Information Systems Approach to Jail Management: Guidance, Development and Use 1633212556, 9781633212558

Citation preview

CRIMINAL JUSTICE, LAW ENFORCEMENT AND CORRECTIONS

INFORMATION SYSTEMS APPROACH TO JAIL MANAGEMENT GUIDANCE, DEVELOPMENT AND USE

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

CRIMINAL JUSTICE, LAW ENFORCEMENT AND CORRECTIONS Additional books in this series can be found on Nova’s website under the Series tab.

Additional e-books in this series can be found on Nova’s website under the e-book tab.

CRIMINAL JUSTICE, LAW ENFORCEMENT AND CORRECTIONS

INFORMATION SYSTEMS APPROACH TO JAIL MANAGEMENT GUIDANCE, DEVELOPMENT AND USE

CHERYL L. COOPER EDITOR

New York

Copyright © 2014 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts of this book based on government reports are so indicated and copyright is claimed for those parts to the extent applicable to compilations of such works. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. Additional color graphics may be available in the e-book version of this book.

Library of Congress Cataloging-in-Publication Data ISBN:  (eBook)

Published by Nova Science Publishers, Inc. † New York

CONTENTS Preface Chapter 1

Chapter 2 Index

vii Running an Intelligent Jail: A Guide to the Development and Use of a Jail Information System Tim Brennan, Dave Wells and John Carr Jail Capacity Planning Guide: A Systems Approach David M. Bennett and Donna Lattin

1 111 197

PREFACE It is clear that virtually all criminal justice organizations, including jails, are driven by information. From initial intake to final release, virtually all key decisions are largely driven by the availability, quality, and careful analysis of data to support the variety of decisions made by jail administrators and personnel. Jails should consider themselves as informationprocessing organizations and active users of information technologies. A precondition of effective management support in the jail system is having access to accurate, high-quality data that can be presented in the appropriate formats. For most jails, this requires a jail management information system (MIS) that is adequate to support all routine inmateprocessing activities. Even when a jail has an adequate MIS, we often see inadequacies in the design of performance measures and inmate-monitoring indexes and, more generally, in quantitative analyses that make use of this information. This book uses many years of the authors’ collective experience in addressing the information technology (IT) infrastructure, database content, and analytical capacities of innumerable criminal justice institutions to develop a guide to the development and use of a jail information system. The book also discusses a jail capacity planning guide. Chapter 1 – It is clear that virtually all criminal justice organizations, including jails, are now driven by the collection, processing, and application of information specific to these settings. With the increasing focus on cost efficiency and the avoidance of wasteful spending, jail administrators must understand the importance of the potential data at their disposal and strategically plan for faster and more effective forms of data collection, storage, and analysis. Running an Intelligent Jail: A Guide to the Development and Use of a Jail Information System encourages jail administrators to consider the design and implementation of a jail management information system (MIS) that is tailored to the specific needs of their institutions, is more cost-effective, and is easier to use. The data they are able to collect, store, analyze, and apply to the correctional setting translates to more effective jail management, more realistic short- and long-term goals, the ability to track trends, a more systematic way to measure performance outcomes for the institution and its staff, and pertinent information on the offender population. These sections provide specific information on the types of data collection and analysis that are required of most jails, and training materials tailored to users’ different skill sets, as well as outlining the steps for implementing a jail management information system, and guidance on how to develop a Request for Proposal and select a vendor.

viii

Cheryl L. Cooper

We have also provided a variety of appendixes, including sample forms and reports, to enhance readers’ understanding of the technology and its many applications and provide the information they will need to move their organizations toward data-driven solutions. Chapter 2 – During the past 30 years, jails nationwide have become crowded in response to policy shifts in the criminal justice system, including the clampdown on driving under the influence, the adoption of mandatory arrests for domestic violence, and the “get tough” approach to many drug crimes. Crowding can create serious management problems, compromising the safety of both inmates and staff. Therefore, it is essential that jurisdictions adopt comprehensive, effective strategies to address the problem of crowding in our nation’s jails. Jails are part of a complex criminal justice system whose policies and practices directly influence total bed need. As such, jail planning cannot be done in a vacuum. Any consideration of future jail bed need must take place within the context of a discussion about how to manage the larger criminal justice system more effectively. Jail planning and system planning are one and the same. Emerging information technology provides us with unprecedented potential for analyzing the dynamics of the complex criminal justice system and forecasting and managing jail capacity needs. This guide describes key population management strategies that have as their foundation the necessity of holding offenders accountable while making judicious use of detention resources. This guide also makes the case for the importance of identifying offenders who pose higher risks and targeting them for the most intensive correctional resources, making available a full continuum of alternatives to jail, relying on evidence-based sanctions and quality treatments, and building in transition and stepdown options from jails. For better or for worse, all local systems will change. The question is not whether, but how, policies will change. We hope that this document will assist jurisdictions as they implement program strategies designed to plan for, respond to, and manage change, while making the most efficient use of existing resources.

In: Information Systems Approach to Jail Management ISBN: 978-1-63321-255-8 Editor: Cheryl L. Cooper © 2014 Nova Science Publishers, Inc.

Chapter 1

RUNNING AN INTELLIGENT JAIL: A GUIDE TO THE DEVELOPMENT AND USE OF A JAIL INFORMATION SYSTEM ∗

Tim Brennan, Dave Wells and John Carr FOREWORD It is clear that virtually all criminal justice organizations, including jails, are now driven by the collection, processing, and application of information specific to these settings. With the increasing focus on cost efficiency and the avoidance of wasteful spending, jail administrators must understand the importance of the potential data at their disposal and strategically plan for faster and more effective forms of data collection, storage, and analysis. Running an Intelligent Jail: A Guide to the Development and Use of a Jail Information System encourages jail administrators to consider the design and implementation of a jail management information system (MIS) that is tailored to the specific needs of their institutions, is more cost-effective, and is easier to use. The data they are able to collect, store, analyze, and apply to the correctional setting translates to more effective jail management, more realistic short- and long-term goals, the ability to track trends, a more systematic way to measure performance outcomes for the institution and its staff, and pertinent information on the offender population. These sections provide specific information on the types of data collection and analysis that are required of most jails, and training materials tailored to users’ different skill sets, as well as outlining the steps for implementing a jail management information system, and guidance on how to develop a Request for Proposal and select a vendor.

∗

This is an edited, reformatted and augmented version of a document, NIC Accession Number 027446, issued by the National Institute of Corrections, dated August 2013.

2

Tim Brennan, Dave Wells and John Carr

We have also provided a variety of appendixes, including sample forms and reports, to enhance readers’ understanding of the technology and its many applications and provide the information they will need to move their organizations toward data-driven solutions. Morris L. Thigpen Director National Institute of Corrections

SECTION 1. WHY DO JAILS NEED TO BECOME INTELLIGENT? Introduction It is clear that virtually all criminal justice organizations, including jails, are driven by information. From initial intake to final release, virtually all key decisions are largely driven by the availability, quality, and careful analysis of data to support the variety of decisions made by jail administrators and personnel. Jails should consider themselves as informationprocessing organizations and active users of information technologies. A precondition of effective management support in the jail system is having access to accurate, high-quality data that can be presented in the appropriate formats. For most jails, this requires a jail management information system (MIS) that is adequate to support all routine inmate-processing activities. Even when a jail has an adequate MIS, we often see inadequacies in the design of performance measures and inmate-monitoring indexes and, more generally, in quantitative analyses that make use of this information. Jail managers should understand the strategic importance of using an MIS to measure a jail’s performance, particularly in today’s fiscal environment. With the increasing focus on cost efficiency and avoidance of wasteful spending, local criminal justice systems, and their jails, must adopt MISs that are based on data-driven decisions and policies and that can be used to measure performance-based outcomes. Influence, or power, in the jail context, is the capacity to mobilize the organization’s energy, resources, information, and staff to support particular goals and outcomes. Most jail managers are aware of the link between knowledge and influence. Leadership influence grows to the extent that the particular leader has both the access to data and the skill to transform it into usable and defensible knowledge. Any exercise of power assumes some desired objective or policy, such as affecting staff and inmates’ behavior or attitudes; marshaling needed resources; increasing the access to information; changing work assignments, processes, or procedures; and proposing specific performance improvements.

Knowledge Is Power The power of knowledge—and its foundation in data—is increasingly central to jails’ organizational processes that involve leadership, planning, directing, and controlling behavior as well achieving better performance outcomes. In witnessing such interactions, jail staff at

Running an Intelligent Jail

3

all levels often observe the exercise of power (legitimate or otherwise) and become aware of its importance for both personal and organizational success. Those jail managers, administrators, and line staff with more data access and greater skill in analyzing and synthesizing data will gain influence, regardless of their position in the bureaucracy, whereas others with less skill will lose their influence. When managers’ influence stems from superior analytic skills and access to the pertinent data, they will typically gather more influence and control. The power of knowledge enhances that manager’s ability to build consensus, set goals, propose actions, and direct the organization’s energies toward the selected goals. Knowledge thus can be used to organize and provide a rationale for most jail procedural and policy decisions, subsequent implementation efforts, and changes in the behavior or attitudes of staff. Senior managers may become less powerful over time as information technology and data-analytic tools evolve. In some instances, senior staff in powerful positions in the jail bureaucracy may become dependent on lower-level employees who have more direct access to the relevant data and have the skills to organize, analyze, and provide the information for the senior manager. To the degree that senior managers lack the necessary skills or are averse to learning them, their power base will erode.

Politics, Power, and Jail Data Knowledge, and the data supporting it, becomes particularly important in decisionmaking situations in which stakeholders must make highly contested choices, often involving a change in policy orientation or resource allocation. Stakeholders exercise their power in these situations to justify the desired results or to ensure that their preferred policy options are accepted and will lead to these results. In most jails, a constant dynamic among senior managers and department heads involves a competition for resources, such as having access to information or a status position, or having priority for specific programs or procedures. Influence and power in these situations is increasingly based on claims of knowledge and are key factors when resolving issues. Thus, power and influence flow to those managers or departments in the jail system who are best able to establish and control the information resources, particularly in jail environments where data-driven decisionmaking (DDDM) is implemented as a strategy.

Becoming an Information-Driven Jail: What Is DDDM in the Jail Context? Data-driven decisionmaking in the corrections field rests on practices—for example, quality control by qualified management, organizational learning, and continuous improvement—that originated mainly in industry and are designed to support both decisionmaking and planning for the future. The goal is organizational improvement by the systematic collection of the types of data that broadly reflect the functions of the organization as a while as well as those of specific departments. DDDM has several main steps, outlined below.

4

Tim Brennan, Dave Wells and John Carr

Step 1: Collect the Appropriate Data It is critical to realize that the types of data collected will vary across different units or departments of a jail and across any particular functional unit (e.g., security, treatment, programming, staff resources). Step 2: Make Sense of the Data In this step, raw data are turned into policy-relevant information or actionable knowledge. Raw data must be analyzed and interpreted to clarify jail processes and to formulate theories to explain the data. Analysis of the raw data can reveal meaningful trends and provide insights that lead to critical information with which planners and managers can compare the merits of different solutions. Two resource issues are critical: 1. Data quality. The first critical issue for many jails is the quality of their data. The accuracy and accessibility of data will vary across jails. Some jails have excellent and well-managed data collection processes; other jails may not value data collection, may overwork staff, or may be more casual about the data collection and verification functions. 2. Analytical capacity and skills. This second task—making sense of raw data and transforming it into useful knowledge—requires some technical skills and training.

Step 3: Apply This Knowledge to Jail Decisions Decisionmakers must rely on pertinent data at their disposal, and on their judgment and background knowledge of the jail, to determine what action(s) to take to resolve specific issues in their jails. Not every jail is successful in transforming its decisionmaking culture into that of a smart, information-driven jail. Several key resources and cultural changes are necessary before this can happen, the most important of which are listed below.

Leadership and Motivation Managers serve as role models for their staff when they adopt behavior and attitudes that support and promote the use of data collection and analysis to drive and bolster their policy decisions and to monitor work performance. Managers should also expect data-driven decisions from other levels of the jail system as new IT resources and enhanced capacities for the collection and analysis of data are introduced and implemented.

The Politics of Leadership Jail administrators need to cultivate or improve their political skills in several major areas where issues typically arise when policies or practices need to change.

Running an Intelligent Jail

5

1. Resistance to change. Any major technological shifts that require new practices or skills or that produce shifts in power or control in management may meet resistance at some or all staffing levels. 2. Interagency information sharing. Informational silos, firewalls, and other obstacles to information sharing have always plagued jail operations. Jail managers must be skilled at coordinating activities and processes throughout their complex criminal justice systems and be capable of engaging meaningfully in information-sharing agreements with courts, law enforcement, state prisons, and probation departments. 3. Resource acquisition for DDDM. Successful IT functions in a complex system are based in having a strong, multifaceted information infrastructure, a well-designed MIS, adequate staffing, and the financial resources to support them. Managers must have the vision, understanding, and political skills to successfully acquire these resources. 4. Support for a cultural shift. Not all jails successfully achieve the cultural shift when they convert to using their database information to drive the decisionmaking process. Top managers must lead the way by consistently emphasizing the value of data, implementing quality control procedures, and setting an example for other managers across the jail system. The new culture must value the importance of data collection and analysis and its application to decisionmaking. Managers must value, give priority, and support to their staff’s creative problem solving and data management expertise. Staff must also learn to appreciate the value of data as the basis for informed decisions at all levels. 5. Support for data sharing and using communication channels. Senior administrators must support the sharing of information across communication channels within their jail and with external agencies. Jail staff across departments must make every effort to share relevant criminal histories, classifications, risk/needs assessments, and demographic data with other agencies that need this critical data for their own decisionmaking processes. Relevant performance-based data should be available to staff at all levels of the jail system. It is particularly critical to share data that are related to the key goals of the particular jail (e.g., inmate safety, staff safety, security breakdowns, efficiency). 6. Use of data to drive planning and policy decisions. Senior administrators and planners need to be receptive to using data analysis as the basis for policy decisions and support their staffs in the use of tools such as analytic forecasting to track jail trends and projections. More broadly, embracing a culture of data-driven decisionmaking is a prerequisite for criminal justice jurisdictions if they are to create more intelligent jail systems with faster, more adaptable data-analytic tools tailored to their own systems’ information needs. Collection and analysis of these data will inform all levels of correctional management decisionmaking, including the monitoring of performance-based goals and outcomes, and planning for the future.

6

Tim Brennan, Dave Wells and John Carr

Summary This book focuses on the design and use of management information systems that are essential to Running an Intelligent Jail. In this context, a management information system should provide the information necessary to manage the jail effectively. An MIS may be regarded as a component of the jail’s internal quality control procedures that support the management in understanding and solving the problems inherent in running a jail. However, MISs are distinct from other IT systems—they can also analyze other information sources, such as visual and verbal data, that are often applicable to operations within the jail. Management information systems can store, retrieve, and analyze vast amounts of data that are specific to their institutions and in an accessible format that informs decisionmaking at all levels of the corrections community.

SECTION 2. MEASUREMENT OF JAIL PERFORMANCE AND KEY CORRECTIONAL POLICIES Introduction Public concern about the efficiency and cost-effectiveness of the jail and other local criminal justice agencies; the increasing legislative demands for data-driven, informed decisionmaking; and emerging calls for performance-driven outcomes should prompt decisionmakers to demand rapid improvements in the implementation of well-designed MIS systems and the more effective use of the information collected and stored by these systems. New policy developments and efficient, cost-effective operations must be linked to performance-based outcomes and goals; such goals should be clear and unambiguous. Monitoring of performance and outcomes provides critical input for policy discussions, planning, budgeting, and the forecasting of future trends and resource needs. The courts or other legislative bodies may also impose legal standards on jails that are based partly on whether they are achieving their goals and projections based on their current outcomes and trends. A well-designed, implemented, and fully utilized MIS will provide the necessary data and documentation to inform this process.

Performance Criteria for a Jail In today’s correctional and budgetary environment, the public demands increased performance, accountability, and reduced costs from correctional agencies. A term that may best describe this initiative is performance-oriented government. The goal of performanceoriented government is to spend scarce resources on services and practices that provide the best results in the most cost-effective way. This cannot be done simply by cutting staff or services but only by implementing systems that increase accountability while focusing on quality, cost savings, and outcomes. Within the jail, the most effective means to achieve this is with a data-driven, informed policy and planning process, implementation and effective use of a well-designed MIS, along with more sophisticated data-collection techniques. With these

Running an Intelligent Jail

7

data, planners and elected officials can better understand the jail’s operations and make the adjustments necessary to meet funding constraints and become a more efficient, costeffective, outcomes-based organization. The establishment of measurable, outcome-based standards also allows for the comparison of performance measures across agencies. Specifying the entire range of data needed to support informed performance- and outcomes-based planning and policy development is difficult. Answering the insightful questions raised by jail managers responsible for making policy decisions is a sensible place to start and cannot be overestimated as a source of insights. Policymakers should obtain perspectives on past trends, present levels, and likely future trends of any practice or problem in the jail and the local criminal justice system. It is also important to distinguish between causes, correlates, and consequences of any jail procedure, problem, or trend. Data collection and having an MIS support this process for each stakeholder in the system. The following section describes several of the performance criteria of a jail and the various goals it should establish.

Overall/Global Jail Performance Criteria Staff and inmate safety. A central role of the jail is to provide valid identification of offenders. This identification relies on carefully collected, individual inmate demographics, and background and risk factor data (e.g., criminal history, past convictions, arrests, past behavior problems). Both inmate and staff safety rely on valid identification, classification, separation, and supervision of inmates. If the jail fails to obtain the appropriate background data, the risk of false-negative classification errors is increased and the truly dangerous offender is seen as a low risk. The courts have also ruled that classification is a primary guarantor of inmates’ right to be reasonably protected from violent assault or the fear of violence—thus reducing the risk of litigation against the jail. Public safety. A second role of the jail is to provide public safety. This requires effective classification, housing, supervision, and inmate management strategies that reduce the risk of escapes, walkaways from work assignments, new crimes committed while on work release, recidivism, and erroneous community placement. A new generation of data-driven offender risk assessment tools has significantly increased the ability of jails, courts, and probation and parole officers to determine an offender’s risk of recidivism or flight after being placed in the community pre- or post-sentence. Protection against liability and protection of inmates’ rights. A third role of the jail is to minimize liability and avoid costly lawsuits and monetary awards. In addition to providing a safe environment, jails must provide a quality of life that ensures access to services and meets the needs of inmates’ medical, dental, mental health, nutrition, and clothing needs. Often, the inadequacy of the physical plant is a confounding factor in minimizing litigation. To monitor performance criteria in these circumstances, it is important to collect data that are specific to the limitations of the facility and could result in litigation, such as inadequate space (crowding, poor cell design), poorly maintained or damaged locks, doors, surveillance cameras, inadequate lighting, lack of access to recreation, and so on. The jail may have good policies and procedures in place for inmate safety and access to services, but the limitations of the facility may hinder the access and thus increase the risk of litigation. Rehabilitation programs and work assignments. Successful jails recognize that an inmate’s incarceration is an opportunity to address that person’s criminogenic risk factors

8

Tim Brennan, Dave Wells and John Carr

(that is, those factors that produce or tend to produce crime or criminality). These include substance abuse, criminal thinking, and lack of employment, education, or housing. As a result, inmates’ access to rehabilitation programs is gaining importance in the field. As reentry initiatives are implemented and begin to take hold in local corrections plans, inmate programs are often initiated in the jail and then continued once inmates are released and reenter the community. This practice is an important component of good correctional policy and may reduce recidivism and save taxpayer dollars. Access to work assignments of lower-risk inmates also supports effective correctional policy—it keeps inmates busy, permits extra time off their sentence for good behavior, and gives participants some additional work experience. Recruitment and retention of staff. High staff turnover can be an indication of low staff morale. Competent and motivated staff creates a more professional, responsive environment and helps ensure a fair, equitable, and efficient jail. Identification of staff training needs and the provision of that training are critical. When jails face fiscal constraints and budget cuts, the first responses are often hiring freezes and staff layoffs. Reduction of staff levels, however, can cost local governments more money in increased lawsuits and may jeopardize public safety. It is important that jail administrators are armed with the information necessary to defend their need for these staff positions by monitoring staff efficiency indexes (e.g., job responsibilities, workloads, sick time, personal leave, administrative leave, overtime).

Unit-Specific Performance Criteria Each unit of the jail has its own information needs. To understand what monitoring indexes are needed by the unit and its manager, ask the question, “For what functions/procedures is the unit responsible?” Follow this question with another, “What performance objectives are we trying to achieve?” Each unit needs to stay informed of its workload and work quality indexes, including error rates and late processing, aggregate breakdowns of work performed, pertinent characteristics of the inmate population, staffing levels, and so on. By collecting, analyzing, and disseminating this information to unit workers and other stakeholders, each unit can be managed by using a data-informed process, analyzing the data, and making adjustments as necessary.

Data Stakeholders High-Level Administrators Administrators who run today’s jails must be much more knowledgeable about the use of data to manage, plan, and budget their operations. The industry has seen a change in the backgrounds of jail administrators from traditional law enforcement to more public administration training and experience, which has led to a greater appreciation and skill sets that are more familiar with data collection and analysis to inform the decisionmaking. Administrators who have implemented MISs in their jails are starting to appreciate and catch up with these technologies. High-level administrators need to accomplish the following goals when collecting and analyzing data:

Running an Intelligent Jail • • • • •

9

Meet global performance requirements (monitor trends and impacts). Monitor workload demands and trends. Monitor work done and services provided. Identify gaps between workload demands and the capacity to meet them. Budget for and acquire the needed resources.

Planners and Policy Analysts Specifying the range of data that are needed to inform policy decisions in jails, and criminal justice systems in general, can be difficult. Policy decisionmakers are responsible for asking intelligent questions so that the appropriate data are collected to address each jail’s issues; their ability to bring their insights to this process cannot be overestimated. Policymakers provide perspectives on past development, present levels, and likely future trends of problems such as inmate population growth, increasing staff workloads, and decreasing resources. Historical trends (e.g., levels of jail crowding over the past 3–5 years) can be useful in clarifying how problems emerge and develop over time. Projections of inmate population growth, and how jails plan to use their resources in the future, provide some lead time during which preventive measures and solutions can be implemented. During this stage of planning, it is important to distinguish between causes, correlates, and consequences of problems in any particular jail system. Additional questions are becoming more relevant for jails to ask, including: • • • • •

How are we currently spending our money and resources? How are the jail’s resources being used? What functions, policies, and programs are still being supported? What resources do we need to acquire to prepare for the future? In the context of the local criminal justice system—including the jail—where should our local criminal justice dollars be spent?

Middle Managers A major challenge for middle managers (e.g. sergeants, shift supervisors, lieutenants) is to develop monitoring indexes that are sufficiently unthreatening and nonintrusive to monitor staff activities, workflow, and inmate management decisions at the individual unit or shift level of a jail’s operations. These data-driven monitoring indexes should be used to assess whether line-level operational goals and compliance with policies and procedures are being achieved. Staff performance and compliance monitoring should routinely be assessed and fed back to line staff through shift or unit meetings and reports, graphs, or other media formats. This process of “managing by the numbers” requires objective measurement of line- and unit-level performance indicators that reflect the various aspects of performance and goal achievement. These indicators must be accurately collected and stored in the jail’s MIS and be accessible by using ad hoc reporting tools, and canned reports and by having dataexporting capabilities through the use of other statistical and reporting software. If such monitoring indexes are not identified and routinely collected (or if they remain unanalyzed), managers can only guess at the degree to which the desired policy and performance goals are being achieved at all levels of the jail system.

10

Tim Brennan, Dave Wells and John Carr

Line Staff Line staff are critical to an effective jail MIS because they are the primary collectors and recorders of the jail system’s data. Often, this is a job requirement without any emphasis, planning, or training in how to maximize use of the data collected. Although simple rosters and reports are commonly used, it is less common to find jails in which line staff are provided with the skills and access to data that are necessary for them to create ad hoc reports or to identify adequately the optimum information needs of the line staff. Line staff often do their work in a vacuum—they are aware of what is occurring in their unit but not what is occurring systemwide (e.g., workloads across units, compliance with policy and procedures, changes in offender characteristics, trends). Staff appreciate any information they can get about the inmate population they are managing as well as any changes in workloads, policy compliance, goals achievement, and performance measures. Providing line staff with this information promotes professionalism and commitment to the goals and achievements of the jail.

Interagency Stakeholders (from the Local System Perspective) Jails are often seen as the hub of local criminal justice systems because they play a critical role in the promotion of public safety and offer an important resource to law enforcement, the courts, and community corrections. All of these stakeholders should be concerned about how the jail’s limited resources are used. It is now widely recognized that, primarily, policy factors dictate the size and makeup (utilization) of a jail’s population and can be affected by several factors, including: • • • • • • • • • • • • •

Crime rates. Arrest rates. Policies on court pre-trials (bail bonds) and sentencing. Policies on arrests by law enforcement. Policies on early releases and less time served for good behavior. Community corrections policies and alternatives to incarceration. Prosecutors’ charging policies. Department of Corrections state transfer policies and practices. Detention policies in local immigration and customs enforcement. Technical violation policies on probation and parole. Court- and inmate-processing delays. Media and public perceptions. State and local politics and legislation.

To understand how the resources of the local jail are being used, members of the local criminal justice system must understand the policies and practices regarding the factors listed above. As can readily be seen, numerous critical stakeholders affect jail operations. It is becoming common for counties to establish a local criminal justice advisory committee of key stakeholders and county leaders to better understand jail operations and develop policies

Running an Intelligent Jail

11

or plans to make the best use of limited jail resources. The work of these advisory groups must be guided by an informed, data-driven process.1

Linkage with Courts A well-designed and implemented jail MIS system gives jail administrators and others the opportunities to coordinate and align their activities with those of the courts by identifying the unsentenced/pre-trial population of the jail by the number of days incarcerated. Such reports can expedite the arraignment, pre-trial release, adjudication, and community corrections process. As a second example, the prosecutor and the courts can use these jail reports, along with other MIS system data links, to “fast-track” felons who are most likely to receive prison sentences because of their current offense or criminal histories. To implement such data-driven decisionmaking processes in jail systems, criminal justice practitioners, treatment providers, county commissioners, planners, and other key players must collaborate. This collaboration has several potential benefits, including: • • • •

Coordination of law enforcement, correctional, and treatment policies across agencies. Efficient and early release of targeted offenders from jail into community corrections programs. Reinforcement of a coordinated system of behavior incentives for offenders. Improved coordination of sanctions and treatment programs with the assessed risks and needs of the offender population.

Summary Using well-designed information systems can inform all stakeholders in the local criminal justice system as well as promoting the most efficient, cost-effective use of the jail’s limited resources and alleviating jail crowding. Data-driven policies and practices should support the development of an integrated system of informed case processing, sentencing, and community corrections so that members of the various offender subpopulations who enter the local criminal justice system are matched with the most appropriate treatments, facilities, and agencies. Well-designed information systems, and the data- driven policies and practices that are derived from them, benefit not only staff who work in the criminal justice system but also the offenders and their communities.

SECTION 3. DATA THAT MOST JAILS ARE REQUIRED TO COLLECT Introduction Identifying and planning for the ongoing information needs of the agency are critical. The various information stakeholders referred to in section 2 have both short- and long-term

12

Tim Brennan, Dave Wells and John Carr

information needs. Jail MIS systems must have the capability to organize and provide the relevant data to support both the short- and long-term needs of their agencies. This typically begins with a consensus on the data-driven objectives of the agency.

Short- and Long-Term Goals for Data-Driven Information and Outcomes Short-term data needs tend to focus on counts and statistical tabulations, whereas longterm data needs may focus more on trend charts, comparisons of aggregate data, and projections from a baseline into the future. The jail’s stakeholders must discover the data needs, performance objectives, and outcomes they are required to monitor over the short and long term. The stakeholders must ask general questions that will identify any specific data requirements, such as the following: • • • • •

How efficient and effective are we? What performance objectives are we trying to achieve? What were the actual outcomes or impacts of our efforts? What are the trends of the jail and its inmate population? What are our objectives/needs for the future?

A better understanding of how data are currently used or ignored in the tracking of routine inmate processing, day-to-day operations, planning, and policy decisions by jail administrators and other stakeholders can help them to develop short- and long-term goals and strategies that are data-driven. Some questions that stakeholders can ask include the following: • • • •

What data (automated or manual) are entering the jail? What kinds of data are entered into our current MIS systems or collected by other means (manual or automated), and are these data in a useable format? What data are accessible from other external information systems? What data are leaving the jail (in reports and via data exports to other MIS systems)?

Consensus among the stakeholders is important in developing a sustainable vision of the future that is data-driven. Having a consensus on the short- and long-term information needs of the jail also reinforces use of the data by the various stakeholders over time. The indexes described in this section are all candidates for time-based sequential monitoring in jails. Regularly scheduled reports that provide timelines, trends, and emerging problems in the jail will assist jail managers and policymakers in making intelligent decisions. A critical task for any jail’s IT and MIS systems is to select data elements to measure key correctional policies and organizational operations that the managers or policymakers wish to monitor. Appendix A consists of a spreadsheet presenting the data elements often included in MIS systems in jails and other correctional agencies. These elements are correlated with the questions to ask, and the management needs to be identified, by determining the goals and objectives of the agency.

Running an Intelligent Jail

13

What Drives a System’s Information Needs? The many information needs of jails and local criminal justice systems can be categorized into four broad areas. In appendix A are examples of both short- and long-term information needs and data elements based on questions asked of correctional managers.

Level 1: Information on Routine Inmate Tracking This category of information focuses on the day-to-day processing and movements of individual inmates within the jail. This information is most useful to the line staff who do the routine maintenance and supervision of inmates; it serves as the basis for all other inmaterelated data that are required to be collected in a correctional facility. Level 1 data are the initiating data when an inmate enters the jail system. These data must be readily available and accurate and will be the primary data on which all other information is based. The line staff learn the identities and characteristics of the inmates and of the inmate population as a whole. Level 2: Information on Daily Operations in Long-Term Inmate Facilities This category of data builds on Level 1 inmate data and addresses correctional operations for those inmates who are incarcerated for the long term. Although related to Level 1 data, Level 2 data generally require wider information gathering within the correctional facility. Housing and operational issues, along with inmate treatment programs and court information, are included in Level 2 data. Too often, data on the inmate population of a correctional facility are simply aggregated into one broad, misleading, and uninformative summary of the total population. However, data on the facility’s inmate subpopulations and their distinguishing characteristics are now used routinely. Managers invariably benefit from a deeper understanding and a more accurate picture of the diversity of the inmate populations in their jails. Analyses of these data, maintained over time, may identify changes in the demographics of their inmate populations, data that may be most useful to the mid-level managers, shift supervisors, and jail administrators. Frequently, Level 2 data are used only by jail personnel and are generally not available to the public. Level 3: Information on Day-to-Day Operations Decisionmaking This category of information focuses on the day-to-day operations of the facility, policy compliance, efficiencies, staff and resource utilization, and system alerts. The questions asked and data collected may not always come from the traditional JMS systems that store inmate data. Data may be stored manually or reside in separate systems but are very dependent on Level 1 and Level 2 data and trends. Many corrections facilities find it challenging to relate data at Level 3 to data at Levels 1 and 2, but it is possible, and effective, if accomplished. Examples of data collection for this category include: • • • • •

Staffing management systems. Payroll systems. Financial systems. Inventory systems. Maintenance systems.

14

Tim Brennan, Dave Wells and John Carr

If the areas responsible for these data are not automated, the data should still be collected even if manual systems are to be used. The collected data impacts decisionmaking for all corrections facilities. This level of information focuses on queries and information to support a multiplicity of planning, policy analysis, forecasting, and budgeting questions. Typically, information analyses at this level are faced by the jail administrators and policymakers. Such data allow administrators to raise queries if they see major changes in inmate populations or offender categories for those entering or being released from the jail. Their queries about these trends guide the type of followup data analyses that are conducted.

Level 4: Information on Public Access and Services These data elements monitor the required services to the public and their access to basic inmate information. This information can generally be gleaned from data collected at Levels 1–3 but may need to be presented differently. These data may not contain information that is relevant to MIS data collection but instead involve the management of information of particular interest to the public and can include relevant corrections statistics. This content needs to be frequently monitored to ensure its accuracy for public consumption.

Summary This section addresses the questions to be asked and the data to be collected in the correctional setting. Information in the form of data can be grouped on the basis of how much detail is required for accurate reporting and presentation. Those stakeholders who are new to data collection and analysis can begin to collect basic inmate information as a first step in the process if the jail does not currently use an automated jail information system. Not all systems currently in use will ask the right questions or will be able to collect data that provide the right answers, such as those outlined in appendix A. However, this information does provide the reader with a broad overview of the capabilities of management information systems in the jail setting.

SECTION 4. DATA USES IN POLICY ANALYSIS AND ORGANIZATIONAL MANAGEMENT Introduction This section addresses the use of data for problem solving in policy analysis in a jail context. A well-established framework for policy analysis is presented as a series of steps or stages that characterize virtually all jail policy problems. An additional theme of this section is the rising importance of data and data-analytic procedures for jails in their role as information-based organizations. From initial intake to final release, virtually all key decisions are driven by the availability, quality, and careful analysis of data to support the variety of sequential decisions made by jail personnel.

Running an Intelligent Jail

15

Data-Driven Decisionmaking DDDM in corrections rests on practices (e.g., quality control by qualified management, organizational learning, and continuous improvement) that have originated mainly in industry and are designed to support both decisionmaking and planning at all levels of the organization. The goal of all of these approaches is organizational improvement by the systematic collection and use of categories of data that broadly reflect the functions of the organization at large as well as specific departments. DDDM has several main steps:

Step 1: Collect the Appropriate Data It is critical to realize that the types of data collected will vary across different units or departments of a jail and for the specific functional unit under consideration (e.g., security, treatment and programming, staff resources). Common data categories include: • •

• •

Input data, such as equipment and costs of the labor, facility, and programs. Outputs or work done, such as treatments provided, classifications completed, number of supervision tasks completed, number of admissions completed, and number of criminal history searches. Results or outcomes data such as escape rates, disciplinary rates, rates of inmate injuries, and staff morale levels. Work quality, such as error rates in data collection, percentage of tasks completed on time, numbers of inmates mis-housed in wrong custody levels, and rates of compliance with various correctional standards.

Step 2: Make Sense of the Data In this step, raw data are turned into policy-relevant information or actionable knowledge. Raw data must be analyzed and interpreted to clarify jail processes and help to explain the data. The analysis process transforms raw data into meaningful trends and insights that yield actionable knowledge so that planners and managers can compare the merits of different solutions. Two resource issues are critical: • •

Data quality. The first critical issue for many jails is the quality of their data. The accuracy and accessibility of the data will vary across jails. Analytical capacity and skills. This second task—making sense of raw data and transforming it into useful knowledge—requires some technical skills and training.

Step 3: Apply This Knowledge to Jail Decisions Decisionmakers must rely on actionable knowledge, combined with their judgment and their knowledge of the jail, to select actions to resolve a specific problem. These actions may include: • • •

Revising the jail’s goals. Assessing inmates’ needs when planning programs. Designing responses to issues of noncompliance.

16

Tim Brennan, Dave Wells and John Carr • •

Evaluating the effectiveness of various programs. Reallocating resources.

Two broad themes reflect how data are used at this stage: 1. Data analysis is used to inform, identify, or clarify critical jail issues or problems. Actions are formulated and taken. 2. Data analysis is used to justify some specific action, policy, or procedural change. These actions may lead to new policy goals, and performance or efficiency objectives based on these statistics. Most jails implicitly reflect the importance of data in the volume of staff resources that are engaged in data collection tasks (intake and booking, records staff, classification staff). However, a critical failure in some jails is the understaffing and work overload of these key departments, which may result in poor quality or incomplete data gathering. This tendency to overload the staff who perform these functions may sabotage DDDM and encourage resistance to the effective use of IT and data-analytic procedures. In such cases, any subsequent policy analysis can be undermined by the lack of adequate data. Both upper level administrators and middle managers must understand DDDM and achieve a high degree of competence in its implementation. Training, mentoring, and external support in running an information-based jail are often required, for example, developing the capacity to analyze data, extracting relevant information from an MIS database, and interpreting the data in tables and graphs.

DDDM and Changes in Business Practices Introducing DDDM into the jail system can prompt many organizational and cultural changes—which can be a prime opportunity for upgrading the jail’s processes of data gathering, storage and analysis, improving data quality, and for developing coordination and cooperation across agencies. All of these are required for successful implementation of an effective MIS. The following changes have been particularly noted in jails: •

•

•

Impact on decisionmaking at line levels. A difficult adjustment for many jail staff, particularly classification and security staff, involves the shift to data-driven decisionmaking. Impact on decisionmaking at policy and managerial levels. Data will become more valuable as it is analyzed, implemented, and presented to staff in support of various policy or planning decisions. The adage “knowledge is power” is best modeled by those managers who can most accurately and persuasively organize background data to support specific policy positions. Impact of performance monitoring on work style and morale. Having performance and results monitoring available for use by jail managers may dramatically change how line staff conduct their work;2 it may also create morale

Running an Intelligent Jail

•

•

17

problems if staff are coerced into using it.3 Systems for monitoring data and performance can also be used to document and describe management performance, enhancing managers’ skills and providing better feedback on their decisions. Impact on competencies, data literacy, and staff skills. In a data-driven jail system, higher levels of data literacy are required. Data literacy implies that jail staff (from line staff to administrators) understand basics of how jail data and an MIS system can be used to monitor day-to-day jail functions, assess performance and outcomes, and detect and analyze emerging or preexisting quality control issues and errors. Impact of highly skilled staff and higher technical competencies. Traditionally, jails have not required the skills to handle MIS systems and data collection and analysis. Jails using MIS data to inform their decisions, in contrast, will require new, diverse, and continuously improving analytical skills. Emerging evidence-based practices, theory-driven assessments, reentry programs, and theory-guided treatment plans will also require more substantial training for the staff and management of certain departments (e.g., treatment staff and treatment directors).

The Seven Stages of Policy Decisionmaking Although the description of organizational change and the politics of information gathering hinted at ways in which jail data can be applied when addressing a jail’s policy issues, it does not provide a roadmap of specific steps to more effective management and policy analysis in jails. However, the next sections describe the seven stages of a wellestablished model of policy decisionmaking that is driven by data (see exhibit 4.1) and that can be applied to solving or improving virtually any policy issue. • • • • • • •

Stage 1: Monitor routine data and detect problems early. Stage 2: Analyze and describe the problem. Stage 3: Understand and frame the problem. Stage 4: Design solutions that address the problem. Stage 5: Evaluate solutions and select one. Stage 6: Implement the solution. Stage 7: Monitor the impact and outcomes of the new policy or program (repeats Stage 1).

Many criminal justice planners and administrators are familiar with these stages, and this policy problem-solving cycle (often with slightly different nomenclature) is included in many college courses in policy analysis. The following approach applies these seven steps to the particular problems of managing a jail when using a data-driven approach.

Stage 1: Monitor Routine Data and Detect Problems Early Monitoring a range of key indicators is the first key strategy for tracking the overall performance and work challenges of a jail as well as detecting potential or emerging problems and trends.

18

Tim Brennan, Dave Wells and John Carr

The MIS information infrastructure of a smart jail includes the careful, systematic monitoring of the jail’s vital statistics over time—for the jail as a whole and for each functional area or department (e.g., booking/intake, security, treatment, and medical). Many of these statistics (e.g., admissions) are monitored in both aggregated and disaggregated form (by gender, type of crime, major offender needs). Thus, jail managers are able to obtain and review routine data on major goals of the jail, policy achievements, and functional status. If any indicator deviates from the correctional goals of the facility, the manager has early warning of emerging problems, can locate organizational trouble spots, and has time to plan appropriate actions.

Exhibit 4.1. Seven Stages of Data-Driven Policy Problem Solving.

Running an Intelligent Jail

19

Exhibit 4.2. Assault Trends.

Data monitoring is critical for identifying problems and placing them on the policy/management agenda. In the absence of clear monitoring data, jail managers may be unaware of or may deny the scope and implications of emerging problems. The availability of critical data (e.g., increasing disciplinary problems) will often determine whether a problem is taken seriously, casually ignored, or placed on the policy agenda. Monitoring data and trend forecasts can powerfully counteract the tendency of some jail managers to avoid problems until they reach critical proportions. When confronted with hard data, managers will have fewer opportunities to avoid emerging problems.

Stage 2: Analyze and Describe the Problem Examples of the monitoring indexes are shown in exhibits 4.2 and 4.3. Although useful for problem recognition, these indexes are generally insufficient to produce the data needed to reveal the scope of the problem. Thus, in the second stage of policy development, managers typically ask for additional data to clarify the emerging problem or issue, gain insight on the extent of the problem, and avoid premature decisions. A reliance on inadequate data when resolving policy issues may lead to wrong conclusions and inappropriate solutions. Thus, several tasks should occur in this second phase.

20

Tim Brennan, Dave Wells and John Carr

Exhibit 4.3. Jail Population Forecasts.

Explore hunches and generate new questions. Jail managers must explore hunches, ask questions, and request additional data to better understand the issues brought up in the first and second stages. These queries will determine what additional data still need to be collected from the offender samples and what additional data elements need to be included to conduct meaningful statistical analyses. It is counterproductive to start formal collection of new data without such preliminary queries.4 The goal is to identify the key factors that caused the problem. Ensure that necessary data are available. Often, the required data are available in the jail MIS or in the databases of other criminal justice agencies, which typically contain a vast array of data elements. Access to the jail’s MIS data is critical to solving most policy problems, so it is important to find out who controls the access. If the necessary data have not been collected, new data can be collected in the short term but may involve additional costs, work hours, or personnel. Collect data that are aligned with the desired goals. Policymakers are responsible for generating intelligent questions based on past developments, the present situation, and likely future trends of any problem (e.g., jail population growth, contraband increases). Their role in aligning the data needs with the jail’s goals cannot be overestimated.

Running an Intelligent Jail

21

Stage 3: Understand and Frame the Problem The third stage of the policymaking process is developing an understanding of the problem. Generally, policymakers feel that they are on safer ground when they can answer “why” questions. Policymakers and administrators usually have their own hunches, preferences, and preconceived notions about the reasons for problems in policy. If the hunches and biases of policymakers are confirmed and a clear picture of the problem is presented, effective policy interventions can be formulated and justified. Policymakers and managers must accurately communicate their questions and hunches to the statistical experts so they can conduct the appropriate analyses, avoid biased interpretations, and provide useful answers. Data-driven explanations provide rational justifications for policies and solutions: Policymakers are more effective when they can justify policy solutions that are logical and coherent, particularly if they are supported by the data and based on a validated model. This becomes critical when attempting to resolve highly contested issues (e.g., whether to build a larger jail or increase the diversion of prisoners to community-based programs). Conversely, advocates of a particular policy option (e.g., increasing pre-trial releases to reduce jail overcrowding) are less able to justify their policy solution if it is not linked to an explicit, testable model or hypothesis. Stage 4: Design Solutions That Address the Problem The fourth stage of the policymaking process—evaluating and comparing policy solutions—builds on the previous stages. The systematic use of data in evaluating and testing potential policy solutions can transform the policymaker’s search from a trial-and-error approach to a focused evaluation of each proposed solution. Consider some common approaches that policymakers use to arrive at policy solutions: Trial and error. Many criminal justice policymakers rely on hunches, untested biases, and a trial-and- error approach. In the case of jail overcrowding, a trial-and error solution might be to expand the jail’s inmate capacity, which often backfires and, within a period of 18–24 months, is again over-crowded. Data-driven model. A validated causal model of the problem is perhaps the best strategy to generate logical and well-designed policy solutions. Following are some of the reasons to use a data-driven model: •

•

Models can pinpoint more specific and testable cause-and-effect linkages that may not be obvious to decisionmakers. Models organize and simplify complex data so that policymakers can more easily visualize the problem and focus on clearer paths to a solution. A complex model of a criminal justice problem may suggest multiple interventions. Each option may include specific policy or program changes that may alleviate some causes of overcrowding (e.g., delays in inmate processing by prosecutors, or judicial decisions to divert low-risk offenders to other facilities).

22

Tim Brennan, Dave Wells and John Carr

Stage 5: Evaluate Solutions and Select One Once policymakers generate possible solutions to a jail problem in their jails and are able to evaluate each proposed solution, they must decide which solution is best, even though several options may appear feasible. Often, certain stakeholders base their decisions on “soft” data or rely mainly on subjective impressions or political considerations, which may result in overlooking more effective solutions or choosing less effective solutions. This style of decisionmaking can incur great costs and impose long-term damage on the correctional system. Before choosing a particular option, the jail policymaker should assess the relative efficiency and cost-benefit data of each policy solution (e.g., determining which inmate subpopulations to divert to other facilities to reduce overcrowding). Common approaches to such appraisals are as follows: •

•

Impact and pilot study analysis. This often entails a smaller scale pilot study to assess the likely impact of a new criminal justice policy or program on the total jail population. Small scale pilot studies can be conducted to assess the impact of a new policy or procedural change. A cost-benefit analysis for each policy option. Jail policy decisionmakers often compare cost-benefit ratios across policy options and then provide guidance on which solution has the best ratio of benefits to costs.

If there is insufficient time to study a policy problem systematically, policymaking fiascos can occur. Too often in corrections and jails, one hears the complaint that “last year’s solution has become the problem.” This emphasizes the fact that poorly considered policy decisions often produce unexpected side effects, poor long-term outcomes, and minimal commitment to a poorly conceived policy by those who are expected to implement it. Premature or thoughtless policies also are vulnerable to challenge from relevant data; thus, policy reversals are a frequent occurrence.

Stage 6. Implement the Solution The next major stage is the implementation of the selected policy option. The application of data to jail policy does not end with the selection of the new policy or program. Effective policy implementation is critical; if done poorly, it can undermine the success of any new policy. A jail administrator’s prime interests typically align with the formal design and intent of the policy and its effective implementation. The administrator requires regular feedback and progress reports to ensure that the new policy is implemented in a manner consistent with the policymaker’s intent. Stage 7. Monitor the Impact and Outcomes of the New Policy or Program The final phase of policy formulation occurs when the new policy is evaluated and results and outcomes are routinely monitored to gauge its success, which is essentially a return to Stage 1—routine monitoring of key outcomes and performance indicators. Thus, the data in this stage focus on monitoring the impact of the new policy or procedure to ensure that it is meeting the intended policy goals.

Running an Intelligent Jail

23

Data also are collected to identify and assess any unanticipated effects of implementing a new program or policy. The policymaker will make decisions regarding this new program or policy on the basis of this evaluation (e.g., to expand, modify, or terminate it). Broad strategies to gather data for this final phase are as follows: Process evaluations. In this approach, data is collected and analyzed to assess the degree of compliance with the new policies or procedures. Policymakers and administrators must be assured that their decisions, policies, and procedures have been properly implemented. •

•

•

Monitoring compliance. In some cases process evaluations can include data retrieved from a jail’s MIS if the data can document actions, behaviors, and inmate profiles indicating whether the appropriate procedures are being followed and the goals are being met. Imposing sanctions for noncompliance. Valid documentation of compliance and noncompliance collected by the agency or system managers can be used in conjunction with a system of sanctions. This system imposes costs for noncompliance and rewards for compliance to the program staff involved in policy implementation. Monitoring data on outputs and performance goals. As noted in Stage 1, any new policy or program must be linked to clear, measurable goals. Ideally, performance goals and outputs are measurable and based on the data. Past achievement of goals, confirmation of trends, and forecasting procedures can provide a baseline for future goal-setting and assessment of outputs.

Impact studies. A second intensive approach is to conduct formal impact evaluation studies to collect data on outcomes and results of a new policy or program. As noted elsewhere, impact studies are typically not part of the routine monitoring of data elements in jails. Instead, they are specialized, experimental designs that are mounted occasionally to provide a detailed evaluation of the impact of a new program (e.g., new drug treatment, cognitive therapy, vocational training). Although impact studies are viewed as valuable sources of information, these data are typically not collected or entered into the jail’s MIS. Routine monitoring of outcomes and performance indicators. As in Stage 1, the manager or administrator will rely on routine monitoring of a large variety of performance indicators to assess the outcomes and impact of any policy or procedure changes. The key factor in determining the effective-ness of a jail’s MIS is whether it contains an appropriate set of data elements to accurately monitor policy outcomes, performance goals, workload trends, demographics of the jail’s population, work quality, and work output. Policymakers can then review these trends and results to assess the impact of their decisions on the jail’s procedures and policies. As noted for Stage 1, the jail’s MIS must provide performance data related to all major policy goals (e.g., inmate safety, inmate health, staff safety, program outcomes). If a required data element is not routinely collected, then an IT team, usually in conjunction with jail managers, must identify the missing data elements and include them as part of the routine data collection procedures of the jail.

