Navigating Term Limits: The Careers of State Legislators 3031394224, 9783031394225

Citation preview

Navigating Term Limits The Careers of State Legislators Jordan Butcher

Navigating Term Limits

Jordan Butcher

Navigating Term Limits The Careers of State Legislators

Jordan Butcher Department of Political Science Arkansas State University Jonesboro, AR, USA

ISBN 978-3-031-39422-5 ISBN 978-3-031-39423-2 (eBook) https://doi.org/10.1007/978-3-031-39423-2 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Paper in this product is recyclable.

For Cole And for all of the lawmakers who so willingly welcomed me in their offices and shared their experiences with me.

Preface

While in my first year of graduate school, I was casually asked, “what do you know about term limits?” My response was simply that I did not know much and so I was instructed to do some looking. What originally started off as a summer research project quickly grew into a research agenda dedicated to exploring the differences between states with term limits and those without. There is a wealth of knowledge that exists about term limits, their history, passage, and early implications. Even though there was once a great deal of research surrounding term limits, it became clear that very little is known about term limits and how they function in the legislature today. Yet, there continues to be a push for term limits and further restrictions on our lawmakers. As it turns out, I did much more than casually look and even wrote my dissertation on the subject. From all of my research about term limits, the most intriguing component has always been legislative careerism. As my work developed, I found that the crux of term limits was their effect on legislative careers. The key behind term limits was to limit legislative careerism, so why are they not assessed through this lens? While I have spent a great deal of time studying state legislatures and term limits broadly, there has always been a need for a narrower focus on legislative careers. Other consequences of term limits mattered more or less based on how careers were shaped under those limits. The further I explored this career-based outcome, the more there was to uncover. This book is a result of that exploration. In the first half, I put long-standing theories

vii

viii

PREFACE

to the test by examining the direct influence of term limits on legislative careers. Following that, I look at the implications of term limits on two key roles that legislators have, committee work and agency oversight. Undoubtedly, the best part of this research process was interviewing 101 different legislators from 4 different states. I owe them all a debt of gratitude, their comments and subtleties gave me a depth of knowledge about the implications of term limits beyond what any reading from my desk could do, and it is my aim that this comes through on the page. The four weeks, in total, that I spent interviewing lawmakers provided an invaluable opportunity to simply be around each capitol and pick up on cues and dynamics both within the legislature and between members. This research presents a small snapshot of what I learned from those members. Looking back, I realize that my early response to knowing nothing about term limits was rather naive. In fact, it seemed that everyone else I came into contact with had an opinion, my friends and colleagues, the students in my classroom, and even lawmakers themselves. I uncovered a love/hate dichotomy with term limits and how people viewed them. The more I learned, the more fascinated I became with term limits and their role in state government. The existence of term limits is no longer merely “good” or “bad” as there is much more to uncover. It is my hope that this research helps to answer some of the questions that circulate around the term limit debate. I have enjoyed writing this book just as I have enjoyed the research. My goal in writing this was to simply share the facts and information about what is happening in the states. This work is for students, scholars, and practitioners; so that they may use this information to form their own opinions of term limits and carry on the research of one of the most unique and beguiling policies in our state legislatures today. Jonesboro, USA

Jordan Butcher

Acknowledgments

There are so many people to thank for this project. First, I want to thank my current institution, Arkansas State University, for believing in this project. I also want to thank my graduate institution, the University of Missouri, for giving me the opportunity to do this research. As well as, the Kinder Institute on Constitutional Democracy and the Institute for Humane Studies for their financial support during my dissertation research. And last, Drury University, which gave me such a strong foundation in the discipline. I want to thank my dissertation committee for their support of this project: James Endersby, Marvin Overby, and Mitchell McKinney. And Peverill Squire, my chair and advisor, who served as a sounding board throughout this entire project and without whom I would not be where I am today. I want to thank a number of other mentors and colleagues who offered help in various capacities: Laron Williams, Jonathan “Vanya” Krieckhaus, Jennifer Selin, Hanna Brant, Cody Drolc, and John Larrick. A special thank you to Marjorie Sarbaugh-Thompson for her advice and encouragement on studying term limits. I would be remiss to not mention some of the excellent student researchers that have hopped on board with this work throughout the years: Noah Haynes, Brandon Turnbull, Morgan Hibbard-Gregg, Adeola Adefelu, Drake Hopkins, and Joel Deere. Morgan Hanley deserves the highest praise for hand-recording every interview that I conducted. This work would not be the same

ix

x

ACKNOWLEDGMENTS

without her. Last, a thank you to all the friends who have offered feedback in conferences and workshops over the years. Without the support of a strong community, this book would not have happened.

Contents

Part I Introducing Term Limits 1

Term Limits and Their Many Theories Data Datasets Legislative Interviews Goals of the Book Looking Forward Bibliography

3 6 7 11 12 14 16

2

The Push and Promise of Term Limits Pushing Term Limits Passing Term Limits Types of Term Limits Unified/Lifetime Limits Oklahoma California Arkansas Divided/Lifetime Limits Michigan Missouri Nevada Unified/Consecutive Limits Nebraska

21 22 26 30 34 34 35 35 36 36 36 37 37 37

xi

xii

CONTENTS

Divided/Consecutive Limits Arizona Colorado Florida Maine Montana Ohio South Dakota Louisiana States Without Term Limits Bibliography

38 38 39 39 40 41 41 42 42 43 45

Part II Changing Careerism 3

Legislative Careerism and Growing Ambition Understanding Ambition Analysis Ambition Pathways Discrete Ambition Static Ambition Progressive Ambition Ambition in Practice Unified/Lifetime Unified/Consecutive Divided/Lifetime Divided/Consecutive Summary Bibliography

4

Declining Turnover and Legislative Leave Turnover and Legislative Leave Analysis Testing Leave Turnover and Leave in the States Unified/Lifetime Unified/Consecutive Divided/Lifetime Divided/Consecutive Summary Bibliography

51 53 58 60 66 71 74 80 81 82 83 84 85 87 89 91 94 100 107 110 111 112 113 114 116

CONTENTS

5

Legislative Leadership Testing Legislative Leadership Experience Apprenticeships Tenure Data and Methods Experience Apprenticeships Tenure Leaders in the States Divided/Lifetime Unified/Consecutive Unified/Lifetime Divided/Consecutive Summary Bibliography

xiii

119 120 122 123 124 125 127 134 138 142 144 146 148 150 151 154

Part III The Implications of Term Limits 6

Changes to Internal Deliberation: Committees The Importance of Committees Committee Changes Under Term Limits Policy Deliberation Gatekeeping Committees in the States Summary Bibliography

159 160 162 164 171 179 187 189

7

Changes to External Relations: State Agencies Legislative Oversight Term Limits and Oversight A Growing Dependency Implications for Bureaucratic Relations Agency Monitoring Agency Reliance Legislators and Agencies Summary Bibliography

193 195 196 198 199 200 203 206 213 215

xiv

8

CONTENTS

Conclusions Reevaluating the Promise of Term Limits Testing Term Limits Navigating Term Limits Bibliography

219 220 221 227 230

Appendix A: State Term Limit Laws

233

Appendix B: Additional Tables

241

Appendix C: Interview Questions

247

Index

255

List of Figures

Fig. 2.1 Fig. 2.2 Fig. 3.1 Fig. 4.1 Fig. 5.1 Fig. 5.2 Fig. 5.3 Fig. 5.4 Fig. 6.1 Fig. 6.2 Fig. 6.3 Fig. 6.4 Fig. 6.5

States with term limits prior to 2022 Different types of term limits Predicted percentage of pursuing each of the three types of ambition Comparing the predicted rate of natural turnover, by chamber Predicted level of experience of the majority party for the lower chamber Predicted level of experience of the majority party for the upper chamber Predicted probability of serving an apprenticeship, by leadership position Predicted Leadership Tenure for Both Legislative Chambers, Pre- and Post-Term Limits Probability of amending a bill on the floor, by chamber Probability of a bill dying in an Education Committee, by chamber Probability of chair committee control responses by experience across states Probability of committee conflict responses by experience across the states Probability of committee reliance responses by experience across the states

28 31 66 99 132 132 136 141 171 179 183 184 185

xv

xvi

LIST OF FIGURES

Fig. 7.1 Fig. 7.2

Predicted time spent monitoring state agencies based on member experience Predicted reliance on state agencies for information based on member experience

210 211

List of Tables

Table 1.1 Table 1.2 Table 1.3 Table Table Table Table

1.4 1.5 2.1 3.1

Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 3.7 Table 4.1 Table 4.2

Examples of single-state case studies on term limits Summary of legislators included in TLS-Careers, by state The number of top leaders in the leadership data, by chamber The number of legislative bills included from each state Characteristics of the interview respondents by state An overview of the 17 states used in this study Defining the different types of ambition in term-limited states Logit analysis for each type of political ambition Multinomial logit of discrete ambition for each chamber (base, reelection) Multinomial logit of static ambition for each chamber (base, reelection) Multinomial logit of progressive ambition for each chamber (base, reelection) Summary of the statistically significant career ambition findings by term limit type Members who feel pressured for time because of term limits Comparing the average rate of total turnover and natural turnover, by chamber Comparing the average rate of total turnover and natural turnover, by chamber (2002–2018)

5 8 9 10 12 23 55 62 68 72 76 79 81 96 98

xvii

xviii

LIST OF TABLES

Table 4.3 Table 4.4 Table 4.5 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table Table Table Table Table

6.6 7.1 7.2 7.3 7.4

Table 8.1 Table B.1 Table Table Table Table

B.2 B.3 B.4 B.5

Multinomial logit analysis of leave decisions in the lower chamber Multinomial logit analysis of leave decisions in the upper chamber (base, staying in office) Sub-sample of state legislators and what they do upon leaving, before and after 2000 Average years of experience for leaders in each state, preand post-term limit TSCS negative binomial regression model of legislative experience post-term limits Logit analysis of leaders who serve apprenticeships post-term limits Average terms of tenure for each state, pre and post-term limits Negative binomial regression model for leadership tenure, post term limits The most influential members in term-limited states Cox estimates of policy deliberation for the lower and upper chambers Cox estimates of policy deliberation within education and Ways and Means committees Logit analysis of amended legislation, by chamber Negative binomial regression models of legislative gatekeeping, count of bills passed Logit analysis of whether bills die in committee, by chamber Ordered logit analyses of committees, scaled response How much time legislators spend on agency monitoring Do legislators rely on outside sources for information? How much legislators rely on state agency workers Ordered logit analysis of legislative monitoring and reliance Testing the many theories of term limits Top positions in state legislative chambers and how they are filled Additional leadership salaries in state lower chambers Additional leadership salaries in state upper chambers Breakdown of interview responses for committees, by state States included in the sub-sample committee analyses

102 105 108 128 130 135 139 140 144 167 169 170 174 177 181 201 203 204 208 223 242 243 244 245 245

PART I

Introducing Term Limits

CHAPTER 1

Term Limits and Their Many Theories

“Moderate the expectation with reality... balance eagerness with how things work” —Ohio Representative (Republican)

Support for legislative term limits has not waned since they first became popular in the early 1990s. Each year a handful of states see new proposals for term limits. For example, North Dakota voters passed a term limit amendment in the 2022 election (Minot Daily News 2022). In September 2019, the New York State Assembly saw a proposal for a four-year term limit (Rochester First 2019), the most stringent term limit to date. At the same time, the Unite for Term Limits organization began to collect signatures for a new proposal in Utah (The Salt Lake Tribune 2019). In March 2018, there was an attempt to pass term limits in the Alabama legislature (Associated Press 2018b). Meanwhile, in April of 2018, Missouri proposed a constitutional amendment to alter the structure of its preexisting term limits (Associated Press 2018a). In the fall of 2020, Arkansas voted to change their term limits again, and they are now on their third plan. Even at the federal level, politicians continue to advocate for congressional term limits. From the former president, Donald Trump, to Florida Senator Rick Scott, term limits remain a popular policy among politicians and citizens (White 2018). The articles in favor of term limits– the studies, the polls, and the advocates–could go on for this entire © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_1

3

4

J. BUTCHER

chapter, but the point is clear that Americans are largely in favor of term limits and the debate over their merit is not going away any time soon. Past arguments in favor of term limits are not dissimilar to current justifications. People tend to agree with term limits because of the proposed changes that coincide with their implementation. Term limits were introduced to the public as a fix for politicians that had become too far removed from constituents. As legislatures became more professional, there was increased capacity and a greater incentive for legislators to serve longer. In other words, term limits were a solution to the careerism that was caused by professionalization (Mooney 2007). Term limits were successful among the public because they promised to remove career politicians (Moncrief et al. 2007). Current advocates for term limits say they are “frustrated with politicians and entrenched power brokers” (Washington Post 2019). Others speak about an “abuse of power” from politicians (Rochester First 2019). Term limits remain popular among the public regardless of state or time. As soon as term limits passed, scholars were already noting their importance. Those who evaluated term limits concluded, “it is no overstatement to say that the term-limits movement... is emerging as one of the most important political developments in this nation in a very long time” (Crane and Pilon 1994, 1). Political scientist Mark Petracca went so far as to say, “term limitation is undoubtedly the twentieth century’s most popular institutional reform as evidenced by both election results and opinion polls” (1994, 57). At the time, national polling showed between 75% and 80% support for term limits (Jacob 1994). Alan Rosenthal summarized that with term limits, “the negative effects are barely perceptible to the naked eye…(and) the odds are that public support for term limits will not erode much” (2007, 222). Business Insider reported just that in 2022, 84% of respondents wanted term limits for the House of Representatives and 83% supported term limits for Senators. Given that there continues to be far-reaching support for term limits, with new proposals regularly, it is important to understand what has occurred in the states with term limits. Rather than focusing on the positive or negative effects of term limits, this study evaluates the predicted changes to legislative careers and the implications of those changes. Evaluating the long-term changes to legislative careers reveals the stringency that term limits bring to the legislature. Institutions are sticky, meaning they experience little change throughout time. The authority of an institution remains with the office, not with the individual people who

1

TERM LIMITS AND THEIR MANY THEORIES

5

Table 1.1 Examples of single-state case studies on term limits State

Prior research

Arizona Arkansas California

Berman (2007) English and Weberg (2007) Cain et al. (2007), Masket and Lewis (2007), Cain and Kousser (2004), Clucas (2003) and Capell (1996) Straayer (2003, 2007) Powell and Jones (2007), Moen et al. (2005) and Moen and Palmer (2003) Sarbaugh-Thompson and Thompson (2017), Sarbaugh-Thompson et al. (2004) and Penning (2003) Farmer and Little (2007) Farmer (2007)

Colorado Maine Michigan Ohio Oklahoma

serve in that office. As such, that authority remains even though people come and go. From the rules in place to the staff, institutions are meant to withstand internal changes as people enter and exit. However, term limits altered the rules of service. As a result, term limits changed the internal structure of the legislature and limited the control that members have over their own careers. While term limits remain a popularly debated issue, the lasting effects on careerism remain unexplored. Early studies focused on the initial effects of term limits and tended to be singular in nature. Table 1.1 provides examples of some of these case studies. In this table, there are 16 different studies about eight individual states. While these case studies increase knowledge about term limits in some states, half of the states have received far less attention. These studies provide the building blocks for understanding the potential changes brought by term limits and reveal some of the changes that occurred within the legislature. The question that remains is, are the changes observed in the case studies attributed to term limits, or are they natural shifts that other legislatures underwent? More recently some scholars have started to examine the broader consequences of term limits. These studies explore the electoral consequences of term limits and how the dynamic of constituents and lawmakers has shifted (Olson and Rogowski 2020). As well as, the dynamic of how a term-limited legislature has altered the relationship between legislators and lobbyists (Strickland and Crosson 2022). There is also an in-depth examination of campaign financing from Lynda Powell,

6

J. BUTCHER

where she examines some of the underlying reasons for why lawmakers do the things that they do (Powell 2012). Although these works have significantly contributed to our knowledge of term limits and their implications, there is a need for understanding underlying career shifts. There are two key components that make this a unique endeavor. First, this is the only comparative approach assessing the majority of states with legislative term limits since their implementation. Second, this examination is the first to capitalize on the unique differences that exist between the types of term limits. In particular, this research demonstrates how the types of term limits result in different behaviors and trends among the members who serve. In addition to comparing 12 of the term-limited states to each other, there are 3 states that do not have term limits, and 2 states that passed term limits but no longer have them. These 17 states are examined collectively in order to understand the effects attributed to term limits. Term limits have long been treated as binary, they do or do not exist. In all other regards, states are treated as being on a spectrum, distinct from each other. This study emphasizes how each type of limit shapes the legislature differently and, as a result, how the implications are felt to varying extents. There are different kinds of term limits. Each comes with its own rules and various nuances, but that is what makes each individual limit important. When treating all term limits as one, there is a great deal of variation that is lost and as a result, it is easy to over- or underestimate the effects of term limits. This comparative analysis is achieved using a collection of original data. In order to assess the effects of term limits and how they vary by state, data was collected on legislator careers, legislative leaders, and committees. Additionally, this research capitalizes on 101 in-person legislator interviews from 4 term-limited states. To this point, there has been no research to this extent on term-limited legislatures, much less about careers. Focusing on the long-term effects of term limits gives insight into legislative institutions and their invariable structure. Most states with term limits have had 30 years to reconcile this great institutional change.

Data For this study, I present several new datasets meant to capture the long-term effects of term limits. Unlike early studies of term limits, these data are not limited to a single cohort of legislators or a single

1

TERM LIMITS AND THEIR MANY THEORIES

7

legislative session. Instead, these data track changes throughout several decades. These data represent different perspectives of the state government for each of the 17 states in this study. There are three original datasets collected from state and legislative records. The datasets include legislative careers, legislative leadership, and legislative bills.1 While the compiled data are original, they are supported by existing information including Squire’s Legislative Professionalization Index (Squire 2017), and Career Opportunity Structures (Squire and Moncrief 2010). These quantitative data are supplemented with interviews from four termlimited legislatures. The interviews contain both open- and close-ended questions allowing for quantitative and qualitative assessments. This mixed-methods approach helps to create a more complete understanding of term-limited legislatures and how they function. Datasets There are four original datasets in this study. The first dataset, TLSCareers, contains all members who have served in the state legislature from the early 1990s through the 2020 election in total there are over 9,500 individual observations. These data track the career movements of legislators from the 17 states and were collected from state legislative records, election records, secretary of state offices, legislative archives, newspapers, obituaries, and more. Table 1.2 illustrates some of the key demographics in the TLS-Careers dataset by state. The table includes all individuals in the dataset for each state, in each chamber. In addition, the table presents the percent of Republican, Male, White, and Black legislators who serve. In all but two states, the majority of members representing each state are from the Republican Party. Males are also more common in the dataset. In the lowest state, Colorado, 64% of the legislators represented in the data identify as male, Oklahoma is the highest with 83% of legislators identifying as male. There is greater racial variation; the states range in the percent of white members from only 44% in Ohio to 99% in North Dakota.2 1 The Legislative Bills raw data are from LegiScan but have been restructured. 2 A member’s race is often difficult to confirm in the context of historical data and not

wanting to mislabel someone, while there are hopes of updating this information in the future it is worth noting that there is missing data here, primarily for Florida, Maine, and Ohio.

8

J. BUTCHER

The percent of Black legislators ranges from 0 (in two states) to 15% in Louisiana. The sub-sample of the 17 states used to test legislative term limits, is representative of the over 7,000 state legislators who serve in any given year. While women and Black lawmakers are underrepresented in reference to the national population, these data are representative state legislatures as a whole. There are only two states that have more Democratic members in the dataset than they do Republican members, but most states were between 50 and 65% Republican and 35–50% Democrat at the time. The Republican lean is expected given the states that have term limits. What is important to note is that the data collected from the states without term limits is similar to the states that do have term limits. The second dataset, TLS-Turnover, is a collection of information from a variety of sources that encompasses information on legislative turnover for each state. The base turnover numbers are from the Council of State Table 1.2 Summary of legislators included in TLS-Careers, by state State

Lower chamber

Arizona 325 Colorado 345 Florida 568 Louisiana 313 Maine 857 Michigan 608 Missouri 727 Montana 521 Nebraska – Ohio 465 Oklahoma 331 South 398 Dakota Non-Term-Limited New Jersey 309 North 247 Dakota Oregon 294 Texas 510 Wyoming 274

Upper chamber

Republican legislators

Male legislators

White legislators

150 166 159 111 198 149 143 218 163 141 150 199

58.5% 51.7 62.7 53.6 45.0 53.5 59.8 59.0 63.2 62.4 62.9 74.0

64.4% 64.2 72.6 78.1 69.1 74.8 76.9 73.3 75.5 75.6 83.8 79.7

72.4% 55.2 54.9* 56.4 86.6* 83.1 88.9 88.8 93.9 44.7* 74.2 96.8

1.3% 2.9 12.4 15.6 0.38 12.7 8.5 0 3.1 7.1 2.9 0.17

117 132

48.4 68.1

76.8 79.7

79.1 99.5

13.1 0

115 81 100

50.4 55.5 75.1

67.7 78.0 78.9

87.3 71.7 95.5

3.4 6.1 0.8

*Denotes significant amount of missing data

Black legislators

1

TERM LIMITS AND THEIR MANY THEORIES

9

Governments’, Book of the States, the specific information on the number of lawmakers termed out from each state comes from TLS-Careers. This dataset also incorporates specific demographic turnover collected from state legislative websites and national and state caucuses. The third original dataset, TLS-Leaders, contains observations of leaders from the same 17 states. This dataset includes top legislative leadership positions in each state. There is a great deal of variation in both the number and title of positions; eight positions from each chamber are included in the dataset.3 These particular positions are the most consistent among the states. Not all states have each of these leadership positions; rather, some only have one top position. Other states have numerous positions with the same title, for example, it is not uncommon for a state to have multiple whips. In total, there are nearly 2,500 individual leaders included from 1990 to 2020.4 Table 1.3 contains the breakdown of the number of different leaders by chamber for all 17 states. The fourth dataset, Legislative Bills, contains information on legislation to understand the role of committees. The raw data are publicly available from LegiScan.5 LegiScan presents the data by state for each year, making comparisons across time and the states difficult. These data were compiled and reorganized in order to make comparisons among the states. These Table 1.3 The number of top leaders in the leadership data, by chamber

Position

Lower

Upper

Speaker/President Speaker/President Pro Tempore Majority Leader Assistant Majority Leader Majority Whip Minority Leader Assistant Minority Leader Minority Whip

186 150 186 139 206 157 148 151

124 206 155 107 119 157 129 117

3 Lower: Speaker, Speaker Pro Tem, Majority Leader, Assistant Majority Leader,

Minority Leader, Assistant Minority Leader, and Majority/Minority Whips; Upper: President/President Pro Tem, Majority Leader, Assistant Majority Leader, Minority Leader, Assistant Minority Leader, and Majority/Minority Whips. 4 Given that members sometimes serve in multiple leadership positions during their service, there is the possibility of repeat individuals. 5 The raw data can be found here https://legiscan.com/datasets.

10

J. BUTCHER

data track bills through the legislature to see how long it takes a bill to get through the legislative process, if it gets through at all. Due to the complex nature of the bill data, this is the dataset that is most likely to have lost observations. The original form of these data are self-reported and are rather inconsistent across the states. This dataset is in its simplest form for ease of use and interpretation. The Legislative Bill data ranges from 2010 to 2018 and tracks each piece of legislation. In total, this dataset contains over 270,000 individual bills from the 17 states. Table 1.4 contains the breakdown of bills by state. While most states fall between 5,000 and 20,000 bills, New Jersey and Texas introduce significantly more bills because of their budgeting process. These data provide information on when bills were introduced and from what chamber. There is also information on how many times the bill was read, when it was read, whether or not it passed the legislature, and whether or not the governor signed it. The data does indicate executive vetoes and legislative overrides. Table 1.4 The number of legislative bills included from each state State

Total bills

Ways and means

Education

Arizona Colorado Florida Louisiana Maine Michigan Missouri Montana Nebraska Ohio Oklahoma South Dakota Non-Term-Limited New Jersey North Dakota Oregon Texas Wyoming

12,808 5,922 18,181 20,927 10,441 16,277 14,172 4,751 7,271 6,340 18,741 5,154

579 114 1,664 794 -

1,270 600 2,179 2,076 1,527 1,011 1,226 349 1,713 515

52,175 3,612 12,445 58,244 2,921

4,229 2,246 -

5,517 6 1,521 5,089 206

1

TERM LIMITS AND THEIR MANY THEORIES

11

In order to assess individual committees, there are sub-samples of bill data from the Ways and Means Committees and from Education Committees.6 These two committees were chosen because of their importance, popularity, and their differences. These two types of committees tend to lean different directions politically and Ways and Means Committees tend to attract more males while Education Committees attract more females. Moreover, Ways and Means Committees are of high value to members because of the prestige but Education Committee work is important because of the direct ties to constituents in every state. As seen in Table 1.4, Education Committees are one of the most consistent committees that exist among the states, whereas Ways and Means is not. Legislative Interviews In addition to the above datasets, this study utilizes face-to-face interviews with legislators who serve in term-limited states. Borrowing from Sarbaugh-Thompson et al. (2004), these interviews reveal information about legislative function, rules and norms, and the day-to-day life of a legislator serving under term limits. The interviews were conducted in 2019 in Missouri, Nebraska, Oklahoma, and Ohio. The four states used in this study cover the four different types of term limits. The types of term limits implemented in these states are vastly different, but allow for a comparison of how different types of term limits alter legislatures differently. Interviews were not conducted in states without term limits as it is hard to pinpoint the effects of something that does not exist; instead, I focus this analysis on the differences across the types of term limits that do exist. The questions used are a variant of the questions originally created by Sarbaugh-Thompson et al. (2004) for their study of the Michigan legislature. There are both open- and close-ended questions. A version of the interview questions can be found in Appendix C.7 As Lynda Powell uncovered, a lawmaker’s personal feelings are often a better measure of influence than what we as social scientists tend to infer from other 6 A list of the committees and states used in this additional analysis can be found in Table B.5 in Appendix B. 7 Nebraska has a slightly different format because they do not have party leaders and there is only one chamber.

12

J. BUTCHER

Table 1.5 Characteristics of the interview respondents by state

Lower Chamber Upper Chambera Democrat Republican Man Woman White Black Multiracial 1st Term 2nd Term 3rd Term 4th Term

Missouri

Nebraska

Ohio

Oklahoma

Total

28 11 11 28 28 11 37 2 0 13 11 6 9

– 15 7 8 9 6 14 1 0 11 4 – –

21 3 12 12 15 9 18 6 0 11 7 3 3

16 7 10 13 14 9 21 1 1 20 2 1 0

65 36 40 61 66 35 90 10 1 55 24 10 12

a Includes Nebraska

methodologies (Powell 2012). Table 1.5 contains a breakdown of some of the characteristics of the 101 members that were interviewed. All interviews occurred in person and took roughly 30 minutes each. All interviews were recorded by hand to allow for a more open dialogue with the respondent. Because of the hand-written notes some quotations used, even when they are in quotes, may be paraphrased. All respondents were guaranteed anonymity for their participation in this study; therefore, any identifying characteristics may be changed to preserve the anonymity of the particular legislator.

Goals of the Book Throughout this chapter, I have pointed to several works from other scholars while noting that term limits have been in place for nearly 3 decades. So, why do we need another study on term limits? The simple answer is time. The longer that term limits have been in place, the greater the observational power. With the additional time, there are several unique contributions to the understanding of term limits and state legislatures. The purpose of this research is not to explain why things are happening but to understand what is happening. This work is less focused on understanding the particular motivations of lawmakers, as that is a rather

1

TERM LIMITS AND THEIR MANY THEORIES

13

difficult task, but to understand what is occurring in the states. To this point, much of our current understanding of term limits is based on theoretical arguments that are now decades old. There is great power in observing how lawmakers behave in office because of the broad implications of their service. This work will provide an overview of how term limits have played out over the last 20+ years by focusing on a few important components of legislative careers. In order to better understand what is happening in the states, it is important to understand that states’ term limits differ in meaningful ways. States with term limits are frequently treated as the same; this mischaracterization of term limits implies that all states are influenced in the same manner (Lazarus 2006; Carey et al. 2000). This is not the case; each term limit type is unique and has a different level of influence on its legislature (Sarbaugh-Thompson 2010; Moncrief and Thompson 2001). More recently, scholars have illustrated the importance of differences among limits, noting that the strength and the type of limit matter and that the variation among states is important (Sarbaugh-Thompson 2010; Cain and Wright 2007). States with stricter limits see more changes; states with longer, more lax limits are less likely to experience any major changes. The last overarching goal of this research is to raise challenging questions. As a warning to the reader, not all of the findings in this text are intuitive. Rather, some results are quite counter to what one might expect. But that is the nature of this project, to understand how lawmakers have come to navigate term limits in the long-term by evaluating their primary target: careers. Term limits are not a one-size-fits-all type of policy and these results will reveal just that. If you are looking for a pro-term limit argument or contrarian takes, you will find neither. Instead, you will find the puzzle of term limits. Including odd results that may lead you to question their merits, and surprising findings that will leave you wondering why you do not support them. While this research is not theoretically nuanced beyond examining the different types of limits, at its core, this research is meant to inform the reader of how theory played out. The implications of this research extend beyond the state legislatures that have term limits. In fact, there are frequently calls for term limits on Congress and there are common discussions over the merits of term limits for federal court judges. Term limits, are already in place for many executives and city councils/boards. The way that we view term limits ought not be limited to the legislature but part of what is so great about our state governments is that they truly serve as laboratories of democracy.

14

J. BUTCHER

Looking Forward In this study, I use new data to assess term limits across the states in a way that has not been done to this point. Rather than relying on any one type of data, this study uses four unique datasets and interviews with legislators who serve under each of the four types of term limits. These data allow for a unique comparative perspective at both the individual and aggregate levels. Term limits, and the constraints they place on the legislature, are understudied. The presence of term limits are not simply an institutional constraint limiting a legislator to an X number of years. Term limits have serious institutional outcomes and reshape the very structure of state government. There is a great deal that can be said about term limits, but conclusions about the long-term effects and the unintended consequences remain unclear. The legislators who served before term limits were implemented are now gone and current members come in knowing that their time is limited. As a result, the legislature must adapt. The extent to which careers, legislative organization, and separation of powers are influenced will be dependent on the preexisting legislative structure and the stringency of the term limit. Term limits will not overrule preexisting institutional structures in all cases. Each legislature is a separate entity with its own rules, norms, and history. Some legislatures will be better equipped to withstand some of the effects of term limits, others will not. Moreover, some term limits will not be strict enough to bring much change to the legislature. There is no greater variation than the institutional constraint of term limits. Present in 15 states, with four different types of limits, there is a great deal to be examined. Chapter 2 presents an overview of the termlimited states. This chapter first presents the arguments for term limits and the promises that were made to voters. Understanding the logic behind term limits and the motivations helps to situate the remainder of this examination. Also in this chapter is a discussion of the different types of term limits and some of the stories of how states got their term limits. In Chapter 3, there is an overview of legislative careers and prior theories of political ambition. In line with Schlesinger (1966), there are three pathways a legislator can choose: discrete, static, or progressive. Term limits, however, create new pathways within each type of ambition. Overall, while members of term-limited legislatures are more progressively ambitious than legislators from states without term limits, the different

1

TERM LIMITS AND THEIR MANY THEORIES

15

pathways that a member chooses to follow are based on legislative incentives and term limit stringency. Much of the research on term limits has been dedicated to understanding how careers have changed, but as it turns out term limit proponents oversold the promise of a citizen legislature. Following the chapter on political ambition, Chapter 4 focuses on a different component of the legislative career, what about those who stay versus those who leave elected office? Lawmakers are becoming increasingly secure in their seats and tend to seek out reelection, leading to decreased turnover and retirements. This chapter looks at what occurs when lawmakers do not seek reelection either because they are forced out by term limits or because they retire. I present an overview of legislative turnover and how natural turnover has begun to decline in term-limited states. This analysis of turnover demonstrates that even under term limits, fewer members are voluntarily leaving office. This leads to the second analysis centered on a member’s decision to retire from office. With no longevity in term-limited states, it was presumed that lawmakers would simply leave office after a short tenure. I uncover that lawmakers are actually less willing to retire and choose to stay in office for as long as possible. Chapter 5 examines the role of leaders in state legislatures to see what has changed under term limits. Leaders are the central actors in state legislatures, as they often have the most experience. I test the long-standing theory that term limits would diminish leadership in state legislatures. This theory has merit, but only in part. This chapter evaluates the changes to the overall legislative experience, the use of apprenticeships, and the tenure of leaders. This chapter uncovers that a term-limited state’s leaders have lengthier tenure and a stronger leadership system in place because of the internal reliance on more senior members. Only in certain cases has the career of a leader diminished. In most states with term limits, the speaker and majority leader have taken on a great deal of power, because they tend to have the most experience. Part two is centered more on the substantive consequences of the changes to legislative careerism under term limits. Chapters 6 and 7 explore two key responsibilities of state legislators to evaluate whether or not term limits have shifted the job of a legislator or if they have learned to adapt. Chapter 6 is focused on legislative committees and the responsibility of lawmakers to create policy. Specifically, Chapter 7 is about the oversight responsibilities of state legislators and how lawmakers interact

16

J. BUTCHER

with state bureaucrats. These are not the only areas affected by changes in legislative careerism, but they are two key roles that a lawmaker in any state must fulfill making these chapters essential for understanding the consequences of term limits. Chapter 6 is focused on internal legislative deliberation and the role of committees in the legislative process. Early scholars expressed concerns that committees, due to time and lack of experience, would no longer be able to deliberate and process legislation efficiently. I examine the length of time it takes legislation to pass to better understand the deliberation process. Now that states have reached a point of equilibrium, committees have found a new system and way to deal with legislation. Focusing on the changes to external relations, Chapter 7 uncovers the changing relationship between legislatures and agencies. Specifically, the decreased monitoring of state agencies and the increased reliance on agencies for information. I rely on interviews from state legislators and how they view their own relationship with state agencies. Currently, legislators appear to rely heavily on bureaucrats without trusting them fully. This study gives insight into the lasting effects of term limits versus the initial implementation effects that have long been the center of scholarship. Importantly, this research emphasizes the variation among the states and how the characteristics of the legislature can alter the effects of term limits. This work is meant to provide a more in-depth look at how term limits look once they have stabilized. Understanding what occurs in termlimited legislatures can help to better inform governments, legislators, and citizens alike. Our knowledge of term limits can likely be extended to other areas of government that frequently hear calls for term limits such as Congress and the Supreme Court. This research attempts to present a balanced, unbiased view of term limits by simply testing preexisting theories of term limits in order to uncover the long-term consequences of term limits on the legislature.

Bibliography Associated Press. 2018a. Missouri constitutional amendment would change term limits. https://apnews.com/article/89a2059fc8004855a1790a5a25a 1a3f9 (accessed July 30, 2018). Associated Press. 2018b. Alabama Senate rejects term limits proposal. https:// www.al.com/news/2018/03/alabama_senate_rejects_term_li.html (accessed July 30, 2018).

1

TERM LIMITS AND THEIR MANY THEORIES

17

Berman, D. 2007. The effects of legislative term limits in Arizona: More churning, more chaos, and a diminished institutional role for legislators. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Cain, Bruce E., and Thad Kousser. 2004. Adapting to term limits: Recent experiences and new directions. Public Policy Institute of California. Cain, Bruce, and Gerald Wright. 2007. Committees. In Institutional change in American politics: The case of term limits, ed. Bruce Cain Karl Kurtz, and Richard G. Niemi, 73–89. Ann Arbor, MI: University of Michigan Press chapter 4. Cain, Bruce, T. Kousser, and K. Kurtz. 2007. Institutional imbalance: The effect of six year limits in California. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Capell, Elizabeth A. 1996. The impact of term limits on the California legislature: An interest group perspective. In Legislative term limits: Public choice perspectives, 67–85. Springer. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 2000. Term limits in state legislatures. University of Michigan Press. Clucas, Richard A. 2003. California: The new amateur politics, 17–32. Lanham, MD: Lexington Books. Crane, Edward H., and Roger Pilon. 1994. An American debate. In The politics and law of term limits, ed. Edward H. Crane. Cato Institute. English, Art, and B. Weberg. 2007. Term limits in the Arkansas General Assembly: A citizen legislature responds. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Farmer, Rick. 2007. The 2004 partisan transition in the Oklahaoma house and term limits. Oklahoma Politics 16: 19–46. Farmer, Rick, and T.H. Little. 2007. Legislative power in the Buckeye State: The revenge of term limits. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Hickey, Walt. 2022. Americans think the government is too old—And wide margins support term limits, age caps, and cognitive tests, an insider/morning consult poll finds. Jacob, Paul. 1994. From the voters with care. In The politics and law of term limits, ed. Edward H. Crane. Cato Institute. Lazarus, Jeffrey. 2006. Term limits’ multiple effects on state legislators’ career decisions. State Politics & Policy Quarterly 6 (4): 357–383. Masket, Seth E., and Jeffrey B. Lewis. 2007. A return to normalcy? Revisiting the effects of term limits on competitiveness and spending in California assembly elections. State Politics & Policy Quarterly 7 (1): 20–38. Minot Daily News. 2022. Poll: Voters back term limits (accessed October 7, 2022).

18

J. BUTCHER

Moen, Matthew C., and Kenneth T. Palmer. 2003. Maine: The cutting edge of term limits. In The test of time. Lexington Books. Moen, Matthew C., Kenneth T. Palmer, and Richard John Powell. 2005. Changing members: The Maine legislature in the era of term limits. Lexington Books. Moncrief, Gary., and Joel A. Thompson. 2001. On the outside looking in: Lobbyists’ perspectives on the effects of state legislative term limits. State Politics & Policy Quarterly 1 (4): 394–411. Moncrief, Gary, Linda W. Powell, and Tim Storey. 2007. “Composition of legislatures.” In Institutional change in American politics: The case of term limits, ed. Bruce Cain Karl Kurtz, and Richard G. Niemi, 22–37. Ann Arbor, MI: University of Michigan Press chapter 4 Mooney, C. 2007. Truncated careers in professionalized state legislatures. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Rochester First. 2019. New bill seeks to impose term limits for state legislahttps://www.rochesterfirst.com/news/archive-state-news/new-billture. seeks-to-impose-term-limits-for-state-legislature/ (accessed September 13, 2019). The Council of State Governments. Various Years. The book of the states. Lexington, KY: Council of State Governments. Olson, Michael P., and Jon C. Rogowski. 2020. Legislative term limits and polarization. The Journal of Politics 82 (2): 572–586. Penning, James M. 2003. “Michigan: The end is near.” In The test of time, ed. Rick Farmer, John David Rausch Jr., and John C. Green. Lexington Books. Petracca, Mark P. 1994. Restoring “the university in rotation”: An essay in defense of term limitation. In The politics and law of term limits, ed. Edward H. Crane. Cato Institute. Powell, L.W. 2012. The influence of campaign contributions in state legislatures: The effects of institutions and politics. Legislative Politics and Policy Making University of Michigan Press. https://books.google.com/books?id=7pZFDw AAQBAJ Powell, R.J., and R. Jones. 2007. Institutional change and legislative term limits in Maine.” In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Rosenthal, A. 2007. Living with term limits. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Sarbaugh-Thompson, Marjorie. 2010. Measuring term limitedness in US multistate research. State Politics & Policy Quarterly 10 (2): 199–217. Sarbaugh-Thompson, Marjorie, and Lyke Thompson. 2017. Implementing term limits: The case of the Michigan legislature. University of Michigan Press.

