114 98 3MB
English Pages 238 [230] Year 2020
Politics and Policy Knowledge in Federal Education Confronting the Evidence-Based Proverb Steven Putansu
Politics and Policy Knowledge in Federal Education “In this important book, Steven Putansu argues cogently that a simplistic adherence to the idea of ‘evidence-based policy’ has negatively impacted how we understand policy and politics. His careful analysis of US federal education policy over 50 years, provides a clear case for embracing ‘evidence-based governance’— a broader and more realistic understanding of the interplay of policy knowledge, decision purposes and politics.” —Jenny M. Lewis, Professor of Public Policy, The University of Melbourne, Australia “Politicians demand that governments produce policy knowledge but then struggle to put that information to good use. With a rare combination of a practical experience, in-depth knowledge of the history of federal programs, and on-point use of scholarly theory and cutting-edge research, Putansu offers a compelling assessment of this struggle, while signposting the factors that make a more realistic marriage of policy knowledge and policymaking possible.” —Donald Moynihan, Professor, McCourt School of Public Policy, Georgetown University, USA “It is rare that scholarship in public administration or political science inform one another. It is rarer still that students of each are informed by the practical acumen of those in government. In Politics and Policy Knowledge in Federal Education, Putansu provides a refreshing exception to these norms, as he skillfully unpacks the complexities and politics of information production and consumption in the US federal government.” —William Resh, Associate Professor, Sol Price School of Public Policy, University of Southern California, USA
Steven Putansu
Politics and Policy Knowledge in Federal Education Confronting the Evidence-Based Proverb
Steven Putansu US Government Accountability Office Washington, DC, USA
ISBN 978-3-030-38394-7 ISBN 978-3-030-38395-4 (eBook) https://doi.org/10.1007/978-3-030-38395-4 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To Bridget O’Brien nothing matches the magic of learning things with you
Foreword
There are many books sitting on the bookshelves in my study that have had a single reading over the years. Each of them captured my attention for a short time but did not remain in my consciousness. Not only did I fail to open them again but I did not view them as intellectually alive. I very much doubt that this will be the fate of Politics and Policy Knowledge in Federal Education: Confronting the Evidence-Based Proverb. Steven Putansu has written a book that should stand as a model for this current generation. He has picked up the agenda established by scholars of policy and evaluation of the end of the twentieth century (especially Carol Weiss) and has written a challenge to reformers, academics, and practitioners of the early twenty-first century to acknowledge the complexity of the world in which we live. Instead of looking for quick fixes built around overly simplistic prescriptions, Putansu’s volume offers a reader a threefold agenda. First, he reminds the reader that decision making inevitably requires the independent and mutual value of ideology, interests, institutions, and policy knowledge. All of these—not simply one or two of them—are required for understanding, influencing, and making decisions. Second, his focus on data, information, and evidence allows him to examine the way that the purpose, design, and quality of policy knowledge can have an impact on and improve decision making. He has emphasized the impact of policy knowledge on two major federal education programs and its role in legislative, policy, and administrative decisions. His analysis
vii
viii
FOREWORD
stands as an implicit criticism of those who argue that the simple collection of information is “objective” and will lead one to the “correct” answer. Third, the volume provides a rich account of 50 years of federal decision making in the United States in two programs: the higher education student loan program and Title I grants in elementary and secondary programs. This historical analysis offers key insights into the goals, program design, and key debates that shape modern federal education policy as they respond to reforms as well as major changes in the policy and political environment. Dr. Putansu’s familiarity with the institutions of US decision making has allowed him to present a perspective that will be extremely useful to practitioners as well as academics and researchers. He concludes the book by offering lessons learned and key considerations for producers of policy knowledge, decision makers who use it, and reformers who attempt to link them together. Beryl A. Radin, Ph.D. Author, Faculty Member, Researcher Washington, DC, USA
Acknowledgements
This book is the product of a combination of research completed and practical experience gained over at least the last 12 years. It started with a paper written in a doctoral class that sought to understand if information from performance management systems served as a replacement for information provided by academics, nonprofits, or even other information from government entities. That paper turned into my Ph.D. dissertation, for which much of the data, information, and evidence in this study was originally collected and analyzed. I began working at the Government Accountability Office (GAO) while writing that dissertation, and my experience working to support a mission if providing objective, non-partisan information to Congress to support better decision making has influenced how this project has evolved over a ten-year career. At GAO, I have been lucky to be deeply involved in the design, methods, and analysis of work across a broad range of policy areas, where we have created policy knowledge for issues with broad bipartisan support, and with strong partisan disagreement, all while working to maintain independence and balance. Lessons from this work have inspired me to reframe, revise, and update this study, with the hope that this combination of scholarship and practical insight can contribute to the continued development and use of policy knowledge to help support robust political decision making, improve government, and serve the public.
ix
x
ACKNOWLEDGEMENTS
With that goal in mind, I want to start by acknowledging other public servants, broadly speaking. At times, this book can be critical of some attempts to bring policy knowledge to the forefront of public service, but my underlying goal is not to tear down their efforts, but rather to join them, and bring a contextually sophisticated approach that better places policy knowledge within the decision making environment. I have interacted with a variety of public servants—elected officials from both parties, their staffs, agency employees, private contractors, professional and technical experts, non-governmental entities, reformers, academics, and others—whose commitment to good governance has consistently inspired me. Every day, they are working to achieve a broad range of goals, within a political environment, while dealing with an onslaught of policy knowledge that often promises to solve all their problems. They often become scapegoats when lofty reforms drain their resources but fail to deliver on unrealistic promises. I hope that this book plays a small part in better defining how the purpose and quality of different sources of policy knowledge can be leveraged across the broad range of constraints they face, to better match policy knowledge to their needs and to improve governance writ large. Personally, I have many other people to thank. Chief among them is Beryl Radin, who chaired my dissertation committee and dedicated dozens and dozens of hours to help me build that first research project. Since then, she has spent countless additional hours mentoring, counseling, and advising me through the theory, framework, and application of evidence in this book. Much of her scholarship framed my original thinking on this topic and subsequent revisions. Her contributions were immense, and my gratitude for them cannot be overstated. I was honored at her agreement to write the preface of the book, as I think she understands the goal of the book as well as anyone else could. I want to thank many people at American University’s School of Public Affairs. This includes many who were initially my professors and then became colleagues when I joined the faculty. Laura Langbein, who may have the broadest combination of policy and program research, and the deepest knowledge of methods, of anyone I have ever met, is a mentor who has provided me insights and ideas that contribute to a wide variety of the thoughts and ideas on this text. These contributions continued well beyond her service on my dissertation committee. Jocelyn Johnston, who was the professor who sparked the first idea that eventually led to this
ACKNOWLEDGEMENTS
xi
book. Alison Jacknowitz, who gave me my first direct experience conducting rigorous and practical policy and program evaluations. Robert Durant, who consistently reminded me how to place my ideas and findings into the broader theory of the field of public administration. David Rosenbloom, who ingrained in me a deep respect for the democraticconstitutional values that shape the decision making environment. Sonja Walti, Vicki Wilkins, and Kim Maloney, who have worked with me to develop and refine my instructional skills. My students, who have brought practical insights and experiences that have opened my eyes to other constraints, interactions, and values of policy knowledge. I also thank my GAO colleagues, starting with Nancy Kingsbury. Dr. Kingsbury served on my dissertation committee, and as the head of the Applied Research and Method team where I work, and has consistently promoted thoughtful, interdisciplinary, and contextually sophisticated approaches to generating practically useful policy knowledge. Thanks to GAO colleagues who have helped me grow in my knowledge of evaluation and evidence, including Valerie Caracelli, Stephanie Shipman, and Martin de Alteriis. Thanks to GAO colleagues who have deepened my methodological skills, including Terry Richardson, Mark Braza, Rebecca Shea, Oliver Richard, Susan Offutt, and many others. Thanks to GAO colleagues who have pushed my knowledge of strategic management, managing for results, crosscutting issues, and implementing evidence-based policy; Chris Mihm, Ben Licht, Adam Miles, Robert Goldenkoff, and others. Thanks to Gene Dodaro, the Comptroller General of the United States, who has led by example with a comprehensive view of the complexities of government, ability to draw insights across bodies of work, and passionate commitment to a government that serves Congress and the American people. Finally, thanks to every other GAO employee I have worked with, as each interaction has contributed to the ideas in this book. I must acknowledge all the scholars who have influenced my thoughts and writing in this book. This includes all those in the 200+ works cited by this work and the hundreds of other books and articles that, while not directly cited, have influenced the material in this book in direct and indirect ways. My deepest appreciation goes to those whose work offered the central frameworks that helped me to frame this research—Carol Weiss, Beryl Radin, Donald Moynihan, and James Q. Wilson. Thank you to all those who, through comments on conference papers, discussions over coffee, and periodic check-ins and encouragement, have kept me committed to
xii
ACKNOWLEDGEMENTS
this goal—Geert Bouckaert, Jenny Lewis, Jonathan Justice, Carolyn Heinrich, Blue Wooldridge, John Kamensky, Jessica Sowa, Alex Henderson, Jaclyn Piatak, Marc Holzer, and many others. Thank you to those who have helped me navigate the publishing process—Bill Resh, Don Kettl, Donald Moynihan, and Tina Nabatchi—and the publishing and editorial team at Palgrave-Macmillan—Michele Chen, John Stegner, and Rebecca Roberts. Finally, a special thank you to the academic peer reviewers (who I do not know) that have provided comments and revisions for the text, and for related materials. Most of all, I have to share my deepest gratitude with friends and family that have supported me, encouraged me, and stood by as I have disappeared for months or years on end while committing to this research project. Thank you to my wife, Bridget O’Brien, for serving as my core moral support throughout this project and all of our life adventures. Thank you for being a sounding board and copy editor on the book, and for being an inspiration and motivation every day. To all my parents, siblings, DC mom, and other relatives, who have shown love and support over many, many years. Thanks to everyone else, too numerous to name, who have shared any part of this journey with me. Finally, thank you to anyone who reads this book, who finds its insights and lessons useful, and who uses it to produce better data, information, and evidence to support and improve decisions, to improve governance, and to improve the lives of others.
Contents
Part I Confronting the Evidence-Based Proverb 1
2
3
Policy Knowledge, Politics, and a Proverb of Decision Making The Evidence-Based Proverb and Policy Knowledge as Panacea A Research Approach to Tackle the Evidence-Based Proverb Organization of the Book References
3 8 17 21 26
No Panacea: The Purposes, Uses, and Limitations of Policy Knowledge Defining a Typology of Policy Knowledge The Purposes and Creation of Policy Knowledge Direct and Indirect Uses of Data, Information, and Evidence Program Design and the Production of Policy Knowledge References
31 32 35 44 45 47
Challenging the Proverb: A Balanced Model for Governance Decisions Politics and Policy Knowledge as Drivers of Decision Making A Balanced Framework of Politics and Policy Knowledge Political Elements of Policy Knowledge References
51 53 59 64 67 xiii
xiv
4
CONTENTS
The Performance Movement and Other Evidence-Based Reforms The Performance Movement: 50 Years of Evidence-Based Reforms Evidence-Based Policy: The Modern Reform Agenda Decision Making in Federal Education Policy References
71 72 85 86 87
Part II Politics and Policy Knowledge in Federal Education 5
The Roots of Modern Federal Education Policy Three Transformations of Education in the United States A Federal Role for Elementary and Secondary Education A Federal Role for Higher Education Implications for the Case Studies References
91 93 100 107 110 111
6
Improving Access to Higher Education HEA Titles and Background Politics and Policy Knowledge at HEA’s Passage Major Eras of Student Loan Programs Summary of the Case References
115 116 119 122 144 146
7
Education for the Disadvantaged ESEA Titles and Background Politics and Policy Knowledge at ESEA’s Passage Major Eras of Title I Grant Programs Summary of the Case References
149 150 154 156 175 178
Part III
8
Conclusions and Implications for Evidence-Based Governance
Contending with the Evidence-Based Proverb The Political-Rational Continuum
183 186
CONTENTS
9
xv
Incremental and Comprehensive Decisions The Implications of Program Goals and Design Quality Versus Purpose: Which Policy Knowledge Gets Used? References
188 190 193 194
Moving Beyond the Evidence-Based Proverb Unanswered Questions and a Research Agenda Creating Policy Knowledge: Purpose and Quality Understanding, Influencing, and Improving Decisions Looking Forward: Federal Education Policy and Evidence-Based Policy References
197 198 201 203
Index
205 208 211
About the Author
Steven Putansu is a pracademic who splits his time as a public servant at the US Government Accountability Office (GAO); a professorial lecturer at American University; and a public management scholar. Steven studies the intersection of evidence and politics in decision making, strategic management practices to enhance evidence-based decisions, and applied work in a variety of policy areas. At GAO, Steven has contributed to products on best practices for evaluating government programs, audit criteria, and frameworks for assessing performance systems, interagency collaboration, human capital, risk management, duplication, and fraud in the federal government. He has applied these and other frameworks in GAO engagements in a range of policy areas, including education, defense, international affairs, nuclear energy, project management, wildfire response, immigration, climate change, veterans affairs, freight rail, and traffic safety. His methodological contributions have ensured rigorous technical standards in hundreds of GAO performance audits, provided policymakers with high-quality information for overseeing federal programs, and ensured the best use of taxpayer dollars, helping to contribute over $260 billion in financial benefits for the federal government. For these contributions, Steven received an Arthur S Flemming award for Leadership and Management in 2017.
xvii
Abbreviations
ACA CAP CBO CCRA CRA CRS DOD DRA ECASLA ECIA ED ESEA ESSA FCRA GAO GED GPRA GPRAMA GSL HEA HEOA HEW IASA LEA MBO MISAA
Patient Protection and Affordable Care Act Cross-Agency Priority Congressional Budget Office College Cost Reduction and Access Civil Rights Act Congressional Research Service Department of Defense Deficit Reduction Act Ensuring Continued Access to Student Loans Act Education Consolidation and Improvement Act Department of Education Elementary and Secondary Education Act Every Student Succeeds Act Federal Credit Reform Act General Accounting Office/Government Accountability Office General Educational Development Government Performance and Results Act Government Performance and Results Act Modernization Act Guaranteed Student Loan Higher Education Act Higher Education Opportunity and Affordability Department of Health, Education, and Welfare Improving America’s Schools Act Local Education Agencies Management by Objective Middle Income Student Assistance Act xix
xx
ABBREVIATIONS
NCLB NDEA NDSL NIE NPR OE OMB PART PIP PMA PMI PPBE PPBS RCT SAFRA SEA US ZBB
No Child Left Behind National Defense Education Act National Defense Student Loan Program National Institute of Education National Performance Review US Office of Education Office of Management and Budget Program Assessment and Rating Tool Productivity Improvement Program President’s Management Agenda Presidential Management Initiatives Planning, Programming, Budgeting, and Execution System Planning, Programming, Budgeting System Randomized Control Trials Student Aid and Fiscal Responsibility Act State Education Agencies United States Zero-based Budgeting
List of Figures
Fig. 1.1
Fig. 1.2 Fig. 2.1 Fig. 3.1 Fig. 4.1 Fig. 6.1 Fig. 6.2 Fig. 6.3 Fig. 6.4 Fig. 6.5 Fig. 6.6 Fig. 7.1 Fig. 7.2
Use of popular evidence-based governance terms through the Corpus of American English, 1950–2008 (Source Google Labs—Books Ngram Viewer [available at http://ngrams. googlelabs.com/]) Use of the phrase “evidence-based” in peer-reviewed journals and US newspapers, 1995–2014 Policy knowledge types, coverage, and quality Dominant models of decision making by scope and type of influence Timeline of performance management reforms HEA decisions during the era of program expansion, 1965–1980 HEA decisions during the era of budget restriction, 1980–1992 HEA decisions during the era of muddling toward reform, 1992–2008 HEA decisions during the era of federal expansion, 2008–2010 HEA decisions during the era of muddling over the federal role, 2010–2015 Summary of major HEA decisions, 1965–2015 ESEA decisions during the era of expanding equity, 1965–1980 ESEA decisions during the era of reversion to the states, 1980–1988
9 10 33 60 73 128 132 139 142 144 146 162 166
xxi
xxii
LIST OF FIGURES
Fig. 7.3 Fig. 7.4 Fig. 7.5
ESEA decisions during the Era of federal accountability standards, 1988–2002 ESEA decisions during the era of returning state flexibility, 2002–2015 Summary of major ESEA decisions, 1965–2015
172 176 177
List of Tables
Table 2.1 Table 4.1 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table 7.1 Table 7.2 Table 7.3 Table 7.4 Table 8.1
Example dimensions of policy knowledge use Key shortcomings of performance management reforms Politics and policy knowledge from HEA’s passage and era of program expansion, 1965–1980 Politics and policy knowledge in HEA’s era of budget restriction, 1980–1992 Politics and policy knowledge in HEA’s era of muddling toward reform, 1992–2008 Politics and policy knowledge in HEA’s era of federal expansion, 2008–2010 Politics and policy knowledge in HEA’s era of muddling over the federal role, 2010–2015 Politics and policy knowledge from ESEA’s passage and era of expanding equity, 1965–1980 Politics and policy knowledge in ESEA’s era of reversion to the states, 1980–1988 Politics and policy knowledge in ESEA’s era of federal accountability standards, 1988–2002 Politics and policy knowledge in ESEA’s era of returning state flexibility, 2002–2015 Ideological changes and case study eras
38 84 124 129 133 140 143 158 163 167 173 187
xxiii
PART I
Confronting the Evidence-Based Proverb
CHAPTER 1
Policy Knowledge, Politics, and a Proverb of Decision Making
Without good data, information, and evidence, government decision makers cannot adequately identify problems, design solutions, or fully understand the impact of government action. This idea can be traced to James Madison’s (1787) argument for the virtue of a large republic over direct democracy in Federalist number 10. He argued in favor of “representatives whose enlightened views and virtuous sentiments render them superior to local prejudices and schemes of injustice.” In Madison’s view, general “enlightenment” was one factor that would help elected officials to make good decisions, and his argument implied that this enlightenment would protect decision makers from abandoning their virtuous sentiments and falling prey to baser, political differences. Madison believed that enlightenment was a necessary, but not sufficient, input into the political process (Bertelli and Lynn 2006). The “opposing forces,” as he described enlightenment and political sentiments, are balanced by elected decision makers in a republic. This book will examine the role that policy knowledge that constitutes enlightenment plays, alongside politics, to influence government action. Politicians, public servants, and governance scholars have attempted to emphasize the value of enlightenment for nearly two and a half centuries. In 1946, Herbert Simon critiqued the study of administration as relying on “Proverbs” that were driven by inconsistent theory and insufficient evidence. He urged scholars and practitioners to place greater emphasis on better defining conceptual terms, focusing on decision making, and building models of theory and practice that could achieve stated © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_1
3
4
S. PUTANSU
results. Consistent with Simon’s call to action, this emphasis has been driven through reform, management practice, and research about the use of policy knowledge—data, information, and evidence about governance (Lindblom and Cohen 1979). These efforts have been associated with a variety of reforms, including performance budgeting (Robinson 2007; West 2011), portfolio budgeting (Meyers 2017), performance management (Hatry 2006), and a range of other calls for evidence-based policy (Haskins and Baron 2011). However, evidence-based reforms have also suffered from a lack of conceptual clarity, insufficient understanding of decision making processes, and limited success in generating models of evidence use that are tailored to specific purposes. The first of these issues is a lack of conceptual clarity: The terms data, information, and evidence have been used in inconsistent, ambiguous, and sometimes inappropriate ways. To avoid confusion, the following will define how the terms will be used in this text. The phrase “policy knowledge” will be used as a broad umbrella term to cover all forms of data, information, and evidence used to inform the crafting and implementation of public policies. The foundation for this study’s definition of policy knowledge derives from Giandomenico Majone’s book, Evidence, Argument, and Persuasion in the Policy Process, which defines information in comparison with data and evidence. In this model, data represent an observation, information represents an interpretation of that observation, and evidence represents the selection and combination of information to make an argument (Ackoff 1989). As an example, these definitions could be applied to studies of the environment. Specifically, the average daily global temperature represents data, the upward trend evident in these data is one kind of information, and the associations between this trend and human factors represent evidence. This study also shows that evidence-based reforms have paid insufficient attention to decision making. Reformers have promised to generate policy knowledge to guide decision making, and these promises wrongly treated enlightenment as a replacement for, rather than a balance against, political forces. Through this flawed assumption, reformers have overlooked important elements of decision making, which has resulted in inconsistent and limited success. Finally, by highlighting the relationships among data, information, and evidence, this text illustrates a critical relationship between the policy knowledge created by reforms and the specific goals it can support. Data provide the basis for the development of information, and information offers insights into potential evidentiary
1
POLICY KNOWLEDGE, POLITICS …
5
analyses. As such, all three elements must be considered in order to understand the potential value of specific sources and types of policy knowledge. This book will use the phrase “evidence-based governance” to refer to the production and use of policy knowledge to support, assess, and understand government, and the phrase “evidence-based reform” to refer to any effort that advocates for greater use of data, information, and evidence. These kinds of reforms are consistent in principle with Simon’s call for a stronger science of public administration. However, this book argues that lack of conceptual clarity, insufficient understanding of decision making processes, and limited success in generating models of evidence use that are tailored to specific purposes have doomed them to join his other Proverbs of Administration. Advocates for evidence-based reform pursue increased production and use of policy knowledge, and increased transparency of its role in governance decisions. For example, starting in the 1960s, there have been repeated attempts to institute performance-based budgeting that explicitly links budget decisions to specific program performance information. At a broad level, these efforts are consistent with Madison’s emphasis on the ability of data, information, and evidence to improve decision making. However, these reforms have often experienced limited and uneven success, and a wealth of research has been dedicated to examining why political decision makers resist or ignore relevant policy knowledge (Gilmour and Lewis 2006a; Olsen and Levy 2004; Radin 2006; Shulock 1999). Both the framing of calls for reform and studies of their limited success reveal the evidence-based proverb at play. Specifically, both have treated the presence of political factors as support for claims that evidence-based reforms have failed, treating the two as an either-or proposition. In performance-based budgeting reforms, one common issue is deciding whether a program that is not meeting performance targets should have its budget reduced, or if this is a signal that the program needs more resources to work. The answer to this question requires consideration of competing goals and alternatives that are influenced by ideological beliefs, engaged interests, and institutional arrangements. Rather than assessing the balance between policy knowledge and these political dimensions, reformers have reduced Madison’s advice to a proverb that policy knowledge can and should supplant politics in decision making. This book will show that reliance on this evidence-based proverb, coupled with restrictive definitions of policy knowledge, has repeatedly led to
6
S. PUTANSU
unfulfilled promises that data, information, and evidence will drive successful governance. Over time, these broken promises may lead decision makers to become more skeptical of policy knowledge, leading paradoxically to decreased trust in and influence of enlightenment. For example, a study of one government-wide performance-based budgeting reform found that congressional appropriations decisions were not significantly correlated with performance results (Gilmour and Lewis 2006a), ultimately leading to the information itself being challenged as political and the effort to generate it being abandoned. This chapter provides a broad introduction to evidence-based governance. It will show that the “evidence” cited in related reform efforts is typically, whether explicitly or implicitly, limited to a single source or type of policy knowledge. It will show that, in each case, researchers and reformers have identified decisions, which they argue should have been guided by policy knowledge from evidence-based reforms, but which instead were driven by some combination of politics and “lesser” data, information, and evidence. This chapter will explain that this limited use of policy knowledge is a clear and predictable result of three aspects of reforms. Specifically, reforms oversimplify the definition of policy knowledge, overpromise on the likely impacts of their reform, and ignore or undervalue the role of politics. As a result, they produce data, information, and evidence that ignore the reality of the political system and lack the contextual depth necessary to identify and successfully support a range of relevant governance decisions. Despite these challenges, the goal of linking policy knowledge to decisions continues to hold promise for improved decision making, and evidence-based reforms have remained popular. However, their uneven and limited success will continue unless the evidence-based proverb is abandoned. Reformers and scholars must adopt a more realistic view of the strengths and limitations of policy knowledge and a more contextually sophisticated view of decision making that recognizes the role and value of political factors. They must abandon a conception of the value of policy knowledge rooted solely in the strength of data, information, and evidence and must incorporate considerations of purpose and relevant political factors. This book attempts to take a step toward this goal by offering frameworks and analyses that demonstrate how researchers and reformers can develop evidence-based reforms that are more responsive to the limitations addressed above. By examining the production, use, and impact
1
POLICY KNOWLEDGE, POLITICS …
7
of policy knowledge in federal education policy over the past 50 years, this study will explore how variation among the types and sources of policy knowledge, decision purposes, and political tensions impacts the decision making process. It will build on these findings to offer a more nuanced understanding of the nexus between policy knowledge and politics in guiding governance decisions. The goal of this study is to explore evidence-based governance broadly and inclusively to better understand whether decision makers are using data, information, and evidence; how they choose which types and sources of policy knowledge to use; and how it impacts the decisions they make. This book will apply a unique framework for systematically studying the role policy knowledge plays in the government decision making process. This framework combines insights from several disciplines, including public administration, political science, policy, evaluation, and organization theory. It will define and assess policy knowledge and politics to challenge promises of policy knowledge as a panacea for decision making, avoid the evidence-based proverb, and move toward a more realistic model of evidence-based governance. Specifically, it will integrate consideration of political factors, including ideology, interests, and institutions, to explain whether and how policy knowledge is used in different contexts (Smyrl and Genieys 2016; Weiss 1989). It will argue that the consideration of these factors is a precursor to successfully advancing comprehensive efforts to enhance the practical use of policy knowledge in decision making and to better understand the processes and outcomes of political decisions. This chapter lays the foundation for this book by providing a more complete definition of policy knowledge, which integrates and synthesizes the various ideals and hierarchies offered by evidence-based reformers. Then, it will conclude by explaining the research approach and content of the remainder of the text. The following section will provide an overview of the myriad of arguments and efforts made to integrate policy knowledge with governance decisions. It will show that these efforts have tended to present decision making as a competition rooted in the proverb that suggests data, information, and evidence can and should supplant politics in guiding decisions. The chapter will go on to show that reforms typically suffer from three key shortcomings. First, they either limit their focus to a single type of data, information, or evidence, or they attempt to enforce a restrictive definition that prioritizes their preferred type. This “one-size-fits-all”
8
S. PUTANSU
approach fails to consider the relative value of different sources and types of policy knowledge. Nor does it acknowledge that such preferences are themselves driven by political factors. Second, they tend to overpromise the ability of this narrowly defined policy knowledge to serve as a panacea for decision making purposes. Finally, they reduce the use of policy knowledge to a proverb that suggests it will supplant politics in decision making. As a result, the importance of interests, ideology, and institutional factors are overlooked. The following section will delve more deeply into the history of the evidence-based proverb and related reforms.
The Evidence-Based Proverb and Policy Knowledge as Panacea The evidence-based proverb can be traced back to Woodrow Wilson’s (1887) paper, “The Study of Administration,” which called for a science of administration to determine how to operate government “with the utmost possible efficiency and the least possible cost.” Wilson argued that this science could be separated out from politics and used to improve governance. While Wilson acknowledged a need to keep “democratic policy very much at heart,” his argument that policy knowledge could and should replace politics took on a life of its own, and helped give rise to the field of public administration as an academic discipline. From this time on, successive iterations of evidence-based reform have often relied on this proverb and have subsequently been challenged by skeptics and underutilized by decision makers operating within the political context. While Madison called it enlightenment, and Wilson called it science, the phrase “evidence-based” only recently became a term of art in governance studies and reforms. Starting in the 1940s, other popular terms for evidence-based governance included policy analysis, performance management, program evaluation, and policy evaluation. As Fig. 1.1 indicates, the relative popularity of these terms has shifted over time, with variations of “evidence-based” becoming the most popular since the late 1990s. The term originated with David Eddy (1984) in the field of medicine and gained traction in 1990 with his publication of several papers in the Journal of the American Medical Association (Eddy 1990; Hoskins 1990). An important aspect of evidence-based medicine was the argument that medical recommendations should only be as strong as the underlying evidence about a case. As such, advocates for evidence-based
1
POLICY KNOWLEDGE, POLITICS …
9
Fig. 1.1 Use of popular evidence-based governance terms through the Corpus of American English, 1950–2008 (Source Google Labs—Books Ngram Viewer [available at http://ngrams.googlelabs.com/])
approaches have established hierarchies of evidence, generally with randomized control trials (RCTs) designated as the gold standard. Beneath this level are various forms of quasi-experimental evidence, including control trials without randomization, studies of single cohorts over time, and time-series designs with pre- and post-intervention measurement. Finally, evidence derived from the opinions of various forms of experts is considered the weakest support for medical decisions. Government reformers associated with the global set of reforms known as New Public Management appropriated this terminology and began using the phrase evidence-based policy in the early 1990s (Gordon 1994; Osborne and Gaebler 1992; Pollitt 2005). A search of peer-reviewed journals revealed that the phrase “evidence-based policy” has grown steadily, with 55 articles referencing the phrase published through the year 2000, 540 more in the following 5 years, 933 between 2006 and 2010, and over 1500 more from 2011 through 2014. Similarly, newspapers began to adopt the phrase in 2002 and had published 47 articles by 2005. From 2006 to 2010, the phrase appeared in 288 more articles, with over 700 more from 2011 through 2014. Several authors have gone so far as to mark the development of evidence-based policy as an indication that we are moving toward an “evidence-based global society.” Figure 1.2 shows this growth in the use of the term. While the use of this specific terminology may be recent, its antecedents were codified into law at least as early as 1902 when Congress passed the Rivers and Harbors Act, requiring cost–benefit analyses for water resource programs. This was followed by the establishment of
10
S. PUTANSU Peer Reviewed Journals
Newspapers
375
Count of Articles using the term "Evidence-Based"
350 325 300 275 250 225 200 175 150 125 100 75 50 25 0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Fig. 1.2 Use of the phrase “evidence-based” in peer-reviewed journals and US newspapers, 1995–2014
legislative, information-producing agencies, including the Congressional Research Service (CRS) in 1914, the General Accounting Office (GAO) in 1921 (now the Government Accountability Office), and eventually the Congressional Budget Office (CBO) in 1974. Further, performancebased budgeting, which lies at the heart of performance-based management, was recommended by the Hoover Commission in February 1949, implemented across the entire federal government in 1965, and codified into law in 1993. To date, cost–benefit analyses, general research, audits, budget analysis, and performance management continued to appear in evidence-based reforms. Similarly, general calls for linking specific rigorous forms of policy knowledge to governance decisions were advocated well before the phrase “evidence-based” took root, perhaps most forcefully in Alice Rivlin’s (1971) Systematic Thinking for Social Action. In fact, as early as 1969, Donald Campbell wrote about the value of treating policy reforms as experiments for research purposes well before RCT was popularized in evidence-based reforms.
1
POLICY KNOWLEDGE, POLITICS …
11
Initially, those using the phrase evidence-based policy simply adopted the approach of the medical profession and its hierarchy of evidence, without direct references to earlier attempts to link policy knowledge with the political decision making process. New Public Management reformers failed to thoroughly consider the context of governance, and adopted the same evidence hierarchy, leading to strong challenges from policy researchers and obstacles to successful use (Radin 2012). For example, Carol Weiss (2002) challenged the reductionist view of the universal superiority of RCT by asking “What to do until the Random Assigner Comes?” By analyzing this question, she highlighted the ethical dilemma of choosing to provide government services to some, but not others, in the name of better evidence. Further, she reveals the difference between the single treatment and effect model common in RCT, and the complexity of the real world where numerous treatments are likely to be applied to a community for a variety of issues, with concurrent effects. Some go even further with their critiques, concluding that RCTs have a limited ability to demonstrate “what works” (Deaton and Cartwright 2018). Weiss (2002) argues that quasi-experiments and qualitative evaluations should also have a role in evidence-based governance. She suggests that theory-based evaluation, driven by a thorough understanding of program assumptions and intentions, could provide more meaningful policy knowledge to decision makers. While a hierarchy of data, information, and evidence is very strong for evidence-based medicine, where there is a clearly defined goal of improved health outcomes for individuals and a directly observable treatment-effect relationship, it is often less relevant to political decision making, where multiple, ambiguous goals and treatments exist in conflict with one another. Weiss’ alternatives to RCT are illustrative of attempts to expand evidence-based policy to a wider variety of policy knowledge. It also represents one faction of a broader balkanization of policy knowledge “camps.” In addition to RCT, Weiss (2001) introduces the potential for evaluation to be the basis for evidence-based policy. Similar calls for alternatives highlight desired primacy of policy analysis and social science research, program monitoring, cost–benefit analysis, performance measurement, expert knowledge, statistics, and economic modeling, among others (Riccucci 2010). In all these cases, reformers deal with specific purposes of this policy knowledge in very limited ways, if at all (Hird 2005; Julnes and Holzer 2001). A common pattern among these various reforms is to start with an argument that one form of policy knowledge is too narrow,
12
S. PUTANSU
then to suggest a similarly narrow replacement while promising it alone can be a panacea for government decisions. Advocates for evidence-based governance often frame their reforms as general grievances over ignored data, information, and evidence. They call for increased development and use of specific and narrowly tailored policy knowledge to address these grievances (Radin 2000b). As a result of their limited, often uncommunicated, definitions of policy knowledge, evidence-based reforms continually repeat the mistakes of previous reforms. They generate data, information, and evidence that are useful only under a narrow set of conditions, while promising it will better meet its promise as a panacea for decision making. Without considering the full range of policy knowledge that can be brought to bear on a decision, or explicitly identifying the limited purposes a specific type or source of policy knowledge is useful for, these reforms are destined to fall short of their promises. The specifics of the decision at hand may also impact which purpose is most relevant to decision makers (Behn 2003; Weiss 1979). Regulators may be more interested in compliance by private entities and may seek out information as a warning of noncompliance. In contrast, appropriators may be more concerned with the costs of implementing regulations and may look for evidence that provides guidance on how to keep costs under control. These two forms of policy knowledge, taken together, may also increase general knowledge of regulatory policies and behaviors, which may change the way future decisions are framed and acted upon. Citing differences in purpose, Harry Hatry (2006) explains that data, information, and evidence may require differential considerations of program inputs, processes, outputs, and/or outcomes. Specifically, one form of policy knowledge may be used to identify a problem, but may not be sufficient to suggest a solution. For example, decision makers may recognize a high unemployment rate as a warning about the economy, but would need more evidence to determine whether jobs programs, trade policies, and/or more education provide meaningful solutions to address that problem. Two results of this approach to evidence-based policy are harmful to the overall goal of better integrating policy knowledge into decision making. First, reformers risk insulting and alienating decision makers by accusing them of ignoring policy knowledge, with article titles like “What will it take for decision makers to care?” (Maynard 2006). Second, when decision makers have to look outside of the reform efforts to find the data,
1
POLICY KNOWLEDGE, POLITICS …
13
information, and evidence they need, or when the policy knowledge produced by a reform does not adequately consider the political context, decision makers begin to abandon these reforms (Kettl 2002; Salamon 2002; Stewart 2009). Rather than addressing the underlying reality that the needs of decision makers vary and no source of policy knowledge can be universally appropriate, new reforms seem to simply adopt new terminology, new preferred sources, and new types of policy knowledge. As a result, they repeat the same cycle of promising and failing to deliver on a panacea and continue to erode trust in policy knowledge. In a seemingly direct attempt to hedge against these criticisms, some evidence-based advocates have acknowledged the realities of the political process, contextual sophistication, and imperfect policy knowledge by using less deterministic phrases, including “informed policy,” “evidenceinformed policy,” and “evidence-aware policy,” among others (GAO 2004; Julnes 2007; Posner and Fantone 2007). This is a step in the right direction, as it more realistically communicates the limitations of data, information, and evidence. However, despite this softer language, reforms continue to, either implicitly or explicitly, rely on narrow concepts of policy knowledge and fail to offer insights into how various forms of policy knowledge may be useful across a range of decision making needs. This book will show that the challenges caused by narrow definitions of policy knowledge are exacerbated by a lack of attention to how decisions are made and the various ways data, information, and evidence can influence political decision making. One key factor that is often overlooked is the continuing importance of politics, even when relevant policy knowledge is readily available (Dubnick 2005; Radin 2002). Relying on a proverb of using policy knowledge to supplant politics—rather than considering political factors as an enduring component of decision making— creates a blind spot in understanding whether and how data, information, and evidence can be used. At the national level, political conflict arises in many forms. For example, in Congress, differences between political parties, the presence of multiple overlapping committees, and competition between the House of Representatives and the Senate all create unique differences in how policy decisions are made (Esterling 2004; Gilmour and Lewis 2006b). Outside of Congress, there is additional complexity resulting from the US separation-of-powers system as the legislative, executive, and judicial branches each influence policy in different ways (Bertelli and Lynn 2006). Evidence-based reforms typically assume that the legislative branch
14
S. PUTANSU
sets policy goals while the executive branch neutrally implements these goals. This assumption is an extension of Wilson’s (1887) “politicsadministration dichotomy” and underlies the evidence-based proverb. Scholarship in the field has shown repeatedly that this dichotomy does not exist in practice (Rosenbloom 1993). Further, the federalist system in the United States results in overlapping authorities at the federal, state, and local levels, which creates additional policy conflicts at the sub-national level (Durant 1999). These conflicts are more obvious than the political implications of policy knowledge. Specifically, decision makers also consider whether the data, information, and evidence given to them are influenced by politics. In doing so, according to Kathleen Hale (2011), they may consider whether the producer is a champion of a particular policy, a supporter, a challenger, or a neutral bystander. For strong supporters and challengers, decision makers may take related policy knowledge as intentionally biased or misleading. As evidence-based reforms are implemented and the narrow definition of policy knowledge becomes clear, they are commonly critiqued for attempting to apply a “one-size-fits-all” approach, and for their failure to consider other types and sources of data, information, and evidence. For example, Lindblom and Cohen in their 1979 text Usable Knowledge compared and contrasted the assumptions, values, and usefulness of “professional social inquiry” with that of the “ordinary knowledge” derived largely from experience and expertise. They showed that decision makers often relied on ordinary knowledge while valuing the more scientific sources only in limited circumstances, despite reformers’ and academics’ contentions that this latter form of policy knowledge was superior. This reveals another limitation of evidence-based reforms, which often looks to academic scholarship while ignoring research produced within the government, as detailed by Arnold Meltsner (1976) in Policy Analysts in the Bureaucracy. Evidence-based reforms implicitly assume unitary goals, easy measurement, simple models of decision making, and a deterministic role of policy knowledge (Radin 2006). These assumptions fail to consider the interconnected, overlapping, and often conflictual relationship among decision makers in a US policy context, which can lead to ambiguous, overlapping, and conflicting goals (Rittel and Webber 1973). As a result, the policy knowledge produced by evidence-based reforms tends to focus narrowly on individual programs, without considering interrelationships and tensions with other programs. They tend to rely on simplifying logic and
1
POLICY KNOWLEDGE, POLITICS …
15
parsimonious models of cause and effect, without contextual depth and consideration of alternative explanations. As a result, when political tensions over these complexities arise, the policy knowledge being produced fails to meaningfully contribute to the decision making process. Policy knowledge can improve the governance process by informing decisions about policy, identifying improvements in planning and implementation, and supporting economy, efficiency, effectiveness, equity, and ethics in governance. When appropriate data, information, and evidence is considered during decision making, it can lead to important and beneficial outcomes. For example, when the Cuyahoga River ignited into flames on June 22, 1969, evidence about water pollution and its causes supported the creation of environmental regulations through the Clean Water Act of 1972. Similarly, as ozone depletion was identified in the 1970s, evidence developed over the ensuing decade served as the basis for the Montreal Protocol, an international agreement to curb the use of chlorofluorocarbons (CFCs). There are numerous other examples from a wide variety of policy areas, including the information on traffic fatalities being used to support requirements for seat belts, information about savings preferences and practices supporting policies for opt-out 401K plans, and evidence on the lack of impact of anti-drug education programs leading to their revision and elimination. Given these examples, it is no surprise that politicians, researchers, journalists, and citizens all routinely call for increased use of data, information, and evidence in decision making to support better outcomes, lower costs, improved accountability, and better policy (Radin 2000a). This popular view of the objectivity and value of policy knowledge, combined with skepticism about “baser political issues,” may help to explain why reformers have not been pushed to address the political realities of government decision making and the limitations of the policy knowledge they have endorsed. Specifically, general acceptance of the potential value of data, information, and evidence is a barrier to a more nuanced discussion of when policy knowledge is valuable and why different types and sources may be of more or less value for a given decision. While the phrase “evidence-based policy” has been used to describe specific kinds of policy knowledge, it has also been used broadly and generally to refer to any call for increased use of data, information, and evidence in legislative, policy, and administrative decision making. In the former case, advocates push for a specific source or type of policy knowledge and generally do not acknowledge or incorporate lessons learned from
16
S. PUTANSU
past reforms. In the latter, they rely on a broader argument for avoiding politics by linking policy knowledge to decisions. This book will acknowledge the parallels between reforms and the value of different types of policy knowledge they have advocated for in order to more fully characterize the use of data, information, and evidence for government decision making. In doing so, it seeks to move from a general view of the value of policy knowledge to a more nuanced presentation of the multiple, varied values of different types and sources of data, information, and evidence. There are three unfortunate results of the poorly defined, contextually lacking, and unrealistically optimistic series of evidence-based reforms. First, lessons learned from each approach are often discarded or forgotten when a new reform is implemented, resulting in repeated expenditures with limited impact. Second, our understanding of the impact of policy knowledge reflects the piecemeal studies of each narrow definition. In many cases, we may perpetuate the belief that decision makers do not use data, information, and evidence, when in reality they may just be using a different type of policy knowledge or using it in a less direct way than anticipated (Moynihan 2008). Finally, reforms that require great expense only to produce policy knowledge that is not suitable for decision makers may reduce their support for data, information, and evidence produced by those reforms. This last impact is particularly troubling as it can result in a condemnation and defunding of government-supported efforts to expand the production and use of policy knowledge. For example, efforts to cut research funding for the National Science Foundation have often been linked to criticisms of unnecessary or frivolous research. Such criticisms are illustrative of the limits of the one-size-fits-all approach, as the rejection of information for one purpose is often framed as a rejection of its use for all purposes. For example, a grant in the 1960s to fund research on American Sign Language was widely criticized, but ultimately cemented as a legitimate language and improved educational and other opportunities for the deaf and hard of hearing. Taken together, these three results further illustrate the paradox that evidencebased reforms may actually have deleterious effects on the trust in and use of policy knowledge. Attempts to link policy knowledge and decision making have been hampered by a reliance on the evidence-based proverb that argues policy knowledge can and should supplant political factors. As described above, the tendency to adopt a “one-size-fits-all” approach to defining policy knowledge results in many scenarios where it is not meaningful to
1
POLICY KNOWLEDGE, POLITICS …
17
a specific purpose and therefore is not considered or used. These reforms produce data, information, and evidence that could be valuable to some decision making. However, they fail to identify appropriate uses for the policy knowledge they create, to consider and adequately value alternative sources, or to consider political factors. The result is a predictable failure to deliver on promises of a panacea for decision making. These failures increase the risk that advocacy for better-informed decision making will paradoxically result in the erosion of trust, and reduced use of, relevant data, information, and evidence. In Chapter 2, this book offers a typology of policy knowledge to address the first of these shortcomings. It then expands this typology to consider variations in the types and sources of policy knowledge and their potential uses. Taken together, this will demonstrate the impossibility of one type or source of information being a panacea for decision making. In Chapter 3, a nuanced decision making framework is be introduced to address the limitations of the evidence-based proverb. This framework offers a stronger lens for analyzing the competing and complementary roles of policy knowledge and political factors in the decision making process. Before turning to those models, the next section lays out the goals and approach that will be used throughout this text.
A Research Approach to Tackle the Evidence-Based Proverb This book systematically assesses and explains the progression of decisions leading to the creation, modification, and current structure of student loan programs under the Higher Education Act (HEA) and Title I Grants to States and Local Education Agencies under the Elementary and Secondary Education Act (ESEA). Specifically, it explores these two historical case studies to consider the types of policy knowledge that have been produced, considered, and used to influence major governance decisions. It will review various approaches that have been used to connect data, information, and evidence to decision making, while thoughtfully illustrating the importance of political context. These two cases illustrate the diversity of decision maker needs for policy knowledge, the importance of political context, and the different ways these opposing forces interact with and balance one another in the decision making process. This study will employ a detailed historical case study approach to demonstrate the conditions under which particular sources and types
18
S. PUTANSU
of policy knowledge can overcome some of the challenges identified in this chapter. This contextually sophisticated approach will enable future reformers to anticipate and account for these challenges in the design and assessment of future evidence-based reform efforts. This book acknowledges that all policy knowledge is not created equal, and that decision makers may require different levels of coverage and quality to support different kinds of decisions. In so doing, it provides a broader view of the range of data, information, and evidence that are potentially relevant to a decision, and insights into how decision makers may link that policy knowledge to specific purposes. This book will examine federal education policy to illustrate and explore the production, use, and impact of different types and sources of data, information, and evidence. It will explain whether policy knowledge gets used, how it is selected and used, and the nature of its impact on policy decisions. By examining the entire history of the two largest federal education programs—student loans under HEA and state grants for primary education from Title I of ESEA—this book will provide insights regarding how policy knowledge is developed, selected, distributed, and used in policy decisions. In 2015, at their 50th anniversary, both of these laws had passed their scheduled reauthorization date and had experienced several one-year extensions. Several aspects of both programs were subject to heated debate, both of which have included different sources and types of policy knowledge. In the case of student loans, the primary disagreements were related to appropriate performance metrics and accounting methods. In the case of Title I, debates were mostly focused on program purpose. By considering these two cases, this book will examine policy knowledge in a variety of forms, and the resulting findings and lessons will be useful in developing and understanding evidence-based reforms, improving their contextual sophistication, and better delineating realistic expectations for the role of policy knowledge in the decision making process. Case selection for this book reflects a number of methodological and logistical decisions. First, in order to enable consideration of two different programs over an extended time frame, the cases were limited to a single broad policy area. For this purpose, I chose programs within a single agency—the Department of Education (ED). There are several reasons for choosing ED. First, the department’s mission to foster the development of knowledge and enable students to make evidence-based decisions in a variety of situations in their everyday lives strongly parallels the focus
1
POLICY KNOWLEDGE, POLITICS …
19
of this study. Further, a tradition of local control in education provides a valuable context for understanding overlapping, ambiguous, and conflictual goals, and their impact on the use of policy knowledge. The decision to focus on ED was also supported by a personal interest in the value of education, which is an area of critical importance in the United States. It ultimately impacts the ability of voters to effectively make democracy work, prepares students for employment, impacts the growth and stability of the US economy, and teaches basic skills that are critical to the general well-being of our society. Finally, and most importantly, personal experience studying education policy provided me with a solid foundation of knowledge on which to examine the cases and increased access to experts and practitioners for interviews and related data collection. To ensure that different sets of interests would be represented in the analysis, both of the key federal laws on primary education are represented, with one case chosen from ESEA, and the other from HEA. Each of the cases has a distinct set of issues that must be confronted by students of both decision making and policy knowledge. These two laws were both passed in 1965, the same year that the first government-wide performance management reform was implemented, and so provide a strong example of programs designed around the same time modern evidence-based advocates started to advance reforms to improve governance decisions. Further, a complete history of these two programs provides an opportunity to assess decisions in a single policy area over an adequate period of time to observe shifts in the political landscape and developments in related data, information, and evidence. By examining the histories and current status of these programs, this study: (1) compares the production and utilization of policy knowledge in the two cases, (2) explores the impact of political factors on decision making, and (3) develops an understanding of the conditions under which policy knowledge is more or less likely to be used. This extended exploration is sufficient to include policy knowledge of varied coverage and quality, in order to better understand the impact of each. This book will examine these two cases using a comparative case study methodology rooted in historical narrative (Agranoff and Radin 1991; Yin 2003). The analysis of these cases builds on two established literatures. The first of these considers the evidence-based proverb, draws on a literature about the production and use of policy knowledge, and provides an approach to considering data, information, and evidence that goes well
20
S. PUTANSU
beyond the typical “one-size-fits-all.” The second considers political decision making and provides a framework for considering how policy knowledge and political factors each contribute to the decision making process, rather than presuming that policy knowledge is a replacement for politics. This study will apply the lessons learned from literature to explore the two cases, compare them, and consider future avenues for evidence-based governance. Consideration of these two broad areas of research requires a synthesis of a wide range of literatures from cognate fields, including public administration, management, policy analysis, evaluation, and political science. The literature has individually examined the role of policy knowledge in decision making from a variety of perspectives, using an assortment of methods, and with a variety of conclusions. Studies from public administration and management tend to focus on the design of measurement systems and implementation of specific evidence-based reforms (Frederickson and Frederickson 2006; GAO 2012; Heinrich 2002; Moynihan 2004, 2006). Policy analysis, evaluation, and education research have focused on methodological approaches and specific policy outcomes (Jones 1983; Nutley et al. 2000; Radin 2013; Stufflebeam 1969). Political science has focused primarily on the decision making process and setting of policy goals (Key 1940; Lindblom 1959; Schick 2008; Wildavsky 1964). What all these literatures have in common is that they tend to focus on narrowly defined time periods and single reform interventions. In doing so, they limit their examinations to the impact of a single policy intervention, a specific evidence-based reform, a particular source of policy knowledge, or a specific legislative, policy, or administrative decision. This analysis will make clear the limitations of historical evidence-based reform efforts and further develop our understanding of the role of policy knowledge in decision making. The historical approach used in this book fills a critical gap in the practical understanding of evidence-based decision making governance by enabling an examination of the evolution of policy interventions. This approach enables an examination of multiple evidence-based reforms, a more complete definition of policy knowledge, and a consideration of successive related political decisions. In so doing, it applies the strategies developed across several disciplines to more completely understand the relationship between contextual issues, policy knowledge, and political decision making. The results of this broad exploration will clearly demonstrate that the best data, information, and evidence are not defined solely by a hierarchy of rigor. While that is one
1
POLICY KNOWLEDGE, POLITICS …
21
aspect of value, the true worth of policy knowledge also depends on its relevance to a specific need and usefulness within a given political context. Given that there are multiple types and sources of policy knowledge, and that they support a variety of purposes, related research must take a nuanced approach that considers these factors. This book will challenge the evidence-based proverb and its unnecessarily constricted view of policy knowledge to supplant politics. It will show that some types of data, information, and evidence are better suited to certain decisions by providing a detailed explanation of the role of policy knowledge within the political context. Although this study covers major decisions—reauthorizations, policy changes, and major budget decisions (those that include substantive policy changes)—throughout the life of both programs, it is not intended to serve as a complete history of all decisions related to these programs. In order to maintain a tenable scope and a consistent level of analysis, decisions made by individual managers or without written records are not included. I have made a significant effort to review source documents when characterizing the intents and opinions of individual decision makers, but these documents are necessarily imperfect reflections of the full range of motivations they may experience. It continues that, while I have endeavored to identify policy knowledge that is consistent with these historical records, it is not always possible to trace individual ideas to their source, as data, information, and evidence are shared formally and informally, in written and verbal exchanges which are often not recorded. The historical case study approach mitigates these limitations because it allows for a consideration of policy knowledge in a variety of scenarios and with a variety of decision makers with multiple observations over time. This also provides an opportunity to examine both the creation of policy knowledge at one point in time and changes in its influence and use over time. The insights and lessons learned from these cases can inform the theory and hypotheses to be explored in future scholarship and in the development of future efforts to enhance the use of data, information, and evidence in decision making.
