining a negotiated plea to avoid the death penalty, compared to defendants with court-​assigned indigent counsel (Harris County did not have a public defender representation in capital cases). He also found that defendants with privately retained counsel had a greater probability of an acquittal at trial. Unah (2011) found that North Carolina defendants with public defenders were 22% more likely to have the death penalty sought by prosecutors than defendants with private representation. A supplemental analysis of Connecticut capital cases by Donohue (2014) also found that defendants represented by public defenders were moderately but significantly more likely to be charged with capital murder and exposed to the death penalty. Regarding death sentencing decisions, Lenza, et al. (2005) examined types of legal representation and found that defendants represented by public defenders were more likely to receive the death penalty than defendants who had assigned or privately retained counsel. Phillips’ (2009b) study also found that no defendant who retained private counsel was given a death sentence. Finally, the study of first-​degree murder convictions in Pennsylvania (from 2000–​2010) by Kramer et al. (2017) found that there were notable differences in death penalty outcomes connected to type of defense attorney representation. The likelihood that the prosecutor would file a notice of intent to seek the death penalty was less likely if the defendant was represented by a public defender than if the defendant was represented by a private attorney or a court-​appointed attorney; this was especially the case in Philadelphia relative to the other 17 counties in the data. However, defendants with public defenders were more likely to receive the death penalty statewide, and defendants represented by private attorneys had extremely low odds of receiving the death penalty. These defense attorney differences, however, varied greatly among counties—​defendants represented by public defenders in 268

Death Penalty Prosecution and Punishment

Philadelphia were substantially less likely to receive the death penalty than cases represented by public defenders in the rest of the state.

Local Variation Death penalty sentences and executions are highly localized. In 2017, 31% of death sentences came from three US counties (Riverside, CA, Clark, NV, and Maricopa, AZ). Further, the 23 executions carried out in 2017 were in just eight states (Texas, Arkansas, Alabama, Florida, Ohio, Virginia, Georgia, and Missouri). Nor is this localization of capital punishment new (Smith, 2012). Only 2% of US counties have accounted for 56% of death sentences since 1976, and all inmates on death rows as of 2013 came from 20% of US counties (Death Penalty Information Center, 2013; Smith, 2012). There is ample empirical support for the proposition that prosecutorial and court decisions regarding the death penalty vary substantially between local court jurisdictions. Some studies suggest that the population size of the court jurisdiction affects the prosecutorial decision to seek the death penalty, but findings are contradictory across the studies. Songer and Unah (2006) found that prosecutors in rural judicial districts in South Carolina were much more likely than those in urban and suburban districts to file for the death penalty. By contrast, Poveda (2006) found that prosecutors in smaller Virginia counties were the least likely to seek the death penalty. In Maryland, Paternoster and Brame (2008) found that prosecutors were much more likely to seek the death penalty in suburban counties than in inner cities. Johnson et al. (2012) noted that, in Delaware, the largest of the three counties accounted for 59% of death sentences in the aggregate. However, one small Delaware county stood out as having more than double the odds of giving a death sentence compared to other counties. A similar study by Hans et al. (2015) also considered “heinousness” in their study comparing judge and jury decision-​making in Delaware in 146 capital cases from 1977–​2007. Hans et al. (2015) found that the odds of a given case receiving the death sentence varied greatly between counties. Donohue (2014) argued that differences in Connecticut county death penalty decisions were due more to prosecutorial ideology than county size. Donohue (2014) found that one county in Connecticut where the prosecutor vehemently supported the death penalty stood out as having a distinctively higher rate of seeking and sustaining death sentences than the rest of the state. Finally, the study of first-​degree murder convictions in 18 Pennsylvania counties (from 2000–​2010) by Kramer et al. (2017) found impressive differences between counties in each of the death penalty outcomes examined. Counties differed in terms of overall likelihood of a prosecutor filing or retracting a death penalty motion. For example, prosecutors in Allegheny County (Pittsburgh) were much less likely to seek the death penalty than prosecutors in the other 17 counties. On the other hand, prosecutors in Philadelphia were much more likely to retract the death penalty than were prosecutors in the other counties. Counties also differed in the effects of defendant and victim race/​ethnicity, and in the effects of type of legal representation, on prosecutor decisions. For example, prosecutors in Allegheny County and Philadelphia were less likely than prosecutors in other counties to seek the death penalty against defendants represented by public defenders. Also, courts in Philadelphia and Allegheny counties were less likely to impose death sentences than were courts in the other counties. Indeed, differences among counties in death penalty outcomes and the effects of other variables on death penalty outcomes were the largest and most prominent differences found in this study.

Methodological Issues Different methods of statistical analysis have sometimes produced contrasting findings regarding race and the death penalty. The major types of analysis found in the literature on disparity in the death penalty over the past 40 years are logistic regression and propensity score methods (see the review by Paternoster, & Brame, 2008). Studies of the death penalty in Maryland and North Carolina provide examples of how these methods can lead to contrasting findings. In their initial analysis using 269

Jeffery T. Ulmer and Lily Hanrath

standard logistic regression, Paternoster and Brame (2008) found a victim race effect in prosecutors’ decisions to file death penalty motions. However, the analysis by Berk et al. (2005) of these same data, using a different statistical approach, did not find such a victim race effect. Paternoster et al. (2008) subsequently re-​analyzed their own data using propensity score weighting and found that prosecutors were 2.3 times more likely to file death penalty motions in cases with white victims than in cases with black victims, thereby confirming their original findings. Earlier, we noted the re-​analysis by Jennings et  al. (2014) of the North Carolina death penalty data, which used a propensity score matching approach.While the original North Carolina analysis by Unah (2011) used traditional logistic regression analysis and found a white victim effect, the re-​analysis by Jennings et al., using propensity score matching analysis, did not find such an effect. Advocates of propensity score methods (PSM) for analyzing death penalty outcomes (e.g., Paternoster & Brame, 2008; Schonlau, 2006) point out that PSM indirectly helps address omitted variable bias due to selection bias by balancing out preexisting differences, thus statistically approximating randomization in an experiment (Apel & Sweeten, 2010; Li et al., 2013). In situations where selection bias might exist, the treatment and control groups might have unequal likelihoods of being selected into the data, and/​or exposed to the outcome of interest. For example, black, white, and Hispanic defendants might differ considerably on many of the important predictors or controls/​ confounds, such as aggravating or mitigating offense circumstances, concurrent convictions, and case characteristics. In other words, these groups might be imbalanced on these variables, and this might be due to selection bias. We cannot directly eliminate such selection bias, but we can try to make “treatment and control” cases as similar, or balanced, as possible on known covariates in the data. PSM thus attempts to replicate experimental design statistically, and, thus, attempts to address such limitations as covariate imbalance, selection bias, and omitted variable bias (Apel & Sweeten, 2010; Li et al., 2013). Propensity score methods attempt to make “treatment” (the comparison category of interest, for example, black defendants) and “control” groups (the group with which the treatment group is compared) similar or “balanced” on known covariates (control variables) and have similar error variance. Propensity score methods may be the best approach to correct for these problems, short of an actual randomized control trial, which clearly would be impossible and unethical in the context of death penalty research. The drawbacks of PSM for analyzing death penalty outcomes include the difficulty of directly examining the effects of multiple predictors at once, which is easily done in logistic regression. Furthermore, it is difficult to examine multilevel effects (i.e.., court or context-​level variables along with individual case or offender variables) on death penalty outcomes. The statistics literature does not provide definitive guidance on how to address multilevel data with propensity score methods, but various options exist (see Li et al., 2013). According to Li et al. (2013), researchers can control for the nesting of cases within larger groupings (like counties or courts) in producing the propensity score (i.e., as a variable in the model that estimates the propensity score). Alternately, Li et al. (2013) describe inverse propensity score weighted regression as an option for examining differences between groupings of cases in PSM. In other words, to address differences among counties in the race and death penalty comparisons, for example, one could estimate inverse propensity weighted logistic regressions that: 1) weight the cases according to their propensity scores that attempt to balance the race groups, and 2) include county dummy variables as predictors of the odds of the various death penalty outcomes in a logistic regression after propensity score weighting. The study of capital cases in Pennsylvania by Kramer et al. (2017) provides an example of this strategy in uncovering between-​ county differences. While past findings have not always been consistent, there are a few overarching patterns that appear in most analyses. In general, death penalty outcomes are more severe for cases involving white victims. Several studies also show that this is especially true of cases involving black defendants and white victims. Socioeconomic status appears to influence the likelihood of a death sentence indirectly, through legal representation. Though research is limited, when defendants have defense attorneys 270

Death Penalty Prosecution and Punishment

with greater capital case competence/​expertise, time, and resources, this appears to greatly reduce the chances of the death penalty being sought or given. Finally, characteristics of individual courts influence outcomes; though whether this is due to county size, prosecutorial ideology, or other factors is still in contention. Despite these consistent findings, the current field of research leaves many gaps, as we describe below.

Theoretical Frameworks As Spohn (2015) characterizes the contemporary study of racial disparity in sentencing, Most researchers now acknowledge that it is overly simplistic to ask whether race and ethnicity matter at sentencing. The more interesting questions … revolve around contexts in which and circumstances under which race and ethnicity influence sentencing and the ways in which disparities accumulate throughout the life course of a criminal case. (p. 78). The combination of the focal concerns and court communities’ perspectives are aimed at exactly these questions. The focal concerns perspective is a predominant framework in the social science literature on sentencing, and criminal justice decision-​making more generally (Steffensmeier, Ulmer, & Kramer, 1998; Kramer & Ulmer, 2009; see review by Ulmer, 2012). It has also recently been applied to studying the effect of race on death penalty decision-​making (Jennings et al., 2014; Kramer et al., 2017). The focal concerns perspective is not a set of hypotheses itself, but rather a heuristic conceptual framework from which research questions/​hypotheses can be derived (Ulmer, 2012). The focal concerns perspective argues that punishment decisions center around three focal concerns: blameworthiness (culpability), protection of the community/​offender dangerousness, and practical consequences and constraints of decisions. The focal concerns perspective argues that both legal and extra-​legal considerations can affect the assessment of defendants and cases in terms of the three focal concerns. But status-​linked attributions and stereotypes can sometimes shape court decision makers’ assessments of defendant blameworthiness, dangerousness/​rehabilitative potential, and/​or practical contingencies and constraints, though they likely do so secondarily to legally relevant factors (Kramer, & Ulmer, 2009). The focal concerns perspective implies, first, that the influence of social statuses like race/​ethnicity may be conditional on other defendant characteristics such as defendant gender and age (Steffensmeier et  al., 1998, 2017), or perhaps defendant criminal history or offense characteristics (Auerhahn, 2007). For murder, the characteristics of victims may also condition different definitions of blameworthiness or community protection (Curry et al., 2004; Curry, 2010) and may entail differing practical constraints or consequences. Prosecutorial decisions as to whether to seek, or retract, the death penalty, as well as court (judge or jury) decisions whether to sentence a defendant to the death penalty certainly consider the blameworthiness/​culpability of the defendant and their potential dangerousness to the community. Indeed, aggravating circumstances that must be specified by death penalty statutes legally codify elements of heightened blameworthiness or dangerousness. It may also be that practical issues such as avoiding the costs of trials and appeals or ensuring convictions by accepting pleas to lesser offenses or lesser penalties are considered in such serious cases as death-​eligible murders. Some scholars have posited that certain ostensibly race-​neutral factors, such as defendant criminal history, may mobilize racial and/​ or gender implicit bias in prosecutors’ decisions to seek the death penalty (Lynch, 2013). Prosecutors, judges, and juries may also see some defendants or especially victims as more sympathetic and worthy of empathy, which may influence their decisions (Mangino & Champion, 2013; Lynch, 2013). Anecdotally, Kramer et al. (2017) described a Pennsylvania District Attorney saying that the decision to seek the death penalty was influenced by whether the case involved a “savory” victim, such 271