24

Tim Brennan, Dave Wells and John Carr

Summary This section focused on the uses of data at all stages of policy decisionmaking in the jail setting. Over the past decade, there has been a gradual improvement in the data collection, storage, retrieval, and the management information systems of most criminal justice agencies. There is rising public concern over the lack of efficiency and cost-effectiveness in criminal justice agencies as well as increasing legislative demands for data management and for outcome data. These concerns, coupled with rapid improvements in information technology and data analysis, should lead to significant improvements in the application of data to policy development and decisionmaking, based on analysis of the vast amounts of information stored in the management information systems of the nation’s criminal justice organizations.

SECTION 5. REQUIRED SKILLS FOR JAIL INFORMATION SYSTEMS Introduction Jails collect tremendous amounts of data about inmates, rosters and headcounts, inmate processing and housing, disciplinary matters, grievances, maintenance, and staffing issues. Yet, a frustrating and disheartening finding is that very little of this costly and useful information is properly captured, retrieved, and analyzed so that it can be used to support management decisionmaking. Successful and proactive jails are showing steady improvements in the use of jail data for planning, process monitoring, resource allocations and, in general, improvements in the ability to explore and understand policy and management decisions at a deeper level. Jail managers and administrators are learning the skills involved in taking a more data-driven approach, which can help them answer a variety of questions regarding jail operations and understand the many factors involved in monitoring, planning, and policy decisionmaking that constitute the complex performance of a jail. For most jail monitoring and management issues, it is not expected that jail staff will become expert statisticians. Although it is true that jails are drowning in data, it appears unrealistic to expect that most jail managers—either senior administrators or middle managers—will have sophisticated statistical training. However, most of the statistical and graphical tools that jail managers find most useful do not require statistical tests of significance or complex analyses. Instead, a manager can develop a substantially improved ability to monitor the workload, work performance, and quality of work done by the department or unit without having to use complex statistical methods. Many jail database and MIS systems incorporate easy-to-use management report formats that are capable of producing most of the rosters, charts, and tables used in this and previous sections. These systems have the ability to do frequency tabulations, simple cross-tabulations, pie charts, timeline graphs, and simple linear projections. High school algebra is all that is needed to understand these simple methods of data aggregation. Many management responsibilities (e.g., monitoring workload) can be readily addressed by these simple methods.

Running an Intelligent Jail

25

There are important roles that more advanced statistical methods can play in many jail policy issues (e.g., systems dynamics to understand jail crowding, nonlinear forecasting models). However, the aim of this section is to illustrate several simple skill sets that are available in most current jail environments and can be used productively by jail staff at all levels of the jail.

What Do You Need to Know to Do Your Job? A key skill that jail employees—from senior administrators to middle managers to line staff—must have is to know what data are required, relevant, and pertinent to their particular jobs. Many jail managers have difficulty specifying what information or data elements they need. They often do not think through the links between their job roles and the kinds of data and statistical procedures that may be most useful. The following two simple steps are recommended that may enable a jail manager to make progress in this first critical task of selecting the necessary variables and designing statistical and data-driven approaches to fit with their own job responsibilities and those of other jail personnel. 1. Clarify job responsibilities. The starting point in identifying the necessary data, for virtually any jail job, is for employees to list their job responsibilities. These are usually specified by the job tasks, goals, and broad job design. The simple question, “What am I responsible for?” can be used to prompt employees to enumerate their responsibilities, job tasks, work outputs, and work goals. This list will typically consist of relevant variables that are directly linked to the basic goals of the person’s job. For example, a key task of security staff is to prevent contraband from entering the jail. Thus, a directly relevant data element is the number of contraband incidents that occur each week or month in the jail. This is measurable and can be collected across time to monitor upward or downward trends. 2. Formulate questions about job responsibilities or goals. A clear list of responsibilities, key tasks, and unit goals can help staff and managers to formulate a list of performance and outcome indicators and related questions as the next step in identifying the precise data elements that may be needed. Listing specific responsibilities of jail personnel often points to many data elements and indicators that are linked to each major policy goal of the jail or of a specific unit. Such lists should enable the administrator or IT staff to identify the data elements that are routinely collected, new data elements that need to be collected, and the kinds of reports, counts, or rosters that are needed to monitor a variety of job goals.

Line Staff Line-staff members do not typically engage in big-picture planning and overall monitoring. However, they clearly depend on an effective information management system for individual data on offenders such as identification, classification, more confidential information, and related information about each specific inmate to guide decisionmaking. These decisions often fall into two categories:

26

Tim Brennan, Dave Wells and John Carr 1. Routine decisions. These are relatively simple, routine decisions that are often needed in real time and relate to classification levels, housing assignments, program and work assignments, transportation arrangements, and security arrangements. All of these decisions, to some degree, depend on assigning valid custody classifications; other decisions can only occur once an inmate’s classification level has been established. 2. Nonroutine or anomalous decisions. These decisions affect a smaller number of anomalous inmates who may be different enough from other inmates that the routine decisions are not sufficient; these inmates present a variety of aggravating or mitigating factors or ambiguities. These introduce nonroutine considerations that often require further information to resolve the anomaly and will usually require supervisory review. Such exceptional cases may be delayed until new information is available to resolve the aggravating issues (such as gang membership) or mitigating issues (such as a greatly diminished criminal career).

Line staff require real-time information about individual inmates from criminal and disciplinary histories as well as other data sources. This immediate availability of data is imperative for line-staff decisionmakers; they are engaged in high-pressure processing of large numbers of inmates that must be completed quickly. However, line staff also may benefit from information on a larger scale, available in information systems software, regarding their overall workload, work tasks completed, goals achieved, and feedback reports on basic indicators of work quality. Such information may help line staff to understand the broader context of the jail and how their own jobs make an important and unique contribution to the larger organizational performance and goals of the jail (see section 1).

Middle Managers Middle managers have a narrower focus than top administrators and will require data mainly to monitor factors for their specific units, such as classification, treatment services, foodstuffs, and transport. Thus, their critical data requirements may contribute to the monitoring of: • • • •

Unit workloads. Work performed by the unit. Work quality indexes of employees within the unit. Achievement of goals and results.

Each of these categories will have multiple secondary measures that may apply to only a subset of these issues.

Upper Management and Policy/Planning Staff Because jail administrators are basically responsible for the overall function and performance of the agency, they have a broader scope of responsibilities than other staff members. The senior staff persons must monitor the big picture and typically have a broader view that covers a range of jail operations, workloads, work performance, work quality, and

Running an Intelligent Jail

27

policy goals. Additionally, these upper level managers must become adept at formulating useful queries regarding various ad hoc policy issues that arise. For example, a useful approach to overall work demands consists of graphs for successive time periods (weekly or monthly) that indicate the total numbers of inmates in the jail. This is a useful proxy measure of overall workload demand on staff. Exhibit 5.1 is an example of a time graph that can be applied to any measure that is counted at regular time intervals. Managers may formulate simple questions that can be linked to specific data elements: •

• •

What caused X to happen (e.g., Has contraband doubled over the past three months)? • When and where did this problem arise (specific locations, times, kinds of inmates involved)? • Do I have any hunches about the causes or correlates of this problem? How strongly and for how long has this trend been developing (e.g., a monthly trend in admission rates or disciplinary actions)? Has anything else changed (e.g., What factors, such as greater percentage of highrisk inmates or new staff, might be linked to staff’s greater use of force)? • What other factors may be correlated with or cause this change (e.g., more overcrowding, higher arrest rates, delays in pre-trial release, or changes in police arrest standards)?

In the jail context, the design of management reports and informative data tables often begins by clarifying the job responsibilities of the manager and the departmental goals of the unit. This should result in queries that point to the kind information or reports that are needed by the jail manager. The manager may set up a series of time graphs and forecasting exercises focusing on key issues in the jail to discover emerging data needs across time for the following purposes: • • • • • •

Forecasting total population trends. Planning for facility bed space to accommodate population trends. Planning resource acquisition (training, equipment). Identifying resources to address mandates for inmates with special needs. Identifying resources for inmates’ rehabilitation and reentry needs. Forecasting future staffing needs.

Formulating Queries by IT Staff In order to effectively use the enormous MIS database systems that exist in most jails, a key skill is the manager’s ability to formulate queries based on issues of concern to management (crowding, performance issues, threats to public safety). Bottom-line management queries will focus attention on a specific aspect or problem or on underlying causes of the issue or concern. However, in many cases, jail managers and administrators approach IT staff with illformed queries regarding a jail policy problem. These preliminary queries must be refined

28

Tim Brennan, Dave Wells and John Carr

and modified before being subjected to data analysis. In many cases, IT staff must secondguess the intent of the jail administrator’s query and what specific data elements are needed and then must translate the query into a specific kind of analysis (e.g. cross-tabulations, drilldown exercises) to explore the problem effectively. In other situations, the jail administrator’s request may represent hunches or guesswork in the form of an open-ended question. In such cases, the IT person must attempt to crystallize the query to address the policy problem. Typically, the IT staff or statistician will be more aware of the types of numerical analyses that may be used; searching the MIS and choosing key data elements can provide useful data tables and reports.

Types of Data Presentation Static Counts and Rosters: Monitoring Amounts and Volumes The simplest management reports traditionally used in jails have been counts. These counts represent volume—for example, a measure of workload for the jail as a whole or in a specific unit to determine whether the jail’s population limit has been exceeded. Rosters are similar to simple counts but are often given as lists of alphanumeric data (inmate names) as a management report in table format. Such lists are tallies of inmates that are in a specific status, stage, or module in the jail. A daily roster is typically maintained to list all inmates held in the jail or in certain housing modules or lists of inmates waiting for some processing event (release, transportation, primary classification). Trend Charts: Monitoring Events and Forecasting Populations Many key events and populations (inmate violent incidents, overall workload, staff morale, staff competencies) must be tracked and monitored across time. Such monitoring is often critical to address management queries about the emergence of trends or how a specific population or event is changing over time. These counts provide the needed numbers to identify trend lines, assist with forecasting, and avert future crises. In addition to monitoring events (e.g., escape attempts, staff sick days), jail managers often wish to monitor levels or trends of a variety of specific inmate population groups (e.g., admissions, inmates needing specific services, percentage of nonviolent inmates in maximum security). Provided such events or subpopulations can be counted, these data can be presented as a chart based on historical trends. For planning purposes, jail managers also must engage in forecasting future resource and staffing needs. The data needed for forecasting typically involve a long sequence of measures of the specific factor being forecast (e.g., grievances, total average daily population). Often, a prior 5- to 10-year history is measured at specific time intervals (see exhibit 5.1). Such time-based number sequences can be used to construct time graphs and then applied with a simple projection technique known as linear regression, typically available in Excel or other software programs. This method can superimpose a linear trend line on the time graph by using simple extrapolation of the historical trend into the future (see exhibit 5.1).

Running an Intelligent Jail

29

However, a basic problem with extrapolation methods is that they assume the future will be similar to the past. However, trend extrapolation methods are all vulnerable to factors such as policy changes or demographic changes in the community.

Exhibit 5.1. Jail Average Daily Population Trend.

Frequency Tabulations: Understanding the Inmate Categories Although counts are useful and necessary for jail data collection, they gain meaning when they can be broken down into subcategories (by ethnicity, gender, age group). These breakdowns, expressed as frequencies or percentages, can help staff to understand the structure of any inmate category (for any subgroup or the total population). These simple frequency calculations can be useful in addressing common management queries, for example, what are the characteristics of the jail population in categories such as current offense, security level, or being sentenced or not sentenced? Cross-Tabulations: Slicing and Dicing the Data Cross-tabulations and frequency counts can be applied sequentially to drill down into the data to answer more specific questions about more narrowly defined subcategories in the jail (e.g., minimum security inmates by current offense).

30

Tim Brennan, Dave Wells and John Carr

For example, a query may include offenses bringing inmates into the jail and which of those are felonies or misdemeanors. This type of query is easily addressed by two sequential cross-tabulation analyses: first, for the overall jail population by offense category and, second, a cross-tabulation of misdemeanors and felonies for each offense subset. Exhibit 5.2 illustrates a cross-tabulation analysis of the jail population. Overall frequency counts had shown that in calendar year 2000, the inmate population consisted of 75.5 percent misdemeanants, 17.5 percent felons, and 7 percent civil offenders. However, of particular interest is the finding that 40 percent (or higher) of the population were incarcerated for probation violations (24 percent) or other court technical violations (e.g., 16 percent for failure to appear).

Exhibit 5.2. Jail Primary Offenses.

Running an Intelligent Jail

31

Exhibit 5.3. Jail Length of Stay.

Another example of drill-down procedures in a study of jail crowding involves a manager’s basic query to know the number of days in jail for various segments of the jail population. This drill-down process produced the graph shown in exhibit 5.3. The results of the drill-down process in exhibit 5.3 established the frequencies or length of stay (LOS) categories. The drill-down first removed the short-term, revolving-door population (i.e., those inmates booked into the holding area and then released within 72 hours). The average LOSs for the remaining jail population was 32 days. A critical discovery was that misdemeanants accounted for 67 percent of all the days served in jail. A second key finding emerged when comparing the number of jail days with the LOSs of different segments of the jail population. Exhibit 5.3 shows that the segment of in-mates incarcerated from 3 to 10 days accounted for 38 percent of all inmates but used only 7 percent of the beds. By comparison, the inmate population who were incarcerated for more than 90 days (8 percent) consumed 40.5 percent of the bed resources. This analysis demonstrates that LOS is a critical contributor to a jail’s population and its crowding problems. A manager can drill down to clarify data on the subpopulations within the jail, but these data would normally be invisible if counts or percentages were used for the total population. The manager can repeatedly cross-tabulate these specific subpopulations with other relevant factors to reveal important data about these subpopulations. Exhibit 5.3 shows that subpopulations with longer LOSs—only a small segment of the total jail population—

32

Tim Brennan, Dave Wells and John Carr

contribute the highest number of days in jail and, as a subpopulation, contribute the most to jail crowding. These drill-down procedures are typically constructed using the management reports module found in most current jail management software. The skills to use these procedures are simple and involve setting up reports that generate the sequenced cross-tabulations to address such queries. Innumerable jail queries about a variety of jail policy problems and subpopulations can be addressed by using such drill-down procedures.

Frequency Counts and Bar Graphs: Simple Data Pictures Presenting data in visual form, such as bar or pie graphs, can have a powerful visual impact. Some simple policy queries can ask for frequency counts on specific subpopulations or event outcomes in the jail. This analysis involves a simple cross-tabulation to compute the frequency of selected categories (e.g., a breakdown of jail population data by current offense) for selected subpopulations. Exhibit 5.4 shows the relative frequencies of current offense categories for a random sample of minimum-security inmates. Exhibit 5.4 shows the results of a drill-down of data on minimum-security inmates that was initiated by the query, “What primary offense categories comprise the minimum-security population?” A simple cross-tabulation is designed to answer such questions. In this case, a cross-tabulation of each offense category (assault, property, fraud) with the jail’s security classifications (maximum, medium, minimum) was conducted.

Exhibit 5.4. Minimum Security Offenses.

Running an Intelligent Jail

33

Exhibit 5.5. County Jail Security Profile.

Pie Charts A second graphical way to present data visually is in a pie chart. These types of visual displays can represent total populations or any relevant subpopulation. Exhibit 5.5 shows the relative frequencies of maximum-, medium-, and minimum-security inmates in a small rural jail; the minimum-security subpopulation is 53.5 percent of the general inmate population. Comparisons Before and After New Policies and Procedures When policy changes are introduced, it is often possible to conduct a comparison before and after the specific outcomes (disciplinary incidents, inmates eligible for GED classes, inmates in various security categories). Changes in percentages of categories across time. Such changes may occur, for example, when jails adopt a new classification system or new processing policies. Often, it is important to assess the changes in the numbers of inmates at the different security classification levels. This analysis uses a simple cross-tabulation of two frequency breakdowns to compare how inmates are classified by the old and the new methods (see exhibit 5.6). This exhibit cross-tabulates the current security classifications (low, medium, and high, in the left column) with the new security classifications (low, medium, and high, in column headings). Exhibit 5.6 shows how offenders are classified by each system and where the two classifications agree or disagree. This exhibit shows that the old system places a

34

Tim Brennan, Dave Wells and John Carr

higher number of detainees into medium security (447), whereas the new system allocates only 222 to medium security, revealing that the new system redistributes many offenders into minimum and maximum security for a more even distribution of offenders in the three custody levels. Agreements between the two classifications are those cases in the diagonal cells of the exhibit, often called the main diagonal, that is, low-low (29), medium-medium (160), and high-high (20) cells. The off-diagonal cells indicate disagreements between the two systems. The totals (or percentages) in the margins of the exhibit provide the numbers of inmates classified as maximum, medium, or minimum security by each system (e.g., 82, 447, and 71 for the old classification system, and 232, 222, and 146 for the new system). Changes in mean values of key performance indicators. In a similar manner, there can be comparisons between old versus new time periods for the mean value of any particular measure (e.g., inmate grievances, rates of disciplinary infractions, staff absenteeism). These before-and-after comparisons may yield useful indicators of the impact of the policy or procedural change. It is important to acknowledge that these simple before-and-after comparisons do not attempt to use a careful experimental design, which is the gold standard against which to assess the impact of any new or experimental change in policy or procedures. Careful experimental designs can sometimes eliminate factors that confound any claims about the impact of a new procedure. However, these experiments are difficult to design in the real world of busy jails; their requirements of random assignment of inmates to experimental and control groups are also difficult to achieve in jails. Thus, tightly structured experimental designs are relatively rare in the jail context. Data dashboard designs for diverse stakeholders. A data dashboard contains a set of critical graphs and charts and a set of key numbers that allow managers to quickly assess the status and performance of the unit or department. (See a sample of a typical data dashboard in exhibit 5.7.) More specific data dashboards are often designed to assess key jail workloads and performance goals for specific departments, each with different concerns and responsibilities. Thus, separate dashboards with different key data indicators are customized for different jail departments and categories of users (e.g., line staff, middle management, administrators, and relevant public groups). The following resources may reflect different stakeholders with diverse responsibilities: Exhibit 5.6. New Inmate Security-Custody Levels Old Custody Levels Low security Medium security High security Totals

New Custody Levels

Totals

Low

Medium

High

29

35

18

82

179

160

108

447

24

27

20

71

232

222

146

600

Running an Intelligent Jail

35

Exhibit 5.7. Sample of Data Dashboard.

•

•

•

Jail administrators. This level may require a data dashboard that displays a broad range of aggregated jail data that cover all major correctional goals (e.g., security, population trends, treatment provision, staff and inmate safety, staff morale, staff training and competence) as well as changes over time in overall jail population characteristics and other relevant policy factors. Such jail-wide indicators should address policy priorities that are measured at regular intervals so that trends and unexpected problems can be discerned quickly. A data dashboard with key data displayed in consistent formats across time should give administrators the information needed for quick identification of performance goals, changes in trends, and work quality levels of all departments. Appendix B shows an example of a jail administrator’s data dashboard. Community/public dashboard. This may focus on indicators of major public concern (e.g., crowding, rehabilitative services, financial value of inmate community service work, goal achievement for public safety, and inmate incapacitation and rehabilitation). Allocations of public funding may be reflected in staffing patterns, training accomplishments, community volunteer services for the jail, and other indicators. Interagency inmate population control committees. Data dashboards of interagency committees may reflect the multiple causes of overcrowding, population

36

Tim Brennan, Dave Wells and John Carr trends and projections, admission categories, release rates by inmate category, and detailed analysis of jail days by the different offender target populations.

Critical indicators for data dashboards. An indicator is a data element that a user may need to monitor to ascertain a jail’s performance relative to several correctional goals (e.g., public safety, staff morale, security risk management). For example, the goal of public safety can be assessed by using indicators such as (1) escapes and walkaways, (2) recidivism rates for violent offenses within a specific time-frame following release, and (3) return to jail for violent offenses. Each goal may have several indicators or data elements that yield information about the performance goal. Each jail should develop its own set of indicators for the major goals of the jail.

Making Predictions with Data: Simple Forecasting Predictive analyses can be applied to any data that are collected at regular intervals (e.g., average daily population, suicide attempts per month, grievances per month by category). If this data stream has a substantial track record over time (e.g., more than a year) and a sufficient number of data points (e.g., 30 successive months or 10 successive years), then it is possible to compute the simple linear trends by using widely available linear regression software (see exhibit 5.1). Linear regression software calculates a straight line that represents the best fit for the string of data points. This line broadly indicates whether the factor that is measured on a regular basis (e.g., ADP) is increasing, decreasing, or static. Of particular importance, the linear regression provides the rate at which the factor is changing for each unit of time (e.g., ADP is growing at the rate of five inmates per month). More generally, forecasting may require very complex analyses when cyclical and nonlinear processes are involved (seasonal cycles, weekly cycles, recurring holiday events). These add a great deal to the complexity and are better handled by professional statisticians. A further complexity that often undermines trend forecasting occurs when the criminal justice system changes its policies, when sentencing policies change, or when a jurisdiction experiences economic or demographic changes. Such changes can have profound impacts on trends that cannot be estimated by relying on prior trends.

Summary This section reviewed a number of simple data analysis procedures that are often used by jail administrators and managers for basic monitoring and management tasks. These procedures do not require the jail personnel to be trained as statisticians. Statistical training substantially enhances a manager’s ability to organize and interpret these data and to use more advanced techniques. However, jail officers are acquiring substantial training in statistics at increasing rates; thus, they will be able to participate in more sophisticated research designs and use more sophisticated methods of data collection and analysis. Note that jail MIS software packages now include useful management report modules, along with Excel, PowerPoint, and other systems that support most of the techniques

Running an Intelligent Jail

37

described in this section. These systems are designed to be easy to use—steady growth is expected in the sophistication and effectiveness of the management reports being produced for jails. This capability will allow jail managers and administrators to monitor most of the key processes and goals of their jails. Having this skill set will also support policymakers in tackling jail policy issues and applying the relevant data to the basic problems that confront jails.

SECTION 6. PLANNING AND DEVELOPING INFORMATION SYSTEMS Introduction This section addresses the important steps involved in planning and developing an information system plan for comprehensive jail data collection systems or smaller, specialized systems. Depending upon the complexity of required functionality and other factors related to implementation, some of these steps may be relatively straightforward. In other instances, each step could be quite involved and may require significant time and resources to complete. However, the use of a method is important to ensure successful implementation of the jail system regardless of size or complexity of the envisioned new jail data collection system.

Pre-Implementation Steps The pre-implementation steps or phases critical to the development of a jail’s information system include the following: • • • • • • •

Developing a strategic plan for the jail and its associated criminal justice agencies. This includes mission statements, goals, and objectives of the agency and/or jail. Identifying and documenting current capabilities for data collection. Identifying additional requirements and desired functions. Developing a strategic plan for the types of data to collect. Analyzing current capabilities versus required or desired capabilities. Assessing funding capabilities. Developing a plan for data collection (automation) and reports that are based on these findings.

These pre-implementation steps are discussed in this section in some detail. The planning stage of a systemwide automation project is the most critical step in the process and is the groundwork necessary for building a project plan to implement new systems or enhancements to current systems. The planning process for jail systems is critical. Historically, a large proportion of systems in both government and the private sector have been total or partial failures. Even systems that have been considered successful have frequently been implemented with significant cost overruns and/or time delays. It is not uncommon for the jail user community to be less than totally satisfied with some of the functions or have concerns about critical

38

Tim Brennan, Dave Wells and John Carr

functions that are missing from the jail’s information system. Although there are many possible causes for a high level of dissatisfaction, poor or inadequate information system planning is frequently the major culprit.

Developing a Strategic Plan for the Agency or Jail Before the development of any policy and procedures that encompass data collection and management, an agency must have a strategic plan in place to guide its staff and operations in both the short and long term. Without this strategic plan, an agency’s purpose and the road forward are not evident to the employees. This strategic plan must be embraced by the public stakeholders the agency serves and by the agency’s funding source. The plan should clarify both short-term objectives and long-term goals. The plan, once developed and approved by the major stakeholders, should be distributed and adopted by the entire agency. Without knowing what indicators will be used to evaluate the agency’s effectiveness, data collection is meaningless. Identifying and Documenting Current Capabilities for Data Collection Identifying the desired future state of an organization’s systems requires a clear understanding of the current status of its systems. The strategic plan should document current manual and automated systems used by the jail or agency. Current systems should be described in detail from both functional and technical perspectives and each system assessed in terms of their strengths and weaknesses. Part of the strategic planning process will identify how these existing systems will be incorporated in the new systems environment. Options to consider include a) continuing to operate “as is,” independent of the new system; (b) maintaining the existing system but providing enhancements, (c) interfacing the existing system with the new system, (d) replacing the existing system altogether; or (e) some combination of the above. During this planning process it is often discovered that small PCbased systems or other niche systems (i.e., specialized inmate classification systems) have been built or purchased and are critical to a business unit’s functioning. These niche systems, as well as manual systems related to the target business areas, should be identified in the strategic plan. Identifying Additional Requirements and Desired Functions This is the fun, creative part of the process. All operations personnel will be able to identify requirements that they need and would desire to perform their jobs better. Managers will be able to identify the reports that they require if they have established the goals and objectives described in the first step of this process. Criminal justice stakeholders and the public will identify what functions and information they will need and what would be desirable to have. These requirements need to be categorized and prioritized so that when funding becomes an issue, higher priority categories are kept and desired but not necessary categories are sacrificed. These are not business requirements; requirements are defined later as part of each project in the plan. The planning process should ensure that new technologies can be adapted and implemented in the jail system as these technologies evolve. For example, technologies such as handheld scanners, RFID tags, and biometrics are becoming more commonplace in the jail environment. The strategic plan must ensure that the implemented system solutions have the

Running an Intelligent Jail

39

flexibility and open architecture to take advantage of new technologies without requiring a significant rewrite or replacement of the deployed application software.

Analyzing Current Capabilities versus Required or Desired Capabilities Technology itself, and its application in the jail environment, is a moving target. Because the time horizon for a strategic plan is typically 3 to 5 years, ensuring that the right information technology is deployed several years in the future becomes more difficult. The problem is further complicated by the investment of time and resources to implement a new technology. Consequently, the strategic plan must ensure that the planned systems have an underlying technical architecture that enables expansion and the use of evolving technologies while protecting the current investment in the system. Assessing Funding Capabilities It is important that the strategic plan identify the estimated level of resources required for each project in the plan. Both hard costs and soft costs should be specified for each project. Hard costs are items related to hardware, software, and services from external vendors. In addition to one-time expenditures for these items, ongoing costs for hardware and software maintenance, training, and related costs should be identified in the system’s budget. Soft costs relate primarily to the personnel time of jail, IT, and other staff within the jurisdiction to develop and implement the jail system(s). One-time funding sources and recurring sources of revenue should be documented in the plan. This is one of the key roles performed by the steering committee. At this early stage, it is difficult with any technology project to ensure that all costs are identified and budgeted as accurately as possible. Contingency funds of up to 20 percent of the project costs is one mechanism used to address unexpected costs as development and implementation evolve. It is not uncommon for project requirements to expand, new legislation to be adopted, or other unpredictable factors to expand the scope of the project during the development phase. Contingency funds provide a means by which to plan for inevitable changes and unknowns in the typical IT project while minimizing the need to procure new monies for the project.

Developing a Plan for Data Collection and Analysis Based on information gathered in earlier steps, the next major step is the development of a plan. This can be a formal, long-term strategic information systems plan (which is preferable) or a project charter specific to each project. If the objective is the phased development and implementation of systems over an extended time frame, an information systems plan for a 3- to 5-year period is appropriate and may encompass a strategic vision for deployment of technology projects. A more limited project charter may be applicable to one specific project. In either case, the contents of the strategic plan and a project charter are similar; the primary difference is the scope and depth of the plan. The heart of the plan is determining the systems, their components, and specific projects and information needs that will be required to address the identified business problems. This is accomplished by prioritizing in the current systems environment; the strategy or approach

40

Tim Brennan, Dave Wells and John Carr

can take several forms. A single, comprehensive system or multiple, interfaced systems may be defined. The advantages and disadvantages of alternative strategies are discussed in more detail in other sections. The information systems strategy will be based on assimilating the identified priorities, relating these to the current systems, and then identifying the systems required to address the priorities. This is typically an iterative process involving the various stakeholders. The system strategy is refined until one or more systems are identified. The systems are then segmented into discrete projects, and a strategy is developed to transition from the current state to the desired future state. Multiyear plans that involve several projects are typically updated on an annual basis to reflect changes and evolving business needs. The plan should identify the factors that are critical to the project’s success. Critical success factors will vary depending on the political climate, current state of automation, and other factors that may be influenced by, but not necessarily under the complete control of, the steering committee. Critical success factors may relate to the availability of funding or budget approval, cooperation of other justice agencies at the local or state level, agreement to major changes in business processes, or several other factors specific to the economic, organizational, and political climate of the jurisdiction in which the jail operates. In the final analysis, a strategic plan for an information system will only be successfully implemented if there is consensus among the stakeholders throughout the organization on the goals, priorities, scope, budget, schedule, and other critical components of the plan. Consensus building is an iterative process facilitated by an active steering committee and involved user groups. Although it is highly unlikely that total consensus can be reached on all aspects of an information system plan, there needs to be agreement on the basic tenets of the plan to mitigate the risks of project delays or failure.

Organizational Structure: The Right Team The cornerstone of the system planning process is to put in place the right organizational structure. A common theme throughout this document is the importance of fully engaging stakeholders. This is certainly the case for information systems planning. The right team must be assembled with appropriate representation of the agency from several levels within the organization, including executive, middle management, and line personnel. Consensus building for the scope, goals, budget, schedule, and other facets of the information system is extremely important. Organizational structures, typically put in place to manage the planning process, include an executive steering committee and more than one user group. The steering committee usually includes jail and IT executives and other management-level representatives from the jurisdiction that are stakeholders in the jail system. This could include representatives from local police agencies, state prisons, budget officers, prosecutors, and other local criminal justice agency stakeholders. The steering committee provides project oversight and addresses policy issues as they occur throughout the system development process. User groups also are a critical component of the organizational structure. Typically, subject matter experts across the disciplines impacted by the jail system are represented in one or more user groups. Depending on the number of disciplines included in the planned system and the size of the jurisdiction, multiple user groups may operate under the auspices of

Running an Intelligent Jail

41

the steering committee. The user groups will be more involved in the definition of requirements, working with the technical development team throughout the design process, testing all components of the system, and guiding project development from the planning stage through implementation.

Priorities in Developing a Jail’s Information System Because jail operations encompass a broad range of functions and information needs— and because changes in an organization can be challenging when new technology is introduced—it is important for the plan to clearly identify the priorities for system development. The basis for prioritization can be one or more of many factors, including but not limited to the following: • • • • • •

Interdependent system functions, requiring some components to be implemented before others. Prioritization of business problems. Ease of implementation and timely retrieval of relevant data. Political priorities. Needing complex interfaces with other systems. Funding limitations or other constraints.

It is important for the steering committee to reach consensus on priorities and to document system development priorities. This aspect of the planning process is critical in finalizing the strategy for the jail’s development and implementation of a new system.

Documentation of Business Requirements Strategic plans should include tasks and activities related to documentation of detailed business requirements that address the identified problems. Requirements must be fully understood and clearly documented early in the system development process. Requirements documentation is often not given adequate time and staff resources to comprehensively complete this task. There is frequently pressure to implement a system in a short time, which may lead to the temptation to jump right to implementation of a system solution without a clear understanding of the requirements. This is a problem unless the agency has previously identified and clearly documented requirements. Even in this instance, when selecting a commercial, off-the-shelf vendor, it is useful to confirm required tasks before testing and implementation. This will ensure that the scope of the project is clear from both the agency and vendor perspectives. Requirements identification is one phase of the jail system development process and is, in fact, a project, in and of itself, that requires management. A project schedule with milestones, identified tasks, assignments, budget, and other components of the project plan must be documented. Depending on the scope of the new jail system, the level of resource

42

Tim Brennan, Dave Wells and John Carr

commitment, and the existing documentation of the current systems environment, this could be a two-month endeavor or it may extend to a year or longer. Frequent interaction with the user groups and oversight by the steering committee will be necessary to manage the requirements identification process effectively. The project team will be held accountable to these groups to provide a quality assurance process. The steering committee will be the final approval authority of the requirements documentation. The plan itself may require updates upon completion of the requirements analysis. It is not uncommon to make some changes in scope, project schedule, and strategic direction as a result of the more detailed analysis in this phase of the development process.

Business Process Re-engineering Business process re-engineering is a term that refers to changes in the procedures and processes for meeting the operational needs of the jail at the time of introduction of new systems and technologies. In terms of strategic planning, it is important to recognize the willingness and level of acceptance within the organization for process change concurrent with implementation of a new system. To limit strategic planning to automation of existing business practices is usually not the best practice. In many cases, efficiencies can be gained by improving work processes rather than simply converting from manual to automated approaches to the same business practices that have been in place for some time. Training issues, resistance to change, and other implementation issues must be addressed when any significant change in the business process occurs.

Information-Sharing Strategy and External System Interfaces Recognition of jail business partners is a critical component of the planning process. Even if information sharing with the courts, law enforcement, district attorney’s office, state corrections, and other agencies is not envisioned in the short-term, systems planning must account for the inevitable sharing and exchange of data through information technology in the future. To ensure that the underlying technical architecture supports system interfaces and information sharing is an important consideration for the strategic plan. Rarely are new jail systems developed without system access by other justice agencies and without interfaces to external systems.

Project Schedule and Timetable The next step in the planning process is to clearly identify the schedule and timetable for each project and project phase in the IT/MIS plan. All major tasks and milestones become part of the project plan. A long-term strategic plan is typically limited to high-level tasks and milestones for each of the plan’s components. Detailed planning for major tasks such as acceptance testing, training, and deployment are deferred. Tasks and activities relevant to IT/MIS projects are discussed in some detail in other sections.

Running an Intelligent Jail

43

Summary It is not uncommon for jail information systems to be planned with only a vague idea of what is really wanted and needed. Stakeholders may reach a consensus that a new computer system is needed without considering the specific business problems that the new system will need to address. The plan or project charter should clearly document the scope of the system from a business perspective. The planning document need not provide detailed requirements; however, each business or system function, problem, and need should be documented. Specifying what is and is not within the scope of the project plan can also be helpful. A clearly defined scope is vital in managing user expectations and controlling the development process.

SECTION 7. IMPLEMENTING INFORMATION SYSTEMS Introduction Jail system implementation is a broad topic that cannot possibly be addressed in an exhaustive manner in this section. Instead, some guidelines, an overview of methods and approaches to ensure effective initial implementation, and the continued evolution of the jail system with functional enhancements and the technical platform are the focus of this section. Implementation is not a single milestone but an ongoing process.

Exhibit 7.1. Major Phases of Implementation and Key Tasks.

44

Tim Brennan, Dave Wells and John Carr

A Four-Phase Model of Implementation This model offers a broad roadmap by which to approach change in the jail’s MIS. The change model can be applied to most situations that require implementation of new technologies, processes, or policies, and it aims to guide managers through such implementation projects. The framework has four broad overlapping and interrelated phases,5 illustrated in exhibit 7.1.

Phase 1: Pre-Implementation The main tasks of this phase include the following: 1. Recognize the initial problem. This task involves presenting a strong justification that a problem exists with the jail’s current systems and/or use of technology and that there is need for change. The staff as a whole, and top management in particular, must understand the deficiencies and rationale, the need for change, and the new vision; otherwise, business as usual will prevail. The reasons behind the change must be clearly communicated. 2. Build a supportive coalition. Change seldom occurs in a jail without strong political support. The ideal is a unified commitment among jail leadership and key stakeholders. To obtain the support of key people who have influence and authority is a priority. If they are not supportive, they may sabotage the IT project. Ideally, such support should be coordinated before the project progresses too far; major stakeholders naturally prefer early involvement in agenda setting and design decisions. 3. Involve a broad base of stakeholders. Any jail-wide IT procedure will typically have broad scope—it may impact multiple jail stakeholders (e.g., security, classification, IT staff). These stake-holders must all be involved; they are more likely to resist if they feel excluded. Incorporation of all key players also offers some direct participation, which typically strengthens their buy-in, their acceptance of the final design, and their commitment. 4. Specify the deficiencies of current jail system performance. The change agent must present a persuasive analysis of the performance deficits of the current IT systems and procedures. 5. Develop a vision of desired goals/benefits. A vision statement of expected benefits provides a sense of direction and motivation. All major stakeholders should agree on the intended benefits of a new system. 6. Develop performance requirements and functions of the new management IT system. This task involves the stakeholders in developing a wish list of ideal performance requirements and specific functions of the new MIS. 7. Mobilize a planning structure to handle the change. This step aims to strengthen the adaptive capacity of the jail. Normal staff jobs are not geared to the management, design, and implementation of change in the IT/MIS design. Thus, new planning structures or committees are usually needed to enhance the adaptive capacity of the jail. These structures may include: • A transition manager for IT (change agent).

Running an Intelligent Jail

45

• An IT steering committee. • An implementation team, including key stakeholders. • A planner to monitor implementation progress. • External IT consultants, as needed. The core transitional team will manage training, planning, design, troubleshooting, coordinating, and maintaining the momentum of the process. Leadership is generally provided by a transition manager. This person must often assume the role of change agent. The selection and skills of this person are critical (e.g., has respect of peers, management and political skills). 8. Review preliminary IT functions and alignment issues. This design task builds on the list of performance requirements and benefits. Preliminary specifications are required to finalize a design for a new system (see section 6). It is impossible to design an appropriate procedure if these specifications are vague. 9. Initiate training and develop competencies with the new software. Major IT changes in jails often require new staff skills and new understandings. For example, a poor understanding of IT functions among staff can result in unrecognized design flaws that can be introduced into the new system by unwary administrators (e.g., gaps in key data elements, inadequate classification methods, poorly designed data screens, unintelligible or missing management reports, inability to produce ad hoc reports). 10. Develop (and continually refine) a project plan. A tentative implementation plan must be developed, maintained, and regularly updated by the transition team. Specific tasks, milestones, and responsibilities must be identified. A critical component is estimating the resources needed to conduct implementation across all phases. The plan should be brief, contain a list of the changes proposed, list why they are important, name who will do them, estimate how long each will take, and determine the sequence in which they are to be completed.

Phase 2: Design This complex phase involves detailed pilot tests and revisions of the initial prototype design of the new jail system procedures, involving the following subtasks: • • • • •

Finalize the system design and performance requirements. Build on preliminary work to specify needs and functionalities of the system. Train staff in the new prototype procedures. Staff must be trained in the new procedures to engage meaningfully in the pilot test. Pilot test the new system, assess whether performance and functionalities have been met, and check the alignment (fit) with the jail. Examine the fit or alignment of the system to the local jail environment using pilot test results, performance testing, and a process analysis in real-life conditions. Make refinements as necessary to achieve the best possible fit with the jail’s needs.

Phase 3: Implementation This phase introduces the new system into the jail’s standard operating procedures. The following tasks are critical:

46

Tim Brennan, Dave Wells and John Carr • • • • • •

Maintain a detailed implementation plan. Develop mechanisms to monitor progress and identify conflicts and glitches. Provide for problem solving and design adaptations as glitches or problems emerge. Allow for continuous planning by emphasizing the continuous, flexible nature of planning and the need to be responsive to the emerging dynamic situation. Transition from the old to the new system (i.e., go live). Standing procedures are often continued while the new system is phased in. Build competence. Successful implementation involves acquiring new skills at requisite levels, and new supervisory procedures may be needed.

Phase 4: Post-Implementation This phase involves consolidation, monitoring, evaluation, and continuous learning from the implementation process. The major tasks are as follows: • •

•

•

•

•

Assess impacts and outcomes of the new procedure and monitor outcomes to answer questions (e.g., “Did the new system reach our goals?”). Evaluate the process to assess the integrity with which staff are using new procedures, as well as their resistance, compliance, goal sabotage, and goal substitution. Make revisions to the system design or procedures as needed. Using postimplementation monitoring, jail managers may identify system features to be modified or added. Conduct debriefing sessions with the transition team to answer questions such as “What has worked well?” “What was difficult?” “What did we learn about change implementation?” Conduct ongoing skills development. The above evaluations may indicate skill deficiencies, a need for new supervision methods, or new statistical reports for jail managers. Provide feedback systems and management reports for all key stakeholders. A new jail system offers a rapidly expanding database with relevant data for all stakeholders to access. Reports should be developed for routine distribution to all units and stakeholders.

Implementing Management Skills Having inadequate or ineffective implementation skills can waste resources, fail to achieve the benefits of a new or improved system and, in some cases, result in the abandonment of the system, with substantial loss of time and financial resources. Furthermore, the software may be perceived as ineffective. This conclusion is clearly misleading if the new software was never implemented effectively, if users did not achieve competency, or there was little fidelity to the original design. Given the continual emergence of new systems and procedures, many jail managers must develop effective skills in planning implementation and must become change leaders. There is a pressing need for clear, systematic implementation strategies to manage system changes in jails.

Running an Intelligent Jail

47

Difficulties of Implementing Change in Jails Experience with jails during implementation of new or changing systems has demonstrated how difficult it is to manage changes in technology and has shown that the success of implementation is often more important than the technical design of the new system. Implementation problems emerge at all phases of innovation and, in some cases, may sabotage the entire effort. In adopting or upgrading information system technologies, remarkable differences exist between jails in the time it takes to achieve implementation as well as competence, function, data quality, and integrity when using the new procedures. Top-down implementation alone cannot force new technologies or procedural innovations onto a jail; it may simply graft superficial changes over deeply rooted attitudes, procedures, and correctional cultures. Several factors contribute to the difficulty of making organizational and procedural IT/MIS changes in jails: 1. No single, standard model of jail technology innovation and implementation exists. Thus, jail managers have no standard strategy to follow when they implement new IT procedures. 2. Reporting software for criminal justice management is usually not designed or documented for easy transfer of data between agencies. Most are tailored to local organizational norms, policies, and procedures. 3. There is a lack of accurate and readable documentation of previous jail system implementation projects that can be used for training. Thus, there is little cumulative development in this topic of implementation.6

Implementing the Pilot Program and the Training Phase This phase focuses on the development of a workable, well-tested design for the new jail system’s features and procedures, initial training strategies, and the completion of a rigorous pilot implementation (or a trial run) to assess the JMS’s achievement of the desired goals, to identify remaining design flaws and omissions of key functions, and identify any further modifications that may be needed. Specific topics covered in this section are user acceptance testing (UAT), system performance testing, training strategies, functional and geographical phasing, and the identification and resolution of system defects during pilot implementation.

User Acceptance Testing UAT is probably the most important level of testing in the implementation process. Typically, the earlier stages of testing, such as unit, system, and integration testing, are conducted by technical and specialized staff. UAT provides end users with the opportunity to test how well the system conforms to and supports actual jail business functions and meets expectations. The basis for UAT is the documented requirements of the new system. Based upon these requirements, test scenarios are defined and specific tests are documented to be used in UAT. Functions (e.g., initial booking, identification, property management, medical

48

Tim Brennan, Dave Wells and John Carr

screening, and classification) are tested individually and then also as an integrated complete process (e.g., the entire intake process, including multiple individual functions as previously specified). A UAT plan should be documented with all of the business scenarios, specific test scripts to support the testing these business scenarios, and expected results for each test script. The test scripts should be comprehensive and identify all of the common variations associated with each business process. It is never possible to test all possible conditions thoroughly; there are simply too many combinations and permutations of intake data that occur over extended time periods. However, the UAT test plan should address all common known variations to ensure that the new system can handle normal variations in business processes. The UAT plan with test scripts and expected results will provide the testers with a basis for the execution of these tests and will report both successful completion and identified problems. Another component of some UAT test plans is so-called bust-the-system testing. The jail system should have a robust design and not fail or abort under abnormal data entry conditions. Bust-the-system testing allows end users to ensure that the system does not fail under any condition of abnormal data entry. In any system, inadvertent user actions will occur on occasion, and the system must be designed to handle these occurrences. UAT test results, not just defects, should be documented in a test results report to provide an audit trail and confirmation that all planned testing has been successfully completed. A reported defect that has been corrected will need to be retested to confirm that the defect has in fact been corrected. Depending upon the quality of the software and the complexity of the system, regression testing could extend UAT test timeframes significantly. Upon completion of UAT, there should be a high degree of confidence that the system meets user functional requirements and expectations and that there is an acceptable level of risk with full deployment of the system.

System Performance Testing System performance testing is another type of test conducted as part of pilot implementation. The purpose of this test is to determine whether the system meets its performance goals. System performance goals consist of both highly technical and userfocused goals. For example, a performance goal might be stated as, “The response time for a booking transaction should be 2 seconds or less with peak load of 100 concurrent users.” Another performance goal might be that data transferred from an interfaced system should be available within five minutes of initial data entry in the original system. Performance expectations should be clearly documented prior to performance testing. Otherwise, meaningful performance testing cannot be conducted. Technical staff will be required to assist with both the identification of system performance goals and a plan to conduct performance testing. Unlike other types of testing that have been discussed, system performance testing is very difficult to conduct in a meaningful way prior to pilot or initial implementation. Although software tools are available for stress and other performance tests, these tools are expensive and sometimes produce misleading results. Performance testing is best conducted during pilot implementation with real users in the real computing environment of the jail. Key system components can be monitored and measured and bottlenecks identified. Based upon performance monitoring, the jail system can

Running an Intelligent Jail

49

be tuned and improved. Once the system is implemented, monitoring system performance becomes an ongoing task that uses system tools that are readily available to measure and report system performance in terms such as the utilization of CPU, memory, storage, network, and other key system components.

Training Strategies Management, local IT staff, and line users must all be trained in the new system procedures and in the ways the system can support jail operations. A training curriculum explains to all staff how JMS technology will meet the many information needs of the jail. This curriculum also includes the following: • • • • • •

Learning the technical procedures of the software and its strengths and weaknesses. Designing management reports. Obtaining critical data from the JMS to support decisionmaking. Meeting legal requirements for collecting objective, high-quality data. Knowing the professional association standards for data quality. Learning and practicing how to use the system to support all jail operations.

In a jail context, there is typically a strong focus on building and maintaining the competency of staff. Major policy, procedural, and technical changes usually require new skills, perspectives, and information. If jail managers poorly understand the roles and functions of IT, they may remain unaware of its capacities or any design flaws in the new jail system. Training plans are essential with any new technology or procedural change; otherwise, current procedures and organizational knowledge may be rendered obsolete. With rapid change, the skill sets of a jail’s staff and its institutional knowledge may deteriorate. Skill building and effective training cannot be ignored. A common training strategy when implementing jail systems, particularly in large organizations, is a train-the-trainers approach. Training staff to be experts in the new jail system and involving them in UAT are important. They train the other jail staff and frequently become the “super-users” who serve as the frontline for ongoing support and technical assistance. This approach can be a cost-effective training strategy that keeps training in-house for new staff and staff on rotation. When the software provider is available, it can support the organization’s training needs, using online tutorials for initial training of new users and for refreshing the skills of existing users.

Implementing the Functional and Geographic Phases Implementing a pilot program can be an effective strategy for phasing in both the functional and geographic components of the new system. In the pilot program, limited functionalities are phased in by subdividing the jail system into manageable components and implementing the program in each component across the system. Geographical phasing, on the other hand, limits the implementation of a jail system to a single facility, or a module

50

Tim Brennan, Dave Wells and John Carr

within the facility, before widespread application of the system throughout the jail organization. There are a number of reasons that an agency may decide that functional phasing is required when implementing a jail system. The pace at which the organization can make changes—and any budget constraints—may limit which jail functions can be implemented in the initial phase. In other instances, further analysis may be necessary before a new system function is well understood and can be implemented. Jail systems sometimes have components that are dependent on each other and thus dictate the order that functions are phased in. External factors also may influence functional phasing. For example, an audit of the accounting system for the commissary and inmate fund may mandate the immediate implementation of a new cashiering system to resolve financial audit issues. Developing an interface may be a lower priority and may be implemented in a later phase. Functional phasing usually makes more sense when implementing a new jail system than making systemwide changes all at once, with its inherent risks. Determine how to implement a pilot program using functional phasing to ensure that all interdependent or linked functions are implemented in the same phase. This avoids situations in which multiple systems, or a combination of manual and multiple systems, are used to complete tasks. Geographical phasing is typically implemented for different reasons than functional phasing. Although budgetary constraints can often play a part, geographical phasing also allows the organization to refine its business processes and introduce new procedures before deploying throughout the jail system. The introduction of new hardware and software technologies may drive the need for geographical phasing. For example, the use of handheld wireless scanners to track inmates’ movements and activities has technological risks. By limiting implementation of a new technology to one part of the jail at a time, the risks are minimized. With geographical phasing, any operational problems or needed enhancements to the jail software can be identified and deployed as part of the pilot implementation. Note, however, that geographical phasing may not be realistic in many situations. The jail jurisdiction may not be large enough to justify this type of phasing. To receive the full benefit of the new system and to avoid expensive parallel operations, full deployment of the new system throughout the jail may be more cost-effective.