Organization of the Book This book will provide detailed explanations of the factors that have influenced whether and how policy knowledge has impacted major decisions for the selected cases. The text is divided into three parts. The first, Confronting the Evidence-Based Proverb, will consider how reform efforts prioritize data, information, and evidence. It will show how different sources
22
S. PUTANSU
of policy knowledge, the purposes they support, and the political context impact decision outcomes. The second, Politics and Policy Knowledge in Federal Education, will use historical analysis of the two education cases to provide a detailed examination of the political decision making process and the role of policy knowledge therein. The third and final part, Conclusions and Implications for Evidence-Based Governance, will draw lessons from the education cases about the different contextual factors that impact whether policy knowledge is used, what sources are used, and how. It will conclude the volume by exploring the broader implications of these findings for current and future evidence-based reformers. The first section summarizes what is known about the role of data, information, and evidence in decision making and provides the foundation for the historical case studies by reviewing what is known about policy knowledge and decision making processes. This chapter has discussed the promises of panacea from evidence-based reformers and their reliance on the evidence-based proverb. Chapter 2 delves further into promises of panacea from evidence-based reforms and shows how different purposes of policy knowledge make these promises impossible to fulfill. Chapter 3 focuses on the proverb of using policy knowledge to supplant politics, showing how established theories of decision making largely consider politics and policy knowledge as substitutes, rather than as dually important inputs into decision making. Chapter 2 will examine common limitations in the definition of policy knowledge and narrow expectations for its use. First, it will introduce a typology of policy knowledge that reflects different levels of coverage and rigor. Then, it will discuss the range of purposes that policy knowledge may serve, along with a range of ways it may influence the decision making process. The chapter will conclude by introducing a bureaucratic typology, developed by James Q. Wilson (1989), which provides insight into how different types and purposes of policy knowledge are encouraged or discouraged based on the design and goals of specific programs. From this framework, a core argument of the book is introduced: The design and purpose of government programs have a direct and lasting impact on the need for, ability to gather, and ultimately use different sources of data, information, and evidence. More specifically, the ability to observe program processes and outcomes also increases the ability to gather comprehensive observations, provide quality analyses, and apply them directly to relevant decisions: “one-size-fits-all” approaches to policy knowledge,
1
POLICY KNOWLEDGE, POLITICS …
23
in turn, are unlikely to deliver on their promises to serve as a panacea for decision making. Chapter 3 examines government decision making in more depth in order to understand how the various source and types of data, information, and evidence impact the process. This chapter challenges the evidence-based proverb by focusing on the dual roles of policy knowledge and political factors in governance decisions. The chapter will consider whether and how policy knowledge impacts decisions, focusing primarily on two competing models: rational-comprehensive and politicalincremental. The chapter will present the strengths and limitations of these models and will focus on the role of policy knowledge in each. It will further explore attempts to refine these models and show that even these refinements fail to fully escape the evidence-based proverb. As a result, these models do not adequately explain the relative strengths and weaknesses of politics and policy knowledge in decision making. Chapter 3 will conclude by outlining an alternative model for decision making, which deconstructs these models into two continua, one from rational to political and the other from incremental to comprehensive. A related framework of politics and policy knowledge is integrated, showing that the “Four-I’s” of Ideology, Interests, Institutions, and Information (Policy Knowledge) help to explain where specific decisions fall along these continua (Weiss 1995). From this model, a second core argument of the book is introduced: While the relative influence of policy knowledge and politics is rooted in program design and purpose, shifts in political factors and policy knowledge offer opportunities to shift that balance, both immediately and for the longer term. Specifically, changes in the contextual environment may lead to decisions that diverge from expectations set by program design and purpose. Further, to the extent that these decisions result in changes to program design and purpose, they may also lead to longer term rebalancing of the relative influence of policy knowledge and politics in future program decisions. The first part concludes with Chapter 4, which briefly examines evidence-based reforms focused on performance, characterized by Beryl Radin (2006) as the “performance movement.” Across 50 years and 10 separate reforms, the performance movement stands out as a prime example of repeated evidence-based reforms (see also Moynihan 2008; Pollitt 2006). Examination of these reforms enables a robust comparison of the purposes and uses of policy knowledge. The chapter will then show how other evidence-based reforms, including those currently championed as
24
S. PUTANSU
evidence-based policy, have made similar promises. It will show how this cycle of iteratively broken promises of a panacea for decision making actually erodes trust in policy knowledge. The second part of the book explores the nexus of policy knowledge and decision making in the context of federal education policy and contains historical information and original research related to the two largest (in terms of expenditures) education programs in the United States. Chapter 5 will set the foundations for the two case studies by detailing the historical context of federal education policy from the colonial era through the modern era of education, which began in 1965 with the passage of HEA and ESEA. The chapter will set out the ideological landscape, institutional norms, interest group dynamics, and information available that represent the Four-I model and show how they define education as a policy area. It will explain how these Four-I’s differed in discussions of higher education as compared to elementary and secondary education. In the former case, low levels of political conflict and generally shared policy knowledge contributed to clearly specified goals and a program design with easily observable processes. In the latter, institutional and ideological conflict, combined with limited policy knowledge, contributed to unclear and ambiguous specification of goals coupled with fragmented and difficult to observe program design. In James Q. Wilson’s specification of program types, this would define student loans as a production program, while Title I would be classified as a coping program. Chapter 5 will then turn to establish expectations for the relative importance of the Four-I’s in decision making relevant to these two programs. Specifically, it will test the hypothesis that production organizations with highly observable designs enable more rigorous collection of data, information, and evidence, while clear goals enable clear, highquality interpretations of that evidence. As a result, the expectation is that policy knowledge will be more easily created, will typically be relevant to common policy decisions, and will tend to play a strong role in decision making. In contrast, the expectations for coping organizations are that high-quality data, information, and evidence are more difficult to produce, are useful for a smaller subset of decisions, and are therefore less central to decision making. Chapter 6 will present the case study for student loan programs, and Chapter 7 will do the same for Title I of ESEA. Each of these cases will begin with a brief summary of the passage of their parent acts and will describe all the titles of each act. Chapter 6 examines the student loan
1
POLICY KNOWLEDGE, POLITICS …
25
programs grouped into five major eras in which the program design and goals remained relatively stable. The 21 major program decisions made throughout the five eras were primarily focused on tweaks to specific implementation practices. Chapter 7 divides the historical narrative of Title I into five eras. In contrast to student loans, the 17 decisions examined in Title I reflect cyclical shifts from targeting program design in one era to focusing on program goals in another. The third and final section synthesizes the evidence from Section Two to comprehensively assess the role of data, information, and evidence in decision making for federal education programs. It explains how policy knowledge of different types and sources was utilized for different purposes in the decision making process, and examines their specific impact on that process. Lessons learned from this examination are used to develop suggestions for future efforts to understand and promote policy knowledge in the political process. Chapter 8 compares the findings from Chapters 6 and 7 to explain conditions that promote policy knowledge as well as those that diminish it. This chapter shows that consideration of all the different sources and types of data, information, and evidence, along with political factors, is required to understand how particular decisions are made. It also demonstrates that program type is a useful predictor for anticipating which types and sources of policy knowledge are likely to be utilized and impacts they are likely to have. In so doing, it demonstrates that the narrow definitions of data, information, and evidence employed by evidence-based reforms have guaranteed their limited usefulness to a wide array of complex governance decisions. Chapter 9 translates the findings from this book into actionable advice for future evidence-based reforms. The depth of analysis presented in this book is instructive for both the decision makers searching for and using data, information, and evidence to support their decisions, and for the analysts, researchers, academics, pundits, think tanks, journalists, and other producers and communicators of policy knowledge. Chapter 9 will also reflect on organized efforts to link research and practice through the National Academy of Public Administration, professional organizations, and other collaborative efforts among scholars. The chapter identifies principles that evidence-based reformers should consider as they design, promote, and assess the success or failure of existing and future evidencebased reforms.
26
S. PUTANSU
By considering these principles and other lessons learned through this study, evidence-based reformers, researchers, and decision makers can better match reform efforts to relevant decisions. By doing so, they can increase the likelihood that relevant policy knowledge will be integrated with political factors into a wider variety of policy areas and decision contexts. Evidence-based reformers can more broadly define their calls for policy knowledge and better consider contextual needs for data, information, and evidence. Researchers can better understand the strengths and limitations of the policy knowledge they provide to decision makers, while decision makers can recognize contextual complexities and the range of policy knowledge that may be used to improve governance. Further, this book provides insights into the wide variety of factors that impact the likelihood of linking policy knowledge and politics to better inform governance decisions. This represents a step away from a naïve, simplistic view of policy knowledge and toward a more nuanced understanding of what it means for decisions to incorporate data, information, and evidence. The book concludes with a research agenda designed to continue advancing this important topic of study and to further enhance the role of policy knowledge in the policy process.
References Ackoff, R. L. (1989). From data to wisdom. Journal of Applied Systems Analysis, 16, 3–9. Agranoff, R., & Radin, B. A. (1991). The comparative case study approach in public administration. Research in Public Administration, 1(1), 203–231. Behn, R. D. (2003). Why measure performance? Different purposes require different measures. Public Administration Review, 63(5), 586–606. Bertelli, A. M., & Lynn, L. E. (2006). Madison’s managers: Public administration and the constitution. Baltimore: Johns Hopkins University Press. Campbell, D. T. (1969). Reforms as experiments. American Psychologist, 24(4), 409. Deaton, A., & Cartwright, N. (2018). Understanding and misunderstanding randomized controlled trials. Social Science and Medicine, 210, 2–21. Dubnick, M. (2005). Accountability and the promise of performance: In search of the mechanisms. Public Performance & Management Review, 28(3), 376– 417. Durant, R. F. (1999). The political economy of results-oriented management in the “neoadministrative state”: Lessons from the MCDHHS experience. American Review of Public Administration, 29(4), 307–331.
1
POLICY KNOWLEDGE, POLITICS …
27
Eddy, D. M. (1984). Variations in physician practice: The role of uncertainty. Health Affairs, 3(2), 74–89. Eddy, D. M. (1990). Practice policies: Where do they come from? JAMA, 263(9), 1265–1275. Esterling, K. M. (2004). The political economy of expertise: Information and efficiency in American national politics. Ann Arbor: University of Michigan Press. Frederickson, D. G., & Frederickson, H. G. (2006). Measuring the performance of the hollow state. Washington, DC: Georgetown University Press. Gilmour, J. B., & Lewis, D. E. (2006a). Does performance budgeting work? An examination of OMB’s PART scores. Public Administration Review, 66(5), 742–752. Gilmour, J. B., & Lewis, D. E. (2006b). Assessing performance assessment for budgeting: The influence of politics, performance, and program size in FY 2005. Journal of Public Administration Research and Theory, 16(2), 169– 186. Gordon, J. (1994). A critique of reinventing government. Spectrum, the Public Policy Journal of the Kennedy School of Government, Harvard University. Government Accountability Office (GAO). (2004). Performance budgeting: OMB’s program assessment rating tool presents opportunities and challenges for budget and performance integration. GAO-04-439T. Washington, DC: GAO. Government Accountability Office (GAO). (2012). A guide for using the GPRA Modernization Act to help inform congressional decision making. GAO-12621SP. Washington, DC: GAO. Hale, K. (2011). How information matters: Networks and public policy innovation. Washington, DC: Georgetown University Press. Haskins, R., & Baron, J. (2011). Building the connection between policy and evidence. In Using evidence to improve social policy and practice (p. 25). Hatry, H. P. (2006). Performance measurement: Getting results (2nd ed.). Washington, DC: Urban Institute. Heinrich, C. J. (2002). Outcomes–based performance management in the public sector: Implications for government accountability and effectiveness. Public Administration Review, 62(6), 712–725. Hird, J. A. (2005). Policy analysis for what? The effectiveness of nonpartisan policy research organizations. Policy Studies Journal, 33(1), 83–105. Hoskins, H. D. (1990). Clinical decision making: Theory vs practice. JAMA, 264(12), 1533–1534. Jones, W. J. (1983). Can evaluations influence programs? The case of compensatory education. Journal of Policy Analysis and Management, 2(2), 174–184. Julnes, G. (2007). Promoting evidence-informed governance: Lessons from evaluation. Public Performance & Management Review, 30(4), 550–573.
28
S. PUTANSU
Julnes, P. D. L., & Holzer, M. (2001). Promoting the utilization of performance measures in public organizations: An empirical study of factors affecting adoption and implementation. Public Administration Review, 61(6), 693–708. Kettl, D. F. (2002). The transformation of governance: Public administration for twenty-first century America. Baltimore, MD: Johns Hopkins University Press. Key, V. O. (1940). The lack of a budgetary theory. American Political Science Review, 34(6), 1137–1144. Lindblom, C. E. (1959). The science of “muddling through.” Public Administration Review, 19(1), 79–88. Lindblom, C. E., & Cohen, D. K. (1979). Usable knowledge: Social science and social problem solving (Vol. 21). New Haven: Yale University Press. Madison, J. (1787). Federalist Paper No. 10, November 22, 1787: 1787–1788. Majone, G. (1989). Evidence, argument, and persuasion in the policy process. New Haven: Yale University Press. Maynard, R. A. (2006). Presidential address: Evidence-based decision making— What will it take for decision makers to care? Journal of Policy Analysis and Management, 25(2), 249–265. Meltsner, A. J. (1976). Policy analysts in the bureaucracy. Berkeley: University of California Press. Meyers, R. T. (2017). Is the US Congress an insurmountable obstacle to any “far sighted conception of budgeting”? Public Budgeting & Finance, 37 (4), 5–24. Moynihan, D. P. (2004). Why and how do state governments adopt and implement “managing for results” reforms? Journal of Public Administration Research and Theory, 15(2), 219–243. Moynihan, D. P. (2006). Managing for results in state government: Evaluating a decade of reform. Public Administration Review, 66(1), 77–89. Moynihan, D. P. (2008). The dynamics of performance management: Constructing information and reform. Washington, DC: Georgetown University Press. Nutley, S. M., Davies, T. O., & Smith, P. C. (Eds.). (2000). What works?: Evidence-based policy and practice in public services. Cambridge: MIT Press. Olsen, R., & Levy, D. (2004). Program performance and the president’s budget: Do OMB’s PART scores really matter. In Association for Policy Analysis and Management’s 2004 Annual Research Conference. Osborne, D., & Gaebler, T. (1992). Reinventing government. Reading, MA: Addison-Wesley. Pollitt, C. (2005, August 31–September 3). Performance measurement and performance management are key elements within most definitions of the NPM. Workshop on Quality and Productivity in the Public Sector European Group for Public Administration Annual Conference (EGPA), Berne.
1
POLICY KNOWLEDGE, POLITICS …
29
Pollitt, C. (2006). Performance management in practice: A comparative study of executive agencies. Journal of Public Administration Research and Theory, 16(1), 25–44. Posner, P. L., & Fantone, D. M. (2007). Assessing federal program performance: Observations on the US Office of Management and Budget’s Program Assessment Rating Tool and its use in the budget process. Public Performance & Management Review, 30(3), 351–368. Radin, B. A. (2000a). The government performance and results act and the tradition of federal management reform: Square pegs in round holes? Journal of Public Administration Research and Theory, 10(1), 111–135. Radin, B. A. (2000b). Beyond Machiavelli: Policy analysis comes of age. Washington, DC: Georgetown University Press. Radin, B. A. (2002). The accountable juggler: The art of leadership in a federal agency. Washington, DC: CQ Press. Radin, B. A. (2006). Challenging the performance movement. Washington, DC: Georgetown University Press. Radin, B. A. (2012). Federal management reform in a world of contradictions. Washington, DC: Georgetown University Press. Radin, B. A. (2013). Beyond Machiavelli: Policy analysis reaches midlife. Washington, DC: Georgetown University Press. Riccucci, N. M. (2010). Public administration: Traditions of inquiry and philosophies of knowledge. Washington, DC: Georgetown University Press. Rittel, H. W., & Webber, M. M. (1973). Dilemmas in a general theory of planning. Policy Sciences, 4(2), 155–169. Rivlin, A. (1971). Systematic thinking for social action. Washington, DC: The Brookings Institution. Robinson, M. (Ed.). (2007). Performance budgeting: Linking funding and results. Basingstoke: Springer. Rosenbloom, D. H. (1993). Have an administrative Rx? Don’t forget the politics! Public Administration Review, 53(6), 503–507. Salamon, L. M. (2002). The tools of government: A guide to the new governance. Oxford: Oxford University Press. Schick, A. (2008). The federal budget (3rd ed.). Washington, DC: Brookings. Shulock, N. (1999). The paradox of policy analysis: If it is not used, why do we produce so much of it? Journal of Policy Analysis and Management, 18(2), 226–244. Simon, H. (1946). The proverbs of administration. Public Administration Review, 6(1), 53–67. Smyrl, M., & Genieys, W. (2016). Elites, ideas, and the evolution of public policy. Springer. Stewart, J. (2009). Public policy values. Springer. Stufflebeam, D. L. (1969). Evaluation as enlightenment for decision-making. In W. H. Beatty (Ed.), Improving educational assessment & an inventory of
30
S. PUTANSU
measures of affective behavior (pp. 41–73). Washington, DC: The Association for Supervision and Curriculum Development, NEA. Weiss, C. H. (1979). The many meanings of research utilization. Public Administration Review, 39(5), 426–431. Weiss, C. H. (1989). Congressional committees as users of analysis. Journal of Policy Analysis and Management, 8(3), 411–431. Weiss, C. H. (1995). The four I’s of school reform: How interests, ideology, information, and institutions affect teachers and principals. Harvard Educational Review, 65(4), 571–593. Weiss, C. H. (2001). What kind of evidence in evidence-based policy? Paper presented at the third international evidence-based policies and indicator systems conference, pp. 284–291. Durham, UK: CEM centre, University of Durham. Weiss, C. H. (2002). What to do until the random assigner comes? In F. Mosteller & R. F. Boruch (Eds.), Evidence matters: Randomized trials in education research (pp. 198–233). Washington, DC: Brookings Institution Press. West, W. F. (2011). Program budgeting and the performance movement: The Elusive quest for the efficiency in government. Washington, DC: Georgetown University Press. Wildavsky, A. B. (1964). Politics of the budgetary process. Boston: Little, Brown. Yin, R. (2003). Applications of case study research. Thousand Oaks, CA: Sage.
CHAPTER 2
No Panacea: The Purposes, Uses, and Limitations of Policy Knowledge
Advocates for evidence-based governance have generally made broad calls for more policy knowledge, or have championed a narrow focus on specific data, information, or evidence. These calls for reform have also commonly featured overly broad promises that the resulting policy knowledge will be a panacea for decision making. This book argues that broad agreement on the importance of policy knowledge enables reformers to overlook important differences in the purposes, uses, and limitations of this knowledge in specific decision making contexts. Further, and counter to the intent of evidence-based reformers, this has actually hindered efforts to increase the role of data, information, and evidence in decision making by ignoring and excluding alternative sources. By emphasizing specific forms of policy knowledge and implicitly devaluing others, evidence-based reforms have contributed to recent retreats from political support for various forms of data, information, and evidence. This includes attempts to defund research grants, restrictions on certain types of data collection, and increasing reliance on “ordinary knowledge” to support decision making (Lindblom and Cohen 1979). This chapter introduces a typology of policy knowledge rooted in the quality of data, information, and evidence. Then, it adds a dimension of purpose, which considers the different aspects of governance that can be focal points for policy knowledge. Finally, it considers the development of policy knowledge as a process, through which decisions about purpose, coverage, and quality shape the production of data, information, and evidence. These additional dimensions further demonstrate the need © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_2
31
32
S. PUTANSU
for evidence-based reformers to be explicit about the qualities and limitations of the data, information, and evidence they produce, and the relevance of alternative types and sources for specific purposes. The chapter then considers the different ways that policy knowledge can impact decisions. Calls for evidence-based policy have promised a panacea for decision making, but this analysis shows that three limitations in the treatment of policy knowledge have consistently limited utility of the reforms. Specifically, reforms have relied on narrow definitions of policy knowledge, failed to consider different purposes it can support, and overlooked the different ways that data, information, and evidence are used. This discussion paves the way for more detailed consideration, in Chapter 3, of the role of policy knowledge, vis-a-vis political forces, in the decision making process.
Defining a Typology of Policy Knowledge To overcome the evidence-based proverb, any examination of the role that policy knowledge plays in the political decision making process must rely on a more complete and nuanced model of data, information, and evidence. This chapter introduces a four-part typology of data, information, and evidence that allows for more complete cataloging, understanding, and examination of the different ways that policy knowledge is produced, considered by decision makers, and ultimately used. Rather than attempt to establish a singular hierarchy of policy knowledge rooted in methodological approaches, this typology is rooted in a consideration of the quality of policy knowledge, as defined by varying levels of coverage and rigor.1 In this context, coverage refers to the representativeness of the data, information, or evidence being considered, and ranges from an individual experience, on the low end, to a complete set of all relevant observations on the high end. Rigor refers to the analytical sufficiency of the policy knowledge, in terms of whether it is valid and reliable for an intended purpose, which generally ranges from ad hoc on the low end, to more systematic on the high end. Rigor is used as a summative term throughout this text to represent the range of practices and methods used to improve the reliability and validity of policy knowledge, including but 1 This typology borrows from, integrates, and updates a variety of definitions offered in Feldman (1989), Lindblom and Cohen (1979), Majone (1989), Shulock (1999), and Weiss (1977).
2
NO PANACEA: THE PURPOSES, USES …
33
not limited to greater qualitative depth, more structured evidence gathering, inclusion of additional quantitative variables, or higher levels of experimental control. As shown in Fig. 2.1, dividing policy knowledge based on coverage and rigor creates four quadrants of data, information, and evidence. Before turning to the implications of this model, an illustrative discussion of the elements of policy knowledge in each quadrant is instructive. The lower left quadrant illustrates a relatively ad hoc approach to generating policy knowledge, with a small and unrepresentative level of coverage. A single data point (i.e., an observation) may be that an agency hires a new employee. Information (i.e., detail) about that person may include characteristics of this employee, such as gender, race, and education. This becomes evidence when an association between the data and information (i.e., an assertion), such as “we hired her because she met the education requirements for the position,” is made. The top left quadrant illustrates higher coverage. For example, this might include data on the last 20 hires (i.e., collection), with information (i.e., description) that presents the proportions of those in each of the demographic categories, and evidence (i.e., correlation) that shows whether education is associated with hiring decisions.
Fig. 2.1 Policy knowledge types, coverage, and quality
34
S. PUTANSU
Using a systematic analytical approach introduces more rigor into the production of data, information, and evidence, which enables the data to be used for more complex purposes. The bottom right quadrant illustrates the addition of quality to the low coverage examples above. Specifically, if the agency assessed the new hire’s education by looking at their resume, this would be of lower rigor than if they confirmed this with an official transcript (i.e., fact). Similarly, the agency may have details collected about gender, race, and education, but they could more rigorously review other relevant characteristics about the individual (i.e., case), including past work experience, references, and so on. Finally, by systematically assessing how each of these characteristics was valued in the hiring decisions, they may develop better evidence (i.e., explanation) of what drove the hiring decision. The rigor dimension can also be added to the higher coverage scenarios, as illustrated in the top right quadrant. When a collection is representative of all observations (i.e., population), it provides a more complete picture of the data. Similarly, having information about the full population (i.e., summary) allows for more thorough descriptive analyses. Finally, when researchers systematically consider potential correlations (i.e., model), they enable an assessment of strength of those relationships while controlling for other factors. For example, the company may review data on all applicants and hires, may compare informational summaries of the demographic characteristics of each, and may develop an evidentiary model to determine the independent associations between those demographic characteristics and hiring decisions. Similar to evidence-based reforms, this typology suggests variation in the relative value of different forms of policy knowledge (high coverage, high rigor policy knowledge should generally be trusted more than low coverage, low rigor). In practice, this means that an individual assertion of education as the reason an employee was hired should be overruled by a detailed explanation rooted in a full case study that offers contradictory evidence, by analysis that shows a negative correlation between education and hiring, or by a model that shows education has no impact, once other factors are controlled for. However, in contrast to evidence-based reforms, this typology does not presume that there is one best type of evidence for all decisions, but rather acknowledges that policy knowledge in all four quadrants could be sufficient and appropriate for specific decision making purposes.
2
NO PANACEA: THE PURPOSES, USES …
35
The appropriate coverage of policy knowledge is rooted in the scope of the decision at hand. For example, government investigations into criminal activity most appropriately focus only on “persons of interest” related to the alleged crime, while the decennial census is required to consider all people living in the United States. Similarly, the sufficiency of rigor of policy knowledge depends on its intended purpose. For example, members of Congress who receive a high number of ad hoc complaints describing a problem may consider that sufficient for them to attempt to step in and serve their constituents through individual casework, while they may seek more systematic and rigorous summary information before pursuing more formal oversight actions. Throughout this text, the typology offered above is used both to comprehensively classify different types and sources of data, information, and evidence presented throughout the book, and to offer deeper insights into whether and how policy knowledge has been used to support specific governance decisions.
The Purposes and Creation of Policy Knowledge Decision making can be understood and improved through detailed understanding of the various sources, types, and uses of policy knowledge (Weiss 1977). Policy and decision making literatures offer an abundance of factors that can influence whether and how individual decision makers receive or observe policy knowledge, whether and how they interpret it, and how this interpretation impacts their perspectives and strategic approach to decision making. This section explores the purposes of policy knowledge and link these purposes to the typology introduced in the previous section. It then shows how the generation of policy knowledge can identify and reinforce these links. Evidence-based reforms are founded on the belief that data, information, and evidence can support decision makers and lead to better outcomes. However, these reforms often do little to detail the kinds of decisions they can support, instead promising a panacea for decision making. In the context of governmental decision making, policy knowledge needs to be provided to several different players in the decision making process: (1) managers and employees can use it to improve the way they function, (2) Congress and the President can use it to improve policy decisions, or (3) citizens can use it to decide their preferences for a particular mix of service provisions and mode of service delivery. For example, an agency
36
S. PUTANSU
department head may be interested in program effectiveness, staff distribution, and cost comparisons, a bureau head may be more concerned with outcomes, and line managers may be focused on inputs and processes. These differences are copacetic with the long-accepted axiom that, in governance, where you stand is where you sit. Philip Joyce (2005) noted that different actors may rely on different forms of policy knowledge about programs, as they are participating in different activities and working toward different final products. Donald Moynihan (2008) provides additional support for this finding, showing that individual roles impact the selection and use of data, information, and evidence. In line with these multiple perspectives, scholars and practitioners have identified a broad array of potential uses of policy knowledge in decision making that vary in both number and detail. They include forty-four potential uses offered by van Dooren (2004), eleven major uses discussed by Hatry (2006), eight managerial uses enumerated by Behn (2003), and four types of managing performance illustrated by Bouckaert and Halligan (2008), among others. In other words, policy knowledge can help improve various aspects of decision making, but only when it is aligned with the needs of decision makers. Further, others note various ways that policy knowledge can be misused, including in the commissioning of work, the process for developing it, or in manipulation of the results (Alkin and King 2017). The choice of what to study has important implications for whether policy knowledge can provide a complete and accurate understanding of a policy or program. For example, if all data, information, and evidence about a program were to focus on economy—the total cost of the policy—then decision makers would not have a way to understand whether the policy effectively achieved goals or resulted in equitable outcomes, or whether these outcomes were being produced efficiently. All of these choices of selection and prioritization mean that the range of policy knowledge available for a given policy, its processes, and its results will be dependent on the interests, priorities, and independence of knowledge producers. This examination of different uses of policy knowledge obviates the impossibility of applying a one-size-fits-all approach to evidence-based policy. No single source of policy knowledge is likely to produce data, information, and evidence reflecting a level of coverage and quality that fulfills the wide array of perspectives and purposes needed for decision
2
NO PANACEA: THE PURPOSES, USES …
37
making. Instead, it shows that when decisions are being made, considering how data, information, and evidence contribute to specific purposes will be critical to ensuring that policy knowledge meets the needs of decision makers. This chapter now turns to the process of creating policy knowledge, in order to show how purposes and needs can be integrated into the development of data, information, and evidence. There are four broad activities required to develop policy knowledge. First, there must be a specification of program goals. Next, relevant data must be collected. Then, data must be analyzed, described, and interpreted in order to turn it into information and evidence. Finally, the resulting policy knowledge must be communicated to decision makers. Production of policy knowledge begins with decisions about what to measure. Producers of policy knowledge, in determining their scope and research questions, must decide where to invest limited resources in order to serve one or more decision making needs. This includes deciding which policies or programs to review, what processes to focus on, and which kinds of results to measure. In generating policy knowledge, any policy (or combination of policies) may be reviewed, and evaluations of those programs may center on clarification of objectives or measures of public opinion, inputs and program costs, implementation, outcomes, or impacts. Goals can be stated in general (strategic) or specific (actionable) terms. For example, either a broad goal of transparency or a more specific goal that records should be publicly available online may be specified. Goals can also have varied timelines, and may be short-term, mediumterm, or long-term. In some cases, efforts to develop policy knowledge may consider all three timeframes, with intermediate outcomes identified at each stage. Finally, goals might be specified at different institutional levels. For example, Congress might set a goal of a balanced budget that impacts the entire federal government, an agency might set a goal of implementing new systems across the entire organization, and a program might set goals about the provision of its specific services. Once the policy and relevant aspects are chosen, the review may proceed to analyze one or more kinds of results, including economy, effectiveness, efficiency, and equity, which may refer to distributional equity, outcome disparities, procedural fairness, and process equity (National Academy of Public Administration 2000). These approaches may consider broad, macro-level uses, such as goal setting, others focus on meso-level management issues, and others at micro-level details of implementation
38
S. PUTANSU
(Roberts 2019). These multiple goals reflect the desires of multiple actors in the decision making process, as different actors have different theories about the programs, understandings of cause-effect relationships, and prioritization of goals. The multitude of ways to specify goals, as described in Table 2.1, highlights the first potential challenge for evidence-based reforms. In order to sort through the wide array of different foci and analytical considerations, stakeholder buy-in is critical. Evidence-based reforms may profess to treat all goals equally, try to prioritize them independently, or seek to prioritize them through interaction with stakeholders, but if the relative importance of goals is disputed, all the resulting policy knowledge may be called into question (Newcomer 2007; Julnes and Holzer 2001; Posner and Fantone 2007; Pollitt 2006; Melkers and Willoughby 2005). In the legislative context, repeated findings suggest that evidence-based reforms are mistrusted, not valued, or completely ignored by decision makers, due in large part to their exclusion from the process of setting goals and metrics (Newcomer 2007; Pollitt 2006; Posner and Fantone 2007). This is true at the state level (Melkers and Willoughby 2005) as well as in a federal context (Posner and Fantone 2007). The enumeration of goals may also suffer from a lack of strategic focus (Posner and Fantone 2007; Joyce 2005), from insufficient specificity to be actionable (GAO 2005; Behn 2003), from inappropriate timelines (Newcomer 2007), or from a focus on the wrong institutional level (Behn 2002, 2003; Pollitt 2006; Talbot 2005; Barnow 1992). Finally, one or more stated goals may be in conflict with others (Wildavsky 1974; Davis et al. 1966). Beyond this, different Table 2.1 Example dimensions of policy knowledge use Dimension
Aspects that policy knowledge might focus on
What to study
Government-wide activities, entire policy areas, one or more programs, specific implementation practices Macro, Micro, and Meso level goals and objectives Prioritization of goals, management, implementation, impacted populations Inputs, outputs, outcomes, impacts History, public opinion/support, economy, effectiveness, efficiency, social equity Policy makers, managers, implementers, oversight bodies, interest groups, the general public
Policy levels Program levels Operations Analytical focus Users
2
NO PANACEA: THE PURPOSES, USES …
39
actors in the process are interested in different goals. The development of policy knowledge must consider which of these goals they intend to contribute to, and how they intend to manage the resulting ambiguity and conflict (Table 2.1). These limitations of goal specification cast doubt on the ability of a single reform effort to effectively produce data, information, and evidence for a variety of decision making needs (Radin 2006). In other words, the reality and complexity of government programs and policies prohibit the use of a one-size-fits-all approach, and assure that no single effort will be a panacea for decision makers. Second, when evidence-based reforms specify goals, they may either prioritize competing goals or ignore certain goals. This occurs most often in cases where individuals who are involved with the program are not given a chance to contribute to the enumeration of goals and can result in a lack of stakeholder buy-in (Newcomer 2007; Posner and Fantone 2007; Melkers and Willoughby 2005; Gilmour and Lewis 2006a, b; Talbot 2005; Behn 2002). Finally, misalignment of goals at institutional levels can lead to a mismatch of incentives (Behn 2003; GAO 2005, 2012; Pollitt 2006). For example, employee goals may be misaligned with program goals, resulting in an organization culture that resists performance information (Durant 1999; Dubnick 2005; Newcomer 2007). The second step in developing policy knowledge is the identification and collection of data. Data falls into four broad categories. The first of these is related to inputs, or the resources utilized by programs and other government activities. Second, data may be gathered about processes used to direct those inputs. Third, data may reflect outputs, or the products or services being produced. Finally, data may be gathered about outcomes, or the measures related to the overall goals of the activity. The quality of policy knowledge is directly related to the rigor, accuracy, and validity of the methods used to identify, generate, and collect data (GAO 2009; Weiss 1977; Lindblom and Cohen 1979). The accuracy of data refers to the extent to which measurement or record of observing the data is correct. For example, if a program is producing widgets on an assembly line, and production amounts are estimated by multiplying a standard number of units produced per hour by the number of hours the production line was active, the resulting figure may not be accurate if, for example, one employee left the line and caused the production rate to fluctuate over the period. In this case, a count of widgets produced as they come off the line could be a more accurate source of data. The validity of
40
S. PUTANSU
information refers to the extent to which the subject of measurement or observation is represented by the data collected. For example, knowing the number of students in one class is not likely a valid measure for the quality of their teacher. Additionally, multiple measures may arguably be valid in some circumstances. For example, one person may believe that the most valid measure of intelligence is a student’s grade, whereas another may believe that standardized test scores are more valid. Some may utilize both measures, whereas others may prefer one over the other. Ultimately, however, if a measure is not valid, the data it produces will not be useful in guiding decisions. The coverage of policy knowledge derives from the degree to which the data are complete and representative of the relevant population, program, or activity being studied. For example, if a decision maker desires information about the average emissions of vehicles in a given area, a sample of hybrid cars would not completely represent the population. To be complete, data would need to be collected on a census or a statistically representative sample of vehicles. A desire to know the national average of automobile emissions would require the entire national population of automobiles of all types to be represented. However, if a decision maker only wanted to compare hybrids and diesel engines, only those two types of automobile would need to be represented for the information to be considered complete. In some cases, multiple sources of data may be required in order to be sufficiently complete and representative. The ability of data to reflect sufficient quality and appropriate coverage is largely dependent on the rigor of the methods used to collect them. Data collection methods refer to the approach used to elicit and capture observations and measures. Methodology, including identification of the relevant population, what instruments are used to record data, where and when data are gathered, and how data are observed, has a direct impact on the data produced. For example, if a statistical sample is developed using a population frame that is not representative of the population of interest, results may be biased. If a survey is administered to a group of individuals in person, it may affect whether respondents are willing to answer certain sensitive questions. If data are gathered during rush hour in Los Angeles, they cannot represent daily driving habits in California. For decision makers to utilize policy knowledge, they must believe that the methodological design is sufficient and appropriate to meet their needs.
2
NO PANACEA: THE PURPOSES, USES …
41
Several limitations of evidence-based reforms emerge during data gathering and measurement activities. First and foremost, it is resourceintensive and technically challenging to gather data that are relevant to stated goals, sufficient and appropriate for understanding, and accurate, valid, complete, and reliable for interpretation of results (Cyert and March 1963). Limitations in data quality are often cited as reasons that data, information, and evidence are ignored (Dubnick 2005; ter Bogt 2004; Julnes and Holzer 2001). This can be exacerbated by reliance on administrative data, since these data are generally produced for other purposes, and therefore often do not meet the standards of data quality outlined above when applied to new areas of inquiry (Frederickson and Frederickson 2006). Similarly, when federal efforts rely on states for data, there are additional difficulties because states may not have standardized, comparable data that can be used with data from other states to produce nationally representative coverage (Majone 1989). The relationships between policy knowledge—as defined by the coverage and rigor of data, information, and evidence—and the range of decision making purposes make it abundantly clear that a single source of data could not possibly serve every need of every decision maker for every type of decision. It is critical that evidence-based reforms consider these relationships as they produce policy knowledge, promote its use, and assess its impact. If they fail to do so, the seemingly ad hoc and adversarial use of policy knowledge is likely to continue. After the enumeration of goals and collection of data, analysis and interpretation of the data are required to produce information and evidence about a given activity. In other words, the data that have been gathered must be processed into a form that directly relates to the goals that have been established. In this process, producers of policy knowledge need to ensure they are meeting their stated goals and define what it means to be “achieving” those goals. They can conduct analysis that reflects program activities and changes in measures of these goals, or they can establish evaluative criteria to generate and test evidence of a relationship among these measures (Radin 2006). Analysis and interpretation of data create additional considerations for the quality and coverage of policy knowledge. The information and evidence produced in this step cannot be of higher quality or broader coverage than the underlying data, but problematic analysis may obfuscate this fact. For example, some producers of policy knowledge may make generalizations from a single case or non-representative collection of cases to
42
S. PUTANSU
an entire population, such as identifying a malfunction in one vehicle as a cause for recalling an entire fleet. They may also be less drastic, including technical manipulations of statistical significance (referred to colloquially as “p-hacking”). In addition, evidence-based reforms may focus on priorities that are driven by political factors, which may result in incomplete, misleading, subjective, or biased results (Weiss 1977; Moynihan 2006; Moynihan 2008). Beyond these problematic analytical practices, producers of policy knowledge may also be subject to heuristic learning processes, where data, information, and evidence processed by an individual are subconsciously filtered through his or her existing knowledge and made to reflect his or her own subjective understandings (Simon 1947). As they produce information and evidence, it will necessarily include some degree of this subjective judgment. Others argue that such judgment becomes a critical part of turning policy knowledge into compelling and useful narratives (Stone 1997). The usefulness of these narratives may vary depending on existing heuristics and chosen points of comparison (Webeck and Nicholson-Crotty 2019), as well as views of the individual providing the policy knowledge (Perna et al. 2019). After completing the first three steps, producers of policy knowledge need to distill their findings into a format that can be used by decision makers (Weiss 1977; Lindblom and Cohen 1979). The method of communicating policy knowledge impacts whether it will be used. If presented too simplistically, it may be rejected for a lack of demonstrated nuance and understanding. However, if presented in too technical a format, it may be rejected for being convoluted or hard to understand (Behn 2002; Burke and Costello 2005; Pollitt 2006). As such, tailoring the communication of the policy knowledge to the specific audience is a critical step in promoting its use. An important component of this step is to explicitly and clearly communicate the limitations and subjective judgments that may occur in the first three steps, as decision makers may disagree about the relative weight of individual goals, the quality and coverage of data, the value of analytical techniques, the appropriateness of selected evaluative criteria, or the neutrality of the resulting policy knowledge (Newcomer 2007; Julnes and Holzer 2001; Posner and Fantone 2007; Pollitt 2006; Gilmour and Lewis 2006a; Talbot 2005; Melkers and Willoughby 2005). Clear communication of the range of decisions made to develop policy knowledge, and the related strengths and limitations of data, information, and evidence
2
NO PANACEA: THE PURPOSES, USES …
43
can enable decision makers to better integrate this knowledge into their deliberations. While producers of policy knowledge can work to communicate their data, information, and evidence in a way that is useful to decision makers, this knowledge often arrives indirectly. Specifically, direct communication of policy knowledge is supplemented by interpretation, combining sources, and communication among stakeholders and decision makers. In The Dynamics of Performance Management, Donald Moynihan (2008) explains that policy knowledge diffuses among decision makers and other stakeholders through a process of interactive dialog. This interactive dialog defines the use of data, information, and evidence as a process of presentation, exchange, and interpretation. Many scholars and advocates have undertaken efforts to promote the dissemination and use of policy knowledge through direct involvement in these dialogs. These efforts have sometimes coalesced into interest groups that advocate directly to decision makers and indirectly through media and other communications outlets for increased use of specific policy knowledge. This may lead to shared understanding, greater agreement, and coordinated action among stakeholders whose interpretations lead to consistent conclusions. However, it may also lead to little more than a restatement of already expressed conflicting opinions among competing preferences, with limited willingness to accept alternative explanations. The considerations and limitations identified in the four steps outlined above have a major influence on whether and how the resulting data, information, and evidence are used, and the use of this knowledge is often a primary factor in whether evidence-based reforms are seen as successful. However, government decisions on what to do, how to do it, and how much it should cost are only partially dependent on policy knowledge (Shulock 1999). Information overload, misleading heuristics, and selective use of policy knowledge can prevent utilization by decision makers (Moynihan 2006, 2008; Tversky and Kahneman 1986; Van de Walle and Boivard 2007). To the extent that these challenges are overcome, there is also significant variation in the ways that decision makers use policy knowledge and the degree of impact they have on the decision outcome.
44
S. PUTANSU
Direct and Indirect Uses of Data, Information, and Evidence Direct and explicit links between policy knowledge and decisions are foundational to evidence-based reforms that promise to be a panacea for decision makers. However, it is often difficult to determine whether and how specific data, information, and evidence contributed to a decision. For example, information on the unemployment rate could represent a warning about economic troubles. When addressing that problem, such as by establishing a new jobs program, decision makers may not explicitly cite the unemployment rate, or any other policy knowledge. This represents a second limitation of evidence-based reforms: even when policy knowledge influences a decision, it may not be a clear and deterministic factor. Further, Moynihan (2008) has demonstrated that multiple and conflicting goals among decision makers may result in compromises and trade-offs that further obscure whether and how data, information, and evidence are used. The policy dialog enables them to consider and prioritize multiple sources and types of policy knowledge before making a decision, but also makes tracking this use more difficult. Consistent with this insight, Majone (1989) has argued that information should not be viewed as deterministic to decisions, but as part of a persuasion that is necessary to achieve desired policy outcomes. Instead, government action is driven by a series of decisions, each of which is the result of the data, information, and evidence available to and utilized by decision makers. This process reflects the personal preferences of decision makers and the preferences of other actors that influence decisions, understanding of decision making rules, expectations about available resources, and specific substantive information about a particular choice (Lindblom and Cohen 1979). Reflective of these varied degrees of persuasion and influence, evidence-based governance reflects variation in how much policy knowledge is considered, and how central it is to decision outcomes. Specifically, decision makers may eschew it all together, may rely on a single source or selection of relevant sources, or may comprehensively consider the full range of relevant policy knowledge before acting. Similarly, the data, information, and evidence they use may only provide context for the decision, may provide insight into the deliberation of decision options, or may be fully deterministic of the outcome.
