DONUT DEVOURERS, FISH FANATICS, POLITICIANS, AND EDUCATORS: FACES AND VOICES OF STATE EDUCATION POLICYMAKING By Rachel Sue White A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Education Policy – Doctor of Philosophy 2017 ABSTRACT DONUT DEVOURERS, FISH FANATICS, POLITICIANS, AND EDUCATORS: FACES AND VOICES OF STATE EDUCATION POLICYMAKING By Rachel Sue White The nature of who is involved in making education policy is rapidly changing. At the local level, an increasingly diverse body of policymakers is in place with the potential for improved representation for marginalized groups. On the other hand, state-level policymakers are being called upon to be more involved in making education policy, and state policymakers are indeed tightening the education policymaking reins. Yet, we know very little about who state education policymakers are and how they go about making education policy decisions. This dissertation draws upon theories of institutionalism, representation, and responsiveness to examine (a) who gains state education policymaking power, (b) whose voices are heard in the state education policymaking process, and (c) the ways in which individual characteristics and institutional structures mediate the voices and victors in the state education policymaking process. I utilize demographic data of state legislators serving on education committees and state board of education (SBE) members across all 50 states, as well as data collected in a national experiment-based survey of state education policymakers, and interviews with 44 state education policymakers located in six states. Although I find that state education policymakers are neither descriptively nor substantively representative, considerable differences exist with regard to who they turn to when making education policy decisions. These results provide important insight into the way citizens think about the avenues through which they could engage, interact, and share important policy-relevant information and experiences with state education policymakers. Copyright by RACHEL SUE WHITE 2017 For the late Sue Vyskocil, who, in the simplest of ways, taught me to be quick to listen, slow to judgment, rich in gratitude; and that a determined woman can do anything she sets her mind to. iv ACKNOWLEDGEMENTS In any foot race, there is a point where you have to mentally decide to either embrace the potential for pain in pursuit of a personal best; or, settle in and be satisfied with mediocrity— running the time that you’ve comfortably run before. Perhaps to my spouse’s chagrin, I typically prefer to run all foot races with a big smile on my face, simply enjoying that I have been given a gift that allows me to both think in solitude and share invaluable time with wonderful people. I never thought that the first race in which I would mentally decide to embrace the potential for pain would not be in a foot race but, rather, the pursuit of a doctoral degree – truly a mental, emotional, and physical marathon. Like many foot races, the pursuit of a doctoral degree should not be done in solitude. I am grateful to all those who were by my side. When running a race, I prefer to have someone at each mile telling me my most recent mile split and, when a split is slow, telling me to step it up. I like to see both familiar, smiling faces as well as strangers serving as impromptu cheerleaders holding encouraging signs that make me smile and forget about the pain for a few minutes. I need friends running the race next to me, experiencing the same ups-and-downs whom I know I can turn to, both when times get tough and in times of celebration. My needs were no different in pursuit of this degree and there are many people who guided me across the finish line for which I wish to thank. First, Rebecca Jacobsen: you were an incredible advisor, mentor, and voice of reason; both a persistent cheerleader and the one telling me to step it up when you knew I needed to be challenged; the person I knew I could trust to prepare me to run the race; and the person I knew I could rely on to discuss the ups-and-downs of life, both in and outside of academia. And, in more v ways than one, you taught me the importance of being tenacious. The research, mentorship, and scholarly opportunities that you provided were invaluable. Thank you. Madeline Mavrogordato: thank you for providing me with research and teaching opportunities beyond anything I imagined. You taught me to be resilient and never give up on my dreams. You believed in me every single step of my PhD journey and gave so graciously of your time. And, as a Wolverine at heart, you are the sole Spartan with whom I enjoy talking about basketball. It is because of you that I grew a small soft spot (albeit very, very, very small one) for Tom Izzo. As we both arrived at Michigan State University five years ago– you as a first year professor and me as a first year graduate student – your visible eagerness to immerse yourself in a new community helped me to do attempt to do the same. Josh Cowen: thank you for your invaluable support, advice, and mentorship—both in writing this dissertation and navigating life after a PhD. Your honesty, professionalism, and approach to research have and will continue to serve as an example of what I strive to be as scholar. Ken Frank and Sarah Reckhow: Thank you for your wonderful mentorship and for your willingness to both lend an ear and provide thoughtful insight into research methods, analyses, and implications. Dr. Michael Sedlak: Without the extensive support and the multitude of incredible opportunities you afforded me, none of this would have been possible. Thank you. Betty Malen: thank you for your honest wisdom and encouraging me to research what I’m passionate about. And, thank you for always providing the advice that I needed at the most opportune times – from the importance of staying true to myself in my career endeavors to vi simply demanding that I “tie a rope” (…around my legs while sitting in my office chair and just finish the darn dissertation!). Those of you with whom I ran the race: it was because of your persistence that I persisted. In particular, I am grateful to Andy Saultz, John Lane, Ben Creed, Alyssa Morley, David Reid, Jason Burns, Kim Jansen, Bryan Beverly, and Hugh Potter. The unending love, care, support, and memories that you’ve provided will never be forgotten. I am also grateful to many other scholars and friends who have shaped my thinking in multitude of ways and provided support throughout this journey, including BetsAnn Smith, Amita Chudgar, Ann Allen, Kristi Bowman, Eric Scorsone, Chloe Kannan, Tim Krupp, Brad Biladeau, Nick Kruse, and Tiffany Johnson. Lastly, I am completely indebted to my family. My parents, Richard and MaryJo White, believed in me from day one. My mother provided a listening ear on the telephone on my drives home from campus or airports. She rode with me on my dissertation journey (literally) and provided many laughs, smiles, and unforgettable memories. I am appreciative for her friendship and grateful that she raised me with humility. In high school, my mom would always come give me a hug at the end of every track race and say, “You may have won, but you’re not even out of breath. Are you sure you tried your hardest?” Well, mom, I have no doubts that, in this race, I tried my hardest. I only hope that I made you proud. My father provided unwavering support and love. His steadfast belief that I could and would succeed is why I did. I only hope that he is half as proud to call me his daughter as I am proud to call him my father. Dad, I definitely did not take a single step of this journey for granted and I hope you know that it is truly because of you that when I got “the chance to sit it out or dance,” I danced. Finally, Justin: what is there to say? No words can express how important it has been to have you by my side these past five years. vii Our hearts and our feet have indeed endured miles of trials and trials of miles. I am grateful for the sacrifices you have made, the encouragement you have provided, your ability to convince me to slow down and take time for us, and – most importantly – your unconditional love and care, even when it was most difficult. You have taught me that love truly is a verb. Oh, and to my trusty fury companions, Gus and Bear: your cuddles and unconditional love were critical. In the book The Giver, Lois Lowry wrote, …now he saw the familiar wide river beside the path differently. He saw all of the light and color and history it contained and carried in its slow-moving water; and he knew that there was an Elsewhere from which it came, and an Elsewhere to which it was going. I cannot thank each of you enough for helping me to see this path differently; for embracing me and the places from which I came; and, most importantly, for encouraging me to look forward to the places to which I am going. viii TABLE OF CONTENTS LIST OF TABLES………………………………………………………………………... xiv LIST OF FIGURES…………………………………………………………………......... xvii CHAPTER I Introduction………………………………………………………………... Overview of the study 1 5 CHAPTER II Education Governance: A Review of the Literature……….……………... Background: Governing America’s K-12 Public Schools………………………... State legislatures………..………...………………………......................... State boards of education………………………...……………………….. State Government Involvement in Educational Matters………………………...... Literature Review: Empirical Research on State Governance of Public Education How has state education policy governance been examined? .……….. State-led education policymaking………………………............... Federally-mandated, induced, and incentivized state education policymaking. ………………………...………………………...... Limitations of Existing Research on State Education Policy Governance………. Complexity of state education policy governance………………………... State education structures and policymaker responsiveness……………… Impacts of general-purpose governance structures……………….. Impacts of special-purpose governance structures………………... State education governance arrangements and policy content as an outcome………………………...………………………...……………….. Conclusion………………………...………………………...……………………. 9 9 10 11 12 15 16 16 CHAPTER III Legal Relationships Between State Government and Public Education……………………………...………………………...………………………...……………… Constitutions, State Legislatures, and the Provision of Education…………………… States with education clauses prior to the adoption of the 1787 U.S. Constitution………………………...………………………...………………………... States that established education clauses upon entering the Union………. Introductory stage………………………...………………………................. Foundational stage………………………....……………………….............. Quiescent stage…………………………...………………………................ States that established education clauses years after entering the Union…. Introductory stage………………………...………………………................. Foundational stage………………………...………………………................ Changing Constitutionally Mandated Education Responsibilities of State Legislatures………………………...………………………...……………………….................. The Role of State Boards of Education………………………......…………………...….... State constitutions and state boards of education………………………............. 18 23 24 29 30 30 32 33 34 34 36 38 39 41 42 42 42 44 45 46 46 Constitutional duties of constitutionally established SBEs……….. Constitutional structures of constitutionally established SBEs…… State statutes and state boards of education………………………...…………… Statutory structures of statutorily established SBEs……………… Statutory structures of constitutionally established SBEs………… Anomalies, abolishers, and never adopters………………………........................ Changes in the Structure, Powers, and Duties of State Boards of Education…….. A Note on Governors and Superintendents of Public Instruction………………… Conclusion………………………...………………………...……………………….................... 47 48 49 50 51 52 52 56 57 CHAPTER IV Who Gains State Education Policymaking Power?................................... Epistemological Perspective and Hypotheses………………………...………….. Sub-structures of SBEs………………………...……………………...... Becoming an elected SBE member: Geographic representation…. Becoming an elected SBE member: Partisanship………………… Becoming an appointed SBE member: Getting noticed………….. Appointers of SBE members………………………....................... Appointed SBE demographic equality………………………........ Sub-structures of education committees in state legislatures……………... Term limits………………………...………………………............ Professionalism………………………...………………………..... Data………………………...………………………...………………………........ State education policymaker demographics………………………............ State population demographics………………………...…………………. Indicators of state education policymaking bodies’ institutional structures Methods………………………...………………………...……………………….. Results: Individual Descriptive Statistics, and State and Institutional Variation…. Gender and race/ethnicity………………………...…………………….. Education levels and professional experience……………………….......... Education levels………………………...……………………….... Professional experience………………………...…………………. State education policymaker demographic characteristic intercorrelations. Political party affiliation ………………………...……………………... SBE member campaign donations………………………........................... Results: Variations in demographics of state education policymakers by state…... Results: Institutional structures and state education policymaker demographics… SBE institutional structures………………………...……………………... Legislative institutional structures………………………........................... Results: State Education Policymakers and Descriptive Representation…………. Discussion………………………...………………………...…………………….. Historically marginalized populations need higher credentials to gain policymaking power………………………...……………………….......... Politics and professionalism increase presence of policymakers from historically marginalized populations………………………...................... Politics and professionalism increase descriptive representation………… 58 60 62 62 62 64 65 65 66 67 68 69 69 70 70 71 74 74 79 79 80 81 85 85 86 94 94 95 97 101 101 103 104 Limitations………………………...………………………...……………………. Conclusion………………………...………………………...……………………. 105 106 CHAPTER V State Education Policymakers, Responsiveness, and Valued Voices……... What Do We Know About Policymaker Responsiveness and Substantive Representation? ……………................................................................................... Theoretical Framework and Hypotheses..……………………...………………… Institutional structures: Ways of coming to power……………………… Institutional structures: SBE selection methods………………………....... Institutional structures: Term limits………………………....………….. Institutional structures: Professionalism………………………................. Individual characteristics, representation, and responsiveness………….. Data………………………...………………………...………………………........ Policy case study: Teacher evaluation policy……………………….......... Survey design………………………...……………………...………….. Randomized control trial: Measuring responsiveness..…………. Forced rating………………………...……………………........... Discrete choice experiment………………………...…………… Least and most important actors………………………...………. Sample………………………...………………………...………………… Target population………………………...………………….......... Accessible population………………………...…………………... Sample population………………………...……………………… Examining response bias……………………….............................. Methods………...………………………..............………...……………………… Are policymakers responsive to public and/or teacher policy preferences?. ………...………………………............................................ When forming their position on an education policy, whose voices do state education policymakers value? ………...………………………........ Ratings of state and local actors………………….......................... Discrete choice experiment…………………….............................. Least and most important actors……………………...................... Results: Are State Education Policymakers Responsive to Public or Teacher Preferences? ………………...………………………...………..………………… Results: Whose Voices are Valued in the State Education Policymaking Process?..................................................................................... ............................. Ratings of importance of actors in informing state education policymakers’ policy positions. …………………….................................. Do state education policymaker characteristics impact ratings of importance of actors in informing policymakers’ teacher evaluation policy positions? ……………………............................ Discrete choice selections……………………............................................ State education policymaker characteristics and discrete choice selections…………………….......................................................... Most and least important actors to inform state education policymakers’ policy positions……………………............................................................ 108 109 111 112 113 113 114 116 117 117 118 119 121 121 122 122 122 123 125 126 132 132 132 133 134 135 135 142 142 143 146 146 154 Are some actors more likely to be nominated alongside others?.... Discussion…………………….............................……………………................... Variation by volume and independence of voice…………………………. Institutional structures and political parties impact the value of external voices…………………….............................…………………….............. Limitations……………………............................................................................... CHAPTER VI Donut Devourers, Fish Fanatics, Politicians, and Educators: Voices in the State Education Policymaking Process ………………………...………………….… Data and Methods……………….…....................................................................... Data…………………….............................……………………................. Methods: Designing the interview protocol…………………..................... Methods: Conducting the interviews…………………............................... Analysis………………….................…………………........................................... Results: Who Gains State Education Policymaking Power? And, How?................ Coming to power………………….................…………………................. Power seekers…………………....................................................... Recruits………………….................…………………................... Goals of state education policymakers…………………............................. No education goals………………….................………………….. Broad education goals………………….......................................... Specific education goals…………………....................................... Most important votes…………………....................................................... Results: Real and Ideal Internal Voices of Teacher Evaluation Policymaking….. Results: Policymakers’ Perceptions of Ideal Internal Voices in State Education Policymaking ………………….................…………………................................. Discussion and Implications for Future Research…………………........................ State education governance structures…………………............................. Weights of internal voices and what that means for external voices……... CHAPTER VII State Education Governance Structures and Policy Outcomes………… Background: Teacher Evaluation Policy…………………..................................... Conceptual Framework………………….................…………………................... Data and Methods………………….................…………………........................... Developing a teacher evaluation policymaking authority devolution index Establishing the legal framework…………………........................ Generating relevant legal measures…………………..................... Conducting preliminary research…………………......................... Formalizing coding procedures…………………........................... Independent variables…………………...................................................... Analysis………………….................…………………........................................... State-by-state variation in teacher evaluation policy authority devolution. State education governance, politics, and teacher evaluation policy authority devolution…………………......................................................... Results………………….................…………………............................................. National trends in teacher evaluation policymaking authority devolution.. 155 157 157 159 160 164 166 166 168 171 172 174 174 174 178 180 180 181 182 183 187 190 196 197 199 201 204 206 208 208 209 210 211 212 214 214 214 215 217 217 Teacher evaluation policy sub-elements: Which decisions are devolved?.. Legislatures and SBES: Devolution differences?........................................ State education governance structures and teacher evaluation policymaking authority devolution………………….................…………. Political and governmental influences on teacher evaluation policymaking authority devolution………………….................................. Legislatures and SBEs: Influences of authority devolution………………. Discussion and Implications for Future Research…………………........................ Conclusion………………….................…………………...................................... 221 222 223 225 227 231 234 CHAPTER VIII Significance of the Study and Future Research ……………………..… 235 Implications………………….................…………………..................................... 241 Conclusion………………….................…………………...................................... 243 APPENDICES………………………...………………………...…………………….... Appendix A: Current State Constitution Education Clauses …………………….. Appendix B: Data and Methods for Examining Legal Relationships between State Government and Public Education…………………..................................... Appendix C: State Board of Education Characteristics by State ………………… Appendix D: Methods of Data Collection for State Education Policymaker Demographic Dataset ………………………...………………………...………… Appendix E: Methods of Collecting State Population Demographics …………… Appendix F: Parity Ratio Distributions ………………………...……...………… Appendix G: Literature Review Methods and Data ………………………...……. Appendix H: Survey Methods………………………...………………………...… Appendix I: Recruitment E-mail Message ………………………...……………... Appendix J: Methods for Analyzing Forced Rating Data………………………… Appendix K: Descriptive Statistics of Rating Question, by Office and Political Party ………………………...………………………...………………………….. Appendix L: Descriptive Statistics and Cross Tabulations of Most and Least Important Actor Selections ………………………...……………………............ Appendix M: Teacher Evaluation Policymaking Authority Devolution Index Methods and Data ………………………...………………………...……………. 246 247 254 256 259 265 267 269 294 307 308 312 314 319 REFERENCES………………………...………………………...………………………... 326 LIST OF TABLES Table 1 Demographics of state education policymakers…………………….…………...... 77 Table 2 Correlations between state education policymaker member demographics………. 82 Table 3 Correlations between state board of education demographic makeup……………. 83 Table 4 Correlations between education committee demographic makeup……..……….... 84 Table 5 State-level summary statistics of state education policymaking body demographics…………………….…………......…………………….…………................... 92 Table 6 Correlations between SBE institutional characteristics and demographic makeup.. 96 Table 7 Correlations between institutional characteristics of state legislatures and demographic makeup of education committee members…………………………………… 97 Table 8 Correlations between demographics of state population and state education policymakers………………………………………………………………………………... 99 Table 9 Coefficients of multivariate regressions to examine relationships between institutional characteristics of state education policymaking bodies and state education policymaking body-to-population parity ratios……………………………………….......... 100 Table 10 Treatment conditions for state education policymakers survey experiment……... 120 Table 11 Comparison of demographics among target, accessible and sample population for survey ……………………………...………………………............................................. 123 Table 12 Comparison of response rates and within-group demographics of accessible and sample populations …………………………………………………………......................... 126 Table 13 Tests of survey response bias …………………………………………………… 127 Table 14 Tests of survey response bias, by control/treatment group …….……………….. 128 Table 15 Response time bias: May vs. September e-mail respondents……………………. 130 Table 16 Response time bias: Initial vs. reminder e-mail respondents……………………. 131 Table 17 Probit and OLS regression results: Responsiveness of state education policymakers to public and teacher policy preferences…….………………………………. 140 xiv Table 18 Relationships between institutional characteristics and policymaker responsiveness…………………….…………......…………………….…………................. 142 Table 19 Multivariate analysis of variance to examine relationship between state education policymaker characteristics and their rating of importance of state and local actors…………………….…………......…………………….…………............................... 145 Table 20 Kolmogorov-Smirnov test to compare distributions of rating data among state education policymakers by office and political party ……………………………………… 145 Table 21 Discrete choice experiment results………………………………………………. 147 Table 22 Descriptive statistics of discrete choice selection and results of Chi-Square tests of independence of discrete choice selection variable by state education policymaker characteristics………………………………………………………….................................. 148 Table 23 Discrete choice experiment analysis: Negative binomial regression results...…… 152 Table 24 Qualitative case study site demographics………………..………………………. 167 Table 25 Qualitative interview sample population demographics, by state..………………. 168 Table 26 State education policymakers most important vote, by office………………….... 184 Table 27 Teacher evaluation authority devolution descriptive statistics, by sub-elements and total...…………………………………………………………....................................... 223 Table 28 Total teacher evaluation authority devolution descriptive statistics……..………. 224 Table 29 Generalized Poisson regression model results…………..………………………. 225 Table 30 Current state constitution education clauses…………..………………………….. 247 Table 31 State board of education characteristics by state…………..……………………………. 256 Table 32 CCES survey data information…………..……………………………………….. 265 Table 33 Integrative literature review data analysis stages, purposes, and descriptions…… 272 Table 34 Literature review sources…………..……………………….…………..………… 274 Table 35 Shapiro-Wilks tests for normal distribution of state- and local-actor rating data… 309 Table 36 Descriptive statistics of state education policymakers’ ratings of level of importance of state- and local-level actors in informing their position on teacher evaluation policy, by office and political party…………..………………………………… 312 xv Table 37 Descriptive statistics of most/least important actor nominations by state education policymakers …………..……………………….…………..……………………. 314 Table 38 Descriptive statistics of clustered most/least important actor nominations by state education policymakers …………..……………………….…………..……………………. 315 Table 39 Cross-tabulations of nominated actors chosen by state education policymakers as least likely to inform their position on teacher evaluation policy…………..………………. 316 Table 40 Cross-tabulations of nominated actors chosen by state education policymakers as most likely to inform their position on teacher evaluation policy…………..………………. 317 Table 41 Cross-tabulations of nominated actors chosen by state education policymakers as most and least likely to inform their position on teacher evaluation policy…………..…….. 318 Table 42 Identifying relevant teacher evaluation policy elements…………..…………… 319 Table 43 Authority devolution index measures, overall and by element, by state………….. 322 Table 44 Data descriptions and sources …………..……………………….…………..…… 324 xvi LIST OF FIGURES Figure 1 Result of Gallup Poll question: How much trust and confidence do you have in……….………………………………………………………..…………………………. 1 Figure 2 Result of 2015 PDK/Gallup Poll question: Which unit of government do you believe should be responsible for deciding which textbooks and teaching methods should be used in America’s K-12 public schools? ………………………………………………. 2 Figure 3 Result of 2015 EdNext Poll question: What level of government should play the biggest role in……………………………………………………..……………………….. 3 Figure 4 Four state education governance and policymaking arrangements……………… 16 Figure 5 Variation in state education governance structures…………………………….... 27 Figure 6 Variation in state education governance structures and policymaking relationships………………………………………………………..………………………. 28 Figure 7 Institutional structures and sub-structures of state education governance………. 61 Figure 8 Demographics of state legislators serving on education committees…………… 76 Figure 9 Demographics of state board of education members……………………………. 77 Figure 10 State board of education gender demographics, by state………………………. 87 Figure 11 State legislators serving on education committees gender demographics, by state……………………………………………………………..………………………….. 88 Figure 12 State board of education racial demographics, by state ……………………….. 89 Figure 13 State legislators serving on education committees racial demographics, by state…….………………………………………………………..…………………………. 90 Figure 14 State board of education political party affiliation demographics, by state……. 93 Figure 15 State board of education members’ political campaign contributions, by state... 94 Figure 16 Results of experimental survey question, by control and treatment group……... 139 Figure 17 State education policymakers’ ratings of state- and local-level actors’ influence on education policy decisions…………………………………………………… 144 xvii Figure 18 State education policymakers’ nominations of most and least important actors in informing position on teacher evaluation policy………………………………………... 156 Figure 19 Proposed framework: Weight of internal voices influences weight of external voices heard in the education policymaking process ……………………………………… 165 Figure 20 Weight of internal voices influences weight of external voices heard in the education policymaking process ………………………………………………………….. 200 Figure 21 State teacher evaluation policymaking authority devolution…………………... 220 Figure 22 State board of education-to-population parity ratio distribution by demographic characteristics ……………………………………………………………………………... 268 Figure 23 Education committee member-to-population parity ratio distribution by demographic characteristics ………………………………………………………………. 269 Figure 24 Most/Least important individual and groups question format – original……….. 295 Figure 25 Most/Least important individual and groups question format – modified……… 296 Figure 26 Discrete choice experiment question format……………………………………. 306 xviii CHAPTER I INTRODUCTION Over the past 15 years, Americans’ confidence in state government has faltered. Between 1997 and 2013, the percent of Americans’ with a favorable opinion of their state government declined at twice the rate (14 percent) as that of local government (7 percent) (Pew, 2013). In 2015, 71 percent of Americans had a “great deal” or a “fair amount” of trust in local government to handle problems, compared to about 62 percent who said the same for their state government (Gallup, 2016; see Figure 1). Of all the problems that state governments are handling, education is surely one: education was the most widely cited policy area in governors’ 2016 State of the State addresses (Maness, Malm & Doroghazi, 2016b); and, for state legislators, education legislation was the second largest priority in 2016 – right behind budget and taxes (Maness, Malm & Doroghazi, 2016a). Figure 1 Results of Gallup Poll question: How much trust and confidence do you have in… The local governments in the area where you live when it comes to handling local problems? The government of the state where you live when it comes to handling state problems? 90 80 70 60 50 Note: Percent represents responses of “great deal” and “fair amount” of trust. Dotted lines are present for years where no data was collected and may not represent accurate between-time trend. 1 At the same time that Americans have become less confident in state government, Americans have called upon state government to be more involved in education. For example, in 2014, over half of Americans (56 percent) said that local school boards should have the greatest influence in deciding what is taught in public schools; in contrast, just 28 percent preferred state government involvement in this facet of education (PDK/Gallup Poll, 2014). Just one year later, the gap between local and state control substantially narrowed: 37 percent of Americans believed that the state government should make education decisions, as opposed to 39 percent for local government (PDK/Gallup Poll, 2015). What is more, within demographic groups, the percent of blacks and Hispanics that believed that the state government should be responsible for textbook and pedagogy decisions was greater than those who believed in local control over these issues (see Figure 2). Results from a 2015 EdNext poll reveal similar results: on average, Americans prefer K-12 public school decisions be made by state governments (Henderson, Peterson & West, 2016; see Figure 3). Figure 2 Results of 2015 PDK/Gallup Poll question: Which unit of government do you believe should be responsible for deciding which textbooks and teaching methods should be used in America’s K12 public schools? Federal — 20 State — 37 17 27 31 37 33 35 Local — 39 5 Don't know — National Total 2015 Blacks 28 32 42 6 8 4 Hispanics 2 Whites Figure 3 Results of 2015 EdNext Poll question: What level of government should play the biggest role in… 35 36 30 30 50 42 46 51 47 18 23 18 20 23 29 Hispanics General Public African Americans Hispanics 46 25 32 African Americans 39 17 General Public 41 General Public Local — 43 15 39 44 Hispanics State — 15 African Americans Federal — Setting educational standards Deciding whether or not a school is failing 46 Deciding how to fix a failing school Recent policy trends suggest that the demands of the average American are being met: state governments have indeed become more active in making education policy (Fusarelli & Cooper, 2009; Henig, 2013; Kirst, 2010). However, this shift comes at a time when local school board demographics have become more representative of historically marginalized groups—or, at least more so than state government. While for most of American history local school boards were comprised of primarily white, middle-class men (Kirst & Wirt, 2009), 44 percent of today’s local school board members are women (Hess & Meeks, 2010). In comparison, less than one in four state legislators is female, a ratio that has increased by less than four percentage points over the past sixteen years (NCSL, 2010). Moreover, minorities make up nearly 20 percent of today’s local school board members; and, in large school districts, nearly one in every three school board 3 members is non-white (Hess & Meeks, 2010). In comparison, just one in nine state legislators is African American or Hispanic (NALEO & NCSL, 2009; NBCSL & NCSL, 2009). In sum, the nature of who is involved in making education policy is rapidly changing. At the local level, an increasingly diverse body of policymakers is in place with the potential for improved representation for marginalized groups. Nonetheless, state-level policymakers have tightened the reins on local education policymakers and expanded their influence over local policy and practice. Some state policymakers continue to support local control, desiring to keep policymaking closer to citizens who “understand better than anyone else their community’s zone of school policy tolerance” (Kirst & Wirt, 2009, p. 31). Yet, pervasive and enduring issues of educational equity have led much of the American public to lose confidence in local education policymakers’ ability to provide equitable access to high quality education (Kirst, 1987). However, the public’s preferences are also nuanced. For example, a review of 40 years of polling data suggests that the public prefers elected local school boards manage and improve local schools while looking to federal and state government to ensure equitable distribution of funding and establish standards for student achievement (Jacobsen & Saultz, 2012). At all levels of governance, responsiveness by government officials to the public remains an underlying democratic value (Key, 1961; Dahl, 1971). While a substantial amount of work has examined local school board governance and responsiveness to public preferences, very little research has examined the responsiveness of state education policymakers within different state education governance structures. As public K-12 education governance in America shifts away from a highly localized, relatively diverse single-purpose government toward an education governance structure involving state policymakers in both single- and general-purpose governing bodies with distinct rules, traditions and norms, is this value of policymaker responsiveness 4 preserved? Who are state education policymakers and whose voices are heard in the state education policymaking process? Overview of the Study Part I of this dissertation lays the historical and theoretical foundations for a study of state-level education governance. Chapter II begins with a review of research on state government involvement in education policymaking, highlighting the ways in which state education governance has been conceptualized. As others have argued (Manna & Harwood, 2011), I find that much work has addressed how state governance has impacted the development and implementation of federal policy (Manna, 2006; Manna, 2011; McGuinn, 2006). Very little research has focused on state-level education policymaking processes. Moreover, the term “state government” has often been applied casually, projecting an image of state government as a solitary, monolithic institution. However, as I explain in Chapter 2, wide variation exists in the institutional structures of America’s state-level formal education policymaking bodies. I argue that this complexity must be taken into account when considering the intersections of state education governance and policymaking. This dissertation focuses on formal state-level education policymakers, defined as statelevel actors in positions of power that are granted the constitutional or statutory authority to review, accept, amend, and/or reject binding education policy in the form of statute or administrative rules. In Chapter III, I provide a historical overview of the origins of state government’s legal responsibilities related to K-12 public education. The purpose of this overview is twofold: (1) to establish grounds for studying state legislators and SBE members as formal state-level education policymakers and (2) to provide a historical and legal understanding 5 of state education governance structures. I draw upon primary legal documents – namely, state constitutions and statutes – to examine the origins of the structures, roles, and education-related responsibilities of state legislatures and SBEs and how they have changed over time. Finally, I provide a description of the wide variation in present-day state education governance structures across the United States. Part II of this dissertation focuses on my empirical analyses of whose voices are heard in the state education policymaking process. Within each chapter, I lay out the theoretical framework(s) that guides how I approach my research questions. I begin Chapter IV by examining demographics and descriptive representation of state education policymakers. Viewing “governance arrangements as serving to institutionalize the preferences of different sets of stakeholders” (McLendon, Hearn, & Deaton, 2006, p. 19), I discuss the importance of considering how the institutional characteristics of state education policymaking bodies may impact who gains state education policymaking power and whose voices are heard by those who gain power. Drawing upon the theory of new institutionalism, I develop hypotheses that motivate my empirical analyses focused on who gains state education policymaking power. Analyzing a unique dataset of demographic characteristics of state legislators serving on education committees and SBE members across all 50 states, I examine state education policymaker demographics. I also explore whether state education policymakers are descriptively representative of their K-12 public education constituents. K-12 public education constituents are defined as those who contribute resources (e.g., time, money) to public schools (i.e., school employees, taxpayers), as well as those who receive services from public schools (i.e., students, parents). Additionally, I examine the ways in which institutional structures the demographics and descriptive representation of state education policymaking bodies. 6 Moving beyond questions of whether state education policymakers look like their constituents, Chapter V explores if state education policymakers act like their constituents. In this chapter, I expand my theoretical framework to recognize that, once in a position of power, state education policymakers have individual agency and can act strategically in the policymaking process. Blending theoretical frameworks of institutionalism and individual agency, I develop hypotheses related to the institutional structures and individual characteristics that may facilitate or impede policy responsiveness. Using survey data, including responses to a randomized control trial survey question, I explore the institutional and individual characteristics that impact state education policymakers’ responsiveness to the public and to teachers. And, because the public and teachers are only two constituencies that policymakers may care about, I also examine the external voices that are most and least influential in state education policymakers’ decisionmaking process and how institutional structures and individual characteristics impact how much value state education policymakers place on difference voices. In Chapter VI, I draw upon interview data with state policymakers across six states – each with a unique education governance structure – to provide a nuanced understanding of who gains state education policymaking power and whose voices are heard in the state education policymaking process. Expanding on Chapter IV, I explore the inward qualities and motivations of state education policymakers and discuss similarities and differences across individuals and education governance structures. Subsequently, I expand upon Chapter V by examining the internal voices that are heard in the state education policymaking process. I then discuss the importance of recognizing the weight of voices internal to the state education policymaker process in order to understand how much weight can be placed on the value of the external voices identified in Chapter V. 7 In Chapter VII, I continue to explore the significance of both internal and external voices in the policymaking process by examining how state education governance structures may impact the policy decisions of state legislatures and SBEs. In particular, utilizing an original and unique database of state teacher evaluation statutes and administrative rules, I analyze the extent to state legislators’ and SBEs’ devolve or retain teacher evaluation policymaking authority. A policy devolution index is presented that quantifies the differences in how states as a whole, as well as state legislatures and SBEs as separate policymaking bodies, delegate authority to local decision-makers responsible for implementing teacher evaluation policies. I supplement this index with other state-level measures to examine how governance arrangements and political contexts influence the level of teacher evaluation policymaking authority devolved from state to local education policymakers Finally, Chapter VIII discusses the implications of this dissertation and lays out future research that should be considered as a result of the findings from this study. Additionally, I discuss both the contributions of this research, as well as the ways in which this research enriches the broader literature on education governance, politics, and representation. 8 Chapter II Education Governance: A Review of The Literature I begin this work with an overview of American K-12 public school education governance, focusing on who is legally responsible for the provision of public education in the United States. I describe how those with education policymaking power have traditionally chosen to devolve that authority to other policymaking bodies at both the state and local level. Subsequently, I review research that has focused on education governance structures and education policy and identify three issues that have largely been overlooked: (a) the complexity of state education governance structures, (b) the state education policymaking process, and (c) the relationships between state education governance structures, representation, responsiveness. Finally, I draw upon research in other social policy areas and consider why it is important to expand our knowledge of these three issues in education policy research. Background: Governing America’s K-12 Public Schools Since the establishment of the first local school board in Bay Colony, Massachusetts in 1826, America’s public schools have represented “the ultimate expression of American democracy,” operated by the people, for the people (Maeroff, 2010, p.1). Through a democratic process, American citizens select members of their local board of education who then govern the schools located in their community, taking public votes on issues dealing with matters such as maintaining and constructing school buildings; hiring, firing, and promoting personnel; selecting a school superintendent; approving the school budget; making student disciplinary decision; and authorizing textbooks for use by students, to name a few. Although this intimately local control 9 of public education has been considered the norm in America, local school boards are creatures of the state, deriving power and authority from law enacted by state legislatures and administrative rules put in place by state boards of education (SBEs). State legislatures. Nearly all fifty state constitutions designate the state legislature1 as the entity responsible for establishing, maintaining, and/or supporting a public education system (see Appendix A). Nonetheless, many state legislatures have deliberately chosen to delegate education policymaking power and authority to local school boards. For example, the North Carolina state legislature, who’s constitution requires them to “provide by taxation and otherwise for a general and uniform system of free public schools […] wherein equal opportunities shall be provided for all students” (N.C. Const., art. IX § 2) enacted statute that states, “it shall be the duty of local boards of education to provide adequate school systems within their respective local school administrative units, as directed by law” (N.C. § 115C-47). Similarly, in fulfilling its constitutional requirement to “provide for the maintenance and support of a thorough and efficient system of free public schools for the instruction of all children in the State between the ages of five and eighteen years,” (N.J. Const., art VIII, § 4, para. 1), the New Jersey state legislature enacted statute mandating local boards of education to carry out a multitude of responsibilities including, but not limited to: setting the school calendar, transporting students to and from school, setting curriculum and selecting textbooks, enacting and enforcing student discipline policies, as well as hiring, firing, promoting, and compensating personnel. Yet, the 1 The term “state legislature” refers to a state’s legislative body, which in some cases is termed the “general assembly” or “general court.” In forty-seven states, this includes an “upper house” (typically, the Senate) and a “lower house” (typically, the House of Representatives). Three states – Georgia, Pennsylvania and Vermont – are comprised of just one, unicameral legislative body. 10 New Jersey Supreme Court ruling in Robinson v. Cahill (1973) succinctly describes the definitive obligation of the state legislature to fulfill their constitutional responsibility: Whether the state acts directly or imposed the role upon local government, the end product must be what the Constitution commands. A system of instruction in any district in the State which is not thorough and efficient falls short of the constitutional command. Whatever the reason for the violation, the obligation is the State’s to rectify it. If local government fails, the State government must compel it to act, and if the local government cannot carry the burden, the State must itself meet its continuing obligation (303 A.2d 273, 1973). Thus, although state legislatures have delegated policymaking authority over many educational matters to local school boards, most state legislatures are vested with the ultimate authority over the provision of education within their state. State boards of education. Some state constitutions call upon an SBE to assist in the governing of a state’s education system. For example, many state legislatures call upon their SBE to enact administrative rules around particular education issues such as standards and assessments. The extent to which state legislatures utilize their SBE varies considerably from state to state. While many state legislatures provide SBEs with broad discretion to enact legally binding rules, others are more prescriptive, establishing strict parameters within which SBEs are allowed to operate. As an example of the former, Massachusetts statute provides its SBE with powers to “establish policies relative to the education of students in public early childhood, elementary, secondary and vocational-technical schools” (MA Title XII § 69 Sec. 1B). Texas is 11 an example of the latter, with statute permitting its SBE to “perform only those duties relating to school districts or regional education service centers assigned to the board by the constitution of this state or by this subchapter or another provision of this code” (TX § 2b-7.102). State Government Involvement in Educational Matters As previously described, state legislatures – also known as general-purpose2 governments – have frequently called upon education-specific governing bodies to carry out educational matters within their state. The motivation for the use of these special-purpose3 governments dates back to the early 1900s when Progressive Era reformers aimed to “take education out of politics.” Seeking to shield educational decisions from elected officials’ parochial and special interests, reformers aimed to turn political issues into matters for administrative discretion by professional educators (Reese, 2002). Confidence in the governing abilities of these local education professionals began to weaken in the 1950s, particularly after the Soviet Union’s launch of the Sputnik space satellite (Manna & McGuinn, 2013). Reasoning that America’s loss to the Soviet Union in the “race to space” was directly related to America’s dismal education system governed by local education professionals, federal and state governments felt pressure to intervene (Manna & McGuinn, 2013; Theobald & Malen, 2000). Americans’ decline in confidence in local school boards continued to accelerate throughout the 2-th century. Time and again, state and federal courts declared that local policymakers unjustly tolerated unequal educational opportunities. Out of the Supreme Court rulings in Brown v. Board of Education I (1954) and II (1955), as well as the Civil Rights Act of 2 General purpose governments, of which legislatures are a part, provide a wide range of functions to the public and operate under broad grants of authority” (Biggs & Helms, 2014, p.199). 3 Special purpose governments, of which school districts are a part, are entities authorized by state law to provide only one or a limited number of designated functions, and with sufficient administrative and fiscal autonomy to quality as separate governments (U.S. Census Bureau, n.d.). 12 1964, state legislatures were called upon to intervene and require local school boards to take remedial action to desegregate schools and ensure that all students had equal access to a quality education. Additionally, at the same time Americans were faced with the 1980-82 recession and an increasing sense of fear of increasing global competition, A Nation at Risk was published. The report assailed schools as producing a “rising tide of mediocrity” that threatened nothing less than “our very future as a nature and a people” (National Commission on Excellence in Education, 1983, para. 1). This report, alongside the economic and social conditions, contributed to education becoming an increasingly important policy issue confronting state and federal governments. For a variety of purposes – national security, social inequities, economic wellbeing – the mounting level of public distrust in local boards of education quickly led to state action on a variety of educational matters traditionally reserved to local school boards. While many state legislatures initiated these reforms, the federal government also played an important role in compelling states to adopt and implement standards, assessment, accountability, and teacher quality policies, as well as finance reform through federal laws and court decisions – most notably, the Elementary and Secondary Education Act (ESEA) and its subsequent revisions, as well as the Brown v. Board of Education I (1954) and II (1955), San Antonio Independent School District v. Rodriguez (1972) and Serrano v. Priest I (1971), II (1976), and III (1977) decisions. For example, between 1983 and 1984, thirty-four state legislatures established high school graduation requirements (Hurst, Tan, Meek, & Sellers, 2003). Throughout the 1980s and 1990s, many state legislatures also enacted K-12 curriculum standards, mandatory student assessments, school accountability systems, centralized school funding systems, and more uniform teacher training and certification requirements (Hurst, Tan, Meek, & Sellers, 2003). 13 Today, state legislatures are more involved than ever in the education policymaking process. While some legislatures utilize their SBE in the policymaking process, reliance upon local boards of education continues to decline. Take, for example, the State of Michigan: in 1996, the state legislature enacted just 11 education-related laws, two-thirds of which addressed statewide student proficiency assessments. By 2004, the state legislature enacted twice as many (24) education-related laws, ranging from school choice, competitive bid school construction projects, alternative governance structures for Detroit Public Schools, use of special education and vocational education bond funds, state assessments, sex education, and teacher hiring, firing, evaluation and certification, to name a few. SBEs in many states are also becoming more active in the policymaking process, either after being prompted by the state legislature or at the SBE’s own discretion; or, in response to federal laws. For example, the Florida SBE enacted 62 new administrative rules in 2014, nearly doubling the number of rules that were enacted in 2008 (38).4 The centralization of education has prompted many researchers to examine the impacts of this shift in education policy governance. The subsequent literature review provides an overview of such research followed by a discussion in which I identify key gaps in the literatures related to (a) the complexity of state education governance structures, (b) the state education policymaking process, and (c) the relationships between state education governance arrangements, representation, and responsiveness. 4 Data compiled by author using publically available data via the web from Michigan Compiled Laws and Florida Administrative Code. 14 Literature Review: Empirical Research on State Governance of Public Education As shown in, state legislatures have several governance options to choose from in their effort to fulfill their constitutional obligations related to the provision of public education. Under configuration A, the state legislature chooses to wield strong policymaking power, enacting rigid statutes. Under configuration B, the state legislature chooses to devolve policymaking power directly to local school boards. Under configuration C, the state legislature chooses to devolve policymaking power to the SBE, who then wields strongly policymaking power, enacting rigid administrative rules. Finally, under configuration D, the state legislature devolves policymaking power to the SBE, who then devolves policymaking power to local school boards. Despite calls to consider these matters more deeply (Manna & Harwood, 2011; Masters, Salisbury, & Eliot, 1964), research examining state governance and education policy has been relatively limited. Of the studies that have been conducted, configurations A and B have been the most commonly researched. Very little research has examined arrangements C and D. As such, this majority of the literature reviewed in this chapter focuses on the first two governance arrangements. 15 Figure 4 Four State Education Governance and Policymaking Arrangements A B State legislature State legislature State Board of Education State Board of Education Local Board of Education Local Board of Education C State legislature D State legislature State Board of Education State Board of Education Local Board of Education Local Board of Education How has state education policy governance been examined? From curriculum standards and standardized testing to teacher licensure and evaluation, states have become “increasingly, and at times aggressively, involved in efforts to influence schools” (Malen, 2003). Origins for this increase in state government involvement lie both in state-led efforts as well as federal mandates, inducements, and incentives. State-led education policymaking. Research focused on state-led efforts to drive educational improvement through policy proliferated in the 1980s when many state governments began to take a heavy-handed approach to regulate, restructure, and reform schools through “results-based” policies (Malen, 2003). A significant portion of research examining the effects of these state-led education policy changes focused on policy implementation. Results from such research found that state-led policy significantly impacted teaching and learning, for better or for worse. For example, policy enacted by state government related to standards and testing was found to effect teachers’ decisions about what content to cover and how to teach it; it even led to 16 some teachers engaging in unethical practices like inflating test scores (Firestone, Fitz & Broadfoot, 1999; Firestone, Mayrowetz & Fairman, 1998; Jacob & Levitt, 2002; Malen, 2003; McNeil, 2000; Smith, 1998; Wilson & Floden, 2001; Wong, Anagnostopoulos, Rutledge, Lynn, & Dreeben, 1999). Other researchers drew upon Weatherly & Lipksy’s (1977) concept of “street level bureaucrats” to emphasize the inherent principal-agent problem between state policymakers and local implementers in a loosely coupled system (Weick, 1976). This research shed light on the ways in which state mandated policy often cannot effectively influence what happens in schools (Fuhrman, 1993; Mavrogordato & White, 2017; O’Day & Smith, 1993; Smith & O’Day, 1991), especially because districts could essentially “use policies that fit their own plans and fight the reforms that did not” (Firestone, Bader, Massel, & Rosenblum, 1992, p. 271). In addition to research on the implementation of state-led education policy, some researchers have examined the political processes that ensue when state governments become involved in the education policymaking process. For example, Tim Mazzoni’s (1974, 1980, 1986, 1988, 1993) work focused Minnesota state government’s education policymaking process constitutes one of the most extensive research projects to examine the political processes of statelevel education policymaking. Additionally, Cornbleth & Waugh (1995) studied the states of California and New York, highlighting power and politics and how they played out with respect to history/social studies curriculum policy enacted by the state government. In their research on California’s math curriculum reforms, Cohen & Hill (2001) also recognized the political processes involved in the policy development and implementation. Finally, Kirst & Wirt (2009) examined the politics involved in state governments’ charter school policy decisions and Buckley & Kuscova (2003) examined the politics of the development of state charter laws. 17 While providing insight into the political processes of state education policymaking, these studies focused on state legislation, bills, and resolutions, rarely incorporating administrative rules developed by SBEs. In fact, the term “state government” in these studies was often used synonymously with state legislatures, largely ignoring the role and influence of the SBE in the policymaking process—save for two exceptions. First, Mazzoni (1993) does recognize that the Minnesota SBE was viewed as an important policymaking arena related to the issue of desegregation in the early 1970s. However, he dismisses much of the influence of the board, explaining that the board had much of its formal authority taken away by the legislature in the late 1970s and early 1980s. Second, Cohen & Hill (2001) recognize the role of the California SBE in the state’s 1980s math curriculum reforms; however, they assert that SBE power was limited due to political traditions that made the SBE reluctant to mandate local educational action. Federally-mandated, induced, and incentivized state education policymaking. Over and above research on state-led education policy implementation and politics, research has extensively examined state government education policymaking in response to federal mandates, inducements, and incentivizes. Similar to research on state-led education policy in the 1980s, research related to state government policymaking in response to federal mandates enacted in the 1970s (e.g., ESEA, 1965) found that policy implementation and outcomes depended upon local conditions and local implementers’ skills and beliefs (Bardach, 1977; Berman & McLaughlin, 1978; Elmore, 1979). A key contribution of this research was the recognition that policy made higher up in the governance hierarchy cannot mandate what matters at the local level (Odden, 18 1991) and that what is actually delivered in the classroom depends ultimately on the individual at the end of the line: the teacher (Weatherley & Lipsky, 1977). The ESEA (1965) also resulted in numerous studies of state and local use of federal education funding (Barkin & Hettich, 1968; Berke, 1974; Berke & Kirst, 1972; Berman & McLaughlin, 1978; McLaughlin, 1975; McLaughlin, 1976; Washington Research Project and NAACP Legal Defense and Education Fund, 1969; Wargo, Tallmadge, Michaels, Lipe, & Morris, 1972;). A limited number of studies explored state education agencies’ (SEA) responses to new federal funds. For example, in addition to examining financial data, Berke & Kirst (1972) conducted case studies in six states, exploring how SEA personnel interpreted ESEA (1965) and viewed their role in implementing the law. Similarly, Murphy (1971) conducted interviews with SEA officials in nine states and Washington D.C. to understand how organizational routines and procedures within SEAs influenced states’ response to federal education reforms. Finally, Milstein (1976) examined the internal and external responses of three SEAs to the influx of state dollars under ESEA. After this initial influx of research associated with ESEA (1965), state-level education policymaking research waned. For example, the Education Amendments Act of 1974 and 1978, as well as the Middle Income Students Assistant Act of 1978 each called upon states governments to help implement federal education policy changes. However, very little research explored state-level education governance and policymaking in response to these federal programs. One exception is McDonnell & McLaughlin’s (1980, 1982) evaluation of Title IV’s consolidated program strategy. Analyzing results of a survey of Title IV program officials, state advisory council members, and public and nonpublic school officials in about 600 school districts, as well as fieldwork in eight state departments of education and 24 school districts, 19 McDonnell & McLaughlin (1982) described how states implemented this federal policy. In this study, the SEA served as the primary state-level entity of focus; state legislators and SBE members were largely ignored. The 1980s brought about some of the initial research focused on state legislators and SBE members responses to federal mandates. Following the passage of the Education Consolidation and Improvement Act (ECIA) of 1982, Darling-Hammond and Marks (1983) explored state policymakers’ plans for ECIA administration. To do so, Darling-Hammond and Marks (1983) conducted interviews in nine states (Alabama, California, Colorado, Georgia, Kansas, Massachusetts, Minnesota, Missouri, and Pennsylvania) with chairpersons of the House and Senate education and appropriations committees or their chiefs of staff; staff member(s) in the governor’s office responsible for education and human services; state department of education officials; and at least one key member of the SBE and/or the Chapter 2 advisory committee. Similarly, Moore, Goertz, & Hartle (1983) examined state administration of federal education programs including those established via ECIA. To do so, the research team examined documentary material from each state and conducted interviews with SEA officials and staff, legislators and their staff, interest group representatives, local school administrators, teachers, and journalists who covered education in eight states (California, Louisiana, Massachusetts, Missouri, New Mexico, New York, Virginia and Wyoming). These studies did incorporate both legislators and SBE members; however, the researchers primarily focused on the implementation of federal education programs, rather than how the state education policymakers came to their decisions. Although empirical research of state legislators and SBEs gained some popularity in the 1980s, research following ESEA reauthorizations of the 1990s largely ignored state legislators 20 and SBE members, focusing instead on local policymakers and, again, SEAs. Subsequent to President Bill Clinton’s signing of the Improving America’s School Act (IASA) of 1994, which required states to develop standards and student assessments aligned to those standards as well as identify schools in need of improvement that are not making adequate yearly progress, numerous scholars provided descriptive analyses of various aspects of IASA. For example, researchers outlined IASA programs related to drug-free and gun-free schools, bilingual education, parent involvement, and in-school health care services, while also providing insight into empirical research evidence that could provide insight into the implementation and possible effects of these new or reformed programs (Modzeleski, 1996; Johnson, 1997; Nakagawa, 2000; Pena, 2000; Wiese & Garcia, 1998; Yell & Rozalski, 2000). Other researchers explored SEA and/or local school districts’ implementation of specific IASA programs, such as the curriculum standards and aligned assessments (Buckles, Schug & Watts, 2001; Sunderman, 2001) as well as Title I, bilingual education, and special education reforms (McLaughlin & Verstegen, 1998). As the federal government tightened their grip on education policy across the nation, however, little to no empirical research explored state legislature or SBE responses to federal education reforms or approaches to education policymaking during the 1990s. With the passage of the No Child Left Behind (NCLB) Act in 2001, seen by some as “represent[ing] the most sweeping intrusions into state and local control of education in the history of the United States” (Becker & Halderman, 2004, p. A1), many researchers examined state-level policymaking efforts as a result of this federal education reform. In line with most empirical research prior to this time, however, state legislators and SBEs were largely ignored as state education policy leaders. The SEA continued to be the primary state-level entity of study. For example, numerous scholars conducted interviews with or surveys of SEA officials to 21 explore SEA capacity and approaches to implementing NCLB, as well as constraints faced by SEAs (Pinkerton, Scott, Buell & Kober, 2004; McMurrer, 2007; Minnici & Bartley, 2007; Minnici & Hill, 2007; Sunderman & Orfield, 2006). Focusing on specific sub-programs of NCLB, Hamann & Lane (2004) examined how SEAs interpreted the comprehensive school reform demonstration program; Nagle, Yunker, & Malmgeren (2006) explored challenges identified by SEA officials related to meeting NCLB’s annual yearly progress requirements for students with disabilities; Gill et al. (2008), Vernez et al. (2009), and Zimmer et al. (2007) highlighted SEA officials’ implementation of NCLB’s school choice and supplemental educational services requirement; O’Day et al. (2009) focused on SEA officials’ implementation of NCLB’s flexibility around the inclusion of students taking alternative assessments in district and state AYP calculations; and Le Floch et al. (2007), as well as Taylor et al. (2010) interviewed SEA officials to understand how they implemented the standards, assessment, and accountability provisions of Titles I and III of NCLB. Additionally, post-NCLB research focused on variation in state government responses to NCLB mandates over time. For example, Carny & Loeb (2002) found that some state governments enacted strong accountability policies that rewarded schools that performed well on state assessments and sent negative signals to those that did not. Researchers also found that state governments continuously found new ways to calculate adequate yearly progress in order to raise the number of schools and districts that met NCLB requirements (Stark Retner et al., 2006), altered definitions of what it means to be a “highly qualified” teacher (Darling-Hammond & Berry, 2006), and, in some cases, enacted state laws placing the state in formal opposition to NCLB (Shelly, 2008). Other studies related to state government responses to NCLB examined political factors that shaped state educational accountability policies and the resulting variability 22 in state accountability policy (McDermott, 2003). However, nearly all of these studies examined state laws, bills, and resolutions, largely ignoring the role of the SBE in the rulemaking process. Although no major changes were made to NCLB for over fourteen years, federal education reform did transpire in the form of the Obama Administration’s Race to the Top (RttT) and ESEA waiver programs. These two programs offered state governments an opportunity to receive additional federal education funding (in the case of RttT) and federal policy implementation flexibility (in the case of ESEA waivers) in exchange for the implementation of federally supported reforms. Again, very little research examined how state legislators and SBE members responded to the RttT and ESEA wavier programs. State-level studies, however, did include Shelly’s (2012) study of the characteristics of state governments that engaged in the ESEA waiver process; Kolbe & King Rice (2012) and Russell et al.’s (2015) examination of SEAs’ RttT applications; Wrabel et al’s (in press) analysis of how political contexts impact state ESEA waiver content; and Wong’s (2015) analysis of SEAs’ ESEA waiver monitoring reports issued by the U.S. Department of Education. Limitations of the Existing Research of State Education Policy Governance While the previously reviewed research has undoubtedly contributed to an understanding of the impacts state government can have on the education policy implementation process and policy outcomes, I identify three gaps in the literature related to how state government is conceptualized and the types of impacts that have been studied related to state education policy governance. 23 Complexity of state education policy governance. State education policy and governance research rarely accounts for the complexities in institutional arrangements related to the governance of state education policy. In particular, education policy literature and research often employs the term “state government” casually, projecting an image of state government as a solitary, monolithic institution – typically, the state legislature or the SEA. For example, in her review of McDermott (2003), stated, “the lack of a strong relationship between ideology and accountability policy makes sense when we consider that many of the choices states have made about assessment policy do not have clear “conservative” and “liberal” sides” (p.172). The terminology here – “the choices states have” – fails to recognize that state-level education governance consists of policymakers that span across both general- and special-purpose policymaking bodies. This terminology also raises a larger issue of ascribing action to “states,” when in fact it is the individuals within the various state education policymaking bodies that are taking policy action. As another example, Stark Retner et al. (2006) stated, Some of the fiercest rebellion against NCLB in year 4 came from state policymakers— both Republicans and Democrats. As of June 2005, bills or resolutions related to NCLB had been introduced in 20 states (National Conference of State Legislatures, 2005a) (p.4). While focusing on actions of state legislatures, SBE member actions – such as policies enacted by California SBE that pushed back against policy elements of NCLB (California State Board of Education, 2015) – have gone largely unnoticed. While formal state education policymakers reside in both state legislatures and SBEs, wide variation exists in the institutional arrangements of these state-level education policymaking bodies. For example, SBEs across America vary in their selection methods (elected, appointed), geographic representation (regional, at-large), membership size, term 24 length, and term limits. Additionally, although the election of state legislators is uniform across all states, there is variation in the frequency with which state legislatures meet, membership and committee size, term limits, and level of professionalism (e.g., salary and benefits, staff and resources). Figure 5 displays the variation of a number of state education policy governance structures. For example, Alabama’s SBE is elected for four-year terms and its legislature has a relatively low professionalism ranking, as measured by the Squire’s (2007) index. In contrast, Pennsylvania’s SBE is appointed for six year terms and its legislature has a relatively high professionalism ranking. The extent to which SBEs are involved in making education policy adds variation to the governance arrangements undergirding each state’s education policymaking process. For example, as previously described, some state legislatures solicit the assistance of SBEs in the policymaking process, while others limit the policymaking power of their SBE. Figure 6 builds upon Figure 5 to display this variation for one education policy – teacher evaluation policy. These variations in policymaking relationships between state legislatures and SBEs may influence the policymaking process. For example, when a state legislature wields strong policymaking power, those involved in the policymaking process are individuals who are popularly elected by local constituents within a legislative district and likely concerned with multiple other policy issue areas, not solely education. In contrast, when a state legislature solicits the assistance of its SBE, policy may be developed by a combination of popularly elected legislators in conjunction with SBE members who could be appointed by the governor or popularly elected statewide and whose sole policy issue area is education. Drawing upon research that reveals the significant influence policymakers’ individual motivations and ambitions, as well as the institutional arrangements in which policy is made 25 (e.g., term limits, selection structures, professionalism, geographic restrictions) has on policymakers’ actions and their ability and willingness to be responsive to public preferences (e.g., Arnold, 1990; Barro, 1973; Carey, Niemi, & Powell, 1998; Fenno, 1978; Ferejohn, 1986; Schlesinger, 1966), I argue that these variations in state education policymaking governance structures impact who gains state education policymaking power and whose voice is heard in the policymaking process. Thus, these more complex education policy governance arrangements should be taken into account when examining state involvement in the education policymaking process. 26 Figure 5 Variation in state education governance structures Elected Michigan (SBE elected, 8-year term; high leg. professionalism) Ohio (SBE mixed selection, 4-year term; high leg. professionalism) SBE Selection Method Alabama (SBE elected, 4year term; low leg. professionalism) 1 0.8 Mixed 0.6 0.4 Appointed 0.2 4 5 6 7 0.5 0.6 8 9 Legislative Professionalism Pennsylvania (SBE apointed, 6-year term; high leg. professionalism) 27 0.4 0.3 0.2 0.1 Figure 6 Variation in state education governance structures and policymaking relationships Elected Michigan (elected, 8-year term, high leg. professionalism) Ohio (mixed selection, 4-year term, high leg. professionalism) SBE Selection Method Alabama (elected, 4year term, low leg. professionalism) 1 0.8 Mixed 0.6 0.4 Appointed 4 5 6 7 Pennsylvania (appointed, 6-year term, high leg. professionalism) 0.6 8 9 0.5 0.4 0.3 0.2 0.2 Legislative Professionalism 0.1 Note: Dotted circles represent states where the state legislature provides the SBE with limited policymaking authority over teacher evaluation. Solid circles represent states where the state legislature provides the SBE with extensive teacher evaluation policymaking authority. 28 State education structures and policymaker responsiveness. The second gap that I identify in the research is related to responsiveness. Although substantial attention has been given to the implementation and outcomes of state education polices, very little research has examined the policymaking process. Given that education is in and of itself, seen as “an expression of democratic political ideas” (Labaree, 1997), the lack of research connecting state education governance arrangements and responsiveness to public education policy preferences is disconcerting. And, if a “key characteristic of a democracy is the continued responsiveness of the government to the preferences of its citizens, considered as political equals” (Dahl, 1971, p. 1), it is important to understand whether citizens are in fact considered political equals in the state education policymaking process. In political science, the concept of “responsiveness” stems from considerations of participatory democracy and political representativeness (Eulau & Karps, 1977; Verba & Nie, 1972). In particular, Pitkin (1967) argues that responsiveness is the critical characteristic of representation, defining representation as “acting in the interest of the represented, in a manner responsive to them” (p. 209). Moreover, Eulau & Karps (1977) contend, “how the representative and the represented interact with respect to the making of public policies lies at the heart of most discussions of responsiveness” (p.242). However responsiveness is defined, “it cannot be denied that policy responsiveness is an important component of representation” (Eulau & Karps, 1977, p. 242). Yet, the relationship between state education policy governance arrangements and policy responsiveness has not been studied. Research in other areas of American domestic policy, however, provides some evidence of the benefits of studying policy responsiveness in conjunction with governance structures. 