1

TERM LIMITS AND THEIR MANY THEORIES

19

Sarbaugh-Thompson, Marjorie, Lyke Thompson, Charles Elder, John Strate, and Richard Elling. 2004. The political and institutional effects of term limits. Springer. Schlesinger, Joseph A. 1966. Ambition and politics: Political careers in the United States. Rand McNally and Co. Squire, Peverill. 2017. A Squire index update. State Politics & Policy Quarterly 17 (4): 361–371. Squire, Peverill, and Gary F. Moncrief. 2010. State legislatures today: Politics under the domes. Rowan & Littlefield. Straayer, J. 2007.Colorado legislative term limits: The worst of both worlds. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Straayer, John A. 2003. Colorado: Lots of commotion, limited consequences. In The test of time. Lexington Books. Strickland, James M., and Jesse M. Crosson. 2022. K Street on main: Legislative turnover and multi-client lobbying. In Political Science Research and Methods, 1–17. The Salt Lake Tribune. 2019. Signature gathering starts for Utah term limits initiative. https://www.sltrib.com/news/politics/2019/09/11/signat ure-gathering/ (accessed September 13, 2019). Washington Post. 2019. Virginians like term limits—But their representatives https://www.washingtonpost.com/opinions/local-opinions/virgin don’t. ians-like-term-limits--but-their-representatives-dont/2019/08/30/dfe5b1c8c915-11e9-be05-f76ac4ec618c_story.html (accessed September 13, 2019). White, Gary. 2018. Gov. Rick Scott pitches term limits, longer congressional schedule during US Senate stump speech in Lakeland.

CHAPTER 2

The Push and Promise of Term Limits

I look forward to term limits... there will be qualified people who come after me. —Nebraska Senator. (Republican)

The pathway to state legislative term limits has been one of learning with numerous successful implementations, several failures, and for some, changes to term limits as they have grown to better understand the implications. Since their passage states have worked to find their new sense of normal for both the legislature and legislators. Each state, with unique politics and diversified institutional resources, will feel the effects of term limits to varying extents. This overview is meant to be a reference point for understanding the term-limited states and the legislators who serve in them. Table 2.1 contains an overview of the 12 term-limited and 5 nonterm-limited states that are used in this study (this excludes Arkansas, California, Nevada, and as of 2023 North Dakota). The table illustrates the differences among the term-limited states while showing the institutional similarities to the states without term limits. The table displays the size of each chamber in the legislature (lower/upper) and the term length for each chamber (lower/upper). Included are two measures of legislative institutions to aid in understanding the differences among the states. First, the professionalization ranking from Squire (2017), as well as the career © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_2

21

22

J. BUTCHER

opportunity categorization from Squire and Moncrief (2010). Professionalization serves as a measure of institutional resources; this ranking often equates to overall legislative capacity and the incentive for members to serve. Career opportunities are a measure of a member’s career prospects in this instance; members serve in legislatures that will either help them to advance to higher office or in dead end states which offer long-term career prospects (as an elected official). Finally, for the term-limited states, the enactment and impact date are listed as well as the limit length (lower/ upper) and term limit type. The similarities and differences between the term limit types will become more noticeable as the point of emphasis changes with each chapter. Term limits are typically treated as dichotomous, a legislature either does or does not have term limits. The purpose of this chapter is to explore term limits beyond their binary existence in order to better understand how term limits have the potential to affect each state legislature and the legislators who serve. States that have term limits are all distinct from one another, aside from the type of term limit; each legislature institutes its own rules and standards. The institutional differences among the legislatures can affect how much influence term limits have on members. This chapter outlines the successes and failures of term limit legislation as well as the types of term limits. In this study, I focus on the comparison of these 12 term-limited states with 5 states that do not currently have term limits. By comparing the 17 states, this study reveals two important aspects relevant to the term limit debate. First, term-limited states are inherently different from each other. Second, by comparing states with and without term limits, the potential differences caused by term limits can be parsed from naturally occurring change.

Pushing Term Limits Term limits were meant to remedy the distrust that citizens felt about selfish, disconnected politicians. Fears grew from the publicity around rampant corruption prevalent in states like Arizona and California where various schemes and deals furthered public distrust around politicians. One such scandal was the AZscam in Arizona, which resulted in FBI raids and numerous prosecutions (Berman 2007). The consensus among the public was that legislators were bad and because of the strong incumbency advantage could not be voted out. Public opinion polls found that when thinking about politicians, voters felt “frustration… dissatisfaction… lack

2

THE PUSH AND PROMISE OF TERM LIMITS

23

Table 2.1 An overview of the 17 states used in this study State legislature

Term

Professionalization Career

Size* Length* Ranka Arizona (AZ)

60/ 2/2 30 Colorado 65/ 2/4 (CO) 35 Florida (FL) 120/ 2/4 40 Louisiana (LA) 105/ 4/4 39 Maine (ME) 151/ 2/2 35 Michigan 110/ 2/4 38 (MI)d Missouri 163/ 2/4 (MO) 34 Montana 100/ 2/4 (MT) 50 Nebraska 49 4 (NE)e Ohio (OH) 99/ 2/4 33 Oklahoma 101/ 2/4 (OK) 48 South Dakota 70/ 2/2 (SD) 35 New Jersey 80/ 2/4 (NJ) 40 North Dakota 94/ 2/4 (ND)f 47 Texas (TX) 150/ 2/4 31 Oregon (OR)h 60/ 2/4 30 Wyoming 60/ 2/4 30 (WY)h

Enacted ImpactLimit

Limit

opportunityb Date

Date

Length* Typec

14

Spring

1992

2000 8/8

D/C

12

Spring

1990

1998 8/8

D/C

15

Spring

1992

2000 8/8

D/C

29

Dead

1995

D/C

41

Dead

1993

2007 12/ 12 1996 8/8

5

Spring

1992

D/L

16

Dead

1992

1998/ 6/8 2002 2002 8/8

45

Spring

1992

2000 8/8

D/C

21

Dead

2000

2006 8

U/C

6

Spring

1992

2000 8/8

D/C

22

Dead

1990

2004 12

U/L

48

Dead

1992

2000 8/8

D/C

20

Spring

47

Dead

19

Dead

23

Spring

1992

1998 6/8

49

Dead

1992

NA

D/ Lg D/C

12/ 12

D/C

D/L

* Numbers are Lower/Upper a Professionalization ranking are from 1 (most professional), to 50 (least professional), see Squire

(2017) b Career

Opportunity structures, for term-limited states, are either dead (few advancement opportunities) or spring (ample advancement opportunities), see Squire and Moncrief (2010) c The limit types are: D/C, Divided Consecutive; D/L, Divided Lifetime; U/C, Unified Consecutive; U/L Unified Lifetime d Michigan amended their term limits in 2022, all analyses are based on the limit type shown here e Nebraska is a unicameral legislature and only has a Senate f North Dakota passed term limits in 2022 to go into effect in 2023 g Oregon law restricted legislative terms in 2 ways: a 6/8 year limit and a total 12 year limit, both forms are under a lifetime ban h The state passed term limits but they were quickly removed, Oregon in 2002 and Wyoming in 2004

24

J. BUTCHER

of confidence…” (Karp 1995, 374). Term limits were the solution, a solution so popular that more than 20 states passed initiatives in a span of four years. Paul Jacob (1994, 22), former Director of U.S. Term Limits, proclaimed that “the American people want term limits for members of Congress, as well as for virtually every person serving in elective office.” Part of Jacob’s job was to push term limits and advocate for them across the states, a job which he did well. Jacob stated (1994, 29), “people support term limits to accomplish two primary goals: (1) rejuvenate the election process and (2) restore a citizen legislature.” The U.S. Term Limits organization was a key actor in the push for term limits. This particular organization has pre-written policies that are pushed to the states; they also offer support in gathering signatures for term limit petitions.1 Citizens feared their elected officials had become too disconnected from the average citizen experience (Jacob 1994). The lack of public support for Congress and state legislatures, made it easier for term limits to be successful as a reform (Mann 1992, 1994). Advocates of term limits presented the reform as a way to save state governments from corruption, as well as an avenue to refresh policy. The frustration with elected officials was channeled into overwhelming public support for term limits. The target being, career politicians. The most obvious intent of term limits was to remove incumbent legislators from office. This removal and increasing turnover would bring additional changes to the legislature. For many, “term limits provide[d] the means of reforming the system, thereby altering the character of representation” (Karp 1995, 374). As Thomas Mann (1994, 84) said, “the crux of term limits is a rejection of professionalism in politics—or legislative careerism.” The push of term limits revolved around legislative careers, as did the support. In sum, careers were considered the “linchpin” for term limits (Mann 1994, 88). Proponents of term limits argued that turnover would push out difficult to beat incumbents and increase electoral competition (Cain 1996; Gerber and Lupia 1996; Grofman and Sutherland 1996; Mann 1992). By removing incumbents credible challengers would come forward and run for office (Coyne and Fund 1992). As a result, elections would be 1 See https://www.termlimits.com/contact-state-legislator/ for more examples of what the organization does to push term limits. (last accessed October 6, 2023).

2

THE PUSH AND PROMISE OF TERM LIMITS

25

more balanced between challengers and incumbents (Mann 1994). Ultimately, by implementing term limits stronger challengers would compete for office and there would be greater electoral competition. Advocates promised that the make-up of the legislature would change and there would be new types of members serving. Under term limits, the legislature would be more diverse (Cain and Kousser 2004; Will 1992). By removing the incumbent white, male legislators there would be seats available for women and minorities. Legislators were also going to be younger and, advocates argued, these new members would seek out more modern policies (Cain and Kousser 2004). These newer, younger members would be more accountable to their constituents because they have not had a chance to become careerist politicians (Mann 1994). Term limits would help state government broadly. The increased turnover would discourage the corruption that citizens feared. Some theories went so far as to say that pushing out senior members, the iron triangle could be dismantled because the relationship between legislators, interest groups, and the bureaucracy would be less stable (Kristol 1993; Polsby 1993). Ultimately, term limits would make legislatures, like Congress, more efficient and members would become more productive (Mann 1994). There were unspoken partisan motivations for term limits as well. Mann (1992, 23) wrote, “we are in the midst of a national campaign to limit legislative terms not because the public is in an open revolt against a rigged electoral system but because the intense ideological and partisan battles of our time have been moved to a new venue.” At the time, term limits were supported by the Republican Party because Democrats had a strong majority in both Congress and state legislatures (Hall 2014; Carey et al. 1998). In fact, the majority of term limit supporters among political elites were from the Republican Party (Polsby 1993). Term limits were supported at the Republican conventions in 1988 and 1992 (Fett and Ponder 1993). In 1994, Republican congressional candidates implemented informal term limits on themselves through the Contract with America (Powell 2008, 33). In theory, term limits would help the underrepresented party (Gilmour and Rothstein 1996; Karp 1995; Mann 1994). Term limits would remove long-seated politicians of one party and allow the minority party to step in and gain a majority. For example, in Maine, term limits were knowingly supported by Republican legislators because of the long-standing Democratic control (Moen et al. 2005). Term limits would effectively diminish

26

J. BUTCHER

the long-standing majority and level the playing field, resulting in a less partisan legislature. “Term limits are an expression of the public’s distrust of power” (Farmer and Green 2007, 13). Term limits promised to bring many positive changes to the legislature; there would be more diversity, competitive elections, and parties would become more equal. As term limits were being implemented, many began to question the promises that proponents had made. After term limits began in Michigan, former Governor John Engler was concerned that the effects of term limits were “oversold” (Penning 2003, 44). Becky Cain (1994, 45) argued that term limits were nothing more than a “smokescreen” attempting to hide larger problems with government. Moreover, advocates largely ignored the potential changes that would occur at the institutional level (Powell and Jones 2007). While some early studies concluded that there were few evident changes, it was not clear what good term limits had brought, if any (Straayer 2003). Not only were these promises left unfulfilled, there were many unintended consequences of term limits. Advocates, focused on turnover, failed to account for the dynamic and demanding job that is serving in the legislature.

Passing Term Limits Term limits have historically been popular among the public, but exist in less than half of the states. Term limits rose to popularity in the early 1990s; at the time, there were 41 states considering term limit legislation (Mooney 2009). Even now, states consider term limits each year. In 2019 alone, 20 state introduced a total of over 40 bills related to legislative term limits.2 In the last 30 years, 20 states ultimately adopted term limits and 15 still have them today. Each of these 15 states is unique, beyond the presence of term limits, which shapes and develops the relationship that the legislature has with these limits. This section outlines the number of

2 Last accessed 2020. https://www.ncsl.org/blog/2019/07/17/only-arkansas-has-pas sed-term-limit-bill-in-2019.aspx.

2

THE PUSH AND PROMISE OF TERM LIMITS

27

successes, the numerous failures to implement term limits as well as the two states who have changed their term limits. Successes For the majority of the states that have them, term limits are a result of citizen initiatives. This is the process where citizens collect signatures to put term limit legislation on the ballot, then voted on by the public. Because the public is largely supportive of term limits, regardless of party, as long as states allow for citizen initiatives there is a high chance of successfully passing legislation. The public push for legislation and the collection of signatures is not typically done by citizens alone, rather, organizations like U.S. Term Limits or state-specific term limits organizations that develop and propose legislation. At the time that term limits were surging, nearly half of the states had a citizen initiative process in place. 14 of the 15 states that passed term limits by initiative still have them today. Figure 2.1 illustrates the current states with legislative term limits. In the states that successfully passed term limits, they are not only guidelines for service, but have constitutional standing. Within a state’s constitution, term limits are labeled as either an electoral qualification or a restriction from running. For example, Florida legislators who serve eight consecutive years are disqualified from seeking reelection.3 In Nevada, term limitations are described as a part of the qualifications for office, i.e., a member is only qualified to run for office if they have not yet termed out.4 Failures There were six states that had passed term limit legislation, but ultimately this legislation failed and the term limits were removed. Massachusetts, Oregon, Washington, and Wyoming all passed term limits in the early 1990s. In these four states, the State Supreme Court tossed out term limits on technical grounds. In the other two states, Idaho and Utah, term limits were repealed by the legislature. Mississippi and North Dakota are the only two states with the citizen initiative process that moved to the ballot and the term limit initiatives actually failed to get the votes needed to pass. Alaska and Illinois also had the citizen initiative process but the 3 Constitution of the State of Florida, Article 6, § 4. 4 Nevada Constitution, Article 4, § 3–4.

28

J. BUTCHER

ME

WA ND

MT

VT

MN

OR ID

NH NY MA RI CT

WI

SD

MI

WY IA

NE

NV UT

CO

IL KS

CA

PA

NJ MD DE

OH IN

WV

VA

KY

MO

NC TN

AZ

OK

NM

SC

AR MS

AL

GA

LA TX FL

AK

HI

Fig. 2.1 States with term limits prior to 2022

legislation never came to a vote because Alaska had informal, voluntary limits and in Illinois it was too difficult to get on the ballot. There were four states that had term limits removed by the court. In Massachusetts, the public referendum passed, but only barely with a 51% vote. The court reached a unanimous agreement 3 years after the passage that the presence of term limits in the state added an additional qualification for office, going against the guidance in the state constitution.5 Oregon had a much higher passage rate of 69%, after the passage the legislature did consider potential amendments to alter the restrictive term limit of 12 years. A former legislator sued the state in 1997 arguing that the limit was a violation of their first and fourteenth amendment rights. The court agreed, initially, but the appellate court did not believe that the limit was in violation of member rights. The question arose again 5 Last accessed October 6, 2023. https://www.nytimes.com/1997/07/12/us/top-mas sachusetts-court-overturns-term-limits.html.

2

THE PUSH AND PROMISE OF TERM LIMITS

29

in 2001–2002 where the court ruled the limits unconstitutional and the provisions were turned over (Taylor 2004). Since this ruling, there have been reoccurring pushes for term limits in the state. The state of Washington went through a similar legal battle based on a violation of the first and fourteenth amendment, the court ultimately rejected the initiative putting a stop to term limits. Wyoming was the last of the states to have their limits removed by the court. One group who filed suit in Wyoming cited a constitutional infringement. Interestingly, they also made the argument that the presence of term limits implied voters did not know how to choose their own representatives, this going against the Republican ideals of government (Hageman and Subson 2004). In 2004, the State Supreme Court ruled that citizens did not have the right to limit legislative terms, thus undoing the law passed 12 years earlier (Press 2004). Idaho and Utah are the only two states that had term limits that were later repealed by the state legislature. Utah’s relationship with term limits is unique, in that they are one of the only states that self-imposed term limits, which were then repealed. In 1994, the Utah legislature implemented term limits out of a fear that a more restrictive term limit would be passed by popular vote. However, in 2003 the legislature repealed the limits. Senator Bramble said the repeal was about the people, “Some are concerned about the going against the voice of the people but the voice of the people is the ballot box,” implying voting as a form of limit (Spangler and Bernick 2003). Idaho has a far more complex history with term limits and the decision to appeal. In 2002, Idaho became the first (and only) state to turn down the citizen-imposed term limits passed in 1994 (Smith 2003). This decision, of course, led to many upset citizens as well as states who were stuck dealing with the implications of term limits that they could not repeal. The term limits legislation in Idaho had endured a long legal battle in the state and it was not until 2001 that the court upheld term limits (Reed 2013). This pushed legislators, who were going to begin to term out of office the very next year, to propose legislation repealing term limits. After more than 90 testimonies between both chambers, the repeal was successfully passed. Governor Kempthorne, however, put a temporary stop to the push and vetoed the legislation (Reed 2013). The repeal was so popular that it was easily overridden by both chambers and went into effect on February 1, 2002.

30

J. BUTCHER

Changes There are some states that passed term limits and have since changed them. California changed their term limits once and Arkansas changed theirs for the second time in 2020. When states change their term limit it not only shifts the process in place but it raises a number of questions for those who serve under the current system and how the state will transition to the new one. Interestingly, when California and Arkansas changed term limits from their original format, they both changed to a similar law but with different year limits. Prior to their change, both states had originally implemented a limit of six years in the lower chamber and eight in the upper (like Michigan). The 2020 change to Arkansas’ term limits restricted the number of years eligible to serve down to 12 and they changed the rules allowing members to return to the legislature after they have termed out of office. Although changes to existing term limit laws are rare, the states that have term limits do continue to see new proposals appear on the ballot or come up for debate in the legislature. For example, in the spring of 2020 the Ohio legislature was considering a change to their term limits. In 2018, Missouri saw a potential piece of legislation to not only reform their term limits but to implement limits on leadership positions. Based on current experiences with term limits, it appears that even the states that already have legislation in place will continue to see new laws proposed to redesign term limits. The most recent change to legislative term limits came in November of 2022 when Michigan voters passed legislation to revamp their previous term limits to 12 years total, to be served in a single chamber, or divided as the lawmaker chooses. Interestingly, the new proposal that passed was tied to disclosure requirements, perhaps helping it to pass with higher margins. Given that this update is so recent, for the purposes of this research, Michigan is still considered a divided/lifetime state as it was during the time this research was conducted.

Types of Term Limits Among the 15 states that still have term limits, there are 4 different variations of term limits. These different rules are in addition to the number of years. Most commonly, states limit terms to either 8 or 12 years. The rules and restrictions that are in place are what differentiate the types of

2

THE PUSH AND PROMISE OF TERM LIMITS

31

term limits. The ability to return to the legislature, in particular, is what sets some of the states apart. This section outlines the different types of term limits while addressing how the variations can affect the members who serve. The four types of term limits can be seen in Fig. 2.2. There are two distinctions that can be made—(1) a limit is either unified or divided, (2) a limit has a lifetime or consecutive ban. The difference between unified and divided limits is how the years are counted. Under a unified limit, members are given a set number of years and they can choose to divide these years between chambers or spend them all in one. Under divided limits, members are given a set number of years for each chamber. Members can serve only the set number of years for that chamber, they cannot be counted for the other chamber. Second, limits have either a lifetime or a consecutive ban. The bans place restrictions on the continuation of service. Members, under a lifetime ban, can never return to office once they have been termed out. A consecutive ban indicates that members can serve the set number of consecutive years, then they must leave, but they may later seek reelection. These seemingly simple differences that exist among types of term limits carry great consequences. While the key difference between unified and divided limits is in how the years are counted the implications extend beyond that. Members behave differently when they can decide where they can spend their time and for how long. Moreover, when they are faced with set restrictions in each legislative chamber they can plan how Lifetime

Consecutive

Unified Year Limit

CA (12), OK (12), AR (16)

NE (8)

Divided Between Houses

MI (6/8), MO (8/8), NV(12/12)

AZ, CO, FL, ME, MT, OH, SD (8/8) LA (12/12)

Fig. 2.2 Different types of term limits (Note Years are [House/Senate] in divided states Nebraska has only one chamber. Arkansas has voted to become a unified/consecutive state)

32

J. BUTCHER

to best use their time, knowing there is more time available in the next chamber. The difference between the consecutive and lifetime bans is more evident. Members serving under lifetime bans cannot return to the legislature forcing them to learn and legislate at an accelerated pace. Those who serve under consecutive bans have more time and are thus able to gain more legislative experience. The ability to return gives the member more flexibility and control over their career. The differences in the types of limits may seem inconsequential but they carry implications for how well the legislature will adapt. With different legislative structures and term limit types, the effects of term limits will vary among the states. Not all legislatures with term limits will feel changes to the same extent. Not all types of term limits are as effective at limiting careerism and legislative experience. Lax limits, like the divided/consecutive limit, are less likely to see changes to committee capacity. Whereas, stricter limits are more likely to experience changes to the legislative process. The institutional structure before term limits—the reliance on leaders and committees—is an indication of how much influence term limits have on the legislature (Little and Farmer 2007). Specifically, those legislatures that relied a great deal on having experienced members would be most affected by the removal of those members. States with lax limits would experience different changes than those with greater restrictions (Sarbaugh-Thompson and Thompson 2017; Little and Farmer 2007). Even when two states have the same type of term limit they will not be influenced in the same way because of their different institutional structures. Notably, term limit influence will vary by level of legislative professionalization. Recall, term limits passed as a way for citizens to retract professionalization reforms. Legislative professionalization is a measurement of time spent in session, member salary, and staff support (Squire 1992b). A more professional legislature has greater staff support, pays their members better, and is in session longer. The changes brought by professionalization strengthened the legislature but also led many members to serve longer. Citizens pushed for term limits because they were unhappy with their professionalized legislatures and long-serving incumbents (Farmer and Green 2007; Mooney 2007; Moen et al. 2005). Given that professionalization was the target, the consequences of term limits vary by level of professionalization. Although term limits do not

2

THE PUSH AND PROMISE OF TERM LIMITS

33

directly influence the individual components of professionalization, there is an evident connection to experience and expertise (Mooney 2007). Given the resources of a more professional legislature, the changes brought by term limits are going to be more evident in highly professionalized states (Farmer and Green 2007). The changes that occur in less professional states will be different than other states, because of the direct connection between citizen legislatures and term limits (Farmer and Green 2007; Mooney 2007). Term limits were intended to be a counterbalance to professionalization, but some of the states that have term limits had not become professionalized—term limits were undoing something that had not come into existence (Smith 2007). Given the added variation in institutional structures, term limits need to be assessed in combination with a state’s level of professionalization (Farmer and Green 2007; Kurtz et al. 2007). Another measure of institutional ability is a state’s career opportunity structure. Career opportunity structure has an influence on how legislators act and work as an incentive to continue serving in office (Squire 1992a). Each legislature has a different career opportunity score indicating if service in that legislature is a dead end, career, or a springboard position (Squire 1988). The specific measurement of career opportunities is based on the advancement potential of legislators who serve in the lower chamber of the state legislature (Squire and Moncrief 2010).6 Term limits were framed around legislative careerism and the distrust of incumbents (Mann 1994). Given the importance of careers to the term limit debate, it is necessary to understand how the legislature’s career trajectories interact with term limits. In the remainder of this chapter, I provide an overview of the 15 termlimited states, emphasizing the 12 that are used in this study. There are also five states without term limits as a control. In addition to term limit type, there are notes about legislative professionalization (Squire 1992b), and career opportunity structures (Squire 1988). These are the two most notable measurements of state legislative institutions and help further current understandings of term limit influence. The following overview is meant to illustrate the differences between the term-limited states and

6 The advancement opportunity score is “the product of the ratio of higher seats to lower house seats and the percentage of higher seats held by former lower house members” (Squire and Moncrief 2010, 110).

34

J. BUTCHER

show that each state will have different experiences with term limits based on term limit type and the pre-existing legislative structure.

Unified/Lifetime Limits There are three states that impose unified/lifetime term limits. California, Oklahoma, and Arkansas all set a singular limit on legislative service. Legislators are not permitted to return once they have termed out of office. Each of these three states has a different institutional structure, which can enhance or hinder the role of term limits. Unified/lifetime limits are considered to be more strict, allowing for less time in total service than divided limits. Additionally, the lifetime restriction harms the ability of members to serve because they are not allowed to return. Oklahoma Oklahoma was also one of the first three states to pass term limits in 1990.7 Former State Senator Gene Stipe has been pointed to as the cause for term limits in Oklahoma. Gene Stipe began serving in the Oklahoma state House in 1948; he then moved to the Senate in 1956.8 Stipe held his seat until 2003, making him the longest serving state representative in the United States at the time. However, in 2003, Stipe faced federal charges for illegally funneling money into a congressional campaign. At the time, Stipe chose to resign his seat; he was again indicted in 2007.9 The end of Stipe’s career was plagued with scandal, which garnered public support for term limits in Oklahoma. Stipe, himself, never actually faced the term limits that were pursued in his honor. Oklahoma instituted a unified 12-year limit with a lifetime ban. A member in Oklahoma can serve in either chamber for up to 12 years. Having a single limit gives members the ability to choose how they divide their time between chambers. A legislator is not permitted to serve in the chamber beyond 12 years, and after terming out legislators are not allowed to return. Members who serve in the Oklahoma legislature have limited advancement prospects and a lower level of professionalization.

7 Constitution of the State of Oklahoma, § V-17A. 8 Former Oklahoma state senator Gene Stipe dies at 85, NewsOK 2012. 9 Former state senator Gene Stipe pleads ‘guilty’ in federal court, Associated Press 2003.

2

THE PUSH AND PROMISE OF TERM LIMITS

35

With few opportunities for advancement, it is more likely that members will stay in the state legislature longer. California California currently has the same type of term limit as Oklahoma, and was one of the first states to pass term limits. California originally had a strict limit allowing for only six years in the House, eight in the Senate, and a lifetime ban (Carey et al. 2000). Term limit support in California was largely a result of the power of the Speaker of the House, Willie Brown, and an FBI corruption sting (Cain et al. 2007; Clucas 2003). The original proposition narrowly passed and began removing members in 1996 (Cain et al. 2007). After the passage of Proposition 28 in 2012, California changed their term limit to a unified 12-year limit and retained the lifetime ban.10 Due to the recent changes in California’s term limit law, this state is not included in this study because not enough time has passed under the new limit to understand the long-term effects. Arkansas Arkansas has also seen changes to its type of term limit. Arkansas first passed term limits in 1992 but changed them in 2014. Arkansas originally had a strict limit of 6 years in the House and 8 in the Senate with a lifetime ban (Carey et al. 2000). Arkansas had a unified limit of 16 years with a lifetime ban. Term limits nearly appeared on the ballot in 2018 but were voted on in 2020. Arkansas now holds the record as the only state to have 3 types of term limits, in 30 years no less. Arkansas’ most recent term limit is a 12 year unified/consecutive limit, which began for members elected in 2021.11 For those serving under the 2014 law, the new 2020 law included an amendment for members to return 4 years after leaving. It is unclear if observed changes would be a result of the prior term limit, or the new term limit, as a result Arkansas is not included in this study.

10 Constitution of California, Article 4, § 2. 11 Constitution of the State of Arkansas, Sect. 2, Amendment 73.

36

J. BUTCHER

Divided/Lifetime Limits There are three states with divided/lifetime limits. Each state with this type of limit has different year restrictions in place. For example, Michigan has one of the shortest limits allowing members only 14 years in total. Nevada, however, has one of the longest limits allowing for a total of 24 years. While Michigan and Nevada have the same type of term limit restrictions, the varying lengths have different effects on the legislature. Michigan Michigan passed a divided/consecutive term limit in 1992.12 Michigan’s strict lower chamber limit of six years went into effect in 1998. A member can serve only three terms (two years each) in the lower chamber, giving legislators only four years in office before they are labeled a lame duck (Sarbaugh-Thompson and Thompson 2017; Sarbaugh-Thompson et al. 2004). Michigan’s upper chamber, restricted to eight years, began terming people from office in 2002. This brings the total length of service in Michigan to 14 years. Michigan legislators have more career opportunities beyond the state legislature, but they also have greater incentive to serve. Michigan is one of the more professional state legislatures; members are well paid and have staff support (Sarbaugh-Thompson and Thompson 2017; Squire 2017). Benefits such as these make it more lucrative for members to stay in the legislature where their position is more secure. For those that are more ambitious, Michigan does provide more career opportunities beyond the legislature. Missouri Missouri passed term limits in 1992, and they went into effect in 2002.13 In Missouri, no one—elected or appointed—can serve more than eight years in each chamber, 16 total. In Missouri, because of the lifetime ban, once a legislator has served eight years in either chamber they can never return to office. Members who serve in Missouri have the benefit of personal staff to help with the workload, which makes it more enticing 12 Constitution of the State of Michigan of 1963, Article 4, § 54. 13 Missouri Constitution, Article 3, § 8.

2

THE PUSH AND PROMISE OF TERM LIMITS

37

for members to serve in the legislature. Due to the limited career opportunities beyond the legislature, members are more likely to serve until they fulfill their term limit. The push for term limits in Missouri began in 1990. Constituents were concerned that the state government was overrun by incumbents and special interests (Richardson et al. 2005). One supporter stated, “people don’t like elites, and they look at [legislators] like they are elites” (Richardson et al. 2005, 178).14 In Missouri, term limits came down to a battle of the numbers. Supporters raised nearly three hundred thousand dollars, those in the opposition raised just over ten thousand (Richardson et al. 2005, 178). In the 1992 election, voters passed term limits with 75% approval (Richardson et al. 2005). Nevada Nevada is the most recent state to implement term limits, with the first cohort terming out in 2010. Nevada’s relationship with term limits began in 1994 when there were two types of term limit proposals on the ballot (Stewart 2006). In 1996, before the limits could be implemented, there was a series of different court cases that restricted the ballot (Stewart 2006). After much debate, Nevada did pass a 12-year limit. At this point, not enough time has passed to evaluate the effects of term limits in Nevada given their lengthy terms. As a result, Nevada is not included in this study.

Unified/Consecutive Limits Nebraska The Nebraska legislature, being unicameral, has its own version of a unified term limit. After much contention, Nebraska passed term limits in 2000, and termed members out in 2006.15 Members of the Nebraska unicameral are allowed to serve up to eight years. After sitting out for one election cycle (four years), members are allowed to return. There are few

14 Original Source: Greg Upchurch. Quoted in Will Sentell, Kansas City Star, August 25, 1992. 15 Nebraska Constitution, Article 3, § 8.

38

J. BUTCHER

opportunities to serve beyond the legislature so it is more common for members to fulfill their allotted terms. In Nebraska, the call for term limits was directed to two men who were serving in the Nebraska Senate. One of those men is Ernie Chambers, who is the longest serving legislator in Nebraska’s history. There are some who cite the push for term limits as being racially motivated, as Senator Chambers was the only black legislator for a long period of time.16 Chambers began serving in 1971, his first term expired in 2009 after term limits went into effect. Ernie Chambers, who holds a great deal of respect in the legislature said, “term limits gutted the legislature as a branch of government” (Schulte 2016). The consecutive term limit allowed Chambers to return in 2013 where he served until terming out again in 2021 at the age of 83.

Divided/Consecutive Limits The most common type of term limit is the divided/consecutive limit. There are eight states that have this limit; seven states restrict members to eight years in each chamber. Louisiana, however, allows for 12 years in each chamber. While this is the most common type of limit, each state is structured differently and passed term limits for different reasons. Arizona In 1991, Arizona legislators nearly imposed term limits on themselves but the proposal ultimately failed. The rising support for term limits among the public came after the infamous AzScam where seven legislators were indicted for bribery by an undercover FBI agent posing as a lobbyist (Berman 2007). After the failed attempt to push term limits in 1991, the Arizona Coalition for Limited Terms (ACLT) pushed for term limits in 1992, which subsequently passed with 74% of the vote (Berman 2007). Arizona’s term limits went into effect in 2000, members can serve four consecutive terms in each chamber, a total of eight years in each chamber.17 Members may return after sitting out one full term (two years). Arizona is one of the smaller state legislatures with only 90

16 These statements come from personal legislator interviews. 17 Arizona Constitution, Article 4, Part 2, § 21.

2

THE PUSH AND PROMISE OF TERM LIMITS

39

members. This smaller size makes it easier to advance to higher office because there are fewer qualified candidates. Arizona is also more professional, which makes it more enticing for members to stay in office for as long as they can. Colorado Colorado was one of the first states to pass term limits, passing in 1990 and implementing in 1998.18 The Colorado Term Limits Commission (CTLC) explicitly said it was their mission to make Colorado a citizen legislature (Straayer 2007). Terry Considine, a businessman and state senator, personally pursued term limits in Colorado. With the help of Club for Growth, a group led by Considine, there was an attempt at legislative instituted term limits (Luning 2022). When this plan failed, Considine went to the public. The initiative in Colorado passed with 71% approval and included limits for the “governor, lieutenant governor, secretary of state, attorney general and treasurer” (Straayer 2003, 61). The move was so popular that in 1994, term limits were extended to all locally elected officials as well (Straayer 2007). In Colorado, legislators can serve no more than eight consecutive years in each chamber. Additionally, if a legislator has, for any reason, served at least half of a term it is counted as a full term. Colorado is a more professional state, so members desire to serve longer. Colorado also presents more career opportunities beyond the state legislature, so ambitious members move from the legislature to other higher offices. Florida Florida passed term limits in 1992 and they went into effect in 2000.19 Term limits passed with 77% approval after numerous attempts to avoid placing the term limit proposal on the ballot (Prier and Wagner 2009; Schraufnagel and Halperin 2006). Florida citizens were upset with their unresponsive legislators (Prier and Wagner 2009). Florida voters liked the idea that term limits would reinvigorate the electoral process; they

18 Constitution of the State of Colorado, Article 5, § 3. 19 Constitution of the State of Florida, Article 6, § 4.

40

J. BUTCHER

were particularly keen on the idea that there would be more electoral competition (Schraufnagel and Halperin 2006). In Florida, members cannot run for reelection if they have served for eight consecutive years. This is the rule for each chamber, allowing for a total of 16 years of service. Florida, like Colorado, is more professionalized and provides personal staff. Members of the Florida legislature also have greater opportunities to serve beyond the legislature and are unlikely to return to the legislature once they have termed out. Maine The push for term limits in Maine evolved from changes that began to occur in the 1950s (Moen et al. 2005). Citizens embraced term limits because they were unhappy with the way the state had changed; many voters thought that the legislature was disconnected with the public (Moen et al. 2005). In fact, The Ellsworth American, a local paper, wrote, “There is a crisis in Maine government… [with] the sitting legislature… [being] an example of how far removed politicians are getting from grass roots concerns of their constituents” (Moen et al. 2005, 35).20 The push in 1993 was a result of three events: first, a budget crisis in 1991; second, a scandal in 1993 with House Speaker John Martin; third, the increasing cost of the legislature (Moen et al. 2005; Moen and Palmer 2003). In 1992, the citizens of Maine pushed an initiative for term limits, which the legislature was supposed to debate in 1993. However, the legislature passed term limits on the leadership as a way of avoiding the original petition (Moen et al. 2005). While there were additional proposals in 1979, 1983, 1989, and 1991, the “Term Limitation Act” finally passed in 1993 by nearly 68% (Moen et al. 2005, 42). Maine is a state that retroactively implemented term limits, which sped up the process of forced removal (Moen and Palmer 2003). Term limits went into effect in 1996, senators and representatives can serve no more than four consecutive terms, eight years in each chamber.21

20 Original source: Brown 1991, “A Crisis in Government.” The Ellsworth American. 17 October. 21 Me. Rev. Stat. Ann., Tit. 21-A, § 553.

2

THE PUSH AND PROMISE OF TERM LIMITS

41

Montana Montana adopted term limits in 1992 and implemented them in 2000.22 Since passage, Montana has seen more than 14 new proposals introduced. In Montana, members are not permitted to serve eight or more years in a 16-year period as a state representative or senator. In other words, a member can serve eight years in one chamber, upon being termed out they cannot return for another eight years. This practice applies for both chambers. This restriction may seem odd, but it does allow for continued service. For example, a member can serve eight years in the lower chamber and then move to serve eight in the upper. By the time service in the upper chamber is complete, the member can then return to the lower chamber. Montana is not considered to be a professional legislature, but members do have increased opportunities to advance beyond the state legislature.23 The state faces additional barriers because they are only in session for roughly 4 months every two years. With a lack of incentive to stay in the state legislature, ambitious politicians are more likely to leave before their term is up and advance to higher office. Ohio Ohio was one of the first states to make moves toward professionalization in the 1970s. As a result, Ohio quickly became one of the most professional legislatures. In direct contrast of professionalization, there was a move in 1992 toward term limits (Farmer and Little 2007). Foreseeing the implementation of term limits, as soon as the proposal went on the ballot legislators pushed Speaker Riffe to put forth a better option for term limits. Speaker Riffe would not propose an alternative. Meanwhile, legislators were concerned that term limits would increase the power of executives and bureaucrats. With no alternative, Ohio passed term limits with nearly 70% of the vote and implemented them in 2000 (Farmer and Little 2007, 44).24 Lawmakers in Ohio cannot serve more than eight consecutive years in either chamber. After terming out, members are allowed to return after one election. Ohio is highly professionalized and offers better career 22 Constitution of the State of Montana, Article 4, § 8. 23 See Squire (2017). 24 Ohio Constitution, Article 2, § 2.