2
NO PANACEA: THE PURPOSES, USES …
45
Program Design and the Production of Policy Knowledge James Q. Wilson’s (1989) typology of bureaucratic organizations provides relevant insights into factors that can influence production and use of policy knowledge. Specifically, producing certain forms of data, information, and evidence can be more or less difficult, depending on the characteristics of the specific program being assessed. In particular, Wilson identifies differences in the ability to measure program activities and outputs related to government efforts. Where both can be easily measured, the bureaucratic type is production; where only action or only outcomes can be measured, the types are procedural and craft, respectively; when neither activity nor outcome can be measured, the type is coping. These four types will be described in more detail below, along with some expectations for the production of policy knowledge and the purposes it may serve. Production organizations are defined as those where both process and outcomes can be measured. In these organizations, Wilson argues, managers can potentially design a system of rules that will lead to an optimal outcome. However, one impact of this may be that those processes and outcomes that are easily measured could lead program managers to de-emphasize work on processes and outcomes that are more difficult to measure. When examining the production of policy knowledge, we might expect to see production organizations whose primary goals are clear and well defined and whose activities and outcomes are observable and measurable. In production organizations, secondary goals may be split into two categories. The first of these is goals whose activities and outcomes are also easily measurable, while the second is those with activities or outcomes that are harder to measure. Production organizations may elevate the importance of measurable secondary goals over those of secondary goals that are difficult to measure. In production organizations, data on inputs, outputs, and outcomes are expected to be readily available and easily converted into relevant information and evidence. Similarly, the tendency for production organizations to limit their focus to these areas suggests that they will be readily able to use this policy knowledge to guide outcomes. Procedural organizations are those where process of the primary goal can be observed but outcomes cannot. In these organizations, managerial focus is on professional expertise. The result of this is that standard operating procedures and rule-following become more important than
46
S. PUTANSU
results. In a decision making context, we would expect policy knowledge to be developed around process goals and data collection to focus primarily on inputs, secondarily on any measurable outputs, and generally not on outcomes. When required to produce outcome data, procedural organizations may attempt to focus on the outcomes of secondary goals, which may be easier to measure. This focus on process is expected to result in a focus on framing policy issues, with limited use to influence decision outcomes. The opposite of this occurs in craft organizations, where the process needed to reach the primary goal cannot be measured but outcomes can. The defining management style of craft organizations is discretion for operators to perform according to professional norms. In a decision making context, craft organizations may be able to produce output and outcome data related to their primary goal, but may have difficulty providing data on inputs and related measures of efficiency. In public organizations, Wilson argues, craft organizations tend toward establishing measures of process that are not directly related to the primary outcome as a means of limiting the opportunity for abuse of discretion. Policy knowledge primarily focused on outcomes is expected to be used to influence bargaining among stakeholders. The last bureaucratic type, coping organizations, can measure neither the process of their operators nor the outcome related to their primary goal. Managers in these organizations must attempt to recruit staff with professional expertise, though they cannot be sure which expertise is appropriate, and must try to establish procedures that will lead to good outcomes without knowing what those outcomes should be. Aside from this, they must intervene during crises, even though they do not know whether the crises are real or persistent. Cohen et al. (1972), in developing their garbage can model of decision making, called this an “organized anarchy.” In a decision making context, we do not expect to see valid and reliable measures of inputs or the primary outcome, but instead to see output measures of secondary goals, potentially viewed as intermediate outcomes related to the primary goals. In this system, the program may alternate its focus among secondary goals over time, as none are ultimately able to fully present progress toward the final outcome. Given these difficulties, any related policy knowledge is expected to be used to frame decisions and influence bargaining among decision makers. Wilson’s typology of bureaucratic organizations can be used to understand the relationship between program design and policy knowledge. In
2
NO PANACEA: THE PURPOSES, USES …
47
order to examine this relationship, the two cases in this book reflect the two ends of this model. Specifically, student loans are most similar to a production program, where measuring is relatively easy, while Title I most closely resembles a coping program, where both process and outcome are difficult to measure. This chapter has shown that evidence-based reforms have often achieved uneven and limited success as a result of overpromising that the policy knowledge they produce will be a panacea for decision making. Chapter 3 shows that these reforms have also been limited by lack of attention to decision making processes and political factors. Evidencebased reforms, in addition to their simplified treatment of policy knowledge, have tended to assume that decisions are either political or rational, treating policy knowledge as a substitute for, rather than supplement to, politics in decision making. Chapter 3 considers how the literature on political decision making has reinforced acceptance of this proverb and offers an alternative model for considering the competing and complementary roles of policy knowledge and political factors in the decision making process.
References Alkin, M. C., & King, J. A. (2017). Definitions of evaluation use and misuse, evaluation influence, and factors affecting use. American Journal of Evaluation, 38(3), 434–450. Barnow, B. (1992). The effects of performance standards on state and local programs. In C. Manski & I. Garfinkel (Eds.), Evaluating welfare and training programs (pp. 277–309). Cambridge, MA: Harvard University Press. Behn, R. D. (2002). The psychological barriers to performance management: Or why isn’t everyone jumping on the performance-management bandwagon? Public Performance & Management Review, 26(1), 5–25. Behn, R. D. (2003). Why measure performance? Different purposes require different measures. Public Administration Review, 63(5), 586–606. Bouckaert, G., & Halligan, J. (2008). Managing performance: International comparisons. London: Routledge. Burke, B. F., & Costello, B. C. (2005). The human side of managing for results. American Review of Public Administration, 35(3), 270–286. Cohen, M. D., March, J. G., & Olsen, J. P. (1972). A garbage can model of organizational choice. Administrative Science Quarterly, 17 (1), 1–25. Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm. Englewood Cliffs, NJ: Prentice-Hall.
48
S. PUTANSU
Davis, O. A., Dempster, M. A. H., & Wildavsky, A. (1966). A theory of the budgetary process. American Political Science Review, 60(3), 529–547. Dubnick, M. (2005). Accountability and the promise of performance: In search of the mechanisms. Public Performance & Management Review, 28(3), 376– 417. Durant, R. F. (1999). The political economy of results-oriented management in the “neoadministrative state”: Lessons from the MCDHHS experience. American Review of Public Administration, 29(4), 307–331. Feldman, M. S. (1989). Order without design: Information production and policy making. Stanford, CA: Stanford University Press. Frederickson, D. G., & Frederickson, H. G. (2006). Measuring the performance of the hollow state. Washington, DC: Georgetown University Press. Gilmour, J. B., & Lewis, D. E. (2006a). Does performance budgeting work? An examination of OMB’s PART scores. Public Administration Review, 66(5), 742–752. Gilmour, J. B., & Lewis, D. E. (2006b). Assessing performance assessment for budgeting: The influence of politics, performance, and program size in FY 2005. Journal of Public Administration Research and Theory, 16(2), 169– 186. Government Accountability Office (GAO). (2005). Performance budgeting: States’ experiences can inform federal efforts, GAO-05–215. Washington, DC: GAO. Government Accountability Office (GAO). (2009). Assessing the reliability of computer-processed data, GAO-09-680G. Washington, DC: GAO. Government Accountability Office (GAO). (2012). A guide for using the GPRA modernization act to help inform congressional decision making, GAO-12621SP. Washington, DC: GAO. Hatry, H. P. (2006). Performance measurement: Getting results (2nd ed.). Washington, DC: Urban Institute. Joyce, P. G. (2005). Linking performance and budgeting: Opportunities in the federal budget process. In J. M. Kamensky & A. Morales (Eds.), Managing for results 2005 (pp. 83–140). Rowman & Littlefield Publishers, Incorporated. Julnes, P. D. L., & Holzer, M. (2001). Promoting the utilization of performance measures in public organizations: An empirical study of factors affecting adoption and implementation. Public Administration Review, 61(6), 693–708. Lindblom, C. E., & Cohen, D. K. (1979). Usable knowledge: Social science and social problem solving (Vol. 21). New Haven, CT: Yale University Press. Majone, G. (1989). Evidence, argument, and persuasion in the policy process. New Haven, CT: Yale University Press. Melkers, J., & Willoughby, K. (2005). Models of performance-measurement use in local governments: Understanding budgeting, communication, and lasting effects. Public Administration Review, 65(2), 180–190.
2
NO PANACEA: THE PURPOSES, USES …
49
Moynihan, D. P. (2006). Managing for results in state government: Evaluating a decade of reform. Public Administration Review, 66(1), 77–89. Moynihan, D. P. (2008). The dynamics of performance management: Constructing information and reform. Washington, DC: Georgetown University Press. National Academy of Public Administrators Social Equity Committee. (2000, November). Standing panel on social equity in governance: Issue paper and work plan. Washington, DC: NAPA. Newcomer, K. E. (2007). How does program performance assessment affect program management in the federal government? Public Performance and Management Review, 30(3), 332–350. Perna, L. W., Orosz, K., & Kent, D. C. (2019). The role and contribution of academic researchers in congressional hearings: A critical discourse analysis. American Educational Research Journal, 56(1), 111–145. Pollitt, C. (2006). Performance management in practice: A comparative study of executive agencies. Journal of Public Administration Research and Theory, 16(1), 25–44. Posner, P. L., & Fantone, D. M. (2007). Assessing federal program performance: Observations on the US Office of Management and Budget’s program assessment rating tool and its use in the budget process. Public Performance & Management Review, 30(3), 351–368. Radin, B. A. (2006). Challenging the performance movement. Washington, DC: Georgetown University Press. Roberts, A. (2019). Bridging levels of public administration: How macro shapes meso and micro. Administration & Society. https://doi.org/10.1177/ 0095399719877160. Shulock, N. (1999). The paradox of policy analysis: If it is not used, why do we produce so much of it? Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(2), 226–244. Simon, H. A. (1947). Administrative behavior. New York: Free Press. Stone, D. A. (1997). Policy paradox: The art of political decision making (Vol. 13). New York: W. W. Norton. Talbot, C. (2005). Performance management. In E. Ferlie, L. Lynn Jr., & C. Pollitt (Eds.), The oxford handbook of public management (pp. 491–517). Oxford: Oxford University Press. ter Bogt, H. (2004). Politicians in search of performance information? Survey research on Dutch Aldermen’s use of performance information. Financial Accountability and Management, 20(3), 221–252. Tversky, A., & Kahneman, D. (1986). Rational choice and the framing of decisions. The Journal of Business, 59, 251–278.
50
S. PUTANSU
Van de Walle, S., & Boivard, T. (2007). Making better use of information to drive improvements in local public services. A report for the Audit Commission. Birmingham: Institute of Local Government Studies, University of Birmingham. van Dooren, W. (2004). Supply and demand of policy indicators: A cross-sectoral comparison. Public Management Review, 6(4), 511–530. Webeck, S., & Nicholson-Crotty, S. (2019). How historical and social comparisons influence interpretations of performance information. International Public Management Journal, 1–24. Weiss, C. H. (Ed.). (1977). Using social research in public policy making (Vol. 11). Lexington, MA: Lexington Books. Wildavsky, A. B. (1974). Budgeting: A comparative theory of budgetary processes. Boston: Little, Brown. Wilson, J. Q. (1989). Bureaucracy: What government agencies do and why they do it (2nd ed.). New York: Basic Books.
CHAPTER 3
Challenging the Proverb: A Balanced Model for Governance Decisions
Evidence-based reforms have often been predicated on the idea that the data, information, and evidence they produced are a direct and definitive driver of policy decisions. These reforms promise policy knowledge as a panacea without addressing the questions of “for what purpose” and “for whom?” Instead, they invoke a proverb that treats policy knowledge as a replacement for politics. This often hinges on the presumption of centralized, unitary decision making. As with the concept of enlightened government, the push for centralized control to improve efficiency is foundational to the republic. While James Madison extolled the protections of the separation-of-powers system, Alexander Hamilton argued for the value of a central government and a central banking system (Bertelli and Lynn 2006). Calls for evidence-based decisions, echoing Hamilton, have asserted and perpetuated use of the proverb while overlooking the competing values of individuals within decision making processes (Frederickson and Frederickson 2006). As shown in the first two chapters, this results in a one-size-fits-all creation of policy knowledge, and unfulfilled promises of a panacea for decision making. Government decisions include which programs to create; how much funding to allocate them; whether or not to continue them; and if they should be expanded, shrunk, eliminated, or replaced by other programs. Within this context, many producers of policy knowledge have noted that their research has answered relevant questions about a program or policy but did not seem to impact the decisions actually being made about that program (Weiss 1979). Instead, decision makers either rely on alternative © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_3
51
52
S. PUTANSU
or conflicting sources, or seem to eschew data, information, and evidence altogether. On the one hand, researchers cite a lack of support, time, and resources for providing actionable research (Cairney 2016). On the other, decision makers are blamed for valuing politics over “quality research” (Lomas and Brown 2009). Alternative types and sources of policy knowledge are often neglected, resulting in a one-size-fits-all approach and a weak link between what is needed for a given decision and the data, information, and evidence produced by the reforms. These reforms also fail to acknowledge how variation in the coverage and rigor of the knowledge produced by their efforts, or the resulting limitations (Weiss 2002). As described in Chapter 2, the public administration and policy literature have shown that information needs vary based on the policy process, program implementation (Melkers and Willoughby 2005), sectoral differences (Askim 2007), cultures (Durant and Legge 1998), country of implementation (Pollitt 2006), and managerial uses (Behn 2003). Thus, decision makers are expected to need multiple types and sources of data, information, and evidence, which depend on the scope of the decision at hand and the urgency of action (Kahneman 2012). This variation in need is multiplied by the needs of different actors who may be interested in different aspects of the same program, and therefore need different policy knowledge to support their deliberations (Feldman 1989). For example, some decision makers may be focused only on the efficiency of a program, while others are concerned primarily with its effectiveness or impacts on equity. Further, for a given decision, the aspects of a program being considered will drive the selection of policy knowledge to use, and the result is that much of the available data, information, and evidence may not be relevant in a specific decision context (Shulock 1999). This chapter builds on the typology introduced and developed in the last two chapters by adding a political dimension to the models. It shows that, beyond failing to consider different purposes of policy knowledge, evidence-based reforms have relied on a proverb of using data, information, and evidence to supplant, rather than complement, political factors. By not considering how policy knowledge reflects political values, the variety of ways it can impact decisions, or the overall complexity of the political decision making process, these reforms have sowed the seeds of their own destruction. As an alternative to this limited approach, this chapter introduces relevant theories of decision making and discusses the different mechanisms through which policy knowledge can have an impact on decision makers.
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
53
It explains the political factors and policy knowledge that are brought to bear in the decision making process and shows how policy knowledge is differentially converted into “usable knowledge” to meet unique purposes (Lindblom and Cohen 1979). As described in Chapter 2, decision makers are more likely to consider data, information, and evidence they believe are relevant to their purposes and of sufficient quality. Further, as the political context and policy focus changes over time, no single source of policy knowledge is likely to meet these needs for every possible decision making opportunity.
Politics and Policy Knowledge as Drivers of Decision Making This chapter outlines two decision making models that are common in the academic literature. First, it highlights the rational-comprehensive decision model that is closely aligned with expectations of evidence-based reformers and has its roots in the works of Max Weber (2013). This model assumes that policy knowledge is the primary factor in decisions and assumes low levels of technical conflict, often with the presumption of a unitary decision maker. The second model is political-incremental and is most closely associated with Aaron Wildavsky (1964). This model assumes a tendency toward the status quo and the dominance of political factors over data, information, and evidence in the decision making process. The chapter summarizes research that demonstrates areas where these models fail to explain decisions, as well as some expansions of the models intended to explain these anomalies. The section shows that, while these models individually and collectively provide important insights into decision making, they fall short of covering the full range of decisions that take place. This study offers an expanded model of decision making that integrates the strengths of these approaches while introducing a framework to counter their limitations. The Rational-Comprehensive Model Advocates of evidence-based policy generally assume a rationalcomprehensive model of decision making, which tends to answer questions of policy and implementation in terms of the “best” way to achieve a given goal. Under this model, policy knowledge can and should be
54
S. PUTANSU
completely deterministic of governance decisions, and decision makers should consider the entire scope of a policy during their deliberations. Rational-comprehensive models are the most definitive example of the proverb of evidence-based policy, suggesting that policy knowledge completely supplants politics in decision making. Evidence-based reforms have often been modeled around this way of thinking and have included requirements to define concrete outcome-based goals, ways to measure progress toward them, and the optimal strategy for achieving that progress (Moynihan 2008). Rational models assume that policy knowledge is never selectively used to support preexisting opinions, but rather that it can supersede beliefs and opinions to support purely evidencebased decisions (Simon 1947). If decision makers use rational-comprehensive processes, they can begin with a shared understanding of their objectives and would be able to objectively agree upon and define goals clearly (Radin 2006). They can find and process all relevant policy knowledge needed to consider every feasible alternative and predict the results of their decisions to make the best decision in a systematic way. Results would be uniformly understood to have one and only one meaning (Stone 1997). In other words, all sources of data, information, and evidence from Chapter 1 would be available, and decision makers would consider each in order to come to complete agreement on the priority of all goals every time a decision was made. When policy knowledge is given primacy in decision making reforms, it reflects a desire to approximate rational-comprehensive models. The ideal type of rational, comprehensive decision making is not strived for or achieved in practice, but is a model of what would exist in an environment where all options could be considered and evaluated accurately (March 2002). Rather than expect that all policy knowledge to be considered, these reforms presume that the provision of more or better policy knowledge will guide decision makers toward increased rationality. Specifically, if reforms can ensure that goals are clear and measurable, policy knowledge often can and does improve the effectiveness, efficiency, and equity of programs (Behn 2003). In the field of public administration, the limits of this model have been evident since Herbert Simon’s publication of Administrative Behavior (1947). In this study, Simon established the concepts of bounded rationality and heuristic decision making. In his model, decision makers take into account the policy knowledge that is available to them, rather than
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
55
all sources. With this limited data, information, and evidence, they use an internalized set of simplified rules to make the decision that seems best, rather than a single objective best decision. This modification has been explored and expanded, showing that information overload, misleading heuristics, and selective use of policy knowledge also prevent rationalcomprehensive decision making from occurring in practice (Moynihan 2006). This means that, rather than considering all policy knowledge in every case, decision makers could selectively rely on relevant data, information, and evidence that contributes something meaningful to their deliberations. This relaxed assumption, however, introduces a new issue, which is that different decision makers may have different heuristics, resulting in disagreements about which policy knowledge is most suitable for a given purpose. This does not mean, however, that all decisions are irrational or uninformed. Rather, the impossibility of a purely rational-comprehensive approach underscores the importance of ensuring that heuristic decision makers are exposed to new data, information, and evidence. By doing this, reformers and producers of policy knowledge can help to generate shared agreement, reduce faulty heuristics, and support improved decision making. Evidence-based reforms have reflected the rational-comprehensive approach and, as a result, have faced challenges in implementation that parallel the limitations of this model. In clear contradiction of the evidence-based proverb, these reforms have continually fallen short of their intent to convert political decision making processes into purely rational ones (Behn 2003; Radin 2006; others). While bounded rationality relaxes the requirements for purely rational-comprehensive decisions, it does not solve the issue of decision makers requiring different types of policy knowledge for different purposes. Because information needs vary by the values and purposes of decision makers, reforms predicated on the use of specific policy knowledge will never be a panacea. The Political-Incremental Model The political-incremental model argues that political factors dominate the decision making process, and that policy knowledge is largely ignored, or otherwise used symbolically to bring meaning to political decisions when conflict is high (Wildavsky 1974). The political-incremental model is routinely cited by critics of the evidence-based reforms as an alternative
56
S. PUTANSU
to its rational assumptions, and as an explanation for why these reforms have not delivered on their promises. The political-incremental models were developed largely in response to efforts to impose performance budgeting, in which policy knowledge is intended to definitively guide allocation decisions. However, incrementalists argue that current budget levels reflect a set of normative compromises and negotiations that have already been resolved, and decisions have a tendency to result in modest changes that maintain the status quo. This inertia is an intentional feature of the system of checks and balances in the United States, which requires agreement among numerous decision makers and provides overlapping authority on a wide range of issues (Wildavsky 1974). Further, these incremental changes are not driven by policy knowledge about optimal allocation amounts, but rather by political agreements among decision makers. Political-incremental theories of decision making have been supported through substantial scholarship that is often traced back to Aaron Wildavsky’s Political Theory of the Budget Process and related critiques of rational-comprehensive based reforms aimed at impacting the budget process (1964), or to Charles Lindblom’s earlier work on the “science of muddling through” (1959). Wildavsky’s original approach to incremental budgeting was founded largely on the premise that the breadth and complexity of government actions were simply too large to be considered comprehensively. This complexity, coupled with the inertia described above, causes all budget considerations to use the previous budget agreement as a starting point. Put simply, in most cases, the current budget is considered a “base,” and any increase or decrease in overall spending is allocated so that each agency gets its fair share, as determined by political factors. In an incrementalist model, there are certain expectations for different decision makers. For example, in the budget process, the House of Representatives appropriations committees play a key role as “guardian” of the budget. The Senate committees tend to serve as a check against the House, handling the appeals of agencies that are unsatisfied with the House committees’ decisions. The President, through the Office of Management and Budget (OMB), starts the process by sending a request of their preferences to Congress. These preferences, it is expected, reflect agency needs and desires to maximize their appropriations. Throughout this mélange of actors, each performs politically to maintain or increase funds available to their preferred policies or programs. More recently, this
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
57
political model of decision making has been undertaken by postpositive researchers (Lynn 1999), who argue that the limitations of policy knowledge require a political solution to decision making. In both cases, these pure political models reflect the same underlying logic as the evidencebased proverb—that decisions are either political or rational, without respecting the valuable role played by both politics and policy knowledge. The incremental model has been utilized and refined for several decades and has been consistent with the outcomes of many political and budget decisions. The theory has provided a sound basis for considering and understanding both policy and budget decisions. It is consistent with Lindblom’s conception of muddling through and Simon’s satisficing, as it explains an environment in which incremental learning leads to incremental changes, and is often considered to be an “equilibrium model.” Political-incremental models assume that this balance is stable and will result in decisions that usually reflect the status quo. However, some decisions are decidedly not incremental (Kingdon 2003). Although this has been acknowledged by incrementalist scholars from the very early days of the theory, including Wildavsky’s acknowledgment of “shift points,” (1974) it has not been thoroughly integrated into the theory. Exceptions to the Dominant Models Various models have been introduced which seek to explain exceptions to the dominant models of decision making (Kingdon 2003). They attempt to develop that explain decisions that would not be possible given either of these models, given that decision opportunities represent “fundamentally ambiguous stimuli” (Cohen et al. 1972). These models generally assume that one of the dominant models represents the usual, or equilibrium, case. Exceptions, therefore, are treated as a shock to this equilibrium. These punctuated equilibrium models were initially championed primarily by Shepsle and Weingast (1981), who posited that the policy environment reflects an economic market, which is a function of demand for a policy and government supply. The equilibrium of this market represents the current political and resource environment and respective agreements. In their model, changes are very minor over time, reflecting an incrementalist mindset. However, in some instances, an exogenous shock to the supply or demand curves can spur a non-incremental change in policy. For example, a natural disaster could cause an increase in homelessness,
58
S. PUTANSU
thereby generating increased demand for government subsidies of housing. Similarly, declaration of war may require shifting of resources that shifts the supply curve for other policies. In some cases, the revelation of a particularly salient piece of policy knowledge, like massive embezzlement of government funds, could punctuate the stability of government programs. This type of decision model is limited, however, because it does not sufficiently explain, other than in crisis situations, how comprehensive policy changes can occur, nor do they explain the scenarios under which policy knowledge may be a primary or secondary driver of a change. One form of punctuated equilibrium model is known as the garbage can model. In the garbage can model, decisions are dependent on the participants involved, their individual problems and solutions, and choice opportunities that allow participants to match their problems to solutions and vice versa (Cohen et al. 1972). All forms of policy knowledge represent either a problem or a solution, which are dumped into a “garbage can” for future use by decision makers. The garbage can model uses a metaphor of streams to explain decisions, including streams of choices, problems, solutions, and energy from participants. If a “choice opportunity” is understood to be an exogenous shock, these models provide a similar outcome to the punctuated equilibrium model, while offering the possibility of implementing a new policy that may change the shape and location of the supply curve, resulting in comprehensive changes to policy. This model, however, does not provide insight into how or why particular data, information, and evidence are produced, or the reasons why some sources may be pulled out from the garbage can for use by decision makers. Further, it essentially only allows for policy knowledge to be used in the case of non-incremental decisions. An alternative model, which allows for slightly more explanatory power, also uses the metaphor of streams. These “policy stream” models are contingent on the existence of exogenous shocks (now called policy windows) to institute new policy ideas, but they more clearly account for these shocks to be a result of the introduction of policy knowledge. This model presents three policy streams—problems, policies, and politics—which can be coupled whenever there is a “policy window” to enact change (Kingdon 2003). In this model, some problems are ignored, others are attached to ineffectual solutions, and some are actually resolved, but non-incremental change only occurs during coupling of the streams, which can occur as the result of an event or through the action of a “policy entrepreneur.” This model
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
59
has an explanatory advantage over the garbage can model because the political stream incorporates some of the aspects of ideology and interests. Further, it allows for rational decisions without major shocks, but only in cases where a policy entrepreneur can find a way to use that policy knowledge to create a coupling of the streams, generally by convincingly connecting a policy to a problem. However, this model still treats data, information, and evidence as an external input into decision making and provides little insight into how and why policy knowledge is produced and why some, but not all, are ultimately used. These two models add explanatory power to the rationalcomprehensive and political-incremental models by providing insights into decisions that are political-comprehensive and rational-incremental. However, none of the four models provides an explanation of decisions that fall somewhere between purely incremental and comprehensive, nor do they explain decisions that result from a mixture of rational and political factors. Graham Allison (1971), in his classic work Essence of Decision, has shown that the rational-comprehensive model reflects situations where a rational actor has decision making authority, while political-incremental models consider organizational processes and government politics. Further, he shows that each of these models is operating simultaneously and, in some circumstances, in opposition to one another. Allison’s insights demonstrate that political decision making is not a question of whether a decision is rational-comprehensive or political-incremental, but rather a question of degrees. The next section offers an expanded framework of decision making that utilizes this insight to enable an examination of these still unexplained decisions.
A Balanced Framework of Politics and Policy Knowledge The sections above showed that rational-comprehensive, politicalincremental, and punctuated equilibrium models each offer valuable insights into government decision making. While these represent the dominant models for understanding the roles of policy knowledge and politics in decision making, they fail to fully integrate these factors into a balanced model. This section offers an integration of the rational-comprehensive and political-incremental models that distinguishes between two continua, one from rational to political and the other from incremental to comprehensive. This framework offers a more
60
S. PUTANSU
Punctuated Equilibrium Models
Rational Comprehensive
PoliticalIncremental
Bounded Rationality
Incremental
Mixed
Comprehensive
nuanced and holistic framework for assessing political decisions by allowing for variation in the extent to which decisions will be rational, political, incremental, or comprehensive, and allowing for a balanced consideration of potential decision outcomes. Figure 3.1 illustrates these continua and shows how the models described so far provide only partial coverage of decision making. The political-rational continuum allows for the full range of policy knowledge influences. This includes purely political, with no consideration of policy knowledge, as well as mixed use to frame debates bargain over interests, or deterministic use to lead decisions. Over a series of publications, Carol Weiss (2001) has examined the non-use of, as well as the factors impacting utilization of, evaluative research. Through these examinations, she has developed a framework of Four-I’s: Ideology, Interest, Institutions, and Information. This model states that political factors
Political
Mixed
Rational
Fig. 3.1 Dominant models of decision making by scope and type of influence
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
61
of ideologies, interests, and institutions, along with the existing knowledge, all influence whether new policy knowledge will be utilized. Specifically, conflict within each factor and across factors impacts whether and how new policy knowledge will be digested and used by decision makers. James Q. Wilson’s bureaucratic types, introduced in Chapter 2, provide some additional expectations for the production and influence of policy knowledge. Specifically, production programs, represented in this study by student loan programs, may be expected to rely on data, information, and evidence to a greater extent than coping programs, represented in this study by Title I. The incremental-comprehensive continuum allows for decisions to range from no incremental changes to programs or their budgets, to comprehensive creation, overhaul, or elimination of programs. Weiss’ Four-I model also offers some expectations for where a particular decision will fall on this continuum. Specifically, as conflict among the Four-I’s increases, decisions are more likely to reflect the status quo. If one factor, or an alignment of factors, is able to dominate the decision making process, comprehensive outcomes may become a more realistic possibility. Integrating the dominant models in this way reveals that many decisions are left unexplained. Viewed individually, or in combination, they fail to provide a unified explanation for what drives decisions. In isolation, explaining decisions at four different ends of these continua requires four different models. Even when all the models are considered together, many decisions that are impacted by both rational and political factors, or that fall somewhere in the middle of incremental and comprehensive, are still unexplained. As noted above, incremental decisions may occur within rational and political models, as well as comprehensive decisions. As such, uncoupling the comprehensiveness of a decision from the political factors and policy knowledge that drives it allows for the full range of decisions to be considered. While some decisions will be incremental, and some may be comprehensive, others will fall in the middle and address multiple, but not all, issues related to a policy area. This makes clear that a more complete approach for explaining decisions is needed, and the following section offers a solution. Each of the political factors can impact the use of policy knowledge. Ideology, which represents people’s core beliefs, can result in the filtering out of data, information, and evidence that contradicts one or more fundamental beliefs of decision makers. For example, Weiss (2001) explains that “no amount of valid data about the positive effects of abortion on
62
S. PUTANSU
the lives of women and children is likely to convince a principled upholder of the right to life to support abortion policy.” Two primary ideological issues underpin consideration of the student loan and Title I case studies. The first ideological issue is federalism. Specifically, there is a tension between local, state, and federal involvement in education policy. Chapter 5 shows that the early history of education was largely defined by a limited federal role, with education policy dictated locally and at the state level. Chapters 6 and 7 show that this tension between levels of government has continued to influence modern federal education policy. The second ideological issue is about the scope of government. The ideological schism here is a tension between the government and the private sector. Specifically, this tension revolves around whether government should be in place to support “efficient markets” or whether the government should be an active participant in setting and achieving non-market goals. These cases explore these and other ideological factors, their impact on decisions, and the resulting utilization of policy knowledge to support those decisions. The factor of interest includes self-interest of potential users, and the agglomeration of those interests in interest groups. Different interests may focus on different parts of a complex decision and may have different, potentially competing, goals. This can also be a limiting factor in the use of policy knowledge, as interests will typically select and distribute data, information, and evidence that supports their positions. The cases in this book examine three broad types of interests: (1) government interests, including state and local governments; (2) policy interests, including affected groups, institutions, and workforces; and (3) private interests, including industry-specific interests and other groups that support the education function outside of direct provision, including students and their families. The case studies will highlight individual interests within this milieu that impacted the use of policy knowledge in decisions about the two programs. While the book generally refers to broad categories of interest groups, this categorization is limited in that in tends toward a monolithic description of interests that obscures differences among these communities. For example, while teachers unions have a variety of interests, some align with student outcomes, others with fair and equitable labor practices, and still others along a variety of dimensions. Some unions may emphasize and prioritize these interests differently. My characterizations throughout this chapter and the rest of the book emphasize the most
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
63
impactful advocacy of groups within these communities at a given point in time. Institutions represent the normal way of doing business, or what Allison (1971) called “organizational processes.” Weiss (2001) argues that different institutional norms can limit the use of policy knowledge. For example, a unitary decision maker will typically utilize knowledge consistent with their own ideology and interests, whereas group decision makers represent a broader set of ideologies and interests, and may therefore rely on broader knowledge. There are two types of institutional norms that impact education policy decisions: programmatic and operational. The first of these refers to institutional norms within the federal government, including separation of powers, congressional procedures, and partisan politics, which impact the process of decision making. Operational institutional norms, on the other hand, refer to the design of the federal program, including its procedures and traditions for providing services, the delivery method of services, and program goals. The case studies will show how these and other institutional norms limit or impact the set of options considered when decisions are made. For example, one institutional issue within Congress is related to the separation of authorization from appropriations. Authorizations are focused on individual policy areas and the details about those programs’ goals and operations, while appropriations are focused broadly across programs and focused on funding. Appropriations have a lot of institutional pressure for support because failure to pass them can result in government shutdowns, while lapsed authorizations generally have less dire consequences. The inclusion of information in Weiss’ framework is important because it acknowledges that each new piece of policy knowledge is created within a context of existing data, information, and evidence. When identical to existing knowledge, it doesn’t necessarily add value; when strongly counter to the existing state of knowledge, it can be resisted by decision makers. In these cases, it must be compelling enough to displace existing knowledge. Differences in the quality of policy knowledge explain many of the pitfalls of evidence-based reforms and can be amplified when conflict exists within and among ideologies, interests, and institutions. These political factors may impact the selection, prioritization, and use of policy knowledge by decision makers to support their preferences. At the same time, these decision makers focus on coverage and rigor to attack or challenge competing sources of data, information, and evidence.
64
S. PUTANSU
Political Elements of Policy Knowledge Throughout this chapter, the focus has been on whether and how policy knowledge and politics interact within the decision making process, and has considered this as an interaction between politics and policy knowledge. It has shown that the proverb of evidence-based policy, which presumes that policy knowledge should supplant politics, fails to accurately describe decision making. However, an important assumption of the neutrality of information has yet to be explored. Specifically, the use of the proverb that elevates policy knowledge over politics is premised on a distinction between facts and values. This distinction between what is and what should be has roots in the Enlightenment philosophies of the eighteenth century and is the basis for most modern scientific methods. However, in the 1950s, two scholars, Dwight Waldo and Herbert Simon, independently and comprehensively challenged this distinction. The resulting scholarship made clear that administration always includes some measure of politics, and that all policy knowledge carries with it a mixture of both facts and values (Waldo 1952a, b; Simon 1952). Despite this work, the politics-administration dichotomy has continued to influence governance scholars, administrators, and politicians to the present day, resulting in reforms predicated on an evidence-based proverb that specific value-neutral policy knowledge can replace political factors in the decision making process. This book offers a more complete definition that accounts for multiple types of data, information, and evidence, reflective of both facts and values. Chapter 2 focused on technical qualities of policy knowledge, including how it is produced, and its quality vis-à-vis sufficient coverage and appropriate scope. Once produced, that policy knowledge enters a political dialog where it is interpreted, shared, and exchanged among decision makers and other stakeholders (Moynihan 2008). This process of exchange can help link disparate sources of policy knowledge to provide more complete understandings for decision makers, but can also be influenced by politically motivated actors, faulty interpretations, and disputes over the independence, representativeness, or accuracy of specific data, information, and evidence. This section examines the political and subjective nature of information production, and the subsequent political dialog that distills and translates it into use for decision making.
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
65
The creators of data, information, and evidence represent a diverse and multidisciplinary collection of individuals and entities with widely divergent motivations, interests, and responsibilities. They include individual academic researchers and evaluators, who may study policy and programs they find personally interesting; programs and offices, who use policy knowledge to manage programs and provide oversight; think tanks and interest groups, who produce policy knowledge with the intent of influencing policy directly; and independent research organizations, who may produce policy knowledge by the direction of or through contract with external entities. Each of these producers may imbue the resulting data, information, and evidence with the subjective values inherent in their political leanings, epistemological preferences, and other conscious and unconscious biases. As described in Chapter 2, one important aspect of this range of producers is that it has an impact on what is studied. However, the independence of these producers also impacts trust in data, information, and evidence. If producers are seen to have a stake in their results that aligns with political factors, they may be distrusted by those with opposing political views. If they have a personal stake in the results (e.g., a financial incentive), this distrust may grow broader. This influences whether and how the information will be introduced into political dialog. These subjective and political aspects of policy knowledge are another important reason why the proverb of supplanting politics is unreasonable. Because the data, information, and evidence used for decision making are imbued with these political elements, the generation, communication, and use of policy knowledge carries a degree of politics with it. While the quality of that information, including sufficient coverage and appropriate rigor, can mitigate some of these political qualities, it cannot eliminate them fully. As such, this interaction between politics and policy knowledge must also be a part of any consideration of decision making. As policy knowledge enters into a policy community, decision makers and other stakeholders influence how it diffuses through that community and whether and how it ultimately gets used. Carol Weiss (1989) suggests that decision makers may introduce and rely on policy knowledge within this dialog in four ways. The first is as support for preexisting opinions. There are two general ways that this purpose is fulfilled. On the one hand, decision makers may intentionally seek to confirm their preexisting beliefs and actively reject policy knowledge that challenges them, thereby symbolically using it only to rationalize a decision they have essentially already
66
S. PUTANSU
made. There is a significant literature on cognitive dissonance and confirmation bias which shows that this use, while often intentional, may also occur on a subconscious level (D’Alessio and Allen 2002; Gordon 1996). For example, a decision maker that supports a program may be naturally inclined to seek out and believe in policy knowledge that detects positive outcomes, while being more skeptical of those that find the reverse. On the other hand, decision makers with incomplete policy knowledge may unintentionally limit their search within the confines of what is already known, and may overlook data, information, and evidence that is contrary to the limited policy knowledge currently available to them (Behn 2002). Cognitive psychology considers this a form of “anchoring.” For example, in some school reforms, a focus on test scores initially anchored decision makers’ focus, causing them to overlook studies that considered graduation rates. Because graduation rates were inversely related to test scores, decision makers ended up with an incomplete and positively biased selection of policy knowledge. The second and third purposes of policy knowledge are warning— which can come in the form of “police patrols” or “fire alarms” (McCubbins and Schwartz 1984)—and guidance. Policy knowledge serves as a warning when it identifies a problem or risk, while guidance provides potential ways to address those identified issues. For example, the Social Security Administration could monitor recipients outside of typical age ranges as a form of warning of potential improper payments and fraud and may use analysis of death records as guidance for identifying recipients whose benefits were not appropriately terminated. Finally, policy knowledge can serve a general purpose of enlightenment, which reflects longer-term learning and development of understanding about the topic being studied. Every use of policy knowledge contributes to enlightenment. In addition, other data, information, and evidence may support enlightenment even when not used directly by decision makers. For example, policy knowledge may introduce new perspectives, develop and reframe policy questions, and explain competing and complementary relationships among topics of study. The first three chapters of this book have demonstrated how overly simplistic definitions of policy knowledge, unfulfilled promises of panacea, and the reliance on the evidence-based proverb have limited the success of these reform efforts. They have offered theoretical development of several frameworks and models for understanding policy knowledge and its role in decision making, vis-à-vis political factors, in a more nuanced
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
67
way. The cases in this book represent the first empirical test of the resulting hypotheses. First, Chapter 4 takes a historical look at evidence-based reforms in order to more comprehensively consider promises of panacea and reliance on the proverb of evidence-based governance. Then, in Section Two of this book, each of the federal education cases considers whether and how the full range of policy knowledge has supported the four purposes of decision making. It also reflects the complexity of the decision making process, and the differential impacts of political factors and policy knowledge. This comprehensive and historical account of successive decisions about the cases enables a robust application of these frameworks that can inform and improve efforts to produce policy knowledge and use it to improve governance decisions.
References Allison, G. T. (1971). Essence of decision: Explaining the Cuban missile crisis. Boston: Little, Brown, and Company. Askim, J. (2007). How do politicians use performance information? An analysis of the Norwegian local government experience. International Review of Administrative Sciences, 73(3), 453–472. Behn, R. D. (2002). The psychological barriers to performance management: Or why isn’t everyone jumping on the performance-management bandwagon? Public Performance & Management Review, 26(1), 5–25. Behn, R. D. (2003). Why measure performance? Different purposes require different measures. Public Administration Review, 63(5), 586–606. Bertelli, A. M., & Lynn, L. E. (2006). Madison’s managers: Public administration and the constitution. Baltimore: Johns Hopkins University Press. Cairney, P. (2016). The politics of evidence-based policy making. Dordrecht: Springer. Cohen, M. D., March, J. G., & Olsen, J. P. (1972). A garbage can model of organizational choice. Administrative Science Quarterly, 17 (1), 1–25. D’Alessio, D., & Allen, M. (2002). Selective exposure and dissonance after decisions. Psychological Reports, 91(2), 527–532. Durant, R. F., Legge, Jr. J. S., & Moussios, A. (1998). People, profits, and service delivery: Lessons from the privatization of British Telecom. American Journal of Political Science, 42(1), 117–140. Feldman, M. S. (1989). Order without design: Information production and policy making. Stanford, CA: Stanford University Press. Frederickson, D. G., & Frederickson, H. G. (2006). Measuring the performance of the hollow state. Washington, DC: Georgetown University Press.
68
S. PUTANSU
Gordon, R. A. (1996). Impact of ingratiation on judgments and evaluations: A meta-analytic investigation. Journal of Personality and Social Psychology, 71(1), 54. Kahneman, D. (2012). Thinking, fast and slow. New York: Macmillan. Kingdon, J. (2003). Agendas, alternatives and public policy. Boston: Longman. Lindblom, C. E. (1959). The science of “muddling through.” Public Administration Review, 19(1), 79–88. Lindblom, C. E., & Cohen, D. K. (1979). Usable knowledge: Social science and social problem solving (Vol. 21). New Haven: Yale University Press. Lomas, J., & Brown, A. D. (2009). Research and advice giving: A functional view of evidence-informed policy advice in a Canadian ministry of health. The Milbank Quarterly, 87 (4), 903–926. Lynn Jr., L. E. (1999). A place at the table: Policy analysis, its postpositive critics, and future of practice. Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(3), 411–425. March, J. G. (2002). A primer on decision making: How decisions happen. New York: Free Press. McCubbins, M. D., & Schwartz, T. (1984). Congressional oversight overlooked: Police patrols versus fire alarms. American Journal of Political Science, 28, 165–179. Melkers, J., & Willoughby, K. (2005). Models of performance-measurement use in local governments: Understanding budgeting, communication, and lasting effects. Public Administration Review, 65(2), 180–190. Moynihan, D. P. (2006). Managing for results in state government: Evaluating a decade of reform. Public Administration Review, 66(1), 77–89. Moynihan, D. P. (2008). The dynamics of performance management: Constructing information and reform. Washington, DC: Georgetown University Press. Pollitt, C. (2006). Performance management in practice: A comparative study of executive agencies. Journal of Public Administration Research and Theory, 16(1), 25–44. Radin, B. A. (2006). Challenging the performance movement. Washington, DC: Georgetown University Press. Shepsle, K. A., & Weingast, B. R. (1981). Political preferences for the pork barrel: A generalization. American Journal of Political Science, 25, 96–111. Shulock, N. (1999). The paradox of policy analysis: If it is not used, why do we produce so much of it? Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(2), 226–244. Simon, H. A. (1947). Administrative behavior. New York: Free Press. Simon, H. A. (1952). “Development of theory of democratic administration”: Replies and comments. American Political Science Review, 46(2), 494–503.
3
CHALLENGING THE PROVERB: A BALANCED MODEL …
69
Stone, D. A. (1997). Policy paradox: The art of political decision making (Vol. 13). New York: W. W. Norton. Waldo, D. (1952a). Development of theory of democratic administration. American Political Science Review, 46(1), 81–103. Waldo, D. (1952b). “Development of theory of democratic administration”: Replies and comments. American Political Science Review, 46(2), 494–503. Weber, M. (2013). The protestant ethic and the spirit of capitalism. London: Routledge. Weiss, C. H. (1979). The many meanings of research utilization. Public Administration Review, 39(5), 426–431. Weiss, C. H. (1989). Congressional committees as users of analysis. Journal of Policy Analysis and Management, 8(3), 411–431. Weiss, C. H. (2001). What kind of evidence in evidence-based policy? Paper presented at the third international evidence-based policies and indicator systems conference, pp. 284–291. Durham, UK: CEM centre, University of Durham. Weiss, C. H. (2002). What to do until the random assigner comes? In F. Mosteller & R. F. Boruch (Eds.), Evidence matters: Randomized trials in education research (pp. 198–233). Washington, DC: Brookings Institution Press. Wildavsky, A. B. (1964). Politics of the budgetary process. Boston: Little, Brown. Wildavsky, A. B. (1974). Budgeting: A comparative theory of budgetary processes. Boston: Little, Brown.
CHAPTER 4
The Performance Movement and Other Evidence-Based Reforms
Evidence-based reforms have attempted to produce policy knowledge that can be used to improve decisions, often without clearly specifying which purposes they serve, which decision makers should utilize it, or which quality and coverage limitations should be considered. As a result, these reforms presume comprehensive consideration of policy knowledge and fully deterministic use of data, information, and evidence that offer a panacea for decision making. In the past 50 years, these attempts have become more formal in their pursuit to collect, interpret, rank, and utilize policy knowledge through standardized systems (Van Dooren et al. 2010).1 During this period, one of the best examples of advocacy for the use of policy knowledge to support decision making is associated with “the Performance Movement” (Radin 2006). In general terms, the performance movement has promised that policy knowledge could improve governance by measuring the extent to which program goals are being achieved, and by using the results to reward or punish programs. This goal often ran afoul of the limitations of policy knowledge, which might be able to tell whether a program was successful, but did little to identify
1 These performance efforts rarely, if ever, adopt analytical strategies to assess the impact of the selected programs, controlling for other factors. As a result, they tend toward generating data and information, but not evidence.
© The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_4
71
72
S. PUTANSU
whether one program was a higher priority than another, or how to compare the successes and failures of programs across different contexts and measures (Key 1940). This chapter focuses on performance efforts from the national government. Since 1965, there have been ten distinctly named federal performance reforms, with individual reforms from almost every presidential administration. This section shows that there have been repeated conflicts over the data, information, and evidence produced by these reforms and the needs of decision makers. In many circumstances, these conflicts led to additional calls for evidence-based governance as opposing parties accused each other of ignoring policy knowledge, repeating in cyclical promises and pitfalls from one administration to the next. While different in name, iterations of reform are largely examples of political repackaging, rather than substantive differences. Carolyn Heinrich (2002) compares the “evidence-based policy” and “performance management” movements and highlights differences in how they produce and value policy knowledge, while noting similar challenges that both face in linking the resulting data, information, and evidence to decision making. This supports the earlier claim that new reforms, when implemented, are often plagued with continued repetition of the same challenges and limitations. However, the reforms have had some unique characteristics and have reflected a gradual evolution in methods and data sources, accompanied by improved skills related to the implementation of performance efforts.