Identifying and Resolving System Defects and Problems Managing the processes of identifying, reporting, and resolving any system defects and problems is ongoing during the testing and use of the JMS. Although initiated during the testing phase, this process of tracking system problems and defects continues throughout the life of the system. System defects are expected to be more extensive during the testing phases of the system’s performance and users’ acceptance. However, testing cannot anticipate all combinations of data and processes. Consequently, issues will continue to be identified throughout implementation of the pilot program. Likewise, with full implementation, issues will continue to be identified as unique situations arise but, in all likelihood, will be less frequent and less severe than in earlier stages. As the system stabilizes and matures, the focus usually shifts from identifying defects to identifying desirable enhancements to the system.

Running an Intelligent Jail

51

Any reported issues should be assessed by the IT team immediately after problems are reported. A good practice is to assign a severity level to each reported problem. A best practice is to ensure that an effective mechanism is instituted to inform higher levels of management of issues that are not addressed or resolved in a timely manner. Users’ perceptions of the system in the early implementation can be negatively influenced if significant issues linger without resolution. Once that negative attitude sets in, it can be difficult to change users’ perceptions and acceptance of the new system. The early stages of implementation are a critical time to ensure, to the extent possible, that users’ expectations are met. “Going live” and changing the procedures for daily operations happen during this phase of implementation of the new system. Some key components and other matters to consider during the implementation phase are highlighted in following section.

Putting the System into Routine Use The new system is put in place, staffed, institutionalized, and used in routine daily jail operations This phase also involves ad hoc problem-solving events; this requires careful planning. Unexpected problems may arise and require immediate solutions and input from a variety of users on the staff. Line staff can often be the first to identify new bugs, software glitches, or other user problems that require immediate attention. Communication with management is critical, and staff users must participate in problem-solving activities to share their knowledge of the workplace and how to best to phase in the new IT procedures.

Setting and Managing Expectations Although the vision and expectations for new IT/MIS projects are primarily developed and communicated during the pre-implementation phase, it is critical that in the busy, stressful stage of implementation, there be recurrent reminders of the benefits and vision behind the new IT system and procedures. All senior and middle managers, as well as line staff, will need regular reminders of the rationale, vision, goals, and justification for the new IT procedures. It is common to forget these reinforcements during the implementation phase. Commitment to the changes from the jail stakeholders can be lost if reinforcement of these messages occurs only at the beginning of IT implementation or if administrators are trying to justify a budget and are not willing to invest in the changes. Multiple communication channels also may provide and celebrate progress reports with all of the involved stakeholders. Top administrators should support the project by requiring review meetings to assess key progress and milestones and to regularly stress the benefits and vision behind new IT projects to jail staff, citizen advisory groups, and other stakeholders.

User Acceptance Four major factors are critical in determining the degree of user acceptance. Supervisors and the transition team must carefully monitor these issues during the implementation phase:

52

Tim Brennan, Dave Wells and John Carr •

•

•

•

Maintaining trust and buy-in with frequent communication. The transition team and senior management must be active in building trust and commitment by using open and frequent communication with line staff and other stakeholders. This may involve many strategies: periodic progress reports, memos, announcements of milestones achieved, and meetings. Line staff and middle managers should be encouraged to raise any questions, concerns, or suggestions and be allowed to participate and provide input. Jail leadership must continue to provide a vision of the direction and benefits of the new system, bolster staff morale, provide rewards, acknowledge milestones, and communicate progress toward the goals. Ease of use. User acceptance is tightly linked to user-friendliness and ease of use. A major component of the pilot implementation is to ascertain whether the new IT procedures and software are efficient enough and easy to use and to resolve any user problems. However, during the implementation phase, the introduction of the new system into routine operations in the jail provides a stricter test of the ease of use. Time and workload demands. The transition team must stay alert and maintain open communication with line staff to identify user problems quickly and to generate new solutions. Common problems include cumbersome screen designs, data scattered across different sources and screens, difficulties in locating the needed data, poorly designed management reports, and too many key-strokes to complete simple tasks. Thus, user acceptance can be influenced if the staff are experiencing workload problems. The possibility exists that in real-life conditions, the IT workload may be excessive. The transition team, as well as the supervisory staff, must be vigilant in monitoring staff workload, errors, signs of stress, and staff complaints during the golive phase. Effectiveness and usefulness of the system. A further feature that influences user acceptance is whether the new IT system is helpful to staff in their work tasks and decisionmaking. If the new system provides effective support and high levels of reliability and validity, it will typically have a high level of user acceptance. A key feature, therefore, is the usefulness of the management reports and rosters provided by the IT system. User acceptance will be high if the scope and range of these reports has a good fit with the information needs of the staff, middle managers, and administrators.

Making Changes in Policies and Procedures During the implementation phase, a sufficiently detailed and updated policies and procedures manual is critical and should describe staff tasks, rules, and new procedures for using a JMS. The transition team and unit supervisors must provide adequate documentation of new policies and procedures to staff who will be using the system. The policies and procedures manual will be thoroughly tested during the go-live phase and, if incomplete, will be updated. The transition manager and the implementation committee therefore must be alert to weaknesses or gaps in the documentation of the new IT procedures. Any deficiencies in the policies and procedures manual may hinder training of staff in the new procedures. Another danger is that jail administrators may fail to assign sufficient staff hours to produce and write an effective MIS policies and procedures manual. These managers

Running an Intelligent Jail

53

may underestimate its value and the time and effort needed to develop, maintain, and update the manual. Another common problem is the shortage of staff with appropriate writing skills. Yet, adequate policies and procedures are critically important in the design and implementation of a new jail system, both in terms of procedure and for liability protection.

Considerations in the Transition and Data Conversion Cutover to a new system is a transition that requires careful planning and coordination of the efforts of both the IT and the jail staff. Typically, cutover involves the movement of data from an existing automated system to a new or improved automated system. To accomplish this in an orderly manner, all of the data from the old system—particularly for inmates in custody and for historical inmate data—must be transferred to the database of the new system. A data-conversion plan and cutover strategy is a best practice that ensures this transition is smooth. There will always be disruptions when a new system, and related policies and procedures, are introduced. However, a well-planned cutover can minimize the degree of disruption. Data conversion and cutover to production should include the following activities: • • • • • • •

Developing a conversion plan and design. Preparing conversion programs and scripts. Completing any required data cleansing and preparation. Performing a simulated conversion and check for errors. Preparing a cutover plan, including contingency planning. Rehearsing the cutover process. Providing adequate technical support staff for the cutover.

Because inmate data derived from prior jail stays is so important, it is usually not advisable to implement a new jail system with no historical data. Because historical data are so necessary in supporting jail decisionmaking, data conversion from prior systems is typically mandatory. If data conversion from the old system is too difficult or expensive, there are less desirable alternatives. In some jurisdictions, the old system is maintained for an indefinite time to mine the historical inmate data until the new system has been operational long enough to be a reliable source of prior-stay data. Implementation of this cutover strategy is more efficient and less cumbersome but provides an alternative to full data conversion.

System Use and Quality Assurance Assuring the quality of the jail’s IT/MIS system is the responsibility of staff supervisors—to ensure high-quality staff training and competency and to identify and resolve any gaps in the staff’s skill sets during the implementation phases. In the early implementation phase, staff must learn new skills to use the new IT procedures. A skills gap is a normal occurrence and may be anticipated at the earliest stages of implementation. Transition teams often develop temporary contingency plans (e.g., extra supervision and

54

Tim Brennan, Dave Wells and John Carr

repeated training) to cope with skills deficiencies early in the process; this must be managed appropriately and may lower staff morale and commitment as well as increasing the liability resulting from user errors. A second and opposite problem as staff become more expert is having a skills surplus, which can occur when the new IT methods are mastered. In some cases, this surplus may lower the quality of work because of boredom, feelings of stagnation, or job impoverishment when most of the major decisions are automated. Supervisors must carefully monitor staff for such problems and take corrective action to reassure and retrain them. Quality is also a function of effective supervision. This may involve monitoring staff, making evaluations, and spot-checking for data and decision errors. Supervisors cannot ignore these issues because they are critical to maintaining high-quality data, analysis tools, and fewer errors in the IT system. Findings from such supervision can be accumulated in statistical reports and provided as feedback to line staff, can guide the assignment of appropriate managers for job performance issues, and can point to corrective actions. A sobering finding is that IT capabilities are often dramatically underutilized in jails. Supervision may also include basic process-evaluation methods during the implementation phase to ensure that staff are positively motivated and are using the new IT procedures correctly. Supervisors must ensure that staff are not undermining IT procedures by streamlining, cutting corners, or engaging in other forms of sabotage, and whether IT capabilities are being used to the fullest extent. Process evaluation is a thorough examination of the manner in which the IT/MIS system is implemented, how competent the staff are, and the overall integrity in using the system. Jails spend an enormous amount of money and time to collect relevant data on inmates and their behaviors and jail operations. The large databases that evolve in busy jails are an enormously valuable resource for managers when monitoring jail operations and performance outcomes, and when analyzing policies. Quality assurance also focuses on the quality of the data entered into the MIS and to minimizing data errors. In the implementation phase, the transition team should meet to identify, discuss, and correct any data quality and data verification issues. Data quality is enhanced when jails verify and spot-check the data regularly; however, staffing shortages and workload demands impose severe obstacles. Some large jail systems include routine data quality control checks on a monthly or quarterly basis, although these quality control measures typically have been externally imposed by court orders.

Post-Implementation Phase The management of implementation does not end with the transition of new procedures into routine operations. Several critical issues emerge following the introduction of a new system. These tasks deal with the following questions: • • • •

Have the new IT procedures achieved the desired goals and outcomes? Are they working as expected? What longer term impacts have the new procedures had on the jail? Is the system a good fit and in good alignment with the jail? Are additional adjustments required? Over the long term, have there been any unexpected forms of resistance, sabotage, or loss of integrity in the staff’s use of the system?

Running an Intelligent Jail

55

Such questions cannot be answered until the new procedures have had a chance to achieve their expected impact and not until the organization and staff have adjusted to the new system. The issues identified in this section are critical in the post-implementation phase.

Developing Feedback Loops Post-implementation feedback loops are needed to identify and resolve system problems. To design an effective IT system, it is critical that all stakeholders are involved and have input, particularly if they are end users. Feedback and participation from staff at all levels should enhance commitment and buy-in across the organization. Specific strategies may be instituted for the following processes: • •

•

Establishing effective mechanisms to collect complaints or weaknesses from IT users across the jail (e.g., security, classification, booking, and intake of inmates). Developing management reports to provide IT users’ feedback to all major units of the jail and accurately monitor workloads, work quality, and trends relevant to that unit. Providing routine reports from unit supervisors to IT staff regarding their unit’s information needs, complaints, and suggestions.

Ongoing Training Strategies With normal rates of staff turnover in jails, and the rapid evolution of IT technology, ongoing training is a necessity. In the post-implementation phase, several strategies may be used to further the training, skills, and competency of the staff who use the IT procedures: •

•

•

•

Systematic job rotation involving the use of IT procedures, coupled with appropriate supervisory reviews, should promote the skills of IT staff and complement their formal training. Supervisor training is also important. Incompetent IT supervisors can severely erode the skills of an IT unit. Conversely, highly trained IT supervisors can substantially upgrade the overall quality and expertise of the IT unit and also provide training to other units in the jail that routinely use IT procedures. As noted earlier, quality assurance and problem-solving groups can be invaluable in identifying competency gaps in the staff and ways to enhance the jail’s IT/MIS capabilities. The IT supervisor can be instrumental in organizing these groups, which identify skills gaps and training needs, and provide a forum for discussion and correction of any IT problems. Error-detection procedures are a critical component of IT/MIS training and skills building. Several strategies are available. Conducting a full review of error-detection procedures is beyond the scope of this document but detailed treatments of this topic are available.7

56

Tim Brennan, Dave Wells and John Carr

Skills development and continuous learning can also be achieved with informating8 feedback, in which relevant management reports are designed for each IT job and are routinely provided to staff (e.g., monthly).

Managing Technology Upgrades Technology upgrades occur throughout the life of the information system. In recent years, the rate of change in information technology has accelerated. IT changes may involve any combination of new or updated hardware and software. Hardware upgrades may involve desktop PCs, back-end servers, or introduction of a new peripheral device, such as a bar code reader or a magnetic card reader. In other instances, an entirely new end-user device (e.g., a wireless handheld PDA) may be introduced. As with hardware, software upgrades occur at several levels. An update to the operating system or tools on either the PC or back-end servers may be required. The database may need to be upgraded to a current version of the relational database management system. New or improved interfaces may require software upgrades. The timing and frequency of IT upgrades should be based on several factors, including the impact on users, required changes to the jail application software, how urgently the changes are needed, and the benefits of the upgrade. As with all jail system activities, the process will require planning, management, and technical and staff support during post-implementation. Consequently, it may make sense to make this process routine by bundling new releases and limiting the frequency of upgrades. There may be occasions, however, when a technology upgrade is required to resolve a problem and must be executed immediately.

Summary As with the private sector, jail organizations are beginning to realize the potential of information technology to improve business processes and reduce costs. The trend is positive; more jail agencies are embracing new IT systems and are using them effectively in their organizations. The caveat is to avoid being on the leading edge of technology or at least ensure that a proof-of-concept process has been successfully completed before deploying the new system.

SECTION 8. REQUESTING PROPOSALS FOR INFORMATION SYSTEM DEVELOPMENT AND SELECTING VENDORS Introduction The request for proposal (RFP) process is the primary mechanism with which jails acquire and implement any major project, such as a new automation system. In this process, a jurisdiction prepares a formal solicitation that is released to the vendor community to obtain proposals for a jail information system. This section describes the best practices involved in

Running an Intelligent Jail

57

RFP development, RFP content, and the vendor selection process. The focus is on best practices to ensure that the jail agency obtains a cost-effective system solution that meets the agency’s needs. An RFP is used when requirements are known but the jail system solution and implementation process may vary, thus requiring the proposer to provide a system solution and the approach to its implementation. Price is important, but proposals also will be evaluated and selected on the basis of other criteria to ensure the most desirable solution for the jail. There may be constraints on the RFP process that are specific to local government procurement rules or other considerations for specific jail system procurements that may dictate a somewhat different RFP process.

Managing the RFP Process Managing the RFP process is simply a continuation of the processes identified in earlier sections for managing the development of jail information systems. Ideally, the RFP process represents the culmination of thorough planning and analysis that have already occurred. The participation of other departments within local government will vary and will be based on the size and structure of the government entity. In any case, it is important that the jail, under the auspices of the steering committee or a related oversight group, manage the RFP process and not defer to other local government entities that have the same stake in the outcome of the RFP process. The steps involved in the RFP process are shown in exhibit 8.1. Although some steps are optional and depend on the degree to which business requirements have already been defined and documented, these steps are usually required, either formally or informally, to select a jail system vendor and establish a contract to implement and maintain the system.

Prerequisites When Preparing RFPs Before preparing the RFP document, there are several prerequisites, as illustrated in exhibit 8.1. Most critical of the prerequisites is having a clear understanding of the scope and documentation of the business functions to be performed by the jail system. How to define the requirements was described in some detail in section 6. This functional requirements document (FRD) provides the starting point for RFP preparation. Because of the passage of time, it may be necessary to augment or modify the functional requirements. The functional requirements will need to be supplemented with technical, operational, and transitional requirements that are not addressed in the FRD. Transitional requirements also should be specified in the RFP. These requirements relate to factors such as data conversion in the transition from the current to the new system. Other transitional requirements may be defined and will relate to issues such as training, system documentation, and deployment of the system. Transitional requirements should be reflected in the tasks and deliverables documented in the statement of work (SOW; discussed in a later section of this section).

58

Tim Brennan, Dave Wells and John Carr

Exhibit 8.1. Management of the Request for Proposals Process.

Components of the RFP Several factors should be kept in mind during preparation of the RFP document. Although not an exhaustive list, these considerations include answers to the following questions: • • • • • • •

How time critical is the implementation of the jail system? What are the minimum mandatory requirements for any vendor? Are there opportunities for improvement in processes and practices? What level of staff resources and expertise will be provided by the jail? Are the implementation constraints well understood? Are expectations of the number of users and system growth documented? Are budget constraints and funding sources identified?

Running an Intelligent Jail

59

The primary RFP components consist of the following:

General Information The general information or introduction section of the RFP usually consists of a combination of information specific to the jail system solicitation and boilerplate information common to all RFPs in the local jurisdiction. The background, purpose, overview, terms and definitions, and minimum mandatory requirements should be developed for the specific RFP. Minimum mandatory requirements serve as pass/fail criteria and provide potential proposers a quick basis on which to determine whether they should respond to the RFP. A sample contract may be referenced in this section and attached to the RFP as an appendix. Other items in this section may provide customized information but will consist primarily of boilerplate terms and conditions that are specific to the local jurisdiction. The procurement and/or contracts staff will play a key role in the development of this and the next section of the RFP. Proposal Submission Requirements This section identifies the specific format and contents expected in the proposer’s response to the RFP. If the vendor’s response does not conform fully to submission requirements, the evaluation score may be significantly reduced or the proposal may be disqualified. Typically included in this section of the RFP are the following components: 1. Proposer capability. Responses will be required to describe the proposer’s qualifying experience as a jail system vendor, provide references, and document the qualifying experience for the proposed project manager and key technical staff. The response also will provide references for sites where the vendor’s system has been installed. Documentation of the company’s financial capability and viability also may be required. 2. Management approach. This component will document how the proposer will manage and execute the project. The vendor will be required to present a detailed workplan including time frames, resource assumptions, and the rationale for staff assignments. The management approach section of the proposal also should include an organization chart, risk mitigation and management, and the proposer’s quality control plan. 3. Proposed system solution. This is a critical section of the response that should explain in detail how the proposed system will address each of the functional, technical, operational, and other requirements named in the RFP. The proposer should present a development, implementation, and support strategy consistent with the tasks and deliverables identified in the SOW. Each specific requirement in the RFP should be acknowledged in the response, including an explanation of how the requirement will be met by the proposed solution. 4. Cost proposal. Proposers must submit a pricing schedule that includes all cost components related to software licensing, development and implementation services, maintenance, and hardware (as applicable). It is important that all cost factors are identified in the proposal to ensure that the total cost of ownership can be assessed as part of the evaluation process. The cost proposal should include the cost of each

60

Tim Brennan, Dave Wells and John Carr task/deliverable as identified in the SOW. A budget narrative that identifies all pricing assumptions should be a required component of the cost proposal. To the extent feasible, costs should be a fixed price to minimize the risk of escalating costs. However, some deliverables (e.g., training and interfaces) will be bid on a time-and-materials basis with an hourly/daily rate and a ceiling price. RFP requirements for withholds or holdbacks (percentage paid only after user acceptance) until successful implementation of the jail system also should be addressed in the cost proposal. 5. Required contract forms. Certain contractually required forms and information must be submitted by the proposer; these vary from one jurisdiction to another. Part of the submission is typically the acceptance of the terms and conditions identified in the RFP or identification of the terms to which the proposer takes exception.

Statement of Work The SOW in the RFP defines the scope of work to be performed by the vendor. All of the required tasks and associated deliverables to develop, implement, and support the jail system are defined in the SOW. The tasks presented in the SOW will vary from RFP to RFP and will be based on the existing system environment of the jail and the level of support required of the vendor to modify, implement, and maintain the proposed jail system. Although the proposed system may require some customization to support the defined functions and/or interfaces, most of the defined services in the SOW will likely consist of those supporting installation, testing, training, the go-live stage, and post-implementation maintenance. The SOW will be based on the defined functional, technical, operational, and transitional requirements. The SOW will be prepared collaboratively by jail and IT staff to ensure that all tasks and deliverables required of the vendor have been clearly defined and are included in proposers’ responses to the RFP. Some typical tasks defined in the SOW for a jail system are as follows: • • • • • • • • • •

Project planning and management. Confirmation of requirements. Installation of software in the test environment. Configuration and/or customization of software and interfaces. System testing and user acceptance testing support. Train-the-trainer and end-user training. Data conversion and data upload. System cutover and go-live support. Final system acceptance. Maintenance and ongoing support.

Most of these SOW tasks are self-explanatory. Inclusion of a final system acceptance is a best practice to ensure that implementation of the jail system is successful. This includes specifying a time frame, such as 60 days, in which the system will need to function with no major defects. Upon completion of this time frame, a hold-back payment would be made to the vendor. A final system acceptance test reduces the risk to the jail of implementation problems and a vendor with less incentive to correct problems after the go-live stage.

Running an Intelligent Jail

61

Each task defined in the SOW should be accompanied by a deliverable. The deliverable represents the product to be provided or the outcome of a completed task, for example, the project plan, the installed software, training and training materials, or the accepted information system. Deliverables typically represent pay points for the vendor and, when well defined, help to avoid disagreements with the vendor about when a task has been successfully completed and accepted.

Requirements Matrix The requirements matrix provides a means of presenting, clearly and concisely, all of the functional, technical, and other requirements in the RFP that the vendor must address. The matrix supports the proposal evaluation process by documenting each requirement using a unique reference number and designating the requirement as mandatory or optional. Additional columns in the matrix provide for a vendor’s response of ‘yes’ or ‘no,’ indicating whether the vendor can meet the requirement, and providing details on how the vendor will meet the requirement with the proposed system. It is important that the requirements be defined as specifically as possible. The completed requirements matrix also supports the evaluation scoring process. A predefined evaluation score sheet for each mandatory and optional requirement is developed as part of the initial evaluation process. As with the SOW, the requirements matrix should be a collaboration between the jail experts and IT staff. A completed requirements matrix helps the proposer and jail project team to avoid misunderstandings. The matrix also confirms the jail system’s capabilities during reference checks, and during demonstrations of the software, as part of the final evaluation and selection process. Evaluation Criteria and the Selection Process This section of the RFP provides the vendor community with a detailed description of the evaluation process to be followed by the evaluation committee and the basis on which to select a vendor. One technique used in the evaluation process is to pass or fail each proposal on the basis of minimum mandatory requirements specified in the RFP. The evaluation also provides specifics on disqualified proposals that have failed to adhere to the required format and contents defined in the proposal. In addition to defining the basis for disqualifying proposals, this section of the RFP identifies the evaluation criteria and the weighting of each criterion. Criteria for a jail system evaluation typically include the following: • • • •

Proposer qualifications. Functional and technical requirements of the proposed system. Approach to the provision of the required services. Cost proposal.

There are no hard and fast rules for weighing the factors for each criterion. Proposer qualifications may be weighted in the range of 20–25 percent. This criterion consists of an assessment of the vendor’s relevant experience and capabilities, based on the verification of references and resumes of the proposed staff. A review of the vendor’s financial capability also may be part of this evaluation.

62

Tim Brennan, Dave Wells and John Carr

The extent to which the vendor meets the mandatory and optional requirements is an important second criterion that may be weighted as much as 40–50 percent of the total evaluation score. This score is determined by analyzing and scoring the response to each specific requirement documented in the requirements matrix. The approach to the provision of required services is a third criterion that may be weighted in the range of 10–20 percent. This scoring is based primarily on the comprehensiveness and methods presented by the proposer in response to the SOW. The last criterion is the cost proposal, which is typically weighted in the range of 30–40 percent of the total evaluation score. Although the cost of the system is clearly important, it is risky to weight the cost too highly to the detriment of other factors, such as how well the system meets the requirements, capabilities, and experience of the vendor and proposed staff. For each cost criterion, maximum points are assigned to the proposal with the lowest overall costs, based on the inclusion of all one-time and recurring cost factors. This ensures that the cost assessment is based on the total cost of ownership.

Evaluating Proposals and Vendor Services The evaluation and selection process is just as critical as the RFP development process in ensuring the successful implementation of a jail system. The vendor—selected to install and support your system for a minimum of several years—becomes an important business partner in the jail’s operations. The first major step is releasing a comprehensive RFP with clear and concise system requirements and services for the jail system. The next major step is to ensure that the best vendor is selected. The evaluation committee should be selected early in the process; it typically consists of five to seven participants. The committee should include a balanced mix of jail experts, technical staff, and procurement staff. Jail representatives on the committee should include jail staff with a history and understanding of the jail system requirements. Members of the user groups previously identified are good candidates for the evaluation committee. There should be representation on the committee by the procurement or contracts unit that was involved in the RFP process and will participate in contract negotiations with the selected vendor. It may also be appropriate to include an evaluation team member from outside the local jurisdiction to provide an objective, external perspective. The evaluation process consists of the following steps: • • • • •

• • • •

Plan the evaluation, including the selection and orientation of team members. Document the detailed evaluation process and criteria in the RFP. Prepare detailed evaluation scoring instruments. Conduct an initial review of proposals to determine any disqualifications. Review the proposals thoroughly and score each proposal, including: • Checking the proposer’s references. • Rating each proposal and completing the evaluation scoring instruments. Reaching consensus on evaluation scores and rank ordering the proposals. If appropriate, interviewing the finalists and seeing demonstrations of the software. Revising scores, based on the interviews and software demonstrations, if applicable. Selecting a vendor and beginning contract negotiations.

Running an Intelligent Jail

63

Any number of variations on this process may be appropriate for a particular jail, depending on the procurement policies and procedures of that jurisdiction.

Finalizing the Evaluation and Selecting a Vendor When evaluation scoring is complete for each qualifying proposal and a consensus score has been reached, proposals are then rank ordered on the basis of composite scores. At this point, a clear winner may be evident and selected. Another option is to select the top two or three proposers and conduct a final evaluation process. In this step, the selected proposers would be given an opportunity to meet with the evaluation committee. During the session, the proposer is interviewed and the software can be demonstrated. Once it is confirmed that the proposed system can meet the RFP requirements, evaluation scores can be adjusted if necessary and the ranking of the proposals can be finalized. The evaluation results should clearly document the strategy for addressing any protests or disputes by the vendor community. Documentation should include completed preliminary scoring instruments and the final consensus scoring documents signed by all members of the committee. The documentation also should include any working documents used in the evaluation process, such as results of reference checks.

RFP Best Practices This section described the RFP development and proposal evaluation processes. This final section identifies and presents, in no particular order, the RFP best practices for procuring a new jail system. Some of these practices have been highlighted in earlier sections; others were mentioned but not necessarily emphasized. Best practices include the following: • • • • • •

•

Include a comprehensive list and detailed description of the functional, technical, and operational requirements in the RFP. Create an RFI or less formal survey that allows reviewers to examine existing system offerings in the marketplace before the RFP was finalized. Identify each requirement as mandatory or optional. Require, in the vendor references, those sites where the software has been installed. Identify all professional services required of the vendor in the SOW, providing clearly defined tasks and deliverables. Include a final acceptance task in the SOW that requires 60 days of operation with no significant defects, and a corresponding hold-back of some funds until this task is successfully completed. Require completion of a standardized, detailed cost proposal, including all one-time and recurring support costs as well as all pricing assumptions, to ensure that total cost of ownership can be assessed.

64

Tim Brennan, Dave Wells and John Carr •

•

•

• • • •

Conduct a bidders conference after the RFP is released, but before the submission of responses, to ensure that all potential proposers understand the proposal submission requirements. Define a system infrastructure in the RFP requirements that is consistent with current standards, such as Web accessibility, open interfaces that provide long-term flexibility, maintainability, and interoperability with third-party tools (e.g., reportwriting tools). Select an evaluation team early in the process that includes members with complementary skill sets, including expertise in jail business functions, IT/MIS systems, and procurement. Consider including an external resource as part of the evaluation team. Evaluate proposals with a balanced approach rather than overemphasizing costs. Prepare an evaluation plan and detailed scoring instruments to ensure consistent, fair ratings by the evaluation team. Consider the use of vendor interviews and software demonstrations before finalizing the vendor selection.

Summary There is no single correct format or specific set of components that applies to all jail systems in all jurisdictions. Most local jurisdictions have specific regulations and exhibits that must be included in all RFPs. Nonetheless, certain components and best practices should be incorporated in the development of any RFP that solicits proposals for a jail system.

APPENDIX A. WHAT DRIVES INFORMATION NEEDS? Level 1. Information on Routine Inmate Tracking Questions to answer Basic Inmate Information Who is being arrested? When are they being arrested? Where are they being arrested? Where do they come from? Where do they live? What is their nationality? A citizen of which country(ies)? Do they have a driver's license?

Information to collect Inmate demographics Dates/Times Address, City, State ZIP Birth City, State, Country Address, City, State ZIP Nationality Citizenship Valid DL number/State

Technique to use

When

Arrest forms/ Interview Arrest forms Arrest forms

Intake

Interview

Intake

Documents/ Interview Interview Interview

Intake

Interview

Intake

Intake Intake

Intake Intake

Running an Intelligent Jail Questions to answer How was inmate positively identified?

Technique to use

When

Livescan/ Fingerprint cards/Facial recognition

Intake

Statute/Description/ Degree/Level Holds/Detainers/Who / Dates No. of warrants/ Agencies Category of charge/Status

Arrest forms/Other paperwork/Databases NCIC

Intake Intake

NCIC

Intake

Policy

Intake

Employer information Address/Telephone number Yes, No, NA

Interview

Intake

Interview

College, high school, other

Interview

Intake/ Classification Intake/ Classification Intake/ Classification

Family/emergency numbers

Interview

Intake

Medical Information Any Injuries?

Type of injury

Intake

Any illnesses?

Type of illness

What is arrestee's medical status? Intake medical screening questionnaire Mental Health Information Any suicide potential?

Medical status

Interview/ Observation Interview/ Observation Interview/ Observation Interview/ Observation

Dates/times of attempts or inclinations determined Dates/times of doctor visits or hospitalizations Drugs prescribed Drug/alcohol usage/amounts

Interview/History

Intake

Interview/History

Intake

Interview/History Interview/History

Intake Intake

Verify history of past arrests/convictions Verify history of past convictions

NCIC/Local court databases NCIC/Local court databases

Intake/ Classification Intake/ Classification

Charge Information What types of charges? Are there holds/detainers? Are there any warrants? What is the primary charge? Status? Employment Information Are you employed? Who? Where? Contact information? Unemployed? Veteran? Disabled? Student? Where? Contacts Contact whom?

Any mental illness reported?

Any psychotropic drugs? Substance abuse (alcohol/ drugs)? Criminal History Past incarcerations? Past convictions?

Information to collect Biometrics

65

(Developed by medical provider)

Interview/Past report

Intake Intake Intake

66

Tim Brennan, Dave Wells and John Carr (Continued)

Questions to answer Past escapes or attempts?

Information to collect Verify history of escape attempts

Incarceration Behavior History Prior incarceration What was prior institutional behavior? behavior Current/prior need to keep Potential enemies in separate from other inmates? custody Was/is inmate affiliated with a Gang name gang ? Family Ties What is your marital status? Single, married, divorced, widowed Do you have children? Number/ages of children Community Ties Do you own a home? Own/rent home or Homeless What is your religious Religion of choice or affiliation? None What is your education level? Highest grade achieved What is your military status? Current or Veteran Inmate Property What personal property came in with inmate? What money came in with inmate? Where is property stored? Was car towed? To where? Whom will you release property to? Who collected property and when? Who released inmate's property? How many people came in with no money? Release Information Who and how do inmates get released? Within what time frame do inmates get released?

Inventory of items surrendered Amount of cash

Technique to use

When

NCIC/Local court databases

Intake/ Classification

Databases/Interview

Classification

Databases/Interview

Intake/ Classification Intake/ Classification

Databases/Interview

Interview Interview

Interview Interview Interview Interview

Intake/ Classification Intake/ Classification Intake/ Classification Intake/ Classification Intake/ Classification Intake/ Classification

Face-to-face inventory Face-to-face inventory Database

Intake

Arrest documents

Intake

Interview

Classification

Receiving staff

Paperwork/Database

Intake

Releasing staff (name, date, time) Zero balance in account

Paperwork/Database

Release

Database

Intake

Inmates released/ Types of release/ Charges Release dates/times

Paperwork/Database

Releasing

Automated or manual entry

Releasing

Property location/Storage type Type of car/Towing company Names of individuals

Intake Intake

Running an Intelligent Jail Questions to answer Who and how many inmates got transferred to other jurisdictions? Who released the inmate? Who transferred the inmate? Were all money and property returned? Trends (If above information can be collected) Total bookings for any time period? Total releases, by type, for any time period? Average daily population? Average length of stay, by type of population? Growth pattern within a timeframe?

67

Information to collect Inmates transferred and dates/times

Technique to use

When

Automated or manual entry

Releasing

Officer release information Officer transfer information Signature of receiving person

Automated or manual entry Automated or manual entry Automated or manual entry

Releasing Releasing Releasing

Level 2. Information on Daily Operations in Long-Term Inmate Facilities Questions to answer Needed Programs What is education level of inmate population? What drug/alcohol use and how much? What treatment programs were attended? Are there domestic violence issues? Are there parental responsibilities? Whom do they live with? Program Completions How many inmates entered each program?

What were the reasons for entering programs? How many inmates completed each program? How many inmates failed to complete programs, and why? How many referrals were made, and what types? What is the capacity of each program?

Information to collect Level of education: High school/vocational/college How much/What type Programs/Dates Charges/Marital history No. of children Address/Responsible party

Technique to use Interview Interview Interview Interview Interview Interview

Admission dates/No. of inmates per program/ Program categories Program requirements/Reasons

Program reports

Types of programs completed/No. of inmates/Dates Types of programs not completed/Dates Referral types/Dates

Program reports

Capacity of each program

Program reports

Program reports

Program reports Program reports

68

Tim Brennan, Dave Wells and John Carr (Continued)

Questions to answer Information to collect How many staff are assigned to each Staff assigned to each program program? Institutional Adjustments and Behavior/Rule Violations What were the major violations and types? Major violations/Types What were the minor violations and types? How many of each type were sustained? How many violations were overruled? How many violations involved contraband? What types of contraband were there? What areas of the jail have high levels of violations? Which staff wrote disciplinary reports, and when? How many inmate assaults were on inmates? How many inmate assaults were on staff? Dietary Needs How many inmates require a religious diet? How many inmates require a medical diet? How many meals were served, by diet category? Facility Movements/Housing What were the movements/activities in the jail with regard to housing? Facility Movements/Events What types of events did inmates request/attend? If events were visitations, who were the visitors? If events were visits by professionals, who were clergy/attorneys/other professionals? How long did the events last (applies to all events)? Commissary How many inmates are indigent? What are the average weekly commissary purchases? What is the amount of revenue in the jail trust account? Inmate Accounting How many inmates have active accounts?

Minor violations/Types Outcome of hearing Outcome of hearing Contraband received/No. of violations with contraband Types of contraband Housing locations of inmates with violations Report writers/Dates and times

Technique to use Program reports

Disciplinary report (DR) process DR process DR process DR process DR process DR process DR process DR process

No. of assaults/Types of violations No. of assaults/Types of violations

DR process

No. of inmates with religious diet/Diet types No. of inmates with medical diet/Diet types No. of meals/Times diet meals served

Interview

Housing locations/relocations and Dates/Times

Data entry all levels

Event types/Dates/Times

Data entry all levels

Visitor names/Relationship to inmate/Address/Dates/ Times Visitor information and Dates/Times Event start/end dates and times

Data entry all levels

Inmate banking balances Types of purchases/Dollar amounts per week Dollar amount of purchases

Inmate banking Inmate banking

All deposits/withdrawals for inmates

Data entry/Report

DR process

Interview Calculation

Data entry all levels Data entry all levels

Inmate banking

Running an Intelligent Jail Questions to answer How much money is spent in the commissary? Types of items purchased? How much money is deposited for inmates, and by whom? How is inmate welfare money spent? What fees are collected, and how much? Housing Assignments How many housing assignments are made per timeframe? How many inmates were mis-housed? Hom many inmates are kept separate from other inmates? How many inmates are in each category of housing? How many inmates are boarded out? How many inmates are boarded by other agencies? What are gang member locations and affiliations? Inmate Grievances What types of grievances were filed?

Information to collect All items from commissary/Dollar amounts Category of each item purchased Depositor names/Deposit amounts Expenditure categories Fee categories and amounts

Technique to use Data entry/Report

No. of inmate housing assignments Housing type/No. of inmates/Housing plan "Keep Separate" information

Report driven

Data entry/Report Data entry/Report Data entry/Report

Report driven Report driven Report driven

Types of grievances/Dates

Paperwork/Data entry Paperwork/Data entry Paperwork/Data entry

Resolution of grievances

How quickly were grievances addressed?

Timeframe for resolution of grievances/Dates

Was a followup screening required? When was physical completed? What types of medical issues are prevalent in the jail? When were issues identified? What types of medication are being distributed? When did each inmate first receive medication? How many doctor/nurse/hospital/lab visits were requested/required? How many outside doctor visits were required?

Data entry/Report

No. of inmates/Housing categories No. of inmates boarded out/Housing categories No. of inmates boarded by other agencies/Housing categories Inmate housing locations/Gang affiliations

How many grievances were sustained?

Health Information When was initial medical screening completed?

69

Report driven Report driven

Report driven

Medical questions designated by provider/Date/ Data-entry person Secondary questions asked Date of physical Categories of medical issues

Interview/ Observation

Dates of diagnoses Types of medication delivered

Data entry/Report Data entry/Report

Date/time of medication

Data entry/Report

Type of visit/Who visited

Data entry/Report

Number of transports to other providers/Dates/Times

Data entry/Report

Data entry/Report Data entry/Report Data entry/Report

70

Tim Brennan, Dave Wells and John Carr (Continued)

Questions to answer Transportation How many inmates were transported outside the facilities, by category? How many inmates were transported to court? Were there any security issues with transports? Who transported personnel? Court Information Who is going to court? How long is their court process taking? Who is eligible for different types of releases? Who was released from court?

Information to collect

Technique to use

No. of inmates transported/Transportation types/Dates No. of inmates transported/ Transportation types/Dates Event information

Report driven

Transport company information and Dates/Times

Report driven

Inmate names/Judge names/Court type Dates of admission/Dates of hearings/Dates of sentences Charge levels/Community ties information/Substance abuse information Inmate names/Release reasons/Dates

Interface with courts Report driven

Report driven Report driven

Report driven

Report driven

Level 3. Information on Day-to-Day Operations Decisionmaking Questions to answer Corrections Staffing How many staff are scheduled, and where? How many staff report to work? How many staff are on leave? What are the absence patterns of staff? What are the scheduling patterns? How many new staff were hired? How many staff retired or were terminated? What was the amount of overtime used, and why? What was the payroll amount and breakdown? Resources How was the budget spent? How many products of each category were purchased? How many products of each category were used?

Information to collect No. of staff scheduled to work/Where No. of staff who report to work Types of leave taken Staff leave/Dates Changes made in scheduling No. of staff hired/When/Job titles No. of staff retired/terminated/When/Job titles Hours of overtime used/Who/When/Why/Job titles Staff rates of pay/No. of hours

Dollar figure of items ordered/received No. of items purchased/Categories of items No. of items in warehouse or in stock

Technique to use Report driven Report driven Report driven Report driven Report driven Report driven Report driven Calculation Calculation

Calculation/ Report Calculation/ Purchase orders Inventory on hand

Running an Intelligent Jail Questions to answer How much was spent on products by each housing unit/facility/agency? Who were the vendors, and what types?

Information to collect No. of items bought/Dollars spent by each housing unit/facility/agency Names of vendors/Types

How much was spent per vendor?

Vendor purchase order totals

Maintenance What were the types of maintenance issues? What is outstanding? What were reasons for maintenance issues? How much was required for repairs? What were the areas requiring facility repairs? How many cells were out of service? How many locks were out of service? What is the equipment inventory? What is the purchase history of the equipment? Preventative maintenance schedules Fleet Maintenance What types of vehicles are in the fleet? What is their state of repair? How many miles were traveled per vehicle? Inspections What were the types and numbers of inspections completed? How many and what type of health Inspections were completed? How many violations were received? Where did violations occur? How long before violations were corrected? Quality Control and Compliance What data were incorrectly entered, when, and by whom? When were data corrected, how, and why? What reports were generated, and who requested them? How many reports were generated past the required timeframe, and by whom?

71 Technique to use Calculation

Purchase orders/Report Purchase orders/Report

Maintenance categories/Current status Causes of maintenance issues

Maintenance tracking Data entry/Report

Cost of materials Locations of maintenance issues

Data entry/Report Data entry/Report

Cell designations Door designations All current equipment over $XXX Dates of equipment purchase

Data entry/Report Data entry/Report Data entry/Report Data entry/Report

Service requirements/Dates performed

Data entry/Report

Makes/Models/Descriptions of vehicles Status of vehicles in use

Purchase orders

Mileage per vehicle

Maintenance requests Observation/ Report

Numbers/types of inspection

Report driven

Numbers/types of inspection

Report driven

Numbers/types of violations/Dates Locations of violations Dates violations were corrected

Report driven Report driven Report driven

Incorrect information entered and validated/Dates/ Data-entry staff Dates correct information entered and validated/ Data-entry staff Dates reports generated/Staff who requested reports Dates reports generated late/Types/Staff who corrected

Report driven Report driven Report driven Report driven

72

Tim Brennan, Dave Wells and John Carr (Continued)

Questions to answer How many reports had to be returned for correction, what types, and by whom? How many strip searches were done? How many headcounts were conducted?

Information to collect No. of reports returned/Types/Writer/ Supervisor

Technique to use Report driven

No. of strip searches No. of headcounts

Report driven Report driven

Level 4. Information on Public Access and Services Questions to answer Public Information Who is in jail?

How do I get them out on bond? What is the process for going to court? What happens while they are in jail? How do I visit? How do I provide funds?

Jail Information How many people are in jail and what types of jail? How many people are booked in? How many people are released? How many people go to state prison? Victim Information How are victims of inmates referred for help? Crime Stoppers How can crimes be reported?

Information to collect Name/Race/Sex/ Date of birth Inmate charges Bond amounts Content: Procedures for release and bond Content: Process for court procedures Content: Summary of what happens to person in custody Content: Visitation schedule and process Content: Money and mail procedures Online money deposits

Technique to use

When

Query Level 1 data

Real time

Query Level 1 data Query Level 1 data Query Level 1 data/ Content online Content online

Real time Real time Static data

Content online

Static data

Link to events online

Real time

Content online

Static data

Content online

Real time

No. of inmates in jail/Type of jail No. of people booked

Report on Level 1 data

Real time

Report on Level 1 data

Real time

No. of people released

Report on Level 1 data

Real time

No. of people to state prison

Report on Level 1 data

Real time

Content: Information on referrals for victims of inmates

Link to VINES online

Real time

Content: How to report information

Application

Real time

Static data

APPENDIX B. SAMPLE OF A DATA DASHBOARD

74

Tim Brennan, Dave Wells and John Carr

APPENDIX C. OVERCOMING THE HURDLES OF JAIL INFORMATION SYSTEMS Frequently, inadequacies of the existing jail information system or lack of a jail data system are catalysts, along with other factors such as changes in technology, for the development of a new or replacement jail system. When identifying the new or enhanced data collection methods, the following hurdles intrinsic to information systems—automated or manual—need to be addressed.

Limited and Missing Functionality One of the major user complaints about existing legacy jail systems9 is that such systems often fail to address one or more of the critical business functions of the jail. Earlier generation jail systems tended to focus solely on core processes such as inmates’ booking, release, and movements. To meet the comprehensive needs of the jail in today’s environment, the jail system may require unique major functions.

Misused and Unused Functionality Another problem with a jail system’s current functionality is when it is available but not used. The system may only address part of what is needed and therefore may require a combination of automated and manual processes to complete the job. In other instances, a lack of training, particularly when using the system requires a high degree of training or there is a high degree of turnover for a specific job, both of which contribute to misused and unused functionality.

Poor Data Quality The current jail system may fail to meet many of the agency’s information needs, particularly at the management level, because of the poor quality of data and inadequate access to it. If the system allows important data elements to be optional rather than mandatory, the usefulness of the captured data for decisionmaking, at both the individual and aggregate levels, may be significantly diminished. For example, on the individual level, if special handling requirements and alerts are not captured, information important to the corrections officers on any subsequent arrests may not be readily available. On the aggregate level, any analysis or assessment of the jail population will be limited and less useful if the volume of missing data is significant. In addition to missing data, there are data quality issues related to poorly coded data elements. If a data element lacks a list of values that address most possibilities, users may revert to code values such as “other” an inordinate amount of the time. Another aspect of poor data quality is the use of free-form text fields rather than coded values. This not only makes it difficult to aggregate and analyze data but it is also an inefficient way to capture data. For

Running an Intelligent Jail

75

example, a data element, such as the reason for release, can consist of a series of values that can be selected from a drop-down list of a free-form text field.

Poor User Interface Many systems still in use have a poor user interface. The user interface consists of the screen displays that the user sees when navigating through the system. Current systems are designed with a graphical user interface (GUI), which provides users with a much more intuitive, easy-to-use system. Because of the wide acceptance and exposure to Web-based Internet applications, the learning curve for new jail staff is shortened significantly with current GUI interfaces of jail systems, in particular, those systems that allow the ordering of screens to match the workflow of the end user.

Lack of Capabilities for Ad Hoc Queries and Reports A common complaint of both line staff and management is their inability to quickly retrieve the needed information from the jail database in a useful format. Although routine periodic reports, such as the court list or the daily booking list, can be predefined and generated in a useful format when required, other reporting requirements simply cannot be predicted in terms of frequency of use or specific data that are needed in detailed summary formats. These types of ad hoc queries and reports have traditionally been difficult to build into jail systems. However, current systems with relational databases and more extensive tools for data retrieval have made it feasible to offer ad hoc report capabilities, which provide a more flexible approach to the generation of reports.

Poor Integration of Data-Capture Technologies A common deficiency with existing jail systems is the failure to take advantage of current data-capture technologies. Scanning of wristbands for booking numbers (rather than keyboard entry), capture and presentation of inmates’ photos on inquiry screens, and the use of a single fingerprint at the time of release to match with the stored fingerprint database are examples of features now available in jail systems but missing from most legacy systems. Newer jail systems are built with an open architecture to support evolving data-capture technologies.

Limited Capabilities for Data Sharing and Data Exchange Another common deficiency of older jail systems is the limited ability to interface and exchange data with other systems that support law enforcement, prosecutors, courts, other justice agencies, and treatment providers. The lack of interfaces with external systems results in inefficiencies that result from redundant entry of data previously captured in another

76

Tim Brennan, Dave Wells and John Carr

system, less timely updates with manual data entry than in a system-to-system data exchange, and reduced data quality when the same data are entered multiple times in different systems. The goal should be the entry of data only once by the originating agency and a data-exchange process to share data of interest with other departments. Newer systems conform to the national justice data-exchange standards to facilitate the exchange of data with the jail’s business partners at the local, state, and federal levels. Bidirectional interfaces are those that include the transmission of data to and from external agencies. Newer jail systems are designed and implemented with interfaces on a near-realtime or scheduled basis (which is unavailable in older systems); this results in improved workflow processes within the justice community.

APPENDIX D. CASE EXAMPLE: CONTRA COSTA JAIL, MARTINEZ, CA Interview with Capt. Sean Fawell, Technical Services; David Pascoe, Field Operations Commander; Dave Spinelli, CAD/RMS/JMS Manager, Technical Services Division; and Sgt. Steve Borbely, Custody Services Bureau Please provide a brief narrative history of your MIS planning and acquisition process. In 1999, the Contra Costa jail went live with its previous JMS system. This system and the vendor support proved to be very costly. There also were significant Y2K problems that would be costly to fix. Even the smallest of change requests (e.g., functional and report outputs) were too costly on an ongoing basis. As a result, the jail began to consider replacement of this system with one that was more cost effective and with customer support that was more flexible and affordable. It was also noted that the skills of the correctional staff did not fit well with the user interface demands of the JMS software. The jail began the planning process, which resulted in the development and release of an RFP. Subsequent to the RFP release, five JMS vendors responded. Just as the RFP responses were coming in, a local software provider, which had previously provided iris recognition software to the department, approached the jail with an offer to design and build a new JMS. The downside, as stated by the jail, was that this vendor had no previous experience with JMS systems. The significant upside was that the vendor offered to build a system, including software, hardware, and database with no up-front cost, and that no payment would be due until the new system went live. It was further determined that the new system could be paid for by simply using the annual software maintenance and support monies required from their current vendor’s contract. Furthermore, the vendor agreed to place software developers inside the jail’s work units to help them understand the functional requirements of the system and to demonstrate proof of concept. Although this process had its merits, perhaps its biggest downside was that it fostered a significant creep in the project’s scope; that is, the functional requirements kept evolving as the process unfolded, even though general requirements were listed in the RFP document. The project unfolded over a 4-year timeframe from 2005 to 2008, when the system went live.

Running an Intelligent Jail

77

Note: The original vendor who offered to build the JMS system was motivated by the desire to enter the JMS market and leverage Department of Homeland Security monies that they believed would be available in the future. During the project, this company was bought by a much larger, international, identity-solutions company that ushered in new priorities and commitments and chose not to focus on continuing in the JMS market. However, the company did finish its commitment to Contra Costa.