42

J. BUTCHER

advancement opportunities. Other opportunities can drive members to seek higher office, but the internal incentives create the desire for members to stay in the state legislature. South Dakota South Dakota adopted term limits in 1992 and implemented them in 2000.25 Term limits passed with 63% approval, which is lower than some other states (South Dakota’s term limit laws under question 2010). In 2008, a term limit repeal went on the ballot and 76% of voters opposed it (South Dakota’s term limit laws under question 2010). Institutionally, South Dakota is different from other divided/consecutive states. South Dakota is not professionalized, nor does it provide career opportunities outside of the legislature.26 If the predictions of term limits are upheld, South Dakota with term limits should not operate all that different from South Dakota before term limits. Louisiana Louisiana is the only state, currently, with self-imposed term limits. While most states have term limits as a result of citizen initiatives, the Louisiana legislature preemptively placed term limits on themselves in 1995.27 In 2007, the first year term limits went into effect; the Louisiana legislature had its highest rate of turnover since the 1970s, then caused by the elimination of multi-member districts.28 Louisiana has the longest divided/consecutive limit, which is likely a result of the limits being self-imposed. No person serving two and half out of three consecutive terms can be reelected to either chamber. In other words, members are limited to three terms in each chamber, or 12 years. Louisiana provides few opportunities to serve beyond the legislature. This, in combination with the potential for 24 years of service, members are likely to stay in the legislature until they choose to retire.

25 Constitution of South Dakota, Article 3, § 6. 26 See Squire (2017). 27 Louisiana Constitution, Article 3, §4. 28 Good idea gone awry? How term limits impact sessions of Louisiana Legislature (2016).

2

THE PUSH AND PROMISE OF TERM LIMITS

43

States Without Term Limits As a point of comparison, there are five additional states included in this study that do not have term limits: New Jersey, North Dakota, Oregon, Texas, and Wyoming. These five states are comparable to the term-limited states—they are of similar size, level of professionalization, and have similar career opportunity structures. Not only are they comparable, but these states cover a range of institutional structures, much like the term-limited states. Additionally, these states account for similar types of leadership, session frequency, and legislative structure. The first comparable state without term limits is New Jersey. The New Jersey legislature has a total of 120 members, which is similar to Ohio and Oklahoma. Members in New Jersey have increased career opportunities to serve outside of the legislature, like Florida. New Jersey also has an average level of professionalization ranking 20th, comparable to states like Nebraska (21) and Oklahoma (22). North Dakota and Texas are two non-term-limited legislatures that meet biennially, like Montana. North Dakota has nearly 150 legislators, which is similar in size to states like Louisiana (144), Michigan (148), and Montana (150). Like South Dakota, there are few outside opportunities for those that serve in North Dakota. Additionally, North Dakota has a lower level of professionalization like South Dakota ranking 47 and 48, respectively. Texas meets biennially and has a large legislature of 181 individuals comparable in size to Maine (186) and Missouri (197). Similar to Missouri, Texas is a dead end state with a moderate level of professionalization ranking 19th. It is at this point, where it is important to share that in 2022 North Dakota joined the ranks of its surrounding state by passing term limits. This legislation passed with more than 60% of the vote, likely because it set term limits for both the governor and the legislature. In the state of North Dakota, lawmakers are now restricted to eight years in each legislative chamber with a lifetime ban (Anderson 2022). While this change is significant for the state of North Dakota, much like the term-limited states that are the center of this examination, it will take more than 16 years before the effects of this passage will really take hold in the state. But, be warned, it will likely take longer than that as members in the House

44

J. BUTCHER

and the Senate are already revamping proposals to extend their new term limits.29 Wyoming and Oregon are the last two states used in this study. These two states are different and important because they once had term limits but they were quickly repealed, these are the “almost states.” These states are important to assess because they nearly underwent the same institutional changes as the states that still have term limits. Oregon is similar in size and career opportunities to Arizona and Colorado. Oregon has 90 members in the legislature, as does Arizona and Colorado has 100. Oregon is more akin to Oklahoma (22) and Ohio (21) when it comes to professionalization ranking as 23. Wyoming, however, is less professional (49) and offers limited career opportunities making it more comparable to Maine (41) and South Dakota (48). In Wyoming, the state Supreme Court upheld that term limits were unconstitutional, this occurred in 2004 before any members had been termed out of office. Wyoming lawmakers were going to be limited to serving only 12 years in a 24-year period in either legislative chamber (this law is similar to that of Montana). On the precipice of forcefully removing lawmakers from office, some lawmakers took the decision into their own hands and resigned prior to term limits becoming effective, even though they never actually became effective. Oregon, however, had members term out of office in 1998 and 2000 before the state Supreme Court ultimately put an end to limits in 2002. In those 2 years, Oregon had 46 legislators term out of office. Oregon’s term limits were unlike any term limits currently in existence. According to the state constitution, members could serve only 6 years in the lower chamber, 8 years in the upper chamber, but no more than 12 years in the legislature.30 Each of these members are accounted for in the dataset. The states of Wyoming and Oregon present a nice complement to the term-limited states and the states without term limits. These states present a view of the early effects of term limits, but never reached a point of equilibrium. In Wyoming, lawmakers dealt with the effects of passing term 29 Last accessed June 7, 2023 https://bismarcktribune.com/news/state-and-regional/ govt-and-politics/north-dakota-house-oks-measures-for-longer-term-limits-more-legisl ative-session-days/article_df26c454-c1bc-11ed-a32e-9f7ba0631066.html and https://csg midwest.org/2022/11/18/a-look-at-how-the-2022-election-results-will-impact-state-pol icy-and-legislatures-in-the-midwest-in-the-year-ahead-and-beyond/. 30 Oregon Constitution, Article 2, § 19.

2

THE PUSH AND PROMISE OF TERM LIMITS

45

limits and their impending implications. In Oregon, lawmakers dealt with the initial surge of term limits pushing senior lawmakers out of office for two years. Yet neither state maintained their term limits long term.

Bibliography Anderson, Tim. 2022. A look at how the 2022 election results will impact state policy and legislatures in the Midwest — in the year ahead and beyond. Council of State Governments. https://csgmidwest.org/2022/11/18/alook-at-how-the-2022-election-results-will-impact-state-policy-and-legislatu res-in-the-midwest-in-the-year-ahead-and-beyond/ (accessed October 6, 2023). Associated Press. 2004. Wyo. Court repeals 1992 term limits law. Fox News. https://www.foxnews.com/story/wyo-court-repeals-1992-termlimits-law (accessed October 20, 2021). Ballard, Mark. 2016. Good idea gone awry? How term limits impact sessions of Louisiana Legislature. The Advocate. https://www.nola.com/archive/goodidea-gone-awry-how-term-limits-impact-sessions-of-louisiana-legislature/art icle_04490330-a89c-5667-8961-5cb7a7f2ab29.html (accessed May 9, 2019). Berman, D. 2007. The effects of legislative term limits in Arizona: More churning, more chaos, and a diminished institutional role for legislators. In Legislating without experience: Case studies in state legislative term limits. Lanham, MD: Lexington Books. Cain, Becky. 1994. Term limits: Not the answer to what ails politics. In The politics and law of term limits, ed. Edward H. Crane. Washington, DC: Cato Institute. Cain, Bruce E. 1996. The varying impact of legislative term limits, chapter 1, 21–36. Cham: Springer. Cain, Bruce E., and Thad Kousser. 2004. Adapting to term limits: Recent experiences and new directions. San Francisco: Public Policy Institute of California. Cain, Bruce, T. Kousser, and K. Kurtz. 2007. Institutional imbalance: The effect of six-year limits in California. In Legislating without experience: Case studies in state legislative term limits. Lanham, MD: Lexington Books. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 1998. The effects of term limits on state legislatures. Legislative Studies Quarterly 23: 271–300. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 2000. Term limits in state legislatures. Ann Arbor: University of Michigan Press. Clucas, Richard A. 2003. California: The new amateur politics, 17–32. Lanham, MD: Lexington Books. Coyne, James K., and John H. Fund. 1992. Cleaning house: America’s campaign for term limits. Washington, DC: Regnery Gateway.

46

J. BUTCHER

Farmer, Rick, and John Green. 2007. Introduction: Accelerating change with term limits. In Legislating without experience: Case studies in state legislative term limits. Lanham, MD: Lexington Books. Farmer, Rick, and T.H. Little. 2007. Legislative power in the Buckeye state: The revenge of term limits. In Legislating without experience: Case studies in state legislative term limits. Lanham, MD: Lexington Books. Fett, Patrick J., and Daniel E. Ponder. 1993. Congressional term limits, state legislative term limits and congressional turnover: A theory of change. PS: Political Science & Politics 26 (2): 211–216. Former Oklahoma state senator Gene Stipe dies at 85. 2012. Gerber, Elisabeth R., and Arthur Lupia. 1996. Term limits, responsiveness and the failures of increased competition. In Legislative term limits: Public choice perspectives, 87–99. Cham: Springer. Gilmour, John B., and Paul Rothstein. 1996. Term limitation in a dynamic model of partisan balance, 145–174. Cham: Springer. Grofman, Bernard, and Neil Sutherland. 1996. Gubernatorial term limits and term lengths in historical perspective, 1790–1990: Geographic diffusion, nonseparability, and the ratchet effect. In Legislative term limits: Public choice perspectives, 279–287. London: Springer. Hageman, Harriet M., and Timothy M. Subson. 2004. Constitutional challenge to Wyoming’s term limit statute. Hall, Andrew B. 2014. Partisan effects of legislative term limits. Legislative Studies Quarterly 39 (3): 407–429. Jacob, Paul. 1994. From the voters with care. In The politics and law of term limits, ed. Edward H. Crane. Washington, DC: Cato Institute. Karp, Jeffery A. 1995. Explaining public support for legislative term limits. The Public Opinion Quarterly 59: 373–391. Kristol, William. 1993. Term limitations: Breaking up the iron triangle. Harvard Journal of Law & Public Policy 16: 95. Kurtz, Karl T., Bruce E. Cain, and Richard G. Niemi. 2007. Institutional change in American politics: The case of term limits. Lanham, MI: University of Michigan Press. Little, Thomas H., and Rick Farmer. 2007. Legislative leadership. In Institutional change in American politics: The case of term limits, ed. Bruce Cain Karl Kurtz, and Richard G. Niemi, chapter 4, 55–72. Ann Arbor, MI: University of Michigan Press. Luning, Ernest. 2022. Over the decades, Colorado comes to terms with term limits. Colorado Politics. https://www.coloradopolitics.com/columnists/ trail-mix-over-the-decades-colorado-comes-to-terms-with-term-limits/article_ ed1d7af4-b118-11ec-8f45-cf046ba2d499.html (accessed October 6, 2022). Mann, Thomas E. 1992. The wrong medicine: Term limits won’t cure what ails congressional elections. The Brookings Review 10 (2): 23–25.

2

THE PUSH AND PROMISE OF TERM LIMITS

47

Mann, Thomas E. 1994. Congressional term limits: A bad idea whose time should never come. In The politics and law of term limits, ed. Edward H. Crane. Washington, DC: Cato Institute. Moen, Matthew C., and Kenneth T. Palmer. 2003. Maine: The cutting edge of term limits. In The test of time. Lanham: Lexington Books. Moen, Matthew C., Kenneth T. Palmer, and Richard John Powell. 2005. Changing members: The Maine legislature in the era of term limits. Lanham: Lexington Books. Mooney, C. 2007. Truncated careers in professionalized state legislatures. In Legislating without experience: Case studies in state legislative term limits. Lanham: Lexington Books. Mooney, Christopher Z. 2009. Term limits as a boon to legislative scholarship: A review. State Politics & Policy Quarterly 9 (2): 204–228. Newson6. 2003. Former state senator Gene Stipe pleads ‘guilty’ in federal court. https://www.newson6.com/story/5e367fa32f69d76f620936d5/formerstate-senator-gene-stipe-pleads-guilty-in-federal-court (accessed March 4, 2019). Polsby, Nelson W. 1993. Some arguments against congressional term limitations. Harv. JL & Pub. Pol’y 16: 101. Penning, James M. 2003. Michigan: The end is near. In The test of time, ed. Rick Farmer, John David Rausch Jr., and John C. Green. Lanham: Lexington Books. Powell, Richard J. 2008. Minority party gains under state legislative term limits. State Politics & Policy Quarterly 8 (1): 32–47. Powell, R. J., and R. Jones. 2007. Institutional change and legislative term limits in Maine. In Legislating without experience: Case studies in state legislative term limits. Lanham: Lexington Books. Prier, Eric, and Kevin Wagner. 2009. Running unopposed: Assessing the impact of term limits on competition in Florida and Maine. Politics & Policy 37 (1): 101–125. Reed, Scott W. 2013. How and why Idaho terminated term limits. Idaho Law Review 50: 1. Richardson, Lilliard E., David Valentine, and Shannon Daily Stokes. 2005. Assessing the impact of term limits in Missouri. State and Local Government Review 37 (3): 177–192. Sarbaugh-Thompson, Marjorie, and Lyke Thompson. 2017. Implementing term limits: The case of the Michigan Legislature. Ann Arbor: University of Michigan Press. Sarbaugh-Thompson, Marjorie, Lyke Thompson, Charles Elder, John Strate, and Richard Elling. 2004. The political and institutional effects of term limits. Cham: Springer.

48

J. BUTCHER

Schraufnagel, Scot, and Karen Halperin. 2006. Term limits, electoral competition, and representational diversity: The case of Florida. State Politics & Policy Quarterly 6 (4): 448–462. Schulte, Grant. 2016. After 10 years, term limits reshape Nebraska Legislature. Smith, Daniel A. 2003. Overturning term limits: The legislature’s own private Idaho? PS: Political Science and Politics 36 (2): 215–220. Smith, M. 2007. It’s all about the turnover: Term limits in citizen legislatures. In Legislating without experience: Case studies in state legislative term limits. Lanham, MD: Lexington Books. South Dakota’s term limit laws under question. 2010. Spangler, Jerry D., and Bob Bernick Jr. 2003. Utah senators vote to repeal term-limits law. Deseret News. https://www.deseret.com/2003/2/25/ 19706340/utah-senators-vote-to-repeal-term-limits-law (accessed October 6, 2022). Squire, Peverill. 1988. Career opportunities and membership stability in legislatures. Legislative Studies Quarterly 13: 65–82. Squire, Peverill. 1992a. Changing state legislative leadership careers. In Changing patterns in state legislative careers, ed. Gary Moncrief and Joel Thompson, 175–194. Ann Arbor: The University of Michigan Press. Squire, Peverill. 1992b. Legislative professionalization and membership diversity in state legislatures. Legislative Studies Quarterly 17: 69–79. Squire, Peverill. 2017. A squire index update. State Politics & Policy Quarterly 17 (4): 361–371. Squire, Peverill, and Gary F. Moncrief. 2010. State legislatures today: Politics under the domes. Lanham: Rowan & Littlefield. Stewart, Michael J. 2006. The evolution and impact of term limits in Nevada. Nevada Legislative Counsel Bureau. Straayer, J. 2007. Colorado legislative term limits: The worst of both worlds. In Legislating without experience: Case studies in state legislative term limits. Lanham, MD: Lexington Books. Straayer, John A. 2003. “Colorado: Lots of Commotion, Limited Consequences.” In The test of time. Lanham, MD: Lexington Books. Taylor, Marjorie. 2004. Background brief on term limits. State of Oregon Legislative Committee Servies. https://www.oregonlegislature.gov/lpro/Publicati ons/2004EJ_Term_Limits.pdf (accessed October 20, 2022). Will, George F. 1992. Restoration: Congress, term limits, and recovery of deliberative democracy. New York: Maxwell Macmillan International.

PART II

Changing Careerism

CHAPTER 3

Legislative Careerism and Growing Ambition

I know that I am a beneficiary of term limits, within that, time goes very quickly. Now, in my 5th year, I feel like I finally understand what it is to be effective as a legislator. —Ohio Representative (Democrat)

Term limits promised to limit legislative careers by forcibly removing long-serving incumbents. In fact, in a handful of states limits were pushed to target specific members. For example, in Maine term limits were targeted at House Speaker John Martin. Martin served more than 30 years in the House and was Speaker for 19 of them (Moen and Palmer 2003). In 1993, Martin’s top legislative aide was caught up in an election scandal to which he pleaded guilty (Moen and Palmer 2003, 49). Even with the election tampering scandal surrounding Martin, he still managed to be reelected. This sequence of events helped with the passage of term limits and John Martin was ultimately removed by term limits, but he was then reelected just one election cycle later (Moen and Palmer 2003). There is a John Martin in every state with term limits, a longserving incumbent who inadvertently helped garner public support for term limits. Yet, many of these members were unaffected by their state’s term limit laws some, like Martin, even returned to office. Political ambition is typically understood as a means for a legislator to achieve his or her goals (Schlesinger 1966). There is, however, a © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_3

51

52

J. BUTCHER

distinction that receives less attention. That is, “ambitions, develop with a specific situation, that it is a response to the possibilities which lie before the politician” (Schlesinger 1966, 8). Yes, ambition is a means to an end, but ambition is a product of a member’s situation within a given institution. Schlesinger (1966) outlined three types of political ambition—progressive, static, and discrete. Legislators who desire a higher office exhibit progressive ambition. In particular, legislators want to serve in a legislature that will “maximize their preferences” (Hibbing 1999, 154). It is well understood that legislators desire to serve in more professional legislatures, which have higher pay and more staff (Hibbing 1999). However, this is altered by the presence of term limits—professional legislatures may offer greater incentives, but term limits put an abrupt end to a legislator’s political career. Proponents of term limits argued that term limits would bring an end to careerist legislators and bring about citizen legislatures, which would focus more on constituents than on political gains (Cain and Levin 1999; Caress 1996). Political ambition is partly dependent on legislative professionalization and career opportunity structures (Moncrief et al. 1992). Members with greater career opportunities are more likely to advance, those with fewer opportunities are less likely to seek higher office (Squire 1988; Schlesinger 1966). Similarly, legislators who served in a more professional legislature would be less likely to be termed out of office because they would leave earlier to seek higher office (Moncrief et al. 1992). But increased political ambition is actually more nuanced than many of these studies suggest. These works tend not to take a comparative approach; therefore any differences between types of term limits remain largely unobserved. A comparative approach strengthens our understanding of political ambition and the differences that are present among the term-limited states. Studies of term limits and legislative careers have revealed many inconsistencies. Before term limits went into effect, the anticipated changes were thought to be limited (Francis and Kenny 1997; Caress 1996; Moncrief et al. 1992). After term limits were passed, the theory then became—members in term-limited states are more ambitious (SarbaughThompson and Thompson 2017; Sarbaugh-Thompson et al. 2010; Lazarus 2006; Carey et al. 2000). It is true that since the implementation of term limits many legislators have had their career forcefully ended. Yet, there is a consensus that term limits have induced greater political ambition across the states (Sarbaugh-Thompson and Thompson 2017; Sarbaugh-Thompson et al. 2010; Carey et al. 2000). Unfortunately,

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

53

existing theories of term limits and careerism are not expansive enough to account for the different types of term limits. Using career movement and legislative service as a way to assess displayed ambition, this analysis can further understanding of legislative careers. A more complete picture of careerism can help to explain membership behavior extending from leadership and committee work to executive branch interactions. While it is nearly impossible to measure one’s political ambition, as it is more of a personality trait. It is possible to examine the careers of legislators to evaluate their career trajectories and ultimate career pathways. Rather than attempting to predict why lawmakers do what they do, this assessment is primarily observational in nature, with predictions of what factors most influence lawmakers. Political career motivations are not linear but are multifaceted and the result of a number of contributing factors. To demonstrate this, I outline a model for understanding careerism incorporating the term limit variations. Following that, I present an evaluation of the different types of political ambition— discrete, static, and progressive. In order to supplement the variability of political ambition and what it means to legislators, there are also legislator interviews to gauge how legislators view their own political aspirations.

Understanding Ambition In this section, I present a model for understanding the career pathways that are observed in the state legislatures. Our previous understandings of a stepping stone approach from the lower to upper chambers of a state legislature and onward to Congress no longer apply. This model not only takes into account the variation of state legislatures, but it is rooted in the career pathways that members are actually carving out for themselves. Now that term limits have reached a point of equilibrium, changes to the states have settled. Consequently, it is easier to observe and understand career movement. The structure of the legislature and the presence of term limits shape the career of a state legislator. When there is only a finite amount of time to serve, the decision to move to a different office is shaped, in part, by that rule. One of the most important changes brought by term limits is that the institution itself now plays a greater role in the career calculus of members. In other words, a certain type of term limit can encourage members to prematurely seek higher office, while members operating under different limits stay until they are termed out of office.

54

J. BUTCHER

Ambition can take many forms, given the different types of term limits and their restrictions; ambition should take on different forms even among term-limited states. While it is difficult to qualitatively assess a lawmaker’s motivations and intentions, it is easier to follow where lawmakers go, from this we can infer their ambition to an extent. This chapter, while centered on the premise of ambition is interested in those career trajectories of members who pursue office. The careers of term-limited legislators are shaped on two distinct levels. There is an extra restriction facing members of term-limited states that members in other states do not face. The first level is the type of term limit, this shapes the different opportunities available to members. The type of term limit creates the different pathways that a legislator may pursue. Some term limits do not allow for certain career pathways. The pathways are based on the strictness of the limit and the ability of members to later return to the legislature. For example, term limits with a lifetime ban do not allow members to return to the legislature once they have termed out of office. Therefore, members in states with a lifetime ban do not have the opportunity to return once they have termed out—their pathway is blocked. The second level is the legislative structure, including professionalization and career advancement opportunities: these are the same institutional factors all legislators face to some extent. Their options, however, are restricted by the type of term limit implemented in the legislature. The career choices that members make are dependent on the incentive to serve, or legislative professionalization (Squire 1992), as well as, the ability to serve in a different position, the career opportunities (Squire 1988). While each state presents a limited number of advancement opportunities, some states offer greater prospects for progressive career movement than do others. Each type of term limit presents opportunities for the different forms of ambition: progressive, static, and discrete (Schlesinger 1966). Table 3.1 presents the original three-part typology from Schlesinger (1966), with an additional breakdown of different variations of each type of ambition. The original three types of ambition will manifest in different ways based on the term limits a member is facing as well as professionalization and career opportunity structures. Looking first at progressive ambition, there are two different ways that members can display this form of ambition. First, there will be some members who advance to a different office before they are termed out.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

55

Table 3.1 Defining the different types of ambition in term-limited states Progressive (Desire to Advance)

Static (Making a Career Out of Service)

Discrete (Serve and Then Leave)

1. Early Advancement Leaving before being termed out 2. Termed Advancement Advancing after being termed out 1. Downward Moving from the upper to the lower chamber 2. Persistent Leaving and returning to the same chamber 3. Teetering Moving back and forth between chambers 1. Retire Voluntarily leaving the legislature 2. Term Out Forcefully removed by term limits

The second form of progressive ambition is termed advancement. With this type of ambition, members are still advancing to higher office but they wait until they have completed their allotted term. Thus, they term out and then advance. The second type of ambition, static ambition, manifests in three different ways. When a member has static ambition, they are making a career out of their legislative service. In order to make a career out of legislative service in term-limited states, members have to move either between chambers or leave and return. The first type of static ambition is downward ambition, this is the act of moving from the upper to the lower chamber. This form of ambition occurs when state senators desire to continue serving but have no option except to move to the lower chamber. Second, there is persistent ambition, this is when members leave for an election cycle but then return to their previous seat once they are allowed (under the term limit provision) to return. Last, there is teetering ambition. Teetering ambition is the act of moving back and forth between chambers to a point where legislative careerism is functionally uninterrupted. Within each type of term limit, it is the differences that direct career movement. States with more career opportunities make it easier for members to seek higher office. As a result, states that offer greater ability to advance have more progressively ambitious members. States with limited opportunities are more likely to have members who display discrete or static ambition. Professionalization, as an incentive, can

56

J. BUTCHER

encourage members to stay in the state legislature longer. Thus, highly professionalized states encourage static ambition, but states with low professionalization can discourage service and encourage discrete ambition. Given the diverse nature of career ambitions and the different pathways at play, I present 4 hypotheses based on each type of term limit and the expected observations. Hypothesis 1: The unified/lifetime limits are strict and do not allow members to return to office once they have been termed out. As a result, career pathways are limited primarily to single-term service.

Members in unified states are less likely to move between chambers, because of the uncertainty that comes with seeking a different office. There are two different pathways that members in unified/lifetime states are likely to take. First, those that display progressive ambition are more likely to advance beyond their current seat by seeking higher office before they term out. Second, those who exhibit discrete ambition will retire from the legislature. Unified limits allow members to serve for a longer period of time in one chamber; those who want to leave public life will choose to retire early rather than waiting to term out of office. Members are far less likely to exhibit static ambition under this type of limit. Instead, members will not move from their seat but will stay until they are forcefully removed. Hypothesis 2: Unified/consecutive term limits cut down the options for members, but do allow members to advance beyond the legislature or return to their previous seat. Given the short terms, members will stay in office until removed by term limits.

The second type of limit is the unified/consecutive limit. Nebraska is the only state in this study with a unified/consecutive limit. In Nebraska, there are three obvious pathways that members may choose to take. First, progressive ambition, while unlikely, will not occur until members term out. For a member to term out they only serve two four-year terms, if a member seeks higher office they likely need this political experience. Progressive ambition is the least likely pathway because of the limited opportunity to advance within the state legislature. The second pathway is for members who exhibit static ambition. Members in Nebraska who desire to serve will exhibit regressive ambition. This type of ambition means that members will leave their legislative seat and then later return to it after sitting out a term. Members who are not ready to leave political

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

57

office, but do not want to advance to higher office will simply return to their previous seat, as allowed by the consecutive limit. The third pathway is for members who exhibit discrete ambition to be termed out and leave public life. Members only have the opportunity to serve two consecutive terms. Members with discrete ambition may have an agenda; they serve their terms and pursue their agenda only to leave public life. Hypothesis 3: Divided/lifetime states have fewer career options available because the lifetime ban restricts the ability of members to return to the legislature.

Members will not move from their seat, unless removed by external factors. There are four primary paths of ambition for these members: termed advancement, downward, persistent, and term out. These pathways assume that members want to maximize their time in office; because of the lifetime ban they cannot return, so they will stay in each position for as long as possible. Members who display progressive ambition will wait until they are termed out of office before pursuing a higher position. Again, this is because ambitious members want to maximize their time in the legislature. Members who display static ambition and want to maintain their legislative service will move downward from the upper chamber to the lower chamber to stay in office. Members of the upper chamber who want to stay in office but have termed out have no choice but to move to the lower chamber. There is also the possibility, in the case of an election loss or other removal, that a member may leave the legislature and then return—an example of persistent ambition. Finally, members with discrete ambition will leave once they have termed out of office; this will typically be restricted to one chamber. Members with discrete ambition are unlikely to seek another term in the legislature. Hypothesis 4: Divided/consecutive term limits are the most lax limit and allow members to serve at their leisure; the ability to freely return to the legislature does not constrain members to serving for their entire term and they can easily have a career similar to politicians in nonterm-limited states.

The final type of term limit, divided/consecutive, is the most common type of limit. It is also the laxest type of term limit. As a result, there are five different pathways that a member might take. First, members who are progressively ambitious are more likely to advance early, or before they

58

J. BUTCHER

term out. This lax limit allows for all three types of static ambition: downward, persistent, and teetering. Members who are in the upper chamber, but have static ambition, are more likely to exhibit downward ambition by moving to the lower chamber. Other members have the ability, under a consecutive ban, to move between chambers. This is called a teetering movement and occurs when members “teeter” between the two state legislative chambers. Finally, those with discrete ambition are more likely to retire than be termed out of office. The lax limits make it easier to serve in the legislature long term, while those who do not have that desire will retire and leave. In sum, a legislator’s career decisions are shaped by the legislature they serve in and the term limits they are constrained by. Career decisions must pass two levels: institutional structure and term limit type. Legislators, in all states, have careers shaped by institutional factors like professionalization and career opportunities. In term-limited states, members face an extra constraint. In most states, term limits restrict career outcomes. Divided/consecutive states, however, have far more career opportunities because of how lax the limits are. This two-level approach to careers helps to explain some of the differences that exist among term limits and the careers of lawmakers.

Analysis The individual-level TLS-Career data is utilized to evaluate the careers of members who have served in the 17 states used in this study. This examination tests the diverse nature of term limits using a multinomial logit. In life, members must make choices about what pathways they will pursue by considering all of the available options. A multinomial logit is a model that accounts for the different choices or pathways that a member can take. This analysis accounts for the different options members have from which to choose including, leaving the legislature, seeking higher office, or even waiting to term out of office. In each analysis, there is a base category, such as seeking reelection, and the results are the likelihood of a member choosing a particular pathway rather than pursuing progressive ambition. While this assessment does not reveal the thought process that members go through each election cycle, the observed career choices serve as a proxy for different forms of ambition.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

59

Each observation is an individual for each year they served.1 Additionally, members do not always leave according to the electoral cycle; given that the purpose of this study is to examine careers it would be misleading to exclude those who do not fall within the typical electoral parameters. Rather than limiting the analysis to electoral variation only, each year that a member serves is included, as members can freely leave a chamber before the end of their term. To account for the added turnover during election years, there is an added control to account for election years.2 There are different dependent variables used in each of the following analyses. Each analysis uses the same independent variables, which are indicators for the different types of term limits. The expected observed behavior will be dependent on the term limit type and the options presented to members. To capture institutional characteristics, each analysis includes controls for professionalization and career opportunity structure. Professionalization is a measurement of a legislature’s time in session, member salary, and staff support (Squire 1992, 2017). Legislative professionalization, a key variable used throughout this text, is a proxy measure for legislative incentives. The benefits or resources provided by the legislature increase the incentive for members to serve, thus they might be more willing to take a less traditional path to stay in office. Career opportunity is a measure of the advancement prospects in the state legislature and an indicator of the opportunities to move beyond the current chamber (Squire and Moncrief 2010). The future political prospects are something that members will weigh heavily when deciding what their legislative careers will look like. Specifically, this measure of career opportunities equates to states that are either “spring boards” into a higher office or “dead ends” with no potential for political advancement. The third type is a “career” legislature where members will work to make a career out of service; this is where lawmakers might, once again, take a unique approach to serve in office. To help account for some of the individual-level variation and some of the additional reasons members might leave, there are four legislatorspecific controls. These additional controls include an indicator for 1 Rather than utilizing data for each session, I choose to look at each individual year for consistency. Not all states have the same election year, so broader national-level trends would be lost. 2 Alternative analyses looking at career movement during election years only yield little difference in the variables that are statistically significant.

60

J. BUTCHER

Republican members because they have been found to be less ambitious at the congressional level (Ang and Overby 2008). Term limits were meant to increase diversity in state legislature, so there are also controls for member race and gender (Sarbaugh-Thompson and Thompson 2017; Straayer 2007; Carey et al. 2000). The fourth legislator-specific control is age, which may impact a member’s decision to stay or go.3 There are also controls to help capture the circumstances that members may leave under. First, there is an indicator for members who were part of the chamber majority as they were exiting. When members are part of the legislative majority, it essentially makes their job easier; for members of the minority party there are more roadblocks to legislating and there are fewer perks for serving, like leadership roles. Second, there is an indicator for members who are part of the governor’s party at the time that they left. Counter to the majority party variable, members may ride the coattails of Governors and seek out higher positions or statewide offices when they are from the same party. Last, there is a control for election year.4

Ambition Pathways The following examination outlines the observed behavior of state legislators, while noting some of the contributing factors for why members act in the manner that they do. To begin, there is a broad overview of career trends using binary indicators for when a member leaves office and displays discrete, static, or progressive ambition. For example, for discrete ambition, any legislator who has retired or termed out of office is coded as (1), all others are not (0). For static movement there is an indicator if a member attempts to stay in the legislature by seeking reelection (1) or leave through either retirement or advancement (0). The final dependent variable is an indicator for progressive ambition, advancing (1) versus not advancing (0).5 Given the binary nature of the dependent variables, there is a logit regression for each variable for each legislative 3 Age was one of the most difficult variables to find for members, given the missing

data, age is not included for all analyses. 4 Louisiana and New Jersey are the only states in the data with odd year elections. 5 Progressive ambition is a measure of whether or not a legislator leaves their seat in

pursuit of higher office, i.e., Congress or state-wide elected position, this measure is not based on whether or not they win that election. Even when a member does not win a higher office, it is necessary to note the personal drive that encouraged them to run.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

61

chamber. Table 3.2 displays the aggregate analysis of the likelihood of state legislators pursuing the different types of ambition.6 To be clear, this analysis does not incorporate the decision made by members to choose one path over another, but the likelihood of experiencing a particular avenue versus all other options. In divided/ consecutive states, members are significantly more likely to pursue all ambition pathways than those who do not face term limits. Recall, the divided/consecutive term limits are the most lax form of a term limit and allow members to return to office. It is also worth noting that this is the most popular type of term limits across the states that have term limits. The divided/life states shape careers differently than the other form of divided limit. The clear differences in the two types of limits indicate that there is something unique in how lifetime bans shape careers. Specifically, there is a lack of evidence supporting the existence of static ambition where divided/life states are no different from those without limits. Progressive movement in these states is significant in both chambers (p < 0.05). In sum, members from divided/life states are more likely to exhibit discrete and progressive ambition than those who serve in states without term limits, but are not all that different from states without term limits when it comes to static ambition. The results for unified/consecutive limits are constrained to a single chamber because Nebraska is a unicameral legislature. Members of the Nebraska legislature are significantly more likely to exhibit discrete and progressive ambition than are members from states without term limits (p < 0.05). Furthering this finding, Nebraska senators are significantly less likely to experience static ambition than states without term limits. The unicameral legislature does limit the static ambition options; members can only leave and return, as two out of three forms of static ambition involve a second chamber. Overall, the unified/consecutive term limit is unique in that members are more likely to exit the legislature and permanently retire from political life or simply advance. The final type of limit is the unified/life limit. This type of limit is significantly different from states without term limits in three different ways. First, members of the lower chamber are significantly more likely to choose a discrete pathway (p < 0.05). Second, members of the lower chamber are significantly less likely to exhibit static ambition, a result of 6 With a small N for some of the indicators, there are some indicator variables that are dropped from this analysis because they are perfect predictors.

Age

Male

White

Republican

Career opportunities

Professionalization

Unified/life

Unified/consecutive

Divided/life

1.059*** (0.175) −0.934 (0.770) 1.801** (0.839) −0.241*** (0.082) 0.100 (0.071) 0.099*** (0.037) 0.031*** (0.002)

0.688*** (0.164) 1.291*** (0.142)

0.726*** (0.134) 1.226*** (0.159) 1.707*** (0.208) 1.044*** (0.157) −2.085*** (0.667) 2.048*** (0.628) −0.270** (0.117) −0.066 (0.099) 0.042 (0.064) 0.002 (0.002) −1.511*** (0.296) 0.271 (3.549) −2.462* (1.287) −0.098 (0.116) −0.158 (0.303) 0.966*** (0.299) 0.069*** (0.010)

1.582*** (0.328) 0.282 (0.426)

Lower

Lower

Upper

Static

Discrete

Logit analysis for each type of political ambition

Divided/consecutive

Table 3.2

2.819*** (0.640) 1.237* (0.643) -0.481 (0.602) −0.331 (0.581) 2.479* (1.487) −3.294* (1.752) 0.043 (0.214) −0.138 (0.373) 0.027 (0.229) 0.001* (0.001)

Upper

1.428*** (0.383) −0.743 (1.487) 1.821* (0.994) 0.085 (0.191) −0.291* (0.152) −0.104 (0.093) 0.001 (0.003)

2.396*** (0.328) 2.122*** (0.406)

Lower

Progressive

3.335*** (0.850) 2.223** (0.936) 3.173*** (1.136) 2.980*** (0.961) 5.094** (2.163) −1.839 (2.228) −0.161 (0.210) −0.018 (0.541) −0.285 (0.397) −0.034** (0.015)

Upper

62 J. BUTCHER

−0.263 (0.205) 0.095 (0.166) −5.770*** (0.810) 14,478

0.549*** (0.129) 0.221** (0.110) −3.907*** (0.246) 14,478

−0.471*** (0.159) 0.160 (0.156) −9.851*** (1.040) 35,770

0.362*** (0.119) 0.073 (0.063) −5.637*** (0.243) 35,770

Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1

Observations

Constant

Same party as governor

Part of chamber majority

Upper

Lower

Lower

Upper

Static

Discrete

0.220 (0.139) 0.182 (0.115) −6.209*** (0.481) 35,770

Lower

Progressive

0.378 (0.343) 0.840** (0.332) −7.481*** (1.455) 14,478

Upper

3 LEGISLATIVE CAREERISM AND GROWING AMBITION

63

64

J. BUTCHER

the lifetime ban.7 Last, members of the lower chamber are more likely to display progressive ambition. All of the significant differences are present in the lower chamber; this finding indicates that unified/life limits have career differences based on the chamber that a lawmaker serves in. The control variables present some interesting findings. Specifically, the two institutional controls do not always work in conjunction. Professionalization appears to only affect the upper chamber, even then, only for static and progressive ambition. Career opportunities appear to have a stronger relationship with discrete movement, but have a negative influence on static movement in the upper chamber. The Republican, White, and male controls only have a significant relationship with progressive movement from the lower chamber. While Republicans are significantly more likely to seek higher office from the lower chamber, White and male lawmakers are significantly less likely to seek higher office. It is also worth noting that male legislators in particular are more likely to pursue static pathways to lengthen their political tenure. Age is statistically significant in each model, but negative only for members of the upper chamber who pursue higher office. This could be a reflection of the demographics of state upper chambers, where members are older and more experienced, making it less likely that older members are going to advance to state-wide or federal positions. There are some important findings when evaluating the big-picture partisan politics surrounding members when they leave office. Members who are part of the chamber majority are significantly more likely to display discrete ambition than those who are not. Conversely, those who are part of the majority party in the lower chamber are significantly less likely to display static ambition, meaning they are less likely to leave and return. The results for the governor are quite different; those who are of the same party are more likely to return to office and to display progressive ambition. Last, it is important to note some of the nuances that exist from chamber to chamber. When looking at the different types of term limits across the different forms of turnover, we start to see that there is not a single pattern for each legislative chamber. In other words, the overall effect of the term limits does vary based on the chamber where a member serves. For example, static ambition increases for those facing a divided/ 7 Results from the upper chamber are dropped from this analysis because of the limited number of observations.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

65

consecutive limit and serving in the lower chamber. However, this relationship is insignificant for members of the lower chamber who face a lifetime ban under a divided limit. Figure 3.1 illustrates the probability of members choosing certain ambition pathways based on the analysis above. This figure contains a graph for each type of ambition for each legislative chamber. Figure 3.1 compares each type of term limit to the states without term limits by presenting the mean and 95% confidence intervals. For each model, institutional controls and Age are held at their means in order to see the substantive effects of each type of term limit. The individual-member demographics are set to represent Republican, White, males, as this is the most common legislator in the dataset. Additionally, the probabilities are set to represent lawmakers who are of the chamber majority and share the same party as the Governor.8 This figure helps to show under what circumstances the term-limited states shape retirement similarly, as well as when the states are different. There is a great deal of variation in the choice to pursue different pathways, this variation is based on both term limit type and the chamber that members serve in. In the upper chamber, each type of term limit is more likely to see discrete ambition. While the divided/lifetime limits are not distinguishable from the other two types of limits, it is clear that each form is statistically different from states without term limits. When it comes to static ambition, divided/consecutive members are more likely to exhibit static ambition, but divided/life and unified/consecutive are not statistically distinguishable from states without limits. Last, progressive ambition in the upper chamber is significantly more likely to occur in divided/ consecutive states. While the other states do have higher rates of progressive ambition, they are not statically different from states without term limits. The lower chamber presents slightly different results. Discrete ambition is more common in the lower chamber among all types of term limits. Each type of term limit has significantly higher percentages of discrete ambition. The difference between the two types of lifetime limits is not statistically different indicating something unique about how lifetime bans shape careers of members in the lower chamber. Static ambition is more likely to occur in those states with divided/consecutive limits, 8 This particular scenario represents a Republican trifecta, which has become increasingly common among the term-limited states.