Jeffery T. Ulmer and Lily Hanrath

as a child, versus an “unsavory” victim, such as a drug dealer. Another prosecutor volunteered that “I know within ten minutes of visiting the crime scene whether I will file for the death penalty,” indicating that his impressions of how emotionally impactful and heinous the crime scene/​murder incident appeared and the perceived egregiousness/​“heinousness” of the murder (e.g., whether there were unintended victims, torture of the victims) determined his decision to seek the death penalty. These factors all likely inform interpretations of the focal concern of blameworthiness, and also perhaps community protection from especially dangerous or predatory offenders. Prosecutors’ personal and political views towards punishment might also play a role in their decision-​making. Prosecutors may also be influenced by their perceptions of evidence strength and other pragmatic features of cases. District Attorneys likely want to rationally maximize their chances to achieve their desired outcomes, and thus act with practical and strategic intentionality in their decisions to file for death in any given case. This all speaks to the focal concern of practical constraints and consequences. Moreover, prosecutors may be pressured by local government officials to be more sparing about when to pursue the death penalty, given the high financial costs of doing so and the strains placed on county budgets. The role of county criminal justice resources thus also relates to practical constraints on death penalty decisions. Later extensions of the focal concerns perspective combine it with the court communities’ perspective and emphasize that the interpretation, prioritization, and indicators of the focal concerns are embedded in and conditioned by local court contexts (Kramer, & Ulmer, 2009).This implies that the effects of key predictors of case processing and sentencing outcomes likely differ between courts. For example, the effects of legally relevant factors such as offense characteristics or criminal history, as well as extra-​legal characteristics such as race/​ethnicity or case processing factors such as mode of conviction (i.e., plea bargaining, trial penalties) or type of defense representation may vary between courts.

Court Communities: Local Court Norms and Socio-​political Influences According to Baumer and Martin (2013), “prosecutors, judges, and juries are products of the broader social and political systems in which they work … there has been a growing recognition that criminal justice decision makers also may be sensitive to broader community conditions in making such assessments” (p. 135). This local variation is equally important, yet understudied, for decisions surrounding the death penalty. According to Paternoster (2011), “there is at least as much if not more variation in the inclination to impose the death penalty within death penalty states as there is between death penalty states (emphasis in original)” (p. 11). The emphasis on variation between local courts has been a feature of the general literature on courts and sentencing for decades. Prominently, the court communities’ perspective depicts courts as “communities” based on participants’ shared workplace, interdependent working relations between key sponsoring agencies such as the prosecutors’ office, judges’ bench, and defense bar, and the court’s relation to its larger socio-​political environment (Eisenstein et al., 1988). Local courts develop distinctive formal and informal case processing and sentencing norms (see Eisenstein et al., 1988; Ulmer, 1997; Ulmer & Kramer, 1998; Ulmer, 2005).This literature argues that the reliance on focal concerns tend to characterize courts and criminal case-​processing decisions generally, but the meaning, relative emphasis and priority, and interpretation of focal concerns is shaped by local court culture (e.g., Ulmer & Johnson, 2017; Kramer & Ulmer, 2009; Spohn & Fornango, 2009; Anderson & Spohn, 2010). Focal concerns interpretations in decisions surrounding capital murder cases would therefore differ according to local court norms, formal and informal prosecutorial policies, and criminal justice resource constraints. For example, prosecutors in a particular county might not seek the death penalty at all, or only seek it in cases seen as most severe or heinous. In other counties, the death penalty might be routinely sought in any first-​degree murder cases where an aggravating factor might be argued, but then retract most of those filings as a plea agreement concession. Or, prosecutors might seek the death penalty in any case where an aggravating factor might be argued and leave the penalty 272

Death Penalty Prosecution and Punishment

decision to the judge or jury. As local prosecutors, judges, and/​or juries weigh their perceptions of the “heinousness” of murders, or the “savoriness” of victims, it is possible that those perceptions would be shaped by defendant or victim race in differing ways. Also, varying county criminal justice resources may also lead to varying practical constraints on death penalty decisions. In some counties, prosecutors might be pressured by local government officials to be sparing about when to pursue the death penalty, given the high financial costs involved for county court budgets. Likewise, as prosecutors and courts prioritize the application of the death penalty based on limited county resources, defendant or especially victim race might differentially shape how they weigh cases as deserving the ultimate punishment. Thus, guidance from the focal concerns perspective on sentencing suggests mechanisms by which the race of defendants and victims might affect death penalty decision-​making (singly or in combination). In addition, the court community perspective implies that death penalty decisions will differ, perhaps substantially, between courts, and the effects of the race/​ethnicity of defendants and/​or victims on those decisions might vary as well. Importantly, this means that we expect death penalty decisions to be made relying on these focal concerns as they are filtered through the lenses of local decision makers and the court community within which they function. These focal concerns interpretations in decisions surrounding capital murder cases would differ according to court community organization, culture, routine practices, and local political conditions and criminal justice resource constraints. Major differences between counties in murder case charging, conviction, and sentencing outcomes would therefore be expected.

New Research Directions Overall, the literature on capital murder case processing and sentencing discussed earlier shows that several areas are relatively under-​researched. Below, we describe new research directions involving prosecutorial discretion, defendant and victim characteristics, defense representation, court socio-​ political contexts, and qualitative research.

Prosecutorial Discretion First, as numerous scholars have pointed out (Ulmer, 2012; Bushway & Forst, 2013; Spohn, 2015), relatively little research focuses on the exercise of prosecutorial discretion, and this is also true of research on death penalty outcomes and broader homicide case processing (Paternoster, 2011). It is particularly important to identify the stages and discretionary decisions in which racial or other disparities may appear. More than four decades ago, Hagan (1974) called for research on where disparities might appear at different stages of the criminal justice system. As Spohn’s (2015) assessment of recent sentencing literature argues, even the most high-​quality studies of sentencing (death penalty or otherwise) are unable to explain “whether disparate treatment was found only at sentencing or accumulated as cases moved through the court process, or whether disparities resulted from decisions made by prosecutors as well as judges” (p. 76). Given their importance in the context of the death penalty, there is a need for more research examining disparity in prosecutorial decisions to charge defendants with capital crimes, to engage in plea bargaining that takes the death penalty off the table, to file a notice to seek the death penalty, and to retract the death penalty option after notice has been filed. In addition, practical constraints and consequences are expected to influence outcomes in murder cases. In the interests of court organizational efficiency and uncertainty reduction, prosecutors and judges may reward people who plead guilty and avoid time-​and resource-​ intensive and uncertain trials. Per the focal concerns perspective, pleading guilty can be also seen as an indicator or remorse, and thus reduced blameworthiness (Kramer & Ulmer, 2009). In short, there is less research on prosecutorial decision-​making in capital cases than on capital sentencing decisions, and future research should correct this imbalance. 273

Jeffery T. Ulmer and Lily Hanrath

Defendant and Victim Characteristics Second, compared to the many studies examining capital murder defendant and victim race, there are relatively fewer studies examining the effect of victim gender and other social statuses, such as social class, age, or education. A  key implication of the focal concerns perspective is that particular constellations of defendant and victim social statuses will likely interact to condition interpretations of focal concerns in capital murder cases. Research should continue to investigate the role of defendant and victim race and ethnicity in prosecutorial charging decisions and decisions to seek the death penalty, as well as capital sentencing decisions. In addition, however, research should apply more focus to other social statuses that might affect the consideration of focal concern and influence capital case processing and sentencing decisions. For example, perhaps defendants accused of killing women, children, or elderly victims are seen as more blameworthy than others, and therefore will receive more severe outcomes. Research should also tease out the direct and indirect ways that socioeconomic status affects capital case processing and outcomes, especially via the quality of defense representation.

Defense Attorneys Third, there are surprisingly few studies comparing different types of defense representation relative to death penalty case processing and sentencing. Type of defense attorney representation has focal concerns implications, as well. On one hand, defendants with private attorneys may be advantaged in that private attorneys may provide more vigorous adversarial counsel, raising more mitigating circumstances that reduce perceptions of blameworthiness, challenging prosecutorial evidence (thus presenting more practical constraints to prosecutorial efforts), etc. On the other hand, defendants with public defenders might have advantages. Public defenders might have substantial insider knowledge as members of the court community (see below), know more ways to negotiate favorable plea bargains, and/​or ways to influence judges or juries to give more lenient sentences. Thus, research should assess two competing propositions. On one hand, because they might provide more vigorous adversarial efforts at mitigating blameworthiness and throw up more practical challenges for prosecutors, private attorneys might provide more favorable outcomes for murder defendants. On the other hand, because they are court community “insiders” and “repeat players,” public defenders might provide more favorable outcomes for murder defendants.

Court Socio-​Political Contexts Fourth, very few studies examine the role of court socio-​political contexts on death penalty decision-​ making.Yet a major theme in the social science literature on sentencing in general is that important differences exist between local courts in sentencing severity, and in the effects of variables like race and ethnicity (see the review by Ulmer, 2012). As discussed earlier, some studies have found considerable variation between counties in death penalty outcomes. However, the decisional context and decision-​making processes of these county-​level actors (most importantly, the prosecutor) remain largely a black box. We submit that any full understanding of the administration of capital punishment is not possible until this local decisional context is more fully mapped out. As noted by LeChance (2016), “Regional variations in customs, statutes, and political structures all imposed significant limitations in trying to tell a national story about the death penalty” (p. 21).This topic is very understudied in terms of murder (capital and non-​capital) in general (see Baumer & Martin, 2013). If the court communities’ perspective directs our attention to the embeddedness of punishment decisions in the contexts of local courts, recent research directs our attention to some key sets of socio-​political contextual factors that may be important in conditioning differences in capital murder case processing and sentencing. First, research has shown that organizational characteristics of the local court community may be important in conditioning differences in case processing and sentencing, 274

Death Penalty Prosecution and Punishment

including the partisan balance of the judges’ bench and prosecutors’ office, social characteristics of judges and prosecutors, and court caseload characteristics (e.g., Johnson, 2006; Wooldredge, 2010; Anderson & Spohn, 2010; King et al., 2010). Second, the political and religious conservatism of the surrounding county may shape the punitiveness of court decisions (Ulmer et al., 2008), especially so for murder (Baumer & Martin, 2013). It is possible that more politically and religiously conservative counties might foster more a greater emphasis on moral blameworthiness (see Baumer & Martin, 2013; Ulmer et  al., 2008), and more punitive use of prosecutorial, judicial, and jury discretion in capital murder cases. Third, responses to crime are indeed influenced by actual crime rates, especially violent crime (Garland, 2001, 2010). Higher violent crime rates tend to mobilize greater public fear and calls for more punitive court action (Garland, 2001). County violent crime rates might be an important factor producing variation in capital murder case processing and sentencing. Furthermore, an important literature has examined racial threat in relation to criminal punishment (see reviews by Ulmer, 2012; Baumer, 2013; Spohn, 2015). Such racial threat may be especially salient and influential in relation to serious murder cases and their handling by courts, given that much public discourse about violent crime is racialized (Bonilla-​Silva, 2014), and public and political discourse about the efficacy of the death penalty is deeply divided by racial divisions (Baumer & Martin, 2013).Thus, local racial or ethnic composition may condition punitiveness in court responses to murder, and may condition the effects of individual defendant and victim race (singly and in combination) on murder case outcomes