29 Impacts of general-purpose governance structures. Scholars in the field of political science have extensively explored the intersection of the institutional arrangements of state legislatures and the policymaking process (Lowi, 1971; Baumgartner & Jones, 1991). For example, Bigelow (2006) examined the circumstances under which over 4,000 state legislators in 30 states were responsive to public preferences related to economic and social policy. Bigelow (2006) found that (a) legislators were most responsive to constituents on highly salient social issues, (b) legislators from electorally safe districts were most responsive to their districts, and (c) more professional legislatures were more responsive to public preferences. Maestas (2000) also examined the relationship between state legislative governance arrangements (i.e., professionalism and progressive ambition) and legislative responsiveness. Maestas’s (2000) measure of responsiveness, however, was not related to a single policy initiative; rather, Maestas (2000) utilized Erickson, Wright, & McIver’s (1993) measure of policy liberalism –an index measure of whether state legislatures have consistently enacted policies that are more or less liberal than state opinion warrants. Similar to Bigelow (2006), Maestas (2000) found more professional state legislatures and legislatures with more opportunities for members to progress to higher office had greater aggregate policy responsiveness. Impacts of special-purpose governance structures. Unlike general-purpose governments that provide a range of public services, special-purpose governments provide “a single or limited services” (Foster, 1997, p. 2). Examples of special purpose governments include school districts, public utilities boards, hospital boards, park and recreation authorities, and transportation authorities. While not as extensive as research on the general-purpose government, some 30 research has examined how institutional arrangements of special-purpose governments influence policy responsiveness. Within the special-purpose governance literature, conflicting hypotheses have emerged related to the policy consequences of specialized institutional arrangements of government for policy responsiveness. In some cases, scholars found that special-purpose governments are less likely to be captured by special interest groups and more likely to deliver efficient policy decisions that are congruent with majority opinion (Bish, 1971; Ostrom, Bish & Ostrom 1988). In contrast, other scholars found that the low visibility of special-purpose governments hindered public participation and created a bias toward special interests that have the capacity to lobby special-purpose government officials (Bollens, 1957; Burns, 1994; Jones, 1942; Macedo & Karpowitz, 2006). Two additional studies have examined policy responsiveness in special-purpose governments. Mullin (2008, 2009) examined special-governments’ responsiveness to public preferences related to local water policy and found that special-purpose governments were more strongly incentivized to respond to the preferences of the median constituent, especially when the special-purpose governments employed electoral rules that promote majority outcomes. In contrast, general-purpose governments, managing multiple policy dimensions, needed to satisfy constituents on the most important policy question and thus sometimes chose to bargain and respond to interest groups on other issues (Mullin, 2008, 2009). Foster (1997) also studied metropolitan regions’ reliance on special-purpose governments and found that specialized governance had a significant and positive influence on resource allocation: metropolitan areas that relied on special purpose governments to provide important public services “devoted 31 proportionately more of total local government expenditures to these services than do metropolitan areas that rely on general-purpose governments” (p. 198). This literature on state governance structures and policy responsiveness is limited in a number of ways. First, the majority of the literature is focused on the local level (e.g., cities and counties), thus making it difficult to know whether findings can be generalized to state-level special-purpose governments such as SBEs. Second, research has not explored how variation in structural arrangements of state-level special purpose governments (e.g., elected vs. appointed officials) may impact policy responsiveness. I am also unaware of any research that focuses on state-level special-purpose governments related to education, and their responsiveness to public preferences. Finally, very little research has simultaneously considered the impacts of specialand general-purpose governance structures. State education governance arrangements and policy content as an outcome. The content included in written education policy are important components that can impact how a policy impacts its intended audience. The specificity of the policy, the amount of flexibility provided to local implementers, and policy consequences or rewards are components of written policy can significantly influence policy implementation and outcomes. Nonetheless, research has generally overlooked the written policy content as an outcome of the state education policy process, instead skipping right to an evaluation of whether the policy was successful or a failure (for exceptions, see Youngs, Odden & Porter, 2003). In contrast to the numerous studies that have examined the institutional arrangements of state legislatures and policy responsiveness, few studies have examined the relationship between the institutional arrangements of state legislatures and the resulting policy content. One exception is 32 a study by Ringquist (1994), which found that states with more professional legislatures, more liberal citizens, and greater issue salience related to environmental quality enacted stronger pollution control regulations. Similar to the general-purpose government literature, very little research has examined the relationship between special-purpose government institutional arrangements and policy content and characteristics. One study, conducted by Berry (1979), did find that special-purpose governments that are more professional and in states where the issue at hand is more salient were more likely to adopt redistributive electric utility rate structures. Conclusion Although scholarship on education governance, politics, and policy is not lacking, there are still several gaps in the research literature, as described above. This study seeks to contribute to this literature by examining how state politics and education governance structures impact who gains sate education policymaking power, whose voices are heard in the state education policymaking process, and how this may impact how a state approaches the education policymaking process. Before these questions can be examined, it is important to have a thorough understanding of the origins and current state of state education governance structures. Thus, in an effort to establish the importance of studying these two formal state-level education, Chapter III includes a discussion of the origins of state governmental bodies’ legal responsibilities related to K12 public education. 33 Chapter III Legal Relationships Between State Government and Public Education With both its silence on the subject of education and the establishment of the Tenth amendment, which expresses the principal of federalism, the United States Constitution reserved the power to educate America’s youth to the states. Although not explicitly stated in the U.S. Constitution, many national and state leaders saw incorporating education with government as critical to the preparation of citizens and necessary for continued support of republican government (Coram, 1791). As such, the basis for state control over education was established in many state constitutions in the early 1800s. In what follows, I first discuss the importance of a constitutionally established responsibility of education for state legislatures, as opposed to one that is statutorily established. Next, I describe the national historical context of constitutional mandates related to education placed upon state legislatures and how those mandates have changed over time. Subsequently, I discuss the role of the state board of education (SBEs) in establishing, maintaining, and overseeing public education and I review how state constitutions and statutes influence the structures, roles, and responsibilities of SBEs. Finally, I examine how the structures, roles, and responsibilities of SBEs have changed over time. Data and methods utilized for this review are detailed in Appendix B. Constitutions, State Legislatures, and the Provision of Education Public education is just one of the many constitutionally established policy issue areas under the purview of state legislatures. Given that the threshold required for an amendment to 34 constitutional law is much higher than that of statutory law, placement of this responsibility within a state constitution is of particular importance. In contrast to constitutional law, statutes and changes to statutes are enacted simply by a majority vote of the legislature and subsequent signature of the governor. Twenty-one states also allow for citizens to bring about a statutory initiative, which empowers citizens to initiate a change in statute by obtaining a predetermined number of petition signatures that can force a public vote (Initiative and Referendum Institute, n.d.). The signature thresholds and standards of statutory initiatives vary widely from state to state: while most states require signatures from somewhere between five and 10 percent of either the total number of qualified voters or the total number votes cast for governor in the preceding elections, the threshold is as low as two percent of the resident population in North Dakota and as high as 15 percent in Wyoming (National Conference of State Legislatures, 2012). If the signature thresholds are met and the initiative is successful at the ballot box, the measure then becomes law. Although all 50 states allows for their state constitution to be amended, the threshold that must be met in order for the change to occur is typically high. For example, most state legislatures are able to propose that a constitutional amendment be placed on the ballot (Tarr & Williams, 2005). Upon passage in the legislature, the majority of citizens of the state must vote to approve the amendment before it is enacted (Tarr & Williams, 2005)5 Additionally, 41 states have an explicit constitutional provision that allows for the assembly of a constitutional convention to review the current constitution and propose changes that then have to be approved by the voters (Tarr & Williams, 2005). Finally, although standards for constitutional initiatives vary, 18 states provide voters with the right to directly amend their constitution by obtaining a 5 Thirty-six states require just one passage through the legislature; 11 states require two; and three states require either passage by the majority in two separate legislative sessions or by a supermajority vote of one session. The Delaware legislature can amend their constitution by a two-thirds vote by two consecutive legislatures; no vote of the people is required. 35 predetermined number of petition signatures that can force a public vote (Tarr & Williams, 2005). There are two other less ways that constitutions can be amended: first, if a court rules that elements of a state constitution are either in conflict with the U.S. Constitution or with rest of the state constitution; and, second, if a statewide commission has been formed – such as those in Florida and North Carolina – and given authority to propose amendments to the constitution (Tarr & Williams, 2005). The placement of a state legislature’s powers and responsibilities related to public education in a state constitution suggests some level of durability: whereas state statutes are amended hundreds of times each and every year, the average number of amendments for a state constitution, over all of American history, is 115 (Bowser, 2015). In what follows, I discuss the education-related constitutional mandates placed on state legislatures, namely how the mandates vary by both state and time of ratification and if and how the mandates have changed over time. I differentiate between states that established education clauses prior to statehood, upon obtaining statehood, and many years after obtaining statehood. States with education clauses prior to the adoption of the 1787 U.S. Constitution. The portion of a state’s constitution that defines the relationship between state legislators and public education is known as the education clause. Prior to the signing of the U.S. Constitution in 1787, many states ratified education clauses within their individual state constitutions. Although a number of state constitutional education clauses established prior to 1787 have stood the test of time, many states modified or even eliminated their education clause after the adoption of the 1787 U.S. Constitution. 36 The language of Massachusetts’ original constitution remains in tact today, with its unique education clause that requires legislators “to cherish the interests of literature and science, and all seminaries of them: especially […] public schools” (Constitution of the Commonwealth of Massachusetts, 1780, Chapter V). Similarly, North Carolina’s original education clause prevailed for nearly 100 years. Like North Carolina, the original constitutions of Pennsylvania and Vermont ordered their state legislatures to establish schools for the “convenient instruction of youth” (Constitution of the Commonwealth of Pennsylvania, 1776, Section 44; Constitution of the State of North Carolina, 1776, Article XLI; Constitution of the State of Vermont, 1777, Section XL). Although the original constitution of the State of Georgia included an education clause, the responsibility of the state legislature was relatively ambiguous, stating, “Schools shall be erected in each county, and support at the general expense of the State, as the legislature shall hereafter point out” (Constitution of the State of Georgia, 1777, Article LIV). After the signing of the 1787 U.S. Constitution, Pennsylvania expedited and expanded the responsibility of its state legislature, stating: “the legislature shall, as soon as conveniently may be, provide, by law, for the establishment of schools throughout the state, in such manner that the poor may be taught gratis” (Constitution of the Commonwealth of Pennsylvania, 1790, Article VI). In contrast, the states of Vermont and Georgia modified their constitutions after the signing of the 1787 U.S. Constitution, leading to the removal of any constitutional requirement that the state legislature establish a public school system. The Constitution of the State of Georgia remained devoid of an education clause from 1789 to 1861. After the start of the Civil War, Georgia reinstituted an education clause, affording the state legislature the power to appropriate money to provide for the education of its people (Constitution of the State of Georgia, 1861, Article II). It was not until 1868, with the initiation of Reconstruction, that the 37 Constitution of the State of Georgia was modified to require its state legislature to establish a public education system (Constitution of the State of Georgia, 1868, Article VI). Vermont, too, shifted from requiring the legislature to establish a school(s) in each town “for the convenient instruction of youth” (Constitution of the State of Vermont, 1777, Section XL) to an education clause that was relatively vague as to the state legislatures roles and responsibilities in establishing schools. Vermont’s revised education clause simply required “a competent number of schools” be “maintained in each town, for the convenient instruction of youth” and “one or more grammar schools” be “incorporated and properly supported, in each county in this State” (Constitution of the State of Vermont, 1793, Section 41st). Vermont reiterated its commitment to local control of schools in 1997 by enacting statute that explicitly states: [O]ne of the strengths of Vermont’s educating system lies in its rich diversity and the ability for each local school district to adapt its educational program to local needs and desires. Therefore, it is the policy of the State that all Vermont children be afford educational opportunities that are substantially equal although educational programs may vary from district to district (16 V.S.A. § 1). States that established education clauses upon entering the Union. Many states established education as a constitutionally mandated responsibility of the state legislature upon entering the Union. The content of these state education clauses varied substantially depending upon the time at which, and the context in which, the state constitution was established. Paul Tractenberg’s (2006) four stages of constitutional education provisions provide an apt structure within which these variations may be discussed. 38 Introductory Stage. Education clauses ratified in state constitutions during what Tractenberg (2006) terms the introductory stage (1776-1834) were established during a time when much debate was taking place about the appropriate or ideal relationship between state governments and the education system. Delaware was the only state during this era to require its state legislature to establish schools in its original constitution. Due to differing views around the constitutionalization of education, the constitutions of the states of New Hampshire, Ohio, Indiana, Maine, and Missouri ratified education clauses focused on the importance of an informed citizenry for the preservation of a free government. The Constitution of the State of New Hampshire (1792), for example, stated, Knowledge and learning generally diffused through a community being essential to the preservation of free government; and spreading the opportunities and advantages of education to various parts of the country being highly conducive to promote this end, it shall be the duty of legislators and magistrates, in all future periods of this government, to cherish the interests of literature and the sciences, and all seminaries and public schools (Sec. LXXXIII). The Constitution of the State of Indiana utilized the same education clause language as the State of New Hampshire; however, the Indiana Constitution (1816) went a step further, requiring the legislature to “from time to time, pass laws as shall be calculated to encourage intellectual, scientifical, and agricultural improvement” and “provide, by law, for a general system of education […] wherein tuition shall be gratis, and equally open to all” (Article 9). Comparably, the Bill of Rights of the Constitution of the State of Ohio (1802) indicated that, “religion, morality and knowledge being necessary to good government and the happiness of mankind,” the “means of instruction shall forever be encouraged by legislative provision” 39 (Article VIIII). Similar to Ohio, the constitutions ratified by Maine, Mississippi, and Missouri recognized that importance of “a general diffusion of knowledge” and “encouraged” the means of education (Constitution of the State of Maine, 1820, Article VIII; Constitution of the State of Mississippi, 1817, Article VI; Constitution of the State of Missouri, 1820, Article VI). Moreover, whereas the original constitutions of Maine and Missouri required the establishment of public schools, Mississippi did not. However, while Maine made it the duty of the state legislature to “require the several towns to make suitable provision, at their own expense, for the support and maintenance of public schools” (1820, Article VIII), Missouri simply required the establishment of one or more schools in each township where poor students would be taught for free. It is unclear whether the legislature was required to fulfill this constitutional mandate, or whether this was left to the local townships. The relatively vague education clause language around the responsibility of the Maine legislature to establish public schools – beyond simply requiring towns to maintain public schools – remains in tact today. Missouri, on the other hand, ratified a new constitution in 1865 that required the legislature to “establish and maintain free schools for the gratuitous instruction of all persons in this state within the ages of five and twenty-one years” (Article IX). Additionally, beginning with the ratification of the state’s second constitution in 1868, Mississippi required its legislature to establish “a uniform system of free public schools” (Article VIII). Finally, it should be noted that constitutions ratified during this introductory era oftentimes afforded state legislatures with flexibility around the time table for the establishment of a public school system, using phrases such as “as soon as practicable and necessary” or “as soon as circumstances will permit” (Constitution of the State of Missouri, 1820, Article VI; Constitution of the State of Indiana, 1816, Article 9). This malleable terminology was eliminated 40 in subsequent constitutions (Constitution of the State of Missouri, 1865, Article IX; Constitution of the State of Indiana, 1851, Article 8). The Foundational Stage. Tractenberg’s (2006) foundational stage (1835-1912) encompasses a period during which the number of states doubled, most states entering the Union established education clauses, and most of those without educational provisions added them to their state constitution (as detailed in introductory stage). During this era, 24 states attained statehood, 22 of which ratified a state constitution that included an education clause requiring state legislative involvement in the provision of public education. State constitutions varied in the extent to which state legislatures were responsible for the provision of public education. Nine states – Michigan, Minnesota, Montana, Nebraska, Nevada, Texas, Wisconsin, Oregon, and West Virginia – required that their state legislature establish a system of public schools. Texas and Wisconsin also required the schools be free; Oregon, Minnesota, Nevada, and Montana also required schools be uniform; Nebraska required schools be thorough and efficient; and West Virginia also required the schools be thorough, efficient and free. Eleven states – Arizona, Alabama, Colorado, Hawaii, Idaho, Kansas, North Dakota, Oklahoma, South Dakota, Utah, and Wyoming – required that their state legislature establish and maintain a system of public schools, with all but Kansas and Hawaii requiring the system be complete, thorough, and/or uniform, and free. Finally, two states – New Mexico and California – ratified relatively vague education language around the responsibility of the state legislature to establish public schools. The constitutions of these two states more closely emulated those in the introductory stage: California simply encouraged the state legislature to promote “intellectual, scientific, moral, and 41 agricultural improvement” (Constitution of the State of California, 1949, Article IX). New Mexico required the establishment and maintenance of a “uniform system of free public schools sufficient for the education of, and open to, all children of school age in the State” (Constitution of the State of New Mexico, 1910, Article 12). It is unclear whether the legislature was required to fulfill this constitutional mandate. Quiescent Stage. The quiescent stage (1930-1950), according to Tractenberg (2006), had only limited, sporadic constitutional activity. During this era, the 49th and 50th states – Alaska and Hawaii – entered the Union. Both states established education clauses in their constitutions, requiring state government involvement in the provision of public education. Whereas Alaska required the state legislature to “establish and maintain a system of public schools,” Hawaii, similar to New Mexico, simply required “the State” to “provide for the establishment, support, control and maintenance of a statewide system of public schools” (Constitution of the State of Alaska, 1956, Article VII; Constitution of the State of Hawaii, 1950, Article X). Again, this language does not clarify the specific role of the state legislature in fulfilling this constitutional directive. States that established education clauses years after entering the union. Finally, a number of states established education as a constitutionally mandated responsibility of the state legislature many years after entering the Union. Introductory Stage. A number of states did not establish constitutional education clauses until several years after attaining statehood. The state constitution of twelve states that attained 42 statehood in the introductory stage – Alabama, Connecticut, Illinois, Kentucky, Louisiana, Maryland, New Jersey, New York, Rhode Island, South Carolina, Tennessee and Virginia – did not include an education clause until many years later. In a handful of states, numerous years passed before the state ratified an education clause to be included in their state constitution: over 30 years in Louisiana, over 50 years in Illinois, over 75 years in Alabama and Maryland, and over 100 years in South Carolina. These late adopters, when they did ratify an education clause, varied in the extent to which state legislatures were responsible for the provision of public education. Whereas Alabama required its legislature to establish and maintain “a liberal system of public schools,” Illinois, Maryland, Louisiana and South Carolina required their state legislatures to provide or establish a system of free public schools (Constitution of the State of Alabama, 1901, Article XIV; Constitution of the State of Illinois, 1870; Article VIII; Constitution of the State of Maryland, 1864, Article VIII; Constitution of the State of Louisiana, 1845, Title VII; Constitution of the State of South Carolina, 1895, Article XI). In contrast, Rhode Island established a much weaker education clause more than 50 years after it entered the Union in which the state legislature was only required to “promote public schools” and “adopt all means which they may deem necessary and proper to secure to the people the advantages and opportunities of education” (Constitution of the State of Rhode Island, 1842, Article XII). Other states built up their education clauses over many years. In particular, Connecticut, New Jersey, and New York entered the union in the late 1780s but did not ratify an education clause as part of their state constitution until 1818, 1844, and 1821, respectively; and, even then, the education clauses in these three states simply reserved state funds and lands for public schools (The Constitution of the State of Connecticut, 1818, Article Eighth; The Constitution of the State of New Jersey, 1944, Section VII; The Constitution of the State of New York, 1821, 43 Article VII). It was not until more than 100 years after statehood –1875 in New Jersey, 1894 in New York, and 1965 in Connecticut – that their state constitutions were amended to require the state legislature to maintain a system of public schools (The Constitution of the State of Connecticut, 1965, Article Eighth; The Constitution of the State of New Jersey, 1875, Section VII; The Constitution of the State of New York, 1894, Article IV). Similarly, Virginia and Kentucky entered the union in 1788 and 1792, respectively, and established a fund for public schools that the state legislature was to maintain in the 1850s. These two states did not require the state legislature to provide for a system of public schools until 1870 in Virginia and 1891 in Kentucky (The Constitution of the State of Kentucky, 1850, Article XI; The Constitution of the State of Kentucky, 1891, Section 183; The Constitution of the Commonwealth of Virginia, 1851, Section 24; The Constitution of the Commonwealth of Virginia, 1870, Article VIII). Foundational Stage. Although Arkansas and Florida both entered the union and ratified an education clause requiring the legislature to provide for a system of public schools during the foundational stage, substantial time lapsed between these two events. The State of Arkansas, which entered the Union in 1836, did not incorporate an education clause into its constitution until 1868. Similar to Missouri and New Mexico, however, the states first and only education clause requires “the State” – not specifically the state legislature – to “maintain a general, suitable and efficient system of free public schools” (Constitution of the State of Arkansas, 1868, Article IX). The State of Florida entered the Union in 1845 but did not enact an education clause until 1868, upon which they required the legislature to provide for the establishment and maintenance of a uniform system of free public schools (1868, Article VIII). 44 Changing Constitutionally Mandated Education Responsibilities of State Legislatures By the mid-1900s, it was well established that state legislatures had a constitutional mandate to establish and, in many cases, maintain a system of free public schools. However, many state education clauses did little to require state legislatures to ensure all students – most notably, minority students – had equal and equitable access to these schools. The Civil Rights movement (1954-1968) resulted in many state’s taking up and ratifying constitutional amendments. Although 12 states already called for the state legislature to maintain and support a “thorough,” “efficient,” or “high quality” system of public education prior to the Civil Rights Era,6 38 states did not require a specific standard of public school system be met for all students. Eight of these states did ratify their constitution to require the state legislature establish a school system that is “thorough,” “efficient,” “high quality,” “adequate” or one wherein equal opportunities are provided for all students.7 As of today, the majority of states8 (30) still do not have a constitutional mandate that specifies a standard of public education the legislature must meet, beyond that of being uniform, free and/or available to all (in many states, it is not necessarily equally or equitably available to all). In fact, Alabama continues to constitutionally provide for separate schools for white and colored children, although the Alabama Democratic Conference chairman in 2012 claims that the “racist language in the constitution” has no effect in the wake of Brown v. Board of Education (Elliot, 2012). 6 Arkansas, Colorado, Delaware, Idaho, Kentucky, Montana, New Jersey, Ohio, Pennsylvania, Texas, Washington, and West Virginia 7 Florida, Georgia, Illinois, Minnesota, Montana, North Carolina, Oregon, and Virginia 8 Alabama, Alaska, Arizona, California, Connecticut, Hawaii, Indiana, Iowa, Kansas, Louisiana, Maine, Maryland, Massachusetts, Michigan, Mississippi, Missouri, Nebraska, Nevada, New Hampshire, New Mexico, New York, North Dakota, Oklahoma, South Dakota, Rhode Island, South Carolina, Tennessee, Utah, Vermont, and Wisconsin 45 The Role of State Boards of Education While state legislatures are general-purpose governments – focusing on a myriad of policy issues – state boards of education (SBEs) are special-purpose governing entities, focused on one issue: education. According to the National Association of State Boards of Education (NASBE), SBEs are intended to serve as an unbiased broker for education decisionmaking, focusing on the big picture, articulating the long-term vision and needs for public education, and making policy based on the best interests of the public and the young people of America (NASBE, n.d.). Serving as a lay body over state education, SBEs are provided powers and charged with various responsibilities through state constitutional and statutory law. In contrast to state legislatures, wide variation exists in the legal home of SBEs powers and responsibilities. In what follows, I discuss the constitutional and statutory mandates placed on SBEs and how the legal venue in which these mandates are placed impacts the extent to which SBEs, their powers, and their responsibilities haven change over time. Additional details of the structure of each state’s SBE can be found in Appendix C. State constitutions and state boards of education. Unlike education-related constitutional mandates placed upon state legislatures, most states established SBEs many years after entering the Union. SBEs were established in 26 state constitutions between 1857 and 1970. Drawing upon Tractenberg’s (2006) four stages of constitutional education provision, 21 SBEs9 9 Alabama, Arizona, California, Colorado, Florida, Idaho, Iowa, Louisiana, Michigan, Mississippi, Missouri, Montana, New Mexico, North Carolina, Ohio, Oklahoma, South Carolina, Texas, Utah, Virginia, and West Virginia 46 made their constitutional debut during the foundational stage (1835-1912), one state10 established its SBE in the quiescent stage (1836-1950 century), and four states11 established SBEs in the more recent rights stage (1950-present). No SBE provisions were ratified in a state constitution during the introductory stage (1776-1834). Finally, while seven states included SBEs in their original constitutions,12 19 states amended their constitutions – often many years after entering the Union – to establish and define the roles and responsibilities of SBEs. Constitutional duties of constitutionally established SBEs. The extent to which current state constitutions establish duties and responsibilities of SBEs varies widely from state to state. In the majority of states with constitutionally established SBEs13 (16), the state legislature is instructed to establish SBE powers and duties, using language such as “the powers, duties, compensation and expenses, and terms of office, of the board shall be such as may be prescribed by law” (Constitution of the State of Arizona, Article 11, Section 3). In contrast, five state constitutions14 establish broad policymaking powers to SBEs, with some oversight by the state legislature. For example, Hawaii’s SBE has “power, as provided by law, to formulate statewide educational policy” (Constitution of the State of Hawaii, 1963, Article X, Section 2). Illinois’s SBE “except as limited by law, may establish goals, determine policies, provide for planning and evaluating education programs and recommend financing” (Constitution of the State of Illinois, 1970, Article X, Section 2). Similarly, Iowa’s constitutionally-established SBE had “full power and authority to legislate and make all needful rules and regulations to Common Schools […] but 10 Georgia 11 Hawaii, Illinois, Kansas, and Nebraska Idaho, Montana, New Mexico, Oklahoma, Utah, and West Virginia 13 Arizona, Colorado, Florida, Georgia, Idaho, Louisiana, Michigan, Missouri, Montana, Nebraska, Ohio, Oklahoma, South Carolina, Texas, Virginia, and West Virginia 14 Hawaii, Illinois, Iowa, New Mexico, and North Carolina 12 47 all acts, rules, and regulations of said Board may be altered, amended or repealed by the General Assembly” (Constitution of the State of Iowa, 1857, Article IX, Section 8). Four constitutionally established SBEs15 are tasked with the “general supervision of public schools.” Finally, California is the only state constitution that establishes, but do not provide for any SBE powers or duties and does not require (but also does not prohibit) the state legislature to assign SBE powers and duties. Constitutional structures of constitutionally established SBEs. In addition to establishing duties and responsibilities, a number of state constitutions establish the structure of SBEs. Of the 24 SBEs established in present day constitutions, seven state constitutions request that the state legislature establish both the size and membership of its SBE. Two state constitutions require SBE members be appointed by the Governor and approved by the legislature; however, the legislature is called upon to establish the size of its SBE. Constitutions in the remaining states with constitutionally established SBEs (15) specify multiple facets of its SBE structure. Both the total member size and total voting member size of these SBEs varies from a minimum of seven members and voting members in Colorado and Florida to a maximum of 19 total members in North Carolina and a maximum of 17 voting members in South Carolina (total member mean=10.7; voting member mean=9.6). Additionally, constitutions in 11 of these 15 states specify the length of SBE members’ terms, which ranges from four to nine years (mean=6.5 years). Ten constitutionally established SBEs specify that the board must be comprised of members who are appointed by the governor and approved by the legislature. Just two of these appointed SBEs must be geographically representative (e.g., by Congressional or Supreme Court districts). Two of the appointed SBEs must also include 15 Alabama, Kansas, Mississippi, and Utah 48 specific individuals, including: the superintendent of public instruction; a local school teacher; a school district administrator; a state university, college or community college president; and a charter school representative. In contrast, seven of states establish elected SBEs: six states require all members be elected and one state (Louisiana) requires about three-quarters of the SBE be elected and one-quarter appointed. Additionally, one state requires partisan SBE elections, one state requires non-partisan SBE elections, and the other five do not stipulate partisanship requirements. State statutes and state boards of education. State boards of education have been established via state statute in 24 states. The first state board of education in the United States was established via statute in Massachusetts in 1787 and the most recent establishment of a SBE via statute transpired in Kentucky in 1990.16 Of those SBEs that acquire powers and responsibilities via state statute, just three were created within the first ten years of the state entering the Union. In contrast to the timing of SBEs established via constitution, the majority of SBEs established via state statute occurred in the introductory (9) and quiescent (11) stages. Just four state legislatures enacted statute establishing an SBE during the foundational stage. Statutes establishing specific duties and responsibilities of SBEs span across both the 24 statutorily established SBEs and the 26 constitutionally established SBEs. These statutory duties and responsibilities vary widely from state to state. The most common duties of SBEs include the development, implementation, and oversight of standards and assessments as well as teacher licensing. Moreover, the structure of the SBE is determined in statute in 24 statutorily established SBEs and nine constitutionally establish SBEs. 16 The Rhode Island State Board of Education was officially created in 2014; however, prior to the establishment of the current SBE, the state had a Board of Regents of Elementary and Secondary Education that functioned much like a SBE and had the ability to enact administrative regulations. 49 Statutory structures of statutorily established SBEs. Both the total member size and total voting member size of the 24 statutorily established SBEs varies from a minimum of seven members and voting members in Alaska, Delaware, New Hampshire, and North Dakota to a maximum of 22 total members and 21 voting members in Pennsylvania. The mean total SBE membership and mean voting membership of statutorily established SBEs is higher than those of constitutionally established SBEs at 11.0 and 10.2 members, respectively. Additionally, statutes in all 24 states specify the length of SBE members’ terms, which ranges from four to seven years (mean=4.9 years). Five-sixths (20) of statutorily established SBEs are appointed: 17 SBEs are made up of members appointed by the governor and approved by the legislature; three are appointed solely by the governor; one is made up of governor appointees that must be chosen from a legislative nominee list; one is a mix of governor appointments and governor appoints with legislative approval; and one is appointed solely by the legislature. Eleven of these appointed SBEs have a geographically representation component, seven of which also have an at-large representative component. Sixteen of the statutorily established appointed SBEs must also include specific individuals, the most common of which are: student representatives, higher education representatives, the superintendent of public instruction, local school board members, and business/industry professionals. Additionally six of these appointed SBEs must have a relatively even distribution of political parties among its members. The remaining one-sixth (4) of statutorily established SBEs are elected. Two states require all members be elected and two states require a mix of elected and appointed members. Three states require regionally-based elections and one state requires the majority of elected SBE members be selected by school 50 district directors in a region. Additionally, one state requires partisan SBE elections, one state requires non-partisan SBE elections, and the other two do not stipulate partisanship requirements. Statutory structures of constitutionally established SBEs. Of the 24 SBEs established in present day constitutions, seven state constitutions request that the state legislature establish both the size and membership of its SBE. Two state constitutions require its SBE be made up of members who are appointed by the Governor and approved by the legislature; however, the legislature is called upon to establish the size of its SBE. The total members and total voting members of these constitutionally established SBEs tends to be larger than those of solely constitutionally or statutorily established SBEs at 12.2 and 11.3 members, respectively. Additionally, when left to the state legislature to determine, the length of constitutionally established SBE members’ terms is, on average, shorter than solely constitutionally or statutorily established SBEs (mean=4.0 years). Finally, nearly two-thirds (6) of constitutionally established, statutorily structured SBEs include appointed members: five of these SBEs are made up of members appointed by the governor and approved by the legislature; and one is appointed solely by the governor. Three of these appointed SBEs have both a geographically representation and an at-large representative component. Two of the constitutionally established, statutorily structured SBEs include elected members: one holds regional elections and one holds a statewide election. The SBE structure in one state (Ohio) is a mix of elected and appointed members. Finally, five of the constitutionally established, statutorily structured SBEs must include specific individuals, the most common of which are the superintendent of public instruction and the governor. 51 Anomalies, abolishers, and never adopters. In the previous discussion of the legal venues of SBEs, two states – Alabama and Iowa – are discussed in both the constitutional and statutorily established SBEs. These anomalies are the result of the states establishing an SBE via constitution in the mid-1800s, abolishing the SBE less than 10 years later, and re-establishing an SBE via state statute in subsequent years (Iowa in 1864; Alabama in 1927). Maine also represents a unique situation in that its SBE was established via state statute in 1846, abolished in 1852, and reinstituted via statute in 1949. New Mexico is distinct from all other states in that an SBE was established in state constitution in 1910; however, in 1958, the SBE was abolished and, in its place, an entity called the Public Education Commission (PEC) was established in the state constitution. The New Mexico PEC functions much like an SBE with ten members elected by the public working with the state department to oversee public education in the state and make policy recommendations to the legislature. Finally, two states – Wisconsin and Minnesota – have never legally established a formal SBE. Changes in the Structure, Powers, and Duties of State Boards of Education An analysis of amendments made to constitutionally and statutorily established SBEs indicates that when an SBE is established in a state constitution, it endures fewer changes to its structure. Across the 26 states that have constitutionally established SBEs, a total of 32 amendments have been made (average of 1.28 amendments per state). Six states17 have never made an amendment to their constitution’s SBE provision; eleven states18 have made just one 17 Arizona, Colorado, Idaho, Illinois, Kansas and Missouri Iowa, Michigan, Mississippi, Montana, Nebraska, New Mexico, North Carolina, Ohio, South Carolina, Utah, and West Virginia. The change made in Iowa is unique in that it resulted in the elimination of a constitutionally established SBE. Iowa then established an SBE via state statute in 1864. 18 52 amendment; four states19 have made two amendments; and four states20 have made three or more amendments. The majority of constitutional amendments to SBEs are related to the makeup and technical aspects of the board (e.g., number of members on the board, who can/must serve on the board, term limits, compensation, election timing). For example, the Constitution of the State of California first established its SBE in 1879 and required that it comprise of the “governor, superintendent of public instruction, President of the University of California, and the professor of pedagogy therein, and the principals of the state normal schools.” In 1885, this language was amended so as to strip the President and the professor of pedagogy from the University of California of their SBE seat. In 1894, the President and professor were reinstituted as SBE members and remained SBE members until 1970 when the Constitution of the State of California was amended to require the legislature to “provide for the appointment or election of the State Board of Education.” As another example, the Constitution of the State of North Carolina established an SBE in 1868 and required that it comprise of the “governor, lieutenant-governor, secretary of state, treasurer, auditor, superintendent of public works, superintendent of public instruction and attorney general.” Amendments made to the state’s constitution in 1971 introduced a new board makeup: “the lieutenant governor, treasurer, and eleven members appointed by the governor, subject to confirmation by the general assembly in joint session.” Because state constitutions are either absent of or incorporate relatively broad language around SBEs duties (e.g., “The duties of the board shall be prescribed by law”), very few 19 Alabama, Georgia, Louisiana, and Virginia. The change made in Alabama is unique in that it constitutes one change that eliminated the SBE (1876) and another change that reinstituted the SBE (1969). Alabama also statutorily established an SBE in 1927. 20 California, Florida, and Texas have each made three modifications; Hawaii has made four. 53 constitutional changes have been made related to the duties of the SBE. However, the duties of many constitutionally established SBEs have indeed changed via state statute. In contrast to constitutionally established SBEs, across the 24 states that have statutorily established SBEs, hundreds of thousands of amendments have been made. A comprehensive historical analysis of every statutory change to an SBE’s structure and every amendment to limit, modify, or expand an SBE’s powers and duties is beyond the scope of this work. However, an examination of SBE-related statutory amendments since the early 1900s suggests that the number and scope of statutory amendments for SBEs is much larger and broader than that of constitutional amendments. As an illustrative example, let’s look at the state of Nevada. When the Nevada state legislature first established its SBE in 1864, the members of the Board consisted of the Governor Superintendent of Public Instruction, and the Surveyor-General (Chapter 52, Statutes of Nevada, 1867). In 1895, the President of the University replaced the Surveyor-General (Chapter 91, Statutes of Nevada, 1895). In 1931, the legislature introduced elected membership: while the governor and superintendent remained on the board, five lay members were to be elected from each of the state’s education supervision districts (Chapter 211, States of Nevada 1931). By 1955, the governor and superintendent were no longer members of the Board and the Board was expanded to eight members, six of whom were elected and two of whom were appointed by the elected members (Chapter 402, Statutes of Nevada, 1955). The Board was expanded to nine members, with the additional member elected from the largest education supervision district in 1969 (Chapter 625, Statutes of Nevada, 1969). In 1971, the legislature required that the two appointed members be representatives from the fields of labor and agriculture (Chapter 651, Statues of Nevada, 1971). In 1973, a fully elective, nine-member board was established (Chapter 54 93, Statutes of Nevada, 1973). This structure stayed in place until 1991 when the SBE was increased from 9 to 11 members and, in 2001, it was decreased to 10 voting members (Chapter 411, Statutes of Nevada 1991; Chapter 218, Statutes of Nevada, 2001). Finally, in 2011, the SBE, once again, became a blended board of seven elected and appointed members, with one voting member being elected from each of Nevada’s four Congressional Districts and three appointed members – one by the Governor, one nominated by the Senate Majority Leader and appointed by the Governor, and one nominated by the Speaker of the Assembly and appointed by the Governor. Within the three appointed members, one of them must be a K-12 teacher at a Nevada public school selected from a list of three candidates provided by the Nevada State Education Association, one must be the parent or legal guardian of a pupil currently enrolled in a public school, and one must be a business person. The SBE was also modified to include four nonvoting members appointed by the Governor: one member of the local school board of trustees, one local school district superintendent, one person from the Nevada System of Higher Education, and one public school student, each nominated by the statewide professional association associated with the individual’s position (Chapter 380, Statutes of Nevada 2011). As the Nevada case illustrates, the number of the changes made to SBE structures via state statute are many; this does not even begin to illustrate the number of changes made each and every year to the powers and duties of SBEs. Because SBE powers and duties are much more likely to be detailed in statute and statute is easier to amend that a constitution, amendments to SBE powers and duties are modified nearly every year in every state. 55 A Note on Governors and Superintendents of Public Instruction Governors can and, in many states, do play a substantial role in the state-level education policymaking process. For example, governors have the power to veto legislation, as well as issue executive orders. However, veto power only applies to bills that the legislature has proposed and executive order powers, which vary from state to state, are primarily limited to emergency situations (e.g., natural disasters, energy crises) and the creation of advisory or investigative committees or commissions (National Governors Association, n.d.). As such, governors lack formal education policymaking authority: any policy or program proposed by a governor must incorporate the legislature and/or state rulemaking agency who must either appropriate funds or enact statute or administrative code. Moreover, a number of governor's have significant informal policymaking power through their authority to appoint a state superintendent or chief state school officer. However, this power, like the veto and executive order power, is not what I define as formal education policymaking power – that is, power to unilaterally enact education laws, rules, and regulations. State superintendents, too, play a role in the state-level exaction policymaking process. In fact, many state constitutions established and specified the roles and responsibilities of the superintendent of public instruction prior to those of the state board of education. In most states, however, the superintendent of public instruction was first established as a position in the Governor’s cabinet. While some present-day state superintendents are still appointed by the governor (15),21 state superintendents are most often appointed by the SBE to serve as the director of the state education agency (22).22 Additionally, about one-quarter of all state 21 The governor of Nevada and Vermont, in making its State Superintendent appointment, must choose of a list of three nominees put forth by the SBE. 22 The state superintendent chosen by the Arkansas SBE must be approved by the governor. Additionally, the Oregon state superintendent is appointed by the Educational Investment Board, not the SBE. 56 superintendents are publically elected. Some state superintendents have a vote on the state board of education and in some cases may serve ass the president of the board and, thus, have an influence over the agenda of the SBE. Regardless of how a state superintendent obtains her power, much like a governor, she does not have authority to single-handedly enact formal policy in the form of statutes or administrative rules. Conclusion This chapter demonstrates that state legislatures, as part of their constitutional responsibility to provide public education, have long held authority to determine education policy. Moreover, SBEs also have the ability to play an important role in the development and enactment of education policy. However, for most of American history, state legislatures have deliberately chosen to reallocate policymaking power to more localized units of government, namely local school boards. Yet, in recent years, scholars have noted the increasing interest and involvement in education among state policymakers (Henig, 2013; Kirst & Wirt, 2009). This chapter affirms the extensive policymaking power that state legislatures and SBEs have over educational matters. While having this legally established policymaking power and being called upon by the American public to become more involved in education policymaking, very little is known about who those with state education policymaking are. Shifting to Part II of this dissertation, I begin my empirical analysis of the demographic characteristics of state education policymakers and whether they are descriptively representative of K12 public education constituents. 57 Chapter IV Who Gains State Education Policymaking Power? One manner in which citizen voices may enter into the political decisionmaking process is through the selection of representatives who mirror the experiences and outward manifestations of a group of people, also known as descriptive representation (Pitkin, 1967). Although descriptive representation by no means leads to policy responsiveness, descriptive representation can “forge bonds of trust” between public officials and constituents, enhancing the “feeling of inclusion” in the policymaking process (Mansbridge, 1999, p. 641). As Pamela Conover (1988) observed in her empirical analysis of differences in policymakers’ views of policies that ensure equitable opportunities for women, The way we think about social groups depends enormously on whether we are part of that group. Try as we might, the political sympathy that we feel for other groups is never quite the same as that which these groups feel for themselves or that which we feel for ourselves (p. 75). A large body of research has shown that descriptive representation does impact the policymaking process. For example, policymakers are more responsive to individuals with similar backgrounds: female policymakers are more likely to be responsive to female policy preferences; and minority policymakers are more likely to be responsive to minority policy preferences (Berkman & O’Connor, 1993; Clark, 2010; Mansbridge, 1999; Mezey, 1994; Skjeie, 1991; Strauss, 1998; Thomas, 1994). Particularly for minority populations, descriptive 58 representation has been shown to increase political engagement such as voting and contacting elected officials (Banducci, Donovan, & Karp, 2004; Barreto, 2007; Barreto, Segura, & Woods, 2004; Rocha, Tolbert, Bowen, & Clark, 2010), increase feelings of trust and reduced feelings of political alienation23 (Gay, 2001; Pantoja & Segura, 2003), and enhance “de facto legitimacy” by making citizens feel as though they themselves were present in political deliberations (Mansbridge, 1999, p. 650). Specific to education, descriptive representation of local school board members has been shown to result in the enactment of more equitable educational policies24 (Meier & England, 1984; Meier & Stewart, 1991). Extant literature reveals that state legislatures are rarely descriptively representative of the broader state population (Clark, 2010; Hardy-Fanta, Sierra, Lien, Pinderhughes, & Davis, 2005; C. Rosenthal, 1998). However, researchers have not yet examined the demographics of legislators serving on policy-specific committees or other state-level policy actors, such as SBE members. As such, we know very little about whether the demographic characteristics of statelevel education policymakers mirror those of their state population. Moreover, although a large body of research has examined the impacts that descriptive representation can have on the policymaking process (Berkman & O’Connor, 1993; Clark, 2010; Gay, 2001; Kathlene, 1994; Mansbridge, 1999; Mezey, 1994; Pantoja & Segura, 2003; Richardson & Freeman, 1995; Skjeie, 1991; Strauss, 1998; Thomas, 1994; Thomas & Welch, 1991), very little research has examined the conditions under which descriptive representation may occur (for exception, see Casellas, 23 Pantoja et al. (2003) focus on one dimension of political alienation: political normalness, defined as “the belief that government officials are violating long-standing norms or rules in an effort to serve some narrow interest (Finifter, 1970, p. 390).” 24 The education policies that Meier & England (1984) examine are (1) teacher hiring practices, with a focus on the hiring of black teacher and (2) “second generation discrimination” policies including the proportion of blacks “enrolled in college, assigned to vocational programs, suspended, dropping out of school, and assigned to special education classes, educable mentally retarded (EMR) classes, gifted programs, and enriched, honors, or advanced classes” (p. 397). 59 2009). Thus, in this chapter, I examine the following questions focused, first, on state education policymaker demographics and, second, descriptive representation: (1) Who gains power to make state education policy? a. How do demographic characteristics of SBE members and state legislators serving on education committee vary among states? b. Within what types of institutional structures do individuals with specific demographic characteristics gain state education policymaking power? (2) Do state education policymakers’ demographic characteristics reflect state population demographics? a. Within what types of institutional structures do the demographics of statelevel education policymakers align with the demographics of the state population? Epistemological Perspective and Hypotheses I approach my study of who gains state education policymaker power through the epistemological perspective of new institutionalism (March & Olsen, 1984). Central to the perspective of new institutionalism is the idea that “political democracy depends not only on economic and social conditions but also on the design of political institutions” (March & Olsen, 1984, p. 738). For example, although SBEs and education committees within state legislatures are arenas in which individual political motivations are played out, they are also “collections of standard operating procedures and structures that define and defend interests” (March & Olsen, 1984, p. 738). This chapter considers two structures and multiple sub-structures that may impact 60 who gains state education policymaker power. These avenues and structures are depicted in Figure 7. Due to the wide variation in SBE sub-structures across the United States, it is challenging to posit hypotheses of how the most broad institutional structures – SBEs and state legislatures – may impact who gains state education policymaking power. Therefore, I focus on the substructures of SBEs and state legislatures. In what follows, I draw upon research and literature from political science and public administration to explain the potential impacts of these institutional structures on who gains state education policymaking power. Figure 7 Institutional structures and sub-structures of state education governance Professionalism Institutional Structures State legislature Term limits Partisan Elected Non-partisan Ways of coming to power State board of education Governor only Appointed Governor + Legislative approval Statewide Legislature only Geographic representation Region-based 61 Sub-structures of SBEs. There are two primary avenues by which an individual can gain state education policymaking power on an SBE: election and appointment. Yet, as displayed in Appendix C, the structures and mechanisms through which an SBE member is elected or appointed vary from state to state. These structures may impact who can and does gain power on an SBE. Becoming an elected SBE member: Geographic representation. Eleven states rely on citizens of the state to elect some or all SBE members. In nine states, members are elected on a regional basis. For example, SBE members in Colorado are elected to represent the state’s seven U.S. Congressional districts. In Ohio, SBE members represent the territory of three contiguous Ohio senate districts. In two states (Michigan and Texas), SBE members are elected statewide. Finally, one state (Washington) requires both public and private school leadership to elect a number of individuals to its SBE. Research suggests that statewide elections favor white males. This is due in part because although “blacks comprise a higher percentage of the population in many large cities, they represent a much smaller share in the states” (Sonenshein, 1990, p. 222). Additionally, Key’s (1949) concept of “friends and neighbors” politics would suggest that regional elections allow for name recognition of candidates that may not be as strong of candidates in statewide elections. When a “home boy” is present, a voter may be much more included to support that candidate given his or her geographic similarities and exposure to each other (Key, 1949). H1A: When compared to states with statewide elected SBEs, SBEs in states with regionally elected SBEs will be more diverse. 62 Becoming an elected SBE member: Partisanship. Yet another point of variation among elected SBEs is partisanship. Six states (Louisiana, Nebraska, Nevada, Ohio, South Carolina, Utah) hold nonpartisan SBE elections and five states (Alabama, Colorado Kansas, Michigan, Texas) hold partisan SBE elections. The partisanship of SBE elections may have consequences for who ultimately gains the SBE member seat. For one, research suggests that having a candidate’s party identification on the ballot provides voters with an important cue that helps them make their voting choice (Aldrich, 1995; Downs, 1957; Rahn, 1993; Schaffner, Streb, & Wright 2001). However, nonpartisan elections may unfairly advantage wealthy candidates: candidates running without the advantage of being able to campaign for votes “on the basis of party identification will have to raise and spend much more money to achieve desired levels of name recognition” (Sherrill, 1998, para. 16). Thus, when SBE elections are non-partisan, those who gain state education policymaking power may be more likely to be wealthy individuals, incumbents, or individuals with name recognition (Schaffner, Streb, Wright, 2001). This is partially what the creators of American nonpartisan elections intended it to do—weaken the political parties that were closely tied to the system of spoils and encourage successful, wellknown citizens “to run for office who would never become candidates under the traditional method since they would be unwilling to become entangled in the ordinary processes of party politics” (Adrian, 1952, p. 771). Additionally, in states with straight ticket voting and partisan SBE elections (Alabama, Michigan, Texas), the “down ballot” effect may result in elected SBE members riding the coattails of a gubernatorial or presidential candidate of the same party (Weber, 1980; Hogan, 2005). The benefits and drawbacks of straight ticket voting have been widely debated. Proponents of straight ticket voting contend that it reduces voters’ cognitive burden and 63 increases electoral participation by speeding the voting process (Downs, 1957; Lupia, 1994: McDermott, 2005). The increase in electoral participation is particularly salient for minority groups: research has shown that the elimination of straight-party voting caused longer wait times at the ballot and more congestion that disproportionately harmed African American voters (Metzger, 2016). H1B: SBEs in states with partisan elections will be more racially diverse than those with non-partisan elections. Becoming an appointed SBE member: Getting noticed. States vary in the ways in which potential SBE appointees are considered. In some states, individuals interested in being nominated as an SBE member by the governor must complete an appointee application. For example, all gubernatorial appointees in the states of Iowa and Ohio must complete an appointment application that provides the governor with information about an applicant’s race, age, gender, disability status, education, employment and experience, citizenship status, and political party affiliation. The Iowa gubernatorial appointment application also requires applicants to describe why they are interested in serving on the SBE. In some states, however, the governor has complete discretion over his or her appointees. In this context, research suggests that appointees are likely to be white males (Riccucci & Saidel, 1997, 2001). And, given that Governor’s have an agenda to be carried out, Governor’s are likely to appoint likeminded individuals that align with his or her political party. Although research has not explored political alignment of governors and SBE appointments, this is often the case with gubernatorial judicial appointments (McLeod, 2012). As previously discussed, while statewide elections favor white males, regional elections may allow for greater diversity of elected SBE members. Given 64 that the majority of elected SBEs are regionally-based, I post that elected SBEs will be more diverse than appointed SBEs. H2A: When compared to states with appointed SBEs, the composition of elected SBEs will be more diverse. Appointers of SBE members. Research suggests that governors may engage in patronage, doling out appointments as rewards for political support (Tolchin & Tolchin, 2011). Therefore, in an effort reduce the politicization of SBE appointments, many states require legislative approval to provide a check on the appropriateness of governor appointees. In the majority of states (35), at least one body of the legislative branch is involved in the nomination or approval of some or all SBE members.25 Research specific to judicial appointments has shown that by having a separate body confirm gubernatorial appointments, the probability of cross-party appointments increases (McLeod, 2012). H2B: When compared to states with governor-only appointed SBEs, SBEs in states with appointed SBE members that must be approved by the legislature will be more diverse. Appointed SBE demographic equality. Another way that SBE appointments vary across states is related to restrictions on the characteristics of appointees. Nine states (Alaska, Delaware, Illinois, Indiana, Iowa, Missouri, Tennessee, West Virginia, and Wyoming) require SBEs be made up of a specific proportion of members from each political party. Two states (New Jersey and Tennessee) have gender equality requirements and two states have racial equality 25 SBE member appointees in one state (New York) requires only legislative approval. Appointees in 34 states require nomination by the governor and approval by the legislature. The governor has the sole authority to appoint all SBE members in Idaho, Indiana, Massachusetts, New Hampshire25, and North Dakota. A limited number of SBE members in South Carolina (1) and Nevada (3) are appointed by the governor. 65 requirements (Arkansas and Tennessee). Finally, some states require geographic representation of SBE members. For example, in Kentucky, four SBE members are appointed at-large and seven are appointed to represent the state’s seven Supreme Court districts. While no research has examined if such restrictions on the characteristics of individuals that a governor can appoint leads to equality of gender, race, and geographic representation, it seems likely that if these structures are in place, an SBE would be more diverse. H2C: States with SBE membership demographic equality requirements will be more diverse than those without such requirements. Sub-structures of education committees in state legislatures. Upon being elected to serve in a state legislature, an individual can gain substantial state education policymaking power by securing a seat on its chamber’s education committee. As an education committee member, an individual serves as one of a limited number of legislators charged with reviewing and deciding the fate of all bills referred to the committee. In essence, members of the education committee serve as the gatekeepers of education policies considered by the legislature (Denzau & MacKay, 1983). The institutional rules defining how a legislator can gain power to serve on an education committee vary from state and state, and between House and Senate chambers. In most states, the leader of the majority party in each chamber has extensive control over the education committee assignment process. For example, when it comes to House or Assembly education committee members, 30 states provide the Speaker of the House (who is the leader of the majority power) with the authority to make education committee appointments. Nine states provide the Speaker of the House with the same authority; however he or she must also consult 66 with the minority leader about minority party committee appointments. In contrast, five states allow the majority party leader to make majority party committee appointments and the minority leader to make minority party committee assignments. In the remaining six states, committee members are appointed by a Committee on Committees, a House District Caucus, seniority, or a mixture of these appointment methods. Similar methods and variations can be seen in the selection of Senate education committee assignments (Erickson, 2016a, 2016b). Some researchers have examined factors that impact the selection of committee chairmanships. For example, research has shown that because legislators tend to view their African-American colleagues as less effective than their non-black colleagues (Haynie, 2002), appointment methods that allow for the selection of committee chairs, as opposed to election, may be advantageous to African Americans (Orey, Overby, & Larimer, 2007). Additionally, research has shown that legislators who align themselves with the winning candidate in the speakership race are more likely to attain committee chairmanships (Capell, 1989). However, very little research has examined factors that impact the selection of all committee members— not just the chair. As such, I turn to research on the ways in which institutional structures of state legislatures may impact who gains access to a position of power in a state legislature and, consequently, who make up the available pool of education committee member candidates. Term limits. Theory suggests that term limits affect the kinds of individuals who run for office. For example, legislatures with term limits may be more attractive to individuals who do not want to make a career out of politics, opening up political office to individuals without “conventional political backgrounds” (Petracca, 1991). Building upon this line of thinking has 67 led others to suggest that term limits will attract citizens of more diverse backgrounds to legislative office (Fund, 1992). H3A: When compared to states without term limits, the composition of education committees in states with term limits will be more diverse. Professionalism. One of the key measures of legislative professionalism is the amount of time legislators are required to commit to serving in the state legislature. Research suggests that more professional legislatures favor Democratic candidates, and thus a more racially diverse legislature (Fiorina, 1994). Fiorina (1994) explains the potential logic behind this finding: in states with legislatures that meet only a few months a year, Republicans, who are a “higherincome, more private-sector party,” are advantaged because they are more likely to have a career and the financial flexibility to absent themselves from their principal occupation for a short period of time (Fiorina, 1994, p. 307). On the other hand, Democratic candidates, who are a “lower-income, more public-sector party” are disadvantaged because, as wage-earners, they must “forgo their incomes when they take leave for legislative service—if leave is even feasible” (Fiorina, 1994, p. 307). In contrast, with a more professional, full-time legislature, legislators must sacrifice outside careers for legislative office (Fiorina, 1994). As a result, Republicans who have “lucrative private-sector careers,” are less likely to enter legislative service; and, for Democratic candidates, legislative service “becomes an attractive alternative career, probably better compensated and more highly regarded than their present careers” (Fiorina, 1994, p. 307). H3B: As a state’s legislative professionalism increases, so too does the diversity of its education committee. 68 Data To examine the demographic characteristics of state education policymakers, descriptive representation, and whether institutional structures of state-level education policymaking bodies impact descriptive representation, I draw upon three data sets: demographic data of state education policymakers, state-level population demographic data, and measures and indicators of state education policymaking bodies’ institutional structures. State education policymaker demographics. I focus my analysis on state legislators who serve on committees concentrated on education policy issues26 and SBE members in office. These two populations represent state-level education policymakers who have the ability to review, accept, reject, amend, and enact formal rules, regulations, and laws related to the provision of education in their state. State education policymaker demographic data, which I collected in the spring of 2016, includes individual-level data of state education policymakers’ gender, race and ethnicity, education level, profession, and political party affiliation. Table 1 provides descriptive statistics of state education policymaker demographic characteristics. I also aggregated these data at the state level to develop statewide demographic dataset that could be used to examine descriptive representation between state education policymakers and state populations. Because research has already established that legislatures are not descriptively representative of their state populations (A. Rosenthal, 1998), I do not aggregate the political party data for education committee members at the state level. Table 5 provides summary statistics of the state-level aggregate data. Methods for the collection of this data are located in Appendix D. 26 State legislators serving on committees focused on the appropriation of funding for education are not included in this analysis. 69 State population demographics. To examine whether state education policymaker demographic characteristics mirror the demographic characteristics of a state’s population, I draw upon a secondary dataset—the 2014 Cooperative Congressional Election Survey (CCES) (Schaffner & Ansolabehere, 2015)— that includes data related to state population demographic characteristics, including, gender, race, Latino origins, education level, and political party. The CCES includes demographic data from a random sample of 56,200 adults across all fifty states. State samples range from a low of slightly fewer than 100 to a high of more than 5,500 adults. I aggregated individual survey respondent data at the state level to create state-level state population demographic data. Appendix E provides details on the CCES survey questions that I utilized and how the CCES survey answer categories mapped onto of the aggregated, state-level demographic characteristic data I collected for state education policymakers. Indicators of state education policymaking bodies’ institutional structures. To examine the relationship between institutional structures of state-level education policymaking bodies and (a) demographic characteristics of state education policymakers and (b) the extent to which state education policymakers are descriptively representative of state populations, I develop indicators of state education policymaking body structures. The SBE indicators include binary indicators of geographic representation of an SBE member (statewide, region-based), the way in which an SBE member gains power (appointed, elected)27, and, for elected SBE members, the type of election (partisan, non-partisan). 27 When conducting analyses at the state-level, I had to make a decision on how to classify five states that have both elected and appointed SBE members (Louisiana, Ohio, Nevada, South Carolina, and Washington). I chose to code states according to the selection method by which the majority of the state’s SBE members were selected. Louisiana (8 elected, 3 appointed), Ohio (11 elected, 8 appointed), Nevada (4 elected, 3 appointed), and South Carolina (16 elected, 1 appointed) were coded as states with elected SBEs. Washington (6 elected, 7 appointed) was coded as a state with an appointed SBE. 70 The data also includes two legislature-specific structure indicators: a binary variable related to term limits and a continuous variable indicating the level of legislative professionalism. For the latter legislature-specific indicator, I utilize Squire’s (2007) measure of legislative professionalism that incorporates three state legislative components of salary/benefits, time demands of service, and staff and resources to create a 0-to-1 index of legislative professionalism. The index number for each state is calculated by dividing the score on each of the three legislative components by the parallel score for Congress for the same year. The scores of the three components are summed and divided by three. The calculation represents a score for each state that can be interpreted as “the percentage of professionalism that [a state’s] legislature had compared to Congress that year” (Squire, 2007, p. 219). I utilize the most recent Squire’s index, which provides a state-level measure of legislative professionalization compared to Congress in 2003. The 2003 Squire’s index ranges from a minimum of 0.027 in New Hampshire to a maximum of 0.626 in California. Methods Focusing on research question one, I begin my analyses of who gains state education policymaking power by exploring descriptive statistics of the demographic characteristics of state education policymakers, both as a whole and broken down by SBEs and education committees. I then utilize correlation analyses, chi-square tests of independence, and analysis of variance (ANOVA) to examine variations in demographic characteristics of SBE members and legislators serving on education committees. I also examine how state education policymaker demographics vary among states. 71 Second, to explore relationships between the institutional structures of state-level education policymaking bodies and who gains power to make state-level education policy, I conduct traditional and point biserial correlations. While point biserial correlations are utilized when correlating continuous demographic variables (i.e., percent of state education policymakers that make up a particular demographic group) with dichotomous institutional structure variables (i.e., SBE elected28, SBE nonpartisan29, SBE region-based, legislative term limits), traditional correlations are utilized when correlating continuous demographic variables with continuous institutional structure variables (i.e., legislative professionalism) (Sheskin, 2011). Finally, to examine the extent to which a state’s education policymaker demographics mirror those of the state population, I conduct correlation analyses of state-level education policymaking body demographics and population demographics. I then examine the relationships between institutional structures of state-level education policymaking bodies and descriptive representation using two multivariate multiple regression models: !"#(!"# − !" − !"#$%&'(") !"#$%& !"#$%!! ) = !!" + !!" !"!#$!!!! + !!" !"#$%& − !"#$%!! + !!" !"!#$%&'($!!! + !!! (1) !"#(!"#$%&'() !"##$%%&& − !" − !"#$%𝑎!"#$ !"#$%& !"#$%!! ) = !!" + !!" !"#$%&&'#()*'&+!! + !!" !"#$ !"#"$%!! + !!! (2) 28 When conducting analyses at the state-level, I had to make a decision on how to classify five states that have both elected and appointed SBE members (Louisiana, Ohio, Nevada, South Carolina, and Washington). I chose to code states according to the selection method by which the majority of the state’s SBE members were selected. Louisiana (8 elected, 3 appointed), Ohio (11 elected, 8 appointed), Nevada (4 elected, 3 appointed), and South Carolina (16 elected, 1 appointed) were coded as states with elected SBEs. Washington (6 elected, 7 appointed) was coded as a state with an appointed SBE. 29 It should be noted that there are a limited number of elected SBEs (10) and, as a result, a small number of nonpartisan SBEs (4). Therefore, the point biserial correlations for these variables should be interpreted with caution. 