The Performance Movement: 50 Years of Evidence-Based Reforms The modern performance movement began on August 25, 1965, when President Lyndon B. Johnson described a new system for defining and prioritizing national goals, determining the most cost-effective way to achieve them, and achieving the best possible return for each dollar spent. Speaking to his cabinet and several heads of US agencies, he said, “The objective of this program is simple—to use the most modern management tools so that the full promise of finer life can be brought to every American at the least possible cost.” President Johnson’s Planning, Programming, Budgeting System (PPBS) 1965 initiative was the first to use a single, systematic approach across the entire national government. Similar efforts have been advocated by almost every president since. These
4
THE PERFORMANCE MOVEMENT AND OTHER …
73
systems have been buttressed, formalized, and entrenched into agencies by Congress, via laws passed in 1993 and 2010. According to the Congressional Budget Office (CBO), most US government spending on performance is absorbed within agency and program budgets as management expenditures. Nonetheless, CBO estimated in 2010 that a legislative effort to modernize performance systems would cost $75 million, above and beyond these absorbed costs, between 2011 and 2015. When implementing performance management systems, many stakeholders have interests and insights that could be represented, but this section focuses on three primary actors: the agencies, Congress, and the president—often represented by the Office of Management and Budget (OMB 2004). Figure 4.1 illustrates a timeline of these reforms and shows both executive and legislative efforts since 1965. This figure demonstrates the lasting allure of performance management, as very few years since 1965 have been without at least one active reform. It shows that Democratic and Republican presidents (represented by dashed lines and white boxes, respectively) have both used this as a management tool, and that each has replaced a prior “failed” reform. In fact, President George H. W. Bush is the only president between 1965 and 2015 who did not implement a government-wide performance measurement reform. However, he did implement an evidence-based reform centered on cost-benefit analysis. The figure also shows that Congress (represented by filled boxes) has institutionalized many practices of the performance movement through legislative action. From the presidential and congressional initiatives described above, the trappings of performance are now clearly evident throughout the federal government. Examples include the existence of new entities to focus on performance, OMB’s office of Performance and Personnel Management PPBS
8/27/1965 – 5/22/1973
Executive
MBO 5/22/1973 - 7/23/1976 PMI 7/23/1976 - 2/27/1977 ZBB 2/27/1977 - 1/20/1981 PIP
2/25/1986 – 1/20/1989
NPR
3/3/1993 – 1/20/2001
PART
1/7/2002 - 1/20/2009
Legislative
PMA GPRA 3/21/1993 - 4/2/2010
7/8/2013 Present
GPRAMA 1965
1973
1977
1981
1986
1989
1993
2001
Fig. 4.1 Timeline of performance management reforms
2009
2013
2015
4/2/2010 Present
74
S. PUTANSU
and the Performance Improvement Council—a collection of Performance Improvement Officers at each of the twenty-four largest US agencies. Examples of activities geared toward performance efforts include (CAP), agency priority goals, and individual program goals that touch every government program. Further, websites like performance.gov—which contains data and information on these goals—reflect the perceived importance of releasing this policy knowledge publicly. Managers across the federal government are now required to meet performance-reporting requirements, intended to justify their existence and funding by proving they can effectively achieve goals while efficiently expending resources. Each agency is required not only to identify goals and metrics of the programs they administer, but also to identify related goals across the federal government and how they might work together to achieve cross-cutting goals (Putansu 2015). Despite these improvements in management and use of performance information, these reforms have fallen short of their promise to fully prioritize and rationalize federal budgeting and management. Evidence-based advocates have made several promises related to these requirements. The primary stated goal of the performance movement has been to produce performance information that will support decision making, particularly in budget decisions. While this book is not focused specifically on performance management reforms, a brief exploration of their history will demonstrate the consistency with which this type of evidencebased reform has failed to consider all relevant policy knowledge and the purposes it can support. These limitations surface throughout the process of producing policy knowledge, with common flaws at the goal setting, data collection and analysis, and communication stages. Planning Programming Budgeting System (PPBS) The immediate precursor to government-wide performance measurement reforms was the implementation of PPBS by the Department of Defense (DOD) during the Kennedy administration. Robert McNamara introduced PPBS to DOD in 1961. Kennedy’s successor, President Johnson, initiated the first government-wide reform by ordering all federal agencies to implement PPBS in 1965. The reform would support the purpose of warning by better identifying national goals, projecting costs years into
4
THE PERFORMANCE MOVEMENT AND OTHER …
75
the future, and measuring performance. It would also support the purpose of guidance by helping choose the most urgent goals and identifying alternative, lower-cost means of achieving outcomes. For PPBS, the Bureau of Budget (which later became OMB) directed agencies to use PPBS to support planning, programming, and budgeting decisions. The agencies, in turn, focused on narrowly defined performance activities, largely directed by the Bureau of Budget, which used PPBS to target programs for budget increases and cuts. Congressional stakeholders were largely not involved in the process of setting these goals (Harper et al. 1969). Evidence suggests that agencies relied primarily on budgetary and program data that were already available to them, and tended to select goals where these data were available, resulting in incomplete coverage over program activities. Additionally, the reform was criticized for producing inconsistent or unreliable data and information, further reducing perceptions of the rigor and quality of the resulting policy knowledge. The Bureau of Budget summarized the agency performance in their congressional budget requests, but this agency-level summary of data provided insufficient depth for congressional deliberations. The Bureau of Budget released a report in 1968 that found agency functions had not been changed by PPBS. By 1973, agencies would no longer be required to use it, and in perhaps its most ferocious obituary, it was declared to have “failed everywhere and at all times” (Wildavsky 1974). The failure of PPBS to fulfill its intended purpose of guidance for congressional budget decisions was thus seen as a complete failure of the system. However, PPBS was continued in DOD and has evolved into the Planning, Programming, Budgeting, and Execution System (PPBE), which is still in use today. It has seen a resurgence in a number of other agencies (West 2011). Its success in these areas has been attributed to its utility for cost projection and performance monitoring at the agency level. This is clear evidence that, while not a panacea for decision making, the reform did have a meaningful and lasting impact on government decisions. Management by Objective (MBO) The end of government-wide PPBS was accompanied by the initiation of MBO, a performance effort championed by President Nixon in 1973. A fundamental tenet of MBO is that the focus would be on objectives
76
S. PUTANSU
rather than on activities. This reform directed agencies to develop objectives that would be reported to OMB, which would then measure performance toward those goals and use it to inform budget requests and as guidance for congressional budget decisions (Sherwood and Page 1976). While MBO’s focus on objective represented broader coverage than PPBS, issues related to data collection and analysis persisted. For example, agencies were aware that congressional priorities were not necessarily in alignment with OMB’s priorities, and submitted trivial objectives to OMB to deflect this conflict, causing the resulting policy knowledge to be deemed irrelevant by decision makers. Further, because goals were not clear and well defined, developing valid and reliable measures of performance was made more difficult, and the performance information produced by OMB was rarely used for decision making due to lack of rigor and poor quality. While OMB communicated the results of MBO to Congress through budget requests, agencies continued to present alternative policy knowledge to Congress for budget decisions, which was more likely to be cited as support for their deliberations. OMB eventually lost interest in MBO, abandoning its implementation by 1976. Presidential Management Initiatives (PMI) President Ford initiated PMI on July 23, 1976, after a series of meetings with department and agency heads. Ford believed that the budget process was a valuable tool for negotiating with agencies, which would serve “to improve the effectiveness of federal programs; to make agency operations more efficient; to foster governmental responsiveness to citizen and program clientele; and to increase operational accountability to the president, the Congress, and the public” (Haider 1979). As part of PMI, agencies would be required to assess their own programs, and these assessments would be communicated during the budget process. However, PMI never fully got off the ground, as a new president was elected approximately four months later, before any relevant policy knowledge could be produced. Zero-Based Budgeting (ZBB) As governor, Jimmy Carter had implemented ZBB in Georgia in 1971. After championing it throughout his presidential campaign, President Carter initiated ZBB and eliminated PMI through memorandum less than
4
THE PERFORMANCE MOVEMENT AND OTHER …
77
six weeks after taking office (Lauth 1978). ZBB required that each agency began with a “base” budget of zero and had to justify its existence as well as every expenditure it would make for each budget. ZBB required measurement of current levels of service that would serve as a comparison tool for budget decisions. Although a study three years after ZBB had been implemented in Georgia found that it had not been responsible for any shifts in resources, Carter proceeded with government-wide implementation as president (Hammond 1980). ZBB was different from the prior three reforms as the intended purpose of warning about poor performance was stripped away, leaving only a focus on guidance about specific budget amounts. ZBB was silent on goal definition, and assumed that goals were already agreed upon and clearly defined. Rather than gathering input from stakeholders, ZBB used existing budget lines to define its coverage. In addition, because the employees inside the agency were forced to justify all of their expenditures, they were incentivized to overstate need, resulting in challenges to the rigor of their analysis (Lee and Shim 1984). Additionally, ZBB required measurement of current levels of service that would serve as a comparison tool for budget decisions, which was problematic for government programs that produce services, like national defense or education, where such outputs are not readily measured. As a result, the quality of related policy knowledge was limited, reducing its utility for decision makers. The lack of stakeholder involvement, combined with challenges in the coverage and rigor of resulting policy knowledge, led to a lack of buy-in among decision makers who did not believe the related data and information were relevant and sufficient. Ultimately, policy changes that occurred during Carter’s presidency were not attributed to ZBB, but rather to changes in political power among Congress and presidential appointees. ZBB ended when Carter left office, as agency staff did not perceive any benefit of the system and discontinued its use. Productivity Improvement Program (PIP) President Reagan implemented PIP by executive order in 1986. PIP was a part of a larger reform, the president’s Management Improvement Program, also known as Reform ’88. The stated goal of PIP was to improve the “quality, timeliness, and efficiency of services provided by
78
S. PUTANSU
the federal government” (Hubbell and Kinghorn 1988) reflecting a guidance purpose. Like previous reforms, PIP also sought to integrate performance measures into the budget-making process. PIP required agencies to submit an annual productivity improvement plan, which identified areas where productivity could be improved. PIP relied on a private sector model of decision making to set goals, led by the presidentially named Grace Commission, a group of private sector CEOs who viewed Congress as a board of directors and assumed the president could and should act as a singular decision maker. They established a government-wide goal of standard cuts to budget totals that was ambivalent to actual agency goals and performance. The recommendation for standard budget cuts without consideration of program goals limited the relevance of the data, information, and evidence that were generated. Agencies worked to identify these cuts while also presenting positive performance narratives in an attempt to protect their programs, and these self-assessments were viewed as subjective and unreliable by decision makers on both sides of the aisle. The communication of policy knowledge from PIP was highly adversarial, as the Grace Commission complained about the role of Congress so much that they titled their report “Congressional Meddling.” Unsurprisingly, PIP was never shown to accomplish its goal of using performance measurement to increase productivity and modify the federal budget, as Congress ignored the policy knowledge provided by the Grace Commission and PIP in their decision making. Government Performance and Results Act (GPRA) During the Clinton administration, Congress took legislative action to promote performance, passing the Government Performance and Results Act (GPRA) in 1993. GPRA’s stated goals require agencies to set strategic objectives, reach performance targets, and report their results publicly. GPRA was designed in the Senate Committee on Governmental Affairs (CGA), largely influenced by the Republican membership. GPRA provided agencies with substantial discretion in specifying goals, collecting data, and providing information, and provided several years for agencies to prepare before official implementation in 1997. Many of GPRA’s original requirements are still in effect today as part of the GPRA Modernization Act (GPRAMA), which is also described below. These requirements create the basis for most current agency performance measurement and
4
THE PERFORMANCE MOVEMENT AND OTHER …
79
reporting. GPRA represents a watershed moment in performance management, as it formalized into law many of the requirements of previous presidential reform efforts. Policy knowledge produced by agencies to comply with GPRA included many sources of data, information, and evidence. It established formal expectations and requirements for agencies to set internal goals and processes, for OMB to have an oversight and management role, and for congressional interests to be represented. It supported all of these goals by establishing a broad framework for agency consideration of performance, which left detailed implementation to the agencies and OMB. As such, the act could be used to generate policy knowledge for a variety of different goals, using a variety of different data collection and analysis approaches, and communicated through a variety of platforms. Elements of these frameworks were present in reform efforts in both the Clinton and George W. Bush administrations, which faced similar challenges to previous reforms, as described below. National Performance Review (NPR) President Clinton, primarily through the direction of Vice President Al Gore, implemented the National Performance Review (NPR) in 1993. This reform attempted to establish a strong presence of the administration in negotiating agency objectives. NPR’s agenda was broad and multipurposed, and included efforts to affect policy, reorganize the bureaucracy, control budgets, empower bureaucrats, improve customer service, and change the ways decisions were made (Gore 1993). NPR highlighted an executive push toward “steering not rowing/” Even in the early implementation of NPR, there was concern that, like previous reforms, it would assume presidential dominance and ignore Congress, characterized by Rosenbloom’s (1993) urge to reformers: “Don’t Forget the Politics!” However, in combination with the GPRA, NPR urged agencies to work with congressional stakeholders and OMB on goal specification. That said, it was “anything but reluctant” to advocate restrictions of “congressional micromanagement,” and at times had an uneasy relationship with OMB, which was not significantly involved in the effort. Kettl and DiIulio (1995) explain that there are really three versions of NPR: (1) a political campaign to attract voters for Clinton, (2) spreading “reinvention gospel,” and (3) actual changes driven from within agencies.
80
S. PUTANSU
NPR was largely focused on areas of measurement and decision making internal to agency operations. As such, it lacked a centralized data collection and interpretation effort, relying instead on agencies to internally generate and use performance information. However, NPR was not without its challenges, as it suffered, like all the other performance efforts, from a consistent inability to measure all the different activities and outcomes being produced by the government, and related quality issues (Gordon 1994). Across the federal government, there were specific areas where this data, information, and evidence provided adequate guidance for improved efficiencies, but the overall thrust of NPR failed to provide these benefits consistently and broadly. Program Assessment and Rating Tool (PART) President George W. Bush revoked Clinton’s NPR reforms via memorandum in 2001, replacing it with the president’s Management Agenda (PMA). While PMA incorporated performance measures into a larger reform package, a more elaborate performance measurement system was enacted through OMB in 2003 and was first revealed as a part of the OMB annual budget in FY2004. This system, the Program Assessment and Rating Tool (PART), was focused on tying budget decisions directly to performance information. Despite rhetoric that PART was an extension of GPRA, that it was a collaborative effort with agencies, and that it was intended to be a supplement to other forms of information, it suffered from many of the same pitfalls of other performance efforts. First, goal specification was largely dictated by OMB, causing unrest among agency and congressional stakeholders. Several sources have highlighted where PART requirements and processes conflict with those laid out in GPRA (GAO 2004; Radin 2000; and others). Key among these differences, OMB established a scorecard for program performance that attempted to standardize and make comparable ratings across programs. However, the evaluative criteria in this scorecard were not meaningful to all the different kinds of programs implemented by the federal government. OMB dictated various definitions of “program” in order to establish the coverage of PART reviews. Ultimately, PART evaluated 1016 government programs, representing nearly 98% of the discretionary budget, over a five-year period. This created a wealth of data on government performance and program goals, but raised several conflicts over whether there
4
THE PERFORMANCE MOVEMENT AND OTHER …
81
was a fair and consistent unit of analysis. Further, the standard criteria raise issues with the rigor of analysis. Specifically, programs where measurement is more difficult were often rated poorly for not having simple, well-defined measures. Finally, OMB examiners were often viewed as subjective, in violation of the assumption of neutral information, making them a distrusted source. While OMB stated explicitly that the information from PART would not be considered alone or be deterministic in decisions, the process was often different in reality. PART scores seem to explain some variation in the president’s budget proposal communicated to Congress, but did not reflect congressional goals. In combination with challenges to the quality of the policy knowledge produced, this may help explain why PART scores were largely unrelated to the budgets passed by Congress during this period (Gilmour and Lewis 2006). GPRA Modernization Act (GPRAMA) In 2010, Congress reauthorized and amended GPRA in order to address certain shortcomings that had been identified. The primary area of changes to GPRA related to the production of information. Specifically, low levels of utilization of GPRA (and PART) information were diagnosed as a problem, and several requirements were added to the law in an attempt to solve it. Although GPRAMA built on the experience of both GPRA and PART, PART was not mentioned explicitly in the legislation. The added requirements include: (1) disclosure of information about accuracy and validity, (2) data on complex, cross-cutting policy areas, (3) quarterly reporting on priority goals on a publicly available website to provide transparency to the public, (4) instilling sustained leadership commitment and accountability for achieving results, and (5) engaging Congress in identifying management and performance issues to address. Like GPRA, this is a broad framework rather than a detailed implementation plan. GPRAMA assigns responsibilities to a Chief Operating Officer and Performance Improvement Officer in each agency and to OMB to improve agency management and performance. By including provisions about multiple and crosscutting goals, GPRAMA offers an opportunity to examine interactions among and conflict between multiple goals, which may increase the number of decision makers who find the information relevant. The legislation continues to place an emphasis on agencies and OMB as the source of information.
82
S. PUTANSU
This will likely continue to result in detailed collection of data by agencies, pressure from OMB to promote executive priorities, and uncertainty about the independence and objectivity of results. Further, because it provides for flexibility in implementation, the information produced may reflect a variety of quality types. As with GPRA, this means that information can be used for a variety of purposes, but essentially guarantees an inconsistent utilization across different programs. President’s Management Agenda The Obama administration was the first to guide implementation of GPRAMA. President Obama’s March 2014 budget outlined the details of a new president’s Management Agenda to achieve this goal. Under Obama, OMB implemented several GPRAMA requirements, including identifying fifteen CAP goals, beginning to develop a comprehensive list of federal programs, issuing guidance related to agency use of data-driven performance reviews, and encouraging public managers to utilize performance information and evaluation in their decision making. This last effort is different from previous presidential reforms, as there is no explicitly stated goal of linking this policy knowledge to budget decisions, but rather a more instrumental attempt to comply with GPRAMA requirements and improve internal agency decision making. One important aspect of the PMA is OMB’s effort to address CAP goals. Each goal has two leaders, one from OMB and another from the agency seen as most responsible for achieving the goal. By centralizing the responsibility for CAP goals, OMB has attempted to ensure appropriate coverage of agency interests. They also facilitated quarterly review meetings and CAP goal plans, which encourage increased interaction intended to improve stakeholder engagement. While a recent study has shown that the benefits of this collaboration are unequal across CAP goals (Putansu 2015), this attempt at creating policy knowledge spanning multiple agency boundaries is an innovation in the performance movement. Under the PMA, agencies were given broad flexibility to lead the development of goals in their strategic plans, with associated data gathering and analysis of related performance metrics. They were encouraged further to produce complementary evaluation information to fill gaps in knowledge that could not be addressed by performance information. Agencies have
4
THE PERFORMANCE MOVEMENT AND OTHER …
83
demonstrated that this policy knowledge, communicated through quarterly performance reviews, has supported a range of internal decisions. Many practices established under the PMA have continued as agencies implement GPRAMA requirements. Assessing the Performance Movement While President Johnson’s original intent with PPBS was simple to describe, it has proven much less simple to achieve. Fifty years and ten performance measurement reforms later, according to the GAO (2011), the US government continues to face many pressing challenges. Despite all these attempts to link performance to the budget, the federal budget is on an unsustainable path. Performance reforms have, in some instances, informed aspects of decision making, but they have typically fallen short of their promises to replace politics in decision making. As shown in Table 4.1, these failures can be attributed, at least in part, by failure to acknowledge the limitations of the data, information, and evidence they produce or the relevance of other sources of policy knowledge. They are also driven by the evidence-based proverb to the extent they haven’t properly accounted for the continued importance of political factors. These challenges reflect two findings about the failures of evidencebased reforms. First, the presence of multiple and often conflicting goals cast doubt on the ability of a single performance effort to effectively produce policy knowledge for all purposes of decision making. In other words, the reality and complexity of government programs and policies prohibit the use of a one-size-fits-all approach. Second, misalignment of goals at institutional levels can lead to a mismatch of incentives in data gathering and analysis. Further, beyond issues of goal conflict, there are limitations associated with data gathering and analysis. Specifically, gathering data that are relevant to stated goals, sufficient and appropriate for understanding performance, and accurate, valid, complete, and reliable for interpretation of results is a resource intensive and technically challenging activity. Limitations in data quality are often cited as reasons for the lack of utilization of performance information. Even the best research design may suffer from problems of bad data. Rittel and Webber (1973) describe these policy issues—where measurement is difficult and multiple desired values are constantly being re-evaluated and balanced against each other—as “wicked problems.” In cases like this, performance efforts often use outputs or intermediate outcomes as a proxy for the desired ultimate
84
S. PUTANSU
Table 4.1 Key shortcomings of performance management reforms Reform
Limitations of policy knowledge
Influence of political factors
PPBS
Stakeholders excluded, limited coverage of program activities Unclear goals, low quality data
Institutional role of Congress overlooked Agency defended their interests directly to Congress using alternate information Easily ended due to lack of institutional development (change in administrations) Not aligned with Institutional process of budget setting in Congress Institutional role of Congress challenged, Conflict between agency interests and ideological preferences of President Goals change with ideological and interest group influence. Timelines not consistent with institutional processes (election cycle) Institutional role of Congress overlooked
MBO
PMI
ZBB
Production of policy knowledge never reached communication stage Inflated estimates of need, current services not reliably measured
PIP
Focus on economy misaligned with effectiveness, efficiency, and equity
GPRA
Initial focus on outputs rather than outcomes, aligned with agency needs more than congressional needs
NPR
Successful for narrowly tailored policy knowledge to solve known problems, less so for strategic decision making Information quality challenged for many programs, as was the definition of program and overall scope Inconsistent production and quality of policy knowledge across agencies and policy areas
PART
GPRAMA
PMA
Inconsistent production and quality of policy knowledge across agencies and policy areas
Institutional role of Congress challenged, institutional processes of agencies not taken into account Cross-agency requirements challenge existing institutional arrangements. Goals change with ideological and interest group influence from external context Competing agency interests may limit implementation of CAP goals
outcome. However, these intermediate targets are not perfect predictors of the ultimate goals and offer competing decision makers with incentives to rely on alternative sources of policy knowledge.
4
THE PERFORMANCE MOVEMENT AND OTHER …
85
Evidence-Based Policy: The Modern Reform Agenda As noted above, the Presidential Management Agenda of the Obama administration emphasized evaluation, in addition to performance reforms, to enhance federal decision making. This is just one of many evidence-based reforms that have gained traction alongside performancecentered efforts. For example, the Digital Accountability and Transparency Act of 2014 created requirements for federal agencies to increase the quality of their financial data, and to improve the availability of that data for analysis. Support for evidence-based reforms continued, and cut across partisan boundaries, as evidenced in 2017 by the bipartisan Commission on Evidence-Based Policy’s unanimous adoption of 22 recommendations to improve policy knowledge in the federal government (Meyer and Mittag 2019). This was followed by the 2019 passage of the Foundations of Evidence-Based Policy Act, which enshrined many of these recommendations into law. This act has a strong focus on the quality and availability of agency data. However, it also created requirements for agencies to comprehensively assess policy knowledge and create plans to develop data, information, and evidence to fill those gaps. In many ways, modern evidence-based reforms have sought to go beyond performance reforms by highlighting the importance of evidence rooted in causal analysis to support management and policy decisions. However, they have elevated a specific kind of causal analysis, rooted in randomized control trials. A recent example of this narrow focus is a recent volume by Haskins and Margolis (2014), which examines the use of policy knowledge in the Obama administration, implicitly and narrowly focusing on aspects most closely related to RCT and evaluation. This text discusses these evaluations as if they were the only needed source of policy knowledge, offers similarly prescriptive promises of improved decision making, while making no mention of previous evidence-based reforms. Further, as with performance reforms, these other evidence-based efforts tend to focus on a subset of program goals, without accounting for the ambiguity, complexity, and conflict among them. This is exemplified by another text (Nussle and Orszag 2014) which presents arguments for increased evidence-based governance. These arguments use broad terminology of data, information, and evidence, but rely on an implicit preference for cost effectiveness measures, without consideration of other forms of policy knowledge.
86
S. PUTANSU
Decision Making in Federal Education Policy Section Two of this book turns to federal education policy, applying the frameworks developed in Section One to the two largest federal education programs—Title I of the Elementary and Secondary Education Act and federal student loans for higher education. These cases provide insight into the production of policy knowledge, its diffusion among the relevant decision makers, its use for different decision making purposes, and the impact of that use. The exploration and comparison of these two cases provide a contextually rich narrative that better demonstrates the range of potential political and rational influences and decision outcomes, and differences in policy knowledge types, purposes, and impacts. This will provide insight into the conditions under which data, information, and evidence can contribute to or lead the political decision making process. Chapter 5—The Roots of Modern Federal Education Policy—outlines the historical context for modern federal education policy generally, as well as for each program individually. This context highlights the important considerations for each program within Weiss’ Four-I framework. Chapter 6—Improving Access to Higher Education—and Chapter 7—Education for the Disadvantaged—present detailed examinations of federal student loans and Title I, respectively. These chapters analyze the policy knowledge produced about and used by decision makers for each program for decisions since 1965. Chapter 6 explores the provision of student loans through federal student aid, as established in the Higher Education Act, from 1965 through 2015. Since student loans are similar to a production program, it is easy to measure both process and outcome, and policy knowledge is expected to have a relatively strong role, relative to political factors. This ease of measurement is also expected to result in a reduced need for contextually sophisticated data, information, and evidence—enabling greater use policy knowledge to impact decision outcomes. Chapter 7 explores Title I of the Elementary and Secondary Education Act, from 1965 through 2015. As this is most similar to a coping program, it is difficult to measure both the process and the outputs of the program, and policy knowledge is expected to have a relatively weak role, compared to political factors. Further, difficulties in measuring will be reflected as a limitation in any resulting data, information, and evidence, limiting its use primarily to use by political actors in support of their preexisting opinions.
4
THE PERFORMANCE MOVEMENT AND OTHER …
87
References Gilmour, J., & Lewis, D. E. (2006). Does performance budgeting work? An examination of OMB’s PART scores. Public Administration Review, 66(5), 742–752. Gordon, J. (1994). A critique of reinventing government. Spectrum, the Public Policy Journal of the Kennedy School of Government, Harvard University. Gore, A. (1993). Creating a government that works better and costs less: Report of the national performance review. Washington, DC: Government Printing Office. Government Accountability Office (GAO). (2004). Performance budgeting: OMB’s program assessment rating tool presents opportunities and challenges for budget and performance integration, GAO-04-439T. Washington, DC: GAO. Government Accountability Office (GAO). (2011). GPRA modernization act provides opportunities to help address fiscal, performance, and management challenges, GAO-11-466T. Washington, DC: GAO. Haider, D. (1979). Presidential management initiatives: A ford legacy to executive management improvement. Public Administration Review, 39(3), 248– 259. Hammond, T. H. (1980). A zero-based look at zero-base budgeting: Why its failures in state government are being duplicated in Washington. New Brunswick, NJ: Transaction Books. Harper, E. H., Kramer, F. A., & Rouse, A. M. (1969). Implementation and use of PPB in sixteen federal agencies. Public Administration Review, 29(6), 623–632. Haskins, R., & Margolis, G. (2014). Show me the evidence: Obama’s fight for rigor and results in social policy. Washington, DC: Brookings Institution Press. Heinrich, C. J. (2002). Outcomes-based performance management in the public sector: Implications for government accountability and effectiveness. Public Administration Review, 62(6), 712–725. Hubbell, L., & Kinghorn, C. M. (1988). President Reagan’s productivity improvement program: A prescription for success? Public Productivity Review, 11(4), 1–10. Kettl, D., & DiIulio, J. J. (Eds.). (1995). Inside the reinvention machine: Appraising governmental reform. Washington, DC: Brookings Institution Press. Key, V. O. (1940). The lack of a budgetary theory. American Political Science Review, 34(6), 1137–1144. Lauth, T. P. (1978). Zero-base budgeting in Georgia state government: Myth and reality. Public Administration Review, 38, 420–430. Lee, S. M., & Shim, J. P. (1984, October). Zero-base budgeting—Dealing with conflicting objectives. Long Range Planning, 17 (5), 103–110. Meyer, B. D., & Mittag, N. (2019). Using linked survey and administrative data to better measure income: Implications for poverty, program effectiveness,
88
S. PUTANSU
and holes in the safety net. American Economic Journal: Applied Economics, 11(2), 176–204. Nussle, J., & Orszag, P. R. (Eds.). (2014). Moneyball for government. Washington, DC: Disruption Books. Office of Management and Budget (OMB). (2004). Rating the performance of federal programs. Budget for fiscal year 2004. Washington, DC. Putansu, S. R. (2015). Cross agency priority goals in the US government: Can directed collaboration be a stepping stone toward politic-centered performance? Policy and Society, 34(1), 25–35. Radin, B. A. (2000). The government performance and results act and the tradition of federal management reform: Square pegs in round holes? Journal of Public Administration Research and Theory, 10(1), 111–135. Radin, B. A. (2006). Challenging the performance movement. Washington, DC: Georgetown University Press. Rittel, H. W., & Webber, M. M. (1973). Dilemmas in a general theory of planning. Policy Sciences, 4(2), 155–169. Rosenbloom, D. H. (1993). Have an administrative Rx? Don’t forget the politics! Public Administration Review, 53(6), 503–507. Sherwood, F. P., & Page, W. J. (1976). MBO and public management. Public Administration Review, 36(1), 5–12. Van Dooren, W., Bouckaert, G., & Halligan, J. (2010). Performance management in the public sector. Oxon: Routledge. West, W. F. (2011). Program budgeting and the performance movement: The Elusive quest for the efficiency in government. Washington, DC: Georgetown University Press. Wildavsky, A. B. (1974). Budgeting: A comparative theory of budgetary processes. Boston, MA: Little, Brown.
PART II
Politics and Policy Knowledge in Federal Education
CHAPTER 5
The Roots of Modern Federal Education Policy
The two cases in this book, Federal Student Loans and Title I Grants for the Economically Disadvantaged, represent the two largest federal expenditures on education. The passage of the Higher Education Act (HEA) and the Elementary and Secondary Education Act (ESEA) in 1965, which created these programs, represents the beginning of the “modern era” of federal education policy. As described in Chapter 3, Weiss’ Four-I framework, which outlines the expected roles of ideology, interests, institutions, and policy knowledge, establishes expectations for the extent to which decisions will be rational or political. Further, Wilson’s typology of bureaucratic organizations provides a framework for understanding the different expected types and purposes of information for each program. This chapter outlines the history of the federal role in education from 1635 through passage of HEA and ESEA in 1965. In doing so, it introduces the dominant ideologies, interests, institutions, and policy knowledge that served as the foundation of each case. Further, it shows how these factors led to the creation of a coping program, in the case of Title I, and a production program, in the case of student loans. This historical-political context is critical to understanding the evolution of these programs after their passage. This chapter will discuss different sources of data, information, and evidence that influenced education policy from the colonial era until 1965, in order to illustrate areas of general agreement and conflict, and to examine the role of policy knowledge in establishing and modifying education policy over this period. By understanding the foundations of these programs, it will become possible to © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_5
91
92
S. PUTANSU
observe changes in the four-I’s over time and to better understand how each contributes to major decisions throughout the programs’ histories. Because 1965 marks the passage of both HEA and ESEA, the two legislative acts that represent the foundation for nearly all federal involvement in education, it represents the beginning of the modern era of federal education policy. Although federal involvement in education stretches back before these acts, state and local governments have always had primacy. As a result, the early history of education shows a limited federal role in elementary, high school, and college education. This chapter discusses the history of US education in two parts. First, it describes the initial expansion of education throughout the United States. Specifically, it discusses the two major transformations in American education related to ESEA: (1) the diffusion of free, publicly funded common schools, and (2) the high school movement. Each transformation is marked by the moment that the proportion of American children receiving education at a given level surpassed 50%. This section also reviews education trends related to HEA. However, there has not been a similar transformation in higher education, which would be marked when more than 50% of young adults earn a college degree. In 2018, 35.1% of people aged 25 and over had earned at least a bachelor’s degree while 45.3% had attained an associate’s degree or greater. This section ends with a discussion of federal precursors to HEA and ESEA, which occurred in the 1940s. These include federal creation of the General Educational Development (GED) program for World War II veterans in 1942 and the publication of the Truman Commission Report in 1947. These represent some of the earliest arguments for expanded federal involvement in primary and post-secondary education. Following this examination of the history of federal involvement in education, this chapter will turn to the passage of each of the federal education laws underpinning the book’s two cases. By examining the passage and contents of each act, the chapter identifies key decision makers, describe the political context surrounding the passage of both bills, and identify any policy knowledge that may have influenced the final outcome. These sections will outline the goals of each program and how the respective programs are designed to achieve them.
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
93
Three Transformations of Education in the United States Despite attempts to empower the national government to promote education during the Constitutional Convention (Farrand 1911), its ultimate absence from the document, and the Bill of Rights, means it was “reserved to the States respectively, or to the people.” As a result, all federal involvement in education begins with the question of the federal role. Specifically, should the federal government provide assistance for a given educational need? As this chapter and the following chapters show, federal education programs and policies often revisit this question of the scope of federal involvement, and the role of state and local governments. Because the Constitution does not grant the federal government direct authority for education, all federal initiatives must work within the context of various state constitutions and systems that offer public education (Kaestle and Lodewick 2007). A further exploration of the history of education in the United States, including examples of the historical federal role, provides important context for understanding modern education policy. This chapter examines this history as it relates to three transformations of education in the United States. Specifically, it will start by tracing the changes that led the United States to develop a locally driven system of common schooling that served a majority of children in the country. Second, it will examine the factors that expanded this majority of attendance to cover completion of high school. Finally, it will discuss the ongoing demographic trend toward a majority of citizens completing a higher education degree. Following discussion of these three transformations, the chapter will turn to the passage of ESEA and HEA, and a discussion of how these acts reflect a departure from the traditional federal role. Toward a Majority of the Population in Common Schools The public relationship with education can be traced back to the earliest days of the republic. In 1635, revenue from a sale of public lands was used to partially fund a school in Boston, the first instance of a “public school” in the United States (Bailyn 1972). In 1648, the first property tax for schools was levied by the city of Denham, Massachusetts (Pulliam 1982). However, the first federal support of education did not occur until about
94
S. PUTANSU
150 years later, nearly ten years after the Declaration of Independence— when the Continental Congress passed the Land Ordinance of 1785— followed two years later by passage of the Northwest Ordinance (Jenkins and Hill 2011). These early laws relied on a coherent Protestant ideology (Kaestle 1983), noting that “as religion, morality, and knowledge are essential to good government, schools and the means of education shall forever be encouraged” (Cubberly 1919). However, this broad statement about equality goals fails to fully explain the broad American consensus on the importance of education, which was also intended to maintain order, provide economic advancement, and ensure a supply of craft labor (Kaestle 1983). During this same period, state laws encouraging public education were passed in Massachusetts, New York, and Connecticut, each with different requirements. In addition to federal and state efforts, the reach of education grew in the early nineteenth century, in part, due to the development of a Lancasterian system of teaching. This system, in which older children could lecture younger children, expanded the ability of school masters to teach larger numbers of students. As schools consolidated under this system, charity schools began to receive a larger share of education money, and eventually became the norm for common schooling (Kaestle 1983). Although education was not in the Constitution, the founding fathers acknowledged the importance of education, particularly in regard to maintaining order in a republic as large as America (Kaestle 1983). In fact, the federal government has provided support to education since at least 1785, as evidenced by the land ordinances passed by the Continental Congress, which provided land for common schools throughout the early days of the republic. Specifically, the ordinance granted land in the form of congressional townships to states as they were admitted to the union, reserving one section of each township for the building of schools. This practice continued with each new state through 1850, when California was admitted to the union and the land ordinance was amended to allot two sections of each township for schools (Goldin 1999). One estimate of the value of this program stated that “in all, the national government has given to the states for common schools […] a total of approximately 132,000,000 acres of public lands.” The value of this land was estimated at $165,000,000, assuming a price of $1.15 an acre (Cubberly 1919). Although there are limitations to this methodology, adjusting for inflation from 1919 to 2018 dollars yields an estimate of nearly $2.4 billion. This
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
95
would indicate an average subsidy for education of approximately $17.9 million per year from 1785 to 1919, in 2018 dollar terms. This is a small amount relative to today’s federal contribution to education, representing about 0.1% of the 2018 budget for Title I grants, which was about $14.9 billion. The amount could be considered negligible when compared to total federal spending on K-12 education, which totaled more than $56.5 billion in 2018. Land for education was the primary form of early federal support for education (Hirschland and Steinmo 2003) through the early nineteenth century. Combined with the locally driven effort to freely and openly provide common education (Heidenheimer 2017), this helped to create one of the most comprehensive and universal educational systems of all nations (Kaestle 1983). Through 1957, the federal role remained very limited, and education policy was primarily driven at the local level by individual school districts (Kaestle and Lodewick 2007). While the federal role has expanded since then, it still pales in comparison with state and local control. In addition to providing substantial lands dedicated to the creation of publicly funded schools, the federal government also indirectly influenced the first transformation of education (Goldin 1999). Prior to the signing of the Constitution in 1789, education was treated as a private good and generally provided in religious settings (Cubberly 1919). The writers of the Constitution, recognizing the differences in religion and sects that composed the colonies, affirmed the right to free exercise of religion and prohibited Congress from establishing a state religion. Educational reformers in the first half of the nineteenth century mounted campaigns against private schooling to increase momentum toward centralization of common schools, but faced resistance. Their opponents preferred neighborhood control, paid schools for students whose families could afford it, and parental discretion over the amount of schooling needed. Widespread consensus about the value of public schooling, however, would ultimately help the reformers to emerge victoriously (Kaestle 1983). By the time of the Civil War, statistics indicated that the majority of children in the United States were attending common school (Goldin 1999). These statistics, however, undercount the proportion of AfricanAmerican students, who were not expected to go to school and were, as a rule, excluded from measurement. This statistical milestone preceded compulsory education laws passed by the states, as the first of these was not passed until 1852 in Massachusetts and others were not adopted by all
96
S. PUTANSU
states until 1918 (Goldin 1999). The compulsory education law in Massachusetts, championed by Horace Mann, and related laws in the other states helped to increase the proportion of students in common school toward the current levels (Kaestle and Vinovskis 1978). However, while enrollment was increasing in absolute terms in the cities, the proportion of children enrolled in urban schools remained consistent over this period (Kaestle 1983). The federal government played a minor role, at most, in this transformation. Education was developed primarily by local governments, and even by the time a majority of students attended common schools, those schools were mostly publicly operated but often only partially publicly funded (Cubberly 1919). Funding and decision making for schools were also driven locally, either by townships or school districts (Pulliam 1982). As population densities grew in cities, social problems became more visible, and this helped reformers in the 1830s–1860s to create urgency for school reforms to help maintain social order. During this period, moral education was a higher priority in schools than intellectual education (Kaestle 1983). Around the same time that compulsory education laws proliferated across the states, another shift was occurring in the philosophical approach to education. Most schools at that time required students to pass tests to move onto the next grade, and the overarching focus on schooling was on attainment of specific knowledge (Gamson 2007). As a result, many students were held back for missing one or more subjects on a test and were never given the opportunity to advance to the next grade level, even if they were skilled at other subjects. According to Gamson, the approach to education underwent a shift, driven by progressive reformers, toward equal opportunity and an acknowledgment of the inherent differences in student abilities. The Education Policies Commission embodied this shift by decreeing the idea of provision of the same education to all people as “naïve egalitarianism” and instead promoted raising “every individual to the highest of their capacities.” This shift established a goal of assuring that each student received an equal opportunity to learn, including the ability to advance in subject areas they excelled, even when unable to pass tests in other subjects at a given grade level. However, differences in test results would be used later to deny transfers intended to encourage desegregation, including tests for “mental maturity” required of students going into the first grade (Reed 2014). Federal involvement continued to evolve during the early period of American education, with important decisions occurring in 1867, 1869,
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
97
and 1939. First, in 1867 Congress expanded the federal government’s role by creating an agency to “collect statistics and facts concerning the condition and progress of education in the several states and territories, and to diffuse information respecting the organization and management of schools and school systems and methods of teaching” (Snyder 1993). To accomplish this, they created a new Department of Education, which was tasked to report on the states in aggregate, which required the states to share disaggregated information with the federal government. The department mostly produced descriptive information about the status of education, but also generated guidance through the identification of good practices. Two years later, Congress revisited this decision and decided to reduce the autonomy of the department but not its mission. It was reduced to the rank of Bureau, renamed the US Office of Education (OE), and placed within the Department of Interior (Cubberly), where it remained until 1939. At this point, it was moved into the Federal Security Agency, where it stayed until the Department of Health Education and Welfare (HEW) was founded in 1953 (Goldin 1999). It remained a part of HEW until 1989, where it was once again created as a separate Department of Education (Radin and Hawley 1989). A Majority of the Population as High School Graduates The second transformation of American education happened much more quickly than the first and was influenced most heavily by changing economics and demographics, rather than by education policy at the local, state, or federal level. During this period, growth in demand for workers led to growth in the number of schools and increased demand for schooling beyond the 8th grade. Between 1910 and 1940, the percentage of youth completing high school expanded greatly, from less than 10% to more than 50% (Goldin 1999). However, the federal government did contribute to this demographic shift indirectly, starting decades earlier, with two initiatives championed by President Abraham Lincoln in 1862. In particular, the Homestead Act incentivized migration and amplified the already growing need for skilled farm laborers while efforts to build a transcontinental railroad, created demand for workers with mechanical and engineering skills. The associated revolution in transportation, in its integration of the nation, served as a touchstone for state involvement in economic and social issues (Kaestle 1983). Additionally, the first baby boomers entered elementary school in the early ’50s, and greatly
98
S. PUTANSU
expanded the number of pupils served from 25.5 million in 1955 to a peak of 34.0 million in 1970. This led to a greater need for schools, teachers, and resources. These demographic changes, coupled with an increased number of students who had completed primary education, drove the movement for children to attend and complete high school (Hansen 1963). During the same time period, high school enrollment grew from 8.0 million in 1955 to 14.7 million in 1970, and peaked a few years later in 1977 at 15.8 million. However, unlike elementary enrollment, high school enrollment eventually surpassed this peak in 1997. The various sources of increased demand for education also increased national support for and financing of education. Further, the launch of Sputnik and the Cold War were driving a greater national desire to expand access to education, and rising federal revenues were increasingly being seen as a way to assist states and local government entities, who were experiencing budget contractions. Further, because investments in schools were reduced during World War II, most public schools were in need of repair, and there were great disparities in the quality of schools across districts, which further heightened the need to quickly increase investments (Goldin 1999). These expansions of federal involvement, however, were met with continual apprehension and a desire to maintain the diffuse local educational control that had persisted for nearly 200 years, highlighting the continuing conflict in education policy between national and local control. The New Deal signaled the most direct federal involvement in education by providing “impact aid” to school districts that were adversely impacted by the presence of government institutions, usually military bases. This funding was based on a formula where, when school districts could show that they would make up to 3% more revenue through property taxes in the absence of the installation, the federal government would compensate them for the difference (Hovey 1965). This formula would be a critical component of the eventual watershed moment in elementary and secondary education policy. Combined with impact aid, the list of federal programs remained short through the late 1950s, limited to vocational education support, subsidies for school lunches, and funding for education of Native Americans (Kaestle and Lodewick 2007). Around this same time period, social and behavioral sciences were making technical advances in the ability to evaluate and understand the interrelationships in the educational system. This led to the development of
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
99
policy knowledge and a better understanding of the links between education, the economy, and the values of American society (Clark 1963; Schultz 1959; Gardner 1961). This increased understanding also led to early studies that generated warnings about various failings of the educational system, and guidance to address them (Conant 1961; Clark 1963; Gardner 1964). These findings increased pressure for education policy but were not initially enough to spur significant federal intervention. The following sections will identify the barriers that had to be overcome for passage of the HEA and ESEA in 1965. It will show how differences in the Four-I’s contributed to differences in program design, which ultimately impacted how data, information, and evidence would be produced, considered, and used to support each. Toward a Majority with a College Degree? The United States continues to move gradually toward a population where a majority completes post-secondary education, a trend that arguably has roots stretching as far back as 1636, when Harvard University (originally Harvard College) was founded (Rudolph 1990). However, in 2018, approximately 35% of the US population age twenty-five or older had a four-year degree or higher, well short of a majority. As with common schooling, the federal role in higher education has been limited from the earliest days of the republic, due to ideological preference for local control. Even when President George Washington recommended the establishment of a national university in his first presidential address in 1790, and set his own money aside to fund such a university, Congress resisted and maintained the states’ power over education (Thomas 2014). The expansion of the federal role in higher education has roots in the Civil War and the Morrill Land Grant Act of 1862. During the Civil War, Congress passed the Homestead Act and an act to construct an intercontinental railroad, which drove expansion of the citizenry westward. Along with these acts, Congress passed the land grant act to provide land to the states for the purpose of establishing colleges of agriculture and mechanical arts (Goldin 1999), which concentrated benefits in the expanding territories. While higher education had globally been reserved to elites across the globe, these early efforts began to make it accessible to a wider demographic in the United States. These expansions were driven partially
100
S. PUTANSU
by population growth, which necessitated improvements in food production, and by the Civil War, which drove an increased need for engineering and military education (Hansen 1963). In 1890, a second land grant act was passed, establishing annual appropriations for land grant institutions, and ensuring continued federal support of higher education. This act also expanded the recipients of the federal benefits by establishing grants for non-white colleges—now known as historically black institutions of higher education (Preer 2012). The federal government also provided additional support for agricultural education in 1887, when a provision in the Hatch Act provided support for experiment stations; in 1914, when Congress set up agricultural extensions; and in 1917, when the Smith-Hughes Vocational Act provided funding to agriculture, industry, and home economics education. This act also created the Federal Board for Vocational Education (Goldin 1999).
A Federal Role for Elementary and Secondary Education As the majority of the population of the United States attended common schools and continued through high school graduation, the federal government began to expand its role into elementary and secondary education (Conant 1959). To do so, it had to recognize the reality of these first two transformations of US education and contend with the dominance of state and local governments. This section will describe the factors that led to ESEA’s passage, and three highly contentious issues that had to be overcome for a successful passage. Specifically, during the 1950s and 1960s, most attempts at education reform were met with resistance related to socialism, religious schools, and racial tensions (Bailey and Mosher 1968). Federal Education Policy as Socialism Congressional attempts to expand the federal role began in earnest in the 1950s, while President Eisenhower was in office. However, education bills during the period (1953–1961) largely failed because of his opposition to the “reds.” Specifically, he argued that federal intervention in education policy would amount to a socialist or communist intrusion on American society that would be seen as a victory for the Soviet Union. With this
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
101
justification, Eisenhower was able to successfully maintain a coalition of opposition against federal efforts to support education. Ironically, the actions of the Soviet Union helped reverse this sentiment. Specifically, as the Cold War continued, the United States and Soviet Union continued to compete for dominance on the world stage. This competition was not limited to military power and nuclear arsenals, but spread to competition over education, economy, and the space race. Advances by the Soviets, particularly in the area of space exploration, led to calls across the United States to strengthen science and math education, and reduced the overall resistance to federal involvement. This created an opportunity for federal expansions in education when President Kennedy took office. Church and State: The Case of Religious Schools President Kennedy’s first attempts at an education bill in 1961 and 1962 failed due primarily to debate over religious schools. Specifically, the question of whether religious schools could receive funds if those funds were earmarked for non-religious purposes. Religious schools, which were predominantly Catholic, were a powerful interest that resisted the passage of any bill that did not provide any means of distributing at least some funds to religious schools. Dichotomous language in the US Constitution—that (1) “Congress shall make no law respecting an establishment of religion” or (2) “prohibiting the free exercise thereof”—has been one of the major sources of ideological conflict in the area of public education. As described above, the early history of elementary and secondary education began with religious schools, which served as the primary common schools during the early years of the nation. The first part of the Constitutional language can be interpreted to mean that the government cannot provide support in any form to religious institutions, including religious schools, while the second can be interpreted to argue for equal federal support of education, regardless of whether it occurs in a public or religious school (Bailey and Mosher 1968). The importance of this distinction before the World War II was limited, as the church generally had resources to run their schools, but inflation and a shortage of teaching staff eventually made this difficult, and the religious schools began seeking public funds for non-religious activities, such as school health and transportation. Because some state constitutions
102
S. PUTANSU
prohibited all direct and indirect aid to religious schools, federal support was of great importance to religious schools. Public policy to define the proper separation of church and state in regard to education developed gradually and unevenly at state and local levels through the early twentieth century, before becoming more of a focus at the federal level (Kaestle 1983). The Supreme Court clarified some of the boundaries of the freeexercise and non-establishment clauses of the Constitution through three cases. In 1947, they ruled in Everson v. Board of Education that the federal government could finance non-religious activities provided by religious institutions under the free exercise clause. Then, in 1948 and 1952, the Court ruled in McCollum v. Board of Education and Zorach v. Clausan, under the establishment clause, that public schools could not provide religious instruction. In President Kennedy’s 1961 and 1962 proposed bills, federal aid was prohibited from going to religious institutions. The National Catholic Welfare Conference successfully lobbied to prevent those bills from being enacted. This obviated a need for a compromise to be made between this group and the National Education Association, who opposed funding religious institutions because their teachers were not a part of the union (Bailey and Mosher 1968). Moving forward from these bills, lawmakers knew that, in order to pass federal legislation related to elementary and secondary education, they would have to include provisions that enabled the use of federal funds by religious schools for non-religious purposes. Separate Is Unequal: Race and Education Following President Kennedy’s assassination, bills sponsored by Democratic Senators Perkins of Kentucky and Morse of Oregon in 1964 failed primarily because of race-related issues. During this time, Democratic Representative Powell of New York offered a standard amendment to all bills which required that federal funds be denied to any jurisdiction that maintained segregation (Fiel and Zhang 2019). Resistance to this amendment by states, localities, and interests helped to maintain congressional opposition that caused the bills to fail (Marshall 2018). The Supreme Court played an active role in the issue of race. Initially, Plessy v. Ferguson affirmed states’ rights to educate African-American students in facilities that were “separate but equal” from those used by Caucasians and was the backbone of the segregation that continued into the
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
103
1950s. Later Supreme Court rulings in the area of higher education began to chip away at this doctrine of separate but equal. For example, Sweatt v. Painter required institutions to accept black students if these students could show that separate facilities were not actually equal. Sweatt v. Painter was one of the precedents cited in the Brown v. Board of Education ruling, which overturned Plessy v. Ferguson and marked the beginning of the end of segregation in elementary and secondary schools. Brown v. Board also represents one of the earliest expansions of the federal role in education policy, which “initiated a new era of federal activism in education and laid the foundation of a policy regime that was to last for approximately thirty years” (McGuinn 2015). The second Brown ruling further expanded the federal role by requiring states to make a “prompt and reasonable start toward full integration.” Reflecting state and local resistance to this perceived federal intrusion, 101 southern members of Congress signed the “Southern Manifesto” and engaged in “massive resistance” to integration (McGuinn 2015). Rufus Miles (1974) credits the Brown decisions for the beginning of the Civil Rights Movement, stating that “although it had a slow fuse, it touched off the Civil Rights Movement in the United States.” One consequence of the resulting fight over segregation was the migration of African-Americans to northern areas and schools, which made the lack of education for those children more visible (Spring 2016). Many African-Americans ended up on welfare rolls and living in slums. Increases in production after World War II only accentuated this inequality as income disparities continued to grow between white beneficiaries of the increased production and systematically disadvantaged African-Americans (Miles 1974). It was during this period that the coalition of disparate civil rights groups, which would eventually help to champion the Civil Rights Act (CRA) of 1964, began to coalesce (Radin 1977; Sanders 2016). The correlation between race, poverty, and low educational attainment became increasingly clear. In response to issues of poverty, many of the nation’s largest cities formed a “Great Cities Program for School Improvement” to better understand the needs of poor and African-American students (Bailey and Mosher 1968). This and other efforts provided a proving ground for attempts to better serve disadvantaged students, and the interlinking of poverty and race was the basis for race-related provisions in education bills in the early ’60s that sought to expand upon state and local efforts.