Beyond Statistical Analyses Finally, there is a dearth of qualitative research examining courtroom workgroup decision-​making for capital murder and court community norms surrounding the decision-​making process. That is, very few studies qualitatively examine how individual prosecutors and their offices make decisions in capital cases, how prosecutors and defense attorneys engage in plea bargaining surrounding aggravating circumstances and the decision to seek the death penalty, and the involvement and perspectives of judges in capital cases. Spohn (2015) declares: There also is a need for qualitative research designed to identify the factors that motivate prosecutors and judges to treat blacks and Hispanics, especially young men, more harshly than Whites.The focal concerns perspective, the most common theoretical perspective used by contemporary sentencing researchers, suggests that harsher treatment results from stereotypes of racial minorities as more dangerous, more culpable, and less amenable to rehabilitation … What is needed is qualitative research designed to document how prosecutors and judges view certain types of defendants and how these perceptions influence their decisions. (p. 91) Future research should answer these calls by examining not only statistical data on murder case outcomes, but qualitative data from interviews of key court community actors on murder case decisions and their embeddedness in local court community norms and contexts. More specifically, the categories of factors below would likely inform the interpretation of the focal concerns, or reflect court community norms, influences, and constraints. They would all be valuable foci for questions in qualitative, in-​depth interviews with prosecutors, judges, defense attorneys, and even jurors. For example, how do criminal history, social status characteristics, resources, motivation, cooperation, demeanor, and other factors influence the perceived blameworthiness and/​or perceived risk and dangerousness of the defendant? How do the characteristics of victims (e.g., what constitutes a “savory” or sympathetic victim) influence the perceived blameworthiness of defendants? Regarding court communities, what formal or informal prosecutorial policies guide capital case decisions? What local norms exist concerning plea bargaining in murder cases (e.g., prevalence of charge reductions 275

Jeffery T. Ulmer and Lily Hanrath

or sentence bargaining). What are practices regarding retracting death penalty filings, especially in exchange for guilty pleas? Lastly, what organizational constraints do prosecutors and/​or judges face in the capital decision-​making process, such as concerns about the costs of capital trials, and do these influence the decisions to seek and impose the death penalty? What are community standards and expectations surrounding punishment and what impact do these have on the decision to seek the death penalty? These are also potential practical constraints and consequences, and potentially important ways that external environments impinge on court community decisions.

Conclusion While both death sentences and executions have been on the decline in recent years, the death penalty continues to have a strong symbolic and practical impact on the US criminal justice system and to reflect racial and socioeconomic inequalities in the broader society. The continued societal impact of the death penalty and its relationship to larger social inequalities is seen in the US Supreme Court’s ongoing attention. Most relevant to our review here, Justice Blackmun’s deep concern with racial and socioeconomic disparity connected to the death penalty is evident from his dissent in Callins v. Collins. More recently, Justice Breyer’s dissents in Glossip v. Gross and Jordan/​Evans v. Mississippi argue that wide geographic disparity in the application of the death penalty constitutes arbitrariness that is barred by Furman v. Georgia. Research on disparity and variation in capital murder case processing and death penalty decisions should therefore continue and should strive to give more contemporary and definitive answers to the research questions posed above. Research on capital murder and death penalty decision-​making also presents an opportunity to delve deeper into important theoretical themes. As we observed earlier, research on death penalty prosecution and sentencing has largely proceeded apart from broader theorizing about courts and sentencing.This is unfortunate, because the study of death penalty decision-​ making presents a valuable chance to apply, test, and extend the focal concerns theoretical perspective and the court communities model, among other theoretical ideas from the literature on courts and criminal punishment. By the same token, the study of the death penalty would benefit from insight from broader theories of sentencing and criminal justice decision-​making. Finally, we recommend more sustained empirical and policy attention to variation between locales in death penalty decision processes and punishment outcomes, and the implications of that variation. Such variation is important because, if capital murder case outcomes are strongly shaped by where defendants are convicted and sentenced, this invokes policy and legal concerns of fairness and arbitrariness. Critics such as Justice Breyer would see place-​based variation in capital murder case processing and sentencing as unwarranted disparity that resembles the kind of “arbitrariness” that the US Supreme Court decried in Furman v. Georgia (1972), and that the guided discretion statutes approved by the Court in Gregg v. Georgia (1976) were designed to reduce. However, others might argue that principles of democracy, local autonomy, and decentralized government mandate the ability of local jurisdictions to fashion punishments as they see fit, at least within broader legal parameters. Clearly, capital case processing and sentencing present a perennial dilemma in which key principles of local communities’ autonomy clash with the principles of uniformity and equality.

Notes 1 Three dissenting opinions from Justices Brennan, Marshall, Blackmun, and Stevens argued that the burden of proof should have been to show there was a risk of disparity rather than the proven fact it occurred in an individual case. Brennan wrote that the Baldus study “reveals that the risk that race influenced McKleskey’s sentence is intolerable by any imaginable standard.” 2 For example, see Donohue’s (2014) description of Connecticut’s statutory aggravating circumstances, and Paternoster et al.’s (2004) description of Maryland’s.

276

Death Penalty Prosecution and Punishment 3 Hidalgo v. Arizona illustrates how the use of aggravating circumstances as justifications for death eligibility may become problematic when aggravators have broad definitions. Looking at first-​degree murder cases in Maricopa County Arizona, the petitioner found that 856 out of 866 of cases were eligible for the death penalty under Arizona’s statutory aggravating circumstances (Hidalgo v. Arizona, 583 U.S. [2018]).

References Anderson, J.M., & Heaton, P. (2012). How much difference does the lawyer make: The effect of defense counsel on murder case outcomes. Yale Law Journal, 122(1), 154–​217. Anderson, A., & Spohn, C. (2010). Lawlessness in the federal sentencing process: A test for uniformity and consistency in sentencing practices. Justice Quarterly, 27(3), 362–​393. Apel, R., & Sweeten, G. (2010). Propensity score matching in criminology and criminal justice. In A. Piquero & D. Weisburd (eds.), Handbook of Quantitative Criminology (pp. 543–​562). New York: Springer. Auerhahn, K. (2007). Just another crime? Examining disparity in homicide sentencing. The Sociological Quarterly, 48(2), 277–​316. Baldus, D.C., Pulaski, C.D., & Woodworth, G. (1983). Comparative review of death sentences: An empirical study of the Georgia experience. Journal of Criminal Law and Criminology, 74(3), 661–​753. Baldus, D.C.,Woodworth, G., & Pulaski, C. Jr. (1990). Equal Justice and the Death Penalty. Boston, MA: Northeastern University Press. Baldus, D.C., Woodworth, G., Zuckerman, D., & Weiner, N.A. (1997–​98). Racial discrimination and the death penalty in the post-​Furman era: An empirical and legal overview with recent findings from Philadelphia. Cornell Law Review, 83, 1638–​1770. Baldus, D.C.,Woodworth, G., Grosso, C.M., & Christ, A.M. (2002–​2003). Arbitrariness and discrimination in the administration of the death penalty: A legal and empirical analysis of the Nebraska experience (1973–​1999). Nebraska Law Review, 81(2), 486–​756. Banner, S. (2002). The Death Penalty: An American History. Cambridge, MA: Harvard University Press. Baumer, E. (2013). Reassessing and redirecting research on race and sentencing. Justice Quarterly, 30(2), 231–​261. Baumer, E., & Martin, K. (2013). Social organization, collective sentiment, and legal sanctions in murder cases. American Journal of Sociology, 119(1), 131–​182. Behnken, M.P., Caudill, J.W., Berg, M.T., Trulson, C.R., & DeLisi, M. (2011). Marked for death: An empirical criminal careers analysis of death sentencing in a sample of convicted male homicide offenders. Journal of Criminal Justice, 39(6), 471–​478. Berk, R.A., Li, A., & Hickman, L.J. (2005). Statistical difficulties in determining the role of race in capital cases: A re-​analysis of data from the state of Maryland. Journal of Quantitative Criminology, 21(4), 365–​390. Blankenship, M.B., & Blevins, K.R. (2001). Inequalities in capital punishment in Tennessee based on race: An analytical study of aggravating and mitigating factors in death penalty cases. The University of Memphis Law Review, 31(4), 823–​973. Bonilla-​Silva, E. (2014). Racism Without Racists: Color Blind Racism and the Persistence of Racial Inequality in America. Lanham, MD: Rowman and Littlefield. Bowers, W.J., & Pierce, G.L. (1980). Arbitrariness and discrimination under the post-​Furman capital statutes. Crime and Delinquency, 26(4), 563–​635. Bright, S.B. (1994). Counsel for the poor: The death sentence not for the worst crime but for the worst lawyer. Yale Law Journal, 103(7), 1835–​1883. Brock, D.E., Sorensen, J., & Marquart, J.W. (1999). Racial disparities in capital punishment in Texas after Penry. The Justice Professional, 12(2), 159–​172. Curry, T. (2010). The conditional effects of victim and offender ethnicity and victim gender on sentences for non-​capital cases. Punishment and Society, 12(4), 438–​462. Curry, T., Lee, G., & Fernando Rodriguez, S. (2004). Does victim gender increase sentence severity? Further explorations of gender dynamics and sentencing outcomes. Crime and Delinquency, 50(3), 319–​343. Death Penalty Information Center. (2013). The 2% Death Penalty: How a Minority of Counties Produce Most Death Cases at Enormous Cost to All. Washington, DC: Author. Retrieved from: https://​deathpenaltyinfo.org/​ documents/​TwoPercentReport.pdf —​—​. (2017). The Death Penalty in 2017: Year End Report. Washington, DC: Author. Retrieved from: https://​ deathpenaltyinfo.org/​documents/​2017YrEnd.pdf —​—​. (2018). Facts about the Death Penalty. Washington, DC: Author. Retrieved from: https://​deathpenaltyinfo. org/​documents/​FactSheet.pdf Donohue, J.J. III. (2014). An empirical evaluation of the Connecticut death penalty system since 1973: Are there unlawful racial, gender, and geographic disparities? Journal of Empirical Legal Studies, 11(4), 637–​696.