72 In model 1, the dependent variables are nine SBE-to-population parity ratios. The independent variables in this model are three SBE structures: selection methods (elected, appointed), geographic representation (region-based, at large), and, for elected SBEs, an indicator of whether the election is partisan or non-partisan. The dependent variables in model 2 are eight education committee-to-population parity ratios. The independent variables include a measure of legislative professionalism (continuous variable) and an indicator of term limits (binary variable). Multivariate multiple regression is appropriate for these analyses because it allows for the modeling between more than one independent variable and more than one dependent variable (Dattalo, 2013; Hartung & Knapp, 2005). The parity ratios were calculated by dividing the (a) the percentage of a particular subpopulation in either the SBE or education committee by (b) the percentage of a particular subpopulation in the entire state, as reported by CCES survey. For example, the SBE-to-population gender parity ratio is calculated by taking the percent of male SBE members in a state and dividing it by the percent of males in the state. If the ratio is less than one, the state’s population has a larger percentage of males than its SBE. If the ratio is greater than one, the state’s population has a smaller percentage of males than its SBE. Distributions of all parity ratios were examined (available in Appendix F) and, as a result of skewness in the distributions of parity ratios, I utilized a log transformation of parity ratios. Legislature-to-population parity ratios developed for analysis include: gender, race/ethnicity, 4-year degree, and advanced degree. SBEto-population parity ratios include those previously listed, as well as a political party affiliation parity ratio. Due to collinearity, I did not include parity ratios for Latino origin and lower levels of education (i.e., high school diploma, some college, or 2-year degree) in the models. The independent variables in the SBE model include SBE institutional structure variables (i.e., SBE 73 elected, SBE nonpartisan, SBE region-based). The independent variables in the education committee model include legislature institutional structure variables (i.e., professionalism, term limits). The previously described methods are appropriate for this initial, exploratory phase of research that aims to describe, show, and summarize data in a meaningful way such that patterns might emerge from the data (Tashakori & Teddlie, 2010). Nonetheless, these generally descriptive statistics do not allow me to make any conclusions beyond the data analyzed; instead, they may be used to simply describe the demographics and descriptive representation of state education policymakers, how those demographics vary amongst states, and the relationships between (a) policymaker demographics and (b) descriptive representation and the institutional structures of state-level education policymaking bodies. Results: Individual Descriptive Statistics, and State and Institutional Variation My first research question asks: who gains power to make state education policy; and, my first sub-questions asks how demographic characteristics of SBE members and state legislators serving on education committee vary among states. To answer these questions, I explore basic descriptive characteristics of state education policymakers both as a whole and broken down by the two state-level education policymaking institutions: SBEs and education committees in state legislatures. I also provide descriptive statistics of how the demographic characteristics across states. Gender and race/ethnicity. In recent years, women and minorities have begun to gain political power at the local, state, and national levels. Between 1995 and 2015, the number of 74 women in Congress nearly doubled (Manning, Brudnick, & Shogan, 2015); there are six times as many women serving in state legislatures today than in 1971 (Kurtz, 2015); and, since the 1970s, the number of female local school board members has nearly doubled (MacManus, Bullock, Pagett, & Penberty, 2006). Additionally, there are three times as many blacks serving in both state legislatures and in Congress today than in 1971 (Kurtz, 2015; Manning & Shogan, 2012). At the local level, black representation on school boards more than doubled between 1973 and 2001 (Joint Center for Political and Economic Studies, 2002). Finally, over the past two decades, the number of Latinos in state legislatures has nearly doubled; and, their presence in Congress has increased by more than 50 percent (Committee on House Administration of the U.S. House of Representatives, 2013; NALEO, 1996; NALEO, 2016). At the local level, Latino representation on local school boards increased 89 percent between 1996 and 2016 (NALEO, 1996; NALEO, 2016). While we know quite a bit about the general and racial demographics of both education-specific policymaking at the local level (i.e., school boards) and general policymaking bodies at the state and federal level (e.g., state legislatures and U.S. Congress), we know very little about education-specific, state-level policymakers. What are the racial and gender demographics of state education policymakers? The proportion of women legislators serving on education committees (26 percent) is nearly identical to that of state legislatures as a whole (25 percent, Kurtz, 2016). However, the proportion of non-white education committee members in state legislatures (17 percent) is marginally larger than the proportion of non-white state legislators as a whole (13 percent, Kurtz, 2015). In contrast, the proportion of Latino state legislators serving on education committees (4 percent) is slightly smaller than the proportion of Latino state legislators as a whole (5 percent, 75 Kurtz, 2015). Nonetheless, as shown in Figure 8, white males make up a considerable majority of education committee members. Figure 8 Demographics of state legislators serving on education committees non-white, male 10% white, female 19% Latino, male 2% non-white, female 7% Latino, female 2% nonLatino female 24% nonLatino male 72% white, male 64% With regard to SBEs, the proportion of non-white and Latino SBE members is nearly identical to that of education committee members in the state legislature (see Table 1). In fact, Chi-square tests of independence reveal that there are no statistically significant differences between state legislators serving on education committees and SBE members with regard to race or Latino origin. While SBEs are largely white and non-Latino, they have a significantly larger female representation than education committees in state legislatures (Figure 9). Chi-square tests of independence confirm this sizeable difference, revealing that there are statistically significantly more females serving on SBEs than on education committees in state legislatures. 76 Figure 9 Demographics of state board of education members non-white, female 9% non-white, male 10% Latino, male 2% Latino, female 3% white, male 43% non-Latino female 44% white, female 38% non-Latino male 51% Table 1 Demographics of state education policymakers Total Gender Male Female Race White Non-White Missing Latino Origin Non-Latino Latino Missing State Education Policymakers Legislators Serving on Education Committees SBE Members 1,740 1,263 477 1,184 (68.1%) 556 (32.0%) 932 (73.9%) 330 (26.2%) 252 (52.8%) 226 (47.2%) 1,437 (82.6%) 302 (17.4%) 1 (0.1%) 1050 (83.2%) 212 (16.8%) 0 (0.0%) 386 (80.9%)30 90 (18.9%) 1 (0.2%) 1,668 (95.9%) 71 (4.1%) 1 (0.1%) 1212 (96.0%) 50 (4.0%) 0 (0.0%) 455 (95.4%) 21 (4.4%) 1 (0.2%) 30 One SBE member’s race and Latino origin were unidentifiable. 77 Chi-Square Tests of Independence between SBE members and legislators serving on education committees p-value !! 70.4 <0.01 1.07 0.301 0.18 0.673 Table 1 (cont’d) Education Level No information available (0) High school diploma (1) Some college, no degree (2) 2-year degree (3) 4-year degree (4) Master's degree (5)32 Doctorate or Professional degree (6) 33 Law Medicine Philosophy Education Profession34 Education Business/Entrepreneurship Health Science/Engineering Law Lobbying/ Consulting Government/ Politics Nonprofit/Advocacy/Interest Group Political Party Affiliation Republican Democrat Other/Not Available 73 (4.2%) 38 (2.2%) 90 (5.2%) 46 (2.6%) 559 (32.1%) 532 (30.6%) 402 (23.1%) 43 (3.4%) 35 (2.8%) 81 (6.4%) 42 (3.3%) 447 (35.4%) 361 (28.6%) 253 (20.1%) 30 (6.3%) 3 (0.6%) 9 (1.9%) 4 (0.8%) 112 (23.5%) 170 (35.6%) 149 (31.2%) 105.4031 7.15 7.45 14.48 8.33 22.62 8.07 24.39 213 (12.2%) 27 (1.6%) 124 (7.1%) 42 (2.4%) 164 (13.0%) 22 (1.7%) 46 (3.7%) 22 (1.7%) 49 (10.3%) 5 (1.1%) 78 (16.4%) 20 (4.2%) 2.39 1.09 84.41 8.81 0.12 0.30 <0.01 <0.01 538 (30.9%) 691 (39.7%) 83 (4.8%) 4 (0.2%) 191 (11.0%) 29 (1.7%) 210 (12.1%) 137 (7.9%) 305 (24.2%) 565 (44.8%) 69 (5.5%) 4 (0.3%) 148 (11.7%) 20 (1.6%) 170 (13.5%) 85 (6.7%) 233 (48.9%) 126 (26.4%) 13 (2.7%) 0 (0.0%) 43 (9.0%) 9 (1.9%) 40 (8.4%) 52 (10.9%) 99.76 48.28 5.73 1.51 2.54 0.20 8.31 8.40 <0.01 <0.01 0.02 0.22 0.11 0.66 <0.01 <0.01 0.01 0.01 934 (53.7%) 727 (41.8%) 79 (4.5%) 733 (58.1%) 527 (41.8%) 2 (0.2%) 201 (42.1%) 200 (41.9%) 76 (15.9%) <0.01 0.01 0.01 <0.01 <0.01 <0.01 <0.01 <0.01 31 Prior to conducting an ANOVA test to examine whether these differences in education among state legislators serving on education committees and SBE members are statistically significant, I examined whether the unequal sample sizes of state legislators and SBE members warranted the use of statistical test that was more robust to heteroskedasticity than one-way ANOVA. The Monte Carlo simulation (simulation number=5000) revealed that the unequal sample sizes did not significantly impact the robustness of the one-way ANOVA (ANOVA reported p=0.000; simulated p=0.000). 32 Master’s degree also includes non-doctoral postgraduate degree such as Education Specialist degree 33 Doctorate and professional degrees also include Doctor of Philosophy, Doctor of Education, Medical Doctor, Doctor of Osteopathy, Doctor of Accounting, Doctor of Business Administration, Doctor of Chiropractic Medicine, Doctor of Dental Surgery, Doctor of Dental Medicine, Doctor of Public Health, Doctor of Strategic Leadership, Doctor of Veterinary Medicine, Doctor of Optometry, and Juris Doctor. Sub-categories may not add to 100 percent due to (a) rounding error, (b) individuals with multiple doctorates were counted in each sub-category for which they held a degree. 34 Categories may not add to 100 percent due to (a) rounding error, (b) individuals with multiple professions were counted in each category for which they had a career. 78 Education levels and professional experience. The educational and professional experiences of state education policymakers are also important demographic characteristics to understand. Research suggests that political preferences differ between educational groups (Bovens & Wille, 2009) and that the least educated are substantively less represented in political offices (Aaldering, 2016). Although some may argue that it is undesirable to have lessereducated state policymakers, under-representation of any group may be undesirable in a representative democracy. For example, a citizen who desires greater emphasis be placed on policies that promote career- and technical education may desire to have in place some state education policymakers with a 2-year technical degree. Research has also shown that the individuals’ past experiences influence decisionmaking (Juliusson, Karlsson, & Gärling, 2005). In a political context, one could imagine that a former teacher-turned-state education policymaker would bring different experiences and knowledge to the policymaking process than a small business owner, doctor, or lawyer. Again, a citizen who desires greater emphasis be placed on policies that promote career- and technical education may desire to have in place some state education policymakers who have experience as electricians, plumbers, or mechanics. So, what are the educational and professional backgrounds characteristics of state legislators serving on education committees and SBE members? Education levels. When compared to the American citizenry – of which 33 percent of hold a Bachelor’s degree or higher (U.S. Census Bureau CPS, 2015b) – state education policymakers are a largely highly educated group of people. Among all state education policymakers, more than 85 percent hold at least a Bachelor’s degree. Even more, nearly one in four state education policymakers hold a doctorate or professional degree, with the majority of 79 these advanced degree holders being lawyers. When educational attainment is broken down by state education policymaking body, I find that a larger proportion of SBE members (67.0 percent) have Master’s and advanced degrees than legislators (48.7 percent) serving on education committees. Additionally, the average education level of SBE members (µ=4.31, σ=1.44) is significantly higher than that of state legislators serving on education committees (µ=4.67, σ=1.52). I also examined whether there are differences among legislators and SBE members with regard to the interaction between (a) education level and (b) gender, race, or Latino origin. I find a statistically significant, albeit modest, interaction between the effects of gender and education level on education policymaking body, F(6, 1728)=1.86, p=0.08. Simple main effects analyses revealed that females with a Bachelor’s degree or higher were significantly more likely to serve on SBEs than education committees (p<0.01 for Bachelor’s, Master’s, and doctorate or professional degrees); however, there were no statistically significant differences between education policymaking bodies for males and females with less than a college degree (high school diploma p=0.69; some college, no degree p=0.30). Additionally, no interaction effects were found between race or Latino identity and education level on education policymaking body (race: F(6, 1727)=0.84, p=0.54; Latino: F(6, 1727)=1.51, p=0.17). Professional experience. Individuals with business/entrepreneurship backgrounds represent the largest proportion of state education policymakers (39.7 percent). Individuals working in the field of education make up the second largest proportion of state education policymakers (30.9 percent), followed by individuals working in government/politics (12.1 percent) lawyers (11.0 percent). Individuals working in the areas of health, science/engineering, 80 lobbying/consulting, and nonprofit/advocacy/interest group make up the remaining 15 percent of state education policymakers. When broken down by state education policymaking body, I find that that SBEs are made up of significantly fewer individuals with careers in government/politics and business and entrepreneurship. In contrast, SBEs are made up of a significantly more individuals with careers in education and nonprofit/advocacy/interest group organizations. There were no statistically significant differences among state legislators and SBE members with careers in health, science/engineering, law, or lobbying/consulting. State education policymaker demographic characteristic intercorrelations. To examine whether some policymaker demographics are associated with others, I examine intercorrelations. As shown in Table 2, non-white state education policymakers are more highly educated than white state education policymakers. Moreover, both female and non-white state education policymakers are more likely to be professional educators. To understand the magnitude of the correlations, the coefficients listed in Table 2 can be interpreted based on explained variation. For example, by squaring the correlation coefficient, I find that 3.6 percent of the variation between male and female state education policymakers is explained by the variation between those who are and are not professional educators. These findings suggest that, when compared to the average white or male state education policymaker, non-whites and females looking to gain power as a state education policymaker may have to acquire additional credentials, namely a higher education or experience as a professional educator. 81 Table 2 Correlations between state education policymaker demographics SBE Female Non-white Latino Professional educator SBE 1.00 Female 0.20*** 1.00 Non-white 0.03 0.10*** 1.00 Latino 0.01 0.08*** 0.43*** 1.00 Professional educator 0.24*** 0.19*** 0.06** 0.01 1.00 Degree category 0.11*** 0.02 0.08*** 0.01 0.24*** Degree category 1.00 ***p<0.01; **p<0.05; *p<0.10 Correlation analyses using the aggregate state level also provide insight into relationships between descriptive characteristics of education legislatures and SBEs (Tables 3 and 4). For example, as the percentage of male education policymakers – both education committee members and SBE members – increases, the percentage of with experience as educators decreases. Additionally, as the percentage of male legislators serving on education committees increases, the percentage of legislators serving on education committees with a Master’s degree declines. Finally, as the percentage of white SBE members increases, the percentage of SBE members with a 4-year degree increases but the percentage of SBE members with an advanced degree actually decreases. These finding suggests that in order to be elected or appointed to an SBE, non-white individuals, on average, must be more highly educated than white individuals. 82 Table 3 Correlations between state board of education demographic makeup % Male % White % Latino % Prof. educator % Dem % HS grad % Some college, no degree % 2year degree % 4-year degree % Master’s degree % Male 1.00 % White 0.11 1.00 % Latino -0.38*** -0.59*** 1.00 % Prof. educator % Democrat -0.29** -0.16 0.12 1.00 -0.08 -0.55*** 0.29** 0.01 1.00 % HS grad 0.03 0.28** -0.12 -0.10 -0.02 1.00 % Some college, no degree -0.01 0.11 0.07 -0.05 -0.20 0.57*** 1.00 % 2-year degree -0.20 -0.07 0.001 -0.07 -0.02 0.02 0.01 1.00 % 4-year degree 0.01 0.39*** -0.32** -0.31** -0.08 -0.08 -0.03 0.001 1.00 % Master’s degree 0.10 -0.06 0.14 0.09 0.06 -0.04 -0.07 0.03 -0.49*** 1.00 % Doctorate or Prof. Degree -0.07 -0.43*** 0.25* 0.27* 0.21 -0.19 -0.26* -0.22* -0.46*** -0.33** ***p<0.01; **p<0.05; *p<0.10 83 % Doctorate or Prof. Degree 1.00 Table 4 Correlations between education committee demographic makeup % Male % White % Latino % Prof. educator % HS grad % Some college, no degree % 2-year degree % 4-year degree % Master’s degree % Male 1.00 % White 0.16 1.00 % Latino -0.27** -0.59*** 1.00 % Prof. educator -0.28** -0.05 0.12 1.00 % HS grad 0.20 0.26* -0.23* -0.14 1.00 % Some college, no degree 0.16 0.21* -0.05 -0.12 -0.03 1.00 % 2-year degree -0.11 -0.04 0.16 -0.03 0.22* -0.04 1.00 % 4-year degree 0.25* -0.07 -0.12 0.03 -0.10 -0.22* -0.13 1.00 % Master’s degree -0.37*** 0.12 0.14 0.22* -0.13 -0.04 0.11 -0.49*** 1.00 % Doctorate or Prof. Degree 0.06 -0.18 0.03 -0.18 -0.20 -0.12 -0.37*** -0.37*** -0.34** ***p<0.01; **p<0.05; *p<0.10 84 % Doctorate or Prof. Degree 1.00 Political party affiliation. As detailed in Appendix C, there is much variation in how an individual gains power on an SBE. While some SBE members are elected, others are appointed. In states with elected boards, some SBE members run in partisan elections, while others are nonpartisan. Additionally, some SBE members must run on a statewide ballot, while others are elected by constituents within a geographically constrained region. In states with appointed SBE members, some provide the governor with the power to select a completely new SBE upon being elected, while others require legislative approval. As a result of these variations, SBE member political party affiliations may be quite different than those of the state legislature and the state population. On the whole, however, I find that SBE member political party affiliations are relatively similar to those of the population: whereas 44 percent of Americans leaned Republican and 45 percent Democrat in the January 2016 Gallup poll, 42 percent of SBE members leaned Republican and 42 percent leaned Democrat. The political party makeup of education committees is proportional to that of the state legislature as a whole. Research has already established that legislatures are not descriptively representative of their state populations (C. Rosenthal, 1998). As such, predictably, I find education committees are also not descriptively representative of their state population. Whereas 44 percent of Americans leaned Republican and 45 percent Democrat in the January 2016 Gallup poll, 58 percent of education committee members were Republican and 41 percent were Democrat. SBE member campaign donations. Some SBE members are also contributors to political party committees and gubernatorial campaigns. Of all SBE members with voting power in 2016, 31 percent donated to political party committees or gubernatorial campaigns. In total, 85 donations reached more than $1 million over the past decade. The average amount of money donated to both political party committee and gubernatorial campaigns is just under $8,000 per SBE member over the past decade. The majority of these donations go to political party committees: while the average amount donated by an SBE member to a gubernatorial campaign was $1,900, the average amount donated to political party committees was nearly $6,000. It should be noted that these data should be interpreted with caution given that the length of time for which campaign contribution data was available varied from state to state and the variation in state campaign finance laws that can make political action committee money difficult to track. Results: Variations in demographics of state education policymakers by state As shown in Figures 10 and 11, the gender composition of both SBEs and state legislatures varies widely from state to state. With regard to SBEs, females comprise the majority of SBEs in states such as California, Rhode Island, Colorado, and Michigan, while males dominate SBEs in states such as New Jersey, Arizona, and Missouri. In contrast, male legislators dominate education committees in nearly all fifty states. Figures 12 and 13 illustrate variation in racial composition among SBE members and state legislators serving on education committees, showing that nearly all education committees and SBEs are made up of majority white policymakers. 86 Figure 10 State board of education gender demographics, by state 87 Figure 11 State legislators serving on education committees gender demographics, by state 88 Figure 12 State board of education racial demographics, by state 89 Figure 13 State legislators serving on education committees racial demographics. by state 90 State education policymakers’ education levels and professions also vary across states. For example, while some states have no education legislators or SBE members with an advanced degree, others have 50 percent or more. Additionally, whereas fewer than 10 percent of some states’ education legislators or SBE members are professional educators, more than 50 percent of education legislators and SBE members in other states are professional educators. Table 5 provides detailed summary statistics of these aggregate demographic characteristics of state education policymakers. Finally, as shown in Figure 14, there is substantial state-to-state variation in SBE members’ political party affiliations: while some states are made up of individuals of one political party, other states are quite evenly split. When it comes to SBE member campaign donations to gubernatorial and political party campaign committees, there is relatively little variation among states (see Figure 15): contributions by SBE members over the past decade are less than $30,000 in most states. However, SBE members in the states of Missouri, New York, Indiana and New Jersey contribute substantially larger sums of money to gubernatorial and political party campaign committees than those in most other states. 91 Table 5 State-level summary statistics of state education policymaking body demographics State Education Policymaking Bodies Male White Non-Latino Education Level High school diploma Some college, no degree 2-year degree 4-year degree Master's degree Doctorate or professional degree Professional Educator Republican State Legislative Education Committees N Mean St.Dev. Min. Max. N Mean St.Dev. Min. Max. N Mean St.Dev. Min. Max. 47 47 47 0.63 0.82 0.04 0.09 0.13 0.06 0.41 0.38 0 0.83 1.00 0.27 50 50 50 0.73 0.83 0.05 0.11 0.16 0.09 0.47 0.20 0 0.94 1.00 0.45 47 47 47 0.52 0.81 0.05 0.18 0.15 0.07 0.14 0.50 0 1.00 1.00 0.30 47 47 47 47 47 0.02 0.04 0.02 0.29 0.32 0.03 0.04 0.03 0.10 0.09 0 0 0 0.04 0.15 0.11 0.18 0.11 0.48 0.58 50 50 50 50 50 0.03 0.06 0.04 0.36 0.29 0.05 0.06 0.04 0.12 0.10 0 0 0 0.08 0.11 0.21 0.26 0.21 0.64 0.50 47 47 47 47 47 0.01 0.02 0.01 0.23 0.36 0.03 0.05 0.03 0.15 0.15 0 0 0 0 0.11 0.14 0.29 0.11 0.56 0.73 47 0.26 0.09 0.08 0.53 50 0.19 0.12 0 0.50 47 0.32 0.15 0 0.65 47 0.37 0.10 0.16 0.62 50 0.25 0.10 0.08 0.44 47 45 0.49 0.52 0.18 0.34 0.13 0 0.82 1.00 Note: Percentages may not add to 100 percent due to rounding. State Boards of Education 92 Figure 14 State board of education political party affiliation demographics, by state West Virginia Oregon New York New Hampshire Massachusetts Hawaii Arkansas Washington Nevada California Montana Delaware Maryland Connecticut Rhode Island Michigan Kentucky Illinois Vermont Ohio Missouri South Carolina North Carolna Virginia Colorado Wyoming Pennsylvania Arizona Maine Indiana Texas Tennessee Alabama New Jersey Utah Nebraska Louisiana Iowa Kansas Wisconsin South Dakota Oklahoma North Dakota New Mexico Mississippi Minnesota Idaho Georgia Florida Alaska 0 .2 .4 .6 .8 State Board of Education Members - Percent Democrat 93 1 Figure 15 State board of education members’ political campaign contributions, by state Missouri New York Indiana New Jersey Michigan Massachusetts Alabama West Virginia Georgia Colorado South Dakota Kentucky Idaho Arizona California Hawaii North Carolina Montana Rhode Island Illinois Kansas Washington Florida Vermont Pennsylvania Massachusets Maine Mississippi Ohio Alaska South Carolina Iowa Maryland Washignton Oklahoma Connecticut Nevada Delaware 0 20,000 40,000 60,000 80,000 Average Total Contribution to Political Party Committee or State Governor Results: Institutional structures and state education policymaker demographics The second sub-question to my overarching question of who gains power to make state education policy asks: under what institutional conditions do individuals with specific demographic characteristics gain state education policymaking power? I find that a number of institutional structures of both state legislatures and SBEs impact the demographic makeup of state education policymaking bodies (Table 3). SBE institutional structures. With regard to SBE institutional characteristics, as displayed in Table 6, I find that appointed SBEs are associated with a greater proportion of females serving on the SBE. In particular, 15 percent of the variation between SBE member gender can be explained by variation in this selection method. Additionally, 12 percent of the 94 variation in appointed SBE member race/ethnicity can be explained by the variation in appointee approval rules: SBEs that require legislative approval of governor appointees are associated with a greater proportion of non-whites serving on the SBE. Non-partisan elected SBEs are associated with a greater proportion of whites serving on the SBE. Moreover, rules that restrict the pool from which appointed SBE members can be selected (i.e., demographic equality requirements) are actually associated with a greater proportion of males serving on the SBE. However, states with demographic equality requirements are also associated with a greater proportion of members with a 2-year degree and fewer educational professionals. Finally, regionally-based SBE elections are associated with SBEs comprised of fewer Latinos and members with 2-year degrees. These findings partially concur with the hypothesis that SBEs in states with partisan elections will be more racially diverse than those with non-partisan elections (H1B); the hypothesis that elected SBEs will be more diverse than appointed SBEs (H2A); the hypotheses that appointed SBEs that require legislative approval will be more diverse than SBEs with sole governor appointments (H2B). These findings provide no support for the hypothesis that SBEs in states with regionally elected SBEs will be more diverse than those with statewide elected SBEs (H1A). Finally, these findings negate the hypothesis that SBE membership demographic equality requirements will be more diverse than those without such requirements (H2C). Legislative institutional structures. With regard to legislative institutional structures, I find that while legislative professionalism has implications for education committee gender and racial makeup, term limits have implications for the diversity of education levels of education committee members serving on the education committee (see Table 7). In particular, education 95 committee members in more professional legislatures are associated with a greater proportion of female and non-white members. Additionally, legislatures with term limits are associated with a greater proportion of education committee members with Bachelor’s degrees and fewer with a two-year degree. These findings provide some evidence that states with more professional legislatures and with term limits are more diverse (H3, H4). Table 6 Correlations between institutional characteristics of SBEs and demographic makeup of SBEs 0.16 Legislative Approval of Gov. Appt. -0.03 Demographic equality requirement 0.36*** 0.30** 0.18 -0.34** -0.05 0.05 -0.14 -0.22* 0.21 0.05 % HS diploma -0.03 0.07 0.20 -0.30* 0.11 % Some college 0.14 0.16 -0.05 0.09 0.30** % 2-year degree -0.03 0.08 -0.25* 0.10 0.07 % 4-year degree 0.24 0.17 0.07 -0.14 -0.03 % Master’s degree -0.06 0.02 0.01 0.16 0.03 % Advanced degree -0.22* -0.20 -0.03 -0.01 -0.09 % Professional Educators -0.09 -0.08 0.17 0.03 -0.24* % Democrat -0.19 -0.21 -0.10 -0.09 -0.04 Elected Nonpartisan Region-based % Male -0.39*** -0.19 % White 0.22* % Latino ***p<0.01; **p<0.05; *p<0.10 96 Table 7 Correlations between institutional characteristics of state legislatures and demographic makeup of education committee members Term Limits Professionalism % Male -0.05 -0.29** % White 0.10 -0.30** % Latino -0.04 0.23* % HS diploma -0.09 -0.18 % Some college -0.18 -0.19 % 2-year degree -0.31** -0.10 % 4-year degree 0.29** -0.07 % Master’s degree -0.07 0.18 % Advanced degree 0.06 0.18 % Professional Educators 0.19 -0.07 ***p<0.01; **p<0.05; *p<0.10 Results: State Education Policymakers and Descriptive Representation The previous analyses provide information related to the demographic characteristics of state education policymakers. But, do the demographics of state education policymakers mirror those of the citizens they represent and serve? In short, yes; but, only for some population demographics. As shown in Table 8, I find a strong, statistically significant, positive relationship between the racial/ethnic makeup of a state and that of both SBEs and education committees: as the percentage of white citizens in a state increases, so too does the percentage of white SBE members and white education committee members; and, as the percentage of Latino citizens increases, so too does the percentage of Latino state education policymakers. There is a strong, 97 positive relationship between a state’s population of Democrats and the percent of Democrats serving on its SBE. I also examine if particular institutional structures are associated with greater descriptive representation of state education policymakers (Table 9). I find a strong negative correlation between elected SBEs and the gender party ratio: elected SBEs are associated with an overrepresentation of female members. In contrast, I find a strong positive correlation between the race parity ratio and both elected SBEs and non-partisan: elected SBEs and non-partisan SBEs are associated with an overrepresentation of white members. With regard to education committee members, I find that legislatures with term limits are associated with a greater proportion of education committee members with bachelor’s degrees and postgraduate degrees than that of the general population. Additionally, higher levels of legislative professionalism are associated with an over-representation of education committee members with postgraduate degrees. 98 Table 8 Correlations between state population and state education policymaker demographics State population % Male % White % Latino % HS diploma % Male -0.26* 0.06 -0.29** 0.15 % Some college, no degree 0.04 0.36*** % 4year degree -0.31** % Postgrad degree -0.33** % White -0.08 0.77*** -0.63*** 0.26* -0.07 0.16 -0.19 -0.34** % Latino 0.31** -0.21 0.67*** -0.27* -0.01 -0.10 0.34** 0.33** % 2-year degree % Dem Education % HS diploma 0.19 0.14 -0.15 -0.07 -0.06 0.33** -0.04 0.06 legislators % Some college 0.11 0.21 -0.16 0.18 -0.16 0.23* -0.21* -0.18 % 2-year degree 0.11 0.07 -0.02 -0.11 -0.31** 0.11 0.24* 0.31** % 4-year degree -0.22* 0.02 0.005 0.02 0.12 0.16 -0.04 -0.32** % Postgrad degree 0.04 -0.17 0.17 -0.06 0.10 -0.37*** 0.06 0.27* % Male -0.18 -0.14 -0.09 0.16 -0.13 -0.23* 0.04 -0.04 -0.02 % White 0.18 0.54*** -0.36*** 0.12 0.13 0.05 -0.13 -0.28* -0.59*** % Latino 0.01 -0.22* 0.48*** -0.10 -0.08 -0.06 0.21 0.13 0.39*** % HS diploma 0.16 0.11 -0.17 0.13 -0.15 -0.16 0.02 0.08 -0.03 % Some college 0.05 -0.02 0.07 -0.09 -0.06 -0.19 0.15 0.28* -0.11 % 2-year degree -0.39*** 0.07 -0.22 0.15 -0.10 0.33** -0.24* -0.21 -0.16 % 4-year degree 0.15 0.26* -0.04 0.07 0.03 -0.07 0.02 -0.10 -0.08 % Postgrad degree -0.18 -0.22* 0.10 -0.10 0.09 0.0003 0.03 -0.02 0.10 % Democrat 0.08 -0.10 0.16 0.15 -0.13 -0.05 -0.05 -0.03 0.48*** SBE members ***p<0.01; **p<0.05; *p<0.10 99 Table 9 Coefficients of multivariate regressions to examine relationships between institutional structures of state education policymaking bodies and state education policymaking body-to-population parity ratios SBE Institutional Structures SBE Member-toPopulation Parity Ratio Education Legislatorsto-Population Parity Ratio Elected Region-based Non-partisan % Male -0.55 (0.15)*** -0.03 (0.22) -0.17 (0.13) % White 1.04 (0.42)** 0.60 (0.61) 0.60 (0.36)* % 4-year degree 0.16 (0.14) 0.07 (0.20) 0.11 (0.12) % Postgraduate degree 0.14 (0.23) 0.31 (0.33) 0.11 (0.20) % Democrat -0.16 (0.12) -0.15 (0.17) -0.09 (0.10) Term limits Professionalism % Male -0.47 (0.38) -0.11 (0.10) % White 0.38 (0.37) -0.08 (0.09) 0.44 (0.17)*** 0.01 (0.04) 0.35 (0.22)* 0.10 (0.06)* % 4-year degree % Postgraduate degree ***p<0.01; **p<0.05; *p<0.10 Legislature Institutional Structures 100 Discussion This chapter sought to provide an understanding of whose voices are likely to be heard in the education policymaking process by examining the demographic characteristics of those who gain state education policymaking power, as well as the extent to which state education policymakers are descriptively representative of the citizens they represent and serve. This chapter also explored the ways in which institutional structures impact who gains state education policymaking power. In general, I find that white males dominate positions of state-level education policymaking power. Moreover, females and non-whites may be held to a higher standard in order to gain state education policymaking power. However, my analyses suggest that the institutional structures of SBEs and state legislatures may either facilitate or impede state education policymaker diversity. Additionally, if we care not only about the demographics of state education policymakers but also whether those demographic are representative of the state population, my findings suggest that institutional structures impact descriptive representation. Historically marginalized populations need higher credentials to gain policymaking power. I find that, when compared to males, female SBE members are more likely to have experience as a professional educator. In addition to experience as a professional educator, nonwhite SBE members are also more likely to have higher levels of educational attainment than white SBE members. These findings suggest that in order to gain state-level education policymaking power, particularly on an SBE, women and non-whites must possess additional credentials as compared to their male, white counterparts. Additionally, these findings indicate that even if historically marginalized populations obtain state education policymaking power, they are likely not representative of their population on other demographic characteristics. For 101 example, the education levels of America’s non-white population are lower than those of the white population (U.S. Census Bureau CPS, 2015b); however, the education levels of non-white state education policymakers are higher than those of white state education policymakers. It is unclear whether theses findings may be the consequence of an inherent inequity in access to positions of education policymaking power or a supply problem. Some research suggests that supply may indeed be the problem: women and blacks are significantly less likely to consider running for state office (Fox & Lawless, 2005). What is more, women and blacks are more likely than men and whites to self-assess as “qualified” or “very qualified” to run for office (Fox & Lawless, 2005). If we believe that “the willingness of a large and diverse group of citizens to seek elective office is necessary for a healthy democracy” (Fox & Lawless, 2005, p.654), perhaps efforts need to be taken to both recruit more diverse candidates to run for positions of education policymaking power and create equitable access to state education policymaking power. The benefits of having more diverse education policymakers are numerous go beyond simply conveying our political system’s level of inclusiveness. For example, if members of traditionally excluded groups have role models to emulate, they may be more likely to consider running for office (Constantini, 1990; Moncrief, Squire, & Jewell, 2001). And, members of historically marginalized groups that decide to run for office can elicit group identification and stimulate voter turnout and political activism (Abney, 1974; Abney & Hutcheson, 1981; Burns, Schlozman, & Verba, 2001). 102 Politics and professionalism increase presence of policymakers from historically marginalized populations. This chapter provides evidence that the institutional structures of state education policymaking bodies do impact who gains state education policymaking power. In particular, more professional state legislatures are associated with a greater proportion of both female education legislators and non-white education legislators. This finding aligns with Fiorina’s (1994) theory that professional legislatures are more conducive to the election of Democrats – the political party with which a majority of females and non-whites identify (Pew Research Center, 2015). Additionally, term limits are associated with fewer education committee members with 2year degrees and more education committee members with 4-year degrees. One potential explanation of this finding is that term limits result in the removal of career politicians. Career politicians are likely older in age, and older Americans have lower education levels (U.S. Census Bureau CPS, 2015a). Therefore, it seems plausible that term limits would lead to the election of younger politicians who are more likely to have higher education levels. Though term limits may indeed lead to a reduction in the average age of state legislators, other research suggests that term limits also makes legislative service more attractive as a second career (Richardson, Valentine & Stokes, 2005). Additional data on state education policymaker age is needed in order to better understand this outcome. With regard to SBE institutional structures, I find that females are more likely to gain policymaking power when SBEs are elected. In contrast, non-whites are more likely to gain policymaking power when SBE are appointed. If SBEs are elected, non-whites are more likely to gain power if the SBE elections are partisan; and, having legislative approval of governor appointees is associated with a larger proportion of non-white SBE members. These findings 103 provide some indication that more politicized state education governance structures result in a greater likelihood of members of historically marginalized populations such as females and nonwhites gaining state education policymaking power. Politics and professionalism increase descriptive representation. The institutional characteristics of state education policymaking bodies not only impact who gains state education policymaking power, but also whether those who gain power are descriptively representative of those who they represent and serve. I find that term limited legislatures lead to an overrepresentation of policymakers with Bachelor degrees when compared to the general population. This, again, may be the result of term limits leading to a lowering of the average age of state legislators and, as a result, an increase in educational attainment levels. Additionally, Latino representation on education committees outpaces Latino representation in the general population as a state’s legislative professionalism increases. This is an interesting finding given other research that suggests that states with low percentages of Latinos and professional legislators “are not good environments for the election of Latino candidates” (Casellas, 2009). Though Fiorina (1994) found that professional legislatures are more conducive to the election of Democrats, Casellas (2009) contends that, although most Latinos are Democrats, Latinos are “political newcomers” that “have not amassed the networks and name recognition needed to win highly desirable seats” (p. 406). My findings suggest that this may not always be the case. Alternatively, my findings may suggest that even if legislative professionalism does negatively impact Latino representation in state legislatures, of those Latinos that do get elected, many gain a position on the state’s Senate or House education committee. 104 Finally, I find that, when compared to a state’s overall population, females are overrepresented on SBEs that are made up of elected, rather than appointed members. Again, this finding provides some indication that more politicized state education governance structures result in a greater likelihood of historically marginalized populations such as females gaining state education policymaking power. Limitations The analyses that make up this chapter are limited in a number of ways. For one, there are many other important characteristics of state education policymakers that were not included in this analysis of descriptive representation. Age, sexual orientation, religious affiliation, disability, geography (e.g., rural, suburban, urban), parental status, among others, are demographic characteristics of state education policymakers that should be explored further. The two latter demographic characteristics are of particular interest given the increasing attention being given to religious protection and free speech, as well as lesbian, gay, transgender, queer and questioning (LGBTQ+) issues in schools is expanding (Allgood, 2016; Arum & Preiss, 2009; Graff & Stufft, 2011; Russell, Horn, Kosciw, & Saewyc, 2010). However, for a variety of reasons, data on state policymakers’ sexual orientation and religious affiliation is much more difficult to identify if one is simply relying on publicly available information. Many of these demographic characteristics are often not directly observable and thus would require selfdisclosure. Therefore, a national poll of state policymakers would likely need to be commissioned in order to obtain this data; and, even then, many state policymakers may still not be willing to disclose this personal information. 105 Another limitation of this analysis is that it explores state education policymaker demographics and descriptive representation at a single point in time. It would be valuable to understand how demographics of state education policymakers have changed over time. Moreover, a longitudinal analysis could provide insight into if and how state and national economic, educational, social and political contexts impact state education policymaker demographics and descriptive representation. Finally, as with any state-level analysis, the analyses must be interpreted with caution due to the sample size being limited to 50 states and even smaller sample sizes of education policymakers within states with different institutional structures. While being attentive to this limitation, the results from this study provide valuable information about who state education policymakers are and how institutional structures may impact who gains state education policymaking power. Conclusion Descriptive representation is an important component of any institution that hinges on participatory politics. Particularly for disadvantaged groups, descriptive representation can “improve the quality of deliberation,” create “social meaning of ‘ability to rule’ for members of a group in historical context where that ability has been seriously questioned,” and increase “de facto legitimacy in contexts of past discrimination” (Mansbridge, 1999, p.141) However, descriptive representation has many weaknesses. For example, research has shown that women policymakers do not see themselves as “acting for” women (Diamond, 1977) and that simply having black policymakers in political office does “not necessarily lead to more representation of the tangible interests of blacks” (Swain, 1993, p. 5). In concurrence with these findings, Iris 106 Marion Young writes, “Having such a relation of identity or similarity with constituents says nothing about what the representative does” (1997, p. 354). Additionally, recent scholars claim that focusing on descriptive representation may over-accentuate the importance of policymaker identity and actually “perpetuate or even aggravate the marginalization of historically disadvantaged groups” (Dovi, 2002, p. 742). Even political theorist Hanna Pitkin (1967), who offers one of the most comprehensive discussions of political representation in her book The Concept of Representation, argues that one’s ability to govern is unrelated to the policymakers’ descriptive characteristics. Moreover, Pitkin (1967) contends that limiting the composition of a legislature to a strictly descriptive standard is too literal an interpretation of representation and does not allow for “any kind of representing as acting for, or on behalf of others” (p. 90). So, do state education policymakers represent the positions of the citizens they represent and serve? Chapter V explores this question. 107 Chapter V State Education Policymakers, Responsiveness, and Valued Voices Simply having a voice present at the policymaking table does not ensure that it is heard when policymakers are making policy decisions. As Hanna Pitkin (1967) contends in her classic work on representation: sharing ascribed features does not automatically lead to sharing perspectives and interests. The previous chapter suggests that state education policymakers are not descriptively representative of a state’s general population (i.e., those impacted by education policy); however, is it possible that non-descriptively representative policymakers’ actions align with the preferences of those that they represent? In Pitkin’s (1967) terms, are state education policymakers substantively representative? And, given that education policy impacts nearly every sector of the population –as users of or contributors to the system – should we expect state education policymakers to be more responsive to particular groups of voters? Pitkin (1976) argues that disputes about the non-representativeness of a policymaking body relate not its demographic composition – but to ‘non-response’ to the voice of the represented. In this chapter, I examine whose voices are heard in the state education policymaking process. In particular, I draw upon teacher evaluation policy as a case study to examine if state education policymakers are responsive to the preferences of two important constituencies: the public and one of the largest voting constituencies in the nation—teachers. Although research suggests that state education policymakers interested in retaining their elected or appointed position will be responsive to those with voting power (Dahl, 1971), it may be that state education policymakers are responsive to neither. Therefore, this chapter also examines whose voices are valued by state education policymakers. In sum, this chapter aims to answer the 108 following questions: (1) Are policymakers responsive to public and/or teacher policy preferences? a. Do individual or institutional characteristics impact state education policymaker responsiveness to public and/or teacher policy preferences? (2) When forming their position on an education policy, whose voices do state education policymakers value? a. Do individual or institutional characteristics impact who state education policymakers nominate as valuable voices in their decisionmaking process? What Do We Know About Policymaker Responsiveness and Substantive Representation? To gain a better understanding of what is known about policymaker responsiveness and substantive representation, I begin with an integrative review of research on substantive representation and policy responsiveness, with a specific focus on state-level government and education policy. Integrative reviews simultaneously incorporate experimental and nonexperimental research in order to more fully understand, critique, and potentially reconceptualize and expand the knowledge base on a topic (Torraco, 2005). Integrative reviews also serve a wide range of purposes, including as a review of theories, evidence, and methodological approaches (Broome, 1993). Literature review methods and data are explained in further detail in Appendix G. Research on substantive representation among state-level policymakers has focused almost exclusively on women, African Americans, and Latinos and whether state policymaker priorities and platforms align with the typical policy priorities of these particular demographic groups. The primary issue areas that this research has encompassed include women’s health, the 109 death penalty, abortion, gun control, and immigration. For example, Dolan and Lynne (1995) conducted a survey of 1,373 women legislators in all 50 states to examine whether the stated priorities of female legislators were “responsive” to women. Similarly, Barrett (1995) surveyed black women in state legislatures, as well as a matched group of black men, non-minority men, and non-minority women to examine their legislative priorities and positions on women and minority policy issues. Building upon the previously described research, my research focuses on a relatively understudied policy issue with regard to substantive representation – education. Additionally, this study explores an issue that is typically understood to be universally beneficial to all groups of people, as opposed to just a single population such as women or minorities. Methodologically, many researchers have drawn upon data from national polls and surveys (e.g., Gallup) to examine policy responsiveness and substantive representation. In some cases, national surveys allowed for the disaggregation of the data at the state level (Erickson, 1976). In other cases, researchers utilized multilevel regression or post-stratification to estimate state policymaker responsiveness to public preferences. A limited number of studies have examined data collected from surveys of state policymakers (Barrett, 1995; Biebel, Nicholson, Williams, & Hinden, 2004; Dolan & Lynn, 1995; Erickson, Wright, & McIver, 1993). Therefore, most research has examined policymaker preference data aggregated at the state or national level. This only provides a measure of whether policymakers, on average, are responsive to public preferences. In contrast, I collected data that allows me to examine individual policymaker responsiveness. Finally, researchers have largely utilized proxy indicators of policymaker responsiveness measured at the institutional level, rather than at the individual level. For example, Hawes (2013) used frequency of contact between school superintendents and local boards of education as a 110 proxy for local government responsiveness. Others have measured responsiveness as whether legislation or administrative rules were enacted that favor or do not favor a particular population (Berkman & O’Connor, 1993; Bratton & Haynie, 1999; Bulluck & MacManus, 1981; Epstein, Herron, O’Halloran, & Park, 2007; Erickson, 1976; Erickson, Wright, & McIver, 1993; Flavin, 2012; Fossett & Thompson, 2006; Lascher, Hagen, & Rochlin, 1996; Lax & Phillips, 2009; McLendon, Mokher, & Flores, 2011; Page & Shapiro, 1983; Reingold, 2000; Reingold & Smith, 2012; Shapiro, 1982; Thomas, 1991). For instance, Berkman and O’Connor (1993) examine the relationship between the percent of women in a state legislature and whether a state legislature passed legislation with a pro-choice position on abortion funding or pro-choice position on parental notification. As an example in the issue area of education, McLendon, Mokher, and Flores (2011) examined whether legislation that was publically supported (in-state resident tuition) was introduced in the Texas state legislature. When using these various proxy measures of responsiveness, it is unclear whether exposure to public or sub-public preferences was the primary influence on policymakers’ decisions. This study examines policymaker responsiveness in a way that compresses the distance between the presentation of public preference data and policymakers’ policy decisions. Theoretical Framework and Hypotheses While institutional rules and norms may influence and constrain who gains state education policymaking power, research and theory suggest that policymakers retain some individual agency and can act strategically during the policymaking process (DiMaggio, 1988; Fligstein, 1997; Giddens, 1976, 1979; Granovetter, 1985; March & Olsen, 1984; Ostrom, 1990, 1991; Seo & Creed, 2002). As such, this chapter expands the theoretical framework from 111 Chapter IV to examine the impacts of both institutional and individual characteristics of state education policymakers on policymaker responsiveness. Institutional structures: Ways of coming to power. As described in previous chapters, there are two ways of gaining formal state education policymaking power: gaining a seat on an SBE or gaining a seat on an education committee in the state legislature. Research focused on special-purpose governance and responsiveness to public and sub-public preferences present conflicting hypotheses. In some cases, scholars argue that special-purpose governments are less likely to be captured by special interest groups and more likely to deliver efficient policy decisions that are congruent with majority opinion (Bish, 1971; Ostrom, Bish & Ostrom 1988). In contrast, other scholars contend that the low visibility of special-purpose governments hinders public participation and creates a bias toward special interests that have the capacity to lobby special-purpose government officials (Bollens, 1957; Burns, 1994; Foster 1997; Jones, 1942; Macedo and Karpowitz, 2006). Henig (2013), too, contends that state legislatures are not invulnerable to domination by privileged interests, but general-purpose government’s “broader scope, visibility, and pluralistic complexity make it relatively less likely that any single interest could prevail and hold its advantage over time” (p. 19). Thus, the current literature is not unified in its understanding of how a policymakers’ location in different state policymaking institutions impacts state-level policymakers’ responsiveness to public preferences. This research will provide additional insight into these conflicting hypotheses. H1A: When compared to education committee members, SBE members will be less responsive to public preferences. 112 H1B: When compared to education committee members, SBE members will be more responsive to teacher preferences. Institutional structures: SBE selection methods. Research indicates that these different selection methods may interest different types of individuals, introduce different behavioral incentives while in office, and influence responsiveness to public preferences (Alesina & Tabellini, 2007; Beasley & Coate, 2003; Burden, Canon, Lavertu, Mayer, & Moynihan, 2013). In the state education policymaking process, this research would suggest that appointed SBE members may be more likely to pursue goals that diverge from the public without fear of losing their position because they are not directly connected to voters. While appointed SBE members are not entirely independent (they are accountable to elected officials, and thus indirectly accountable to the voters), the addition of an intermediary (i.e., the appointer) may weaken voters’ ability to hold appointed SBE members accountable and perhaps result in appointed SBE members being less responsive to public preferences. In contrast, elected SBE members might have policy goals that better match the preferences of the electorate than those of appointed SBE members because they have been selected, ostensibly, because of their views on education. H1C: When compared to elected SBE members, appointed SBE members will be less responsive to public preferences Institutional structures: Term limits. Theory suggests that term limits affect the kinds of individuals who run for office and the behavior of individuals (Carey, Niemi & Powell, 1998). More specifically, term limits (a) force politicians to push their priorities more quickly because their period of service is limited and (b) result in politicians being less concerned with the needs 113 of their district and more concerned with the state as a whole resulting in a great propensity to follow their conscience (Carey, Niemi & Powell, 1998). H1D: When compared to non-term limited state education policymakers, term limited state education policymakers will be less responsive to public preferences. Institutional structures: Professionalism. Squire’s (2007) measure of legislative professionalism reveals the wide variation that exists among state legislatures’ staffing and resources, time demands, and salary and benefits. Each of these components that influence the level of legislative professionalism influence the policy decisionmaking process. For example, legislatures that offer salary and benefits often retain policymakers for longer periods of time, creating a more experienced policymaking body (Squire, 2007). Similarly, legislatures that require policymakers to serve year-round increase the amount of time for policy development and deliberation (Squire, 2007). SBEs, on the other hand, do not vary in professionalism as much as state legislatures: while some SBEs are compensated for their service, no state has a full-time SBE and nearly all SBE members have a separate full-time career outside of their service on the SBE. As it relates to education policymaking, one might conjecture that a more professional legislature may be more responsive to public preferences given that they have more time to learn about, understand, and respond to public preferences. Research indeed contends that the professionalization of policymakers encourages responsiveness for numerous reasons. First, professional policymakers may be inclined to respond neutrally and competently to competing interests by applying expert knowledge and skills in accordance with professional norms and standards rather than responding based, say, on political party lines (Kearney & Sinha, 1988). 114 Second, professionalism may make the policymaking process more democratic because authority “is less a power to command than to coordinate” in organizations that are dependent on the activities of experts (Redford, 1969, p. 159); thus, more professional policymakers are encouraged to exercise independent judgment and be innovative and creative in problem solving (Kearney & Sinha, 1988). H1E: Education committee members in states with more professional legislatures will be more responsive to public preferences. Additionally, it has been argued that professionalism of policymakers facilitates interaction between experts and politicians because more professional policymakers are “better able to understand their respective needs, strengths, and weaknesses, and to form mutual expectations accordingly” (Kearney & Sinha, 1988, p. 576). If teachers are perceived as experts, perhaps education committee members in more professional legislatures may be more responsive to their preferences. H1F: Education committee members in states with more professional legislatures will be more responsive to teacher preferences. Nonetheless, research has shown that more professional policymakers frequently advance their own notions of what is good for the people rather than seeking and responding to the interests and demands of the general public (Hummel, 1987; Miewald, 1978; Mosher, 1982). Thus, it is unclear, particularly in the case of education policy, whether more professional legislatures are more or less responsive to public preferences. This research will contribute to an understanding of this relationship. 115 Individual characteristics, representation, and responsiveness. Rooted in cognitive psychology, research has aimed to understand how policymakers arrive at their policy choices. General decisionmaking research has identified factors including past experience (Juliusson, Karlsson & Garling, 2005), cognitive biases (Stanovich & West, 2008), and age and individual differences (de Bruin, Parker & Fischoff, 2007) that influence the choices people make. In extending this research to policymakers, it has been suggested that policymakers’ cognitive biases influence their policy decisions (West et al., 2008). Political orientation is one individual characteristic that been shown to impact individuals’ decisions. While researcher have not yet examined the relationship between political orientation and responsiveness to education policy preferences, research at the national level suggests that while all individual members of Congress may move in response to common public opinion, Democrats tend to move over time with public opinions, while Republicans have been shown to be insensitive to changing public opinion (Przeworski, Stokes, & Manin, 1999). H2A: State education policymakers with a liberal political orientation will be more responsive to public preferences than those with a conservative orientation. Finally, research suggests that Democrats would be more responsive to teacher policy preferences. If state education policymakers are indeed motivated to retain their position of power (Dahl, 1989; Downs, 1957; Mayhew, 1974; Arnold, 1990), Democrat policymakers are likely to align with the preferences of individuals who make up the most powerful interest group in education: teacher unions (DeBray-Pelot & McGuinn, 2009). Indeed, research shows that teachers unions are quite successful at getting their favorite candidates – almost exclusively Democrats – elected, both at the state and local level (Moe, 2003; Moe, 2011). 116 H2B: State education policymakers with a liberal political orientation will be more responsive to teacher preferences than those with a conservative orientation. Data To examine if state education policymakers are responsive to the policy preferences of both the general public and teachers, this chapter employs data collected from an online survey of state legislators serving on education committees and SBE members administered in the spring and fall 2016. I also combined survey response data with data of both individual and state education policymaking body characteristics utilized in Chapter IV. The purpose of my survey was to gather information from a sample of state education policymakers for the purpose of understanding (a) whether state education policymakers are responsive to public and teacher preferences about an education policy and (b) the voices that are valued (or not valued) by state education policymakers when forming education policy positions. To examine these issues, I focus on one education policy as a case study: teacher evaluation policy. Policy case study: Teacher evaluation policy. Teacher evaluation policy was selected on the basis of its salience, particularly for state education policymakers. Under former President Barack Obama, two federal government programs provided incentives for state policymakers to require that school districts use student assessment scores in teacher evaluations. From 2010 to 2011, the U.S. Department of Education (USDOE) enticed state education agencies (SEAs) to develop innovative teacher evaluation systems through Race to the Top (RttT), a competitive grant program designed to encourage and reward states for enacting education policies that the administration believed will improve student achievement and increase school system 117 productivity and effectiveness (USDOE, 2009). The RttT competition required states to submit an application that outlined how it planned to develop and implement each of the Obama Administration’s education reforms. States were awarded a significant number of points based on the development and implementation of “rigorous, transparent and fair evaluation systems” that incorporated measures of student growth (USDOE, 2009). Additionally, in 2011, the Obama Administration proclaimed that “the unintended results of NCLB’s strictest requirements are hindering further school reforms and innovations” and, thus, the administration intended to develop “a new partnership where the federal role is to support innovation and reform in the states” through a federal waiver process (USDOE, n.d.). Contending that NCLB “did little to elevate the teaching profession or recognize the most effective teachers,” the USDOE incorporated teacher evaluation policy reform into an ESEA flexibility waiver application. To fulfill Principle 3 of the application –“Supporting Teachers, Leaders and Local Innovation” – SEAs were to describe how they would develop and implement teacher evaluation systems that use multiple measures of performance including student growth based on state assessment as a significant factor (USDOE, 2012). Partly as a result of these federal inducements, at the time the survey was administered, all states required the use of state assessment data for use in teacher evaluations for those teachers in subject areas in grades in which state assessments are administered. Survey Design. Relying on single question to make an inference about a large construct – such as “responsiveness” – may make it difficult to uncover complexity (Converse & Presser, 1986). Therefore, I generated questions that would provide “multiple measures” of the issue of focus. In designing my survey, I developed one experimental question and three additional 118 questions that aimed to understand the groups and individuals that these policymakers draw upon when making education policy decisions, I chose four different question formats that each aim to “cast light at different angles” (Converse & Presser, 1986, p. 47). Additional details on the survey question design and limitations are available in Appendix H. Randomized control trial: Measuring responsiveness. Whether state education policymakers are responsive the policy preferences of (a) the public at large and (b) teachers may be a function not only of policymaker observables such as background, office, or party affiliation, but also characteristics and experiences of the policymaker that cannot be observed. To mitigate this concern, I designed an experimental survey question. In particular, I randomly assigned state education policymakers to one of three (one control, two treatment) conditions and then asked them whether they would be inclined to vote in support of or in opposition to a teacher evaluation policy change. State education policymakers assigned to the control condition were asked whether they would vote in favor or opposition to the policy change with no other information provided. State education policymakers assigned to treatment condition A were provided with additional data related to public preferences on the policy change; and, those assigned to treatment condition B were provided with additional data related to teacher preferences on the policy change. Due to the relevancy of the issue, data for the treatment conditions had the benefit of being factually true—they were drawn from Question 25 of the 2014 Education Next/Harvard PEPG Survey.35 The exact language of the experimental survey question, as seen by state education policymakers, is displayed in Table 10. 35 The full 2013 EducationNext/PEPG survey is available at http://educationnext.org/the-2013-education-nextsurvey/ 119 Table 10 Treatment conditions for state education policymakers survey experiment Control Student achievement data from state standardized tests is a required element in many teachers’ performance evaluations and, thus, tenure decisions. However, in recent months, some consideration has been given to the elimination of this requirement. Treatment A (public preference - oppose) Would you be inclined to vote in favor of or in opposition to a policy that would remove the required use of state assessment data in teacher evaluation and tenure decisions? Student achievement data from state standardized tests is a required element in many teachers’ performance evaluations and, thus, tenure decisions. However, in recent months, some consideration has been given to the elimination of this requirement. According to a recent survey conducted by Harvard University researchers, a majority of the public SUPPORTS the inclusion of student progress on state tests in teacher tenure decisions. Treatment B (teacher preference - support) Would you be inclined to vote in favor of or in opposition to a policy that would remove the required use of state assessment data in teacher evaluation and tenure decisions? Student achievement data from state standardized tests is a required element in many teachers’ performance evaluations and, thus, tenure decisions. However, in recent months, some consideration has been given to the elimination of this requirement. According to a recent survey conducted by Harvard University researchers, a majority of teachers OPPOSE the inclusion of student progress on state tests in teacher tenure decisions. Would you be inclined to vote in favor of or in opposition to a policy that would remove the required use of state assessment data in teacher evaluation and tenure decisions? Because the version of the survey question that each state education policymaker received was chosen at random, any differences in responses can be attributed directly to the treatment (public or teacher pressure). The two treatments – public opposition to and teacher support for the policy change – are theorized to be pertinent political incentives faced by state education policymakers. Many policymakers may aim to be responsive to their electorate (Dahl, 1971). Policymakers may also find it advantageous to be responsive to teachers, particularly 120 because teachers unions constitute one of the largest donors to political campaigns (Camera, 2014). Forced rating. The first question in this multiple measures approach was designed to force state education policymakers to rate the importance of different state- and local-level actors in informing their position on teacher evaluation policy. Survey respondents were asked to rate 11 state-level actors and eight local-level groups and actors on a scale from 1 (not important) to 5 (very important). The actors included in the survey (see Appendix J, Table 35) were those that research has shown can have an impact on the policymaking process Discrete choice experiment. The second item presented state education policymakers with two actors and forces respondents to select the actor that is most likely to inform their teacher evaluation policy position. This type of discrete choice experiment (DCE) is frequently used in marketing and health care. DCEs present survey respondents with samples of hypothetical scenarios (choice sets) drawn a priori from all possible choice sets (Amaya-Amaya, Gerard, & Ryan, 2008; Louviere, 1991, 2001). In my survey, respondents were asked, “Thinking about your position on statewide teacher evaluation policies, consider the following sets of two viewpoints. When the positions of these two groups/individuals differ, whose is more likely to inform your position?” Respondents were presented with 28 choice sets. The choice sets control for order effects and context effects: the order in which the survey respondent viewed the paired items was randomized; and, each individual item was paired with every other item an equal number of times. The items that were included were: business organizations, education advocacy 121 groups, education research organizations, personal beliefs, school leadership groups, teacher labor unions, the governor, and your constituents. Least and most important actors. The final question item asked state education policymakers to select the two most important and two least important actors in informing their position on teacher evaluation policy. Because policymaking doesn’t take place in a vacuum and more than one group can influence a policymakers’ decision at once, this question allowed respondents to discriminate between the actors of greatest and least importance in informing Sample Target population. The target population for this research was a purposive sample of 1,670 formal state education policymakers, defined as individuals with constitutional or statutory responsibility and power to make education policy. More specifically, I targeted SBE members and state legislators serving on a K12 education policy committee36 in 47 states37 in February 2016. While some SBE members serve ex-officio—such as student members, the governor, lieutenant governor, and the state superintendent—my target population was limited to only those state policymakers with voting power. Descriptive statistics for the target population can be found in Table 11. 36 Unless they also served on the K12 policy committee, state legislators serving on K12 education appropriations committees were not included in this study 37 Three states – New Mexico, Minnesota and Wisconsin – do not have a formal state board of education and thus were not included in the study target population. 122 Table 11 Comparison of demographics among target, accessible and sample population for survey Target Population Sample Population 1670 Accessible Population 1605 (96.1%) Male Female Missing 1133 (67.8%) 537 (32.2%) 0 (0.0%) 1093 (68.1%) 513 (31.9%) 0 (0.0%) 125 (62.8%) 74 (37.0%) 1 (0.5%) White Non-White Missing 1383 (82.8%) 286 (17.1%) 1 (0.0%) 1335 (83.2%) 270 (16.8%) 1 (0.0%) 169 (84.9%) 30 (15.1%) 1 (0.5%) Latino Non-Latino Missing 58 (3.5%) 1611 (96.5%) 1 (0.0%) 56 (3.5%) 1549 (96.5%) 1 (0.0%) 9 (4.5%) 190 (95%) 1 (0.5%) Legislature SBE Missing 1192 (71.4%) 478 (28.6%) 0 (0.0%) 1192 (74.2%) 414 (25.8%) 0 (0.0%) 91 (45.5%) 108 (54.0%) 1 (0.5%) Elected Appointed Missing 1295 (77.5%) 375 (22.5%) 0 (0.0%) 1295 (80.7%) 311 (19.4%) 0 (0.0%) 152 (76.0%) 48 (24.0%) 0 (0.0%) Democrat Republican Independent or Not Available 934 (53.7%) 726 (41.7%) 80 (4.6%) 855 (53.5%) 669 (41.9%) 73 (4.6%) 80 (45.0%) 97 (48.5%) 13 (6.5%) Total Gender 200 (12.5%) Race/Ethnicity Latino Office Selection Method Political Party Accessible population. Upon ascertaining my target population, I began to identify my accessible population, which included all formal state education policymakers with a publicly available e-mail address. The e-mail identification process included four steps. First, I attempted to retrieve e-mail addresses directly on the state government sponsored web pages of the SBE and state legislature. When e-mail addresses were not available via this source, I conducted an Internet search and visited web pages that may have provided the policymakers e-mail address (e.g., campaign websites, social media pages, publications, conference rosters, interest group contact lists). If this search did not produce contact information, I searched for the individual on 123 the state’s campaign finance disclosure database. In some cases, individual e-mail addresses were provided for policymaker campaign committees. The final attempt to identify a policymaker’s e-mail address included sending an e-mail message to an administrator or secretary within their governing body to request an e-mail address. I was unable to identify 64 (3.8 percent) state education policymakers’ e-mail addresses. Descriptive statistics for the accessible population are found in Table 11. Upon identifying my accessible population, I randomly assigned state education policymakers into three groups that would receive the survey with the experimental survey question wording of one of the three control or treatment question wording (discussed in depth below). Aside from the variation in the experimental survey question wording, all other questions in the survey were the same across all policymakers. The random assignment of individuals into groups took place within clusters that represented the three state policymaking bodies: SBE, Senate, and House of Representatives. State policymakers were invited to participate in the online survey via Qualtrics. A copy of the e-mail that was sent to state education policymakers is provided in Appendix I. The first wave of recruitment e-mails was sent to the accessible population in May 2016. Ninety-eight state education policymakers completed the survey within one week of the first email. A reminder e-mail was sent two weeks later to those who had not completed the survey and who had not opted out of receiving further e-mails about the survey. An additional seventy-three state education policymakers completed the survey within one week of the reminder e-mail. In this first wave of survey distribution, I received a total of 58 auto-reply e-mail messages (4 percent). Additionally, 94 e-mail messages (6 percent) were returned for failure to find the policymaker because the e-mail address was either incorrect or no longer in service. All returned 124 e-mails were reviewed to ensure no human errors were made in the input of the e-mail address and corrected as necessary. Because many state legislators do not meet year round, a second round of e-mails were sent in September 2016. A reminder e-mail was sent one week later. The purpose of this second round of e-mails was to recruit individuals who may not have been in legislative session in the spring but who were in session in the fall. Thirty-eight state education policymakers completed the survey within one week of the first e-mail and an additional 21 completed the survey within one week of the second e-mail. In this second wave of survey distribution, I received a total of 55 auto-reply e-mail messages (4 percent). Additionally, 94 e-mail messages (6 percent) were returned for failure to find the policymaker because the e-mail address was either incorrect or no longer in service. Again, all returned e-mails were reviewed to ensure no human errors were made in the input of the e-mail address and corrected as necessary. Sample Population. A total of 200 individuals participated in the survey, resulting in a response rate of 13 percent. Within control and treatment groups, survey response rates were 8 percent for the control, 16 percent for treatment A, and 14 percent for treatment B (see Table 12). Of those who participated in the survey, 73 percent completed all questions in the survey. Individual survey item response rates varied, although the experiment-based question had the highest response rate among non-demographic questions (86 percent). 125 Table 12 Comparison of response rates and within-group demographics of accessible and sample populations Control Accessible Sample Population Population 534 41 (7.7%) Total Gender Male Female Missing Race/Ethnicity White Minority Missing Latino Non-Latino Latino Missing Office Legislature SBE Missing Selection Method Elected Appointed Missing Political Party Democrat Republican No Party /Missing Treatment A Accessible Sample Population Population 535 85 (15.9%) Treatment B Accessible Sample Population Population 536 74 (13.8%) 345 (64.6%) 189 (35.4%) 0 (0.0%) 21 (51.2%) 20 (48.8%) 0 (0.0%) 372 (69.5%) 163 (30.5%) 0 (0.0%) 56 (65.9%) 28 (32.9%) 1 (1.2%) 376 (70.2%) 160 (29.9%) 0 (0.0%) 48 (64.9%) 26 (35.1%) 0 (0.0%) 445 (83.3%) 88 (16.5%) 1(0.2%) 36 (87.8%) 5(12.2%) 0 (0.0%) 442 (82.3%) 93 (17.4%) 0 (0.0%) 71 (83.5%) 13 (15.3%) 1 (1.2%) 447 (83.4%) 89 (16.6%) 0 (0.0%) 62 (83.8%) 12 (16.2%) 0 (0.0%) 511 (95.7%) 22 (4.1%) 1 (0.2%) 40 (97.6%) 1 (2.4%) 0 (0.0%) 518 (96.8%) 17 (3.2%) 0 (0.0%) 79 (92.9%) 5 (5.9%) 1 (1.2%) 519 (96.8%) 17 (3.2%) 0 (0.0%) 71 (96.0%) 3 (4.5%) 0 (0.0%) 395 (74.0%) 139 (26.0%) 0 (0.0%) 17 (41.5%) 24 (58.5%) 0 (0.0%) 399 (74.6%) 136 (25.4%) 0 (0.0%) 36 (44.7%) 46 (54.1%) 1 (1.2%) 398 (74.3%) 138 (25.8%) 0 (0.0%) 36 (48.7%) 38 (51.4%) 0 (0.0%) 426 (79.8%) 108 (20.2%) 0 (0.0%) 33 (80.5%) 8 (19.5%) 0 (0.0%) 434 (81.1%) 101 (18.9%) 0 (0.0%) 55 (64.7%) 30 (35.3%) 0 (0.0%) 434 (81.0%) 102 (19.0%) 0 (0.0%) 64 (86.5%) 10 (13.5%) 0 (0.0%) 196 (40.3%) 268 (55.1%) 22 (4.5%) 20 (48.8%) 147 (41.5%) 4 (9.81%) 176 (40.5$) 242 (55.6%) 17 (3.9%) 36 (42.4%) 44 (51.8%) 5 (5.9%) 190 (41.9%) 244 (53.9%) 19 (4.2%) 41 (55.4%) 29 (39.2%) 4 (5.4%) Examining response bias. I conducted two-sample t-tests to examine whether the demographics of those who participated in the survey were significantly different from those that did not (Table 13). The following groups were underrepresented among survey respondents: males, legislators, elected officials, and Republicans. There were no significant differences among respondents and non-respondents with regard to race/ethnicity. Additionally, within treatment conditions, legislators and Republican policymakers were underrepresented among all three groups; and, elected officials were underrepresented within treatment (Table 14). 