104
S. PUTANSU
In 1961 and 1962, Kennedy’s proposed bills included a set of experimental programs that emphasized the “educationally disadvantaged.” They included a provision for development of curriculums related to disadvantaged pupils, and a requirement for coordination between private and nonprofit agencies to better serve the educationally disadvantaged (Bailey and Mosher 1968). Although both bills failed due to opposition from religious schools, the ongoing civil rights movement continued to put pressure on the federal government to take action, spurred by events like the protests in Birmingham, Alabama, where televised attacks of protesters by police dogs spurred a public outcry and further strengthened the coalition of civil rights groups (Miles 1974). The assassination of President Kennedy in 1963 increased these feelings even further (Radin 1977). The Kennedy administration had set the stage for an omnibus education bill to synthesize the distinction between general and categorical aid (Kaestle and Lodewick 2007). This strategy of avoiding general federal aid but focusing on special needs was continued in an education bill sponsored by Senators Perkins and Morse in 1964. This bill continued the attempts to address the needs of civil rights groups, by incorporating the impact aid formula from New Deal programs into the funding of education, based on income levels. This formula was modified to include poverty, welfare, and employment in another bill proposed by Senator Morse in 1965 (McAndrews 2006). This central omnibus education bill approach would ultimately be the foundation for successful passage of ESEA in 1965 (Kaestle and Lodewick 2007). Although all of these bills failed, the impact aid formula would ultimately be the basis for federal funding of ESEA Title I, supported strongly in 1965 by the Commissioner of Education, Francis Keppel. One major factor impacting the passage of ESEA in 1965 was that, unlike all of these previous bills, the need for the Powell Amendment had been eliminated because Title VI of the CRA of 1964 (Public Law 88-352) codified its requirements and applied them to all federal funding (Miles 1974). The CRA also authorized the Commissioner of Education to arrange for support for institutions of higher education and school districts. By enabling the agency to provide in-service programs for assisting instructional staff in dealing with problems caused by desegregation, the CRA removed the need to include such a provision in ESEA. By essentially resolving the primary debate over race and segregation separately through the CRA,
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
105
future education bills no longer needed to incorporate race-specific provisions. The CRA was supported by a diverse group of interests, consisting of nearly eighty organizations representing a variety of racial, economic, and religious positions (Radin 1977). The Passage of ESEA ESEA ultimately passed in 1965 on the heels of the CRA of 1964, incorporating the impact aid formula to address issues of race and poverty and allowing funds for religious schools for non-religious purposes. The Leadership Conference on Civil Rights was a coalition that overcame many conflicting opinions among disparate civil rights groups in order to support the CRA. However, this coalition would be difficult to maintain in the process of passing ESEA (Radin 1977), resulting in a need to address issues of race indirectly. As a result, the use of the impact aid formula served as a necessary tool to address the needs of disadvantaged students. It indirectly served the interests of civil rights groups, because framing the funds as an issue of race would have led to increased political conflict that could have prevented the law from passing (Bailey and Mosher 1968). Further, while section 605 of ESEA established a general prohibition against payments for religious instruction, various aspects of the law included direct and indirect distribution of funds to religious schools and their students. In addition, the bill attempted to address ideological concerns about federal encroachment by declaring that “no federal official could exercise supervision over the curriculum, administration, or personnel of any institution or school system or over the selection of any instructional materials” (Thomas and Brady 2005). ESEA contained a myriad of provisions and goals related to improving education, which attempted to address the concerns of a myriad of state interests, teachers’ unions, civil rights groups, and religious interests. Since passage more than fifty years ago, the multiple goals in the original bill have persisted and expanded through numerous reauthorizations, amendments, and extensions. These goals reflect a wide assortment of critical societal needs, including the economy, system of democratic government, criminal justice system, nutrition, health, art, and culture. Through the lens of Weiss’ Four-I’s, ESEA represented an ideological victory for the belief that the national government should be involved in elementary and secondary education. Even among those who voted in
106
S. PUTANSU
favor of ESEA, there was a sense that the limitations of the law—including imperfect targeting, ambiguous implementation requirements, and a precarious balance of education interests—were not sufficient to dislodge the need for national intervention to help disadvantaged students (Eidenberg and Morey 1969). Policy knowledge relevant to the passage of ESEA mainly argued that intervention in schools was needed to help disadvantaged students, rather than specific guidance about how to intervene. This data, information, and evidence were primarily used by decision makers whose ideologies and interests already supported this result. The other three I’s all played a more significant role in the passage of ESEA. First, unified Democratic control of the government with supermajorities in both houses, and Johnson’s election on the Great Society platform represented both the strength of the Democratic Party. This aligned with ideologies that had historically supported federal involvement, as well as an institutional alignment of the branches of government that eased passage of the major reform. Furthermore, Senator Morse’s usage of the impact aid formula as the basis for distributing aid used an institutionally established practice to implement a new program, and compromises with state, educational, civil rights, and parochial interests cemented the deal. In sum, a mix of political coalitions and successful strategies, employed by a collection of key decision makers, ultimately overcame the divisiveness of socialism, religion, and race, and helped secure ESEA’s passage. ESEA has numerous, often conflicting, goals related to educational attainment, and Title I reflects many of those conflicts. It is difficult to know the qualities that make for a good teacher, or the measures that ensure a good education, as it relates to this broad variety of goals. Without the ability to define or measure processes or outcomes, Title I can be characterized as a coping program under Wilson’s typology. The processes needed to improve educational outcomes for disadvantaged students are uncertain and difficult to measure. Further, the breadth, variety, and complexity of multiple goals inherent in the program make results similarly difficult to measure. As established in the previous chapter, this means that policy knowledge is likely to play a secondary role, relative to political factors, in future decisions related to the program. In Chapter 7, all major legislative decisions about Title I from 1965 through the present day are examined in detail. The chapter describes whether and the extent to which the relative influence of the Four-I’s is stable or changing, and will group periods of relative stability into eras.
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
107
All changes are examined together to illuminate the production, dissemination, and use of policy knowledge in the policy process. The chapter shows—consistent with the expectations established in Chapter 2—that the relatively high levels of political conflict combined with the resulting coping-type program design, result in relatively low use of data, information, and evidence coupled with generally incremental decisions. Further, it will present exceptions, when decisions are either rational or comprehensive, and provide detailed contextual insights into why they occur.
A Federal Role for Higher Education By the turn of the twentieth century, the federal role in higher education was mostly limited to the various land grants, and still less than 1% of the US population held four-year degrees (Archibald 2002). That number would reach nearly 4% by 1950. In 1946, President Truman appointed a commission to consider the problems facing higher education in the United States, and the result was one of the earliest explicit calls for direct federal support of and involvement in education policy. The 1947 Truman Commission Report represents the first pivot from the early era of indirect federal support and institutional-level control of education toward the modern era of increased direct federal involvement in setting and pursuing national education goals. The Truman Commission Report was focused on issues of higher education and declared two basic roles for higher education: to “insure equal liberty and equal opportunity to differing individuals and groups”— regardless of status, race, creed, color, or sex—and “to enable the citizen to understand, appraise, and redirect courses, men, and events as these tend to strengthen or weaken their liberties.” Further, it called on the higher education community to become the means for every citizen to “carry his education, formal and informal, as far as his native capacities permit” (President’s Commission on Higher Education 1947). To this end, the report offered guidance to reduce college costs, establish a system of federally funded scholarships and fellowships, and eliminate racial and religious discrimination. The report also made recommendations about the content of curriculum but stopped short of suggesting direct federal involvement in any curriculum development. The move toward the modern era of federal involvement continued with several federal initiatives passed during World War II. The first of these initiatives was the creation of the GED program, which allowed
108
S. PUTANSU
World War II veterans who had not earned a high school diploma to pursue secondary education degrees. This was followed by the Servicemen’s Readjustment Act of 1944 (the GI Bill), which provided veterans with subsistence allowances, tuition fees, and supplies for higher education. These two bills were strongly linked, as the purpose of GED exams was to certify veterans returning from World War II so that they could take advantage of the benefits provided in the GI Bill (Tyler 2005). In 1952, the GI Bill would also be extended to veterans of the Korean War (Goldin 1999). Despite increasing federal involvement in education, the GI Bill easily won support, in large part because it was framed as an antidepression measure to protect the economy (Olson 1973). While the GI Bill was limited in scope to veterans, and although education was not its primary purpose, it did set a precedent for the federal government to promote higher education. The spending and incentives it offered ultimately allowed over 2 million World War II veterans to attend college (Olson 1973). These were the first of several federal initiatives that were, in part, intended to spur competition in math and science with the Soviet Union during the Cold War. Six years later, the US government, for the first time, provided financial aid in the form of loans to students in an attempt to improve educational opportunity. Specifically, the National Defense Student Loan Program (NDSL) was created as part of the National Defense Education Act (NDEA) of 1958. The NDSL allowed loans to be distributed to new and continuing college students, as well as to graduate students (Carlson 1959). As a result of NDSL, student loans became more accepted as a form of financial aid and provided a precedent for mixing loans and direct aid (Ihrke 1962). NDEA was passed, in part, to address a critical need to compete with the Soviet education system, especially after the launch of Sputnik in 1957 (Rudolph 2011). The NDEA originally also included a provision for aid to developing institutions, which was removed before the bill passed, but which would resurface as part of HEA in 1965. Though small in size, NDSL program established a programmatic framework for expansion of federal student aid in the following decade. Further, NDEA’s success established the potential positive influence that federal education legislation could have on the economy and educational attainment overall (Flattau 2007).
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
109
While these initiatives were geared primarily toward veterans, similar changes were being driven by demographic shifts that made higher education a more important topic to the American public. Specifically, the percentage of Americans pursuing and receiving college degrees was growing exponentially. According to analysis of census documents by Archibald, the number of young adults in college nearly doubled—likely due to the impact of the GI Bill—from 12.5 to 22.2% between 1946 and 1960 (Archibald 2002). Similarly, the percentage of Americans with a bachelor’s degree continued to rise from 4% in 1950 to nearly 16% in 1998. As mentioned above, this number had increased to 35.1% by 2018. This signals a continued and growing importance of higher education in the American economy, which has been, and continues to be, accompanied by increased federal interest and involvement. The Passage of HEA Growing interest in higher education came to a head in 1965 with the passage of the HEA, which represented the culmination of a shift from the early federal role in higher education, which was defined by institutional aid, to the current era of federal involvement, defined largely by student aid (Archibald 2002). Title IV of HEA is the basic authorizing legislation for a variety of such student aid programs, including student loan programs. Student loans are the largest source of federal funds in higher education (Shohov 2004). These loans have a clearly defined goal of increasing enrollment in higher education by helping students to pay for school. The conflict and congruence of Weiss Four-I’s—ideology, interests, institutions, and information—at the time of HEA’s passage highlight the political context for student loans. This large expansion of the federal role, and the resistance from states, universities, and private industry, demonstrates the continuing ideological tension surrounding the proper role of the federal government, and the power of a unified government to overcome this tension. Further, the shift from institutional aid to student aid represents a shift in power from education providers to students and families, and underscores continuing tension between these and other groups. This tension, however, was mitigated by a reliance on existing educational and finance institutions, and a general acceptance of policy knowledge guidance showing both that higher education provided significant returns to several goals and that the primary obstacle to enrollment
110
S. PUTANSU
was insufficient financial resources. The minor nature of political conflicts, combined with general agreement on policy knowledge that was relevant, trusted, and accurate, resulted in a student loan program with a clear primary goal and well-defined implementation processes. The primary goal of student loans is to improve access to higher education. Further, both the process of issuing loans and the output of college enrollment are relatively easily measured, so the student loan programs can be considered a production type under Wilson’s typology of bureaucratic organizations (Wilson 1989). As established in the previous chapter, this means that there is a relatively high potential for the program to be considered comprehensively in decision making processes. In Chapter 6, the student loan program is examined in detail, including all major decisions from 1965 through 2015. Legislative decisions that occur in times of relative stability will be grouped into eras, which will be examined together to examine the production, dissemination, and utilization of policy knowledge in decision making. The chapter shows—consistent with the expectations established in Chapter 2—that the relatively low levels of political conflict combined with a production type program design, result in relatively high use of policy knowledge. Further, it will present exceptions, including decisions that are politically driven or incremental in nature, and provide detailed contextual insights into why they occur.
Implications for the Case Studies This chapter has provided a historical background on federal education policy, which provides important contextual information for the cases that are examined in the following chapters. While the first section of this chapter has broadly introduced the ideologies, interests, institutions, and policy knowledge that impact federal education policy, the sections on the passage of HEA and ESEA expanded these to provide a comprehensive account of the political context, program design, and bureaucratic type of each program. These historical accounts show that the two programs, passed within months of each other, contended with very different political realities. The passage of HEA was impacted by minor ideological and interest factors, which were overcome without the need for serious compromise. The program was designed in a conceptually simple way, with the process of distributing loans and collecting payments, along with the clearly
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
111
defined primary goal of increased enrollment both being straightforward to measure, as in production type programs. In contrast, the passage of ESEA was highly contentious with multiple ideological conflicts, numerous competing interests, strong institutional resistance, and limited useful information. The program design reflects this conflict, with vague guidelines for implementation processes and multiple, often conflicting, goals. As a result, neither process nor goals are simple to measure, in line with expectations for a coping type program. The frameworks described in the first three chapters establish expectations for student loans to have relatively higher use of rational processes and comprehensive decisions, and for Title I to have mostly political and incremental decisions. The next two chapters trace the evolution of political factors and policy knowledge throughout the history of both student loans and Title I. They examine major legislative decisions that have occurred throughout these programs’ histories, and trace the evolution of the programs through consideration of the impacts of information and political factors. Examination of the ensuing 50-year histories of these programs provides a robust set of evidence to better understand when policy knowledge is brought to bear on policy decisions. Because these cases represent the expectations for the most and least likely to rely on policy knowledge under Wilson’s typology of bureaucratic organizations, they should illustrate the range of ways it is used, and its impact on policy outcomes. The grouping of decisions into eras of relative stability allows for consideration of incremental and comprehensive decisions for each case. The historical narratives trace policy knowledge throughout the policy process, from its introduction, to its dissemination, and ultimately to its use, or lack thereof. Examining the quality of policy knowledge in conjunction with this account of its use provide insights into the needs of decision makers and the conditions under which different types of data, information, and evidence have impacted decision making for these two cases.
References Archibald, R. B. (2002). Redesigning the financial aid system: Why colleges and universities should switch roles with the federal government. JHU Press. Bailyn, B. (1972). Education in the forming of American society: Needs and opportunities for study (Vol. 3). W. W. Norton & Company.
112
S. PUTANSU
Bailey, S. K., & Mosher, E. K. (1968). ESEA: The office of education administers a law. Syracuse University Press. Carlson, T. E. (1959). Guide to the National Defense Education Act of 1958 (No. 553). US Department of Health, Education, and Welfare, Office of Education. Clark, K. B. (1963). Educational stimulation of racially disadvantaged children. in Passow, A. H. (1963). Education in depressed areas, 142–162. New York: Bureau of Publications, Teachers College, Columbia University. Conant, J. B. (1959). The american high school today: A first report to interested citizens. NY: McGraw-Hill. Conant, J. B. (1961). Slums and suburbs: A commentary on schools in metropolitan areas. New York: McGraw-Hill. Cubberly, E. P. (1919). Public education in the U.S. Boston: Houghton Mifflin Co. Eidenberg, E., & Morey, R. D. (1969). An act of Congress: The legislative process and the making of education policy. Norton. Farrand, M. (1911). The records of the federal convention of 1787. Yale University Press. Fiel, J. E., & Zhang, Y. (2019). With all deliberate speed: The reversal of courtordered school desegregation, 1970–2013. American Journal of Sociology, 124(6), 1685–1719. Flattau, P. E. (2007). The National Defense Education Act of 1958: Selected outcomes. Institute for Defense Analyses, Science & Technology Policy Institute. Gamson, D. (2007). Historical perspectives on democratic decision making in education: Paradigms, paradoxes, and promises. Yearbook of the National Society for the Study of Education, 106(1), 15–45. Gardner, J. (1961). Learning throughout life. The American Journal of Economics and Sociology, 20(3), 239–240. Gardner, J. (1964). “Report of the president’s task force on education”, 1964. Austin, TX: LBJ Presidential Library. Goldin, C. (1999). A brief history of education in the united states (NBER Working Paper, Historical Series no. 119). Hansen, W. L. (1963). Total and private rates of return to investment in schooling. Journal of Political Economy, 71(2), 128–140. Heidenheimer, A. J. (2017). Education and social security entitlements in Europe and America. In Development of welfare states in Europe and America (pp. 269–304). Routledge. Hirschland, M. J., & Steinmo, S. (2003). Correcting the record: Understanding the history of federal intervention and failure in securing US educational reform. Educational Policy, 17 (3), 343–364. Hovey, H. A. (1965). United States military assistance: A study of policies and practices. Praeger.
5
THE ROOTS OF MODERN FEDERAL EDUCATION POLICY
113
Ihrke, D. F. (1962). Evaluation of the effectiveness of the National Defense Education Act of 1958 (Doctoral dissertation, Northern State Teachers College). Jenkins, C. J., & Hill, M. H. (2011). Role of federal government in public education: Historical perspectives. Washington, DC: League of Women Voters, 110. Kaestle, C. F. (1983). Pillars of the republic: Common schools and American society, 1780–1860 (Vol. 154). Macmillan. Kaestle, C. F., & Lodewick, A. E. (2007). To educate a nation: Federal and national strategies of school reform. University Press of Kansas. Kaestle, C. F., & Vinovskis, M. A. (1978). From apron strings to ABCs: Parents, children, and schooling in nineteenth-century Massachusetts. American Journal of Sociology, 84, S39–S80. Marshall, J. P. (2018). The Mississippi civil rights movement and the kennedy administration, 1960–1964: A history in documents. LSU Press. McAndrews, L. J. (2006). The era of education: The presidents and the schools, 1965–2001. University of Illinois Press. McGuinn, P. (2015). Schooling the state: ESEA and the evolution of the US department of education. RSF: The Russell Sage Foundation Journal of the Social Sciences, 1(3), 77–94. Miles, R. E. (1974). The Department of Health, Education, and Welfare (No. 39). Praeger. Olson, K. W. (1973). The GI bill and higher education: Success and surprise. American Quarterly, 25(5), 596–610. Preer, J. (2012). America’s historically black colleges and universities: A narrative history, from the nineteenth century into the twenty-first century. The Journal of American History, 99(1), 311–312. President’s Commission on Higher Education. (1947). Higher education for American democracy: A report of the president’s commission for higher education. New York: Harper & Brothers. Pulliam, J. D. (1982). History of education in America. Columbus, OH: Charles E. Radin, B. A. (1977). Implementation, change, and the federal bureaucracy: School desegregation policy in HEW, 1964–1968. New York: Teachers College Press. Radin, B. A., & Hawley, W. D. (1989). The politics of reorganization. New York: Pergamon Press. Reed, D. S. (2014). Building the federal schoolhouse: Localism and the American education state. Oxford University Press. Rudolph, F. (1990). The American college and university: A history. University of Georgia Press. Rudolph, J. L. (2011). More than science and sputnik: The National Defense Education Act of 1958. The Journal of Southern History, 77 (4), 1044.
114
S. PUTANSU
Sanders, C. R. (2016). “Money talks”: The Elementary and Secondary Education Act of 1965 and the African-American freedom struggle in Mississippi. History of Education Quarterly, 56(2), 361–367. Schultz, T. W. (1959). Investment in man: An economist’s view. Social Service Review, 33(2), 109–117. Shohov, T. (Ed.). (2004). Federal student loans. Nova Publishers. Snyder, T. D. (Ed.). (1993). 120 years of American education: A statistical portrait. DIANE Publishing. Spring, J. (2016). Deculturalization and the struggle for equality: A brief history of the education of dominated cultures in the United States. Routledge. Thomas, G. (2014). The founders and the idea of a national university: Constituting the American mind. Cambridge University Press. Thomas, J. Y., & Brady, K. P. (2005). Chapter 3: The Elementary and Secondary Education Act at 40: Equity, accountability, and the evolving federal role in public education. Review of Research in Education, 29(1), 51–67. Tyler, J. H. (2005). The General Educational Development (GED) credential: History, current research, and directions for policy and practice. Review of Adult Learning and Literacy, 5, 45–84. Wilson, J. Q. (1989) Bureaucracy: What government agencies do and why they do it (2nd ed.). Basic Books.
CHAPTER 6
Improving Access to Higher Education
When President Johnson stood in front of the crowd at Southwest Texas State College on November 8, 1965, prepared to sign the Higher Education Act (HEA), he delivered a speech that summarized the act’s primary goal: “It means that a high school senior anywhere in this great land of ours can apply to any college or any university in any of the fifty states and not be turned away because his family is poor” (Johnson 1965). All four titles of the original act were geared toward providing various forms of financial assistance with a goal of expanding access to higher education (Cervantes et al. 2005). The guaranteed student loan (GSL) program— one of many programs created—is supposed to contribute to this goal by providing financial support to students, enabling a larger number to enroll. This chapter provides an overview of the change to the federal role in higher education that accompanied passage of the law. Using this as a starting point, the chapter goes on to explore five eras that represent the major reauthorizations, appropriations, and budget decisions that impacted student loan programs since they were established in 1965. In order to understand the role of policy knowledge in the decision making process and test the expectations derived from Wilson’s (1989) typology of bureaucratic organizations, this chapter applies Weiss’ Four-I framework to major legislative decisions initiated during the fifty-year period following the bill’s passage. First, the landscape of political factors and policy knowledge will be introduced for each era. Then, individual decisions in each of the eras will then be analyzed to understand the specific © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_6
115
116
S. PUTANSU
roles played by these factors and the relative dominance of each. Finally, the summary of each era will conclude with a consideration of the factors that contribute to the selection and use of data, information, and evidence to support the decision making process.
HEA Titles and Background HEA established the federal government’s role in higher education policies by establishing a goal of eliminating price barriers to access. This legislation was passed with minimal fanfare and was greatly overshadowed by the Elementary and Secondary Education Act (ESEA), which passed the same year. These two acts marked the beginning of an expansion of the role of the national government relative to state and local governments, which has continued for over fifty years. This section provides some information on the major aspects of each of the HEA titles, which provide context needed to understand the trade-offs that occurred when the act was originally passed and as it was revised over the years. Title I of HEA provides for grants to states for “community service programs of colleges and universities.” This title also created the National Advisory Council on Extension and Continuing Education for oversight of these community service programs. Title I was considered to be an extension of general aid as offered previously in the Morrill Land Grants, but some lawmakers expressed concern that it would be effective in rural areas but suffer in urban ones, as had been the case with the land grants (Sorber 2018). Specifically, these grants had been most effective in areas where there was sufficient land to provide for the construction of educational institutions, whereas providing land in concentrated urban areas was thought to be more difficult (NASULGC 1995; Stein 2017). Around this time, Congress was undergoing its own demographic shift as a result of the Supreme Court ruling in Reynolds v. Sims, which established the requirement of “one man, one vote” in 1964, requiring seats in the House of Representatives to represent similar populations and reducing the power of rural areas within state politics. However, rural interests remained a majority for the passage of HEA, and the concern of urban representatives was not strong enough to garner a change to this provision in Title I.
6
IMPROVING ACCESS TO HIGHER EDUCATION
117
Additional general aid was offered in Title II, which provided grants to institutions for libraries and library staff training. Again, Title II maintained state discretion in assigning funds. Title III provided aid to developing institutions, which were primarily African-American institutions in the south. Passage of Title III demonstrates a microcosm of the politics of race that occurred between 1958 and 1964. The program had originally been offered as a part of NDEA, but did not have sufficient support to pass because of the fear that it could promote integration. Specifically, a program that could make African-American institutions “more equal” would have to start from the proposition that the schools were not equal, and admitting inequality might strengthen arguments to end segregation by race. In 1965, though the text of the title had not changed significantly, the original supporters had become the opposition, because they feared it could prevent integration by assisting African-American institutions, and potentially improving them to a point where judicially imposed integration would become unnecessary. Title IV, the main focus of this case, represents the first generally available aid program for postsecondary students, made up of grants and scholarships on the one hand and loans on the other. Grants were geared toward the neediest students, while loans initially targeted the middle class. In the original HEA debates, Republicans opposed the use of grants in favor of tax credits. These credits were opposed by Democrats, who claimed they were regressive in nature and who cited an estimate from the Treasury Department that said their costs would be too high. The tax credits were considered regressive because they would primarily benefit individuals who had sufficient tax obligations to benefit from the credit, which would not be true for the neediest families, who pay low or no tax due to their low income. Guaranteed Student Loans (GSLs) were proposed by President Johnson and targeted toward the middle class as a way to justify Educational Opportunity Grants for needy students. Because students generally don’t have significant incomes or credit histories, banks would not give loans without some protection from default, even if they were not from the neediest families. President Johnson offered this benefit for middleincome families so that he could build additional support for his grant programs. The American Banking Association argued that GSLs were not needed, that existing guarantees were sufficient, and that the government could not improve other shortcomings. They also warned of the possibility of government takeover and said that the interest rate was too low for
118
S. PUTANSU
the program to be profitable. Similarly, private guarantee agencies argued that people would stop using their guarantees and move to the federal program, effectively killing their businesses. These concerns were mitigated by a promise from President Johnson that the federal government would only guarantee loans if no other guarantor was available. Further, some members of Congress argued that direct loans, as established in National Defense Education Act (NDEA), could be more appropriate than GSLs. Johnson, however, was convinced that direct loans would not receive sufficient support for passage, and that his preferred grant programs could be prevented as a result. By convincing banking and guarantee interests to support the GSL approach, he ensured that HEA could pass with grants intact. Johnson also succeeded in getting a parental income limit for interest subsidies, ensuring that the wealthiest families would not qualify for GSL. Families over this limit would have to prove need in order to receive interest subsidized loans. Two other important programs under Title IV were the federal work-study program, which was moved from the Office of Economic Opportunity, and the National Defense Student Loan (NDSL) program, which was brought in from NDEA. From 1965 until today, federal student aid has continued to utilize both grants and loans. Title V supported the teacher corps and teacher education. Democratic Senators Ted Kennedy of Massachusetts and Gaylord Nelson of Wisconsin sponsored Title V after they had failed to get these programs into ESEA. This is because K-12 educators were a major player in the interest group politics of ESEA, while they were relatively uninvolved in the passage of HEA. Title VI provided funds for technical equipment and faculty development for undergraduate instruction, and Title VII contained graduate and postsecondary improvement programs, including grants for graduate and legal students, as well as grants to institutions to provide financial assistance to graduate students. The following section summarizes all of the ideological, interest-based, institutional, and informational factors that shaped HEA and led to this specific set of Titles. The passage of HEA represented a straightforward and indirect involvement by the federal government, in the form of allocation of federal funds with limited requirements or intrusion on policy. Title IV, in particular, was structured simply as an effort to provide funds to match the needs of students, thereby increasing their ability to enroll in institutions of higher education. In order to run the student loan program, the federal government was only required to take two basic actions. First,
6
IMPROVING ACCESS TO HIGHER EDUCATION
119
they have to set eligibility criteria to determine the amount of funding needed for each student. Second, they need to distribute those funds, in this case, through the banks participating in the GSL program. The GSL program’s primary outcome is the level of enrollment in higher education. In both process and outcome, these metrics are relatively straightforward to measure. Because both the process of distributing loans and the output of increased enrollment are easily measured, student loan programs can be thought of as a “production organization” in Wilson’s typology of bureaucratic organizations. As noted in Chapter 2, of Wilson’s four types (Production, Process, Craft, and Coping), the production type of organization—due to the ease of measuring both process and results—is expected to be associated with increased production of data, information, and evidence relative to other program types. Further, decisions about the program are expected to be more strongly influenced by policy knowledge. In 1964, the year prior to HEA, the federal government spent approximately $31.8 million ($267.0 million in 2019 dollars) on grants and loans through existing education programs. To fulfill Johnson’s promise, this annual spending figure multiplied over thirteen times within five years, reaching $517.5 million ($3.4 billion in 2019 dollars) by 1970, and $10.5 billion ($20.6 billion in 2019 dollars) by 1990. In 1990, congressional budget rules changed, which also changed how the costs of the GSL program were tracked. This change makes it difficult to compare costs before and after 1990, because the increase to $148.3 billion by 2010 ($174.5 billion in 2019 dollars) is partly driven by the change in the way the amount was calculated.
Politics and Policy Knowledge at HEA’s Passage As described in Chapter 5, HEA’s passage was driven by a combination of institutional partisan alignment and interest group involvement, which largely circumvented serious ideological tensions. Policy knowledge played a supporting role in the decision, by linking the lack of financial means as an explanation for decisions not to pursue and enroll in higher education. The passage of HEA represents a comprehensive policy change marked by a significant expansion of the federal role that lies well outside of political-incremental theories of decision making. Punctuated equilibrium models also fall short in explaining the outcome. Aspects of the
120
S. PUTANSU
decision, like Johnson’s major electoral victory, may represent an exogenous shock to the environment. Further, it could be argued that policy entrepreneurs took ideas from the “garbage can” to take advantage of this shift. However, both of these frameworks provide a partial explanation of the compromises that went into getting the bill passed. Instead, the decision is best understood by a comprehensive reflection on the ideological tensions, major interests, institutional factors, and policy knowledge that constituted the policy environment in 1965. The following section will summarize this environment, before turning to successive decisions about the program. The two ideological tensions that persist across the history of the student loan program relate to the preferred role and size of the federal government. The first ideological disagreement is related to the proper role of the government relative to private sector responsibilities. In general, one side of this disagreement believes that optimum enrollment in higher education will be reached in a competitive equilibrium, which is best achieved through private sector markets. The other side of this debate claims that the federal government has a role in increasing access to higher education because various public externalities result from a well-educated populace, justifying increased government intervention. Similarly, the ideological tension over the size of government balances those who believe that the powers of the central government should be minimized and those who argue that centralization of power in the federal government can lead to enhancements in economy, effectiveness, efficiency, and equity that benefit the public. There are six major types of interests vying for influence in student loan programs. Perhaps the most obvious interest group is students, the direct beneficiaries of the program, and their families, who often share a role in financing higher education. Over the fifty-year history of student loans, this group has been treated sometimes as a single interest, but other times there has been a tension between lower-income students and middle- to high-income students. Institutions of higher education, including public and private universities, community colleges, and trade schools, are also an interest that influences decisions about student loans. Third, non-governmental organizations have a range of advocacy interests, including promoting racial and gender equality, supporting particular fields of study, and creating employment opportunities. The other three interests are all parts of the system that implements student loan
6
IMPROVING ACCESS TO HIGHER EDUCATION
121
programs, including banks that finance the loans, agencies, such as Sallie Mae, that guarantee them, and government agencies that administer them. Finally, states were also a major interest in student loans during the early years of the student loan program. The primary institutional factor that has impacted decisions about the student loan program is whether the decision is being made in the context of a reauthorization or as part of an appropriations process. While there is some overlap in these two institutional processes, they are associated with a different set of interests and goals. Specifically, in a reauthorization context, the six major interests described above have a greater influence on the decision because in an appropriations context, these interests have to compete with interests in other policy areas that are also vying for a share of the budget. For this reason, both reauthorizations and major appropriations decisions are explored in this case. The other institutional factor is tension between the executive and legislative branches. This is often hard to separate from other political factors, because when Congress and the president are of the same party, their general agreement on ideology and interests may result in reduced friction between the branches. In contrast, when either or both the House and the Senate contain a majority party different from the president, it may magnify institutional tensions between the branches of government. It is important to note, however, that this relationship between party affiliation and tension between the branches is not perfect, as the separationof-powers system creates natural tensions between Congress passing laws and appropriating funds on the one hand and the president executing those laws on the other. Considerations of data, information, and evidence for student loans broadly fit into Wilson’s characterization of process and outcome. On the process side, policy knowledge is generated on alternative implementation options and issues of fairness in process. For outcomes, it focuses on the effectiveness of the programs, their efficiency (or cost-effectiveness), and equitable outcomes. At HEA’s passage, relevant policy knowledge was mostly limited to student and family incomes and their relationship to college enrollment rates and had a strong emphasis on equity issues. As shown in Chapter 5, and explored in detail above, all four elements from Weiss’ framework had some level of influence over the passage of HEA. However, some of these factors had a larger impact than others. Specifically, the strong institutional alignment reflected in the Democratic Party control of the presidency and Congress. Similarly, the influence of
122
S. PUTANSU
interest groups was limited by this shared institutional power and ideological alignment, leading interest groups to negotiate the margins of the bill, rather than defining how best to frame the federal approach. Similarly, the use of policy knowledge largely reflected a partisan agreement on framing and an attempt to demonstrate rationality, rather than a technical approach to identifying a problem and choosing a specific solution. The passage of HEA was thus politically dominated and comprehensive, reflecting elements of the punctuated equilibrium, policy stream, and garbage can models of decision making. However, the chapter now turns to five major eras of HEA and shows that neither these models, nor the other dominant decision making models, consistently explain successive decisions about student loans. Specifically, this chapter examines all 20 major decisions over the 50 ensuing years and shows that 17 of these 21 decisions are not well explained by existing models. Instead, they reflect a mixture of political factors and policy knowledge, or they fall in between the extremes of the incremental-comprehensive continuum.
Major Eras of Student Loan Programs While the passage of HEA was primarily driven by the institutional factors of unified government under the Democratic Party, other aspects of ideology, interests, and policy knowledge remained important. The law represented a victory for ideologies of public intervention and strong centralized government; it provided significant benefits to interests, including students, schools, states, and banks, while creating more responsibility and power for the office of federal aid within the Department of Health, Education, and Welfare (HEW) (Cervantes et al. 2005). Over the following fifty years, major decisions about student loans continued to focus on enrollment information and cost as important outcomes, and decision makers continued to work within the boundaries established by various interests. The major decisions from HEA’s enactment through the end of 2015 can be grouped into five major eras: (1) Initial Expansion, (2) Budget Restriction, (3) Muddling Toward Reform, (4) Federal Expansion, and (5) Muddling over the Federal Role. Era 1: Program Expansion (1965–1980) President Johnson was committed to expanding access to higher education, particularly for low-income students. To him, the centerpiece of the
6
IMPROVING ACCESS TO HIGHER EDUCATION
123
HEA was the grant programs established under Title IV to support the neediest students, while the student loan programs were a necessary element to satisfy interests and get the bill passed (Nevins 1962). Built on compromise among competing interests, the student loan program developed over the ensuing 50 years with little change to its goals or conflict over its existence. The first era of HEA begins with the 1965 passage and continues through 1980, covering three reauthorizations (1968, 1972, and 1976) and one modification offered outside of the normal reauthorization process. This era of HEA reflected a continued institutional dominance of the Democratic Party, with large majorities in Congress that effectively stymied resistance to the program, even under Republican presidents, ensuring that ideological challenges to federal intervention would continue to weaken. The institutional environment also began to shift after passage of HEA. Specifically, the creation of the GSL program created guarantee agencies and other government machinery to work and advocate on behalf of the program, and to ensure its continued reach and effectiveness. This period, however, saw a major shift in the role of interests in decisions about the GSL program. The most dramatic of these shifts is evidenced by the growing dominance of banking institutions in decisions about the program. Banks involved in the NDEA prior to HEA’s passage had an incentive to lobby for the GSL program at passage, but this incentive increased significantly once the profits from the GSL program began to take hold. Banks, attempting to increase and protect their profits, mounted increasing advocacy efforts to increase eligibility for the program and increase interest rates paid by student borrowers. Similarly, institutions of higher education similarly worked to increase access to these loans, leading to increases in enrollment and revenues. The expansion of power in these organized interests began to crowd out the voices of other interests, including civil rights groups and state interests. Despite an increase in the production and volume of information available, the landscape of policy knowledge remained relatively stable during this period. Studies and other published policy knowledge primarily maintained a focus on encouraging college enrollment, and the development of related information on program costs. Variations in the policy knowledge focused on aspects of social equity, including the relative
124
S. PUTANSU
Table 6.1 Politics and policy knowledge from HEA’s passage and era of program expansion, 1965–1980 Ideology
Interests Institutions
Policy knowledge
Democratic Party control of Congress mitigated ideological conflicts and stymied Republican presidential requests for budget cuts Banks and program beneficiaries grew in influence, largely displacing civil rights groups that helped ensure passage Loan guarantee agencies created at passage grew in importance, highlighted processes of the loan programs Sallie Mae created, facilitating secondary loan market Processes : Alternating focus on eligibility criteria, program cost, and availability of funds Outcomes: Consistent focus on reducing financial barriers to access Key values: Economy, effectiveness, equity
access of different disadvantaged groups, procedural fairness in application and repayment processes, and overall distributional fairness of GSL funds. The Four-I’s of HEA’s first era are summarized in Table 6.1 The first reauthorization of HEA occurred in 1968, before many elements of the law were fully implemented. The reauthorization increased GSL spending, raised interest rates, and established minimum monthly repayment amounts for students, authorized advances for guarantee agencies, and gave guarantee agencies the authority to grant deferments. Changes to Title IV were driven largely by interest group advocacy and were mostly incremental in nature. Banks were a primary player in this decision and benefitted directly from expanded GSL spending, higher interest rates, and minimum monthly payments. President Johnson also attempted to set more stringent income limits for GSL, in order to push the balance in student federal aid more toward grants, but was unsuccessful due to opposition from the banks. Guarantee agencies also played a role in the decision, by providing information showing that enrolled students were missing loan payments. They presented this information to Congress with an asserted causal explanation: Enrolled students were unable to also work and therefore could not afford to pay the loans while in school. They explained that when students missed these payments, the guarantee agencies were forced to make payouts to the banks, which increased cost to the government. These agencies were able to successfully leverage this policy knowledge to secure the authority to grant deferments.
6
IMPROVING ACCESS TO HIGHER EDUCATION
125
Although institutional conflict was not a major factor in the decision, it did nearly derail the reauthorization. Specifically, Senator Morse (DOR) raised opposition to the possibility that NDEA direct loans would be phased out in favor of the guaranteed loan approach, threatening to filibuster the reauthorization. Morse argued that direct loans, like those in NDEA, should not be phased out too quickly, out of concern that private credit streams would dry up. The Johnson administration promised that the expansion would not eliminate federally available credit, which would still be available if private funds were not. This was sufficient to convince Morse, and he retreated from his strong opposition to the GSL expansions. While the relative dominance of interests and the small modifications to eligibility and interest rates reflect a political-incremental decision, aspects of the reauthorization do not fit the model. Specifically, the introduction of policy knowledge by guarantee agencies was used to support a broader creation of previously unavailable deferment authority. As a result, this decision reflects a combination of political factors and policy knowledge not anticipated by existing decision making models. The 1972 reauthorization of HEA occurred under a Democratic Congress and Republican President Richard Nixon. This reauthorization created two new grant programs, which gave states grants to pay up to 50% of state-run student grant programs, and changed the formulas used to calculate the distribution of grant funds. Although the 1972 reauthorization was primarily focused on grants for the neediest students, it also included incremental changes to the student loan program, including additional increases to the annual and aggregate loan limits for GSL, and the creation of Sallie Mae for a secondary loan market. The institutional power and ideological dominance of the Democratic Congress are made obvious by the strong focus on grant programs for low-income students and reduced emphasis on grant programs targeted to middle-income students and families. This represented a successful attempt to make grants the core of the federal effort, in order to reduce the need for poorer students to use GSL. A primary justification for this expansion was a 1969 report from HEW’s Assistant Secretary of Planning and Evaluation, Alice Rivlin, which presented evidence that income was still the primary determinant of whether a student would enroll in college. However, the ideological preference for grants was consistent with the Democratic Party position well before the HEW report, and their political dominance was a clear driver of this decision.
126
S. PUTANSU
Although the 1972 reauthorization was primarily focused on grants for the neediest students, it included some changes to the student loan program. Specifically, it included additional increases to the annual and aggregate loan limits for GSL, and the creation of Sallie Mae for a secondary loan market, two changes advocated for intensely by banking interests. Following this reauthorization, loan costs would modestly increase from $1.48 billion in 1972 to $1.61 billion in 1976, while grant costs would more than double, from $2.93 billion to $6.62 billion. While the 1972 reauthorization did little to change the GSL program directly, the creation of Sallie Mae represented a substantial new program, driven primarily by political interests. The secondary loan market created new avenues for banks to generate profit from the sale and trade of student loan debt and represented a broader market incentive for the support of student loans. In addition to the increasing interest of banks, this had an impact on the ideological tensions between the Republican and Democratic parties. Specifically, while initial opposition to federal intervention fueled resistance to federal financial aid, the creation of a market for loans led to ideological support for the GSL program for proponents of private markets. Changes to student aid were very minor in the 1976 reauthorization, reflecting a typical case predicted by political-incremental models. Once again, GSL increased borrowing limits and increased the family income ceiling, reflecting the continued alignment between the power of Democrats in Congress, who desired expanded access to higher education, and banking and higher education interests, who benefited from the programs. In the next two years, loan costs increased 35% from $1.61 billion to $2.18 billion while grant costs decreased 8% to $6.09 billion. While authorizations in 1968, 1972, and 1976 tended to favor the neediest students, middle-income families were starting to become a more vocal interest. This led to a major policy shift in 1978. The 1976 reauthorization had raised the family income limit for subsidized loans to $25,000. However, the median income in the United States had risen to nearly $20,000 by 1978, and middle-income families began complaining that they were being squeezed out of the GSL program. In response, lawmakers passed the Middle Income Student Assistance Act (MISAA) of 1978, which removed all family income limits from GSL. This represented a victory for the banking interests, which desired a larger pool of customers with higher incomes. Over the intervening four years, the number of GSL loans increased from 1 million to 3.1 million, which came
6
IMPROVING ACCESS TO HIGHER EDUCATION
127
with vastly increased costs to the federal government. Between 1978 and 1980, loan costs nearly doubled to $4.13 billion. While MISAA was a result of direct advocacy from middle-income families, the actual policy adopted was impacted directly by the use of policy knowledge. Specifically, in response to the complaints about being squeezed out of the GSL programs, Republican lawmakers in the majority Democratic Congress introduced an education tax credit. Democrats, who supported broader benefits for middle-income families, raised objections based on equity concerns. Specifically, they referenced a report produced by the US Treasury that highlighted a high regressivity in tax credits, which tended to benefit wealthier families more, while not directly meeting the needs of middle-income families. Although opponents did not cite any evidence of the effectiveness of the GSL program in closing these gaps, or differences across income levels, they successfully resisted tax credits in favor of the GSL expansion. This comprehensive change to the program thus represents a mix of political and policy knowledge drivers. The five decisions that were made during HEAs first era represented a steady expansion of the benefits of GSL driven primarily by institutional stability, interest group advocacy, and moderate use of policy knowledge focused on enrollment. Although data, information, and evidence played a role in nearly every decision during this era, it never dominated the decision making process. As shown in Fig. 6.1, three of the five decisions made during this era do not fit into the dominant decision making models. As such, a balanced examination of political factors and policy knowledge provides a more complete picture of decisions made during this period. Era 2: Budget Restriction (1980–1992) After MISAA removed family income limits for student loans, the cost of federal student aid skyrocketed, and information about those costs became a primary justification for decisions about the program for more than a decade. Whereas spending on loans had grown from about $1.5 billion annually to about $4.5 billion in the 1970s, it jumped to almost $12 billion by 1990. During this era, major decisions were limited to two HEA reauthorizations and two appropriations-related laws, which all primarily attempted to slow or halt this growth in expenditures. The
S. PUTANSU
Punctuated Equilibrium Models
1978 MISAA
Rational Comprehensive
Mixed
1965 Passage
1972 Reauthorization
Incremental
Comprehensive
128
PoliticalIncremental
1968 Reauthorization
Bounded Rationality 1976 Reauthorization
Political
Mixed
Rational
Fig. 6.1 HEA decisions during the era of program expansion, 1965–1980
information that guided most of this discussion was the net cost of student loan programs and attempts to cut costs fell into two categories: (1) restricting eligibility—including reinstating family income limits, expanding needs testing, and limiting borrowing eligibility to the determined need and (2) decreasing benefits to certain interests—including creating alternative unsubsidized loan programs, reducing interest subsidies, and increasing interest rates. The second era of HEA saw several changes in the political environment, including increased ideological tension, institutional conflicts, and weakened organized interests. The primary focus of policy knowledge shifted from equitable access and enrollment to improved economy achieved through cost reductions. As described above, the first era was largely characterized by ideological consensus that framed issues and
6
IMPROVING ACCESS TO HIGHER EDUCATION
129
enabled the use of policy knowledge to address specific priorities. In contrast, the second era saw increased political tensions and policy knowledge used as the primary framing for decisions. Specifically, shared concerns about increased student loan costs were cited by a variety of ideological, interest, and institutional perspectives and used as a basis to form coalitions around specific program changes. While information on access, program success, and equitable outcomes continued to be produced, it received little attention in decision making. Table 6.2 summarizes the balance among the Four-I’s throughout the budget restriction era. The 1980 reauthorization was the first to explicitly recognize the massive growth in loan spending and push for changes in HEA to address it. However, despite the growing importance of reducing costs, the Democratic Party control of the presidency and Congress led to relatively incremental changes, including minor restrictions in loan eligibility and maximum borrowing limits. It also introduced two program changes advocated for by banking interests. Specifically, it raised the GSL interest rate from 7 to 9%, which banks argued would reduce interest in the loans and slow program growth. However, no policy knowledge was cited to estimate how responsive student borrowers would be to this change, nor how much bank profits would increase as a result of the change. The 1980 reauthorization also created the parent loans for undergraduate students program in an effort to create a repayment system with lower default risk, since it was assumed that parents would already be working. Yet again, there was no data, information, or evidence cited explicitly in the record to support this argument. During the debates for the 1980 reauthorization, decision makers had acknowledged that a driving force in the burgeoning cost of student loans had been the removal of family income Table 6.2 Politics and policy knowledge in HEA’s era of budget restriction, 1980–1992 Ideology
Interests Institutions Policy knowledge
Republican Presidents and increasing power in Congress highlight ideological preference for reduced federal role, increased attention on reducing spending Established interests circumvented through changes to institutional process Increased use of appropriations process to enact changes Processes : Increased focus on total costs Outcomes: Reduced attention on patterns of access Key values: Economy
130
S. PUTANSU
limits, but higher education, middle-income, and banking interests coalesced to prevent these limits from being re-instituted. Despite continued ideological, institutional, and interest support for the program, concerns about rising costs drove these incremental restrictions of the program. After the reauthorization, the newly elected President Reagan and a Republican Senate passed the Omnibus Budget Reconciliation Act of 1981, which re-instituted a family income ceiling at $30,000 for interest subsidies and expanded other loan programs, to reach new populations of independent students, as well as graduate and professional students. While interests had resisted more significant changes during the 1980 reauthorization, the use of the broader institutional process of appropriations weakened their influence. They were drowned out by the much broader collection of ideologies and interests that participate in budget discussions, which leveraged cost concerns to lobby for decisions to reduce federal involvement in the program. As a result, the inclusion of family income limits represented a much more substantial change to the program. However, the budget restrictions in both the 1980 reauthorization and the 1981 Omnibus only moderately slowed the growth of loan costs. These costs increased twofold from $4.13 billion in 1980 to $8.81 billion in 1986. Because these costs are formula-based, many other program budgets were cut to make up the difference. This includes grant programs, whose budgets fell from $6.75 billion to $5.36 billion over the same period. When reauthorization was due in 1986, the continued dominance of small government ideology combined with information on rising costs continued to drive decisions about student loans. The 1986 reauthorization took several steps to reduce spending and improve program economy, including expanding needs testing to all students and capping the amount a student could borrow to their demonstrated need. As with the 1981 omnibus, this decision reflected a combination of politics and policy knowledge resulting in a moderate change to program eligibility and scope. Despite the new restrictions, costs continued to increase, reaching nearly $12 billion by 1990. In 1990, Congress moved to change how budgeting and accounting would be done for federal loan programs, including GSL. When HEA passed in 1965, congressional budgeting rules made the guaranteed approach attractive, because the full cost of giving a direct loan was required to be placed on the budget in the year the loan was made,
6
IMPROVING ACCESS TO HIGHER EDUCATION
131
whereas GSL only had to record costs of defaults and interest subsidies. Thus, although the government would eventually recoup most of the costs (actually turning a profit in most cases), the budget would show a net loss for the full value of direct loans issued each year, while GSL only recorded costs of interests payments and defaults only as they happened. While budget agencies and outside researchers had long expressed concerns about this approach resulting in unfunded liabilities, political and institutional expedience won the day. Continued research into the costs of guaranteed loan programs revealed the impact of this budgeting approach, including reports in the late 1980s from the General Accounting Office (GAO) and the Congressional Budget Office (CBO): which showed that the full costs of the loans weren’t being accounted for (GAO 1977; Rivlin 1961). Specifically, if students were missing payments, this cost was not being accounted for until the student reached a full default, at which point the loss was much larger. These findings led to passage of the Federal Credit Reform Act (FCRA) in 1990, which required all loans (including GSL) to account for all longterm expenses they would accrue, and not just current interest payments and defaults (Redburn 1988). Policy knowledge from CBO and GAO was a primary driver in enacting FCRA, which was made easier by President George H. W. Bush’s decision to make the act a part of the 1990 appropriations process. These changes started to impact the official accounting of student loans, increasing their recorded cost by more than fifty percent from 1990 to 1993. A continued expansion of student loans, combined with the implementation of this new accounting approach, resulted in a doubling of the recorded cost of student loans, from about $12 billion in 1990 to about $24 billion in 1995. The four decisions that were made during HEA’s second era reflected repeated efforts to contain the overall costs of the program and to ensure that all costs were being captured. Figure 6.2 summarizes these decisions by decision type, as defined by the rational-political and incrementalcomprehensive continua. This shows that only one of the four decisions is readily explained by the dominant decision making models, while the other three all fall in the middle of these continua, reflecting the importance of a balanced consideration of politics and policy knowledge in understanding how these decisions were made.