277

Jeffery T. Ulmer and Lily Hanrath Doyle, J.M. (2013). The lawyers’ art: “Representation” in capital cases. Yale Journal of Law & the Humanities, 8(2),  1–​34. Eisenstein, J., Flemming, R.B., & Nardulli, P.F. (1988). The Contours of Justice: Communities and Their Courts. Lanham, MD: University Press of America. Foglia, W.D. (2003). They know not what they do: Unguided and misguided discretion in Pennsylvania capital cases. Justice Quarterly, 20(1), 187–​211. Garland, D. (2001). The Culture of Control: Crime and Social Order in Contemporary Society. Chicago, IL: University of Chicago Press. —​—​ . (2010). Peculiar Institution: America’s Death Penalty in an Age of Abolition. Cambridge, MA: Harvard University Press. General Accountability Office. (1990). Death Penalty Sentencing: Research Indicates Pattern of Racial Disparities. Washington, DC: Author. Gross, S.R., & Mauro, R. (1984). Patterns of death: An analysis of racial disparities in capital sentencing and homicide victimization. Stanford Law Review, 37(1), 27–​153. Hagan, J. (1974). Parameters of criminal prosecution: An application of path analysis to a problem of criminal justice. Journal of Criminal Law and Criminology, 65(4), 536–​544. Hans, V.P., Blume, J.H., Eisenberg, T., Hritz, A.C., Johnson, S.L., Royer, C.E., & Wells, M.T. (2015). The death penalty: Should the judge or the jury decides who dies? Journal of Empirical Legal Studies, 12(1),  70–​99. Hindson, S., Potter, H., & Radelet, M.L. (2006). Race, gender, region and death sentencing in Colorado, 1980–​ 1999. University of Colorado Law Review, 77, 549–​594. Holcomb, J.,Williams, M., & Demuth, S. (2004).White female victims and death penalty disparity research. Justice Quarterly, 21(4), 877–​902. Jennings, W.G., Richards, T.N., Smith, M.S., Bjerregaard, B., & Fogel, S.J. (2014). A critical examination of the ‘white victim effect’ and death penalty decision-​making from a propensity score matching approach: The North Carolina experience. Journal of Criminal Justice, 42(5), 384–​398. Johnson, B.D. (2006). The multilevel context of criminal sentencing: Integrating judge and county-​ level influences. Criminology 44(2), 259–​298. Johnson, S.L., Blume, J.H., Eisenberg,T., Hans,V.P., & Wells, M.T. (2012).The Delaware death penalty: An empirical study. Iowa Law Review, 97,  1–​41. Keil, T.J., & Vito, G.F. (1995). Race and the death penalty in Kentucky murder trials: An analysis of post-​Gregg outcomes. Justice Quarterly, 19(4), 633–​672. Kramer, J.H., & Ulmer, J.T. (2009). Sentencing Guidelines: Lessons from Pennsylvania. Boulder, CO: Lynne Rienner Publishers. Kramer, J.H., Ulmer, J.T., & Zajac, G. (2017). Capital Punishment Decisions in Pennsylvania, 2000–​2010: Implications for Racial, Ethnic and Other Disparate Impacts. Harrisburg, PA: Pennsylvania Interbranch Commission on Gender, Racial, and Ethnic Fairness. Retrieved from: www.justicecenter.psu.edu/​research/​projects/​files/​ the-​administration-​of-​the-​death-​penalty-​in-​pennsylvania-​pdf LeChance, D. (2016). Executing Freedom: The Cultural Life of Capital Punishment in the United States. Chicago, IL: The University of Chicago Press. Lenza, M., Keys, D., & Guess, T. (2005). The prevailing injustices in application of the Missouri death penalty (1978 to 1996). Social Justice, 32(2), 151–​166. Li, F., Zaslavsky, A.M., & Landrum, M.B. (2013). Propensity score weighting with multilevel data. Statistics in Medicine, 32(19), 3373–​3387. Lynch, M. (2013). Institutionalizing bias: The death penalty, federal drug prosecutions, and mechanisms of disparate punishment. American Journal of Criminal Law, 41(1), 91–​131. Mangino, M.T., & Champion, D.R. (2013). Prosecutorial discretion and the death penalty: An integral perspective. Journal of Criminology and Sociology, 2, 426–​437. Marshall,T. (1986). Remarks on the death penalty made at the judicial conference of the second circuit. Columbia Law Review, 86(1),  1–​8. Ohio Joint Task Force to Review the Death Penalty. (2014). Final Report and Recommendations. Columbus, OH: Ohio State Supreme Court. Retrieved from: www.supremecourt.ohio.gov/​Boards/​deathPenalty/​resources/​ finalReport.pdf Paternoster, R. (1983). Race of victim and location of crime: The decision to seek the death penalty in South Carolina. The Journal of Criminal Law and Criminology, 74(3), 754–​785. —​—​. (1984). Prosecutorial discretion in requesting the death penalty: The case of victim-​based discrimination. Law and Society Review, 18(3), 437–​478. —​—​. (2011). Capital punishment. In M. Tonry (ed.), The Oxford Handbook of Crime and Criminal Justice (pp. 1–​22). New York: Oxford University Press.

278

Death Penalty Prosecution and Punishment Paternoster, R. & Brame, R. (2003). An Empirical Analysis of Maryland’s Death Sentence System with Respect to the Influence of Race and Legal Jurisdiction. College Park, MD: Department of Criminology, University of Maryland. Paternoster, R., & Brame, R. (2008). Reassessing race disparities in Maryland capital cases. Criminology, 46(4), 971–​1008. Paternoster, R., Brame, R., Bacon, S., & Ditchfield, A. (2004). Justice by geography and race: The administration of the death penalty in Maryland, 1978–​1999. Margins:The University of Maryland Law Journal of Race, Religion, Gender, and Class, 4,  1–​98. Paternoster, R., Soltzman, L., Waldo, G. P., & Chiricos, T. G. (1983). Race of victim and location of crime: The decision to seek the death penalty in South Carolina. Journal of Criminal Law and Criminology, 74, 701–​731. Peterson, N., & Lynch, M. (2012). Prosecutorial discretion, hidden costs, and the death penalty: The case of Los Angeles county. The Journal of Criminal Law & Criminology, 102(4), 100–​143. Phillips, S. (2009a). Status disparities in the capital of capital punishment. Law and Society Review, 43(4), 807–​837. —​—​. (2009b). Legal disparities in the capital of capital punishment. Journal of Criminal Law and Criminology, 99(3), 717–​756. Poveda, T.G. (2006). Geographic location, death sentences and executions in post-​Furman Virginia. Punishment & Society, 8(4), 423–​442. Radelet, M.L., & Pierce, G.L. (1985). Race and prosecutorial discretion in homicide cases. Law & Society Review, 19(4), 587–​621. Richards, T.N., Smith, M.D., Jennings, W.G., Bjerregaard, B., & Fogel, S.J. (2014). An examination of defendant sex disparity in capital sentencing: A propensity score matching approach. American Journal of Criminal Justice, 39(4), 681–​697. Royer, C.E., Hritz, A.C., Hans,V.P.,Wisenberg,T.,Wells, M.T., Blume, J.H., & Johnson, S.L. (2014).Victim gender and the death penalty. Cornell Law Faculty Publications, 838, 429–​467. Schonlau, M. (2006). Charging decisions in death-​eligible federal cases (1995–​2000): Arbitrariness, capriciousness, and regional variation. In S.P. Klein, R.A. Berk, & L.J. Hickman (eds.), Race and the Decision to Seek the Death Penalty in Federal Cases (pp. 95–​129). Santa Monica, CA: RAND. Smith, R.J. (2012). The geography of the death penalty and its ramifications. Boston University Law Review, 92, 227–​289. Songer, M.J., & Unah, I. (2006). The effects of race, gender, and location on prosecutorial decisions to seek the death penalty in South Carolina. South Carolina Law Review, 58, 161–​210. Spohn, C. (2000).Thirty years of sentencing reform:The quest for a racially neutral sentencing process. In Policies, Processes and Decisions of the Criminal Justice System (Vol. 3, pp. 427–​501). Washington, DC: US Department of Justice. —​—​. (2015). Race, crime, and punishment in the twentieth and twenty-​first centuries. Crime and Justice, 44(1),  49–​97. Spohn, C., & Fornango, R. (2009). US Attorneys and substantial assistance departures:Testing for interprosecutor disparity. Criminology, 47(3), 813–​847. Steffensmeier, D., Ulmer, J.T., & Kramer, J. (1998).The interaction of race, gender, and age in criminal sentencing: The punishment cost of being young, black, and male. Criminology, 36(4), 763–​797. Steffensmeier, D., Painter-​Davis, N., & Ulmer, J.T. (2017). The intersectionality of race, ethnicity, gender, and age in criminal punishment. Sociological Perspectives, 60(4), 810–​833. Suni, E.Y. (1990). Capital punishment in Missouri: Recent developments in the interpretation and administration of the death penalty. UMKC Law Review, 58(4), 523–​580. Ulmer, J.T. (1997). Social Worlds of Sentencing: Court Communities under Sentencing Guidelines. Albany, NY: State University of New York Press. —​—​. (2005).The localized uses of federal sentencing guidelines in four U.S. district courts: Evidence of processual order. Symbolic Interaction, 28(2), 255–​279. —​—​. (2012). Recent developments and new directions in sentencing research. Justice Quarterly, 29(1),  1–​40. Ulmer, J.T., Bader, C., & Gault, M. (2008). Do moral communities play a role in criminal sentencing? Evidence from Pennsylvania. The Sociological Quarterly, 49(4), 737–​768. Ulmer, J.T., & Johnson, B.D. (2017). Organizational conformity and punishment: Federal court communities and judge-​initiated guideline departures. Journal of Criminal Law and Criminology, 107(2), 253–​292. Ulmer, J.T., & Kramer, J.H. (1998). The use and transformation of formal decision-​making criteria: Sentencing guidelines, organizational contexts, and case processing strategies. Social Problems, 45(2), 248–​267. Unah, I. (2011). Empirical analysis of race and the process of capital punishment in North Carolina. Michigan State Law Review, 610–​658. Vito, G.F., Higgins, G.E., & Vito, A.G. (2014). Capital sentencing in Kentucky, 2000–​2010. American Journal of Criminal Justice, 39(4), 753–​770.

279

Jeffery T. Ulmer and Lily Hanrath Williams, M.R., & Holcomb, J.E. (2001). Racial disparity and death sentences in Ohio. Journal of Criminal Justice, 29(3), 207–​218. Williams, M., Demuth, S., & Holcomb, J. (2007). Understanding the influence of victim gender in death penalty cases: The importance of victim race, sex-​related victimization, and jury decision-​making. Criminology, 45(4), 865–​892. Wooldredge, J. (2010). Judges’ unequal contributions to extralegal disparities in imprisonment. Criminology 48(2), 539–​567. Zatz, M. (2000). The convergence of race, ethnicity, gender, and class on court decision making: Looking toward the 21st Century. In Policies, Processes and Decisions of the Criminal Justice System (Vol. 3, pp. 503–​552). Washington, DC: US Department of Justice.

280

PART V

Current Controversies

13 RETHINKING THE ROLE OF CRIMINAL HISTORY IN SENTENCING Rhys Hester, Richard S. Frase, Julia Laskorunsky, and Kelly Lyn Mitchell1

Introduction An offender’s prior record is often considered a legitimate basis for increasing his or her sentence, with criminal history penalty enhancements treated as “legal characteristics.” These enhancements carry a strong intuitive appeal, and in a field concerned with whether illegitimate factors like race are associated with punishment, criminal history premiums have been taken as a given. Yet laws and policies that impose hefty increases in punishment on the basis of past crimes need careful rethinking. Criminal history enhancements, while perhaps intuitively appealing on first consideration, are coupled with a number of substantial deleterious collateral consequences (contributing, for example, to mass incarceration and racial disproportionality without contributing to public safety). In sentencing guidelines jurisdictions, the specifics of these policies vary greatly, suggesting a lack of cohesive agreement over the most appropriate approach to prior record premiums, and highlighting the ad hoc development of these policies in the first place. In this chapter we delve into these issues, reviewing recent research on criminal history enhancements and identifying important issues in need of further study. We focus on sentencing guidelines systems because criminal history enhancements are more visible (and probably more substantial) in those systems.2 Roberts (1997) once quoted Shakespeare in this context for the proposition that: “Men’s evil manners live in brass: Their virtues we write in water.” In the 20 years since Roberts’ (1997) seminal article on the subject, courts have imposed untold years of prison time on offenders attributable—​not to the nature or circumstances of a current offense—​but because of the criminal record a person brought with them into a courtroom.These criminal records live in perpetuity, and sentencing policies, particularly in sentencing guidelines jurisdictions, solidify past behavior by imposing steep increases in penalties on the basis of prior record, quite effectively etching past transgressions into criminal system policy outcomes. Offenders with extended prior records are disproportionately minority, and they also tend to be older. Having a record is a signal of increased likelihood of reoffending compared to a first-​time offender, but as repeat offenders accumulate extensive records and advance in age, they also enter the desistance process and become ever nearer to the end of their criminal careers. Robust criminal history enhancement policies largely ignore the aging-​out phenomenon, instead imposing increasing prison costs for declining public safety benefits. In recent years scholars have highlighted some of these problems associated with punitive criminal history policies (e.g., Frase et al., 2015; Hamilton, 2015; Hester et al., 2018). Researchers have 283

Hester et al.

also uncovered tremendous variation in the way jurisdictions use information about a person’s criminal history at sentencing: in some contexts, prior record can be the primary driver of a person’s sentence, while in others it might hardly count at all.The extensive variation in criminal history policies begs the question of purpose. As we review, despite the intuitive appeal, legal philosophers have struggled to articulate a compelling retributive justification for the kind of repeat offender aggravation policies found in guidelines states. As for utilitarian purposes of punishment like deterrence and incapacitation, the empirical evidence suggests that adding to already long prison sentences is a poor deterrent; it may even lead to increased criminality, thus lessening or negating any incapacitation benefits. Criminal history enhancements have been neglected among sentencing scholars, but thankfully that is changing. In this chapter we review the recent research on the subject and ultimately conclude that criminal history sentencing policies are in need of a careful rethinking. In particular, we encourage policymakers and independent researchers to engage in empirical assessments of their criminal history policies. We begin our treatment of criminal history with a section detailing the adverse consequences of prior record policies. Next, we provide an overview of the jurisdictional variation in policies and discuss the struggles to put forth a compelling justification for prior record enhancements. We then discuss the few examples of empirical evaluation of prior record policies from the federal system, Pennsylvania, and Minnesota. We conclude by highlighting the value of this sort of empirical research and discussing some unanswered questions and avenues for future inquiry.