126 It is difficult to determine how the response rate of my survey compares to other surveys of state legislators and SBE members. First, there are no nationwide surveys of SBE members to which a comparison can be made. With regard to legislators, Maestas, Neeley, & Richardson (2003) reported that the average response rate across 33 studies of surveys with state legislators between 1975 and 2000 was 45.7 percent. However, only eight of these studies were surveys that covered all 50 states and only five included more than 2,000 legislators in the sample. Additionally, only one article in Maestas and colleagues’ (2003) sample was published after 1996. Therefore, very little is known about state-level political elite response rates in the current environment. Table 13 Tests of survey response bias Female Non-white SBE member Appointed Democrat Respondents (n=200) M SD 0.37 0.48 0.15 0.36 0.54 0.50 0.24 0.43 0.48 0.59 Non-respondents (n=1,405) M SD 0.31 0.46 0.17 0.38 0.21 0.41 0.16 0.37 0.43 0.50 127 Two-sample t-test t p -2.91 <0.01 1.42 0.16 -10.46 <0.01 -2.33 0.02 3.81 <0.01 Table 14 Tests of survey response bias, by control/treatment group M 0.49 0.12 SD 0.51 0.33 Control Nonrespondents (n=486) M SD 0.34 0.48 0.17 0.38 0.59 0.50 0.23 0.42 -5.39 <0.01 0.55 0.50 0.19 0.39 -7.65 0.20 0.46 0.40 0.51 0.18 0.42 0.38 0.49 -0.19 1.66 0.85 0.10 0.35 0.55 0.48 0.50 0.15 0.42 0.36 0.50 -3.48 2.11 Respondents (n=47) Female Non-white SBE member Appointed Democrat Two-sample t-test Respondents (n=97) t 0.15 1.18 p 0.03 0.24 M 0.33 0.15 SD 0.47 0.36 Treatment A NonTwo-sample respondents t-test (n=436) M SD t p 0.29 0.45 -1.78 0.08 0.18 0.39 0.91 0.36 128 M 0.35 0.16 SD 0.48 0.37 Treatment B NonTwo-sample trespondents test (n=453) M SD t p 0.29 0.45 -1.61 0.11 0.17 0.37 0.52 0.60 <0.01 0.51 0.50 0.22 0.41 -5.22 <0.01 <0.01 0.04 0.14 0.41 0.34 0.50 0.15 0.44 0.36 0.50 0.07 2.84 0.94 0.01 Respondents (n=83) While multiple rounds of contacts were necessary to prompt state education policymakers to complete the survey, this technique could also result in response bias. For example, those who completed the survey in the first round may be different than those who completed the survey in the second round; or, those who completed the survey after receiving the first e-mail may be different than those who completed the survey after receiving a reminder e-mail. To examine this, I compared state education policymakers’ demographics and their response to the experimental survey question – the question with the highest item response rate– for those who (a) completed the survey in the spring versus fall round of e-mails and (b) completed the survey in response to the first versus the reminder e-mail (see Tables 15 and 16). This comparison revealed minor differences in the demographic characteristics of survey respondents at different times. The largest difference between first versus second round e-mail responders is that more female state education policymakers responded in the first round of e-mails. Additionally, state legislators serving on education committees that responded in the second round were slightly more responsive to the preferences of the group for which data was presented to them in the experimental question. Differences between those who responded to an initial e-mail and those who responded to the survey after receiving a reminder e-mail were minimal. 129 Table 15 Response time bias: May vs. September e-mail respondents Female Non-white SBE member Democrat Responsiveness All state education policymaker responders Difference Overall 1st 2nd between Average round round 1st and 2nd round 0.37 0.45 0.28 0.17 0.15 0.88 0.83 0.05 0.54 0.55 0.48 0.07 0.48 0.33 0.18 0.15 0.44 0.53 0.62 -0.09 State legislators serving on education committees Difference Overall 1st 2nd between Average round round 1st and 2nd round 0.30 0.35 0.23 0.12 0.15 0.88 0.77 0.11 0.54 0.25 0.29 -0.04 0.42 0.49 0.74 -0.25 *Note: Only includes survey respondents in the two treatment group 130 State board of education members Overall Average 1st round 2nd round 0.44 0.15 0.43 0.46 0.52 0.87 0.40 0.56 0.32 0.89 0.07 0.50 Difference between 1st and 2nd round 0.20 -0.02 0.33 0.06 Table 16 Response time bias: Initial vs. reminder e-mail respondents All state education policymakers respondents Female Non-white SBE Democrat Responsiveness Overall Average Initial Reminder 0.37 0.15 0.54 0.48 0.44 0.37 0.90 0.26 0.55 0.44 0.82 0.32 0.56 Difference between initial and reminder -0.07 0.08 -0.06 -0.01 State legislators serving on education committees Difference Overall between Initial Reminder Average initial and reminder 0.30 0.32 0.33 -0.01 0.15 0.89 0.80 0.09 0.54 0.27 0.25 0.02 0.42 0.51 0.65 -0.14 *Note: Only includes survey respondents in the two treatment groups 131 State board of education members Overall Average Initial Reminder 0.44 0.15 0.43 0.46 0.43 0.92 0.27 0.60 0.52 0.83 0.38 0.49 Difference between initial and reminder -0.09 0.09 -0.11 0.11 Methods Are policymakers responsive to public and/or teacher policy preferences? Analysis of response data to the experimental survey question begins with descriptive statistics of policymaker support for or opposition to the policy change by control and treatment groups. Subsequently, I test the relationship between exposure to public or teacher pressure and state education policymaker responsiveness, first, using ordinary least squares (OLS) regression. The outcome variable takes on a value of 0 for those who support the teacher evaluation policy change and a value of 1 for those who oppose the policy change. There may be some concern about using OLS with a binary dependent variable. For example, a probit model could accommodate for the unboudedness problem in the binary outcome (Aldrich & Nelson, 1984; Long, 1997). However, Blunch & Gupta (2014) argue that OLS may be appropriate, especially if most of the independent variables are discrete and heteroskedasticity-robust standard errors are used in place of regular standard errors (Wooldridge, 2010). As such, I report the results of both probit regression models and ordinary least squares (OLS) regression models. I report robust standard errors for all models. Model 1 is the most simplified, testing whether either treatment has an impact on state education policymakers’ teacher evaluation policy position. Model 2 incorporates state education policymakers characteristics, as well as an indicator of whether the policymaker resides in the legislature or SBE. Finally, models 3 and 4 build upon model 2 by incorporating a treatment-bypolitical party interaction term. All models include Census region fixed effects. When forming their position on an education policy, whose voices do state education policymakers value? The analysis of the experimental survey question response data aims to 132 answer the questions: do state education policymakers listen to the voices of (a) the public at large and (b) teachers? And, are some policymakers more responsive than others? However, some research suggests that policymakers have a strong status quo bias (Gilens, 2005); and, specifically at the state level, some policies consistently reflect opinion majorities while, for others, “even clear supermajority support seems insufficient for policy adoption” (Lax & Phillips, 2009, p. 383). Therefore, it may be that education policymakers are relatively unresponsive to both the public and teachers. As a result, I analyze data collected from three additional survey questions designed to provide an understanding of whose preferences are most and least influential in shaping the decisions of state education policymakers. Ratings of state and local actors. With multiple continuous dependent variables and multiple discrete independent variables, I begin my analysis with a multivariate analysis of variance (MANOVA) to examine the relationship between state education policymaker demographic and institutional characteristics and their ratings of the importance of state and local actors in informing their position on statewide teacher evaluation policy. The dependent variables include individual state education policymakers’ 0-to-5 ratings of 11 state actors and eight local actors. The independent variables include policymaker gender, race/ethnicity, political party affiliation, office, and Census region. The null hypothesis (!! ) is that a state education policymaker’s characteristics have no effect on their rating of an actor’s importance. The alternative hypothesis (!! ) is the ratings of an actor’s importance in informing policymaker’s position on statewide teacher evaluation policy varies based on state education policymaker characteristics. 133 Additionally, I examine descriptive statistics and conduct nonparametric analyses of the ratings given to actors by policymakers with different demographic characteristics and institutional homes to explore differences among state education policymakers’ ratings of state and local actors. Given that rating data for a number of state and local actors is not normally distributed and that groups may not differ in mean rating but may differ in distribution of ratings, I utilize two-sample Kolmogorov-Smirnov (K-S) tests for equality of distribution functions (Conover, 1971). Appendix J provides additional details on the appropriateness of nonparametric modeling to analyze this data. Discrete choice experiment. To examine whether state education policymakers more frequently nominate some actors over others as important to informing their teacher evaluation policy position, I examine descriptive statistics of the frequency with which each of eight actors was selected over all other actors. I also conduct correlation analyses to examine relationships between state education policymaker characteristics and the frequency with which they select each actor. Finally, I conduct negative binomial regressions to examine whether the correlations previously identified hold when controlling for a multitude of state policymaker characteristics. The relationship between the expected number of times each actor was selected over all other actors, !! , and a set of education policymaker demographic characteristics is !! = !0 + !1 !"#$%"1 + !2 !"!#ℎ!"#2 + !3 !"#$%&'(3 + !4 !"#4 + !5 !"#$%&'() !"#"!5 + !! 134 (3) I report the results of the negative binomial analyses for two models: model one does not include Census region fixed effects; model two does include Census region fixed effects. Reported standard error terms are robust and clustered by Census region. Least and most important actors. Finally, to examine which actors policymakers perceive as most and least important in informing their teacher evaluation policy position, I examine descriptive statistics and correlations of policymakers’ most and least important actor nominations. In this exercise, I group the actors into five categories (see Appendix L, Table 2). Additionally, to understand if there is a relationship between (a) two actors being nominated as most important, (b) two actors being nominated as least important, and (c) one actor being nominated as most important and one as least important, I conduct Pearson Chi-Square tests and examine cross-tabs. Results: Are State Education Policymakers Responsive to Public or Teacher Preferences? Among all state education policymaker survey respondents, 55 percent indicated that they would vote in favor of a policy change that would remove the required use of state assessment data in teacher evaluation and tenure decisions. Forty-four percent indicated that they would oppose this policy and less than 2 percent did not provide a response. As shown in Figure 16, among state education policymaker respondents in the control group, 60 percent indicated that they would be inclined to vote in favor of a policy change that would remove the use of state assessment data in teacher evaluation and tenure decisions. When provided with information that the majority of the public supports the inclusion of student progress on state tests in teacher evaluation and tenure decisions (treatment A)—signaling that 135 the majority of the public would prefer that a policymaker vote in opposition to the policy change—slightly fewer state education policymakers (59 percent) supported the policy change. In contrast, when provided with information that the majority of teachers oppose the inclusion of student progress on state tests in teacher evaluation and tenure decisions (treatment B)— signaling that the majority of teachers would prefer that a policymaker vote in favor of the policy change—even fewer state education policymakers (49 percent) supported the policy change. These descriptive statistics suggest state education policymakers may be responsive to pressure from teachers related to this education policy issue but may not be responsive to pressure from the public at large. Statistical analyses provide a nuanced understanding of these descriptive results. Table 17 reports the OLS regression results examining the impact of public and teacher preference data on state education policymakers’ positions on the teacher evaluation policy change. Model 1reports only the treatment effects and shows that, on the whole, exposure to public or teacher preference data did not impact state education policymakers’ position on the teacher evaluation policy change. However, as shown in the coefficient on the party-by-treatment interaction term in model 2, when taking into account state education policymaker characteristics, political party does appear to have a modest impact on the position state education policymakers take on the teacher evaluation policy change. To examine this political party impact further, I include a treatment-by-political party interaction term in models. As shown in model 4, there is a significant interaction between the teacher preference treatment and political party: for Democrats, exposure to teacher preferences to oppose the policy change led to less opposition to the policy change than those with no exposure to teacher preferences. Finally, to further explore the statistically significant interaction 136 between political party and teacher preference treatment, I run model 1 separately for Republicans and Democrats. As shown in the right-most four columns, I find that there is not a statistically significant relationship between policymakers’ opposition to the policy change and exposure to teacher preferences. Instead, I find that Republicans are more likely to oppose the policy change when exposed to data that teachers would prefer they support the policy change. It should be noted that Republicans were an underrepresented group among survey respondents; therefore, the results of this analysis may underestimate this effect. It is possible that state education policymakers’ non-responsiveness may simply be the result of the survey question design: policymakers were presented with public or teacher preference data from a national poll, not a state or local poll. Similar to how American voters express frustration with Congress but are okay with their own representative (Mendes, 2013) or how Americans believe their local schools are doing well, but the nation’s schools are performing poorly, perhaps policymakers feel as though the preferences of voters and teachers in their constituency preferences are different than the preferences of American voters and teachers at large. Analysis of text from the open-ended response question that followed this experimental question provide some evidence that this is not the case. Just under half of the survey respondents (47.5 percent) provided some commentary about their choice to either support or oppose a change in teacher evaluation policy. Not a single respondent mentioned that the preferences provided to them via the Harvard/PEPG poll data were not reflective of voters or teachers in their constituency. Instead, those who were in favor of removing student assessment scores from teacher evaluation typically commented on three issues: (a) the importance of taking into account multiple measures of student achievement in a teacher’s evaluation, (b) the difficulty of taking into account student backgrounds, and (c) the 137 validity of the use of standardized assessments in teacher evaluation. With regard to multiple measures, one respondent wrote, “Although I'm in favor of using state assessment data in teacher evaluation and tenure decisions, I do believe that this should be one of many other components used to evaluate teachers. Thus, I would also favor a cap on the percentage that state assessment results can be used to evaluate teachers.” Another respondent stated, “The NFL doesn't draft a player on a one day performance neither should we. Teachers should be evaluated on a whole host of items which deals [sic] with out comes [sic] and student growth. Not on whether or not their students are good test takers!” Similarly, those who focused on student backgrounds made statements such as, “There are simply too many variables that enter into a child's ability to learn to make tests a reliable source of teacher evaluation.” Finally, one respondent’s concern about validity and reliability was stated as: “There are many teachers in situations where statewide testing does not provide a valid measurement of their job performance.” Those who were in opposition to removing student assessment scores from teacher evaluation focused on the importance of teacher accountability for student achievement. As some respondents wrote: “Teachers should be able to produce achievement, measure by growth, in all students regardless of the student's situation or background”; “The concept of including an indicator of student performance in teacher evaluations is a critical component of education system and practitioner accountability”; and, “Unfortunately, poor-performing teachers have been left in classrooms too long, even until they retire. If principals won't hold them accountable, then some other authority should.” 138 Figure 16 Results of experimental survey question, by control and treatment group Support 0% 10% 20% Oppose 30% 60.0% 40% 50% 60% 70% 80% 40.0% Baseline 58.8% 41.2% Treatment A - Public pressure to support 48.6% Treatment B- Teacher pressure to oppose 139 51.4% 90% 100% Table 17 Probit and OLS regression results: Responsiveness of state education policymakers to public and teacher policy preferences Model 1 Probit OLS All State Education Policymakers Model 2 Model 3 Probit OLS Probit OLS Model 4 Probit OLS Probit OLS Probit OLS Treatment A: Exposure to public preference (oppose) 0.06 (0.24) 0.02 (0.09) 0.23 (0.28) 0.08 (0.10) 0.13 (0.35) 0.05 (0.13) 0.19 (0.27) 0.07 (0.10) -0.09 (0.38) -0.03 (0.14) 0.16 (0.42) 0.16 (0.14) Treatment B: Exposure to teacher preference (support) 0.28 (0.25) 0.11 (0.10) 0.32 (0.27) 0.12 (0.10) 0.31 (0.27) 0.11 (0.10) 0.65 (0.32)** 0.23 (0.12)** -0.08 (0.42) -0.03 (0.16) 0.77 (0.39)** 0.25 (0.13)** -0.49 (0.22)** -0.17 (0.08)** -0.57 (0.29)* 0.20 (0.11)* -0.19 (0.27) -0.07 (0.10) 0.19 (0.42) 0.07 (0.16) -0.79 (0.42)* -0.28 (0.15)* 0.03 (0.21) 0.32 (0.33) -0.15 (0.21) x -0.38 0.10 187 0.01 (0.08) 0.12 (0.13) -0.05 (0.08) x 0.36 0.14 187 x -0.61 0.07 84 x 0.27 0.13 90 x -0.34 0.11 91 x 0.40 0.20 97 Democrat Treatment A x Political party Teacher B x Political party Female Nonwhite SBE Census region Constant !! N x -0.28 0.01 200 x 0.39 0.01 200 0.05 (0.21) 0.25 (0.32) -0.15 (0.20) x -0.24 0.09 187 0.01 (0.08) 0.09 (0.12) -0.05 (0.08) x 0.41 0.12 187 0.05 (0.21) 0.27 (0.33) -0.14 (0.21) x -0.21 0.09 187 ***p<0.01, ** p<0.05, *p<0.10 140 0.01 (0.08) 0.10 (0.13) -0.05 (0.08) x 0.42 0.12 187 Democrats Republicans Findings from the analysis of this experimental survey question provide no support for the hypothesis that, when compared to legislators, SBE members will be less responsive to public preferences or more responsive to teacher preferences (H1A, H1B). Additionally, these findings provide no support for the hypothesis that state education policymakers with a liberal political orientation will more responsive to public preferences than those with a conservative orientation (H2A). Finally, these findings negate, and completely flip on its head, the hypothesis that state education policymakers with a liberal political orientation will be more responsive to teacher preferences than those with a conservative orientation (H2B). Finally, correlation analyses shown in Table 18 show that the majority of sub-structure institutional structures of SBEs and legislatures are not associated with policymaker responsiveness to public and teacher preferences. One institutional structure of legislatures – term limits – is modestly associated policymaker responsiveness: policymakers in without term limits are more responsive to public preferences. This finding provides modest supports the hypothesis that term limited state education policymakers are less responsive than those who are non-term limited (H1D). However, hypothesis related to differences in responsiveness among elected and appointed SBE members and among legislators in more professional legislatures are not confirmed (H1C, H1E, H1F). 141 Table 18 Relationships between institutional characteristics and policymaker responsiveness Elected SBE No Leg Term Limits ! p n ! Statistic Reported Responsive to public preferences !! p 0.16 0.69 46 2.82* 0.09 Responsive to teacher preferences 0.34 0.34 38 0.17 0.68 n Leg. Professionalism ! p n 85 0.09 0.87 85 72 -0.28 0.66 72 ***p<0.01, **p<0.05, *p<0.01 Results: Whose Voices are Valued in the State Education Policymaking Process? Results from the experimental survey question indicate that exposure to public and teacher preference data does not result in state education policymakers voting in line with those preferences. In fact, when exposed to teacher preference data, Republican state education policymakers were more likely to go against teacher preferences. So, who, if anyone, does inform state education policymakers’ policy positions? Ratings of importance of actors in informing state education policymakers’ policy positions. As shown in Figure 17, when asked to rate the level of importance of various statelevel actors in informing their position on teacher evaluation policy, state education policymakers indicated that important actors included statewide school leadership organizations (e.g., school boards, superintendents, principals), other state officials (e.g., state department of education staff), and education research organizations (e.g., think tanks, universities). Governors, political party leaders, and statewide religious organizations received the lowest ratings. At the local-level, state education policymakers indicated that local schoolteachers, school leaders, and 142 constituents were important to informing their position on teacher evaluation policy. The groups that received the lowest ratings were local business groups, local government leaders (e.g., mayor, city council), and local teacher labor unions. Do state education policymaker characteristics impact ratings of importance of actors in informing policymakers’ teacher evaluation policy positions? As shown in Table 19, the differences in ratings of state and local actors among Democrats and Republicans are statistically significant. Differences in ratings of actors among education committee members and SBE members are approaching statistical significance. Are there particular actors that are driving these differences? Which actor’s ratings vary the most among education committee members and SBE members; and, among Democrats and Republicans? As shown in Table 20, there are statistically significant differences in ratings among SBE members and education committee members for (a) other state officials and (b) constituents that called or wrote. Further examination of these differences indicates that SBE members rated other state officials higher than education committee members; and, education committee members rated constituents that called or wrote higher than SBE members. There are also statistically significant differences among Democrats and Republicans for the ratings of both state and local teacher labor unions, statewide racial/ethnic group, local schoolteachers, and local education advocacy groups. When compared to Republicans, Democrats had a higher average rating of each of the aforementioned groups. Descriptive statistics of local and state-actor ratings by office and political party are available in Appendix K. 143 Figure 17 State education policymakers’ ratings of state- and local-level actors’ influence on education policy decisions 1 (least important) 0% 2 3 10% State level groups and individuals Statewide school leadership organizations Other state officials Education research organizations Constituents throughout your state Education advocacy/special interest groups or leaders Statewide teacher labor union organization Statewide business organizations Statewide racial or ethnic group-related organizations The governor Your political party leaders Statewide religious organizations Local level groups and individuals Local school teachers School leaders in your constituency Your constituents Constituents that called or wrote you on the issue Local policy/advocacy groups Local business groups Local teacher labor unions Local government leaders 144 4 20% 5 (most important) 30% 40% 50% 60% 70% 80% 90% 100% Table 19 Multivariate analysis of variance to examine relationship between state education policymaker characteristics and their rating of importance of state and local actors Source Model Female Non-white Democrat SBE member Region (base=New England) Wilk’s Lambda 0.05 0.87 0.75 0.57 0.71 0.21 F 1.21 0.61 1.37 3.11 1.66 0.91 p 0.03 0.89 0.17 <0.01 0.06 0.75 Table 20 Kolmogorov-Smirnov Test to compare distributions of rating data among state education policymakers by office and political party State actors Local actors Constituents throughout your state The governor Other state officials Your political party leaders Statewide teacher labor unions Statewide school leadership groups Statewide business organizations Statewide education advocacy/special interest groups Statewide religious organizations Statewide racial/ethnic groups Education research organizations Your constituents Local government leaders Local school teachers Constituents that called or wrote you Local teacher labor unions School leaders in your constituency Local business groups Local education policy or advocacy groups 145 D 0.07 0.15 0.23 0.15 0.11 0.08 0.11 0.07 0.05 0.16 0.01 0.17 0.13 0.17 0.21 0.12 0.10 0.07 0.17 SBE p-value 0.98 0.26 0.02 0.26 0.59 0.93 0.67 0.98 1.00 0.19 0.79 0.18 0.48 0.14 0.04 0.57 0.76 0.99 0.14 Democrat D p-value 0.05 1.00 0.06 0.99 0.02 1.00 0.09 0.91 0.48 <0.001 0.15 0.27 0.05 1.00 0.16 0.22 0.14 0.40 0.41 <0.001 0.22 0.03 0.06 1.00 0.18 0.12 0.29 0.002 0.07 0.97 0.48 <0.001 0.16 0.20 0.07 0.98 0.29 0.002 Discrete choice selections. Rating questions may sometimes result in positivity bias or a bias away from extreme ratings (Graves et al., 2009). Thus, the results of the discrete choice experiment provide a nuanced understanding of the actors most likely to influence state education policymakers’ teacher evaluation policy position. State education policymakers most frequently chose school leadership groups above all other actors. Education research organizations also emerged as an actor that state education policymakers most frequently selected to inform their teacher evaluation policy position. Governors, business organizations, and teachers unions were chosen at the lowest frequency as actors that inform a state education policymaker’s position on teacher evaluation policy. Responses for each discrete choice set are available in Table 21. State education policymaker characteristics and discrete choice selections. The selection of some actors over others varies by state education policymaker characteristics. Democrats more frequently chose teachers unions and education advocacy groups; and, Republicans more frequently chose business groups and the governor. There were also statistically significant differences among the discrete choice selections of education committee members and SBE members. SBE members more frequently selected education advocacy groups, school leadership groups, and the governor. In contrast, education committee members more frequently selected constituents. Descriptive statistics of discrete choice selections, as well as Chi-Square analysis results are available in Table 22. 146 Table 21 Discrete choice experiment results Choice Set 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Group A Education advocacy groups Education advocacy groups Education advocacy groups Education advocacy groups Education advocacy groups Education advocacy groups Education advocacy groups Education research orgs Education research orgs Education research orgs Education research orgs Education research orgs Education research orgs School leadership groups School leadership groups School leadership groups School leadership groups School leadership groups Teacher unions Teacher unions Teacher unions Teacher unions Your constituents Your constituents Your constituents Personal beliefs Personal beliefs Business orgs Frequency (Percent) 40 (25.5%) 31 (20.0%) 103 (66.5%) 56 (35.7%) 54 (34.2%) 112 (72.3%) 115 (72.3%) 69 (44.2%) 114 (72.2%) 93 (58.5%) 86 (55.5%) 136 (85.4%) 130 (83.3%) 121 (76.1%) 102 (64.2%) 87 (54.7%) 133 (84.2%) 135 (85.4%) 39 (25.0%) 35 (22.6%) 84 (53.5%) 83 (53.9%) 67 (41.6%) 130 (82.3%) 131 (82.9%) 127 (79.9%) 142 (89.3%) 83 (53.6%) 147 Frequency (Percent) Education research orgs 117 (74.5%) School leadership groups 124 (80.0%) Teacher unions 52 (33.6%) Your constituents 101 (64.3%) Personal beliefs 104 (65.8%) Business orgs 43 (27.7%) The governor 44 (27.7%) School leadership groups 87 (55.8%) Teacher unions 44 (27.9%) Your constituents 55 (41.5%) Personal beliefs 69 (44.5%) Business orgs 23 (14.6%) The governor 26 (16.7%) Teacher unions 38 (23.9%) Your constituents 57 (34.9%) Personal beliefs 72 (45.3%) Business orgs 25 (15.8%) The governor 23 (14.6%) Your constituents 117 (75.0%) Personal beliefs 120 (77.4%) Business orgs 73 (46.5%) The governor 71 (46.1%) Personal beliefs 94 (58.4%) Business orgs 28 (17.7%) The governor 27 (17.1%) Business orgs 32 (20.1%) The governor 17 (10.7%) The governor 72 (46.5%) Group B Table 22 Descriptive statistics of discrete choice selection and results of Chi-Square tests of independence of discrete choice selection variable by state education policymaker characteristics Descriptive Statistics Variable Teacher union selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican Democrat Education advocacy groups selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican Democrat Education research organizations selection count Gender Male Female Race White Non-white Office Legislature n 200 125 74 169 30 91 108 M 1.88 1.72 2.16 1.73 2.73 2.13 1.68 SD 2.24 5.35 4.41 4.77 5.79 5.67 4.43 Min. 0 0 0 0 0 0 0 Max. 7 14.50 0.04 11.67 0.14 5.30 0.64 48.99 <0.01 7.49 0.34 7.59 0.38 15.76 0.02 12.91 0.06 11.32 0.12 4.42 0.81 10.59 0.15 7 7 7 6 7 7 97 90 200 0.91 3.03 2.56 2.88 5.11 1.79 0 0 0 7 7 7 125 74 2.51 2.66 3.16 3.24 0 0 7 6 169 30 2.49 3.03 3.06 3.69 0 0 7 6 91 108 2.25 2.83 3.06 3.15 0 0 6 7 97 90 200 2.37 2.81 3.72 3.05 3.32 2.45 0 0 0 6 7 7 125 74 3.63 3.92 5.32 7.12 0 0 7 7 169 30 3.73 3.80 6.00 6.03 0 0 7 7 91 3.23 5.82 0 7 148 Chis-Square Tests of Independence p !! Table 22 (cont’d) State board of education Political party Republican Democrat School leadership groups selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican Democrat Constituents selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican Democrat Personal beliefs selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican 108 4.17 5.75 0 7 97 90 200 3.78 3.69 3.95 5.38 6.40 2.43 0 0 0 7 7 7 125 74 169 30 91 108 3.92 4.04 4.04 3.57 3.49 4.36 5.72 6.15 5.97 5.15 6.05 5.39 0 0 0 0 0 0 0 0 0 7 7 7 125 74 3.58 2.99 5.76 5.05 0 0 7 7 91 108 3.58 3.18 6.78 4.61 0 0 0 0 5.92 5.49 2.65 0 0 0 7 7 7 125 74 3.73 3.54 6.65 7.59 0 0 7 7 169 30 3.70 3.40 6.95 7.28 0 0 7 7 91 108 3.35 3.92 7.56 6.39 0 0 7 7 149 7.26 0 0.29 17.16 0.02 8.30 0.30 7.28 0.39 4.01 0.76 14.49 0.04 6.61 0.47 3.16 0.89 5.24 0.55 11.06 0.14 3.38 0.85 7 7 3.44 3.34 3.64 3.84 7.05 7 7 97 90 200 97 0.79 7 7 6.33 5.35 2.38 5.68 5.43 4.03 7 7 4.00 3.90 3.34 3.38 3.23 0.33 7 7 97 90 200 169 30 8.26 7 Table 22 (cont’d) Democrat Business organizations selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican Democrat Governor selection count Gender Male Female Race White Non-white Office Legislature State board of education Political party Republican Democrat 90 200 125 74 169 30 91 108 3.52 1.54 1.74 1.22 1.60 1.20 1.37 1.69 6.59 1.63 3.10 1.73 2.67 2.44 2.50 2.74 0 0 0 0 0 0 0 0 7 6 0 0 0 6 6 7 125 74 1.59 1.09 3.42 2.31 0 0 7 7 169 30 1.42 1.33 3.08 2.99 0 0 7 6 97 90 1.86 0.92 3.48 2.27 0 0 0 0 0.28 8.55 0.17 18.29 <0.01 7.91 0.27 2.86 0.88 17.98 0.01 21.20 <0.01 6 6 2.98 1.94 1.75 2.46 3.33 7.59 6 5 1.98 1.06 1.40 1.01 1.74 0.26 6 5 97 90 200 91 108 8.14 6 7 7 7 When controlling for other characteristics of state education policymakers with a negative binomial regression, these results still hold (Table 23): state education policymaker’s office and political party significantly impact their discrete choice selection. While Democrats more frequently selected education advocacy groups and teachers unions, Republicans more frequently selected business organizations and the governor. And, SBE members more frequently selected education advocacy groups, education research organizations, school leadership groups, and the governor. When taking into account other state education policymaker 150 characteristics, the positive, significant relationship between education committee members and the selection of their constituents no longer holds. Given that both political party and office had a significant impact on state education policymakers’ actor selection, I also examined whether there were interaction effects among office and political party. The only actor for which the office-by-political party interaction was significant was education advocacy groups: Republican education committee members, when compared to Republican SBE members, less frequently selected education advocacy groups (!=0.28, p=0.02). 151 Table 23 Discrete choice experiment analysis: Negative binomial regression results Education advocacy groups Model 1 Female Non-white Democrat SBE member Education level West North Central West South Central Mountain Pacific Model 2 Model 1 Model 2 Teacher labor unions Model 1 Model 2 Your constituents Model 1 Model 2 Personal beliefs Model 1 Model 2 Business organizations The governor Model 1 Model 2 Model 1 Model 2 -0.01 -0.03 0.09 0.05 0.02 -0.002 0.03 -0.03 -0.15 -0.17 -0.06 -0.10 -0.31* -0.33* -0.22* -0.17 (0.10) (0.10) (0.10) (0.11) (0.11) (0.24) (0.23) (0.15) (0.15) (0.19) (0.19) (0.20) (0.22) (0.13) (0.14) 0.08 0.10 -0.02 -0.02 -0.15 -0.16 -0.06 0.16 -0.03 -0.04 -0.07 -0.05 0.26 0.03 0.06 0.07 (0.15) (0.12) (0.19) (0.20) (0.24) (0.25) (0.18) (0.13) (0.28) (0.26) (0.18) (0.18) (0.28) (0.35) (0.27) -0.49*** (0.11) 0.20* 0.32*** 0.01 -0.07 0.06 0.12 1.27*** 1.51*** 0.04 0.16 -0.04 0.02 -0.59** -0.58* (0.22) 0.53*** (0.13) (0.10) (0.16) (0.14) (0.15) (0.14) (0.28) (0.30) (0.16) (0.14) (0.15) (0.14) (0.28) (0.33) (0.09) 0.29*** 0.32*** 0.28*** 0.29*** 0.25*** 0.26*** 0.01 0.11 -0.05 -0.03 0.18 0.22* 0.62*** 0.59** 0.23* 0.20* (0.07) (0.07) (0.09) (0.08) (0.08) (0.08) (0.16) (0.18) (0.15) (0.15) (0.14) (0.13) (0.25) (0.27) (0.13) (0.13) -0.04 -0.05 -0.03 -0.03 -0.03* -0.04 -0.08* -0.10* -0.05** -0.05** 0.01 -0.004 -0.08** -0.07* -0.07 -0.06 (0.04) (0.04) (0.02) (0.02) (0.02) (0.03) 0.05 (0.06) (0.02) (0.02) (0.05) (0.05) (0.04) (0.05) (0.04) (0.04) X -1.16*** South Atlantic East North Central Model 1 School leadership groups (0.11) Census Region (0=New England) Middle Atlantic East South Central Model 2 Education research organizations -0.12*** 0.76*** 0.28*** 0.35*** -0.10** 0.28** (0.04) (0.03) (0.05)** (0.10) (0.05) (0.11) (0.12) -0.002 0.11* 0.20*** 0.004 0.24*** 0.17*** 0.36*** -0.03 (0.05) (0.07) (0.06) (0.06) (0.06) (0.05) (0.13) (0.07) X X X X X X 0.28*** 0.34*** 0.29*** 0.11 0.35*** 0.45*** (0.04) (0.05) (0.05) (0.08) (0.06) (0.05) 0.67*** 0.33*** 0.25*** 0.29*** 0.36*** 0.44*** (0.03) (0.05) (0.04) (0.10) (0.05) (0.06) 0.51*** 0.41*** 0.34*** 0.83*** 0.54*** 0.47*** (0.06) (0.05) (0.06) (0.17) (0.05) (0.05) 0.39*** (0.15) 0.20 (0.20) 0.28 (0.24) 0.66*** -1.71*** (0.02) X 0.13* (0.08) -0.27*** (0.09) 0.12** (0.06) 0.62*** 0.26*** 0.40*** 0.25*** 0.64*** 0.24*** (0.04) (0.6) (0.05) (0.08) (0.6) (0.06) (0.17) 0.35*** (0.07) 0.41*** 0.41*** 0.43*** 0.52*** 0.43*** 0.53*** 0.26** -0.32*** (0.02) (0.03) (0.03)* (0.07) (0.04) (0.03) (0.12) (0.03) 0.19** 0.20** 0.17 -0.17 0.17* 0.26*** 0.80*** -0.23* (0.08) (0.09) (0.13) (0.09)** (0.11) (0.08) (0.13) (0.13) 152 Table 23 (cont’d) Constant 0.85 0.50** (0.20) (0.21) 1.29*** (0.14) 1.01*** 0.09 1.39*** (0.13) 1.12*** (0.16) 0.25 -0.17 (0.32) (0.39) 1.51*** (0.16) 1.12*** (0.11) 1.22*** (0.22) 0.90*** 0.64*** 0.30 (0.23) (0.14) 0.1 6 (0.18) 0.8 5 (0.16) (0.26) /lnalpha -2.33 (0.78) -3.99 (3.38) -1.39 (0.48) -1.61 (0.55) -1.56 (0.54) -1.75 (0.61) 0.04 (0.20) -0.04 (0.19) -1.12 (0.47) -1.33 (0.52) -0.81 (0.31) -0.92 (0.33) alpha 0.10 (0.08) 0.02 (0.06) 0.25 (0.12) 0.20 (0.11) 0.21 (0.11) 0.17 (0.11) 1.05 (0.21) 0.96 (0.18) 0.33 (0.15) 0.27 (0.14) 0.45 (0.14) 0.40 (0.13) 187 187 187 187 187 187 187 187 187 187 187 187 N *p<0.10, **p<0.05, ***p<0.01 153 187 -0.26 (0.18) 0.88*** (0.21) 0.88*** (0.20) -0.78 (0.39) -1.00 (0.49) 0.77 (0.14) 0.46 (0.18) 0.37 (0.18) 187 187 187 Most and least important actors to inform state education policymakers’ policy positions. The analysis of the discrete choice experiment question data largely focuses on the actors that influence state education policymakers when policymakers are faced with just two actors at a time. However, policymaking does not take place in a vacuum and policymakers are likely influenced by more than one actor at a time. To fully explore the relationship between difference sources of influence, I examine the actors that state education policymakers selected as most and least important in informing their position on teacher evaluation policy. As shown in Figure 18, state education policymakers most frequently selected local schoolteachers as one of the most important groups in informing their position on teacher evaluation policy. Much like previous findings, Democrats were more likely than Republicans to select local teachers as the most important group (! ! 1 = 3.72, ! = 0.05). State education policymakers most frequently selected statewide religious groups and political party leaders as the least important groups in informing their position on teacher evaluation policy. There were no differences in the selection of these groups as least important by any state education policymaker characteristics. Taken as a whole, the selection of statewide teacher unions, governors, and statewide education advocacy groups as important informants were indiscriminant: while some state education policymakers find these groups to be the most important in informing their teacher evaluation policy position, others find them to be the least important. Consistent with previous findings, it was political party and office differences that drive these findings. When broken down by political party, Democrats were more likely to select teacher labors unions (! ! 1 = 29.9, ! < 0.01) and education advocacy groups (! ! 1 = 4.86, ! = 0.03) as most important; and, Republicans were more likely to select the governor as most important (! ! 1 = 7.07, ! < 154 0.01). When broken down by office, SBE members were more likely to select the governor as most important (! ! 1 = 5.21, ! = 0.02). Are some actors more likely to be nominated alongside others? There were statistically significant relationships between the nominations of particular actor clusters as least and most likely to inform state education policymakers teacher evaluation policy position (least likely: ! ! 12 = 128.7, ! < 0.001; most likely: ! ! 12 = 69.9, ! < 0.001). There is substantial overlap between state policymakers’ selection of actors in the “other elected officials” cluster, “educators and their representatives” cluster, and “other interest groups” cluster for the groups least likely to inform their position. With regard to state education policymakers’ nominations of actors most likely to inform their position, there is considerable overlap between the selection of actors in the “constituents” and “educators and their representatives” clusters, between the “other elected officials” and “educators and their representatives” clusters, and between the “educators and their representatives” and “education research and advocacy” clusters. There are also statistically significant relationships between state education policymakers’ nomination of most important and least important actors (! ! 16 = 26.7, ! = 0.05). In particular, state education policymakers that nominated actors in either the “other elected officials” or “other interest groups” cluster as most important tended to nominate an actor in the “educators and their representatives” cluster as least important. Descriptive statistics on the number of nominations for each actor, number of nominations for the clusters of actors, and cross-tabulations of group selections as most and least important are available in Appendix L. 155 Figure 18 State education policymakers’ nominations of most and least important actors in informing position on teacher evaluation policy Number of Nominations Most Important 80 70 60 50 40 30 20 10 0 156 Least Important Discussion Variation by volume and independence of voice. Although the public is the primary funder of a state’s public education system and the electorate for many state education policymakers, I find that constituent voices are seemingly absent from the state education policymaking table. This finding aligns with a long line of research in political science that suggests that “state legislators are not over-wrought by an inclination to seek out the views of constituents on which policy alternatives should be selected” (Uslaner & Weber, 1979, p. 580). Second, as both one of the largest voting constituencies in the nation and the actors closest to the education process, teacher voices are only heard by some policymakers; and, they are heard in a way that does not lend favor to them. These findings provide some support for what has been termed “the war on teachers” (Giroux, 2013; Goldstein, 2015). Giroux (2013) contends that, unlike past educational reform movements, recommendations that emerge in current education debates ignore the “intelligence, judgment, and experience that teachers might offer in such a debate” (pp. 3-4). However, the results of my study suggest that it is more complex than this. A holistic look at the results of this study reveals that all policymakers value individual, local teacher voices; however, teacher union voices are not universally valued. This finding provides support for Eric Hanushek’s (2011) proposition that there “is not a war on teachers en masse.” Hanushek argues that what we are seeing is “a war on the blunt and detrimental policies of teachers’ unions,” whose interests “aren’t the same as the interests of children, or even of most teachers.” This study cannot explain why Republicans who were exposed to teacher preferences were less likely to align with them. We also cannot infer whether this behavior is the result of a war on teachers, a war on teachers unions, or something else. As such, additional research – 157 particularly qualitative research that allows the researcher to explore questions of why and how – is needed to provide a greater understanding of policymakers thought processes and behaviors during the education policymaking process. The findings of this study may also be tightly coupled with the policy issue examined – teacher evaluation policy. It may be that, rather than turn to those who are personally impacted by the policy, policymakers prefer to turn to individuals that they deem as “outside” authorities or experts in the area of teacher evaluation – education research organizations and school leaders who will be implementing teacher evaluation. Some evidence of this is found from both the ratings and DCE selections of state education policymakers: state education policymakers both gave high ratings to and frequently selected school leadership groups and education research organizations above all other actors as important to informing their position on teacher evaluation policy. At the same time, teacher labor unions, governors, and business groups, on average, received low ratings and were the least frequently selected actors within the DCE. It is also possible that, given pressure from the federal government to implement a statewide teacher evaluation system, state education policymakers felt that they were unable to be responsive to any constituency. Future research is necessary to understand whether these findings hold when other education policies – both related and unrelated to teachers’ work – are considered by state education policymakers. For example, future research could examine the value of difference voices when state education policymakers consider other teacher-specific policies such as teacher education or certification requirements. Additionally, research could examine whether different voices are valued when considering policies related to students such as seclusion and restraint policies, bullying, or graduation requirements; or, when considering policies related to school governance and management such as school choice policies or school 158 calendar restrictions. By broadening our understanding of how policies focused on different actors, we may be able to develop a more generalizable framework for whose voices are valued by which state education policymakers when policies focus on different K12 public education entities and individuals. Institutional structures and political parties impact the value of external voices. Although some actors emerged as important informants to all state education policymakers, the voices that were valued by state education policymakers varied along political and institutional lines. Democrats consistently valued teacher union and education advocacy group voices to a greater extent than did Republicans. Legislators consistently placed higher value on the voices of constituents. The voices valued by SBE members were less consistent across the multiple measures in the survey. Results from the 1-to-5 rating of actors showed that SBE members, when compared to legislators, gave statistically significant higher ratings to other state officials (e.g., state education agency staff). In contrast, results from the DCE indicated that, when compared to legislators, SBE members more frequently selected education advocacy groups, school leadership groups, and the governor. This finding may suggest that although SBE members may place a high rating value on the voices of state officials, when forced to choose between two informants, SBE members are more likely to listen to education advocacy groups, school leadership groups, and the governor. In essence, this enigma of high ratings but low selection frequency in the DCE may reflect the differences in state education policymakers desires and actions. Evidence for a broader disconnect between the desires and actions of all state education policymaker may be found in the results of the most/least important selection exercise. While the 159 experimental question revealed that state education policymakers are not responsive to the policy preferences of the general public or teachers, state education policymakers often placed local schoolteachers and local constituents in the category of one of the most important actors in informing their teacher evaluation policy position. This finding may very well be the result of respondent social desirability bias: policymakers may have rated local school teachers and constituents highly and even selected them as one of the most important groups in informing their policy position because they believe that this is “good behavior.” However, when forced to choose between the preferences of local schoolteachers or constituents and those of other actors like school leadership groups, education advocacy groups, or the governor, the latter win out. This finding could be indicative of the immense influence that interest groups and the executive office can have on state education policymakers when the rubber meets the policymaking road. Limitations The findings in this chapter are limited by the small sample size and survey response rate. While the results of this study should be interpreted with caution, the demographics of survey participants were quite representative of those of all state education policymakers. This should provide some solace that the results of this study are not spurious. The findings of this study are not generalizable to other education policy issues. I only examine one policy issue – teacher evaluation. Given the highly politicized environment in which this policy has been debated in recent—a context in which states were induced to take action on teacher evaluation policy by the federal government’s Race to the Top and ESEA wavier opportunities—the results of this study may be reflective of state education policymakers response to those federal inducements, rather than to their constituencies. More research is 160 needed to examine the responsiveness of other state-level education policy issues. Such research would not only to provide a better understanding of whether the findings of this research hold across education policy issues but, such research could also examine whether whose voices are valued in the state education policymaking process vary for top-down and bottom-up education policy proposals. The research is also limited in the inferences that can be made from its findings. The goal of this research is to examine the responsiveness of state education policymakers to the preferences of two key constituencies – the public and teachers – and to understand the ways in which institutional and individual characteristics impact whose voices are heard in the state education policymaking process. While I explore whose voices are heard, the results do not warrant generalization about whose voices should be heard. And, more importantly, this research does not purpose to place any value on the voices of any one group. For example, my research suggests that voices of actors like school leadership groups, education advocacy groups, and education research organizations were highly valued by state education policymakers when forming their teacher evaluation policy position. Political science research suggests that having the voices of interest groups at the policymaking table may actually serve the public interest. For one, interest groups often have a thorough understanding of how the policy decisionmaking process works and how to effectively use their resources; and, partly due to the credibility and policy expertise that interest groups have with legislators, interest groups have more access to legislators than ordinary citizens have (Berry, 1997). Moreover, the doors of policymakers may be open wider to leaders of interest groups, who often represent a large constituency of people, than to an ordinary citizen (Berry, 1997). However, others have suggested that the most successful interest groups are the ones that represent the narrowest interests and, as a result, do 161 not represent most people (Berry, 1997). Additionally, some have argued that because a primary activity of interest groups is contributing money to political candidates, representatives may be focused on returning favor to interest groups while abandoning constituent interests (Berry, 1997). Strengths and weaknesses for valuing constituent voices could also be debated. From Jean-Jacques Rousseau to John Stuart Mill and Robert Dahl, theorists have provided a strong case for a connection between public policy and public preferences as a critical component of representative democracy. However, A. Rosenthal (1998) warns of the dangers of excessive emphasis on public opinion. Since the public is often ill-informed about issues and their connection with larger public goals, A. Rosenthal (1998) argues that public opinion may lack coherence. Moreover, given that the public often mobilizes around narrow issues, it may be that acting in accordance with public opinion could result in uncoordinated, unworkable policies that fail to meet the larger needs of society (A. Rosenthal, 1998). Thus, perhaps a Burkian38 style of policymaker responsiveness— in which an elected representative should seek and consider the advice of his constituents, but it should be but one input into his decisionmaking—is desirable. Additionally, although responsiveness by government officials to the public remains an underlying democratic value, it is important to keep in mind that democratic responsiveness cannot be measured solely by the alignment of citizens’ policy preferences and policymaker actions (Pitkin, 1967; Powell, 2003). Key to this argument is that it is empirically difficult to assess what citizens want (Powell, 2003). In this research, policymakers were presented with preferences of the public and teachers collected via a survey. Surveys, however, may raise questions for policymakers such as: how informed are the respondents? Moreover, surveys on a 38 As in, political theorist and philosopher Edmund Burke. 162 single policy issue do not capture policy trade-offs across issues that pervade the policymaking process and questions are typically worded in ways that do not capture the full complexities of specific policies (Powell, 2003; Miller & Stokes, 1964). As a result of these limitations in quantitative data, qualitative data, I turn to qualitative data to explore state education policymakers approach to the education policymaking process. More specifically, I examine whose internal – and, subsequently, external –voices are valued by state education policymakers. 163 Chapter VI Donut Devourers, Fish Fanatics, Politicians, and Educators: Voices in the State Education Policymaking Process While quantitative analyses in Chapters IV and V expand our knowledge of the outward appearances and actions of state education policymakers, demographic and survey data are unable to provide deeper insight into the more inward qualities of state education policymakers. In particular, quantitative analyses revealed that who gains state education policymaking power and whose voices are heard by state education policymakers vary between and within policymaking bodies; and, the latter – whose voices are heard – also varies by state education policymakers’ political persuasions. Yet, quantitative data cannot answer the question of “how?” How do politics and education governance structures impact who gains education policymaker power and whose voices are heard in the policymaking process? Moreover, quantitative analyses provided insight into the outside voices that are valued by state education policymakers. However, discussion of these results assumed that all state education policymaker voices carried equal weight in the education policymaking process. Yet, this may not be the case. As discussed in Chapters II and III, the voices of legislators and SBE members may not be given equal weight in the policymaking process. For example, as shown in Figure 19, if SBE members value school leaders’ voices but the voices of SBE members carry relatively little weight in the internal education policymaking process, then the voices of school leaders will likely carry little weight in the policymaking process. This model could be extended to different political party affiliations: if Democrats value the voices of teachers but the voices of Democrats carry little weight in some state’s education policymaking processes, then the voices 164 of teachers will likely also carry little weight in the policymaking process. As such, it is also important to understand whose voices are heard within state education governance structures. Figure 19 Proposed framework: Weight of internal voices influences weight of external voices heard in the education policymaking process Policymaker Characteristic: Office Legislator Valued External Voice Internal Voice Weight Local teachers High + SBE Member Low School leaders The purpose of this chapter is twofold: (1) to explore the inward qualities of state education policymakers to better understand how and why they came to their position of power, and (2) to examine how state politics and education governance structures may impact whose voices are heard in the policymaking process – both internally and externally. To answer these questions, I turn to state education policymakers themselves, meeting them on their own turf to learn more about their experiences, goals, perceptions, and beliefs about the education policymaking process. 165 Data and Methods This study employs a multiple case study (Yin, 2009) approach to explore how state education policymakers came to their positions of power and how they approach the education policymaking process. This approach allows for the investigation of a contemporary phenomenon – in the case of this research, how policymaking power is gained and acted upon – “within its real-life context when the boundaries between the phenomenon and context are not clearly evident” (Yin, 2009, p.13). The primary unit of analysis in my case study is the individual policymaker that is embedded within a policymaking body (i.e., state legislature or SBE), which is rooted within the context of a larger state education governance structure. Theory development was an integral part of the design phase of this research (Yin, 2009). In particular, existing research provided a rich theoretical framework from which I selected my case sample. Focusing on the role of institutional structures, as described in previous chapters, I maximized variation among my case study sites based upon state education governance structures. In particular, I sought out interviews with state education policymakers located in six states, each with a slightly different education governance structure (see Table 24). Data. Data for this study are comprised of transcripts from semi-structured interviews with a convenience sample of state education policymakers located in six states collected from March to October 2016. A total of 44 interviews were conducted: 19 with state legislators serving on a Senate or House education committee, 22 with state board of education (SBE) members, and 3 with non-voting SBE members or staff. Interviews ranged from 45 to 120 minutes. 166 Table 24 Qualitative case study site demographics State SBE Structure Duckburg Mix of region-based governor appointments and at-large legislature appointments Elsewhere Region-based governor appointments, approved by legislature (includes demographic equality requirement) Hennepin Region-based governor appointments, approved by legislature Hogsmeade Mix of region-based and at-large governor appointment, approved by legislature Maycomb Statewide, partisan election Narnia Mix of region-based, non-partisan election and at-large governor appointments, approved by legislature Legislature Structure Legislative Term professionalism Limits Low No Medium No Low No Medium No High High Yes Yes Informants were first recruited from late March to early April 2016. A second round of recruitment took place in early June 2016. My target sample consisted of all state education policymakers in my six case study states—a total of 261 state education policymakers (67 SBE members, 174 legislators serving on education committees). I contacted each state education policymaker via e-mail to invite them to participate in the study and offered the option of a faceto-face, video, or telephone interview. For those that did not respond to my initial e-mail, I sent a follow-up e-mail one and two weeks later, and again in early June. In total, 17 percent (44) of my target sample agreed to participate in an interview. Response rights were substantially higher for SBE members (39 percent) than for legislators (11 percent). Nonetheless, as shown in Table 25, I conducted at least one interview with a state legislator serving on an education committee and at least two SBE member interviews in all six of states. Interviewees were not provided with any incentive for participation. When looking across Tables 24 and 25, it is evident that state legislators in states with a higher professionalism 167 index were more likely to participate in the study. Given that legislators in these states are likely in their office with staff that are able to check e-mail messages on a regular basis, it is perhaps unsurprising that their response rate was higher. Table 25 Qualitative interview sample population demographics, by state Number of Interviewees State Duckburg Elsewhere Hennepin Hogsmeade Maycomb Narnia Legislators serving on education committees 3 2 1 1 5 7 SBE members (including staff) 2 3 6 2 6 7 Interview Demographic Characteristics Percent Percent Percent Male White Democrat 60% 100% 40% 80% 100% 40% 67% 83% 67% 100% 100% 50% 55% 91% 55% 57% 93% 57% Of those that were not included in my sample, three confirmed an interview and then were not able to complete the interview; and, 11 responded to my invitation to indicate that they were unable to take part in the study. The remaining members of the target sample (204) were non-responsive to recruitment efforts. Methods: Designing the interview protocol. All interviews were conducted using a semi-structured interview protocol. Semi-structured interviews are often used in policy research (Harrell & Bradley, 2009) and consist of a guide with questions and topics that must be covered. While the questions are standardized, semi-structured interviews provide the interviewer with some discretion about the order in which the questions were asked and allow for probing (Harrell & Bradley, 2009). This allows for the data to be collected in a more conversational environment (Harrell & Bradley, 2009). 168 The interview protocol focused on soliciting responses to both descriptive and structural questions. Additionally, allow my interviewees to become more comfortable with the interview process, I designed a funneled protocol in which broad questions were followed by more pointed questions (Harrell & Bradley, 2009). I began the interview with descriptive questions to gain an understanding of how state education policymakers came to their position of power and how they viewed their role in making education policy. In the second half of the interview, I engaged the interviewee in structural questions to help me understand the state education policymakers’ perceptions and beliefs related to the education policymaking process and whose voices are and should be heard. Upon initial development of the semi-structured interview protocol, I sought input from colleagues and mentors that had experience with conducting semi-structured interviews. Additionally, I conducted two pilot studies. First, in spring 2015, I conducted a pilot study with a former SBE member and a former staff member of a state legislator. The feedback I received from these efforts was taken into consideration when revising my interview protocol for my second pilot study. In summer 2015, I conducted a second pilot study that involved interviews with 10 statelevel education policy actors in a single state. In an effort to obtain meaningful feedback on the strengths and weaknesses of the interview protocol, I chose to interview a broad range of actors that interact with formal state education policymakers. Although I did interview three state legislators serving on education committees, I also interviewed a member of the Senate nonpartisan fiscal analysis service, an individual serving as legal counsel and policy advisor to the House or Representatives, a member of the Senate majority policy office, a legislative liaison to the governor, and three chief of staff for legislators serving on education committees. 169 The purpose of the second pilot study were many: learning about the field time necessary to complete the interview, refining my research questions to understanding the types of probes that may be necessary, and examining whether proposed questions are inappropriate or too complicated (vanTeijlingen & Hundley, 2001). At the conclusion of each interview, I asked pilot study participants for feedback to identify ambiguities, difficult questions, or questions that I should have asked (Chenail, 2011). All pilot study interviews with digitally recorded and transcribed. I reviewed transcripts to assess whether each question provided meaningful information that was necessary to understanding the phenomenon of interest (Chenail, 2011). The second pilot study also served as a learning tool to help me understand how to present oneself appropriately to policymakers as a researcher (Kim, 2010). I strove to guide my qualitative research with by an ethic of caring in a culturally appropriate way, which required continuous reflection about my role as a researcher (Hill, 2006). After each pilot interview, I journaled about how both my demographics and my position as a researcher may have shaped the interview process (Hill, 2006). As one illustrative example, I wrote about one pilot interviewee’s perception of me as a researcher: Jack39 was explaining that if there were multiple classrooms of 27 kids in a grade and the grade lost five kids, the school would be asked to cut a teacher. I nodded and said, “Right, and it is hard to do that since those kids may be spread out among multiple classrooms.” Jack looked up and, without hesitation, said, “Wow! You are a lot smarter than you look…and…well it’s not that you don’t look smart…it’s just…” Jack fumbled with his paper a bit. I wanted to move the interview along so I moved on to the next question. This, among many other reflections, helped me to both prepare me for uncomfortable topics that may arise in future interviews and reflect on how I dialogue with political elites. 39 To protect anonymity, all names have been replaced with pseudonyms. 170 Additionally, in reviewing my pilot study journal entries, I came to understand that while I staunchly emphasized my role as a neutral researcher, sometimes my detached attitude hindered me from building rapport with the interviewees. Although it may be that engaging in non-topical discussions with interviewees is seen as inefficient and unnecessary to understanding the phenomenon of interest, I found that allowing the interviewee to elaborate on topics that they were passionate about signaled to the interviewee a sense of care and respect. Additionally, providing the interviewee with the opportunity to talk about their interests allowed me as the interviewer to become comfortable with presenting myself in a culturally appropriate way in order to facilitate trust. For example, once I became comfortable “chit-chatting” with the interviewee, I noticed that the way the interviewee sat and spoke changed too. This reflection guided the way I conducted future interviews. Methods: Conducting the interviews. All participants were provided with the option of a face-to-face interview in a location convenient to them. Compared to telephone interviews, face-to-face interviews offer a number of advantages. For one, face-to-face interviews allow the interviewer to observe visual cues (Garbett & McCormack, 2001) and nonverbal data (e.g., emotion, gestures, actions) (Burnard, 1994; Fontana & Frey, 2005). Face-to-face interviews also lessen potential distractions of participants by activities in their environments (McCoyd & Kerson, 2006; Opdenakker, 2006). Finally, telephone interviews may reduce rapport, probing, and in-depth discussion (Opdenakker, 2006; Sweet, 2002). Therefore, I desired to conduct faceto-face interviews with as many policymakers as possible to establish trust and rapport with the hope that a positive experience would be shared with colleagues, who would then be willing to participate in an interview. For those who could not participate in a face-to-face interview, I 171 offered the option of either a videoconference or telephone interview. In total, 50 percent (22) of the interviews were completed face-to-face, 45 percent (20) via telephone, and 5 percent (2) via videoconference. As another way of establishing trust and rapport with my informants, I provided all interviewees with the option of having the interview digitally recorded. All but two informants agreed to be digitally recorded. For those interviews that were not digitally recorded, I took detailed notes and composed a post-interview memo immediately after the interview concluded. All digitally recorded interviews were transcribed verbatim. To gain intimate familiarity with my data and complete analytic memos along the way, I chose to transcribe the interview data myself (Merriam & Tisdell, 2016). Finally, all interviewees were offered an opportunity to member check – that is, the interview could review their interview transcript for accuracy or to clarify any thoughts or ideas expressed during the interview (Merriam, 1998). When offered this option at the end of the interview, nearly all interviewees expressed thanks, but did not take up the opportunity; however, many interviewees did express interest in seeing the final written research product. Analysis My analysis of coded interview transcripts focused on discerning patterns of thought among state education policymakers (Yin, 2009). Informed by both the theoretical frameworks and quantitative analysis findings in Chapters IV and V, the qualitative analysis of interview transcript data focused on variations in thoughts among state education policymakers of different political persuasions that were embedded in distinct policymaking bodies with assorted institutional structures. My unit of analysis was individual themes – that is, the expression of a 172 thought or idea – rather than physical linguistic units (e.g., word, sentence, paragraph) (Minichiello et al., 1990). Coding of interview transcripts was both deductive and inductive. In particular, my first round of coding was deductive, driven by the theories that drove my research questions (Strauss & Corbin, 1990). Subsequently, I engaged in a second round of coding that reflected inductive coding in which I allowed patterns and themes to emerge (Strauss & Corbin, 1990). To preserve the uniqueness of each individual but generate cross-cutting theories, I followed a within- and cross-case synthesis of the interview transcripts (Miles, Huberman & Saldana, 2014; Yin, 2009). In short, coded transcripts were first analyzed within and then across state education policymakers (Miles, Huberman, & Saldana, 2014; Yin, 2009). The within-case analysis focused on becoming intimately familiar with each state education policymaker (Eisenhardt, 1989). The across-case analysis then focused on understanding the themes and patterns among state education policymakers with similar or different individual characteristics embedded within various institutional structures. Throughout the analysis phase, I strove to establish grounds for credibility, dependability, and conformability. As recommended by Lincoln & Guba (1985), I had prolonged engagement in the field (20 months from pilot to completion of data collection); triangulated data, both among interviewees and between interviewees and documents relevant to state governance structures and the policy of study; and engaged in negative case analysis, peer debriefing, and member checking. 173 Results: Who Gains State Education Policymaking Power? And, How? Findings from my quantitative analyses revealed little variation in the outward characteristics of state education policymakers: the majority of state education policymakers are white businessmen; although, SBEs are more representative of women. However, we still know very little about the inward qualities of state education policymakers. For example, how did state education policymakers come to their position of power? What are state education policymakers’ goals and priorities? And, in what ways do politics and state education governance structures impact who gains education policymaking power? Coming to power. While there is wide variation in why and how state education policymakers in my study came to power, most articulated that they did not seek out this power. The ways in which state education policymakers came to power were diverse, ranging from those interested in education with no aspirations of policymaking power who were recruited to their position of power, to those simply interested in politics and policy who were ascribed to an education committee by political party leaders. In what follows, I separate how individuals describe their ascendancy to power into two camps: power seekers and recruits. As described below, even within these two groups there is wide variation in the types of power sought out, the ways in which individuals are recruited, and the politics that permeates the processes by which state education policymaking power is gained. Power seekers. Those seeking policymaking power were primarily legislators. However, most legislators sought to simply gain policymaking power, not education policymaking power. Many education committee member interviewees described how they came to their position of 174 power by listing other political positions they held: county commissioner, township trustee, city councilmember, local school board member. In running for state office, 15 of the 19 education committee member interviewees did not state that they sought out a seat on the education committee; rather, most discussed how they were assigned to serve on the education committee by political party leaders. These individuals – whom I describe as “recruits” – are described in further detail in the subsequent section. Of the few legislators that who were motivated to run for office to be involved in education policymaking, all were Democrats. These legislators frequently drew on their their perceived education expertise or experiences and felt compelled to run for office in hopes of making an impact on education, as illustrated in the following statements: Senate Bill [X] kind of started gritting my teeth here in [Narnia] and I had an opportunity in 2012 to retire and to run for the [Narnia] legislature and I did just that (Democrat, Legislator, Narnia). That is when all the cuts started to be made and I could see the quality going down. I became the PTA [parent-teacher association] president for the entire school district and that was one of my jobs is to talk to lawmakers in [city] and really try to understand how this wasn’t a local issue, it was a state issue on how we fund our schools and the priority we put on funding our schools. I thought just going to [city] wasn’t enough so I wanted to be one of them. And I won. (Democrat, Legislator, Maycomb). I really ran for office because I wanted to work on improving educational experiences for kids and wanted to provide more equitable opportunities and I kind of viewed it as a social justice back. I grew up in a community and a household that really valued public education and I wanted to make sure that all kids have the same opportunities that I had. (Democrat, Legislator, Maycomb). I taught [many] years in the public school system. Testing came along and I wasn’t pleased with how the testing system was getting put together. I thought it was unfair and more of a “gotcha” system rather than a “help ya” system. So I got involved with the teachers union as a representative for my building and then worked with the teachers union to help shape policies to negotiate. […] Teachers were really struggling to get a good system of accountability. It was always a top-down process and never was I asked what I thought about what needed to happen to build a new education system for modern times. So there was a gentleman named [Jim Chrysler], who was a minority leader from 175 [city name] who was looking for someone to run who lived in his House district and I lived in his House district. So at the end of the day I agreed to run. (Democrat, Legislator, Narnia) Just one Republican legislator indicated that (s)he sought out a seat on the education committee. This individual felt compelled to be involved in education policy of his perceived policy expertise. One of the committees that I requested to be on was education […] I was made chairman of the education committee and that is because I am kind of…I do read a lot of policy information and want to change a lot of policy so that is pretty much where that is coming in to. I want to change the system, so to speak. I’ve written about it and criticized it and I’ve got my own ideas about how it can be fixed. (Republican, Legislator, Narnia) Although less common than legislators, some SBE members also sought out their position of power – all but one of which was elected. Similar to the previous quote from the Republican legislator, many SBE members felt their perceived expertise – which was not always in education – would be useful on the SBE, as can be seen in these three illustrative quotes from elected SBE members: An opportunity came to run for state board of education and as a statewide political leader focused on education policy it sort of certainly made sense since it is so important for me from a political…and from what I think are important policy enablers of economic growth and economic opportunity. (Democrat, SBE Member, Elected, Macomb) I’ve always been very involved in the community working with women and children and different organizations so I decided to run. I thought it’d be a good fit. […] I am just very familiar with that whole process and I knew that within the district, even though it was a very large district, my name ID [identification] would be fairly strong. So that is kind of what led me on the path and brought me to where I am. (Democrat, SBE Member, Elected, Narnia) I started thinking, well I could never get elected to anything locally so running statewide might be the possibility. So I gave it a try. […] Folks appreciated what I had to say and my experience as a teacher. (Republican, SBE Member, Elected, Maycomb) These last two quotes provide some insight into the politics that were frequently discussed among elected SBE members who sought out their position of power. As the third SBE 176 member explained, running for a statewide office provided an opportunity to gain education policymaking power that perhaps couldn’t have been obtained locally. The second SBE member quote also discussed the importance of name recognition. Other elected SBE members elaborated on the politics of pursuing SBE power. In particular, the politics of running for an elected SBE position often involved substantial sums of money, as well as political party and interest group persuasion. As one SBE member explained, You have to ask for money and you have to raise $50,000…at least, if not more. Out of that money that you get, you have to give some to the party. You have to help other people that are running. (Democrat, SBE member, Elected, Maycomb) This SBE member – elected in a partisan race – also described the important mediating role political party leaders play in the pursuit of education policymaking power: The principals of the party get together and they decide who is going to win […] The first time I ran, [the principals of the party] didn’t even let me go through the process. I said, “I want to run” but they don’t let you go through the process if you aren’t going to be the nominee. So I said, “Well, just let me do it so I just know what it entails. Maybe I’m not even interested.” But they don’t let you. They don’t let you. They aren’t going to waste their time. (Democrat, SBE member, Elected, Maycomb) Finally, as research suggests is true of down-ballot races (Weber, 1980; Hogan, 2005), SBE members elected in a partisan race discussed the importance of riding the coattails of either a governor or president: I got the nomination and then I immediately went down in the [President] landslide. And then two years later, a couple of county chairs encouraged me to give it a go again and so I did and I rode the [Governor] landslide. (Republican, SBE member, Elected, Maycomb) In 2012, President Obama was at the top of the ticket and when you have a powerful president at the top of the ticket, most of the time all of the ticket wins up and down. So that year I ran and I won on his coattails. (Democrat, SBE member, Elected, Maycomb) 177 Recruits. Most legislators sought out general policymaking power; how they came to serve on the education committee was typically through recruitment by political party leaders. As the following quotes illuminate, education committee members’ social networks with political party leaders played an important role in their selection to the education committee. Additionally, legislators were recruited to serve on the education committee for a wide variety of reasons – from demeanor and passion to geographic representation. I still wasn’t exactly sure I wanted to [be on the education committee] just because of the climate, but our [political party] leader, who is a close friend of mine, asked me. He said “[Jamie], I really need you to be on the education committee. We have to find a different message for democrats. We can just be against everything.” So that was sort of my task this year. (Democrat, Legislator, Duckburg) I think it was my third year, the president of the Senate said, “We need you on the education committee because there is nobody from our part of the state.” He was from northwestern [part of Narnia] as well. So, we needed somebody to sort of represent the interest of our schools and so I said, “Well if I have to, I will.” (Republican, Legislator, Narnia) After I served for two years, the chairman was retiring and he came to me and said, “I would like to recommend you.” Of course I had only been in the legislature two years but I have been passionate about education so I was blown away by that and so he did recommend me and I was appointed and I’ve been doing it ever since. (Republican, Legislator, Hogsmeade) The dirty secret of education chair is that nobody really wants to do it. I told the Speaker I would do it because knowing the committee…there was [sic] three of us that had been on the committee the whole time we’ve been here and the first couple chairs that we had were very political, very partisan, very angry. And I came at it a different way. (Republican, Legislator, Maycomb) Similar to education committee members, political party leaders and social networks played a role in the recruitment of SBE members – particularly for elected SBE members. Special interest groups were much more prevalent in the SBE member recruitment process. Additionally, SBE member recruiters varied among those who were elected and appointed. While the teachers union played a central role in the recruitment of elected SBE members, the 178 governor and his staff and advisors were key to the recruitment of appointed members. And, particularly among governor-recruited appointees, individuals’ social networks played a substantial role in them gaining this state education policymaking power, as is evident in the following quotes: So Governor [Langley] gets elected and he calls me and says, “You want to be on the school board?” And I said, “I’d rather be on the [Hennepin] wildlife and fisheries if I had to answer, since you asked.” And he said, “Well, we’re talking about school board.” So, that is how I came about it. (Republican, SBE member, Appointed, Hennepin) I was an executive with [Company name], which is a major employer in [Hennepin], and had a responsibility for a number of things, but kind of public relations and human resources were part of that. And the governor, my first appointment was related to that function at [Company name]. (Democrat, SBE member, Appointed, Hennepin) The governor asked me if I wanted to be in the cabinet. I decided not to and then in late December I got a call from [the Governor’s policy advisor] saying, “The governor wants to know if you wouldn’t be in the cabinet would you be willing to serve on the state board of education?” I thought about it and I said, “Sure.” (Republican, SBE member, Appointed, Narnia) The governor, who was a friend of mine at that point, had his office contact me and wanted me to go on the State Board of Education. Frankly, after [many] years I was kind of like I really don’t want to keep going back to [city]. So I actually said, “No, I’m not going to do it.” So the governor appointed another person […] but the Senate wouldn’t confirm him so his term just evaporated […] So at that point, I got called back from the governor’s office who said, “Since they wouldn’t confirm him, I need somebody that they will confirm. Since you’ve been in the legislature for [many] years, they would have a hell of time not confirming you… and you have a reputation.” So I ended up agreeing and I got appointed to the board and in fact they did confirm me. (Democrat, SBE member, Appointed, Hogsmeade) As is evident in these quotes, as well as many other views previously expressed, not all who obtained state education policymaking power were interested in education or policymaking – or at least most did not act on those desires without the nudging of someone in their social network. As the next section shows, this theme was common among both SBE members and legislators alike and as a result, very few state education policymakers came to their position of power with specific education goals they wanted to accomplish. 