S. PUTANSU
Punctuated Equilibrium Models
Incremental
Rational Comprehensive 1990 FCRA
1981 omnibus
Mixed
Comprehensive
132
1986 Reauthorization
PoliticalIncremental
Political
1980 Reauthorization
Bounded Rationality
Mixed
Rational
Fig. 6.2 HEA decisions during the era of budget restriction, 1980–1992
Era 3: Muddling Toward Reform (1992–2008) Although policy knowledge was a primary driver of decision made in HEA’s second era, continually increasing costs led some decision makers to accept that current strategies could not successfully control the growth of the program. As they attempted to institute broader reforms, the influence of organized interests rose up to resist change. As a result, the third era of HEA is characterized by increasing conflict within and among the Four-I’s that largely prevented substantive changes to the student loan program. During this third era, divided government led to increased ideological tensions and resulted in failure to reauthorize the program as scheduled.
6
IMPROVING ACCESS TO HIGHER EDUCATION
133
Table 6.3 Politics and policy knowledge in HEA’s era of muddling toward reform, 1992–2008 Ideology
Interests
Institutions Policy knowledge
Divided government throughout most of the era, with high conflict, including several veto threats and Congressionally led action Banks, students, and higher education interests continue to be important, but their interests become less aligned and more individual Reauthorization schedule not followed, alternative decision making arenas become dominant Processes : Continued focus on costs and cost-effectiveness of direct loans Outcomes: Access, graduation, and default rates all increase in importance Key values: Economy, effectiveness
Instead, decision makers had to use institutional processes, like extensions and appropriations decisions, to keep the program active. Organized interests became increasingly vocal and effective during this period, successfully resisting major program changes. While policy knowledge focused on costs remained central to decision making discussions, these interests also pressed Congress to consider broader information on the effectiveness of GSL programs. Table 6.3 summarizes this alignment of the Four-I’s during HEA’s third era. Shortly after the FCRA was passed, an analysis conducted by President Bush’s Department of Education suggested that a direct loan program, similar to the NDEA, would be more cost-effective than the guaranteed approach. Responding to this research, a direct loan program was considered in the 1992 reauthorization deliberations, representing the first consideration of a major design change in the program since 1965. The 1992 reauthorization of HEA proposed four major changes to the program. First, financial need calculations would be modified to expand the scope of the program. Second, unsubsidized loans would be offered for the first time. Third, maximum loan limits would be further increased, again expanding the program. Finally, direct loans would be introduced as a pilot program. The first three of these modifications were supported broadly by entrenched interests, which cited a study showing that increases in higher education costs were, in many cases, outpacing individual loan amounts (Archibald 2002). All three changes received bipartisan support given a
134
S. PUTANSU
clear indication that direct loans would be more economical and better control the rising costs of student loans. However, an effective resistance was mounted through several political factors. Ideologically, President Bush opposed direct loans as an expansion of the federal role and as an intervention in private markets, leading him to resist the creation of a direct loan program, even though the relevant policy knowledge had come from his own administration. This ideological opposition was buttressed by resistance from banks, who fought to protect their student loan profits. Institutions of higher education, in contrast, supported the direct loan program, putting them at odds with the banks. While both interests had supported growth in the program over time, the schools felt that they would receive better terms from loans offered by the government. President Bush threatened to veto the bill to stop direct loans, but Congress, controlled by the Democratic Party, fought to maintain the program. Citing information from the Department of Education and GAO about the expected savings from a direct loan program, they created a direct loan pilot program, which was limited to a maximum of 300 institutions, representing less than 5% of total student loan volume (Hannah 1996). While Congress had proposed a comprehensive change to the program, conflict among political factors and policy knowledge worked to moderate the scope of that change. As a result, this decision reflects a mix of politics and a significant, but not comprehensive change to the program. The 1992 compromise over direct loans brought financial aid into the national spotlight as Governor Bill Clinton began campaigning in favor of a direct loan program in 1992. The topic was featured in his campaign ads and in presidential debates leading up to the 1992 elections. Building on his victory, which resulted in unified Democratic Party government, President Clinton expanded the pilot through the 1993 Omnibus Reconciliation Act. The act allowed for a major phasing in of direct loans by 1997. The ideological and institutional alignment of the Democratic Party enabled them to act to make a more comprehensive change to the program, based on the studies of the relative cost of direct loans. The planned phase-in of the direct loan program, however, did not run on schedule. The guaranteed loan program had accumulated an industry of lenders who profited from guaranteed loans without facing the risks of loan default. As the program was being phased in, private banks improved their service levels and product offerings, and worked to prevent direct
6
IMPROVING ACCESS TO HIGHER EDUCATION
135
loans from gaining the expected share of the student loan market. Meanwhile, loan costs after the massive 1992 reauthorization continued to skyrocket, increasing from $13.47 billion in 1992 to $31.35 billion in 1998.1 While student loan programs continued to increase in cost, middleincome families continued to complain about being “priced out” of higher education. Despite two earlier failed attempts to enact tax credits for higher education, and outlining the regressivity of such an approach, the idea was introduced again in 1997. This time, proponents of a small government ideology, who tend to view expenditures through tax credits as a benefit (reduction in government) rather than a cost, worked together to support enactment of the Taxpayer Relief Act of 1997. The Department of the Treasury estimated that these credits would cost $40 billion over five years, approximately equal to the cost of all federal education grant programs, but this estimate was largely ignored in the political debate. The Democratic President, Bill Clinton, threatened to veto these credits, but the Republican-controlled Congress inserted them into the omnibus appropriations bill. Since this bill had followed a government shutdown, which President Clinton did not want to repeat, he chose not to use the veto. This represents a comprehensive decision that was framed by policy knowledge about access, but mostly driven by political factors. The political strength of the Republican Congress, however, waned later in President Clinton’s second term. When HEA was slated to be reauthorized again in 1998, political tensions surrounding direct loans continued to act as a barrier, preventing action directed at alleviating concerns about rising costs. As a result, the 1998 reauthorization was fairly limited in scope. No new loan programs were created, and there were no eligibility changes, no increases in loan limits, and no new decisions about direct loans—maintaining the status quo (Burd 1998). While there were no major changes to the program, the debates over the 1998 reauthorization did introduce a new framing of student loans, driven by policy knowledge. Specifically, several reports showed increasing rates of student loan default, raising bipartisan concerns about the program. By 1990, defaults had increased to over 20% of all loans, and the debate surrounding the 1998 reauthorization attempted to ascribe causes 1 Part of this increase is attributable to the phasing in of FCRA accounting rules in 1990. However, it is not possible to clearly separate out which portions were attributable to program growth versus accounting differences.
136
S. PUTANSU
to this increase. Some decision makers claimed that high default rates were evidence that some schools were not providing education of sufficient quality to place students into jobs that could allow them to repay their loans. Another group argued that default rates were high because banks had set repayment amounts too high for the entry-level positions that most students obtain after college. Others argued that default occurred before students could find a job. Still, others argued that default rates were concentrated in students who had dropped out of college and actually represented a different problem all together. Without clear evidence favoring these explanations, the only action taken was to extend the grace period for repayment from 90 to 120 days. HEA was slated for its next reauthorization in 2004. However, the response to the September 11 attack, the wars in Iraq and Afghanistan, and hurricane Katrina delayed congressional action. Finally, as a result of the passage of No Child Left Behind, which is described in the next chapter, education policy was largely focused on elementary and secondary education. As a result, the HEA reauthorization process would span over 5 years, including four sessions of Congress, and a series of 14 Congressional extensions through 2008. Throughout this period, one of the major issues of consideration was that of increased college costs. This issue was highlighted by a 2002 study of college costs and prices. In 2003, Republican Representative Buck McKeon proposed the creation of a college affordability index. This would require tuition increases to be limited to twice the Consumer Price Index, or to lose eligibility for federal assistance. McKeon, acknowledging advocacy efforts of the American Association of State Colleges and Universities, eventually dropped the college affordability index from the bills, but threatened to return it if tuition continued to increase. The next point of contention in HEA reauthorization was the issue of graduation rates. The Education Trust and a broad coalition of other members of the higher education community mounted a lobbying effort based on the Education Trust’s 2004 report “A Matter of Degrees.” This report presented ways to improve the completion rate of students in higher education, including highlighting equity issues related to the poorest students and to racial and ethnic minorities. One thing these groups argued was that federal money for access to education failed to address a deeper problem of college completion. McKeon responded to this effort with a letter that accused the higher education community of being self-interested and working against the interests of students and
6
IMPROVING ACCESS TO HIGHER EDUCATION
137
their parents. Specifically, he viewed their efforts as an attempt at rentseeking, to receive institutional funding, potentially at a cost to direct aid to students. Movement toward HEA reauthorization began in earnest in 2005, when President George W. Bush nominated Margaret Spellings to be the new Secretary of Education and proposed a budget that expanded funding for Pell grants, increased eligibility limits for loans, and cut lender subsidies. This reflected an ideological compromise, which was driven by a need for bipartisan support in the Senate, where Republicans had only a two-member majority. Putting these HEA changes into the budget effectively removed much of the Title IV debate from the reauthorization process and moved it into the budget reconciliation process, increasing pressure on the Senate to pass the Deficit Reduction Act (DRA), and weakening the ability of banking interests to resist cuts to their subsidies. In addition to President Bush’s push for expansions of financial aid, the act also included a change to the loan consolidation formula, which was driven by findings from a government-wide evidence-based reform. Specifically, the Program Assessment Rating Tool (PART) discovered that a step in the consolidation calculation resulted in an unintentional reduction in interest rates for some students. By modifying the loan formula, the program would be able to avoid the unintended loss of revenues (Ginsburg et al. 1997). The combination of changes in the DRA reflected a mix of impacts from political factors and policy knowledge, but represented only incremental changes. The DRA had reauthorized only a part of HEA, and decision makers continued to work toward a full reauthorization over the next three years, starting with the College Cost Reduction and Access (CCRA) Act of 2007. As with the decisions made in 1992, 1997, and 1998, the passage of CCRA occurred during a conflictual ideological period, with a President and Congress of opposing political parties. Amid this increased ideological and institutional tension, the inclusion and participation of various interest groups gained momentum. Student and university groups pressed the Democratic Congress to reduce interest rates and create easier methods for repayment, while cutting subsidies to banks. Reflecting opposition of banking interests, President W. Bush threatened to veto the bill, but ultimately signed it after it passed both houses with large majorities. The CCRA was driven primarily by these political factors, with scant policy knowledge reflected in debates. It offered moderate changes that provided permanent funding to Pell Grants for the following six
138
S. PUTANSU
years, further lowered subsidies to private lenders, lowered interest rates for students, and instituted income-based repayment and loan forgiveness options for students. The third, and final, decision to fully reauthorize HEA was completed near the end of President W. Bush’s second term, with passage of the Higher Education Opportunity and Affordability Act (HEOA) of 2008, which fully reauthorized HEA until 2014. This legislation was focused mainly outside of Title I, and included several hotly contested issues, including graduation rates, teacher licensing, increased regulations surrounding the lender-institution relationship, and credit transfer policies. Although defaults were not addressed in the bill, the debate surrounding them motivated schools to create a voluntary system of accountability for defaults. This action by higher education interests successfully mollified decision makers, who then kept mention of defaults out of the final bill (Lowry 2009). Another event that added some pressure to finish was when Senator Kennedy (D-MA) became ill and had to relinquish his chair position of the Education Labor Committee to Barbara Mikulski (DMD), who pushed to pass the bill in his name. As had happened with passage of NCLB after the September 11 attacks, this created a moment of institutional alignment that aided in the bill’s passage. While both policy knowledge and political factors shaped this decision, the relatively high levels of conflict ensured that changes were incremental. The seven decisions that were made during HEAs third era reflect limited changes to the program stemming from strong political tensions throughout the period. Figure 6.3 summarizes these decisions in terms of their decision type, as defined by the rational-political and incrementalcomprehensive continua. This shows six of the seven decisions reflect either a mix of political factors and policy knowledge, a scale that is between incremental and comprehensive extremes, or both. The remaining decision is the only one that neatly fits into one of the established models of decision making. Era 4: Federal Expansion (2008–2010) A month after HEOA was passed, the stock market crashed, marking the beginning of the Great Recession. The resulting crisis, reframing of government priorities, and impending elections, combined to create a significant upheaval of the political environment. This ultimately led to a
Mixed
Punctuated Equilibrium Models
2007 Reauthorization
Incremental
Comprehensive
6
PoliticalIncremental
IMPROVING ACCESS TO HIGHER EDUCATION
1997 taxpayer relief act
139
Rational Comprehensive
1993 omnibus
1992 Reauthorization
1998 Reauthorization
2005 omnibus
Political
Mixed
Bounded Rationality
Rational
Fig. 6.3 HEA decisions during the era of muddling toward reform, 1992–2008
unified federal government with President Barack Obama and a majority of Democrats in both houses of Congress. As the financial crisis was blamed largely on banking and financial interests, student loan lenders were significantly hobbled in their ability to continue to advocate for GSL changes. During this period, policy knowledge centered on cost reduction and improved access gave way to a focus on effectively stimulating the broader economy. Table 6.4 summarizes the Four-I’s during this era. Immediately following the market collapse, students, schools, NGOs, and the Department of Education combined to express concern that loans would still be available for the 2008–2009 school year; while banks, guarantee agencies, and Sallie Mae all requested additional capital for lending (Santo and Rall 2010). Further, the presidential elections created pressure on the Republican President to respond quickly, despite ideological concerns about the proper role of the central government vis-à-vis the
140
S. PUTANSU
Table 6.4 Politics and policy knowledge in HEA’s era of federal expansion, 2008–2010 Ideology Interests Institutions Policy knowledge
Unified Democratic government mitigates conflict Financial crisis limits influence of banking interests and strengthens students and higher education interests Financial Crisis creates need to reinforce existing institutional processes Processes : Cost-effectiveness of direct loans highlighted Outcomes: Reduced attention on access Key values: Economy, Efficiency
private sector. The combination of these political pressures led to three decisions from 2008 through 2010, which ultimately combined into a dramatic expansion of the federal role. The first decision of the era was passage of the Ensuring Continued Access to Student Loans Act (ECASLA), which was signed into law in October 2008. This act gave the government the authority to purchase student loans from private banks. This ensured that banks would have sufficient capital to continue issuing new student loans for the school year (Santo and Rall 2010). Through these purchases, the federal government was directly providing capital to lenders and taking ownership of the debt accrued by students. While the combination of policy knowledge and political factors contributed to only a moderate change in the government role, the passage of the act represented a first step toward a broader expansion. The second major change to the student loan programs was passed through a broader action to stimulate the economy. Specifically, the 2009 passage of the American Recovery and Reinvestment Act infused federal spending into a broad array of federal policies and areas of interest. Among these, the unified government under President Obama included increased spending on Pell grants, student loans, direct loans, and government purchases of existing student loans. This represented another moderate expansion of the federal role, which amplified the reach of ECASLA. The most comprehensive change to student loans happened the following year. Higher education interests and student interests increasingly cited studies of the cost-effectiveness of direct loans, supplemented by information on the success of ECASLA in sustaining the school lending market. This policy knowledge, combined with strong ideological and
6
IMPROVING ACCESS TO HIGHER EDUCATION
141
institutional alignment in Congress, and the weakened advocacy position of banks, created strong support for further expansion of federal direct loans. The case for the savings associated with direct loans was seen as so strong, that a conversion of all student loans to direct loans was used to buttress support for the passage of major healthcare reform. Specifically, President Obama was facing stiff opposition to passage of the Patient Protection and Affordable Care Act (ACA), in large part due to the costs associated with the law. In response, the Student Aid and Fiscal Responsibility Act (SAFRA) was incorporated into the ACA. CBO estimated that the takeover of the guaranteed loan programs by the Direct Loan program would generate $68.7 billion in savings over 10 years, which reduced the cost projections for the ACA (Baum and Payea 2011). Further, to reduce ideological opposition to government control of the funds, banks and NGOs were given authority as loan servicers, so that the federal government would not have total control over the student loan system (Negri 2013). The passage of the ACA represented the most comprehensive change to student loans since they were created in combination with the CBO estimate of potential savings. The three decisions that were made during the fourth era of HEA reflect a major expansion of the federal role in providing student loans. Figure 6.4 summarizes these decisions in terms of their decision type, as defined by the rational-political and incremental-comprehensive continua. As shown in the figure, none of these three decisions are fully explained by dominant decision making models, but rather require a balanced and contextual assessment of the middle sections of these continua. Era 5: Muddling Over the Federal Role (2010–2015) Following the major changes in 2008, 2009, and 2010, political deadlock caused by partisan divide between the executive and legislative branches has mostly prevented major changes from being enacted. HEA reauthorization hearings began in 2013, and topics discussed included increasing school costs, Pell grants, and accreditation, in addition to ideas to consolidate individual grant and loan programs or provide free tuition instead of distributing grants and loans. As of the fiftieth anniversary of HEA’s passage (November 2015), the act had been extended but not fully reauthorized. The relative stability of the policy since 2010 reflects the culmination of the reform era and the beginning of an era of managing the expanded
S. PUTANSU
Punctuated Equilibrium Models
Incremental
Mixed
Comprehensive
142
2010 ACA
Rational Comprehensive
2009 ARRA
PoliticalIncremental
2008 Reauthorization
Bounded Rationality
Political
Mixed
Rational
Fig. 6.4 HEA decisions during the era of federal expansion, 2008–2010
federal role. The political landscape has changed dramatically as it relates to student loans. Specifically, banks and guarantee agencies have lost significant advocacy power, and the interests of students and higher education institutions are now negotiating more directly over their relationship with the government lender. Whereas ideological divisions over the size of government continue, the argument over public versus private financial aid has shifted to an ideological debate over the appropriateness of the government to issue debt and charge interest to students. The flow of policy knowledge is increasingly focused on the impacts on student borrowers, including defaults, long-term finances, and inequities across lending groups. This shift in the Four-I’s is outlined in Table 6.5. In 2013, interest rates were set to be pegged to a financial market rate as a result of changes enacted in 2010. This change would have doubled the interest rate being paid by loan holders. Following intervention by
6
IMPROVING ACCESS TO HIGHER EDUCATION
143
Table 6.5 Politics and policy knowledge in HEA’s era of muddling over the federal role, 2010–2015 Ideology Interests Institutions Policy knowledge
Divided government and high levels of conflict Banks largely eliminated as an interest, students, loan servicers, and higher education interests gain strength Growing focus on institutional arrangements for provision of direct loans Processes : Costs of the program continue to be a focus Outcomes: Costs to students and Default rates are highlighted Key values: Economy, equity
student advocacy groups, who offered studies that predicted likely default rate increases that could result from the increased interest rates, Congress acted to delay the change for one year by passing the Bipartisan Student Loan Certainty Act (Chen and Wiederspan 2014). This represented a move to prevent change in the program, driven by a mix of policy knowledge and political interests. Continued ideological and institutional conflict, however, delayed any other specific changes to the program. After a similar debate in 2014, a compromise was made in the budget to tie the interest rate to the ten-year Treasury interest rate instead of the financial market rate. This represents a compromise between the parties that served the Democratic Party’s interest in maintaining lower rates, while also serving the Republican Party’s ideological preference of pegging interest rates to a relevant proxy that is guided by economic markets (Chen and Wiederspan 2014). During the debates over this change, both parties cited the earlier FCRA and offered arguments in favor of different benchmarks for interest rates. Republican lawmakers preferred a market rate, consistent with their preference for more private sector control, while Democrats preferred a short-term federal rate. The 10-year Treasury rate was a compromise that represented a federal source of funds, but which is heavily traded in private markets. As with the 2013 act, this represented an incremental change that balanced policy knowledge and political factors. The two incremental decisions that were made during the fifth era of HEA represent minor changes that underscore a broad political conflict and focus on other policy areas. Figure 6.5 summarizes these decisions in terms of their decision type, as defined by the rational-political and
S. PUTANSU
Punctuated Equilibrium Models
Rational Comprehensive
Incremental
Mixed
Comprehensive
144
PoliticalIncremental
Political
2013 Extension Bounded Rationality 2014 Extension
Mixed
Rational
Fig. 6.5 HEA decisions during the era of muddling over the federal role, 2010– 2015
incremental-comprehensive continua. For both of these decisions, a combination of political factors and policy knowledge influenced the final outcome, contrary to the expectations of dominant decision making models.
Summary of the Case This chapter has presented the legislative history of HEA and has used Weiss’ information utilization framework to explore the relative importance of the Four-I’s from the law’s passage in 1965 through its fiftieth anniversary in 2015. Examples of the use of policy knowledge in decision making occurred in each era, resulting in changes to program formulas, interest rates, and program targeting. Interestingly, each source of policy
6
IMPROVING ACCESS TO HIGHER EDUCATION
145
knowledge is easily grouped thematically in each era, which demonstrates some shared idea of framing and relevance during each period. For passage and the first era of expansion, a shared goal of increasing enrollment resulted in a focus on policy knowledge targeted to identifying how funds could improve enrollment. In the second era, administrative data on program costs took center stage, along with reports focused on budget and accounting from the CBO and GAO. In the third era, policy knowledge on costs continued to be cited, but was overtaken by political factors. In the fourth era, policy knowledge about economic stimulus took center stage and was used to overcome longstanding political divides. Finally, through 2015, other forms of policy knowledge on college costs, access issues, and optimal interest rate setting had gained traction (Favero and Rutherford 2019). The decisions made in each of these eras varied in the degree to which they were driven by that policy knowledge. For example, the expansion of loans was highly aligned with a demonstrated relationship between student financial status and enrollment. At the other extreme, tinkering with eligibility rules and interest rates was driven strongly by consideration of cost information, but these solutions were largely guided not by study of the optimal values, but rather by advocacy from impacted banking interests. This chapter has also shown that political factors were important throughout HEA’s history, and even resisted clear and compelling estimates of cost for fifteen years in the case of direct loan programs. Figure 6.6 shows the number of decisions falling on different parts of the rational-political and incremental-comprehensive continua. While dominant models of decision making presume that decisions are either political or rational, only 6 decisions occurred over this 50 year period that fit this expectation. The other 15 were better explained by considering both policy knowledge and political factors. When compared against the incremental and comprehensive continuum, the expectations of these models are further challenged. Specifically, only 4 of these 21 decisions fit into the dominant models along both continua. In summary, the evidence presented in this chapter demonstrates that, from their initial passage in 1965 until the present day, student loan decision processes often utilized and relied on pertinent policy knowledge that directly impacted the decision outcomes. However, it also shows that political factors, particularly interests, played a significant role throughout the program’s history. This shows that, between HEAs passage and the
S. PUTANSU
Incremental
Mixed
Comprehensive
146
Punctuated Equilibrium Models
3 Decisions
1 Decision
2 Decisions
PoliticalIncremental
1 Decision
6 Decisions
0 Decisions
Bounded Rationality 6 Decisions 0 Decisions
2 Decisions
Political
Rational Comprehensive
Mixed
Rational
Fig. 6.6 Summary of major HEA decisions, 1965–2015
twenty-one major decisions about student loans over the past 50 years, the political-incremental model is reflected in two decisions, while punctuated equilibrium and rational-comprehensive models explain one each. This leaves seventeen more decisions which none of the established models fully explain and which demonstrate the value of a balanced and contextual approach to analyzing governance decisions.
References Archibald, R. B. (2002). Redesigning the financial aid system: Why colleges and universities should switch roles with the federal government. Baltimore and London: Johns Hopkins University Press. Baum, S., & Payea, K. (2011). Trends in student aid, 2011. Trends in Higher Education Series. New York: College Board Advocacy & Policy Center.
6
IMPROVING ACCESS TO HIGHER EDUCATION
147
Burd, S. (1998, October 16). The higher education amendments of 1998: The impact on college and students. Chronicle of Higher Education, A39. Retrived from www.chronicle.com. Cervantes, A., Creusere, M., McMillion, R., McQueen, C., Short, M., Steiner, M., et al. (2005). Opening the doors to higher education: Perspectives on the Higher Education Act 40 years later. TG (Texas Guaranteed Student Loan Corporation). Chen, R., & Wiederspan, M. (2014). Understanding the determinants of debt burden among college graduates. The Journal of Higher Education, 85(4), 565–598. Favero, N., & Rutherford, A. (2019). Will the tide lift all boats? Examining the equity effects of performance funding policies in US higher education. Research in Higher Education, 61(1), 1–25. General Accounting Office (GAO). (1977). Problems and needed improvements in evaluating office of education programs, HRD-76-165. Washington, DC: GAO. Ginsburg, A., Rhett, N., & Cavataio, M. (1997). Strategic planning and performance management: The US Department of Education’s experience. US Department of Education. Hannah, S. B. (1996). The Higher Education Act of 1992: Skills, constraints, and the politics of higher education. The Journal of Higher Education, 67 (5), 498–527. Johnson, L. B. (1965). Remarks at Southwest Texas State College upon signing the Higher Education Act of 1965. American Presidency Project. Santa Barbara, CA. Retrieved from http://www.presidency.ucsb.edu/. Lowry, R. C. (2009). Reauthorization of the federal higher education act and accountability for student learning: The dog that didn’t bark. Publius: The Journal of Federalism, 39(3), 506–526. National Association of State Universities and Land Grant Colleges (NASULGC). (1995). The land grant tradition. Washington, DC: NASULGC. Negri, K. (2013). Mortgaging the American dream: The misplaced role of accreditation in the federal student loan system. Fordham Law Review, 82, 1905. Nevins, A. (1962). The state universities and democracy. Urbana-Champaign: University of Illinois Press. Redburn, F. S. (1988). Never lost a penny: An assessment of federal deposit insurance. Journal of Policy Analysis and Management, 7 (4), 687–702. Rivlin, A. (1961). The role of the federal government in financing higher education. Washington, DC: The Brookings Institution. Santo, G. F., & Rall, L. G. (2010). Private student loan financing in an era of needs and challenges. The Journal of Structured Finance, 16(3), 106–115.
148
S. PUTANSU
Sorber, N. M. (2018). Land-grant colleges and popular revolt: The origins of the Morrill Act and the reform of higher education. Ithaca and London: Cornell University Press. Stein, S. (2017). A colonial history of the higher education present: Rethinking land-grant institutions through processes of accumulation and relations of conquest. Critical Studies in Education, 1–17. Wilson, J. Q. (1989). Bureaucracy: What government agencies do and why they do it (2nd ed.). New York: Basic Books.
CHAPTER 7
Education for the Disadvantaged
President Johnson set the tone for passage of the Elementary and Secondary Education Act (ESEA) with his War on Poverty and passion for education, but it was his Commissioner of Education, Francis Keppel, who skillfully constructed the house of cards needed to pass it (Bailey and Mosher 1968). As described in Chapter 4, President Johnson framed ESEA in terms of equality of opportunity, but it was Commissioner Keppel who had to balance several strong competing interests, address the issues of race, religion, and reds, and craft a bill that would serve Johnson’s goals of equality while also building a foundation for deeper conversation about education quality (Graham 2011). This chapter provides an overview of the expansion of the federal role in elementary and secondary education and a description of the original ESEA titles. It explores four eras that represent the major reauthorizations and appropriations decisions that have impacted the Title I grant programs since they were established in 1965. As with the previous case, the chapter applies Weiss’ Four-I framework to these eras in order to better understand the specific roles and relative strength of ideology, interests, institutions, and policy knowledge in decisions over the program’s first fifty years. For each era, it will consider the landscape of political factors and relevant data, information, and evidence available to decision makers. Then, it will consider each decision made during the era to better understand how these factors interact to support governance decisions. Finally, each era will be summarized to examine where each decision falls on the political-rational and incremental-comprehensive continua, in order © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_7
149
150
S. PUTANSU
to assess the strength of the Four-I approach relative to dominant decision making models.
ESEA Titles and Background ESEA established the federal government’s role in elementary and secondary education by establishing a goal of free appropriate public education for all children. With great effort, Titles I through V of ESEA were passed within eighty-nine days of Johnson’s reelection, and the act was highly visible and accompanied by great fanfare when signed into law. This section describes major aspects of Title I and all the other titles of the law. Title I, known as “Better Schooling for Educationally Deprived Children,” was and still is the largest single federal source of funding for elementary and secondary education, comprising more than 80% of the authorized funds from ESEA. Funding was originally determined based on the impact aid formula, which was calculated by multiplying the percentage of low-income children in each school district by the “federal percentage” of state averages for per pupil expenditures. This funding was administered through the states to individual school districts. Congress established the qualifications for “low-income” and the “federal percentage.” Any districts with at least 3% eligible children were originally eligible for Title I grants, and the formula relied on Census data to determine eligibility, removing the need for means testing by school officials. This formula was strategically important, as it resulted in funds being distributed in the largest amounts to urban areas and rural areas, reviving the Roosevelt coalition of northern cities and the rural south. Although funding was distributed based on district demographics and not controlled by the states, the State Education Agencies (SEAs) were given broad discretion to approve proposals from the districts before funds could be dispersed, placing them into a stronger role in defining policy at the local level. However, this broad discretion was embedded in a poorly defined goal of ensuring that “size, scope, and quality” of the proposals were “appropriate,” as defined by the judgment of SEAs. In addition, Title I requires an evaluation, at least annually, by districts to ensure that the programs effectively meet the needs of educationally deprived children. Title I struck a balance on the religious school issue by requiring local public schools to include children enrolled in private and religious schools in their programs. For example, children could bus
7
EDUCATION FOR THE DISADVANTAGED
151
from religious schools to participate in after-school programs. In addition to representing an expanded federal role in education policy, the use of states to approve and distribute funds inadvertently resulted in an increase of their role. While state control had traditionally been secondary to Local Education Agencies (LEAs), this began a shift from local to stronger state authority. Titles II through V were included in ESEA because Title I did not satisfy all of President Johnson’s policy objectives, but also because various interest groups would oppose the act if their own interests were not met. Johnson engaged in client politics to win over these interests through provisions in the other titles (Lowi 1972). Title II provides grants for “the acquisition of school library resources, textbooks, and other printed and published materials for the use of children and teachers in public and private elementary and secondary schools.” Title II provided $100 million in grants to states over five years. In cases where states have specifically prohibited distributing funds to private schools, the Commissioner of Education was given the authority to use state funds to make materials of the same kind provided to public schools in the state to private schools. Although Title I served private school interests by requiring action from public schools, Title II ensured that religious schools would become direct recipients of federal aid. The Office of Education (OE) stated that this did not violate the establishment clause, but that statement was initially challenged by the Council of Chief State School Officers, the American Civil Liberties Union, and other interest groups, who believed that this clause was an inappropriate expansion of federal power over the states. This growing disagreement reflected the growing tension between African-American and religious interest groups that had previously worked together to support the Civil Rights Act of 1964. Ultimately, the strength of support from the National Catholic Welfare Council, among others, as well as the inclusion of benefits for civil rights groups, tempered this initial opposition. Title III, “Supplemental Educational Centers and Services,” was the ultimate realization of President Kennedy’s proposals for experimental programs, and it was also funded $100 million over five years. This title was framed primarily as a response to the literature on failing schools, with a goal of discovering ways to improve educational outcomes. Like Title I, Title III required LEAs to provide services to private school students, in addition to those enrolled in their own schools. These legislative attempts to develop more information have been a hallmark of all ESEA legislation,
152
S. PUTANSU
highlighting an endogenous relationship between the production of data, information, evidence, and decision making. This is important to note, as the search for policy knowledge about program effectiveness and equity started in the very first days of the law (Keppel 1966). Title IV provided federal funds for educational research and training to various research organizations for the development of educational research. Title IV initially provided $100 million in grants over five years for research projects, an early acknowledgment of the desire for policy knowledge pertinent to education decisions to be available and utilized in decision making. The states, however, were wary of this provision, as it further diluted their ability to control education policy and expanded federal influence. Despite this, the linkage of Title III and IV in a single bill created support that overcame this resistance. Furthermore, Title V, which provided funds ($25 million in the first year) to SEAs to “strengthen their leadership resources” and assist them in establishing and improving “programs to identify and meet the educational needs of states,” provided a benefit to states that further tempered their opposition to other parts of the bill. Commissioner Keppel had proposed this Title in order to garner additional support from the states. Title VI included general provisions, including the provisions limiting federal authority and prohibiting funds for religious instruction discussed in the previous section. A year after ESEA was passed, Commissioner Keppel authored a book in which he described all the threats to the bill, among them the challenge of discussing quality, writing “There was a lack of information. Without a reporting system that alerted state or federal authorities to the need for support to shore up educational weakness, programs had to be devised on the basis of social and economic data” (Keppel 1966). From Keppel’s comments, it is clear that there was some difficulty in measuring the outcome of Title I programs from its earliest days. While the federal government had asserted itself to reduce educational disparities, the lack of policy knowledge about educational outcomes resulted in a program design with lofty, but hard to measure goals, rooted in providing funds without a strong understanding of how to best use them. The rest of this chapter shows that, as other information was gathered following ESEA’s passage, educational outcomes remained extremely hard to define and agree on and thus remain difficult to measure. At its most basic, the ability to measure particularized knowledge, critical thinking, and understanding within each student is difficult to test. Beyond that
7
EDUCATION FOR THE DISADVANTAGED
153
basic level, there is no agreement on which particularized knowledge is most important, whether the goal of “knowing” is sufficient, or if other goals—such as employment potential, good citizenship, and the ability to understand complex policy issues that will be influenced by their elected officials—are as important. Further, the strong tradition of local control in education means that every state, district, and locality may approach these questions differently. Beyond the issues associated with measuring learning, education also faces the challenge of measuring the process of teaching. The difficulty arises from issues of scope and teaching approach. First, the scope of the education system would require an immense amount of resources to measure it in its entirety (Conant 1963). In 2012, total enrollment in elementary and secondary education in the United States was 49.7 million students, and they were taught by 3.7 million teachers. Measuring the output of 3.7 million teachers would be relatively simple if all teaching was performed the same way, but each teacher may have preferred teaching styles, adapted for students who may require varied techniques to improve their learning outcomes (Hmelo-Silver 2004). Further, these adaptations often occur at multiple times throughout the year, as classroom composition may change (e.g., new students). The combination of individualized teaching approaches and the scope of teaching greatly hinder the ability to measure the process of teaching in each classroom. Further, variation in policies across states and localities results in additional difficulties measuring process and outcome at a national level. Difficulty measuring both process and outcomes associated with Title I leads its implementers to work in much the same way as a “coping organization” under Wilson’s typology of bureaucratic organizations.1 As noted in Chapter 2, measurement difficulties on both of these dimensions may create limitations in any policy knowledge that becomes available. These measurement difficulties are expected to be associated with increased influence of political factors on decision making. In the 1963–1964 school year, total federal spending on elementary and secondary education was $900 million ($7.4 billion in 2019 dollars), and this figure more than doubled to $2.0 billion in 1965–1966 ($15.8 billion in 2019 dollars), the first year of ESEA implementation. 1 For simplicity, this chapter uses the phrase “Title I” to refer to the entire set of organizations and programs that, in combination, contribute to the requirements described in Title I of ESEA.
154
S. PUTANSU
Federal spending on K-12 education has continued to grow in nominal terms since passage, but has not always kept pace with inflation. The next five years of ESEA were marked by rapid growth, which averaged 18.6% per year in real terms, and federal spending had reached $4.5 billion by 1971–1972 ($27.6 billion in 2019 dollars). Over the next seven years, in real terms, federal spending continued to grow, but by a relatively smaller 2.8% average rate of growth, reaching $6.3 billion ($28.4 billion in 2019 dollars) in 1975–1976. This trend reversed over the next five years—at an average rate of 6.9% per year, in real terms—and spending fell to $8.6 billion (only $21.2 billion in 2019 dollars) in 1983–1984. Growth was fairly consistent—averaging about 5.7% per year—through the 2000–2001 school year, when it reached $29.1 billion ($42.2 billion in 2019 dollars). But it spiked in the following three years to $41.9 billion in 2003–2004 ($56.9 billion in 2019 dollars), an average growth of 11.6% per year, then remained fairly stable over the next four years before spiking again 2008–2009 and 2009–2010, where it jumped to $76.0 billion ($89.4 billion in 2019 dollars).
Politics and Policy Knowledge at ESEA’s Passage As explained in Chapter 5, the passage of ESEA was driven primarily by strong institutional alignment of the president and both houses of Congress, which enabled decision makers to consider and balance ideological and interest factors. While some policy knowledge was cited during the debate, its impact was minor. Like HEA, the passage of ESEA represents a comprehensive policy change, driven strongly by politics, reflecting some expectations from the punctuated equilibrium, garbage can, and policy stream models. This section summarizes the ideological tensions, major interests, institutional factors, and information that influenced ESEA’s passage, before turning to successive decisions about the program. The ideological tensions between public and private education, a strong and weak government, and centralized power versus state and local authority, each buttressed by a related interest group, were drivers of the specific terms of the act, which represented a balance among each of the three political factors. The public versus private tension was highlighted by religious interests that worked to make sure federal funding for nonreligious activities could be distributed to religious schools. The weak versus strong government tension was made obvious by restrictions on
7
EDUCATION FOR THE DISADVANTAGED
155
the federal government’s ability to set curriculum requirements, driven by SEAs, LEAs, and educators. Finally, the long tradition of local control in education resulted in SEAs and LEAs lobbying to limit this expansion of the federal role to distributing funds. Most importantly, ESEA represented a major expansion of the federal role, largely driven by a concept of equity that was supported strongly by civil rights groups (Kaestle 2016). Institutional factors provided an opportunity for these ideological and interest group factors to take action. Specifically, a unified government with Democratic Party control of the presidency and Congress, relaxed tensions between the executive and legislative branches. The major interests at ESEA’s passage remained important to decisions about Title I programs over the ensuing fifty years. These include civil rights groups and religious schools—which represented the interests of race and religion during ESEAs passage—as well as educators, the states, and local governments. Throughout this case, civil rights groups are defined as groups that have an advocacy interest in equality related to race, gender, or other protected classes. When ESEA passed, these groups were particularly strong, as they had coalesced to advocate for passage of the Civil Rights Act. Religious schools are defined as the schools themselves as well as related religious interests. Educators include organizations that represent the interests of principals, teachers, and their unions, all of whom provide education to students or have an interest in student outcomes as well as fair labor practices. State interests include central state offices with an interest in managing and distributing federal funds and in regulating education within their SEAs. Finally, in this case LEAs are the school districts and district officials, which have the most direct management of individual schools. At ESEA’s passage, which occurred around the same time as HEA, President Johnson and the unified Democratic Party control of Congress mitigated ideological and institutional tensions. For ESEA, this institutional alignment was particularly critical because it eased resistance to a central focus on public schools, expanding the federal influence on education, and federal redistributive efforts, and made it possible to pursue these ideological preferences while winning the support of relevant interest groups. Throughout the case, whether changes to Title I are being made in the context of a reauthorization or as part of an appropriations process are another institutional factor that will be considered. At ESEA’s passage, no single source of policy knowledge was particularly influential, but a major driving factor was common belief about
156
S. PUTANSU
the inequality inherent in the US education system. Two forms of policy knowledge were cited in support of ESEA. First, information reflecting the inequity of the US education system reinforced support for the preexisting opinion of the decision makers. Second, data from the impact aid formula that DOD had implemented to provide funds to schools near military bases were used as a benchmark for Title I funds. As demonstrated in Chapter 5, the passage of ESEA was driven strongly by the political factors of ideology, interests, and institutions, with a limited role for policy knowledge. Where HEA was dominated by a relatively few interest groups, the passage of ESEA required efforts to satisfy a much broader collection of interests (Casalaspi 2017). In the next section, this chapter turns to the four major eras of ESEA, examining all 17 decisions that occurred from 1965 through 2015. The analysis of this case shows that 8 of 17 decisions are explained by dominant decision making models, while the other 9 fall in the middle of either the political-rational or incremental-comprehensive continua.
Major Eras of Title I Grant Programs The passage of ESEA, as described in Chapter 5, represented major victories for the ideological supporters of a public role in education, a strong central government, and federal influence on education policy. These victories were achieved through a combination of institutional factors related to a unified government controlled by the Democratic Party, which utilized policy knowledge to advocate for their existing ideological beliefs and created a tenuous balance of interests achieved through inclusion of several policies to engage their support and calm their opposition. Over the following fifty years, major decisions about Title I follow a similar theme—factors of ideology and interests remain powerful, while policy knowledge is mostly used to frame problems or offer support for preferred policy interventions. These major decisions can be grouped into four major eras: (1) Expanding equity, (2) Reversion to the States, (3) Federal accountability, and (4) State flexibilities. The following sections explore the major changes to the law over this period. Era 1: Expanding Equity (1965–1980) President Johnson wrote that the passage of ESEA was one of his greatest accomplishments, a major step toward ending segregation, furthering the
7
EDUCATION FOR THE DISADVANTAGED
157
goals of the Civil Rights Act, and a first step in his “War on Poverty.” As described in Chapter 5, this decision was driven primarily by political factors and represented a comprehensive change to the federal role in education, consistent with a punctuated equilibrium model of decision making. However, federal involvement in K-12 education remained contentious. Conflict and compromise among ideologies and interests led Title I to be replete with multiple and conflicting goals. As a result, the program was continuously challenged, amended, and expanded throughout the 50 years covered by this case study. In fact, almost immediately after the bill was passed, Congress took up a new debate over whether the benefits of Title I should be targeted to the poor or targeted to all children at risk for school failure (Stein 2004). The first era of ESEA begins with the 1965 passage and continues through these early debates and subsequent amendments in 1965, 1966, 1967, 1970, 1974, and 1978. Throughout the fifteen years covered by ESEA’s first era, the Democratic Party retained strong majorities in both chambers of Congress. The first three major decisions in this era occurred while Johnson was still president and continued the strong coordination of civil rights and teachers’ interests. This was followed by one decision each under presidents Nixon, Ford, and Carter. The united Congress was a driving force in these decisions, maintaining ideological support for the program even during the Republican presidencies. However, Republican support for smaller federal government and more state authority helped to nurture and represent state interests, which became more powerful throughout the second half of this era. During this period, policy knowledge supplemented the strength of ideological, interest, and institutional alignment. Political factors were in relative agreement that both funding and outcomes of education were unequal, but data, information, and evidence used throughout the period were targeted toward better defining these inequities. Table 7.1 summarizes these Four-I’s for ESEA’s first era. Through its first three years, Title I was amended three times. Each of the three amendments to ESEA represented expansions of eligibility and funding for Title I. For the 1965 amendments, Title I was amended to provide additional funds for disabled students. In 1966, benefits were extended to children in foster homes and juvenile institutions, and the definition of low income was expanded to include more students. Benefits were also extended to cover children of migratory parents. For the 1967 amendments, the poverty threshold was increased, leading to coverage of more students from low-income families. In each of these cases,
158
S. PUTANSU
Table 7.1 Politics and policy knowledge from ESEA’s passage and era of expanding equity, 1965–1980 Ideology
Interests Institutions
Policy knowledge
Democratic Party control of Congress mitigated ideological conflicts, and stymied Republican and state opposition to expansion Civil rights and advocacy groups for disadvantaged populations remain powerful, state interests generally weak Creation of institutional arrangements to support programs, including National Commission on School Finance (for oversight), National Center for Education Statistics (to assess impact), and the Department of Education (to centralize education functions) Processes : Unequal school funding. Defining and determining eligibility Outcomes: Inequitable outcomes, state behaviors, program impacts Key values: Equity
specific groups representing the interests of newly eligible students were at the center of the debate, providing policy knowledge showing that each group faced systematically reduced education funding and systematic disadvantages in educational outcomes. The institutional dominance of the Democratic Party, which was already supportive of expanded equity, helped enable these interests to successfully press for these changes. During this first three year period, state interests began to successfully challenge expansions of federal authority. For example, in debates over the 1967 amendments, states successfully convinced Congress to use annual census numbers, rather than the decennial census, to identify the number of disadvantaged students, to institute special incentive grants to provide more money to the states, and to guarantee minimum grant amounts without need for appropriations. States also pushed for expanded use and flexibility of funding during this period. Originally, Title I funds were capped at 30% of the total state education budget, with only 1% of outlays allowed to be used on state administrative costs. As part of the 1967 amendments, states successfully lobbied to have the cap increased from 30 to 50% of the state education budget. All three of these decisions reflected the political-incremental approach to decision making, with debates dominated by political factors and resulting in small tweaks to program goals and processes. However, a common theme of these debates was a concern about the lack of data, information, and evidence to understand the extent
7
EDUCATION FOR THE DISADVANTAGED
159
of the problems ESEA sought to address, or the effectiveness of Title I in addressing them. As a result, the 1967 amendments required the OE, through the National Institute of Education (NIE) to develop a report on the most promising compensatory education programs receiving Title I funding. State influence continued to grow under Republican presidents, beginning with the 1970 reauthorization. Once again, states drove a change in the data used for the funding formula, pressing Congress to fully eliminate the use of the census, instead allowing state- and school-level data to be used. They also successfully pushed to expand the use of administrative funds from 1% of outlays to 1% of allocations.2 At the same time, Democratic control of Congress helped to ensure a commitment to reducing inequities. Schools, teachers, and advocacy groups complained that states were abusing their flexibility by reducing their own budgets as they took in federal funds, negating the intent of the law. To address this issue, Congress sought to create the National Commission on School Finance which would serve in an oversight role to ensure that federal funds “supplement, not supplant” federal funds, ensuring that the federal intent to expand education was being met by the states. The Republican President, in an attempt to uphold ideology in favor of state authority, threatened to veto the act, arguing that supplemental funding had not been demonstrated to improve educational outcomes. Congress responded to this threat by including grant funding for planning and evaluations into the effectiveness of Title I programs. Under Republican President Ford, the institutional tension continued to be a primary driver of decisions, resulting in limited substantive changes to Title I in the 1974 amendments. Specifically, citing a lack of evidence about its effectiveness, Ford suggested Title I should be eliminated altogether. This concern about evidence was further highlighted during a series of congressional hearings in which none of the parties testifying could speak to the actual impact of Title I. Again, the Democratic Congress responded not by eliminating the program, but by pressing for
2 State funds are first authorized, which sets the total amount that can be spent. These
authorized funds are then allocated through budget and planning processes, to determine where states intend to spend them. Finally, these allocated funds are spent as outlays. Typically, allocations will plan to use the full authorization, but outlays may not match allocations. This is particularly true in early stages of implementation, while a program is still developing the needed structure to manage the program.