Adverse Consequences, Variation, and Justifications We begin our treatment of criminal history with a discussion of some of the adverse consequences that flow from robust prior record premium policies. These collateral effects include that criminal history enhancements: 1. Strain prison bed prioritization and increase the size and expense of an already overburdened prison system through the incarceration of non-​violent offenders. 2. Confound offense-​based punishment proportionality by imposing more serious penalties on offenders guilty of less serious crimes on the basis of their past crimes. 3. Exacerbate racial disparities where, as appears to be universal in the United States, on average African-​Americans have more extensive prior records that whites. We expound on each of these and provide some examples of the extent to which prior record policies contribute to these problems. Given these collateral consequences, it seems imperative that prior record-​related polices be well thought out and justifiable pursuant to one of the commonly recognized theories of punishment. But prior record policies do not appear to be very well thought out. And justifying these policies on either retributive or utilitarian grounds proves more difficult than one might first imagine. Instead, prior record policies appear to be based not on careful reasoning or empirical research, but “gut intuition.” As a result, what counts towards a prior record score varies tremendously across jurisdictions, including what characteristics of a prior record are counted (e.g., misdemeanors, juvenile offenses, offenses committed while under community supervision) and how these items are weighted. How much a score enhances sentence severity (what we call the “magnitude” of prior record enhancements) also varies considerably, with prior record premiums ranging from less than 200% to well over 1,000%: in some jurisdictions one’s prior record can roughly double a person’s sentence recommendation while in others it increases it more than ten-​ fold. In this part we survey this variation and present the attempts and critiques to justify prior record premiums. 284

Criminal History in Sentencing

Adverse Consequences Criminal history can have a number of unintended and potentially undesirable consequences, and those consequences are more problematic the larger the magnitude of the enhancement (that is, the greater the impact of criminal history on sentencing decisions related to prison commitment and prison duration). This part summarizes the potential adverse impacts of these enhancements. Prior record scores identify many non-​violent offenders for increased punishment. This results in a prison population that is both larger and filled with a greater number of non-​violent offenders. Moreover, because of the correlation between offender race and criminal record, prior record scores also exacerbate racial disparities, sending more African-​Americans to prison. Here, we discuss four unintended adverse consequences of criminal history enhancements.

Undercutting Prison Use Priorities and Increasing the Size of Prison Sentencing guidelines permit policy makers to set priorities in the use of limited and expensive prison resources.These priorities usually reflect the ideals that prison should be reserved for offenders who commit the most serious offenses or who have the longest criminal records. In particular, guidelines systems are often constructed to give higher priority to incarceration of violent offenders, and to use community-​based sanctions for property and other non-​violent offenders. Consider Minnesota’s main sentencing guidelines grid, provided in Figure 13.1, as an example. The presumptive sentence for offenders who are convicted of a severity level 8 or higher offense is prison, regardless of the offender’s criminal history score. By contrast, the presumptive sentence for offenders who are convicted of an offense at severity level 1 is probation except for those who have very long criminal records, earning a criminal history score of 6. Criminal history works within this structure in two ways. At high-​severity levels, where prison is presumptive, criminal history serves to lengthen the prison stay. At low-​severity levels, where probation is presumptive, criminal history determines when an offender’s criminal record will cause the sentence to cross the dispositional line, converting a probation sentence to a prison sentence. Criminal history can undercut prison priorities by imprisoning more low-​level and non-​violent offenders. These offenders often have high recidivism rates and tend to accumulate higher criminal history scores (Langan & Levin, 2002). Criminal history also increases sentence length for offenders at all levels of offense severity, and the longer sentences in turn result in an aging prison population. However, many offenders with higher criminal history scores are already older and well past their peak offending years, or they will reach that past-​peak age well before they finish serving their enhanced prison terms (e.g., Piquero, Farrington, & Blumstein, 2007; Sampson & Laub, 1995; Hirschi & Gottfredson, 1983). Arrest data and criminological research have repeatedly demonstrated an age-​crime curve in which the frequency of criminal behavior tends to peak in the teens and early 20s, and declines steadily thereafter. By the time offenders have reached their 40s and 50s their rates of offending have substantially declined, and most of them have desisted entirely. Even the small subset of “chronic,” “career,” or “life course persistent” offenders display this pattern and eventually desist, albeit more slowly. Moreover, since the medical and other costs of incarcerating aging offenders can be quite high, their incarceration becomes increasingly less cost-​effective as they grow older. High-​magnitude criminal history enhancements thus are likely to contribute to an aged, low-​r isk, high-​cost inmate population. Not only is the prioritization of prison beds undermined through criminal history enhancements, but the size and expense of prisons is increased. The majority of sentencing guidelines systems are tasked with managing prison and other correctional resources, and avoiding prison overcrowding (Watts, 2017; see also Frase & Hester, forthcoming a). Frase and Hester (forthcoming c) estimated prison bed impacts of criminal history using data from three well-​established guidelines states: Kansas, Minnesota, and North Carolina. We focused on lower severity offenders who are identified for prison due to their more extensive prior records; 285

Hester et al.

Figure 13.1  Minnesota Standard Sentencing Grid

unlike the high-​severity offenders who would be recommended for prison regardless of their prior record, these lower severity offenders would have been recommended for a non-​prison sentence if they were first-​or low-​history offenders.We call these individuals “Zone 2” offenders and found that they were responsible for 36% of all prison commitments in Kansas, 51% in Minnesota, and 18% in North Carolina.We estimated a one-​year fiscal impact of these Zone 2 offenders to be $28 million in Kansas, $160 million in Minnesota, and over $131 million in North Carolina.These findings illustrate that criminal history impacts vary greatly across jurisdictions, and that the fiscal costs can be substantial, particularly if many of these Zone 2 offenders pose little threat to public safety in the form of potential for violent victimization. Taken together, these findings suggest that more modest criminal history policies could reduce the size and expense of prisons by reducing the number of aging non-​violent offenders who are sent to prison. The result should be the mitigation of mass incarceration with no compromise to public safety. 286

Criminal History in Sentencing

Decreasing the Proportionality of Sentence Severity Relative to Offense Severity Sentencing proportionality—​ the concept that more severe crimes should receive more severe sanctions—​is an important goal of most sentencing guidelines. Proportionality is important not just under a retributive theory of punishment, but also for reasons of effective crime control—​penalties that increase in proportion to crime seriousness tend to match punishment costs with expected deterrence, incapacitation, and other crime control benefits, while also sending valuable standard-​ setting and norm-​reinforcing messages about the relative seriousness of different crimes (Frase, 2005, 2012). Under sentencing guidelines, recommended and imposed sentence severity depends primarily on two factors: the severity of the conviction offense and the magnitude of the offender’s criminal history score. Accordingly, the greater the magnitude of prior record enhancements, the less the sentence depends on the severity of the conviction offense, thus lowering the proportionality of punishment relative to the crime being sentenced. For example, the Minnesota grid (Figure 13.1) shows that the presumptive sentence for an offense at severity level 8 with a criminal history score of 0 is a 48-​ month prison sentence. But if the offender instead has a criminal history score of 6, the presumptive sentence is 108 months in prison, which is 60 months longer. In this example, a greater proportion of the sentence for the offender with a criminal history score of 6 is attributable to criminal history (60 months) than to the severity of the offense (48 months). The issue is one of degree: focusing on proportionality concerns, the question is to what extent the punishment should be driven primarily by the current offense with comparatively modest criminal history enhancements, versus having the punishment determined in substantial part by criminal history at the expense of the current offense. Proportionality can also be undercut when, due to the design of sentencing guidelines grids, low-​ level offenses carry more severe punishment than higher-​severity offenses because of criminal history enhancements. For example, in Minnesota the highest-​history offenders convicted at severity level 1 on the main grid shown in Figure 13.1 (e.g., fleeing a police officer) have a recommended executed prison sentence of 19 months, which is more severe than the recommended 18-​month suspended sentence specified for the lowest-​history offenders convicted at severity level 5 (e.g., residential burglary), which is four severity levels higher on the grid. Of the Minnesota main grid’s 77 total cells, only 18 cells (23%) are what Frase and Hester (forthcoming a) define as “fully proportionate,” meaning that the recommended sentences for offenders convicted in those cells are more severe than all recommended sentences at lower offense severity levels on the grid and are also less severe than all recommended sentences at higher levels on the grid. Notably, most of these fully proportionate cells are at the top of the Minnesota grid, where few offenders are sentenced (only 2% of Minnesota offenders fall into the 18 fully proportionate main grid cells mentioned above).

Increasing Racial Disproportionality in Prison Inmate Populations Another goal of many guidelines systems is to reduce racial disparities in sentencing. But prior research, from both guidelines and non-​guidelines jurisdictions, has found consistent racial differences in prior record (Frase, 2009). Some of these differences in prior record have been linked to differences in policing and charging practices (Hannah-​Moffat, 2013). However acquired, more extensive criminal records result in higher criminal history scores under sentencing guidelines, and this in turn increases the racial disproportionality of prison populations beyond the levels that would result solely from racial differences in the nature and severity of conviction offenses. Criminal history scores cause higher proportions of African-​American offenders to be sentenced to prison rather than probation (Frase, & Hester, forthcoming b) and also affect decisions such as sentence length and the likelihood of receiving a favorable departure from a guidelines recommendation (Frase, 2009). All told, the combined effects of prison commitment and sentence length decisions work to the substantial disadvantage of African-​American offenders. 287

Hester et al.

Research from Minnesota, for example, shows that in recent years the percentage of African-​ American individuals receiving an executed prison term has been almost 50% greater than the percentage for white individuals, and well over half of this racial difference is due to African-​American individuals having higher criminal history scores (the remainder is due to differences in offense severity and eligibility for mandatory prison terms) (Frase & Hester, forthcoming b). In addition, the average length of executed prison sentences for African-​American Minnesota offenders is about 20% longer than the average for white offenders. These racial differences can be expressed in a single measure that reflects both prison commitment and prison length decisions: “prison months,” with probation sentences counted as zero months. In 2012–​14 the average prison months imposed on African-​American offenders in Minnesota was 73% greater than the average for white offenders (16.7 vs. 9.7, respectively). About half of the racial differences in executed prison durations and prison months were due to African-​Americans having higher criminal history scores. Thus, racial impact should be an important consideration in the construction of the criminal history dimension of sentencing guidelines, and should be regularly reviewed as the policies that undergird criminal history are revisited and adjusted over time.