66

J. BUTCHER

Upper-Discrete

Upper-Static

0

.015 .01

0

0

.01

.005

.02

Probability (%) .05 .1

.03

.15

.04

.02

Upper-Progressive

DL

UL

.15

DL

UC

NL

DC

DL

UL

Lower-Progressive

.01 0

0

0

.005

.02

.01

Probability (%) .05 .1

DC

Lower-Static

.015

Lower-Discrete

NL

.04

DC

.03

NL

NL

DC

DL

UL

NL

DC

DL

UL

NL

DC

DL

UL

Fig. 3.1 Predicted percentage of pursuing each of the three types of ambition (Note NL [No Limit], DC [Divided/Consecutive], DL [Divided/Life], UC [Unified/Consecutive], UL [Unified/Life])

but is also present in states with no limits and divided/life limits. Finally, progressive ambition is significantly more likely to occur in all of the termlimited states, but this difference is less noticeable for states with divided/ consecutive term limits. Members from the lower chamber seek higher office at an increased rate in states with term limits, rather than those without. The differences between the types of term limits, however, are not distinguishable. Discrete Ambition Discrete ambition is observed when members leave office in lieu of staying or seeking another elected office. For the purpose of this analysis, discrete ambition is displayed by members in two ways: first, when a member elects to retire; second, when a member terms out and chooses not to return

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

67

to the legislature. The first, retirement, is an active choice made by the legislator to leave. While terming out of office is not a choice, but a forced decision. The difference between retiring and terming out is that those who term out of office are ineligible to seek reelection, whereas those who retire willingly pass up that opportunity. It is well understood within political science that a member’s primary goal is to win reelection (Mayhew 1974), but what do members do when their opportunities to do so are constrained or simply unavailable? To assess the choice between the two types of discrete ambition, retirement and terming out, I use a multinomial logit model (MNLM) comparing these options versus staying in the legislature for a full term to, presumably, seek reelection.9 Now, not all members will leave office or seek reelection, which is why this model also accounts for both static and progressive ambition (broadly speaking). The use of the MNLM allows for the comparison of multiple opportunities, much like members must in real-life weigh the career pathways that are in front of them. Those who do not seek reelection are distinct from those who try to stay or seek out a higher office. Table 3.3 contains the results of discrete ambition compared to other avenues of ambition. For each column, the pathway of interest, such as, retiring or terming out of office, is compared to the likelihood of seeking reelection. Given the sheer number of opportunities available, the interpretation of the MNLM model can become exhaustive, so I highlight the most important results focusing on discrete ambition. Recall from the aggregate analysis, discrete ambition was more likely to occur in all term-limited upper and lower chambers. This analysis reveals that when comparing the choice of staying in office to seeking reelection, members are not more or less likely to retire or term out of office. These results carry few exceptions, those in the lower chamber of divided/ consecutive states are significantly less likely to retire. In divided/lifetime upper chambers members are significantly less likely to prematurely retire than run for office. This result is due to the nature of serving under a lifetime ban, knowing that their time serving in the upper chamber is really the last opportunity to serve in the state legislature, given they cannot 9 Recall, each observation is individual/year but staying in office each year implies pursuit toward staying in office or seeking reelection. Whereas some early scholars predicted that under term limits members might leave the legislature or retire strategically before the end of their term (see Moen et al. 2005; Carey et al. 2000).

Same party as governor

In chamber majority

Age

Male

White

Republican

Career opportunities

(0.483) 19.073*** (0.527) 18.439*** (0.524) 1.472 (1.143) 0.690 (1.019) −0.284*** (0.085) 0.229** (0.115) 0.157*** (0.037) 0.043*** (0.004) 0.271*** (0.084) 0.068

(0.146) −0.099 (0.281) 0.204 (0.140) −2.876*** (0.873) 1.844***

(0.669) −0.272* (0.146) 0.041 (0.157) 0.051 (0.064) 0.014** (0.006) 0.451**

(0.192) −0.066

(0.225) 0.339*

(1.320) −0.259 (0.203) −0.168 (0.259) 0.890*** (0.310) 0.063*** (0.013) −0.052

(0.408) 0.736* (0.409) −1.592*** (0.318) −0.648 (3.191) −1.862

(0.138) 0.138

(1.006) 0.075 (0.178) −0.266* (0.143) −0.097 (0.097) 0.003 (0.003) 0.210

(0.365) 2.125*** (0.431) 1.285*** (0.404) −0.489 (1.378) 1.464

2.386***

(0.206) 0.302*

(0.880) −0.258 (0.197) −0.015 (0.136) 0.126 (0.110) 0.011 (0.018) 0.402*

(0.227) −0.373 (0.484) −0.097 (0.215) −1.668 (1.259) 1.343

(0.243) 0.173

(2.446) −0.171 (0.195) 0.242 (0.201) 0.117 (0.109) 0.012 (0.021) 0.070

(0.570) 1.871*** (0.658) 0.760 (0.717) 3.302 (2.954) −3.847

1.969***

1.672***

−0.084

18.410***

0.135

Term out

Progressive

Retire

Static

Retire

Term out

Upper

Lower

Multinomial logit of discrete ambition for each chamber (base, reelection)

Professionalization

Unified/life

Divided/life

Divided/ consecutive

Table 3.3

(0.227) 0.349

(1.775) −0.471** (0.232) −0.027 (0.430) 0.142 (0.247) 0.011 (0.021) 0.241

(0.809) 1.659** (0.730) −12.538*** (1.191) 2.519* (1.324) −3.621**

3.419***

Static

(0.293) 0.927***

(1.684) −0.025 (0.192) 0.161 (0.525) −0.249 (0.402) −0.035** (0.017) −0.027

(0.458) 0.659 (0.561) 0.722** (0.302) 7.532*** (2.339) −5.687***

2.154***

Progressive

68 J. BUTCHER

(0.199) 1.853*** (0.235) −11.028*** (1.199) 34,968

(0.116) 1.780*** (0.216) −7.112*** (0.552) 34,968

(0.172) 1.448*** (0.258) −5.048*** (1.081) 14,032

(0.063) 2.088*** (0.216) −25.996*** (0.646) 34,968

(0.106) 1.694*** (0.233) −5.038*** (0.576) 34,968

Standard errors in parentheses (by state) ***p < 0.01, **p < 0.05, *p < 0.1 Unified/consecutive term limits are dropped from this analysis due to the lack of observations

Observations

Constant

Election year

Retire

Progressive

Term out

Retire

Static

Upper

Lower

(0.134) 2.311*** (0.402) −6.881*** (1.501) 14,032

Term out (0.256) 1.724*** (0.230) −8.197*** (1.407) 14,032

Static

(0.323) 2.231*** (0.265) −6.816*** (1.113) 14,032

Progressive

3 LEGISLATIVE CAREERISM AND GROWING AMBITION

69

70

J. BUTCHER

return, it is unsurprising that members would instead choose reelection at a higher rate. While the statistically significant findings are lacking, this may actually carry greater substantive significance. The analysis in Table 3.3 contains the observed movement of lawmakers during each election cycle. This evidence suggests that counter to what we believe to be true for members of Congress and other state legislatures, members, in most cases, are not far more likely to stay in office than they are to retire. Although prior research has uncovered that there are fewer retirements in term-limited legislatures (Butcher 2021), these findings reveal that while that may be the case, the difference between choosing to retire or seek reelection is insignificant. Perhaps the most interesting distinction in these results are that individual and party factors appear to matter more in the lower chamber than the upper. Not only does this say something about the individual-level demographics of members, but also about the relationship that lawmakers may have with their party. Those who are in the upper chamber are often assumed to have higher levels of experience, which can play a role in their relationship with the party/party elites. Whereas in a lower chamber, members are new, less-experienced, and have an increased reliance on the party that can then pressure members to pursue particular positions. The broader findings include that members are, in many states, significantly more likely to seek static or ambitious avenues to staying elected. Again, we start to see the nature of experience come out when looking at the results from the lower chamber as compared to the upper. Aside from election year, the control variables offer little added explanation. Professionalization and career opportunities work together. As legislative professionalization declines so does the rate of retirement in term-limited states. Meanwhile, greater career opportunities encourage members to leave the legislature in pursuit of other opportunities. Republicans, in particular, are less likely to retire, as are older members. Race and gender appear to vary somewhat but reveal no underlying trends. Similarly, party ties to the governor reveal no discernible patterns based on chamber or career pathway, whereas those in the chamber majority do appear to leave at higher rates. Overall, the observed career patterns imply that members are less likely to choose discrete pathways, but are more likely to pursue other static and progressive pathways.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

71

Static Ambition The second type of ambition is static ambition, which implies that a member is attempting to make a career out of legislative service. Traditionally, members would seek reelection to make a career out of service but under term limits that is not possible. There are three types of static ambition unique to states with term limits: downward, persistent, and teetering. Downward ambition is when members move from the upper to the lower chamber. Persistent ambition is when a member leaves the chamber and returns to the same seat taking a time-out period. Teetering ambition is when a member alternates between legislative chambers, going back and forth. Once again, the baseline for this analysis is the decision to seek reelection. For some lawmakers, static ambition or finding a way to continue serving in the legislature may be viewed as an easier option than seeking higher office but will not always be possible. The results of the MLMN for static ambition can be seen in Table 3.4. Given that downward static ambition is defined by movement from the upper chamber to the lower chamber, there are no estimates for downward movement for those who serve in the lower chamber. The presence of different types of static ambition varies by both type of term limit and legislative chamber. Given the number of possibilities and outcomes using the MNLM, I want to draw attention to a couple of important distinctions. The first results to note are the differences that exist between the states with unified and divided term limits. This classification of unified versus divided tends to receive less attention, but as noted in this analysis it matters how the years of service are organized. In the states where members face divided term limits, members appear to take a more flexible approach to service by returning, leaving, and weaving all throughout the legislature. In states with unified term limits, there is far less movement, given that members have the ability to serve most (or all) of their time in a single chamber, it looks like they tend to stay there. The long-term experience that can be gained from serving in a legislature with a unified term limit is unmatched in the divided states. Members make a calculated decision choosing to stay where they have greater experience (i.e., greater institutional power) rather than to seek out a position in the other chamber. Second, the distinction between the lifetime and consecutive bans for divided term limits manifests different career trajectories. In fact, states

Unified/life

(0.787) −0.284** (0.115) 0.177*** (0.062) 0.065** (0.032) 0.412*** (0.125)

(2.557) −0.335 (0.504) 0.566 (0.480) 0.877** (0.447) −1.173*** (0.207)

(0.131) 0.961*** (0.164)

0.590*** (0.195) −1.067 (0.783) 1.103

(0.874) 0.594 (1.289)

(0.296) 0.322 (0.339)

0.523***

Discrete

−10.267*** (1.418) 0.026 (4.612) 1.526

3.191***

Teetering

1.246***

Persistent

Lower

(0.114)

(0.880) 0.082 (0.159) −0.356*** (0.130) −0.071 (0.101) 0.318***

1.544*** (0.334) −0.942 (1.082) 2.632***

(0.304) 2.202*** (0.343)

2.298***

(0.205)

(0.476)

(2.271) 0.589 (0.417) −0.454 (0.536) 0.146 (0.244) −0.679

(0.663) 0.275 (0.511) 0.925 (1.655) −3.547

(1.180) −15.034*** (1.160) 1.477 (2.260) −2.328 (1.814) −0.491** (0.200) −0.058 (0.431) 0.154 (0.317) 0.202

(0.491) 1.795*** (0.656) 0.517

2.447***

Persistent

(0.581) 1.162* (0.614) -15.214***

2.494***

Progressive Downward

Upper

Multinomial logit of static ambition for each chamber (base, reelection)

−1.357*** (0.254) Professionalization −2.261 (2.656) Career −2.032** opportunities (0.889) Republican 0.159 (0.128) White −0.168 (0.342) Male 0.823*** (0.248) In chamber −0.257* majority (0.138)

Unified/ consecutive

Divided/life

Divided/ consecutive

Table 3.4

(0.328)

(2.682) 0.175 (0.172) 1.340 (0.858) 0.313 (0.401) −0.645**

(1.519) −14.614*** (1.464) 4.659** (2.079) −3.000

(1.087) -14.419*** (1.270) -15.182***

3.711***

Teetering

(0.139)

(0.454) −0.351** (0.152) 0.059 (0.105) 0.131** (0.059) 0.515***

(0.151) 0.471*** (0.130) −1.849*** (0.560) 1.524***

(0.114) 0.998*** (0.127) 1.227***

0.548***

Discrete

(0.330)

(2.032) −0.276 (0.228) 0.153 (0.540) −0.238 (0.330) 0.483

(1.124) 3.163*** (0.977) 5.297** (2.382) −1.404

(0.880) 2.378** (0.972) 3.333***

3.375***

Progressive

72 J. BUTCHER

(0.291) 0.244 (0.491) −10.498*** (1.342) 46,059

(0.159) 1.051*** (0.149) −6.253*** (0.696) 46,059

(0.070) 1.738*** (0.179) −4.443*** (0.223) 46,059

0.059

−0.747**

0.164 (0.078) 1.779*** (0.200) −7.309*** (0.430) 46,059

0.205*** (0.210) 1.883*** (0.260) −7.352*** (0.901) 16,803

0.291

Progressive Downward

Upper Teetering

(0.245) 0.054 (0.463) −6.194*** (0.839) 16,803

(0.232) 0.237 (0.468) −10.250*** (1.560) 16,803

−0.505** 0.162

Persistent

(0.102) 1.697*** (0.236) −4.480*** (0.251) 16,803

0.209**

Discrete

Robust standard errors in parentheses (by state) ***p < 0.01, **p < 0.05, *p < 0.1 Age is excluded from this analysis because of collinearity, excluding age from the model results in nearly 10,000 more observations

Observations

Constant

Election year

Same party as governor

Discrete

Teetering

Persistent

Lower

(0.304) 2.053*** (0.279) −10.818*** (1.142) 16,803

0.839***

Progressive

3 LEGISLATIVE CAREERISM AND GROWING AMBITION

73

74

J. BUTCHER

with divided/lifetime term limits are similar to those states with unified term limits. For some, this result may be unsurprising given how lax divided/consecutive term limits are. Those who serve under divided/ consecutive term limits have more opportunities given the flexibility to oscillate between chambers and it seems members are well aware of that ability. In divided/consecutive states, members of the lower chamber are significantly more likely to leave and return or teeter between chambers than they are to seek reelection. The results for the upper chamber are similar, with the inclusion of an increased likelihood of moving down from the upper to lower chamber. Of these patterns, the only one that holds true for those from divided/life states are moving down and returning in the upper chamber. Given this discussion, it is also important to understand what other factors may influence members pursuing these pathways. For those who leave and return in the upper chamber professionalization appears to be an underlying motivator. In a similar vein, when states have greater career opportunities present in the lower chamber, members are significantly less likely to leave and return to the lower chamber, as there are other positions available. Both party measures appear to affect a member’s willingness to continue serving in the legislature, but there are no strong trends. Those who are part of the chamber majority of the lower chamber are significantly less likely to pursue these less traditional career avenues, likely because the upper chamber is of the same party affording them greater advantages to advance. The results of this analysis on static ambition reveal two things about careers. Substantively, legislative careerism, or the efforts of long-term service have been redefined by term limits. However, it is also worth noting that the arguments surrounding term limits, to cut career politicians, have not come to fruition. Rather, those politicians who wish to serve in the state legislature will find ways of doing so, even if that means regularly starting over in the other chamber. Overall, static ambition varies by chamber and limit type. The most consistent finding is that lawmakers follow different pathways dependent on what chamber they serve in. Progressive Ambition The third type of ambition that members can display is progressive ambition. Progressive ambition, the act of seeking higher office, is the most commonly studied type of ambition. Given the nature of term limits,

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

75

there are two primary ways a member may display progressive ambition. First, members can prematurely seek higher office by leaving before they have completed their allotted term, i.e., early. Second, members can term out of office and then seek higher office, dubbed termed. Much like the observations made about static and discrete ambition, members are going to choose to seek higher office early as a result of the circumstances surrounding their position in the institution. Once again, a MLMN is used to compare the decision to seek higher office with the decision to seek reelection. As before, static and discrete ambition are included in the analysis for comparison. Members who display progressive ambition are actively making the choice to not leave elected service, even when they are unable to seek reelection. The results presented in Table 3.5 display more consistency across chambers and term limit types than what is seen with the other ambition pathways. The differences that do exist appear to be rooted in the chamber where the member serves. In the lower chamber, members are significantly more likely to seek either early or termed methods of higher office; those who serve in the upper chamber are not any more likely to leave office early than those who do not face term limits. For each type of term limit, lawmakers are significantly more likely to seek out at least one progressive ambition pathway versus reelection. In both forms of divided states, members are more likely to seek higher office both prematurely and after terming out in the lower chamber. The difference between the divided states becomes more noticeable when looking at trends in the upper chamber, where divided/consecutive limits see more members seeking higher office early and after terming out, but divided/life term limits only see the boost after members have termed out of office. This is where the important distinction between lifetime and consecutive bans comes into play. Knowing time is finite, members from states with a lifetime ban are less willing to give up their secure seat to seek a higher position; they do not do so until their allotted term has been filled. In unified/life states, members in both chambers are significantly more likely seek higher office before terming out of office (p < 0.05), but not before terming out simply because they cannot. In the lower chamber, displays of early ambition are significantly more likely to occur, while there is no significant relationship in the upper chamber. Overall, given the equation at hand, it is unsurprising that the term-limited legislatures primarily see members waiting to term out to seek higher office, rather

(0.129)

(0.308)

(0.142) 1.255*** (0.145)

0.924*** (0.178) −0.802 (0.785) 1.582** (0.766) −0.289*** (0.105) 0.136** (0.068) 0.109** (0.043) 0.029*** (0.002) 0.378***

(1.026) 4.020*** (1.029)

(0.325) 1.703*** (0.414)

0.676***

−12.668*** (1.439) 0.974 (1.554) −0.190 (1.372) 0.191 (0.206) 0.038 (0.249) −0.282** (0.138) 0.028*** (0.008) 0.106

4.327***

1.948***

(0.168)

−1.519*** (0.286) 0.660 (3.549) −2.620** (1.270) −0.093 (0.140) −0.218 (0.295) 0.964*** (0.307) 0.070*** (0.010) −0.387**

(0.334) 0.283 (0.437)

1.601***

Static

(0.606)

(1.043) −0.543 (1.222) 5.751*** (1.978) 1.067 (2.381) −0.067 (0.429) 0.511 (1.005) −0.008 (0.351) −0.004 (0.018) 0.255

(1.149) 2.245** (0.943) 4.930 (3.195) −3.557 (2.733) −0.217 (0.262) −0.146 (0.520) −0.384 (0.623) −0.062*** (0.014) 0.385 (0.323)

(0.603) 15.063*** (0.581) 17.182***

15.477***

Termed

(0.826) 1.499 (0.981) 0.149

2.991***

Early

Discrete

Early

Termed

Upper

Lower

Multinomial logit of progressive ambition for each chamber (base, reelection)

1.440*** (0.385) Professionalization −1.279 (1.479) Career opportunity 2.359** (1.006) Republican −0.007 (0.196) White −0.373 (0.244) Male 0.007 (0.108) Age −0.009*** (0.003) In chamber 0.305** majority (0.138)

Unified/life

Unified/ consecutive

Divided/life

Divided/ consecutive

Table 3.5

(0.129)

(0.198) 0.718*** (0.153) −1.811*** (0.680) 1.660*** (0.596) −0.279** (0.133) −0.000 (0.100) 0.130** (0.065) 0.003 (0.002) 0.460***

(0.130) 1.164*** (0.152) 1.381***

0.682***

Discrete

(0.204)

(0.585) −0.415 (0.569) 2.398* (1.403) −3.306** (1.684) 0.039 (0.211) −0.123 (0.383) 0.081 (0.221) 0.002 (0.002) −0.274

(0.638) 1.292** (0.644) −0.530

2.829***

Static

76 J. BUTCHER

0.033 (0.182) 1.452*** (0.208) −11.066*** (1.172) 35,770

0.193

(0.122) 1.998*** (0.269) −6.812*** (0.539) 35,770

Robust standard errors in parentheses (by state) ***p < 0.01, **p < 0.05, *p < 0.1

Observations

Constant

Election year

Same party as governor (0.076) 1.898*** (0.191) −6.441*** (0.254) 35,770

0.005 (0.166) 0.903*** (0.191) −10.218*** (1.018) 35,770

0.128

Static

(0.487) 1.836*** (0.385) −6.241*** (1.170) 14,478

0.917*

Early

Discrete

Early

Termed

Upper

Lower

(0.281) 2.981*** (0.564) −26.027*** (1.834) 14,478

0.616**

Termed

(0.099) 1.850*** (0.259) −4.875*** (0.289) 14,478

0.201**

Discrete

(0.164) 0.954*** (0.182) −6.200*** (0.813) 14,478

0.104

Static

3 LEGISLATIVE CAREERISM AND GROWING AMBITION

77

78

J. BUTCHER

than reelection. Again, when a member is barred from staying in office, their options are rather limited. What is of substantive importance here is that rather than retiring, where there were limited trends, members are seeking out higher office when they are unable to run for reelection. The broader picture of legislature diversity indicates that professionalization and career opportunities are still key factors to this relationship. Legislative professionalization contributes to a significant increase in termed advancement from the upper chamber (p < 0.05) because members like to stay in the legislature for as long as possible. Once again, the individual traits of members reveal little insight into broader demographic trends. What is perhaps of interest is that (in the lower chamber), older members are significantly less likely to leave office early for a higher position, but more likely to wait to term out. In other words, younger legislators who have more time for a lengthier career are more likely to seek higher office earlier on in their career. Collectively, this analysis of legislative careers reveals a great deal of variation among the states that have term limits. See Table 3.6 for an overview of the statistically significant findings from the examination of term limits and career ambition, as compared to the decision to seek reelection. For example, unified term limits offer fewer ambition pathways, whereas, divided limits offer far more opportunities to members. In particular, the divided/consecutive limit allows the greatest freedom to members and presents more potential career pathways than any other limit. This table not only illustrates the differences among the states but sheds light on the consistencies that exist among chamber type and across states with similar rules (i.e., unified versus divided or lifetime versus bans). Those who serve in states with lifetime bans see greater differences between chambers, whereas consecutive states see similar opportunities in either chamber. In unified/lifetime states, there is a greater variation within the lower chamber; this shifts to the upper chamber for divided states. Overall, members of upper chambers follow a wider array of career pathways that are not always displayed by members of the lower chamber. While there are similarities among the states in terms of discrete movement, and even progressive ambition, there are unique differences in how members from the states pursue static ambition.

+Teeter, +Persistent, + Downward +Early +Termed

−Teeter, −Downward +Termed

+Retire −Persistent, −Downward, −Teeter +Early, −Termed −Retire +Downward, +Persistent, − Teeter + Early +Termed

Note +/ −Sign indicates direction of significance. Italics indicate significance in just the upper chamber, bold indicates significance in the lower chamber only

Divided between houses

Consecutive

Lifetime

Summary of the statistically significant career ambition findings by term limit type

Unified year limit

Table 3.6

3 LEGISLATIVE CAREERISM AND GROWING AMBITION

79

80

J. BUTCHER

Ambition in Practice The preceding section reveals a great deal about the ambition pathways that legislators tend to pursue in different term-limited states. These analyses, however, reveal little about the actual mindset of legislators. Are legislators who face term limits thinking about the next office? Do they feel pressured by term limits to get work done? It is one thing to see trends in legislative careers across the states; it is another to understand why members make the career decisions that they do. One way to capture the careerist nature of serving under term limits is to ask members if they feel pressured by term limits, spurring a conversation about whether they feel rushed, prepared, or overwhelmed by their position. These interviews reveal distinct differences among the states in how lawmakers view their careers; indeed, the different types of term limits cause members to view their careers differently. To gauge how members feel about their jobs in the legislature, they were asked questions about how they feel about the job. Lawmakers were asked, “Do you feel pressured for time because of term limits?” The responses were recorded as a simple dichotomy, yes (1), or no (0). The responses to this question indicate the career ambitions of a member and whether they will continue to seek office in order to achieve their goals. Table 3.7 contains the number of members who feel pressured for each state; the descriptive statistics illustrate the stark differences across the states. In Missouri, the majority of respondents said that they feel pressured for time because of term limits. In contrast, there is a nearly equal split from the respondents in Nebraska, half feel pressured for time and the other half do not. In Ohio, while many members do feel pressured, many others do not. In Oklahoma, however, the majority of respondents do not feel pressured for time because of term limits. The different states all have slight variations in whether or not they feel pressured for time; this is likely due to a combination of the type of legislature and the type of term limits members serve under. Questions such as whether members felt pressure in their jobs and if they planned to run again helped to reveal how positively members felt about their role in the legislature and how much they liked their job. The question is, do the results from the interviews line up with the observations from state legislative careers? The remainder of this chapter is dedicated to understanding how members view their careers and how that connects to observations of their expressed career pathways.

3

Table 3.7 Members who feel pressured for time because of term limits

LEGISLATIVE CAREERISM AND GROWING AMBITION

81

State

Yes (%)

No (%)

Total

Missouri (Divided/life) Nebraska (Unified/consecutive) Ohio (Divided/consecutive) Oklahoma (Unified/life) Total

22 (58) 8 (53) 14 (58) 7 (32) 52%

16 (42) 7 (47) 10 (42) 15 (68) 48%

38 15 24 22 99

Unified/Lifetime Hypothesis 1 predicted that in unified/lifetime states, there would be fewer observations of premature movement. Due to the lifetime ban, members would primarily serve in a single chamber. Specifically, those who serve in the upper chamber are more likely to find pathways to continue serving; only those in the lower chamber leave prematurely for higher office. Hypothesis 1 holds up in reference to the decline in static movement. With the lifetime ban in place and a set number of years, few members find an opportunity to return to office once they leave. In Oklahoma, members adhere to a unified limit designating a set number of years to be distributed among both chambers. The format of their type of term limit appears to cause members to think about their careers differently. Most of the members that were interviewed in Oklahoma were split on whether or not they were pressured for time, but few talked about prospects of advancement. Rather than speaking about advancement to a higher office, many members spoke about their 12 years as a lump sum in the office that they currently hold. A Democratic member said that they do feel pressured for time, “because I know the things that I hope to accomplish and putting a time limit on it, even with the grand assumption that I will come back and be reelected, is concerning. . . you have to get everything, your hopes and dreams, done in 12 years. That’s it.” Conversely, a Republican representative shared that they do not feel pressured for time “because I could care less if I am here more or not and as long as I am helping I will be here.” Unlike some of the other term-limited states, members from unified/ lifetime legislatures are significantly more likely to see all three types of

82

J. BUTCHER

ambition, but this does not hold up when comparing that decision to stay in the same seat. This result indicates that while members do experience more variety in their careers than those who do not face term limits, within unified/lifetime states few members take alternative pathways. In conversations with members some felt no pressure while others felt immense amounts of pressure, not about the job but about running for office. One woman shared that she did feel pressure but, “I think the real pressure for time... is the compressed time of the legislature.” A firstterm member remarked, “I haven’t necessarily felt (pressure) yet but I am already getting to a point to be worried and thinking of the next session and running again.” They continued, “That is always in the back of your mind.” One of the youngest members in the lower chamber shared that she felt no pressure, “A lot of people do this later in life, I flipped the script and now this is a stepping stone in my public service.” Unified/Consecutive Under the unified/consecutive ban, present only in Nebraska, there are limited opportunities available for lawmakers. Hypothesis 2 predicted that unified/consecutive term limits would limit the options of members to advance and that instead members would stay until they are forced from office. This hypothesis holds up to the observational test where static ambition is a rare occurrence with only three lawmakers who have returned after terming out. In Nebraska, legislators are permitted to serve a total of 8 years making ambitious lawmakers even less likely to retire than in states with lengthier terms. Without having another chamber to move to, legislators have very little say over their own careers with the only positive movement being terming out and seeking higher office. Many Nebraska legislators did not speak about the careerist aspect, but instead when asked if they would return after terming out of office, many responded that they would not or were on the fence due to extraneous factors (i.e., age, spouse, children). Similarly, members were less inclined to say that they felt pressured for time to accomplish what they needed. There were few contrary voices in Nebraska. In regard to ambition, one senator explained that with term limits “you get more people focused primarily on the next step rather than someone that can spend the time to focus on making changes.” This sentiment comes out in the observed results with few observations of static movement but an increase in the

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

83

number of those who term out of office. There were a couple of individuals who did share that they would be interested in returning, one who has already termed out once before. But as one member put it succinctly, “I look forward to term limits. . . there will be qualified people who come after me.” Divided/Lifetime Hypothesis 3, centered on the effects of divided/lifetime term limits predicted that the strictness of the limit would severely limit service and members would not leave office until forced to. In states with a divided/ lifetime term limit the career pathways that members display are divided by the chambers where members serve. In some ways, this does appear to be the case as there is declined static movement, meaning members seek traditional reelection avenues over those less traditional strategic moves. However, when it comes to progressive ambition, lawmakers are likely to leave in both chambers and pursue higher office. As one Missouri representative stated, “people get sworn in and are already looking down the road and they can be difficult to work with.” This member then continued, “people are in constant campaign mode and looking toward the future move... (term limits) restrict their actions so that they are looking for a future job.” In both chambers, lawmakers are significantly more likely to prematurely leave the legislature in pursuit of higher office. Yet, only in the upper chamber are legislators less likely to term out of office before seeking a higher position. Another important take away for the divided/lifetime limits are the significant findings for progressive ambition. The divided term limit emphasizes the division between chambers, with the upper being more prestigious as shown by the members who work hard to stay in the upper chamber. Yet, in both chambers lawmakers are more likely to wait to advance until they term out of office. In these interviews, there was never a discussion about not seeking reelection or another position; rather, it was an implied understanding that yes, of course, they would run. Legislators from Missouri were also the most likely to make statements about career ambition and to admit feeling pressured for time because of term limits. One member remarked, “The second you get to the house, you think about running for the senate.” Another member explained, “You can plan (for) when a seat is open.” Overall, Missourians, even those who agreed with their term limits, felt they were too short to get the job

84

J. BUTCHER

done. Most members who were interviewed spoke of how quickly the time passes and a number of others spoke about the need to be constantly campaigning. Despite all of that, most did express interest in continuing to serve in elected office. Of all of the members that were interviewed in Missouri only one candidly responded, “I wouldn’t do this again.” Divided/Consecutive Last, hypothesis 4 centered on the divided/consecutive term limits. Given how lax these term limits are, it was expected that members would display a wider array of career movement given the lack of constraints facing them. The popular divided/consecutive term limit is the only type of limit that sees significant changes for all forms of static and progressive ambition. Divided/consecutive legislatures see all forms of static ambition. Teetering, unique to states with divided/consecutive limits, is present in both chambers. This is a trait of having a divided limit, which encourages members to shift between chambers in order to continue serving and is allowed by the consecutive ban. Lawmakers originating from both chambers are significantly more likely to bounce between chambers, indicating chamber of origin matters less than personal drive. Whereas, persistent and downward ambition are likely pathways for those who serve in the upper chamber. Again, this is the nature of a consecutive ban: if a legislator can serve their “time out” in the Senate, then they will choose that over leaving elected office to return to the lower chamber. Last, in both chambers members are more likely to leave the chamber before the end of their allotted time in order to pursue higher office. Regardless of the avenue in which they do so, legislators in states with a divided/consecutive term limit are more likely to pursue progressive ambition pathways than their counterparts in non-term-limited states. In Ohio it is far easier for members to have lengthy legislative careers. With the lax consecutive limit, members are able to bounce back and forth between legislative chambers in order to continue serving in the legislature. This strategy is well known among members who serve in the Ohio legislature. One first-term member remarked, “(term limits are) loose enough for people to still be career politicians and come back in full force, which was not really the intent.” Similar to other states, the results about feeling pressured for time are mixed: some are pressured by term limits, others simply by time.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

85

This pressure starts early in a member’s career. When asked if there was anything else they would like to share a first-term member offered up that they wish they had known how much they would feel pressured for time because of term limits. A third-term representative said not only do they feel pressured by elections but, “I am always running!” A different member said that they wish they had known the pressure that they would feel because of term limits. At the end of the day, the members who serve in Ohio are well aware of the flexibility afforded by their term limits. A state senator, who previously served in the house, said that they do not feel pressured because, “the fact is I could run for another office and do the same job.”

Summary This chapter explores state legislative career ambitions by type of term limit to better understand how term limits have influenced the careers of state legislators. Term limits were popular with the voters largely because of their promise to the curb careerism of legislators like Speaker John Martin from Maine. However, early examinations of term limits and their suspected influence on legislative careers revealed a lack of consensus among early scholars. It was unclear as to what extent, if any, term limits would influence legislative careers. As term limits began to go into effect, scholars argued that legislators were now more politically ambitious. There was a lack of consistency in the results of the tests of these earlier theories centered on retirement and ambition. The observations within this chapter reveal that more restrictive limits shape careers to a greater extent. These results have several unique implications for how we view not only careers but the partisan aspects of state government. What might be of most interest to the members themselves is that, depending on the states where they serve, they might have vastly different political careers, regardless of personal aspirations. Legislative rules dictate members’ careers far more than they likely know. Substantively, it is important to note that the party plays a limited role in the careers of members. Our current political climate is one of heightened polarization and tenuous relations across the aisle, but being in the majority or minority party does not seem to drive the observed behavior of legislators. The factors that tend to influence members are those that are institutional such as professionalization and career opportunities. Contrary to popular belief, these

86

J. BUTCHER

results indicate that the institution is a better indicator for observed career behavior than party. Looking at the responses from state legislators, it is clear that the pressure for time varies by state and term limit type. In Missouri, members feel greater pressure than their counterparts do in Oklahoma. Missourians have less time to serve in a single chamber with only eight years; in Oklahoma a member can spend all twelve years in a single chamber. In Nebraska, there is an even split among respondents who feel pressure and those who do not. This is in part the result of the fact that Nebraska senators can return to the Unicameral once they term out, as well as the true citizen nature of the Nebraska legislature. The same is true for Ohio members, who do not feel pressured knowing that they have the option to continue serving in the legislature at any time. Careers in term-limited state legislatures have turned out to be more complex than initially anticipated because of the variation in types of limits. Therefore, theories examining term limits and their outcomes must be expansive. To put it simply, more time and more data were needed to understand the effects of term limits. The timing of studies of term limits greatly matters because of the long-standing career patterns. For example, with static ambition, it takes multiple cycles of members shifting careers and exploring previously undeveloped career pathways to know what is occurring. Observations of returning, teetering, and moving downward are going to increase with time, not discontinue as one might have expected. One such example is the new law in Arkansas which now allows members who served under previous versions of term limits to return once again. Term-limited state legislatures differ from states without term limits, and from each other. Assessing the different ambition pathways that members choose to follow reveals the complexity of the changes brought by term limits. The observations within this chapter about the choices that legislators make open up a new understanding of how careers operate. Substantively, these results are important because they demonstrate that the institution plays a far greater role in a member’s career than previously accounted for. However, it is the broader implications that deserve greater attention. Term limits were meant to end legislative careerism and this analysis demonstrates that term limits have done just that, but only in some states. In other states, term limits have simply opened the door for more strategic politicians by simply adding an additional step to attaining a lengthy political career.

3

LEGISLATIVE CAREERISM AND GROWING AMBITION

87

This chapter presents a broad examination of the different forms that legislative careers can take on. This initial test of legislative careers demonstrates that careerism is more dynamic under term limits than anticipated. Not only are there distinct trends among states with the same type of term limit; there are differences between legislative chambers. This examination of legislative careers, as they appear today, helps to increase our understanding of term limits and how they actually work. The following chapters further explore both the substantive and normative effects of term limits on legislative careerism.

Bibliography Ang, Adrian U-Jin, and L. Marvin Overby. 2008. Retirements, retentions, and the balance of partisan power in contemporary congressional politics. The Journal of Legislative Studies 14 (3): 339–352. Butcher, Jordan. 2021. Rethinking retirement: Ambition in term-limited legislatures. The Journal of Legislative Studies 27 (2): 285–299. Cain, Bruce E., and Marc A. Levin. 1999. Term limits. Annual Review of Political Science 2 (1): 163–188. Caress, Stanley M. 1996. The impact of term limits on legislative behavior: An examination of a transitional legislature. PS: Political Science & Politics 29 (4): 671–676. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 2000. Term limits in state legislatures. University of Michigan Press. Francis, Wayne L., and Lawrence W. Kenny. 1997. Equilibrium projections of the consequences of term limits upon expected tenure, institutional turnover, and membership experience. The Journal of Politics 59 (1): 240–252. Hibbing, John R. 1999. Legislative careers: Why and how we should study them. Legislative Studies Quarterly 24 (2): 149–171. Lazarus, Jeffrey. 2006. Term limits’ multiple effects on state legislators’ career decisions. State Politics & Policy Quarterly 6 (4): 357–383. Mayhew, David R. 1974. Congress: The Electoral Connection. Yale University Press. Moen, Matthew C., and Kenneth T. Palmer. 2003. Maine: The cutting edge of term limits. In The test of time. Lexington Books. Moen, Matthew C., Kenneth T. Palmer, and Richard John Powell. 2005. Changing members: The Maine legislature in the era of term limits. Lexington Books. Moncrief, Gary F., Joel A. Thompson, Michael Haddon, and Robert Hoyer. 1992. For whom the bell tolls: Term limits and state legislatures. Legislative Studies Quarterly, 37–47.

88

J. BUTCHER

Sarbaugh-Thompson, Marjorie, John Strate, Kelly LeRoux, Richard C Elling, Lyke Thompson, and Charles D Elder. 2010. Legislators and administrators: Complex relationships complicated by term limits. Legislative Studies Quarterly 35 (1): 57–89. Sarbaugh-Thompson, Marjorie, and Lyke Thompson. 2017. Implementing term limits: The case of the Michigan legislature. University of Michigan Press. Schlesinger, Joseph A. 1966. Ambition and politics: Political careers in the United States. Squire, Peverill. 1988. Career opportunities and membership stability in legislatures. Legislative Studies Quarterly, 65–82. Squire, Peverill. 1992. Legislative professionalization and membership diversity in state legislatures. Legislative Studies Quarterly, 69–79. Squire, Peverill. 2017. A Squire index update. State Politics & Policy Quarterly 17 (4): 361–371. Squire, Peverill, and Gary F. Moncrief. 2010. State legislatures today: Politics under the domes. Rowan & Littlefield. Straayer, J. 2007. Colorado legislative term limits: The worst of both worlds. In Legislating without experience: Case studies in state legislative term limits. Lexington Books.