160
S. PUTANSU
the production of new policy knowledge. Specifically, they created the National Center for Education Statistics and required the Commissioner of Education to provide “such technical and other assistance as may be necessary to State educational agencies to enable them to assist local educational agencies in the development and application of a systematic evaluation of programs in accordance with the models developed by the Commissioner.” More significantly, however, a coalition of state and education interests, citing information on the distribution of funds, argued that the Title I funding formula was exacerbating inequities. Specifically, because the formula was responsive to state spending on education, it tended to provide more federal funds to the wealthiest areas, including urban centers. In response, formulas for Title I grant programs were all revised to reduce this effect and distribute a larger proportion of funding to poorer areas, especially in rural communities. Both the 1970 and 1974 amendments reflected moderate changes to the Title I program. However, both cases were prompted by particular policy knowledge, and limited by political conflict. In 1970, an Office of Education report on reductions in state appropriations led to the creation of the “supplement, not supplant” rule. Similarly, in 1974, information on federal fund allocation led to significant revisions to grant formulas. These decisions do not meet the expectations of dominant models. Specifically, they do not comprehensively assess all aspects of the programs, nor do they rely on established heuristics. Instead, they show that changes to policy were impacted by information, but that the extent of change was limited by political tensions, resulting in a moderate, rather than incremental or comprehensive change. The election of Jimmy Carter resulted in a reduction in ideological and institutional tensions that had slowed the expansion of ESEA in 1970 and 1974. However, continued concern about the effectiveness of Title I continued to play a role. Specifically, in early hearings, several decision makers continued to consider elimination of Title I, citing uncertainty about its impacts. One such area of focus was the common practice of “pullouts”—Title I students were removed from their classroom for special instruction, which some argued improved outcomes for those students and others argued was disruptive to both recipients and non-recipients of Title I funds (Glass and Smith 1977). However, as reauthorization started to take shape, a series of reports were issued by the NIE, satisfying the 1967 requirement for the OE to
7
EDUCATION FOR THE DISADVANTAGED
161
evaluate compensatory education. These evaluations, led by Paul Hill, essentially halted discussions about eliminating Title I and refocused conversations around the evaluations themselves. Reports were issued in succession, and each cataloged one or more of the many diverse goals associated with Title I funds. For each goal, the evaluations provided some means of measurement and some discussion of the relationship with other goals (Jones 2003). These reports further identified several ways to improve future studies, and these recommendations were echoed by the General Accounting Office (GAO 1977). While the NIE reports did not resolve the identified issues and tensions, they did provide decision makers with a list of topics to discuss and attempt to address. As a result, the 1978 amendments included several efforts to improve data, information, and evidence. First, it authorized the use of school-level enrollment data to demonstrate need, which was expected to improve the targeting of funds. Further, it provided additional funding if 5000 students or 20% of students were identified as low income. Finally, it supported the creation of school-wide Title I programs, which authorized the use of Title I funds for all students, in schools with 75% or more students who were low income. This represented a substantial break from the early practice of targeted-assistance programs, which had to go directly to individuals and helped perpetuate the practice of pull-outs. This decision showed an important combination of political factors and policy knowledge. The seven decisions that were made during ESEAs first era represented the creation of a federal role in elementary and secondary education, and the expansion of that role in pursuit of greater equity among students. Beyond the scope of Title I, the partisan alignment and dominance of an ideology in favor of strong government were symbolically made permanent. In 1980, President Carter pressed for and successfully negotiated the establishment of the US Department of Education (ED). This had important implications for Title I, as the department-level status for ED now elevated the importance of the federal role for future decisions (Radin and Hawley 1989). The creation of ED also represented the institutionalization of many of the federal programs and policies that were authorized through HEA, ESEA, and other education statutes (McGuinn 2006). As shown in Fig. 7.1, the first four decisions of this era fit into the predictions of political-incremental and punctuated equilibrium models, whereas the other three decisions fall outside of the dominant decision models.
S. PUTANSU
Punctuated Equilibrium Models 1965 Passage
Incremental
Mixed
Comprehensive
162
1978 Amendments
Rational Comprehensive
1970 and 1974 Amendments
PoliticalIncremental Bounded Rationality 1965, 1966, 1967 Amendments
Political
Mixed
Rational
Fig. 7.1 ESEA decisions during the era of expanding equity, 1965–1980
Era 2: Reversion to the States (1980–1988) The election of President Reagan and a Republican-dominated Senate in 1980 signaled a reversal in the ideological bent of the national government. Reagan, who had campaigned on a promise of eliminating ED, began attempting to reduce the federal role in education and dismantle some of the equity provisions in Title I in favor of state flexibility and control (McGuinn 2005; McDermott 2011). Unlike with Presidents Nixon and Ford, President Reagan was able to have a more direct influence on education policy through the Republican-controlled Senate and through strategic use of the appropriations process. State interests continued to have a strong influence during this period and were joined by increasingly vocal religious school interests. These political factors were dominant during the next three major decisions related to ESEA, constituting the era of reversion to the states. However, the introduction of
7
EDUCATION FOR THE DISADVANTAGED
163
Table 7.2 Politics and policy knowledge in ESEA’s era of reversion to the states, 1980–1988 Ideology
Interests Institutions Policy knowledge
Republican Presidents and increasing power in Congress highlight ideological preference for reduced federal role, increased attention on reducing spending Resurgence of state and religious interests Movement toward use of block grants and state flexibility Processes : Minor focus on costs Outcomes: Introduction on test scores becomes more important Key values: Effectiveness
new policy knowledge, related to the United States’ international competitiveness in math and reading, began to shift the topics considered by decision makers from a focus on equity to one of effectiveness. Table 7.2 provides a summary of these Four-I’s during this era. President Reagan began taking action to reduce federal control education early in his term. Because ESEA was not up for reauthorization, he leveraged the 1981 appropriations process as a vehicle for passing the Education Consolidation and Improvement Act (ECIA). This act consolidated 42 ESEA programs into just seven, with reduced federal regulations for states (Kimbrough and Hill 1983). Title I grants were not consolidated, but the definition of disadvantaged student was relaxed, states were given increased flexibility to design their own evaluations, and federal spending was reduced. Each of these changes represented a moderate reduction in federal role and influence and was justified by ideological support for state’s rights and a limited federal role, without direct reference to policy knowledge related to the design or effectiveness of Title I. The ideological and institutional shift that resulted from Reagan’s election can thus be viewed as a “shock” to the political equilibrium that was sufficient to drive through a comprehensive change to ESEA overall, but only a moderate change to Title I specifically. The Secretary of Education under President Reagan, Terrel Bell, had come into some conflict with the administration during the 1981 amendment process after speaking out in favor of public schools and against the elimination of ED. He appointed a National Commission on Excellence in Education to study the performance of public schools, in the hopes that positive findings would limit future calls for radical education reforms (McGuinn 2006). Reagan ordered the Commission to “focus on five fundamental points that would bring excellence back to education,” which
164
S. PUTANSU
included religious instruction, tax credits for families who send children to parochial schools, vouchers, parental responsibility, and a call to abolish ED. The Commission did not follow either Bell or Reagan’s guidance, but completed the study in 1983 and released it under the title A Nation at Risk: The Imperative for Educational Reform (National Commission on Excellence in Education 1983). In a surprise to Reagan and his Secretary of Education, the report reviewed test scores of American students in math and reading, among other areas, and compared these results to other nations. The study concluded that American students lagged behind their international counterparts in nearly every educational topic. This, warned the report, could have detrimental impacts on economic productivity, technological development, and the overall influence of the United States on the world stage. President Reagan did not take the report seriously. At a ceremony at the White House Rose Garden, he received the report publicly, praising “its call to end federal intrusion on education,” despite the fact that the report made no such recommendation. Reagan went on to promise to “work in the months ahead for passage of tuition tax credits, vouchers, educational savings accounts, voluntary school prayer, and abolishing the Department of Education,” none of which had been endorsed in the report (McGuinn 2006). Congress, too, seemed to disregard the findings. In the 1984 amendments to ESEA, responding to private sector interests pressing for school choice, they permitted for-profit organizations to participate in grant programs if they can: (1) make a significant contribution to attaining the objectives of the Act and (2) provide substantially equivalent education at a lesser cost or provide services and equipment not available in public institutions. This represented a shift from state interests to private market interests. However, the immediate impact of this provision did not substantially alter the implementation of Title I, but was seen instead as a moderate change. By 1988, the Democratic Party had regained control of the Senate, weakening the ideological influence of President Reagan. During this time, the results of A Nation at Risk had penetrated to political dialog and were a central debate topic in the 1988 presidential election cycle. George H. W. Bush, President Reagan’s Vice President, relaxed the centrality of state control as part of the Republican education platform, instead focusing on calls to improve the performance of US students overall. The 1988 amendments were passed during the election
7
EDUCATION FOR THE DISADVANTAGED
165
cycle and paid homage to A Nation at Risk by setting a broad goal of cultivating school improvement and excellent programs, and by allowing for states to initiate “program improvements” in cases where Title I was not having the desired impact. While this represented a formal recognition of the study, it did not result in significant substantive changes to the program. The only other change instituted through the 1988 amendments was the removal of a matching requirement for LEAs to use Title I funds for school-wide programs. This change met the interests of educators and civil rights groups, who desired a broader use of Title I funds, but also was consistent with Reagan’s state-rights ideology, as it removed a federal requirement and provided additional flexibility to the states. Given the centrality of education to the presidential elections, these amendments represented only minor changes to the law and reflected a careful balance of attention to political factors and policy knowledge. The three decisions that were made during ESEA’s second era reflected an emphasis on ideological preference for state control of education and a shift from a focus on student disadvantages and on program impacts to a focus on the proper role of the federal government in implementation. While initial efforts under Nixon drove some comprehensive changes to ESEA’s grant structures, the impacts on Title I were relatively more moderate, reflecting some degree of tension between the Democratic Party in the House and the Republican Senate and Administration. As shown in Fig. 7.2, none of these three decisions fits the dominant models of decision making. Era 3: The Era of Federal Accountability Standards (1988–2002) As President George H. W. Bush came into office, A Nation at Risk had become a central component of debates over ESEA. The issues it identified had started being explored by state governments and through the National Assessment of Education Progress. During the 1988 election cycle, both political parties and all interests expressed agreement that there was a serious problem with American education (McGuinn 2015). As such, questions about student performance became central to considerations of ESEA for the five decisions over this 14-year long era. Despite the centrality of this policy knowledge, resistance and tensions among ideological and interest group factions resulted in high levels of conflict and resistance to substantial policy changes. Specifically, although the Republican and Democratic parties largely agreed on a need for an increased
S. PUTANSU
Punctuated Equilibrium Models
Mixed
Rational Comprehensive
1981 and 1984 Amendments
Incremental
Comprehensive
166
PoliticalIncremental
1988 Amendments
Bounded Rationality
Political
Mixed
Rational
Fig. 7.2 ESEA decisions during the era of reversion to the states, 1980–1988
federal role, conflict between the parties led to institutional resistance to change, and civil rights interests expressed reticence to trust Republican efforts. Table 7.3 provides a summary of these Four-I’s during this era. President Bush’s efforts cemented important groundwork for the development of a national agenda of education reform, starting with introduction of legislation to Congress in the first (and again the second) year of his presidency. These bills introduced new Republican-backed educational interventions centered on school competition, choice, and accountability (McGuinn 2006). However, partisan animosity, exacerbated by educators and civil rights groups’ distrust of the Republican Party, combined to limit the extent of changes to the law, and both bills failed in Congress. President Bush attempted to grow more support for these reforms by inviting Governors from all fifty states to attend
7
EDUCATION FOR THE DISADVANTAGED
167
Table 7.3 Politics and policy knowledge in ESEA’s era of federal accountability standards, 1988–2002 Ideology
Interests
Institutions Policy knowledge
Divided government throughout most of the era, with high conflict within federal government and between the federal government and states State and LEA resistance to federal education goals, increased pressure from religious schools, private schools, and civil rights organizations Creation of basic goals and standards, strong federal oversight Processes : Diminished focus on process related measures Outcomes: Test scores begin to dominate (including achievement gaps) Key values: Effectiveness, Equity
the “Charlottesville Summit” to discuss state preferences for education reforms. In 1991, President Bush introduced a plan for education reform, titled America 2000, which reflected goals established in the Charlottesville Summit. However, the plan met with opposition from both sides of the aisle, with Democrats assaulting it as “education strip mining” that robbed existing programs to pay for new ones, while conservative Republicans continued to resist on ideological principle against federal involvement. No legislation resulted from America 2000, as political pressures amounted to no more than incremental changes in 1992. President H. W. Bush was ultimately defeated in the next election cycle by Bill Clinton, before being able to institute any kind of substantive education reform. The political stalemates that resulted in a status quo for Title I during these nearly four years reflect the expectations of the political-incremental model. During the 1992 election campaign, Bill Clinton repeatedly declared that “We need to overhaul America’s public education system from top to bottom.” Although he and President Bush had large areas of agreement on education policy, Clinton was successful in highlighting vouchers and block grants as evidence of Bush’s “Radical Republican” ideology. However, Clinton continued to support the focus on standards, tests, and choice that had been developed during Bush’s presidency. After the election, which put Democrats in unified control of the government, President Clinton resurrected America 2000 as Goals 2000 and used his partisan advantage to pass the law, even though it was opposed by the NEA
168
S. PUTANSU
and other education agencies on the left, and by the conservative Republicans who opposed increased federal involvement on the right. Clinton’s support, along with his pledge to make standards voluntary, ultimately helped Goals 2000 become law. President Clinton’s Secretary of Education, Richard Riley—the former Governor of South Carolina and chair of the Charlottesville Summit— claimed that Goals 2000 was “a vehicle to facilitate the implementation of a state-defined reform strategy, not as a new free-standing federal program with its own rules and regulations […] virtually no strings were attached to the use of Goals 2000 funds” (Schwartz, et al. 2000). This reauthorization, Public Law 103-382, renamed ESEA as the Improving America’s Schools Act (IASA) and delivered on Clinton’s promise to overhaul Title I. Congress rewrote the law to ensure that Title I programs were in line with the standards-based education reforms taking place in general education. These changes were designed to ensure that educationally disadvantaged students were educated according to the same high standards that states were establishing for all students, in order to make the system more equitable (McDermott 2011). To accomplish this purpose, Title I was modified to require states to: “ensure high standards for all children and align the efforts of states, local educational agencies and schools to help children served under this title to reach such standards.” Other changes to the law included provisions intended to offer children an “enriched and accelerated educational program” that included school-wide programs, requirements for “effective instructional strategies” and “challenging academic content” and improved quality of instruction (National Research Council 1999). A key provision of the act was intended to improve accountability by using state assessment systems. This provision was expressed as a quid pro quo for schools and teachers, who would get greater decision making authority and flexibility in exchange for this increased responsibility for student testing performance (Romano 1999). For a state to receive Title I funds, it was required to develop assessments based on the state content and performance standards. These assessments were not intended to make decisions about individual children, but rather to keep track of how well districts and schools were enabling Title I students to meet standards (Jenkins and Hill 2011). School-wide programs were required to identify individual students having difficulty mastering any of the identified standards and provide the identified students with timely and effective assistance. School-wide and
7
EDUCATION FOR THE DISADVANTAGED
169
targeted-assistance programs had to demonstrate sufficient gains in performance, as measured by state assessments, for all students being served, including students with disabilities and students who are low-income or have limited English proficiency (Rivera and Stansfield 2004). Though IASA passed with support from moderate Republicans and Democrats, it faced serious opposition from the far left and far right. Resistance from the latter group ultimately resulted in removal of enforcement mechanisms from the law and led to a reduced authority to sanction states and districts that failed to comply. These changes gave a high degree of flexibility to SEAs to define their own standards, self-determine if they were in compliance with the new mandates, and resist accountability to the federal government. As a result, there was minimal change to education practices (McGuinn 2006). Because many of these goals and requirements had already been adopted individually at the state level, it could be argued that IASA represents little more than a federal stamp on state efforts. However, the promulgation of specific national goals and requirements represents a comprehensive shift in the declared federal role in education policy. The changes made in 1994 represent the first direct federal response to A Nation at Risk and reflect a delicate balance of political factors and policy knowledge not anticipated by the dominant decision making models. About a month after the passage of IASA, the unified Democratic Party control of government was ended with a sweeping electoral victory for Republicans in both the House and Senate, engineered by Newt Gingrich (R-GA) and the “Contract with America.” As Republicans took office the following January, a common thread through the contract was the imposition of restrictions on federal authority. In that vein, the Republican Congress lobbied several attacks at President Clinton’s education policy, including severe cuts to education in the 1995 budget. President Clinton vetoed that budget, citing cuts to education as a reason and leading to a very public shutdown of the federal government. Following the government shutdown, the tension between the President and Congress continued. However, Clinton’s reelection in 1996 was framed as an indicator of broad support for his education policies and led to fewer Republican attacks on IASA, and even some modest expansions of the programs (Meier 2004). Most of these proposals were narrow, targeted education programs that were not related to Title I. However, for the reauthorization of ESEA in 1999, Clinton used his State of the Union address to reinforce his support for strengthening the standards
170
S. PUTANSU
and accountability measures of ESEA. In response, Republicans introduced the “Academic Achievement for All Act,” which became known as the “straight A’s” (Calzini and Showalter 2009). The Republican proposal was largely similar to Clinton’s, but it increased federal spending by a larger amount, focused more on teacher improvement, and turned categorical grants into block grants. Democrats in Congress responded with a third plan, called the “Public Education Reinvestment, Reinvention, and Responsibility Act” or “the Three R’s,” which integrated many aspects of both plans but added on federal rewards and punishments for state performance toward educational goals. Ultimately, the divide over the three plans, along with pressure from the 2000 presidential election, led Congress to table the reauthorization, hoping that the election results would inform which of the plans to follow (McGuinn 2006). As with the stalemates under President Bush, the 1999 stalemates reflect the expectations of the political-incremental model. During the presidential elections in 2000, George W. Bush successfully took the helm of the debate over education, mostly mirroring President Clinton’s talking points and effectively preventing Al Gore from raising strong education-related campaign challenges. In fact, many aspects of Bush’s plan were taken directly from President Clinton and his New Democrat Progressive Policy Institute (McGuinn 2006). Bush’s announcement of support for many of these initiatives likely played a role in his election victory over Al Gore in 2000, as he co-opted many of the ideas before Gore had stated a platform. In addition, state reforms in Texas while Bush was governor added to his credentials as an education advocate, and reports of their results were used to support his proposals for increased testing and accountability. President Bush ultimately proposed a national strategy for education, but left Congress to work out the details, inviting the Democrats who had sponsored the Three R plan to an education policy summit held before his inauguration, in which he invited bipartisan work on a bill, based on his 28-page strategy. Among the bipartisan compromises made in drafting the bill was President Bush’s decision to drop his proposals of vouchers for private and parochial schools early in the process, the agreement to implement a pilot program to demonstrate the potential of block grants, rather than a wholesale replacement of categorical grants with block grants, and massive increases in federal spending, championed by Senator Ted Kennedy (D-MA). The centerpiece of Bush’s proposal and the most conflictual part of the bill were the testing provisions, which
7
EDUCATION FOR THE DISADVANTAGED
171
required the same annual yearly progress that had been included in IASA, but which has far sharper teeth, in terms of enforcing rewards and punishments related to the achievement of that progress. As with previous attempts to assert federal requirements, testing provisions had the potential to derail the bill. However, following the attacks on September 11, 2001, there was a great desire among elected officials to pass positive legislation and show unity. Likely due to the broad agreement on a need for accountability, the first domestic bill to be voted on (after the PATRIOT Act) was Bush’s proposal, No Child Left Behind (NCLB). In perhaps the decision most reflective of a policy window or punctuated equilibrium, NCLB passed by wide margins in the House (381-41) and the Senate (87-10). The four decisions that were made during ESEA’s third era reflected a definitive shift in the federal role, and the most comprehensive change to the policy since its passage in 1965. The passage of NCLB can be seen as a watershed moment for the impact of A Nation at Risk. This report, originally misquoted and buried in 1983, had slowly but surely reframed the debate about the federal role of education from disagreement over whether to be involved to broad agreement on a role that overcame traditional partisan and ideological battle lines. This created a convergence between the Democratic Party, which moved from promoting expanded federal influence as a means to improve equity, and the Republican Party, which retreated somewhat from its ideological preference for state and local control of education. Mutual acceptance of the importance of educational effectiveness ultimately served as a foundation for this comprehensive change to the law. During this era, as shown in Fig. 7.3, three of four decisions would be explained by dominant decision models. However, these models are limited in their ability to explain the long-term acceptance of A Nation at Risk or its ability to reframe political positions, particularly as reflected in the passage of IASA. Era 4: Returning State Flexibility (2002–2015) Although standards were required under IASA during the Clinton administration, the Bush administration’s implementation of these requirements through NCLB was quite different than Clinton’s had been. While Secretary Riley had made a strong effort to promote the “no strings attached” nature of federal funds, President Bush’s Secretary of Education, Rod Paige, took a different approach. Shortly after the bill was passed, Paige
S. PUTANSU
Punctuated Equilibrium Models 2002 NCLB
1994 IASA
Rational Comprehensive
Incremental
Mixed
Comprehensive
172
PoliticalIncremental Bounded Rationality 1992, 1999 stalemates
Political
Mixed
Rational
Fig. 7.3 ESEA decisions during the Era of federal accountability standards, 1988–2002
met with thirty state education chiefs. In this meeting, he explained that they would all be held to the letter of the law, that noncompliance would not be tolerated, and that no waivers would be granted (McGuinn 2015). This stance on the federal role, and the focus on using test scores as the primary source of information in education decisions, mobilized education interests, state officials, and others into strong opposition. Broad opposition made appeals to ideological opposition to federal involvement, equity concerns, labor issues, and private markets. As these interests produced and shared increasing amounts of policy knowledge demonstrating process flaws and negative consequences of NCLB, ideological opposition to NCLB grew on both sides of the aisle. However, continued institutional tension throughout this era, largely defined by divided government, constrained decision makers from taking affirmative action to ease these
7
EDUCATION FOR THE DISADVANTAGED
173
Table 7.4 Politics and policy knowledge in ESEA’s era of returning state flexibility, 2002–2015 Ideology Interests Institutions Policy knowledge
Broad agreement about overstep of the federal role across federal partisan boundaries Resurgence of interests, including states, civil rights organizations, teachers unions, and parents organizations Executive branch flexibility and waivers, alternative funding Processes : Focus on standards, unfunded mandate Outcomes: Focus on graduation and drop out rates, long-term achievement Key values: Effectiveness, equity
requirements. As a result, ESEA would not be reauthorized even once from the passage of NCLB in 2001 to its fiftieth anniversary on April 7, 2015. The political factors and policy knowledge that dominated this area are summarized in Table 7.4. The Bush administration’s decision to strictly enforce NCLB standards and use a one-size-fits-all approach led to stiff resistance from states, teachers’ unions, and civil rights groups (McGuinn 2015). This included passage of laws in Colorado granting local districts the freedom to opt out of NCLB, and passage of a law in Utah refusing to accept NCLB requirements. At the same time, each type of major interest complained about various provisions of the law. This included a myriad of civil rights groups, who have seen that the burdens of corrective action and other unintended consequences of the law have been disproportionately faced by minorities (Hirschland and Steinmo 2003). By 2004, state legislatures had begun to declare that NCLB was an unconstitutional intrusion on state rights, groups across the education community argued that it was underfunded, and schools complained that it was being administered unfairly and inflexibly (Hirschland and Steinmo 2003). By the end of 2004, a total of thirty-one states had introduced bills seeking greater flexibility in administration of NCLB, increased funding for implementation, or limitations in state participation (Manna 2006). In its initial response to these complaints about the federal requirements, the Bush administration responded by threatening to stop providing federal funds to states that were not in compliance, and Secretary Paige went so far as to call the National Education Association a “terrorist organization” (McGuinn 2015). However, growing disillusionment with the law continued, concurrent with studies that began to find flaws in
174
S. PUTANSU
and attack the Texas model, and as a result, Secretary Paige was replaced with long-time Bush adviser Margaret Spellings. Spellings announced a “new common sense approach” to implementation, which limited state requirements to four “bright lines”: annual testing, disaggregated data, teacher quality improvements, and dissemination of information to parents (Evans 2019). While not a legislative decision, this represented a moderate change to the program, driven by pressure from political factors. Opposition to NCLB continued through the end of President Bush’s administration and into the administration of President Obama. However, Obama was initially focused on the financial crisis and initially attempted to address these concerns through financial stimulus, rather than reform to NCLB. As part of his stimulus package, President Obama created the Race to the Top program, which offered grants and funding similar to many aspects of Title I. The political expedience of needed stimulus helped this moderate supplemental funding to temporarily address concerns raised by the interests. This influx of funding, combined with a nationwide focus on the financial crises and healthcare reform, temporarily took NCLB off of the national stage. However, because the program itself had not changed, problematic processes continued, and negative consequences continued to build, resulting in continued pressure to grant states flexibility from NCLB requirements. President Obama moved to further address these concerns by granting states waivers from the requirements of NCLB. In a speech announcing the first waivers, he said, “The goals behind No Child Left Behind were admirable, and President Bush deserves credit for that. Higher standards are the right goal. Accountability is the right goal. Closing the achievement gap is the right goal. And we’ve got to stay focused on those goals” (Obama 2011). In this vein, he required states accepting waivers to develop new state goals aligned with the requirements of NCLB. A key element of this NCLB flexibility is that NCLB metrics were limited to reading, math, and graduation metrics, while flexibility waivers allow these metrics to be supplemented by multiple measures, including science, social studies, writing, ACT/SAT scores, AP courses, IB courses, dual enrollment, industry certification, college enrollment, remediation rates, and dropout rates (Vinovskis 2015b). Each of these elements has been raised by interests groups who have provided information related to potential negative consequences associated with overlooking them. As
7
EDUCATION FOR THE DISADVANTAGED
175
such, these waivers represent a mixture of political factors and policy knowledge, and a moderate change to NCLB. ESEA was scheduled to be reauthorized in 2007 but was extended during the 2008 election cycle. Following the election of President Obama, ESEA was extended for several more years without reauthorization due to competing priorities, controversy about aspects of the law, and because institutional factors, interests, and ideological tensions have resulted in too much disagreement to overcome (Vinovskis 2015a). Congress held hearings in 2013, 2014, and 2015 to reauthorize ESEA, but had not reauthorized the act as of its fiftieth anniversary.3 The three decisions that were made during ESEA’s fourth era reflected a retreat from the expanded federal role adopted by NCLB. Political interests and policy knowledge combined throughout this period to raise challenges to NCLB requirements, implementation, and outcomes, and to challenge the major expansion of the federal role (Gamson, McDermott, and Reed 2015). As shown in Fig. 7.4, two of the three decisions during this period were more than incremental, but less than comprehensive, and neither fit into the dominant decision making models. The other decision, however, more accurately reflected the expectations of the political-incremental model.
Summary of the Case This chapter has presented the legislative history of Title I from its passage in 1965 through its 50-year anniversary. It has applied Weiss’ information utilization framework and described the relative importance of the factors of ideology, interests, institutions, and information throughout the four eras of ESEA. The evidence presented throughout the chapter has made it clear that political factors including institutions and major interest groups have had the most direct impact on Title I decisions, which alternately supported different ideologies. The decisions throughout the last fifty years, however, varied in how data, information, and evidence were used and whether they impacted the final outcome. For example, in the case of A Nation at Risk, the report was initially cited inaccurately and used to support the president’s existing
3 The Act would be authorized at the end of 2015. This and other developments outside the case study period are discussed further in Chapter 9.
S. PUTANSU
Mixed
Punctuated Equilibrium Models
2009 Race to the Top
Incremental
Comprehensive
176
PoliticalIncremental
Rational Comprehensive
2012 Waivers
Bounded Rationality 2005 Rules Change
Political
Mixed
Rational
Fig. 7.4 ESEA decisions during the era of returning state flexibility, 2002–2015
opinions. Eventually, however, it became the foundation for all arguments about the need for broader federal involvement. Policy knowledge, when cited, was generally used to establish a problem or to support preexisting opinions about preferred solutions, but rarely directly aligned with the substance of major decisions. For example, A Nation at Risk was used to establish a problem and define technical scope, and it was repeatedly cited by parties on both sides of important debates, but it did not provide information on solutions to those problems. Alternately, evidence from Texas school reforms was used to support President Bush’s preexisting preference for the reforms outlined in NCLB, and studies of NCLB’s impacts were introduced by oppositional interest groups, but related decisions were strongly impacted by these political factors.
7
EDUCATION FOR THE DISADVANTAGED
177
Incremental
Mixed
Comprehensive
However, this chapter has also shown that policy knowledge has changed the nature of the problem being considered, impacting the strategies employed by ideological proponents and major interests. For example, the evaluations conducted by Paul Hill at NIE did not specifically identify or prioritize problems and solutions, but rather outlined the variety of outcomes impacted by Title I funds. Similarly, A Nation at Risk informed debates and decisions from 1992 to 2001, but took until 2002 to drive a more comprehensive change to Title I. This chapter has also shown that political factors were important as several points throughout ESEA’s history, even dominating over clear conflicting information at the passage of NCLB, when later used to show that the focus on testing had several negative unintended consequences. Figure 7.5 shows all 17 decisions from this case based on where they fall on the political-rational and incremental-comprehensive continua of
Punctuated Equilibrium Models
3 Decisions
Rational Comprehensive
2 Decisions
3 Decisions
PoliticalIncremental
1 Decision
3 Decisions
Bounded Rationality
Mixed
Rational
5 Decisions
Political
Fig. 7.5 Summary of major ESEA decisions, 1965–2015
178
S. PUTANSU
the balanced decision making framework. Dominant models suggest that decisions will be either political or rational, and 13 of 17 decisions in this case are consistent with that expectation. Further, 5 of those 13 fail to meet the second expectation of the models, of either an incremental or comprehensive change. As such, only 8 of the 17 decisions are fully explained by dominant models along both continua, while 9 fall outside the expectations of these models. In summary, the evidence presented in this chapter shows that, from the initial passage of Title I in 1965 until its fiftieth anniversary, decision processes about the program often utilized and relied on pertinent information, but generally in the framing of debates and winning political advantages, rather than leading decisions. Instead, these decisions were driven primarily by political factors, as predicted by Wilson’s typology of bureaucratic organizations (1989). However, this chapter also shows that, in a few cases, policy knowledge resulted in distinct and long-lasting changes to design approaches and desired goals of the program, which ultimately reframed the alignment of these political factors. Five of the seventeen decisions in this case provide support for a political-incremental model of decision making, and two others reflect the expectation of punctuated equilibrium models. However, the remaining ten are only fully understood through a balanced consideration of political factors and policy knowledge.
References Bailey, S. K., & Mosher, E. K. (1968). ESEA: The Office of Education administers a law. Syracuse: Syracuse University Press. Calzini, C., & Showalter, P. (2009). The increase of federal influence on educational policy in the United States: A pathway to national standards and testing. Journal of Cross-Disciplinary Perspectives in Education, 2(1), 1–13. Casalaspi, D. (2017). The making of a “legislative miracle”: The Elementary and Secondary Education Act of 1965. History of Education Quarterly, 57 (2), 247–277. Conant, J. (1963). The education of American teachers. New York: McGraw-Hill. Evans, R. (2019). Fear and schooling: Understanding the troubled history of progressive education. New York: Routledge. Gamson, D. A., McDermott, K. A., & Reed, D. S. (2015). The Elementary and Secondary Education Act at fifty: Aspirations, effects, and limitations. RSF: The Russell Sage Foundation Journal of the Social Sciences, 1(3), 1–29.
7
EDUCATION FOR THE DISADVANTAGED
179
General Accounting Office (GAO). (1977). Problems and needed improvements in evaluating office of education programs, HRD-76-165. Washington, DC: GAO. Glass, G. V., & Smith, M. L. (1977). “Pull out” in compensatory education. Washington, DC: Department of Health, Education and Welfare. Graham, H. D. (2011). The uncertain triumph: Federal education policy in the Kennedy and Johnson years. Chapel Hill: University of North Carolina Press Books. Hirschland, M. J., & Steinmo, S. (2003). Correcting the record: Understanding the history of federal intervention and failure in securing US educational reform. Educational Policy, 17 (3), 343–364. Hmelo-Silver, C. E. (2004). Problem-based learning: What and how do students learn? Educational Psychology Review, 16(3), 235–266. Jenkins, C. J., & Hill, M. H. (2011). Role of federal government in public education: Historical perspectives. Washington, DC: League of Women Voters, 110. Jones, L. V. (2003). National assessment in the United States: The evolution of a Nation’s report card. International handbook of educational evaluation (pp. 883–904). Dordrecht: Springer. Kaestle, C. (2016). Federalism and inequality in education: What can history tell us? The dynamics of opportunity in America (pp. 35–96). Cham: Springer. Keppel, F. (1966). The necessary revolution in American education. New York: Harper & Row. Kimbrough, J., & Hill, P. T. (1983). Problems of implementing multiple categorical education programs. Santa Monica, CA: Rand Corporation. Lowi, T. J. (1972). Four systems of policy, politics, and choice. Public Administration Review, 32(4), 298–310. Manna, P. (2006). Control, persuasion, and educational accountability: Implementing the no child left behind act. Educational Policy, 20(3), 471–494. McDermott, K. A. (2011). High-stakes reform: The politics of educational accountability. Washington, DC: Georgetown University Press. McGuinn, P. (2005). The national schoolmarm: No Child Left Behind and the new educational federalism. Publius: The Journal of Federalism, 35(1), 41–68. McGuinn, P. (2015). Schooling the state: ESEA and the evolution of the US Department of Education. RSF: The Russell Sage Foundation Journal of the Social Sciences, 1(3), 77–94. McGuinn, P. J. (2006). No Child Left Behind and the transformation of federal education policy, 1965–2005. Lawrence: University Press of Kansas. Meier, D. (2004). Many children left behind: How the No Child Left Behind Act is damaging our children and our schools. Boston: Beacon Press. National Commission on Excellence in Education. (1983). A Nation at Risk: The Imperative for Educational Reform—A Report to the Nation and the Secretary
180
S. PUTANSU
of Education, United States Department of Education. Washington, DC: The Commission. National Research Council, Wise, L. L., Noeth, R. J., & Koenig, J. A. (1999). Evaluation of the voluntary national tests, year 2. Washington, DC: National Academy Press. Obama, B. (2011, September 23). Remarks by the president on No Child Left Behind flexibility. Washington, DC: White House East Room. Radin, B. A., & Hawley, W. D. (1989). The politics of reorganization. New York: Pergamon Press. Rivera, C., & Stansfield, C. W. (2004). The effect of linguistic simplification of science test items on score comparability. Educational Assessment, 9(3–4), 79–105. Romano, C. (1999). Title I: Tools for ensuring quality educational opportunities. Washington, DC: The Federation for Children with Special Needs. Schwartz, R., Robinson, M., Kirst, M., & Kirp, D. (2000). Goals 2000 and the standards movement. Brookings Papers on Education Policy, 3, 173–214. Stein, S. J. (2004). The culture of education policy. New York: Teachers College Press. Vinovskis, M. A. (2015a). Using knowledge of the past to improve education today: US education history and policy-making. Paedagogica Historica, 51(1– 2), 30–44. Vinovskis, M. A. (2015b). From A Nation at Risk to No Child Left Behind: National education goals and the creation of federal education policy. New York: Teachers College Press. Wilson, J. Q. (1989). Bureaucracy: What government agencies do and why they do it (2nd ed.). New York: Basic Books.
PART III
Conclusions and Implications for Evidence-Based Governance
CHAPTER 8
Contending with the Evidence-Based Proverb
The first section of this book reviewed the value of policies knowledge, expectations for its use in decisions, and limitations of evidence-based reforms. Chapter 1 explained that the combination of narrowly defined policy knowledge with broad promises of panacea for decision making has caused such efforts to suffer critiques about failure to deliver on these promises. Chapter 2 expanded on the limitations of narrowly tailored policy knowledge to explain how well-known limitations of evidence, and the need to consider the quality and purpose of information, are often overlooked by advocates for evidence-based reform. Chapter 3 further examined the limitations of models of decision making, which tend to assume that decision making is either political or rational, with limited ability to consider joint roles of each. Chapter 4 demonstrated how over-commitment to a single source of policy knowledge, problematic expectations for whether and how such knowledge would be used to support decisions, and failure to consider political factors has resulted in the iterative failure of similar reforms for more than fifty years. Taken together, these chapters demonstrate that existing models of decision making behavior are subject to a range of critiques, and that they must be integrated to better reflect the reality of decision making. The critiques raised throughout the first section are not new, nor are they unique to studies of policy decisions. As noted throughout the first three chapters, many scholars and practitioners have studied, discussed, and explained the limitations of various forms of evidence, assumptions of
© The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_8
183
184
S. PUTANSU
decision making, and limitations in our ability to integrate policy knowledge and decisions (Cairney 2016; Radin 2006; Stone 1997; others). However, these critiques have, themselves, often been embedded within the proverb of evidence-based policy making, either contending that policy knowledge is ineffective in overcoming political factors, inadequate to support specific decisions, or itself only a reflection of political preferences. These critiques are often embedded within narrowly scoped reviews of data, information, and evidence. In many cases, the unit of analysis is either an individual source of policy knowledge or an individual decision. In the former case, they seek to understand why the specific policy knowledge is or is not reflected in the relevant decision outcomes. In the latter, they seek to understand the processes and factors, including policy knowledge, that drive decision outcomes. While these approaches surface legitimate critiques of the general limitations of policy knowledge and constraints of the political environment, they do not offer comprehensive insights into how those limitations and constraints interact when multiple sources of evidence and iterative decisions are considered. To address that limitation, the first section introduced a framework for studying decisions that conceptualizes politics and policy knowledge as jointly important in decision making, rather than assuming one supplants the other. The second section applied this framework to the roots of federal education policy and to fifty years of major legislative decisions about the two primary federal efforts to support US education: student loans and Title I grants. This chapter reviews the broad historical evidence from Chapters 6 and 7 and compares it to four expectations developed in the first section of this book. The first comparison examines the evidence-based proverb, which presumes that policy knowledge can and should supplant politics in decision making. Fundamentally, this frames politics and policy knowledge as an either-or proposition, with decisions either reflecting rational or political processes. Common refrains calling for increasing policy knowledge, making evidence-based decisions, and supporting only “what works” from rational perspectives are met with a chorus of reminders that there are multiple, often-conflicting goals to consider, that policy knowledge is not a panacea, and that politics matters (Key 1940). The cases in this book have shown that, in reality, decisions usually reflect a combination of politics and policy knowledge (Head 2016).
8
CONTENDING WITH THE EVIDENCE-BASED PROVERB
185
The second comparison relates to the scope of decision making along an incremental-comprehensive continuum. Whether politically or rationally grounded, dominant models of decision making offer explanations for when decisions will be more or less comprehensive. Under the political-incremental models, the status quo is seen as a mutually agreedto compromise among political factors, and decisions are expected to be incremental with few exceptions. However, punctuated equilibrium models, such as the garbage can (Cohen et al. 1972) and policy stream (Kingdon 2003) models, allow for exogenous shocks to upset the equilibrium and lead to more comprehensive changes. In rational-comprehensive models, these comprehensive changes occur when a full consideration of evidence suggests such a change is needed. However, bounded rationality presumes that understanding of policy knowledge is rooted in hardto-change heuristics, which limit the likelihood of major changes and tend to result in incremental changes. The cases examined in this book suggest that most decisions have both incremental and comprehensive components. This suggests that both political-incremental and boundedly rational models of decision making overstate policy stability. Further, neither major exogenous shocks nor comprehensive consideration of policy knowledge are required for more comprehensive changes to occur. The third comparison contrasts the two cases and examines whether program design can help to explain whether and how policy knowledge is balanced against political factors in decision making. It tests the hypothesis that production agencies will be characterized by more frequent inclusion and impact of policy knowledge in decision making processes, relative to coping agencies. The case of student loans shows that policy knowledge had a substantial role in decisions, potentially owing to the broader agreement on desired program outcomes and design. This tendency toward a strong role for policy knowledge was only challenged when decisions turned toward broad design changes and resurfaced in discussions about program goals. In contrast, the case of Title I demonstrated a more politically dependent use of policy knowledge, with exceptions only occurring for policy knowledge that reframed discussions of program goals and design. Finally, this chapter concludes by considering evidence for the cases against expectations for the quality of policy knowledge. As shown in Chapter 1 and elaborated in Chapter 4, modern evidence-based reforms
186
S. PUTANSU
have pressed for a hierarchy of evidence consistent with the medical evidence, which places randomized control trials as the “best” source of evidence. This is consistent with the performance movement, which similarly promoted performance information as the primary or best driver of political decision making. However, as elaborated in Chapter 2, hierarchies that are dependent on the method of information, and not on the purpose of that information, fail to meet the practical needs of decision makers. While randomized controlled trials and other behavioral and experimental methods provide strong evidence of causal relationships, they often lack the contextual sophistication, lived experience, and other data, information, and evidence that can improve the quality of decisions. Evidence from this book suggests that relevant policy knowledge comes from a wide variety of sources and methods, with a range of coverage and rigor. This book examined twenty one major legislative decisions that occurred during five broad eras for student loans, and seventeen decisions across four eras for Title I. Table 8.1 provides a timeline for these eras from 1965 through 2015, along with the partisan control of the presidency, House, and Senate during each. The remainder of this chapter examines and compares characteristics of these decisions in order to better understand how ideology, interests, institutions, and policy knowledge interact to influence governance decisions.
The Political-Rational Continuum The foundation of many evidence-based reforms is a proverb that policy knowledge can and should supplant politics in decision making. As made clear from the first section of this book, this has led to a schism in research on policy making, with politically driven theories placed in contrast to rational models. From the perspective of political models, policy knowledge is an appendage of political factors, used symbolically in support of preexisting beliefs, but not directly responsible for decisions (Krehbiel 1991). Rational models, on the other hand, presume that political differences are largely a reflection of inadequate or incomplete policy knowledge. As such, politics can be overcome by the right data, information, and evidence. When reviewing individual decisions, nearly half (15 of 38) support political models of decision making. Specifically, political factors explain all the major aspects of five student loan decisions and ten Title I decisions. For example, in the 1998 HEA reauthorization, decision makers
8
CONTENDING WITH THE EVIDENCE-BASED PROVERB
187
Table 8.1 Ideological changes and case study eras Dates
Ideological balance President
Senate
House
1963–1968*
Johnson (D)
1969–1976 1977–1980*
Nixon/Ford (R) Carter (D)
1981–1986
Reagan (R)
1987–1992 1993–1994*
Reagan/ Bush (R) Clinton (D)
1995–2000
Clinton (D)
2001–2006*
W. Bush (R)
2006–2007*
W. Bush (R)
2008–2010*
Obama (D)
2011–2014
Obama (D)
2015
Obama (D)
64–36 (D) 57–43 (D) 61–39 (D) 54–46 (R) 55–45 (D) 57–43 (D) 55–45 (R) 51–49 (R) 55–45 (R) 59–41 (D) 53–47 (D) 54–46 (R)
247–187 (D) 243–192 (D) 292–143 (D) 241–191 (D) 254–178 (D) 259–176 (D) 227–207 (R) 220–210 (R) 233–202 (R) 255–178 (D) 242–192 (R) 247–188 (R)
Student loan era
Title I era
Expanding equity
Program expansion
Reversion to states Accountability standards
Budget restriction
Muddling toward reform
Returning state flexibility
Federal expansion Muddling over federal role
*These time periods had single-party control of the presidency and both houses of Congress
explicitly acknowledged the protection of banking and university interests in their decision to avoid major changes to student loan programs, despite clear evidence of the improved cost-effectiveness of direct loans. Similarly, despite broad agreement on the need for federal education standards, the 1992 and 1999 reauthorizations of ESEA both avoided the issue due to ideological tensions among decision makers. These two cases offer much less support for rational models of decision making, which provide clear explanations for only one student loan decision. Specifically, the only decision about student loans that was clearly defined entirely by information was passage of the Federal Credit Reform Act (FCRA) of 1990. This act focused on the accounting methods used to record federal lending expenditures and was pursued outside of HEA
188
S. PUTANSU
authorization and appropriations decisions. For Title I, no decisions were clear deterministic responses to policy knowledge. With FCRA, the focus of the decision was on narrow technical implementation issues, and not on overall program goals or management. This example suggests that policy knowledge is unlikely to be the sole contributor to decisions, particularly in decisions about major program issues. As such, in the vast majority of cases, treating policy knowledge as a replacement for politics would result in failure to influence or understand decisions. While sixteen decisions seem to be explained fully by either political or rational inputs into decision making, the remaining twenty two require both politics and policy knowledge in order to understand the outcome. This includes fifteen student loan decisions and seven Title I decisions. For example, the passage of the Affordable Care Act, which transformed all student loans into direct loans, required a combination of ideological and institutional alignment represented by Democratic Party control of Congress and the presidency. It also benefitted from reduced influence of banking interests after the financial crisis of 2008 and compelling policy knowledge demonstrating the improved cost-effectiveness of direct loans. Similarly, while A Nation at Risk (National Commission on Excellence in Education 1983) eventually convinced both political parties about the need for accountability, political tensions prevented action on this need for several major decisions, and eventually led to the adoption of voluntary accountability standards in 1994. The full adoption of federal requirements did not take place until, following the attacks of September 11, which helped political divisions to set overcome tensions and reach a quick compromise. In all sixteen cases of mixed political-rational drivers, an assumption that decisions were purely political or rational would lead to an incomplete understanding. This demonstrates that a balanced framework for studying decision making, like the Four-I model used throughout this book (Weiss 1995), is critical to a complete understanding of the decision making process.
Incremental and Comprehensive Decisions In both political and rational models, major changes to policy are expected to be relatively rare. In political models, policy stability is expected as a reflection of an equilibrium among political factors who have reached a given compromise position. In rational models, large decisions are only expected when there are significant changes to policy knowledge.