Jurisdictional Variation Next, we document some of the tremendous variation among jurisdiction in how criminal history scores are constructed. This variation should be interesting to scholars and practitioners for at least two distinct reasons. First, the variations create opportunities for research. For example, as discussed below, some jurisdictions include prior misdemeanor offenses while others do not, creating the potential for investigation into whether consideration of prior misdemeanors carries any crime control benefits. Second, the variation reflects the haphazardness of prior record approaches. There is anecdotal evidence that prior record policies were largely adopted on intuition (Frase & Roberts, forthcoming; Tonry, 2010). The lack of cohesiveness in criminal history score construction lends support to the view that prior record policies fail to command a clear and cohesive line of thought. Consequently, many elements of criminal history scores may not contribute to the punishment purposes that would support additional time in prison—​in some cases much more time in prison—​on the basis of prior record. Punishment deprives individuals of liberty and significantly affects a person’s future, as well as that of their children, family, and communities. Given the gravity of liberty deprivation coupled with the unintended negative consequences of increased costs and exacerbated racial disparities, it seems imperative that prior record premiums be carefully thought out and tailored to effect legitimate punishment goals. The variation is one indication that prior record policies fall short of this bar. As Figure 13.1 (above) again illustrates, the sentencing matrix provides a simple and easy-​to-​use guide for distilling information in what can be a complex decision-​making arena. But the heuristic simplicity of the grid masks the deep complexity of the concept of criminal history. The intricacies become apparent when surveying how jurisdictions count aspects of past behavior for current punishment. As enumerated below, at its core, criminal history always includes some measure of prior felony convictions; misdemeanor convictions and juvenile adjudications are usually captured as well. Additionally, criminal history often includes other factors that are tangentially related to prior offending, such as prior probation revocations, and prior incarcerations (see Frase et al., 2015). All of these factors come together to determine the defendant’s criminal history score or category. The following are common variants of criminal history scoring rules. 1.

Defining a “Prior” Conviction: Most guidelines systems base criminal history on prior convictions rather than prior charged conduct; however, “the definition of a ‘prior’ conviction in guidelines systems ranges from very narrow (only convictions entered before commission of the current 288

Criminal History in Sentencing

2.

3.

4.

5.

6.

7.

crime or crimes being sentenced) to very broad (any crime sentenced before or during the current sentencing event, regardless of the order in which the ‘prior’ and the current crimes were committed)” (Frase, 2015). Prior Offense Weighting: All guidelines jurisdictions employ some type of offense weighting as a means of accounting for the quantity and quality of prior offending. For example, felonies are typically weighted more heavily and result in a higher criminal history score or category than misdemeanors, and more serious felonies will result in a higher criminal history score or category than less serious felonies. Jurisdictions typically apply weighting schemes based upon the severity of the prior offense (e.g., four points for Class A felonies), the punishment imposed for the prior offense (e.g., one point for each executed sentence of incarceration), the type of offense (e.g., person versus property), or a combination of attributes (e.g., a moderate record might be defined as three to five convictions at a certain severity level) (Hester, 2015). Misdemeanors: Sentencing guidelines may limit the misdemeanors that are counted for criminal history purposes (i.e., only the most serious or all but traffic offenses), and it may take several misdemeanors to reach the equivalent criminal history value of one less serious felony (e.g., four misdemeanors equal one point). A few jurisdictions have also built in special considerations such as not counting misdemeanors used to enhance the current offense to a felony or excluding from the criminal history score misdemeanor convictions for which the individual was not represented by an attorney (Mitchell, 2015). Juvenile Adjudications: Every guidelines jurisdiction except North Carolina includes juvenile adjudications in the criminal history calculation. About a third of jurisdictions count juvenile adjudications in the same manner as adult convictions. But the majority of jurisdictions discount juvenile adjudications by capping the number of points that can be derived from this source, limiting the kinds of offenses for which juvenile adjudications will be counted (i.e., only the most serious felonies), or limiting the inclusion of juvenile adjudications by the offender’s age (i.e., if the offender is 23 or younger) or the amount of time that has elapsed since the adjudication (Mitchell, 2015). Decay and Gap: About half of guidelines systems incorporate decay or gap rules that serve to wash out or eliminate prior offenses from the criminal history calculation if the offenses are very old or if the individual achieved a crime-​free existence for a specified number of years. The Federal Sentencing Guidelines offer an example of a decay rule. Prior convictions are no longer counted in the criminal history score calculation after 10 or 15  years, depending on the length of sentence for the prior offense. The Washington Sentencing Guidelines offer an example of a gap rule. Prior Class B felony convictions are not counted in the criminal history score if, since the last date of release from confinement, the offender has spent ten consecutive years in the community without a new conviction. Most jurisdictions start the clock running on the decay or gap period when the offender is discharged from the sentence, but a few start the clock earlier at sentencing or upon release from incarceration (Mitchell, 2015). Patterning: About two-​thirds of sentencing guidelines jurisdictions incorporate “patterning” rules wherein priors that are similar to the current offense are weighted more heavily, thereby further enhancing criminal history for repeat offenders. For example, in North Carolina, one point is added to the criminal history score if all of the elements of the present offense are included in any prior offense (Roberts, 2015). Custody Status: Most guidelines provide for a sentence enhancement if the offender was under some form of custody status (i.e., on probation or parole, or in jail) at the time of the offense. The most common strategy is to assign additional criminal history points, but some guidelines treat custody status as an aggravating factor for departure. For those jurisdictions that add criminal history points, custody status may carry as much or more weight than an additional felony; in other jurisdictions, it carries only a fraction of the weight of a felony (Roberts, 2015). 289

Hester et al.

In addition to all of the variation in the rules described above, there is one additional layer of structural variation in the way that criminal history is represented in the guidelines. Thirteen jurisdictions use a point-​based system in which points are assigned for each of the various criminal history factors, and the criminal history score is the sum of these points. The remaining five jurisdictions take a categorical approach in which the criminal history dimension is divided into categories based upon the number and severity of prior offenses (e.g., three or more person felonies) (Frase et al., 2015). Figure  13.1 shows the point-​based system used in Minnesota; Figure  13.2 shows the categorical approach used in Oregon. Both the point-​based and categorical approaches account for the seriousness of prior offenses, but with the potential for very different results. When a point-​based system is used, higher criminal history scores usually result from very lengthy criminal records or from shorter criminal records consisting of very serious crimes that are weighted more heavily. For repeat offenders, the criminal history score builds gradually as the individual accumulates a criminal record. Progression across the point-​based criminal history categories is relatively modest, generally increasing by one to two categories at a time (and thus, one to two columns on the grid), depending upon the weighted values of the prior crimes. For lower severity crimes, at some point an individual’s criminal history will cause movement across the dispositional line, and an offense that would ordinarily garner a probation sentence will garner a prison sentence solely because of the criminal history score. For example, under the Minnesota grid (Figure 13.1), the shaded cells indicate that probation is the appropriate disposition and non-​shaded cells indicate that prison is the appropriate disposition. The dark bolded line around the shaded cells represents the dispositional line, and this runs through the rest of the grid, establishing the boundary between presumptive probation and presumptive prison sentences. At severity level 7, for a first-​time offender with a criminal history score of 0, the presumptive sentence is probation. But for an offender with a criminal history score of 3, the presumptive sentence is a 54-​month executed prison sentence. Under this scheme, even a long record of very low-​level offenses can accrue a significant criminal history score, resulting in sentences of confinement for crimes that might ordinarily garner probation sentences. For example, at severity level 1, the presumptive sentence for an offender with a criminal history score ranging from 0 to 5 is probation whereas the presumptive sentence for an offender with a criminal history score of 6 is a 19-​month executed prison sentence. When a categorical system is used, the defendant’s progression through the criminal history categories is less linear. A first-​time defendant will always be placed in the lowest criminal history category (e.g., Category I on the Oregon grid shown in Figure 13.2). But a second-​time offender could fall into any number of categories, depending on the severity of the prior offense. In the Oregon example, if the defendant’s prior offense was a Class A misdemeanor, the defendant’s criminal history would increase from Category I to H, only a one-​step move. If the prior offense was a non-​person felony, the defendant would move from I to G, a two-​step increase. But if the prior was a felony person offense, the defendant would move from Category I to D—​a full five-​step move—​even if the current offense is much less serious than the prior. On its face, the progression of sentences across the Oregon grid appears to increase incrementally. In reality, an offender whose criminal history consists of a person or very serious offense will leapfrog over criminal history Categories I through E, and the sentence will be drawn from the cells in Categories D through A. The resulting sentence will represent a significant—​and perhaps disproportional—​increase from the punishment received by the first-​time offender in Category I. On the other hand, the categorical approach also prevents a low-​level repeat offender from attaining the highest criminal history categories. In the Oregon example, offenders who never commit a person offense will remain in categories I through E indefinitely, regardless of the number of offenses on their criminal records. In this way, the categorical approach serves to cap the sentence for low-​level offenders. Jurisdictions also vary tremendously in the magnitude by which criminal history enhancements impact sentencing outcomes, which include both how likely a person is to receive a jail or prison 290

Criminal History in Sentencing The Oregon Sentencing Guidelines Grid Crime Seriousness 11 10

C

ro Term

G

8

61– 65 35– 40

56– 60 29– 34

51– 55 27– 28

46– 50 25– 26

41– 45 23– 24

39– 40 21– 22

37– 38 19– 20

34– 36 16– 18

7

31– 36

25– 30

21– 24

19– 20

16– 18

180 90

180 90

180 90

180 90

25– 30 15– 16

19– 24 13– 14

15– 18 11– 12

13– 14 9– 10

10– 12 6– 8

180 90 180 90

180 90 120 60

180 90 120 60

180 90 120 60

10– 11

8– 9

120 60

120 60

120 60

120 60

120 60

120 60

120 60

120 60 90 30

120 60 90 30

120 60 90 30

120 60 90 30

120 60 90 30

120 60 90 30

90 30 90 30

90 30 90 30

90 30 90 30

90 30

90 30

90 30

90 30

90 30

90 30

90 30

90 30

90 30

6 5 4 3 2 1

S

225– 196– 178– 164– 149– 135– 129– 122– 120– 269 224 194 177 163 148 134 128 121 5 58– 121– 116– 111– 91– 81– 71– 66– 61– 70 65 60 Years 90 80 130 120 115 110 66– 72 41– 45

9

Ma e ar t

3 Years 3 Years

2 Years

11 2 Years

18 Mos.

12 Mos.

6 Mos.

2 Years

1 Year

A

The o ender s criminal histor includes three or more erson adult con ictions or u enile ad udications.

B

The o ender s criminal histor includes t o erson con ictions or u enile ad udications.

C

The o ender s criminal histor includes one adult con iction or u enile ad udication or a erson elon and one or more adult con iction or u enile ad udication or a non erson elon .

D

The o ender s criminal histor includes one adult con iction or u enile ad udication or a erson elon ut no adult con iction or u enile ad udications or a non erson elon .

E

The o ender s criminal histor includes our or more adult con ictions or non erson elonies ut no adult con iction or u enile ad udication or a erson elon .

F

The o ender s criminal histor includes t o or three adult con ictions or non erson elonies ut no adult con iction or u enile ad udication or a erson elon .

G

The o ender s criminal histor includes our or more adult con ictions or Class misdemeanors one adult con iction or a non erson elon or three or more u enile ad udications or non erson elonies ut no adult con iction or u enile ad udication or a erson elon .

H

The o ender s criminal histor includes no adult elon con iction or u enile ad udication or a erson elon no more than t o u enile ad udications or non erson elonies and no more than three adult con ictions or Class misdemeanor s.