CHAPTER 4

Declining Turnover and Legislative Leave

If you boil down to it, I don’t like (term limits) because I’ve always thought that if someone is doing a good job for their district why make them leave? If you don’t like what they are doing join a campaign and vote to get them out. —Oklahoma Representative (Republican)

Jerry Ellis, a Democrat from Oklahoma, was first sworn into the Oklahoma House of Representatives in 2002 where he served until 2008. In 2008 Ellis advanced to the state senate, and he ran for and secured reelection in 2012. In the state of Oklahoma legislators are limited to 12 years of service; this can be divided in whichever way a member chooses but cannot be extended beyond 12 years. As a result of this strict limit, Senator Ellis was forced to resign his seat mid-term in 2015. Knowing that he would be unable to fulfill his legislative term, Ellis choose to seek reelection in 2012 anyway. Lawmakers, like Ellis, who choose to seek reelection knowing that a long term in office is impossible, are unique to term-limited states. This desire to continue legislative service, despite the restrictions, is telling of the new careerism under term limits. Rather than serving under the citizen-legislature model, some members seek out avenues to stay longer. Much of legislative careerism is centered on the moves lawmakers make: when they advance, when they retire, and where they go. The © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_4

89

90

J. BUTCHER

previous chapter explored the different options that term-limited legislators have presented to them and examined whether members leave the legislature prematurely. There is another perspective of legislative careerism that can reveal just as much about members and the legislature that they serve in, that is, when do members stay? Knowing that time is finite and a career in a term-limited legislature is not what it once was, one would think members would leave office. A citizen legislature was the goal of term limits after all. Rather than seeking out a higher office or leaving for a local position, or even retiring altogether, why do some members stay in office and continue to seek reelection? In this chapter, I explore the importance of the legislators who stay in office with an emphasis on comparing those who face term limits and those who do not. I begin by examining the rate of turnover between states with term limits and those without. I then turn to look at whether or not lawmakers are more likely to seek reelection and why they leave the legislature in both contexts. In the previous chapter, I uncovered the dynamic career pathways of term-limited state legislators and how they can vary based on the state they serve in. This is continued with an examination of one of the most studied components of the legislature: turnover rates. It was evident that turnover rates were declining in state legislatures before the implementation of term limits but then spiked around the time they became effective (Moncrief et al. 2004; Francis and Kenny 1997). The very intention of term limits was to increase legislative turnover, therefore higher rates should not be surprising (Berman 2007; Powell 2007; Lazarus 2006; Moncrief et al. 2004; Cain and Levin 1999; Francis and Kenny 1997; Cain 1996). However, based on the analysis in the preceding chapter, it is clear that there are avenues available for members to pursue lengthy legislative careers and term limits did not inherently curb careerism. The question is, are legislative careers different for those who serve in termlimited states? Turnover rates are an aggregate indication of legislative careerism signifying the difference between those who stay and those who leave, making this the ideal assessment for understanding the broader differences between states with and without term limits. In order to test the relationship between term limits and turnover, I distinguish between two types of turnover: natural and artificial. Artificial turnover is the direct effect of term limits where individuals are forcefully removed from office. It is artificial as it inflates turnover rates when given the choice, those same members may have retained their seats otherwise. Then there is turnover that is naturally occurring in all legislatures,

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

91

i.e., election loss, retirement, death, or seeking higher office. While early works provided insight into the potential changes in legislative turnover, they did not encompass the variation in types of term limits (Cain and Kousser 2004; Francis and Kenny 1997). There is a lack of emphasis on individuals who choose to stay in the legislature versus those who leave. This has been overlooked, in part, because measurements of turnover tend to be in the aggregate. It is important to understand who leaves the legislature and the motivating factors for why they do so. In this chapter, I provide an update on legislative turnover and how it has changed in states with term limits. It is easier to see legislative changes when the different types of turnover are disaggregated. Specifically, focusing on the decision to stay or go reveals trends that have previously been unobserved. When a legislator retires, resigns, or waits to term out, these are all telling of their career ambitions and how they view the office that they serve in. The different ways that a legislator leaves office can say just as much about their career as the decision to not leave.

Turnover and Legislative Leave The fear among early scholars was that term limits would dramatically increase turnover, thus harming the institution. Moncrief et al. (2004) found that in term-limited states turnover was increasing, but that the turnover was not the same across the different types of term limits. There was the potential that term limits would lead to decreased turnover because they would alter the electoral process, resulting in a decline in electoral competition. Francis and Kenny (1997) predicted that the direct turnover effect of term limits would be nearly 36%, whereas the indirect effect of term limits on turnover would be greater than 50%. Overall, turnover was anticipated to increase and it would be even higher in the states with stricter limits. While these works provide great insight into the potential changes in legislative turnover, they do not encompass the variation that has occurred. Just how much would term limits alter turnover in the long run? That has remained unclear for the last 15 years, however, this update finds that turnover rates are not as high as once predicted. Building from these two works, I take the approach that there is a difference in the artificial turnover that is produced by term limits and turnover that is naturally occurring. I present four hypotheses aimed at testing the effects

92

J. BUTCHER

of term limits and how they have helped to shift the end of a legislator’s time in office. The most obvious trend when looking at turnover rates is that turnover will be higher in states with term limits. The fear of term limits and the lessening control that members would have over their own careers caused a number of lawmakers to leave prematurely, which in turn increased the rate of turnover before term limits were even implemented (Cain and Kousser 2004). When looking at the rate of total turnover, without deciphering natural versus artificial causes, the states with term limits will see more members exit the legislature and more new members arrive. This particular result should not be a surprise as the very intention of term limits was to stimulate turnover and remove entrenched incumbents from office. Hypothesis 1: Turnover rates in term-limited legislatures will appear to be higher than in states without term limits. There is the artificial turnover, produced directly by term limits, and the natural turnover, which is how we typically think of turnover. The artificial turnover produced by term limits can be separated from the natural turnover. Natural turnover is what is seen in other states; this includes leaving because of death, resignation, election loss, seeking higher office, and retirement. Once the artificial turnover caused by term limits is accounted for, it becomes easier to assess why lawmakers leave office. Given that term-limited legislators do continue to pursue lengthy careers, there should be a decrease in the amount of natural turnover for members. To be clear, to an extent the different forms of turnover cannot be separated. However, it is clear when a member leaves a chamber because of hitting their limit as opposed to leaving prematurely, that is the key distinction in this measurement difference. Along the same lines, the artificial turnover, or terming out of office, will be the primary reason that lawmakers in term-limited states leave office. This forced out is going to be the primary way in which members leave office, further illustrating that it can be misleading to look at turnover as a whole. Hypothesis 2: The implementation of term limits decreases the amount of natural turnover. Hypothesis 2a : Artificial turnover, terming out of office, is the primary cause of turnover in term-limited states.

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

93

It is important to not only differentiate between natural and artificial turnover but to understand the causes of natural turnover and the changes that have occurred under term limits. In the previous chapter, there was support for the theory that term limits have allowed members to take on lengthy careers and circumvent legislative rules. Does this remain true for when and how a legislator leaves office? Put another way, are the reasons that a lawmaker facing term limits leaves a legislature the same as a lawmaker who does not face such restrictions? For much of legislative history, retirements have been the primary way in which members leave office. Retirement encapsulates a complete retirement from the legislature, where former members typically return to their private life away from the public eye. While older members do tend to retire at greater rates, members choose to retire for a variety of reasons. Lawmakers will retire from elected office for family or personal reasons (Blair and Henry 1981), if they no longer like the job (Moore and Hibbing 1998), or if the benefits are lacking the incentive to stay (Francis and Baker 1986; Rosenthal 1974). Given the levels of careerism displayed in Chapter 3 for lawmakers in term-limited states, I anticipate declining levels of retirement for lawmakers in term-limited states in comparison to non-term-limited states. Knowing that time is finite, a legislator may be more inclined to forgo retiring for reasons such as pay or personal satisfaction because serving in the legislature is such a unique, yet limited opportunity. Hypothesis 3: Term limits encourage lawmakers to serve out their terms, reducing the rate of retirement. Hypothesis 3a : There will be fewer retirements in states with term limits than those without. Following retirements, electoral defeat is a predominant cause of lawmakers exiting office. Now, this particular form of leave has varied greatly based on other factors related to elections, such as legislative professionalization which increased the incumbency advantage in highly professionalized states. Electoral rules also have the ability to strengthen incumbent security; for example, incumbents in single-member districts are less likely to lose (Habel 2008, 486). Early examinations of term limits in Michigan and California revealed that incumbents had become less likely to lose reelection (Sarbaugh-Thompson et al. 2004). This has become the trend in term-limited states as challengers are less willing to

94

J. BUTCHER

enter an election versus an incumbent and instead wait for the seat to open up (Carey et al. 2000). In sum, incumbents in term-limited states who are still eligible to run for their seat are safer than incumbents in states without term limits (See also Sarbaugh-Thompson and Thompson 2017). Hypothesis 4: The implementation of term limits decreases the rate of election loss as incumbents become more secure. Each hypothesis is assessed by chamber in order to have a micro-level view of leave in the states. When taking the macro-level approach by looking at the entire legislature, there may be nuances, such as a stronger relationship, unique to a specific chamber that are overlooked. There will be fewer losses in states with term limits than those without.

Analysis There are two central analyses used to test the above expectations. First, there is an examination of legislative turnover in the states. This examination of turnover is focused not only on the rate of turnover before and after the implementation of term limits but also is meant to distinguish between the natural and artificial causes of turnover. The second key analysis is related to understanding why lawmakers leave the legislature. This analysis assesses the reasons lawmakers leave in states with term limits in comparison to states that do not have term limits. These tests demonstrate just how much legislative turnover has changed since the implementation of term limits and present the primary sources of leave in these legislatures. This analysis of turnover uses an original dataset, composed of data from multiple sources. This Turnover dataset takes a broader approach, where each observation is a chamber/year. While most examinations of turnover focus on electoral turnover, as that is when most people leave, given that term limits can result in strategic early retirement, it is important to include each year in the analysis. This yearly rate of turnover is also referred to as, total turnover because it includes all members who leave each year, regardless of the reason. Distinct from that, there is naturally occurring turnover which occurs when people leave the legislature due to election loss, resignation, or even death. This is called natural turnover because it is part of the natural pattern of legislative careers where there are reelections and individuals who actively choose not to

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

95

seek reelection. This is distinct from artificial turnover, which is in reference to the members who are forced from office by term limits. This forced action is artificial because it is out of the hands of the member to choose whether or not to seek reelection, rather it is forced by the institution. The base turnover numbers, or total turnover, are collected from The Book of the States, these data range from 2002 to 2018.1 The Book of the States, put together by the Council of State Governments, provides the percentage of legislative turnover from each state for every election. These data are supplemented by term limit information from TLS-Careers and The National Conference of State Legislatures (NCSL). There is demographic data that captures the number (and percent) of seats held by each party, as well as by gender and race.2 Institutional controls, professionalization, and career opportunities, are included in the analysis to account for individual state effects. Professionalization has been found to serve as a “buffer” to external electoral pressures (Berry et al. 2000), whereas career opportunities are an indicator of what other political options a member may have available. Each analysis controls for the number of seats occupied by Whites, males, and Republicans as this is the modal demographic in the states. This first examination is dedicated to understanding legislative turnover and how it has changed. I seek to answer two questions about legislative turnover. First, did term limits increase the rate of turnover? Second, if those who are “terming out of office” are accounted for, did term limits actually increase turnover? The key to the analysis is in the distinction between natural turnover and artificial turnover. I begin by examining total turnover, also known as observed turnover, which includes all members leaving regardless of cause. Following that, I present an assessment comparing the naturally occurring turnover to the artificial turnover caused by term limits. This second analysis is meant to uncover the nature of term limits and whether or not turnover, as a result of other factors, is actually increasing or decreasing. To begin, I assess the rate of total turnover, comparing states with term limits and those without. The dependent variable, total turnover, is 1 2002 is a point in time where most, but not all, states have removed the first cohort of term-limited legislators. 2 These data were collected from state legislative websites, NCSL, The National Black Caucus of State Legislators, and the Center for American Women and Politics.

96

J. BUTCHER

the percentage of seats that changed in each chamber. In other words, this measure is the percentage of members leaving the chamber divided by the total number of seats in the chamber (this rate is calculated for each calendar year). For reference, Table 4.1 includes a comparison of the average rate of natural versus total turnover. The natural turnover accounts for the artificial turnover produced by term limits. In order to distinguish natural turnover from artificial turnover, the total number of members who are termed out of office is removed from the rate of total turnover. As indicated by Table 4.1, there is a noticeable decline in the rate of turnover for states that have term limits when excluding those who term out of office. The differences between total and natural turnover across both chambers are noticeable, but there appears to be a greater effect in the upper chamber of the legislature. The states that do not have term Table 4.1 Comparing the average rate of total turnover and natural turnover, by chamber Lower chamber

Upper chamber

State

Total

Natural

Total

Natural

Arizona Colorado Florida Louisiana Maine Michigan Missouri Montana Nebraska Ohio Oklahoma South Dakota Non-Term-Limited New Jersey North Dakota Oregon Texas Wyoming

22.45% 18.17 16.72 9.02 20.80 23.01 17.76 19.71 – 16.22 15.78 20.92

17.75% 11.04 9.90 5.71 14.26 11.25 9.56 13.71 – 10.46 10.64 15.55

20.86% 16.13 13.97 9.92 20.00 15.79 14.72 16.00 14.99 14.44 14.58 21.68

13.33% 6.79 5.67 6.67 12.14 3.51 5.49 6.82 4.96 6.26 9.31 16.30

10.96 8.76 14.22 10.55 14.71

6.76 8.26 10.20 7.02 10.39

All percentages are the average rate for each state from 2002–2018, including non-election years. Natural turnover excludes the artificial turnover caused by term limits

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

97

limits only have natural turnover because there is no other factor forcefully removing members from office. While there is no comparison between states with and without term limits and their rates of artificial turnover, the rates of natural turnover can be compared among all states. At first glance, term-limited states appear to have higher turnover than non-termlimited states but they actually have similar rates of natural turnover. In fact, there are 8 term-limited legislatures that have similar rates of natural turnover in the upper chamber to the lowest non-term-limited state. In the lower chamber, only 2 states surpass the rate of natural turnover of states without limits and only 3 states in the upper chamber have higher rates of turnover. Using pooled time series data, I estimate a regression to compare the average rates of total and natural turnover3 . Each observation is the rate of turnover for a single legislative chamber in a year. Recall, hypothesis 1 is aimed at examining the differences in the types of turnover and that the high turnover rates in term-limited states are primarily attributed to lawmakers leaving involuntarily at the end of their allotted term. The following analysis tests to what extent term limits significantly alter the rate of turnover. Looking at Table 4.2, it appears that term limits do significantly increase the rate of turnover in all states, but this is only for total turnover. However, when looking at the rate of natural turnover alone, there does not appear to be a significant relationship between having term limits and the rate of turnover for any of the states. This finding not only provides support for hypothesis 1, that term-limited states only appear to have higher turnover, but it demonstrates the lack of difference in natural turnover rates. By removing legislators who are forced from office, it is easier to see how little term limits actually affect turnover in terms of legislators leaving office organically. Conversely, term limits have a prominent role in forcefully pushing legislators from office and significantly increasing the rate at which members are forced to leave. Even with the findings in Table 4.2, it can be difficult to surmise just how similar the term-limited legislatures are to states without term limits. Figure 4.1 illustrates the predicted rate of natural turnover, which allows for a comparison of the states. This illustration of the natural turnover in

3 Given that the percent of turnover is bounded, OLS is favorable over a count model.

98

J. BUTCHER

Table 4.2 Comparing the average rate of total turnover and natural turnover, by chamber (2002–2018) Total turnover

Divided/Consecutive Divided/Life

Lower

Upper

Lower

Upper

6.117*** (2.024) 12.422*** (3.578)

0.537 (1.926) 1.915 (3.393)

9.460** (4.366) −17.128 (13.368) 13.743 (10.328)

8.158*** (2.005) 8.689** (3.582) 9.165* (5.047) 7.803* (4.300) −6.924 (13.476) 6.399 (10.184)

2.911 (4.195) −16.526 (12.687) 9.293 (9.781)

1.317 (1.259) −1.995 (2.270) −2.930 (3.214) 1.635 (2.795) −11.389 (8.560) −1.917 (6.425)

−0.769*** (0.248) 2.830*** (0.397) −1.009** (0.502) 9.293*** (3.009) 270 16

−1.226*** (0.324) 1.680*** (0.364) −0.352 (0.570) 8.217*** (3.013) 284 17

−0.463** (0.183) 2.056*** (0.294) −0.860** (0.374) 10.873*** (2.851) 270 16

−0.445** (0.202) 1.075*** (0.228) −0.180 (0.372) 10.868*** (1.895) 271 17

Unified/Consecutive Unified/Life Professionalization Career Opportunity Turnover by Demographic Male Republican White Constant Observations Number of States

Natural turnover

Standard errors in parentheses ***p < 0.01, ** p < 0.05, * p < 0.1

the states, excluding members who are forced out by term limits, demonstrates that there is not a statistically significant difference in the rate of turnover between states with term limits and those without. Additionally, the difference in the rate of turnover between the chambers within a given type of limit is insignificant. The test of legislative turnover reveals that the rate of natural turnover in states with term limits is marginally different from states without term limits. This assessment of legislative turnover reveals that there is more that needs to be considered when evaluating the effects of term limits than just the rate of turnover. When evaluating the rate of turnover alone, it would appear that term limits are not only effective at expelling

DECLINING TURNOVER AND LEGISLATIVE LEAVE

99

0

5

Pr(Turnover) 10 15

20

4

Div/Cons

Div/Life Lower Chamber

Uni/Cons

Uni/Life

No Limit

Upper Chamber

Fig. 4.1 Comparing the predicted rate of natural turnover, by chamber (Note Nebraska has a unicameral legislature)

members but that they have significantly increased the rate of turnover in the states that have them. When parsing turnover, however, it becomes clear that the difference in turnover rates is attributed to term limits forcefully removing members from office, as there the difference in naturally occurring turnover, like retiring or losing an election, is insignificant. This not only contributes to our understanding of turnover but brings to light new information about the legislators who serve in term-limited states. These results have implications that extend far beyond what is revealed in this chapter alone. For example, prior research on legislative turnover has uncovered that there is a relationship between higher rates of turnover and lobbying (Strickland 2023), further term limits affect that electoral relationship that lawmakers have with their voters (Olson and Rogowski 2020; Titiunik and Feher 2018).4 In particular, there is not a “new breed” of legislators serving in term-limited legislatures. Term limits do not appear to have reinforced the idea of citizen legislatures, this is particularly true in the legislatures that are more professionalized. 4 See also Strickland (2023) and Fouirnaies and Hall (2022).

100

J. BUTCHER

In order to understand exactly what is driving the turnover rates and why members end up leaving, the next section takes an in-depth look at the different reasons individuals leave legislative office in order to further uncover any differences among the states.

Testing Leave In order to better understand how term limits have altered legislative careerism, particularly in the form of turnover, it is necessary to compare the decision to seek reelection with the decision to leave the legislature. Legislative leave is just as telling of a member’s career as staying in office. A legislature’s appeal and rules may be more or less attractive for members to serve for a long time. Put another way, some legislatures encourage members to retire and others encourage members to stay. Lawmakers leave office for a variety of other reasons, sometimes these reasons are out of their control, some legislatures offer greater electoral security for members, others provide incentives to seek higher office, and some see more resignations. Although it was anticipated that term limits would hinder legislative careers and bring in a “new breed” of less ambitious legislators, I argue that term limits shape careers differently than anticipated. Notably, termlimited legislators want to serve for as long as possible, as a result, there will be a decline in alternative forms of leave as members choose to stay until they term out of office. In addition to this desire to continue serving in office, lawmakers in term-limited states have greater electoral security because competitive challengers are more likely to wait for an open election. The following analyses revisit the TLS-Careers data, which includes information on how members left office. At the most basic level, there were 5 ways a member might leave office: seeking higher elected office, retiring, resigning, losing reelection, or terming out of office. For this analysis, higher office indicates an individual seeking (but not necessarily securing) a higher elected position; these include the state upper chamber, a state-wide office, or the United States Congress. Retiring indicates that a member left elected office and has not pursued another elected position (within 2–4 years of leaving office). The indicators for leave resign and lose indicate when a member resigns their seat prior to completing a term or loses reelection, respectively. Last, when a member is coded as terming

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

101

out of office, it means that they fulfilled their allotted term and are leaving because they are unable to run for reelection. With term limits in place, it is expected that there will be lower retirement rates, but the issue at hand is whether term limits encourage lawmakers to serve out their entire term as opposed to leaving early. I test the different forms of leave using a multinomial logit analysis to compare the different options that members have. Each option is compared to seeking and securing reelection. The multinomial logit model allows for a comparison of the possibilities that members face, much like how members have multiple options facing them in reality. To adequately gauge this, the results in Tables 4.3 and 4.4 are for members who have served 12 years or less, to allow for a comparison between states with and without term limits. Table 4.3 contains the results of the leave decisions that members of the lower chamber make for each type of term limit. While the coefficients are not directly interpretable, they do indicate whether each type of limit is statistically different from states without term limits. For example, in the lower chamber in states with lifetime limits, lawmakers are not significantly less likely to seek higher office over seeking reelection than their non-term-limited counterparts. However, those who serve under a divided/consecutive term limit are more likely to seek higher office over reelection than those who do not face term limits. In term-limited states lawmakers who have served less than 12 years are significantly more likely to term out of office. While this result seems rather intuitive, it is the substantive significance that matters. Of all of the potential reasons to leave, terming out is the only dependent variable that is positive and statistically significant for all types of term limits. This particular finding offers partial support for hypothesis 3 that term limits encourage members to serve out their term rather than leaving prematurely. Some of the other interesting findings include institutional and individual-specific effects. An increase in legislative professionalization leads to fewer lawmakers retiring or seeking higher office. Interestingly, Republicans are more likely to leave for a higher office over seeking reelection and are less likely to lose reelection. Whereas males are less likely to lose an election or term out of office. Partisan ties appear to matter less than individual legislator demographics. Perhaps counter to prior beliefs members who are part of chamber majority and are in the same party as the governor are more likely to resign. Thinking of the various reasons

Constant

Same Party as Governor

Part of Chamber Majority

Male

Republican

White

Career Opportunity

Professionalization

Unified/Life

Divided/Life

Retire 0.124* (0.070) 0.024 (0.124) 0.202 (0.154) −4.557*** (0.536) 0.790** (0.348) −0.053 (0.117) 0.004 (0.070) −0.038 (0.071) −0.006 (0.070) −0.088 (0.064) −2.511***

Higher office 0.287*** (0.091) 0.170 (0.152) −0.010 (0.241) −1.060* (0.632) 3.201*** (0.459) −0.351*** (0.130) 0.232*** (0.090) −0.034 (0.088) −0.105 (0.089) 0.178** (0.082) −4.436*** 0.114 (0.133) −0.569** (0.227) −0.691* (0.358) 0.790 (0.929) 1.208* (0.656) −0.793*** (0.177) 0.028 (0.137) 0.165 (0.141) 0.324** (0.140) 0.236* (0.127) −4.824***

Resign

Multinomial logit analysis of leave decisions in the lower chamber

Divided/Consecutive

Table 4.3

−0.046 (0.087) −0.722*** (0.150) −1.145*** (0.234) −0.557 (0.666) −1.963*** (0.435) 0.171 (0.149) −0.217** (0.086) −0.186** (0.088) −0.060 (0.088) 0.034 (0.084) −2.838***

Lose

3.162*** (0.222) 3.545*** (0.235) 2.837*** (0.263) 0.736 (0.503) 1.160*** (0.394) 0.199* (0.107) −0.018 (0.062) –0.041 (0.064) −0.013 (0.063) −0.006 (0.057) −6.621***

Term out

102 J. BUTCHER

(0.169) 34,723

(0.201) 34,723

(0.289) 34,723

Resign (0.208) 34,723

Lose (0.259) 34,723

Term out

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1 Note Nebraska (Unified/Consecutive) does not have a lower chamber. The analysis is restricted to members who have served 12 years or less (Base category: staying in office)

Observations

Retire

Higher office

4 DECLINING TURNOVER AND LEGISLATIVE LEAVE

103

104

J. BUTCHER

why a lawmaker might resign, this result is likely picking up on the trend of those members who resign early for other positions. Those members who are in the same party as the governor are also more likely to seek higher office, and greater career opportunities increase resignations. Table 4.4 contains the results of the multinomial logit regression for the upper chamber of each legislature. As expected, there are differences between the chambers. Most notably, the role of professionalization appears to affect members in the upper chamber to a different extent. For example, in states with higher levels of professionalization members of the upper chamber are significantly more likely to seek higher office; the opposite was true for members of the lower chamber. Similar to the lower chamber, heightened professionalization does decrease the rate of retirement but it also decreases the number of lawmakers who ultimately term out of office. Substantively, these findings demonstrate that members from the upper chamber who serve in a more professional legislature are more likely to seek higher office (i.e., state office or Congress) and are also less likely to term out of office. There are a number of other results that are unique to the upper chamber. For example, there is a significant decline in retirement for members who serve in divided/life and unified/consecutive legislatures. It appears that for states with divided/lifetime term limits members are more apt to term out and are significantly less likely to leave for any other reason. Broadly, most of the differences between chambers appear to be a result of the prestige of serving in the upper chamber. To put it succinctly, serving in the upper chamber is a better position and members are less inclined to leave. Perhaps the most unique finding from Table 4.4 is how different the unified/lifetime states are by chamber. This is likely a result of the unified limit, members from states with a unified/lifetime term limit are more likely to stay in the same chamber throughout their careers. Having this particular mindset has created a division in leave between chambers. In the lower chamber, those who face a unified/lifetime term limit are less likely to leave as a result of resignation or election loss. In the upper chamber, however, members are more likely to resign. These results signify that there is something unique about the combination of having a unified number of years to serve while also having a lifetime ban on that service. This examination of legislative leave offers evidence in support of hypotheses 2, 3, and 4. Specifically, the decline in natural turnover and high levels of artificial turnover indicates that terming out of office is the

Constant

Same Party as Governor

Part of Chamber Majority

Male

Republican

White

Career Opportunity

Professionalization

Unified/Life

Unified/Consecutive

Divided/Life

Resign 0.058 (0.203) −0.670* (0.374) 0.417 (0.518) 1.133** (0.451) 1.792 (1.331) 1.849* (1.119) −0.505* (0.288) 0.054 (0.216) 0.177 (0.228) −0.235 (0.230) 0.620*** (0.195) −5.120***

Retire 0.245** (0.121) −0.583** (0.265) −1.190*** (0.429) −0.706 (0.532) −2.956*** (0.887) −0.348 (0.641) 0.233 (0.266) −0.136 (0.142) −0.134 (0.140) 0.167 (0.153) −0.047 (0.120) −2.761***

Higher office −0.421 (0.309) −1.130** (0.498) −0.962 (0.919) 0.501 (0.642) 5.472*** (1.938) 0.018 (1.574) −0.133 (0.474) −0.225 (0.329) −0.139 (0.329) 0.639* (0.363) −0.003 (0.292) −5.935*** 0.217 (0.183) −1.107*** (0.408) −1.153** (0.453) −0.393 (0.555) −0.926 (1.304) −2.948*** (0.937) 0.401 (0.387) 0.119 (0.207) −0.625*** (0.185) −0.234 (0.225) 0.016 (0.181) −2.910***

Lose

Multinomial logit analysis of leave decisions in the upper chamber (base, staying in office)

Divided/Consecutive

Table 4.4

(continued)

3.708*** (0.452) 4.192*** (0.478) 4.469*** (0.532) 3.856*** (0.563) −1.449* (0.797) 2.067*** (0.727) −0.022 (0.189) 0.121 (0.133) −0.086 (0.115) −0.029 (0.140) −0.111 (0.096) −6.686***

Term out

4 DECLINING TURNOVER AND LEGISLATIVE LEAVE

105

(continued) Retire (0.343) 11,210

Higher office (0.700) 11,210

Standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1 Note Analysis is restricted to members who have served 12 years or less

Observations

Table 4.4

(0.475) 11,210

Resign (0.484) 11,210

Lose

(0.520) 11,210

Term out

106 J. BUTCHER

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

107

primary form of leave for lawmakers in term-limited states. Further, the implementation of term limits significantly decreased the retirement rate in states with divided term limits, where members face a shorter time in office. The turnover results are further bolstered by the declining rates of election loss in states with term limits. While the decline in election loss is not significant in all of the states, it is significant for states with divided/ lifetime and unified/consecutive term limits. The decline in loss supports the idea that term limits bring a stronger incumbency advantage, which decreases the high rate of turnover.

Turnover and Leave in the States In this chapter, I have demonstrated that members staying in office is just as important as members leaving office by evaluating legislative turnover and the different causes of leave. In term-limited legislatures, members are more likely to seek reelection to stay in office. Rather than choosing to leave early, lawmakers stay in office until the end of their allotted term. This section is devoted to increasing our understanding of legislative careers beyond what members do in office. Namely, when do members leave, and where do they go? The TLS-Leaders data provide a sub-sample of lawmakers allowing for an in-depth look at where members go once they leave office. Further, the legislative interviews with members reveal the mindset of legislators when making these particular career decisions. Table 4.5 displays some of the most common pathways that members take upon leaving office. This table is important for 2 key reasons. First, this table illustrates the differences within the state before and after 2000, which is when states had passed term limits and were beginning to see their effects. Second, this table allows for a comparison between the termlimited states and the average of non-term-limited states. This table is comprised of lawmakers who served in a leadership capacity during their time in the legislature but were not currently serving in office at the time the data was collected. Given the nature of service in a state legislature, it is not always easy to find what members do upon leaving office. This information, however, is easier to attain for leaders who were presumably more invested and had higher-profile careers.5

5 For example, this information comes from a variety of sources including news articles, obituaries, LinkedIn profiles, Twitter, Facebook, and even business web pages.

108

J. BUTCHER

Table 4.5 Sub-sample of state legislators and what they do upon leaving, before and after 2000 Retire from politics

Elected/ Appointed

Private law

Lobby/ Consult

State (# of obs.)

pre

post

pre

post

pre

post

pre

post

Arizona (116) Colorado (117) Florida (149) Louisiana (25) Maine (86) Michigan* (74) Missouri* (100) Montana (114) Nebraska (13) Ohio (107) Oklahoma* (160) South Dakota (156) No Term Limits(Avg.)

39% 54 55 33 42 47 54 57 100 55 80 55 56

31% 40 30 35 37 43 38 37 60 36 55 43 47

42% 39 45 50 58 47 31 38 0 45 15 27 37

68% 50 58 53 58 43 54 59 40 56 36 55 50

6% 7 0 17 0 0 8 0 0 0 0 15 4

0% 8 5 6 3 2 4 0 0 8 6 2 2

3% 0 0 0 0 7 0 5 0 0 5 3 6

1% 3 7 6 2 13 4 2 0 1 4 3 1

* Denotes states with a lifetime ban

The post-legislative career options presented in Table 4.5 do not encompass all of the possibilities of members but present some of the most common avenues with politically tangential careers. Each column presents the percentage of leaders from that state who followed a particular path after leaving office (within a 2-year grace period, as some members would wait for an open seat or position). Those in the first column, retire from politics, are members who both retired from public life entirely or left for another career outside the realm of politics. For example, some members left to work in insurance, become adjunct teachers, and one even returned to their prior career as an ophthalmologist. There are many states in Table 4.5 that see higher percentages of individuals leaving for an elected or appointed office than retiring from political life. For these purposes, elected and appointed positions include other positions in the state legislature, Congress, or state-wide agency heads/commissioners like the State Insurance Commissioner. This particular column includes those who ran for an office as well as those who won. Those who ran, but did not necessarily win, are included because the focus is more so on the intention rather than the success. Next,

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

109

there are those who left for private law, a non-elected/non-appointed position. While a number of individuals left for the county, city, and state-level attorney positions, some returned to their own law firms or joined an existing firm. The last politically adjacent column included is for members who left the legislature and began either lobbying or consulting on political/policy matters. It is worth noting some of the vast differences in what members do after leaving office. There were many individuals who sought out countylevel positions or ran for state-wide positions. A number of members were appointed to different boards by Governors, for example, Dan Metelsky (OH) was appointed to the state Lottery Board. Others do retire and seek out somewhat unconventional pathways, at least for what we think former legislators might be doing. Ken Gagnon (ME) has listed on his Facebook page that he was a “Driver at Uber” and a “Substitute Teacher,” among a number of other jobs. Perhaps, one of the most interesting retirement plans was from Patrick “Pat” Colwell, also of Maine. Pat and his brother, Bob, started a band called Soul Sensations. It is worth noting that Pat Colwell is the former Speaker of the Maine House of Representatives, as advertised on their band website. There are a couple of important trends that need to be addressed in Table 4.5 before looking at some specific examples. First, in the post-term limit era, all but two of the term-limited states have a lower percentage of lawmakers retiring from politics than the non-term-limited average. Second, all but one state see equal or higher rates of members seeking out elected or appointed positions in the post-term limit era. Comparatively, states without term limits do see a slightly higher percentage in retirement and a lower percentage of members seeking out elected or appointed positions post-term limits. Third, states with consecutive term limit bans see the highest percentage of lawmakers seeking other elected/appointed positions. Last, it appears that another change brought by term limits is an increase in the number of individuals seeking careers as lobbyists or consultants; while these percentages are rather low, the increase from 0% for many states is rather dramatic. Below, there are some specific examples of members leaving state legislatures while including insight from current lawmakers.

110

J. BUTCHER

Unified/Lifetime Similar to the other term-limited states, there is no difference in the rate of naturally occurring turnover in states with unified/lifetime limits. States with unified/life term limits are unique because they allow a member to serve all of their years in a single chamber. In the case of Oklahoma, members can spend 12 years in one chamber. This is an important distinction because it allows members to continue to seek reelection in a safe seat. In the lower chamber, members who serve are significantly less likely to lose reelection and resign. In the upper chamber, however, members are less likely to retire and instead opt to either resign or term out of office in order to leave the legislature. A unique trait about unified/ lifetime term limits is the low number of individuals seeking higher office, members are no more or less likely to seek higher office from either chamber. In sum, members are able to safely secure reelection and tend to wait out their term limit or, in the upper chamber, resign from the legislature. These patterns become evident when talking to legislators about their careers. Oklahoma legislators openly discuss their stressors with the time that it takes to learn how to do the job. One member said, it takes 4– 6 years just to figure out what you are doing, at which point members have already spent half of their time. Because it takes years to understand the legislative process and gain footing, members want to stay to push their agendas. However, very few members talk about future ambitions, their service to the legislature is enough. One member openly shared they are just trying to do what they can because “they are not going anywhere after this.” While members are not actively seeking out ambition pathways, as seen in Chapter 3, it is important to understand that legislators, even in a restrictive lifetime state, choose to stay for as long as possible. Once members leave they appear to take on a path that is a result of their lifetime ban. Less than half of the legislative leaders have a continued career in politics, even though there are more members seeking out these opportunities than before term limits. Instead, members are more apt to retire from political life. For example, Cal Hobson who served 16 total years in the legislature, only to term out of office, became a teacher and a writer. Retiring as a public official, but still being involved in politics, James Williamson termed out of the legislature and later worked for the President Pro Tempore of the Oklahoma State Senate.

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

111

There were some noticeable changes to leave for legislators in Oklahoma. Before term limits, most members in Oklahoma retired. This rate dropped by 25% post-term limits. Other changes include an increase in the number of legislators who return to private law after leaving, and the percentage of members seeking elected/appointed political posts doubled post-term limits. These results indicate that even though Oklahoma appears to be an outlier compared to the other term-limited states, the legislators who serve there have experienced some career changes with the implementation of term limits. Unified/Consecutive Nebraska is the only state with unified/consecutive term limits and its members have a unique approach to term limits. The lawmakers in Nebraska note a lack of time, but when asked if they would return to office (which is permitted under the consecutive ban), most members respond that they would not. One member clarified, “I’m not a career politician, I’m a farmer.” Now, it is difficult to pinpoint the different influences that members face and what drives their career choices, it could be the non-partisan legislature, the unicameral chamber, or the lengthy 4year term. No matter the cause, it is important to note the consistency among the interview respondents who show a lack of desire to return to the legislature after a 4-year break, but no member acknowledged that returning for a second term was unlikely or out of the question. The feelings described by members are evident in the quantitative analysis of legislative leave. Not only are members significantly less likely to lose reelection, but they are also less likely to retire from their seat. Instead, the most common way for members to leave the Nebraska legislature is to wait and term out of office. Overall, under the unified/ consecutive term limit, there is a lack of interest in utilizing the alternative pathways described in Chapter 3 but members do desire to seek reelection and stay until they have fulfilled their term limit; they just do not show an interest in returning beyond that. As seen in Table 4.5, before term limits 100% of the leaders in the sample retired from political life. Only after term limits was there an increase in the number of people seeking elected/appointed positions, but this still falls well below the average in non-term-limited states. The positions that members pursued included 2 running for the United States House of Representatives, 1 for governor, and 1 actually returning back

112

J. BUTCHER

to the Nebraska Senate shortly after leaving. In Nebraska, which is a deadend state (see Squire 1988) with only one chamber, there are few options for members to stay in politics. Given the circumstances that members face, these results are unsurprising. Divided/Lifetime Those who serve under a divided/lifetime term limit have the flexibility of being able to serve in both chambers while facing the restrictive lifetime ban on service. Yet, turnover rates in states with a divided/life term limit are not statistically different from states without term limits. In the lower chamber, legislators are significantly less likely to retire, resign, or lose reelection and instead leave once they term out. In the upper chamber, lawmakers are significantly more likely to wait and term out of office and are less likely to seek higher office, retire, resign, or lose reelection. According to the legislators interviewed much of the drive to continue to seek reelection comes from the lack of time. Many lawmakers cited a great deal of pressure as a result of a personal desire to meet legislative goals with little time to do so. No members responded hesitantly or were doubtful about running for and securing reelection. Instead, members talked about their time in office as the entire 8 years, as if reelection were a given. It is assumed that members will stay for their entire term. As one member put it, “the job is incredibly high-paced and addictive as hell.” The divided limit in Missouri provided the opportunity for more members to have prior service and greater experience. Of the members that were interviewed 8 members had previously served and 4 had termed out of one legislative chamber before moving to the next. Broadly speaking, Missouri has fewer members retire than those states without term limits but they are also on the lower end of states who have individuals seek out other elected offices among the term-limited states. Given the lifetime ban in Missouri, many individuals who want to continue a political career are forced to either state-wide positions or local offices. Edward Quick who was termed out in 2004 after 20 years of service became a county commissioner. Steven Tilley, Peter Kinder, Ken Jacob, and Gracia Backer all ran for lieutenant governor. Madia Coleman ran to become the Mayor of St. Louis and Brandon Ellington served on the Kansas City Council. With limited options, lawmakers had to look at state-wide and district-level positions to stay involved in politics.