179 Goals of state education policymakers. Although some state education policymakers sought out an opportunity to impact education policy and others were recruited, most had relatively broad educational goals, some had no goals, and just a few had education-specific goals upon gaining education policymaking power. No education goals. When asked what goals they had upon obtaining a state policymaking power, legislators serving on education committees rarely expressed educationspecific goals. This finding may not be all that surprising given the wide array of issues that state legislators must consider as part of a general-purpose government. As one policymaker explained, “In the legislature, we have to be generalists.” (Democrat, Legislator, Elsewhere). As such, the majority of education committee members listed non-education goals across a broad spectrum of policy areas, including: job growth, deregulation, insurance reform, communitypolice relations, prison reform, water quality, medical delivery system, senior transportation, unemployment compensation. Other legislators expressed even broader, more ideological goals. For example, one education committee member confidently stated, “One is to do what is right; and second is to establish truth in [Duckworth] government” (Republican, Legislator, Duckworth). Although some SBE members indicated that they did not have specific goals upon gaining a seat on the SBE, unlike legislators, they were not focused on other policy areas. Instead, these SBE members were either focused on politics or figuring out the role of an SBE member. For example, one SBE member stated, “What I wanted to get done is take every opportunity to avoid being political” (Republican, SBE member, Appointed, Duckworth). Similarly, another SBE member focused on policy strategy. He explained, 180 [The previous chairman’s] vision of the board was that it was sort of a sounding board for anybody and everybody who wanted to come and talk about anything. And so if, for example, you and your husband had organized a bake sale at a county school or something and it had been very successful, you might decide to call the chair of the board and say, “We’d like to let you know how we ran a successful bake sale”— and [the chairman] would be inclined to let them come on the agenda to talk about the bake sale. So we would have two-day meetings, and they would be filled with interesting but really almost non-policy kind of issues. […] I said that my vision for the state board of education is to deal with, instead of 10 miles wide and an inch deep, I think there are three or four major policy issues that we are supposed to…my thought is we dig deep into those things and make sure the state board’s voice is being heard on those kinds of serious policy issues and not anybody who just wants to come talk about anything. I think we should celebrate the bake sales…we should celebrate the bake sale probably at the local level […] I was going to be the next guy in line to be the chair, I made it crystal clear that when I was a chair, there were going to be fewer and fewer of these feel good meetings and that I wanted members of the board and the staff to work tightly with each other on really significant policy areas. (Democrat, SBE member, Appointed, Hogsmeade) Yet others SBE members with non-education specific goals focused simply on figuring out their role both as a member of the SBE and the role of the SBE in general. As shown in the quotes below, all of these individuals were Democrats: I really just felt like I needed to orient myself and get more information about how the board operated. (Democrat, SBE Member, Appointed, Hennepin) You know, just understanding all of this took a lot of time and it wasn’t just an orientation. It took a lot of just board meetings and experience both positive and negative to really understand what is our role. (Democrat, SBE member, Appointed, Elsewhere) I spent the first year just learning. Like I kept my mouth shut and I just learned as much as I could. And now I have kind of honed in on certain platforms but at the time, I just wanted to support public education. (Democrat, SBE member, Elected, Maycomb) Broad education goals. When asked about their goals upon obtaining office, the relatively small number of education committee members who mentioned education – both Democrats and Republicans – made relatively broad statements like, “I wanted to focus on what 181 I call research- or evidence-based education policy—things that are known to produce results” (Democrat, Legislator, Duckworth); “I wanted to look at how we fund our schools” (Republican, Legislator, Narnia); “I wanted to expand these opportunities for these kids. I wanted to work on the equity thing” (Democrat, Legislator, Maycomb); and “My goal was to push the teachers union along to be a little more reform-minded or a little more innovative in their support for reforms” (Democrat, Legislator, Duckworth). Somewhat surprisingly, most SBE members had near equivalently broad education goals, which included: “Trying to really connect state policy regulations and direction toward real educators” (Democrat, SBE member, Elected, Narnia); “To be a teacher’s voice” (Democrat, SBE member, Elected, Maycomb); “Student achievement. Student achievement. Student achievement (Republican, SBE member, Appointed, Elsewhere); “Providing all the education for all kids despite their background” (Democrat, SBE Member, Elected, Narnia); and Help drive the policy agenda that is most effective in increasing improving educational outcomes, closing achievement gaps, and raising the bar for increasing levels of educational attainment. (Democrat, SBE member, Elected, Maycomb) Specific education goals. Just four state education policymakers – all located in one state, Narnia – stated education-specific goals. Two SBE members – both Republicans – focused on creating a more comprehensive and user-friendly school report card accountability system. Two legislators in Narnia – one Democrat and one Republican –also listed charter accountability as a goal of theirs upon taking office. 182 Most important votes. Although state education policymakers may not have come to their position of power with specific education goals, it may be that, upon gaining power, they found education policy to be an important part of their policymaking work. The informants of this research study suggest that this was not the case. As shown in Table 26, when asked about their most important vote since serving in the legislature, just three of the 19 education committee members identified an education policy. Upon further prompting to identify their most important education vote, all education committee members listed an education vote. While SBE members were much more likely to initially nominate an important education-related vote, a number of SBE members contended that they had not taken an important vote. This pattern was particularly salient among appointed SBE members, as shown in the following statements: I will tell you this, I don’t think I have made a vote that I believe is going to change outcomes in students. […] A lot of it is just mundane, nonsense that you got to go through because state law says you got to approve this stuff…real boring…damn…a lot of mundane stuff. (Republican, SBE member, Appointed, Hennepin) Other than just sort of run of the mill voting, there was nothing really important, honestly. (Democrat, SBE member, Appointed, Hennepin) I don’t think there was a whole lot…I mean honestly there wasn’t all that much that was memorable. (Democrat, SBE member, Appointed, Elsewhere) 183 Table 26 State education policymakers most important vote, by office State Board of Education Members Elected • • • • • • Standards (2) Required employment of school nurses, PE teachers, social workers (2) Charter accountability Social justice, civil rights of students State takeovers/school turnaround Chief State School Officer/State Superintendent selection Appointed • • • • • • Curriculum standards (6) None Teacher quality School report cards Mandatory attendance Early literacy Legislators Social Issues • Ban on gay marriage (2) • Medicaid expansion • Child abuse of deaf children • Performance-based pay for healthcare professionals • Religious Freedom Restoration Act (RFRA) • Medical marijuana • Human trafficking Labor Issues • Right to Work • Union wage setting for government works Environmental Issues • Water quality Education Issues • Teacher union collective bargaining • School finance • Choice and charter schools In many ways, these qualitative findings build upon my quantitative findings that showed that state education policymakers rarely rely upon their personal beliefs when making education policy decisions. These qualitative interviewees may provide some insight into why this may be the case. The majority of my interviewees did not seek out education policymaking power, did not have specific education policy goals, and did not find education-related issues to be among their most important votes. As such, it may be that state education policymakers do not rely solely on personal beliefs during the education policymaking process. Moreover, these qualitative findings, which suggest that political party leaders play an important role in recruiting state legislators to serve on the education committee, bring into question the results of Chapter 184 V’s rating, DCE, and most/least likely survey questions that suggest that political party leaders were among the least valued voices in the education policymaking process. This qualitative interview data also provides a broader understanding of the descriptive representation of state education policymakers. While many legislators sought out more general power, Democrats were more likely to seek education policymaking power and more likely to have broad education goals. In contrast, three out of four policymakers with education-specific goals were Republicans. Nonetheless, the majority of legislators serving on education committees did not seek out education policymaking power; and, even after serving on the education committee, did not find votes related to education issues to be among their most important votes. As members of general-purpose government, education committee members serve on numerous other committees and must vote on a myriad of policy issues. As such, this more generalist approach is likely expected. When compared to education committee members, the ways in which SBE members came to obtain education policymaking power are much more diverse. In most cases, those who sought out education policymaking power were elected SBE members. While often selfmotivated to run for office, political party leaders played a critical role in their ascent to power. Again, this finding is of particular interest given the results of Chapter V’s rating, DCE, and most/least likely survey questions suggested that political party leaders were among the least valued voices in the education policymaking process. In contrast, SBE members who were recruited to their position of power were typically appointed; however, some recruits were also elected SBE members. Key differences emerged among elected and appointed recruits: while appointed recruits often gained power through their connections to the governor’s office, elected 185 recruits typically gained power through their connections to special interest groups – in particular, teachers unions. When it comes to most important votes, elected SBE members often listed specific education-related votes that were of particular importance to them. This is perhaps not surprising given that elected SBE members often were motivated to seek out a position of education policymaking power; or, if they were recruited to run for election, SBE members often put together a campaign platform. In contrast, appointed SBE members had greater diversity in their goals and priorities. Although many SBE members felt curriculum standards were an incredibly important part of their work, a number of appointed SBE member also felt that their votes were relatively unimportant. Taken holistically, these findings reveal the wide variation in how and why state education policymakers come to their positions of power. In what follows, I shift toward examining nuances of whose voices are heard in the state education policymaking process. I focus on whose internal voices are valued in the policymaking process, as well as how state education policymakers perceive of the weight of different internal voices involved in the state education policymaking process. Building upon my survey research, I focus on teacher evaluation policy. I begin with a discussion of state education policymakers’ perceptions of whose internal voices are heard in the teacher evaluation policymaking process. I then examine state education policymakers’ beliefs about the ideal balance (or imbalance) of internal voices in the teacher evaluation policymaking process. Finally, I begin to expand the scope of policy that has permeated this dissertation, moving beyond teacher evaluation policy and toward a more general understanding of state education policymaking. I discuss state education policymakers’ 186 beliefs about the ideal balance (or imbalance) of voices that should be heard in the state education policymaking process. Results: Real and Ideal Internal Voices of Teacher Evaluation Policymaking Across all six states, all but two state education policymakers identified the state legislature as the primary driver of recent teacher evaluation policy reforms. In contrast, SBEs were seen simply as a rubber stamp, or non-actors. Oh in my view, [the SBE role was] very little. Most of it drove out of the legislature and the governor’s office. (Republican, SBE Member, Appointed, Narnia) I don’t know that [the SBE was] greatly involved in the actual policy. There was a lot of discussion about it. I think my view would be that it tended to come more from the governor and the legislature. (Republican, SBE Member, Appointed, Elsewhere) To say that [the SBE] was involved in designing it might be a stretch. (Republican, Legislator, Hennepin) It is more like [the SBE] take a set of recommendations by the [SEA] staff that are presented to them and they vote on it and they think they’ve just made a policy. But I have never seen them really come up with a policy on their own, which isn’t a bad thing because I don’t really trust them to come up with a good policy. (Republican, Legislator, Narnia) Well [the SBE] supported [the legislature’s] policy but I mean the state board of education in Maycomb, I hate to say this because I know many of them really well and I like them…they are basically an advisory body. Their powers…they have no statutory powers in terms of actually being able to force something to happen. (Democrat, Legislator, Maycomb) Whether policymakers felt that the legislature’s heavy-handed role in making teacher evaluation was appropriate sometimes varied; however, variations were not consistent across partisan lines or office. Some policymakers – both Republicans and Democrats, SBE members and legislators – focused on democratic accountability as the rationale for legislative involvement in making teacher evaluation policy. 187 Oh absolutely [the legislature should be involved in making teacher evaluation policy]! Because, you know, accountability to have high quality teachers working with children. Accountability to taxpayers who are paying the dime…paying for the system. I think that is all legitimate stuff. (Democrat, SBE Member, Elected, Narnia) I believe fundamentally that the elected will of the people, through the legislature, is the most direct means of developing education policy in the state. They can be directly accountable more so than the indirect accountability of a [state] board [of education] made up of people who are appointed by the highest elected official of the state and a group of people who are politically unknown. (Republican, SBE Member, Appointed, Narnia). That is what we are elected to do and that is why we have an education committee and we deal with education issues. (Republican, Legislator, Duckburg). I think the legislature plays a part – we can set the general parameters. Because it is a statewide…and we are giving billions of dollars supporting education and the public can and does demand an explanation…we aren’t just throwing those away. (Democrat, Legislator, Narnia) Other state education policymakers felt that legislative involvement was necessary to have the force of law. The state legislature does have a role to play because a lot of these things require the force of law so, you know, that is something that is the role they play. (Democrat, SBE Member, Appointed, Hennepin). Ultimately, you need the legislature to codify it in order to actually have teeth. So the legislature would have needed, regardless of the state role of the board if it had been greater or the state department of education if it had been greater, the legislature would have had to codify the policies anyways in order for them to actually mean something. (Democrat, Legislator, Maycomb). Differences in perspective emerged among those state education policymakers who felt the legislature was too involved in making teacher evaluation policy. However, again, there were no distinct differences among political parties or office. For example, a diverse group of state education policymakers expressed beliefs that legislators were not experts in education and, therefore, should have limited involvement in making teacher evaluation policy. 188 We have non-education professionals telling educators how to run schools. That was one of my other goals was to get the legislature out of education that is not their baby. We have a state board of education – that is their job. Why are we implementing all these policies into statue when best practices change all the time – to codify that and tie our hands is just bad policy and teacher evaluations is one of those. (Democrat, Legislator, Maycomb). I think [the legislators] are too much involved in all things related to education that they know very little about. It is whoever gets their ear first and I find that quite distressing (Democrat, SBE Member, Appointed, Hennepin). I think the legislature should simply have sort of the broader outline of the policy – we want an evaluation…maybe some of the, we want to evaluate their teaching effectiveness, the knowledge of the subject matter, and whether we want it locally or we want it done by some more objective, external way. But what that evaluation is and how it works and all those sort of things, it would be good if the state board of education and the state superintendent and the department of education would put that together because they are the ones who understand it (Republican, Legislator, Narnia). As this last quote shows, some state education policymakers felt that the SBE should have a larger role in making teacher evaluation policy. As one legislator said, “The state board, I think, should have the predominant role in making the policy in that because they should be the education experts and they have all the experts working for them” (Republican, Legislator, Narnia). However, one appointed SBE member had an alternative view – seeing SBE members as non-experts in teacher evaluation policy. I don’t think that state board members necessarily have the expertise to do that. I don’t think they’ve been trained in that area. We are just sort of…you see a lot of folks that have come from local school boards…they are business people in the state and this is sort of an appointment by somebody who knows the political insider. It doesn’t mean that their hearts aren’t into education. It doesn’t mean that they don’t think that education is really important. It doesn’t mean that they aren’t smart. They are. They are very bright people. Sometimes they just haven’t been involved in education so much so…when you bring an issue like that to us and throw it on the table and say, “Is this a good evaluation?” I’m not sure. We would tend to say to our experts in the Department, what do you think? Where are the limitations? What should we change? And then give our advice back to the state legislature. (Republican, SBE Member, Appointed, Elsewhere). 189 Finally, some state education policymakers – again, both Democrats and Republicans, both legislators and SBE members – couched their preference for less legislative involvements less in legislators’ lack of education expertise and more on shifting control back to local experts. [The legislature] can set the guardrails but within those parameters, I think there is a lot more that needs to be left to the local districts to decide how their teachers are doing based on local variables such as socioeconomic standing. (Democrat, Legislator, Narnia) I don’t think it is always to put the solution down in law all the time but it is our job to put our framework down on how to get to the solution. So maybe saying that we need a teacher evaluation system. […] It could be appropriate to say that we will implement a teacher evaluation system and it will be used for these purposes. I don’t think the implementation and detail should be written on the wall – we set the overall course and then the map and directions on how to get there is left to the districts (Democrat, Legislator, Maycomb). I would be fine if school systems came up with a system that actually evaluated their teachers. The state legislature wouldn’t need to be involved (Republican, SBE Member, Appointed, Narnia). What we do when we make law is that we don’t make evaluation policy, we just tell them that they need to do it at the local level and it should be a common framework (Republican, Legislator, Hogsmeade) Results: Policymakers’ Perceptions of Ideal Internal Voices in State Education Policymaking Moving beyond teacher evaluation policy, there were distinct differences among legislators and SBE members and along political party lines with regard to the ideal weight of voices from individuals within formal state education policymaking bodies. First, similar to teacher evaluation policy, both Republicans and Democrats saw the legislature’s involvement in education policymaking as an important component of democratic accountability – a way to ensure the voice of the people and interests of the state are pursued. So because the legislature represents the people, the public, the taxpayers, the parents, as well as the educators, there is definitely a role for that voice to be heard….the voices of all those people to be heard in the legislature. (Republican, Legislator, Narnia) 190 The state legislature has a role in determining the general parameters of what is valued in that particular state. (Democrat, Legislator, Narnia). [Legislators] represent the people. They absolutely should [be involved] and I would think that good policy should come…I mean it comes from hard work by state legislators who have done their homework with teachers and legislators and parents and school boards and all of those folks that have an interest in education. Yeah, absolutely. That is where good things can happen for people who work hard. (Republican, SBE Member, Appointed, Elsewhere). Education is a huge part of our budget – 40 to 50 percent – so the legislature has to play a significant role in that but I think it should be based upon a carefully thought out plan, a vision of where do we want to go with education in this state and then we put that into law (Republican, Legislator, Narnia) Where Democrats and Republicans differed is that Republicans – both SBE members and legislators – were more likely to prefer limited legislative involvement in the education policymaking process. This preference is perhaps related to ideological roots—conservative beliefs in limited government involvement. As the following quotes exemplify, Republicans preferred that the legislature set broad frameworks, appropriate funding, and whenever possible, decentralize policymaking. I would say the legislature has a responsibility with regard to standards and with evaluation and assessment…certainly, in my opinion, they should have no role or very little role in terms of curriculum and the actual classroom instruction. They just should not be involved in that. That should be local. (Republican, SBE Member, appointed, Elsewhere) You need to give your local districts support. You need to give them guidelines. You need to make sure that they play within the bounds. You know there are some things that, statewide, that need to be…we need to be on the same page. But at the same time, I want to give the local LEAs authority to move quickly. The smaller you are, the quicker you move. (Male, White, Republican, Legislator, Hennepin) [The legislature is] making the plans behind [the framework]. If you want to use the analogy of the skyscraper, [the legislature is] the engineer that tells you what to build, the board of education puts the framework together, and then the local districts finish it off. (Republican, Legislator, Elsewhere) It falls on us to make sure we have the policy that is going to allow us to be the most successful as a state in educating our kids in our public schools. So that has to be our 191 role. But, many people like creating laws; but if you don’t need a law, why create one? (Republican, Legislator, Hogsmeade) I would make it as limited as possible. I would make a system that would be Milton Friedman based. (Republican, Legislator, Narnia) In an ideal world, I would say the state legislature’s role would be to appropriately fund public schools. But as far as policy…eh. (Republican, SBE member, Appointed, Narnia) Similar to findings specific to teacher evaluation policy, some Democrats were supportive of a more limited role for the state legislature; however, this preference was associated with a perception of legislators as non-experts in education. State education policymakers with this preference were primarily SBE members. Leave it to the experts […] Now because you are a lawmaker you always have input. You always have the right to suggest. You always have the right to pass legislation. But even before you recommend legislation or recommend a course of action, sound the experts – which is the staff of your state board of education or your staff of your department of education. (Democrat, SBE Member, Appointed, Hennepin) If we have a [state] board of education, I would think the legislators would be requesting information constantly. Why would legislators assume they know what is best for schools and teachers and students on their own? (Democrat, SBE Member, Appointed, Hogsmeade) As these last two quotes exemplify, some state education policymakers – primarily Democrats, but also some Republicans – expressed preferences for a much stronger voice and more proactive role for the SBE in the state education policymaking process. But I would say that a state board of education should have more policymaking authority and the ability to really set the direction of public education but do so in a way that creates stability (Democrat, SBE Member, Elected, Maycomb) 192 I think they need to be the education experts in the state and they need to have authority to make policy and they should be the ones determining the leadership of the department of education, which should be outside the political ideological realm. (Democrat, SBE Member, Elected, Maycomb) I would say that the state board of education should take a look at the research and the evidence from other states as to what is the most effective as well as the most efficient education system and ask the legislature to adopt in…not in minute detail…but in sufficient detail that the state board could put that system into operation. (Republican, Legislator, Narnia). While still emphasizing the importance of the SBE, other state education policymakers – primarily Republicans, but also a limited number of Democrats – felt that the SBE played an important role as a cheerleader, advisor to the legislature, and a watchdog, “updating and creating policy based on things that are in the best interest of students” (Democrat, SBE Member, Elected, Narnia). Others elaborated, I would like to have a bigger policy role [making education policy] than we have now […] I think that we have a responsibility to represent education in the state of [Elsewhere] and show folks that this is what we are doing and we want to help you highlight programs that you are doing in your schools that are very good and very effective- how can we show that? We can visit it. We can talk about it. We can bring the press with us. We can show folks you are doing a great job in your community. (Republican, SBE Member, Appointed, Elsewhere) I think in the broadest sense, their role is to make sure that every child receives the highest quality education possible. And it is kind of hard then for me to identify that they need to do this, they need to do this, you know. […] I think we see ourselves as kind of a sounding board…as promoting…maybe a cheerleader for things we think are important kind of thing. (Republican, SBE Member, Appointed, Elsewhere) So our role has really become one of a watchdog if you will and then pulling out things that, you know, this is a mistake, you really should change this, we agree with that. (Democrat, SBE Member, Elected, Maycomb) I think you have to have somebody that is in oversight that implements the state policy and makes recommendations back to the general assembly. They should be the experts. (Republican, Legislator, Hennepin) Still other state education policymakers advocated for the SBE to have little to no involvement in making education policy. Again, consistent with conservative beliefs in limited 193 government, Republicans tended to support some SBE involvement in developing frameworks or models while leaving the majority of the decisionmaking to locals. Well there may not be a role…there may be a role where they are just you know coming up with model curriculum or coming up with model standards that a local school can adopt if they wish or not (Republican, Legislator, Narnia) The role of the state board is to manage the basic structure…you know you build a skyscraper and you have all these pylons that make up the framework and then everything else covers that up. To me, the state board should act as the pylons. They build the frame and then they let the local districts fill it in the best way to meet the needs of the local needs (Republican, Legislator, Elsewhere) Finally, a number of state education policymakers – both SBE members and legislators, and primarily Republicans – were dismissive of the SBE having any role in the state education policymaking process. Get rid of it. I’m sorry. I just haven’t found much use for the state board […] I think that the board is usually only as good as the executive you have that is appointing them. And I have seen much more thought and I guess true interviewing, scrutiny go into who is being appointed to the board because if you are just going to appoint people because they are your friend and you own them a political favor and it is prestigious to be on the board, then they’ll show up and they’ll eat your cookies and drink your punch and take their little stipend for mileage and stuff like that and do the “rah rah” thing and “oh yeah that sounds good.” But to get somebody who really understands, that is really willing to roll up their sleeves and work, that’s…we don’t have but maybe one person that I know of on that board right now that does that. And so for me, if you are going to be on board, you should be willing to do work. (Republican, Legislator, Hogsmeade) I think it is a horrible system. I used to think it was innocuous but now I think it is a bad system. I would be in favor of wiping out the state board of education (Republican, SBE Member, Appointed, Narnia). Feelings of discontent with the SBE were particularly salient in Narnia – a state with a SBE that has both elected and appointed members. One legislator’s description of the state’s education system summed up many other legislator and SBE member views: Our system is a dysfunctional process. The electoral process— I mean we have elected some real idiots. We had one [individual] who was elected [more than once] who was 194 basically a hermit that…you know, didn’t have any connection to schools […] This [individual] would come to state board meetings and first thing [the individual] would do would be to walk over to the donut tray and stuff [her/his] pockets with donuts—to the point that they quit serving them. But, [the individual] got elected, not once but twice because [the individual] had a name that sounded somewhat familiar in [Narnia]…and absolutely no contribution whatsoever to the board of education. [The individual] never opened [her/his] mouth because if [he/she] did, I don’t know what [he/she] would have said. I wasn’t there. I seen it from the outside. So I mean that is the worst case. But I know just looking at the elected members that how they got elected…A couple of them that are pretty good are, frankly, the ones who were backed by the teachers unions. They put the money in it and made sure people knew who they were and elected them. Those are fairly valuable members of the board because they do know what they are talking about. But then if you look at the appointed members, I am not sure what criteria are used for appointees but very rarely are they coming to us with the kind of experience that they have to have and they look to the governor to tell them how to vote and…you know I’ve just heard far too many times that the governor placed his phone calls and told them what to do. (Republican, Legislator, Narnia) Yet another legislator described Narnia’s system “in theory” and “in reality.” As many others in the state of Narnia concluded, the mixed system results in a lot of gridlock and frustration over how to make the system “work.” Well I think there is the theory and then maybe there is current practice. [In Narnia,] it was decided that the governor can appoint several members of the board so the governor’s voice would be heard in the state board of education; but the majority would be elected and their terms would overlap so that a new governor could not just appoint new members of the board and thereby take over the education system. That was the theory. In practice it has been that if a governor is determined, the governor actually can—with pressure and other means—get [his/her] people appointed and get a few favorably elected members. Then, the governor’s point prevails. As a result, you get these deep splits within the state board as to what education policy should be and it becomes more partisan than it should. In theory, it should work really well. In practice, at least with the personalities here in Narnia, it has become somewhat dysfunctional. So the question goes, well alright if the hybrid board isn’t working, just wait it out and see if personalities change? Or, do you go to an all elected board? Or, do you go to an all appointed board? Or, is there something else that could make a hybrid board work? (Republican, Legislator, Narnia) Despite feelings of frustration with the current system, at least one SBE member and one legislator – both Republicans – in Narnia had different views of what the SBE structure should look like: one desired a board of individuals who “do education policy,” the other, education practitioners. 195 I would have a policymaking board that actually knew how to do policy that was directed by the superintendent or the Governor or the [legislature] that says, “Here is the idea. Here is what we’d like to do. Now how do you go about doing it?” Now will it have some consequences? Yes. But it would have to be written around…this is not a political position. This is not payback for something that you helped me do. This is that you have to be qualified to actually do education policy. Those people exist. They may not be the most politically adept…but they actually know how to do policy and so those would be the people that I would have. (Republican, SBE Member, Appointed, Narnia) I think if you are going to have a state board, it should be composed of people who are selected for their knowledge, not because they want to run for the state school board. So I would put together a board that consists of, you know, four teachers, four principals, two superintendents, a couple of higher ed folks, a couple of early childhood ed folks, and put the policy recommendation in their hands. […] I like people making policy on education that aren’t going to change when the governor changes or when someone’s term expires. I think that brings much more stability and it brings reality to what we are doing in education policy. But we are far cry from having that. (Republican, Legislator, Narnia). Discussion and Implications for Future Research This research begins to address a major gap in the literature by providing empirical evidence of how and why state education policymakers come to their positions of power; and, upon gaining power, the internal voices that are valued in the state education policymaking process. While the study is limited to insights from voluntary informants in six states – each with a slightly different education governance structures – the research provides timely empirical and conceptual understandings of the state education policymaking process. Current trends suggest that those with state education policymaking power are becoming increasingly important to the policymaking process: the American public is increasingly calling upon state government to be involved in education matters (PDK & Gallup, 2015); the recent reauthorization of the Every Student Succeeds Act (ESSA, 2016) shifts much decisionmaking power back to the states (Klein, 2016); and, recent change in federal executive power have resulted in a shift in the education policy approach of the executive branch in which it is believed that “state and local leaders are best equipped to address the unique challenges and opportunities they face, not the federal 196 government” (DeVos, 2017). As such, understanding who state education policymakers are and their motivations, goals, and approaches to the education policymaking process are valuable empirical insights. State education governance structures. The findings from this chapter suggest that institutional structures may not have substantial influence on the internal qualities of those who gain education policymaking power. Very few of the education policymakers interviewed as part of this study aspired to positions of education policymaking power and, even upon gaining power, very few found education-related votes to be their most important. One exception is that elected SBE members tended to have more education-specific goals and saw education votes as particularly important to their work as an education policymaker. This finding is likely not surprising given that legislators are part of a general-purpose government, while SBE members are part of a special-purpose government focused solely on education issues. Yet, this finding may also speak to the types of people who are willing to endure the process of running for elected office in a special-purpose government: perhaps only those that are passionate about the issue for which they have oversight – education – are willing to endure the political processes of campaigning for a position on an elected SBE. Or, this finding may connect to research that shows that politicians seek particular positions in order to build a base of support to bid for higher office at some point in the future. Also known as ambition theory (Schlesinger, 1966), it may be that in an attempt to move to higher office, elected SBE members are more keen to focus on a specific policy record or message and exude specific conduct while in office. This research study only examined a limited number of elected SBE members in two states; however, to better understand the motivations of SBE members, 197 future research could examine a greater number of SBE members across the 12 states with elected SBE members. Specific to states with a mix of elected and appointed SBE members, this research suggests that the mixed state education governance structure may result in a contentious and even antagonistic education policymaking arena. Legislators and SBE members – both appointed and elected – in Narnia discussed dissatisfaction with the current structure: many appointed SBE members felt that the elected members only represented parochial interests or those of the teachers union; many elected SBE members felt appointed members were pawns of the governor’s office—“the appointed people don’t talk to anybody but the governor,” one elected SBE member in Narnia stated; and, some legislators expressed reservations about the qualifications of both elected and appointed SBE members. While this mixed structure may seemingly blend the best of both elected and appointed boards, in reality, this did not seem to be the case. Elected SBE members felt the need to be responsive to public preferences, as stated by one elected SBE member: I feel like many of us elected people feel like we have a different mandate, that we have to answer to a different audience and that is the people that elected us – the citizens, the parents, the teachers, and the taxpayers. (SBE member, Democrat, Elected, Narnia) In contrast, appointed SBE, although responsive to the vision of the governor, are “removed from the partisan ideological whims of the day” (SBE member, Democrat, Elected, Maycomb). Yet, my research suggests that this combination of elected and appointed SBE members may bring out the politics of both worlds, which don’t necessarily play well together. Across the nation, there are four states with mixed elected/appointed SBE structures. Future research could examine whether this politically contentious environment is generalizable to all those with mixed structures or, to use the words of the previously quoted Narnia legislator, “is there something 198 else that could make a hybrid board work?” Are there circumstances under which individuals that are responsive to the electorate and to the governor can work together? Weights of internal voices and what that means for external voices. With regard to teacher evaluation policy, I find widespread agreement across SBE members and education committee members in all six states that the teacher evaluation policymaking process was driven by the state legislature. This finding has implications for my findings in Chapter V: if state legislators’ voices carried more weight in the teacher evaluation policymaking process, then the external voices that state legislators valued likely had more influence on the making of teacher evaluation policy. Although quantitative analysis of my experimental survey question revealed that there were no significant differences in education committee members’ and SBE members’ responsiveness to public and teacher preferences, their responses to the rating, DCE, and most/least likely questions reveal other the difference in the valuation of voices among education committee members and SBE members. In Figure 20, I list the top three actors whose voices were valued by education committee members and SBE members based on the three nonexperimental survey questions analyzed in Chapter V. While Chapter V discussed many of the statistically significant differences in whose voices are valued by state education policymakers, there is still quite a bit of overlap among the voices considered to be valuable among SBE members and education committee members. Qualitative results suggest that the internal voices of legislators were the loudest when it came to the teacher evaluation policymaking process. This then suggests that the external voices that education committee members likely carry more weight than those of SBE members. Yet, nearly all of the voices that SBE members valued were also valued by education committee members, save for one group: education research 199 organizations. As such, when it comes teacher evaluation policymaking, this logic would suggest that education research organizations were on the outside looking in. While they may have the ear of SBE members, they may not be able to influence the policymaking process if the SBE’s voice is the education policymaking process is relatively quiet. If education research organizations aspire to impact the education policymaking process, this finding underscores the importance of understanding whose internal voices are heard in the education policymaking process. Figure 20 Weight of internal voices influences weight of external voices heard in the education policymaking process Education committee members SBE members + - Local school leaders (ML, DCE, R) Local teachers (ML, R) Constituents (DCE, R) Statewide school leadership groups (ML, R) Education research organizations (ML, DCE) Other state officials (ML) Personal beliefs (DCE) *Note: Parentheticals of valued external voices represent survey question with statistically significant finding (R=rating question; ML=most/least likely question; DCE=discrete choice experiment) 200 Chapter VII State Education Governance Structures and Policy Outcomes As previously demonstrated, wide variation exists in the governance arrangements undergirding each state’s education policymaking process. While some state legislators solicit the assistance of their state boards of education (SBEs) in the policymaking process, others limit the policymaking power of their SBE. Additional variation exists in the institutional arrangements of the individual policymaking bodies that make up a state’s larger education policy governance arrangements. Recognizing the complexity of state education governance systems is particularly important for two reasons. First, as demonstrated in Chapters IV, V and VI, the demographics, backgrounds, experiences, goals, and preferences of those who gain power in a state legislature often differ from those who gain power in an SBE. And, state legislators and SBE members value different voices when forming their education policy positions. Second, the institutional norms, rules and traditions related to who gains power and how policy decisions are made vary within state legislatures and SBEs. State legislatures are made up of a group of individuals, popularly elected by local constituents within a geographically constrained legislative district, and often concerned with multiple other policy issue areas, not solely education. In contrast, SBEs are often composed governor appointees or representatives elected on a statewide or regional basis, whose sole policy issue area is education. Moreover, the frequency and length of education policy deliberation vary amongst SBEs and state legislatures. SBEs, recently described as “inconspicuous stewards convening in sparsely attended, daylong meetings where they debate education policy” (Burnett, 2016), often meet just a few times a year 201 for a relatively short amount of time. In contrast, most state legislatures meet more frequently and for longer periods of time. Additionally, as my research demonstrates, although legislators and SBE members both value the voices of groups such as local school leaders, local school teachers, school leadership groups, there were differences in the extent to which these voices were sought out among different governance bodies: SBE members frequently relied on education research organizations, education advocacy organizations, and the governor to inform their policy decisions; in contrast, legislators often relied on local constituents. These differences within and between state legislatures and SBEs matter for whose voice is heard in the policymaking process: just as no political action takes place without having different effects on different groups’ values, who has and wields power over the education policymaking process is contentious because some voices will be heard and others will not; some participants will win, others will lose. For example, when education policy is developed in the state legislature, research suggests that, in general, legislators will advance their own notions of what is good for the people rather than seeking and responding to the interests and demands of the general public (Hummell, 1987; Miewald, 1978; Mosher, 1982). Moreover, term limited legislators are often even less inclined to follow the demands of voters and more inclined to follow their conscious in making policy decisions (Carey, Niemi, & Powell, 1998). When it comes to education policy, my research suggests that, while state legislators give high ratings to constituents as informant of their policy positions, legislators were not responsive to constituents. Legislators also did not rank their personal beliefs very highly when it comes to developing their teacher evaluation policy position. Rather, legislators tended to be responsive to individual voices of teachers and school leaders. Moreover, legislators of different political parties were responsive to different interest groups. 202 In contrast, research presents conflicting hypotheses for the SBE policymaking process. In some cases, scholars have argued that “special-purpose governments” such as an SBE are less likely to be captured by special interest groups and more likely to deliver policy decisions congruent with majority opinion (Bish, 1971; Ostrom, Bish, & Ostrom, 1988). Yet, other scholars contend that the low visibility of special-purpose governments hinders public participation and creates a bias toward special interests that have the capacity to lobby specialpurpose government officials (Bollens, 1957; Burns, 1994; Macedo & Karpowitz, 2006). My research suggests that, in the case of state-level special purpose education government, SBE members are likely to listen to the voices of interest groups like education research groups, education advocacy groups and school leadership groups, as well as the interests of the governor. Moreover, research suggests that different selection methods will influence whose voice comes into the policymaking process: SBEs made up of appointees, when compared to elected SBEs, may be more likely to pursue policies that diverge from public preferences without fear of loosing their positions (Alesina & Tabellini 2007; Beasley & Coate 2003; Burden, Canon, Lavertu, Mayer, & Moynihan 2013). While appointed SBEs are not entirely independent (they are accountable to elected officials, and thus indirectly accountable to the voters), the addition of an intermediary (i.e., the appointer) may weaken voters’ ability to hold appointed SBEs accountable and perhaps result in appointed SBEs being less responsive to public preferences. In contrast, elected SBEs might have policy goals that better match the preferences of the electorate than those of appointed SBEs because the members have been selected, ostensibly, because of their views on education. Additional variation may exist amongst statewide and regionally elected SBEs: the policy goals of popularly elected SBEs may align more closely with the preference of the political majority and/or those who are more likely to pay attention to the SBE 203 elections, which are often placed at the very bottom of the ballot. Regionally elected SBEs, on the other hand, may allow for minority policy preferences to be represented in education policy discussions. My quantitative analyses suggests that these SBE selection structures do not have a statistically significant impact on SBE members’ responsiveness to public or teacher preferences; nor, do they have an effect on whose voices SBE members turn to in forming their policy position. However, the qualitative analyses suggest that elected and appointed SBE members both come to the education policymaking process with different goals and expectations, and have different preferences for whose internal voices are heard in the state education policymaking process. Given that state-level education governance structures seem to influence who gains power, and whose internal and external voices are valued in the education policymaking process, I now seek to understand how who is making state education policy, and the governance structures in which they reside, impact policy decisions through the lens of one area of education policy: teacher evaluation. Background: Teacher Evaluation Policy Drawing upon research that indicates that teachers play a critical role in the educational lives of their students, increasing emphasis has been placed on the need to identify which teachers are most effective in the classroom (Darling-Hammond 2000; Jordan, Mendro, & Weerasinghe 1997; Marzano, Pickering, & Pollock 2001; Wright, Horn, & Sanders 1997). Prior to the 1990s, many states enacted teacher evaluation policies requiring districts to simply evaluate teachers. Decisions around the details of such evaluations, however, were typically reserved for local school boards. As some states began to implement more rigorous whole-school 204 accountability systems in the 1990s, and with the passage of the NCLB Act in 2001, state governments began to experience increasing pressure to improve teacher quality through more uniform, standardized teacher evaluation policies. Thus, many states adopted more stringent, rigorous teacher evaluation policies and procedures, many of which required the incorporation of student achievement on standardized tests. State-level efforts to develop and implement more uniform teacher evaluation systems have rapidly expanded, with a significant amount of fiscal and human capital being allocated to the development of complex value-added models, rigorous standards-based teacher observation protocols, and multifaceted comprehensive teacher evaluation systems that incorporate multiple measures of student achievement and teacher professional practice. Although some of these efforts were initiated at the state level, the federal Race to the Top (RttT) competition and the U.S. Department of Education’s Elementary and Secondary Education Act (ESEA) waivers induced much of the recent state action around teacher evaluation policy. The majority of research related to teacher evaluation post-RttT and ESEA waivers has focused on either the ways in which new teacher evaluation policies have impacted teaching as a profession (Milner 2013), the variation in teacher evaluation policies among districts and states (Steinberg & Donaldson 2016), or the validity and reliability of the quantitative data, often based on student standardized assessment scores (Baker et al., 2010; Bill & Melinda Gates Foundation, 2010; 2013; Buddin & Croft, 2014; Chetty, Friedman, & Rockoff, 2013; 2014; Hanushek & Rivkin, 2010; Rothstein, 2009). Moreover, researchers have examined whether teacher evaluation reform efforts have improved teacher quality in all schools, finding that while some states have struggled to fully support the new systems, others have seen some success and are en route to developing a means of supporting districts in teacher evaluation work (McGuinn, 2012). 205 Very little research, however, has examined teacher evaluation policy reform from a governance perspective. Although researchers have recently begun to examine how teacher evaluation policies vary from state to state (Steinberg & Donaldson, 2016), we have very little knowledge of who is making the policies and how this impacts the outcome of the policymaking process. Moreover, almost all state-level education policy research, in general, has characterized the state education policymaking as taking place within single, monolithic institution, examining either the enactment of state statutes by the state legislature (Young, Wang, & Lewis, 2016) or policy action taken by state education agencies (SEAs) and SBEs in response to federal policy change (Wrable, Saultz, Polikoff, McEachin, & Duque, 2016) (for exceptions, see Jimerson & Childs 2015; Loeb, Miller, & Strunk 2009). This trend applies to state teacher evaluation policy as well: while some research has examined variation in state teacher evaluation statutes (Collins & Amrein-Beardsley, 2014; Dixon, 2011; Steinberg & Donaldson, 2016), others have focused solely on teacher evaluation policy actions taken by SEAs and SBEs (Pennington, 2014) (for exceptions, see Baker, Oluwole, & Green, 2013). Yet, state education policymaking is complex, involving multiple state policymaking bodies with various norms and motivations. As such, education policy research must acknowledge the significant, interrelated roles of both state legislatures and SBEs in making education policy decisions and how this may influence the policymaking process and outcomes. Conceptual Framework I seek to understand how who is making education policy, and the governance structures in which they reside, impacts the state-level teacher evaluation policy outcomes. The policy outcome that I focus on is the extent to which state education policymakers – namely, state 206 legislators and SBE members – either retain policymaking authority over teacher evaluation policy elements or devolve policymaking authority to a subordinate policymaking body. To guide this work, I first draw on Krause’s (2010) state legislative delegation framework. As Krause (2010) explains, legislators may base decisions to delegate authority to executive branch agencies, of which an SBE can be considered, for a number of reasons. For example, by not delegating policymaking authority, legislators can “ensure against tyranny by a singular executive” and, given their re-election motivation, ensure the will of the people is heard (Krause 2010, p. 5). On the other hand, legislators may have an incentive to delegate responsibility to the SBE, particularly if legislators seek to (1) increase education policy expertise, (2) increase the amount of time and resources spent on other legislative issues since education oversight may be costly with little electoral return, (3) mitigate policy responsibility, allowing blame for unpopular policies or outcomes to be placed on the SBE (especially when the opposition party controls the SBE and/or executive branch), and (4) mitigate problems related to policy instability and lack of credible policy commitment (Krause 2010). Another devolution option that state legislators is to remove the SBE from the policymaking process and, instead, devolve policymaking authority directly to the local school districts. Research suggests that a state government may decentralize directly to a local government for a variety of reasons. For one, decentralization may promote policy innovation, allowing states to serve as “laboratories of democracy” where several different policies can be considered simultaneously (Rosen, 1999). Additionally, it is thought that local governments, being closer to the people, will be more responsive to the needs and preferences of those directly using and affected by the services (Oates, 1972). Oates’s (1972) decentralization theorem 207 establishes that, on the grounds of economic efficiency, “to maximize overall social welfare requires that local outputs vary accordingly” (p. 1122). Finally, SBEs also have policymaking authority devolution options, if they are indeed delegated policymaking authority. If a state legislature devolves policymaking authority to an SBE, they can decide to retain that authority or devolve it to local school districts. However, research and current theoretical frameworks such as those of Krause (2010) and Oates (1972) are limited in two ways. When applied to the focus of this research – legislative and SBE policymaking authority devolution – the frameworks reveal little about SBE action and what SBEs actually do with the policy discretion that they are granted by the legislature, and the resulting policy outcomes. Moreover, Krause’s (2010) legislative delegation theory, coupled with public administration scholarship, has focused primarily on “what bureaucrats do with their discretion, and much less on the conditions whereby their discretion is expanded or restricted” (Krause, 2010, p. 24). This research aims to address these two gaps by examining state-level education policymaking that takes into account both SBE and legislative education policymaking authority decisions, and the political and governance conditions under which these decisions are made. Data and Methods Developing a teacher evaluation policymaking authority devolution index. In the American federal system of governance, policy devolution can be understood as a transfer of specific powers or authority from a superior government to a subordinate government (Kincaid, 1998). It is through this lens of authority devolution that I examine differences in how teacher evaluation policymaking authority varies in states with different education governance 208 structures. To do so, I first created a teacher evaluation policymaking authority devolution index that measures the extent to which state legislatures and SBEs devolve specific powers or authority around teacher evaluation policy to local school districts. I also developed two subindices that measure the extent to which state legislatures and SBEs – as separate, formal education policymaking bodies – devolve teacher evaluation policymaking authority. Because state legislatures legally establish SBEs, I assert that state legislatures are the strongest state policymaker who can choose to devolve authority either to an SBE or local school districts. This relationship is not reciprocal. SBEs cannot choose to devolve policymaking authority to state legislatures. SBEs may obtain policymaking authority from the state legislature and, upon receipt, choose to either retain that policymaking authority or devolve it to local school districts. The data I use to developed the teacher evaluation policymaking authority devolution indices includes publically available state statute and administrative code related to teacher evaluation in all 50 states collected in the summer of 2014. Reports, publications and other information or training materials provided on SBE an SEA websites served as secondary sources and provided additional insight into state teacher evaluation policy. Teacher evaluation laws, rules and regulations in all 50 states were entered into a qualitative analysis tool (TAMS Analyzer) and coded and analyzed the texts following Anderson, Tremper, Thomas, & Wagenaar’s (2012) process for measuring law variation, as subsequently described. The process was iterative, with “one or more steps being repeated as discoveries at one stage expose inadequacies of constructs developed at a previous stage” (Anderson et al., 2012, p. 7). Establishing the legal framework. The first step I took to create a measure of teacher evaluation policymaking authority devolution was identifying a legal framework of interest. 209 Although ESEA waivers and RttT opportunities tasked SEAs with detailing state teacher evaluation systems, these applications are neither the sole nor a binding source of law or regulation for most state teacher evaluation policies. These applications have obvious importance and, in some cases, state statute and administrative code actually refers to or reiterates teacher evaluation practices described in an SEA’s RttT or ESEA waiver application. However, multiple other sources exert equal if not greater influence on teacher evaluation. Nearly every state has enacted statute detailing teacher evaluation system requirements that must be implemented in every school district. Moreover, many SBEs have enacted administrative rules and regulations that either echo or build upon state statute related to teacher evaluation. Each of these legal frameworks – statutes, administrative code, state ESEA waivers and RttT applications, and other regulatory documents – yield teacher evaluation policy levers. Thus, I utilized these documents in the development of my teacher evaluation policymaking authority devolution indices. Generating relevant legal measures. Next, I identified specific teacher evaluation policy elements for which authority may be devolved by state legislatures and/or SBEs to local school districts. The initial list of teacher evaluation system elements originated from a systematic literature review of research related to teacher evaluation across the United States published between 2010 and 2014. In conducting this literature review, I began with a targeted ProQuest search of peer-reviewed books, conference proceedings and papers, reports, and scholarly journals that included the term “teacher evaluation” in the abstract. This initial search yielded 755 results. Subsequently, I limited the search to include only sources that focused on United States K-12 education policy. For example, I eliminated sources that focused on higher education faculty evaluation or nurse preceptor evaluation. I also eliminated sources that focused on 210 teacher evaluation in countries outside of the United States. This modified search yielded 31 results. Next, I reviewed each of the 31 sources and eliminated 21 sources that were unrelated to United States K-12 teacher evaluation policy. For example, one eliminated source focused on whether music teachers’ self-evaluation instruction impacted music students’ self-evaluation, while another focused on the extent to which math teachers were aware of the National Council of Teachers of Mathematics teaching standards documents. Finally, I created a research memo for each of the 10 remaining sources, focusing on the elements of teacher evaluation policy that were examined or discussed. I also reviewed these 10 sources for additional sources that were not identified in the ProQuest search. Thirteen additional sources were identified in this process. This systematic literature revealed teacher evaluation policy elements shown in Table 1 in Appendix M. Conducting preliminary research. In addition to the a priori teacher evaluation policy elements that were identified in the literature review, the process of identifying teacher evaluation policy elements was iterative, allowing for additional elements to be added or clarified throughout a preliminary review of state statute, rules, regulations, and other policy resources. Teacher evaluation policy elements that emerged or were clarified from this iterative process included: variation in state and/or local assessment use, variation in weighting levels of data components, and posting of evaluation practices on district websites. Upon combining these lists of teacher evaluation policy elements, I grouped the elements into five categories, as shown in Table 1 in the Appendix M. 211 Formalizing coding procedures. While traditional legal research typically produces narrative descriptions of how laws differ in their text or meaning, the measurement of law utilized in this study examines how laws vary on specific policy elements. To operationalize this policy element variation, I identified relevant elements of teacher evaluation policy that varied among states. Subsequently, I developed both a textual and a numeric coding scheme to capture teacher evaluation policy variation. The development of the coding schemes were informed by Carnoy & Loeb’s (2002) approach to developing a “scale of accountability,” which captured degrees of state external pressure on school to improve student achievement. The numeric coding scheme that informed the teacher evaluation policymaking authority devolution index occurred at two levels. The first level of coding identified which teacher evaluation policy elements were put in place by the state legislature via statute, and which by the SBE via administrative rule. If the element was enumerated in both statute and administrative code, I conducted additional archival legal research to identify the order in which the statute or administrative code was enacted. I then coded the policy element in line with the policymaking body that first enacted the policy. The second level of coding identified the extent to which teacher evaluation policy elements were devolved. This coding occurred in two forms. For some variables, a dichotomous coding scheme of the presence or absence of an identified teacher evaluation policy element was appropriate. I used this level of coding for teacher evaluation policy elements that, upon review of all 50 state policies, were either present or absent. For example, for the sub-category of “observation frequency”, I coded a policy that mandated that classroom observations take place two times a year with a 1, and a policy that mandated classroom observations take place but did not mandate classroom observation frequency with a 0. 212 For other variables, a weighted measure of the level of authority devolved to school districts around an identified teacher evaluation policy elements was appropriate. This coding occurred with teacher evaluation policy elements that, upon review of all 50 state policies, incorporated varying degrees of policymaking authority. A higher weight indicated strict state authority over the teacher evaluation policy element and a lower weight indicated substantial local flexibility. For example, in the “weighting” category, I coded a policy that mandated that state assessment data must account for 40 percent of the final teacher evaluation score with a code of 1.5; a policy that mandated that state assessment data must be between a range of weights – for example, between 25 and 40 percent – with a code of 1; a policy that mandated that state assessment data must account for a “substantial” portion of the final teacher evaluation score with a code of 0.5; and a policy that did not mandate any data weighting with a code of 0. I summed the dichotomous and weighted coding for each state which, ultimately, informed the 0 to 17.5 teacher evaluation policymaking authority devolution index measure, as displayed in Table 2 in Appendix M. A higher index number indicated less teacher evaluation policymaking authority devolution to local school districts while a lower index number indicated greater authority being devolved to allow local school districts to shape teacher evaluation policy to their local context. I also developed indices specific to the two formal state policymaking bodies – the state legislature and SBE – that captured the extent to which each entity devolved teacher evaluation policymaking authority to local school districts. The overall teacher evaluation authority devolution index and the teacher evaluation policy sub-elements indices, as well as the SBE and state legislature specific devolution indices are the outcome variables examined in this study. 213 Independent Variables To examine the ways in which state education governance arrangements and politics influence teacher evaluation policymaking authority devolution, my primary predictor variables of interest are as follow: a binary indicator of the SBE’s selection method, a binary indicator of the Chief State School Officer’s (CSSO) selection method, a binary indicator of the presence of SBE term limits, a binary indicator of the presence of legislative term limits, and a continuous variable that offers an index of legislative professionalism (Squire, 2007). Additionally, I take into account state government political ideology using a single variable of state government political ideology developed by Berry, Rinquist, Fording, Hanson, & Clarner (2010). Berry et al.’s (2010) aggregated measure is ideal for two reasons. First, Berry et al. (2010) suggest that static measures of political ideology do not capture longitudinal variation and change over time. Moreover, research that uses solely political party in power as an indicator of state political context ignores the complex decision-making process that involves members of all political parties (Barry et al., 2010). Finally, I include education-related political measures, including a measure of teacher union strength, the state share of local school district revenue in 2011, as well as the average number of students per district. Additional variable specification details are available in Table 3 of Appendix M. Analysis State-by-state variation in teacher evaluation policy authority devolution. To examine the general variation in state teacher evaluation policymaking authority devolution, I produce a visual representation of the teacher evaluation policymaking authority devolution index variation among all fifty states. The visual representation incorporates symbols to show 214 how legislative and SBE authority devolution varies from the overall teacher evaluation policymaking authority devolution index. Additionally, I examine the variation in teacher evaluation policymaking authority devolution statistically using Hartigan’s dip test. Hartigan’s dip test is used to measure multimodality in a sample by “the maximum difference, over all sample points, between the empirical distribution function and the unimodal distribution function that maximizes that maximum difference” (Hartigan & Hartigan, 1985, p. 1). If the teacher evaluation policymaking authority devolution index formed a unimodal distribution, this would indicate that the majority of states devolve some authority over teacher evaluation policy to local school districts with fewer states devolving all or no authority. However, if the distribution of the teacher evaluation policymaking authority devolution index is not unimodal, this would indicate that there is more variance among states devolving all, some and no authority. Finally, to examine whether state legislatures and SBEs vary in the extent to which they devolve teacher evaluation authority differently, I employ a two-sample t-test on the state legislature- and SBEspecific teacher evaluation overall and sub-element policymaking authority devolution index variables. State education governance, politics, and teacher evaluation policy authority devolution. I apply generalized Poisson regression models (GPRMs) to conduct exploratory analyses of state legislatures’ and SBEs’ teacher evaluation policymaking authority devolution. GPRMs are utilized because the small sample size would likely bias the results when using either generalized linear modeling or maximum likelihood estimation. Moreover GPRMs are ideal due to the nature of the outcome variable being a count variable (Wang & Famoye, 1997). The distributions of many of the teacher evaluation sub-elements’ policymaking authority devolution 215 index had a substantial number of zeros; thus, I applied the Vuong (1989) test for model selection to identify whether it was more appropriate to utilize zero-inflated Poisson or zeroinflated binomial regression models. The Vuong test indicated that these alternative models were not appropriate for any of the outcome variables. Moreover, I examined both the estimated dispersion parameter and the Akaike information criterion statistics for each model to examine whether a standard Poisson model (PRM) should be utilized, as opposed to the GPRM. The estimated dispersion parameters from the GPRM were each negative, indicating underdispersion. Moreover, the GPRM has a smaller AIC value than the standard PRM, indicating that the GPRM is preferred to the standard PRM. I also conducted diagnostics to examine whether particular states serve as outliers and would therefore disproportionately impact the analyses. The primary diagnostic examined was Cook’s distance, which measures the aggregate change in the estimated coefficients when an observations left out of the estimation. This diagnostic assessment revealed that California and Hawaii were outliers; thus, these two states were excluded from subsequent analyses. In describing the results of these analyses, incident rate ratios, which are equal to the exponentiated coefficients, are reported for ease of interpretation: an incident rate ratio greater than one indicates that an increase in the independent variable results in an a higher teacher evaluation policymaking authority devolution index number (i.e., less devolution to local school districts) with the number to the right of the decimal signifying the extent of that increase. On the other hand, an incident rate ratio of less than one indicates that an increase in the independent variables results in a lower teacher evaluation policymaking authority devolution index number (i.e., more devolution to local school districts). In this case, subtracting the 216 number to the right of the decimal from one provides an understanding of the extent of that decrease Results In this section, I present findings from my analysis of the variation in teacher evaluation policymaking authority devolution decisions among state governments as a whole, as well as state legislatures and SBEs as separate policymaking institutions. Initially, I explore nationwide trends in the extent to which teacher evaluation policymaking authority is devolved to local school districts while also providing insight into the extent to which formal state education policymaking bodies contribute to these decisions. I also provide insight into the variation in the elements included in state teacher evaluation policies. Next, I statistically examine the extent to which state-level teacher evaluation policymaking authority devolution differences are associated with state education politics and governance arrangements. Finally, I examine whether particular political or governance arrangement factors differentially impact state legislatures’ and SBEs’ teacher evaluation policymaking authority devolution decisions. National trends in teacher evaluation policymaking authority devolution. As shown in Figure 21, there is substantial variation in the extent to which state governments devolve teacher evaluation policymaking authority to local school districts. While some state governments devolve very little teacher evaluation policymaking authority to local school districts, others provide substantial teacher evaluation policymaking authority to local districts. In addition to the Figure 21 visual, statistical tests also confirm substantial state-to-state variation in teacher evaluation policymaking authority devolution levels. Hartigans’ dip test for 217 unimodality rejects the null hypothesis that the distribution of the devolution authority index is unimodal (p <0.001). The distribution of the index measure resembles a trimodal distribution with a number of states leaving much policymaking authority to local districts, and others enacting highly prescriptive teacher evaluation policy. While the primary shade of the state shown in Figure 21 is associated with the overall index measure of teacher evaluation policymaking authority devolution, the map also incorporates an indicator of how much authority local districts are afforded specifically from the state legislature via statute and SBE via administrative rules, as is represented by the shade of a diamond and the triangle, respectively, within each state. For example, in Maine, West Virginia, Massachusetts, South Dakota, Nebraska, Louisiana and North Carolina, among others, both state legislatures and SBEs provide local school districts with some or limited teacher evaluation policymaking authority; however, when the laws, rules and regulations from both of these state policymaking entities are pooled together, locals are ultimately left with little local policymaking authority. States with only one alternate shaded statutory or administrative code symbol are those in which one state policymaking body – the state legislature or SBE – was the primary driver in limiting the amount of teacher evaluation policymaking authority provided to local school districts. For example, in Pennsylvania, the SBE provides significant policymaking authority to local districts; however, because there is no alternate shaded diamond, this indicates that the diamond color would be the same as the overall state color. Thus, in Pennsylvania, the state legislature is the governing body that primarily limits local districts’ teacher evaluation policymaking authority. In contrast, the Illinois legislature provided local districts with some local policymaking authority; however, because there is no alternate shaded triangle, this 218 indicates that the triangle shade would be the same as the overall state shade. Thus, in Illinois, the SBE is the governing body that primarily limited local school districts’ teacher evaluation policymaking authority. It should be noted that states with no alternate colored statutory or administrative code symbol are those in which both statute and administrative code either provide local districts with significant teacher evaluation policymaking authority (i.e., Arizona, California, Connecticut, Delaware, Georgia, Iowa, Minnesota, Montana, New Hampshire, North Dakota, Tennessee, Vermont, Wyoming) or with little to no local authority (i.e., New York, New Jersey). Additionally, four states – Alabama, Mississippi, Rhode Island and South Carolina –have other regulatory, binding documents outside of state statute and administrative code that restrict local authority, which explains why both state statute and administrative code provide substantial local authority but the overall local authority over teacher evaluation policymaking is quite limited. 