I

The o ender s criminal histor does not include an u enile ad udication or a elon or an adult con iction or a elon or Class misdemeanor .

elonies in an com ination o

elonies in an com ination o adult

Figure 13.2  Oregon Sentencing Grid and Criminal History Categories

sentence, and the length of incarceration sentence imposed. Elsewhere we have compared guideline jurisdiction magnitudes in detail (see Frase et al., 2015; Hester et al., 2018).We found that, comparing first-​offender categories with maximum criminal history categories in 12 US grid-​based guidelines systems, high criminal history offenders were given recommendations over six times longer than first 291

Hester et al.

offenders on average. In some jurisdictions the multiplier was as low as about two times, and at the other extreme as high as over 14 times. In other words, in the most pronounced case, a person’s prior record results in a more than 1,300% increase in the sentence imposed and on average it is over a 500% increase. As we expressed in Hester et al. (2018, p. 218): The average multiplier of 6.4 means that, if a first offender is recommended for some particular crime to 18 months in prison, the high-​history offender attracts a recommended sentence of 115 months for that same offense. The additional 97 months is attributable to the second offender’s prior record—​offenses for which he has already been found guilty, been sentenced, and satisfied his debt to society. Yet those past satisfied debts account for 84% of his total sentence. As this part demonstrates, there are myriad policy choices embedded within the criminal history score or category. Given the weighty consequences, and the wide variation and magnitudes of criminal history enhancements, one might revisit the question of purpose: why should an offender receive increased punishment for a current offense on the basis of his or her prior offenses? Presumably the answer is an extension of why we punish at all, as expressed through the traditional justifications for punishment—​retribution and utilitarian (crime control) theories like rehabilitation, deterrence, and incapacitation. Articulating and expounding a compelling justification for robust prior record enhancements has proven difficult, however.

Justification Issues The trouble with prior record enhancements is that they carry a strong intuitive appeal but are ultimately more difficult to square with a theory of punishment than the initial intuition might suggest. Elsewhere we have given a fuller survey of the various attempts to justify criminal history enhancements (see Hester et al., 2018). Here we provide a brief overview.

Retributive Rationales First, legal theorists and philosophers are divided over whether criminal history enhancements are justifiable on retributive, desert-​based grounds. For retribution, to enhance a current penalty of the basis of prior behavior, the past behavior must have some bearing on current culpability: the offender must somehow be more blameworthy for the current crime because he or she has been convicted of crimes in the past, or else have inflicted more harm in the current crime because of the past convictions (Hester et al., 2018; see also von Hirsch, 1993). One set of retributive theorists asserts that, on this basis, past behavior has no bearing on current culpability and thus there is no retributive basis for prior record enhancements (e.g., Fletcher, 1978; Singer, 1979; Davis, 1985; Bagaric, 2014; Duff, 2001; Dagger, 2012). Under this view, punishment for a crime should be tailored to factors related to the circumstances of that crime and the harm caused, and an offender’s past simply has no bearing on the present concerns. As Tonry (2010) has observed, it is not sufficient that a recidivist offender is viewed as a worse person. The law punishes bad behavior, not bad character. While this view is parsimonious and has some intellectual appeal, it cuts against deep-​seated intuition and public and policy-​maker opinions (Roberts, 2015), and would appear to prevent even first-​offender mitigation (Frase and Roberts, forthcoming; Hester et al., 2018). While we agree that criminal history enhancements under most current guidelines practices need rethinking and scaling back, we do not go so far as to assert that the prior record should be censored for the current sentencing. Those who argue prior record is relevant for current sentencing can be divided into two camps—​ one that accepts a continuous aggravation of punishment that increases roughly in step with the 292

Criminal History in Sentencing

criminal record, and a second camp that focuses on some appropriate mitigation from full desert for first-​time offenders (or those who may not technically be first-​time offenders but whose records are not extensive). Among the aggravation theorists, there have been numerous efforts to articulate and expound on the theory that gets around the mere bad character problem noted above (i.e., that it’s not sufficient that the prior record merely demonstrates the individual is of a worse character—​ something about the prior record must make something about the current offense worse) (see, e.g., Roberts & von Hirsch, 2010; Tamburrini & Ryberg, 2012). For example, some have asserted that a person who has previously been convicted is on heightened notice not to commit a crime (Lee, 2009). While this might carry more appeal for some obscure offenses, it remains uncompelling to some that a person needs a sanction in order to be put on sufficient notice that they should not commit most crimes like aggravated assault, burglary, robbery, etc. Another effort is omission theory, borrowed from a narrowly applied rule in tort law, which suggests that the repeat offender’s culpability is greater because he or she has neglected to take the necessary action to remediate the criminal tendencies in light of the earlier crimes and sanctions (Lee, 2009). Given the prior punishment, the offender has a heightened duty to avoid criminal activity. But as we observe in Hester et al. (2018), this theory would seem to apply best to those whose criminal patterns are similar; its singular focus on human agency is also likely inconsistent with at least some of the theories of crime causation; and finally, some observers remain unconvinced that prior record changes a person’s duty to obey the law, at least where criminal law is clearly promulgated and understood. (See Frase & Roberts, forthcoming and Hester et al., 2018 for a review and critique of some additional aggravation theories.) Perhaps more compelling is von Hirsch’s (2017) theory of progressive loss of mitigation (PLM). Under this theory, it is not that recidivists have elevated desert, but rather that offenders with no (or a modest) prior record are not fully blameworthy. As the prior record accumulates, sentence mitigation is progressively withheld so that after a second or third or fourth chance, the repeat offender receives his or her full desert. This theory squares with the notion of giving first offenders a second chance and opportunity for rehabilitation. It is not, however, reflective of any current sentencing guidelines policies. Rather than “flattening out” after a few offenses, most guidelines systems impose a roughly monotonic increase in recommended severity across the grid. In the highest criminal history category a “cap” is imposed; but before that point, and with at least five and as many as ten categories, the claim of reduced blame seems unpersuasive.

Utilitarian Theories Prior record policies could find better theoretical footing in utilitarian justifications. Past behavior is one of the best predictors of future behavior, so theories aimed at preventing future crime might have good reason to incorporate prior conduct as an indicator of the appropriate disposition or dose of a current sentence. Unfortunately, at least as currently implemented, prior record policies do not seem well supported by utilitarian theories in current practice. First, current evidence suggests that longer prison terms do not specifically deter individuals and may even have criminogenic effects. This is the takeaway from a number of recent studies that use a diverse array of data and methodological approaches. For example, Meade et al, (2013) used multivariate logistic regression and propensity score matching to estimate the effect of time served on the odds of recidivism for a sample of Ohio offenders on post-​release supervision. They found no effects of lengthier prison terms until the offender had served over five years. Offenders who served more than five years (most of whom had in fact served more than ten years) had lower odds of recidivism than offenders who served shorter terms. Meade et al. note, however, that the results, while consistent with specific deterrence for longer prison terms, could also be the result of incapacitation and maturation. Offenders who served prison terms of five or ten years would have been incapacitated for longer through peak crime years and thus would have been closer to aging out of crime upon release. 293

Hester et al.

Meade and colleagues concluded that since offenders who served between two and five years were not at higher risk of recidivism than offenders with shorter sentences, sentences in the two-​to five-​ year range could be reduced without compromising public safety. In another recent innovative approach, Nagin and Snodgrass (2013) used an instrumental variable method that keyed in on the different sentencing tendencies of Pennsylvania judges. Since cases within a county are randomly assigned to judges, and since judges differ in punitiveness, the researchers were able to examine whether offenders who served longer prison sentences (doled out by more punitive judges) had different likelihoods of recidivism than those who received shorter sentences.They found no evidence of a crime reducing effect of incarceration dispositions on reoffending (as measured by re-​arrest) at various follow-​up periods of one, two, five, and ten years. Cochran, Mears, and Bales (2014) examined differences in the effectiveness of four different types of sanctions (probation, intensive probation, jail, and prison) in reducing recidivism. They too used propensity score matching, which allowed them to isolate the treatment effects of the type of sanction. They found that more intensive sanctions were associated with higher rates of recidivism, thus supporting the notion that the effect of prison is criminogenic rather than deterrent. Mitchell and colleagues (2016) used a regression discontinuity design to study the impact of sanction type on recidivism among Florida offenders. Their analytic approach capitalized on the exogenous assignment of guidelines scores to offenders. Punishment recommendations are based on score cut-​points, meaning respondents near the cutoff share similarities on observed and unobserved measures, approximating random assignment. Mitchell et al. found that imprisonment had no effect, or possibly a small adverse effect, on the likelihood of reconviction compared to non-​prison sanctions. On the whole, these mixed findings underscore Nagin, Cullen, and Jonson’s (2009) earlier synopsis that while the “scientific jury is still out … the great majority of studies point to a null or criminogenic effect of the prison experience on subsequent offending” (p. 178). Further, while there appears to be general agreement that the existence of a criminal justice system serves as some general deterrent (Doob & Webster, 2003; Nagin, 1998), there is a lack of evidence that increasing already stiff penalties continues to yield beneficial general deterrent returns. This was the conclusion of the National Academy of Sciences report on mass incarceration (Travis, Western, & Redburn, 2014), as well as other assessments and reviews of the deterrence literature (see Cullen, Pratt, & Turanovic, 2016; Doob & Webster, 2003; Durlauf & Nagin, 2009; Nagin, 2013). There is of course an incapacitation effect of prison, so increasing prison terms on the basis of criminal history scores will result in some crime savings. However, as we have argued elsewhere (Hester et al., 2018), non-​selective incapacitation as an across-​the-​board policy for linear criminal history enhancements is a poor theoretical fit. Incapacitation has been conceived of as a way to remove the most violent and high-​rate offenders for a substantial portion of their criminal careers (Zimring & Hawkins, 1995). Prison is a costly endeavor, but murder, rape, and other violent crimes are also costly to society, so if policies were able to adequately identify (in advance) serious and prolific offenders, the benefits could outweigh the costs. But criminal history policies indiscriminately recommend increases for all offenders with prior records. Adding a few months or years of prison for non-​violent offenders is a costly way to reduce crime and fails to deliver the primary thrust of incapacitation theory, which is the long-​term removal of the worst offenders. Moreover, attempting to use incapacitation as a justification for across-​the-​board criminal history enhancements exposes a more global problem with criminal history enhancements in general: these policies largely ignore the findings and implications of much of the criminological literature on issues like desistence and the life course.