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

113

The tale of Michigan is similar to what occurred in Missouri, but there is a higher percentage of legislative leaders who went on to secure a lobbying or consulting position. In Michigan there were fewer individuals who pursued local-level positions, one example was Kwame Kilpatrick who became the Mayor of Detroit. The most prominent pathways were for members to pursue the state senate or a seat in the United States Congress. The focus on higher positions versus more local is unsurprising given that Michigan is a highly professionalized legislature whereas Missouri is considered to be more of a citizen legislature. Mark Schauer, Mike Rogers, Andrew Raczkowski, and Michael Bishop pursued congressional seats and 10 others sought a seat in the Michigan Senate after term limits. Divided/Consecutive The most common type of term limit is the divided/consecutive term limit. Given that this is also the laxest limit, legislator careers in these states are most similar to lawmakers who do not face term limits. Natural turnover is not statistically different between states with divided/ consecutive term limits and states without term limits. Similarly, there is not a significant difference in the rate of retirements or resignations from either chamber or the rate of election loss in the upper chamber. Unique to divided/consecutive states, lawmakers from the lower chamber are significantly more likely to seek higher office over seeking reelection. However, members are also more likely to wait until the end of their term to leave, this is true for both chambers. There are more states with divided/consecutive limits than any other state, which results in a great deal of variation in how and why members leave. Much of the variation in the divided/consecutive states can be attributed to differences in legislative professionalization and the career opportunities that are offered to members. For example, Ohio is a highly professionalized state so there is a stronger incentive for members to serve. During the interview process in Ohio, I had the opportunity to speak to three members who had previously been termed out of office but had since returned. Interview respondents from Ohio did not express any intentions to not seek reelection, instead most discussed the continual practice of seeking reelection and their limited time in office. Members discussed reelection as it was a given, which is evident in the results above. A state senator who had previously served in the house summarized the

114

J. BUTCHER

motivations for his political career as “accountants like to do their jobs, I like to do mine.” While lawmakers in Ohio expressed interest in continuing to serve the states of Arizona, Montana, and Maine had the highest percentage of lawmakers seeking elected or appointed office after leaving. In Arizona, there were a number of members who left to seek another seat in the state legislature bouncing back and forth between chambers (as referenced in Chapter 3). In fact, among the leaders in the sub-sample, roughly 20 of the over 100 individuals ran for a seat in the state legislature shortly after leaving, this would not even be possible under most circumstances in a state with a lifetime term limit ban. Maine appears to have a more even distribution among the positions that people seek out after leaving the legislature, while some return to the legislature there were also members who ran for governor and who ran for Congress. Other divided/consecutive states like Florida and Colorado also saw a dramatic decrease in the percent of retirements post-term limits, while experiencing an increase in the number of individuals seeking out elected or appointed positions. Louisiana also saw these changes but to a lesser extent. The difference in Louisiana is the length of the term, rather than the more common 8 years, Louisiana terms are 12 years for each chamber. It appears this lengthy tenure resulted in few changes to a member’s post-legislative career. It is also worth noting that sometimes the careers members follow after leaving the legislature are not so clear. For example, Daniel Lederman of South Dakota announced he was resigning his seat to spend more time with his family but he then became a lobbyist. Overall, the divided/consecutive states simply do not see the changes to turnover that other term-limited states do but they do see variation in what members choose to do upon leaving office.

Summary This chapter is dedicated to exploring a lesser-known aspect of legislative careerism, turnover, and legislative leave. This assessment of turnover and leave from the term-limited states presents a new way of understanding legislative careers. Not all lawmakers spend their careers seeking out higher office and, as seen in Chapter 3, only under unique circumstances will members leave and return to the legislature. The easiest and most common option for legislators to serve is to seek reelection, but that is not always an option. I evaluate the changes to leave in two key

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

115

ways, first, I examine aggregate turnover rates among the states; second, I assess the different reasons why lawmakers leave the legislature. This line of inquiry is then followed by examining a sub-sample of legislators and what it is they do upon leaving the legislature. Extending the work from Chapter 3, I utilize the TLS-Career dataset to examine the different reasons why legislators leave office. In both chambers, legislators from term-limited states are less likely to retire, resign, and lose reelection versus securing reelection than legislators from states without limits. Rather than leaving office early or choosing to advance to higher office, most legislators choose to stay in their seats for as long as possible. This finding goes against many early predictions of term limits, which stated that the “new breed” of legislators would lack the desire to make a career out of service. Yet, as the turnover analysis reveals, when removing term limits from the equation, there is little difference between turnover rates and leave for states with and without term limits. This test of legislative leave and the long-term trends under term limits gives new insight into an under-appreciated area of legislative careerism. In comparison to other studies of careerism understanding why members leave has received the least amount of attention. Leave is quite normative in nature as members leave for all types of reasons and has minimal effect on the subject of interest, the legislature. The substantive importance is far greater than first realized. The careers of members and how they end can reveal a great deal about the legislature and its institutional design. Retirements used to be the most common way that members would leave the legislature. After a long career serving in the legislature members, having had their fill of service, would retire from public life and not seek out other positions. With term limits, legislators have changed the way that their membership turns over. Decreasing the rate of retirement is one of the most important unintended consequences of term limits because it implies that political ambition did not decrease and moves the state farther from the idealized citizen legislatures. When looking at what members do upon leaving the legislature it appears that this mindset remains, post-term limits more lawmakers are seeking out other positions within the political realm than previously. Looking at the results from the leave analysis and the trends in responses from state legislators, it is clear that most term-limited legislators do not enter office with the intention of leaving prematurely. In fact, there are some legislators who enjoy serving so much that they continue

116

J. BUTCHER

to run for reelection forgetting that they are banned from doing so by term limits. Others, like Jerry Ellis from Oklahoma, run for reelection knowing they will be forced to leave prematurely because they simply enjoy the job. Term limits are not the deterrent to legislative careerism that many had hoped for, rather, members choose to stay in office for as long as possible. This has resulted in terming out of office becoming the primary form of leave, decreasing the rate of natural turnover. There are important discrepancies between those states with a lifetime ban and those with a consecutive ban. Much like the findings from Chapter 3, there are clear differences for the more stringent lifetime cases. Here, legislators are significantly less likely to lose an election and tend to stay in a single chamber until they fulfill their term. Taken together, it appears that those who serve in a state with a lifetime ban are beginning to demonstrate drastically different career trajectories (and career endings) than the members who merely face a consecutive ban on service. The second half of this chapter revealed that even in those circumstances where members do leave, members are now staying in politics at higher rates. In the post-term limits world fewer members are retiring from political careers, even after leaving the legislature, and more are opting for careers in politics or politically adjacent fields. While yes, term limits do remove lawmakers from office and members are being termed out these results are one-sided. Even though members are forced to leave substantively term limits do not follow through on what proponents had promised. There are now higher percentages of members seeking elected positions and becoming lobbyists than before term limits were passed. The other side of members leaving is that the passage of term limits and members cycling through the legislature quickly may have just created a larger pool of local politicians and lobbyists with political experience. In short, just because a member leaves, it does not mean they are gone.

Bibliography Allen, Linnea. 2022. Former Oklahoma senator shares story of gambling addiction to help others. Berman, David. 2007. Legislative climate. In Institutional change in American politics: The case of term limits, Chapter 4, 107–118, ed. Bruce Cain Karl Kurtz and Richard G Niemi. Ann Arbor, MI: University of Michigan Press. Berry, William D., Michael B. Berkman, and Stuart Schneiderman. 2000. Legislative professionalism and incumbent reelection: The development of institutional boundaries. American Political Science Review 94 (4): 859–874.

4

DECLINING TURNOVER AND LEGISLATIVE LEAVE

117

Blair, Diane Kincaid, and Ann R. Henry. 1981. The family factor in state legislative turnover. Legislative Studies Quarterly 6: 55–68. Cain, Bruce E. 1996. The varying impact of legislative term limits, chapter 1, 21–36. Cham: Springer. Cain, Bruce E., and Thad Kousser. 2004. Adapting to term limits: Recent experiences and new directions. San Francisco, CA: Public Policy Institute of California. Cain, Bruce E., and Marc A. Levin. 1999. Term limits. Annual Review of Political Science 2 (1): 163–188. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 2000. Term limits in state legislatures. Ann Arbor: University of Michigan Press. Fouirnaies, Alexander, and Andrew B. Hall. 2022. How do electoral incentives affect legislator behavior? Evidence from U.S. state legislatures. American Political Science Review 116 (2): 662–676. Francis, Wayne L., and John R. Baker. 1986. Why do US state legislators vacate their seats? Legislative Studies Quarterly 11: 119–126. Francis, Wayne L., and Lawrence W. Kenny. 1997. Equilibrium projections of the consequences of term limits upon expected tenure, institutional turnover, and membership experience. The Journal of Politics 59 (1): 240–252. Habel, Philip D. 2008. The consequences of electoral institutions for careerism. Legislative Studies Quarterly 33 (3): 471–494. Lazarus, Jeffrey. 2006. Term limits’ multiple effects on state legislators’ career decisions. State Politics & Policy Quarterly 6 (4): 357–383. Moncrief, Gary F., Richard G. Niemi, and Lynda W. Powell. 2004. Time, term limits, and turnover: Trends in membership stability in US state legislatures. Legislative Studies Quarterly 29 (3): 357–381. Moore, Michael K., and John R. Hibbing. 1998. Situational dissatisfaction in Congress: Explaining voluntary departures. The Journal of Politics 60 (4): 1088–1107. Olson, Michael P., and Jon C. Rogowski. 2020. Legislative term limits and polarization. The Journal of Politics 82 (2): 572–586. https://doi.org/10.1086/ 706764. Powell, Richard J. 2007. Executive-legislative relations. In Institutional change in American politics: The case of term limits, Chapter 4, 134–147, ed. Bruce Cain Karl Kurtz, and Richard G. Niemi. Ann Arbor, MI: University of Michigan Press. Rosenthal, Alan. 1974. Turnover in state legislatures. American Journal of Political Science 18: 609–616. Sarbaugh-Thompson, Marjorie, and Lyke Thompson. 2017. Implementing term limits: The case of the Michigan Legislature. Ann Arbor, MI: University of Michigan Press.

118

J. BUTCHER

Sarbaugh-Thompson, Marjorie, Lyke Thompson, Charles Elder, John Strate, and Richard Elling. 2004. The political and institutional effects of term limits. Cham: Springer. Squire, Peverill. 1988. Career opportunities and membership stability in legislatures. Legislative Studies Quarterly 13: 65–82. Strickland, James M. 2020. The declining value of revolving-door lobbyists: Evidence from the American states. American Journal of Political Science 64 (1): 67–81. Strickland, James M. 2023. The contingent value of connections: Legislative turnover and revolving-door lobbyists. Business and Politics 25 (2): 152–172. The Council of State Governments. Various Years. The book of the states. Lexington, KY: Council of State Governments. The Term-Limited States. 2015. http://www.ncsl.org/research/about-state-legisl atures/. Titiunik, Rocio, and Andrew Feher. 2018. Legislative behaviour absent reelection incentives: Findings from a natural experiment in the Arkansas Senate. Journal of the Royal Statistical Society 181 (2): 351–378.

CHAPTER 5

Legislative Leadership

It is a top-heavy structure with the Speaker of the House, in terms of power. —Missouri Representative (Republican)

During the 2019 legislative session, Larry Householder became the Speaker of the Ohio House of Representatives. To most, Speaker Householder would not seem different from most other state speakers; he is an experienced politician who knows how to control his chamber. In fact, Householder previously served two terms as speaker beginning in 2001. What is different is that he termed out of office in 2004, returned to the legislature in 2017, and reassumed the speakership in 2019. By curbing legislative careerism, term limits were predicted to harm legislative leadership, making it more difficult for leaders to gain experience and assume authority over rank-and-file members. Speaker Householder’s story tells a different tale; Householder, being one of the most experienced members in the chamber, rose to the speakership because his inexperienced colleagues depended on his expertise. Legislative leaders are some of the most powerful actors in state government because of their ability to compete and bargain with the governor in ways that rank-and-file members cannot (Rosenthal 1990). Strong leaders are able to push their party’s agenda while guiding their fellow members (Jewell and Whicker 1994; Rosenthal 1990). With the © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_5

119

120

J. BUTCHER

passage of term limits, scholars speculated that the role of leaders would diminish, making it more difficult for leaders to push an agenda or direct members. Term limits, by altering the institution, would change the leadership structure via apprenticeships, tenure, and experience, which would subsequently diminish the power of leaders. In this chapter, I test these ideas to evaluate whether the role of formal leadership has been diminished by term limits or if leadership has actually become more centralized. While the work assessing state legislative leadership is limited, scholars do agree that the role of leadership is rooted in the structure of the legislature (Clucas 2007, 2001; Squire and Hamm 2005). Leaders rely on the institution as a source of power. Changes to the institution, such as increasing professionalization, can add or subtract leadership power (Clucas 2007). The imposition of term limits on state legislatures has received less attention than other factors related to a member’s career. Yet term limits have the ability to alter leadership, which influences members and legislative function. In this chapter, I focus on three potential changes to leadership that would be most affected by the implementation of term limits. First, there are a number of institutional factors, including term limits, that can shape the level of experience that a member has before becoming a leader. Second, the amount of time a member has in office can limit their ability to serve apprenticeships in lower leadership positions. Finally, the institution can shape how long leaders stay in their positions, based on incentives and finite time. By forcefully cutting off legislative service, term limits inherently alter these three components. While the long-standing theory has been that term limits would diminish leadership within state legislatures, this test of term limits reveals that states that afford lengthier tenure have a stronger leadership system in place because of the internal reliance on senior/experienced members.

Testing Legislative Leadership What is known about leadership within the legislature? Members have been found to rise to leadership positions out of a sense of ambition (Maestas 2003; Moncrief 1999). Beyond this, there has been a lack of attention on leadership organization within the legislature (Moncrief 1999; Freeman 1995; Jewell and Whicker 1994; Squire 1992a). The lack of research on leadership is due to the variation among the states. The

5

LEGISLATIVE LEADERSHIP

121

states not only have different term limits, but each has its own unique leadership structure. Before delving into the analysis of how term limits have altered legislative leadership, it is important to understand how each state has a different established leadership structure. Legislative leadership in the states is even more diverse than the types of term limits, with each state having its own leadership design and rules. Leadership positions are all different and leaders come to their position in different ways. Some states have a clear hierarchy and place greater emphasis on seniority while others do not (Squire 1992a). For example, the Wyoming legislature tends to follow a particular line of succession, vice president, majority floor leader, and then speaker.1 In the senate, some states have the Lt. Governor serve as president but for others it is a member of the chamber. Other differences are in the number of leadership positions, the way that leaders are selected, the potential for additional compensation, and whether or not there are limits placed on the individuals who serve in a leadership position (formal or informal). These differences are based on the needs of both the legislature and its members (Squire 1992a). The variation in rules and positions have made it difficult to make generalizable statements about leadership. The first difference among state legislative leadership is in the positions. Who are the chamber leaders and how extensive is the rest of the leadership? The states in this study range from having no party leaders to having an extensive majority and minority office with a clear hierarchy.2 This expanded form of leadership not only includes assistant party leaders or deputy leaders but can often include an army of party whips. Not only are the number of leadership positions different across the states; the names are different as well. There are majority leaders or majority floor leaders, or sometimes a state will have both. Some states utilize assistant minority leaders, others have deputy leaders, and more rare are the caucus leaders. There are sometimes speakers pro tempore or presidents pro tempore, but some states do not use these roles. The Speaker and party whips remain relatively consistent throughout the states compared to the other changes. In addition to the different leadership positions, there are three important factors that vary for leaders. First, the selection process for how a

1 While this is tradition, it is not always maintained (information provided by NCSL). 2 See Table B.1 in the appendix for an overview of top positions and how they are

filled.

122

J. BUTCHER

rank-and-file member becomes a leader. The most common methods of selection are an election by the entire chamber or caucus, or an appointment by the chamber leader. The second factor that varies greatly is the potential for additional compensation for leaders. A handful of states do not offer additional compensation for leaders, some states offer extra pay just for top leaders, while others have compensation for all leaders.3 Extra compensation varies from a few dollars a day (Wyoming) to a 100% increase in a member’s base salary. Third, there can be formal (or informal) limits placed on leadership positions. Placing a restriction on how long a member can serve alters the use of apprenticeships and the tenure of leaders. In addition to all of these differences, the implementation of term limits inherently changed legislative leadership by altering three components of service. Knowing the intention of term limits, scholars predicted that there would be dramatic shifts to leadership. First, legislators would have less experience and rise to leadership positions early. Second, the limited amount of time would change the use of apprenticeships and the development of leadership experience. Third, legislators who rise to leadership positions would only be able to serve for a short amount of time cutting the tenure of leadership positions. Collectively, these changes have the potential to create larger problems within the legislature with the potential to diminish the role of leaders in the chambers where they serve. Experience Term limits, by their very definition, cut the tenure of members affording them less time to gain experience. The extent to which leaders influenced the legislature prior to term limits matters—the more the institution relied on its leaders, the greater the influence of term limits (Little and Farmer 2007; Kousser 2005). Early research centered on the increased rate of turnover and the negative implications for both leaders and the legislature (Capell 1996). Turnover would not only increase but new leaders would have less experience (Little and Farmer 2007). Those who face term limits have less authority because they only have so much time available (Little and Farmer 2007; Kousser 2005).

3 Information on additional leadership salaries can be found in the appendix, Tables B.2 and B.3.

5

LEGISLATIVE LEADERSHIP

123

Given the nature of term limits, it is expected that leaders should have less legislative experience. Prior to term limits, it was uncommon that a rank-and-file member would immediately be placed in a leadership position during their first term (Squire 1992a; Simon 1987). Post-term limits, studies found that members who desired a leadership position were able to move into a position quickly, without knowing much about the role they would be filling (Carey et al. 2000). With limited time in office, there was increased competition for leadership positions in a system where most competitors lack necessary legislative experience (Little and Farmer 2007). The extent to just how much experience a member has prior to becoming a leader is dependent on the type of term limit, as seen in hypothesis 1a and 1b . Hypothesis 1a : In states with more restrictive term limits, the overall legislative experience of leaders will be lower than those who do not serve under term limits. Hypothesis 1b : States with divided/consecutive term limits afford greater time in the legislature and therefore will not be different from states without term limits.

Apprenticeships Before term limits, apprenticeships were viewed as an institutional norm and a stepping stone to becoming a top leader. Members serve apprenticeships in lower leadership positions so that they are better equipped to take on the highest-ranking position of speaker or president. Lucinda Simon (1987) found that seniority was an important criterion to become a leader and most legislative leaders had prior leadership experience, either as a committee leader or an assistant leader. Freeman (1995, 370) found that by the 1990s, nearly 100 percent of speakers had prior experience as a committee chair. Apprenticeships, however, made it a lengthy process to get to a leadership position; once there, leaders would serve for a short period of time before leaving office (Squire 1992a, 193). As term limits were being implemented, there was a great deal of uncertainty around whether apprenticeships were still important to the leadership process. One early study argued that term limits killed apprenticeships (Peery and Little 2003). In contrast, Jennifer Drage Bowser and her colleagues (2003) found that lines to the speakership were not

124

J. BUTCHER

formalized before or after the implementation of term limits. Additionally, Rosenthal (2007) observed that apprenticeships had been declining in states without term limits and that this would likely be amplified in states with term limits. With increased competition and limited time in office, it is difficult for members to serve apprenticeships and have time to climb the leadership hierarchy. Alternatively, term limits may not make a difference at all and states were simply already moving away from the practice. It can be very appealing to rank-and-file members to become a leader, the authority that comes with being a leader encourages more ambitious members to rise through the ranks. Legislators from term-limited states have less time to serve, suggesting that there may not be enough time for lawmakers to serve lower-level leadership positions before rising to become the party or chamber leader. Term limits are not only likely to shift the timeline for members to become a leader, but the changes result in placing greater emphasis on some positions and less on others. In particular, it would be of utmost importance for chamber leaders to have prior leadership experience, as the inexperienced members would need someone to guide them.4 In contrast to these early theories, I present three hypotheses that center on the use of apprenticeships in term-limited states. Hypothesis 2a : Chamber leaders in all states, term limits or not, will be the most likely to serve apprenticeships. Hypothesis 2b : With limited time to serve in office, and hopes of advancement, term-limited legislators are more likely to serve an apprenticeship in order to gain experience than members in non-term-limited states. Hypothesis 2c : Given the lax nature of divided/consecutive term limits, there is less pressure to quickly advance through the ranks of leadership. These states will be most similar to states without term limits.

Tenure Historically, leaders served for only one or two terms, but with legislative modernization, members began to serve longer and leadership 4 Speakers, in particular, have been found to have more power in term-limited states because members rely on them more (Shay 2020).

5

LEGISLATIVE LEADERSHIP

125

turnover decreased (Squire 1992a; Chaffey and Jewell 1972). Overall, the increased incentives of legislative service drove members to serve for longer and stay in leadership positions for longer periods of time. Similar to Congress, tenure equals power and without tenure, members are not able to develop the expertise necessary to make policy (Powell et al. 2007; Coyne and Fund 1992; Rosenthal 1992). More specifically, members with less tenure place less of an emphasis on legislation (Sarbaugh-Thompson and Thompson 2017). Hypothesis 3a : Members in term-limited states have less time to serve, therefore they will serve in their leadership position for less time than members in non-term-limited states. Given the goal of term limits—to expel legislators—scholars have long suspected that leaders should have shorter tenure in term-limited states (Little and Farmer 2007; Kousser 2005; Reed and Schansberg 1996; Squire 1992a). Increasing turnover rates brought speculation over how leaders would be affected. With term limits, leaders are no longer able to build up tenure because of the high rate of institutional turnover (Little and Farmer 2007; Kousser 2005; Bowser et al. 2003; Squire 1992a). Early on, scholars argued that leaders would serve shorter tenures in their position (Kousser and Straayer 2007; Little and Farmer 2007; Kousser 2005), the experience of leaders would decline (Little and Farmer 2007; Kousser 2005), and speakers would not be as powerful as they once were (Kousser and Straayer 2007; Little and Farmer 2007; Kousser 2005). These early findings led to singular tests about the time that legislators are able to serve in leadership positions.

Data and Methods This analysis relies on the TLS-Leaders dataset, which is a subset of data examining leaders from 17 states. These data represent some of the top leaders in the states and give insight into how leadership experience and tenure have changed under term limits. These data contain over 2,400 leaders and how long they have served in the legislature, as well as in their position as a leader. For the quantitative analysis, I also use information collected from the Council of State Government’s Book of the States. In particular, information regarding leadership selection and additional compensation comes from the Book of the States in order to better understand the rules and incentives surrounding leaders. To account for

126

J. BUTCHER

other institutional factors, I rely on legislative professionalization5 (Squire 1992b) and the career opportunities (Squire 1988). These data do include some repeat members, each observation is for each position a member served. While some of the 2,400 observations are single entries, many members have 2 entries or more. For example, Elizabeth Mitchell served in multiple leadership roles in both chambers of the Maine legislature, as such Mitchell has four separate entries in the dataset. This method of recording leaders allows for the following analyses as it becomes easier to track the movement of an individual and whether they have served an apprenticeship or how much experience they have gained. There are three distinct measures used to evaluate leadership changes under term limits. First, is an assessment of political experience. The experience variable is measured as the number of years a member served in the legislature prior to serving in a particular leadership position. Second, there is an assessment of apprenticeships. This asks the question of whether or not a member served in a lower leadership position to gain the necessary skills and expertise to serve as a higher-ranked leader. For example, if a Majority Leader previously served as a majority whip, they would be coded as (1) but if a president pro tempore had not previously served as a whip or party leader, they would be coded as a (0). There are often informal rules that dictate whether or not members must serve an apprenticeship or follow a certain line of succession, but because these are informal and not codified, they can be difficult to track. Last is an examination of tenure. This measure is specific to the leadership position that a member is currently serving in and is a count of how many years they have served in the position. Included in each model are controls that can further current understandings of leaders and their experiences in the legislature. Each model includes an indicator for those legislators who serve in a top leadership position (speaker or president/president pro tempore).6 Leaders in termlimited states have less experience than those who do not serve under term limits, but it should stand to reason that chamber leaders should have more experience than party leaders. There is also a binary indicator 5 In state legislatures, the institutional structure shapes the pathways to becoming a leader matching trends of legislative professionalization (Squire 1992a, 176). 6 The highest internal senate official is the president pro tempore in Michigan, Missouri, North Dakota, Oklahoma, South Dakota, and Texas.

5

LEGISLATIVE LEADERSHIP

127

for members serving in the majority party; while this is not tied to a specific political party, it is important to understand if a leader is part of the majority or minority. Members of the majority party tend to enjoy certain advantages that members of the minority party do not; there are also different internal structures in place for the majority and minority parties that can alter a member’s experience as a leader.

Experience The first component affecting leadership is the overall level of legislative experience that leaders possess. One of the most important qualifications to becoming a leader is the level of experience that a member has (Freeman 1995; Simon 1987). But members need the time and experience of serving in the legislature before rising to become a leader (Freeman 1995). Prior to term limits, leaders had already spent seven or more years in the chamber before rising to the rank of the speaker (Squire 1992a). Term limits, however, can cut an individual’s service off in as little as six years, making it impossible to have seven years of legislative experience. In this section, the changes in overall legislative experience are evaluated to assess how different types of term limits have affected the level of legislative experience that leaders possess. I explore the relationship between term limits and legislative experience by chamber because theories of legislative careerism would suggest senators would have more experience than members of the lower chamber. For the lower chamber, the variable of interest is the number of years a member served in the chamber prior to being elected a leader. For the upper chamber, the dependent variable includes how many years a member served in the chamber prior to becoming a leader and how many years they served in the lower chamber. Both measures capture how much legislative experience a particular member has before assuming their position as a leader.7 Table 5.1 contains the average number of years of legislative experience that leaders had upon taking their position in the same chamber. The table contains the average experience pre- and post-term limit in order to compare changes. The earliest states began to remove legislators 7 Years are used to best account for legislators who enter the legislature at abnormal times for reasons such as an appointment or special election. This also translates better across chambers where there is greater inconsistency in the length of terms.

128

J. BUTCHER

from office in 1996/1998, whereas Nebraska and Louisiana followed 10 years later. Given that the states implemented term limits at their own pace, there is no one cut-off point. Instead, this measure is of the average number of years of experience before and after term limits.8 The states that do not have term limits are cut off at 2000, as this is when most states implemented term limits. This table represents the number of years a member spends in the chamber before becoming a leader, this is not representative of all legislative service for members who have served in multiple chambers. Looking at Table 5.1, there is an evident decrease in experience as states began to implement term limits. In the lower chamber, states drop Table 5.1 Average years of experience for leaders in each state, pre- and postterm limit State

Arizona Colorado Florida Louisiana Maine Michigan Missouri Montana Nebraska Ohio Oklahoma South Dakota New Jersey North Dakota Oregon Texas Wyoming

Pre-term limit

Post-term limit

Lower

Upper

Lower

Upper

6.8 4.9 6.5 14.8 5 9.8 6.3 9.7 – 12.2 6.5 5.6 6.1 8.9 8.4 17* 11

9.7 9.2 13 16.6 7.9 7.4 10.8 11.5 8.4 10.1 9.8 8.3 11.8 11.2 11.8 12.5 9.4

3.3 3.9 4.5 8.9 4.1 3.2 3.8 3.5 – 4.4 5.3 3.1 7.3 8.5 5.2 13 6.8

7.5 6.8 9.2 17.6 8 8.7 8.9 7.8 4.6 9.6 6.1 5.1 12.1 11 12.8 16.3 12.8

For senators who have served in both chambers, prior legislative experience is included (*From 1990 to 2002, there were only 2 speakers, no other individuals were included)

8 Michigan is the only state that has two separate cut-off points, members began to term from the lower chamber in 1998 and the upper in 2002.

5

LEGISLATIVE LEADERSHIP

129

by at least two years in experience. Some states see a decrease in experience by half. Michigan sees ones of the most dramatic shifts going from 9.8 years of experience in the lower chamber to just over 3. This is not surprising, however, given that Michigan allows representatives to serve a mere 6 years in the lower chamber. The short terms in office make 9 years of legislative experience impossible. Oklahoma sees the least dramatic shift and this is likely a result of the unified lifetime limit which allows members to spend up to 12 years in a single chamber. What is worth noting is the consistency seen in the states without term limits. Even with exogenous political shifts occurring from 1990 to 2020, there is very little change in the levels of experience of legislative leaders. While there is a slight decrease in Wyoming, this decrease can partially be attributed to their short-lived term limits. New Jersey and North Dakota saw very little variation in years of experience, while Texas and Oregon saw a shift to more experienced upper chambers. Overall, experience among leaders is decreasing in term-limited states while non-term-limited states have been able to retain leaders with far more legislative experience. These results should not be surprising as term limits inherently cut off the length of time that a member can serve, but it remains important to understand the extent to which experience is cut off. Given the design of term limits, it is expected that there will be a decline in overall legislative experience but it is important to understand what this decline means, in which chamber it occurs, and who tends to be the most affected. To account for the pooled time series cross-sectional nature of the data, I utilize random-effect negative binomial regression. This model, in turn, accounts for the dispersion of the dependent variable among the states. While the coefficients of this model are not directly interpretable, they do give insight into the relationship between experience, term limits, and the other explanatory variables. Table 5.2 illustrates the relationship between term limits and legislative experience. There are a couple of important initial takeaways. First, prior experience appears to be correlated to what chamber a leader is serving in. Second, the added elements of either being in the majority party or being a top leader contribute to the years of experience a member had before serving in a leadership role. The findings in Table 5.2 suggest that there is a key difference in the number of years of experience that members in the upper chamber are able to gain; this is significantly hindered in states that have term limits. Whereas in the lower chamber, leaders appear to have significantly more experience than their counterparts in other states.

130

J. BUTCHER

Table 5.2 TSCS negative binomial regression model of legislative experience post-term limits

Divided/consecutive Divided/life Unified/consecutive Unified/life Majority party Speaker President/President Pro Tem Professionalization Career opportunity Constant Observations Number of states

Lower Chamber

Upper Chamber

0.821*** (0.314) 2.206*** (0.747)

−0.0422** (0.088) −0.0482*** (0.154) –1.171*** (0.294) –0.799*** (0.172) 0.080* (0.046)

1.642*** (0.560) –0.096** (0.045) 0.504*** (0.056) –4.056** (1.855) 4.064*** (1.315) 0.999*** (0.347) 896 16

0.254*** (0.055) 1.003* (0.537) –0.239 (0.426) 1.726*** (0.149) 769 17

Standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1

The contrast in legislative experience between chambers is intriguing and presents a sort of puzzle. While overall legislative experience is known to be decreasing, the experience of term-limited lawmakers is not so simple. For example, in divided/consecutive states members have more experience upon assuming a leadership role; this remains true for divided/ life and unified/life term limits. The most likely explanation for this is that because lower chambers tend to be the larger governing body, there are more individuals to choose from for leadership positions. Knowing there is already a loss of institutional knowledge, those who serve in termlimited states are more likely to select the most experienced lawmakers to serve as leaders. Thus, rather than arguing that term-limited leadership has far greater experience, these results indicate that it is the most experienced legislators who are leaders. Beyond the initial examination, there are some important findings that illustrate other connections to the experience of leaders. For those who serve in the lower chamber professionalization appears to play a significant role in the level of experience of leaders. Interestingly, as a legislature professionalizes the experience of leaders declines. One argument could be that members from more professional legislatures are more likely to view their service as a career and therefore may be more likely to seek

5

LEGISLATIVE LEADERSHIP

131

out other elected offices. As seen in the upper chamber, professionalization leads to more experienced leaders. This analysis also indicates that top chamber leaders (the speaker and president/president pro tempore) are significantly more likely to have extensive experience. This is substantively important, particularly for term-limited states where overall experience is declining due to term limits, it is essential to legislative function that the top leaders have significantly more experience. Perhaps the most interesting result is the relationship with experience and majority party. While this relationship is significant in both legislative chamber, it diverges in sign. In the lower chamber, leaders who belong to the majority party have significantly less experience. Likely because there are higher numbers of new and inexperienced members in the lower chamber. In the upper chamber, members of the majority party have more legislative experience than those who are not in the majority. This could simply be that the leaders in the lower chamber tend to advance to the upper chamber at a more rapid pace where they can enjoy greater advantages as part of the majority party. Because of the complexity of the negative binomial regression model, there is some additional interpretation needed to decipher the estimated effect of each variable. The probabilities for the negative binomial regression model are rooted in the assumption that the random effect is zero. To assess the importance of each variable from Table 5.2, professionalization and career opportunity are held at their mean values, and chamber leaders are set at (0), while majority party status is set at (1). Figures 5.1 and 5.2 present estimates for the predicted level of experience for each type of term limit, as well as states without limits, in the post-term limits era. These figures illustrate the predicted level of experience and the 95 percent confidence intervals for each type of term limit by legislative chamber. In the lower chamber, a member of the majority party who is serving in a state without term limits likely has very little experience. Members from lifetime states are more likely to have experience, holding all else constant. For each scenario, there is a consistent trend that members in term-limited states have 2 or more years of experience. When a member serves in a state without term limits, they have just over a year of experience before taking on a leadership position. In a divided/lifetime state, members from the majority party have nearly 4 years of experience before taking on a leadership role. This relationship between experience and the

132

J. BUTCHER

Div/Cons

Div/Life

Uni/Life

No Limit 1

2

3 Legislative Experience

4

5

Fig. 5.1 Predicted level of experience of the majority party for the lower chamber

Div/Cons

Div/Life

Uni/Cons

Uni/Life

No Limit 0

.5

1 1.5 Legislative Experience

2

Fig. 5.2 Predicted level of experience of the majority party for the upper chamber

5

LEGISLATIVE LEADERSHIP

133

type of legislature is likely a result of the competition that exists for leadership positions. In term-limited legislatures, members have less time to make a name for themselves, making leadership positions a highly valued commodity that are reserved only for senior members who are on their way out. In the upper chamber, leaders from states with term limits are less likely to have experience. For example, members who serve in a unified/ consecutive legislature and serve in the majority party have about 1 year of experience in the chamber before taking on a leadership role. Comparatively, those who do not have term limits tend to have far more experience. Members from the majority party who serve in a state without term limits do not enter a leadership position until they have roughly 2 years of experience. Members from unified/lifetime states are predicted to have the least amount of experience upon assuming a leadership role. The support for the experience hypotheses is mixed, as there is a clear division between chambers. Overall, these analyses reveal mixed support for hypothesis1a , where the differences between states with and without term limits vary by chamber. There is some, albeit weak, support for hypothesis 1b , that divided/consecutive limits in particular are closer aligned to patterns seen in states without term limits. Broadly speaking, divided/consecutive states appear to have trends most like that of states without term limits. Conversely, states with more strict unified term limits, which restrict movement between chambers, are more likely to have inexperienced leaders and the overall trends are distinct from states without term limits. Taking experience into account, term limits do appear to alter legislative organization and leadership structures. Term limits significantly decrease the level of legislative experience that leaders in the upper chamber have, while actually increasing the level of experience of lower chamber leaders. This particular finding is perhaps more indicative of the organization of the house rather than the individuals in the leadership positions. In the lower chamber, the members with the most experience are those who rise to the highest office, indicating underlying changes to term-limited legislatures, including an institutional dependency on more senior members.