Planning for Your New Jail MIS 1. What was your planning process? Although there was no formal goal or vision statement developed by the jail, the primary objective was to acquire a more efficient and cost-effective JMS system. The jail did not adopt a formal planning process for the design and implementation of a new JMS. The impetus for the change mainly came from middle management in the technology and custody services divisions. The internal project work teams were not established until after the decision was made to work with the identity-solutions vendor that offered to build a custom system. 2. How did you engage stakeholders (both internal and external to the agency) in the planning process, and who were they? The primary approach was to examine the various workflows in the jail, starting with booking and then moving from there out to other inmate processing functions. As the major functions were identified, work groups representing staff from each unit were identified to work with the software developers. Work groups included clerical, civilian, line, and supervisory staff. Middle management took the lead in the project and was held accountable for the project’s success. No formal, high-level steering committee was developed. Higherlevel administration was not significantly involved in the project nor did it contribute to identification of their JMS needs. Some effort was made to engage the courts in the design process but with little success, as there was no big-picture vision for the project and thus no perceived stake in it by external agencies. Thus, no other external stakeholders were involved in the project. 3. Did you have or engage the appropriate political support (e.g., sheriff, county commissioners, courts) both internal and external to the agency? There was very little political support for this project as it was primarily engineered and managed by mid-level management (e.g., captains from technical support, custody services, and central identification departments and their support staffs). 4. How did you identify your data/information needs? The systems data needs were principally identified by the software developers who worked directly with each unit (e.g., booking, medical, classification, property) of the jail. Data needs were driven by the work processes of each unit and by the reports each unit perceived they needed. 5. How did you identify your functional requirements? Technical and custody services management met with the internal work groups before the meetings with the vendor to discuss the limitations of the current JMS system and the desired improvements. Management did not want the work groups to simply re-engineer the functionality of the current system and just replicate current practices. They discussed more flexible

78

Tim Brennan, Dave Wells and John Carr software navigation and user interfaces, and identified old processes that could be dropped from the new system. The software vendor then met with the work groups to develop the functional requirements. They then reviewed the requirements with the technical and custody services management. PowerPoint slides of workflows and screen mock-ups were used to facilitate the process. 6. What happened during the planning phase that you did not expect? Working with little or no money allocated to the project and adapting to the lack of interest and participation from outside stakeholders were the biggest issues. They were “kind of winging it” through the planning process, so there were consequently not a lot of initial expectations. Upper administration did not engage in the process either positively or negatively. 7. What planning processes were most useful? On their own initiative, the vendor researched state legal requirements that pertain to the jail and its reporting, and assisted in bringing that information to the design discussions. It also was noted by the mid-management project team that “the facility is not the backbone of corrections operations; it is the JMS system.” 8. What were the pitfalls in the planning process? In addition to those previously mentioned, perhaps the biggest pitfall, ironically, was the fact that the JMS system was being developed and offered at no cost (paying for it after going live, with software maintenance and support monies already allocated to the previous vendor). The fact that there was no cost allowed the top administration and other stake-holders to be less engaged in the process. Because no new money was allocated, it became difficult to implement system change requests and to grow the system further after going live. 9. How did you deal with the obstacles? Issues were dealt with as they arose by the management team. Fortunately, the administration and other jail stakeholders outside the process did not dig in their heels on any issues. The funding obstacle was addressed by timing the new JMS to go live at the end of the old system’s maintenance contract year, so no new money had to be allocated. Members of the work groups and the mid-management project team changed over time as people moved to new positions. These changes meant that the experience and knowledge of those members left the project. 10. What would you have done differently? The project team identified several things that they would have done differently with the project: a. Appoint a high-level steering committee that involved the sheriff and other key stakeholders both inside and outside the agency. b. Engage a system integrator to pull together all of the related MIS vendors for information sharing. c. Involve the jail administration more in the process, and have them take some ownership of the project. d. Establish a dedicated JMS transition team, and keep core user groups intact through the completion of the project. e. Involve the county administration in the process to pull together external stakeholders (despite concerns of the risk of their getting too involved in the process).

Running an Intelligent Jail

79

f.

Educate the county police departments before the project to get their buy-in for the remote booking functionality built into the system; it had not been used by the arresting officers. g. Institute better controls over the change process so development could proceed more effectively. 11. What has been gained with the new system? The new system provides a more accurate inmate identification/confirmation process that uses iris-scanning for booking and release. The workflow processes and components can be modified to accommodate change better, especially in the area of booking. Most of the processes can be performed inhouse and without the cost or delay of going through the vendor. The JMS user interface is more intuitive and easier to learn than the previous system. The system provides an easier reporting process and our data are more accessible. Cost savings included no individual user license costs, no Oracle database maintenance, reduced system maintenance, and increased stability of the SQL server. 12. List examples of data outputs and their impact on the jail and stakeholders. The new Contra Costa JMS system offers both canned reports and an ad hoc report feature that uses Crystal Reports software. However, the ad hoc reporting feature is only used by the technical services staff when custom reports are requested. System users can change a limited number of report select-and-sort parameters (e.g., date ranges, facility, gender) on the ad hoc reporting screen. Rosters must be exported to Excel or Acrobat to get row and column totals. The system offers numerous routine reports for event scheduling activities (e.g., court actions, transportation, pending classifications, visitation, due date for release, temporary release, and inmates still in the booking station after X hours). The system produces various facility counts, including a current population head count, number of inmates booked, released, and so on. It offers various inmate notices, including fee agreements, property receipts, program status reports, required counseling notices, and participant rights letters. Other reports include work assignments, rosters, schedules, and arrests/ charges rosters. Accounting reports include financial transactions, account balances, check registers, and inmate cash authorizations. Some of the aggregate historical reports offered by the system include arrest by agency in the past month/year, bookings in the past month/year, classification scores, ICE/ Fed holds in the past month/year, Taser use, incidents of battery in the past week, and population demographics in the past year.

Exemplary Use of Technology in the Jail Environment: The Contra Costa County Sheriff The Contra Costa County Sheriff has been using a system referred to as ARIES (Automated Regional Information Exchange System) since 2003. The system shares jail data with all agencies in Contra Costa County. It represents the innovative use of information technology in several respects; more than 5,000 users and 61 agencies participate in the system. Although several features and functions could be cited, the electronic Probable Cause Declaration (PCD) system is described here. The electronic PCD system is a good example of

80

Tim Brennan, Dave Wells and John Carr

interagency data and document exchange as well as the use of a data dashboard in the jail. This module has been in operation since April 2007. The system enables law enforcement officers to create draft PCDs and submit them electronically and concurrently to both the duty judge and the jail. The jail is informed, on a close-to-real-time basis, of the current status of the PCD process for recent bookings. The judge reviews the draft PCD submitted electronically and can approve, deny, or return the PCD to the law enforcement agency for more information. Each PCD is logged into the system and its current status is maintained throughout the process. The judge can access the PCDs from home during off-hours by using an encrypted Internet connection. The documents can then be e-signed and transmitted electronically. The status of the PCD is available to the jail for relevant information on inmates at any time through a regularly refreshed dashboard. A ticking clock informs all three involved parties, including the originating law enforcement agency, the judge, and the jail, of the amount of time remaining for a PCD to be issued before the inmate is released from jail custody. If an approved PCD is not available within the statutory time limits (typically 48 hours), the jail intake staff will release the inmate at that time, based on the status displayed on the online dashboard. This implementation of technology meets a business need for several justice agencies from both a functional and a technical perspective. Documents and data are exchanged electronically in an efficient manner and with timely notification to the jail by using a data dashboard.

APPENDIX E. CASE EXAMPLE: KENT COUNTY JAIL, GRAND RAPIDS, MI Interview with Capt. Randy Demory Please provide a brief narrative history of your MIS planning and acquisition process. We started planning and analyzing our needs in March 2001. The RFP went out, and we selected our vendor by the end of the year. Development work started in 2002 and go-live implementation was in June 2004.

Planning for Your New Jail MIS 1. What was your planning process? We had an implementation team with representatives from several divisions from the jail, courts, and IT. We probably should have solicited more involvement from detectives and investigators because they are big users of the data. 2. How did you engage stakeholders (both internal and external to the agency) in the planning process, and who were they? The County had already formed a stakeholder group around the topic of criminal justice system computer integration, and they had already been meeting for a while by the time the sheriff’s department got the money to move forward with the jail MIS project. The jail MIS project was

Running an Intelligent Jail

3.

4.

5.

6.

7.

81

made a recurring agenda item for those regular meetings, and our issues were added at the subcommittee level as well. The committee had members from the circuit court, some district courts, community corrections, county administration, IT, the prosecutor’s office, juvenile detention, and law enforcement. Within the jail, we had a newsletter, updated monthly, that kept people informed and solicited input. The jail’s project manager also met frequently with the concerned parties, including the Office of the Sheriff. Did you have or engage the appropriate political support (e.g., sheriff, county commissioners, courts) both internal and external to the agency? Yes, we had their full support. How did you identify your data/information needs and functional requirements? Our jail had been automated, at least in part, since 1985, with older mainframe technology. We were able to build on that when we decided to move to a client server solution. We surveyed a number of other large jails to see what they had and were happy with, and what they wish they had. We went to conferences and looked at JMS display booths. We got our own users together and created detailed descriptions of the functionality they desired. I would mention that one problem we discovered with surveying other jails is that the person I would call would be the administrator because I am an administrator. The administrator frequently tended to put the best possible face on their JMS and would claim to be happy with it unless it was just a wreck. One of our lower level staff would talk to lower level staff in that same jail and get a totally different story. We found out that we had to drill down to the lower levels to discover how the JMS was really working. (You might want to keep that in mind as you read the responses to these questions.) What happened during the planning phase that you did not expect? I was not quite prepared for the amount of time that it took to accomplish, partly because we went into such great detail and the JMS developers were willing to make many of the changes we requested. Our planned interface between the JMS and the court records system was never successfully developed, and I did not expect that. The main thing that I did not expect, although it was not part of the planning phase, is that immediately after we went live, the parent company laid off the project manager that we had worked closely with for more than 3 years. Ownership of our JMS has changed hands twice since then. What planning processes were most useful? Actually, it was the very laborious process of documenting every single aspect of our previous system. We went through every single field, each field length, every code, every user profile, the details of each table in the old database, and more. We looked at data, tried to find where the bad data were, and what could be done in the next system to clean it up and prevent the same thing from happening with the new system. It all was very detail-oriented but gave us a very solid foundation to build on. What were the pitfalls in the planning process? The amount of staff time required in the whole planning and implementation process was phenomenal. It was difficult to get everyone involved who probably needed to be involved because we all had other jobs and responsibilities as well. Another pitfall was working through the unrealistic expectations of some of our staff people for the new JMS system. The vendor or the project team would meet with users and ask them what they would like

82

Tim Brennan, Dave Wells and John Carr

8.

9.

10. 11.

12.

to see in the new system. Many people essentially wanted a computer that would read their minds when they came into the room and do all of their work for them, and then turn itself off at the end of the day. In many instances, I, the project manager for the jail, had to come back in after a meeting with accounting, or medical, or court security and bring them back down to the real world. Part of this, too, was around the topic of “automated versus manual,” which I discuss in more detail in the concluding paragraphs. How did you deal with the obstacles? It all boils down to time. Many of the technical obstacles or difficulties could be resolved once everyone in the decisionmaking process was educated on the topic until they finally understood what the technicians were talking about. That takes time, but it is needed. Some people in positions of power or influence would attempt to block something or push something through without having a good understanding of what that meant. Once we were able to educate them so they actually understood the technology, the human obstacles went away. As it relates to technology or hardware, we were beaten by some obstacles simply because we did not have the money, in the end, to buy everything we would have liked. For example, we wanted to greatly expand our delivery system for jail reports (i.e., Crystal Reports and otherwise) and make them available at the desktop level in the prosecutor’s office, courts, community corrections, and so on. We also wanted some of the data dashboard features that were just coming out in 2003–2004. We identified a Crystal enterprise solution but, in the end, we had to throw in the towel after it turned out to be just too expensive. What would you have done differently? Given the scale of the project and the fact that it included software, hardware, a network, and about 12 interfaces, we probably should have given more serious consideration to hiring an outside project manager to guide us through the process. Did you upgrade/enhance your present system or purchase a new one? We purchased a new one. How are you using the system (line staff, middle management, administrators, and other stake-holders)? We have modules for booking, classification, visitation, basic medical, a housing unit floor log, case notes, charge tracking, visitation, work release, report writing (and hearings and administrative approval), and property. We have a little functionality for programs but not much. Line staff and middle-managers use the JMS every day for almost all functions in their jobs. We do not have much functionality specifically for the top administrators. Outside of the courts and community corrections, our main outside users are detectives, and their main access point is through a Web interface. What has been gained with the new system? One big thing is the advantages in the report writing. For example, with the new system, we went through and attached code tables to every conceivable field that we could, so the uniformity of data entry would help with our selection and query processes. The whole system approach was built with a view toward getting the data back out—when and how we wanted it. The second big thing we gained, paradoxically, was an expansion of free-text fields that allow us to put in as much narrative detail as we want. The case notes are an example of this, and the staff now love to enter case notes for all kinds of interesting details. In many cases, we will have both a code table and a free-text narrative field for the

Running an Intelligent Jail

83

same thing. For example, for classification overrides, the classification staff can select a code that describes the reason for the override and can add an expanded freetext narrative to provide more detail. We enter data in the same way for inmates’ tattoos, floor log entries, moves, and so on. This combination of codes only, free-text narratives only, and using code and/or free text has come to mean a lot for our data abilities.

Data Capacity and Data Use Analytical Capacity 1. How did you specify a report generation procedure for the vendor? We gave the JMS vendor a stack of reports from our previous JMS system that we wanted replaced and improved upon. We went over the vendor’s proposed reports and made improvements. They created some Crystal Reports for us. We trained a number of our people in Crystal Reports, and then the new JMS vendor provided us with some training in writing Crystal Reports, using our own database table structure. The tables are pretty complicated so, in some instances, our vendor created views to simplify the Crystal Report Generator. They also enabled us to send almost every report or query result to a file or to Excel, so that helps. In a few instances, we have used Microsoft Access as the tool to query our database, and that works very well for the staff people who know how to use it. The new JMS does not have its own report generator tool. 2. Who was involved in specifying performance requirements for report generation? We had a JMS implementation team made up of people with different responsibilities in the jail. 3. Did the report generator live up to your requirements? We are happy with using Crystal Reports and Excel, and we do fine with that, along with the vendor’s JMS canned reports. Now that we have used it for a while, I would like to see if we could get some more views of the database to simplify some things for us, but ownership of the parent company of our JMS has changed hands twice and the new owners are not as good to work with. 4. What staff in the jail have the competency to set up/build ad hoc reports? We started with three corrections staff—one in community corrections and two IT staff—but now, 6 years later, because of promotions, job changes, and layoffs, we are down to one in corrections, one in community corrections, and one in IT. Routine Monitoring Procedures 1. Can the MIS produce charts to monitor key outcomes (e.g., disciplinary rates per month) across time? The JMS does have limited ability to produce charts, but we export to Microsoft Excel for all of our charts and graphs. 2. Does the MIS offer procedures to monitor trends in the jail? The JMS does not have this ability beyond allowing us to select by date range, but we generally export to Excel and go from there. For the more detailed trend analysis, our community corrections guy uses SPSS.

84

Tim Brennan, Dave Wells and John Carr 3. Does the MIS offer procedures to monitor classification percentages (e.g., maximum, medium, and minimum security) across time? Again, only to the extent it allows us to select data by date range. We take daily snapshots and export to Excel instead of relying on the JMS to recreate a historical build of data. The JMS does store its own snapshots, but we would still want to get it into Excel to do any work.

Coding in the MIS 1. Who sets up your coding configuration for new factors to be monitored (e.g., sexual assaults, grievance coding)? The JMS administrator at the jail. 2. Who set up your original canned/out-of-the-box coding configurations? The JMS implementation team. Canned Reports 1. Who set up your initial set of canned reports? We had our JMS implementation team, made up of people with different responsibilities in the jail. 2. Are the canned reports meeting the information needs of key stakeholders in the jail? Generally, as supplemented by the ad hoc reports. a. Security monitoring? Yes. b. Drug use in the jail? Yes. c. Identification and coding of security incidents? Yes. d. Safety of inmates? Yes. e. Safety of staff? Yes. f. Disciplinary order (e.g., misconduct rates/types, grievances, staff use of force)? Yes. g. Services and treatments provided? No. h. Program activities of inmates? No. Information Needs of Specific Divisions and Managers Which departments in the jail monitor the following information categories? 1. Work demands across time. Jail administrators and senior administrators in the Office of the Sheriff. 2. Work done (daily, weekly, monthly). Sergeants, classification officers, medical and mental health staff, and video court staff. 3. Work quality (quality indexes). Intake sergeants monitor errors made in check-in and booking. The records supervisor monitors errors made in entering court paperwork. In these two cases, the supervisors are not typically using any computeraided search to conduct quality checks; they are simply responding to reported errors. The classification supervision does conduct a monthly audit that is guided by a number of reports that reflect quality, such as override rate, housing plan compliance, and AFIS verifications. 4. Work outcomes (meeting your goals, achieving selected outcomes). The jail sergeants and lieutenants use the JMS to make sure the deputies are meeting security outcomes like block/housing checks. We have a set of performance measures that we

Running an Intelligent Jail

85

report to the Office of the Sheriff quarterly, and the JMS assists with the collection of those data. The Jail Population Management committee uses JMS data to make sure we stay on track with our jail population projections.

Ad Hoc Policy Queries 1. Do senior administrators make ad hoc queries regarding management/policy problems? I am a senior administrator and I do, but I am the only one. Other administrators with the sheriff’s office occasionally request data from me, and outside entities such as county administration frequently do. 2. What happens when a senior administrator needs answers for an ad hoc public policy issue? I take care of it. 3. Does the jail have staff to routinely conduct ad hoc policy queries? We did until the county started going through budget cuts and we started laying people off. 4. How do you feel about the analytical and reporting capacity of the JMS/MIS system? Overall, I would give it about 80 percent. The table structure is very complicated, and that limits one’s ability to extract things sometimes unless one is very skilled with query writing. Data Quality Assurance and Error Detection Does your system have built-in data quality, omission, or error-detection features? We have the ability to set any field as a required field, so that takes care of some, but not all, of the omission problems. We do not have any true error-detection features beyond input masks to require date/time fields to be correct. For some things, we decided to allow staff to have some leeway or freedom when it comes to data entry, but it may not have been the best move. For example, with scars, marks, and tattoos, we have code table fields for location, type, color, and a couple of other things, and also a narrative notes field to describe the tattoo. None of the fields are required, so the booking people are free to fill all, part, or none of the fields that describe the tattoos. It seems like they do not want to be bothered by it, and if we make one of those fields required, they just skip the whole thing.

Other Jail MIS Issues One of the things that we probably made a wrong choice on, in the beginning, was to not link our address fields with dispatch’s database of true or correct addresses in the county. When dispatch types in an address, the database will alert them if it is not a “true” address. At the time, we thought it would complicate our lives because a certain number of people are from out of town, but in retrospect, I wish we would have done it, so at least we would have confidence in a certain number of addresses. That brings up another thought—that is, I have discovered that the single biggest users of our jail data (besides us) are the detectives in our agency and the surrounding agencies. There may have been some things we would have done slightly differently if I had known that, going into the project, to the degree that I know it now. We do not give the outlying agencies direct access to our JMS, but we did provide a Web interface that allows them access to nearly everything they want, and it is very popular with them. Another thing that I did not

86

Tim Brennan, Dave Wells and John Carr

think through, to the degree I wish I had, is this whole thing of “primary charge” or “driving charge.” If an inmate has five charges, how do you decide which is the charge that is the one principally holding an inmate? Or, for reporting purposes, people always want to know how many inmates are in jail for drugs or assault. We built in a logic model to calculate and flag the top charge for “in-custody inmates” based on severity, bond, and so on, and I am happy with that. The part I failed to comprehend was how frequently the primary charge changed upon release, and the last active charge at the final moment before release might be a minor civil charge instead of the serious violent charge that actually was the principal charge holding the inmate for several months. So, our JMS did not store what the primary charge was after release for the majority of the time the inmate was here in jail. The vendor wrote us an “expression” that we can include when we write Crystal Report queries, and it works pretty well, but it greatly increases processing time, as it has to recalculate the primary charges all over again. In a similar vein, I wish we would have had a canned report that would allow us to search by charge a bit more easily, particularly for inmates out of custody. There is a way to do it in our new JMS, but it is not as clean as I would like. We have a Crystal Report that I use, but it is not widely available to everyone. One area that was underdeveloped in our initial launch is a module for inmate programs. There is very minimal functionality in there now, and we always thought we would come back and enhance it, and we just never did. Now that we are doing much more inmate program work and getting very active in the whole inmate reentry initiative, we are missing it and having to turn to outside third-party solutions to meet this need. I would much rather have all of this functionality in our JMS. Another philosophical issue that informed our team’s development discussions back in 2003, but still pops up from time to time, is the discussion between automatic data entries versus manual entries. What do you want the computer to fill in automatically, and what do you want a human to fill in, even though the computer could do it if you told it to? For example, our housing unit floor log is called the Daily Journal. We want the housing unit officer to know what is going on in their housing unit and be accountable for everyone who is off the floor, so we thought it would be best if the housing unit officer was required to make a journal entry when an inmate left the floor for a visit, went to court, or something else. After a few years of doing that, we decided that it was tying officers to the computer too much and we would rather free them up to be with the inmates, so we opted for an automatic entry. Our one concession was to leave the ending date/time entry up to the officer, so now the computer automatically logs the event and the starting time, but the officer is required to enter the end time in the computer by hand when the inmate returns to the housing unit. That was our compromise. That same automatic versus manual discussion comes up from time to time in the context of many other modules, including classification.

APPENDIX F. EVALUATING YOUR JMS SYSTEM SUPPORT AND USABILITY FEATURES Introduction The following information is intended to assist in the evaluation of features in both the jail management information system (JMS) currently in use and in the JMS system(s) to be

Running an Intelligent Jail

87

considered when replacing the current one. The first involves evaluating the comprehensiveness, functionality, and usability of your agency’s current JMS components, and identifying areas for improvement to support the running of an intelligent jail. The second goal of this section is to provide a guide for assessing the comprehensiveness, functionality, and usability of new JMS systems to be designed or considered for purchase. The various functions and outputs of the JMS, listed in the evaluation instruments, also may be useful in developing system specifications for Requests for Proposals (RFPs) of new JMS systems. Note: The functions, outputs, and reports listed in the assessment guide are examples and are not intended to be all-inclusive. You may wish to add additional functions and outputs to the guide’s assessment list.

Using the Self-Assessment and Inventory Instruments In preparing to use the JMS Self-Assessment and Inventory Guide that follows, select a group of 6–10 system users and stakeholders who represent various units and organizational levels within the jail, including information technology staff. These staff may handle inmate processing in the following areas: • • • • • • • •

Booking/intake. Medical/mental health. Classification. Housing. Transportation. Scheduling. Work or program assignment. Release.

Staff to represent the organization may include the following: • • • • •

Line staff. Shift supervisor. Administration. Planners and staff responsible for budgets. Outside stakeholders (e.g., courts).

For each system, collect hard copies of input and output documents and screens, navigation screens, relevant manuals, data dictionaries, and code tables. Hold assessment sessions, using the System Assessment and Inventory Guide (see exhibit F.1). Each member of the assessment group scores the appropriate sections of the assessment instruments separately. The group discusses the reasons for any poor assessments and any differences in scores.

88

Tim Brennan, Dave Wells and John Carr

Instrument-Scoring Guidelines for JMS Features Using the assessment guide that follows (see exhibit F.1), this section describes how to score your assessment of the JMS system’s data support, functionality, timely access, data quality (integrity), and ease of the user interface. Exhibit F.1. Data Support Inputs, Functionality, Timeliness, and User Interface Score

Level

3

Comprehensive

2 1

Adequate Insufficient

3

Good

2

Fair

1

Poor

3 2

Good Fair

1

Poor

Definition Data Support Inputs To what extent do the data inputs designed in the system adequately support the information needs and processes of the jail? Are there data holes in the system (data gaps that the system does not capture? To what extent are data complete (provision in the system to capture the data but data are missing)? Data inputs are comprehensive and meet all or most expectations/requirements. Data inputs are adequate: they do not impair effectiveness of system. Data inputs are inadequate: they seriously impair effectiveness of the system. Functionality To what extent are the data inputted and stored in the system organized in an efficient and useful manner to support inmate management and agency decisions? Do routine automated system edits prevent missing data? Do the edits use available automated information, coded fields, logic matrices, etc., to prevent errors? System organizes case-processing and decision support data in an efficient, effective manner and readily displays data as needed by the user. System edits prevent missing data and prevent as many inaccurate entries as possible, based on coded fields and automated logic. Some data are organized in an efficient and effective manner, and some decision support data are readily displayed. This prevents most missing data and prevents some erroneous data entry. Keyed data are not efficiently organized and presented to the user in support of decisions. There are little or no checks for missing data or erroneous data entry. Timeliness How timely is the information provided for the immediate task at hand? Examples include data-entry clerks keying in offense information instantly, classification staff accessing criminal history, prior classification history, disciplinary history to complete the classification instrument, etc. Assess the ease of accessing information from prior bookings. In some systems, data from prior bookings are just a click or two away, but in other systems you may have to back all the way out of screens, look up historical book numbers, copy and paste them into additional screens, etc. Meets all or most expectations and requirements. Meets some requirements and does not seriously hamper the immediate task. Much of the data are not provided in a timely manner, which seriously impairs the efficiency of the system.

Running an Intelligent Jail Score

Level

3

Good

2

Fair

1

Poor

89

Definition User Interface Are the data screens easily understood and do they follow the workflow? Is the system easy to navigate and move between systems and screens? Do the input screens automatically fill in all available data? Does the system minimize or eliminate redundant steps and data entry? Is the screen layout logical? Is the flow intuitive? Are there shortcuts that the experienced user can use that speed tasks up? Can the user add codes to the code table or drop-down lists without calling tech support? Can the user set a field as mandatory or add a default value without asking for an enhancement? System is easily understood, screens are well organized and navigable, system is well integrated with the workflow, and data fields are automatically populated where appropriate. System is understandable and relatively easy to train to, follows workflow relatively well, some data fields are populated automatically, and system is relatively easy to navigate. System is not very understandable and does not adequately follow the workflow. System is not very easy to navigate, and few if any fields are automatically populated.

Tabulating the Assessment Scores for JMS Features Exhibit F.2 is a sample guide for a system assessment. This guide provides a format for the rating of the system reviewers’ findings; the assigned numerical values are calculated to arrive at a single score for the system being assessed. To use the guide, total the scores in each column and row, and enter the results at the end of each column and row. Count the total number of assessment items scored in the column. If all of the items in the example are scored, the total is 64. The number of items scored should be the same for each column. Enter that total on the line, “total items scored,” under each column. To compute the total average column score, divide the total of the column scores by the total number of items scored (the average column scores will be between 1 and 3). Row scores may also be totaled to provide insight into the adequacy of each system component; scores range from 5 (minimum) to 15 points (maximum). Total row scores of 10 to 15 indicate adequate to good system functions. To compute the overall system assessment score, sum the total column scores and enter that total in the ‘total column score’ field under the total row score. Enter the total number of items scored in the assessment on the ‘total items scored’ line (this should be the total items scored in the first column times 5; this will be 64 if all items in the example are scored). Divide the total of the column scores by the total number of items scored to get the overall JMS functions score (the overall score will be between 1 and 3). If all items in the assessment were not scored, this will need to be acknowledged during your final assessment of the system’s comprehensiveness. Instrument-Scoring Guidelines for Outputs and Reports Outputs and reports require considerations that are different from those of the overall systems and should be addressed separately using their own rating scale (see exhibit F.3). This exhibit shows a guide for output/report system assessment that is similar to the system assessment guide in exhibit F.1. The form provides a format for the reviewers’ findings and provides a single score for the output/ report system’s capabilities and comprehensiveness. The following section describes how to assess the JMS system’s reporting availability, user interface, comprehensiveness, timeliness, and data quality (integrity).

Intake: Positive Identification 1. Master System ID Search 2. Automated Fingerprint Match 3. Physical Characteristics 4. Automated NCIC/CCH Search Booking: Positive Identification 1. Inmate Demographics (*auto-populate with existing static data) 2. Background Data* 3. Identify Keep-Separates* 4. Identify Detainers/Warrants LEIN Initial Medical/Mental Screening 1. Initial Medical Screening 2. Initial MH & Suicide Risk Screening 3. Previous MH Brought Forward

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Exhibit F.2. System Assessment and Inventory Guide

4. Current MH Inventoried 5. Medical Status Alerts 6. Previous History Brought Forward 7. Previous Suicide Assessments and Attempts Brought Forward 8. Current History Inventoried 9. MH Status Alerts Booking 1. Inmate Demographics (*auto-populate with existing static data) 2. Background Data* 3. Offense Information 4. Identify Keep-Separates* 5. Identify Detainers/Warrants LEIN Initial Medical/Mental Health Screening Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

1. Initial Medical Screening 2. Initial MH & Suicide Risk Screening 3. Previous MH Brought Forward 4. Current MH Inventoried 5. Medical Status Alerts 6. Previous History Brought Forward 7. Previous Suicide Assessments and Attempts Brought Forward 8. Current History Inventoried 9. MH Status Alerts Time Computations 1. Sentence Dates, Sentence Lengths, Concurrent/ Consecutive, Statutory Minimum/Maximums 2. Time Credits at Intake 3. Automated Ongoing Time Credit Adjustments

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Exhibit F.2. (Continued)

Security Classification 1. Integrated Criminal History 2. Inventories Current Offense Seriousness, Disciplinary History, Escape History, Gang Status, Age, Substance Abuse, and Previous Incident History 3. Other Instrument Risk Factors 4. Automated Recommended Classification Assignment 5. Documents Override Assignment 6. Automated Inmate Notice Alerts and Events Tracking 1. Automated Alerts 2. Manual Alerts Posted (e.g., gang member, special diet, hold/detainer) 3. Mis-housed 4. Past Projected Release Date

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

5. Due for Court 6. Due for Classification/Review 7. Sentenced 8. Visitations, Sick Call, Pill Pass, etc. Program Needs Assessment 1. Previous Assessed Needs and Treatment History Brought Forward 2. PSI Information Brought Forward or Inventoried 3. Criminal History (e.g., domestic abuse, drug/ alcohol offenses) Inventoried 4. Current Needs Assessment Inventoried Transport Assignments 1. Automatically Matches Inmate to Appropriate Transportation List

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Exhibit F.2. (Continued)

2. Alerts for Transport Issues (e.g., Medical Holds, Keep-Separates) 3. Automatically Schedules Transport 4. Documents Transport Activity Housing Assignments 1. Automatically Inventories Appropriate Available Beds, Based on Housing Policies 2. Automatically Warns of Keep-Separate 3. Documents Housing Assignment and Any Housing Policy Override Program/Work Assignments 1. Matches Programs to Assessed Needs 2. Tracks Program Openings and Automatically Schedules Enrollment 3. Tracks Program/Work Assignment Schedules 4. Tracks Enrollments, Terminations, Termination Reasons

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

5. Automatically Tracks and Posts Earned Good Time Reclassification 1. New Mitigating/Aggravating Circumstances (e.g., Detainers, Disciplinaries, New Needs, Program/Work Outcomes Inventoried) 2. Automated Recommended Classification Assignment 3. Documents Override Assignment 4. Automated Inmate Notice Community Placement and Reentry 1. Early Release/Pretrial Release Eligibility Determination Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Exhibit F.2. (Continued)

2. Inventories Community Placement Risk Predictors (e.g., Criminal History, Escape, Probation/ Parole Violations, Work/Education Histories; Substance Abuse Severity, Social Stability, Criminal Cognitions, Age) 3. Risk and Needs Assessment 4. Reentry Plan Release 1. Release Date Confirmations 2. Positive Identification 3. Automated Victim Notification Registration Alert 4. Automated Offender Sex Registration Alert 5. Detainer/Holds Alert 6. Release Date and Reason Other

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

1. Flexible Notes and Comments Throughout System 2. Automated Checks for Data Quality/ Missing Data 3. Workload Driver TOTAL COLUMN SCORES

TOTAL ITEMS

AVERAGE COLUMN SCORE (Total column score divided by total items scored) To Compute Overall Score: 1. Add total column scores for inputs, functionality, timeliness, integrity and interface. 2. Then add total items together for each column. 3. Then divide the total column score by the total items to get the overall JMS features score.

Total Row Score

User Interface (Are the screens easily understood, uncluttered, easy to navigate, coded fields?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Data Integrity (Are data reliable and accurate?) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Timeliness (Information for process at hand provided to user in timely manner) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Functionality (Information efficiently organized, edits prevent missing/ erroneous data) 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

JMS-Related Features (Note: Function examples below are not intended to be all-inclusive. Additional functionality to be assessed may be added as desired.)

Data Support Inputs 3 = Comprehensive, 2 = Adequate, 1 = Insufficient

Exhibit F.2. (Continued)

Total of Column Scores Total Cumulative Items Overall Score

Running an Intelligent Jail

99

Exhibit F.3. Availability, User Interface, Comprehensiveness, Timeliness, and Data Integrity Score

2 1

Level

Definition Availability Are these outputs or reports available with your system?

Yes No

3

Good

2

Fair

1

Poor

3

Good

2

Fair

1

Poor

3

Good

User Interface Are the outputs or reports easy to generate? Can they be easily and quickly modified if needed? (e.g., changedata items, reorder the report, change or select parameters/date ranges, change output format from counts to frequencies) The output and/or report is easy to generate, flexible, and offers the user the ability to select or change parameters, reorder data, change formats, or create supporting ad hoc queries quickly. The output/report is moderately easy to generate and (if applicable) offers some flexibility in changing parameters, reordering data, supporting ad hoc queries, etc. The output/report is not easily generated and offers no flexibility in modifying the content or format. Comprehensiveness Do the data captured in the system provide adequate coverage of the information needed for the output or report content? Is some necessary information for the output or report missing or not available? Complete coverage and content availability of all necessary information to produce an informative output/report is available and meets most user expectations. Adequate coverage of most necessary information is available to produce an informative output/report and does not seriously degrade the comprehensiveness of the output/report. Adequate information/data in the system is not available to produce an informative report; lack of comprehensiveness severely limits the usefulness of the output/report. Timeliness How timely is the output/report provided to the user for the immediate task at hand, e.g., Are outputs/reports on inmates who are due for court transfer immediately available when staff are ready to schedule the event? Is the appropriate output/report available for inmates due for classification/reclassification? Are management and strategic-planning reports produced in a timely manner? The output/report is provided in a timely manner and meets all or most of the time requirements of users.

100

Tim Brennan, Dave Wells and John Carr Exhibit F.3. (Continued)

Score 2

Level Fair

1

Poor

3

Good

2

Fair

1

Poor

Definition The output/report timeliness meets some user requirements and does not seriously hamper the immediate task at hand. The output/report is not provided in a timely manner and seriously impairs the efficiency or need for the task at hand. Data Integrity Are the data in the output/report accurate and reliable? Are they often missing? Are the data continuously kept current? High level of confidence in the quality, accuracy, and reliability of the data needed for the output/report. Moderate level of confidence in the quality, accuracy, and reliability of the data needed for the output/report. Low or suspect level of confidence in the quality, accuracy, and reliability of the data needed for the output/report.

Tabulation of the System Assessment Score for Outputs and Reports Referring to exhibit F.4, for the first column, total the response scores (each rated 1 or 2) for the available items. Divide the total number of items (rows) assessed by the ‘total column score’ to determine the percentages of outputs and reports your system currently provides. For the remaining four self- assessment columns (user interface, comprehensiveness, timeliness, and data integrity), score only those items that were identified as available (having a score of 2). Add the total scores for each column and enter that score on the ‘total column score’ line under each column, including the ‘total row score’ column. Count the total number of assessment items scored in each column (74, if all items in the example are scored). The number of items scored should be the same for each column. Enter that total on the ‘total items scored’ line under each column. To compute the ‘total average column score’, divide the ‘total column score’ by the total number of items scored (the average column scores will be between 1 and 3). To compute the assessment score for the system’s outputs and reports, add each of the four ‘total column scores’ and enter the total on the ‘total column score’ line under the ‘total row score’. Enter the cumulative total of items scored in the assessment on the ‘total items scored’ line. (This should be the total items scored in each of the four columns, which is 74 if all items in the example are scored.) Divide the total of the four column scores by the total number of items scored to get the overall outputs/reports score (between 1 and 3). If all items in the assessment were not scored, this will need to be acknowledged in the final assessment of your system’s comprehensiveness.

Automated Internal/External MIS System Interfaces and Data Exchanges 1. State/Local Courts 2. Probation/Parole 3. NCIC 4. Triple I 5. State CCH 6. Local Law Enforcement 7. INS Case Processing/Inmate Tracking 1. Inmate labels (e.g., bar code IDs) 2. Schedules/rosters of inmates for classification review, reclassification, court appearance, etc. 3. Schedules of inmates to process for legislatively mandated initiatives, such as DNA, sex offender notification, victim notification 4. Work logs for staff to complete, such as classification forms

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Available 2 = Yes 1 = No

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

User Interface (Outputs/ reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Table F.4. System Assessment and Inventory Guide

5. Rosters of program vacancies and inmates awaiting program assignments 6. Preprinted fingerprint cards 7. Rosters of inmates available for movement 8. Housing alerts and vacancy reports 9. Mis-housed and keep-separate rosters/reports 10. Medical intake summary 11. Pre-parole reports on inmate’s institutional adjustment Operations Control 1. Classifications overdue for review 2. Inmates mis-housed today 3. Distribution of classification workload and overrides, by classification officer and facility 4. Relation of inmate program assessed needs as compared to program/work assignments 5. Use of program resources: program vacancies, program utilization 6. Percentage of population with zero fund balance

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Available 2 = Yes 1 = No

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

User Interface (Outputs/ reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Table F.4. (Continued)

7. Inmate Management Plan Performance/Progress Reports 8. Inmates in administrative segregation, by reason/days 9. How many inmates in disciplinary segregation, by days and infraction type 10. How many inmates in medical cell/unit, by days 11. Exception reports, including: a. Data accuracy b. Data omissions c. Inmate releases without victim notification, sex offender notification, etc. 12. Number of inmates assigned to special housing, by type and facility 13. Security profile of the jail today, by gender 14. Cells offline today, by reason and days offline Management Control 1. Use of staff resources: caseloads and unit productivity 2. Use of bed resources: patterns of bed vacancies, bed shortfalls, bed misuse, beds offline

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Available 2 = Yes 1 = No

User Interface (Outputs/ reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

3. Use of transportation resources: transports, by type/week/month/year; patterns of empty seats 4. Checked property lost in past quarter/past year 5. Inmate fund deposits, debits, fund balance for past quarter/past year 6. Total inmate processing fees collected in past quarter/past year 7. Disciplinary infractions rates per 100 inmates in past month/year, by classification level 8. Rates of inmate-on-inmate assaults in past month/past year, by classification level 9. Rates of inmate-on-staff assaults in past month/past year, by classification level 10. Total trust account revenues in past quarter/past year 11. Rates of sexual assaults on inmates in past month/past year, by classification level 12. Aggregate medical/MH needs of population in past quarter/past year, by type

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Available 2 = Yes 1 = No

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

User Interface (Outputs/ reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Table F.4. (Continued)

13. Aggregate assessed program needs of population, by type 14. Utilization of programs by assessed need/termination reasons, and month/year 15. Work assignments versus program capacity, by month/year 16. Facility capacity report: Actual capacity versus design capacity and lawful capacity 17. Security profile of population, by month/year/gender 18. Aggregate release reasons, by month/year 19. Number of inmate and staff grievances, by month/year/reason 20. Number of released inmates who are homeless, by month/year 21. Cells offline, by month/year/reason/days offline 22. Aggregate number detained past authorized release date, by days past due/reason/month/year

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Available 2 = Yes 1 = No

User Interface (Outputs/ reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

23. Percentage of population for delayed release due to forfeiture of good-time status, by reason/days delayed/month/year 24. Projections of resource needs: trends in stock and flow populations, by resource variables, e.g.: a. Trends in numbers and types of special needs inmates b. Trends in distribution of security classification types c. Trends in education levels 25. Frequency of staff shortages, by reason/month/year 26. Security profile of the jail, by week/month/year 27. Inmates boarded for a fee in past month/past year, by type; total fees generated 28. Recidivism rates of inmates booked in past month/past year 29. INS/ICE holds not boarded for a fee, by month/year

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Available 2 = Yes 1 = No

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

User Interface (Outputs/reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Table F.4. (Continued)

Strategic Planning 1.Total bookings and releases by month/year 2.Average length of stay, by year 3.Average daily population, by month/year/ security level 4.ADP housed less than 72 hours 5.ADP greater than 72 hours 6.Rates of mis-housings, by security levels/ average days mis-housed/year 7.Recidivism rates of inmate population, by assessed needs if available 8.What are the inmate target populations for early release consideration? 9.How many times did population exceed functional capacity or court-mandated capacity? 10.Internal Policy Simulations (e.g., changes in classification variables) 11.Population Forecasting 12.Staffing Analysis/Projections 13.Legislative Impact Analysis (e.g., mandatory sentences, determinate sentencing)

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Available 2 = Yes 1 = No

User Interface (Outputs/reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

TOTAL COLUMN SCORES

TOTAL ITEMS SCORED

AVERAGE COLUMN SCORE (Total column score divided by total items scored) To Compute Overall Score: 1. Add total column scores for interface, functionality, timeliness, integrity and comprehensiveness. 2. Then add total items together for each column (excluding the first column, “Available”). 3. Then divide the total column score by the total items to get the overall outputs/reports score.

Total Row Score

Comprehensiveness (Information needs coverage) 3 = Good, 2 = Fair 1 = Poor

Data Integrity (Are data reliable, accurate?) 3 = Good, 2 = Fair 1 = Poor

Available 2 = Yes 1 = No

Timeliness (Outputs/reports provided to user in timely manner) 3 = Good, 2 = Fair, 1 = Poor

Outputs and Reports (Note: Output/report examples below are not intended to be all-inclusive. Additional outputs and reports may be added as desired.)

User Interface (Outputs/ reports easy to generate, easy to modify/ create) 3 = Good, 2 = Fair, 1 = Poor

Table F.4. (Continued)

Total of Column Scores Total Cumulative Items Overall Score

Running an Intelligent Jail

109

Row scores also may be totaled to provide insight into the adequacy of each system output or report relative to the interface, timeliness, data integrity, and comprehensiveness. Component scores will range from 12 (maximum) to 4 points (minimum). Total row scores between 9 and 12 indicate an adequate to good output or report.

APPENDIX G. “MEASURING WHAT MATTERS” KENT COUNTY CORRECTIONAL FACILITY ANNUAL STATISTICAL REPORT Access the report at http://nicic.gov/Library/027347.

REFERENCES Bennett, Dave & Donna, Lattin. (2009). Jail Capacity Planning Guide: A Systems Approach. Washington, DC: U.S. Department of Justice, National Institute of Corrections. Brennan Tim. (1999). “Implementing Organizational Change in Criminal Justice: Some Lessons From Jail Classification Systems.” Corrections Management Quarterly, 3(2), 11-27. Brennan, Tim & James, Austin. (1997). “Classification of Female Inmates: A Review of Key Issues For Jail Managers.” Aurora, CO: U.S. Department of Justice, National Institute of Corrections Information Center. Brennan, Tim, Breitenbach, M. & Dieterich, W. (2010). “Unraveling Women’s Pathways to Serious Crime.” Perspectives: Journal of American Probation and Parole Association. Brennan, Tim & Dieterich, W. (2007). “New York State Division of Parole: COMPAS Reentry Pilot Study.” Psychometric Report. Northpointe Technical Document. Brennan, Tim & Dave, Wells. (March 1991). Policy Making in Criminal Justice: The Use of Hard Data at Each Stage of the Policy Process. Traverse City, MI: Northpointe Institute. Brennan, Tim, Dave, Wells & Alexander, J. (2004). Enhancing Prison Classification Systems: The Emerging Role of Management Information Systems. Washington, DC: U.S. Department of Justice, National Institute of Corrections. Elias, Gail. (2007). How to Collect and Analyze Data: A Manual for Sheriffs and Jail Administrators, 3d ed. Aurora, CO: U.S. Department of Justice, National Institute of Corrections. Garb, Howard N. (1998). Studying the Clinician: Judgment Research and Psychological Assessment. Washington, DC: American Psychological Association. Hall, Andy. (1985). Alleviating Jail Overcrowding: A Systems Perspective. Monograph. Washington, DC: U.S. Department of Justice, National Institute of Justice. Janis, Irving & Lon, Mann. (1977). Decision Making. New York: The Free Press. MTC Institute. (2006). “Measuring Success: Improving the Effectiveness of Correctional Facilities.” Centerville, UT: Management & Training Corporation. Retrieved May 30, 2006, from http://www.mtctrains.com/institute/publications/MeasuringSuccess.pdf. MTC Institute. (2007). “Contracting Prison Opportunities: A Plan to Improve Performance.” Centerville, UT: Management & Training Corporation. Retrieved April 22, 2013, from http://www.mtctrains.com/public/uploads/1/2010/10/ContractingForSuccessReport.pdf.

110

Tim Brennan, Dave Wells and John Carr

Quinsey, V. L., Harris, G. T., Rice, M. E. & Cormier, C. A. (1998). Violent Offenders: Appraising and Managing Risk. Washington, DC: American Psychological Association. Reason, James T. (1990). Human Error. New York: Cambridge University Press. Schoech, Dick. (1982). Computer Use in Human Services: A Guide to Information Management. New York: Human Sciences Press. Walton, Richard E. (1989). Up and Running: Integrating Information Technology and the Organization. Boston: Harvard Business School Press. Wells, Dave & Brennan, Tim. (1995). “Jail Classification: Improving the Link to Intermediate Sanctions.” Corrections Today (February).

End Notes 1

See Jail Capacity Planning Guide: A Systems Approach, by D. Bennett and D. Latin (Washington, DC: U.S. Department of Justice National Institute of Corrections, November 2009). 2 Walton, Richard E. 1989. Up and Running: Integrating Information Technology and the Organization (Boston: Harvard Business School Press). 3 Schoech, Dick. 1982. Computer Use in Human Services: A Guide to Information Management (New York: Human Sciences Press). 4 Elias, Gail. 2007. How to Collect and Analyze Data: A Manual for Sheriffs and Jail Administrators, 3d ed. (Aurora, CO: U.S. Department of Justice, National Institute of Corrections). 5 Adapted from Up and Running: Integrating Technology and the Organization, by Robert E. Walton (Boston: Harvard Business School Press,1989). 6 Brennan, Tim. 1999. “Implementing Organizational Change in Criminal Justice: Some Lessons From Jail Classification Systems.” Corrections Management Quarterly 3(2):11–27. 7 Reason, James T. 1990. Human Error (New York: Cambridge University Press). 8 Informating is the process that translates descriptions and measurements of activities, events, and objects into information. Through this process, these activities become visible to the organization. 9 In information technology, a legacy system is one that is no longer supported, cannot be changed or updated, and usually is not Web-based.

In: Information Systems Approach to Jail Management ISBN: 978-1-63321-255-8 Editor: Cheryl L. Cooper © 2014 Nova Science Publishers, Inc.

Chapter 2

JAIL CAPACITY PLANNING GUIDE: A SYSTEMS APPROACH ∗

David M. Bennett and Donna Lattin FOREWORD During the past 30 years, jails nationwide have become crowded in response to policy shifts in the criminal justice system, including the clampdown on driving under the influence, the adoption of mandatory arrests for domestic violence, and the “get tough” approach to many drug crimes. Crowding can create serious management problems, compromising the safety of both inmates and staff. Therefore, it is essential that jurisdictions adopt comprehensive, effective strategies to address the problem of crowding in our nation’s jails. Jails are part of a complex criminal justice system whose policies and practices directly influence total bed need. As such, jail planning cannot be done in a vacuum. Any consideration of future jail bed need must take place within the context of a discussion about how to manage the larger criminal justice system more effectively. Jail planning and system planning are one and the same. Emerging information technology provides us with unprecedented potential for analyzing the dynamics of the complex criminal justice system and forecasting and managing jail capacity needs. This guide describes key population management strategies that have as their foundation the necessity of holding offenders accountable while making judicious use of detention resources. This guide also makes the case for the importance of identifying offenders who pose higher risks and targeting them for the most intensive correctional resources, making available a full continuum of alternatives to jail, relying on evidence-based sanctions and quality treatments, and building in transition and stepdown options from jails.

∗

This is an edited, reformatted and augmented version of a document, NIC Accession Number 022722, issued by the National Institute of Corrections, dated November 2009.