The Need to Rethink Criminal History Enhancements The concept of the grid as a tool for guiding judicial discretion at sentencing originated with the work of Don Gottfredson and Leslie Wilkins in the 1960s developing parole release guidelines 294

Criminal History in Sentencing

(Gottfredson, Wilkins, & Hoffman, 1978; Gottfredson et  al., 1975; see also Hester, 2018b). After developing guidelines for back-​end release decisions, the same team of researchers turned their attention to a complementary approach aimed at the front-​end sentencing decisions (see Hester, 2018b). From the outset, they were inspired by the success of their federal parole efforts and the utility of the Cartesian coordinate system which allows one to identify a point of interest on a map using just two points of information (see Wilkins et al., 1978, p. 15). The grid approach took hold in the early “Albany” approach guidelines which were conceived of as descriptive (rather than prescriptive) guidelines that would be based on current practices, but which would then constrict recommended sentences to the central 80% of current sentencing, thus trimming away the outlying 20% of sentences and offering more uniformity and consistency in sentencing (Hester, 2018b). One significant difference between parole decisions and sentencing decisions which the research team became aware of is that while paroling revolves around one decision (whether to release or not), sentencing involves two decisions: (1) whether to incarcerate and (2) if so, for how long. Evidence suggests that, as Gottfredson,Wilkins, and colleagues constructed sentencing guidelines in several pilot jurisdictions, it became apparent that the factors informing these two decisions did not always overlap. As Hester (2018b) recounts, while the guidelines architects considered using two different decisional instruments, this was met with strong opposition by practitioners who had enough concerns over implementing a single instrument. Accordingly, the grids were constructed on the empirics of the in/​out decisions (since whether any deprivation of liberty was considered the most pressing concern); sentence durations were then overlaid onto the in/​out grids, even though the roughly linear duration numbers did not comport with prior practices (Hester, forthcoming). For instance, in Cook County the sentence length numbers did not conform to the expected pattern and thus “mapping the length of sentences was not predictive in the statistical sense. Thus, the guideline model became an ‘experience’ table” (Calpin, Kress, & Gelman, 1982, p. 76). In Maricopa County (Arizona), investigation of the factors predicting past sentencing practices showed that all but two out of 21 criminal history-​related variables were related to the decision to incarcerate, but only four of these were related to the sentence length decision at a statistically significant level (Calpin, Kress, & Gelman, 1982; see also Hester, 2018b). Maryland followed the same approach to constructing guidelines and encountered similar issues; accordingly, the statistical analyses of past practices were augmented by “a good deal of collective judgment” (Carrow et al., 1985). Overall, while the efforts to empirically construct descriptive sentencing guidelines based on past practices “proved ‘inconclusive’ and ‘disappointing’ ” (Calpin, Kress, & Gelman, 1982), the allure of an abscissa and an ordinate3 making powerful reductions of complex constructs prevailed, and criminal history as one of the two primary determinants of both the in/​out and length decisions was fixed. On the heels of the descriptive guidelines approach came the first statewide sentencing commissions and guidelines. Although some of these replaced the descriptive Albany approach with a prescriptive approach, the two-​way matrix was retained and remains the modal system for implementing sentencing guidelines (see Frase et al., 2015; Hester, 2018b). As these early commissions created the earliest guidelines, it appears that decisions about how the criminal history dimension would be constructed and would operate were mostly ad hoc and based on intuition.Tonry (2010), noting that he consulted for many of the earliest guidelines commissions, describes how deliberations over offense gravity rankings were often methodical and protracted, but how, by contrast, decisions over criminal history were typically ad hoc and based on intuition and “back-​of-​an-​envelope” calculations. (The one exception to this is the US Sentencing Commission, which we discuss in detail below.) Hester’s quantitative and qualitative findings from South Carolina, a non-​guidelines state, may also provide some insight into how sentencing guidelines have driven criminal history outcomes. Hester and Sevigny’s (2016) and Hester and Hartman’s (2017) analysis of felony sentencing in South Carolina provide a rare example of a statewide sentencing study in a non-​guidelines state (the overwhelming majority of sentencing research in the past decades has come from guidelines jurisdictions where prior record policies are built into the guidelines). Both Hester and Sevigny (2016) and Hester and 295

Hester et al.

Hartman (2017) found that for the in/​out decision in this non-​guidelines state, the offender’s prior record was highly salient, one of the most predictive factors of the outcome. But for the sentence length decision, depending on the methodology adopted (a hierarchical linear model in Hester and Sevigny and a truncated negative binomial count model in Hester and Hartman) the impact of prior record was minimal or non-​existent. Subsequent qualitative interviews of the state’s trial judges by Hester (2017, 2018b) suggested that while an offender’s criminal history was often relevant, the way the non-​guidelines judges processed prior record was nuanced. Several of the judges interviewed expressed the desire to give offenders a second chance, but if a person had several convictions, the record signaled a perceived need to incarcerate. As one judge said, “I always feel like the defendant should be given a first chance … [a]‌nd maybe probation a second time … [b]ut by the third time, you’re about ready to stop messing around with this and let them spend some time in jail” (Hester, 2018b, p. 20). But the judges also emphasized that the relevance of the prior record was not mechanical. Judges variously noted that factors like similarity in offenses, patterns, and time between offenses provided important context in an overall judgment of whether the offender had been making strides towards successful living. Several judges also noted that the criminal history was less important for prison length determinations; as one judge said: “Prior record goes not as much to the severity as the certainty … You know a lot of times there’s not a lot of difference between 15 and 20 years.” These findings help explain why in the prior empirical analyses, criminal history was such a strong predictor of the in/​out decision and a much weaker predictor of the sentence length decision.They might also provide insight into why in the early guidelines efforts the prison length practices did not map cleanly onto an in/​out grid. The implications would be that in an effort to provide a single instrument to provide uniformity and consistency in sentencing, guidelines may have actually unwittingly imposed overly mechanical prison length criminal history multipliers. In summary, from the outset, guideline criminal history scores appear to have been less than carefully conceived. In fact, in many cases decisions related to prior record ran counter to empirical findings. With the benefit of several decades of experience, we now know that robust criminal history policies carry substantial unintended collateral consequences. Efforts to justify prior record enhancements have also developed and matured. From a social science perspective we are unable to say that criminal history enhancements lack a retributive basis, but the review of the literature demonstrates that the retributive rationale for prior record enhancements lacks a convincing consensus perspective. Although there appears to be general support for some novice offender mitigation, articulation of retributive grounds to support the sort of steady aggravation found in most guidelines schemes remains lacking. As for utilitarian justifications, research findings on the limited deterrent and incapacitation effects of prison sanctions continue to grow in number and sophistication. Perhaps the time is ripe for rethinking of criminal history policies. As a starting point, policymakers should grapple with the retributive limits of prior record. Is past behavior relevant to current culpability? If so, are there inherent limits on the magnitude of criminal history aggravation? For instance, should the amount of punishment attributable to prior behavior that has already led to conviction and punishment ever eclipse the portion of the sentence attributable to non-​prior-​record-​related culpability? In other words, is there an inherent principle that the past should never count more than the present, or that the prior record enhancement should, at maximum, be less than double the penalty recommended for the first offender? Whatever the answers to those retributive-​based questions, jurisdictions should also engage in empirical analysis of prior record policies—​to understand the costs and impacts of policies; to investigate whether criminal history designations differentiate among different levels of recidivism risk; and to identify which elements of prior record are associated with utilitarian benefits and which have no bearing on the likelihood of reoffending. In the next part we turn to the empirical research that has been done in the federal system for decades and two recent examples from Pennsylvania and Minnesota. 296

Criminal History in Sentencing

Empirical Findings Rethinking criminal history enhancement policies requires a foundation of empirical research. In this part we review the US Sentencing Commission’s exemplary efforts to evaluate the risk-​predictive accuracy of the federal criminal history score, and report on similar research from Pennsylvania and Minnesota. We then take note of several recurring methodological and normative issues, and some important unanswered questions.

The Federal Guidelines The US Sentencing Commission has conducted extensive research on the criminal history provisions of the federal guidelines. Much of this research has sought to assess how effective those provisions are in identifying offenders with lower or higher odds of committing new offenses once they are released back into the community. Unlike almost all other guidelines systems, the federal criminal history score [CHS] was expressly designed to predict recidivism risk and serve punishment goals related to offender risk, although the score was also based in part on the view that repeat offenders are more culpable (retributive punishment goal) and on the belief that enhanced punishment is needed for reasons of deterrence (USSC 2004a). Most of the CHS components were based on two validated risk assessment instruments: the US Parole Commission’s Salient Factor Score [SFS], and an instrument developed by a private research group (USSC 2004a, 2005). To assess the predictive value of the CHS, along with other predictors of recidivism, the Commission studied a sample of offenders sentenced in fiscal year 1992, and a second, larger group of offenders released on probation or from prison in calendar year 2005. The four reports from the 1992 cohort (USSC 2004a, b; 2005, 2010) focus primarily on criminal history. Some of the same criminal history and recidivism issues are also examined, and the findings mostly corroborated, in four reports based on the 2005 cohort (USSC 2016b, 2017a, b, c). Most of these reports also consider how recidivism rates vary with other legal and extra-​legal factors.4 The remainder of this part will emphasize the Commission’s research on criminal history and recidivism, while briefly noting other findings that have important policy or scientific implications.

The Federal Criminal History Score Under Sec. 4A1.1 and related provisions of the guidelines (USSC 2016a), the CHS had four components until 2010, and now has three (as a direct result of the commission’s research (USSC, 2010), the third of the following components was dropped): 1) one to three points (depending on the severity of the sentence imposed) are added to the score for each prior federal or state-​court conviction;5 2) two points are added if the current offense was committed while under post-​conviction criminal justice supervision, incarceration, or escape status [“status” points]; 3) prior to 2010, two points were added (or one point, for offenders who already had status points) if the current offense was committed within two years of release from a custody sentence of 60 days or more [“recency” points]; 4) one point is added for each prior crime-​of-​violence conviction (up to a maximum of three points, for this item) that did not receive any points under item (1) above because that offense was sentenced with other crimes and treated as a single sentence.6 The resulting point total places an offender in one of six criminal history categories: CHC-​I = 0 or 1 point; CHC-​II = 2 or 3 points; CHC-​III = 4, 5, or 6 points; CHC-​IV = 7, 8, or 9 points; CHC-​V = 10, 11, or 12 points; CHC-​VI = 13 or more points.

The 1992 and 2005 Offender Cohorts, Recidivism Measures, and Analytic Methods The first cohort consisted of a sample of 6,062 defendants sentenced in fiscal 1992; the second cohort included 28,519 offenders released into the community in calendar 2005. Both cohorts excluded 297

Hester et al.

non-​US citizens, due to incomplete information about their prior convictions. Recidivism by the 1992 offenders was measured during their first two years of release on probation or from prison; approximately 20% of offenders sentenced to prison were excluded because they did not have at least two years on post-​prison release (USSC 2004a, Exhibit 1). The 2005 offenders were followed for up to eight years after being released (no citizen offenders needed to be excluded). The 1992 study used two recidivism measures: 1) the “primary” measure included the first of the following to occur within two years after release: reconviction for a new offense; re-​arrest (with no further information about the final disposition); and revocation of probation or post-​prison release; 2) the second measure only counted the first reconviction. For the 2005 offenders, three measures of recidivism were used: 1) re-​arrests (number of unique arrest dates, including arrest for a new crime or for violation of release conditions); 2) reconviction; and 3) incarceration (commitment to prison or jail based on conviction for a new crime or revocation of release for an earlier conviction). Analytic methods applied to the 1992 cohort included the percentage of offenders recidivating, area under-​the curve (AUC) analysis, and survival analysis. For the second cohort, published findings have been limited almost entirely to percentages of offenders who recidivated, overall and within categories of various predictor variables (occasionally, two variables at once); sometimes survival analysis is also employed.

Major Findings Accuracy. In both offender cohorts recidivism rates by all measures rose steadily with increases in the guidelines criminal history category [CHC], and the greatest increases were between CHC-​I and CHC-​II; for example, the re-​arrest rate for 2005 offenders was 33.8 for CHC-​I, rising to 54.3 for CHC-​II—​58% higher. At CHC-​VI the 2005 rate was 80.1, yielding an average increase from CHC-​ II to CHC-​VI of 6.45percentage points, but the increase from IV to V was only 3.1 points, and from V to VI, 2.3 points; in relative terms, the percentage increases from CHC-​II to III and from CHC-​III to IV were 17 and 18%, respectively, while the increased from CHC-​IV to V and CHC-​V to VI were only 4 and 3%. Yet much more severe sentences are recommended as offenders move from IV to V, and from V to VI. For the 1992 cohort, AUC analysis found criminal history points to a more accurate predictor than CHC, by both recidivism measures (USSC, 2004a, Exhibits 6 to 8). Some components of the CHS were found to have strong predictive value, while others had little if any. For 2005 offenders, re-​arrest rates were highest and time to re-​arrest was shorter for those with 2-​point and 3-​point priors7 than for offenders with only 1-​point priors, and these differences generally held within each offender age category (USSC, 2017b) (total CHS and other risk-​related factors were not controlled for, however). For 1992 offenders, a comparison of CHS components (USSC, 2005, Exhibit 2) found that all were significantly related to time to recidivism (Chi Square p