134

J. BUTCHER

Apprenticeships The second anticipated change produced by term limits is a diminished use of apprenticeships. In this context, an apprenticeship is when a leader first serves in a lower-level position in the legislature. For example, a majority leader first serves as a majority whip, then assistant majority leader, before finally becoming the majority leader. As legislatures modernized, apprenticeships became an important component (Squire 1992a; Simon 1987) and members began to serve long apprenticeships before rising to more prominent positions (Squire 1992a; Chaffey and Jewell 1972). Apprenticeships are currently understood within the context of time, or how many terms a legislator can serve. In states without term limits, members are more likely to serve apprenticeships because they have the time to do so. In term-limited legislatures time is finite; leaders are not able to wait until their third or fourth session to assume a leadership position. This is amplified for positions such as the majority and minority leader where both the majority leader and the speaker are likely in their last term, so there is nowhere for them to advance within the institution. If a member needs to serve an apprenticeship prior to becoming majority leader and then speaker, they often must be elected as a committee chair or party whip during their first term. Term limits speed up the legislative process for all members but especially those who desire leadership positions. In this section, I evaluate to what extent term-limited legislatures utilize apprenticeships by examining whether or not leaders have previously served in another leadership position. In addition to the term limit indicators, there are indicators for the top four party positions: chamber leader, chamber leader pro tempore, party leader, and assistant party leader. There are also controls for members of the majority party and for a member’s overall legislative experience (years). Table 5.3 includes the results of a time series logit analysis for each legislative chamber. This analysis not only evaluates the relationship between the different types of term limits and the likelihood of a member serving an apprenticeship, but it also reveals the leadership positions that are more or less likely to serve an apprenticeship. The results in Table 5.3 indicate that there is a difference in the importance of apprenticeships among the term-limited states. Notably, the states that have lax term limits are more likely to have leaders serve

5

LEGISLATIVE LEADERSHIP

135

Table 5.3 Logit analysis of leaders who serve apprenticeships post-term limits

Divided/consecutive Divided/life Unified/life Chamber leader Chamber leader pro tem Party leader Assistant party leader Majority party Legislative experience Professionalization Career opportunity Constant Observations Number of states

Lower chamber

Upper chamber

0.564* (0.295) 0.300 (0.552) 1.618*** (0.579) 3.098*** (0.389) 1.661*** (0.410) 2.232*** (0.347) 0.978** (0.388) -0.126 (0.225) 0.076*** (0.027) -3.630** (1.799) 2.673* (1.465) -3.851*** (0.592) 896 16

0.757* (0.445) 1.537* (0.787) 2.531*** (0.879) 3.503*** (0.484) 1.597*** (0.476) 2.775*** (0.396) 1.690*** (0.409) -0.293 (0.227) 0.179*** (0.034) -6.235** (2.642) 3.950* (2.105) -4.519*** (0.799) 760 16

Standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1

an apprenticeship. Lawmakers who serve in divided/consecutive and unified/lifetime states are significantly more likely to serve an apprenticeship than those who do not face term limits. There is a division between chambers for states that have a divided/lifetime term limit. In the lower chamber, apprenticeships do not appear to be emphasized any differently from states without term limits. However, in the upper chamber, legislators are more likely to serve an apprenticeship (p < 0.1). This analysis also indicates which leaders are more likely to serve an apprenticeship. In both chambers, each of the leadership positions is more likely to have served an apprenticeship, with party whips being the base category. Interestingly, there is no difference between the majority and minority parties in serving an apprenticeship. There is statistically significant relationship between lawmakers with more experience and serving an apprenticeship, but this is an endogenous relationship where one factor begets the other. Last, the two institutional controls indicate that greater career opportunities encourage apprenticeships in both chambers, while professionalization decreases the likelihood of utilizing apprenticeships. In order to compare the use of apprenticeships across the different types of term limits, I estimate the probability of the various leadership positions serving an apprenticeship. Figure 5.3 presents the probability of

136

J. BUTCHER

four different positions serving an apprenticeship by chamber. For each version of term limits, the left line is for the lower chamber and the right is for the upper chamber. At first glance, it is evident that term-limited states have a higher probability of leaders serving an apprenticeship. Across the board, there is some level of consistency by type of term limit, but there is less consistency by leadership role. One might assume that the top leader, or chamber leader, has served an apprenticeship before taking on the highest-ranking position. It appears that this is not always the case. Now, in term-limited states it does appear that chamber leaders are more likely to serve an apprenticeship in the upper chamber than the lower chamber. Perhaps not surprising, the party leaders appear to be more likely to serve an apprenticeship than pro tem. In support of hypothesis2a , chamber leaders have a higher probability of serving an apprenticeship in all states and in both chambers. The overlapping confidence intervals indicate that there is not a difference in

Chamber Leader

0

.2

.2

.4

.6

.8

Apprencticeships 1 .4 .6 .8

1

Pro Tempore

No Limit Div/Cons Div/Life

Uni/Life

No Limit Div/Cons Div/Life

Uni/Life

Assistant Party Leader

Apprencticeships 0 .2 .4 .6 .8 1

0 .2 .4 .6 .8 1

Party Leader

No Limit Div/Cons Div/Life Lower Chamber

Uni/Life Upper Chamber

No Limit Div/Cons Div/Life Lower Chamber

Uni/Life Upper Chamber

Fig. 5.3 Predicted probability of serving an apprenticeship, by leadership position

5

LEGISLATIVE LEADERSHIP

137

the use of apprenticeship for top leaders. The differences, however, are notable for the other positions. For pro tempore leaders, there is a great deal of overlap for states with divided limits and states without term limits. Yet states with unified/lifetime term limits have a much higher probability of pro tem leaders having previously served in another leadership role. For the top two positions in all states, it is more likely for legislators to have first served in another position; this is also where the differences between states with and without term limits are less pronounced. There is some support for hypothesis 2b that term-limited legislators are more likely to serve an apprenticeship. For example, party leaders in term-limited states are more likely to have served in another leadership position. In states with divided/lifetime term limits party leaders in the upper chamber are significantly more likely to have served an apprenticeship than those in states without term limits. This finding holds true for both chambers of states with unified/lifetime term limits. The findings for assistant party leaders are unique for two reasons. First, the lifetime term-limited states are distinguishable from states without term limits. Second, this is the position where differences between chambers appear to become more prominent, where leaders in the upper chamber have a higher probability of having first served in another role. The most consistent result is that the more stringent the term limits, the more likely that a leader served an apprenticeship. While this finding may be surprising, it has important substantive implications. Even though it may be counterintuitive that the most stringent term limits are the ones that have an embedded system of apprenticeships, it ought to be encouraging. The most stringent term limits are those with the least amount of time or with strict lifetime bans, meaning the level of experience a member can gain is quite restricted. But what this means is that even though experience, as a whole, is limited those who serve in leadership roles have at least some form of leadership experience. It is almost as if the legislature has structured it in such a way that the least experienced chambers provide avenues for the most experienced leaders. Despite early fears that apprenticeships would no longer be utilized in term-limited states, this test of term limits reveals that apprenticeships remain a staple in the states. This finding is particularly true for states that allow members to serve for a longer period of time, while simultaneously encouraging members to quickly seek out positions of importance. The time component of the term limit is essential to affording lawmakers enough time to actually gain experience and serve in multiple positions.

138

J. BUTCHER

Legislators in states with lifetime limits have a higher probability of their leaders serving in other leadership positions. States with divided/ consecutive term limits remain similar to those states that do not have term limits.

Tenure The third anticipated change produced by term limits was a decline in leadership tenure, or how long a member serves in their position. This section is centered on the differences in length of tenure between states with term limits and states without term limits. The variable of interest is the length of time served in a given leadership position. Here, tenure is measured in terms rather than singular years, because members are often selected for a leadership position by term not a calendar year. Specifically, the tenure variable is a count of the terms served in the top leadership positions. As Schlesinger (1966) noted, political time is different than regular time and is based on electoral terms. Table 5.4 presents the average number of terms that a leader serves pre-and post-term limit by chamber. Looking at Table 5.4, there appears to be little variation in the average tenure length among the states pre- and post-term limits. For the most part, legislatures have remained consistent in the tenure of leaders with most in the lower chamber serving just under two terms and those in the upper chamber serving between half of a term and one and a half terms. There are few instances among the states where tenure decreased; rather, in most instances, there were (minimal) increases to how long leaders serve in their positions. In order to better understand this, I test the hypothesis that termlimited leaders serve shorter tenure in office. The independent variables are the term limit indicators, in total there are four indicators. There are also indicators for each leadership role because different institutional norms can restrict the terms of certain leaders. In addition to the institutional control variables, there is an indicator if the leader is part of the majority party. There is also an indicator for those positions that have additional compensation for leaders (1) versus those that do not (0). Last, I control for legislative experience, this is a count of years of experience prior to serving in the position.

5

139

LEGISLATIVE LEADERSHIP

Table 5.4 Average terms of tenure for each state, pre and post-term limits Pre-Term Limit State Arizona Colorado Florida Louisiana Maine Michigan Missouri Montana Nebraska Ohio Oklahoma South Dakota New Jersey North Dakota Oregon Texas Wyoming

Lower 1.2 0.96 0.94 1 1.2 1.5 1.2 1.1 – 1.2 1.2 1.2 1.6 0.8 1.1 2.8 1.1

Post-Term Limit Upper 1.1 0.55 0.55 1.3 1.1 1.2 0.79 0.71 0.83 0.73 0.78 1.3 1.4 1.1 1.1 0.3 0.5

Lower 1.4 1.4 0.96 1.1 2.4 1.7 1.7 1.3 – 2.2 1.5 1.4 1.3 0.76 1.8 N/A 1.2

Upper 1.5 1.6 0.54 1 1.6 1.6 0.92 0.91 0.83 0.92 1.1 1.3 0.88 0.66 0.79 0.26 0.52

Table 5.5 contains the results of a negative binomial regression model for leadership tenure post-term limit implementation.9 At first glance, it does not appear that leadership tenure is significantly different in states with term limits from states without term limits. With the exception of unified/life limits, where in the lower chamber leaders serve in their position for significantly less time. Overall, term limits do not appear to bring any significant changes to leadership tenure. The only noticeable difference is in the upper chamber of unified/lifetime term limits where leadership tenure does decrease. Interestingly, speakers have significantly longer tenure but chamber leaders in the senate do not. Additional differences in the lower chamber include decreased tenure for members of the majority party and a decrease in tenure where there is additional compensation. Expectantly, increases in professionalization lead to lengthier tenure.

9 The negative binomial regression model does a better job at predicting tenure than the Poisson regression model due to over dispersion (Long 1997).

140

J. BUTCHER

Table 5.5 Negative binomial regression model for leadership tenure, post term limits

Divided/consecutive Divided/life Unified/consecutive Unified/life Chamber leader Chamber leader pro tem Party leader Assistant party leader Majority party Additional compensation Legislative experience Professionalization Career opportunity Constant Observations

Lower chamber

Upper chamber

−0.106 (0.132) −0.035 (0.158)

−0.155 (0.237) −0.163 (0.317) −0.445 (0.335) −0.509** (0.242) −0.003 (0.142) −0.207 (0.157) 0.030 (0.082) 0.063 (0.074) 0.044 (0.072) −0.003 (0.155) −0.035*** (0.013) 0.208 (1.130) −1.219 (0.803) 0.483* (0.261) 744

−0.077 (0.151) 0.211*** (0.071) −0.057 (0.104) 0.092 (0.077) −0.039 (0.050) −0.166*** (0.050) −0.183** (0.073) 0.0001 (0.014) 0.997*** (0.354) −0.306 (0.437) 0.224 (0.175) 880

Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1

In contrast to hypothesis 3a , the preceding analysis indicates that there is very little difference in leadership tenure between states with term limits and those without. Predicted estimates of leadership tenure are included to better understand the differences between the different types of limits: pre- and post-term limits. Looking at this relationship before and after term limits will allow for an evaluation of whether or not term limits have shifted tenure or if this particular component is unchanged. Figure 5.4 illustrates the predicted number of terms that a top leader will serve (institutional factors are held at the mean). The figure on the left contains the predictions for the lower chamber, on the right are the predictions for the upper chamber.

5

Lower Chamber

141

.4

.6

.6

.8

.8

1

1

1.2

Tenure 1.2

1.4

1.4

1.6

1.6

1.8

1.8

Upper Chamber

.4 No Limit

LEGISLATIVE LEADERSHIP

Div/Cons Pre Limit

Div/Life

Uni/Life

Post Limit

No Limit Div/Cons Div/Life Uni/Cons Uni/Life Pre Limit

Post Limit

Fig. 5.4 Predicted Leadership Tenure for Both Legislative Chambers, Pre- and Post-Term Limits

Figure 5.4 illustrates the shift in predicted leadership tenure moving from not having term limits to having term limits for each legislative scenario. These predicted values are once again for members who serve in the majority party, all other control variables are held at their mean. Beginning with the lower chamber, the results for states without term limits illustrate the effect of time on leadership tenure. Even though there are likely some shifts caused by term limits, the results for states without term limits illustrate that there is also a time component. Before the implementation of term limits in other states, those who did not have term limits served in a leadership position for an average of 1.4 terms; this declined to under 1.2 post term limits. For the states that implemented term limits the shift in tenure is more noticeable. For majority party members in divided/consecutive states, members once had 1.4 terms of experience in a particular role, not indifferent from states without term limits. However, with term limits in

142

J. BUTCHER

effect this declines to just one term in a particular position. The results for divided/life and unified/life term limits are similar, with noticeable declines in tenure post-term limits. Overall, there is a lack of support for the hypothesis that term-limited legislatures would see lower levels of turnover than their non-term-limited counterparts. Focusing on the predicted values, the predicted number of terms in a particular role are not distinct between the term-limited and non-term-limited states. Given the similarities between states with and without term limits, there is not support for the theory that term limits reduce the rate of tenure for leaders.

Leaders in the States The preceding analyses revealed inconsistencies in both theory and practice. Scholars anticipated that legislatures would come to deemphasize apprenticeships, leaders would serve in their position for less time, and leaders would have less legislative experience. The results of these examinations vary widely and are largely dependent on the type of term limit. The legislative experience of leadership does have a noticeable decline but only in the upper chamber. While states do appear to have declined the use of apprenticeships for some positions, term-limited states tend to utilize them more than states without limits. The worry over decreasing tenure in any given leadership role does not appear to have come to fruition; in fact, members in term-limited states serve for the same length of time as those in states without limits. The decreased level of experience has the potential to affect leaders and how they function in the legislature. Term limits, by prematurely cutting off legislative careers, have created a dichotomy between formal and informal leadership. The type of term limit determines the role of leadership, while also encouraging/ discouraging informal leaders based on either legislative or professional experience. The results above illustrate that leaders serve for the same amount of time in their position as those in states without term limits. However, there are differences in how much experience leaders in termlimited states have. Understanding how rank-and-file members view their leaders can help to explain why leaders in the lower chamber have more experience, while those in the upper chamber have significantly less experience. Perhaps in the lower chamber, only the most experienced are selected for leadership positions, whereas in state senates, because legislators are more equal, who is in a leadership role matters less.

5

LEGISLATIVE LEADERSHIP

143

Many of the observed changes in the previous analyses result in arguably unobservable outcomes. It is easy to count the number of years that a member serves; it is far more difficult to understand what those years mean to others who serve in the legislature. Leadership tenure, or lack thereof, can have many implications for leaders and the members that follow them. When a member serves in a top position during their last term, they do not exercise the same level of authority as members in states without term limits do. Those term-limited leaders do not have the same ability to follow through or persuade rogue members, because they will be gone by the end of the session. This section attempts to capture some of those unobservable effects by examining whom legislators view as influential and how much they think their leaders matter. To gauge the influence of legislative leaders relative to informal leaders, members were asked, “Thinking about the legislature, who are the most influential members?” When members asked for a definition of influence, they were told it was however they wished to define it and whoever they believed was influential. Members who asked were told they could list either specific members or positions, whichever they were more comfortable doing. The responses vary by state and member, but there are discernible trends for each state. Table 5.6 contains the aggregate responses of who members believe is the most influential person in their legislature. This table includes the three most common responses. First, are the majority party leaders, this includes the speaker and president/president pro tem. This response is rooted in the office, rather than the individuals who hold that office. Second, some respondents named multiple individuals who they viewed as influential including: speaker, majority leader, minority leader, assistant majority and minority leader, and committee chairs. The final category is other; this category does not include either party or chamber leaders. For example, a handful of Oklahoma legislators pointed to the freshman class as being the most influential while in Nebraska members pointed to Senator Ernie Chambers as having the most influence. There are clear themes in each state when it comes to influence. In Missouri, influence appears to be vested in the majority party and others. Legislators either view power as being vested in the speaker and majority party or in other individuals who do not serve as a party leader. In Nebraska, there is a clear trend toward individuals being the most influential. Only one respondent stated that the majority party (in the formally non-partisan body) held the most influence, whereas eleven members

144

J. BUTCHER

Table 5.6 The most influential members in term-limited states Lower chamber

Upper chamber

Total

State

Majority party

Multiple leaders

Other

Majority party

Multiple leaders

Other

Missouri Nebraska Oklahoma Ohio Total

11

6

10

6 9 26

7 5 18

2 5 17

4 1 2 1 8

1 2 3 1 7

4 11 2 1 18

36 14 22 22 94

named individual colleagues. In Ohio nearly half of the respondents cited the majority party as being the top leader. The split responses suggest a differentiation based on party, those of the majority party find the majority party to be the most influential. In Oklahoma, influence is again somewhat removed from the majority party, instead respondents were more willing to name multiple leaders from both parties. Additionally, there are a handful of individuals who think others outside of leadership are the most influential. Each state has a slightly different leadership structure, which can be difficult to capture in the above analyses. In the remainder of this section, a sub-sample of four states and their leaders are evaluated in-depth. Using interviews from the states, this section evaluates the role and importance of leaders in term-limited states. Influence in Nebraska appears to be very decentralized, whereas influence is highly centralized with the speaker and majority party in Missouri. Within the Ohio legislature, the majority party is highly centralized under the majority party leadership; those in the minority party seek out other influential leaders. In Oklahoma, however, power seems to remain with multiple party leaders, not just the speaker or president pro tempore. This section captures the feelings of actual legislators and how they view their leaders. Divided/Lifetime Missouri was one of the states that had a long-serving speaker prior to the passage of term limits. Bob Griffin served as Speaker of the General Assembly from 1980 to 1996, totaling 16 years as Speaker and 26 years of service. Ultimately, Griffin who is the longest serving speaker in state

5

LEGISLATIVE LEADERSHIP

145

history resigned his position due to a federal corruption investigation and even briefly spent time in prison before having his sentence commuted.10 What is worth noting is that Griffin served as Speaker for the same period of time that members are restricted to for a lifetime of service. Because of term limits, Griffin will forever remain the longest-serving speaker in state history. In the Missouri legislature responses suggest a highly centralized power structure with slight variations between chambers. Specifically, members in the lower chamber consistently view the speaker as highly influential. In the upper chamber the leadership holds a great deal of power, but senators tend to feel more equal with their leaders. Overall, there tends to be agreement that top leaders and party leaders hold the most influence. As one member put it succinctly, “speakers and presidents are most influential.” In the lower chamber, many members found the speaker to be the most influential. A second-year representative said simply, “well, of course, it’s the speaker.” One member said the speaker is influential because he says what goes and what does not. A more senior legislator explained that the legislature has a “top heavy structure with the Speaker of the House in terms of power.” One representative stated, the speaker is the most powerful, “by virtue of the power we give them,” adding, “I didn’t realize how much power the speaker has.” A freshman representative remarked, “We do not have as much say as I thought we did.” They continued by saying that everything has to go through leadership. The same representative said that the party leaders are the most influential it is, “not even a question.” Another freshman representative went so far as to say, “The system is flawed” because of the immense amount of power the speaker has. A more senior representative argued that the speaker was important because they have the ability to decide what goes to the floor and what does not. A different member responded to the question of influence saying, “the Speaker of the House, absolutely,” because of the control they have over the legislature. A state senator, who had previously served in the House, stated that the president, floor leader, and committee chairs have a lot of control, but “all senators will get a say.” They went on to say the house is different, not everyone will get a say. The same senator said that power in the 10 Last accessed October 10, 2023. https://missouriindependent.com/2021/07/07/ bob-griffin-longest-serving-missouri-house-speaker-dies-at-age-85/

146

J. BUTCHER

house is vertical, whereas in the Senate all members get a say regardless of who the floor leader is. A representative, in reference to the Senate, said “(the) Senate is designed to force the majority to listen to the minority.” A member pointed out that this is in the design of the Senate, as the Missouri Senate allows for a filibuster. One senator said the most influential members are the ones you trust the most, that you should value character over position. Others agreed that influence is entirely relationship-based. A representative stated that the most influential members are the ones “who don’t forget why they’re here.” In the same vein, a different member said that influence is based on character. A different perspective, offered by a senior legislator, “anybody who really works at it can be influential.” Outside of leadership, several members find the Budget Committee and Appropriations Committee chairmen to be very powerful. Others said the more experienced members are the most influential, as are members with greater expertise. Established friendships and relationships also influence members. Another individual argued the most influential members are the ones who are able to work across the aisle behind the scenes. A different representative noted those with influence are “not always those in leadership positions. Depending on the issue, different members are the most effective and influential.” There are a select number of outlier members who do not think any one person was particularly influential. For example, one representative said, “As all members are elected by voters, they share the same and are equal. Positions give you more latitude to do many things, the only one with greater power is God.” Another said “The most influential members are the ones who have been productive in life, not career politicians or those successful in private enterprise.” And finally, one senator said the most influential are those that know how to do the job, “20 percent of people understand the game,” the rest of the legislature does not. Unified/Consecutive Although Nebraska is excluded from much of the analysis in this chapter, it is still important to understand the relationship that members have with leadership. Nebraska has only one formal leadership role, the Speaker of the Senate. While there is a president pro tempore in the Senate, this role is less formalized and difficult to find information on. Historically, members did not serve in speakership role for an extensive period of time.

5

LEGISLATIVE LEADERSHIP

147

A senator’s term in Nebraska is 4 years, but most speakers before term limits would only serve as speaker for a single 2-year session amounting to half of a term. Legislators in Nebraska have a different experience with leadership and who they view to be important. First, Nebraska is unique because they do not have majority and minority party leaders. The lack of party leaders dissipates the central structure that is seen in many other legislatures. Given that Nebraska has one chamber and one elected leader, it would be easy to assume the speaker holds a great deal of power. In Nebraska, however, legislators do not view the speaker as one of the more important figures. As one member shared, “No one is more remarkable than anyone else, truly, and I think that is a result of term limits.” When asked about influence, few members cited the speaker as having a great deal of influence. Only one respondent listed the speaker as having the most influence. In fact, one member cited leadership as having the least knowledge. Instead, members listed other individual members, or even multiple members (rarely including the speaker) as being the most influential. It is clear that Nebraska state senators do not view the speaker as a powerful actor. Rather, they tend to value those members with more experience. One senator stated they felt people with the most institutional knowledge were the most influential. In particular, Senator Ernie Chambers has earned the respect of many of his colleagues. When asked, “Who is the most influential member?” many senators pointed out Ernie Chambers saying, “Ernie Chambers is the one I’d day is truly influential,” and “outside of the legislature (in the media) Senator Chambers is the most influential.” Other members stated that they felt all of the senators were pretty equal. If anything, the most influential member changes with time and subject. Two senators remarked that the most influential members are the ones who have earned trust on both sides of the aisle. Another stated, the influential members are the ones who get stuff done. Outside of member characteristics, a couple of senators noted they felt appropriations and revenue chairmen were the most influential. In Nebraska, it appears that the power of the speaker pales in comparison to those members who have more experience or are better at communicating across the aisle. The question is, is this a Nebraska phenomenon linked to their non-partisan unicameral legislature? Or do term limits actually limit the role of the speaker? Then there is the added

148

J. BUTCHER

dynamic of not having any formal party leadership. It is, at this point, difficult to parse out the different pieces of the Nebraska legislature and the changes that they bring, however, these results are likely due to a combination of all of the institutional factors. Unified/Lifetime Before the implementation of term limits, it was not uncommon to have members of the Oklahoma legislature serve for a single term in a leadership role, but there are some members who served either lengthy terms or numerous consecutive positions. Herbert Rozell happens to be one member who served for an extensive period of time served at the Assistant Majority Leader for 20 of his 28 years in office before he finally termed out in 2004. Upon his exit from the legislature there was a resolution thanking Rozell for his service which included the following statement: STATE OF OKLAHOMA, 2nd Session of the 49th Legislature (2004) “WHEREAS, one of the consequences of the term limits initiative will be to deprive this body of some of its unique characters, of whom Senator Herb Rozell is certainly one, as evidenced by his unceasing questioning during committee meetings, and his tireless and relentless work when presiding to ensure the smooth and orderly flow of legislation. These qualities, along with his fairness and sense of order, will be missed when his service concludes.” In the Oklahoma legislature there is a far more extensive leadership structure. In the House, there are 37 individuals with a leadership position, 26 in the majority party, and 11 in the minority party.11 The speaker has 25 leaders directly under him. In the Senate, there are 17 leadership positions, 12 in the majority, and the remaining five in the minority. This gives the president pro tempore control over 11 other leaders. The extensive leadership structure presents a different dynamic than what is seen in Nebraska. A unique characterization came to light in Oklahoma. In responding about influence, members typically stated that [blank] was the most “powerful.” Members were, across the legislature, equating influence with power unprompted; this is not something that was as evident in the other states. One legislator commented that those in power have the most

11 As of the 2019 session, according to the Oklahoma State Legislature Website.

5

LEGISLATIVE LEADERSHIP

149

influence, especially with the super majority that exists in Oklahoma. A different legislator said, the “ones with the most power are not the most inspiring or helpful, but they are the most influential.” It was common for Oklahoma legislators to point to leadership as the most influential. One legislator stated that power is vested in leadership but trickles down to committee chairs, this speaks to the extensive leadership structure in Oklahoma. A respondent even commented on how centralized the leadership is in the legislature. Another member stated that leaders can give and take access, making them more influential. A different legislator said, “outside of leadership I don’t think anyone has really been able to stand out as influential.” The Speaker in particular was noted as being influential to some. Some of the comments about the Speaker include, “Speaker, of course,” “Speaker. Absolutely,” and “Speakers have a tremendous amount of power in both the House and Senate.” There were some other positions that were noted as being influential by some legislators. In particular, the Appropriations Committee chair, the floor team, and the majority whip. One member said, “People are influential because of their position,” citing party leaders as being more influential. One person responded that what makes a member influential is that they are respected by their peers. Members cited different important traits that they look for; these were centered on personal relationships and hard work. For example, one member responded, “People are influential because they are hard-working and committed, when they speak others listen.” Several members also valued outspoken members and senior legislators. The different conversations that I had with legislators reveal that institutional rules and culture have a greater influence on leadership than term limits. The variation in the importance of leadership does not appear to be directly connected to term limits, but rather a product of the longstanding institutional culture. For example, in Nebraska it is not the Speaker that is most influential, but individual members. This difference is not a product of term limits, but is a result of the non-partisan legislature. Similarly, an institution that vests greater authority in leadership will have stronger leaders. This is a dichotomy that exists regardless of term limits.

150

J. BUTCHER

Divided/Consecutive Ohio has a unique situation with its current leadership. The Speaker of the House during my interviews was Larry Householder.12 Householder is unique because he took advantage of the lax term limits present in Ohio. Householder first began serving in 1996. He was the speaker from 2001 until he was termed out in 2004. Only in a divided/consecutive state, like Ohio, can a previously termed out speaker return to office and resume the role of Speaker of the House. Householder has more experience as both a representative and a speaker than what would be expected in term-limited states. Given that Householder has more experience, the expectation is that he would carry more influence than a lesser experienced speaker. The Ohio House leadership is pre-disposed to have great influence because of the formal rules that favor their authority. One representative said the top leaders are the most influential—people in leadership positions are voted there by the body, this increases influence through positional authority. Another noted, “structurally, its the leadership.” One freshman member asserted, “the speaker is in charge.” Others pointed to specific roles, noting that the speaker and president are the most influential because they are the policy gatekeepers or that the leadership teams are great guides through the legislative process. Within the Senate, one member said that the president is the most influential at moving legislation, but all members of the Senate are potentially equally influential. In a legislature like Ohio, where members are permitted to return, senior members are able to exert a great deal of authority. The knowledge and expertise that senior members have gained throughout their time helps them to influence and shepherd the new members. One of the more senior representatives serving said that leadership and seniority make someone influential, regardless of party. They continued saying, leadership is the most influential, “(there is a) Republican revolving door with the Senate, more senior members and more influence.” A different representative said, “it varies. . . we have a wide breadth of expertise and members with different backgrounds. . . people are influential in terms of experience.” One senior legislator, with prior experience in both legislative chambers, said the most influential members are the ones who have been 12 For a completebiography on Speaker Householder, http://www.ohiohouse.gov/ larry- householder/biography

5

LEGISLATIVE LEADERSHIP

151

around the longest. The same member also gave a warning, “not to rely upon leadership, listen to what they have to say but do not depend on it.” For some members of the Ohio legislature the most influential members depend on the topic at hand. Several members pointed to those who are in the know as having the most experience. The members who pay attention and have their hands in policy are able to influence other members. When it comes to specific policies, it is the members who have studied and looked at the legislation who are able to exert influence. Other members stated it is those who connect well with others, a high “emotional IQ,” that are able to influence others. A more senior representative said that influential members are simply the ones who discuss openly and freely based on relationships, expertise, and backgrounds. The conversations that I had with different members brought to light the unique differences that exist from state to state. While some members largely value their defined leadership roles, other states place more emphasis on those senior, experienced members. By understanding the different ways that members may view their leaders, formally and informally, it becomes easier to understand some of the trends seen in the quantitative analysis. In several ways, the changes seen were not as stark as anticipated, but that is likely a result of the shifting emphasis on leaders in some legislatures to those that simply have more experience. Whereas in other states, like Missouri, that emphasis on formal leadership roles has deepened.

Summary Leaders are important to state legislatures because, having the backing of a majority of the members, they are better equipped to deal with external actors. Leaders not only guide and direct rank-and-file members, but they are typically among the most experienced members in the chamber. With the passage of term limits, it was predicted that leadership would be most effected in the areas of legislative experience, apprenticeships, and position tenure. This chapter sheds light on the changes that were brought by term limits and what term limits left unchanged. The first, and perhaps most obvious way, that term limits were to impact leadership was a decline in overall legislative experience. Term limits cut off legislative service, therefore legislators would have less experience when they assume a leadership role. Early scholars predicted that

152

J. BUTCHER

term limits would put members with limited legislative experience into top leadership positions too quickly. As predicted, leaders in term-limited states do have less experience than leaders in states without term limits but only in the upper chamber. However, those who serve in the lower chamber have greater experience. This result is likely an indicator that in term-limited states, leaders are the most experienced members in the chamber. This analysis provides mixed results for the theory that term limits do result in leaders who have less experience, the implications of which are still unknown. The second expected change was a decline in the use of apprenticeships. In other words, there would no longer be a hierarchical order to attain a top leadership position and top leaders would not have served in any proceeding leadership positions. This theory, similar to the lack of experience, is derived from a of lack of time. With term limits, there simply would not be time to serve an apprenticeship. Contrary to previous research and existing theories, I find that term-limited states in several instances actually rely on apprenticeships more than states without term limits. Further, looking specifically at each leadership position, those who serve in term-limited states are more likely to serve an apprenticeship. This finding is strong and significant for chamber leaders in particular. Last, it was predicted that under term limits, leaders would serve in their position for a shortened period of time. The increased turnover caused by term limits meant leaders would be unable to serve in their position for an extended period of time. This theory, however, does not hold up to empirical testing. Surprisingly, there is no statistically significant difference in the length of position tenure between states with term limits and states without limits. The only exception is in the upper chamber of unified/lifetime limit states; this finding is amplified when examining the predicted tenure of the top legislative leader. Overall, there is not support for the theory that term limits reduce the tenure of legislative leaders. Collectively, these findings present inconsistencies in how leaders from term-limited states are viewed. Leaders from term-limited states have more experience in the lower chamber, less in the upper chamber. Those who serve under term limits are more likely to partake in the practice of apprenticeships than those who do not face term limits. Moreover, tenure does not vary significantly among the states. Those who face term limits do not have exceedingly different tenure experiences. Yet members from term-limited states do have significantly less experience. In fact, legislative

5

LEGISLATIVE LEADERSHIP

153

leaders in term-limited states have less than half the experience of those who do not face term limits. It is possible that term limits only changed the level of experience members are able to attain. These findings indicate that members are less likely to serve an apprenticeship out of a desire to assume leadership roles early in their career. While, leaders stay in those positions for the same amount of time as members in states without limits, the early rise to the position contributes to the declining experience. The interview responses from state legislators increase current understandings of the relationship that lawmakers have with their leaders. When legislators were asked who they think is influential in the legislature, answers varied among the states revealing just how much the importance of formal leadership varies by the legislature and chamber. In Nebraska, members do not tend to identify the speaker as an influential member. Rather, Nebraska senators tend to value experience and expertise over titles. In Oklahoma, power appears to include, but is not limited to, the majority party. Party leadership carries more power than the speaker alone. In Missouri, many members acknowledged the hierarchical structure that places the House speaker in charge. On the Senate side, the president pro tempore retains a great deal of power but members feel more equal. In Ohio, leadership remains powerful but this is rooted in the leaders’ experience. With the passage of term limits, scholars feared that leadership would lose too much power and their role would ultimately diminish. Now that term limits are at the equilibrium stage, it does not appear that the role of leaders has actually diminished. Rather, term-limited legislatures have worked to adapt to the lessened expertise by establishing a stronger hierarchy and putting the most experienced leaders in the top positions. Apprenticeships have become increasingly more prevalent in states with term limits. Members serve in those positions for the same length of time as members in states without term limits. The role of leaders has not diminished; if anything, leaders have become more powerful as less experienced individuals get elected. The exception to this is Nebraska. There, lawmakers do not appear to invest much power in their speaker, but this could be a product of being non-partisan and unicameral, and not because of term limits. Given the examinations in this chapter, leaders are just as influential as ever in states with term limits. If anything, newer, inexperienced members are more reliant on leadership. Take the case of Ohio; Speaker Householder was one of the most influential members in the legislature, in part,

154

J. BUTCHER

because of his prior service. More experienced members have the ability to exert greater authority over new inexperienced members. The speaker and president (pro tempore), as well as their top party leaders, hold a great deal of authority. In the cases of Oklahoma and Missouri, nothing happens in the legislature without the approval of the speaker or majority floor leader. It appears that term-limited states have leaders that have just as much control, or possibly more, than states without term limits.

Bibliography Bowser, Jennifer Drage, Rich Jones, Karl T. Kurtz, Nancy Rhyme, and Brian Weberg. 2003. The impact of term limits on legislative leadership. In The Test of Time: Lexington Books. Capell, Elizabeth A. 1996. The impact of term limits on the California legislature: An interest group perspective, 67–85. In Legislative term limits: Public choice perspectives. Springer. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 2000. Term limits in state legisla- tures. University of Michigan Press. Chaffey, Douglas Camp, and Malcolm E. Jewell. 1972. Selection and tenure of state legislative party leaders: A comparative analysis. The Journal of Politics 34 (4): 1278–1286. Clucas, Richard A. 2001. Principal-agent theory and the power of state house speakers. Legislative Studies Quarterly, pp. 319–338. Clucas, Richard A. 2007. Legislative professionalism and the power of state house leaders. State Politics & Policy Quarterly 7 (1): 1–19. Coyne, James K., and John H. Fund. 1992. Cleaning House: America’s campaign for term limits. Regnery Gateway. Freeman, Patricia K. 1995. A comparative analysis of speaker career patterns in US state legislatures. Legislative Studies Quarterly, pp. 365–376. Jewell, Malcolm E, and Marcia Lynn Whicker. 1994. Legislative Leadership in the American States. The University of Michigan Press. Kousser, Thad. 2005. Term limits and the dismantling of state legislative professionalism. Cambridge University Press. Kousser, Thad, and John Straayer. 2007. Budgets and the policy process. In Institutional change in American politics: The Case of Term Limits, ed. Bruce Cain Karl Kurtz, and Richard G. Niemi. Ann Arbor, MI: University of Michigan Press chapter 4, pp. 148–164. Little, Thomas H., and Rick Farmer. 2007. “Legislative leadership”. In Institutional Change in American Politics: The Case of Term Limits, ed. Bruce Cain Karl Kurtz, and Richard G. Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4: 55–72.

5

LEGISLATIVE LEADERSHIP

155

Long, Scott J. 1997. Regression models for categorical and limited dependent variables. Sage Publications. Maestas, Cherie. 2003. The incentive to listen: Progressive ambition, resources, and opinion monitoring among state legislators. Journal of Politics 65 (2): 439–456. Moncrief, Gary F. 1999. Recruitment and retention in US legislatures. Legislative Studies Quarterly, pp. 173–208. Peery, George, and Thomas H. Little. 2003. Views from the Bridge: Legislative leaders’ perceptions of institutional power in the stormy wake of term limits. In The Test of Time: Lexington Books. Powell, Lynda W., Richard G. Niemi, and Michael Smith. 2007. Constituent attention and interest representation. In Institutional change in American politics: The case of term limits, ed. Bruce Cain Karl Kurtz, and Richard G Niemi. Ann Arbor, MI: University of Michigan Press chapter 4, pp. 134–147. Reed, W Robert, and D Eric Schansberg. 1996. Impact of congressional tenure restriction on spending. In Legislative term limits: Public choice perspectives. Springer, pp. 101–116. Rosenthal, A. 2007. Living with term limits. In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Rosenthal, Alan. 1990. Governors and legislatures: Contending powers. CQ Press. Rosenthal, Alan. 1992. The effects of term limits on legislatures: A comment. In Limiting legislative terms, ed. Gerald Benjamin, and Michael J. Malbin. Congressional Quarterly Washington, DC, pp. 205–208. Sarbaugh-Thompson, Marjorie, and Lyke Thompson. 2017. Implementing term limits: The case of the Michigan Legislature. University of Michigan Press. Schlesinger, Joseph A. 1966. Ambition and politics: Political careers in the United States. Rand McNally and Co. Shay, Laine P. 2020. Do term limits “Limit” the speaker? Examining the effects of legislative term limits on state speaker power. State Politics & Policy Quarterly. Simon, Lucinda. 1987. The climb to leadership-career paths and personal choices. Journal of State Government 60 (6): 245–251. Squire, Peverill. 1988. Career opportunities and membership stability in legislatures. Leg- islative Studies Quarterly pp. 65–82. Squire, Peverill. 1992a. Changing state legislative leadership careers. In Changing patterns in state legislative careers, ed. Gary Moncrief, and Joel Thompson. The University of Michigan Press, pp. 175–194. Squire, Peverill. 1992b. Legislative professionalization and membership diversity in state legislatures. Legislative Studies Quarterly, pp. 69–79. Squire, Peverill, and Keith E Hamm. 2005. 101 chambers: Congress, state legislatures, and the future of legislative studies. Ohio State University Press.

PART III

The Implications of Term Limits

CHAPTER 6

Changes to Internal Deliberation: Committees

You cannot depend on some committee chairs because they too have their own agenda. —Missouri Representative (Republican) You need to rely on committees because members cannot know everything. —Missouri Representative (Democrat)

When it comes to policymaking in the states, legislative committees are frequently considered to be the central actor. Committees are the powerhouse where the work gets done, which often means many committees and numerous committee assignments for lawmakers. During the 2018 regular session the Missouri legislature had 85 committees this included standing, joint, and special committees. During this session, the 163 members of the Missouri House of Representatives introduced 1,705 bills. Given these numbers, members must sit on multiple committees, propose their own legislation, and deal with constituent needs all within a shorter time in office due to term limits. The question is, what are the effects on the ability to deliberate and pass legislation? Of the bills proposed, only 73 were “truly agreed and finally passed,” meaning they would become law. What happened to the other pieces of legislation? Of the 1,705 bills that were introduced, only 245 were sent on to the Senate meaning that most did not survive the committee process in the House. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_6

159

160

J. BUTCHER

Following the early predictions of term limits, this chapter takes an indepth look at how changes to legislative careerism have altered committee deliberation and policy outcomes. Committees are known for their expertise in policymaking and their importance to the legislative process has been well established (Hamm et al. 2011, 2006; Francis 1989; Francis and Riddlesperger 1982; Basehart 1980). Expertise, however, is dependent on both institutional and membership stability (Makse 2017; Hamm and Moncrief 1982; Basehart 1980). The introduction of term limits, which altered membership stability, has long been thought to affect committees. Under term limits, the collective limited experience of members dampens the capacity of the committee to produce well-considered legislation. Like Congress, states use committees to manage the workload and to make policies by performing three important functions (Cain and Wright 2007). First, committees generate information and form expertise in a particular area. Second, committees deliberate on legislation in an effort to get the best possible policy outcome. Finally, committees are gatekeepers, meant to keep poor legislation from reaching the floor. The potential changes to legislative committees would have broader implications than any other change. Under term limits, scholars feared that committees would struggle with deliberation and gatekeeping, while the legislature would pass record amounts of legislation. Have these fears come to fruition, or have committees learned to adjust to term limits with time? This chapter serves as a test of the changes to internal deliberation by examining the work of legislative committees. Namely, I evaluate the consequences of legislative careerism as it pertains to committees and whether lawmakers can maintain their legislative duties.