112

David M. Bennett and Donna Lattin

For better or for worse, all local systems will change. The question is not whether, but how, policies will change. We hope that this document will assist jurisdictions as they implement program strategies designed to plan for, respond to, and manage change, while making the most efficient use of existing resources. Morris L. Thigpen Director National Institute of Corrections

INTRODUCTION Available beds in any correctional facility have a tendency to become filled—no matter what the size of the facility. Jail crowding is a symptom of the policies and practices of the larger criminal justice system. The changes that have occurred in the nation’s jail population during the past 30 years provide evidence that policy shifts alone can bring about dramatic changes in the demand for jail beds. Furthermore, reductions in crime do not necessarily translate into reduced demand. There is no correlation between crime rates and incarceration rates. Instead, unstated, and often unexamined, policies across adjudication decision points in the criminal justice system largely drive jail bed usage. The number of incarcerated individuals has increased significantly over the past three decades. Yet planners in the 1970s would not have been able to foresee the nationwide policy shifts that would fuel this growth: the clampdown on drunk driving; the adoption of mandatory arrests for domestic violence; the crack cocaine problem and, some years later, the methamphetamine epidemic; and the “get tough” approach to most drug crimes. Nor could they have anticipated the expansion of mandatory minimum sentencing policies, the adoption of three-strikes laws, the restrictions on judicial discretion, or the combined effect of the deinstitutionalization of persons with mental illness and the lack of community-based resources to serve them. Jail planning must be shaped by an understanding of the interactive effects of criminal justice system policies, its practices, and the availability of alternative programs. Assessing the efficiency of the criminal justice system first and then taking steps to optimize resources can postpone the overflow of a new facility for many years, typically saving a county hundreds of thousands of dollars at the very least. Today, construction costs for new jail facilities can be upwards of $100,000 per bed. Building costs, however, account for only a fraction of total expenditures. On average, construction costs for a new county jail represent only 10 percent of overall operating costs over a 30-year period. A plan addressing the intricate nature of the criminal justice system (a system master plan) can help jurisdictions manage limited jail resources and adapt to changing circumstances for years to come. The system master plan is a comprehensive strategy for addressing the many factors that drive jail demand.

Jail Capacity Planning Guide: A Systems Approach

113

Today’s Jail Jails play an essential role in the criminal justice system. Unlike prisons, which serve only a sentenced population, jails accommodate a broader category of individuals with shorter lengths of stay—usually up to 1 year, but in some jurisdictions up to 2 years. Jails are short-term correctional facilities operated primarily by counties, but also by some cities and states. They detain individuals awaiting trial and incarcerated offenders who have been sentenced to jail, are awaiting transfer to prison, or are serving time on a probation or parole violation. In some localities, jails also house inmates held under state or federal jurisdiction. They may also temporarily detain juveniles, persons waiting for a mental competency examination or transfer to a psychiatric facility, and defendants awaiting transfer to another county. Some county jails also accommodate an overflow of state inmates; others hold, through contract, federal and Immigration and Customs Enforcement prisoners. More recently, jails are being used to hold juveniles on remand status. In localities without adequate community-based resources, jails are still used to house individuals being held for non-criminal reasons such as detoxification or mental health stabilization. In the end, the jail population is a function of two factors: the number of admissions and the average length of stay, both of which greatly affect jail crowding (see exhibit I–1).

The Upward Trend During the 1990s alone, jail and prison populations almost doubled. Today, U.S. jails hold more than 700,000 inmates. The growth in the nation’s jail population has been part of an overall surge in the correctional population (prison, jail, and probation) over the past 20 years. Skyrocketing expenditures for corrections have accompanied the steady growth in the demand for more jails as cities and states rushed to construct more bedspace. Surprisingly, at the same time that the number of incarcerated individuals is at an unprecedented level, the reported crime rate is at a 40-year low. Although crime rates are in decline almost everywhere, the annual growth rate in the jail population continues its steady upward trend. The Sentencing Project reports that: Between 1991 and 1998, those states that increased incarceration at rates less than the national average experienced a larger decline in crime rates than those states that increased incarceration at rates higher than the national average. . . . Since 1998, 12 states experienced stable or declining incarceration rates, yet the average decrease in crime rates in these states was the same as in the 38 states in which rates of imprisonment increased.1

That there is no correlation between crime rates and incarceration rates can be seen in the experiences of three different states: California, New York, and Texas (see exhibit I–2).

114

David M. Bennett and Donna Lattin Exhibit I–1. Factors That Influence Jail Population

Number of Admissions to Jail County population Number of law enforcement officers Booking and cite-and-release policies County booking fee policy Availability of prebooking alternatives (detoxification and crisis centers) Access to comprehensive pretrial services Failure-to-appear rate and warrant policy Pretrial failure-to-appear investigation and court return procedures Pretrial supervision, monitoring, and tracking Violation of supervision rate and policy Juveniles certified as adults to stand trial State policy transferring inmates to other counties Contracts with other agencies Courtesy holds for other agencies Availability of alternative sanction and diversion options Quality of system intervention Politics and the media

Average Length of Stay Access to timely pretrial assessment Early appointment of counsel Pretrial release options Bonding policy Pretrial bond review procedures Early case resolution procedures Charge and plea-bargaining policies Local case-processing times Diversion and deferred sentence options Availability of jail alternatives (treatment, work release, etc.) Eligibility criteria for jail alternatives Sentencing mandates Sentence length Stepdown options from jail to alternative facilities/programs Prevailing philosophy regarding punishment versus treatment

Exhibit I–2. Comparative Change in Incarceration and Crime Rates: 1991–2001 State California New York Texas

Incarceration Rate (%) 42 11 139

Crime Rate (%) –42 –53 –34

Several factors have been fueling the increase in the jail population, but the fluctuation in reported crime has been a weak contributor. On the other hand, shifts in sentencing policies have had a profound effect on custody resources. Sentencing policies for drug offenses, for example, represent one of the more significant factors contributing to higher incarceration numbers over the past several decades. The number of drug offenders in prison and jail increased from 40,000 in 1980 to more than 450,000 today. The effect on jails is dramatic: The proportion of jail inmates constituted by individuals with a drug charge or conviction increased from 1 in 10 in 1983 to 1 in 4 in 2002.2 Within this upward trend, growth rates differ by groups. The adult female jail population has increased by an average of 7 percent annually over the past 10 years, whereas the adult male jail population has grown by a slower 4.2 percent.3

Jail Capacity Planning Guide: A Systems Approach

115

Finally, one cannot speak of the upward trend in the growth of the jail population without speaking of the downward trend in institutional housing for persons with mental illness. The United States has experienced an astounding 95-percent drop in the rate of admission to state mental hospitals over the past 50 years. In 1955, psychiatric hospitals housed 558,239 patients with severe mental illness; by 1994, this number had been reduced to 71,619.4 Administrators surveyed about the increasing demand on jails cited as primary factors the large number of bookings for drug offenses and violent crime, longer jail sentences, an increase in probation violators, and the increasing fallout from crowded prisons.5

Consequences of Jail Crowding Crowding can create serious management problems and can compromise the safety of inmates and staff as the jail environment becomes increasingly volatile. Crowding can be measured in the lack of flexibility that comes with court-ordered limits (caps) on the jail population. The dynamics of a jail, with unpredictable inputs and daily fluctuations in population, require management flexibility in the form of a few empty beds. Because of this, a jail is at capacity before reaching its design limits. Beds have to be set aside for classification (a male prisoner cannot be housed in a female bed, nor can a maximum-security prisoner be housed in a minimum-security bed), and sufficient beds need to be set aside to handle the population during peak periods. A crowded jail can result in the loss of system integrity. This occurs when inmates are turned loose from the jail through “forced releases.” It does not take long for this to become common knowledge. In some jurisdictions, defendants routinely ask jail staff at the time of booking how soon it will be before they are “forced released” back to the streets. Forced releases are not an acceptable method for controlling jail crowding; they are evidence of the failure of the criminal justice system. This method of release not only requires the jail manager to assume something of a judicial role in shortening the time served on court-ordered sentences but it also does nothing to protect the community or interrupt the costly cycle of failures yet to appear. Moreover, in jurisdictions where forced releases have become the norm for managing the jail population, the failure-to-appear rate has increased exponentially. In fact, national data indicate that defendants released from jail on forced release are more than twice as likely as those released with pretrial conditions and supervision to have a bench warrant issued because of a failure to appear in court.6 Lane County, OR, confirmed these findings in an examination of its own failure-toappear rates. In 2002, it found that 22 percent of circuit court defendants who were forced released from the Lane County Jail failed to report to court. This contrasted with a 10-percent failure-to-appear rate for defendants exiting the jail through the custody referee (pretrial release office) with a pretrial release agreement.7 Sometimes criminal justice systems respond to crowding by making abrupt changes in policy or practice. This can be a formal decision, as when a district attorney stops prosecuting all nonviolent misdemeanors because of a lack of resources, or, as in the case of Los Angeles County, when crowding leads to a decision to limit jail bookings to violent misdemeanors and felony defendants for whom a high bond has been set.8 The system also informally modifies its behavior in response to a lack of jail resources. In one jurisdiction, law enforcement officers shifted to a “cite and release” policy for most drunk drivers in response to lack of jail

116

David M. Bennett and Donna Lattin

space; in another jurisdiction, judges modified sentence length downward. The integrity of the criminal justice system is also compromised when, because of inadequate jail space, there is no guarantee that the sentence rendered will be the sentence served. At a national level, the loss of integrity of the criminal justice system can be seen in the shrinking proportion of the jail population made up of sentenced inmates and the corresponding increase in the proportion of pretrial inmates. Since 1990, the relative percentage of pretrial offenders in jails has increased from 51 percent to 60 percent.9 Common inefficiencies in adjudication that extend the time needed to resolve the case contribute to this number. In some jurisdictions, the number of beds available for the sentenced population is so low that this group makes up less than 10 per cent of the total jail population—that is, the sentenced population is literally being squeezed out. The end result is a system left to go through the motions of dispensing justice without the means to impose it in the manner a judge has ordered. To address the problem of crowding, some counties have resorted to boarding inmates in other counties. Counties in Michigan did this when, over a 2-year period, at least 10 counties declared jail crowding emergencies. Even then, crowding continued, despite a 17-percent increase in county jail capacity and a 20-percent drop in arrests over a 6-year period.10 In some cases, crowding can even lead to system fragmentation. For example, some municipalities break away from county facilities to seek funding for their own jails. This is an imperfect solution because it can increase system redundancies and costs and exaggerate case-processing disparities. There is another cost to crowding: the cost to the victims. For every inmate released from jail early, a victim is affected. For every inmate released prematurely who is then rearrested for another crime, additional victims are created.

Getting Out Ahead of the Problem Planning often begins with a crisis. The best planning, however, starts long before the jail is overflowing. After Hurricane Katrina, a congressional committee was formed to examine the problems that led up to the levee break. The comprehensive analysis that followed was impressive, resulting in 800,000 pages of documentation and testimony from 250 experts and witnesses. Many questions were addressed: How could we have anticipated this? How should information have been shared? Who was in charge? How do we better predict and plan for future events? These same questions are relevant to projecting and planning future jail capacity. The goal is to get out ahead of the problem, not to wait until the existing jail is so dangerously crowded or antiquated that local officials either lose control (through court imposition of population caps) or are left with no choice but a rushed and poorly planned response. Jurisdictions must make decisions determining capacity before crowding occurs. If a jail is planned to accommodate the addition of bunks and the infrastructure is in place to manage the additional population (dayrooms are sized appropriately, enough support areas are available in the units, etc.), the addition of bunks may not result in jail crowding. However, if the addition of bunks results in a violation of the classification system or the housing plan, then steps need to be taken to bring the jail into compliance through the

Jail Capacity Planning Guide: A Systems Approach

117

implementation of well-planned alternatives as the planning process for new beds is completed. The same problem applies to budget-based crises. Once the cutting back of resources begins, the opportunity to draft and enact systemwide changes in jail management without drastically disrupting the running of the broader system is curtailed. At that point, stopgap measures are often a county’s only realistic option. This leads to piecemeal improvement of the criminal justice system that does not further long-term planning objectives. It is always preferable for a county to undertake a needs assessment of its criminal justice system and begin implementing its desired systemwide changes before a crisis develops. Comprehensive planning is meant to design strategies that are balanced and that anticipate future needs.

A New Paradigm For better or worse, all local criminal justice systems will change. The question is not whether but how policies will change. Jail capacity planning will always be an imperfect art—one that must include a comprehensive and continuous analysis of key variables of the criminal justice system. The methodology presented in this guide challenges jurisdictions to take a systems approach to planning based on available data. However, this guide also presents another challenge: to set a goal not just to meet demand but also to make a difference. It argues for a results-based paradigm for planning. The jail will always be a scarce and expensive resource. The best knowledge should inform the policies that dictate how it is used. A body of research now exists that can provide a framework for taking a fresh look at how jail beds are used. The research makes the case for the importance of 1) identifying the higher risk offender for the most intensive corrections resources, 2) making available a full continuum of alternatives to jail, 3) relying on evidencebased sanctions and quality treatment, 4) building in transition planning and stepdown options from jail, and 5) adopting a positive emphasis on change. This research challenges the old notion that a jail bed is the only place, or the best place, to meet the goal of punishment and community protection. Instead, it makes the case for making treatment the norm, backed by the certainty of a short jail sanction, and for allocating both jail time and treatment intensity based on an offender’s risk to the community. What has been learned strengthens the argument for the judicious use of the detention resource. It argues for a new paradigm that views jail as the alternative. The drug court movement reflects this research; it has introduced a new conceptual framework that reasserts the primacy of treatment and redefines the system’s response to failure. The community corrections center is another example of a new way of thinking about the central mission of the criminal justice system, one that makes reducing future crime a central goal. The challenge is to make effective use of limited corrections resources by instituting jail population management strategies that hold offenders accountable while taking a constructive approach that promotes public safety.

118

David M. Bennett and Donna Lattin

SECTION I. GETTING STARTED Jail Planning Forecasting is the science and art of predicting the future. The science involves the use of objective and tested methods to track trends. The art involves the ability to imagine courses of action that might alter those trends. Jail forecasting relies on tracking a modest set of variables. These include county population, incarceration rates, crime, adult misdemeanor and felony arrests, jail admissions (ADM), average length of stay (ALOS), and average daily population (ADP) of important subsets of the jail population. By mapping changes in these variables over time, a picture emerges that allows counties to make assumptions about future jail capacity need if criminal justice policy remains unchanged. This requires systematic planning, and the best jail planning, like the best planning for a flood, occurs well in advance of an emergency. The first order of business in jail planning is acknowledging the limitations of the task. Anticipating future demand is a difficult endeavor when attempting to plan several years in advance, let alone 10 years or more. Jail forecasting models, like all such models, are only as good as the data that go into them. In the end, no method of forecasting can predict the future perfectly. Each local criminal justice system has its own complex and dynamic characteristics that influence jail capacity planning. Outside factors that cannot be controlled or predicted will affect future demand. Policies change, new laws are passed, financial resources wax and wane. Furthermore, the capacity-driven nature of most jails makes jail capacity planning difficult. In most cases, available jail space tends to fill. Surveys of jails reveal that facilities that had been expected to be adequate for 10 or 15 years were filled in half that time or less.

Most Jails Operate at Capacity Because available jail beds tend to fill quickly and most jails operate at capacity, jail planning is challenging. Unlike most businesses, which experience an ebb and flow of demand, or some (like hotels) that actually have vacancies, jails tend to have a full house. This presents a unique challenge for jail capacity planning. Pent-up demand in jurisdictions with crowded jails can result in changes in criminal justice system practices (e.g., police no longer booking certain offenders, prosecutors no longer filing particular offenses) that challenge the task of gauging actual demand. Incarceration Rates Continue to Rise Not only do most jails operate at capacity, but incarceration rates have steadily increased, as have the total number of adults under supervision in both custodial settings and noncustodial set tings such as probation. The increasing reliance on incarceration to respond to criminal activity can lead to a spiral of ever-increasing costs and facility demands unless counties can better manage criminal justice system conditions. Exhibit 1–1 shows increases in incarceration rates. The graph compares U.S. incarceration rates with regional incarceration rates for the Northeast, Midwest, South, and West for the years 1978, 1983, 1988, 1993, 1999, and 2005. The data are from the Bureau of Justice Statistics’ Census of Jails.11 Unfortunately, the Bureau collects national data only every 5 or 6 years.

Jail Capacity Planning Guide: A Systems Approach

119

Source: Bureau of Justice Statistics, Census of Jails (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, various years). Note: As this document went to press, the Bureau of Justice Statistics released updated jail statistics. For 2008, the reported national incarceration rate was 258 per 100,000 population. (BJS, Jail Inmates at Midyear 2008 - Statistical Tables, Todd D. Winton and William Sabol, Ph.D., March 2009.) Exhibit 1–1. Increase in U.S. Incarceration Rate, by Region: 1978–2005.

In 1978, the national incarceration rate was 76 per 100,000 population, as compared with 54 for the Northeast, 49 for the Midwest, 98 for the South, and 100 for the West. The last year for which national data are available is 2005, when the national rate was 252 per 100,000 population, as compared with 178 for the Northeast, 187 for the Midwest, 341 for the South, and 235 for the West.

No Relationship between Crime and Incarceration Rates There is no direct relationship between crime rates and incarceration rates. Crime rates today are at the lowest levels in 30 years and yet the number of inmates incarcerated in jails and prisons is at an all time high. That there is no direct relationship between the two can be seen in states where prison growth has leveled out or declined yet the crime rate continues to drop. Incarceration rates reflect crime policies more than levels of criminal activity. As a result, jail size is not a function of the crime rate in a community. A higher number of beds per capita does not bring about less crime; a lower number of beds per capita does not result in more crime. There is no relationship. Jails Alone Cannot Address Unmet Need Not only do crime rates not correspond to jail size, but crime statistics reflect only the level of offending that is detected. When considering that half of the violent offenses that occur are not reported (and that almost half of those crimes that are reported do not result in an arrest), one realizes that jails are, at best, responding only to a subset of actual need. Measuring this unmet or unmeasured need for the purposes of jail planning is difficult. In addition, most jails operate within systems that have built-in failure rates. Take, for example, the population that comes into contact with the criminal justice system and then

120

David M. Bennett and Donna Lattin

drops out. In California today there are approximately 2.67 million un-served arrest warrants. If the state apprehended even a modest number of these individuals, its total jail capacity would be over whelmed.12 For example, if only 10 percent of the 285,000 felony warrants in California resulted in jail incarceration, the state would need an estimated 28,000 additional beds to meet the demand. It is as true for jail planning as it is for planning within the larger criminal justice system that planned workload is based on the presumption that, as has been historically true, not every one will show up.

How Not to Plan a Jail Certainly there are times when crowding is so extreme that little doubt remains that more capacity is needed. Too often, an agreement that “something must be done” leads inevitably to the conclusion that “something bigger must be built.” Indicators of major system breakdown do of course require immediate attention. The following conditions seriously compromise the criminal justice system: • • • • • • • •

Federal population caps imposed on jails. Pending lawsuits. Large numbers of inmates being held in other jails. Jail environments that are dangerous for inmates and staff. Law enforcement “locked out” from booking some arrestees. High disciplinary rates. Suicides. Sentences not served fully.

Insufficient jail capacity can result in jurisdictions resorting to immediate, emergency measures to manage the jail population, rather than thoughtfully considering all available options. In effect, the “keys to the jail” are turned over to the sheriff as release schedules become dictated by jail crowding and not by court order. When jurisdictions are forced to release inmates outside of a pretrial court order—short of having inmates serve their full sentence—the integrity of the sys tem is compromised. At the same time, the argument for jail expansion too often turns on an incomplete argument or a weak premise. Taken on their own, the four arguments that follow are not adequate justifications for jail expansion.

The Overflow Argument Management of a jail through emergency (forced) releases is a sign of system failure. However, this is not, in itself, a sufficient argument for more jail beds. Only within a larger context can counties understand the significance of forced-release numbers. Consider the following questions: • •

Does the county have a comprehensive pretrial services program? How quickly are the courts effecting release decisions?

Jail Capacity Planning Guide: A Systems Approach • •

121

How many of those released from jail on forced releases were booked into jail on failure-to-appear warrants? What programs are in place to improve appearance rates?

The Rate Comparison Argument Although it is interesting to know how one county ranks in comparison with other counties in terms of incarceration rates, jail beds per 1,000 population, and other factors, this kind of rate comparison taken alone does not provide a sufficient argument for additional beds. It does not reveal local differences in demand or practice that can drive differences in rates. The “Let’s Build to Solve It” Argument It is understandable that counties want jail capacity to address unsolved facility issues. For example, a problem that many counties have is the time it takes them to transfer sentenced defendants from jail to state prison. However, before counties plan beds to address system-related problems, new efforts must be made to resolve these problems. The projected number of jail beds needed will be much lower if based on a negotiated 5-day transfer to prison than if based on a 43-day average wait in jail before transfer, a waiting period that is not unusual in some jurisdictions. The “Jail Data Tell the Whole Story” Argument Jail planning must not rely solely on jail data. To do so is to institutionalize current practices by assuming the existing system is operating in an optimal fashion. Jail data alone can never reveal larger system issues. For example, in most jails, defendants awaiting trial make up the largest population, but case-processing efficiency and release policies, which are larger system issues, influence the number of those awaiting trial. Jail planning should be done within a systems approach, one driven by a broader question: To what extent would modifications to the existing criminal justice system affect jails and future capacity need?

A Systems Approach Because county jail populations are constantly changing, jail planning is not a one-time process. The process of developing a master plan for man aging the jail population includes determining how efficiencies in the criminal justice system can be realized, what alternatives to jail should be in place, and how jail beds (existing beds or new beds) will be used. The plan also serves to guide decisions about the types of beds to build, in-custody program space, transition services, and other needs along the custody-to-community continuum. The plan includes assessing the efficiency of the criminal justice system first and then planning jail and program space needs based on the system modifications that the assessment suggests. Following such a process can postpone the overflow of a facility by decades. The approach to jail capacity planning presented in this guide is based on the collection of the following information:

122

David M. Bennett and Donna Lattin • • • •

Jail snapshot data. Case-processing analysis. Jail and county population trends. System assessment.

The data gleaned from the jail snapshot, case-processing analysis, jail and county population trends, and system assessment inform the development of a system master plan, a collection of strategies designed to make the most efficient use of existing resources and to manage change. Analyzing these data involves some level of statistical analysis, which this guide covers in section 5. For additional information about using statistical analysis in correctional administration, readers are encouraged to refer to the National Institute of Corrections’ How To Collect and Analyze Data: A Manual for Sheriffs and Jail Administrators, Third Edition.13 The collected data will guide the selection of a jail forecast, a statistical model that represents the upper and lower limits of expected demand at a point set years into the future. The selection of a jail forecast depends on the degree to which a county can implement tested strategies to manage growth. This will require the county to consider the effect of different policy choices. For example, a county may wish to consider the following options: • • • • • •

Taking inmates to a detoxification facility to sober up instead of booking them into jail. Implementing plans to decrease pretrial failure-to-appear rates. Reducing the average time between booking and case disposition by half for incustody misdemeanants. Expediting immigration hold cases and out-of-county cases held in the jail. Expanding diversion options for nonviolent offenders. Imposing community-based sanctions on low-risk probationers with a technical violation instead of sentencing them to time in jail.

In most significant endeavors, planning is often key to ensuring a desired outcome. The ultimate goal is establishing a strategy that facilitates continuous system assessment through ongoing data collection and policy review. The best way to prepare to plan for jail capacity is to assemble a planning team and select a planning staff or consultant who will be responsible for completing all the elements of the jail capacity planning process, including the gathering of data and their analysis and assessment of the state of the current criminal justice system.

Assemble a Planning Team The complex and interrelated nature of the criminal justice system makes it important that each county have a jail capacity planning team or criminal justice coordinating committee (CJCC) before undertaking jail capacity planning. The CJCC provides the forum for jail planning and acts as the center of ongoing data tracking, sys tem monitoring, policy review, and program implementation.14 The CJCC’s responsibilities are as follows:

Jail Capacity Planning Guide: A Systems Approach • • • • • •

123

System planning: Development of a system master plan. Ongoing review of system data. Policy review and development. Development of standards and quality control. Coordination of system information. Community education.

The CJCC should include key decisionmakers in the criminal justice system, including (at a minimum) the police chief, chief judge, sheriff, prosecuting attorney, public defender, court administrator, county commissioner, juvenile director, and probation and parole director. These individuals are the policymakers and, as such, they should sit on the committee rather than send their designees. At the same time, jurisdictions should invite other key system players (e.g., victim advocates and representatives of the chiefs of police, alternative programs, and mental health and substance abuse services) to participate as committee members.

MEMBERSHIP OF A SAMPLE CRIMINAL JUSTICE COORDINATING COMMITTEE The membership of one county’s criminal justice coordinating committee included the following representatives:* • • • • • • • • • • • • • •

The chief judge. The county district attorney (chief prosecutor). A public defender or defense attorney. A county commissioner. A health/mental health director. A city council member or mayor. A representative of the state police department (nonvoting). A police chief selected by police chiefs in the county. The county sheriff. A state court judge. A director of community corrections. A county juvenile department director. One or more lay citizens. A city manager or other city representative.

* Adapted from Guidelines for Developing a Criminal Justice Coordinating Committee, by Robert C. Cushman (Washington, DC: U.S. Department of Justice, National Institute of Corrections, 2002), NIC Accession Number 017232, www.nicic.gov/Library/017232, accessed January 8, 2009.

Full participation in the CJCC is necessary for successful system assessment and jail planning. It is important that members from all parts of the system be involved in the process. The CJCC is to become familiar with the data outlined for collection in this guide:

124

David M. Bennett and Donna Lattin • • •

Jail snapshot data. Case-processing data. Jail and county population trends.

The CJCC will want to review the availability of the data and then decide on a plan for approaching the system assessment.

Select a Planning Staff or Consultant Jurisdictions vary widely with respect to the time and expertise needed to conduct a system assessment. Some rely on in-house staff, while others seek outside assistance. Certainly, few counties are in a position to conduct a facility needs assessment or perform the architectural work needed to complete the final stages of detention planning alone, and many would benefit from a one-time system review by an outside system planner. Counties considering hiring an expert will be confronted with the choice of what kind of expert to employ. Although many jurisdictions rely on architects to assess capacity needs, the methodology presented here argues instead for a system planner—one who operates independently of the “design and build” part of a project. Criminal justice planners bring a unique perspective to the task. They offer a broad and nuanced perspective that allows them to assess system functioning, offer models from other jurisdictions, and demonstrate how shifts in policy and programs can affect capacity needs. They should bring special expertise in areas such as pretrial services, early case resolution procedures, program alternatives, and evidence-based practices, to name a few areas. They can help decisionmakers understand not only what is needed, but also what is possible. If an outside expert is used, the county’s own employees should be involved directly in the data collection and analysis phase of the work to gain expertise in these areas. Whether a county employs in-house staff to conduct this work or engages a contractor, it is imperative that local elected officials and policymakers understand the jail capacity planning process— what goes into it, its potential, and its limitations. Review and Assess the State of the System As used here, the term “system assessment” refers to an assessment of the criminal justice system. To the casual observer, the criminal jus tice system may not be an “organized assembly.” There is no central authority overseeing its components and no single mission guides its course. Often, there is no integrated system of information to inform its activities, and planning, when it occurs, is more often than not conceived in isolation and implemented without broad consultation. Yet it is clear that this collection of individuals and entities is interdependent and inextricably interconnected. Minor changes in one part of the criminal justice system can have dramatic effects in other areas. For example, placing a few more police officers on the street can have a significant ripple effect, affecting the jail and the workload of all other players in the criminal justice system. The CJCC will need to design a system assessment to review the key policies and practices over which counties have control and that significantly affect jail population. These areas include:

Jail Capacity Planning Guide: A Systems Approach • • • • • • • • • •

125

Prebooking options. Pretrial release services. Jail classification and other reoffense risk instruments. Adjudication policies and practices. Diversion options. Sentencing alternatives. Program adherence to evidence-based practices. Sanction policies and programs. Jail stepdown, reentry, and discharge planning. Data availability and integration.

For each area, make an attempt to examine policies, review available data, assess need, and understand the effect it has on jails. The CJCC will need to discuss how to approach this task, what information to collect, and how to conduct the analysis. This guide provides more information to help frame this discussion in sections 3–5. Finally, by way of preparation, the CJCC can also identify outstanding local issues that merit special attention as part of a larger system review. For example, a county may want to capture more detailed information about a particular population (e.g., drunk drivers), a specific point in the adjudication process (e.g., sanctions and revocations), or a certain type of alternative pro gram (e.g., a work-in-lieu-ofjail program).

SECTION 2. SYSTEM ASSESSMENT: JAIL POPULATION MANAGEMENT STRATEGIES An assessment of jail population management strategies should be an integral component of any jail forecast. The key to the long-term management of a jail is developing practices that allow the county not only to react to change, but also to influence and shape that change. The county will want to develop an assessment strategy for examining the larger criminal justice system, with a particular focus on areas over which a jurisdiction can exercise control and that have a direct impact on the jail. Following are examples of areas the county may wish to examine: • • • • • •

Case decisionmaking. Processing efficiency. Booking alternatives. Diversion and sentence options. Program effectiveness. Data availability and use.

Together, jail snapshot data, case-processing data, jail and county population trend data, and a formal assessment of the criminal justice system inform the development of a system master plan. The system master plan in turn guides the selection of a jail forecast. The degree to which a county commits to implementing new programs, policies, and procedures will help refine the selection of the jail forecast. This approach is in contrast to other approaches that

126

David M. Bennett and Donna Lattin

focus simply on jail capacity and trends with the sole purpose of expanding facilities. This guide takes the reader through the steps of data collection and an assessment of the criminal justice system. It ends with a focus on the selection of a jail forecast.15 At a minimum, jurisdictions should focus on the key population management strate gies outlined in this section: • • • • • • • • • •

Booking decisions. Pretrial release decisions and services. Classification and objective risk assessment. Adjudication policies and practices. Diversion options. Sentencing alternatives. Adherence to evidence-based practices. Sanction policies and programs. Jail stepdown, reentry, and discharge services. Routine examination of system data.

Each strategy includes examples of the related data this guide identifies for collection and offers examples of the relationship between the strategy and jail management. Jurisdictions will want to supplement those data with a qualitative review of their system. In the end, this review can take many forms:16 • • • • • • •

Review of local policies (booking, bail, sanctions, etc.). Conformance to standards (e.g., National Association of Pretrial Services standards). Inventory of existing programs. Analysis of outcome data. Review of existing and pending state legislation. Review of new and pending state corrections policies. Review of available incentives and sanctions.

Booking Decisions Police availability, cite-and-release policies, booking fees, and the availability of community alternatives all affect jails. Too often, jails become the only option for law enforcement officers, especially when jurisdictions try to resolve issues involving persons who suffer from mental disorders or who are social inebriates. The jail is not an appropriate or proportionate response to mental illness and public intoxication, conditions that are better treated as public health issues. Creating prearrest and prebooking options in the community has helped some localities divert these groups from jail. Examples of Data Recommended for Collection • •

Incarceration rates (source: jail Bookings by arresting agency (source: case-processing study).

Jail Capacity Planning Guide: A Systems Approach

127

Jurisdictions can use the data this guide recommends for collection as a starting point. They may use the data to discuss cite-and-release policies and a continuum of prebooking alternatives and then develop programs and policies to divert low-risk inmates who would be better served in a nonjail setting.

Pretrial Release Decisions and Services Pretrial services programs are an indispensable component of an efficient criminal justice sys tem. Expediting release and reducing pretrial failures and rearrest are two ways a fullservice pretrial services program supports jail population management goals. Pretrial staff supply the courts with accurate information about a defendant to inform decisionmaking, support the early appointment of defense counsel, identify diversion candidates, monitor pretrial jail inmates to facilitate timely bail reviews, and monitor, track, and supervise pretrial defendants. Pretrial programs are a fundamental contributor to enlightened population management. The results are compelling: Research shows a direct correlation between jail crowding and the amount of coverage a pretrial services program offers. Jurisdictions with pretrial services programs that operate 24 hours a day, 7 days a week are less likely to have crowded jails (i.e., those that exceed capacity).17 However, a well-managed jail is associated not only with access to pretrial services but also with the timeliness of those services. Jurisdictions with pretrial services pro grams that interview defendants before their initial court appearance are less likely to have a jail that exceeds its rated capacity.18 The role pretrial program staff play in the early assignment of defense counsel also has an effect on the jail. According to one study, defendants not represented by an attorney at their initial appearance are less likely to be released on their own recognizance and more likely to have an unaffordable bail set, which contributes to higher detention rates.19 One county that documented the positive effect of a pretrial program was Montgomery County, MD. After the first year of program operation in the early 1990s, the county measured decreases in the average number of jail days for pretrial defendants, reductions in failure-to-appear rates for defendants released on pretrial supervision (the lowest rates in 5 years), and low rearrest rates.20

EXAMPLES OF DATA RECOMMENDED FOR COLLECTION • • • • •

Percentage of inmates in pretrial status (source: jail snapshot). Pretrial release rate (source: case-processing study). Pretrial release type (source: case-processing study). Pretrial failure-to-appear rate (source: case- processing study). Rearrest rate (source: case-processing study).

Additional information to collect as part of a qualitative review may include the availability of services for verification, bail review, a failure-to-appear investigation, identification of diversion candidates, and a review of policies and procedures.

128

David M. Bennett and Donna Lattin

Pima County, AZ, put together a fast-track pro gram to monitor the pretrial jail population by providing routine bail review for defendants not released at their initial appearance. The program involved the collection of additional information that could form the basis of a release plan. The county gave pretrial staff the authority to sched ule bond hearings. Experts credit the program with reducing the felony pretrial jail population by 20 percent.21

Classification and Objective Risk Assessment Reliance on an objective assessment of the level of risk an inmate poses ensures the rational allocation of jail resources. A reliable, validated classification instrument is also a population management tool in that it provides objective information that jurisdictions can use to modify risk thresholds in response to population pressures. Classification instruments are used across the criminal justice system, from pretrial release decisionmaking and jail classification to probation supervision and sanction decisions. Criminal justice systems use classification and risk-assessment instruments in the following contexts: • • • • • • • • •

Jail classification. Pretrial risk assessment. Determining the risk of offender recidivism. Forced release. Sentencing guidelines that take into account the offender’s probability of reoffending. Level and intensity of supervision required. Structured sanction guidelines. Jail stepdown. Assessment of offender risk and need for program referral.

Jail planning and alternative program planning must go together. Analyzing in-custody and probation (community supervision) populations by risk can help counties determine how many resources to fund. A starting point might be to assess whether high-risk/high-needs offenders are accessing programs. In the case of jail classification, an objective assessment of risk can help ensure that inmate management decisions are consistent and reliable. It can also help spare jails the costs associated with overclassification (which leads to more intense management and longer stays) and litigation (an objective classification scheme provides more legal defensibility). Other benefits include improved efficiency and public safety. Defendant information that police, jail, and pretrial release staff collect is vital to helping jurisdictions make quick and informed release/ detention decisions that promote efficiency in the criminal justice system, effective jail management, and, ultimately, public safety. In fact, national data indicate that pretrial programs that rely on subjective assessments of risk are more than twice as likely to have a jail that exceeds its capacity than pretrial programs that rely exclusively on an objective risk assessment.22

Jail Capacity Planning Guide: A Systems Approach

129

EXAMPLES OF DATA RECOMMENDED FOR COLLECTION •

Jail inmate classification (source: jail snapshot).

Note: Some jurisdictions will want to consider a separate and more detailed analysis of their jail classification instrument as part of this study.

The recommendations for data collection and analysis presented in this guide provide an opportunity to review the use and application of risk-assessment instruments across the criminal justice system. Objective risk assessment should also be used to guide decisions about diversion, supervision intensity, and the response to noncompliance.

Adjudication Policies and Practices The adjudication of defendants includes decision points, such as the charging and filing decision, that influence case outcome and, ultimately, the jail. The case-processing study this guide outlines affords jurisdictions an opportunity to look at this process in great detail, offering them, in most cases, the first indepth examination of their case- processing system. The case-processing analysis examines many factors, including the appointment of defense counsel (which contributes to the speed of case resolution and the final outcome), policies on immigration and other types of jail holds, and overall adjudication outcomes.

Appointment of Counsel Although mounting workloads and declining budgets are a reality for most components of the criminal justice system, underfunded public defender systems have direct implications for jail crowding. When jurisdictions cannot guarantee immediate access to defense services, both jails and defendants are adversely affected. A short age of public defenders can be particularly acute in state-funded public defender systems; the state does not have the same interest as a local jurisdiction in adequately funding the public defender. When needed, the jurisdiction must appoint a public defender quickly. Once appointed, the public defender must confer with the defendant before the first appearance or arraignment and be prepared to participate in an expedited case resolution process. Assignment of the defense counsel is but one of many issues to address as part of the case-processing analysis. Early Case Resolution Swift case disposition depends on the timely receipt of police reports, the early assignment of counsel, upfront screening, quick sharing of discovery, and the early offer of pleas. Lack of such procedures affects jails directly. The effect can be seen in the average length of stay (ALOS) for defendants who remain in custody pending dis position and in defendants released “time served” because the time available for a sentence has been spent awaiting the verdict. Time-served releases often reflect system processing failures. Too often, defendants are detained pending adjudication only because the system is not prepared to adjudicate the case. When this happens, defendants take up scarce jail beds and jurisdictions delay for as long as

130

David M. Bennett and Donna Lattin

60 days those cases that could have been resolved in 10 days with a meaningful intervention. These cases have no option but time served. For many jurisdictions, an early case resolution (ECR) program is the answer. Fully implemented, an ECR program can produce significant results. It is a central management tool for improving the efficiency of the criminal justice system and managing the jail population. The objective of such a program is to provide a process for expediting the resolution of cases. The process begins with defendants having early, meaningful consultation with counsel and ends with prosecutors and public defenders meeting on a routine basis to review new cases to determine which ones might be resolved at arraignment. Given that most cases are resolved through entry of a plea, a mechanism needs to be in place for exchanging discovery, arranging negotiations between counsel, and accepting pleas at the earliest stages of prosecution. It is inadvisable to house for long periods those pretrial defendants who will ultimately plead guilty. An ECR program benefits jurisdictions in the following ways: • • • • • • •

Reduces ALOS. Reduces crowded dockets. Reduces jail overcrowding. Reduces case-processing times. Reduces the number of pretrial defendants. Reduces time invested in less serious cases. Reduces barriers to jail programs available only to sentenced inmates.

Jurisdictions will want to use the data this guide recommends for collection and other information gleaned from policy and procedure assessments to explore the use of ECR as a population management strategy. To achieve expedited case resolution goals, jurisdictions will need to implement new upfront adjudication procedures. Adequate system resources must also exist to support this effort. A sufficient number of prosecutors, public defenders, judges, and other support staff must be in place to realize the orderly administration of justice. Moreover, a pretrial program is essential to this front-end work. Jurisdictions should take all of this into account in the assessment. An ECR program results in savings in jail bed days through the timely resolution of cases. Jurisdictions that have adopted this approach have measured significant reductions in jail impact.

Positive Effect on the Jail In 2003, Orange County, FL, implemented an ECR program in response to jail crowding. It took a series of actions to achieve the timely processing of inmates, including the assignment of a permanent judge to conduct first appearance hearings. Within a short period of time, the number of inmates processed in first appearance hearings increased from 77 per day to 93 per day. The result was that the average daily population of the county’s jail dropped 15 percent, from 4,000 inmates to 3,413 inmates. In 2005, the National Association of Counties honored Orange County with an Innovative Programs Award for the effect the county’s “meaningful first appearance pro gram” had on its jail.23

Jail Capacity Planning Guide: A Systems Approach

131

Lucas County, OH, implemented elements of an ECR program in the form of a special prosecutors’ unit dedicated to immediate screening of felony warrantless arrests. In the program, prosecutors review a case with the arresting officer and make an immediate filing decision. As a result, the county either drops about 20 percent of its cases immediately or reduces the cases to misdemeanors. According to a Lucas County prosecutor, this approach has resulted in a decrease in the jail population.24 Washoe County, NV, implemented an ECR program that involves a coordinated effort at ECR by the judiciary, public defenders, and the district attorney. As a result, the time for case resolution decreased from an average of 8 weeks to 6 weeks for certain offenses. The county releases the offender the day the case is resolved through diversion, drug treatment, or another alternative. This ECR program reduced the ECR population demands on the jail.25 Lee County, FL, now resolves a quarter of its felony cases through ECR, and Washington County, OR, resolves a third of its cases through this program. Monroe County, NY, took an approach to change that it credits with postponing the building of additional jail bedspace. The approach included developing pretrial and posttrial alternatives, expediting cases, and improving case management. One specific effort involved expediting the completion of presentence investigations for in-custody cases. According to the county, this effort succeeded in reducing completion time for presentence investigations from 4 weeks to 2 weeks, saving 4,319 jail bed days in 1 year.26 Other examples of expedited case processing include a project in Maricopa County, AZ, where the county focused on expediting the adjudication of probation/parole violation hearings. This resulted in a 43-percent reduction in the aver age time for case resolution and an associated decrease in the average daily population of jail inmates.27

EXAMPLES OF DATA RECOMMENDED FOR COLLECTION • •

•

• • •

Legal representation, by attorney type (source: case-processing study). Case-processing times from booking to filing, from filing to arraignment, from arraignment to disposition, from disposition to sentencing, and so forth (source: case-processing study). Case-processing times, by charge type (domestic violence, property, drug, public order, traffic, etc.) and charge class (felony/misdemeanor) (source: caseprocessing study). Felony filing rate (source: case-processing study). No-complaint rate (source: case-processing study). File attrition rate (i.e., the percentage of defendants charged with a felony who were convicted of a felony) (source: case-processing study).

Diversion Options Diversion options offer lower cost and effective interventions for low-risk offenders. The lack of use of meaningful diversion to treatment is easily seen in the rates of release, rearrest, and return that plague jails. For example, nearly 80 percent of inmates booked into the Los Angeles jail had previously been in jail or prison. After release from jail, nearly 62 percent

132

David M. Bennett and Donna Lattin

were rearrested within 2 years. Of those rearrested, 42 percent were picked up within 3 months of release.28 This recidivism and the uncounted number of associated new crime victims show the current failure of the criminal justice system.

Drug Diversion Legislation Drug diversion programs come in many different forms, and some states, such as California with its Proposition 36, have passed legislation to mandate diversion programs for particular types of offenders (e.g., nonviolent second-and third-time drug offenders or those with probation violations linked to drug use). The California statewide diversion program has proved to be an effective measure in reducing the demand for custody resources. Incarceration costs decreased dramatically: Jail costs for defendants who completed the diversion program were measured 30 months after they entered the program and found to be 41 percent lower than the jail costs for defendants who never entered the program.29 Other court-based diversion options include mental health courts, domestic violence courts, courts specializing in cases involving driving under the influence, and community courts. Community service and work-related diversion programs are additional options. Drug Courts The success of drug courts has made a strong case for diversion programs (many drug courts now also function as postplea sentencing options). These programs work to address the underlying problems associated with criminal activity, and they have repeatedly been shown to halt further offender entry into the system. A recent analysis of the Multnomah County (OR) Drug Court (the second oldest program in the nation) tracked 11,000 offenders eligible for drug court over a 10-year period. The study found significantly reduced recidivism for drug court participants for up to 14 years after they entered the program as compared with eligible offenders who did not participate. The program reduced the incidence of rearrest by nearly 30 percent for drug court participants. In the final analysis, costs for drug court participants were $1,392 less than the costs for “business as usual” cases.30 Mental Health Services Diversion services are also an appropriate option to consider for defendants suffering from mental illness. Jail discharge planning also holds promise. Providing immediate and concrete assistance is the key. For example, a recent study found that offenders with mental illness who were released from jail with Medicaid benefits to assist with payment for continued community mental health treatment had, on average, 16 percent fewer sub sequent detentions over the following year than those who were released without Medicaid.31 Mental health courts and assertive case management programs also demonstrate a real potential to decrease demands on jails and hospitals. A new study on the San Francisco Behavioral Health Court found that, 18 months after program completion, participants had a 39-percent lower risk of being arrested for a new offense and a 54-percent lower risk for committing a violent offense than a comparable group booked into the jail who did not participate in the specialty court.32

Jail Capacity Planning Guide: A Systems Approach

133

EXAMPLES OF DATA RECOMMENDED FOR COLLECTION Disposition Type: • •

Percentage of defendants diverted to nonjail program (source: case-processing study). Nature of diversion program (source: case-processing study).

Jurisdictions will want to supplement these data with a more indepth examination of their continuum of diversion options, including a look at eligibility criteria, number of defendants served, policies on incentives and sanctions, and outcomes.

Sentencing Alternatives The Bureau of Justice Assistance advises that jurisdictions “view jails as but one alternative in a continuum of graduated responses to criminal conduct.…The availability of alternatives, treatment options, and other resources is a powerful tool in decreasing jail populations.”33 Jails must be part of a system of alternatives that allows counties to move inmates to less expensive community-based options as inmate classification and inmate behavior allow. These posttrial alternatives range from work-release facilities or community corrections centers (CCCs) to community work crews. Jurisdictions that have incorporated a CCC into the facility plan can also use the CCC as a release valve for the jail, allowing low-risk inmates to serve all or part of their sentences in a less restrictive, program-oriented set ting. This alternative lowers system costs while providing another treatment- and employment- based option.

EXAMPLES OF DATA RECOMMENDED FOR COLLECTION • • •

Percentage of offenders sentenced to prison, jail, probation, or other alternatives (source: case-processing study). Average length of sentence imposed, by type (source: case-processing study). Jail inmates, by charge or conviction type (source: jail snapshot).

Additional information to collect includes program inventory and capacity, eligibility criteria, average length of stay, sanction policies, costs, and outcomes. Continuums of diversion, treatment, supervision, and sanctions must be in place to address the varied and complex issues that offenders present. Foremost is the need for a continuum of quality alcohol- and drug-treatment programs. When offered as an alternative to jail, these programs save dollars and lives. A body of research demonstrates that a punitive approach to addiction does not lower criminal behavior and that quality treatment, coupled with case management and swift sanctions, can significantly reduce recidivism. Substance abuse treatment, in its many forms, shows repeatedly that it is a good investment. A study of

134

David M. Bennett and Donna Lattin

44 treatment programs in California across 13 counties revealed that every dollar spent on drug treatment yielded $7 in savings for the local criminal justice system.34 Work and transition programs that bridge the gap between jail and the community must be a central feature of any criminal justice system. Jails are costly to build and operate. For this reason, safe and proven alternatives to jail must be a standard feature of any jail plan.

Adherence to Evidence-Based Practices Not all programs are created equal. Research shows that the programs that achieve the greatest reductions in recidivism share common characteristics. These include targeting offenders at higher risk for recidivism, focusing on known predictors of criminal behavior, and having well-qualified staff deliver cognitive-behavioral interventions of adequate duration and intensity. Recent research demonstrates the cumulative benefit of evidencebased practices. Researchers have empirically linked the following approaches to a significant reduction in recidivism, which, in turn, has a positive impact on the jail: • • • • • • • • • •

Targeting offenders at higher risk for recidivism for the most intensive services. Providing treatment of at least 3 months’ duration. Using a reoffense risk-assessment instrument. Varying treatment intensity by risk level. Expediting entry into treatment. Ensuring treatment continuity. Delivering cognitive-behavioral programs. Ensuring swift, not severe, sanctions. Providing treatment of sufficient intensity. Offering incentives and rewards for progress.

To have the greatest influence on recidivism, jurisdictions should clearly link an offender’s risk level to the length of supervision and services they provide. The significance of targeting the higher risk offender is made clear in research studies showing that not only do services for this population provide the greatest public safety return but also that intensive services delivered to low-risk offenders can actually increase recidivism (some researchers speculate that this may be explained by bringing low-risk offenders into prolonged contact with higher risk offenders).35 Particular treatment models are crucial for realizing long-term gains. Cognitivebehavioral approaches that address criminal thinking and help individuals understand triggers for addictive/ criminal behavior have proven most effective, providing a benefit to the criminal justice system (after accounting for treatment costs) of $10,299 per individual.36 Providing a balance of treatment and supervision achieves the greatest reductions in recidivism. Surveillance-only or sanction-only approaches demonstrate no positive outcome. A review of 23 control-group studies of surveillance-oriented, intensive supervision showed zero positive effect.37 The Washington State Legislature recently com missioned research into evidence-based practices. The study found that although intensive supervision alone yielded no reductions in

Jail Capacity Planning Guide: A Systems Approach

135

recidivism, such supervision combined with a treatment- oriented approach resulted in an almost 22-percent reduction in recidivism. After accounting for the cost of supervision and treatment, the legislature estimates the cost benefits per individual to be $11,563.38 Although most research on evidence-based prac tices has focused on the properties of effective programs, recent research is paying attention to the quality of the interaction between staff and the offender. Evidence suggests that the quality and nature of the interaction is as important as the program itself.39 The State of Oregon is phasing in a law (SB 267) that makes funding for corrections programs contingent on the delivery of evidence-based practices. In the first year of implementation, the state expects that it will designate 25 percent of county funding for “best practices” and that this allocation will increase each year. Even for defendants at the highest risk levels, jail alone is only a temporary stopgap measure for preventing repeat criminal behavior. Given research findings regarding the effectiveness of treatment, it is time to rethink the approach that views treatment as an alternative; treatment should instead be considered the norm. The ulti mate goal is a system in which jail becomes the alternative.

EXAMPLE OF DATA RECOMMENDED FOR COLLECTION •

Percentage of offenders admitted who have previous bookings (source: caseprocessing study).