219 Figure 21 State teacher evaluation policymaking authority devolution State Government Teacher Evaluation Policymaking Authority Devolution Significant local policymaking authority Some local policymaking authority Limited local policymaking authority Little to no local policymaking authority Level of authority via administrative code Level of authority via statute 220 Teacher evaluation policy sub-elements: Which decisions are devolved? Are there particular elements of teacher evaluation policy that are more frequently devolved to local school districts and others that state legislatures and SBEs are more likely to retain control over? The most common mandate handed down from state governments to school districts is related to the frequency of teacher evaluations (i.e., annually, bi-annually, variations for probationary and/or high-performing teachers), with 36 state governments specifying this in statute or administrative code. Additionally, more than half of all state governments require districts to adopt specific rating categories to be used to classify teacher performance (e.g., highly effective, effective, developing, ineffective). Half of all state governments also require districts to implement a teacher evaluation system that incorporates professional practice and student achievement and/or growth. Three state governments require solely professional practice and do not mention student achievement and/or growth; 12 state governments require solely student achievement and/or growth and do not mention professional practice; and professional practice and student achievement and/or growth are not specified in 10 state governments’ laws. While more than half of state governments mandate the frequency with which classroom observations must occur (25) fewer mandate a specific length of time for the classroom observation itself (13). Finally, approximately one in four state governments require conferences between teachers and their evaluator, as well as the use of state teacher standards in evaluating teachers. With regard to data elements, 24 state governments require local districts to use a statespecified or state approved professional practice observation tool or rubric when evaluating teachers. Moreover, 21 state governments require local school districts to use a measure of student growth that is based upon a state specified or state approved student assessment. About 221 half of those states (10) require school districts to incorporate a method of measuring student growth that is specified by the state, while the other half allow locals to determine how student growth on the state assessment is measured. Just 14 states require local districts to incorporate local assessment results into a teacher’s student growth measure. Finally, in determining the weight each teacher evaluation element will hold in a teacher’s overall evaluation, 20 state governments provide specific weights for these components while eight states require either a range or relatively vague description about the weights of the components (e.g., “a significant portion must be based on student growth” or “student growth must be taken into consideration”). Twenty-two states do not specify teacher evaluation element weights that local school districts must consider in calculating a teacher’s overall evaluation score. These findings suggest that technical teacher evaluation policy decisions, such as evaluation frequency, rating categories, and evaluation elements to be taken into account (e.g., professional practice, student achievement and/or growth) are often determined by state legislatures and SBEs. Discretion over more detailed teacher evaluation policy decisions, such as the methods used to measure and incorporate professional practice and student achievement into a summative evaluation rating and the weights that these measures have, is often left to local school districts. Legislatures and SBEs: Devolution differences? To examine whether state legislatures and SBEs vary in the extent to which they devolve teacher evaluation authority, I conducted twosample t-tests of teacher evaluation policymaking authority devolution index variables specific to state legislatures and SBEs. No statistically significant differences exist between overall levels of 222 teacher evaluation policymaking authority devolved by state legislatures and SBEs (t(49)= 0.07, p=0.95). Moreover, no significant differences (at the !=0.05 level) were found among state legislatures and SBEs with regard to devolution of authority related to any of teacher evaluation policy sub-elements. While the overall levels of teacher evaluation policymaking authority devolution do not differ among state legislatures and SBEs, descriptive data provide insight into the differences among teacher evaluation policymaking authority devolution levels for state legislatures and SBEs (Table 27). Similar to the two-sample t-tests, the descriptive data reveals very little variation among state legislatures and SBEs with regard to devolution of teacher evaluation policymaking as a whole or its sub-elements. Table 27 Teacher evaluation authority devolution descriptive statistics, by sub-elements and total State Board of Education State Legislature Mean Median Process 2.04 2.00 Std. Dev. 1.69 Mean Median 2.04 1.00 Std. Dev. 2.23 Data & Measurement 0.50 0.00 0.89 0.86 0.00 Weighting 0.49 0.00 0.88 0.49 Oversight 0.11 0.00 0.31 Approval/Assurance 0.20 0.00 Total 3.38 2.00 Combined Total Mean Median 3.32 3.50 Std. Dev. 2.04 1.28 1.14 1.00 1.26 0.00 0.89 0.88 0.50 1.07 0.07 0.00 0.25 0.22 0.00 0.41 0.39 0.26 0.00 0.44 0.40 0.00 0.48 3.30 3.33 1.50 4.02 6.46 6.50 3.92 State education governance structures and teacher evaluation policymaking authority devolution. Given the variation among state governments’ overall levels of teacher evaluation policymaking authority devolution, are there particular contexts in which this authority devolution is more or less likely to occur? For example, are elected SBEs more likely to devolve policymaking authority than appointed SBEs? Are SBEs or legislatures with term 223 limits less likely to devolve policymaking authority than those that do not experience term limits? Both exploratory (Table 28) and statistical data analyses (Table 29) of these relationships provide insight into the political and governance arrangement factors are related to the extent to which teacher evaluation policymaking authority is devolved to local school districts. Table 28 Total teacher evaluation authority devolution descriptive statistics N Mean Median St. Dev. Mean Median St. Dev. State Board of Education St. Mean Median Dev. 11 34 6.32 6.68 6.50 6.50 3.71 3.97 4.23 3.60 3.00 2.00 3.88 3.29 2.82 3.21 1.00 1.50 4.07 3.89 13 35 6.15 6.64 7.00 6.50 2.95 4.15 5.04 2.99 6.50 2.00 3.47 3.26 1.50 3.87 0.00 2.00 2.07 4.24 12 6.08 5.00 4.80 2.83 1.50 3.53 4.83 3.50 4.28 11 7.18 7.00 4.41 2.55 1.00 3.05 3.95 3.50 4.01 13 4.96 5.50 2.59 2.96 2.00 2.61 1.92 0.00 2.59 12 8.00 8.50 3.03 5.79 6.50 3.70 2.38 0.00 4.32 12 6.17 6.25 3.24 2.79 2.00 2.85 3.38 3.00 3.23 12 5.63 6.00 3.33 3.96 2.00 4.09 0.88 0.00 1.28 12 6.92 6.50 4.28 3.92 4.50 3.27 3.71 2.25 4.36 12 7.33 7.25 4.60 3.50 2.00 3.61 4.96 4.25 4.93 12 6.54 6.25 4.55 3.13 1.50 3.78 2.46 0.00 4.03 12 8.04 8.00 2.95 4.38 3.75 3.30 4.00 3.75 3.80 11 5.59 6.00 3.52 3.27 2.00 3.12 3.14 3.50 3.29 13 5.85 5.00 4.11 3.38 2.00 3.64 3.31 2.00 4.61 Combined Total SBE selection process SBE elected SBE appointed Leg. term limits Leg. term limit Leg. no term limit Leg. professionalism 1 to 25th percentile (0.03 to 0.11) 26 to 50th percentile (0.12 to 0.14) 51 to 75th percentile (0.15 to 0.20) 75 to 99 percentile (0.21 to 0.48) State gov’t ideology 1 to 25th percentile (0 to 10.25) 26 to 50th percentile (10.69 to 34.39) 51 to 75th percentile (35.44 to 72.44) 75 to 99 percentile (72.82 to 91.45) % state funding 1 to 25th percentile (29 to 37%) 26 to 50th percentile (38 to 43%) 50 to 75th percentile (44 to 52%) 76 to 99th percentile (53 to 82%) 224 State Legislature Table 28 (cont’d) Avg. district size 1 to 25th percentile (0-2,000 students) 26 to 50th percentile (2,001-4,000 students) 51 to 75th percentile (4,001-6,000 students) 76 to 99th percentile (6,000 + students) CSSO selection process CSSO elected CSSO appointed Union strength Strongest (Tier 1) Strong (Tier 2) Average (Tier 3) Weak (Tier 4) Weakest (Tier 5) 12 3.88 2.75 3.56 1.75 1.00 2.44 2.17 1.50 2.72 12 7.42 7.25 3.56 4.00 3.00 2.96 4.33 3.50 4.62 12 8.25 7.50 3.61 5.17 6.00 4.01 3.46 0.00 4.88 12 6.50 6.25 3.59 3.25 1.25 3.45 2.96 1.50 3.13 12 36 6.79 6.42 7.00 6.50 4.18 3.78 3.88 3.43 3.75 2.00 3.53 3.40 3.29 3.21 1.00 2.00 4.22 3.85 8 10 11 9 10 6.81 5.65 6.41 6.94 6.85 6.75 5.75 6.50 6.50 7.75 4.12 3.04 4.74 4.21 3.56 3.50 3.40 2.91 3.61 4.35 1.50 3.00 2.00 2.00 4.00 4.37 3.05 2.70 3.65 3.89 2.25 2.25 4.77 4.50 2.15 0.50 0.00 6.00 3.00 0.50 3.16 3.62 4.16 4.99 3.08 Table 29 Generalized Poisson regression model results Combined Total State Legislature IRR (std.error) IRR (std.error) SBE elected 0.97 (0.23) 1.01 (035) Leg. term limit 0.91 (0.19) 1.46 (0.47) Leg. professionalism 3.34 (3.52) 158.98 (291.99)*** State gov’t ideology 1.00 (0.003)* 1.01 (0.01) % state funding 1.30 (1.09) 8.03 (10.72)* CSSO elected 1.09 (0.25) 1.31 (0.45) Avg. district size 1.22 (0.11)** 1.42 (0.23)** Union strength 1.08 (0.08) 1.10 (0.13) Constant 0.58 (0.58) 0.01 (0.02)** /lnphi -2.67 (0.38) -1.89 (0.35) phi 0.07 (0.03) 0.16 States (n) 45 45 **p<0.01; **p<0.05; *p<0.10. State Board of Education IRR (std.error) 0.55 (0.41) 0.36 (0.21)* 0.01 (0.03)* 1.02 (0.01)* 1.44 (4.18) 0.76 (0.56) 0.96 (0.22) 1.60 (0.40)* 1.24 (3.26) -0.88 (0.230) 0.41 45 Political and governmental influences on teacher evaluation policymaking authority devolution. A state’s average school district size has a significant, positive relationship with a state’s teacher evaluation policymaking authority devolution index (p=0.02). For example the average authority devolution index in states with less than 2,000 students per district is 3.9, 225 compared to an average index of 6.5 in states with more than 7,500 students per district. This finding suggests that state governments with few, large school districts, such as North Carolina and Florida, devolve less authority to local school districts than states with many small school districts. If we take consolidation of schools and districts as an indicator of centralization of authority over education issues (Strang, 1987), this finding suggests that states that have expanded control over education issues in general are also more also more likely to expand control over teacher evaluation policy. State government ideology also has a positive, statistically significant relationship with teacher evaluation policymaking authority devolution: states with a more liberal ideology devolve less authority to school districts than more conservative state governments (p=0.08). The average authority devolution index of states in the 25th percentile on the government ideology scale (more conservative) is 6.2, with a maximum index of 11.5, compared to an average index of 7.3 and a maximum of 14.5 for states in the 75th percentile on the government ideology scale (more liberal). There may be several possible explanations for this finding. One possibility is that more conservative state governments espouse traditionally conservative values, including local control of schools (Bartolomé, 2008; Shelly, 2011). Alternatively, it may be that more liberal state governments are more likely to recognize potential value in establishing a rigorous statewide teacher evaluation system. For example, school districts with less time, money or human resources may have little capacity to develop a robust teacher evaluation system – one that will provide meaningful differentiation of teachers that will ostensibly lead to more equitable distribution of high quality teachers among the most underserved students. Thus, more liberal state governments may be more likely to believe they are working towards more equitable distribution of teachers with the establishment of a rigorous statewide teacher evaluation system. 226 Moreover, perhaps more liberal state governments are of the mindset that local control over school policy may facilitate the capture of resources by local elites that face reduced political competition (Bardhan & Mookherjee, 2005). In the case of teacher evaluation policies, this could mean that locally developed policies are likely to reflect the preferences of the local elites, while the voices of those with fewer resources are drowned out. As such, more liberal state governments may advocate for a statewide teacher evaluation system to ostensibly reduce issues of inequity in the policy development process. Legislatures and SBEs: Influences of authority devolution. Similar to a state’s overall teacher evaluation policymaking authority devolution, state government ideology is significantly related to state legislatures’ and SBEs’ teacher evaluation policymaking authority devolution levels, with more liberal state governments devolving less teacher evaluation policymaking authority (legislature, p=0.11; SBE, p=0.10) Furthermore, the average school district size is significantly related to state legislatures’, but not SBEs’, teacher evaluation policymaking authority devolution level (legislature, p=0.02; SBE, p=0.89). Additional noteworthy relationships also emerge with regard to legislature- and SBE-specific teacher evaluation policymaking authority devolution. With regard to state legislatures’ overall teacher evaluation policy authority devolution levels, SBE term limits, legislative professionalism, and state share of school funding are also significantly related to legislative teacher evaluation policymaking authority devolution decisions. State legislatures with non-term limited SBEs devolve 50 percent more teacher evaluation policymaking authority than those with elected SBEs (p=0.14). There are several possible explanations for this finding. One possibility is that state legislatures see non-term 227 limited SBEs as more knowledgeable and experienced and, thus, are more likely to call upon the SBE members for expertise, resulting in less strict state statute around teacher evaluation policy. Alternatively, the interpretation of this finding may be dependent upon the historical relationships between, and political party alignment of the state legislature and SBE. For example, if the state legislature’s majority party aligns with that of the governor, who is often tasked with making SBE appointments, the state legislature may be more likely to call upon the SBE, regardless of term limits, to develop more specific teacher evaluation via the administrative rulemaking process. However, in states where the political parties of the state legislature and governor do not align but the SBEs are not term limited, perhaps the majority party in the legislature still matches the majority party of the SBE as a result of previous governor appointments. In each of these cases, state legislative authority devolution levels would be higher because their involvement in the teacher evaluation process would be relatively minimal. Conversely, it is possible that when a state legislature’s majority party and the governor’s majority party do not align and the SBEs are not term limited, the state legislature may limit or remove the SBE from the teacher evaluation policymaking process and, instead, directly devolve teacher evaluation policymaking authority to local school districts, again, resulting in a lower teacher evaluation policymaking authority index. More professional state legislatures also devolve less teacher evaluation policymaking authority to local school districts (p<0.01). Oftentimes, legislative professionalism is construed as an indicator of greater governance capacity (Thompson & DiIulio, 1998). If this is indeed the case, these results provide some evidence that legislators in more professionalized legislative institutions do see themselves as better equipped to formulate more extensive policy, and act upon that by enacting more restrictive state policy. Finally, state legislatures in states with more 228 centralized school funding systems devolve less teacher evaluation policymaking authority to local school districts (p=0.10). This finding suggests that state legislatures that have a tighter grip on education funding purse strings are also likely to tighten their grip on education policymaking. Regarding SBEs’ teacher evaluation policymaking authority devolution, state legislative term limits, legislative professionalism, and union strength also significantly influence the extent to which SBE’s devolve teacher evaluation policymaking authority. SBEs in states with legislative term limits have an authority devolution index approximately 71 percent lower (i.e., they devolve more authority to local districts) than those with non-term limited legislatures (p=0.035). This relationship between one state policymakers’ authority devolution and the characteristics of another state policymaking institution may occur for a variety of reasons. One possibility is that term limited legislatures, who may be less concerned about reelection (Grofman, 1996), are much more likely to enact strict teacher evaluation policies since they may not be concerned with ramifications of dissatisfied voters during the next election. Similarly, term limited legislators may seek recognition aside from their constituents to curry favor with those who can provide rewards and opportunities after legislative service ends (Kesler, 1990; Montgomery, 1990), which may entice term limited legislators to enact stricter teacher evaluation policies to advance their own future career prospects. Yet another possibility is that term limited legislatures are in a political environment in which executive branch power is intensified (Baker & Hedge, 2013). As such, perhaps state legislatures in these environments are more likely to enact strict teacher evaluation policies in an attempt to increase its power over the executive branch, who may be the appointer of SBE members. In any of these instances, an SBE may choose not to, or may not be called upon, to enact statewide teacher evaluation policies. 229 Conversely, it is possible that legislative term limits reduce the extent of knowledge and experience of state legislators and result in less policy productivity (Olds, 2011). Less experienced and knowledgeable legislators may rely more heavily on their SBEs to enact administrative rules around teacher evaluation. If this were the case, these findings would suggest that SBEs in states with legislative term limits who are called upon by state legislatures are more willing to provide local districts with greater teacher evaluation policymaking authority. SBEs in states with more professional state legislatures devolve less teacher evaluation policymaking authority to local school districts (p=0.09). Again, this relationship between one state policymakers’ authority devolution and the characteristics of another state policymaking institution may occur for different reasons. If legislative professionalism is construed as an indicator of greater governance capacity (Thompson & DiIulio, 1998), these results may suggest that more professionalized legislatures may either draw upon the expertise of the SBE, who then enacts strict administrative rules; or, more professionalized legislatures may explicitly call upon the SBE to enact more strict administrative rules. Finally, SBEs in states with weaker teacher unions devolve less teacher evaluation policymaking authority to local school districts (p=0.06). Given that teachers unions have long been in control of local teacher evaluation policies through the collective bargaining process (Cowen & Strunk, 2015), this finding provides evidence that states with stronger teacher unions may be better positioned to utilize their financial and human resources to advocate for local control of teacher evaluation. Moreover, this finding provides support for one of the aforementioned conflicting hypotheses: in the case of education, low visibility of special-purpose governments may hinder public participation and create a bias toward special interests that have 230 the capacity to lobby special-purpose government officials (Bollens, 1957; Burns, 1994; Macedo & Karpowitz, 2006). Discussion and Implications for Future Research With the passage of the Every Student Succeeds Act (ESSA) in December 2015, state and local policymakers ostensibly gained newfound authority over policies, practices and processes taking place in their elementary and secondary schools. As a result, school leaders, policymakers and researchers have begun working to understand and potentially seize upon the flexibility that the reformed federal policy ostensibly provides to states and local school districts. However, state-level teacher evaluation policies have been codified in nearly every state in response to previous federal education law. The analyses presented in this chapter provides empirical evidence that prior to ESSA state education policymakers embraced a wide array of approaches to teacher evaluation policymaking: while state legislatures in some states devolved substantial policymaking authority to local school districts to develop and shape teacher evaluation policymaking to their local context, other state legislatures and SBEs – sometimes in conjunction with one another, sometimes in silos – enacted a more uniform, standardized state teacher evaluation system specified in state-level policy. In addition to the overarching state-bystate variation, the results of this study have at least three specific implications for policymakers, education practitioners, and researchers of state-level policymaking. First, the quantitative results presented in this chapter suggest that there is a relationship between state education policymaking bodies’ devolution of teacher evaluation policymaking authority and state-level political and governance arrangements. For one, more professional legislatures are associated with less policymaking authority devolution to local school districts. 231 Moreover, legislatures in states with term limited SBEs, as well as SBEs in states with termlimited legislatures are both associated with greater policymaking authority devolution to local school districts. Given that the institutional arrangements of one state education policymaking body seem to influence the decisions of another policymaking body, these results underscore the importance of incorporating both state legislatures and SBEs into the study of the education policymaking process. Future research should focus on the interplay among and relationships between these state policymaking bodies and the ways in which politics and institutional structures may influence the other’s policy decisions. Second, results from this study show that political decisions related to education made by states in the past are associated with current state policy decisions. These results suggest that in a more flexible ESSA context, local school districts in states that have a history of centralization, either through a more centralized school funding system or the consolidation of school districts, may find that their state legislatures and SBEs are less likely to dismantle previously established statewide teacher evaluation policies. On the other hand, in states with more decentralized funding systems and less emphasis on school district consolidation, states legislatures and SBEs may be more willing to return to their modus operandi, providing school districts with greater authority to determine if, how, and why they want to evaluate their teachers. Third, more liberal state government ideology is interestingly associated with less devolution of state teacher evaluation policymaking authority to local school districts. This finding may not, on the surface, come as a surprise: decentralization has historically been associated with conservative ideologies, and centralization with liberal ideologies. However, support for pro-labor groups traditionally lean more liberal while conservatives have been viewed as “anti-union” (Boyle, 1998; Dark, 1999; Francia, 2006; Lichtenstein, 2002). Given the 232 stronghold that teachers unions have had on teacher evaluation policies and practices through the collective bargaining process, it is peculiar that more conservative state governments devolve more authority to local school districts, where the teachers unions are likely influential in policy decisions. On the other hand, more liberal state governments are more likely to retain teacher evaluation policymaking authority, instead enacting more uniform, statewide systems of teacher evaluation. One possible explanation for this finding is that, with the recent proliferation of rightto-work laws in states with conservative leadership, state education policymakers may have been more willing to devolve teacher evaluation policymaking decisions to local school districts once the strength of the teachers union began to diminish under right-to-work laws. Yet, findings from this study also suggest that this may only take hold when state legislatures are the primary teacher evaluation policymaker. When it comes to SBE policymaking authority devolution decisions, stronger teacher unions are associated with greater devolution of teacher evaluation policymaking authority to local school districts. These findings warrant further investigation. Is it simply states with stronger teachers unions are more likely to elect SBE members that align with liberal policy views and, as a result, devolve more authority to the local school districts where the teachers unions can potentially shape local teacher evaluation policy? Or, does the potentially low visibility of SBE, in comparison to the state legislature, provide an advantage to special interest groups, such as teachers unions, that have the capacity to lobby SBE members? This quantitative analysis cannot provide a complete picture of how this complex process plays out, suggesting that further qualitative research is necessary to better understand the political complexities of the state education policymaking process. 233 Conclusion Important questions related to the ways in which state education politics and governance influences the state education policymaking process arise from these findings. It is clear that teacher evaluation policy is no longer a policy established solely at the local level, primarily by local teachers unions through their role as a collective bargaining unit. New state-level actors located in distinct state-level political institutions with different norms, rules, and traditions have entered the teacher evaluation policymaking space. Research thus far has only examined the variation in state teacher evaluation policy elements (Steinberg & Donaldson, 2016). This study builds upon this work by providing an understanding of who state-level teacher evaluation policymakers are and of how decisions to devolve policymaking authority are a culmination of both state legislatures and SBEs. By incorporating both legislatures and SBEs in research related to the state policymaking process, as well as examining the two education policymaking bodies as separate entities, we can broaden the conversations about the role that each body plays in the policymaking process and estimate the determinants of state policymakers’ decisions to devolve policymaking authority, both at the state-level in general, as well as the level of each individual policymaking body. 234 Chapter VIII Significance of the Study and Future Research James Reichley, author of States in Crisis, wrote: [T]he problems, the tribulations, the opportunities for future development of state government are of utmost importance to every American because the American political system has, to a great extent, been based and constructed on the fact of the existence of the states. If the states should fail to meet the challenges that are being put to them during the seventh decade of the twentieth century, the only alternatives, short of chaos, would be an enormous expansion of the resources and authority of local government or acceleration of the present trend toward complete assumption of governmental responsibility by the national government … [which] would, carried to its logical extreme, almost surely undermine if not destroy the political freedom, regional diversity, and social and economic flexibility which have thus far characterized the American system (Reichly, 1964, p. 256-257). When it comes to education, policymakers in state legislatures and SBEs have long held the responsibility to meet the challenges of educating the citizens of their state. Yet, for most of American history, authority over most education policies and practices has been devolved to local school boards. In recent years, however, the American public has shifted toward a preference for more state education government involvement in education policy. Even more 235 recently, the federal government has shifted some policymaking power back to the states – and there is no indication that this trend is going to slow down. As state education policymakers are both being asked to be more involved in making education policy and actively taking more control of education policy, I aimed to provide a more thorough understanding of who state education policymakers are and how they approach the education policymaker process. In particular, this study empirically examined the following questions: (1) Who gains power to make state education policy? a. How do demographic characteristics of SBE members and state legislators serving on education committee vary among states? b. Under what institutional conditions do individuals with specific demographic characteristics gain state education policymaking power? (2) Do state education policymakers’ demographic characteristics reflect state population demographics? a. Under what institutional conditions do the demographics of state-level education policymakers align with the demographics of the state population? (3) Are policymakers responsive to public and/or teacher policy preferences? a. Do individual or institutional characteristics impact state education policymaker responsiveness to public and/or teacher policy preferences? (4) When forming their position on an education policy, whose voices do state education policymakers value? a. Do individual or institutional characteristics impact who state education policymakers nominate as valuable voices in their decisionmaking process? 236 (5) What are the internal qualities of state education policymakers? More specifically, how do state education policymakers come to their positions of power and what are their goals? (6) How do state education policymakers approach the education policymaking process and what are state education policymakers preferences related to whose internal voices carry weight in the state education policymaking process? (7) How do state education governance structures impact formal state education policymaking bodies approach to education policymaking in the form of policy devolution? While many studies have examined state education policy implementation and how state governments respond to federal education policy reform, the majority of these studies ascribe policy action to “the state” or “state government,” and/or largely ignore the complexity of the state education policy and governance. The results of this study suggest that state education governance research should embrace a more comprehensive understanding of the complexity of the actors that are involved in state education policymaking and the ways in which institutional structures and individual characteristics of state education policymakers impact the policymaking process. This dissertation employed a rich mixed-methods approach to understanding who state education policymakers are, whose voices are valued by state education policymakers, and whose internal voices are heard in the state education policymaking process. This study also examined the impacts that institutional structures of formal state education policymaking bodies have on who gains power, whose voices are heard, and what decisions are made in the 237 policymaking process. To my knowledge, this is the first study to (a) expand descriptive representative to include comprehensive review of policymakers in both general- and specialpurpose governing bodies, (b) examine policymaker responsiveness specific to education policy issues, (c) examine the intersections of institutional structures, individual policymaker characteristics, descriptive representation, and substantive representation, and (d) explore the impacts of institutional structures and policymakers’ approaches to and decisions in the education policymaking process. Drawing upon institutional and agency theory, I examine the ways in which the characteristics of education policymaking body structures and individuals influence the education policymaking process. First, the theory of new institutionalism suggests that the structures and institutions in which actors reside have particular norms and rules that can influence the behaviors of the actors (March & Olsen, 1984). As such, this dissertation argues that the structures of state legislatures and SBE members impact who gains state education policymaking power, whose voices are valued and heard in the policymaking process, and what policy alternatives are considered. However, I also recognize that individuals within these institutions do retain some agency in their decisions (DiMaggio, 1988; Fligstein, 1997; Giddens, 1976, 1979; Granovetter, 1985; March & Olsen, 1984; Ostrom, 1990, 1991; Seo & Creed, 2002). Therefore, this study also posits that individual state education policymaker characteristics also play a role in shaping whose voices are valued in the state education policymaking process and what policy alternatives are considered. Among the institutional structure variables, I find that both the macro structures (i.e., SBEs, state legislatures) and the micro structures (e.g., within SBE, selection processes, geographic representation, equality requirements; within legislatures, professionalism, term 238 limits) impact who gains state education policymaking power. For example, females and individuals with experience as professional educators are more prominent on SBEs than education committees. Within SBEs, appointed SBEs are comprised of more females; and appointed SBEs that require legislative approval, as well as partisan elected SBEs are comprised of more non-white members. With regard to education committee membership demographics, more professional legislatures are associated with a greater proportion of females and non-white members. Moreover, legislative term limits are associated with more education committee members with a Bachelor’s degree and fewer with a two-year degree. Finally, elected SBEs are associated with an overrepresentation of female members; and legislatures with term limits are associated with an overrepresentation of education committee members with bachelor’s degrees and legislatures with higher levels of legislative professionalism are associated with an overQualitative data provides evidence that those who gain state education policymaking power may not be all that different in one aspect: motivation. Most state education policymakers did not seek their position of power and, even upon gaining a position of power, did not perceive their work in education to be of great value or purpose. One institutional structure – selection method of SBEs – did emerge as a point of variation in that elected SBE members were more likely to seek out a position of education policymaking power and have more specific education goals. I also find that macro structures impact whose voices are heard and valued in the education policymaking process. For example, I find that while education committee members were more likely to value the voices of constituents, SBE members valued the voices of other state officials, such as state education agency staff, education advocacy groups, education research organizations, school leadership groups, and the governor. 239 With regard to individual state education policymaker characteristics, one demographic characteristic was found to consistently influence state education policymakers’ responsiveness to different voices in the policymaking process: political party. In particular, after exposure to the policy preferences of teachers, Republican state education policymakers were more likely to vote in opposition to teacher policy preferences; and, Democrat state education policymakers were no more likely to support teacher policy preferences. Additionally, while Democrats often valued the voices of education advocacy groups and teachers unions, Republican were more likely to value the voices of the governor and business organizations. Nonetheless, both Republicans and Democrats valued the voices of school leadership groups, local schoolteachers, and local school leaders. Qualitative analyses also revealed variation in whose voices were valued based on political party. For example, Democrats were much more likely to advocate for a strong, proactive SBE in the policymaking process, and a more limited role for the state legislature. Political party leaders were also perceived as essential partners in gaining state education policymaking power – particularly by elected SBE members. Moreover, I found that the ideological leaning of SBEs and legislatures is associated with state education policymakers’ approach to policy, particularly as it relates to policymaking authority devolution. In particular, more liberal state legislatures and SBEs devolve less teacher evaluation policymaking authority. Yet, the quantitative rating, DCE, and most/least important data suggested that political party leaders were not seen as highly valued voices in the policymaking process. As such, future research should examine this misalignment between (a) qualitative interview data and the RCT survey question responses and (b) rating, DCE, and most/least important survey question responses with regard to the importance of political party leaders in the state education policymaking process. 240 Implications As local school board demographics have become more representative of historically marginalized groups, the public is calling upon state education policymakers to be more involved in making education policy. Yet, very little is known about who state education policymakers are and their approach to the education policymaking process. Descriptive results that describe state education policymaker demographics illustrate the lack of diversity among state education policymakers. For example, 44 percent of today’s local school board members are women (Hess & Meeks, 2010). I find that just 32 percent of formal state education policymakers are women. Moreover, minorities make up nearly 20 percent of today’s local school board members; and, in large school districts, nearly one in every three school board members is non-white (Hess & Meeks, 2010). In comparison, 17 percent of state education policymakers are non-white. Not only are state education policymakers non-representative of the general population, they are also non-responsive to the policy preferences of both the general population and those who are intimately impacted by many of their education policy decisions – teachers. Yet, state education policymakers do value the voices of many different constituencies, including school leaders, schoolteachers, school leadership groups, and education research organizations. While the value of the voices of these individuals and groups vary among state education policymakers located in distinct institutions and with particular individual demographics, this research provides some optimism that these voices are valued. These findings have implications for both education policy and educators. In particular, as state education policymakers are becoming more involved in making education policy they are increasingly impacting the ways schools operate. As such, it is important they those making policy have a clear understanding of the implications of their actions. One way in which this is 241 possible is by encouraging those with a wide array of formal and informal experiences in and with education – beyond simply being a student – to pursue and obtain state education policymaking power. This study suggests that there are some institutional structures that could be put in place to attempt to increase the diversity of those making state education policy. Moreover, the U.S. Department of Education (2014) recently reported that minorities now outnumber whites among the nation’s public school students for the first time. As America’s school-age population becomes more diverse, it will become increasingly important that state education policymakers recognize the different educational experiences of students of various backgrounds and the ways in which policies will impact the day-to-day experiences of the students in their state. For example, as state legislators and SBE members grapple with issues such as how to provide a rapidly increasing student population of English learners (ELs) with an adequate and equitable education or how to ensure that LGBTQ+ students feel safe in their school environment, it is likely beneficial to have individuals with first-hand experience involved in the policymaking process. Having such individuals serving as education policymakers is one way to accomplish this. Yet, we cannot expect educators or those with first-hand experience related to the issues state education policymakers are tackling to pursue a position as a state education policymaker. Moreover, simply having a more diverse state education policymaking does not guarantee that (a) policymakers will be substantively representative of the populations they identify with and (b) the voices of policymakers with diverse perspectives will carry weight in the policymaking process. As such, it is imperative that those with experiences and expertise valuable to the education policymaker process understand both who education policymakers are, who is likely to listen to their voice, and how to effectively engage in the education policymaking process. The 242 results of this study suggest that individual schoolteachers and school leaders are voices valued by state education policymakers – and, perhaps some policymakers in particular institutions and with particular personal characteristics may be more or less responsive. Nonetheless, many postsecondary degrees for schoolteachers and school leaders are not required to and do not provide educators with the necessary tools to engage in the state policymaking process. A review of the websites of the graduate programs among U.S. News and World Report’s Top 20 Graduate Schools of Education (2017) reveals that in many programs, policy studies and the politics of state education policymaking (if available) are optional elective courses for future educators and school leaders. Those seeking to engage in the policymaking process could benefit from knowing who education policymakers how and how they approach the education policymaking process in that this information may help educators and school leaders tailor their approach to state policymakers in different state education policymaking bodies. While not all educators and school leaders will have the time or be interested in engaging in the state education policymaking process, educators should have equitable access to knowledge and tools to understand how to become engaged and share valuable experience with those charged with making education policy that impacts the ways schools operate. Conclusion As state education policymakers are increasingly called upon to make education policies that impact students’ day-to-day schooling experiences and the daily operation of schools, it is important to understand who is making these policies, whose voices are heard in this policymaking process, and how state education policymakers approach the policymaking process. This dissertation set out to explore these questions, as well as to better understand the 243 institutional structures and individual characteristics that may make some faces more or less present among state education policymakers, some voices more or less heard by state education policymakers, and some decisions by state education policymakers more or less likely. Results suggests that both institutional structures such as whether an SBE is elected or appointed or whether a education committee member is part of a more professional legislature impact who gains access to education policymaker power and whose voices are heard in the policymaking process. Moreover, I find that simply being located in one state education policymaking body over another results in different approaches to the devolution of education policymaking authority. The results of this study make an important contribution to the relatively non-existent knowledge base related to the myriad of formal actors in the state education policymaking process. Nonetheless, the findings of this study lead to new questions that warrant further investigation. Future research could expand the examination of state education policymaking processes to other education issues beyond teacher evaluation policy. For example, research could explore education issues that are both brought to the attention of state policymakers from a grassroots approach compared to those that federal policymakers require state policymakers address. Additionally, research could further explore why state education policymakers’ value local schoolteacher voices but place relatively low value on teacher union policy preferences and are unresponsiveness to the preferences of teachers at large. Knowledge from this research could potentially provide educators with an understanding of different ways to effectively approach state education policymakers to share their experiences and expertise throughout the policymaking process. Finally, qualitative research should be expanded to additional states. While this study focused on maximizing variation in education governance structures, it would 244 also be beneficial to explore similarities and differences among state education policymakers’ and their approach to the education policymaking process among states with similar education governance structures. 245 APPENDICES 246 Appendix A Current State Constitution Education Clauses Table 30 Current state constitution education clauses State Constituti onal Location Alabama Article XIV Alaska Article VII Arizona Article XI Arkansas Article 14 California Article IX Colorado Connecticut Article IX Article Eighth Delaware Article X Education Clause Language The legislature shall establish, organize, and maintain a liberal system of public schools throughout the state for the benefit of the children thereof between the ages of seven and twenty-one years. Separate schools be provided for white and colored children, and no children of either race shall be permitted to attend a school of the other race. The legislature shall by general law establish and maintain a system of public schools open to all children of the State The legislature shall enact such laws as shall provide for the establishment and maintenance of a general and uniform public school system Intelligence and virtue being the safeguards of liberty and the bulwark of a free and good government, the State shall ever maintain a general, suitable and efficient system of free public schools and shall adopt all suitable means to secure to the people the advantages and opportunities of education. The Legislature shall encourage by all suitable means the promotion of intellectual, scientific, moral, and agricultural improvement. The general assembly shall, as soon as practicable, provide for the establishment and maintenance of a thorough and uniform system of free public schools throughout the state, wherein all residents of the state, between the ages of six and twenty-one years, may be educated gratuitously. There shall always be free public elementary and secondary schools in the state. The general assembly shall implement this principle by appropriate legislation. The General Assembly shall provide for the establishment and maintenance of a general and efficient system of free public schools, and may require by law that every child, not physically or mentally disabled, shall attend the public schools, unless educated by other means. 247 Table 30 (cont’d) Florida Article IX Georgia Article VIII Hawaii Article X Idaho Article IX Illinois Article X Indiana Article 8 Iowa Article IX Kansas Article 6 Section 183 Article VIII Kentucky Louisiana The education of children is a fundamental value of the people of the State of Florida. It is, therefore, a paramount duty of the state to make adequate provision for the education of all children residing within its borders. Adequate provision shall be made by law for a uniform, efficient, safe, secure, and high quality system of free public schools that allows students to obtain high quality education. The provision of an adequate public education for the citizens shall be a primary obligation of the State of Georgia. Public education for the citizens prior to the college or postsecondary level shall be free and shall be provided for by taxation. The State shall provide for the establishment, support and control of a statewide system of public schools free from sectarian control, a state university, public libraries and other such educational institutions as may be deemed desirable, including physical facilities therefor. The stability of a republican form of government depending mainly upon the intelligence of the people, it shall be the duty of the legislature of Idaho, to establish and maintain a general, uniform and thorough system of public, free common schools. A fundamental goal of the People of the State is the educational development of all persons to the limits of their capacities. The state shall provide for an efficient system of high quality public educational institutions and services. Education in public schools through the secondary level shall be free. Knowledge and learning, generally diffused throughout a community, being essential to the preservation of a free government; it shall be the duty of the General Assembly to encourage, by all suitable means, moral, intellectual, scientific, and agricultural improvement; and to provide, by law, for a general and uniform system of Common Schools, wherein tuition shall be without charge, and equally open to all. The general assembly shall have power to abolish or reorganize said board of education, and provide for the educational interest of the state in any other manner that to them shall seem best and proper. The general assembly shall encourage, by all suitable means, the promotion of intellectual, scientific, moral, and agricultural improvement. The legislature shall provide for intellectual, educational, vocational and scientific improvement by establishing and maintain public schools, educational institutions and related activities which may be organized and changed in such manner as may be provided by law. The General Assembly shall, by appropriate legislation, provide for an efficient system of common schools throughout the State. The legislature shall provide for the education of the people of the state and shall establish and maintain a public educational system. 248 Table 30 (cont’d) Maine Article VIII Maryland Article VIII Massachusett s Chapter V Michigan Article VIII Minnesota Article XIII Mississippi Article 8 Missouri Article IX A general diffusion of the advantages of education being essential to the preservation of the rights and liberties of the people; to promote this important object, the Legislature are authorised, and it shall be their duty to require, the several towns to make suitable provision, at their own expense, for the support and maintenance of public schools. The General Assembly, at its first session after the adoption of this constitution, shall, by law, establish, throughout the State, a thorough and efficient system of free public schools, and shall provide by taxation, or otherwise, for their maintenance. Wisdom, and knowledge, as well as a virtue, diffused generally among the body of the people, being necessary for the preservation of their rights and liberties; and as these depend on spreading the opportunities and advantages of education in the various parts of the country, and among the different orders of the people, it shall be the duty of legislators and magistrates, in all future periods of this Commonwealth, to cherish the interests of literature and the sciences, and all seminaries of them: especially the university at Cambridge, public schools, and grammar schools in the towns; to encourage private societies and public institutions, rewards and immunities, for the promotion of agriculture, arts, sciences, commerce, trades, manufactures, and a natural history of the country; to countenance and inculcate the principles of humanity and general benevolence, public and private charity, industry and frugality, honesty and punctuality in their dealings; sincerity, good humor, and all social affections, and generous sentiments among the people. Religion, morality and knowledge being necessary to good government and the happiness of mankind, schools and the means of education shall forever be encouraged. […] The legislature shall maintain and support a system of free public elementary and secondary schools as defined by law. Every school district shall provide for the education of its pupils without discrimination as to religion, creed, race, color or national origin. The stability of a republican form of government depending mainly upon the intelligence of the people, it is the duty of the legislature to establish a general and uniform system of public schools. The legislature shall make such provisions by taxation or otherwise as will secure a thorough and efficient system of public schools throughout the state The Legislature shall, by general law, provide for the establishment, maintenance and support of free public schools upon such conditions and limitations as the Legislature may prescribe. A general diffusion of knowledge and intelligence being essential to the preservation of the rights and liberties of the people, the general assembly shall establish and maintain free public schools for the gratuitous instruction of all persons in this state within ages not in excess of twenty-one years as prescribed by law. 249 Table 30 (cont’d) Montana Nebraska Article X Article VIII Nevada Article 11 New Hampshire Sec. LXXXIII New Jersey Section IV New Mexico Article 12 New York Article XI It is the goal of the people to establish a system of education which will develop the full educational potential of each person. Equality of educational opportunity is guaranteed to each person of the state. The legislature shall provide a basic system of free quality public elementary and secondary schools. It shall fund and distribute in an equitable manner to the school district the state’s share of the cost of the basic elementary and secondary school system. The legislature shall provide for the free instruction in the common schools of this state of all persons between the ages of five and twenty-one years. The legislature shall encourage by all suitable means the promotion of intellectual, literacy, scientific, mining, mechanical, agricultural, and moral improvements, and also provide for a superintendent of public instruction and by law prescribe the manner of appointment, term of office and the duties thereof. The legislature shall provide for a uniform system of common schools, by which a school shall be established and maintained in each school district at least six months of ever year. Knowledge and learning generally diffused through a community, being essential to the preservation of a free government; and spreading the opportunities and advantages of education the various parts of the country being highly conducive to promote this end, it shall be the duty of legislators and magistrates, in all future periods of this government, to cherish the interests of literature and the sciences, and all seminaries and public schools; to encourage private and public institutions, rewards and immunities for the promotion of agriculture, arts, sciences, commerce, trade, manufactures, and natural history of the country; to countenance and inculcate the principles of humanity and general benevolence, public and private charity, industry and economy, honesty and punctuality, sincerity, sobriety, and all social affections, and generous sentiments among the people. The legislature shall provide for the maintenance and support of a thorough and efficient system of free public schools for the instruction of all children in the State between the ages of five and eighteen years. A uniform system of free public schools sufficient for the education of, and open to, all the children of school age in the State shall be established and maintained. The legislature shall provide for the maintenance and support of a system of free common schools, wherein all the children of this State may be educated. 250 Table 30 (cont’d) North Carolina Article I & Article IX North Dakota Article 8 Ohio Article VI Oklahoma Article I Article I: The people have a right to the privilege of education, and it is the duty of the State to guard and maintain that right; Article II: Religion, morality and knowledge being necessary to good government and happiness of mankind, schools, libraries, and the means of education, shall forever be encouraged. The General Assembly shall provide by taxation and otherwise for a general and uniform system of free public schools, which shall be maintained at least nine months in every year, and wherein equal opportunities shall be provided for all students A high degree of intelligence, patriotism, integrity and morality on the part of every voter in a government by the people being necessary in order to insure the continuance of that government and the prosperity and happiness of the people, the legislative assembly shall make provision for the establishment and maintenance of a system of public schools which shall be open to all children of the state of North Dakota and free from sectarian control. The legislative requirement shall be irrevocable without the consent of the United States and the people of North Dakota. The legislative assembly shall provide for a uniform system of free public schools throughout the state, beginning with the primary and extending through all grades up to and including normal and collegiate courses. In all schools instruction shall be given as far as practicable in those branches of knowledge that tend to impress upon the mind the vital importance of truthfulness, temperance, purity, public spirit, and respect for honest labor of every kind. The legislative assembly shall take such other steps as may be necessary to prevent illiteracy, secure a reasonable degree of uniformity in course of study, and to promote industrial, scientific, and agricultural improvements. The principal of all funds, arising form the sale, or other disposition of lands, or other property, granted or entrusted to this State for educational and religious purposes, shall be used or disposed of in such manner as the General Assembly shall prescribe by law. The General Assembly shall make such provisions by taxation, or otherwise, as, with the income arising from the school trust fund, will secure a thorough and efficient system of common schools throughout the state; but no religious or other sect, or sects, shall ever have any exclusive right to, or control of, any part of the school funds of this state. Provision shall be made by law for the organization, administration, and control of the public school system of the state supported by public funds. The legislature shall establish and maintain a system of free public schools wherein all the children of the state may be educated. The legislature shall provide for the compulsory attendance of some public or other schools, unless other means of education are provided, of all the children in the State who are sound in mind and body, between the ages of eight, and sixteen years, for at least three months in each year. 251 Table 30 (cont’d) Oregon Article VIII Pennsylvania Article III Rhode Island Article XII South Carolina Article XI South Dakota Article VIII Tennessee Article XI Texas Article VII The legislative assembly shall provide by law for the establishment of a uniform and general system of common schools. […] The Legislative Assembly shall appropriate in each biennium a sum of money sufficient to ensure a system of Equalization Grants to public education meets quality goals established by law, and publish a report that either demonstrates the appropriation is sufficient, or identifies the reasons for the insufficiency, its extent, and its impact on the ability of the state’s system of public education to meet those goals. The General Assembly shall provide for the maintenance and support of a thorough and efficient system of public education to serve the needs of the Commonwealth. The diffusion of knowledge, as well as of virtue, among the people, being essential to the preservation of their rights and liberties, it shall be the duty of the general assembly to promote public schools and public libraries, and to adopt all means which it may deem necessary and proper to secure to the people the advantages and opportunities of education and public library services. The General Assembly shall provide for the maintenance and support of a system of free public schools open to all children in the State and shall establish, organize, and support such other public institutions of learning, as may be desirable. The stability of a republican form of government depending on the morality and intelligence of the people, it shall be the duty of the Legislature to establish and maintain a general and uniform system of public schools wherein tuition shall be without charge, and equally open to all; and to adopt all suitable means to secure to the people the advantage and opportunities of education The state of Tennessee recognizes the inherent value of education and encourages it support. The General Assembly shall provide for the maintenance, support and eligibility standards of a system of free public schools. A general diffusion of knowledge being essential to the preservation of the liberties and rights of the people, it shall be the duty of the Legislature of the State to establish and make suitable provision for the support and maintenance of an efficient system of public free schools. 252 Table 30 (cont’d) Utah Article X Vermont Article 41 Virginia Article VIII Washington West Virginia Article IX Wisconsin Section X Wyoming Article 1 & Article 7 Article XII The legislature shall provide for the establishment and maintenance of the state’s education system including a public education system, which shall be open to all children of the state Laws for the encouragement of virtue and prevention of vice and immorality, ought to be constantly kept in force, and duly executed; and a competent number of schools ought to be maintained in each town unless the general assembly permits other provisions for the convenient instruction of youth: and one or more grammar schools shall be incorporated and properly supported, in each county in this State. The General Assembly shall provide for a system of free public elementary and secondary schools for all children of school age throughout the Commonwealth, and shall seek to ensure that an education program of high quality is established and continually maintained. It is the paramount duty of the State to make ample provision for the education of all children residing within its borers, without distinction or preference on account of race, color, caste, or sex. The legislature shall provide for a general and uniform system of public schools. The public school system shall include common schools, and such high schools, normal schools, and technical schools as many hereafter be established. But the entire revenue derived from the common school fund and the state tax for common schools shall be exclusively applied to the support of the common schools. The Legislature shall provide, by general law, for a thorough and efficient system of free schools. […]The Legislature shall foster and encourage moral, intellectual, scientific and agricultural improvement The legislature shall provide by law for the establishment of district schools, which shall be as nearly uniform as practicable, and such schools shall be free, and without charge for tuition to all children between the ages of four and twenty years, and no sectarian instruction shall be allowed therein. Article 1: The right of the citizens to opportunities for education should have practical recognition. The legislature shall suitably encourage means and agencies calculated to advance the sciences and liberal arts; Article 7: The legislature shall provide for the establishment and maintenance of a complete and uniform system of public instruction, embracing free elementary schools of every needed kind and grade, a university with such technical and professional departments as the public good may require and the means of the state allow, and such other institutions as may be necessary. […] In none of the public schools so established and maintained shall distinction or discrimination be made on account of sex, race or color. 253 Appendix B Data and Methods for Examining Legal Relationships between State Government and Public Education I begin by reviewing each state’s original constitution, specifically focusing on the education clause(s). Constitutions were then tracked forward in time to identify all amendments to each state’s education clause. I began my search for each state’s original and amended constitutions by exploring primary sources available from the HathiTrust digital library, state government websites, and state historical and archival websites. Subsequently, I utilized Westlaw, an electronic legal database, to verify that I accurately identified all modifications to state constitution education clauses. For those states that did not establish an SBE in their state constitution, I utilized Westlaw to examine the history of state statute that established an SBE. The text of both constitution- and statute-based education content was compiled into a single spreadsheet to then be analyzed. Arkin (1999) suggested that content analysis can be applied to legal documents, which includes both constitutions and statutes. Drawing upon Krippendorf (1980), who wrote “humanities and the social sciences, including efforts to improve the political and social conditions of life, is concerned with symbols, meanings, messages, their functions, and effects” (p.9), Arkin (1999) argued, [w]ithin social organizations the right to use a particular channel of communications is regulated and whatever data one obtains in such contexts, they reveal what an institution deems permissible (p. 47) […] [l]aws certainly fit the criteria of being documents that are regulated and whose primary purpose is political and social improvement (p. 67). 254 Constitutions and statutes are written products of human communication (Rossman & Rallis, 2003). Therefore, I engage in content analysis to objectively and systematically examine these legal documents for both patterns as well as a descriptive understanding of the origins and changes over time in the legal relationships between state government and public education. 255 Appendix C State Board of Education Characteristics by State Table 31 State board of education characteristics by state Elected State Legal Establishme nt Number of Voting Members Region Based 8 At large Partisan Gov erno r Onl y Legislative Involvement (nomination or approval) Alabama Statute 9 9 Alaska Statute 7 7 7 Arizona Constitution 11 11 11 Arkansas Statute 9 9 9 California Constitution Region Based At large Partisan Term Length Term Limits Distribution equality requirements 4 4 3 5 8 1 7 Political party 1 Race Determined by the state legislature Statute 11 11 Colorado 10 4 Constitution 7 7 Connecticut Statute 14 Delaware 11 14 4 Statute Florida 7 7 7 Georgia Constitution 7 7 7 Constitution 10 10 10 Hawaii Constitution Statute Number of Members Appointed 7 Partisan Determined by the state legislature 11 5 2 6 Political party 4 10 7 x 9 9 256 6 3 3 3 (consecutive) Table 31 (cont’d) Idaho Constitution Statute Illinois Determined by the state legislature 8 8 Constitution Statute 9 9 Statute 9 9 Iowa Statute 10 9 Constitution 5 Determined by the state legislature Indiana Kansas 7 10 10 6 9 7 2 8 2 10 10 Louisiana Statute 4 Political party 6 Political party Partisan 4 12 11 11 Nonpartisan 4 3 4 3 (consecutive) 1 5 2 4 2 5 2 11 Maine Statute 11 9 9 Maryland Statute 12 12 11 Massachusetts Statute 11 11 Michigan Constitution 10 8 Mississippi Constitution 9 9 9 Missouri Constitution 8 8 8 8 7 Determined by state legislature 11 3 4 11 Constitution 8 7 Constitution Montana 8 9 8 Partisan 8 7 9 9 Statute Nebraska Nevada New Hampshire New Jersey New Mexico Constitution 9 8 8 Statute 11 7 4 Statute 7 7 Statute Constitution (advisory only) 13 13 10 10 Statute Political party 7 Nonpartisan Nonpartisan 17 4 3 4 7 5 13 10 Partisan 17 257 13 2 4 6 4 17 (legislature only) New York Political party Determined by state legislature Statute Kentucky 4 5 3 3 Gender Table 31 (cont’d) North Carolina Constitution 19 13 North Dakota Statute 7 7 Ohio 6 Constitution Statute Oklahoma 11 21 19 11 Constitution 8 3 6 8 6 Determined by the state legislature Nonpartisan 8 4 Determined by the state legislature Statute 7 7 6 Statute 9 7 7 4 Pennsylvania Statute 22 21 17 6 Rhode Island Statute 8 8 8 3 9 4 Oregon South Carolina South Dakota 2 (consecutive) Constitution 17 17 Statute 9 9 Nonpartisan 16 5 1 4 2 (consecutive) 3 (successive) 1 Political party, gender, race, age Tennessee Statute Texas Statute Utah Vermont Virginia 11 10 15 15 15 Determined by the state legislature Nonpartisan Statute 19 15 Statute 11 9 Constitution 9 15 16 14 Constitution 12 9 13 6 4 4 10 6 1 (consecutive) 4 2 (consecutive) x 9 Statute Statute 5 Partisan Determined by state legislature Washington Wyoming 9 Determined by the state legislature Constitution Statute West Virginia 9 Constitution 9 5 (elected by school district directors) 1 (elected by board of directors of all private schools) 2 (consecutive) 11 7 4 9 9 11 258 7 4 6 Political party 1 Political party Appendix D Methods of Data Collection for State Education Policymaker Demographic Dataset To develop the state education policymaker demographic dataset, I first visited the Senate and House of Representatives40 education committee websites of all fifty states, as well as the websites of all 47 SBEs.41 I collected the first and last names of each state education policymaker with voting power.42 A total of 1,742 state education policymakers were accounted for in this dataset. State education policymakers’ biographies available on state legislature and SBE websites served as the primary data source of individual demographic characteristics. Secondary sources of data included Vote Smart, as well as news and media sources. Vote Smart is a nonprofit, non-partisan research institution founded in 1992 that maintains an online repository of local, state and federal candidates for public office in the United States. Vote Smart data is currently compiled and managed by 23 full-time employees and as many as 70 student interns at Drake University in Des Moines, Iowa and other local colleges and universities. When demographic data was not directly available on state legislature and SBE websites or Vote Smart, I conducted an Internet search of the individual state education policymaker’s name, state and office (e.g., “Senator John Smith Virginia”; “State board of education member Jill Smith Maine”) and examined news articles, personal and campaign webpages, and social media pages to obtain needed demographic data. 40 Also known as the “general assembly,” “state assembly,” or “house of delegates” Minnesota, New Mexico and Wisconsin do not have a formal SBE. 42 Some SBE members serve as ex-officio members (e.g., student members, governor, lieutenant governor, state superintendent) and do not have voting power. These individuals were not included in the dataset. 41 259 With regard to state education policymaker gender, when an individual identified with the male or female gender in their biography located on either the state legislature or SBE website, the individual was assigned to the gender category of male (0) or female (1). If no gender was specified on the state legislature or SBE website, I turned to both the Vote Smart database and an Internet search and assigned the individual the gender they were most frequently referred to in the news and/or the gender to which Vote Smart assigned to them. Similar processes were followed when compiling state education policymakers’ race and Latino origin. I first ascribed an individual to the racial category of either non-white (1) or white (0) and to the Latino origin category of either Latino (1) or non-Latino (0) based on information provided in their state legislature or SBE website biography. When this information was not available, I explored Internet news sources and Vote Smart. For example, any individual that was listed on Vote Smart as a member of a black or Hispanic legislative caucus was given non-white or Latino status. Similarly, any individual that was referred to as African American, Asian, Native American, Pacific Islander, Hispanic, Latino, or Puerto Rican – among others – in a news article was given non-white or Latino status, as applicable. It should be noted that all Latino’s were identified as non-white. I collected educational attainment levels and profession data using a similar process as previously described. Educational attainment was broken into seven categories: high school diploma; some college, no degree; 2-year degree; Bachelor’s degree; Master’s degree; and doctorate or professional degree. If educational attainment information was unknown after reviewing an individual’s biography, Vote Smart, and conducting an Internet search, an individual was assigned to an eighth category of “no data.” I created a categorical variable, ranging from 0 (no data) to 8 (doctorate or professional degree). Individuals with data that 260 indicated they completed high school but did not provide any other educational attainment data were assigned the category of “high school degree.” Individuals with data that indicated they attended college but did not provide an indication that they completed a degree were assigned the category of “some college, no degree.” Individuals with data that indicated they completed a technical degree, certificate program, or Associate’s degree were assigned the category of “2year degree.” Finally, individuals with data that indicated they completed a Bachelor’s degree; a Master’s or specialist degree (e.g., EdS); and a doctorate or advanced degree were assigned to those categories. For those who completed a doctorate or advanced degree, I also collected detailed data on the type of degree, namely whether the individual received an advanced degree in education (EdD), law (JD, LLM), medicine (e.g., DC, DDS, DMD, DO, DVM, MD, OD), or philosophy (PhD). Data on state education policymakers’ primary profession was placed into one of eight inductively developed categories: education, business/entrepreneurship, health, science/engineering, law, lobbying/consulting, government/politics, and nonprofit/advocacy/interest group. Individuals assigned to the education category include those who’s day-to-day work takes place inside of a school, college or university, including teachers, professors, counselors, administrators, human resources and administrative staff, speech language pathologists, athletic directors, and school nurses. Substitute teachers and individuals working for non-profit organizations that work with education institutions were not included in this category. An individual must have served in this capacity for more than five years to be assigned this category. The business/entrepreneurship category includes individuals in any business profession such as realty, marketing, insurance, journalism, business management, and farming. The health category includes individuals involved in health professions such as nurses 261 and doctors, as well as those involved in health finance, health administration and pharmacy. The law category includes individuals that practice law as their primary profession. The lobbying/consulting category includes individuals who self-identified as consultants or lobbyists or as “government relations” for specific consulting or lobbing organizations. The government/politics category incorporates those involved in government work at local, state and federal levels in all branches of government, as well as military personnel and police. Individuals who listed no other profession listed other than their current position as an elected official were ascribed to the government/politics category. Finally, the nonprofit/advocacy/interest group category includes individuals that work for a non-profit organization or those who work for an advocacy/interest group organization and are not directly involved in lobbying. This category also includes individuals working in religious organizations as well as individuals who’s primary self-described profession is as a volunteer for non-profit or advocacy groups (e.g., ParentTeacher Association). I created a binary variable for each of these professions (0=not associated with profession, 1=associated with profession). Individuals were allowed to be associated with more than one profession if they were part of the profession for more than five consecutive years. For example, an individual who was in the military for 10 years and then became a teacher for 15 years was assigned to both the education and government/politics profession categories. Finally, I compiled data on the political party affiliations and political campaign contributions of SBE members only. Governors appoint the majority of SBE members across the nation. Scholars of political science have suggested that governors may “plumb appointments for their political allies” (Kousser & Phillips, 2012, p.4). Similarly, research has found a positive relationship between campaign contributions and a governor’s ability to appoint officials, suggesting that individuals who are interested in influencing the governor’s appointment 262 decisions can be expected to contribute financially to the campaign (Abrams, 1981). In the case of SBEs, this may lead one to hypothesize that governors will appoint individuals who align with their political party and/or contribute financially to their campaign—either directly to the governor’s campaign or to the political party committees that provide support to the governor’s campaign. Therefore, I include both political party affiliations and political campaign contributions as a descriptive characteristic of SBE members. In contrast to SBE members, legislators serving on education committees are always elected by local constituents. Research has already established that legislatures are not descriptively representative of their state populations (Rosenthal, 1998). Thus, because the political party makeup of education committees are proportional to that of the state legislature as a whole, I did not collect data on education committee member political party affiliations. Moreover, no research to date indicates that political campaign contributions play a role in the legislative committee assignment process. Rather, scholars of political science have argued that majority party leaders select legislative committee members and leaders that they believe will ensure favorable action on policies that align with the party leader’s goals and priorities (Liebowitz & Tollison, 1980). Researchers have also shown that assignment to state legislative committees is related to seniority and legislators’ margin of victory in the previous election (Hedlund & Patterson, 1992). Therefore, political party affiliation and political campaign contribution data was not collected for this group of education policymakers. To compile data on SBE members’ political contributions, I drew upon state-level campaign finance databases. I used the contributor search option to collect all political donations made by SBE members to either a sitting governor or directly to a political party for all available 263 years in a state’s campaign finance database.43 I collected data on the name of the governor or political party committee, as well as the amount of the contribution and the date on which the contribution was made. To identify the political party affiliations of SBE members, I first examined whether the SBE member had a declared political party either in running for office or in their biographical information listed on the SBE website. For those SBE members whose political party was not readily available in this way, I conducted a search of political campaign contributions. I first explored state-level campaign finance databases to identify whether a SBE member donated to primarily Republican or Democrat candidates or political parties and categorized the SBE member accordingly. If a SBE member donated to members of both parties, I examined both the number of contributions made to candidates of each party and the SBE member was categorized with the political party affiliation to which they donated most often. If no data was available at the state-level, then I explored campaign finance data from the Federal Election Commission and categorized the SBE member based on her federal political campaign donations. Finally, if the campaign finance search still yielded no results, I conducted an Internet search of SBE members to identify indicators of political party (e.g., serving as a political party convention delegate, ran for or held another office with a political party affiliation). If this search yielded no results, I did not classify the SBE member as being affiliated with any political party. 