8
CONTENDING WITH THE EVIDENCE-BASED PROVERB
189
In the case of rational-comprehensive decision models, new policy knowledge can displace what is known, resulting in quick policy changes (Hale 2011). In bounded rationality models, this new knowledge must contend with established heuristics and are more difficult to dislodge. Whether relying on a political or rational model, major changes are expected only when there is a major shock that causes some element of the system to shift. Of the thirty-eight decisions examined in this book, sixteen conform to expectations for incremental decision making. Seven of these fall squarely into political-incremental models, where the programs remained stable, reflecting continued alignment of political factors. For example, strong polarization and ideological conflict in Congress led to the delay of HEA reauthorization, with lawmakers instead passing extensions of the existing program. This is similar to cases in Title I, such as the 1965, 1966, and 1967 amendments, which marginally expanded Title I along the same partisan lines that supported original passage of the act. The other nine incremental decisions were influenced by political factors, in combination with policy knowledge. Beyond the sixteen clearly incremental decisions, ten more represent comprehensive changes associated with major shocks to the environment. For example, the passage of HEA and ESEA were both comprehensive expansions of federal power driven by landslide Democratic Party victories when Johnson was elected president. For HEA, this most closely reflected a garbage can model of decision making, where preexisting policy solutions were drawn together based on the balance of power among political interests. In the case of ESEA, the policy streams metaphor is more apt. Specifically, Commissioner Keppel can be viewed as a policy entrepreneur. Taking advantage of the window of opportunity, he worked tirelessly to bring together a coalition of interests and options to “join the policy streams” and push the bill through Congress. Of the other eight comprehensive decisions driven by exogenous shocks, one more was politically driven, one was driven by policy knowledge, and six were driven by a mix of both. For example, passage of MISAA in 1978 required a combination of policy knowledge demonstrating access issues faced by students from middle-income families with the backing of interest groups that supported the elimination of family income caps. Taken together, these sixteen incremental and ten comprehensive decisions represent over two-thirds of all the decisions examined in this book. The other twelve decisions, however, do not conform to the expectations
190
S. PUTANSU
of the dominant models. Specifically, these decisions represent moderate and major program changes without an associated exogenous shock. Five of these decisions are primarily political, and the other seven reflect both politics and policy knowledge. For example, HEA was reauthorized in 1981 through an omnibus appropriations act. There were no major new sources of policy knowledge discussed and no major shift in the balance of power among political factors. Similarly, by 2012, ideological and interest opposition had grown in a consistent resistance to NCLB, and partisan gridlock had prevented substantial changes to Title I grants (Meier 2004; Vinovskis 2015). However, by granting waivers to NCLB requirements, President Obama was able to enact a substantial change without a major exogenous shock to the environment. While these twelve decisions are not well-explained by dominant models of decision making, the balanced model used in this book provides additional insights. For example, the 1981 HEA reauthorization did not require a major shock to the policy environment, but is rather explained by a shift in the institutional approach to passing the bill. Specifically, by passing it through the appropriations process, the bill was placed in the control of different congressional committees, with a broader set of dedicated interests, and this political strategy was sufficient to enact a major change. Similarly, President Obama used executive authority to start granting NCLB waivers in 2012, effectively sidestepping the entrenched political interests. In both of these cases, a careful consideration of political factors reveals strategic behaviors that can lead to comprehensive changes without a significant external change in environment. This further demonstrates the value of this contextually sophisticated approach to understanding the decision making process.
The Implications of Program Goals and Design In Chapter 2, this book introduced Wilson’s (1989) typology of bureaucratic organizations. It predicted that student loans, which have easily measured goals and outcomes consistent with a production organization, would more readily utilize policy knowledge. In contrast, Title I, which has processes and outcomes that are difficult to measure, consistent with a coping organization, would have a higher incidence of political decision making. The decisions examined in this book generally support these hypotheses, but the relationship is not as strong as anticipated. As described above, both programs had very few decisions driven entirely by
8
CONTENDING WITH THE EVIDENCE-BASED PROVERB
191
policy knowledge, so this section compares purely political decisions to those with at least some significant use of policy knowledge. For the twenty one student loan decisions made between 1965 and 2015, five were driven primarily by political factors, while the other sixteen (roughly 75%) were significantly influenced by policy knowledge. During the same period, Title I underwent seventeen major decisions, ten of which were primarily political, while the other five (roughly 30%) were significantly influenced by policy knowledge. This provides some support for the hypothesis that easily measured processes and goals will be associated with higher rates of information use. However, these results warrant further review. Student Loan programs were a clear example of a production program when they were created in 1965. Language in the statute makes clear that enrollment in higher education was a shared, clearly defined goal, and that the process of distributing loan money was agreed to be a sufficient process for attaining that goal. However, the continuous influence of ideology, interests, and institutions demonstrates that these clear elements of program design were actually a reflection of a subjective political agreement. As a result, shifts in the political environment, which enabled other political viewpoints to gain a foothold, began to shift elements of the program (Ginsburg et al. 1997). Further, the use of direct loans provides a strong illustrative example of the continued importance of political factors. Specifically banking interests and a free market ideology combined to resist direct loans, despite clear evidence that they would be more costeffective than guaranteed loans. The conversion of student loans into direct loans in 2010 also reflected a larger shift in program design, which had become a part of the political dialog in the 1990s, has grown in use ever since. For example, questions of student loan defaults and college accountability have surfaced, creating questions about whether enrollment is a sufficient and appropriate goal for the program. During this time, various interests have suggested that graduation, employment, and income are outcomes that should also be considered. The Federal Student Aid office has resisted formal adoption of the measures, arguing that the act of providing loans doesn’t have a direct effect on these outcomes. In effect, this potentially shifts student loans from a production to a craft organization and could heighten the importance of political factors. Immediately after the passage of ESEA, Commissioner Keppel authored a book in which he, in no uncertain terms, explained that there
192
S. PUTANSU
was no evidence on what outcomes Title I was expected to improve, or what processes could be followed to achieve them. In other words, the program provided an archetypal example of a coping organization. The passage of ESEA and several immediately following amendments each required states and the federal government to produce additional policy knowledge to clarify these purposes and processes. Lack of data, information, and evidence continued to be a problem, leading many legislators to suggest eliminating the program altogether. This changed in the 1970 and 1974 reauthorizations, when the NIE evaluations of Title I provided in-depth information on various goals and related processes that could be supported by Title I funds. While the information did not solidify any particular target outcome or process, it did enable further dialog to drive ensuing Title I decisions. Similarly, the coping nature of the program was relaxed as the findings of A Nation at Risk took root. Specifically, as agreement coalesced around using standardized tests as a relevant outcome, the IASA and NCLB reauthorizations of ESEA reflected more of a craft organization, with more influence of policy knowledge related to these scores, which persisted through the end of the case study in this book. However, in late 2015, just beyond the scope of this review, Congress passed an ESEA reauthorization titled the Every Student Succeeds Act (ESSA), which made NCLB waivers permanent, relaxed the focus on test scores, and reverted Title I to a more coping-type program. Although this comparison of cases does not find a deterministic link between program design and the relative importance of policy knowledge in decision making, it does find support that such a relationship exists. The findings suggest that political agreement on desired goals and processes can enable a clear determination of relevant policy knowledge, which can influence routine decisions. It also shows that this broad agreement may change over time, resulting in shifts in the decision making environment. In contrast, when political tensions result in ambiguous goals and processes, these cases suggest that policy knowledge can be used to reframe debates and expand the policy dialog (Moynihan 2008; Schlaufer et al. 2018). This reframing may change the alignment of ideologies, interests, and institutions, leading to changes in program design, which may subsequently impact whether and how additional policy knowledge is used in decision making. These findings demonstrate that a balanced model of decision making must be considered dynamically,
8
CONTENDING WITH THE EVIDENCE-BASED PROVERB
193
taking into account historical alignment of politics and policy knowledge, as well as shifts that have occurred among driving factors.
Quality Versus Purpose: Which Policy Knowledge Gets Used? Of twenty three decisions where policy knowledge played a significant role, only two relied on data, information, and evidence typically associated with the top of “hierarchies of evidence” proposed by reformers (Lynn 1999). Specifically, detailed cost comparisons of direct versus guaranteed loans represented broad coverage of student loans estimated using highly rigorous methods. This information was ultimately influential in the decision to switch to direct loans, but took over 15 years between initial publication and the policy change. Similarly, the NIE evaluations under Paul Hill, which enabled decision makers to have a robust dialog about Title I goals, also had broad coverage and high rigor. These evaluations provided a framework for policy discussions, but ultimately did not lead to a specific decision outcome. When decision makers relied on policy knowledge for many student loan decisions, they were typically focused on simple cost projections, narrow eligibility claims, and high level information on other outcomes, such as default rates. In most of these instances, this information was combined with ideological support or communication from related interests, who framed the results in terms of political interests. Similarly, for the Title I decisions where policy knowledge played a significant role, the focus was on simple comparisons of achievement, state spending totals, or other narrow outcome measures highlighted by interest groups. An interesting result of this analysis is that three of the most influential sources of policy knowledge all resulted from federal evidence reforms, consistent with Meltsner’s expectations for policy analysts within the bureaucracy (1976). Specifically, cost comparisons for direct loans were developed by the Department of Education and confirmed by the Government Accountability Office (1977), while the NIE evaluations were directed by Congress in an ESEA amendment, and A Nation at Risk was developed by a Commission created by the Secretary of Education. This reaffirms that well designed evidence-based reforms have the potential to positively influence decision making. In contrast, only one of these decisions used policy knowledge that came from a performance management
194
S. PUTANSU
reform. This suggests that many of the criticisms about these sources usefulness for legislative decisions continue to be warranted, and that this policy knowledge may be better suited to internal agency use. The cases in this book, and the analysis in this chapter, have demonstrated the practical and realistic limitations of the evidence-based proverb and dominant models of decision making. The results show that program goals and design can shape the interaction of politics and policy knowledge in decision making, and that simple presumptions about the quality of policy knowledge fail to deliver on promises to remove politics from decisions. These findings suggest that a balanced and contextually sophisticated approach toward producing, using, and studying the role of policy knowledge in decision making will be critical to more successful reform efforts. The following chapter considers the implications of this evidence and how these lessons can be incorporated into future research, practice, and reform of evidence-based governance.
References Cairney, P. (2016). The politics of evidence-based policy making. Springer. Cohen, M. D., March, J. G., & Olsen, J. P. (1972). A garbage can model of organizational choice. Administrative Science Quarterly, 17 (1), 1–25. General Accounting Office (GAO). (1977). Problems and needed improvements in evaluating office of education programs, HRD-76-165. Washington, DC: GAO. Ginsburg, A., Rhett, N., & Cavataio, M. (1997). Strategic planning and performance management: The US department of education’s experience. US Department of Education. Hale, K. (2011). How information matters: Networks and public policy innovation. Georgetown University Press Head, B. W. (2016). Toward more “evidence-informed” policy making? Public Administration Review, 76(3), 472–484. Key, V. O. (1940). The lack of a budgetary theory. American Political Science Review, 34(6), 1137–1144. Kingdon, J. (2003). Agendas, alternatives and public policy. Boston: Longman. Krehbiel, K. (1991). Information and legislative organisation. University of Michigan Press. Lynn, L. E., Jr. (1999). A place at the table: Policy analysis, its postpositive critics, and future of practice. Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(3), 411–425.
8
CONTENDING WITH THE EVIDENCE-BASED PROVERB
195
Meier, D. (2004). Many children left behind: How the no child left behind act is damaging our children and our schools. Beacon Press. Meltsner, A. J. (1976). Policy analysts in the bureaucracy. University of California Press. Moynihan, D. P. (2008). The dynamics of performance management: Constructing information and reform. Georgetown University Press. National Commission on Excellence in Education. (1983). A nation at risk: The imperative for educational reform: A report to the nation and the secretary of education, United States Department of Education. Washington, DC: The Commission. Radin, B. A. (2006). Challenging the performance movement. Georgetown University Press. Schlaufer, C., Stucki, I., & Sager, F. (2018). The political use of evidence and its contribution to democratic discourse. Public Administration Review, 78(4), 645–649. Stone, D. A. (1997). Policy paradox: The art of political decision making (vol. 13). W. W. Norton. Vinovskis, M. (2015). From a nation at risk to no child left behind: National education goals and the creation of federal education policy. Teachers College Press. Weiss, C. H. (1995). The four I’s of school reform: How interests, ideology, information, and institutions affect teachers and principals. Harvard Educational Review, 65(4), 571–593. Wilson, J. Q. (1989). Bureaucracy: What government agencies do and why they do it (2nd ed.). Basic Books.
CHAPTER 9
Moving Beyond the Evidence-Based Proverb
The comparative case studies in this book provide support for four key insights about the roles of data, information, and evidence in decision making. First, the cases demonstrate that these forms of policy knowledge are rarely a clear and deterministic replacement for politics, in clear opposition to the evidence-based proverb. Second, they show that dominant models of decision making, which tend to presume incremental decision making except in cases where there have been external shocks to the environment, fail to explain many decisions. Specifically, tension, conflict, and compromise among the factors of ideology, interests, institutions, and policy knowledge explain some comprehensive decisions that occurred without significant external shocks. Third, the cases show that alignment among politics and policy knowledge may translate to clear and agreed upon goals and processes, which shape program design (Wilson 1989). In turn, program design is associated with the relative influence of politics and policy knowledge, with increased use of policy knowledge when goals and processes are clear. Finally, for these two cases, decision makers did not rely on “gold standard” types of evidence that are typically praised as being at the top of “evidence hierarchies.” Instead, these decision makers relied on policy knowledge whose purpose, coverage, and rigor were aligned with their preferences and needs. Taken together, these findings suggest that the magnitude and complexity of the federal government do not lend itself to one-size-fits-all solutions to policy problems (Daviter 2019), to single types of evidence, or to apolitical decisions. This reinforces Key’s (1940) observation that © The Author(s) 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4_9
197
198
S. PUTANSU
politics is always required when it comes to deciding on and prioritizing goals. Decisions are not driven by a permanent definition of static goals that are clearly prioritized, but rather by a conflictual process within and among the factors of ideology, interests, institutions, and policy knowledge. This leads to continuously readjusting goals, broken and reformed alliances, changes to program designs, and policy actions that range from muddling through to sweeping evidence-based changes (Lindblom 1959). This continuous shifting among factors demonstrates the value of considering all Four-I’s to better influence and understand governance decisions. These findings do not parallel post-positivist critiques of evidencebased policy that conflate the continued importance of politics with an outright dismissal of the value of policy knowledge (Lynn 1999; Rosenbloom 1993). In fact, they highlight that appropriate policy knowledge often has significant impacts on legislative decision making that drive both incremental and comprehensive policy changes. As such, this chapter concludes the text by considering what can be done to produce policy knowledge that is useful for decision making, and what can be done to understand and influence whether and how it is used. Before turning to these lessons learned, however, it considers areas where additional data, information, and evidence could improve upon what has been derived from these cases and introduces a research agenda for future work that can help address some of these limitations.
Unanswered Questions and a Research Agenda This book provides a breadth and depth of evidence on each of the two cases that far exceeds typical studies of policy knowledge and evidencebased policymaking. Review of two fifty-year histories enables the tracking of ideology, interest, and institutions, along with data, information, and evidence, in a way that has not been done before. However, these two cases represent a narrow coverage of federal government programs. The cases are based on comprehensive review of testimonial and documentary evidence surrounding the thirty-seven decisions analyzed in this case, but they do not provide an experimental setting to assess causal relationships, detailed estimation of correlations and relationships, or other standards of rigor that add value to our understanding. To further develop and test the Four-I model of decision making, future research is needed on a broader variety of policy areas, and the consideration, use, and impact of policy
9
MOVING BEYOND THE EVIDENCE-BASED PROVERB
199
knowledge must be examined through these other methodological lenses (Weiss 1979). Moreover, the findings shared in this book represent a small subset of the data, information, and evidence gathered while working on this project and present a necessarily simplified depiction of decision making for both cases. First, this study was scoped to examine major legislative decisions. While a large amount of research has focused on performance budgeting and other associations between policy knowledge and appropriations, this study sought to expand consideration to broader policy decisions. However, it overlooks several executive-led, implementation, and program-management decisions that continuously occurred throughout these programs’ histories (Vinovskis 2015). Related studies on performance management, evaluation use, and implementation suggest that these decisions are similarly impacted by both politics and policy knowledge (Lindblom and Cohen 1979; Riccucci 2010; Shulock 1999; and others). Further, in some of these arenas, types of evidence that were not observed being used in these cases, such as RCTs, may have greater usefulness. As such, systematic study to apply this framework would enable development of a more complete understanding of the relative influence and impact of different sources and types of policy knowledge on those decision contexts. An emerging body of work on administrative burdens also suggests that individuals who are impacted by federal programs may make decisions that are impacted by ideology, interests, institutions, and policy knowledge (Herd and Moynihan 2019). Specifically, complex administrative burdens may introduce learning costs that require individuals to seek out relevant policy knowledge in order to access services and benefits. For example, student loan programs require completion of the Federal Application for Student Aid, which has often been cited as a hurdle to applying for and receiving student loans. Aspects of federal policy that require engagement with federal entities introduce institutional constraints. For example, Title I policy originally required only children in disadvantaged groups to receive Title I services, which created a need for “pull-outs” that were shown to have adverse effects on their education (Glass and Smith 1977). Removing this constraint by allowing school-wide services to schools that had more than 75% of students from disadvantaged groups eased this institutional constraint. Individual decisions in accessing benefits may be influenced by direction from a variety of interests, which may promote or dissuade take-up of federal programs. For example, banking
200
S. PUTANSU
interests, who no longer benefit from guaranteed student loans, may now attempt to provide private student loans, which can reduce the number of students who access federal direct student loans. Finally, personal ideological beliefs can influence whether an individual decides to participate in a given program. For example, parents who believe in religious instruction may send their children to private religious schools. As a result, these children may miss out on benefits of Title I funding or require additional burdens (such as busing to a public school), in order to experience these services. In each case, fully understanding the range of political factors and policy knowledge that can contribute to or reduce these administrative burdens may enhance the public’s ability to benefit from these federal programs. Finally, this study focused primarily on the decisions made about each of the two cases examined, given a particular alignment of ideology, interests, institutions, and policy knowledge. It did not seek to explain how policy knowledge is developed, what characteristics of policy knowledge tend to matter, or how the political factors impact the production of policy knowledge. A few decisions throughout the cases, such as the inclusion of requirements for evaluations of Title I, do indicate that political factors may seek out policy knowledge, which is then produced under direction. However, significant amounts of data, information, and evidence are also created voluntarily by agencies, interest groups, and academic researchers. A substantial body of literature on factors that influence publication in peer-reviewed journals, links between research interests and the topics they study, and other political rationales suggest that the Four-I’s contribute to which policy knowledge is produced and made accessible to decision makers (Cairney 2016; Hale 2011; Stone 1997; Weiss 1995; and others). Direct and systematic study of these impacts, in order to understand how they align with or depart from the Four-I’s in the broader decision making context, could further strengthen the production of policy knowledge that is both useful and used by decision makers. This may also bring additional insight into ways that misuse of policy knowledge can be avoided (Alkin and King 2017). While beyond the scope of this study, some of the events in these cases raised questions that warrant further attention. Among the most interesting of these relates to A Nation at Risk (National Commission on Excellence in Education 1983). Specifically, this study was controversial from the moment the commission was formed, and then the results were
9
MOVING BEYOND THE EVIDENCE-BASED PROVERB
201
initially ignored, mischaracterized, and failed to influence decision making processes. Over time, however, the results of the study took on a broad national narrative about US competitiveness in education, which ultimately reframed Title I debates, changed political alignments, and led to some of the most significant changes the program experience during its first fifty years. Studying this single source of policy knowledge, and how it diffused through the decision making process, would likely lend insight into what made it so powerful, which could translate to other studies. Moreover, comparative studies of policy knowledge that have had similar impacts in other contexts could help to reveal how policy knowledge can be targeted to broad changes in goal specification. Another interesting question relates to how decision makers deal with uncertainty and contradicting information. Throughout the cases, decision makers relied primarily on relatively narrow selections of policy knowledge, and the core findings of these studies were usually not challenged. This changed after the passage of NCLB, when various interests and researchers began to offer evidence of unintended consequences, problematic NCLB data, and alternative measures of success. In this case, Congress did not act on this contradictory information for several years. Instead, President Obama began to grant waivers in 2012, and Congress didn’t formalize those waivers until after the time period covered by these cases. Because this was the only occurrence of this type of policy knowledge “debate” observed across the cases, it cannot support broader conclusions about decision making. Future research, however, could examine whether differences in policy knowledge reflect the different political perspectives in the decision making environment, a known gap in knowledge, or intentional misinformation produced in conjunction with political factors. Further, these analyses can examine conditions under which decision makers choose to ignore, refute, or contradict particular policy knowledge, or the individuals that produce it (Esterling 2004; Perna et al. 2019). While these and many other questions would benefit from additional research, the results of these cases provide important lessons for producers of policy knowledge and evidence-based reformers.
Creating Policy Knowledge: Purpose and Quality The cases in this book offer no support for the proverb that policy knowledge can supplant politics in decision making. When political conflict and uncertainty is high, as it was for Title I, political factors dominated many
202
S. PUTANSU
decisions. When decision makers support conflicting goals, evidence of effectiveness in achieving one or the other cannot resolve this disagreement. In some cases, as with A Nation at Risk, information may lead decision makers to focus on alternative goals, where their agreement may be higher. However, as with NCLB, this reframing may only be temporary. In contrast, when political tensions are relatively low, allowing for shared goals to be articulated, policy knowledge can play a stronger role. However, even in this scenario, political factors remain important to decision making. If nothing else, this fundamental limitation in the ability of policy knowledge to be deterministic in a variety of decision contexts should convey a sense of humility in producers of policy knowledge and evidencebased reformers. Advocates for policy knowledge must abandon the idea that any single hierarchy of evidence can represent the best solution for all decisions and that data, information, and evidence can be a panacea for decision makers. The successful inclusion of policy knowledge in decisions can be achieved by shedding the false proverb of evidence-based policy, and focusing more on specific purposes, appropriate quality, and effective diffusion of policy knowledge (Berry and Berry 2018). The policy knowledge used throughout the cases in this book served three broad purposes. In the case, where decision makers were focused on an agreed goal, such as increasing ability to afford college, policy knowledge could be used for the purpose of leading decisions. Specifically, analysis about implementation strategies and eligibility could lead to direct and specific recommendations to enact change. In many cases, this type of policy knowledge was aligned with and diffused by either ideological, interest, or institutional processes. During levels of higher conflict, policy knowledge was utilized more for negotiating decisions. In these cases, some data, information, and evidence made similar recommendations for particular actions, reflecting a likely intention to lead decisions. However, such policy knowledge was often ignored or countered by other political factors and policy knowledge, with decisions ultimately being driven by the outcome of political processes, including compromise. In the case of the NIE evaluations of Title I, this was accomplished by studying a comprehensive selection of outcomes of importance, laying out the tensions among them, and allowing decision makers to consider these various outcomes in the political dialog (GAO 1977). Finally, policy knowledge sometimes was used in framing decisions. In the case of direct loans, policy knowledge on cost-effectiveness reframed the existing focus on college
9
MOVING BEYOND THE EVIDENCE-BASED PROVERB
203
enrollment. For Title I, A Nation at Risk introduced standardized tests as a central focus, quieting the political dialog around other education goals. By assessing the decision making environment prior to conducting research, producers of policy knowledge can attempt to discern whether the environment is suited toward purposes of leading, negotiating, or framing and can make coverage and scoping decisions to target those purposes. In order to enable this thoughtful consideration of purpose, reformers should avoid specifying the type of evidence to be produced in ways that are not context-dependent. If political factors are aligned in support of clear goals, the most influential policy knowledge may be leading decisions toward that goal or framing decisions to consider alternatives. If political tensions are higher, policy knowledge may lend its weight to one or more political factions, may offer broader tradeoffs, or may consider alternative framings. Rather than assuming that quality is subject to a universal hierarchy, producers of policy knowledge should tailor the coverage and rigor of their studies to the purpose, level of conflict, and specific research questions under consideration (Rivlin 1971; Schattschneider 1960). For example, as conflict over goals and processes increases, additional coverage and rigor may provide a stronger influence for negotiations. However, when goals are very clear and processes already well understood simpler approaches may be sufficient and appropriate. By taking a contextually sophisticated approach to considering the purpose and quality of data, information, and evidence production, evidencebased advocates will be better positioned to generate policy knowledge that is useful for decision makers (Shulock 1999). However, to ensure that this information is actually used, they must also work to ensure that this policy knowledge enters the policy dialog and is given appropriate consideration in the decision making process (Moynihan 2008).
Understanding, Influencing, and Improving Decisions Policy knowledge can enter the policy dialog in a variety of ways, and evidence-based advocates would benefit by pursuing diffusion strategies tailored to the purpose(s) they are hoping to support. For the purpose of leading when decision makers have shared goals, awareness of policy
204
S. PUTANSU
knowledge might be sufficient to encourage use. This might include providing the results of the studies to key decision makers. When conflict is high, use is more likely if targeted to specific users (Henry and Mark 2003). This can be done by working with decision makers’ constituents, relevant interest groups, and others who might help to impact the negotiation process. For instances in this book where policy knowledge served a framing purpose, the information decision went to a broader audience than decision makers. Many researchers are now joining groups, for example, that disseminate research through blogs, podcasts, newspaper articles, and other venues to reach a wider audience. Through these efforts, they may be better positioned to drive broader framing of policy debates in areas where goals and processes are not as clear or universally supported (Schlaufer et al. 2018). The need for humility, mentioned above, also comes into play when trying to bring policy knowledge into the decision making process. Researchers must be thoughtful and intentional about the research questions they attempt to answer, how they answer these questions, and what purposes their answers may support (Julnes 2007). They also have to be aware of how each of these decisions might be perceived by different decision makers who are driven by tensions within and among the Four-I’s. While no source of policy knowledge can provide definitive data, information, and evidence about every relevant outcome and their interrelationships, or a single utility-maximizing guide to decision making, they must not use scoping decisions as an excuse to overlook important ideological, interest, and institutional considerations. They must be active participants in the policy dialog, working to explain omissions, limitations, and caveats, while still highlighting the positive impact their policy knowledge could have (King and Alkin 2018). Evidence-based advocates must accept, as a matter of course, that the policy knowledge they produce will be considered, understood, and valued differently by a variety of decision makers (Hird 2005). By actively considering how ideology, interest, institutions, and other policy knowledge have shaped the decision making environment, they can better position themselves to reach and influence a wider range of decision makers. These efforts, if successful, will improve their ability to link policy knowledge to decision outcomes.
9
MOVING BEYOND THE EVIDENCE-BASED PROVERB
205
Looking Forward: Federal Education Policy and Evidence-Based Policy The cases in this book covered the first fifty years of both cases. As of October 2019, while this book is being written, four more years have passed. As mentioned above, the ESEA was once again reauthorized in 2015, as the Every Student Succeeds Act (ESSA). This reauthorization reflected the strong interests of parents, teachers, and states in resisting strict federal accountability standards, and granted significant flexibility back to state and local control. HEA still has not had a full reauthorization since 2008, although bills are currently being debated. During this period, the debate around student loans has become highly visible and has shifted from questions of cost and access, to a focus on the longerterm impacts of debt-burdened students, accountability, and other measures (Lowry 2009; Negri 2013). Meanwhile, the relaxed federal role has persisted since passage of ESSA, with little movement to reauthorize the act. For student loans, agreement on goals and program design has all but disappeared. The political landscape has reoriented to a much stronger focus on the federal role, student outcomes, and issues of better linking funding amounts to future earnings, while addressing excessive amounts of debt that have accrued over time. Because these goals are more difficult to measure and because they have not been a central focus of previous decisions, there may be an opportunity for policy knowledge to help frame the policy dialog and inform negotiations. The result of this next reauthorization, however, may depend on the impending presidential elections, which may serve to realign or reinforce existing political divisions. Policy knowledge is critical to understanding the effectiveness, efficiency, and equity of government actions (NAPA 2000). For the simplest decisions, it can confirm the existence of problems, identify opportunities for improvement, suggest potential alternative approaches and compare their likely results, assess the success of given interventions, and help promote accountability. Modern governance is rarely this simple. Since the end of the World War II, the role of government has expanded to contend with increasingly complex, difficult, and interrelated policy problems (Rittel and Webber 1973); it has used a broader set of policy tools to identify and address those problems; and it has increasingly relied on nonprofit and private workforces to implement those tools. These profound changes in the scope, purpose, and operation of government have led to
206
S. PUTANSU
increased need for policy knowledge, as reflected in iterative evidencebased reforms over the past fifty years, each aimed at increasing the production and use of policy knowledge to better understand the impacts of government action and to improve decision making (Alkin and King 2016; Feldman 1989). Support for these reforms crosses partisan boundaries, as evidenced in 2017 by the bipartisan Commission on EvidenceBased Policy’s unanimous adoption of 22 recommendations to improve policy knowledge in the federal government, and the 2019 passage of the Foundations of Evidence-Based Policy Act, which enshrined many of these recommendations into law. The Evidence Act includes several provisions that align well with a balanced model of decision making. It requires federal agencies to regularly assess the data, information, and evidence available to them, to understand where additional policy knowledge may be needed, and to develop learning agendas to prioritize these needs. The act suggests that agencies consider a portfolio of evidence, which includes different types and sources of policy knowledge. For example, agencies may combine performance information, program evaluations, and random control trials to develop a comprehensive understanding of how program activities relate to program goals. If agencies take advantage of these nuances to develop contextually sophisticated policy knowledge, they may make more consistent and demonstrable progress in supporting the use of policy knowledge, relative to past reforms. This study offers a contextually rich examination of decision making for two programs over an extended period of time. Highlighting both the stability and malleability of the Four-I’s, it shows that these factors are mutually interdependent and important for decision making. In addition, the depth of this study is unique to the literature on political decision making and offers actionable insights for researchers and reformers. Specifically, by considering a production and a coping program, this method offers insights into a range of political decisions and information uses that goes beyond typical examinations of single decisions or programs. By considering the complex and interactive nature of decisions, reformers and researchers should be able to improve whether and how policy knowledge is being utilized for political decisions (Hale 2011). Over time, this contextually sophisticated approach should enable reformers to abandon the evidence-based proverb in favor for a more thoughtful approach to
9
MOVING BEYOND THE EVIDENCE-BASED PROVERB
207
advocating for policy knowledge to support decisions. Open acknowledgment of the mutual importance of all Four-I’s can improve whether policy knowledge is produced, what purpose it serves, and whether and how decision makers can and will use it. Keeping in mind the need for humility, this book’s findings are not definitive and should serve as a reminder to be thoughtful, rather than as guidance for how to structure future reform efforts. Insights about the interactions among ideology, interests, institutions, and policy knowledge represent just a first step toward a broader conversation about political decision making. Systematic application of this framework in other contexts is a critical next step. Future research is needed to confirm these findings, expand to other program types and policy areas, and develop more parsimonious models of information production and use that can help gauge the impact of specific information efforts. Ultimately, these studies can confirm, expand, and modify the way practitioners and decision makers receive, interpret, and use policy knowledge. Clarification of these processes will enable scholars and reformers to improve the design and usefulness of the data, information, and evidence they produce and advocate for. Knowledge derived from these activities, in turn, may have a greater influence on decision outcomes. Interaction between politics and policy knowledge occurs continuously, as demonstrated throughout both cases in this book. Nearly every decision examined included some shift in the relative influence of ideology, interests, institutions, and policy knowledge. Generally, established program design mitigated the impact of these shifts and led to incremental changes in the respective programs (Termeer and Dewulf 2019). Greater impacts often required a large exogenous shock to the environment, but in many cases reflected shifting influence within the existing Four-I’s. Reforms that carefully consider how policy knowledge fits into this environment may therefore have a better chance of driving systematic change than dominant models of decision making have anticipated. Evidencebased advocates may be able to drive such changes through more casespecific decisions about the purpose, coverage, and scope of the policy knowledge they produce. Additionally, this approach may mitigate use of simplistic and problematic assumptions that all evidence will be a valid substitute for political factors. This book will hopefully serve as an additional bridge between scholars, practitioners, and decision makers who ascribe to politically and rationally based theories of decision making. The results of this book show not only
208
S. PUTANSU
that both camps have merit, but that they are both integral to a complete understanding of the interactive decision making process. However, the relative importance of these groups may depend, at least in part, on the clarity of the goals and processes of specific government programs being examined. Enlightenment, as Madison called it, may come in the form of leading, negotiating, or framing the policy dialog. Evidence-based governance, therefore, should be flexible enough to pursue any one of these three goals. New evidence-reforms, built on these principles, can cumulatively increase the capacity and ability to develop and leverage policy knowledge. Decision makers, in turn, may more effectively integrate data, information, and evidence with the other political factors in support of effective, efficient, and equitable governance.
References Alkin, M. C., & King, J. A. (2016). The historical development of evaluation use. American Journal of Evaluation, 37 (4), 568–579. Alkin, M. C., & King, J. A. (2017). Definitions of evaluation use and misuse, evaluation influence, and factors affecting use. American Journal of Evaluation, 38(3), 434–450. Berry, F. S., & Berry, W. D. (2018). Innovation and diffusion models in policy research. Theories of the policy process (pp. 263–308). New York: Routledge. Cairney, P. (2016). The politics of evidence-based policy making. Dordrecht, The Netherlands: Springer. Daviter, F. (2019). Policy analysis in the face of complexity: What kind of knowledge to tackle wicked problems? Public Policy and Administration, 34(1), 62–83. Esterling, K. M. (2004). The political economy of expertise: Information and efficiency in American national politics. Ann Arbor, MI: University of Michigan Press. Feldman, M. S. (1989). Order without design: Information production and policy making. Stanford, CA: Stanford University Press. General Accounting Office (GAO). (1977). Problems and needed improvements in evaluating office of education programs, HRD-76-165. Washington, DC: GAO. Glass, G. V., & Smith, M. L. (1977). “Pull out” in compensatory education. Boulder: University of Colorado. Hale, K. (2011). How information matters: Networks and public policy innovation. Washington, DC: Georgetown University Press. Henry, G. T., & Mark, M. M. (2003). Beyond use: Understanding evaluation’s influence on attitudes and actions. American Journal of Evaluation, 24(3), 293–314.
9
MOVING BEYOND THE EVIDENCE-BASED PROVERB
209
Herd, P., & Moynihan, D. P. (2019). Administrative burden: Policymaking by other means. New York: Russell Sage. Hird, J. A. (2005). Policy analysis for what? The effectiveness of nonpartisan policy research organizations. Policy Studies Journal, 33(1), 83–105. Julnes, G. (2007). Promoting evidence-informed governance: Lessons from evaluation. Public Performance & Management Review, 30(4), 550–573. Key, V. O. (1940). The lack of a budgetary theory. American Political Science Review, 34(6), 1137–1144. King, J. A., & Alkin, M. C. (2018). The centrality of use: Theories of evaluation use and influence and thoughts on the first 50 years of use research. American Journal of Evaluation, 40(3), 431–458. Lindblom, C. E. (1959). The science of “muddling through”. Public Administration Review, 19(1), 79–88. Lindblom, C. E., & Cohen, D. K. (1979). Usable knowledge: Social science and social problem solving (Vol. 21). New Haven: Yale University Press. Lowry, R. C. (2009). Reauthorization of the federal higher education act and accountability for student learning: The dog that didn’t bark. Publius: The Journal of Federalism, 39(3), 506–526. Lynn, L. E., Jr. (1999). A place at the table: Policy analysis, its postpositive critics, and future of practice. Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(3), 411–425. Moynihan, D. P. (2008). The dynamics of performance management: Constructing information and reform. Washington, DC: Georgetown University Press. National Academy of Public Administrators Social Equity Committee. (2000, November). Standing panel on social equity in governance: Issue paper and work plan. Washington, DC: NAPA. National Commission on Excellence in Education. (1983). A nation at risk: The imperative for educational reform—A report to the nation and the secretary of education, United States department of education. Washington, DC: The Commission. Negri, K. (2013). Mortgaging the American dream: The misplaced role of accreditation in the federal student loan system. Fordham Law Review, 82, 1905. Perna, L. W., Orosz, K., & Kent, D. C. (2019). The role and contribution of academic researchers in congressional hearings: A critical discourse analysis. American Educational Research Journal, 56(1), 111–145. Riccucci, N. M. (2010). Public administration: Traditions of inquiry and philosophies of knowledge. Washington, DC: Georgetown University Press. Rittel, H. W., & Webber, M. M. (1973). Dilemmas in a general theory of planning. Policy Sciences, 4(2), 155–169.
210
S. PUTANSU
Rivlin, A. (1971). Systematic thinking for social action. Washington, DC: The Brookings Institution. Rosenbloom, D. H. (1993). Have an administrative Rx? Don’t forget the politics! Public Administration Review, 53(6), 503–507. Schattschneider, E. E. (1960). Semi-sovereign people. New York: Holt, Rinehart and Winston. Schlaufer, C., Stucki, I., & Sager, F. (2018). The political use of evidence and its contribution to democratic discourse. Public Administration Review, 78(4), 645–649. Shulock, N. (1999). The paradox of policy analysis: If it is not used, why do we produce so much of it? Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(2), 226–244. Stone, D. A. (1997). Policy paradox: The art of political decision making (Vol. 13). New York, NY: W.W. Norton. Termeer, C. J., & Dewulf, A. (2019). A small wins framework to overcome the evaluation paradox of governing wicked problems. Policy and Society, 38(2), 298–314. Vinovskis, M. A. (2015). Using knowledge of the past to improve education today: US education history and policy-making. Paedagogica Historica, 51(1–2), 30–44. Weiss, C. H. (1979). The many meanings of research utilization. Public Administration Review, 39(5), 426–431. Weiss, C. H. (1995). The four I’s of school reform: How interests, ideology, information, and institutions affect teachers and principals. Harvard Educational Review, 65(4), 571–593. Wilson, J. Q. (1989). Bureaucracy: What government agencies do and why they do it (2nd ed.). New York: Basic Books.
Index
A Agency(ies), 10, 18, 33–35, 37, 56, 72–83, 85, 97, 104, 118, 121, 123–125, 131, 139, 142, 160, 168, 185, 194, 200, 206 Appropriation, 5, 6, 21, 37, 56, 57, 61, 63, 74–78, 80–83, 95, 98, 100, 115, 119, 121, 124, 129–131, 133, 135, 137, 143, 145, 149, 155, 158–160, 162, 163, 169, 188, 190, 199 Authorization, 18, 21, 24, 63, 85, 91–94, 99–102, 104–106, 109–111, 115, 117–119, 121– 127, 129–131, 133, 135–138, 140, 141, 144, 145, 149, 152, 154–161, 163–165, 168–171, 173, 175, 177, 178, 186–193, 201, 205, 206
B Budget. See Appropriation Bureaucratic type. See Program design
C Charity school. See School choice Civil rights, 33, 34, 101–107, 117, 149, 155, 157, 158, 166, 174 College. See Higher education institutions Commission on Evidence-based Policy, 85, 206 Common school. See Public school Communism, 100, 106, 149 Compulsory school. See Public school Congress, 9, 13, 35, 37, 56, 63, 73, 76–79, 81, 95, 97, 99, 100, 103, 106, 116, 118, 121, 123–126, 129, 130, 133–137, 139, 141, 143, 150, 154, 155, 157–159, 162–166, 168–171, 175, 186, 188, 189, 192, 193, 201 D Data. See Policy knowledge Decision making bounded rationality, 54, 55, 185 garbage can, 46, 122, 189
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 S. Putansu, Politics and Policy Knowledge in Federal Education, https://doi.org/10.1007/978-3-030-38395-4
211
212
INDEX
policy stream, 122 political-incremental, 23, 53, 56, 59, 158, 178, 189 rational-comprehensive, 23, 53, 54, 59 Democratic Party, 106, 121–123, 125, 129, 134, 143, 155–158, 164, 165, 169, 171, 188, 189 Department of Education (ED), 18, 19, 97, 133, 134, 139, 151, 158–164, 177, 192, 193, 202 Direct loans, 118, 125, 131, 133–135, 140, 141, 143, 187, 188, 191, 193, 202
E Economy, 12, 15, 36–38, 84, 99, 101, 105, 108, 120, 124, 128, 129, 133, 140, 143 Effectiveness, 4, 6, 8, 12–17, 20, 23, 24, 36–42, 44–47, 51, 52, 54–59, 61, 63, 65, 71, 76, 77, 81–83, 85, 86, 92, 96, 105–108, 110, 111, 116, 118–121, 123– 125, 127, 129, 130, 132–136, 140, 142, 143, 152, 153, 157, 159–161, 163–165, 167–171, 173, 174, 185, 188, 191–194, 200–202, 204, 205, 207 Efficiency, 8, 15, 37, 38, 46, 51, 52, 54, 72, 77, 120, 121, 191, 205 Elementary and Secondary Education Act (ESEA) Education Consolidation and Improvement Act (ECIA), 163 Every Student Succeeds Act (ESSA), 192, 205 Improving America’s Schools Act (IASA), 168, 169 No Child Left Behind (NCLB), 171, 173, 177
Equity, 15, 36, 37, 52, 54, 62, 94, 120, 121, 123, 124, 127–129, 143, 149, 152, 155, 156, 158, 161, 162, 167, 168, 171–173, 205 Evaluation. See Policy knowledge Evidence. See Policy knowledge Evidence-based, 5, 8, 10, 18, 19, 54, 74, 203, 204
F Federalism, 17, 41, 60, 62, 94, 96–99, 102, 103, 107, 109, 115, 116, 121, 122, 125, 150–153, 155, 158, 159, 162, 163, 165–169, 173, 174, 192, 205 Foundations of Evidence-Based Policymaking Act, 85, 206 Four I model, 24, 61, 198
G Government Accountability Office (GAO), 10, 13, 20, 38, 39, 80, 83, 131, 134, 145, 161, 193, 202 Government Performance and Results Act (GPRA), 78–82 Government Performance and Results Act Modernization Act (GPRAMA), 78, 81–83 Guaranteed Student Loans (GSL), 115, 117–119, 123–127, 129–131, 133, 139, 200
H Health, Education, and Welfare (HEW), 97, 122, 125 High school. See Public school Higher Education Act (HEA)
INDEX
College Cost Reduction and Access (CCRA), 137 Deficit Reduction Act (DRA), 137 Ensuring Continued Access to Student Loans Act (ECASLA), 140 Federal Credit Reform Act (FCRA), 187 Middle Income Student Assistance Act (MISAA), 127 Student Aid and Fiscal Responsibility Act (SAFRA), 141 Higher education institutions, 92, 99, 107–110, 115, 116, 120, 121, 123, 125, 126, 133, 136–138, 140, 142, 143, 145, 174, 187, 191, 202 Historically black colleges and universities. See Land grant
I Ideology. See Political factors Impact aid, 98, 104–106, 156 Information. See Policy knowledge Institutions. See Political factors Interests. See Political factors
L Land grant, 107, 116
M Management by Objective (MBO), 75, 76 Manager, 21, 36, 38, 45, 46, 74, 82
N National Defense Education Act (NDEA), 108, 117, 118, 123, 125, 133
213
National Defense Student Loans (NDSL), 108, 118 National Institute of Education (NIE). See Department of Education (ED) National Performance Review (NPR), 79, 80
O Office of Education (OE). See Department of Education (ED) Office of Management and Budget (OMB), 56, 73, 74, 76, 79–82
P Pell grant, 137, 140, 141 Performance budgeting. See Performance management Performance management, 4, 8, 10, 19, 56, 72–75, 79, 193, 199 Performance movement, 23, 71–74, 82, 186 Planning, Programming, Budgeting System (PPBS), 72, 74–76, 83 Policy analysis. See Policy knowledge Policy dialog, 44, 192, 203–205, 208 Policy knowledge, 3–8, 10–18, 23–26, 35, 36, 39, 42–47, 51, 54–63, 66, 67, 71, 72, 74–76, 78–80, 82, 83, 85, 86, 91, 92, 99, 106, 107, 109–111, 115, 116, 119–125, 127, 129, 131, 133, 135, 137–140, 142, 143, 145, 149, 150, 152–154, 156–161, 163, 165, 167, 169, 173–175, 177, 178, 184–192, 197–207 Political factors, 5–8, 13, 16, 17, 19, 20, 23–26, 36, 37, 39, 40, 42, 45–47, 52, 53, 55, 56, 59–67, 73, 76, 79, 82, 83, 86,
214
INDEX
91, 94, 97, 98, 100–106, 109– 111, 115–138, 140, 142–145, 149, 151–167, 169, 172–178, 183–193, 197–205, 207, 208 President Bush, George H.W., 73, 131, 134, 164, 165, 167, 187 Bush, George W., 79, 80, 137, 138, 170, 187 Carter, Jimmy, 76, 77, 161, 187 Clinton, Bill, 78, 79, 134, 135, 167–171, 187 Eisenhower, Dwight, 100 Ford, Gerald, 76, 157, 159, 162, 187 Hoover, Herbert, 10 Johnson, Lyndon B., 72, 74, 83, 106, 115, 117, 118, 122, 124, 125, 149, 151, 155, 156, 187, 189 Kennedy, John F., 74, 101, 102, 104, 151 Lincoln, Abraham, 97 Madison, James, 3, 5, 51 Nixon, Richard, 75, 125, 157, 162, 187 Obama, Barack, 82, 85, 139–141, 174, 175, 187, 190, 201 Reagan, Ronald, 77, 130, 162–164, 187 Truman, Harry, 107 Washington, George, 99 Wilson, Woodrow, 8, 14 Private school. See School choice Program Assessment Rating Tool (PART), 80, 81, 137 Program design, 5–7, 12, 16, 19, 22–25, 31, 34, 36, 37, 39, 42, 45–47, 51, 54, 55, 59, 63, 64, 71, 74, 78–81, 85, 91, 94, 99, 100, 106, 107, 110, 111, 119, 121, 123, 124, 129, 133, 140,
143, 145, 149, 152, 158, 160, 163, 167, 173, 174, 178, 184, 185, 188, 190–192, 194, 197, 198, 200–208 Public school, 93–98, 100–102, 108, 115, 150, 151, 155, 163, 200
R Random control trial. See Policy knowledge Reform, 4–14, 16, 17, 19–21, 23, 26, 31, 32, 35, 39, 47, 51, 52, 54–56, 64, 66, 71–75, 77, 79, 80, 82, 83, 85, 96, 100, 106, 132, 141, 163, 166–168, 174, 176, 183, 193, 194, 206, 207 Religious schools, 94, 95, 100–102, 104–106, 149–151, 154, 155, 162, 167, 200 Republican Party, 143, 166, 171 Research. See Policy knowledge
S School choice, 94, 151, 164, 167, 170 Segregation, 96, 102–104, 117, 156 Socialism. See Communism Student, 18, 19, 40, 62, 94–96, 98, 103, 105, 106, 108, 109, 117, 118, 120–126, 130, 131, 133, 136–140, 142, 143, 145, 151–153, 155, 157, 158, 160, 161, 163–165, 168, 169, 189, 199, 200, 205 Student loans, 18, 24, 25, 47, 86, 91, 108–111, 120–122, 126, 127, 130, 131, 134, 135, 140–142, 146, 184–186, 188, 190, 191, 193, 199, 200, 205
INDEX
T Teacher, 40, 62, 98, 102, 105, 106, 118, 138, 151, 153, 155, 159, 167–170, 173, 174, 205 Title I, 17, 18, 24, 25, 47, 61, 62, 86, 91, 95, 104, 106, 111, 116, 149–152, 155–165, 167–169, 174, 175, 177, 178, 184–186, 188–193, 199–203 Truman Commission Report, 92, 107
215
U University. See Higher education institutions V Vouchers. See School choice Z Zero-based Budgeting (ZBB), 76, 77