The data this guide recommends for collection will reveal some recycling of defendants through the system, often as a result of no intervention or ineffective interventions. Jurisdictions should adopt quality control reviews for their programs and formalize the collection of outcome data. The planning effort should include a discussion about achieving evidence-based standards for programs.

Sanctioning Policies and Programs The success or failure of treatment and supervision programs in the community strongly affects the jail. A continuum of graduated sanctions allows offenders to move up and down a range of graduated punishments based on the severity of their offense or violation and their level of risk. The use of structured sanctions not only saves jail beds but is supported by research. Research shows that recidivism is not reduced when incarceration is the sole intervention nor is reduced recidivism correlated with time spent in jail. Deterrence programs alone do not reduce recidivism.40 Although the overall effect size for treatment programs is 10 percent (an average 10percent reduction in recidivism across programs), the reduction in recidivism for all types of surveillance- oriented interventions is zero. Intensive-supervision deterrence programs had no

136

David M. Bennett and Donna Lattin

effect on recidivism, and programs like Scared Straight and electronic monitoring produced a 5- to 7-percent increase in recidivism, respectively.41 Research demonstrates that it is the swiftness and certainty of the sanction that is important, not the severity. The Oregon Department of Corrections showed this in a study that matched offender groups by risk level and then tracked outcomes for groups that received sanctions of different types and length. The study found similar rates of reconviction for highrisk offenders, regardless of time in jail. For most medium-risk offenders, higher recidivism was associated with longer stays in jail. Overall, low-cost community sanctions yielded lower reconviction rates.42

EXAMPLE OF DATA RECOMMENDED FOR COLLECTION •

Percentage of inmates with failures to appear in court (source: case-processing study).

Jurisdictions will also want to review sanction policies and the continuum of options available.

Jail Stepdown, Reentry, and Discharge Services The manner in which jurisdictions release inmates affects the rate at which they return to jail. Constructive opportunities for transition to the community can interrupt the cycle of rearrest and return. Discharge planning can help stabilize those who suffer from mental illness and ensure ongoing linkages for those at a high risk of recidivism. Stepdown options with movement to a minimum-classification/work release facility reduce jail costs while addressing community reintegration and rehabilitation goals. The reality of the modern jail is that it must function as an integral part of a larger community network. Preparing offenders to reenter society upon their release—whether after days or months of incarceration—benefits the community. The failure of the criminal justice system to prepare offenders for reentry is evident in the number of persons who repeatedly cycle through the system. Most jails are well aware of the dramatic effect this population— sometimes dubbed “frequent fliers”—can have on both the community and the jail, yet the extent of the impact can be surprising. Multnomah County, OR, determined that the most frequent of “frequent fliers,” who comprise approximately 4 percent of the jail population at any time, accounted for 26 percent of annual bookings. Some may surmise that many of those who suffer from mental illness also have substance abuse problems. This was the conclusion in a study that examined “frequent fliers” recycling through emergency rooms. A Washington State analysis found that 56 percent of those who visited emergency rooms 31 or more times in 1 year had been diagnosed as having both mental illness and alcohol or drug abuse problems.43 A jail must be part of a larger network of services. These services must extend from the community through the jail and on to a continuum of criminal justice programs and community alternatives.

Jail Capacity Planning Guide: A Systems Approach

137

Community Corrections Centers One approach to reentry is to plan for a CCC as part of jail planning. A CCC is a minimum- security residential facility that offers a structured, supervised living environment for the transition from jail to the community. It provides a lower cost option that allows inmates to serve their sentence in a minimum-security setting while maintaining employment and having the benefit of a range of programs. The principal goal of the CCC is to facilitate a successful transition back to the community. Center staff design individual case plans to address conditions of supervision, court orders, treatment needs, community safety, and victim restitution. Issues addressed include employment, life skills, and substance abuse. At the same time, the challenge is to keep from widening the net— to make sure that the CCC serves as a substitute for jail time and does not turn into an expanded custody resource. Courts base individualized plans for offenders on their risk level and needs and on the anticipated length of their stay at the CCC. For residents with short stays (less than 2 weeks), the principal goal is to connect the individual to treatment before release. For those with longer stays, the goal is to work with the resident to find employment, engage in treatment, and move into drug- and alcohol-free housing upon exit. Jurisdictions can realize some of the concepts of a CCC even if they do not have an actual building. Strafford County (Dover), NH, the recent recipient of a National Association of Counties award, has an excellent community corrections department. One of the functions of the department is to work with custody staff to develop a reentry plan for each sentenced offender. Sentenced prisoners who complete essential programming while in custody can earn an early stepdown from jail and complete the remainder of their sentence in the community. Impact on Jails CCCs have demonstrated good outcomes. In Washington County, OR, a 215-bed CCC serves a diverse population, including inmates transitioning from jail or prison, offenders serving direct sanctions, and persons undergoing short-term stabilization. The overall success rate, measured by successful completion, is 89 percent. Of the 11 percent who are unsuccessful, only 1 percent of the failure is due to the commission of a new crime. The success of the Washington County CCC is also demonstrated by its residential treatment program, which houses approximately 30 residents. This program has been evaluated and ranked in the top 8 percent of programs nationwide for adherence to evidence- based practices.44 A CCC also costs less to construct than a jail. By definition, a CCC is a minimumsecurity facility that has dormitory-style housing. One organization estimates that building costs for a CCC are one-third less than those for a jail (Rosser International, personal communication, September 9, 2007). Moreover, a CCC’s operating costs are usually less, depending on whether it is civilian run. In Washington County, OR, the cost per resident per day of operating a CCC run by civilians from the community corrections department is $65— significantly less than the $109.46 cost per inmate per day of operating a government facility.45 Like CCCs, day-reporting programs have a good track record. In Hampden County, MA, a pre- release facility serves inmates who are within 6 months of release. Inmates reside at an alternative facility and work in the community. In a move to reduce jail crowding even more, the county added a day-reporting component. This program serves offenders serving shorter

138

David M. Bennett and Donna Lattin

sen tences and pretrial defendants released with a condition to report, and functions as a stepdown from prerelease. The program not only has saved jail beds for those who need them most and reduced the costs of holding inmates, but also has improved the chance of successful community reentry for individuals who have earned the opportunity to participate in it.46 Examples of jail cost savings can also be found in jurisdictions that incorporated dayreporting centers into their continuum of services. Several months after opening a reporting center, the jail in Franklin County, PA, reported its lowest number of inmates in 4 years. This achievement reportedly allowed the county to build a smaller jail, resulting in savings of $10 million in construction costs.47 In Davidson County, TN, the sheriff sought grant funds to start a day-reporting center in a move to alleviate jail crowding. Designed for nonviolent offenders, the center’s programrich environment gives “someone an option to turn his or her life around in a positive manner,” according to the sheriff. The county judges the program a success, with a per diem rate of one-third of the jail residents actively completing programs.48 The judicial branch of the State of Connecticut established the Office of Alternative Sanctions to expand alternative programs. It developed day-reporting centers as part of this approach. These centers are community-based alternatives to jail for defendants with more serious offenses, who need more structure than straight probation provides. Participants report to these centers during the day and are under house arrest at night. One study estimates that this program saves Connecticut the cost of 700 jail beds each year.49

EXAMPLE OF DATA RECOMMENDED FOR COLLECTION •

Percentage of offenders who serve a full sentence in jail (sentence type/sen- tence length) compared with the percentage who exit to an alternative facility (source: case-processing study).

Additional information to collect includes the percentage of offenders stepping down from jail to lower level security, the percentage receiving discharge planning, the percentage exiting jail with medications or prescriptions (if needed), the percentage exiting jail with a referral/appointment to a community agency or case manager, the percentage transported from jail to stable housing, and the percentage exiting jail who receive followup.

Routine Examination of System Data Jail population management depends on access to good information. Receiving feedback about trends and performance in the criminal justice system is vital to guiding change. Criminal jus tice systems must track trend and performance data, but they must also monitor the quality of their efforts and track outcomes. Decisionmakers must have information about the long-term effect of interventions so they can answer fundamental questions such as what it cost and whether it made a difference.

Jail Capacity Planning Guide: A Systems Approach

139

Jurisdictions need to develop the analytical capacity to allow policymakers to routinely examine criminal justice system data. They should work toward an information system that allows for the linkage and integration of separate system databases. Some jurisdictions are developing data warehouses to integrate data across separate systems. This kind of capability allows for a more sophisticated and ongoing system analysis. This guide is designed to help jurisdictions structure the baseline system data and jail data they need to further this goal. Collecting these data will help jurisdictions pinpoint the weak links in their data collection system.

SECTION 3. THE JAIL SNAPSHOT Local system policies and practices that influence who is booked into jail and how long they stay largely drive jail bed usage. Because of this, assessing the need for New beds must begin with an understanding of who is in custody and how the county is currently using the jail within the context of the larger criminal justice system. This understanding comes from two valuable planning tools: jail snapshot data and system case-processing data (see section 4). This section focuses on jail snapshot data. Compiling data on the local jail and criminal justice system is fundamental to any longterm planning exercise. By revealing local practices, these data help decision- makers see “the big picture.” Framing the data helps frame the debate. With good data in hand, decisionmakers are in a better position to identify practices that affect the jail, observe patterns, and determine where to focus further study. More importantly, jail and system data pro vide a reference point, or baseline, against which to measure change. The real value lies in comparing jail and system data over time: to track and observe the effects on the jail of modifications to policy and practice. Over time, this repository of local system data will become an important source of information with which to inform and shape criminal justice decisionmaking. The first step toward this goal is to examine how a county currently uses the jail.

The Jail Snapshot A single jail snapshot offers a window into one moment in time. By itself, it can raise interesting questions but, like the individual frames that make up a movie, the real picture comes into view by collecting and overlaying multiple shots of the same scene. Only then does a story emerge. Daily jail snapshots are averaged over a 1-month period to produce a monthly composite average. Comparing monthly data composites over time can then reveal changing dynamics in the criminal justice system.50 A jail snapshot allows a jurisdiction to accurately describe how it is using its jail. A jurisdiction should be able to answer questions such as: • •

How many individuals are in jail? What is the relative proportion of misdemeanor versus felony charges/convictions?

140

David M. Bennett and Donna Lattin • • • • • •

What percentage of jail beds does the jurisdiction devote to the sentenced population and to the pretrial population? For what reasons are inmates in jail for holds? What is the impact of probation violators on the jail? Which charges are most frequently represented? How many inmates are awaiting transfer to state prison and how long have they been waiting? What is the length of stay by reason for detention or by charge?

The goal is to develop an automated routine that produces a single daily jail snapshot captured at the same time each day. The snapshot collects profile, legal status, and time-incustody data for each inmate.

A Data Standard for Local Planning The jail snapshot presents more detail than data from the Bureau of Justice Statistics. The data classify prisoners as either pretrial or posttrial. The jail snapshot, in contrast, adds a third category: holds. That is because jail snapshot data serve a different purpose. Not all persons in jail are awaiting trial or serving a sentence, and this guide acknowledges that jurisdictions need more detailed information than this for long-term planning. Holds include those prisoners who will not be processed by the local criminal justice system. U.S. marshal pretrial prisoners are one such group. These prisoners are not awaiting local trial, but they do take up local jail space, and whether the federal prisoners are awaiting trial or not is of no concern. A county must simply know how many persons will take up space in the jail. The same logic applies to immigration holds, probation holds, and holds for other jurisdictions. The approach this guide describes allows jurisdictions to isolate and quantify probation and parole violation cases, thereby facilitating both a separate examination of immigration violators or other federal prisoners held in jail and a breakout of parole violators or sentenced inmates awaiting transfer to state prison. Distinguishing these subcategories of probation and parole violators is important because state and federal prisoners have nothing to do with the local criminal justice system, but can take up a significant number of jail beds. In some jurisdictions, the state backs up prisoners in local county jails to manage prison crowding. For local planning purposes, it is crucial to be able to separate these prisoners from the total jail population and to identify the policies that govern their being housed in a local jail. Within a recommended framework, each jurisdiction will want to customize specific variables that reflect local practices and terminology. One jurisdiction’s district court is another jurisdiction’s circuit court. Jurisdictions may also want to expand the variables to allow a closer look at a particular inmate subpopulation: juveniles, the mentally ill, inmates on psychotropic medications, individuals held because of public inebriation, and so forth. Jail Snapshot Variables Jail snapshot variables fall into several broad categories. Within those categories, jurisdictions will want to tailor the data to meet local system issues and priorities, to construct a jail snapshot according to what the local situation dictates, and to reflect the data that are available.

Jail Capacity Planning Guide: A Systems Approach

141

Legal Status Because a case may have more than one legal status (i.e., pretrial, posttrial, holds), a hierarchy is used to assign cases to a particular category. The most significant charge keeping each inmate in custody dictates assignment. For each separate legal status, the jail snap shot breaks out more specific information, for example, charge class (misdemeanor/felony) and charge type (person, property, narcotics, drunk driving, public order, traffic) information. For inmates in the hold category, the breakout is limited to charge type. The purpose of the hold category is to discretely capture inmates who have ties to another jurisdiction (e.g., other county holds, federal holds, immigration holds), who are awaiting transfer to prison or another facility, or whose principal reason for incarceration is a violation of probation or parole supervision. Some of the inmates whom counties are holding for other jurisdictions will also have local charges pending. This group merits separate review because of the effect they have on the jail and because a hold makes these inmates subject to different management considerations. For example, a hold may limit an inmate’s eligibility for either a pretrial release, a posttrial alternative program, or restricted participation in a work program. Although it is important to acknowledge both the hold and any new charge, local considerations should dictate the hierarchy for classifying holds; the key is consistency in the selected approach. There are three options for classifying holds: 1. Hold as priority. In this method, counties first prioritize the inmates by hold status and then by whether they are “hold only” or “hold plus local charge(s).” 2. Local status as priority. In this method, counties first prioritize inmates by local status and adjudication stage (pretrial/posttrial) and then by “holds” or “no holds” for each category. 3. Charge severity as priority. In this final method, inmates are first categorized by the severity of the charge against them, with felony charges taking precedence over holds (i.e., an inmate with both a felony charge and a hold would be counted as a felony). For inmates with misdemeanor charges, the hold would take precedence. Counties then break out both groups further to show local charges with holds and holds with local charges.

Inmate Profile Though profile information may include descriptors such as employment status, level of education, etc., it should always include age, gender, and race. The number of races classified should reflect a locale’s racial diversity. Moreover, a separate accounting for ethnicity can be important for certain populations (such as the Hispanic population) that are often not fully identified within broad race categories. The jail snapshot should also indicate whether an inmate is a local resident. For metropolitan areas, this information helps show the extent to which neighboring jurisdictions have an impact on the county. For counties with a large tourist trade or a university population, the jail snapshot can reveal seasonal patterns that affect the system. Additionally, the jail snapshot should capture time-in-custody information. This will provide a useful indicator for gauging, over time, the relative effect that different populations have on the jail.

142

David M. Bennett and Donna Lattin

It is important to note that time-in-custody data are not the same as average length of stay (ALOS), the latter being derived by calculating an average based on time from jail booking to jail exit for each inmate (pretrial/posttrial/hold). Time in custody includes only the duration of detention at one point in time, the length of time an inmate has been in jail on a charge when the jail snap shot is taken. (Note: ALOS data related to jail planning are discussed in section 6, where they are captured as part of a historical trend analysis.) As a measure of duration, time-in-custody data provide a useful indicator for examining jail usage and exploring the system questions the data suggest. The kind of jail and systems analy-sis demonstrated in the sidebar “Case Study”— an analysis made possible by converting good data into more useful information—needs to be an integral part of jail and criminal justice system management. Tracking such information over time allows a county to explore measures that can be taken to reduce lengths of stay.

CASE STUDY A jurisdiction tracking time-in-custody data observes a steady decrease in the number of pretrial felons in jail and a corresponding increase in time in custody for this group. This observation leads to an examination of possible causes, and two possible explanations are advanced for further review: 1. Case processing has slowed, resulting in delays in release. 2. The success of a new pretrial supervision program in providing a release option for low-risk defendants has resulted in a more concentrated population of highrisk inmates who would be expected to have longer stays. A look at system data can help in evaluating the theories advanced above, leading to the following questions: Is the system experiencing more delays and continuances? What do the program data indicate about the number and type of pretrial jail releases over this same period?

Analyzing Jail Snapshot Data Analyzing jail snapshot data may provoke a series of questions about the state of the local criminal justice system. Look at the small segment of a jail snapshot from the sample county depicted in exhibit 3–1. This figure selects the hold population, highlights probation violators within the hold population, and then pinpoints misdemeanor probation violators, illustrating how a comprehensive jail snapshot provides a multilayered analysis. The overlay of levels of data allows for a detailed analysis of discrete jail populations, which is critical for intelligent jail planning. Presenting the data initiates what is meant to be an ongoing discussion about the system dynam ics behind the numbers and begins the process of generating policy questions to guide further review. Over time, the ongoing comparison of jail snapshots will enable a jurisdiction to track patterns.

Jail Capacity Planning Guide: A Systems Approach

143

Exhibit 3–1. Jail Snapshot of Inmate Status in a Sample Jail.

Exhibit 3–1 demonstrates the significant influence the hold population has on a jail and highlights the effect of supervision violation cases. This jail snapshot shows the inmate status of 539 inmates in the sample county jail. On this particular day, there are 182 individuals on hold status and 113 individuals (more than 20 percent of the total jail population) on a probation or parole violation. Digging deeper illuminates the real effect of technical violations. Though not shown here, the analysis of felony probation violations shows a similar outcome, with technical violations comprising the vast majority of violations. Also not shown in this snapshot are the time-in-custody data that provide another overlay to the analysis. Exhibit 3–1 can be analyzed even further. Digging deeper into the misdemeanor probation violation group reveals the number of bench probation cases as opposed to traditional supervision cases. Within the bench probation group, one can examine the percentage of inmates who are in jail because they did not comply with a court-ordered financial condition, and so on. This multidimensional approach allows for a more refined analysis. An analysis of the “hold” snapshot might raise the following questions: • • •

• •

To reduce impact on the jail, what would be needed to expedite the resolution of local charges for inmates who also have holds? What is the policy for responding to technical violations, and what other options should be considered? What is the role of the financial condition, and how is it monitored? For instance, are payment schedules put in place, and is there an immediate response short of a summons at the first nonpayment? Or does the system simply wait until the term of probation is about to expire and then issue a bench warrant for nonpayment of fines? Are some of the individuals who have been booked into jail on violations simply being processed for entry into the work-release center? If so, what is the system cost? The percentage of pretrial inmates is somewhat lower than national statistics report. It is likely that this reflects, in part, the local jail crowding policy that gives

144

David M. Bennett and Donna Lattin sentenced cases precedence over pretrial ones. How can an objective and validated pretrial risk assessment be developed and integrated into a comprehensive jail management strategy?

Putting a jail snapshot in place is not just another information collection routine done for the sake of more data. It is an important management and planning tool, and the analysis should be integrated into the ongoing management of the jail and the system. This analysis should be a perpetual exercise that involves key decisionmakers and the local criminal justice coordinating committee. Ultimately, the answers to the policy issues these kinds of data identify will show the local willing ness or ability to address system issues, which, in turn, will affect jail bed forecast scenarios and thereby dictate the size and cost of future detention facilities. Although all local criminal justice systems operate within the constraints of things they cannot control, jail and system data provide the detail jurisdictions need to affect positively those things that they can manage.

Automating the Snapshot Automating the gathering of jail snapshot data will provide jurisdictions with consistent, ongoing information for analysis. The foundation of this is a written codebook that defines all terms and written procedures that detail quality control protocols, with instructions on how to check data accuracy and investigate data anomalies. The manual snapshot process is laborious. Consequently, those jurisdictions that must rely on manual data collection at the jail are, of necessity, restricted in the amount of data that can be collected. It is not unusual, for example, for a manual snapshot to take the better part of a day to collect. By the time the analysis has been completed, circumstances may have changed. Not only is this terribly time consuming, but it also limits the amount of information that counties can make available for analysis. A manual approach that might yield one snapshot per month is no match for an automated system that can capture data on a daily basis and then present it as a monthly composite. The difference is not only of degree, but also of kind: composite data are more reliable. A snapshot provides a static moment in time within a dynamic system. As such, there is no guarantee that a jail snapshot taken today is representative of the bigger picture that emerges over time. Automation is important for more reasons than convenience: It is crucial to the development of reliable data upon which counties can base sound decisions. Confidence in good information is the basis for rational planning.

Jail Alternative Facility Data In addition to the jail, a jurisdiction should create snapshots that will illustrate the utilization of alternatives to jail facilities. These might include forest work camps or, as in the case of our sample county, a community corrections center.

Jail Capacity Planning Guide: A Systems Approach

145

A separate accounting of those in alternative programs outside facilities is also encouraged. In doing this, counties must take care not to double count: persons fulfilling their sentence on electronic monitoring, for example, should not be counted as inmates. On the other hand, the jail snapshot can show those serving their sentence on an intermittent schedule of checking into jail (i.e., “weekenders,” “part-timers”) as a subset of the sentenced population. For alternative facilities, an accounting of the population housed should, at a minimum, be generated on a monthly basis. This should include the average daily population, status of the individuals served (pretrial, posttrial, hold), as well as charge class (felony/misdemeanor) and charge type.

Jail Classification Study A jail classification study is another type of analysis that sheds light on jail usage. It is designed to assess jail classification protocols and criteria and to analyze infraction rates by classification type as a measure of validity—the latter accomplished by collecting data on a representative sample of jail inmates. A jail forecast should include a realistic inmate classification plan. Without this, a county can either overbuild (construct too many maximum-security beds) or underbuild (construct too few). Inmate classification drives the current and future demand for jail bed types, housing unit configuration, and program needs. And, it is another tool in identifying populations who might be good candidates—because of low risk—for jail alternatives.

Notes: Study sample includes only inmates incarcerated more than 4 days. The average number of days served was 43. Detail may not add to total becuase of rounding. Exhibit 3–2. Percentage of Inmates Released and Total Bed Days in a Sample County, by Number of Jail Days Served.

146

David M. Bennett and Donna Lattin

JAIL SNAPSHOT EXERCISE The following exercise should provide a starting place for developing a jail snapshot. As designed, it provides only basic information but enough to allow a county to set up a routine and begin examining patterns of use. Developing a comprehensive routine would then be the next step. For this exercise, document all inmates in custody on one day according to the data outline. Repeat this exercise on the same hour and day once a month for each of the next 6 months, or once a quarter for the next year, depending on the time required and resources available. The sample should include all inmates who were in the jail(s) on each sample day. Note: To calculate time in custody, document the number of days in custody for each inmate from booking to the date of the snapshot. Employ a hierarchy that ranks cases by the most serious charge: If the defendant is charged with more than one offense, record only the most serious charge and charge type. Each jurisdiction should define all terms used according to local law and common usage.

Data Outline 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Sample number: number each inmate in your sample beginning with 1. Inmate name or identification number. Day and year booked. Age. Race/ethnicity. Gender. Current address. Arresting agency. Court of jurisdiction. Legal status: pretrial, posttrial, hold, other. Hold type: parole, probation, federal, immigration, state, other county, other, none. 12. Charge class: felony, misdemeanor. 13. Charge type: person, property, narcotics, drunk driving, public order, traffic.

For those jails without a formal, objective classification scheme, the first step is adopting one. One example is the Northpointe Decision Tree model.

Jail Release Study A jail release study is another type of analysis that can shed light on jail bed usage. This involves tracking a sample of inmates from the point of release from jail, and then tracking the case back in time through the adjudicatory stages to document profile and case-processing information.

Jail Capacity Planning Guide: A Systems Approach

147

A jail release study can demonstrate jail population dynamics, showing the relative impact of different categories of offenders, offenses, and classification and admission types. It shows how a small group of inmates (those with longer average lengths of stay, for example) can have a significant effect. The weakness of this methodology is that its retrospective focus limits the analysis to information already available in the jail or official record—which is often inadequate. As a corollary to other data collection efforts, however, this approach can contribute to jail and alternative program planning. Exhibit 3–2 from the sample county jail shows sample jail release data. It demonstrates the tremendous effect that a small number of inmates with long stays can have on the jail. Although only 14 percent of the inmates in the sample were detained for 90 days or more, this group consumed 56 percent of total bed days. In the end, the value of any type of jail snapshot is realized over time, as comparative analyses reveal shifting demands and broad patterns in jail usage. This broader perspective allows for flexible management and forms the basis for ongoing planning. It is now time to turn to an analysis of the larger criminal justice system, with a look at how to conduct a case-processing study.

SECTION 4. CASE-PROCESSING STUDY With a jail snapshot routine in place, a county can examine how its custody resources are being used and turn its attention to a broad analysis of its criminal justice system. Mapping the current state of system functioning (“where we are”) is essential before beginning the process of jail planning (“whither we are tending”) because the two are inextricably linked. This section uses an inmate sample study to describe a method for examining the workings of the criminal justice system through the collection of case-processing data. The detailed data this study generated provide not only a rich baseline for analysis but also the framework for a qualitative review of system policies, programs, and practices. Several areas of emphasis— alternatives to jail, pretrial practices, and court case-processing efficiency (all of which affect jail usage)—are addressed in this section. Jurisdictions can also supplement the analysis of local case-processing data with other system data—crime rate trends and court filing data, for example. In examining these areas, the goal is to develop a systematic approach to jail management that acknowledges the complex dynamics that affect its use. The data collection and analysis outlined here can be conducted with in-house staff or with the assistance of a consultant. A case-processing study tracks a sample of cases through the criminal justice sys tem from arrest through disposition and provides a picture of the multiple factors that affect the jail. It also provides system baseline data that help reveal the efficiency and effectiveness of the local criminal justice system. A case-processing study generates data about the timeliness of case processing for both misdemeanor and felony cases. These data are crucial, for the more efficient a criminal justice sys tem can be in processing cases, the fewer beds will be necessary and the amount of bedspace constructed will last longer. Expedited case processing results in fewer defendants held before trial and can reduce the incidence of failure. There is no question that the more efficient a jurisdiction’s caseprocessing system is, the more efficient is its jail. There are additional benefits. For example,

148

David M. Bennett and Donna Lattin

victims and witnesses can appreciate a more streamlined process that offers fewer continuances and delays. In addition to case-processing times, the data gathered in a case-processing study can provide answers to other important questions, including: • • • • • • • • • •

What is the profile of the individuals booked into jail? Which offense types have the greatest impact on jail bed days? What impact do individuals with multiple prior bookings have on the jail? What types of release are used for inmates released before trial, and what are the failure to-appear and rearrest rates by type of release? What is the processing time from booking to disposition? How does this differ between misdemeanants and felons? How many defendants have an assigned public defender? What is revealed about case filing and disposition? To what extent are alternative sentences used? What types of sentence options are used? What is the average sentence length by offense or by misdemeanor/felony?

Conducting a Case-Processing Study The steps in organizing a case-processing study are as follows: 1. 2. 3. 4. 5.

Define the sample population. Determine the variables to be studied. Determine the size of the study sample. Collect the data. Analyze the data.

Each of these steps is discussed in the sections below.

Step 1: Define the Sample Population A case-processing study takes a sample of defendants booked into the jail on a new charge and tracks them through the system to disposition. (Each booking identified for the sample should be a new local arrest; individuals booked into jail for violating release conditions are not tracked.) Bench warrants, sentenced inmates, out-of-county cases, and probation-violation-only arrests are also excluded. However, individuals on active probation supervision are tracked as part of the analysis. In this way, the study provides the most direct look at how cases without multiple extenuating factors move through the criminal justice system. To ensure that the sample cases are processed fully during the study period, a sample is collected from several points over the previous year or two. The goal is to have a sample of recent cases, most of which have been resolved by the criminal justice system. Skip a quarter or two and then select a random sample at even intervals. For example, a jurisdiction may select a sample from jail bookings that occurred in the months of January, April, July, and

Jail Capacity Planning Guide: A Systems Approach

149

October. This will ensure that the sample is not influenced by seasonal variations. Selecting the sample from jail bookings will also ensure that it properly represents cases from throughout the county. The sample should include cases from multiple areas of the county and from different times of the year. Such variations of random sampling will ensure that your data accurately reflect the overall trends of a county rather than the nuances of one particular jurisdiction or one particularly violent period of time.

Step 2: Determine the Variables to be Studied Questionnaires represent one type of data collection instrument. Those used in caseprocessing studies typically gather defendant profile data (age, gender, employment, etc.), pretrial release type, key case-processing dates, sentence type and length, and failure-toappear and rearrest data for the period from pretrial release to case disposition. Each jurisdiction will develop a data col lection instrument by selecting a list of variables. However, terms and definitions will vary between jurisdictions, so it is important that each jurisdiction customize its variables based on relevant terminology and available data. Each jurisdiction should include in its data collection instrument variables that address its own unique circum stances. Once the proper terms are identified and the availability of data confirmed, a code book should be drafted that describes all terms and abbreviations. Exhibit 4–1 presents a list of key variables to include in a case-processing study. Where there are automated databases, a county may be able to supplement the list of variables. For example, the criminal justice coordinating committee (CJCC) in Utah County, UT, was also able to track inmates in the study sample into treatment and then document their exit status. Other counties may be able to collect more detailed profile information about mental health status, intoxication at booking, and so forth. The variables that are initially selected should be finalized based on input from the CJCC. Doing so helps ensure that the study can answer questions and address issues important to the policy board that will ultimately be making recommendations about jail capacity. Step 3: Determine the Size of the Study Sample Sample size depends on the annual number of pretrial jail bookings. The goal is to draw a sample that is representative of who comes to jail before trial; that is, the sample should reflect an accurate picture of what the jail encounters in a year of operation. For example, if the jail booked in misdemeanor inmates at a ratio of five misdemeanors to one felony, the sample should reflect the same ratio. A second characteristic required of a sample is reliability; in other words, if one drew several samples from a jail, what would be the odds that they would all tell the same story? The degree to which samples are the same is a measure of their reliability. For example, if a jurisdiction has a significant number of felony bookings that are reduced to misdemeanors or dismissed before disposition, it becomes necessary to oversample the felony population so that the felony sample will remain statistically valid as the cases proceed to court disposition. Oversampling ensures that a baseline population will be large enough to account for these variations and maintain an appropriate number of cases for sampling.

150

David M. Bennett and Donna Lattin

KEY VARIABLES TO INCLUDE IN A CASE-PROCESSING STUDY The list below gives an example of the kinds of data collected in a case-processing study. Appendix B features an additional listing of variables.

Detention Center 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Defendant identification number. Date of birth (month/day/year). Booking date. Release date. Education level. Employment. Charge (circle one): Domestic violence Person Property Charge class (circle one): Felony Misdemeanor Charge degree (circle one): First Second Third Number of charges. Arresting agency. Total bail amount. Release type.

Circuit and County Court 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

First appearance date. Filing date. Arraignment date. Disposition date. Sentencing date. Number of administrative hearings. District attorney charge class. District attorney charge degree. District attorney charge. Attorney type. Disposition type. Disposition charge class. Reason for nonconviction. Bond amount. Release type. Sentence type and length. Failure to appear: yes/no. Rearrest: yes/no. Prior arrests.

Drug

Public order

Traffic

Jail Capacity Planning Guide: A Systems Approach

151

Exhibit 4–2. Recommended Jail Sample Size Pretrial Admissions 1,000 2,000 5,000 10,000 25,000 50,000 or more

Recommended Sample Size 516 696 880 964 1,023 1,045

Note: The sample size is based on a 95-percent confidence level and ±3-percent confidence interval (see Creative Research Systems, www.surveysystem.com/sscalc.htm#top).

Exhibit 4–2 presents recommendations for sample size based on annual jail admissions. Most samples do not need to exceed about 880 inmates. However, if resources permit, a larger sample is preferable because as a sample’s size increases, so does its reliability. Note that some jurisdictions have a limited budget for data col lection and it is better to have some data than no data at all. Exhibit 4–2 is merely a guide to deter mining sample size and not a rigid rule. Today some criminal justice systems have achieved a level of electronic data integration that allows a full year of data to be analyzed, negating the need for a sample. This, of course, is the ideal.

Step 4: Collect the Data Once the sample has been identified, the work of collecting the data begins. Many local criminal justice systems have computerized databases. However, it is unlikely that all data will be auto mated. Rearrest and failure-to-appear data often require a case-by-case check, for example. In such instances, data collectors will need to be trained. The best approach is to enlist the help of individuals who already work within the sys tem. Because the data may have to be drawn from several sources and/or from several agencies, cooperation among all system sources and agencies involved is essential. Pretesting the data collection instrument with a trial run of 20 to 30 cases at the beginning of the study is indispensible. Some data elements will likely need to be changed. It may be that information purportedly found in files is not reliably available. A data collection instrument is easier to adjust at this stage than after the data collection is fully under way. Data can be collected with paper forms and then entered into a spreadsheet. The data will ultimately be transferred to a statistical software program for more detailed analysis. Appendix C features a list of sample calculations. At a minimum, the software used for the analysis should be capable of these basic calculations below: • • •

Means, modes, medians, and standard deviations. Frequency counts and distributions. Cross tabulations of two or three variables.

The county will want to enlist the assistance of someone with a basic knowledge of statistical software systems such as the Statistical Package for the Social Sciences (SPSS) or SYSTAT. Although a county can contract for such assistance, the goal should be to develop

152

David M. Bennett and Donna Lattin

in-house expertise for this kind of analysis. Appendix C features an additional list of sample calculations that jurisdictions may use in a case-processing study.

DETERMINING THE SIZE OF THE STUDY SAMPLE IN THE SAMPLE COUNTY The sample county jail books more than 18,000 inmates a year. Even though a few inmates inevitably will be excluded for incomplete data, the size of this inmate population should be sufficient to provide a sample large enough both to be reliable and to be drawn in a manner that fairly represents seasonal changes in the population. Although this county could have used a larger sample (a good idea if possible), it chose to draw a sample of 800 inmates. To be representative, a sample must be drawn from weeks and months that offer a balanced picture by accounting for seasonality and other expected variations. To draw a sample effectively, the jail must go far enough back in time that almost all the inmates in the sample period will have been processed through the criminal justice system by the time data collection begins. For this reason, skipping two quarters and then taking a sample from each of the four previous quarters is usually recommended, depending on case-processing times, to ensure that the cases will have been resolved within the study period. The sample of 800 defendants booked into the sample county jail was taken from different days and months over a 9-month period. Each defendant identified for the sample was a new local arrest. Bench warrants, out-of-county cases, and probation violation-only arrests were excluded. The sample includes the first 100 misdemeanor bookings and the first 100 felony bookings from the months below. To account for potential variations related to the week of the month and the day of the week, a different week and day were chosen to begin the data collection for each month: • • • •

April 10, 2004. July 23, 2004. October 16, 2004. January 3, 2005.

Step 5: Analyze the Data Initial Categorization After the data have been collected and entered into a statistical software program, the initial categorization of the data can begin. First, calculate the frequency distribution for each key variable identified (e.g., age, gender, race, charge class (felony or misdemeanor), legal status (pretrial or posttrial), and so forth (see exhibit 4–1). A frequency distribution quantifies the number of times a given variable occurs in a set of data. Calculating frequency distributions allows the analyst to make certain that no errors have occurred in the coding of the data or in its entry into the computer. Because coding errors almost always occur, frequency distributions should be calculated to “clean” the data before any further analysis is

153

Jail Capacity Planning Guide: A Systems Approach

performed. Calculating frequency distributions also provides an opportunity to discover how much information is missing for each of the variables in the dataset. After the data have been cleaned, scan the frequency distributions for basic information about the sample. Examine profile data, court processing time, failure rates, and other information about the system. Examine the data on a given variable either individually (e.g., what types of sentences were received) or in relation to other variables. The “cross tabulation” function of a statistical software program allows examination of the degree of association between variables. Exhibit 4–3 shows the relationship between charge class and selected other variables. Note that within each variable category, the numbers for each subgroup sum to the category total. It is this capability for cross tabulation that creates the potential to ask nuanced questions of a complex system. Exhibit 4–3. Relationship between Charge Class and Selected Variables for Defendants Booked Into a Sample County Jail Variable Age (years)* All inmates 18–24 25–34 35–44 45 or older Type of release All types Recognizance 10% bail Forced Sentence type All types Prison Jail Community corrections Diversion Probation Fine Rearrest All inmates Rearrested Not rearrested *

Misdemeanor Number of Inmates Percent

Felony Number of Inmates

Percent

391 107 136 83 65

100 27 35 21 17

386 106 132 103 45

100 27 34 27 12

309 214 52 43

100 69 17 14

230 117 53 60

100 51 23 26

299 0 91 24 76 97 11

100 0 30 8 25 32 4

199 50 27 53 1 68 0

100 25 14 27 1 34 0

309 53 256

100 17 83

230 77 153

100 33 67

The average age of inmates charged with misdemeanors is 33.1; the average age of inmates charged with felonies is 31.7. Note: Detail may not add to total because of rounding.

154

David M. Bennett and Donna Lattin

Case-Processing Analyses The breadth of system data collected in the case-processing study lends itself to varied analyses. These analyses can help frame discussions about system policies, the availability and use of alternative-to-jail programs, and the quality of supervision and treatment services. One population grouping, for example, that allows for specific analysis is the pretrial dataset. One can compute what the average court processing times might be for this group. Further dividing the sample into pretrial inmates who leave the jail before their case is concluded and those who stay in jail throughout their pretrial processing time can be constructive. Comparing these two timelines ought to reveal valuable information about the efficiency of court processing, the pre trial release rate, and whether incarcerated pre trial inmates are on a faster track for having their cases handled or not. Analyzing the decisionmaking that occurs in the criminal justice system is also instrumental for determining the efficiency of the system. For example, examining the data on the charge a defendant was booked on, the charge filed by the prosecutor, and the charge for which the defendant was ultimately convicted—along with the resultant impact on the custody status of the defendant—reveals the cumulative influence of prosecutorial and judicial discretion on jail capac-ity needs. With the information obtained thus far, it may be possible to identify points at which expediting the processing of defendants might help the local criminal justice system gain jail space. Remember, the degree to which a jail is full is a function of both admissions and time in custody. Court appearance data for study sample defendants on pretrial release can provide valuable information. The data on who failed to appear in court for their cases and who were rearrested while waiting for their cases to move to disposition can help determine which subgroups of the jail pretrial population can safely be released on their own recognizance or on supervised pretrial release, which defendants should be held until disposition, and what the time to release is. Taken together, this information gathered from the pretrial dataset provides a framework for assessing the potential of a county’s pretrial services program to mitigate pressures on the jail. Followup questions that can be asked about the program in this regard include: • • • • • • •

What percentage of defendants booked receive a pretrial interview? Is there staff coverage 24 hours a day? Is an objective and validated risk instrument used? In what percentage of cases is the judge presented with verified background information at the time of a defendant’s first appearance? Does the program provide supervision and court date notification? Does the program provide ongoing review of the jail pretrial population for bail review consideration? Does the program work to return failure-to-appear cases to court in lieu of issuing a warrant?

Further analysis will provide more detailed information about sentenced inmates, including age, sex, vocational and educational background, and sentence type and sentence length. These data can be the starting point for discussions about the local continuum of alternatives to jail, specifically, how these alternatives are used and what might be added or

Jail Capacity Planning Guide: A Systems Approach

155

enhanced to provide additional nonjail sentencing or sanctioning options. Questions raised might include: • • • • • • •

What prebooking services are in place to pre vent use of the jail for detoxification or stabilization of mental health crises? What diversion resources are in place? What alternative-to-jail facilities exist? What community-based resources are in place? What kinds of treatment or supervision services exist for specialized populations (e.g., offenders with mental disabilities, female offenders, sex offenders, and others)? What programs are available for inmates in custody? Which community-based programs can be accessed as a sanctioning option?

The study data can also provide information about the adjudicatory process for both felony and misdemeanor cases in the local criminal justice system. For example, it will be helpful to know what percentage of new felony arrests later pled to misdemeanor-level offenses. Revealing the overall pattern of prosecution practices, charge attrition (see exhibits 4–4 and 4–5), and case-processing times can be the start of discussions with the prosecutor’s office to find a way to expedite case processing or explore a different approach for handling and disposing of particular offense categories, again saving jail space. It can also lead to discussions about local policies and how they govern case processing. Questions might include: • • • • • • •

What policies guide law enforcement citation versus booking decisions? What policies dictate pretrial release? Are there policies to allow efforts to return failure-to-appear cases to court before issuing a warrant? What are the eligibility policies for diversion or other alternatives? To what extent do probation officers have discretion to sanction an offender without returning the case to court? Are there mandatory sentencing policies or local sanction policies that dictate the type or length of sentence? Are there policies to allow a sentenced offender to be moved along a custody/ community program continuum?

Exhibit 4–4. Relationship between Age and Charge Class for Defendants Booked into a Sample County Jail Age 18 to 24 25 to 34 35 to 44 45 or older

Misdemeanor Number Percentage 107 27% 136 35% 83 21% 65 17%

Felony Number Percentage 106 27% 132 34% 103 27% 45 12%

156

David M. Bennett and Donna Lattin

Exhibit 4–4. Relationship between Charge Attrition and Charge Class for Defendants Booked into a Sample County Jail

All charge attrition Booking District attorney filing Disposition

1,038 391 336 311

86 93

80

920 396 310 214

78 69

Percentage of Dispositions from Booking

Percent*

Number

Felony

Percentage of Dispositions from Booking

Percent*

Charge Attrition

Number

Misdemeanor

54

* The percentages are of the previous event. For example, 86 percent of the misdemeanants arrested had charges filed.

Finally, data such as the percentage of admissions made up of offenders already on supervision and the rearrest rates for those released from jail before case disposition can lead to discussions about the effectiveness of local interventions that serve to reduce recidivism and returns to jail. These discussions should be informed by the research on practices shown to reduce risk and decrease recidivism. There is now a body of literature about these best practices that takes as its starting point the use of objective and validated risk assessment tools. It suggests that the most intensive resources be reserved for offenders at the highest risk of reoffending. An example of a program that embodies many of these principles is drug court. Drug courts divert nonviolent, substance-abusing offenders from prison and jail into treatment. The concept embodies many variations, but is premised on judicial monitoring, a team approach to case review, swift sanctions, and positive incentives—to name just a few. Following are some questions that can be asked to generally assess adherence to best practices: • • • • • • •

Is an objective and validated risk instrument used to allocate supervision and treatment resources? Are the most intensive interventions reserved for the higher risk offender? Do interventions focus on individual risk factors associated with criminal behavior (substance abuse, employment, peers, attitudes, etc.)? Are programs of sufficient duration (3 to 9 months at minimum) and intensity (structuring 40 to 70 percent of an offender’s time)? Does the program employ sanctions in a manner consistent with reducing recidivism (swift ness, rather than severity, is important), and does it use positive enforcement? Are interventions specialized or tailored to particular populations (e.g., women, persons with mental disabilities)? Are staff trained in interaction/communication styles that have been shown to foster positive change in offenders?

Jail Capacity Planning Guide: A Systems Approach

157

SAMPLE COUNTY FINDINGS • • • • • • • •

Thirteen percent of felony defendants and 8 percent of misdemeanants had six or more prior bookings into the sample county jail since 2000. Forty percent of the sample had not completed high school. Twenty percent of the defendants were intoxicated at booking. Twenty-six percent of felony defendants released pretrial were rearrested awaiting disposition of their case, higher than the national average. Seventy-eight percent of filed felony cases resulted in a conviction, higher than the national estimate of 64 percent. Ninety-three percent of local felony dispositions were guilty pleas. The average time from booking to disposition for felony defendants was 4 months (lower than the national average). The average duration of a local jail sentence for felony offenders was 4.3 months.

Summary This section barely touches on the numerous questions that can be asked of caseprocessing data. As questions are raised about the data and as discussion follows, it is normal to go through several iterations in bringing results back to the CJCC for inspection and discussion. Particular agencies will have questions of the data that can entail a more detailed analysis or further review. The sample county study yielded valuable system information. It revealed, for example, that high pretrial failure rates affected the jail significantly and that case-processing times were reasonable but could be improved. A case-processing study may expose issues that merit more indepth analysis, warranting the undertaking of additional specialized research. In the sample county, for example, two other studies were conducted: 1) an analysis of the processing and disposition of probation violation cases and 2) an analysis of the transition of inmates from jail to the community corrections center. Both were undertaken after early data analysis illuminated these two areas as having a direct effect on jail bed usage. Case-processing data and more qualitative sys tem assessments allow a jurisdiction to become familiar with the workings of its criminal justice system and to build on this knowledge to develop strategies for reducing the size of the inmate population and/or controlling the rate of growth. The case-processing study in the sample county resulted in more than 80 system recommendations, with a focus on the development of an expedited case disposition program, discussion about the establishment of a comprehensive pretrial program, attention to the use of nonjail sanctions, and a review of the policies that governed the movement of inmates from jail to the community corrections center. The value of this systematic approach is the ability to prioritize areas of greatest concern—or of greatest potential—in developing jail management strategies. The set of local strategies adopted for inmate management and control of the growth of the inmate population represent only one part of a county’s system master plan.

158

David M. Bennett and Donna Lattin

SECTION 5. JAIL CAPACITY PLANNING OVERVIEW There is good reason why weather fore casts do not usually exceed 5 days: The reliability of any prediction weakens with time. For jail planning, this realization is tempered by the fact that facility planning and construction together constitute a lengthy and expensive proposition. For this reason, most jurisdictions base their jail forecasts on 10or 20-year horizons. Within that range, jurisdictions can map out a plan in 5-year increments, if desired. Just as there is no standard time horizon in which to forecast the weather, there also is no set historical timeframe from which to collect data. However, jurisdictions cannot rely on data from shorter periods (less than 5 years, for example) with confidence. The rule for collecting jail data should be to go back as many years as time and accuracy permit. Ideally, jurisdictions will collect at least 20 years of historical data.

The Jail Capacity Forecast The jail capacity forecast builds on an analysis of three types of data, which this guide refers to as “jail usage variables”: • • •

Admissions (ADM) rate. Average length of stay (ALOS). Average daily population (ADP).

The following two adjustment factors further refine the jail capacity forecast: • •

Peaking factor. Classification factor.

DATA USED IN A JAIL FORECAST Jail Usage Variables • • •

Admissions rate. Average length of stay. Average daily population.

Adjustment Factors • •

Peaking factor. Classification factor.

Local and State Incarceration Rates Finally, the jail capacity forecast is checked against local and state incarceration rates to determine where local practice falls within historical and regional contexts.

Jail Capacity Planning Guide: A Systems Approach

159

The first step for a jurisdiction should be to collect historical data for jail usage variables for as many years as data are available. In some states, jails are required to submit monthly or annual reports to their state jail inspection office or to the state department of corrections. When historical local data are not available, it can pay dividends to retrieve this information from these agencies, if possible. If data on all three variables are not available, not all is lost. The relationship between the three jail forecast variables (ADM, ALOS, and ADP) allows a planner with only two variables on hand to arrive at the third by applying a simple mathematical formula. Understanding how to use this formula can come in handy because many jurisdictions do not have complete data. Most jails, for example, will have ADP data for each year for the past 5 to 20 years. Some will have ADM data for that same period. If two of the variables are known, then the third can be calculated using the formulas given below.

Admissions Jail admissions (ADM) provide an indicator of workload, and over time, they help gauge changing pressures on a jail. Factors that influence the number of admissions include the population of the jurisdiction, police resources, the availability of prebooking alternatives (including the use of citations, summons, or detoxification/crisis intervention centers), the pretrial failure rate, the supervision violation rate, and program effectiveness. Locating records of jail admissions is the starting point for developing a jail capacity forecast. This information allows a jurisdiction to track the demands on its facility over time. Jurisdictions with an alternative custody facility such as the sample county’s community corrections center would track ADM, ALOS, and ADP data for the second facility as well. The sample county would need to collect this data for both the jail and the community corrections center, because together they form a continuum of sentence options. Appendix D features a list of data from the com munity corrections center in the sample county. Historical jail admissions data collected for the jail capacity forecast, which examines trends over time, provide a broader look at ADM than either the case processing study or the jail snapshot. Admissions Data to Collect Calculate the total number of individuals booked into the jail each month for as far back as this information is available, but for at least the past 5 years. Exclude those who enter only for fingerprinting (“print and mug”) and sex offenders reporting to the jail to register. Next, calculate the total daily admissions for each month over the same period. The sum of the month’s daily admission counts should equal the number obtained for the total admissions for the month. Count admissions only; do not count releases and do not count individuals who were brought to the jail but not booked. Do not count as new admissions those who went to court and were returned to jail on the same day, even if their status changed. If the jurisdiction booked an individual more than once during the month, count each new admission to the jail separately. Count inmates serving an intermittent sentence, such as a weekend sentence, as an admission for their first booking only. If possible, break down the monthly count by other inmate characteristics such as age grouping, gender, charge, and mental health status. The sum of the inmates in each of these categories must equal the monthly total. When sorting inmates by charge, classify those who

160

David M. Bennett and Donna Lattin

have multiple charges according to the most serious charge, using the same hierarchy of charges applied in the jail snapshot: domestic violence, other person, property, narcotics offense, drunk driving, public order offense, or traffic. If the data are available, it may be useful to contrast admissions for felons and nonfelons or pretrial and posttrial inmates in each year, for as many years as are available. This provides a more detailed analysis of the trends present within discrete subpopulations, which may in turn reveal particular system dynamics. Planners might discover, for example, that felony admissions have been constant, but that nonfelony admissions ebb and flow with the priorities of local law enforcement.