43 There is wide variation in the date ranges for which contributor data can be found on state campaign finance databases. In some states, data was available beginning in 1991, while other states only housed data back to 2011. I collected campaign contributions of SBE members for all years for which data was available. 264 Appendix E Methods of Collecting State Population Demographics I drew from survey respondents’ answers to the CCES survey questions listed in Table 32. Using the CCES “pre election state name” variable, I aggregated individual survey respondent data at the state level to create state-level state population demographic data. I then mapped CCES survey answer categories onto the categories of the aggregated, state-level demographic characteristic data I collected for state education policymakers. Table 32 CCES survey data information CCES Survey Question Are you male or female? What racial or ethnic group best describes you? Are you of Spanish, Latino, or Hispanic origin or descent? CCES Survey Answer Options Male (1) Female (2) Skipped (8) Not asked (9) White (1) Black (2) Hispanic (3) Asian (4) Native American (5) Mixed (6) Other (7) Middle Eastern (8) Skipped (98) Not Asked (99) Yes (1) No (2) Skipped (8) Not Asked (9) CCES Survey Total Responses 26283 29917 0 0 41419 6971 3895 1365 414 1193 865 78 0 0 1764 50381 159 3896 265 State Education Policymaker Demographic Data Match Male Female No data White Non-white No data Latino Non-Latino No data Table 32 (cont’d) What is the highest level of education you have completed? From whom did you vote for President in 2012? No HS (1) High school graduate (2) Some college (3) 2-year (4) 4-year (5) Post-grad (6) 1493 15703 12917 5349 13969 Skipped (8) Not Asked (9) Barack Obama (1) Mitt Romney (2) Someone else (3) Did not vote (4) Don’t recall Not Asked (9) 0 0 22569 18041 1469 144 662 0 6769 266 No matched category High school diploma Some college, no degree Associate’s/technical degree/certificate Bachelor’s degree Master’s degree + Doctorate/Professional degree No data No data Democrat Republican Other/Not available No matched category No matched category No data Appendix F Parity Ratio Distributions Figure 22 State Board of Education-to-Population Parity Ratio Distribution by demographic characteristic Density 0 .5 1 Density 012 State board of education-to-population parity ratio distribution by demographic characteristics 1 White 1.2 1.4 0 .5 1.5 2 2.5 Male 4 Latino 6 8 0 .1 .2 .3 .4 .5 Highest education level: High school diploma Density 0246 2 Density 024 0 0 .5 1 1.5 2 2.5 Highest education level: 4-year degree 1.5 0 .2 .4 .6 .8 1 Highest education level: 2-year degree Density 0 .1.2 .5 1 Highest education level: Some college Density 0 .5 1 0 2 4 6 8 10 Highest education level: Postgraduate degree Density 0.2.4.6 0 1 Density 0 510 .8 Density 0.2.4.6 .6 .5 1 1.5 Democrat 2 2.5 267 Figure 23 Education Committee Member-to-Population Parity Ratio Distribution Education committee member-to-population parity ratio distribution by demographic characteristics 0 Density 2 4 Density 0 .5 1 1.5 by demographic characteristic .6 .8 White 1 1.2 1 2 2.5 Density 5 0 0 0 5 10 15 0 .2 .4 .6 Education level: High school diploma 0 1 2 Education level: 2-year degree .8 Density 0 1 2 3 Density 0 .5 1 1.5 Latino .2 .4 .6 .8 Education level: Some college 1 3 0 Density 0 .1 .2 .3 Density .5 1 0 .5 1.5 Male Density .2 .4 .4 1 1.5 2 Education level: 4-year degree 2.5 0 2 4 6 Education level: Postgraduate degree 268 8 Appendix G Literature Review Methods and Data Who do state education policymakers listen to when making policy? I began examining this question by reviewing previous literature that focuses on substantive representation by state policymakers related to education issues. To avoid systematic bias and error in my review, I follow data analysis strategies put forth by Whittemore and Knafl (2005), as described in Table 33. To begin, I conducted a targeted ProQuest search of peer-reviewed research that included the term “substantive representation” or “policy responsiveness” in the abstract and that was carried out in the United States. This initial search yielded 5,917 pieces of work. I then excluded the following document types: commentaries, reviews, news articles, editorials, correspondence, and market research. This modified search yielded 500 pieces of research. Next, I reviewed the title, abstract, classification codes, and subject terms for each of the 500 articles and identified 64 articles that focused specifically on substantive representation or responsiveness to constituency preferences by policymakers. The other 436 articles were eliminated from the literature review because either (a) the study did not include data from the United States, or (b) as opposed to a focus on descriptive representation or political responsiveness of policymaking bodies, the study focused on market responsiveness; responsiveness to economic trends; private sector or non-profit organization responsiveness to consumers or market trends; student or teacher responsiveness to reform, intervention or parent needs/desires; or responsiveness of local, state, regional, and federal courts. Within the 64 relevant articles, 15 focused on substantive representation of state policymakers, three of which 269 focused on substantive representation of state bureaucratic agencies and 12 focused on substantive representation of state legislatures. Additionally, four of the 64 relevant articles focused on substantive responsiveness by policymakers on education policy issues (two focused on K12 policymakers at the local level; two focused on higher education policymakers at the state level). In discussing the data collection stage of a literature review, Randolph (2009) asserts, “electronic searches lead to only about ten percent of the articles that will compromise an exhaustive review” (p. 7). Thus, I followed Randolph’s (2009) suggestion of conducting a snowball-like search that included searching the references of the articles that were retrieved, determining which of those seem relevant, finding those, reading their references, and repeating the process until a point of saturation is reached – “a point where no new relevant articles come to light” (p. 7). This snowball method led to the identification of 76 additional articles focused on policy responsiveness or descriptive representation, 27 of which were focused on state policymakers and one of which was focused on education policy issues. Table 34 provides a list of all literature that was reviewed. The first column includes my classification of each source’s governance level, as well as whether the source focuses on (a) education as a specific policy issue or (b) responsiveness or substantive representation of bureaucratic agencies. If neither notation is provided, then the source focused on policymakers in a legislative branch on a noneducation issue. I conducted two complementary literature reviews focused on policymakers’ responsiveness to public opinion data. I began with a targeted ProQuest search of peer-reviewed research that included the terms “polls” and “policymaking” in the abstract and were carried out in the United States. This initial search yielded 65 results. I then excluded the following 270 document types: newspapers, trade journals, magazines, wire feeds, and reports. This modified search yielded 26 results. I reviewed the abstracts of each of these sources and eliminated those that did not focus on public opinion polls and policymaking, resulting in a total of six relevant sources. The snowball method led to 11 additional relevant sources. Of the 17 relevant sources, just one was focused on state policymakers use of polling data and none were focused on the use of polling data around education-specific issues. While conducting the previously described search, I noticed that the term “public opinion” was used quite frequently when discussing public polling data. Therefore, I conducted a second targeted ProQuest search of peer-reviewed research that included the terms “public opinion” and “policymaking” in the abstract and were carried out in the United States. This search yielded 67 results. I then excluded articles in trade journals, magazines, and newspapers. This modified search yielded 53 results. I reviewed the abstracts of each of these sources and eliminated those that did not focus on public opinion polls and policymaking, resulting in a total of four relevant sources. The snowball method led to 49 additional relevant sources. Of the 53 relevant sources, nine focused on state policymakers use of public opinion data and none were focused on the use of public opinion data around education-specific issues. 271 Table 33 Integrative literature review data analysis stages, purposes, and descriptions Data analysis stage Problem identification stage Purpose Description Clearly identify the problem that the review is addressing and the review purpose The main problem I am interested in studying is whether state policymakers advance the policy preferences that serve the interests of the represented. The purpose of my literature review is to better understand the ways in which scholars have studied the relationships between public preferences and policymakers’ policy decisions. In particular, I am interested in (a) how have scholars studied this relationship at the state-level? and (b) how have scholars studies this relationship as it relates to education-specific policy issues. Literature search stage Identify the maximum number of eligible primary sources, using at least two to three stages (Jadad et al., 1998; Conn et al., 2003b). Clearly document the literature search process, including search terms, the database used, additional search strategies, and the inclusion and exclusion criteria for determining relevant primary sources. To identify the maximum number of eligible sources, I utilized ProQuest, which is a database of scholarly publications with abstracting and indexing that enables strategic acquisition, management and discovery of information collections. I conducted targeted searches of the following terms: (a) substantive representation OR policy responsiveness, (b) polls AND policymaking, and (c) public opinion AND policymaking. The searchers were limited to (a) resources with a focus in the United States and (b) academic resources such as journal articles, books, book chapters, academic conference papers, and dissertations. Upon retrieving the sources from these ProQuest searches, I reviewed the title, abstract, classification codes, and subject terms of each source and identified those that were indeed focused on the topic of interest. Next, in a snowball sampling-like method, I reviewed the reference page of all relevant sources to determine if other sources should be reviewed. Additional details of these processes are provided in literature review methods section. 272 Table 33 (cont’d) Data evaluation stage The authenticity, methodological quality, informational value, and representativeness of available sources is considered (Kirkevold, 1997) The sample that was evaluated for this integrative review included empirical and theoretical sources. These sources were coded according to three criteria relevant to this review: governance level focus (three levels: local, state, and federal), bureaucratic focus (two categories: yes and no), and education focus (two categories: yes and no). A source could be coded for multiple criteria. Data analysis Source data is ordered, coded, categorized, and summarized into a unified and integrated conclusion about the research problem (Cooper, 1998) The sample that was analyzed for this integrative review included all empirical sources focused on (a) state-level policymaking bodies (including bureaucratic agencies) or (b) education policy issues. I reviewed each of these sources, identifying the data and methods used and the researchers’ conclusion. Presentation Report conclusions of integrative review in table or diagrammatic form. Provide explicit details from sources and evidence to support conclusions, demonstrated in a logical chain of evidence, allowing the reader of the review to ascertain that the conclusions of the review did not exceed the evidence (Oxman, 1994). Presentation of literature review results listed in text. 273 Table 34 Literature review sources Topic Title Author(s) “Substantive representation” and “policy responsiveness” literature Federal Substantive representation for Hispanics: Santos, A. Explaining Congressional support for Hispanic issues Federal Patterns of representation: Dynamics of Wlezien, C. public preferences and policy Federal Inequality and democratic Gilens, M. responsiveness Federal Beyond women, children, and families: Boles, J.L., & Gender, representation, and public Scheurer, K. funding for the arts Federal Are female House members still more Frederick, B. liberal in a polarized era? The conditional nature of the relationship between descriptive and substantive representation? Federal Dynamic representation(s): Federal Nicholson-Crotty, S., criminal justice policy and an alternative Peterson, D. A. M., dimension of public mood & Ramirez, M. D. Federal Partisan moods: Polarization and the Ura, J.D., & Ellis, C. dynamics of mass party preferences R. Federal Women, earmarks, and substantive Schulze, C. representation Federal Representing Latinos: Examining Wallace, S.J. Descriptive and Substantive Representation in Congress 274 Publication Venue Year of Publication Type of Publication 2000 Journal Harvard Journal of Hispanic Policy 2004 Journal The Journal of Politics 2005 Journal Public Opinion Quarterly 2007 Journal Social Science Quarterly 2009 Journal Congress & the Presidency 2009 Journal Political Behavior 2012 Journal The Journal of Politics 2013 Journal 2014 Journal Journal of Women, Politics & Policy Political Research Quarterly Table 34 (cont’d) Federal Federal Federal Federal Federal Federal Federal Federal Federal Federal Federal Evaluating Asian American representation through bill sponsors and cosponsors from 1999 to 2009 The symbolic and substantive representation of LGB Americans in the US House The dynamics of earmark requests for the women and men of the US House of Representatives A Legislature or a legislator like me? Citizen demand for collective and dyadic political representation Democracy's denominator: Reassessing responsiveness with public opinion on the national policy agenda Constituents' responses to descriptive and substantive representation in Congress Economics and politics: The case of sugar tariff reform Racial differences in information, expectations, and accountability Cold War, public opinion, and foreign policy spending decisions: Dynamic representation by Congress and the President Personal attributes and Latino voting behavior in Congress Opportunities and expectations: The gendered organization of legislative committees in Germany, Sweden, and the United States Chaturvedi, N.S. 2015 Journal Journal of Asian American Studies Hansen, E. R., & Treul, S. A. 2015 Journal The Journal of Politics Schulze, C., & Hurvitz, J. 2016 Journal Journal of Women, Politics & Policy Harden, J.J., & Clark, C. J. 2016 Journal American Politics Research Barabas, J. 2016 Journal Public Opinion Quarterly Jones, P.E. 2016 Journal Social Science Quarterly Ellison, S.F., & Mullin, W.P. Griffin, J.D., & Flavin, P. Cox, D.G., & Duffin, D. L. 1995 Journal 2007 Journal Journal of Law and Economics The Journal of Politics 2008 Journal Congress & the Presidency Rocca, M.S., Sanchez, G. R., & Uscinski, J. Bolzendahl, C. 2008 Journal Social Science Quarterly 2014 Journal Gender & Society 275 Table 34 (cont’d) Federal Weighing the alternatives: Preferences, parties, and constituency in roll-call voting Pursuing women’s' interests in partisan times: Explaining gender differences in legislative activity on health, education, and women's health issues The impact of women in Congress Ramey, A. 2015 Journal The Journal of Politics Swers, M.L. 2016 Journal Journal of Women, Politics & Policy Dodson, D.L. 2006 Journal Women and representation: A different view of the district Quasi-experimental design, constituency, and advancing women's interest: Reexamining the influence of gender on substantive representation The difference women make: The policy impact of women in Congress Women in the club: Gender and policy making in the Senate How women legislate Gerrity, J.C., Osborn, T., & Mendez, J.M. MacDonald, J.A. & O'Brien, E.E. 2007 Journal New York, NY: Oxford University Press Politics & Gender 2011 Journal Political Research Quarterly Swers, M. 2002 Book Swers, M.L. 2013 Book Thomas, S. 1994 Book Federal Enlarging representation: Women bringing marginalized perspectives to floor debate in the House of Representatives. Cramer Walsh, K. 2002 Book Chapter Federal Do differences matter? Women members of Congress and the Hyde Amendment Transforming Congress from the inside: Women in committee Levy, D., Tien, C., & Aved, R. Norton, N.H. 2002 Journal Chicago, IL: University of Chicago Press. Chicago, IL: University of Chicago Press New York, NY: Oxford University Press In C.S. Rosenthal (Ed.), Women Transforming Congress. Normal, OK: University of Oklahoma Press Women & Politics 2002 Book Chapter Speaking as women: Women and floor speeches in the Senate Osborn, T., & Mendez, J.M. 2010 Journal Federal Federal Federal Federal Federal Federal Federal Federal Federal 276 In C.S. Rosenthal (Ed.), Women Transforming Congress. Normal, OK: University of Oklahoma Press Journal of Women, Politics & Policy Table 34 (cont’d) Federal Speaking for the underrepresented in the House of Representatives: Voicing women's interests in a partisan era Speaking out: An analysis of democratic and republican women-invoked rhetoric of the 105th Congress Representing women: Congresswomen's perception of their representational roles Pearson, K., & Dancey, L. 2011 Journal Politics & Gender Shogan, C. 2001 Journal Women & Politics Carroll, S. 2002 Book chapter Frederick, B. 2010 Journal Frederick, B. 2011 Journal Guinier, C.L. 1994 Book Cameron, C., Epstein, D., & O'Halloran, S. Lublin, D. 1996 Journal American Political Science Review Federal Gender and patterns of roll-call voting in the Senate Gender turnover and roll call voting in the US Senate The tyranny of the majority: Fundamental fairness in representative democracy Do majority-minority districts maximize substantive black representation in Congress? The paradox of representation In C.S. Rosenthal (Ed.), Women Transforming Congress. Normal, OK: University of Oklahoma Press Congress and the Presidency Journal of Women, Politics & Policy New York: The Free Press 1997 Book Federal Black faces, black interest Swain, C. 1995 Book Federal Sizing up the Senate: The unequal consequences of equal representation Sheer numbers: Critical representation thresholds and women's political representation Acting for women: Is what legislators say, what they do? Lee, F.E., & Oppenheimer, B.I. Beckwith, K., & Cowell-Myers, K. 1999 Book 2007 Journal Princeton, NJ: Princeton University Press. Cambridge, MA: Harvard University Press. Chicago, IL: University of Chicago Press Perspectives on Politics Dodson, D. 2001 Book Chapter Federal Federal Federal Federal Federal Federal Federal Federal 277 In S. Carroll (Ed.), The impact of women in public office. Bloomington, IN: Indiana University Press Table 34 (cont’d) Federal Support for women's interests in the 103rd congress: The distinct impact of congressional women The effects of gender on congressional behavior and the substantive representation of women Female legislators and women's rights agenda: From feminine mystique to feminist era Dolan, J. 1997 Journal Women & Politics Vega, A., & Firestone, J.M. 1995 Journal Legislative Studies Quarterly Wolbrecht, C. 2002 Book Chapter Are women more liberal than men in the U.S. Congress? Should blacks represent blacks and women represent women? A contingent yes. Latino representation: It's direct and indirect Latinos and substantive representation in the U.S. House of Representatives: Direct, indirect or nonexistent? Latino representation in the U.S. Congress: To what extent are Latinos substantively represented? Welch, S. 1985 Journal Mansbridge, J. 1999 Journal In C.S. Rosenthal (Ed.), Women Transforming Congress. Normal, OK: University of Oklahoma Press Legislative Studies Quarterly Journal of Politics Kerr, B., & Miller, W. Hero, R.F., & Tolbert, C. 1997 Journal 1995 Journal Casellas, J.P. 2002 Conferenc e Paper Federal Trends in substantive and descriptive minority representation, 1974-2000 Epstein, D., O'Halloran, S. 2006 Book Chapter Federal Sex and voting in the US House of Representatives: 1961-1975 Francovic, K. 1977 Journal Federal Federal Federal Federal Federal Federal Federal 278 American Journal of Political Science American Journal of Political Science Paper presented at the Annual Meeting of the Southern Political Science Association. Savannah, GA In D. Epstein, R. de la Garza, S. O'Halloran, & R. Pildes (Eds.), The Future of the Voting Rights Act. New York, NY: Russell Sage Foundation Press American Politics Quarterly Table 34 (cont’d) Federal Indelible effects: The impact of women of color in the US Congress Bedolla, L.G., Katherine, T., & Wong, J. 2005 Book Chapter Federal Are women more likely to vote for women's issue bills than their male colleagues? Women in Congress: Do they act as advocates for women's issues? Swers, M.L 1998 Journal Swers, M.L. & Larson, C. 2005 Book Chapter Federal Rethinking women's substantive representation 2008 Journal Federal Critical mass theory and women's political representation Black political representation: An examination of legislative activity within US House Committees The effect of black Congressional representation on political participation Congress in black and white: Race and representation in Washington and at home. Oversight: Representing the interests of blacks and Latinos in Congress Diversity in political institutions and Congressional responsiveness to minority interests The politics of presence Celis, K., Childs, S., Kantola, J., & Krook, M.L. Childs, S., & Krook, M.L. Gamble, K.L. In S. Thomas, & C. Wilcox (Eds.), Women & Elective Office: Past, Present, and Future. New York, NY: Oxford University Press Representation 2008 Journal Political Studies 2007 Journal Legislative Studies Quarterly Gay, C. 2001 Journal Grose, C. 2011 Book American Political Science Review Cambridge: Cambridge University Press Minta, M.D. 2011 Book Minta, D., & Sinclair-Chapman, V. 2012 Journal Phillips, A. 1995 Book Federal Federal Federal Federal Federal Federal Federal 279 In S. Thomas, & C. Wilcox (Eds.), Women & Elective Office: Past, Present, and Future. New York, NY: Oxford University Press Legislative Studies Quarterly Princeton, NJ: Princeton University Press Political Research Quarterly Oxford: Oxford University Press Table 34 (cont’d) Federal Federal Federal Federal (bureaucracy) Federal (bureaucracy) Federal (bureaucracy) Federal (bureaucracy) Federal (bureaucracy) Federal (bureaucracy) Federal (international comparison Federal (international comparison) Federal (International comparison) The color of representation: Congressional behavior and black interests Black faces in the mirror: African Americans and their representatives in the US Congress Going home: Black representatives and their constituencies Policy outputs and bureaucracy. The roles of need, demand, and agency structure Formal procedures, informal processes, accountability, and responsiveness in bureaucratic policy making: An institutional policy analysis Are regulators responsive? Whitby, K.J. 1997 Book Ann Arbor: University of Michigan Press Tate, K. 2003 Book Princeton, NJ: Princeton University Press Fenno, R.F. 2003 Book Robins, B.J. 1983 Journal Chicago, IL: University of Chicago Press Urban Affairs Review West, W.F. 2004 Journal Public Administration Review Vibeke, L.N. 2006 Journal Law & Policy Race, gender, and government contracting: Different explanations or new prospects for theory? Staff of the people? Assessing progress in descriptive representation under the Obama Administration Technical management in an age of openness: The political, public, and environmental forest ranger Political institutions and the opinionpolicy link Fernandez, S., Malatesta, D., & Smith, C. R. Villalobos, J.D. 2013 Journal Public Administration Review 2011 Journal Race, Gender & Class Anderson, S.E., Hodges, H.E., & Anderson, T.L. Wlezien, C., & Soroka, S. N. 2013 Journal Journal of Policy Analysis and Management 2012 Journal West European Politics Social policy responsiveness in developed democracies Brooks, C., & Manza, J. 2006 Journal American Sociological Review Why do welfare states persist? Brooks, C., & Manza, J. 2006 Journal The Journal of Politics 280 Table 34 (cont’d) Federal (international comparison) Federal (international comparison) Federal & state State State State State State State State State State Assessing need in the United States, Germany, and Sweden: The organization of welfare casework and the potential for responsiveness in the "Three Worlds" In the defense of women: Gender, office holding, and national security policy in established democracies Latino representation in state houses and Congress Gender and roll call voting in the New York State Assembly Black substantive representation in state legislatures from 1971-1994 Administrative responsiveness to the disadvantaged: The case of children's health insurance The behavior and success of Latino legislators: Evidence from the states Estimating the effect of redistricting on minority substantive representation Income inequality and policy representation in the American states Black legislative politics: Examining the issue of voting rights in the postHurricane Katrina period Minority policies and political participation among Latinos: Exploring Latinos’ response to substantive representation The realities of electoral reform Jewell, C. J. 2007 Journal Law & Policy Koch, M.T., & Fulton, S. A. 2011 Journal The Journal of Politics Casellas, J. 2011 Book Barnello, M.A. 1999 Journal Cambridge: Cambridge University Press Women & Politics Owens, C.T. 2005 Journal Social Science Quarterly Fossett, J., & Thompson, F. J. 2006 Journal Bratton, K.A. 2006 Journal Journal of Public Administration Research and Theory Social Science Quarterly Epstein, D., Herron, M.C., O'Halloran, & Park, D. Flavin, P. 2007 Journal Journal of Law Economics & Organization 2012 Journal American Politics Research Hoston, W.T. 2012 Journal The Social Science Journal Jeong, H.O. 2013 Journal Social Science Quarterly Stephanopoulos, N. O., McGhee, E. M., & Rogers, S. 2015 Journal Vanderbilt Law Review 281 Table 34 (cont’d) State State State State State State State State State State State Interest groups or the public interest: Why do we regulate health occupations? Legislating without constraints: The effect of minority districting on legislators' responsiveness to constituency preferences Does the new VRA Section 5 overrule Georgia v. Ashcroft? Contrasting party dynamics: A three decade analysis of the representation of Democratic versus Republican women state legislators Agenda-setting and legislative success in state legislatures: the effects of gender and race How women represent women: Political parties, gender, and representation in state legislatures Representing women: Sex, gender, and legislative behavior in Arizona and California When women lead: Integrative leadership in state legislatures African American legislators in the American states The color of their skin or the content of their behavior? Race and perceptions of African American legislators Emerging influentials in state legislatures: Women, blacks and Hispanics Graddy, E. 1991 Journal Gay, C. 2007 Journal Epstein, D.L., & O'Halloran, S. Elder, L. 2008 Journal 2014 Journal New York University Annual Survey of American Law The Social Science Journal Bratton, K.A. and Haynie, K.L. 1999 Journal Journal of Politics Osborn, T.L. 2012 Book New York, NY: Oxford University Press Reingold, B. 2000 Book Rosenthal, C.S. 1998 Book Haynie, K.L. 2001 Book Haynie, K.L. 2002 Journal Chapel Hill, NC: The University of North Carolina Press New York, NY: Oxford University Press New York: Columbia University Press Legislative Studies Quarterly Nelson, A.J. 1991 Book 282 Journal of Health Politics, Policy and Law The Journal of Politics New York, Praeger Table 34 (cont’d) State State State State State State State State State State State Representation is not enough: The limits and possibilities of minority inclusion and influence in representative policymaking bodies in the American states Policy responsiveness to the black electorate: Programmatic versus symbolic representation Racial fairness in legislative redistricting Preuhs, R.R. 2001 Conferenc e Paper Paper presented at the American Political Science Association’s Annual Meeting. San Francisco, CA Bullock, C.S., & MacManus, S.A. 1981 Journal American Politics Quarterly King, G., Bruce, J., & Gelman, A. 1995 Book chapter The policy priorities of African American women in state legislatures Do women legislators matter? Female legislators and state abortion policy Representing women: Women state legislators as agents of policy-related change Barrett, E. 1995 Journal Berkman, M., & O'Connor, R. Carroll, S. 1993 Journal In P.E Peterson (Ed.) Classifying by race. Princeton, NJ: Princeton University Press Legislative Studies Quarterly American Politics Research 2001 Book Chapter Exploring gender differences in state legislators' policy preferences Does their difference make a difference? The impact of women on public policy in the Arizona legislature The impact of women on state legislative policies The impact of gender on the activities and priorities of state legislators: Legislative styles and policy priorities Reshaping the agenda: Women in state legislatures Poggione, S. 2004 Journal In S. Carroll (Ed.), The Impact of women in public office. Bloomington, IN: Indiana University Press Political Research Quarterly Saint-Germain, M. 1989 Journal Social Science Quarterly Thomas, S. 1991 Journal Journal of Politics Thomas, S., & Welch, S. 1991 Journal Western Political Quarterly Dodson, D.L., & Carroll, S.J. 1991 Book New Brunswick: Center for the American Woman and Politics, Eagleton Institute Rutgers University 283 Table 34 (cont’d) State Women in state legislatures: Feminist identity and legislative behaviors Critical mass theory revisited: The behavior and success of token women in state legislatures Agenda setting and African-American women in state legislatures Dolan, K., & Lynne, F.E. Bratton, K.A. 1995 Journal American Politics Quarterly 2005 Journal Politics & Gender Bratton, K.A., Haynie, K.L. & Reingold, B. 2006 Book Chapter The intersection of race and gender on representation - black women legislators' impact on legislation Welfare policymaking and intersections of race, ethnicity, and gender in US state legislatures Does women's representation in elected office lead to women-friendly policy? Analysis of state-level data Linking women's descriptive and substantive representation Brown, N. 2010 Dissertati on Reingold, B., & Smith, A.R. 2012 Journal In C. Hardy-Fanta (Ed.), Intersectionality and Politics: Recent research on gender, race, and political representation in the United States. New York, NY: Hayworth Press PhD dissertation. Rutgers, The State University of New Jersey American Journal of Political Science Caiazza, A. 2004 Journal Women & Politics Cowell-Meyers, K., & Langbein, L. 2008 Conferenc e Paper State The geography of gender power: Women in state legislatures Norrander, B., & Wilcox, C. 1998 Book Chapter State (bureaucracy) The responsiveness of state mental health authorities to parents with mental illness Biebel, K. Nicholson, J., Williams, V., & Hinden, B.R. 2004 Journal Presented at the annual meeting of the Midwestern Political Science Association. Chicago, IL. In S. Thomas & C. Wilcox (Eds.), Women in Elective Office: Past, Present, and Future. New York, NY: Oxford University Press Administration and Policy in Mental Health State State State State State State 284 Table 34 (cont’d) State (bureaucracy) State (education) State (education) State & local Local Local Local Local Local Local Political control and bureaucratic autonomy revisited: A multi-institutional analysis of OSHA enforcement The preparation of highly qualified teachers for English language learners: Educational responsiveness for unmet needs Legislative agenda setting for in-state resident tuition policies: Immigration, representation, and educational Access Racial differences in the employment policy of state and local governments: The case of male workers The county perspective on intergovernmental relations and government capacity Representation and local policy: Relating county-level public opinion to policy outputs The distributional Impact of greater responsiveness: Evidence from New York towns Policy responsiveness in local government: Adoption of smoke-free policies in U.S. counties Varied responsiveness to immigrant community growth among local governments: Evidence from the Richmond, Virginia, Metropolitan Area Local government-sponsored community development: Exploring relationships between perceptions of empowerment and community impact Kim, D. 2008 Journal Cadiero-Kaplan, K. Rodríguez, J.L. 2008 Journal McLendon, M.L., Mokher, C.G, & Flores, S.M. Singell, L.D. 2011 Journal American Journal of Education 1991 Journal Southern Economic Journal Streib, G., & Waugh, W.L. 1994 Journal International Journal of Public Administration Percival, G.L., Johnson, M., & Neiman, M. Sances, M.W. 2009 Journal Political Research Quarterly 2016 Journal The Journal of Politics Ybarra, V.D., & Krebs, T.B. 2016 Journal State & Local Government Review Rissler, G.E. 2016 Journal State & Local Government Review Glaser, M.A., Denhardt, K.G., & Grubbs, J.W. 1997 Journal American Review of Public Administration 285 Journal of Public Administration Research and Theory Equity & Excellence in Education Table 34 (cont’d) Local Local elected women and policymaking: Movement delegates or feminist trustees? Boles, J.L., & Carroll, S. 2001 Book Chapter Local (Bureaucracy) The temporal, social, and responsiveness dilemmas of public personnel administration Personnel rules and reform in an unreformed setting Citizen involvement efforts and bureaucratic responsiveness: Participatory values, stakeholder pressures, and administrative practicality Immigrant bureaucratic incorporation: The dual roles of professional missions and government policies Structural choices and representational biases: The post-election color of representation Representation, salience, and responsiveness to Latino student performance The politics of Latino education: The biases of at-large elections Klay, W.E. 1993 Journal Stein, L. 1994 Journal Yang, K., & Callahan, K. 2007 Journal Marrow, H.B. 2009 Journal American Sociological Review Meier, K.J., Juenke, E.G., Wrinkle, R. D., & Polinard, J.L. Hawes, D.P. 2005 Journal American Journal of Political Science 2013 Journal Journal of Latinos and Education Leal, D., MartinezEbers, V., & Meier, K.J. Dye, T.R. 2004 Journal Journal of Politics 1990 Journal Cato Journal Manza, J. 2002 Journal American Politics Research Hong-Hai, L. 2006 Journal Public Administration Review Minta, M.D. 2012 Journal Politics & Gender Local (bureaucracy) Local (bureaucracy) Local (bureaucracy) Local (education) Local (education) Local (education) Theoretical Theoretical Theoretical Theoretical The policy consequences of intergovernmental competition A democratic polity? Three views of policy responsiveness to public opinion in the United States Representative bureaucracy: Rethinking substantive effects and active representation Gender, race, ethnicity, and political representation in the United States 286 In S. Carroll (Ed.), The Impact of women in public office. Bloomington, IN: Indiana University Press International Journal of Public Administration Review of Public Personnel Administration Public Administration Review Table 34 (cont’d) Theoretical Theoretical Potentiality and representation: The link between descriptive representation and participation in the United States Preferable descriptive representation: Will just any woman, black or Latino do? "Polls", "Public opinion," and "Policymaking" literature Federal Assessing the Nation's Diet: Limitations of the Food Frequency Questionnaire Federal Don't blame the public for failed health care reform Federal Private polls and presidential policymaking: Reagan as a facilitator of change. Federal Variation in White House polling operations: Carter to Clinton Federal Lyndon Johnson and the public polls Federal Presidential decision making and opinion polls Federal Polling as a means toward presidential autonomy: Emil Hurja, Hadley Cantril, and the Roosevelt Administration Federal Counting the public in: Presidents, public opinion, and foreign policy Federal Public opinion and US foreign policy decisionmaking Federal The rise of presidential polling: The Nixon White House in historical perspective Uhlaner, C.J. 2012 Journal Politics & Gender Dovi, S. 2002 Journal American Political Science Review Briefel, R.R, & Flegal, K.M. Jacobs, L.R., & Shapiro, R.Y. Murray, K.S. 1992 Journal 1995 Journal 2006 Journal American Dietetic Association Journal of Health Politics, Policy and Law Public Opinion Quarterly Murray, K.S., & Howard, P. Altschuler, B. Beal, R.S., & Hinckley, R.H Eisinger, R.M., & Brown, J. 2002 Journal Public Opinion Quarterly 1986 1984 Journal Journal Public Opinion Quarterly Annals of the AAPSS 1998 Journal International Journal of Public Opinion Research Foyle, D.C. 1999 Book Graham, T.W. 1994 Book Chapter Jacobs, L.R., & Shapiro, R.Y. 1995 Journal New York, NY: Columbia University Press In D.A. Deese (Ed.), The new politics of American foreign policy. New York, NY: St. Martin's Public Opinion Quarterly 287 Table 34 (cont’d) Federal Politicians don't pander: Political manipulation and the loss of democratic responsiveness Going public: New strategies of presidential leadership. (3rd Ed.) Jacobs, L.R., & Shapiro, R.Y. 2000 Book Chicago, IL: University of Chicago Press Kernell, S. 1997 Book Federal Who leads and who follows/ US Presidents, public opinion, and foreign policy. Shapiro, R.Y., & Lawrence, R.J. 1996 Book Chapter Federal The pulse of the people: Franklin D. Roosevelt and the gauging of American public opinion Variation in White House polling operations: Carter to Clinton Health reform impasse: The politics of American ambivalence toward government Early American Perceptions of Korea and Washington's Korea Policy, 1882-1905 Amplifying public opinion: The policy impact of the US environmental movement Constituency opinion and Congressional policymaking: The Reagan defense buildup Democratic frustration in the Anglo American polities: A quantification of inconsistency between mass public opinion and public policy Changing public policy: The impact of public opinion, antiwar demonstrations, and war costs on Senate voting on Vietnam War motions, 1964-73 Steele, R.W. 1974 Journal Washington, DC: Congressional Quarterly Press In B. L. Naco, R.Y. Shapiro, & P. Isernai (Eds.) Decisionmaking in a glass house. Lanham, MD: Rowman & Littlefield Journal of Contemporary History Murray, K.S., & Howard, P. Jacobs, L.R. 2002 Journal Public Opinion Quarterly 1993 Journal Journal of Health Politics, Policy and Law Johnson, A.S. 2011 Journals Korea Journal Agnone, S.L. 2007 Journal Social Forces Bartels, L.M. 1991 Journal American Political Science Review Brooks, J.E. 2007 Journal Western Political Quarterly Burstein, P., & Freudenberg, W. 1978 Journal American Journal of Sociology Federal Federal Federal Federal Federal Federal Federal Federal 288 Table 34 (cont’d) Federal Federal Federal Federal Federal Federal Federal Federal Federal Federal Federal Presidential responsiveness and public policymaking Lumpers and splitters: The public opinion information that politicians collect and use Presidential responsiveness to public opinion Cohen, J.E. 1997 Book Druckman, J.N., & Lawrence, L.R. 2006 Journal Druckman, J.N., & Lawrence, L.R. 2009 Book Chapter Constituency opinion and Congressional behavior: A reexamination of the MillerStokes data Leading the public to war? The influence of American public opinion on the Bush Administration's decision to go to war in Iraq Public opinion and the common defense: Who governs military spending in the United States? The Congressman and his public opinion polls Numbered voices: How opinion polling has shaped American politics Reading public opinion: How political actors view the democratic process People, polls, and policymakers: American public opinion and national security The health of nations: Public opinion and the making of health policy in the US and Britain Erickson, R.S., 1978 Journal Foyle, D.C. 2004 Journal International Journal of Public Opinion Research Hartley, T., & Russett, B. 1992 Journal American Political Science Review Hawver, C. 1954 Journal Public Opinion Quarterly Herbst, S. 1992 Book Herbst, S. 1998 Book Hinckley, R.H. 1992 Book Chicago, IL: University of Chicago Press Chicago, IL: University of Chicago Press New York, NY: Lexington Books. Jacobs, L.R. 1993 Book 289 Ann Arbor, MI: University of Michigan Press Public Opinion Quarterly In G.C. Edwards & W.G. Howell (Eds.), Oxford Handbook of the American Presidency. Oxford, UK: Oxford University Press American Journal of Political Science Ithaca, NY: Cornell University Press Table 34 (cont’d) Federal Public opinion and President Clinton's first year: Leadership and responsiveness Jacobs, L.R., & Shapiro, R.Y. 1995 Book Chapter Federal Lyndon Johnson, Vietnam, and public opinion Paying attention to foreign affairs: How public opinion affects presidential decisionmaking Public opinion and gay rights: Do members of Congress follow their constituents' preferences? Jacobs, L.R., & Shapiro, R.Y. Knect, T. 1999 Journal 2010 Book Krimmel, K., Lax, J.R., & Phillips, J.H. 2011 Conferenc e Paper Public opinion and public policy, 19801993 Decision-making in a glass house: Mass media, public opinion, and American and European foreign policy in the 21st century Effects of public opinion on policy Monroe, A.D. 1998 Journal Nacos, B.L., Shapiro, R.Y., & Isernia, P. (Eds.) 2000 Book Boston, MA: Rowman and Littlefield Page, B.I., & Shapiro, R.Y. Quirk, P.J 1983 Journal 2009 Journal American Political Science Review The Forum Rottinghaus, B. 2007 Book Shapiro, Y. 1982 Shapiro, R.Y., & Jacobs, L.R. 2002 Dissertati on Book Chapter Federal Federal Federal Federal Federal Federal Federal Federal Federal Politicians do pander: Mass opinion, polarization, and law making The provisional pulpit: Modern Presidential leadership of public opinion The dynamics of public opinion and public policy Public opinion, foreign policy, and democracy: How Presidents use public opinion 290 In S.A. Renshon (Ed.), The Clinton Presidency: Campaigning, governing & the psychology of leadership. Boulder, CO: Westview Press. Presidential Studies Quarterly University Park, PA: Penn State University Press Paper presented at the annual meeting of the Midwest Political Science Association. Chicago, IL Public Opinion Quarterly College Station, TX: Texas A&M University Press Ph.D. Dissertation, University of Chicago In J. Manz, F.L. Cook, and B.I. Page (Eds), Navigating public opinion: Polls, policy, and the future of American democracy. Oxford, UK: Oxford University Pres. Table 34 (cont’d) Federal Foreign policy and public opinion Shapiro, R.Y., & Page, B.I. 1994 Book Chapter Federal The sometime connection: Public opinion and social policy. Sharp, E.B. 1999 Book Federal The impact of public opinion on US foreign policy since Vietnam: Constraining the colossus Public opinion and public policy: The case of Kennedy and civil rights Out of touch: The presidency and public opinion The governmental process: Political interests and public opinion. Polls, priming, and the politics of welfare reform Sobel, R.S. 2001 Book Stevens, D. 2002 Journal Towle, M.J. 2004 Book Truman, D.B. 1951 Book Weave, R.K. 2002 Book Chapter Patterns of representation: Dynamics of public preferences and policy The relationship between public opinion and public policy Wlezien, C. 1995 Journal Wlezien, C., & Soroka, S.N. 2007 Book Chapter The recoil effect: Public opinion and policy making in the United States and Britain Jacobs, L.R. 1992 Journal Federal Federal Federal Federal Federal Federal Federal (international comparison) 291 In D.E. Deese (Ed), The Politics of American Foreign Policy. New York, NY: St. Martin's Press. Albany, NY: State University of New York Press New York, NY: Oxford University Press Presidential Studies Quarterly College Station, TX: Texas A&M University Press New York, NY: Knopf In J. Manz, F.L. Cook, and B.I. Page (Eds), Navigating public opinion: Polls, policy, and the future of American democracy. Oxford, UK: Oxford University Press. Journal of Politics In R.J. Dalton, & H. Klingemann (Eds), The Oxford Handbook of Political Behavior. Oxford, UK: Oxford University Press Comparative Politics Table 34 (cont’d) Federal (international comparison) Federal & State State State State State State State State State State Local Theoretical Theoretical Public Opinion, Domestic Structure, and Foreign Policy in Liberal Democracies Risse-Kappen, T. 1991 Journal World Politics Public Opinion and American Democracy Public opinion and penal policymaking: An examination of constructions, assessments, and uses of public opinion by political actors in New York State Direct democracy and the link between public opinion and state abortion policy The relationship between public opinion and state policy: A new look based on some forgotten data Statehouse democracy: Public opinion and policy in the American states Shapiro, R.Y. Brown, E.K. 2011 2009 Journal Dissertati on Public Opinion Quarterly ProQuest Dissertations and Theses Arceneaux, K. 2002 Journal Erickson, R.S. 1976 Journal State Politics & Policy Quarterly American Journal of Political Science Erickson, R.S., Wright, G.C., & McIver, J.P. Lascher, E.L., Hagen, M., & Rochlin, S.A. Lax, J. & Phillips, J.H. Norrander, B. 1993 Book Cambridge, UK: Cambridge University Press 1996 Journal Journal of Politics 2009 Journal 2000 Journal American Political Science Review Political Research Quarterly Norrander, B., & Wilcox, C. 1999 Journal Policy Studies Journal Weber, R.E., & Shaffer, W.R. Watson, D.J., Juster, R.J., & Johnson, G.W. Key, V.O. Lasswell, H.D. 1972 Journal 1991 Journal Midwest Journal of Political Science Public Administration Review 1961 1941 Book Book Gun behind the door? Ballot initiatives, state policies, and public opinion Gay rights in the states: Public opinion and policy responsiveness The multi-layered impact of public opinion on capital punishment implementation in the American states Public opinion and policymaking in the states: The case of post-Roe abortion policy Public opinion and American state policymaking Institutionalized Use of Citizen Surveys in the Budgetary and Policy-Making Processes: A Small City Case Study Public opinion and American democracy Democracy through public opinion 292 New York, NY: Knopf Menasha, WI: Banta Table 34 (cont’d) Theoretical Theoretical Theoretical Theoretical A democratic polity? Three views of policy responsiveness to public opinion in the United States Public opinion and popular government Manza, J., Cook, F.L. 2002 Journal American Political Research Lowell, A.L. 1913 Book Polling, policy and public opinion: The case against heeding the voice of the people Public opinion and popular government Weissberg, R. 2002 Book New York, NY: Longmans, Green, and Co. New York, NY: Palgrave Macmillan Weissberg, R. 1976 Book 293 Englewood Cliffs, NJ: Prentice-Hall Appendix H Survey Methods Selecting a mode of data collection: Online survey To examine if state education policymakers are responsive to the policy preferences of both the general public and teachers, I carried out an online survey of state legislators serving on education committees and SBE members in the spring and fall 2016. The decision to conduct an online survey, as opposed to a mail or telephone survey was made as a result of both resource constraints and a review of research on survey techniques, as well as surveys with government officials. The primary benefits of online surveys are that they present both time and cost savings for the researcher (Clatyon & Werking, 1998; Dillman, 2000). Online surveys may also be more visually attractive (Fisher & Herrick, 2010) and provide respondents with the flexibility to complete the survey at their own convenience, which may increase likelihood of participation (Sax, Gilmartin & Bryant, 2003). Moreover, online surveys may reduce a respondent’s perceived costs simply because, unlike mail surveys, “there are no envelopes to open, no pencils to find, and no need to remember to put it back in the mail” (Fisher & Herrick, 2012, p. 149). Online surveys are also limited in a number of ways. For one, not everyone may have access to the Internet (Gjestland, 1996). I argue that this weakness does not impact my study because state policymakers are typically given e-mail addresses and provided with Internet access by their state government. Another limitation of online surveys is that respondents may have “concerns about confidentiality that discourage participation” (Sax, Gilmartin & Bryant, 2003, p. 410). To reduce such concerns, I included a statement explaining survey participant 294 confidentiality in both the e-mail inviting policymakers to participate in the survey as well as on the first page of the survey itself. Finally, technical troubles may arise when Internet browsers (e.g., Internet Explorer, Mozilla Firefox, Google Chrome) or mediums (e.g., laptop, desktop, iPad or notebook, cell phone) are not equal in terms of their capacity to view the survey. This may cause surveys to have different appearance from respondent to respondent (Dillman, 2000) or may result in complications that prevent some individuals from even accessing the survey (Smith, 1997). I am aware of one such technological issue because four state policymakers contacted me and identified an issue they had with a drag-and-drop feature on one survey question (see Figure 24). With some browsers, when the survey respondent dropped a second response into the single answer box, the second response covered the first and the respondents were unsure if both items were recorded. All four of these respondents were part of the control group. Therefore, I created an alternate survey for the four respondents in which the drag-and-drop question was reformatted as a checkbox question, as shown in Figure 25. 295 Figure 24 Most/Least important individual and groups question format - original Drawing on the lists above, please drag-and-drop to the appropriate boxes the two most important and two least important individuals or groups in informing your position on teacher evaluation policy. Items Two Most Important Individuals or Groups Constituents throughout your state Your constituents The governor Local leaders (e.g., mayor, city council) Other state officials (e.g., state board members, state education agency) Two Least Important Individuals or Groups Your political party leaders Statewide teacher labor organizations Local school teachers Statewide school leadership groups School leaders in your constituency Statewide business organization Local business groups Statewide education advocacy/special interest groups or leaders Local policy/advocacy groups Statewide religious organizations Statewide racial or ethnic-related organizations Education research organizations (e.g., think tanks, universities) Other______ Other _____ 296 Figure 25 Most/Least important individual and groups question format - modified Drawing on the lists above, please check the appropriate boxes for the two most important and two least important individuals or groups in informing your position on teacher evaluation policy. Constituents throughout your state Your constituents The governor Local leaders (e.g., mayor, city council) Other state officials (e.g., state board members, state education agency) Your political party leaders Statewide teacher labor organizations Local school teachers Statewide school leadership groups School leaders in your constituency Statewide business organizations Local business groups Statewide education advocacy/special interest groups or leaders Local policy/advocacy groups Statewide religious organizations Statewide racial or ethnic-related organizations Education research organizations (e.g., think tanks, universities) Other __________ Other __________ Two Most Important Individuals or Groups Selection 1 Selection 2 ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ Two Least Important Individuals or Groups Selection 1 Selection 2 ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ The distribution of an online survey can be difficult due to high levels of spam protection on state government e-mail servers. Moreover, because state policymakers likely receive many email messages each day, my survey invitation e-mail may have been easily lost or ignored (Hoonakker & Carayon, 2009; Alperin & Schultz, 2003). These issues are likely to affect 297 nonresponse rates and nonresponse bias. Therefore, when discussing my data, I strove to be as transparent as possible: I report the response rate as the total number of state education policymakers that submitted a survey response divided by the total number of e-mails sent; I also report the number of e-mails sent, the number of e-mails that were returned for failure to find the addressee because the address was either incorrect or no longer in service, as well as the number of auto-response messages. Finally, when describing my sample, I compare demographics of respondents and non-respondents and report any statistically significant differences in demographics among these two groups. Social desirability bias may also arise when collecting online survey data from politicians. Social desirability bias – that is, “the tendency to present oneself in a favorable light” (Groves et al., 2009, p.168) – may lead to state education policymakers over-reporting what they believe are socially approved behaviors and under-reporting what they believe are socially disapproved behaviors. Survey respondents may also repeatedly endorse items regardless of content (acquiescence), expend little effort in the interpretation and answering of questions (satisficing), avoid extreme response options, or exaggerate their answers (Caslyn & Winter, 1999; Grandy, 1998; Krosnick, 1999) Finally, caution must be taken in any type of mail or online survey because staff members, rather than state policymakers themselves, may fill out the surveys (Maestas, Neeley, & Richardson, 2003). However, as suggested by Herrera, Herrera, and Smith (1992), this should not be a serious problem in this study since staff persons “should know the [policymaker’s] thinking fairly well and probably share similar perspectives and level of knowledge about the issue” (p. 190). 298 Survey Design Survey question and answer design: Grammatical forms. In developing my survey questions and answers, I aimed to reduce the number and extent of seven comprehension problems that often present themselves in surveys: grammatical ambiguity, excessive complexity, faulty presuppositions, vague concepts, vague quantifiers, unfamiliar terms, and false inferences (Tourangeau, Rips, & Rasinski, 2000). One way in which I aimed to limit these problems is by conducting a field test of the survey with members of my dissertation committee who have experience conducting surveys of individuals in positions of power. I also conducted a field test of the survey in which a retired state legislator and two current legislative staffers completed the survey and provided feedback on question and answer wording, formatting, and order. As a result of these efforts, I modified one survey question to reduce a faulty presupposition problem related to a survey question that asked survey respondents to rate state and local actors in terms of their level of importance in informing policymakers’ positions on teacher evaluation policy. The first iteration of the questions read, “Thinking about your most recent vote on teacher evaluation policy, how important were each of the following in determining your action? Please rate each of the following on a scale from 0 (not important) to 5 (very important).” Upon further review of the question, two issues arose. First, the use of the word “recent” was problematic because state policymakers may interpret this term differently – some may define “recent” as within the past week, while others may define “recent” as within the past four years. Second, the use of the word “vote” assumed that both legislators and SBE members took a vote related to teacher evaluation policy. Given that the purpose of this question is to understand the actors that may play a role in state education policymakers’ decisionmaking 299 process, I chose to focus on a policymaker’s position on a policy issue, rather than an actual vote. Upon identifying these limitations, the question was modified to read, “Thinking about your position on teacher evaluation policy, how important were each of the following for informing your position? Please rate each of the following on a scale from 0 (not important) to 5 (very important).” In addition to considering various grammatical forms and clarifying the meaning of words in phrases in my survey, I also considered individual question and answer design for specific questions, as subsequently described. Survey question and answer design: Measuring responsiveness. In designing my questions and answer choices for the survey item that aimed to measure whether state education policymakers are responsive to public and teacher preferences about teacher evaluation policy, I chose to use a closed question with a forced answer. More specifically, I forced respondents to choose whether they would either vote in favor of or in opposition to a policy that would remove the required use of state assessment data in teacher evaluation and tenure decisions. In designing this question, I thought critically about issues that may arise with the use of a hypothetical question. The policy change that I asked policymakers to consider may not have been an active policy issue in some states. Asking respondents to imagine how they would vote on a specific issue if it were to be considered might require respondents to use “imaginative projection” and may require respondents to take additional time “to mull over the possibilities” (Converse & Presser, 1986, p. 23). Moreover, because such hypothetical questions provide no assurance that respondents will feel the full force of the political reality of their decision (Converse & Presser, 300 1986), it is possible that survey respondents were agnostic about their position if the policy decision was non-consequential. Nonetheless, because policies considered by state education policymakers vary from state to state and time to time, I chose to use a hypothetical question in an effort to standardize the question being considered across all policymakers (Converse & Presser, 1986). I chose to force respondents to select whether they would support or oppose a change in teacher evaluation policy. Although closed questions force people to choose among offered alternatives instead of answering in their own words, closed questions are more apt to communicate the same frame of reference to all respondents (Converse & Presser, 1986). The choice to present policymakers with a forced response, close-ended question was made because I wanted to emulate the political process in which policymakers have two options to vote on a piece of policy: “yay” or “nay.” There is not an “I don’t know” voting option. It is possible for a policymaker to abstain from voting; however, abstention is often discouraged because constituents may interpret a poor participation record as evidence of poor representation (Cohen & Noll, 1991). Finally, I included an optional, open-ended follow-up question to provide policymakers with the opportunity to offer any additional information about their answer choice. The purpose of this follow-up question was three-fold: to provided a space for respondents to pinpoint terms or words that were particularly troublesome, to identify respondents whose understanding of the question was limited, and as a qualitative aid to interpretation of respondent answers (Converse & Presser, 1986). 301 Survey question and answer design: Multiple measures. Relying on single question to make an inference about a large construct may make it difficult to uncover complexity (Converse & Presser, 1986). Therefore, I generated questions that would provide “multiple measures” of the issue at hand. In designing my questions and answer choices for three questions that aimed to understand the groups and individuals that these policymakers draw upon when making education policy decisions, I chose three different question formats that each aim to “cast light at different angles” (Converse & Presser, 1986, p. 47). The three questions that I developed did not include a “don’t know” option. In the 1980s, studies concluded that “don’t know” or “no opinion” options should be used to filter out respondents who don’t know much and thus can’t have an attitude (Converse & Presser, 1986). For example, if a survey respondent were to be asked a question in such a way as to suggest that they ought to have an opinion on the matter, they “may to wish to appear foolishly uniformed and may therefore give arbitrary answers” (Krosnick, 2002). Thus, it has been argued that offering a “don’t know” response option reduces pressure on the survey respondent (Krosnick et al., 2002) and allows the respondent to feel as though their beliefs were accurately represented (Andrews, 1984). However, research also found that “don’t know” responses are least likely to be used by people who have more formal education (Bishop, Oldenick, & Tuchfarber, 1980; Schuman & Presser, 1981), who know more about the topic in question (Converse, 1976; Faulkenberry & Mason, 1978; Rapoport, 1981, 1982), who are more interested in the topic (Krosnick & Milburn, 1990; Rapoport, 1982; Wright & Niemi, 1983), who are more exposed to information on the topic (Krosnick & Milburn, 1990; Wright & Niemi, 1983), who feel they have a greater ability to understand the topic (Krosnick & Milburn, 1990), and who feel others are interested in knowing their opinions on the topic (Francis & Busch, 1975; Krosnick & 302 Milburn, 1990). Moreover, research has found that survey respondents often utilize “don’t know” response options because they are seen as the “safe” answer (Dillman & Christian, 2005; Krosnick, 1999) or an “escape hatch” (Krosnick et al., 2002). However, if pushed to offer opinions, people will offer meaningful ones (Krosnick et al., 2002). Because state education policymakers are likely to be at least somewhat interested in education topics, be exposed to more information on education topics, and feel that others are interested in knowing their opinions on education issues, I chose to omit a “don’t know” option in my survey questions. Forced rating. The first question in this multiple measures approach was designed to force state education policymakers to rate the importance of different state- and local-level actors in informing their position on teacher evaluation policy. Survey respondents were asked to rate 11 state-level actors and eight local-level groups and actors on a scale from 1 (not important) to 5 (very important). The actors included in the survey (see Appendix J, Table 35) were those that research has shown can have an impact on the policymaking process When presenting survey respondents with a list such as the one presented for this question, one must be wary of priming the respondent to think about the importance of one group or individual when providing an answer about the importance of subsequent groups or individuals. Goodhue (1998) has argued that adjacent questions often propel respondents to use anchoring and adjustment in deciding upon their answer. For example, in my survey, a respondent may provide a response for the first actor independently, and then make minor adjustment in answering about the importance of the remaining actors. As a result, the error terms related to the level of importance for subsequent actors should be correlated, inflating the apparent reliability without increasing the actual reliability. To reduce this question order effect 303 in my survey, I randomized the order in which the actors were presented to respondents. The randomization of groups and individuals took place within level. The rationale for this withinlevel randomization is based on research by Davis and Venkatesh (1996), who contend that by constantly shifting from construct to construct, respondents suffer “output interference” or confusion and frustration, which impedes their ability to accurately retrieve all but the early constructs. Because I ask state education policymakers about the influence of both state and local actors, I chose to randomize groups and individuals within these two levels to reduce confusion or frustration from having to mentally shift from actors among different levels. Finally, respondents may talk to individuals or groups that they could not find on the precompiled lists of state and local groups and individuals (Harlacher, 2016; Pope, Boleman, & Cummings, 2005) Therefore, I presented respondents with two blank boxes in which they could input the name of an “Other” actor and rate their level of importance in informing their position on teacher evaluation policy. There are multiple limitations to this scale-rating question. First, the choice of the number of items in a rating scale is one that is debated among survey methodologists. Although research suggests that fewer scale points lose information, more scales tend to produce cognitive overload for the survey respondent (Graves et al., 2009). Yet others have found that “the number of available response categories does not materially affect the cognitive structure derived from the results” (p. 323), which seems to suggest that the number of response categories has little effect on the results obtains (Preston & Colman, 2000). Nonetheless, research that suggests that the human mind has a span of apprehension capable of distinguishing among seven different items plus or minus two (Miller, 1956). Moreover, Preston & Colman (2000) suggest that different scales may be best suited to different purposes. For example, when respondents are time limited, 304 shorter scales, which are likely to be perceived by respondents as relatively quick and easy to use, may be more appropriate to reduce respondent frustration and demotivation (Preston & Coleman, 2000). Drawing upon this research, as well as other studies that report higher reliabilities for five-point scales (Jenkins & Taber, 1977; Lissitz & Green, 1975; McKelvie, 1978; Remmers & Ewart, 1941), I chose to use a five-point scale. A second limitation of this scale-rating question is that respondents may shy away from the negative end of a scale, producing positivity bias; or, respondents may avoid the most extreme answer categories (Graves et al., 2009). Third, the consequences of the use of middle alternative (e.g., a 3 on a 5-point scale) answer option continues to be debated. On one hand, Moser and Kalton (1972) suggested that having a “middle alternative” presents “a risk in suggesting a noncommittal answer to the respondent” (p. 344). On the other hand, Rugg and Cantril (1944) argued that a middle alternative “provides for an additional graduation of opinion (p. 33). Given this disagreement, Converse and Presser (1986) suggest that surveys that include a question with a middle category be followed by a question with an intensity item. Therefore, I follow my rating item with two forced response intensity items. Discrete choice experiment. The first forced response intensity item presents state education policymakers with two actors and forces respondents to select the actor that is most likely to inform their teacher evaluation policy position. This type of discrete choice experiment (DCE) is frequently used in marketing and health care. DCEs present survey respondents with samples of hypothetical scenarios (choice sets) drawn a priori from all possible choice sets (Amaya-Amaya, Gerard, & Ryan, 2008; Louviere, 1991, 2001). In my survey, respondents were asked, “Thinking about your position on statewide teacher evaluation policies, consider the 305 following sets of two viewpoints. When the positions of these two groups/individuals differ, whose is more likely to inform your position?” Respondents were presented with 28 choice sets, The choice sets control for order effects and context effects: the order in which the survey respondent viewed the paired items was randomized; and, each individual item was paired with every other item an equal number of times. The items that were included were: business organizations, education advocacy groups, education research organizations, personal beliefs, school leadership groups, teacher labor unions, the governor, and your constituents. Figure 26 provides an abbreviated sample of the question format. Figure 26 Discrete choice experiment question format School leadership groups Education advocacy groups Your constituents Personal beliefs Business organizations The group/individual that is most likely to inform my position is: ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ ☐ Teacher labor unions School leadership groups The governor Education research organizations The governor Least and most important actors. The second forced response intensity item asked state education policymakers to select the two most important and two least important actors in informing their position on teacher evaluation policy. Because policymaking doesn’t take place in a vacuum and more than one group can influence a policymakers’ decision at once, this question allowed respondents to discriminate between the actors of greatest and least importance in informing their position on teacher evaluation policy (Flynn & Marley, 2013). 306 Appendix I Recruitment E-mail Message Dear [Salutation FirstName LastName], My name is Rachel White and I am a doctoral candidate of education policy at Michigan State University. I am conducting research on how state policymakers – namely, state legislators and state board of education members – make education policy decisions. With the recent passage of the Every Student Succeeds Act (ESSA), we know that state policymakers like yourself will play a pivotal role in the decisionmaking processes around education issues important to your state. However, as researchers, we often try to make inferences about politics and the policymaking process without hearing from experts like you. Thus, this survey is designed to try to learn more about how state policymakers make education policy decisions. This will greatly contribute to an improved understanding of the importance of state education governance structures in the education policymaking process. To take the confidential, 5-minute survey, please click the link below. Take the Survey [Survey hyperlink] Or copy and paste the URL below into your internet browser: [Long link] All of your responses (as well as your decision to participate in this study) will be confidential so that only the principal investigator and co-principal investigator will have access to the individual data. Moreover, each participant is linked to a unique ID and, thus, no personally identifiable information is linked to the data. Participation in this study is completely voluntary. You are free to decline to participate, to end participation at any time, or to refuse to answer any individual question without penalty. If you have further questions, you may contact me at whitera3@msu.edu or (810) 569-2682. If you would like to talk with someone else about any dimension of the research, you may contact the Michigan State University Human Subjects Committee, 207 Olds Hall, Michigan State University, East Lansing, MI 48824 or irb@msu.edu. I genuinely respect and have great appreciation for the work that you are doing to improve the educational opportunities for children in your state. I look forward to your participation in the survey. If you wish to talk further about the survey or would be interested in taking part in a follow-up interview, please do not hesitate to contact me (whitera3@msu.edu). Thank you for your time and consideration. All the best, Rachel S. White 307 Appendix J Methods for Analyzing Forced Rating Data I chose to use nonparametric analytic methods to analyze the rating data for two reasons. First, there has been much debate in the methodological literature around whether Likert-type question responses can be quantitatively analyzed. For one, responses to Likert-type questions tend to be skewed due to extreme response style (ERS) and acquiescence response style (ARS) of the respondents. ERS is the tendency to choose only the extreme endpoints of a Likert scale (Hurley, 1998) and ARS is tendency to provide positive answers, rather than negative answers, with items regardless of item content (Van Herk, Poortinga, & Verhallen, 2004). Moreover, when compared to full-labeling of response categories, labeling only the ends of the rating scale – which I utilized in this survey question – has been shown to increase the likelihood of ERS (Moors, Kieruj & Vermunt, 2014). These patterns of skewness are present in the responses to some of the Likert-type questions in this study. Shapiro-Wilks test44 results (Table 35) indicate that the distribution of the ratings of the following groups are non-normal: statewide constituents, other state officials, political party leaders, statewide school leadership groups, statewide religious organizations, your local constituents, local school teachers, and school leaders in your constituency. 44 The Shapiro-Wilks test null hypothesis is that the variable follows a normal distribution. Rejection of the null hypothesis indicates that the data is not normally distributed. 308 Table 35 Shapiro-Wilks tests for normal distribution of state- and local-actor rating data State Actors Local Actors Constituents throughout your state The governor Other state officials (e.g., state education agency staff) Your political party leaders Statewide teacher labor unions Statewide school leadership groups Statewide business organizations Statewide education advocacy/special interest groups Statewide religious organizations Statewide racial/ethnic groups Education research organizations (e.g., universities, think tanks) Your constituents Local government leaders (e.g., mayor, city council) Local school teachers Constituents that called or wrote you Local teacher labor unions School leaders in your constituency Local business groups (e.g., chamber of commerce) Local education policy or advocacy groups 309 n 167 160 170 164 168 171 167 168 165 169 168 155 164 164 168 162 166 168 165 W 0.97 1.00 0.97 V 3.30 0.43 4.26 Z 2.72 -1.90 3.31 p-value <0.01 0.97 0.97 0.99 0.04 1.00 1.00 4.31 0.92 7.72 0.31 0.32 3.33 -0.20 4.66 -2.66 -2.63 <0.01 0.58 <0.01 1.00 0.95 0.99 0.01 6.69 1.00 0.39 4.33 -0.01 -2.18 <0.01 0.50 0.97 0.002 4.15 0.99 3.23 -0.01 <0.01 0.02 0.99 1.00 0.03 1.00 10.45 1.64 0.51 9.11 0.23 5.35 1.13 -1.55 5.04 -3.39 <0.01 0.13 0.94 <0.01 1.00 0.11 -5.04 1.00 <0.01 1.00 0.99 0.51 1.00 Second, methodologists have debated whether the data obtained from responses to Likerttype questions should be considered interval or ordinal, with most agreeing that responses from Likert-type questions are at least ordinal (Goldstein & Hersen, 1984). In the case of this research, there is indeed a clear ordering of the variables: a rating of one indicates least important and a rating of five indicates most important. Whether responses from Likert-type questions are interval, however, depends upon whether one believes that the difference between a rating of a one and a two is equal to the difference between a rating of a two and a three, a three and a four, and a four and a five. In analyzing the responses to the Likert-type questions, I do not make this assumption of equal differences between ratings; rather, I treat the response data as ordinal. Nonetheless, additional consideration must be given to the ordinal data because “Likert scaling presumes the existence of an underlying (or latent or natural) continuous variable whose value characterizes the respondents’ attitudes or opinions” (Clason & Dormody, 1994, p. 31). As such, the discrete nature of the responses cannot be ignored. For example, suppose I want to compare the ratings of teachers unions among legislators and SBE members. I observe a means of 3.1 and 2.9, respectively. According to a t-test, the ratings of education research organizations among these two populations are not significantly different (t(165) = 1.08, p=0.28). However, this test ignores the discrete nature of response options in the Likert-type question. This issue is evident upon examining the data as both counts and percentages among the five response categories: among legislators, 10 (14 percent) rated teachers unions with a value of one, 13 (18 percent) with a value of two, 15 (21 percent) with a value of three, 25 (35 percent) with a value of four, and 9 (13 percent) with a value of five; among SBE members, 20 (21 percent) rated teachers unions with a value of one, 17 (18 percent) with a value of two, 24 (25 percent) with a value of three, 19 310 (20 percent) with a value of four, and 15 (16 percent), with a value of five. The mean scores in both groups are quite similar; however, the distribution of the scores within each group differs. 