HELP WITH STATISTICAL ANALYSIS FROM NIC Much of the advice in this guide relies on readers’ understanding that legislative factors, community involvement, and the workings of the criminal justice system all contribute to determining what constitutes arrest and how inmates, once arrested, are tried. The outcome of these cases determines length of stay in jail, which in turn determines future jail capacity. Reliance on numbers and statistical data is only one aspect of analysis. A thoughtful approach balances numerical data with other, less quantifiable data. This guide discusses both forms; yet, some may require additional help in understanding the numerical aspects. For those readers, How To Collect and Analyze Data: A Manual for Sheriffs and Jail Administrators, Third Edition, may help. How To Collect and Analyze Data demystifies statistics while providing tips on efficient data collection, data analysis, and organizing data for interpretation. County officials, other agencies, and the public must understand the data. To help, the manual provides step-by-step instructions for using statistical data to improve an organization’s efficiency, find support for funding initiatives, and make informed decisions about obtaining data, storing them, accessing them, and applying methods for interpreting them. Readers will find explanations of management techniques, fundamental concepts in mathematics and statistics, and ways to maximize the potential of information systems. The appendixes include a glossary of technical terms, an annotated bibliography, sample forms for data collection, and tables for determining sample sizes and generating random numbers for use in sample selection.

Average Length of Stay ALOS data can be a measure of system efficiency reflecting, for example, the time needed to move from booking to pretrial release or from booking to case disposition. ALOS data can also indicate the seriousness of offenses that jurisdictions prosecute through the local jail or that reflect changes in policy. A legislative directive to shift inmates from the prison to jails, for example, can significantly influence jail ALOS. Other factors that affect length of stay in jail include the availability of pretrial release options, case-processing times, and access to sentence alternatives.

Jail Capacity Planning Guide: A Systems Approach

161

ALOS data can also show how jails, unlike prisons, are processing centers. The majority of individuals booked into jail will be in and out within the first few days. The sample county has data to demonstrate this fact (exhibit 5–1). In the sample county jail, 67 percent of all individuals booked in 2004 remained in custody for 4 days or less. A summary table of annual ALOS can be assembled to allow for cross-year comparison of the data. Note that ALOS differs from the time-in-custody measure captured in the jail snapshot and from the adjudication times measured in the case processing study. Exhibit 5–2 shows the ALOS in the sample county over different periods.

CALCULATING AVERAGE LENGTH OF STAY The average length of stay (in days) for a given year equals the annual average daily population multi plied by 365, divided by the total number of admissions in that year: (ADP x 365)/ADM = ALOS

Average Daily Population The jail population is made up of inmates with varying profiles. These include legal status (pre trial, posttrial, hold), age (adults and juveniles), gender, charge class, classification, and other characteristics (mental illness, intoxication, etc.). ADP is a general indicator of jail bed need, representing the degree to which a facility operates at capacity. ADP is a direct function of ADM and ALOS. Examined over time, ADP will show the degree to which the jail is operating under pressure. Most jails take a daily count of all inmates. When collecting data, make sure the count is taken at the same time each day. For example, use the count taken at midnight or 6 a.m. for each day of the month. Collect data for as many years as are available, but for at least the past 5 years. Exhibit 5–1. Average Length of Stay for All Bookings in a Sample County: 2004 Average Length of Stay (hours) Less than 24 24–96 96 or more

Percentage of Inmates 32 35 33

Exhibit 5–2. Average Length of Stay in a Sample County Period 10-year average 5-year average 2005 2006

Average Length of Stay (days) 12.8 14.6 14.3 14.7

Note: Data represent an average of the jail and the community corrections center.

162

David M. Bennett and Donna Lattin

Ideally, monthly ADP data will be available. Collect the daily inmate count and add the counts to determine the monthly total. If possible, break down the monthly count by the status of the inmates counted. The sum of the inmate days in each of these categories must equal the monthly total. If the monthly ADP is not available, then the annual ADP would be used. Section 3 discusses the compilation of ADP with relation to developing a jail snapshot. Ultimately, a formalized jail snapshot is a way to capture ADP data accurately. Record each inmate in only one category, al though some inmates will have multiple charges. In all cases, when there is the likelihood of local prosecution, classify the inmate under the local classification: pretrial or posttrial. When defining the status by charge category, use the most serious charge. For example, count a defendant admitted on both felony and misdemeanor charges as a pretrial felon. Where data are available, further identify detention days for pretrial and posttrial inmates by type of offense. Felon and nonfelon are important categories to include. Jurisdictions can also tailor the data collection to reflect more indepth break downs (e.g., driving while intoxicated, traffic, misdemeanor).

CALCULATING AVERAGE DAILY POPULATION The average daily population for a given year is the number of admissions in that year multiplied by the annual average length of stay, divided by 365 (days in a year): (ADM x ALOS)/365 = ADP

Forecast Adjustment Factors Criminal justice planners have usually factored in a percentage of the total number of jail beds when developing their jail capacity forecast, so as to absorb peaks that occur throughout the year and manage fluctuations in the number of inmates in different classification categories. Often referred to as “rated beds above the operational capacity used,” these percentages have ranged anywhere from 10 to 20 percent. Jurisdictions commonly use these numbers to identify when a jail is becoming crowded. Some jurisdictions experience large population changes during certain seasons (e.g., resort communities) and require a larger margin of additional beds, whereas others see very little change throughout the year. There is more than one way to determine how many beds a jurisdiction will need to manage peaking and classification fluctuations. When the Jail Capacity Forecast Workbook was completed, the authors decided that it would be more accurate to split peaking and classification factors so that jurisdictions could complete a more accurate assessment. The resulting assessment was born out of a thoughtful process, not one suggesting that administrators simply add more jail beds to the forecast. This guide employs the same methodology with minor changes.

Peaking Factor The forecast methodology anticipates facility demands based in part on an analysis of changes in ADP; however, during peak periods—traditionally weekends, the end of the

Jail Capacity Planning Guide: A Systems Approach

163

month, and the summer months—jail populations climb. The jail must be prepared to have space available during such peak periods. Adjustments for peak periods are made by going back to several years during which the jail had not yet reached capacity. These years provide the high population counts needed to calculate a peaking factor for each year.

CALCULATING A PEAKING FACTOR Document peak counts for jail inmates. Obtain the three highest daily population counts (“peaks”) during each month of the past 3 years. If the jail was operating at capacity or was under a cap, then take the data from the most recent 12-month period when the jail was not at capacity. Record this count data by day, month, and year. Obtain the annual peaking factor for each of the 3 years. For each month, take the sum of the three peak daily population counts. Divide this number by three to obtain the peak daily population count for that month (P). Then add the 12 peak daily population counts and divide the total by 12 to obtain the average peak count for the year: (P1 + P2 + P3 + P4 + P5 + P6 + P7 + P8 + P9 + P10 + P11 + P12)/12 = Average Annual Peak Daily Population Count Divide this number by the annual ADP to obtain the peaking factor for the year: Average Annual Peak Daily Population Count/ Annual ADP = Annual Peaking Factor Obtain the 3-year average peaking factor. Calculate the sum of the three annual peaking factors (APFs) and divide this number by three to obtain the average peaking factor for the 3-year period: (APF1 + APF2 + APF3)/3 = 3-Year Average Peaking Factor Obtain the projected peaking factor. Multiply the 3-year average peaking factor by the projected average daily population for each of the forecast years to obtain the projected peaking factor: (3-Year Average Peaking Factor) x (Projected Annual ADP) = Projected Peaking Factor

Classification Factor Classification, a second adjustment factor, takes into account the flexibility needed to separate populations by characteristics such as gender, risk level, mental health, physical health, and disciplinary segregation. The classification factor pro vides for those times when the number of inmates in a classification category exceeds the number of beds available for that classification. It creates a planning cushion that allows for the jail’s need to have a few open beds within each classification category available at all times for new inmates. There is no one percentage or number that will work for every jurisdiction, as each jurisdiction is unique. Nor is there a single formula that can assure a jurisdiction that it will

164

David M. Bennett and Donna Lattin

build space for just the right number of additional beds. One rule of thumb might be to apply a classification adjustment factor for each of the primary classification categories—that is, to select a specific number to apply to each different classification category. This decision will depend on the facility size, type of inmate housing unit (direct supervision, podular remote, dormitories, etc.), gender separation factors, and the number of housing units dedicated to each classification category. A different method for determining a classification factor is to consider the number of classification categories housed in the jail. This method is possibly more valid than applying a general percentage factor or assigning the same number to each of the primary classification categories. For example, if a jail holds primarily medium-security, post-sentenced male inmates, only a small percentage of additional beds may be needed to accommodate temporary classification issues. However, if the jail is a “full service” facility that holds a mix of male and female inmates, inmates with mental illness, and pretrial and posttrial inmates, the percentage of additional beds allowed for is likely to be much higher. Each jurisdiction must decide for itself the appropriate number of beds needed to accommodate the numbers for each classification, keeping gender separation issues and special management needs in mind. Much of that decision will depend on how many different classification categories are allowed to mix in housing areas and how many housing units there are. Accepting that forecasting is not an exact science, and that efforts to manage the jail population are most important, the goal is not to increase the number of jail beds unnecessarily, but rather to use a conservative approach for setting aside additional beds to handle both the peaking factor and the classification factor.

County Population Trends County population is an especially important variable to study in relation to jail admissions.51 Tracking population growth rates helps anticipate future demands on the jail; admissions per 100,000 county population (i.e., the number of admissions per 100,000 residents of the county) provides a rate that allows for the examination of trends in jail bookings. As a county’s population grows, the number of admissions most likely will increase; however, the admissions rate may remain constant. The U.S. Census Bureau collects county population data every 10 years. The Bureau surveys all households to determine the number and ages of people in the nation. The Bureau then breaks the data out by state and county. Many county and/ or state planning departments will already have reviewed and interpreted the findings.

DATA TO COLLECT: COUNTY POPULATION TRENDS AND PROJECTED GROWTH FIGURES Collect actual population figures for each year, going as far back as average daily population is avail-able. Additionally, break out totals by gender, race, and age groups where available. Using official U.S. Census data, document the county’s projected popu-lation for 5, 10, 15, and 20 years into the future.

Jail Capacity Planning Guide: A Systems Approach

165

Tracking population trends by age cohort and other characteristics adds a greater level of detail to the forecast and can help a county look at broad population trends (e.g., Is the population expected to get older or younger over time?). Adding this information comes with a note of caution: Population trends do not always follow expected courses, and levels of criminality may not always conform to the expected age hierarchy (where crime diminishes with age). Tracking county population by age cohorts can also reveal interesting information about the types of services that a jurisdiction needs to plan for, but such tracking is not necessarily a reliable tool for predicting changes in crime rates or in jail demand. A study that examined crime trends nationally and in California concluded that with respect to the role of changing age demographics: It appears that the crime rate decrease in the early 1980s was largely driven by demo graphics; the number of juveniles (17 years of age and under) and youths (18 to 24 years of age) in peak crime-prone ages decreased markedly. In contrast, it seems that the crime rate decline from 1991 to 1999 had very little to do with demographics since the number of individuals in crime-prone ages changed very little from year to year.52

Statistical Models Used in Forecasting Forecasting is a method for translating past experience into estimates of future need. The method a jurisdiction chooses depends on the experience of the planner and data availability. Jurisdictions employ several statistical methods in jail forecasting, including regression (causal) models, rate and ratio models, jail exit analysis, and time-series models.

Regression (Causal) Models Regression models attempt to reveal the many variables that have influenced the jail and then speculate how causal relationships between selected variables would affect future demand. Regression is a statistical technique that selects a number of independent variables (e.g., crime, arrests, filings, population) and conducts a regression routine to determine the relationship between variables and their predictive strength. The objective is to determine the extent to which discrete variables can be applied to the task of forecasting future jail need. The goal is to construct a set of “what if” scenarios to examine the relative influence of different variables on fluctuations in jail population. The appeal of using a regression model for jail forecasting is that it allows a jurisdiction to analyze the interactive effect of different variables and then manipulate them to imagine different futures. The problem with using a regression model is that it can provide a false sense of precision. Forecasting jail bed needs will never be an exact science. The multiple and shifting factors that affect a jail are too numerous to be captured completely to reliably predict future need. In large part, this is because jails are influenced by events and policies over which a jurisdiction has little control. For example, questions like the following might arise: • •

What would be the effect of adding more police officers or more judges to the jurisdiction? How might one quantify the impact of more pretrial resources?

166

David M. Bennett and Donna Lattin • •

What could be the expected benefits of programs that better conform to “best practices”? How would a change in sentencing policy affect the jail?

Regression models also suffer from other shortcomings, such as a lack of available data. Jurisdictions often lack the detailed retrospective data they need for this kind of analysis. Another shortcoming is complexity: The need to use complex statistical techniques puts this method beyond the expertise of many jurisdictions. Nevertheless, there is real value in being able to ask “what if?” questions of the criminal justice system and to engage in planning that proposes different outcomes. Large-scale academic studies can be informative in efforts to understand the degree to which various factors can predict changes in jail populations. One study found, for example, that reported crimes and court filings had minimal usefulness in predicting future jail growth.53 The study also found that variations in total arrests (especially violent arrests) and changes in the demographic makeup of the county population were associated with changes in jail demand. As administrators analyze and compare data, their awareness of the limitations of their predictions based on that data will be crucial.

Rate Analysis Methods A rate analysis uses a jurisdiction’s rate of incarceration, rate of admissions, or crime rate as the basis of the forecast. When compared over a span of years, these data provide an indicator against which to measure change. A rate is calculated by dividing the total of a given data element by each 100,000 people in the jurisdiction’s population. The result is referred to as the “rate per 100,000 population.” Jurisdictions with a total population of less than 100,000 divide by each 10,000 people to calculate the rate per 10,000 population. Any model based on a single rate analysis is appealing because of its simplicity. It stimulates discussion of straightforward questions like “How do the incarceration rates of the county compare to other similar counties?” or “What can one infer from changes in rates of admission?” Yet, the problem is that focusing on a single indicator, such as a rate, ignores the complexity of the criminal justice system. The complex and shifting nature of change within a criminal justice system does not lend itself to forecasts based on an analysis of one or two variables. To do so is to ignore the complex nature of the criminal justice system and can offer a false sense of precision. Experience teaches that it is most often those variables outside the control of the jurisdiction that drive jail population growth—in particular, county population growth and changes in criminal justice policy. Thus, the winds of policy change can blow off course any forecast that relies too much on a single factor. For example, any analyst in the 1970s who fore cast juvenile detention bed need into the 1990s based only on the growth rate of the general juvenile population ages 10–17 would have been off course considerably. Although the national population of juveniles remained relatively constant from 1970 to 1998, juvenile court caseloads more than doubled during that same period.54 Although a forecast that relies exclusively on a rate analysis is not recommended, looking at rates of change does have an important place in a comprehensive approach to jail forecasting. A rate analysis provides perspective. Plotting change over time provides a reference point for forecasting that can help identify trends and highlight areas for research.

Jail Capacity Planning Guide: A Systems Approach

167

Exhibit 5–3. Admissions Rate in a Sample County Period 10-year average 5-year average 2005 2006

Admissions Rate (inmates per 100,000 population) 3,200 3,700 3,900 3,600

Suppose, for example, that a county alarmed by a steep upward trend in jail admission begins to question whether reliance on the jail has increased. Examining admission rates can help the county determine whether the increase is related to population growth. An example from the sample county demonstrates this. Exhibit 5–3 reveals that although total admissions in the sample county went up steeply over the past 8 to 10 years (from around 12,000 inmates to more than 18,000 inmates; see exhibit 5–2), the admissions rate, measured at different intervals, changed less significantly. The conclusion is that the increase in jail admissions has more to do with county population growth than changes in booking policy.

INCARCERATION RATES Historical and projected incarceration rates are useful as broad indicators of local, state, regional, and national trends, but they should not be incorporated into any jail scenario. Incarceration rates are influenced by policies and enforcement strategies out side the easy control of jails, so to plan on the basis of incarceration rates would be to consign jail planning to the vagaries of political winds. Instead, jail planning should acknowledge incarceration rates but then focus on those factors over which the criminal justice system has more direct control.

Ratio Model In the context of corrections, the ratio model (sometimes called a stock/flow analysis) looks at the relationship between the number of individuals who move in and out of a jail and the length of time they spend in custody. This model uses a ratio to measure growth in jail admissions as compared with releases to forecast future ADP, which the model posits as the central mea sure of past and present demand on the jail. The ratio model suffers from the same weakness as the rate model: oversimplification. In a ratio analysis, the more admissions increase (or releases dwindle), the more likely it is that the arrest/release ratio will be greater than 1. Conversely, the more admissions decrease (or releases increase) the more likely it is that the arrest/release ratio will be less than 1. Admission/release ratios can be misleading. If a jail has been at capacity for some time, there will be no fluctuation in ratios to track. This lack of variability, however, might not reflect actual demand. Full jails often lead to a situation where admissions decrease in an artificial manner or releases are increased to keep the jail population at or near the upper space limit. The ratio model has other weaknesses. Jails frequently do not keep accurate admissions and release records for more than 4 or 5 years. Attempting to forecast population levels for 15

168

David M. Bennett and Donna Lattin

or 20 years into the future based on only 4 or 5 years of data may not make sense. Furthermore, because the ratio model uses the same formula used for computing compound interest, the results can be wildly unrealistic once arrest/release ratios go above averages of 1.

Jail Exit Analysis Sometimes a jail exit analysis is used to calculate lengths of stay and examine other inmate-processing variables. The problem with this method is that it is based on a single point in time and so may not be representative of jail usage. Additionally, an exit analysis may underestimate lengths of stay. By design, this method may capture a disproportionate number of individuals who had short stays in jail. However, apart from forecasting, a jail exit analysis does have value in providing another perspective on jail bed usage, especially when coupled with classification information. It is useful for understanding the dynamics of the jail population in terms of jail bed days used, for revealing the relative impact of inmates in different classification categories, and for exploring the release potential of groups the jurisdiction con siders to be low risk. Time-Series Model The recommended approach to jail forecasting, the time-series model, tracks several jail population variables. A study comparing forecasting models applied to prison populations found that the time-series model was superior to the others, including a regression model, on measures of predictive accuracy.55 The time-series model assumes that, similar to predicting recidivism, past behavior is the most reliable predictor of future behavior. The model looks at data on past jail use, measured at set intervals, to plot broad trends that are used to map future demand. With a historical/time-series model, jurisdictions can produce a range of alternative capacity scenarios by altering assumptions about the rate of change of discrete variables (ADM and ALOS). Graphing these results enables them to plan with in the curve represented by trend lines that are an objective extension of the past. At the same time, data collected from the broader system analysis can be used to raise “what if” questions about how potential shifts in policy or practice might affect the jail. The degree to which a jurisdiction can identify system actions that hold potential for mitigating jail demand and commit to making them happen will dictate whether to select the low end or the high end of the jail capacity forecast. Time-series models make use of different types of analysis that vary in complexity, from moving averages and weighted moving averages (exponential smoothing) to simple mathematical and basic statistical models. This guide uses the latter approach to track trends in major indicators (county population, ADM, and ALOS) and to chart both the direction and the rate of change to forecast future conditions.

Putting It All Together: Forecast Scenarios Forecasts test different capacity assumptions by modeling changes in the principal factors that drive jail populations. Forecasting calls on county planners and key decisionmakers to speculate about the robustness of observed trends (i.e., the likelihood that they will continue)

Jail Capacity Planning Guide: A Systems Approach

169

and to gauge their confidence in initiating measures to effect change. Forecast scenarios are built on historical jail trend data and different classification and peaking factors. Each scenario represents a different future based on changes in underlying assumptions. Forecast scenarios test different capacity assumptions by modeling changes in the principal factors that drive jail populations. The goal is to develop a range of scenarios that represent an upper and lower limit of expected demand at a point set years in the future. In selecting a forecast model for a general audience, such as criminal justice practitioners, jurisdictions should choose one that is simple for the task but sufficient to the need. The best model is one that maximizes accuracy and minimizes bias. Altering different variables modifies forecast out comes. Exhibit 5–4 shows how changing the sample county’s admissions rates affects bookings. Forecast scenarios factor in not only different admission rates, but also assumptions about ALOS. The sample county examined two different average lengths of stay: 15 days and 17 days. Exhibit 5–5 shows two forecasts for the year 2030. For purposes of comparison, both examples assume an admissions rate of 4,000 per 100,000 population but adjust the ALOS. The ALOS select ed for constructing scenarios is based on the local trend analysis of changes in ALOS. Jurisdictions can construct scenarios to reflect either the status quo or changes in different variables. The cumulative effect of the changing variables on the estimated jail bed need in exhibit 5–5 is evident. Using isolated variables based on historical data to map different future scenarios allows jurisdictions the flexibility of testing the effect of different assumptions on future capacity needs and discussing the underlying factors that drive those assumptions. Based on the trends in jail and population data, jurisdictions can develop jail forecast scenarios that represent a range of upper and lower limits of expected demand. Jurisdictions may base each one on different assumptions about jail use indicators, rates of population growth, and peaking and classification factors. To complement this, a jurisdiction can also model various policy choices to consider how specific changes in policy might mitigate the anticipated demands on the jail. The caveat is that because jails reside within complex and ever-changing systems, a change modeled in one policy may quickly be neutralized by another change unanticipated by the model. The best system assessment and the most accurate forecast cannot account for all the factors involved in determining bedspace demand. Therefore, jurisdictions are best served by examining a range of forecasts based on different assumptions when selecting a scenario on which to base jail planning. However, the extent to which the county commits to making changes in the criminal justice system and implementing strategies to manage the jail population will dictate which end of the scenario range it can comfortably select. Exhibit 5–4. Three Admissions/Bookings Scenarios for a Sample County for 2030 Projected Admissions Rate (inmates per 100,000 population) 4,000 4,500 5,000

Bookings 31,256 35,467 39,408

170

David M. Bennett and Donna Lattin

Exhibit 5–5. Two Average Length of Stay Scenarios for a Sample County for 2030 Scenario A: ALOS = 15 days Forecasted need: 1,296 beds + Peaking factor: 1,425 beds + Classification factor: 1,453 beds

Scenario B: ALOS = 17 days Forecasted need: 1,468 beds + Peaking factor: 1,615 beds + Classification factor: 1,643

Note: Both scenarios are based on an admissions rate of 4,000 per 100,000 population.

Forecasting encourages a review of assumptions about policies and practices and can be a catalyst for change. It should not be a once-a-decade exercise. Ongoing review of forecast assumptions against measured developments and trends allows a jurisdiction to continuously fine-tune its planning and refine its population management techniques.

SECTION 6. THE JAIL CAPACITY FORECAST: A COUNTY EXAMPLE Jail capacity forecasts depend on the availability and quality of local data. The forecasts contained in this guide are no exception. For the sample county, much information was available for the previous 21 years. Jail admissions, average length of stay (ALOS), and average daily population (ADP) data were available from 1986 to 2006. Corresponding data for the community corrections center (CCC) were available from 1999 to 2006. The county annualized its jail data for 2006 on the basis of the first 9 months, and it annualized its 2006 CCC data on the basis of the first 7 months.56 Attempts to obtain older data proved impossible—the records simply did not exist or were not reliable. The state’s Office of Economic Analysis pro vided an estimate of the forecast county population. As useful as these numbers may be in constructing a picture of what is to come, they will not aid the county unless it reaches a consensus about criminal justice system policy for the next 20 years. Analyses of the data culminate in several scenarios, each of which suggests an alternative future in terms of jail bed demand. No one policy scenario is the “right” scenario. In the end, it is up to county decisionmakers to select the view of the future that best represents what they believe to be the most likely direction, based on all the information at hand, and then plan for jail bedspace on that basis. The county collected historical data for both the jail and the CCC. This section presents the jail data; CCC data are presented in appendix D. As the county considered the forecast for future jail space, it combined its data from both facilities for the years 1999 through the present. In the end, local decisionmakers received several scenarios of total corrections space to help them determine the percentage of beds to be allocated to the jail and the CCC.

Jail Data Admissions Plotting admissions to the jail provides a visual display of the increasing demand on the local facility. Exhibit 6–1 shows that total bookings into the county jail increased from 7,268 in 1986 to 18,388 in 2006, an increase of 153 percent. In contrast, the admissions rate (the

Jail Capacity Planning Guide: A Systems Approach

171

number of inmates per 100,000 population) shows a lower, 40-percent increase over the same period (see exhibit 6–2), revealing the contribution of county population growth to admissions. In fact, the rate analysis shows a fairly constant admissions rate for at least the preceding 8 years (see “Overall Admissions Rate” below).

Average Length of Stay Exhibit 6–3 shows ALOS from 1986 to 2006. ALOS in 1986 was 6.9 days. In 2006 ALOS was 11 days, a 69-percent increase since 1986. Exhibit 6–3 shows the pressure on the jail from 1994 to 1997, when ALOS dropped due to severe increases in jail admissions. Once new jail beds became available in 1998, ALOS rose and remained fairly stable for the next 4 years. Average Daily Population Exhibit 6–4 presents ADP for the county jail from 1986 to 2006. ADP was 137 inmates in 1986. It rose modestly for 2 years and then remained constant until new beds became avail able in 1998. ADP in 2006 was 556 inmates, a 306-percent increase since 1986. With ADP at 556 inmates, there is no question that the jail, which has a rated capacity of 572 inmates, is operating above capacity, as there are not enough beds to properly classify or house inmates. Rate Analysis Overall Admissions Rate Exhibit 6–2 shows the rate of admissions to the county jail per 100,000 county population from 1986 to 2006. Because virtually all individuals admitted to the CCC are first booked into the jail, exhibit 6–2 also represents the overall admissions rate for the county. In 1986, the jail’s admissions rate was 2,603 inmates per 100,000 county population. By 2006, the rate had risen to 3,637 inmates per 100,000 population, a 40-percent increase.

Note: The number of admissions increased 153 percent between 1986 and 2006. Exhibit 6–1. Annual Jail Admissions in the Sample County: 1986–2006.

172

David M. Bennett and Donna Lattin

Note: The admissions rate increased 40 percent between 1986 and 2006. Exhibit 6–2. Admissions Rate in the Sample County: 1986–2006.

Note: The average length of stay increased 69 percent between 1986 and 2006. Exhibit 6–3. Average Length of Stay in the Sample County Jail: 1986–2006.

Note: The average daily population increased 306 percent between 1986 and 2006. Exhibit 6–4. Average Daily Population in the Sample County Jail: 1986–2006.

Jail Capacity Planning Guide: A Systems Approach

173

Crime Rate Exhibit 6–5 compares crime rates (the number of crimes committed per 100,000 county population) and admissions rates for the period 1991 to 2003. Although there was some relationship between the crime rate and the admissions rate in 1991–92 and 2001–03, exhibit 6–5 shows that these two rates do not rise and fall together consistently. Incarceration Rate Exhibit 6–6 shows the incarceration rate (the number of inmates per 100,000 county population) from 1986 to 2006. For the years 1999 to 2006, the number shown is the combined rate for the jail and the CCC. The incarceration rate rose from 49 inmates per 100,000 population in 1996 to 147 inmates per each 100,000 population in 2006, a 200-percent increase. The dramatic increase in the incarceration rate in 1998 corresponds to the county’s new jail bed capacity. (This steep upward slope in 1998 is also evident in exhibits 6–3 and 6–4.)

Exhibit 6–5. Comparison of Sample County’s Admissions Rate and Crime Rate: 1991–2003.

Note: The incarceration rate increased 200 percent between 1986 and 2006. Data for 1999 through 2006 include both the jail and community corrections center populations. Exhibit 6–6. Incarceration Rate in the Sample County: 1986–2006.

174

David M. Bennett and Donna Lattin

Note: The county population increased 81 percent between 1986 and 2006. Exhibit 6–7. Population of the Sample County: 1986–2006.

County Population Actual County Population: 1986–2006 Exhibit 6–7 shows the actual county population for each year from 1986 to 2006. In 1986, more than 279,000 individuals resided in the county. Since then, the population has risen steadily, and the county estimates that 505,528 individuals lived in the county in 2006, an 81-percent increase across the period. Forecasted Increase in County Population: 2006–2030 Exhibit 6–8 shows the forecasted increase in the county population from 2006 to 2030 as provided by the state’s Office of Economic Analysis. The county population in 2006 was 505,528 residents, and the county expects the population to grow to 788,162 by 2030, a 56percent increase. When data are available, planners may choose to categorize future population trends by age and gender cohorts, remaining mindful that predicted changes may not occur and that the assumptions attached to those cohorts may not hold.

Jail Capacity Forecasts The primary factors driving jail population in the sample county have been steady increases over the years in admissions and ALOS. Admissions rates can be examined for different time intervals: • • • • • •

1986–2006 (past 20 years): 3,300. 1996–2006 (past 10 years): 3,700. 2001–2006 (past 5 years): 3,700. 2004: 3,600. 2005: 3,900. 2006: 3,600.

Jail Capacity Planning Guide: A Systems Approach

175

Based on rates of growth in admissions over time, the sample county used three different admissions rates for their forecasts: 4,000 inmates per 100,000 population, 4,500 inmates per 100,000 population, and 5,000 inmates per 100,000 population. ALOS can be viewed over different time intervals: • • • • • • •

1986–2006 (past 20 years): 10.4 days. 1996–2006 (past 10 years): 12.8 days. 1998–2006 (past 8 years): 14 days. 2001–2006 (past 5 years): 14.6 days. 2004: 15.7 days. 2005: 14.3 days. 2006: 14.7 days.

The past 8 years are particularly relevant because that period is the only one for which the county has data that represent the combined ALOS of the jail and the CCC. The county used two estimates of ALOS to forecast jail bed need for the year 2030: 15 days and 17 days.

Adjustments: Peaking and Classification Factors Having an expected ADP for each of the jail forecast scenarios does not mean the county should have that number of jail beds available. Because these are daily averages, the county’s plans should include allowances for the days in a given year when the jail population surges above the average because of normal fluctuations in admissions and releases. This situation is similar to a storm drain system. A storm drain sits empty most of the year; how ever, it needs to be large enough to handle peak runoff from a summer thundershower or melting snow from the mountains. Jail populations are like the water filling the drain. During peak periods—traditionally weekends, the end of a month, and in summer— jail populations climb. A jail, like a storm drain, needs to be large enough to handle its peak periods.

Note: The forecasted increase in the county population between 2006 and 2030 was 56 percent. Exhibit 6–8. Forecasted Increase in the Population of the Sample County: 2006–2030.

176

David M. Bennett and Donna Lattin

Exhibit 6–9 shows the peaking factor for the sample county. The jail first identified its three highest population days each month for 2003, 2004, and 2005. The jail then determined each month’s average peak population and compared it with the annual ADP to develop the peaking factor. The 3-year average was 2.5 percent, which was rounded up to 10 percent for the jail forecasts. Ten percent was selected to be conservative, since the forecasts are made for a time interval greatly exceeding that for which the peaking factor was developed. The county also used a second factor, classification, to allow for the jail’s daily need to have a few open beds available for new inmates within each classification category. In a jail of the size under study in the sample county, a reasonable classification adjustment factor would be seven beds for each of the four primary classification categories—high/maximum security, close custody, medium custody, and minimum custody. (Note: Counties will have to determine what is reasonable or appropriate for their own facilities’ management strategies.) For the sample county that means increasing its estimate of the number of beds needed in each year of the planning cycle by 28 beds per year.

Jail Forecast Scenarios for 2030 Varying the assumptions for the principal factors that contribute to jail demand is key to developing a jail forecast scenario. These assumptions work to set the upper and lower parameters for future capacity needs. Exhibit 6–10 presents jail forecast scenarios for the sample county in the year 2030 based on two estimates of ALOS (15 days and 17 days) and on three estimated admission rates (4,000 inmates per 100,000 population, 4,500 inmates per 100,000 population, and 5,000 inmates per 100,000 population). For each estimated ALOS and admission rate, the exhibit shows the ADP, the number of beds necessary to handle peak periods, and the number of beds necessary for classification purposes. That is, each jail forecast scenario in exhibit 6–10 is formed on the basis of slightly different assumptions about what the county is likely to experience in the future with regard to jail capacity demands. The scenarios represent different possibilities—a relatively stable rate of growth, a moderate rate of growth, and a more aggressive rate of growth—for the years between 2006 and 2030.

Exhibit 6–9. Peaking Factor for the Sample County.

177

Jail Capacity Planning Guide: A Systems Approach

Exhibit 6–10. Sample County Jail Capacity Forecast for Year 2030, by Average Length of Stay and Admissions Rate ALOS and Admissions Rate 4,000 4,500 5,000 ALOS = 17 days 4,000 4,500 5,000

Average Daily Population 1,296 1,458 1,620 1,468 1,652 1,835

Total Beds Necessary For Peak For Populations* Classification† 1,426 1,454 1,604 1,632 1,782 1,810 1,615 1,817 2,019

1,643 1,845 2,047

Incarceration Rate‡ 164 185 206 186 210 233

* Calculated as ADP x 1.1 (for this peaking factor example). † Calculated as (ADP x 1.1) + 28. ‡ Calculated as (ADP/Projected County Population in 2030) x 100,000. Notes: Admission and incarceration rates calculated as number of inmates per 100,000 county population. Estimated county population in 2030 is 788,162. Incarceration rates provided for purposes of comparison only; jail forecast scenarios are not based on suggested incarceration rates.

For purposes of comparison only, exhibit 6–10 also presents incarceration rates for each combination of ALOS and admission rate. (Note that jail forecast scenarios are not developed on the basis of incarceration rates.) The exhibit shows how the incarceration rate changes as the ALOS and admission rate change: the larger the jail population, the higher the incarceration rate. Before selecting one scenario over another, a county must anticipate changes within the criminal justice system that might increase demand and then discuss the extent to which the county can implement system changes to mitigate future demand. An effective approach is to start with a midrange projection (4,500 in exhibit 6–10) and then discuss the factors of the criminal justice system that could affect that projection positively or negatively.

CRIMINAL JUSTICE SYSTEM FACTORS THAT CAN AFFECT A MIDRANGE PROJECTION Factors that might increase need for more jail capacity: • • •

Potential for changes in state prison admission policies. Discontinuation of booking fee requirements. Acceleration of population growth.

Changes that might mitigate need for more jail capacity: • • •

Implementation of full-service pretrial program. Development of early case resolution program. Adoption of risk assessment to guide treatment allocation.

178

David M. Bennett and Donna Lattin

Both underestimating and overestimating capacity needs come with their own set of implications. Underestimating future need for beds can result in crowded and sometimes unsafe conditions. Overestimating the need can result in poorly managed public dollars and could even widen the net (i.e., bringing into the criminal justice system individuals who previously would never have entered), thus leading to longer sentences for offenders who might have been diverted had fewer beds been available. In the end, counties should make all jail forecast decisions in conjunction with discussions that ask critical questions about the assumptions guiding both policy and practices and that explore the extent to which a county can put in place the strategies needed to manage change. Forecast scenarios provide a framework within which to have these discussions and to consider a plan for the future of a jurisdiction’s criminal justice system.

Bed Allocation Custody space needs are not just a function of population trends and crime rates. System policies that dictate jail bed use also determine custody space needs. As such, jurisdictions planning how to best allocate beds between a jail and an alternative facility such as a CCC should conduct their planning on the basis of both population and policy considerations. The extent to which local policies allow particular inmates (as determined by classification, charge, and other factors) to serve all or part of their sentences in an alternative facility sets the parameters for this discussion. Jurisdictions can perform additional analyses to help guide their decisions about the distribution of custody beds among the jail, the CCC, and other nonsecure correctional options that are available today in the graduated array of sanctions and services. Conducting a risk-and-needs study is one approach to planning the allocation of beds between a jail and a CCC. A risk-and-needs study uses a validated tool (i.e., data collection instrument) for analyzing risk (e.g., risk of recidivism) to examine a sample of inmates. The results should help frame a discussion about how to man age a jail population by managing risk. A risk and-needs analysis can help a jurisdiction explore how it might manage inmates across a custody continuum to improve offender outcomes. The analysis can also help a jurisdiction examine how risk-based policy decisions translate into cost savings. Consider the following: • • • •

What would happen if a jail moved all inmates who had served half their sentence to a CCC? What would happen if a jail exited all low-risk inmates to sheriff work crews and day reporting? What would happen if a jail exited all medium-risk inmates to a CCC? What would happen if a jail exited all high-risk inmates to a CCC that offered substance abuse treatment?

Several risk tools are available for such an analysis. However, regardless of the instrument, a jurisdiction will want to ensure that it is well tested and valid. Once a jurisdiction settles on a particular risk tool, it will then select a random sample of sentenced

Jail Capacity Planning Guide: A Systems Approach

179

inmates. The sample should be of adequate size to ensure significant results. At a minimum, it should include sentenced misdemeanants, sentenced felons, and felony technical violators. The risk tool generates a risk score for each inmate in the sample. It also collects demographic information (age, gender, etc.) and information about sentence length, jail classification, and individual needs (i.e., risk factors). Substance abuse, mental health, vocation, and education are the risk factors most commonly included. Collecting this information involves a review of each inmate’s official record and a short interview with the inmate. Some risk tools collect data on an extensive array of risk factors. However, a shorter list is sufficient for this study. The risk factors selected depend on the tool a jurisdiction uses and the jurisdiction’s interest in assessing particular areas of risk. With the information obtained from a risk-and-needs study, a jurisdiction can test different scenarios for allocating beds between a jail and a CCC. Findings from one county risk-and-needs study include the following: • • •

Eighteen percent of jail inmates were at low risk for recidivism. Forty-eight percent of the inmates were at moderate risk for recidivism. Thirty-four percent of the inmates were at high risk for recidivism.

The analysis also revealed the factors that correlate with high risk. Compared with the low-risk group, high-risk offenders in the county: • • • •

Had four times the number of prior arrests. Were less than half as likely to have a job. Were three times as likely to have used drugs at the time of their current offense. Had a greater effect on the jail in terms of bed days used.

In discussing possible facility housing scenarios based on risk, the objective is not to exclude the high-risk offender from access to the CCC. Instead, the purpose of knowing the risk level is to improve inmate management along a custody continuum. For example, the Washington County, OR, CCC excludes from its facility only inmates who either have less than 2 weeks remaining on their sentence or have the highest jail classification level (once reclassified, these inmates may be reconsidered for CCC eligibility). All other sentenced inmates are eligible to move from jail to the CCC, with exclusions made on a case-by-case basis. A focus on risk and needs, along with local policies regarding the discretion for jail managers and corrections directors to move offenders along a custody-to-community continuum, can help guide a jurisdiction’s allocation of resources. Ultimately, the planning of resource allocation should be guided by a philosophy that views jail as the alternative—an approach that views the jail as the option of last resort and that plans to use the jails as but one option along a broad continuum. Jail scenarios are just a starting point, and projections are, at best, estimates of what is likely to occur in coming years. Should a jurisdiction’s decisionmakers wish to alter any of their scenarios, they can do so by adjusting the key indices of jail use: county population trends, admission rates, expected ALOS, and the peaking and classification factors. Adjusting these indices will yield different estimates of the required number of jail beds. The process of

180

David M. Bennett and Donna Lattin

estimating future space needs should be an active exercise, one that is updated as conditions change. Creating scenarios is not a one-time exercise, and neither is the implementation of strategies to manage growth. The average time from planning a jail to opening a new facility can be 4 years or even longer. Over this period, jurisdictions should analyze and implement their recommendations in the criminal justice system. Doing so allows for the kind of continuous fine-tuning needed to manage the existing jail population while constructing a new jail and ensures that the new facility is not full on the day it opens. If the necessary changes that the county has recommended do not occur, then more jail beds than those predicted might be necessary. Left uncontrolled, jail populations continue to grow, filling and overfilling whatever facilities counties construct in response to such growth and leaving no alternatives for managing a jail population other than to expand facilities every few years. An approach that emphasizes active management, on the other hand, may make it possible to prolong the sufficiency of newly constructed jail space for a longer period, giving a county time to explore and try out the many viable alternatives to construction that have become available in recent years. In any case, planned jail bed demand is bound to create “sticker shock” when the cost of building a new or expanded facility is finally calculated. However, realizing the cost can itself be the impetus for exploring changes in the criminal justice system that support reducing admissions or ALOS.

SECTION 7. PLANNING FOR ONE EMPTY BED A good plan executed right now is far better than a perfect plan executed next week. —General George S. Patton57

The key to the long-term management of a jail and other corrections resources is the implementation of a system master plan: a set of policy and program strategies that will enable a jurisdiction not only to react to change but also to influence and shape the course of that change.

System Master Plan The development of a system master plan should provide the foundation for any jail capacity forecast scenario. A system master plan is a strategic plan for the future: It outlines principles and practices designed to make the most efficient use of existing resources and manage change. Once implemented, the strategies developed in the system master plan can help manage a jail toward the goal of “one empty bed” and forestall the need for more jail beds. A system master plan should, at the mini mum, address the following areas: • • •

Prebooking options. Pretrial release services. Classification and use of objective risk assessment.

Jail Capacity Planning Guide: A Systems Approach • • • • • • •

181

Adjudication policies and practices. Diversion options. Sentencing alternatives. Program adherence to evidence-based practices. Sanction policies and programs. Jail reentry and discharge planning. Data availability and integration.

In examining these areas, a jurisdiction should refer to its case-processing and jail snapshot data, its policies and procedures, qualitative reviews of its programs, and any other data it believes would inform its decisionmaking process. The sample county’s system master plan contains more than 80 recommendations that address a broad spectrum of system policies and practices. In some cases, the county took action before the plan was complete. One presiding judge, for example, made the commitment to assign a single judge to handle front-end court proceedings as part of a broader initiative to streamline processing and adopt early case resolution (ECR) practices. In the broadest terms the sample county could group its recommendations by the two factors that most directly influence the jail: admissions and average length of stay (ALOS). The sample county’s recommendations to reduce admissions were as follows: • • • • •

Establish a comprehensive pretrial program. Fund local detoxification services. Increase the use of nonjail sanctions. Develop diversion options for the mentally ill. Reduce the use of jail for probation violations through a structured sanction policy.

The sample county made the following recommendations for reducing ALOS: • • •

Implement an ECR program. Reduce the time between citation for a probation violation and the hearing. Expedite the movement of inmates from the jail to the community corrections center (CCC).

These are but a sample of the measures jurisdictions might take to improve system efficiency. In the end, the degree to which jurisdictions implement these changes will dictate whether a jurisdiction selects the high end or low end of its projected need in a jail forecast scenario.

Selecting a Jail Forecast Scenario Recommendations proposed in a system master plan should represent jail management strategies that, if implemented, will go a long way toward mitigating jail crowding and optimizing outcomes. Jurisdictions can fashion scenarios to test the influence of changes in policies and practices. Policy scenarios allow a jurisdiction to use its jail capacity forecast not

182

David M. Bennett and Donna Lattin

only to plan for the future but to shape it as well. These scenarios can be a useful tool for speculating about the ramifications of different policy choices. One example from a study of the Los Angeles County Jail shows the power different policy choices wield. This study simulated how different eligibility policies for jail diversion would affect the jail population. The county found that when the population eligible for diversion from jail was limited to inmates who had been incarcerated for a nonviolent crime and who had no previous jail incarceration, approximately 11 percent of the sample in custody would qualify for diversion. However, when the criteria for diversion eligibil ity were expanded to include inmates who had a previous jail incarceration but no prison incarceration, the population of inmates eligible for diversion rose to more than 53 percent (although a quarter of these inmates were felony drug offenders).58 The policy choices of the criminal justice system and achievement of efficiencies and improvements within the system can reduce demand on a jail. However, because jails are complex and dynamic systems, the effect of a single system modification is difficult to quantify. Moreover, although making changes can help a jurisdiction moderate its jail population growth, shifts in policies and practices over which the jurisdiction has no control can undermine these improvements. For example, the sudden employment of more police officers or a new state “zero tolerance” policy for probation violators can reverse gains in holding down jail population growth. Conversely, one California county found that a new state pol icy rendered forecast jail bed demand unreliable: Proposition 36, which diverts nonviolent drug users from jail to treatment, led to less growth in the jail population than anticipated. (Overall, however, jail admissions in California are at their highest, even with Proposition 36.) Jail planning also cannot fully anticipate the interactive factors of a complex system. For example, a successful strategy for reducing the number of offenders on absconded status can swell caseload size and negatively affect a jail. Likewise, a jurisdiction might expect an early case resolution program to reduce the case dis missal or decline rate. No model is yet sensitive enough to examine the interactive effects of all changes to the criminal justice system. To alter a plan on the basis of changes anticipated for one particular target group ignores the complexity of the criminal justice system. Jurisdictions that have attempted to model changes in jail usage based on a single policy change are often surprised when jail demand quickly exceeds or falls short of modeled changes. This is because another system change that has an opposite effect can quickly neutralize the initial policy change. For this reason, this guide focuses on population management strategies, which are flexible tools for managing change. Jail planning is not a precise science. In fact, planning methodologies based on a high level of detail might only serve to give a false sense of precision. One county that developed a forecast based on modeling jail admissions for more than 20 types of criminal offenses was quickly thrown off target by larger changes in system policy, proving that sometimes it is better to be “generally right” than “precisely wrong.” For these reasons, jurisdictions are advised to conduct jail forecasts by using historical patterns of demand as their starting point. Historical changes in jail admissions and ALOS set the parameters. Anticipated rates of future growth in county population set the pace. Finally, the jurisdiction’s level of confidence in implementing system measures to manage the jail population and improve system efficiency helps determine which end of the range of

Jail Capacity Planning Guide: A Systems Approach

183

projected demand for jail beds it should choose. Before selecting one jail forecast scenario over another, a jurisdiction must decide the extent to which it can effect changes to the criminal justice system and mitigate future demand on the jail.

CRIMINAL JUSTICE SYSTEM FACTORS THAT AFFECT THE CHOICE OF A JAIL FORECAST SCENARIO Factors Associated with Increased Jail Capacity Need • • • • •

Pent-up demand (as measured by indicators such as housing inmates in other jails, truncating sentences, and increasing citation rates). Degree of uncertainty about state policy changes. Elimination of booking fee requirement. Anticipated increase in the number of police officers. Anticipated increase in county population.

Factors Associated with Decreased Jail Capacity Need • • • • • • •

ECR programs. Pretrial services. Community corrections centers. Jail stepdown options. Elimination of booking fee requirement. Quality diversion and sentencing options. Risk-based programs.

Focusing on population management strategies encourages a jurisdiction to adopt measures that allow flexible and continuous jail management. In contrast, approaches that attempt to model single policies are vulnerable to changes in those policies. This can limit the usefulness of modeling efforts. This is not to say there is no value in modeling the effect of policies or practices. This kind of planning can help demonstrate the relative influence of different changes and can serve as a catalyst for change. The sample county considered the jail mitigation scenarios listed in exhibit 7–1. Exhibit 7–1. Outcomes of Various Jail Scenarios in Mitigating Average Daily Population Scenario Reduce transfer time from jail to the CCC by 27 days Reduce pretrial failure rate by 50 percent Reduce case-processing time by 3 days Divert low-risk inmates with mental illnesses from jail

Estimated Reduction in Average Daily Population 109 29 43 33

In the end, the main focus of jail planning should be on the adoption of population management strategies that can reduce pressures on the jail by managing change. Three

184

David M. Bennett and Donna Lattin

examples of these are pretrial programs, which manage the front end of the jail, thereby structuring release decisionmaking; early case resolution programs, which address the movement of cases through the jail; and jail stepdown options, which move certain inmate populations out to work crews, day reporting centers, or CCCs, according to individual levels of risk and need. The extent to which a jurisdiction is committed to making changes to the criminal justice system will dictate which end of the forecast range of options it can select. A jurisdiction that outlines a bold plan that addresses major drivers of jail bed demand can more confidently plan on the low end of projected future capacity need. The sample county concluded that implementation of key recommendations would allow it to base its plans on the low end of the continuum. Each jurisdiction confronted with the need for long-term jail capacity planning can use this opportunity to influence change in the criminal justice system toward improved jail management and better offender outcomes.

Presenting the Forecast Results Once the forecast is complete, planning carefully how to present the results can pay dividends. If the jurisdiction has a criminal justice coordinating committee, the committee should be aware of the details of the jurisdiction’s data collection efforts. Optimally, the committee also will have had a voice in shaping the direction taken during the data collection and in providing input into tracking the analysis. Including key elected and appointed officials in the final review of the data is particularly important. After the planning team has analyzed the data, a representative of the team should visit each official individually to discuss the findings specific to that official’s agency. This is not only a matter of courtesy but is also essential to ensuring that the data are correct and that the jurisdiction has interpreted them appropriately. Jurisdictions may want to consider scheduling a full-day symposium to present system data to a full contingent of criminal justice system representatives. Not only is this an opportunity to reveal data about jail usage and system case-processing dynamics, but it can also be an interactive forum. Presenters can ask the audi ence to help derive the assumptions that drive the selection of different jail forecast scenarios. The symposium may be followed up with additional presentations to criminal justice-related commissions. If the members of the jurisdiction’s elected governing body (e.g., the County Board of Commissioners) do not all attend the symposium, the jurisdiction may wish to offer them a separate presentation. A jurisdiction may also present its findings to other interested parties and community groups. Doing so is important for laying the ground work for citizen understanding of the issue. Jurisdictions should incorporate the feedback from all these meetings into a final report: the system master plan. This report will include the jail and case-processing data, jail forecast scenarios, analysis, and final recommendations. Once the system master plan is issued, jurisdiction officials will want to decide how to proceed. They may form subcommittees to begin addressing the plan’s recommendations. If the jurisdiction decides to proceed with planning a jail facility, it will make arrangements to bring in experts in jail operations, staffing, facility design, site selection, and cost analysis. It will also begin a conversation with voters.59

Jail Capacity Planning Guide: A Systems Approach

185

Making Adjustments Facility planning takes place on shifting ground. Jail planning (based on 10to 20-year forecasts) drives building plans that proceed as jail demand continues to fluctuate and, in some cases, veers off course completely. For this reason, a jurisdiction should reexamine its jail projections as the criminal justice system adopts new policies or passes new laws. In some cases, changes in county population will also require a review. In all cases, the forecast of jail and alternative facility needs should be an ongoing process, just as the review of the programs and policies of the criminal justice system should be a continuous endeavor. Jurisdictions should view the system master plan as a starting point on which to build as time and circumstances dictate. Only through commitment to ongoing review can a jurisdiction get ahead in its planning and consider how to forestall the continued building of facilities to increase jail beds.

Planning for Results Jail usage is driven by the policies of the criminal justice system, judicial decisions, sentencing law, available alternatives, case-processing efficiency, and program effectiveness. Each of these factors can be examined or measured. For this reason, jail capacity planning should be grounded in a systems approach.

APPENDIX A. JAIL SNAPSHOT: DATA VARIABLES The data needed to complete a jail snapshot analysis include a daily census or roster of all inmates held at all facilities included in a study. If a jurisdiction has a work-release facility, it can develop separate snapshots for the jail and the work-release facility. Jurisdictions should consider the following guiding principles when developing a jail snapshot: • •

• •

•

A “snapshot” of the jail population consists of information that describes a jurisdiction’s incarcerated population at one point in time. Most jails take a census that includes all inmates. For this exercise, take a sample at the same hour on the same day of the week once a month for each of the next 6 months. The sample should include all inmates who were in the jail on each sample date. For the purposes of this exercise, if a defendant is charged with more than one offense, record only the most serious charge type and status. Capture time in custody by noting the number of days each inmate has been in custody. For an aggregate time in custody, divide the total number of days by the number of inmates or by inmate status (e.g., pretrial, sentenced). Add any additional demographic variables where the data are available. For example: • Residence. • Employment. • Education level.

186

David M. Bennett and Donna Lattin • •

Serious mental illness: yes/no. Prescribed psychotropic medication: yes/no.

Data to Collect for a Jail Snapshot The variables listed below are only an example of the types of information to collect for a jail snapshot. Each jurisdiction should tailor the data collected to the needs of its own system. 1. Inmate number: Beginning with 1, number each inmate included in the sample. If there are 200 inmates in the census, the data sheet should contain the numbers 1–200. 2. Inmate name or jail identification number. 3. Status: Enter the appropriate letter as given below: P = Pretrial S = Posttrial H = Hold O = Other 4. Charge type: Use the following codes to indicate the most serious charge or conviction for which the jurisdiction is holding the inmate. DV = Domestic violence* PE = Person PR = Property DR = Drug DU = Drunk Driving PO = Public Order TR = Traffic *Domestic violence is a subset of person crimes. 5. Charge class: F = Felony M = Misdemeanor 6. Hold type: If the jail recorded the status as a hold, select the type of hold. If there are other hold types in the jurisdiction, develop a set of codes that fit the jurisdiction’s needs. Then assign each code a number. 1 = Federal hold 2 = State hold 3 = Other jurisdiction 4 = [Enter name of additional hold] 5 = [Enter name of additional hold] 7. Year booked: Enter the last two digits of the year an inmate was booked. 8. Day booked: day/month/year 9. Arresting agency: Develop a code for the jurisdiction by assigning a number to each agency (police, sheriff, or other entity) that arrests and brings defendants to the jail. 0 = Unknown 1 = [Enter name of agency] 2 = [Enter name of agency] 10. Court of jurisdiction. 11. Age: [Enter age in years]

Jail Capacity Planning Guide: A Systems Approach

187

12. Sex: M = Male F = Female 13. Race/ethnicity: Record the appropriate code according to local categories: W = White B = Black H = Hispanic I = American Indian A = Asian O = Other U = Unknown 14. Classification assignment: Close Maximum Medium Minimum

APPENDIX B. CASE-PROCESSING STUDY: SAMPLE VARIABLES The terms and definitions for the inmate sample variables used in the case-processing study will vary with each jurisdiction. Once a jurisdiction identifies the proper terms and confirms the availability of data, it should draft a code manual that describes all terms and abbreviations. Note: The variables listed below are only an example of the types of information a jurisdiction needs to collect. Each jurisdiction should tailor the data collected to the needs of its own system.

Data to Collect for a Case-Processing Study Detention Center 1. Defendant identification number. 2. Date of birth: month/day/year. 3. Place of birth: (state, region, other state, other country). 4. Age at booking. 5. Sex (circle one): male/female. 6. Race/ethnicity: Record the appro priate category according to local classifications. 7. Residence (circle one): • Current jurisdiction • Other jurisdiction within same state • Other state 8. Residence length (circle one): • Less than 1 year • 1–3 years • 3 or more years 9. Homeless (circle one): yes/no

188

David M. Bennett and Donna Lattin 10. 11. 12. 13.

14. 15. 16. 17. 18. 19.

20. 21. 22. 23. 24. 25.

26.

Marital status. Driver’s license state. Employment. Level of education (circle one): • No high school • Some high school • High school graduate • Education beyond high school Gang affiliation. Booking date. Booking time. Release date. Release time. Charge at time of booking (circle one): • Domestic violence • Person offense • Property offense • Drug offense • Public order offense • Traffic offense Charge class (circle one): felony/misdemeanor. Charge degree (circle one): 1, 2, 3. Number of charges. Arresting agency (list major law enforcement agencies). Total bail amount. Pretrial release type (circle one): • Released on own recognizance • Supervised release • Bond • Supervised release plus bond • Forced release • No release Posttrial release type: • Time served • Case dismissed • Released from probation • Released to other agency

Court Data 27. First appearance date. 28. Filing date. 29. Arraignment date. 30. Disposition date. 31. Sentencing date. 32. Number of administrative hearings.

Jail Capacity Planning Guide: A Systems Approach 33. 34. 35. 36.

37.

38. 39. 40. 41. 42.

43. 44. 45.

46.

47. 48. 49. 50.

Prosecutor charge class (capture the charge at the time of filing). Prosecutor charge degree. Prosecutor charge. Attorney type (circle one): • Public defender • Other appointed counsel • Retained • None Disposition type (circle one): • Guilty • Not convicted • Pending Guilty verdict type (pled, found). Disposition charge class. Disposition charge degree. Disposition charge. Reason for nonconviction (circle one): • Prosecutor decline • Dismissed • Not guilty • Pending Bond amount. Release type (circle one): pretrial, posttrial. Posttrial release type (circle one): • No complaint • Dismissed • Time served • Transported • Court order Sentence type (circle one): • Prison • Jail • Community corrections center • Diversion (e.g., drug court, mental health court) • Probation • Fine Sentence length. Failure to appear: yes/no. Rearrest: yes/no. Previous bookings (circle one): • 0 • 1–2 • 3–5 • 6–10 • 11 or more

189

190

David M. Bennett and Donna Lattin

APPENDIX C. SAMPLE DATA ANALYSIS CALCULATIONS Jurisdictions can analyze the data collected in both the jail snapshot and the caseprocessing study in numerous ways. A full report on the data may include hundreds of charts and graphs displaying the data from different analytic perspectives. This will require the services of a good analyst. Specific examples of data calculation discussed in the section on the case-processing study are listed below. These examples provide only a sample of the kinds of analysis jurisdictions can conduct to reveal system functioning.

Data Analysis Calculations for a Case-Processing Study • •

•

• • • •

Pretrial release rate: Divide the number of inmates released prior to case disposition by the number of inmates booked into jail. Pretrial failure-to-appear rate: Calculate the number of inmates re leased before trial who failed to make a court appearance during the period between release and case disposition. Then divide the number of inmates who failed to appear by the number released. Pretrial rearrest rate: Calculate the number of inmates released before trial who were rearrested during the period between release and case disposition. Then divide the number of inmates who were rearrested by the number released. Felony filing rate: Divide the number of felony cases filed by the number of felony cases booked. No-complaint rate: Calculate the number of cases for which the prosecutor filed no formal charges. Then divide the number of these cases by the number booked. File attrition: Subtract the number of no-complaint cases from the number of cases filed. Case-processing times: Calculate the average time between different points in the adjudication process. The average processing time is calculated by taking the total process times between two points and dividing it by the number of cases.

APPENDIX D. SAMPLE COUNTY COMMUNITY CORRECTIONS CENTER DATA The following data show the additional information collected for the community corrections center (CCC) in the sample county. Jurisdictions with a jail and an alternative facility need to track the same forecast variables for both.

Jail Capacity Planning Guide: A Systems Approach

191

Community Corrections Center Overall Admissions Exhibit D–1 shows the number of admissions into the CCC from 1999 to 2006. In 1999, there were 1,813 admissions into the facility. Administrators anticipated 2,050 admissions in 2006, resulting in a 13-percent increase over the period. Admissions by Resident Admission Status Exhibit D–2 shows the number and percentage of admissions into the CCC for 2006 by type of admission. In 2006, there were 1,985 admissions to the CCC. Of these, 9 percent were inmates serving a sanction, 75 percent were sentenced offenders, and the remaining 16 percent were T-lodgers—short-term “transition” lodgers. T-lodgers are offenders who are 1) reentering the community from prison, 2) in a CCC bed as a stabilization measure after release from jail, or 3) serving short-term direct sanctions from a probation officer (this category includes inmates who have been released from the state prison and do not have a residence).

Exhibit D–1. Community Corrections Center Admissions: 1999–2006.

Note: A T-lodger is a short-term “transition” lodger who is reentering the community from prison, occupying a CCC bed as a stabilization measure after release from jail, or serving a short-term direct sanction from a probation officer. Exhibit D–2. Distribution of Community Corrections Center Admissions, by Type of Admission: 2006.

192

David M. Bennett and Donna Lattin

Overall Average Length of Stay Exhibit D–3 shows the average length of stay (ALOS) at the CCC from 1999 to 2006. In 1999, the ALOS was 29.2 days. By 2006, the ALOS had increased to 33.3 days, a 14-percent increase over the period. Average Length of Stay by Type of Admission Exhibit D–4 shows the 2006 ALOS at the CCC categorized by the type of admission. In 2006, the overall ALOS was 33.3 days; for sanctioned inmates, the ALOS was 25.8 days; for sentenced inmates, the ALOS was 38.9 days; and for T-lodgers, the ALOS was 18 days.

Note: Average length of stay increased 14 percent between 1999 and 2006. Exhibit D–3. Average Length of Stay by Type of Admission.

Overall Average Daily Population Exhibit D–5 shows the average daily population (ADP) at the CCC from 1999 to 2006. In 1999, there were an average 145 individuals at the CCC. By 2006, that number had grown to 187, a 29-percent increase over the period. Average Daily Population by Type of Admission Exhibit D–6 shows the ADP of the CCC in 2006 by type of admission. Overall, 6 percent of CCC residents were sanctioned inmates, 85 percent were serving a sentence, and the remaining 9 percent were T-lodgers.

Jail Capacity Planning Guide: A Systems Approach

193

Exhibit D–4. Average Length of Stay at the Community Corrections Center, by Type of Admission: 2006.

Exhibit D–5. Average Daily Population at the Community Corrections Center: 1999–2006.

Note: A T-lodger is a short-term “transition” lodger who is reentering the community from prison, occupying a CCC bed as a stabilization measure after release from jail, or serving a short-term direct sanction from a probation officer. Exhibit D–6. Distribution of Community Corrections Center Admissions, by Type of Admission: 2006.

194

David M. Bennett and Donna Lattin

ABOUT THE AUTHORS Author Biographies David M. Bennett David Bennett has more than 30 years of experience in addressing jail overcrowding issues and criminal justice system reform. As a consultant, he has advised more than 250 jurisdictions in 40 states regarding the development of system-based solutions to jail population management. Before beginning his consulting career, he established and directed the Pretrial Services Department of Salt Lake County Criminal Justice Services in Utah. Mr. Bennett has also worked on several nationwide jail projects. Under the sponsorship of the National Institute of Corrections and the American Justice Institute, he served as lead trainer and helped set the agenda for the federal government’s first jail overcrowding seminar in 1981. He also participated in the development of the Law Enforcement Assistance Administration’s jail management guidelines, which have been recommended to state and local officials since 1978. Mr. Bennett is a coauthor of the first Jail Capacity Planning Workbook published by the U.S. Department of Justice. Mr. Bennett is adept at working with jurisdictions to achieve comprehensive solutions to jail overcrowding. The hallmark of his work is the development of individualized, researchbased jail population management plans for counties. As part of this work, Mr. Bennett has successfully implemented pretrial service programs, early case resolution protocols, meaningful alternatives to incarceration, and improved information systems in jurisdictions throughout the United States. At the forefront of his consulting practice is using innovation and integrity to develop jail population management plans. Donna Lattin Donna Lattin began her career as a community corrections manager and legislative policy analyst. Currently, as a consultant, she specializes in the assessment of local criminal justice systems and jail alternatives. Ms. Lattin develops system master plans based on best practices and evidence-based research, helping jurisdictions develop comprehensive strategies to optimize jail resources and enhance their capacity to effect positive change. Ms. Lattin has worked around the country in the design and refinement of pretrial services, in-custody programs, specialty courts, mental health and treatment services, risk assessments, and supervision and sanction alternatives.

End Notes 1

2

3 4

Ryan S. King, Marc Mauer, and Malcolm C. Young, Incarceration and Crime: A Complex Relationship (Washington, DC: The Sentencing Project, 2005), pp. 3–4, emphasis added; www.sentencingproject.org/ Publications.aspx?IssueID=2. Paige M. Harrison and Allen J. Beck, Prison and Jail Inmates at Midyear 2004 (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, 2005), www.ojp.usdoj.gov/ bjs/abstract/pjim04.htm. Harrison and Beck, 2005. E. Fuller Torrey, Out of the Shadows: Confronting America’s Mental Illness Crisis (New York, John Wiley & Sons, 1997).

Jail Capacity Planning Guide: A Systems Approach 5

195

National Institute of Justice, NIJ Survey of Jail Administrators (Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice, 1995). 6 Brian A. Reaves and Jacob Perez, Pretrial Release of Felony Defendants, 1992 (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, 1994), www.ojp.usdoj.gov/bjs/ abstract/nprp92.htm. 7 Public Safety Coordinating Council, “FTA Task Force Report,” Lane County, OR, November 2004 (August 2005 Update). 8 Joan Petersilia, Susan Turner, and Terry Fain, Profiling Inmates in the Los Angeles County Jails: Risks, Recidivism, and Release Options, final report for U.S. Department of Justice, Office of Justice Programs, National Institute of Justice, 2000. 9 Harrison and Beck, 2005. 10 Pretrial Services Resource Center, “Governor’s Task Force in Michigan Points to Importance of Enhancing Pretrial Services to Address Crowding,” The Pretrial Reporter, April/May 2005. 11 Bureau of Justice Statistics, Census of Jails (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, various years). 12 S. Cohen and J.W. Kohls, Do the Crime, Do the Time? Maybe Not in California (Sacramento, CA: California State Sheriffs’ Association, 2006), www.calsheriffs.org/Documents/do_the_crime,_ do_the_time.pdf, accessed February 16, 2008. 13 Gail Elias, How To Collect and Analyze Data: A Manual for Sheriffs and Jail Administrators, 3d ed. (Washington, DC: U.S. Department of Justice, National Institute of Corrections, 2007), www.nicic.gov/ Library/021826, accessed March 26, 2009. 14 For more detail on this topic, see Robert C. Cushman, Guidelines for Developing a Criminal Justice Coordinating Committee (Washington, DC: U.S. Department of Justice, National Institute of Corrections, 2002), www.nicic.gov/ Library/017232, accessed January 8, 2009. 15 J.M. Taylor, “Pell Grants for Prisoners,” The Nation, January 1993, p. 90. 16 Bureau of Justice Assistance, A Second Look at Alleviating Jail Crowding: A Systems Perspective (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Assistance, 2000) describes similar impacts. 17 John Clark and D. Alan Henry, Pretrial Services Programming at the Start of the 21st Century: A Survey of Pretrial Services Programs (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Assistance, 2003). 18 Clark and Henry, 2003, p. 11. 19 “Maryland Lawsuit Targets Lack of Legal Representation at Bail-Setting Hearings,” The Pretrial Reporter 32(5), October/November 2006. 20 Faye S. Taxman, Pretrial Services Unit: An Evaluation of First Year Experiences (Rockville, MD: Criminal Justice Coordinating Council, 1992). 21 Shelby Meyer and Kim Holloway, “The Fastrack Program,” Federal Probation 57(19):36–41, 1993. 22 Clark and Henry, 2003. 23 National Association of Counties, NACo Achievement Award Program, 2005. 24 Bureau of Justice Assistance, 2000. 25 David Bennett, unpublished data. 26 Jail Utilization System Team, Total Quality Management: Comprehensive Community Based Corrections Program (Monroe County, NY: Jail Utilization System Team, 1994). 27 Bureau of Justice Assistance, 2000. 28 Joan Petersilia, Profiling Inmates in the Los Angeles County Jails: Risks, Recidivism, and Release Options (Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice, 2000). 29 “SACPA Cost Analysis,” University of California, Los Angeles, April 5, 2006. 30 Michael W. Finigan, Shannon M. Carey, and Anton Cox, “Impact of a Mature Drug Court Over 10 Years of Operation: Recidivism and Costs,” unpublished final report, 2007, www.ncjrs.gov/pdffiles1/nij/grants/ 219225.pdf, accessed March 26, 2009. 31 Joseph P. Morrissey, “Medicaid Benefits and Recidivism of Mentally Ill Persons Released from Jail,” unpublished final report for University of North Carolina, Chapel Hill, 2006, www.ncjrs. gov/pdffiles1/nij/ grants/214169.pdf, accessed March 26, 2009. 32 Dale E. McNiel and Renée L. Binder, “Effectiveness of a Mental Health Court in Reducing Criminal Recidivism and Violence,” American Journal of Psychiatry 164:1395–1403, 2007. 33 Bureau of Justice Assistance, 2000, p. 64. 34 The California Treatment Outcome Project, final report, 2002. 35 C.T. Lowenkamp and E.J. Latessa, “Increasing the Effectiveness of Correctional Programming Through the Risk Principle: Identifying Offenders for Residential Placement,” Criminology & Public Policy 4(2):263–90, 2005. 36 Steve Aos, Marna Miller, and Elizabeth Drake, Evidence-Based Public Policy Options to Reduce Future Prison Construction, Criminal Justice Costs, and Crime Rates (Olympia, WA: Washington State Institute for Public Policy, 2006a).

196 37

38 39

40

41 42 43

44 45

46

47

48

49

50

51 52

53

54

55

56 57

58

59

David M. Bennett and Donna Lattin

Steve Aos, Marna Miller, and Elizabeth Drake, Evidence-Based Adult Corrections Programs: What Works and What Does Not (Olympia, WA: Washington State Institute for Public Policy, 2006b). Aos, Miller, and Drake, 2006a. Scott T. Walters, Michael D. Clark, Ray Gingerich, and Melissa L. Meltzer, Motivating Offenders To Change: A Guide for Probation and Parole (Washington, DC: U.S. Department of Justice, National Institute of Corrections, 2007). D.A. Andrews, Craig Dowden, and Paul Gendreau, “Clinically Relevant and Psychologically Informed Approaches to Reduced Re-offending: A Meta-analytic Study of Human Service, Risk, Need, Responsivity and Other Concerns in Justice Contexts,” unpublished man uscript, Carleton University, 1999. Aos, Miller, and Drake, 2006b. Oregon Department of Corrections, “Structured Sanction Study,” September 2002 (paraphrased). Washington State Department of Social and Health Services, Research and Data Analysis Division, Frequent Emergency Room Visits Signal Substance Abuse and Mental Illness, 2004. Washington County, OR, Community Corrections, “Biennium Plan 2005–2007,” unpublished paper. The Washington County CCC per-day rate for the small number of residents involved in in-house residential treatment is $75 per day (John Hartner, Community Corrections Director, Washington County, OR, personal communication, 2007). The higher cost is accounted for by services associated with the extensive treatment program. Richard McCarthy, The Hampden County Day Reporting Center: Three Years’ Success in Supervising Sentenced Individuals in the Community (Hampden County, MA, Sheriff’s Department, 1990). BI Incorporated, “Strengthening the County Criminal Justice System: Implementing a Day Reporting Center Program that Reduces Jail Overcrowding, Gives Judges Effective Sentencing Options, and Makes Probation Stronger,” www.bi.com/pdfs/case_study/BI_CS_franklin.pdf, 2008. Karla Crocker, Davidson County’s Day Reporting Center: An Effective Alternative (Longmont, CO: National Institute of Corrections Information Center, 2000), www.nicic.org/ Library/period173, accessed March 26, 2009. Justice Education Center, Inc., Longitudinal Study: Alternatives to Incarceration Sentencing Evaluation, Year 3 (Hartford, CT: Justice Education Center, Inc., 1996). Burton Stevenson, The MacMillan Book of Proverbs, Maxims and Familiar Phrases (New York and Chicago: MacMillan, 1968). For the sake of brevity, the term “county” refers to all jurisdictions that maintain jails. Leonard A. Marowitz, Why Did the Crime Rate Decrease through 1999? (And Why Might it Decrease or Increase in 2000 and Beyond?): A Literature Review and Critical Analysis (Sacramento, CA: California Department of Justice, Criminal Justice Statistics Center, 2000), www.nicic.gov/Library/016837, accessed March 13, 2009. William D. Bales, Palm Beach County, Jail Population Plan: 2001 to 2010 (Florida Department of Corrections, Bureau of Research and Data Analysis, 2003). Jeffrey Butts and William Adams, Anticipating Space Needs in Juvenile Detention and Correctional Facilities (Washington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention, 2001). Lin Bin-Shan, Doris MacKenzie, and Thomas R. Gulledge Jr., “Using ARIMA Models To Predict Prison Populations,” Journal of Qualitative Criminology 2(3):251–264, 1986. Annualized data were calculated using the following formula: (total for n months/n) x 12. Roy Rosenzweig, “Scarcity or Abundance? Preserving the Past in a Digital Era,” American Historical Review 108(June):754, 2003, n. 47. Joan Petersilia, Profiling Inmates in the Los Angeles County Jails: Risks, Recidivism, and Release Options (Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice, 2000). For a description of how to engage the public in the process of jail planning, see Gail Elias, Building Community Support for New Jail Construction (Washington, DC: U.S. Department of Justice, National Institute of Corrections, 2006), www.nicic.gov/Library/021328, accessed February 24, 2009.

INDEX A abuse, 65, 70, 94, 133, 179 access, vii, 2, 3, 7, 8, 10, 14, 20, 42, 46, 74, 80, 82, 85, 88, 127, 129, 138, 160, 179 accessibility, 4, 15, 64 accountability, 6 accounting, 50, 82, 134, 135, 141, 145, 152 adaptations, 46 adjustment, 16, 102, 158, 163, 164, 176 administrators, vii, 1, 2, 3, 4, 5, 8, 11, 12, 13, 14, 16, 17, 21, 23, 24, 25, 26, 27, 34, 35, 36, 37, 45, 51, 52, 82, 84, 85, 162, 166 ADP, 36, 107, 118, 158, 159, 161, 162, 163, 167, 170, 171, 175, 176, 177, 192 adults, 114, 118, 161 age, 29, 129, 131, 141, 149, 152, 153, 154, 159, 161, 164, 165, 174, 178, 179, 186 agencies, 5, 6, 11, 12, 16, 20, 24, 37, 40, 42, 47, 56, 69, 75, 76, 77, 79, 80, 85, 114, 151, 157, 159, 160, 188 agency decisions, 88 aggregation, 24 alcohol use, 67 American Psychological Association, 109, 110 appraisals, 22 architects, 124 arraignment, 11, 129, 130, 131 arrest(s), viii, 7, 10, 27, 65, 74, 79, 111, 112, 116, 118, 119, 120, 131, 138, 147, 148, 150, 152, 155, 160, 165, 166, 167, 168, 179, 186 assault, 7, 32, 86 assessment, 23, 61, 62, 74, 87, 88, 89, 100, 114, 117, 121, 122, 123, 124, 125, 128, 129, 130, 134, 162, 169, 194 attitudes, 2, 3, 4, 47, 156 audit, 48, 50, 84 authority, 42, 44, 124, 128

automation, 37, 40, 42, 56 avoidance, vii, 1, 2 awareness, 166

B background information, 154 bail, 10, 126, 127, 128, 150, 153, 154, 188 banking, 68 bargaining, 114 barriers, 130 base, 3, 22, 44, 137, 144, 158, 169, 184 behaviors, 23, 54 benefits, 22, 44, 45, 46, 51, 52, 56, 128, 130, 132, 136, 147, 166 BI, 196 bias, 169 biased interpretation, 21 BJS, 119 boils, 82 bonds, 10 boredom, 54 breakdown, 32, 70, 120 budget cuts, 8, 85 Bureau of Justice Assistance, 133, 195 Bureau of Justice Statistics, 118, 119, 140, 194, 195 bureaucracy, 3 business function, 47, 57, 64, 74 business partners, 42, 76 business processes, 40, 48, 50, 56 businesses, 118

C CAD, 76 candidates, 12, 62, 127, 145 cash, 66, 79

198

Index

catalyst, 170, 183 categorization, 152 category a, 30, 163, 187 causal relationship, 165 Census, 118, 119, 164, 195 challenges, 17, 117 Chicago, 196 children, 66, 67 city(s), 64, 109, 113 classes, 33 classification, 7, 16, 25, 26, 28, 33, 38, 44, 45, 48, 55, 77, 79, 82, 83, 84, 86, 88, 99, 101, 102, 104, 106, 107, 115, 116, 125, 128, 129, 133, 136, 145, 146, 147, 161, 162, 163, 164, 168, 169, 176, 178, 179 climate, 40 clothing, 7 cocaine, 112 coding, 84, 152 cognitive therapy, 23 collaboration, 11, 61 color, 85 combined effect, 112 commercial, 41 communication, 5, 51, 52, 156 community(s), 6, 7, 8, 10, 11, 21, 29, 35, 37, 56, 61, 63, 76, 81, 82, 83, 112, 113, 115, 117, 119, 121, 122, 123, 126, 128, 132, 133, 134, 135, 136, 137, 138, 144, 155, 157, 159, 160, 161, 162, 170, 173, 179, 181, 184, 190, 191, 193, 194 community service, 35 competition, 3 compilation, 162 complement, 55, 169 complexity, 36, 37, 48, 166, 168, 182 compliance, 9, 10, 13, 15, 23, 46, 84, 116 composites, 139 computer, 43, 80, 82, 84, 86, 152 computing, 48, 168 conference, 64 configuration, 84, 145 consensus, 3, 12, 40, 41, 43, 62, 63, 170 consolidation, 46 construction, 112, 138, 158, 180 consulting, 194 consumption, 14 contingency, 53 control group, 34 control measures, 54 conviction, 114, 133, 157, 186 cooperation, 16, 40, 151 coordination, 11, 16, 53 correlation, 112, 113, 127

cost, vii, 1, 2, 6, 8, 11, 22, 24, 37, 49, 50, 57, 59, 60, 62, 63, 76, 77, 78, 79, 116, 131, 135, 136, 137, 138, 143, 144, 178, 180, 184, 196 cost benefits, 135 cost saving, 6, 138, 178 cost-benefit analysis, 22 counsel, 114, 127, 129, 130, 189 counseling, 79 CPU, 49 creep, 76 crimes, viii, 7, 72, 111, 112, 119, 166, 173, 186 criminal activity, 118, 119, 132 criminal behavior, 133, 134, 135, 156 criminal justice policy, 21, 22, 118, 166 criminality, 8, 165 crises, 28, 117, 155 CT, 196 culture, 4, 5 curriculum, 49 cycles, 36

D danger, 52 data analysis, 5, 24, 28, 36, 157, 160 data availability, 165 data collection, vii, 1, 4, 5, 8, 13, 14, 15, 16, 17, 23, 24, 29, 36, 37, 38, 74, 122, 124, 126, 129, 139, 144, 147, 149, 151, 152, 160, 162, 178, 184 data gathering, 16 data transfer, 48 database, vii, 5, 16, 24, 27, 46, 53, 56, 75, 76, 79, 81, 83, 85 database management, 56 defects, 47, 48, 50, 60, 63 defendants, 113, 115, 121, 127, 128, 129, 130, 131, 132, 133, 135, 138, 142, 147, 148, 152, 154, 157, 186 deficiency(s), 44, 46, 52, 54, 75 deinstitutionalization, 112 demographic change, 29, 36 demographic data, 5 demonstrations, 61, 62, 64 Department of Homeland Security, 77 Department of Justice, 109, 110, 119, 123, 194, 195, 196 deposits, 68, 72, 104 depth, 39 detainees, 34 detection, 55, 85 detention, viii, 81, 111, 117, 124, 127, 128, 140, 142, 144, 162, 166 deterrence, 135

199

Index detoxification, 113, 114, 122, 155, 159, 181 diet, 68, 93 directors, 17, 179 discrete variable, 165, 168 disposition, 122, 129, 131, 147, 148, 149, 154, 156, 157, 160, 190 dissatisfaction, 38 distribution, 34, 46, 106, 152, 178 district courts, 81 diversity, 13, 141 DNA, 101 domestic violence, viii, 67, 111, 112, 131, 132, 160 draft, 80, 117, 187 drug abuse, 136 drug courts, 132 drug offense, 114, 115 drug treatment, 23, 131, 134 drugs, 65, 86, 179 dynamic systems, 182

E early warning, 18 education, 8, 66, 67, 106, 123, 179, 188 educational background, 154 eligibility criteria, 133 emergency, 65, 118, 120, 136 employees, 3, 25, 26, 38, 124 employment, 8, 133, 137, 141, 149, 156, 182 employment status, 141 enemies, 66 energy, 2 enforcement, 10, 80, 120, 156, 167 engineering, 42 environment(s), 2, 3, 6, 7, 8, 25, 38, 39, 42, 45, 48, 60, 74, 115, 120, 138 epidemic, 112 equipment, 15, 27, 71 error detection, 85 ethnicity, 29, 141, 146, 187 evidence, viii, 17, 111, 112, 115, 117, 124, 125, 126, 134, 135, 137, 181, 194 evidence-based practices, 17, 124, 125, 126, 134, 135, 181 evolution, 43, 55 execution, 48 exercise, 2, 3, 125, 139, 144, 146, 170, 180, 185 expenditures, 39, 112, 113 experimental design, 23, 34 expertise, 5, 55, 58, 64, 124, 152, 166 exports, 12 exposure, 75

F fear, 7 federal government, 194 feelings, 54 felon, 162 fidelity, 46 financial, 5, 35, 46, 50, 59, 61, 79, 118, 143 financial condition, 143 financial resources, 5, 46, 118 firewalls, 5 flaws, 45, 47, 49 flexibility, 39, 64, 99, 115, 163, 169 flight, 7 fluctuations, 115, 162, 165, 175 force, 27, 47, 84 forecasting, viii, 5, 6, 14, 23, 25, 27, 28, 36, 111, 118, 164, 165, 166, 168 forecasting model, 25, 118, 168 formula, 159, 163, 168, 196 fraud, 32 freedom, 85 frequency distribution, 152, 153 frequent fliers, 136 funding, 7, 35, 37, 38, 39, 40, 58, 78, 116, 129, 135, 160 funds, 39, 63, 72, 138

G goal-setting, 23 good behavior, 8, 10 grants, 195 graph, 27, 28, 31, 118 Grievances, 69 grouping, 159 growth, 9, 37, 58, 112, 113, 114, 115, 119, 122, 157, 166, 167, 175, 176, 180, 182 growth rate, 113, 114, 166 guidance, vii, 1, 22 guidelines, 43, 128, 194 guiding principles, 185 guilty, 130, 157, 189

H health, 23, 71, 123, 132 high school, 65, 157, 188 hiring, 8, 82, 124 Hispanic population, 141 historical data, 53, 158, 159, 169, 170 history, 7, 15, 28, 62, 65, 66, 67, 71, 76, 80, 88

200

Index

hotels, 118 housing, 7, 8, 24, 26, 28, 68, 69, 71, 82, 84, 86, 103, 115, 116, 137, 138, 145, 164, 179, 183 hub, 10 human, 82, 86 Hurricane Katrina, 116 hypothesis, 21

I ICE, 79, 106 ID, 90 ideal, 44, 151 identification, 7, 25, 35, 41, 42, 47, 48, 60, 77, 79, 127, 146, 150, 186, 187 identity, 77 immigration, 10, 122, 129, 140, 141, 146 Immigration and Customs Enforcement, 113 imprisonment, 113 improvements, 2, 6, 24, 77, 83, 182 in transition, viii, 111, 117 incarceration, 7, 10, 66, 112, 113, 114, 118, 119, 120, 121, 135, 136, 141, 158, 166, 167, 173, 177, 182, 194 incidence, 132, 147 independent variable, 165 individuals, 66, 112, 113, 114, 120, 123, 124, 134, 138, 139, 140, 143, 145, 148, 151, 159, 161, 165, 167, 168, 171, 174, 178, 192 industry, 3, 8, 15 information sharing, 5, 42, 78 information technology(s), vii, viii, 2, 3, 24, 39, 42, 56, 79, 87, 110, 111 information-processing organizations, vii infrastructure, vii, 5, 18, 64, 116 injury(s), 15, 65 inmate-processing activities, vii, 2 INS, 101, 106 inspections, 71 institutions, vii, 1, 6 integration, 47, 80, 125, 139, 151, 181 integrity, 46, 47, 54, 88, 89, 98, 100, 108, 109, 115, 116, 120, 194 interagency committees, 35 interface, 50, 75, 76, 79, 81, 82, 85, 88, 89, 98, 100, 108, 109 interoperability, 64 intervention, 114, 130, 135, 159 intoxication, 126, 149, 161 investment, 39, 133 iris, 76, 79 isolation, 124

issues, 3, 4, 9, 13, 15, 16, 17, 20, 21, 24, 26, 27, 42, 45, 50, 51, 54, 55, 57, 67, 69, 70, 71, 74, 78, 81, 121, 125, 126, 129, 133, 140, 144, 149, 157, 164, 194

J jail administrators, vii, 1, 2, 8, 11, 12, 13, 14, 26, 36, 52 jail bed need, viii, 111, 161, 165, 169, 175 jail capacity planning guide, vii jail information system, vii, 14, 43, 56, 57, 74 jail management information system, vii, 1, 2, 86 jail system, vii, 2, 3, 4, 5, 9, 10, 11, 13, 16, 17, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 53, 54, 56, 57, 58, 59, 60, 61, 62, 63, 64, 74, 75, 76 job performance, 54 judicial branch, 138 judiciary, 131 justification, 44, 51 juveniles, 113, 140, 161, 165, 166

L law enforcement, 5, 8, 10, 11, 42, 75, 80, 81, 114, 115, 126, 155, 160, 188 laws, 112, 118, 185 layoffs, 8, 83 lead, 3, 4, 5, 9, 16, 19, 24, 41, 77, 116, 118, 155, 156, 167, 194 leadership, 2, 44, 52 learning, 3, 24, 46, 56, 75 Lee County, 131 legislation, 10, 39, 126, 132 level of education, 141 light, 145, 146 literacy, 17 litigation, 7, 128 living environment, 137 local government, 8, 57 love, 82 low risk, 7, 145, 168, 179

M major decisions, 54 majority, 86, 143, 161 man, 121, 178, 196 manual systems, 14, 38 mapping, 118 marital status, 66

201

Index marketplace, 63 Maryland, 195 materials, vii, 1, 60, 61, 71 mathematics, 160 matrix, 61, 62 matter, 40, 112, 184 measurement(s), 9, 110 media, 9, 114 Medicaid, 132, 195 medical, 7, 18, 47, 65, 68, 69, 77, 82, 84, 103, 104 medication, 69, 186 melting, 175 membership, 26, 123 memory, 49 mental disorder, 126 mental health, 7, 84, 87, 113, 123, 132, 149, 155, 159, 163, 179, 189, 194 mental illness, 65, 112, 115, 126, 132, 136, 161, 164, 183, 186 mentoring, 16 messages, 51 methodology, 117, 124, 147, 162 metropolitan areas, 141 Microsoft, 83 military, 66 MIS, vii, 1, 2, 5, 6, 7, 9, 10, 11, 12, 14, 16, 17, 18, 20, 23, 24, 27, 28, 36, 42, 44, 47, 51, 52, 53, 54, 55, 64, 76, 77, 78, 80, 83, 84, 85, 101 mission, 37, 117, 124 misuse, 103 models, 4, 118, 124, 134, 165, 166, 168 modifications, 47, 121, 139 modules, 28, 36, 82, 86 momentum, 45 morale, 8, 15, 16, 28, 35, 36, 52, 54 motivation, 44 multidimensional, 143 multiple factors, 147

N narcotics, 141, 146, 160 narratives, 83 nationality, 64 NIJ, 195 nutrition, 7

O obstacles, 5, 54, 78, 82 offenders, viii, 7, 11, 21, 25, 30, 33, 111, 113, 114, 116, 117, 118, 122, 128, 131, 132, 133, 134, 135,

136, 137, 138, 147, 155, 156, 157, 178, 179, 182, 191 Office of Justice Programs, 119, 194, 195, 196 officials, 7, 116, 124, 160, 184, 194 OH, 131 omission, 85 operating costs, 112, 137 operating system, 56 operations, 5, 6, 7, 8, 9, 10, 12, 13, 24, 26, 38, 41, 49, 50, 51, 52, 54, 62, 78, 184 opportunities, 11, 19, 58, 136 organ, 88 organizational learning, 3, 15 organize, 3, 12, 16, 21, 36 overlay, 142, 143 oversight, 40, 42, 57 overtime, 8, 70 ownership, 59, 62, 63, 78, 83

P parallel, 50 parole, 7, 10, 102, 113, 123, 131, 140, 141, 143, 146 participants, 8, 62, 132 pass/fail, 59 payroll, 70 performance indicator, 9, 22, 23, 34 performance measures, vii, 2, 7, 10, 84 permit, 151, 158 personal communication, 137, 196 physical health, 163 pilot study, 22 planning decisions, 16 platform, 43 police, 27, 40, 79, 118, 123, 124, 128, 129, 159, 165, 182, 183, 186 policy choice, 122, 169, 182 policy issues, 17, 19, 25, 27, 37, 40, 144 policy options, 3, 22 policy problems, 14, 20, 32, 85 policymakers, 12, 14, 21, 22, 37, 123, 124, 139 politics, 10, 17 population control, 35 population group, 28, 154 population growth, 9, 20, 164, 166, 167, 169, 171, 177, 182 potential benefits, 11 predictive accuracy, 168 preparation, 53, 57, 58, 125 primacy, 117 primary data, 13 principles, 156, 180 prisoners, 21, 113, 137, 140

202

Index

prisons, 5, 40, 113, 115, 119, 161 private sector, 37, 56 probability, 128 probation officers, 155 problem solving, 5, 14, 46 problem-solving, 17, 51, 55 processing variables, 168 procurement, 57, 59, 62, 63, 64 professionalism, 10 professionals, 68 program outcomes, 23 program staff, 23, 127 programming, 4, 15, 137 progress reports, 22, 51, 52 project, 37, 38, 39, 40, 41, 42, 43, 44, 45, 51, 56, 59, 61, 76, 77, 78, 79, 80, 81, 82, 85, 124, 131 proposition, 158 protection, 7, 53, 117 prototype, 45 psychiatric hospitals, 115 psychotropic drugs, 65 psychotropic medications, 140 public administration, 8 public concern, 24, 35 public health, 126 public policy, 85 public safety, 7, 8, 10, 27, 35, 36, 117, 128, 134 punishment, 114, 117

Q qualifications, 61 quality assurance, 42, 55, 85 quality control, 3, 5, 6, 15, 17, 54, 59, 123, 135, 144 quality of life, 7 query, 28, 30, 31, 32, 82, 83, 85 questionnaire, 65

R race, 141, 152, 164 random assignment, 34 random numbers, 160 rate of change, 56, 168 rating scale, 89 ratio analysis, 167 real time, 26 reality, 129, 136 real-time basis, 80 recidivism, 7, 8, 36, 128, 132, 133, 134, 135, 136, 156, 168, 178, 179 recidivism rate, 36

recognition, 19, 65, 76 recommendations, 129, 149, 151, 157, 180, 181, 184 recreation, 7 recycling, 135, 136 reform, 194 regression, 28, 36, 48, 165, 168 regression model, 165, 168 regulations, 64 rehabilitation, 8, 27, 35, 136 rehabilitation program, 8 reinforcement, 51 reliability, 52, 100, 149, 151, 158 rent, 66 repair, 71 requirements, 9, 12, 26, 34, 37, 38, 39, 41, 42, 43, 44, 45, 47, 48, 49, 57, 58, 59, 60, 61, 62, 63, 64, 67, 71, 74, 75, 76, 77, 78, 81, 83, 88, 99, 100, 177 researchers, 134 Residential, 195 resistance, 5, 16, 42, 46, 54 resolution, 47, 51, 69, 114, 124, 129, 130, 131, 143, 177, 181, 182, 184, 194 resource allocation, 3, 24, 179 resource utilization, 13 resources, viii, 2, 3, 4, 5, 9, 10, 11, 15, 16, 27, 31, 34, 37, 39, 41, 45, 46, 58, 102, 103, 111, 112, 113, 114, 115, 117, 122, 128, 130, 132, 133, 146, 147, 151, 155, 156, 165, 179, 180, 194 response, viii, 48, 59, 61, 62, 100, 111, 115, 116, 117, 126, 128, 129, 130, 143, 180 response time, 48 restitution, 137 restrictions, 112 revenue, 39, 68 rewards, 23, 52, 134 rights, 7, 79 risk(s), viii, 5, 7, 8, 11, 21, 27, 36, 40, 48, 50, 59, 60, 78, 111, 117, 122, 125, 126, 127, 128, 129, 131, 132, 133, 134, 135, 136, 137, 142, 144, 154, 156, 163, 177, 178, 179, 180, 183, 184, 194 risk assessment, 7, 126, 128, 129, 144, 156, 177, 180, 194 risk factors, 7, 156, 179 risk management, 36 robust design, 48 rules, 52, 57, 61 runoff, 175

S sabotage, 16, 44, 46, 47, 54 SACPA, 195 safety, viii, 5, 7, 23, 35, 36, 111, 115, 137

203

Index sanctions, viii, 11, 23, 111, 117, 122, 125, 126, 133, 134, 135, 136, 137, 156, 157, 178, 181, 191 savings, 79, 130, 134, 138 scarce resources, 6 school, 24, 67, 188 science, 118, 164, 165, 182 scope, 19, 26, 39, 40, 41, 42, 43, 44, 52, 55, 57, 60, 76 scripts, 48, 53 seasonal changes, 152 seasonality, 152 security, 4, 5, 15, 16, 18, 25, 26, 28, 29, 32, 33, 34, 35, 36, 44, 55, 70, 82, 84, 106, 107, 115, 137, 138, 145, 164, 176 segregation, 103, 163 sentencing, 10, 11, 36, 107, 112, 114, 122, 131, 132, 155, 166, 183, 185 servers, 56 services, 6, 7, 9, 14, 26, 28, 35, 39, 59, 60, 61, 62, 63, 77, 79, 114, 120, 121, 123, 124, 125, 126, 127, 129, 132, 134, 136, 138, 154, 155, 165, 178, 180, 181, 183, 190, 194, 196 sex, 101, 103, 154, 155, 159 sex offenders, 155, 159 sexual assaults, 84, 104 shape, 125, 139, 180, 182 shock, 180 shortage, 53 showing, 24, 134, 147 side effects, 22 signs, 52 smoothing, 168 society, 136 software, 9, 26, 28, 32, 36, 39, 45, 46, 47, 48, 49, 50, 51, 52, 56, 59, 60, 61, 62, 63, 64, 76, 77, 78, 79, 82, 151, 152, 153 solution, 17, 21, 22, 41, 57, 59, 81, 82, 116 specifications, 45, 87 spending, vii, 1, 2, 9 stability, 79 stabilization, 113, 137, 155, 191, 193 staff members, 25, 26 staffing, 5, 8, 24, 27, 28, 35, 54, 184 stakeholders, 3, 8, 10, 11, 12, 14, 22, 34, 38, 40, 44, 45, 46, 51, 52, 55, 77, 78, 79, 80, 84, 87 standard deviation, 151 state(s), 5, 10, 38, 40, 42, 71, 72, 76, 78, 113, 115, 119, 120, 121, 122, 123, 126, 129, 132, 135, 140, 142, 146, 147, 158, 159, 164, 167, 170, 174, 177, 182, 183, 187, 188, 191, 194 state mental hospitals, 115 state planning, 164 Statistical Package for the Social Sciences, 151

statistics, 14, 16, 18, 36, 119, 143, 160 stock, 106 storage, vii, 1, 16, 24, 49 strategic planning, 38, 42 stress, 48, 51, 52 structure, 29, 40, 44, 57, 83, 85, 138, 139 structuring, 156, 184 style, 16, 22, 137 subgroups, 154 substance abuse, 8, 123, 136, 137, 156, 178 substitution, 46 success rate, 137 suicide, 36, 65 suicide attempts, 36 supervision, 7, 13, 15, 46, 53, 54, 84, 114, 115, 118, 127, 128, 129, 133, 134, 135, 137, 141, 142, 143, 148, 154, 155, 156, 159, 164, 194 supervisor(s), 9, 13, 52, 53, 55, 84, 87 support staff, 77, 130 surplus, 54 surveillance, 7, 134, 135 system analysis, 139, 168 system planning, viii, 38, 40, 111

T target, 36, 38, 39, 107, 182 target population(s), 36, 107 Task Force, 195 team members, 62 teams, 53, 77 technical assistance, 49 technical change, 49 technical support, 53, 77 techniques, 6, 36, 160, 166, 170 technology(s), viii, 2, 38, 39, 41, 44, 47, 49, 50, 55, 56, 74, 77, 80, 81, 82 testing, 21, 41, 42, 45, 47, 48, 50, 60, 169 threats, 27 time frame, 39, 59, 60, 66 time periods, 27, 34, 48 time-frame, 36 tracks, 147, 148, 168 trade, 141 training, vii, 1, 4, 8, 10, 15, 17, 24, 27, 35, 36, 39, 42, 45, 47, 49, 52, 53, 55, 57, 60, 61, 74, 83 transactions, 79 transmission, 76 transport, 26 transportation, 26, 28, 79, 104 treatment, 4, 11, 13, 15, 17, 18, 26, 35, 67, 75, 114, 117, 131, 132, 133, 134, 135, 137, 149, 154, 155, 156, 177, 178, 182, 194, 196

204

Index

trial, 11, 21, 27, 47, 113, 114, 121, 140, 147, 148, 149, 151, 154, 161, 190 triggers, 134 troubleshooting, 45 turnover, 8, 55, 74

violent crime, 115 vision, 5, 12, 39, 44, 51, 52, 77 vocational training, 23 voters, 184

W U United, 115, 194 United States, 115, 194

V vacancies, 102, 103, 118 vacuum, viii, 10, 111 valve, 133 variables, 25, 106, 107, 117, 118, 140, 148, 149, 150, 151, 153, 158, 159, 165, 166, 168, 169, 185, 186, 187, 190 variations, 48, 63, 149, 152, 156, 166 vehicles, 71 vein, 86 victims, 72, 116, 132, 148 violence, 7, 150, 186, 188 violent arrests, 166

Washington, 109, 110, 119, 123, 131, 134, 136, 137, 179, 194, 195, 196 waste, 46 water, 175 weakness, 147, 167 welfare, 69 witnesses, 116, 148 workers, 8 workflow, 9, 75, 76, 79, 89 workload, 8, 9, 23, 24, 26, 27, 28, 52, 54, 102, 120, 124, 159 workplace, 51

Y yes/no, 150, 186, 187, 189 yield, 15, 34, 36, 144, 179