311 Appendix K Descriptive Statistics of Rating Question, by Office and Political Party Table 36 Descriptive statistics of state education policymakers’ ratings of level of importance of state- and local-level actors in informing their position on teacher evaluation policy, by office and political party n State actors Constituents throughout your state The governor Other state officials Your political party leaders Statewide teacher labor unions Statewide school leadership groups Statewide business organizations Statewide education advocacy/speci al interest groups Office State legislators Median Mean St.Dev. n SBE members Median Mean St.Dev. n Political Party Republicans Democrats Median Mean St.Dev n Median Mean St.Dev 72 4.0 3.7 1.1 95 4.0 3.8 1.0 78 4.0 3.7 1.1 79 4.0 3.7 1.0 66 3.0 2.5 1.1 94 3.0 3.0 1.3 75 3.0 2.8 1.2 76 3.0 2.8 1.2 72 4.0 3.5 1.0 97 4.0 4.0 1.0 80 4.0 3.8 1.1 79 4.0 3.8 1.0 69 2.0 2.3 1.1 95 2.0 2.1 1.1 76 2.0 2.1 1.1 78 2.0 2.2 1.1 72 3.0 3.1 1.3 95 3.0 2.9 1.4 79 2.0 2.4 1.2 78 4.0 3.7 1.1 73 4.0 4.0 0.9 97 4.0 3.9 1.1 80 4.0 3.8 1.1 80 4.0 4.1 0.9 71 3.0 3.0 0.9 96 3.0 3.0 1.2 79 3.0 2.9 1.1 78 3.0 3.0 1.1 71 3.0 3.3 1.1 96 3.0 3.2 1.1 79 3.0 3.0 1.1 78 4.0 3.5 1.1 312 Table 36 (cont’d) Statewide religious organizations Statewide racial/ethnic groups Education research organizations Your constituents Local government leaders Local school teachers Constituents that called or wrote you Local teacher labor unions School leaders in your constituency Local business groups Local education policy or advocacy groups 73 2.0 1.9 0.9 92 2.0 1.9 1.1 77 2.0 2.0 1.0 78 1.0 1.8 1.0 73 3.0 2.7 1.2 96 3.0 3.0 1.3 80 2.0 2.3 1.2 79 3.0 3.3 1.0 72 4.0 3.6 0.9 95 4.0 3.9 0.9 78 3.0 3.5 0.8 79 4.0 3.9 0.9 69 4.0 4.0 1.0 86 4.0 3.7 1.1 74 4.0 3.8 1.1 72 4.0 3.9 1.0 72 3.0 2.8 1.0 91 3.0 2.6 1.1 77 3.0 2.4 1.1 77 3.0 2.9 1.0 71 4.0 4.3 1.0 93 4.0 4.0 1.0 76 4.0 3.9 1.0 78 5.0 4.4 1.0 72 4.0 3.8 1.1 95 3.0 3.3 1.0 79 4.0 3.4 1.1 78 4.0 3.6 1.0 67 3.0 3.1 1.2 94 3.0 2.7 1.3 75 2.0 2.7 1.1 76 4.0 3.6 1.1 72 4.0 4.2 0.9 93 4.0 4.0 1.0 77 4.0 3.9 1.0 78 4.0 4.2 1.0 73 3.0 3.1 1.1 95 3.0 3.0 1.1 79 3.0 2.9 1.1 80 3.0 3.1 1.1 70 3.0 3.3 0.9 95 3.0 3.0 1.1 77 3.0 2.8 1.0 78 3.5 3.5 0.9 313 Appendix L Descriptive Statistics and Cross Tabulations of Most and Least Important Actor Selections Table 37 Descriptive statistics of most/least important actor nominations by state education policymakers # Most Important Nominations 3 68 2 20 43 39 43 42 34 9 1 3 8 1 9 2 2 Statewide religious groups Local school teachers Political party leaders Statewide teacher unions Statewide school leadership groups Other state officials Local school leaders Your constituents Education research organizations Governor Statewide racial/ethnic groups Statewide business organizations Statewide education advocacy/interest groups Local government leaders Constituents throughout your state Your local business groups Local education advocacy groups 314 # Least Important Nominations 71 4 69 39 2 5 1 1 9 24 29 24 13 19 3 8 6 # Total Nominations 74 72 71 59 45 44 44 43 43 33 30 27 21 20 12 10 8 Table 38 Descriptive statistics of clustered most/least important actor nominations by state education policymakers Sub-group Statewide teacher labor organizations Local school teachers Statewide school leadership groups School leaders in your constituency Political party leaders The governor Local government leaders Other state officials Statewide racial/ethnic groups Statewide business organizations Local business groups Statewide religious organizations Education research organizations Statewide education advocacy/interest groups Local education advocacy groups Your constituents Constituents throughout your state Cluster Most Important Least Important Total Nominations Educators and their representatives 126 43 169 Other elected officials 45 102 147 Other interest groups 8 111 119 Education research and advocacy groups 40 27 67 Constituents 49 4 53 315 Table 39 Cross-tabulations of nominated actors chosen by state education policymakers as least likely to inform their position on teacher evaluation policy Constituents Constituents Other elected officials Educators and their representatives Education research and advocacy groups Other interest groups Total 0 0 0 0 0 0 Other elected officials 2 0 0 0 0 2 316 Educators and their representatives 0 20 1 0 0 21 Education research and advocacy groups 0 5 6 1 14 26 Other interest groups 2 59 14 0 1 76 Total 4 84 21 1 15 125 Table 40 Cross-tabulations of nominated actors chosen by state education policymakers as most likely to inform their position on teacher evaluation policy Constituents Constituents Other elected officials Educators and their representatives Education research and advocacy groups Other interest groups Total 0 0 0 0 0 0 Other elected officials 9 0 0 0 0 9 317 Educators and their representatives 32 22 0 0 0 54 Education research and advocacy groups 4 6 24 0 1 35 Other interest groups 1 1 3 0 3 5 Total 46 29 27 0 1 103 Table 41 Cross-tabulations of nominated actors chosen by state education policymakers as most and least likely to inform their position on teacher evaluation policy Constituents Other elected officials Least Educators and their representatives Likely Education research and advocacy groups Other interest groups Total Most Likely Constituents Other Educators and Education elected their research officials representatives and advocacy groups 0 25 19 13 1 25 10 13 4 81 30 20 2 30 11 0 1 3 3 1 8 164 73 47 318 Other interest groups Total 31 37 90 30 3 191 88 86 225 73 11 483 Appendix M Teacher Evaluation Policymaking Authority Devolution Index Methods and Data Table 42 Identifying relevant teacher evaluation policy elements Teacher Evaluation Policymaking Authority Devolution Index Element Teacher Evaluation Policy Elements Identified in Literature Frequency of evaluation Evaluator feedback/conferences Classroom observations Process Elements Frequency and length of observations Standards and evaluation of professional practice Evaluation of student achievement and/or growth Rating categories Literature and Research to Support Policy Element Inclusion Herlihy et al., 2014; Hull, 2013; Marzano & Toth, 2013 Barker, 2014; Cowen & Fowles, 2013; Herlihy et al., 2014; Hull, 2013; Darling-Hammond, 2012; Goe, Biggers & Croft, 2012; Danielson, 2010 Amrein-Beardsley & Collins, 2012; Barker, 2014; Bill & Melinda Gates Foundation, 2010; Collins & Amrein-Beardsley, 2014; Cowen & Fowles, 2013; Darling-Hammond, 2012; Goe, Biggers & Croft, 2012; Harrison & CohenVogel, 2012; Herlihy et al., 2014; Hull, 2013; Lavigne, 2014; Rockoff & Speroni, 2010; Sartain, Stoelinga & Brown, 2011; Taylor & Tyler, 2011 Herlihy et al., 2014; Hull, 2013; Marzano & Toth, 2013 Amrein-Beardsley & Collins, 2012; Barker, 2014; Darling-Hammond, 2012; Goe, Biggers & Croft, 2012; Herlihy et al., 2014 Amrein-Beardsley & Collins, 2012; Barker, 2014; Berliner, 2014; Collins & Amrein-Beardsley, 2014; Gabriel & Lester, 2013; Harrison & Cohen-Vogel, 2012; Herlihy et al., 2014; Konstantopoulous, 2014; Lavigne, 2014; Rockoff & Speroni, 2010; Welner, 2013; Harrison & Cohen-Vogel, 2012; Herlihy et al., 2014; Lavigne, 2014 319 Table 42 (cont’d) Student learning, achievement and growth data (as measured by state and/or local assessments) Amrein-Beardsley & Collins, 2012; Barker, 2014; Berliner, 2014; Bill & Melinda Gates Foundation, 2010; Burnett, Cushing & Bivona, 2012; Collins & Amrein-Beardsley, 2014; Darling-Hammond, 2012; Darling-Hammond, Amrein-Beardsley, Haertel & Rothstein, 2012; Gabriel & Lester, 2013; Glazerman, Loeb, Goldhaber, Staiger, Raudenbush & Whitehurst, 2010; Goe, Biggers & Croft, 2012; Harrison & Cohen-Vogel, 2012; Herlihy et al., 2014; Hull, 2013; Konstantopoulous, 2014; Lavigne, 2014; Marzano & Toth, 2013; Data & Rockoff & Speroni, 2010; Welner, 2013 (variation in state and/or local Measurement assessment identified in iterative coding process) Elements Methods of measuring Amrein-Beardsley & Collins, 2012; Berliner, 2014; Collins & Amreinstudent achievement (e.g. Beardsley, 2014; Darling-Hammond, 2012; Gabriel & Lester, 2013; Harrison & value added, student Cohen-Vogel, 2012; Herlihy et al., 2014; Hull, 2013; Konstantopoulous, 2014; growth) Lavigne, 2014; Rockoff & Speroni, 2010; Professional practice Amrein-Beardsley & Collins, 2012; Barker, 2014; Herlihy et al., 2014; Rockoff observation rubric/tool & Speroni, 2010 data Multiple measures of Amrein-Beardsley & Collins, 2012; Barker, 2014; Berliner, 2014; Collins & effectiveness in Amrein-Beardsley, 2014; Darling-Hammond, 2012; Gabriel & Lester, 2013; summative evaluation Goe, Biggers & Croft, 2012; Harrison & Cohen-Vogel, 2012; Herlihy et al., (with various weightings) 2014; Hull, 2013; Konstantopoulous, 2014; Lavigne, 2014; Marzano & Toth, 2013 (various types of weightings identified in iterative coding process) Weighting of Amrein-Beardsley & Collins, 2012; Lavigne, 2014 Weighting Elements components within professional practice rubric Weighting of various Collins & Amrein-Beardsley, 2014; Gabriel & Lester, 2013; Harrison & assessment within Cohen-Vogel, 2012; Lavigne, 2014 student growth/learning component 320 Table 42 (cont’d) State oversight of local implementation School district posting of State Oversight teacher evaluation Elements framework and information to public website State State approval/ assurance Approval/Assurance required flexibility from Elements state model. Herlihy et al., 2014 Identified in iterative coding process Herlihy et al., 2014 321 Table 43 Authority devolution index measures, overall and by element, by state State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Process Elements S 0 3 1 4 1 4 2 1 3 5 0 2 4 3 0 3 3 3 4 1 2 5 1 2 0 A 0 2 1 0 0 3 5 2 0 5 5 3 7 2 1 1 0 1 5 4 3 0 0 0 2 T 4 5 2 4 1 4 2 3 3 0 0 3 7 3 1 4 3 4 7 4 4 5 1 5 2 Data & Measurement Elements S A T 0 0 1 0 0 0 0 0 0 2 0 2 0 0 0 0 2 0 0 2 0 0 0 0 2 0 2 2 3 0 0 4 0 0 2 2 1 3 3 2 0 2 0 0 0 0 1 1 0 0 0 1 1 2 0 3 3 0 1 1 0 4 4 1 0 1 0 0 0 0 0 1 0 0 0 Weighting Elements S 0 0 0 0 0 1.5 0 0 1.5 1.5 0 0 0.5 0.5 0 0 1.5 1.5 0.5 0.5 0 0.5 0 0 0 A 0 1.5 0 0 0 0 0 0 0 1.5 0 1.5 1.5 0 0 1 0 1.5 1.5 0 0 0 0 0 0.5 T 0 1.5 0 0 0 1.5 0 0 1.5 0 0 1.5 2 0.5 0 1 1.5 1.5 1.5 0.5 0 0.5 0 1 0.5 322 State Oversight of Implementation S A T 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1 0 1 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0.5 0.5 0 0 0 0 0 0 0 0 0 1 0 1 0.5 0 0.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 State Approval/ Assurance of Implementation S A T 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 1 0 1 0 1 1 0 0 0 0.5 0 0.5 0 0 0 0 0 0 0 1 1 Index S 0 3 2 6 1 6.5 3 1 8.5 8.5 1 2 5.5 5.5 0 3 6.5 6 5.5 1.5 2 7 1 2 0 A 0 3.5 0 0 0 0 0 2 0 9.5 11.5 7.5 12 0 1 4 0 3.5 9.5 6 7 0 0 0 3.5 T 5 6.5 2 6 1 6.5 3 3 8.5 11.5 12.5 7.5 12.5 5.5 1 7 6.5 8 12.5 6.5 8 7 1 7 3.5 Table 43 (cont’d) Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York 0 2 5 0 1 0 2 0 1 2 6 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 1.5 0 0 0 3.5 0 0 0 3.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 1 1 0 0 2 7.5 0 1 2 6.5 0 1 4 10.5 0 3 0 5 7 4 5 8 4 6 1 0 2 4 1 3 4 1 3 0 0 3.5 1 2.5 3.5 1 2.5 3.5 0 0 0 0 0 1 0 0 1 1 0 1 0 1 1 1 1 1 5 0 11.5 14 8.5 14.5 North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia 1 1 4 4 2 3 0 0 1 0 2 1 0 3 5 3 7 0 0 0 3 0 4 0 0 0 4 3 0 0 0 6 1 1 4 4 5 3 5 5 1 0 5 4 1 3 5 7 0 0 2 1 0 4 0 0 0 0 0 0 0 1 1 2 0 0 0 0 0 0 1 0 3 0 0 1 0 0 0 2 0 0 2 1 0 4 3 3 3 0 0 1 1 1 1 2 0 0 1.5 2.5 0 3.5 0 0 0 0 0 0 0 0 0.5 1.5 0 0 0 0 0.5 0 0 0 0 0 0 0 0 0 0 1.5 0 0 1.5 2.5 0.5 3.5 3.5 1.5 0 0 0 0 0 0.5 0.5 1.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 1 1 0 0 0 1 0 0 1 1 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0.5 0 1 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 1 1 1 0 0 1 0 1 0 1 0 0 0 0 0.5 0 1 1 8.5 7.5 2 11.5 0 0 1 0 2 1 0 4 8 6.5 Wisconsin Wyoming 0 0 3 1 3 1 0 0 1 0 1 0 0 0 1.5 0 1.5 0 0 0 0 0 0 0 0 1 1 1 1 1 0 1 12 8.5 13. 5 1 0 0 0 3.5 0 5 1 3 1 4 5 0 0 0 10. 5 6.5 2 323 2 2 8.5 7.5 5.5 11.5 12.5 11.5 4 1 5 6 2 4.5 8 11.5 6.5 2 Table 44 Data descriptions and sources Variable Type Variable Specifications N=50 !=0.18; !=0.12; min=0.23; max=0.63 Legislative professionalism Continuous State Government Political Ideology Continuous N=50 !=41.79; !=31.59; min=0; max=91.45 Chief State School Officers/State Superintendent Electoral Structure Dichotomous Appointed=0 Elected=1 N=50 Appointed=38 Elected=12 Source/Information A measure of legislative professionalism for legislatures that incorporates professionalism components of salary/benefits, time demands of service, and staff and resources. Source: Squire, P. (2007). Measuring state legislative professionalism: The Squire Index revisited. State Politics and Policy Quarterly, 7, 211-227. A measure of the ideology of a state’s political leaders which utilizes the roll call voting scores of state congressional delegations, the outcome of congressional elections, the partisan division of state legislatures, the party of the governor, and various assumptions regarding voters and state political elites. The “NOMINATE” state government ideology scores, which rely on “Common-Space” congressional ideology scores to construct the measure of state party ideology (available at http://voteview.com/basic.htm) are utilized, as opposed to the “ADA/COPE” measures, which rely on the Americans for Democratic Action and the AFL-CIO Committee on Political Education (COPE) data regarding state legislators voting records. Source: Berry et al. (2010). Measuring citizen and government ideology in the American states: A re-appraisal. State Politics and Policy Quarterly, 10, 117-135. Available at https://rcfording.wordpress.com/state-ideology-data/ A binary indicator of whether a state’s chief state school officer is popularly elected or appointed. Source: http://www.nasbe.org/wpcontent/uploads/State-Education-Governance-2013-State-by-StateMatrix1.pdf 324 Table 44 (cont’d) State Board of Education Electoral Structure Average school district size Dichotomous Appointed=0 Elected=1 N=50 Appointed=37 Elected=11 Continuous (natural log) State share of school revenue (2011) Continuous (percentage) N=50 !=10,237 !=28,923; min=344; max=201,374 N=50 != .47; !=0.12; min=0.29; max=0.83 Teacher Union Strength Categorical (1 = strongest to 5=weakest) N=50 !=2.98; !=1.42; min=1; max=5 A binary indicator of whether a state’s board of education is popularly elected or appointed. Source: http://www.nasbe.org/wpcontent/uploads/State-Education-Governance-2013-State-by-StateMatrix1.pdf An indicator of the average size of a school district in the state, calculated by taking the log of the total number of students in the state divided by the total number of school districts in the state. Source: http://www.gbnx.com/district.php. An indicator of the extent to which funding for K12 education is centralize using data on the state share of school district revenue in 2011. Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data (CCD), "National Public Education Financial Survey," 1989-90 through 2011-12. A measure of teacher unions’ strength according to the power and influence of their state-level unions. Source: Northern, A.N., Scull, J. & Zeehandelaar, D. (2012). How strong are U.S. teacher unions? A state-by-state comparison. Dayton, OH: Thomas B. Fordham Institute. 325 REFERENCES 326 REFERENCES Aaldering, L. (2016). Political representation and educational attainment: Evidence from the Netherlands (1994-2010). Political Studies. Doi: 10.1177/0032321715622788. Abney, F.G. (1974). Factors related to negro turnout in Mississippi. Journal of Politics, 36(40, 1057-1063. Abney, F.G., & Hutcheson, J.D. (1981). Race, representation, and trust: Changes in attitudes after the election of a black mayor. Public Opinion Quarterly, 45(1), 91-101. Abrams, B.A. (1981). Political power and the market for governors. Public Choice, 37(3), 521529. Adrian, C.R. (1952). Some general characteristics of nonpartisan elections. American Political Science Review, 46(3), 766-776. Aldrich, J. (1995). Why parties? Chicago, IL: University of Chicago Press. Alesina, A. & Tabellini, G. (2007). Bureaucrats or politicians? Part I: A single policy task. The American Economic Review, 97(1), 169-179. Allgood, I. (2016). Faith and freedom of religion in U.S. public schools: Issues and challenged facing teachers. Religious Education, 111(3), 270-287. Alperin, D.J., & Schultz, D. (2003). E-democracy: Legislative-constituent communications in Minnesota and Wisconsin. Paper presented at the Annual Meeting of the American Political Science Association, Philadelphia, PA. Amaya-Amaya, M., Gerard, K., & Ryan, M. (2008). Discrete choice experiments in a nutshell. In M. Ryan, K. Gerard, & M. Amaya-Amaya (Eds.), Using discrete choice experiments to value health and health care (pp.13-46). Dordrecht, Netherlands: Springer. Anderson, E., Tremper, C., Thomas, S., & Wagenaar, A.C. (2012). Measuring statutory law and regulations for empirical research. PHLR Methods Monograph Series. Andrews, F.M. (1984). Construct validity and error components of survey measures: A structural modeling approach. Public Opinion Quarterly, 48(2), 409-442. Arkin, J.H. (1999). A content analysis of state legislative responses to educator assault. Doctoral Thesis, The College of William and Mary, Williamsburg, VA. Arnold, D. (1990). The logic of Congressional action. New Haven, CT: Yale University Press. 327 Arum, R., & Preiss, D. (2009). Law and disorder in the classroom. Education Next, 9(4). Retrieved from http://educationnext.org/law-and-disorder-in-the-classroom/ Baker, B.D., Oluwole, J., & Green, P.C. (2013). The legal consequences of mandating high stakes decisions based on low quality information: Teacher evaluation in the Race-to-theTop era. Education Policy Analysis Archives, 21(5). Baker, E.L., Barton, P.E., Darling-Hammond, L., Haertel, E., Ladd, H.F., Linn, R.L., …, Shepard, L.A. (2010). Problems with the use of student test scores to evaluate teachers. Washington, D.C.: Economic Policy Institute. Baker, T.J., & Hedge, D.M. (2013). Term limits and legislative-executive conflict in American states. Legislative Studies Quarterly, 38, 236-258. Banducci, S.A., Donovan, T., & Karp, J.A. (2004). Minority representation, empowerment, and participation. The Journal of Politics, 66(2), 534-556. Bardach, E. (1977). The implementation game: What happens after a bill becomes a law. Cambridge, MA: MIT Press. Bardhan, P., & Mookherjee, D. (2005). Decentralizing antipoverty program delivery in developing countries. Journal of Public Economics, 89(4), 675-704. Barkin, D., & Hettich, W. (1968). The Elementary and Secondary Education Act: A distributional analysis. Saint Louis, MO: Institution for Urban and Regional Studies, Washington University. Barreto, M.A. (2007). Si se puede! Latino candidates and the mobilization of Latino voters. American Political Science Review, 101(3), 425-441. Barreto, M.A., Seguara, G.M., & Woods, N.D. (2004). The mobilizing effect of majorityminority districts on Latino turnout. American Political Science Review, 98(1), 65-75. Barrett, E. (1995). The policy priorities of African-American women in state legislatures. Legislative Studies Quarterly, 20, 223-247. Barro, R. (1973). The control of politicians: An economic model. Public Choice, 14, 19-42. Bartolomé, L.I. (Ed.). (2008). Ideologies in education: Unmasking the trap of teacher neutrality. New York, NY: Peter Lang Publishing. Baumgartner, F., & Jones, B.D. (1991). Agenda dynamics and policy subsystems. The Journal of Politics, 53(4), 1044-1074. Beasley, T., & Coate, S. (2003). Elected versus appointed regulators: Theory and evidence. Journal of the European Economic Association, 1(5), 1176-1206. 328 Becker, J., & Halderman, R.S. (2004, January 24). Virginia delegates denounce Bush education law. The Washington Post, pA1, A4. Berke, J. (1974). Answers to inequity: An analysis of the new school finance. Berkeley, CA: McCutchan Publishing. Berke, J., & Kirst, M. (1972). Federal aid to education. Lexington, MA: Heath. Berkman, M.B., & O’Connor, R. (1993). Do women legislators matter? Female legislators and state abortion policy. American Politics Quarterly, 21(1), 102-124. Berman, P., & McLaughlin, M.W. (1978). Federal programs supporting educational change, Vol. VIII: Implementing and sustaining innovations. Santa Monica, CA: Rand. Berry, J.M. (1997). The interest group society (3rd ed). New York, NY: Longman. Berry, W. (1979). Utility regulation in the states: The policy effects of professionalism and salience to the consumer. American Journal of Political Science, 23, 263-277. Berry, W.D., Rinquist, E.J. Fording, R.C., Hanson R.L., & Klarner, C.E. (2010). Measuring citizen and government ideology in the American states: A re-appraisal. State Politics and Policy Quarterly, 10, 117-135. Biebel, K., Nicholson, J., Williams, V., & Hinden, B.R. (2004). The responsiveness of state mental health authorities to parents with mental illness. Administration and Policy in Mental Health, 32(1), 31-48. Bigelow, N.S. (2006). Representation in state legislatures: Searching for responsiveness in an age of polarization. (Unpublished doctoral dissertation). University of Maryland, College Park, MA. Bill & Melinda Gates Foundation. (2010). Learning about teaching: Initial findings from the Measures of Effective Teaching (MET) Project. Seattle, WA: Author. Bill & Melinda Gates Foundation. (2013). Ensuring fair and reliable measures of effective teaching. (MET Project Policy and Practice Brief). Seattle, WA: Author. Bish, R.L. (1971). The public economy of metropolitan areas. Chicago, IL: Markham. Bishop, G.F., Oldendick, R.W., & Tuchfarber, A.J. (1980). Experiments in filtering political opinions. Political Behavior, 2(4), 339-369. Bollens, J.C. (1957). Special district governments in the United States. Berkeley: CA: University of California Press. 329 Bovens, M., & Wille, A. (2009). Diploma democracy: On the tensions between meritocracy and democracy. Utrecht and Leiden: Netherlands Organisation of Scientific Research – Contested Democracies. Bowser, J.D. (2015, September). Constitutions: Amend with care. State Legislatures Magazine. Retrieved from http://www.ncsl.org/research/elections-and-campaigns/constitutionamend-with-care.aspx Boyle, K. (1998). Organized labor and American politics: 1894-1994, the labor-liberal alliance. Albany, NY: SUNY Press. Bratton, K.A., & Haynie, K.L. (1999). Agenda setting and legislative success in state legislatures: The effects of gender and race. Journal of Politics, 61(3), 658-679. Broome, M.E. (1993). Integrative literature reviews for the development of concepts. In B.L. Rodgers & K.A. Knafl (Eds.) Concept Development in Nursing (2nd ed). Philadelphia, PA: W.B. Saunders Co. Brown v. Board of Education of Topeka (I), 347 U.S. 483 (1954) . Brown v. Board of Education of Topeka (II), 349 U.S. 294 (1955). Buckles, S., Schug, M.C., & Watts, M. (2001). A national survey of state assessment practices in the social studies. The Social Studies, 92, 141-146. Buckley, J., & Kuscova, S. (2003). The effects of institutional variation on policy outcomes: The case of charter schools in the states. (Report No. 79). National Center for the Study of Privation in Education. Buddin, R., & Croft, M. (2014). Recent validity evidence for value-added measures of teacher performance. ACT Research & Policy Issue Brief (September). Bulluck, C.S., & MacManus, S.A. (1981). Policy responsiveness to the black electorate. American Politics Quarterly, 9, 357-368. Burden, B.C., Canon, D.T., Lavertu, S., Mayer, K.R., & Moynihan, D.P. (2013). Selection method, partisanship, and the administration of elections. American Politics Research, 41(6), 903-936. Burnard, P. (1994). Phone interviewing as a means of data collection: Lessons learned and practical recommendations. Forum: Qualitative Social Research, 2(2), Art. 7. Burnett, D. (2016). State school boards feel new urgency to flex muscles. Education Week 35(26). 330 Burns, N. (1994). The formation of American local governments: Private values in public institutions. New York, NY: Oxford University Press. Burns, N., Schlozman, K.L., & Verba, S. (2001). The private roots of public action: Gender, equality, and political participation. Cambridge, MA: Harvard University. California State Board of Education. (2015 January). Elementary and Secondary Education Act: Proposed Amendments to the Accountability Workbook for 2015. Retrieved from https://www.documentcloud.org/documents/1509588-sbe-nclbrequestedamendments011415.html Camera, L. (2014, Oct 22). Teachers unions to spend more than ever in state, local elections. Education Week, 34(10), 1,18. Capell, E.A. (1989). Patterns of committee membership and leadership in the California legislature: 1967-1988. Presented at the annual meeting of the Western Political Science Association, Salt Lake City, UT. Carey, J., Niemi, R.G., & Powell, L.W. (1998). The effects of term limits on state legislatures. Legislative Studies Quarterly, 23, 271-300. Carnoy, M., & Loeb, S. (2002). Does external accountability affect student outcomes? A crossstate analysis. Educational Evaluation and Policy Analysis, 24(4), 305-331. Casellas, J.P. (2009). The institutional and demographic determinants of Latino representation. Legislative Studies Quarterly, 34(3), 399-426. Caslyn, R.J., & Winter, J.P. (1999). Understanding and controlling response bias in needs assessment studies. Evaluation Review, 23(4), 399-417. Chenail, R.J. (2011). Interviewing the investigator: Strategies for addressing instrumentation and research bias concerns in qualitative research. The Qualitative Report, 16(1), 255-262. Chetty, R., Friedman, J.N., & Rockoff, J.E. (2013). Measuring the impacts of teachers II: Teacher value-added and student outcomes in adulthood (NBER Report No. w19424). National Bureau of Economic Research. Chetty, R., Friedman, J.N., & Rockoff, J.E. (2014). Measuring the impacts of teachers I: Evaluating bias in teacher value-added estimates. American Economic Review, 104(9), 2593-2632. Civil Rights Act of 1965. Pub. L. 88-352. 79 Stat. 241. 2 July 1964. Clark, C.J. (2010). Unpacking descriptive representation: Examining race and electoral representation in the American states. (Unpublished doctoral thesis). University of Iowa, Iowa City, IA. 331 Clason, D.L., & Darmody, T.J. (1994). Analyzing data measured by individual Likert-type items. Journal of Agriculture Education, 35(4), 31-35. Clayton, R.L., & Werking, G.S. (1998). Business survey of the future: The World Wide Web as a data collection method. In M.P. Couper, R.P. Baker, J. Bethlehem, Z.F. Clark, J. Martin, W.L. Nicholls II, & J.M. O’Reilly (Eds.). Computer Assisted Survey Information Collection. (pp. 543-562). New York, NY: John Wiley & Sons. Cohen, D.K. & Hill, H.C. (2001). Learning policy: When state education reform works. New Haven, CT: Yale University Press. Cohen, L.R., & Noll, R.G. (1991). How to vote, whether to vote: Strategies for voting and abstaining on Congressional roll calls. Political Behavior, 13(2), 97-127. Collins, C., & Amrein-Beardsley, A. (2014). Putting growth and value-added models on the map: A national overview. Teachers College Record, 116(1), 1-34. Committee on House Administration of the U.S. House of Representatives. (2013). Hispanic Americans in Congress, 1822-2012. Washington, D.C.: U.S. Government Printing Office. Conn, V.S., Valentine, J.C., Cooper, H.M., & Rantz, M.J. (2003). Grey literature in metaanalyses. Nursing Research, 52(4), 256-261. Conover, P.J. (1988). The role of social groups in political thinking. British Journal of Political Science, 18(1), 51-76. Conover, W.J. (1971). Practical nonparametric statistics. New York, NY: John Wiley & Sons. Constantini, E. (1990). Political women and political ambition: Closing the gender gap. American Journal of Political Science, 34(3), 741-770. Converse, J.M. (1976). Predicting no opinion in the polls. Public Opinion Quarterly, 40(4), 515530. Converse, J.M., & Presser, S. (1986). Survey questions: Handcrafting the standardized questionnaire. Sage University Paper series on Quantitative Applications in the Social Sciences (series no. 07-073). Beverly Hills, CA: Sage Publications. Cooper, H. (1998). Synthesizing research: A guide for literature reviews (3rd ed.). Thousand Oaks, CA: Sage Publications. Coram, R. (1971). Political inquiries: To which is added, a plan for the general establishment of schools throughout the United States. Wilmington: Andrews and Brynberg. Cornbleth, C., & Waugh, D. (1995). The great speckled bird: Multicultural politics and educational policymaking. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. 332 Cowen, J.M., & Strunk, K.O. (2015). The impact of teachers’ unions on educational outcomes: What we know and what we need to learn. Economics of Education Review, 48, 208-223. Dahl, R.A. (1971). Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press. Dahl, R.A. (1989). Democracy and its critics. New Haven, CT: Yale University Press. Dark, T.E. (1996). The unions and the Democrats: An enduring alliance. Ithaca, NY: Cornell University Press. Darling-Hammond, L. (2000). Teacher quality and student achievement: A review of state policy evidence. Educational Policy Analysis Archives, 8(1). Darling-Hammond, L., & Berry, B. (2006). Highly qualified teachers for all. Educational Leadership, 64(3), 14-20. Darling-Hammond, L., & Marks, E.L. (1983). The new federalism in education: State responses to the 1981 Education Consolidation and Improvement Act. Santa Monica, CA: RAND. Davis, F.D., & Venkatesh, V. (1996). A critical assessment of potential measurement biases in the technology acceptance model: Three experiments. International Journal of HumanComputer Interaction, 45(1), 19-45. DeBray-Pelot, E., & McGuinn, P. (2009). The new politics of education: Analyzing the federal education policy landscape in the post-NCLB era. Educational Policy, 23(1), 15-42. DeBruin, W.B., Parker, A.M., Fischhoff, B. (2007). Individual differences in adult decisionmaking competence. Journal of Personality and Social Psychology, 92(5), 938-956. Denzau, A.T., & MacKay, R.J. (1983). Gatekeeping and monopoly power of committees: An analysis of sincere and sophisticated behavior. American Journal of Political Science, 27, 740-761. DeVos, B. (2017, March 29). Prepared remarks to the Brookings Institution. Retrieved from https://www.ed.gov/news/speeches/us-secretary-education-betsy-devos-preparedremarks-brookings-institution Diamond, I. (1977). Sex roles in the State House. New Haven, CT: Yale University Press. Dillman, D.A. (1991). The design and administration of mail surveys. Annual Review of Sociology, 17, 225-249. Dillman, D.A., & Christian, L.M. (2005). Survey mode as a source of instability in responses across surveys. Field Methods, 17(1), 30-52. 333 DiMaggio, P.J. (1988). Interest and agency in institutional theory. In L. Zucker (Ed.), Institutional patterns and organizations (pp.3-22). Cambridge, MA: Ballinger. Dixon, A. (2011). Focus on teacher reform legislation in SREB states: Evaluation policies. Atlanta, GA: Southern Regional Education Board. Dolan, K., & Ford, L.E. (1995). Women in state legislatures. American Politics Quarterly, 23(1), 96-108. Dovi, S. (2002). Preferable descriptive representatives: Will just any woman, black or Latino do? American Political Science Review, 96(4), 729-743. Downs, A. (1957). An economic theory of democracy. New York, NY: Harper. Education Consolidation and Improvement Act (ECIA) of 1982. Pub. L. 97-35. 13 August 1981. Eisenhardt, K.M. (1989). Building theories from case study research. Academy of Management, 14(4), 532-550. Elementary and Secondary Education Act [ESEA] of 1965. Pub. L. 89-10. 79 Stat. 27. 11 April 1965. Elliott, D. (2012, November 1). Alabama racist language measure draws unexpected foes. NPR Morning Edition. Retrieved from http://www.npr.org/2012/11/02/164107184/ala-racistlanguage-measure-draws-unexpected-foes Elmore, R.F. (1979-80). Backward mapping: Implementation research and policy decisions. Political Science Quarterly, 94(4), 601-616. Epstein, D., Herron, M., O’Halloran, S., Park, D. (2007). Estimating the effects of redistricting on minority substantive representation. Journal of Law, Economics, and Organization, 23(2), 499-518. Erickson, B. (2016a). Appointment of house or assembly committees. Washington, D.C.: National Conference of State Legislatures (retrieved via personal correspondence). Erickson, B. (2016b). Appointment of standing committee chairs and members. Washington, D.C.: National Conference of State Legislatures (retrieved via personal correspondence). Erickson, R.S. (1976). The relationship between public opinion and state policy: A new look based on some forgotten data. American Journal of Political Science, 22, 25-36. Erickson, R.S., Wright, G.C., & McIver, J.P. (1993). Statehouse democracy: Public opinion and policy in the American states. Cambridge, MA: Cambridge University Press. 334 Eulau, H., & Karps, P.D. (1977). The puzzle of representation: Specifying components of responsiveness. Legislative Studies Quarterly, 2(3), 233-254. Faulkenberry, G.D., & Mason, R. (1978). Characteristics of nonopinion and no opinion response groups. Public Opinion Quarterly, 42(4), 534-543. Fenno, R.F. (1978). Home style: House members in their districts. Boston, MA: Little, Brown and Co. Ferejohn, J. (1986). Incumbent performance and electoral control. Public Choice, 50, 5-25. Fiorina, M.P. (1994). Divided government in the American states: A byproduct of legislative professionalism? American Political Science Review, 88(2), 304-316. Firestone, W.A., Bader, B.D., Massel, D., & Rosenblum, S. (1992). Recent trends in state educational reform: Assessment and prospects. Teachers College Record, 94(2), 254-277. Firestone, W.A., Fitz, J., & Broadfoot, P. (1999). Power, learning, and legitimation: Assessment implementation cross levels in the United States and United Kingdom. American Educational Research Journal, 36(4), 759-793. Firestone, W.A., Mayrowetz, D., & Fairman, J. (1998). Performance-based assessment and instructional change: The effects of testing in Maine and Maryland. Educational Evaluation and Policy Analysis, 20(2), 94-113. Fisher, S.H., & Herrick, R. (2012). Old versus new: The comparative efficiency of mail and internet surveys of state legislators. State Politics & Policy Quarterly, 13(2), 147-163. Flavin, P. (2012). Income inequality and policy representation in the American state. American Politics Research, 40(1), 29-59. Fligstein, N. (1997). Social skill and institutional theory. American Behavioral Scientist, 40(4), 397-405. Flynn, T.N., & Marley, A.A.J. (2013). Best-worst scaling: Theory and methods. In S. Hess & A. Daly (Eds), Handbook of choice modeling. Northampton, MA: Edward Elgar Publishing. Fontana, A., & Frey, J. (2005). The Sage handbook of qualitative research. Thousand Oaks, CA: Sage. Fossett, J., & Thompson, F.J. (2006). Administrative responsiveness to the disadvantaged: The case of children’s health insurance. Journal of Public Administration Research and Theory, 16(3), 369-392. Foster, K.A. (1997). The political economy of special-purpose government. Washington, D.C.: Georgetown University Press. 335 Fox, R.L., & Lawless, J.L. (2005). To run or not to run for office: Explaining nascent political ambition. American Journal of Political Science, 49(3), 642-659. Francia, P.L. (2006). The future of organized labor in American politics. New York, NY: Columbia University Press. Francis, J.D., & Busch, L. (1975). What do we know about “I don’t knows.” Public Opinion Quarterly, 39(2), 207-218. Fuhrman, S. (1993). Designing coherent education policy: Improving the system. San Francisco, CA: Jossey-Bass. Fund, J.H. (1992). Term limitation: An ideal whose time has come. In G. Benjamin & M.J. Malbin (eds.). Limiting legislative terms. Washington, D.C.: CQ Press. Gallup. (2016). Americans still more trusting in local over state government. Retrieved from http://www.gallup.com/poll/195656/americans-trusting-local-state-government.aspx Gallup. (n.d.) Party affiliation. Retrieved from http://www.gallup.com/poll/15370/partyaffiliation.aspx Garbett, R., & McCormack, B. (2001). The experience of practice development. An exploratory telephone interview study. Journal of Clinical Nursing, 10, 94-102. Gay, C. (2001). The effect of black congressional representation on political participation. American Political Science Review, 95(3), 589-602. Giddens, A. (1976). New rules of sociological method. London, UK: Hutchinson. Giddens, A. (1979). Central problems in social theory: Action, structure and contradiction in social analysis. Los Angeles, CA: University of California Press. Gilens, M. (2005). Inequality and democratic responsiveness. Public Opinion Quarterly, 69(5), 778-796. Gill, B., McCombs, J.S., Naftel, S., Ross, K.E., Song, M., Harmon, J. & Vernez, G., Birman, B., Garet, M., & O’Day, J. (2008). Title I school choice and supplemental education services under No Child Left Behind: Progress toward implementation (RB-9334-EDU). Santa Monica, CA: RAND Corporation. Giroux, H. (2011). Neoliberalism’s war against teachers in dark times. Cultural Studies Critical Methodologies, 13(6), 458-468. Gjestland, L. (1996). Net? Not yet. Marketing Research: Magazine of Management and Application, 8(1), 26-29. 336 Goldstein, D. (2015). The teacher wars: A history of America’s most embattled profession. New York: Anchor Books. Goldstein, G., & Hersen, M. (1984). Handbook of psychological assessment. New York, NY: Pergamon press. Goodhue, D.L. (1998). Development and measurement validity of a task-technology fit instrument for user evaluations of information systems. Decision Sciences, 29(1), 105138. Graff, C., & Stufft, D. (2011). Increasing visibility for LGBTQ students: What schools can do to create inclusive classroom communities. Current Issues in Education, 14(1). Grandy, J. (1998). Response bias in a survey of Asian-American and white science and engineering students. Journal of Women and Minorities in Science and Engineering, 4(1), 1-13. Granovetter, M. (1985). Economic action and social structure: The problem of embededness. American Journal of Sociology, 91(3), 481-510. Grofman, B. (Ed.) (1996). Legislative term limits: Public choice perspectives. Norwell, MA: Kluwer Academic Publishers. Groves, R.M., Fowler, F.J., Couper, M.P., Lepkowski, J.M., Singer, E., & Tourangeau, R. (2009). Survey methodology (2nd Ed.). Hoboken, NJ: John Wiley & Sons, Inc. Hamann, E.T., & Lane, B. (2004). The roles of state departments of education as policy intermediaries: Two cases. Educational Policy, 18(3), 426-455. Hanushek, E. (2011). The “war on teachers” is a myth. Hoover Digest, 1. Retrieved from http://hanushek.stanford.edu/publications/%E2%80%9Cwar-teachers%E2%80%9D-myth Hanushek E.A., & Rivkin, S.G. (2010). Generalizations about using value-added measures of teacher quality. The American Economic Review, 100(2), 267-271. Hardy-Fanta, C., Sierra, C.M., Lien, P., Pinderhughes, D.M., & Davis, W.L. (2005). Race, gender, and descriptive representation: An exploratory view of multicultural elected leadership in the United States. Paper presented at the Annual Meeting of the American Political Science Association (Sept 1-4, 2005) Washington, D.C. Harlacher, J. (2016, January). An educator’s guide to questionnaire development (REL 2016108). Washington, D.C.: U.S. Department of Education, Institute of Education Sciences, National Center for Education Evaluation and Regional Assistance, Regional Education Laboratory Central. Retrieved from http://ies.ed.gov/ncee/edlabs. 337 Harrell, M.G., & Bradley, M.A. (2009). Data collection methods: Semi-structured interviews and focus groups. Santa Monica, CA: RAND Corporation. Hartigan, J.A., & Hartigan, P.M. (1985). The dip test of unimodality. The Annals of Statistics. Hawes, D.P. (2013). Representation, salience, and responsiveness to Latino student performance. Journal of Latinos and Education, 12(2), 87-107. Haynie, K.L. (2002). The color of their skin or the content of their behavior? Race and perceptions of African American legislators. Legislative Studies Quarterly, 27(2), 295314. Hedlund, R.D., & Patterson, S.C. (1992). The electoral antecedents of state legislative committee assignments. Legislative Studies Quarterly, 17(4), 539-559. Henderson, M.B., Peterson, P.E., & West, M.R. (2016). The 2015 EdNext Poll on School Reform. EducationNext, 16(1). Henig, J. (2013). The end of exceptionalism in American education: The changing politics of school reform. Cambridge, MA: Harvard Education Press. Herrera, C.L., Herrera, R., & Smith, E.R.A.N. (1992). Public opinion and congressional representation. Public Opinion Quarterly, 56(2), 185-205. Hess, F.M., & Meeks, O. (2010). Governance in the accountability era. The National School Boards Association, The Thomas B. Fordham Institute and the Iowa School Boards Foundation. Hill, M.L. (2006). Representin(g): Negotiating multiple roles and identities in the field and behind the desk. Qualitative Inquiry, 12(5), 926-949. Hogan, R.E. (2005). Gubernatorial coattail effects in the state legislative elections. Political Research Quarterly, 58(4), 587-597. Hoonakker, P., & Carayon, P. (2009). Questionnaire survey nonresponse: A comparison of postal mail and internet surveys. Journal of Human-Computer Interaction, 25(5), 348373. Hummel, R.P. (1987). The bureaucratic experience (3rd ed.). New York, NY: St. Martin’s Press. Hurley, J.R. (1998). Timidity as a response style to psychological questionnaires. Journal of Psychology, 132(2), 202-210. Hurst, D., Tan, A., Meek, A., & Sellers, J. (2003). Overview and inventory of sate education reforms: 1990 to 2000. Washington, D.C.: U.S. Department of Education Institute of Education Sciences. 338 Initiative and Referendum Institute, (n.d.). State I&R. Retrieved from http://www.iandrinstitute.org/states.cfm Jacob, B.A., & Levitt, S.D. (2003). Rotten apples: An investigation of the prevalence and predictors of teacher cheating . The Quarterly Journal of Economics, 11(3), 843-877. Jacobsen, R., & Saultz, A. (2012). The Polls – Trends: Who should control education. Public Opinion Quarterly, 76(2), 379-390. Jadad, A.R., Moher, D., & Klassen, T.P. (1998). Guides for reading and interpreting systematic reviews: How did the authors find the studies and assess their quality? Archives of Pediatric and Adolescent Medicine, 152(8), 812-817. Jenkins, G.D., & Taber, T.D. (1977). A Monte-Carlo study of factors affecting three indices of composite scale reliability. Journal of Applied Psychology, 62, 392-398. Jimerson, J., & Childs, J. (2015). Signal and symbol: How state and local policies address datainformed practice. Educational Policy, 31(1). Johnson, D. (1997). Putting the cart before the horse: Parent involvement in the Improving America’s Schools Act. California Law Review, 85(6), 1757-1801. Joint Center for Political and Economic Studies. (2002). Black elected officials: A national roster. Washington, D.C.: Author. Jones, V. (1942). Metropolitan government. Chicago, IL: University of Chicago Press. Jordan, H.R., Mendro, R.L., & Weerasinghe, D. (1997). Teacher effects on longitudinal student achievement: A report on research in progress. Paper presented at the CREATE Annual Meeting, Indianapolis, IN. Juliusson, E.A., Karlsson, N., & Gärling T. (2005). Weighing the past and the future in decision making. European Journal of Cognitive Psychology, 17, 561-575 . Kathlene, L. (1994). Power and influence in state legislative policymaking: The interaction of gender and position in committee hearing debates. American Political Science Review, 88(3), 560-576. Kearney, R.C., & Sinha, C. (1988). Professionalism and bureaucratic responsiveness: Conflict or compatibility? Public Administration Review, 48(1), 571-579. Kesler, C. (1990). Bad housekeeping: The case against congressional term limits. Policy Review, 53, 20-25. Key, V.O. (1949). Southern politics in state and nation. New York, NY: A.A. Knopf. 339 Key, V.O. (1961). Public opinion and American democracy. New York, NY: A.A. Knopf. Kim, Y. (2010). The pilot study in qualitative inquiry: Identifying issues and learning lessons for culturally competent research. Qualitative Social Work, 10(2), 190-206. Kincaid, J. (1998). The devolution tortoise and the centralization hare. New England Economic Review, May/June, 13-52. Kirkevold, M. (1997). Integrative nursing research – an important strategy to further the development of nursing science and nursing practice. Journal of Advanced Nursing, 25(5), 977-984. Kirst, M. (1987). Who should control out schools: Reassessing current policies. Paper prepared for the Breckenbridge Forum for the Enhancement of Teaching (Trinity University, San Antonio, TX, August 18-21, 1987) Kirst, M. (2010). The politics and policy dynamics of K-12 reform from 1965 to 2010: Implications for changing postsecondary education. Research Priorities for Broad-Access Higher Education. Retrieved from https://cepa.stanford.edu/sites/default/files/Evolution%20of%20K-12_9_2_10.pdf Kirst, M. W., & Wirt, F.M. (2009). Political dynamics of American education (4th ed.). Richmond, CA: McCutchan Publishing Corporation. Klein, A. (2016, January 5). Under ESSA, states, districts to share more power. Education Week. Retrieved from http://www.edweek.org/ew/articles/2016/01/06/under-essa-statesdistricts-to-share-more.html Kolbe, T., & King Rice, J. (2012). And they’re off: Tracking federal Race to the Top investments from the starting gate. Educational Policy, 26(1), 185-209. Kousser, T., & Phillips, J.H. (2012). The power of American Governors: Winning on budgets and losing on policy. New York, NY: Cambridge University Press. Krause, G.A. (2010). Legislative delegation of authority to bureaucratic agencies. In R.F. Durant (Ed.), The Oxford Handbook of American Bureaucracy (pp. 521-544). New York, NY: Oxford University Press. Krippendrof, K. (1980). Content analysis: An introduction to its methodology. Beverly Hills, CA: Sage. Krosnick, J.A. (1999). Survey research. Annual Review of Psychology, 50, 537-567. Krosnick, J.A. (2002). The causes of no-opinion responses to attitude measures in surveys: They rarely are what they appear to be. In R.M. Groves, D.A. Dillman, J.L. Eltinge, & R.J.A. Little (Eds.), Survey nonresponse (pp. 88-100). New York, NY: Wiley. 340 Krosnick, J.A., & Milburn, M.A. (1990). Psychological determinants of political opinionation. Social Cognition, 8(1), 49-72. Krosnick, J.A., Holbrook, A.L., Berent, M.K., Carson, R.T., Hanemann, W.M., Kopp, R.J., Mitchell, R.C., Presser, S., Rudd, P.A., Smith, V.K., Moody, W.R., Green, MC., & Conaway, M. (2002). The impact of “no opinion” response options on data quality: nonattitude reduction or an invitation to satisfice? Public Opinion Quarterly, 66(3), 371-403. Kurtz, K. (2015). Who we elect: The demographics of state legislatures. State Legislatures Magazine (December). Retrieved from http://www.ncsl.org/research/about-statelegislatures/who-we-elect.aspx Labaree, D. F. (1997). Public goods, private goods: The American struggle over educational goals. American Educational Research Journal, 34(1), 39-81. Lascher, E.L., Hagen, M.G., & Rochlin, S.A. (1996). Gun behind the door? Ballot initiatives, state policies, and public opinion. Journal of Politics, 58(3), 1250-1256. Lax, J.R., & Phillips, J.H. (2009). Gay rights in the states: Public opinion and policy responsiveness. American Political Science Review, 103(3), 367-386. Le Floch, K., Martinez, F., O’Day, J., Stecher, B., Taylor, J., & Cook, A. (2007). State and local implementation of the No Child Left Behind Act: Volume III—Accountability under NCLB: Interim report. Washington, D.C.: U.S. Department of Education, Office of Planning, Evaluation and Policy Development, Policy and Program Studies Service. Lichtenstein, N. (2002). State of the union: A century of American labor. Princeton, NJ: Princeton University Press. Liebowitz, A., & Tollison, R. (1980). A theory of legislative organization: Making the most of your majority. Quarterly Journal of Economics, 94, 261-77. Lincoln, Y.S., & Guba, E.G. (1985). Naturalistic Inquiry. Beverly Hills, CA: Sage. Lissitz, R.W., & Green, S.B. (1975). Effect of the number of scale points on reliability: A Monte-Carlo approach. Journal of Applied Psychology, 60, 10-13. Loeb, S., Miller, L., & Strunk, K.O. (2009). The state role in teacher professional development and education throughout teachers’ careers. Education Finance and Policy, 4(2), 212228. Lopez, S.J. (2012). Americans’ views of public schools still far worse than parents’. Gallup. Retrieved from http://www.gallup.com/poll/142658/americans-views-public-schools-farworse-parents.aspx 341 Louviere, J.J. (1991). Best-worst scaling: A model for the largest difference judgments. Working paper. University of Alberta. Louviere, J.J. (2001). Choice experiments: An overview of concept and issues. In J. Bennett, & R. Blamey (Eds.), The choice modeling approach to environmental valuation. Northampton, MA: Edward Elgar Publishing. Lowi, T. (1971). The politics of disorder .New York, NY: Basic Books. Lupia, A. (1994). The political dynamics of urban voting behavior. American Journal of Political Science, 33(1), 150-174. Maestas, C., Neeley, G.W., & Richardson, L.E. (2003). The state of surveying legislators: Dilemmas and suggestions. State Politics and Policy Quarterly, 3(1), 90-108. Macedo, S., & Karpowitz, C.F. (2006). The local roots of American inequality. PS: Political Science & Politics, 39(1), 59-64. MacManus, S.A., Bulluck, C.S., Padgett, K., & Penberthy, B. (2006). Women winning at the local level: Are county and school board positions becoming more desirable and plugging the pipeline to higher office? In L.D. Whitaker (Ed.), Women and elections: Outsiders or insiders (pp. 117-136). Upper Saddle River, NJ: Prentice Hall. Maeroff, G.I. (2010). School boards in America: A flawed exercise in democracy. New York, NY: Palgrave McMillan. Maestas, C. (2000). Professional legislatures and ambitious politicians: Policy responsiveness of state institutions. Legislative Studies Quarterly, 25(4), 663-690. Malen, B. (2003). Tightening the grip? The impact of state activism on local school systems. Educational Policy, 17(2), 195-216. Maness, R., Malm, L. & Doroghazi, L. (2016a). 2016 predictions: What will state legislatures focus on this year? Multistate Insider. Retrieved from https://www.multistate.com/insider/2016/01/2016-predictions-what-will-statelegislatures-focus-on-this-year/ Maness, R., Malm, L. & Doroghazi, L. (2016b). Trends in gubernatorial priorities in 2016. Multistate Insider. Retrieved from https://www.multistate.com/insider/2016/02/trends-ingubernatorial-priorities-in-2016/ Manna, P. (2006). School’s in: Federalism and the national education agenda. Washington, D.C.: Georgetown University Press. Manna, P. (2011). Collision course: Federal education policy meets state and local realities. Washington, D.C.: CQ Press. 342 Manna, P., & McGuinn, P. (Eds.). (2013). Education governance for the twenty-first century: Overcoming the structural barriers to school reform. Washington, D.C.: The Brookings Institution. Manna, P., & Harwood, T. (2011). Governance and educational expectations in the U.S. states. State Politics & Policy Quarterly, 11(4), 483-509. Manning, J.E., & Shogan, C.J. (2012). African American members of the United States Congress: 1870-2012. Washington, D.C.: Congressional Research Service. Manning, J.E., Brudnick, I.A., Shogan, C.J. (2015). Women in Congress: Historical overview, tables, and discussion. Washington, D.C.: Congressional Research Service. Mansbridge, J. (1999). Should blacks represent blacks and women represent women? A contingent “yes.” The Journal of Politics, 61(3), 628-657. March, J.G., & Olsen, J.P. (1984). The new institutionalism: Organizational factors in political life. The American Political Science Review, 78(3), 734-749. Marzano, R.J., Pickering, D.J., & Pollock, J.E. (2001). Classroom instruction that works: Research-based strategies for increasing student achievement. Alexandria, VA: ASCD. Masters, N.A., Salisbury, R.H., & Eliot, T.H. (1964). State politics and the public schools: An exploratory analysis. New York, NY: Alfred A. Knopf. Mavrogordato, M., & White, R.S. (2017). Reclassification variation: How policy implementation guides the process of exiting students from English learner status. Educational Evaluation and Policy Analysis. DOI: 10.3102/0162373716687075. Mayhew, D. (1974). Congress: The electoral connection. New Haven, CT: Yale University Press. Mazzoni, T.L. (1974). State policymaking for the public schools of Minnesota. Columbus, OH: Educational Governance Project. Mazzoni, T.L. (1980). Deciding state school policy in Minnesota. St. Paul, MN: University of Minnesota, Department of Educational Administration. Mazzoni, T.L. (1986, April) The choice issue in state school reform, the Minnesota experience. Paper presented at the Annual Meeting of the American Educational Research Association, San Francisco, CA. Mazzoni, T.L. (1988). The politics of educational choice in Minnesota. In W.L. Boyd & C.T. Kerchner (Eds.), The politics of excellence and choice in education (pp. 217-230). New York, NY: Falmer. 343 Mazzoni, T.L. (1993). The changing politics of state education policy making: A 20-year Minnesota perspective. Educational Evaluation and Policy Analysis, 15(4), 357-379. McCoyd, J.L., & Kerson, T.S. (2006). Conducting intensive interviews using email: A serendipitous comparative opportunity. Qualitative Social Work, 5,389-406. McDermott, K.A. (2003). What causes variation in states’ accountability policies? Peabody Journal of Education, 78(4), 153-176. McDermott, M. (2005). Candidate occupations and voter information shortcuts. Journal of Politics, 67, 201-219. McDonnell, L.M., & McLaughlin, M.W. (1980). Program consolidation and the state role in ESEA Title IV. Santa Monica, CA: Rand Corporation. McDonnell, L.M., & McLaughlin, M.W. (1982). Education policy and the role of the states. Santa Monica, CA: Rand Corporation. McGuinn, P.J. (2006). No Child Left Behind and the transformation of federal education policy, 1965-2005. Lawrence, KS: University of Kansas Press. McGuinn, P. (2012). The state of teacher evaluation reform: State education agency capacity and the implementation of new teacher evaluation systems. Washington, D.C.: Center for American Progress. McKelvie, S.J. (1978). Graphic rating scales: How many categories? British Journal of Psychology, 69, 185-202. McLaughlin, M. (1975). Evaluation and reform: Elementary and Secondary Education Act of 1965, Title I. New York, NY: HarperCollins. McLaughlin, M. (1976). Implementation of ESEA Title I: A problem of compliance. Teachers College Record, 77(3), 397-415. McLaughlin, M., & Verstegen, D. (1998). Increasing regulatory flexibility of special education programs: Problems and promising strategies. Exceptional Children, 64, 371-384. McLendon, M.K., Hearn, J.C., & Deaton, R. (2006). Called to account: Analyzing the origins and spread of state performance-accountability policies in higher education. Educational Evaluation and Policy Analysis, 28(1), 1-24. McLendon, M.K., Mokher, C.G., & Flores, S.M. (2011). Legislative agenda setting for in-state resident tuition policies: Immigration, representation, and educational access. American Journal of Education, 117(4), 563-602. 344 McLeod, A. (2012). The party on the bench: Partisanship, judicial selection commissions, and state high-court appointments. Justice System Journal, 33(3), 262-274. McMurrer, J. (2007, December). NCLB Year 5: Choices, changes, and challenges: Curriculum and instruction in the NCLB era. Washington, D.C.: Center on Education Policy. McNeil, L.M. (2000). Contradictions of reform. New York, NY: Routledge. Meier, K.J., & England, R.E. (1984). Black representation and educational policy: Are they related? The American Political Science Review, 78(2), 392-403. Meier, K.J., & Stewart, J.J. (1992). The impact of representative bureaucracies: Educational systems and public policies. American Review of Public Administration, 22(3), 157-171. Merriam, S.B. (1998). Qualitative research and case study applications in education. San Francisco, CA: Jossey-Bass. Mendez, E. (2013). Americans down on Congress, OK with own representative. Gallup. Retrieved from http://www.gallup.com/poll/162362/americans-down-congress-ownrepresentative.aspx Merriam, S.B., & Tisdell, E.J. (2016). Qualitative research: A guide to design and implementation (4th ed.). San Francisco, CA: Jossey-Bass. Metzger, K. (2016). Metzger Report. Prepared report for Mich. State A. Philip Randolph Inst. v. Johnson, No. 16-2071. 6th Cir. Aug. 17, 2016. Mezey, S.G. (1994). Increasing the number of women in office: Does it matter? In E.A. Cook, S. Thomas & C. Wilcox (Eds.), The year of the woman: Myths and realities. Boulder, CO: Westview Press. Miewald, R.D. (1978). Public administration: A critical perspective. New York, NY: McGrawHill. Miles, M.B., Huberman, A.M., & Saldana, J. (2014). Qualitative data analysis: A methods sourcebook (3rd ed.). Thousand Oaks, CA: Sage. Miller, G.A. (1956). The magical number seven plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81-97. Miller, W.E., & Stokes, D. (1963). Constituency influence in Congress. American Political Science Review, 57, 165-177. Milstein, M.M. (1976). Impact and response: Federal aid and state education agencies. New York, NY: Teachers College Press, Columbia University. 345 Minichiello, V., Aroni, R., Timewell, E., & Alexander, L. (1990). In-depth interviewing: Researching people. Melbourne: Longman Cheshire. Minnici, A., & Bartley, A.P. (2007, March). State implementation of supplemental education services under the No Child Left Behind Act. Washington, D.C.: Center on Education Policy. Minnici, A., & Hill, D.D. (2007, May). Educational architects: Do state education agencies have the tools necessary to implement NCLB? Washington, D.C.: Center on Education Policy. Modzeleski, W. (1996). Creating safe schools: Roles and challenges, a federal perspective. Education and Urban Society, 28(4), 412-423. Moe, T.M. (2003). Political control and the power of the agent. Paper presented at the annual meeting of the Midwest Political Science Association, Chicago, IL. Moe, T.M. (2011). Special interest: Teachers unions and America’s public schools. Washington, D.C.: Brookings Institution Press. Moncrief, G.F., Squire, P., & Jewell, M.E. (2001). Who runs for the legislature? Upper Saddle River: Prentice Hall. Montgomery, P. (1990). Should congressional terms be limited? Common Cause, July/August, 31-33. Moore, M.T., Goertz, M.E., & Hartle, T.W. (1983). Interaction of federal and state programs. Education and Urban Society, 15, 452-478. Moors, G., Kieruj, N.D., & Vermunt, J.K. (2014). The effect of labeling and numbering of response scales on the likelihood of response bias. Sociological Methodology, 44(1), 369299. Moser, C.A., & Kalton, G. (1972). Survey methods in social investigation. New York, NY: Routledge. Mosher, F. C. (1982). Democracy and the public service (2nd Ed). New York: Oxford University Press. Mullin, M. (2008). The conditional effect of specialized governance on public policy. American Journal of Political Science, 52(1), 125-141. Mullin, M. (2009). Governing the tap: Special district governance and the new local politics of water. Cambridge, MA: The MIT Press. 346 Murphy, J. (1971). Title I of ESEA: The politics of implementing federal education reform. Harvard Educational Review, 41(1), 35-63. Nagel, K., Yunker, C., & Malmgren, K.W. (2006). Students with disabilities and accountability reform: Challenges identified at the state and local levels. Journal of Disability Policy Studies, 17(1), 28-39. Nakagawa, K. (2000). Unthreading the ties that bind: Questioning the discourse of parent involvement. Educational Policy, 14(4), 443-472. National Association of Latino Elected and Appoint Officials [NALEO]. (1996). National directory of Latino elected officials. Washington, D.C.: NALEO Education Fund. National Association of Latino Elected and Appoint Officials [NALEO]. (2016). National directory of Latino elected officials. Washington, D.C.: NALEO Education Fund. National Association of Latino Elected Officials [NALEO] and the National Conference of State Legislatures [NCSL]. (2009). 2009 Latino legislators. Retrieved from http://www.ncsl.org/research/about-state-legislatures/latino-legislators-overview.aspx National Association of State Boards of Education. (n.d.). States boards of education. Retrieved from http://www.nasbe.org/about-us/state-boards-of-education/ National Black Caucus of State Legislatures [NBCSL] and NCSL. (2009). African-American legislators 2009. Retrieved from http://www.ncsl.org/research/about-statelegislatures/african-american-legislators-in-2009.aspx 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, D.C.: Author. National Conference of State Legislatures [NCSL]. (2010). Women in state legislatures: 2010 legislative session. Retrieved from http://www.ncsl.org/legislatorsstaff/legislators/womens-legislative-network/women-in-state-legislatures-2010.aspx National Conference of State Legislatures. (2012). Initiative petition signature requirements. Retrieved from http://www.ncsl.org/research/elections-and-campaigns/signaturerequirements.aspx National Governors Association. (n.d.). Governors’ powers and authority. Retrieved from https://www.nga.org/cms/home/management-resources/governors-powers-andauthority.html No Child Left Behind [NCLB] Act of 2001, Pub. L. 107-110. 8 January 2002. Oates, W.E. (1972). Fiscal federalism. New York, NY: Harcourt Brace Jovanovich. 347 O’Day, J.A. & Smith, M.S. (1993). Systemic reform and educational opportunity. In S.H. Fuhrman (Ed.), Designing coherent education policy: Improving the system (pp. 250312). San Francisco, CA: Jossey-Bass. O’Day, J., Elledge, A., Le Floch, K.C., Taylor, J., & Anderson, L. (2009). State and local implementation of the No Child Left Behind Act: Volume V—Implementation of the 1 percent rule and 2 percent interim policy options. Washington, D.C.: U.S. Department of Education, Office of Planning, Evaluation and Policy Development, Policy and Program Studies Service. Odden, A. (Ed.) (1991). Education policy implementation. New York, NY: SUNY Press. Olds, C. (2011). Unexpected consequences? Bicameral policymaking shifts through term limits. Foresight, 13(6), 48-69. Opendakker, R. (2006). Advantages and disadvantages of four interview techniques in qualitative research. Forum: Qualitative Social Research, 7(4), Art. 11. Orey, D., Overby, L.M., & Larimer, C.W. (2007). African-American committee chairs in American state legislatures. Social Science Quarterly, 88, 619-639. Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective action. New York, NY: Cambridge University Press. Ostrom, E. (1991). Rational choice and institutional analysis: Toward complementarity. American Political Science Review, 85(1), 237-243. Ostrom, V., Bish, R., & Ostrom, E. (1988). Local government in the United States. San Francisco, CA: Institute for Contemporary Studies Press. Oxman, A.D. (1994). Checklists for review articles. British Medical Journal, 309, 648-651. Page, B.I., & Shapiro, R.Y. (1983). Effects of public opinion on policy. The American Political Science Review, 77(1), 175-190. Pantoja, A.D., & Segura, G.M. (2003). Does ethnicity matter? Descriptive representation in legislatures and political alienation among Latinos. Social Science Quarterly, 84(2), 441460. Pena, D.C. (2000). Parent involvement: Influencing factors and implications. The Journal of Educational Research, 94(1), 42-54. Pennington, K. (2014). ESEA waivers and teacher-evaluation plans: State oversight of districtdesigned teacher-evaluation systems. Washington, D.C.: Center for American Progress. Petracca, M.P. (1991). The poison of professional politics. Cato Institute Policy Analysis No 151. 348 Pew Research Center. (2013). State governments viewed favorably as federal rating hits new low. Retrieved from http://www.people-press.org/2013/04/15/state-govermnents-viewedfavorably-as-federal-rating-hits-new-low/ Pew Research Center. (2015, April). A deep dive into party affiliation: Sharp differences by race, gender, generation, education. Washington, D.C.: Author. Phi Delta Kappan [PDK] & Gallup. (2014). The 46th annual PDK/Gallup poll of the public’s attitudes toward the public schools. Retrieved from http://pdkpoll2015.pdkintl.org/wpcontent/uploads/2015/08/PDK_Poll46_2014.pdf PDK & Gallup. (2015). The 47th annual PDK/Gallup poll of the public’s attitudes toward the public schools. Retrieved from http://pdkpoll2015.pdkintl.org/wpcontent/uploads/2015/10/pdkpoll47_2015.pdf Pinkerton, E., Scott, C., Buell, B., & Kober, N. (2004, January). From the capital to the classroom: Year 2 of the No Child Left Behind Act. Washington, D.C.: Center on Education Policy. Pitkin, H.F. (1967). The concept of representation. Berkeley, CA: University of California Press. Pope, P., Boleman, C., & Cummings, S. (2005). Questionnaire design: Asking questions with a purpose. College Station, TX: Texas A&M AgriLIFE Extensions Service. Retrieved from http://agrilifecdn.tamu.edu/od/files/2010/04/Questionnair-Design-Publication-E-227.pdf Powell, G.B. (2003). On democratic responsiveness. Paper presented at The Quality of Democracy: Improvement or Subversion Conference, Stanford University. Preston, C.C., & Coleman, A.M. (2000). Optimal number of response categories in rating scales: Reliability, validity, discriminating power, and respondent preferences. Acta Psychologica, 104, 1-15. Przeworski, A., Stokes, S.C., & Manin, B. (Eds.). (1999). Democracy, accountability, and representation. Cambridge, UK: Cambridge University Press. Rahn, W.M. (1993). The role of stereotypes in information processing about political candidates. American Journal of Political Science, 37(20), 472-496. Randolph, J.J. (2009). A guide to writing the dissertation literature review. Practical Assessment, Research and Evaluation, 14(13). Available online: http://pareonline.net/getvn.asp?v=14&n=13 Rapoport, R.B. (1981). The sex gap in political persuading: Where the ‘structuring principle’ works. American Journal of Political Science, 25(1), 32-48. 349 Rapoport, R.B. (1982). Sex differences in attitude expression: A generational explanation. Public Opinion Quarterly, 46(1), 86-96. Redford, E. (1969). Democracy in the administrative state. New York, NY: Oxford University Press. Reese, W.J. (2002). Power and promise of school reform: Grassroots movements during the Progressive Era. New York, NY: Teachers College Press. Reichley, J. (Ed.) (1964). States in crisis: Politics in ten Americans states, 1950-1962. Chapel Hill, NC: The University of North Carolina Press. Reingold, B. (2000). Representing women: Sex, gender, and legislative behavior in Arizona and California. Chapel Hill, NC: University of North Carolina Press. Reingold, B., & Smith, A.R. (2012). Welfare policymaking and intersections of race, ethnicity, and gender in U.S. state legislatures. American Journal of Political Science, 56(1), 131147. Remmers, H.H., & Ewart, E. (1941). Reliability of multiple-choice measuring instruments as a function of the Spearman-Brown prophecy formula. Educational and Psychological Measurement, 35, 319-324. Riccucci, N.M., & Saidel, J.R. (1997). The representativeness of state-level bureaucratic leaders: A missing piece of the representative bureaucracy puzzle. Public Administration Review, 57(5), 423-430. Riccucci, N.M., & Saidel, J.R. (2001). The demographics of gubernatorial appointees: Toward an explanation of variation. Policy Studies Journal, 29(1), 11-22. Richardson, L., & Freeman, P. (1995). Gender differences in constituency services among state legislators. Political Research Quarterly, 48(1), 169-179. Richardson, L.E., Valentine, D., & Stokes, S.D. (2005). Assessing the impact of term limits in Missouri. State and Local Government Review, 37(3), 177-192. Ringquist, E.J. (1994). Policy influence and policy responsiveness in state pollution control. Policy Studies Journal, 22(1), 25-44. Robinson v. Cahill, 303 A.2d 273 (1973). Rocha, R.R., Tolbert, C.J., Bowen, D.C., & Clark, C.J. (2010). Race and turnout: Does descriptive representation in state legislatures increase minority voting? Political Research Quarterly, 63(4), 890-907. Rosen, H.S. (1999). Public finance (5th ed.). McGraw-Hill. 350 Rosenthal, A. (1998). The decline of representative democracy: Process, participation, and power in state legislatures. Washington, D.C.: Congressional Quarterly Inc. Rosenthal, C.S. (1998). When women lead: Integrative leadership in state legislatures. New York, NY: Oxford University Press. Rossman, G.B., & Rallis, S.F. (2003). Learning in the field: An introduction to qualitative research (2nd Ed.). Thousand Oaks, CA: Sage. Rothstein, J. (2009). Student sorting and bias in value-added estimation: Selection on observables and unobservables. Education Finance & Policy, 4(4), 537-571. Rugg, D., & Hadley, C. (1944). The wording of questions. In C. Hadley, Gauging Public Opinion (pp. 23-50). Princeton, NJ: Princeton University Press. Russell, J.L., Meredith, J., Childs, J., Stein, M.K., & Prine, D.W. (2015). Implementing education policy through organizational networks: An analysis of state Race to the Top applications. Educational Evaluation and Policy Analysis, 37(1), 92-112. Russell, S.T., Horn, S., Kosciw, J., & Saewyc, E. (2010). Safe schools policy for LGBTQ students. Society for Research in Child Development Social Policy Report 24(4). San Antonio Independent School District v. Rodriguez, 411 U.S. 1. (1972). Sax, L.J., Gilmartin, S.K., & Bryant, A.N. (2003). Assessing response rates and nonresponse bias in web and paper surveys. Research in Higher Education, 44(4), 409-432. Schaffner, B., & Ansolabehere, S. (2015). Cooperative Congressional Election Survey [CCES] common content, 2014. doi: 10.7910/DVN/XFXJVY, Harvard Dataverse, V3. Schaffner, B.F., Streb, M., & Wright, G. (2001). Teams without uniforms: The nonpartisan ballot in state and local elections. Political Research Quarterly, 54, 7-30. Schlesinger, J.A. (1965). The Politics of the Executive. In H. Jacob and K. N. Vines (Eds.). Politics in the American States: A Comparative Analysis (pp. 207-37). Boston, MA: Little, Brown. Schlesinger, J.A. (1966). Ambition and politics: Political careers in the United States. Schuman, H., & Presser, S. (1981). Questions and answers in attitude surveys: Experiments on question form, wording and context. New York, NY: Academic Press. Serrano v. Priest (I), 5 Cal.3d 584 (1971). Serrano v. Priest (II), 18 Cal.3d 728 (1976). 351 Serrano v. Priest (III), 20 Cal.3d 25 (1977). Seo, M., & Creed, D. (2002). Institutional contradictions, praxis and institutional change: A dialectical perspective. Academy of Management Review, 27(2), 222-247. Shapiro, R.Y. (1982). The dynamics of public opinion and public policy. Doctoral Dissertation, University of Chicago. Shelly, B. (2008). Rebels and their causes: State resistance to No Child Left Behind. Publius: The Journal of Federalism, 36(3), 443-459. Shelly, B. (2011). Money, mandates, and local control in American public education. Ann Arbor, MI: The University of Michigan Press. Shelly, B. (2012). Flexible response: Executive federalism and the No Child Left Behind Act of 2001. Educational Policy, 26(1), 117-135. Sherrill, K. (1998). The dangers of non-partisan elections to democracy. Social Policy, 28(4), 1522. Sheskin, D.J. (2011). Handbook of parametric and nonparametric statistical procedures (5th ed.). Boca Raton, FL: Chapman & Hall/CRC. Skjeie, H. (1991). The rhetorical of difference: On women’s inclusion into political elites. Politics and Society, 19(2), 233-263. Smith, B. (1998). It’s about time: Opportunities to learn in Chicago’s elementary schools. Chicago, IL: Consortium on Chicago School Research. Smith, C.B. (1997). Casting the net: Surveying an Internet population. Journal of Computermediate Communication, 3, 77-84. Smith, M.S., & O’Day, J.A. (1991). Systemic school reform. In S. Fuhrman & B. Malen (Eds.), The politics of curriculum and testing (pp. 233-268). Bristol, PA: Falmer. Sonenshein, R.J. (1990). Can Black candidates win statewide elections? Political Science Quarterly, 105, 219-241. Squire, P. (2007). Measuring state legislative professionalism: The Squire Index revisited. State Politics and Policy Quarterly, 7(2), 211-227. Stanovich, K.E., & West, R.F. (2008). On the relative independence of thinking biases and cognitive ability. Journal of Personality and Social Psychology, 94, 672-695. 352 Stark Retner, D., Scott, C., Kober, N., Chudowsky, N., Chudowsky, V., Joftus, S., & Zabala, D. (2006). From the capital to the classroom: Year 4 of the No Child Left Behind Act. Washington, D.C.: Center on Education Policy. Steinberg, M.P., & Donaldson, M.L. (2016). The new educational accountability: Understanding the landscape of teacher evaluation in the Post-NCLB era. Education Finance and Policy 11(3), 340-359. Strang, D. (1987). The administrative transformation of American education: School district consolidation, 1938-180. Administrative Science Quarterly, 32, 352-366. Straus, A.L., & Corbin, J.M. (1990). Basics in qualitative research: Grounded theory procedures and techniques. Thousand Oaks, CA: Sage. Strauss, J.E. (1998). Women in Congress: The difference they make. (Unpublished doctoral dissertation). Northwestern University, Chicago, IL. Sunderman, G.L. (2001). Accountability mandates and the implementation of Title I schoolwide programs: A comparison of three urban districts. Educational Administration Quarterly, 37(4), 503-532. Sunderman, G.L., & Orfield, G.O. (2006). The unraveling of No Child Left Behind: How negotiated changes transform the law. Cambridge, MA: The Civil Rights Project at Harvard University. Swain, C.M. (1993). Black faces, black interests: The representation of African Americans in Congress. Cambridge, MA: Harvard University Press. Sweet, L. (2001). Telephone interviewing: Is it compatible with interpretive phenomenological research? Contemporary Nurse, 12, 58-63. Tarr, G.A., & Williams, R.F. (Eds.). (2005). State constitutions for the twenty first century: Volume 1: The politics of state constitutional reform. New York, NY: State University of New York Press. Taylor, J., Stecher, B., O’Day, J., Naftel, S., & Le Floch, K.C. (2010). State and local implementation of the No Child Left Behind Act: Volume IX—Accountability under NCLB: Final Report. Washington, D.C.: U.S. Department of Education, Office of Planning, Evaluation and Policy Development, Policy and Program Studies Service. Theobald, N.D., & Malen, B. (Eds.). (2000). Balancing local control and state responsibility for K-12 education: 2000 Yearbook of the American Education Finance Association. New York, NY: Routledge. Thomas, S. (1991). The impact of women on state legislative policies. Journal of Politics, 53(4), 958-976. 353 Thomas, S. (1994). How women legislate. New York, NY: Oxford University Press. Thomas, S., & Welch, S. (1991). The impact of gender on activities and priorities of state legislators. Western Political Quarterly, 44(2), 445-456. Thompson, F.J., & DiIulio, J.J. (Eds.). (1998). Medicaid and devolution: A view from the states. Washington, D.C.: The Brookings Institution. Tolchin, M., & Tolchin, S.J. (2011). Pinstripe patronage: Political favoritism from the clubhouse to the White House and beyond. Boulder, CO: Paradigm Publishers. Torraco, R.J. (2005). Writing integrative literature reviews: Guidelines and examples. Human Resources Development Review, 4(3), 356-367. Tourangeau, R., Rips, L.J., & Rasinski, K. (2000). The psychology of survey response. Cambridge, UK: Cambridge University Press. Tractenberg, P. (2006). Education. In G.A. Tarr, & R.F. Williams (Eds.). State constitutions for the twenty-first century, Volume 3. New York, NY: SUNY Press. United States Census Bureau Current Population Survey [CPS]. (2015a). Annual social and economic supplement. Retrieved from https://www.census.gov/hhes/socdemo/education/data/cps/ United States Census Bureau CPS. (2015b). Educational attainment in the United States: 2015. Retrieved from https://www.census.gov/content/dam/Census/library/publications/2016/demo/p20578.pdf United States Department of Education [USDOE]. (2009, November). Race to the Top program executive summary. Washington, D.C.: Author. USDOE. (2012, June 7). ESEA Flexibility. Washington, D.C.: Author. USDOE. (2014, February). Projections of education statistics to 2022: Forty-first edition. Washington, D.C.: Author. U.S. News and World Report. (2017). Best education schools. Retrieved from https://www.usnews.com/best-graduate-schools/top-education-schools/edu-rankings Ulsaner, E.M., & Weber, R.E. (1979). U.S. state legislators’ opinions and perceptions of constituency attitudes. Legislative Studies Quarterly, 4(4), 563-585. Van Herk, H., Poortinga, Y.H., & Verhallen, T.M.M. (2004). Response styles in rating scales: Evidence of method bias in data from six EU countries. Journal of Cross-Cultural Psychology, 35(3), 346-360. 354 vanTeijlingen, E.R., & Hundley, V. (2001). The importance of pilot studies. Social Research Update, 35. Retrieved from http://sru.soc.surrey.ac.uk/SRU35.html Verba, S., & Nie, N.J. (1972). Participation in America: Political democracy and social equality. New York, NY: Harper and Row. Vernez, G., Naftel, S., Ross, K.E., Le Floch, K.C., Beighly, C., Gill, B., Birman, B., Garet, M.S., & O’Day, J. (2009). No Child Left Behind educational options: Availability expands, but participation remains low. Santa Monica, CA: RAND. Vuong, Q.H. (1989). Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica, 57(2), 307-333. Wang, W., & Famoye, F. (1997). Modeling household fertility decisions with generalized Poisson regression. Journal of Population Economics, 10(3), 273-283. Wargo, M.J., Tallmadge, G.K., Michaels, D.D., Lipe, D., & Morris, S.J. (1972). ESEA Title I: A reanalysis and synthesis of evaluation data from fiscal year 1965 through 1970. Palo Alto: CA: American Institutes for Research. Washington Research Project and NAACP Legal Defense and Education Fund. (1969). Title I of ESEA: Is it helping poor children? Washington, D.C.: U.S. Department of Health, Education & Welfare, Office of Education. Weatherly, R., & Lipsky, M. (1977). Street-level bureaucrats and institutional innovation: Implementing special-education reform. Harvard Educational Review, 47(2), 171-197. Weber, R.E. (1980). Gubernatorial coattails: A vanishing phenomenon. State Government, 53, 153-156. Weick, K. (1976). Educational organizations as loosely coupled systems. Administrative Science Quarterly, 21, 1-9. Wiese, A., & Garcia, E.E. (1998). The Bilingual Education Act: Language minority students and equal educational opportunity. Bilingual Research Journal, 22(1), 1-18. West, R. F., Toplak, M. E., & Stanovich, K. E. (2008). Heuristics and biases as measures of critical thinking: Associations with cognitive ability and thinking dispositions. Journal of Educational Psychology, 100, 930-941. Whittemore, R., & Knafl, K. (2005). The integrative review: Updated methodology. Journal of Advanced Nursing, 52(5), 546-553. Wilson, S.M., & Floden, R.E. (2001). Hedging bets: Standards-based reform in classrooms. In S.F. Fuhrman (Ed.), From the capital to the classroom: Standards-based reform in the states (pp. 193-216). Chicago, IL: National Society for the Study of Education. 355 Wong, K. (2015). Federal ESEA waiver as reform leverage: Politics and variation in state implementation. Public: The Journal of Federalism, 45(3), 405-426. Wong, K.K., Anagnostopoulos, D., Rutledge, S., Lynn, L., & Dreeben, R. (1999). Implementation of an educational accountability agenda: Integrated governance in the Chicago public schools enters its fourth year. Chicago, IL: Department of Education and Irving B. Harris Graduate School of Public Policy Studies, The University of Chicago. Wrabel, S., Saultz, A., Polikoff, M., McEachin, A., & Duque, M. (2016). The politics of the ESEA waivers. Educational Policy. Wright, J.R., & Niemi, R.G. (1983). Perceptions of issue position. Political Behavior, 5, 209223. Wright, S.P., Horn, S.P., & Sanders, W.L. (1997). Teacher and classroom context effects on student achievement: Implications for teacher evaluation. Journal of Personnel Evaluation in Education 11, 57-67. Yell, M. L., & Rozalski, M. E. (2000). Searching for safe schools: Legal issues in the prevention of school violence. Journal of Emotional and Behavioral Disorders, 8(3), 187-196. Yin, R.K. (2009). Case study research: Design and methods. Thousand Oaks, CA: Sage. Young, I.M. (1997). Deferring group representation. In I. Shapiro & W. Kymlicka (Eds.), Ethnicity and group rights. New York, NY: New York University Press. Young, T.V., Wang, Y., & Lewis, W.D. (2016). Explaining how political actors gain strategic positions: Predictors of centrality in state reading policy issue networks. Educational Policy, 30(6), 799-819. Youngs, P. Odden, A. & Porter, A.C. (2003). State policy related to teacher licensure. Educational Policy, 17(2), 217-236. Zimmer, R., Gill, B., Razquin, P., Booker, K., Lockwood, J.R., Vernez, G., Birman, B.F., Garet, M.S., & O’Day, J. (2007). State and local implementation of the No Child Left Behind Act: Volume I—Title I school choice, supplemental educational services, and student achievement. Washington, D.C.: U.S. Department of Education, Office of Planning, Evaluation and Policy Development, Policy and Program Studies Service. 356