The Importance of Committees There has been limited work done on the intricacies and function of committees in state legislatures. What is known is that with time and growth in United States government, there has been an increased reliance on the committee system (Francis and Riddlesperger 1982). Committees not only serve as policy experts but are key deliberators who evaluate the substance, feasibility, amendments, and alternatives to legislation (Rosenthal 2007). Committees are the legislative decision-makers in most states and it is important to evaluate the long-term effects of term limits on this particular component of the legislative function.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

161

Committees are not only important for the institution, but they are important for individual members. Legislators spend at least half of their time doing committee work (Francis and Riddlesperger 1982). Individual members have their personal policy preferences and pursue particular committee assignments based on their personal experiences and occupational expertise (Makse 2017; Hamm et al. 2011; Francis 1989). Other members rely on committees for their level of expertise, which strengthens policymaking and benefits the entire legislature by being a source of information (Hamm et al. 2011; Basehart 1980). Committees can also be used by leadership for rewarding loyal members and enhancing member reelection prospects. The legislative committees that exist in the states differ from congressional committees in several key ways. State legislative committees typically do not follow the same seniority rules that are seen in Congress, instead, members are typically placed on committees by leadership (Squire and Moncrief 2019; Jewell and Whicker 1994). Each state has its own unique committee system and while there are many similarities the states implement different rules. One difference across the states is in the rules that influence party composition on committees. For example, in some states, the majority party is not required to consult with the minority party when making committee assignments. In other states, the majority party must accept minority party recommendations. States do not follow a universalistic rule, nor is there a singular party rule implemented across the states. In some states both parties have equal representation, in others, they do not. The rules regarding committee makeup are different for each legislative institution. The movement to modernize legislatures did more than increase the pay and number of staffers, it also helped to establish and organize committee systems in state legislatures (Rosenthal 1990). Legislatures became stronger organizations, but in order to compete with the governor there needed to be a place for legislators to become policy experts—this led to the creation of stronger committees. Committees are important to the legislature because of their three key duties. First, committees are the central place for policy deliberation and they work on legislation to achieve the best possible outcome. Second, one of the most important roles of the committee is to be a gatekeeper. Committees are supposed to keep poor legislation from making it out of committee. The last role incorporates the first two, committee members are meant to be

162

J. BUTCHER

information experts; they should know more about bills and policies that come before their committee than other legislators. The work that committees complete within these three capacities makes them of vital importance to the policymaking process that occurs in the legislature. Committee expertise, however, is unlikely to exist without membership stabilization (Basehart 1980). This membership stability influences the work that is done in the legislature Makse (2017), for example, found that membership stability and reorganization influence the number of bills that are passed. When members have greater expertise they are better equipped for committee work with a lower start-up cost. Careerist legislators, a result of increased professionalization, were serving on committees for long periods of time gaining knowledge and issue expertise. Turnover poses a major threat to the development and maintenance of committee capacity because members are not in office long enough to gain experience (Francis 1989; Basehart 1980; Rosenthal 1974). The problem is, turnover is the goal of term limits. Consequently, term limits threaten committee capacity.

Committee Changes Under Term Limits By increasing membership turnover, term limits alter how legislative committees function. Under term limits, initial research found that the three roles a committee is meant to fulfill (as suggested by Cain and Wright 2007) are strained. Committees are “less independent, respected, and active” and are no longer viewed as autonomous actors (Kousser 2005, 101). Internally, much of the committee power was diverted to the majority caucus, which removes bipartisan deliberation (Rosenthal 2007). Instead of being self-sufficient, committees were forced to rely heavily on outside actors for information. In anticipation of term limits, scholars speculated that the loss of institutional memory would negatively influence committees. Schrag (1996) observed that in California, committees were no longer producing objective reports, instead they were merely citing lobbyists’ interests. Meanwhile, the importance of committee assignments was enhanced because members knew they would only be in office for a short time and they would exploit their committee work to make a name for themselves (Carey et al. 2000). Yet new members came to the legislature not knowing the rules, processes, or how things worked (Carey et al.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

163

2000). The concern among scholars was that the increased number of new members would influence the way that committees handle business. In order to perform their committee duties, members must become information experts. Members tend to serve on committees that are related to their careers or training, but this is not always the case. Legislators need to know the ins and outs of the subjects under their committees’ jurisdiction. Term limits were argued to have negatively impacted the level of information that committee members enjoyed (Cain and Wright 2007). Without knowledge and experience, committee members struggle with the other roles that are expected of them—knowing the issues and mastering procedure (Cain and Wright 2007; Kousser and Straayer 2007; Powell et al. 2007). Given the importance of committees to the institutional structure of the legislature, there are two ways to evaluate their effectiveness. Committee work can be evaluated at the macro-level by evaluating policy output from the legislature as a whole. Or, committees can be assessed at the micro level by looking at the process and output of each individual committee. Each method gives insight into how committees perform in the legislative process. For this study, evaluating the role of committees at the macro-level gives insight into how committees work together to create legislation through deliberation and gatekeeping. A micro-analysis of the Ways and Means Committee and the Education Committee gives insight into how committees play their part in pushing through good legislation or cutting bad legislation. These two committees were chosen because of their importance, popularity, and their differences. These committees tend to lean in different directions politically, with Education more left and Ways and Means more right. Moreover, Ways and Means is traditionally a more prestigious committee but Education Committee work is important because of the direct ties to constituents in every state. The variation in the number of committees, as well as the policy area of each committee, makes it difficult to do a comparative analysis of each committee in each state. Looking at a subset of the Ways and Means and Education Committees provides some consistency for the purposes of evaluation. In the following three sections, I evaluate three key changes to legislative committees. First, there is a multi-part assessment of policy deliberation, including an examination of how long bills take to get through the legislative process and reach the governor’s desk. This follows the work of Kurtz, Cain, and Niemi (2007) and Sarbaugh-Thompson

164

J. BUTCHER

et al. (2004) that found legislation was rushed through the committee process after the implementation of term limits. Second, there is an overview of gatekeeping, which centers on the number of bills that are introduced, the number of bills passed in the legislature, and how many bills are ultimately signed by the governor. Similar to research done by Cain and Kousser (2004), where they examined the number of bills assigned and passed by committees. Last, I turn to legislator interviews to see how members view committees in the legislature and to what extent the value of committees has changed. This final assessment is much like the research conducted by Kurtz, Cain, and Niemi (2007) where the authors drew implications from broad interviews about legislative work. Collectively, these three analyses provide insight into how much, if at all, term limits have changed the process and power of committees. Have committees failed to do their job, or have they learned to adapt to term limits? Policy Deliberation Committees are supposed to use their expertise to deliberate legislation. Yet preliminary studies have found that with term limits there is less deliberation over legislation as bills are quickly passed through the committee process with fewer amendments (Cain and Wright 2007; Kousser and Straayer 2007). Term limits also reduce the capacity for complex legislation because of a lack of focus on long-term legislative goals (Baker and Hedge 2013; Kousser 2005). This hinders the ability of the committee to deliberate, but it does not exist to the same extent in all states (See Kousser 2005). Looking at the concerns of scholars and policymakers, this examination uses proxy measures to ascertain how long a committee deliberates over legislation and how thoroughly they deliberate. The following assessments evaluate the length of time that a bill sits in a committee and how many amendments are made to a bill once it makes it to the floor. To assess the length of deliberation there is a proxy measure for how long a bill was in the legislature beginning on the day that it was introduced. While there are a number of legislative effects that are not accounted for in this measure this does address the concern of scholars that legislation would move too quickly. Second, there is an assessment of how many amendments legislation incurs once it goes to the floor of the chamber. While amendments are not solely an indicator of deliberation, a high number of amendments can insinuate that the original draft of

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

165

the bill is lacking in some capacity (See also Cain and Kousser 2004). It should be the goal of a committee to successfully create legislation that members of the chamber would support, otherwise, it should not be reported from the committee. Admittedly, these measures of policy deliberation are imperfect and the measures themselves are quite crude. However, this assessment is a direct response to the fears and speculations that term-limited legislatures would pass legislation quickly (Berman 2007a; Cain et al. 2007). Aside from sorting through committee minutes for each subject, there are a few other ways to assess the extent to which policy was deliberated. Notably, there are a number of extraneous factors such as the length of time that lawmakers are in session, and partisan control over committees. Additionally, some states pass larger comprehensive bills, other states pass smaller more succinct pieces of legislation. In short, yes, there are potential alternative explanations but this proxy measure does illustrate variation among the states and addresses the concerns of term-limit opponents. Committees that send a great deal of legislation to the floor are not performing well (Rosenthal 1974, 28). Specifically, committees that are less adept at deliberation are more likely to send legislation, in its original form, to the floor. Quickly moving a bill to the floor implies that the committee has not deliberated the piece of legislation, applied their policy expertise to create better legislation, or thought through alternatives and amendments. The length of time that a policy is in committee is a sufficient proxy measure for the length of deliberation. Current theory holds that states with term limits are more likely to send bills directly to the floor because they do not have the time or capacity to deliberate legislation (see Carey et al. (2000) and Cain and Wright (2007)). With a finite amount of time in office, members prefer to pursue their own policy agenda rather than to deliberate others’ legislation. Examining the speed at which bills are rushed through the process gives insight into the deliberative process within the legislature. The changes to policy deliberation under term limits can be estimated using a Cox proportional hazard model. This model estimates the length of time a bill was in the legislature, from “introduction” to the final floor votes that send it to the governor. Once the bill leaves the legislature its time in the institution is over, regardless of what the governor does with the

166

J. BUTCHER

legislation.1 The estimates from the Cox model are the expected hazard ratios (Box-Steffensmeier and Jones 2004). When looking at policy deliberation, the “hazard” indicates the length of deliberation, a hazard of less than one indicates less risk (increasing the amount of time), while a hazard of greater than one means increased risk and therefore approaching the hazard more quickly (decreasing the amount of time). I test policy deliberation within the legislature with two separate analyses. The first analysis is focused on the entire legislative chamber, under the assumption that policy is assigned to committees and that committees are the actors who perform the bulk of legislative work. A secondary analysis focuses on two important committees across the states, Education and Ways and Means. Education Committees tend to be more prevalent among the states, whereas specific “Ways & Means” committees are less common. These committees both review a great deal of legislation, as well as endure a great deal of strife. This sample of two committees allows for greater insight into the different types of committees and to what extent they deliberate on policy. Table 6.1 contains the Cox estimates of policy deliberation for upper and lower chambers, a measure of how long bills are in the legislature. The hazard ratios indicate the length of time a bill is in action based on each type of term limit. The risk of hazard is higher in some term-limited states and lower in others. In divided/consecutive states there is a stark contrast between chambers. In the lower chamber, the rate of deliberation decreases by 24 percent. In the upper chamber, however, the hazard rate is increased by nearly 50 percent. The findings across chambers are more consistent in divided/life and unified/life states. In a divided/life state both chambers lower the hazard rate, nearly 75 percent in the lower chamber and 40 percent in the upper chamber. In a unified/life state, however, bills are retained 20 percent longer in the lower chamber and over 100 percent in the upper chamber. These results are similar to unified/consecutive term limits which increase the hazard by nearly 50 percent. Those states with an increased hazard have bills for less time and therefore spend less time deliberating. Only divided/life states have legislation for lengthier periods of time in both chambers. It is worth noting the larger institutional and partisan effects on deliberation as well. States that are highly professionalized appear to spend 1 There is the exception for the potential of a veto override vote, but this is a rare occurrence.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

167

Table 6.1 Cox estimates of policy deliberation for the lower and upper chambers

Divided/consecutive Divided/life

Lower chamber estimates

Upper chamber estimates

0.765 (0.0344) 0.259 (0.0100)

1.492 (0.0747) 0.576 (0.0222) 1.479 (0.0930) 2.894 (0.160) 0.421 (0.0155) 0.305 (0.0159) 0.00152 (0.000537) 51.98 (17.14) 14,781

Unified/consecutive Unified/life Republican legislature Divided government Professionalization Career opportunity N

1.301 (0.0524) 0.206 (0.00729) 0.455 (0.0204) 0.000208 (6.63e-05) 176.1 (52.91) 17,135

Robust standard errors in parentheses. Coefficients are expressed as hazard ratios

less time on legislation, indicating that less professional legislatures spend more time on legislation. This result holds across both chambers. Additionally, Republican-led chambers decrease the rate of turnover, by nearly 80 percent in the lower chamber. This result is of particular importance given that many of the term-limited states are Republican-dominate. While there is clearly some partisan aspect to this, the relationship does not appear to be driven by party alone. The results of the broader partisan and institutional factors are not closely aligned with the results for term-limited states, holding all things equal. Overall, the findings in Table 6.1 reveal that term limits do not inherently remove deliberation from the policymaking process as some suspected. In fact, in states with stricter limits, the bills are retained in the chamber for longer periods of time. What is most striking is the difference in the potential for deliberation in divided/consecutive states. The division between the two chambers may appear counter-intuitive but is likely a result of more experience in the upper chamber. In the lower chamber,

168

J. BUTCHER

where there are newer members legislation is deliberated longer. Meanwhile, in the upper chamber where there are more experienced members, less time is spent deliberating. The stark contrast that exists within a single type of term limit speaks to the importance of legislative experience in the policymaking process. The second way to assess policy deliberation is by looking at how long an individual committee deliberates legislation. The committees used in this assessment are the Education Committee and the Ways and Means Committee. As seen in the last analysis, the length of time that a committee has a bill indicates the level of deliberation. While this measure does not reveal how much time a committee actually spent on an individual bill, it does indicate how long the bill was with the committee. If a bill is with a committee for a maximum of one day, then it is unlikely the committee spent much time deliberating a bill. Even if a bill is a high priority, one day is simply not enough time to examine a piece of legislation in its entirety. When a bill is with a committee for 10 days or even 20 days it is more likely that the committee is performing its due diligence by analyzing the bill and speaking to relevant experts. The Cox hazard model for this assessment is centered on the length of time a bill was in committee, from the day it was “referred” to the committee to the day it was “reported” from the committee. Of course, not all bills are reported from the committee; the bills that die in committee are censored in the data. Table 6.2 contains the Cox estimates of policy deliberation in Education and Way and Means Committees, by chamber.2 The most interesting finding is the clear difference in deliberation between the two types of committees. For Education Committees term limits increase the rate of deliberation. In contrast, in Ways and Means Committees the rate of deliberation is decreased. In comparison with states without limits, all Education Committees deliberate longer while Ways and Means Committees push legislation through quickly. These results may be concerning because when legislation is moved quickly committees are not doing a good job deliberating legislation, then it is more likely that the chamber will later add amendments. Part 2 This particular analysis is limited to a subset of the states. There are six states used for the Ways and Means committee analysis and there are fourteen for the Education Committee. A table of the states and committees can be found in the appendix, Table B.5.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

169

Table 6.2 Cox estimates of policy deliberation within education and Ways and Means committees Education committee

Divided/ consecutive Divided/life Unified/life N

Ways and means committee

Lower chamber estimates

Upper chamber Lower chamber estimates estimates

Upper chamber estimates

0.167

0.294

4.390

(0.0254) 0.352 (0.0714) 0.269 (0.0396) 1,455

(0.0616) 0.148 (0.0316) 0.460 (0.0945) 1,551

(0.530) 3.966 (0.534)

2.202 (0.225)

979

400

Robust standard errors in parentheses. Coefficients are expressed as hazard ratios

of the job of a committee is to deliberate legislation and think of potential amendments and alternatives. It is necessary for the committee to work through additional amendments as they are viewed as policy experts. However, when committees are rushed or inexperienced they are less likely to deliberate on potential amendments and as a result, after the bill is referred to the floor it has amendments added by other members. Table 6.3 contains the results of a logit analysis for legislation that has amendments added by the chamber. In divided/consecutive states there are more amendments added to legislation. There are two potential explanations for this finding. This can be a result of poor deliberation because legislation moved quickly through the upper chamber. Or it can be a result of credit claiming, where each member tried to take part in legislation. In divided/life states there is a lack of distinction from states without term limits. Both types of unified limits are where the differences reside. In states with unified term limits it is significantly less likely that there will be amendments added to the legislation. Interestingly, both types of unified limits push legislation through at a faster rate, signifying less deliberation overall. Professionalization appears to be the most important legislative indicator for amendments. Those states that are the most professionalized, which tend to have more expertise, are significantly less likely to add amendments to legislation after it has been through the committee process.

170

J. BUTCHER

Table 6.3 Logit analysis of amended legislation, by chamber

Divided/consecutive Divided/life

Lower chamber

Upper chamber

1.019** (0.408) 0.074 (0.543)

0.749* (0.434) −0.108 (1.142) −2.313*** (0.574) −3.986*** (0.453) −0.902 (0.661) −0.339 (0.393) 0.130 (0.764) −6.748* (3.769) −1.307 (1.582) 0.564 (1.150) 109,028

Unified/consecutive Unified/life Republican legislature Republican governor Republican legislature X governor Professionalization Career opportunities Constant Observations

−1.057*** (0.407) −1.321** (0.531) −0.564 (0.410) 0.583 (0.648) −8.822*** (3.390) −0.457 (1.764) 0.629 (1.127) 161,354

Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1

This brief overview of added amendments helps to tell a larger story about the policy process in the states. The differences that exist between the types of term limits can be seen in Fig. 6.1, which represents the probability of amending a bill (including 95 percent confidence intervals). The unified term limits are unique and carry a great deal of the variation that has been seen in previous scholarship. These unified legislatures are more stringent and offer limited time to serve and build a reputation, while severely cutting experience. The lack of experience has resulted in struggles with legislative deliberation in ways that states with divided limits experience only minimally. This assessment of policy deliberation reveals that, much like the preceding chapters, the long-term effects depend on the type of term limit. In some instances, like unified term limits, the length of time that a bill takes to move through the legislature has decreased. The reduced

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

171

Upper Chamber

-.1

0

0

.05

Probability .1

Probability .1

.2

.15

.2

.3

Lower Chamber

NL

DC

DL

UL

NL

DC

DL

UC

UL

Fig. 6.1 Probability of amending a bill on the floor, by chamber (Note NL [No Limit], DC [Divided/Consecutive], DL [Divided/Life], UC [Unified/ Consecutive], UL [Unified/Life])

rate of deliberation is not absolute; there are some instances where termlimited legislatures are actually increasing the rate of deliberation but this is largely attributed to the lower chamber. While there is support for diminished deliberation in some term-limited states, and on Education Committees, there is more of a decrease in policy deliberation than would be expected. There was a fear that term limits would usher in a new era and legislators would not take the time to deliberate good policy. This analysis demonstrates a great deal of support for this theory for those states that have unified term limits. Gatekeeping Changes brought on by term limits affect policy expertise and deliberation; as a result, committees are less equipped to maintain their

172

J. BUTCHER

gatekeeping responsibility. In tandem with the lack of deliberation, prior research argues that committees no longer serve as gatekeepers because they do not have the necessary expertise. There are two signs that can be taken to indicate a less capable committee. First, weaker committees report a higher number of bills (Rosenthal 1974, 24). A higher number of bills indicate that members are not selective about legislation. Second, a less capable committee is pushing out bills that either fails or is edited moving forward. A study from Kurtz et al. (2007, 82), found that after term limits the rate of bills dying in the legislature declined by 13 percent for the lower chamber and 5 percent for the upper. In addition to Rosenthal’s (1974a) assessment of committee workloads there were a number of scholars who feared the role of “gatekeeping” or keeping bad policy out would no longer be a priority. There was a fear that as a result of bills passing through committees quickly; committee policy recommendations would not be taken as seriously on the floor in term-limited states (Berman 2007b; Cain and Wright 2007; Powell 2007). This assessment again relies on proxy measures as gatekeeping is difficult to measure, because bills die in the legislative process for many reasons. There are two components to this gatekeeping analysis. First, there is an assessment of how term-limited states gatekeep relative to states without term limits. This assessment looks at the number of bills that are introduced in the legislature and the number of bills that are ultimately signed by the governor. In order to separate out partisan effects there are controls for divided government and for a Republican legislature. The second assessment looks at the role of committees in gatekeeping by evaluating the number of bills that are killed. Gatekeeping, of course, can occur elsewhere in the legislative process, not just within a committee. This first assessment gives an overview of the legislature’s broad gatekeeping authority. Table 6.4 contains the results of a negative binomial regression comparing states with term limits to states without term limits. This analysis looks first at the number of bills introduced in order to understand how states differ in the frequency of bills as a starting point. There are two key gatekeeping functions included, the number of bills passed by the legislature and the number of bills signed by the governor. The data in this analysis have been aggregated to the state level by session year for each chamber. Each observation is the number of bills in a chamber for each session, by state. In order to

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

173

account for some of the state by state variation, where specific institutional norms might affect committee work, some of the models below include state-fixed effects. Table 6.4 reports the change in log-likelihood for each variable. For example, the coefficient for divided/consecutive legislatures is −0.936 and is statistically significant. This is interpreted as, legislatures that go from no term limits to divided/consecutive limits the expected log count of bills proposed in the lower chamber decreases by 0.936. Similarly, each state with term limits has a decrease in the expected log count of bill proposals in both chambers and this is statically significant for each chamber except the upper chamber of unified/life states. Overall, term-limited legislatures tend to introduce fewer pieces of legislation than non-term-limited states. This initial piece is important to our understanding of gatekeeping and the frequency of bills that make it to the governor’s desk. Focusing on columns 2 and 4, the added state-fixed effects appear to strengthen the relationship of the base model. When accounting for the differences that exist between the states, the results not only hold, but the effects are stronger. That is, with one exception, States with divided/consecutive term limits actually see a weakened relationship when accounting for each individual state. So much so that bill passage in the upper chamber is no longer significant. However, unified/life term limits, previously lacking significance now see a notable decline in bills passing in the upper chamber. When it comes to the legislature actually passing legislation, passed bills, the initial trends deviate. Counter to the strong, significant results described above there are no significant trends in the base model for bills passed by both chambers. Instead, the relationship appears to be primarily dependent on institutional effects such as, professionalization, and career opportunities, which are quite weak. On the one hand, it is surprising that the individual chamber results do not hold up when assessing the entire legislature. On the other hand, it ought to be encouraging that there are no significant differences between the states with and without term limits when it comes to total legislative productivity. However, when state-fixed effects are added to the model this no longer holds. When the fixed effects are included in the model, divided/consecutive states pass far more legislation than their counterparts in others states, but members from all other term-limited states pass significantly less legislation.

Robust standard errors by state

log(α)

Constant

Career opportunities

Republican legislature X Republican governor Professionalization

Republican governor

Republican legislature

Unified/life

Unified/consecutive

Divided/life

−1.723*** (0.326) −0.066 (0.356) −0.261 (0.245) 0.290 (0.413) 16.977*** (1.067) −9.863*** (1.441) 6.935*** (0.452) −1.775*** (0.140)

−0.482*** (0.168) −2.113*** (0.307)

−0.936** (0.394) −1.756*** (0.540)

−0.934* (0.499) 0.618* (0.321) 0.672* (0.348) −1.446*** (0.531) 11.128*** (3.559) −5.512** (2.178) 7.082*** (0.704) −0.791*** (0.254) *

Lower chamber

Lower chamber −0.730** (0.348) −1.840*** (0.472) −1.010* (0.569) −0.419 (0.428) 0.575* (0.323) 0.972*** (0.215) −1.730*** (0.399) 10.440*** (3.158) −4.630** (1.833) 6.389*** (0.599) −0.955*** (0.240) *

Upper chamber −0.198 (0.170) −2.086*** (0.319) −1.943*** (0.436) −1.026*** (0.329) 0.225 (0.370) 0.297 (0.246) −0.532 (0.403) 16.009*** (1.078) −7.838*** (1.456) 5.932*** (0.467) −1.758*** (0.138)

Upper chamber 2.068 (1.503) −2.240 (1.765) −2.232 (1.830) −0.721 (0.977) 0.903 (0.989) −1.312 (1.420) 0.004 (2.316) 18.130* (10.395) −15.411* (9.202) 4.849*** (1.305) 1.036*** (0.320) *

Passed bills 3.445*** (0.428) −4.559*** (0.916) −5.350*** (1.007) −2.480*** (0.668) −1.526 (1.042) −0.524 (0.717) 0.850 (0.923) 29.865*** (2.543) −28.892*** (3.357) 6.333*** (1.168) 0.325*** (0.109)

Passed bills

Negative binomial regression models of legislative gatekeeping, count of bills passed

Divided/consecutive

Table 6.4

4.281 (9.547) −7.784*** (2.091) −31.730*** (6.695) −27.926*** (3.054) −0.200 (11.372) −8.210 (16.438) 3.600 (24.794) 32.456 (49.277) −37.870 (38.500) 9.529 (20.634) 2.652*** (0.477) *

Bills signed by the governor

174 J. BUTCHER

99

***p < 0.01, **p < 0.05, *p < 0.1

State-fixed effects Observations

Lower chamber * 99

Lower chamber

104

Upper chamber * 104

Upper chamber

203

Passed bills * 203

Passed bills

203

Bills signed by the governor

6 CHANGES TO INTERNAL DELIBERATION: COMMITTEES

175

176

J. BUTCHER

What is perhaps more interesting are the control variables accounting for some of the variation among legislatures. Having both a Republican governor and a Republican legislature does have a significant effect on bill introductions, but not when accounting for each individual state. Professionalization appears to be one of the most important factors affecting both legislative introductions and the rate of passage in the legislature this finding is consistent across all legislative models. In contrast, there is no statistically significant relationship between professionalization and bills signed by the governor. In both chambers of the legislature, the expected log count of bill introductions is positive and statistically significant for those legislatures that are more professional. This finding remains true for bills passed by the legislature, with an expected log count increase of 18, and 29 for the fixed effects model. One possible sign of gatekeeping error is when a unified government (both legislative chambers and the governor under the control of one party) has a governor who does not sign the legislation. This sample size spans nearly a decade of policy and multiple governors in order to mitigate any potential gubernatorial effects. In divided/life states the expected log count of bills signed by the governor decreases by eight. In unified/ consecutive and unified/life the expected log count of bills signed by the governor dramatically decreases by 31 and 27, respectively. The high rate of bills being denied by the governor, out of the already limited number of bills, indicates that the legislature may not be doing its job of gatekeeping. Instead, the task of keeping out poor legislation is being relegated to the governor. In order to get a better understanding of this the following analysis, I revisit the Education and Ways and Means Committees to test their gatekeeping abilities, by assessing whether committees kill legislation or simply pass it through. The amount of bills that are shuffled through or are killed in committee is not time-dependent; a logit analysis is sufficient to reveal the differences among the states. For this analysis, each unit is an individual bill in either the Education or Ways and Means Committee. Table 6.5 contains the results of a logit analysis for bills that die in committee. Due to the constraints of the self-reported data, there are gaps but this analysis does provide insight into the gatekeeping function of two key committees. Given the limited number of states included in this analysis state-fixed effects are not accounted for. Looking at Table 6.5, there are clear differences between states with term limits and those without, but the level of significance varies. In the

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

177

Table 6.5 Logit analysis of whether bills die in committee, by chamber Education lower chamber Divided/ consecutive Divided/life

−1.715** (0.720) −0.071 (0.788)

Unified/ consecutive Unified/life Republican legislature Republican governor

Education upper chamber −1.418** (0.666) −1.183* (0.712) 0.546

Ways and means Ways and means lower chamber upper chamber −1.395*** (0.040) 1.599*** (0.069)

−1.727*** (0.000)

−2.077*** (0.423) 0.316

(0.697) −1.197*** (0.432) 0.845

1.973***

0.184***

(0.825) 2.235***

(0.870) 2.315***

(0.072) 3.471***

(0.000)

(0.736) −4.228***

(0.070) −4.609***

(0.963)

(0.074)

−0.486 (7.366) 0.345

−16.887*** (0.390) −0.129

(3.285) 2.064** (0.856) 10,785

(0.288) 4.783*** (0.048) 7,572

(0.710) Republican −3.355*** legislature X Republican (0.937) governor Professionalization 0.559 (8.929) Career −1.297 opportunities (3.816) Constant 2.695*** (0.972) Observations 14,020

1.703*** (0.000) 2,054

Robust standard errors by state in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

committees present in divided/consecutive states, there are significantly fewer bills that die in committee. For divided/life states, fewer bills die in the committees in the upper chamber. In the lower chamber, however, education bills do not die at a different rate than in states without term limits. Meanwhile, the Ways and Means Committee is significantly more likely to kill pieces of legislation, and this is the only committee that is more likely to do so. Aside from the differences between states with and without term limits, there are unique differences based on party. There is a significant effect of having both a Republican legislature and governor;

178

J. BUTCHER

this finding is evident in each model. The assessment of bills that die in committee reveals that when there is a difference between states with and without term limits; term-limited legislatures are less likely to gatekeep. It is worth noting the importance of legislative professionalization in this particular analysis. Contrary to nearly every other analysis in this work, professionalization is not a significant factor in the work that is performed on state Education Committees. This is likely due to the salience over the importance of education policy because education policy is an important component of each state; the work gets done regardless of broader institutional resources. This, however, is not at all the case for Ways and Means committees where professionalization has a significant outcome on legislation. In highly professionalized legislatures, where members often have more time and support, bills are significantly less likely to die in committee. Figure 6.2 illustrates the predicted probability of a bill dying in the Education Committee for each type of term limit. Again, there is little variation here between those states with term limits and those without. This figure displays the percent chance of a bill dying in committee for each term-limit type (all other values are held at their mean). In the lower chamber, the probability of a bill dying is not statistically different among the four types of states. In the upper chamber, there is greater variation but the probability of a bill dying is not statistically different among each of the categorizations. This analysis of gatekeeping reveals an interesting dichotomy present in term-limited legislatures. Counter to what scholars initially feared, there is no evidence that a term-limited legislature introduces more legislation than a state without limits. Yet, committees are less likely to kill bills. In fact, both legislative chambers introduce and kill significantly fewer bills than in states without term limits. Interestingly, this carried over to fewer passed bills by the legislature and fewer governor approvals for states with divided/lifetime and unified/consecutive term limits. Divided/consecutive term limits continue to see different trends than the other states with more bills being passed by the legislature but no difference in the rate at which the governor approves bills.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

Upper Chamber

0

0

.2

.5

.4

Probability

Probability .6

1

.8

1

1.5

Lower Chamber

179

NL

DC

DL

UL

NL

DC

DL

UC

UL

Fig. 6.2 Probability of a bill dying in an Education Committee, by chamber (Note NL [No Limit], DC [Divided/Consecutive], DL [Divided/Life], UC [Unified/Consecutive], UL [Unified/Life])

Committees in the States In order to understand committees and how they have adapted to term limits, it is important to understand how members view committees. This includes the legislation that is pursued by committees, the deliberation and gatekeeping that occurs, and whether or not members trust the legislation pushed through committees. The preceding analyses provide a great deal of insight into the legislative process by evaluating the work of committees and the rate of bill passage. However, these studies do not indicate any level of importance to the members. Legislative interviews allow for an in-depth examination of how lawmakers view committees and remove some of the guesswork and inferences made when using large-N analyses. Reliance on information is key, do members trust what their

180

J. BUTCHER

own committees do? In this section, I use legislator interviews to answer whether or not lawmakers value the work of committees. There are three questions that I use to gauge the value of committees. Respondents were asked to answer these questions based on their personal experiences of serving on committees.3 In order to capture the committee dynamics, lawmakers were first asked, “How much control would you say the chair has over the work of the committee?” Also meant to capture the dynamics of the committee members were asked, “How much conflict is there on the committee?” Responses range on a scale of 1-none, 2little, 3-some, 4-a lot, and 5-most.4 Last, in order to gauge the value of the committee, members were asked about what they would do if they were facing an issue and needed further information, specifically they were asked if a bill was coming out of a particular committee, “how much do you rely on the committee chair for information?” When asked about committees it was not uncommon for the respondents to make additional comments, this open-ended approach allows for further examination of the role of committees under term limits. Table 6.6 contains an ordered logit analysis of the interview questions for Ohio, Missouri, Oklahoma, and Nebraska. Each model represents one of the three questions gauging the amount of control a chair has over the committee, the amount of conflict on a committee, and how much members rely on a committee for information. Each state cites high levels of committee chair control over the work of the committee (in comparison with Nebraska). Respondents from Ohio and Missouri report higher levels of conflict on the committees that they serve on. The divided term limits present in Ohio and Missouri can make it more difficult for members to form bonds and the shorter; divided limits cut down on the time members have in office both of these factors can contribute to a less collegial atmosphere. What is perhaps most interesting is that each state is significantly less likely to rely on the committee chair for information when making policy decisions. While respondents agree that

3 The breakdown of responses for each of these questions can be found in the appendix, Table B.4. 4 Respondents were shown a scale and asked to rank on that scale; members who did not know were recorded as a zero. In the few instances where members answered between two numbers (i.e. 3.5), answers were rounded down because of their uncertainty that the number was any higher.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

181

committee chairs tend to have greater control over committee proceedings, with little conflict, members do not actually count the information as reliable. In order to better understand how serving under different types of term limits can lead respondents to view committees differently, I present an overview of how responses to each of these questions vary based on the level of experience that a member has. These responses are based on White, Republican men from each state, and vary by the number of years of experience. It is important to keep in mind that of these states only in Ohio and Nebraska are members able to gain more than 12 years of experience, and even though it is possible in Nebraska it is rare. Whereas in Missouri members are limited to 8 years in each chamber and Oklahomans are limited to 12 years total. Given the limitations on service, Table 6.6 Ordered logit analyses of committees, scaled response

Ohio (Divided/consecutive) Missouri (Divided/life) Oklahoma (Unified/life) Legislative experience Republican White Male τ1 τ2 τ3

Committee control

Committee conflict

Committee reliance

0.482** (0.188) 1.400*** (0.164) 1.059*** (0.228) −0.049* (0.027) −0.535** (0.226) −0.999*** (0.370) 0.505 (0.635) −4.217*** (0.979) −1.931*** (0.396) 0.443 (0.606)

0.327* (0.191) 1.341*** (0.079) −0.065 (0.064) 0.075** (0.032) −0.426 (0.784) 0.553 (0.457) −1.165 (1.235) −1.946*** (0.594) 0.539 (0.430) 3.331*** (0.798) 5.032*** (0.646) 84

−1.206*** (0.161) −0.851*** (0.044) −0.701*** (0.112) 0.028 (0.044) 1.234*** (0.345) −0.310 (0.615) −0.508* (0.260) −3.003*** (0.485) −1.205*** (0.415) 0.618* (0.325)

τ4 Observations

98

Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1

87

182

J. BUTCHER

it is important to direct attention to the changes in predicted responses between 0 and 8 years. One indicator of committee performance is how centralized the committee is and whether or not the committee chair is able to maintain control. In the preceding analysis, respondents from each state cited committee chairs as having a great deal of control. Figure 6.3 breaks down the predicted response to the question of committee chair control by the state and across the legislative experience. There are several trends among the states. First, in no instance did members report committee chairs as having no control, excluding this from the model. In each state the probability of a member reporting that a chair has “most” control declines with experience, this decline occurs with 2 years of experience in Missouri and 8 years of experience in Ohio. The probability of those who report committee chairs as having “little” control over the committee does not vary by experience; this is largely consistent across the states. In other words, those who do not believe the committee chair has little control are unlikely to be persuaded otherwise. Conversely, the probability of a member responding that a chair has “some” or “a lot” of control is likely to increase with legislative experience. While this finding is present in each state, it is less prominent than the decline seen in the probability of responding “most.” Figure 6.4 illustrates the predicted responses of a member citing committee conflict and how it varies by level of experience. The probability of responding that a committee has “a lot” or “most” conflict does not vary based on experience and remains close to zero for each state. In three of the states, the response to committee conflict is predominately “little” and this does not vary by experience. In Missouri, however, the probability of a respondent answering “little” declines after 2 years of experience, at which point “some” increases. In Missouri, there is a low probability of members citing no conflict. In the remaining states “none” is a higher probability for those who lack experience, but at around 2 years in Ohio and 6 in Oklahoma and Nebraska, there is a shift, and the probability of a legislator responding “none” declines. While the probability of responding that committees have “most” or “a lot” of conflict is unlikely, this figure does illustrate that as members become more experienced they do cite greater conflict than when they first arrived. The last question that helps to gauge committees and their importance is, how much do members rely on the committee for making decisions? Figure 6.5 presents the predicted responses to this question for each

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

183

0

Committee Control 0 .1 .2 .3 .4 .5

.1 .2 .3 .4 .5

6

0

2

4

6

8

10 12 14 16 18

0

2

4

6

8

10 12 14 16 18

Divided/Consecutive, OH 0 .1 .2 .3 .4 .5

Committee Control 0 .1 .2 .3 .4 .5

Divided/Life, MO

0

2

4 6 8 10 12 14 16 18 Prior Experience (years)

Little

Some

0

2

4 6 8 10 12 14 16 18 Prior Experience (years)

A lot

Most

Fig. 6.3 Probability of chair committee control responses by experience across states (Note The lack of responses to the “none” category result in zero predictions for that category)

state based on legislative experience. In comparison with the predictions for the other two questions, this varies the most among the states. The most consistent result among the states is the probability of a member responding “none;” the probability is minimal and consistent across each state. For three states members are most likely to respond “some” and this will vary little based on experience. While the analysis in table 6.6 indicates a decline in reliance on committees for information, there is a clear deviation among the states for those who respond “little” and “a lot.” In Oklahoma, as members gain more experience the likelihood of responding “a lot” increases while responding “little” decreases. Missouri sees a similar trend to Oklahoma, as members gain experience the probability of responding to relying on information “a lot” increases. Ohio sees this trend as well but sees a more dramatic shift as members with less experience were more likely to respond with “little” reliance on committees.

.6

J. BUTCHER

0

0

.2

.2

.4

.4

.6

184

0

2

4

6

8

0

10 12 14 16 18

2

4

6

8

10 12 14 16 18

Divided/Consecutive, OH

0

0

.2

.2

.4

.4

.6

.6

Divided/Life, MO

0

2

4 6 8 10 12 14 16 18 Prior Experience (years)

None

Little

0

Some

2

4 6 8 10 12 14 16 18 Prior Experience (years)

A lot

Most

Fig. 6.4 Probability of committee conflict responses by experience across the states

However, the shift seen in Ohio does not occur until members have nearly 10 years of experience, which means they have served in more than one chamber. Nebraska is unique in that there is little variation in predicted responses as members become more experienced. Even though each form of term limit is unique and there are clear differences when it comes to deliberation and gatekeeping, there are trends that separate the term-limited states from states without limits. Rooted in the inexperience and constant variability of the legislature caused by consistently adding large freshman classes there is a lack of experience that shifts committee work. Namely, among states with term limits the same legislation tends to be reintroduced frequently because it is passed on by members who are leaving. In Missouri, a divided/lifetime state, one member said that legislators keep repeating the same arguments because people have not been around long enough to remember. A

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

185

0

.2

.4

Committee Reliance 0 .1 .2 .3 .4

.6

6

0

2

4

6

8

0

10 12 14 16 18

2

4

6

8

10 12 14 16 18

Divided/Consecutive, OH

.1

.2

.3

.4

Committee Reliance 0 .1 .2 .3 .4

.5

Divided/Life, MO

0

2

4 6 8 10 12 14 16 18 Prior Experience (years)

None

Little

0

2

4 6 8 10 12 14 16 18 Prior Experience (years)

Some

A lot

Fig. 6.5 Probability of committee reliance responses by experience across the states (Note The lack of responses to the “most” category result in zero predictions for that category)

legislator from Ohio remarked, “it’s like groundhog day, [the legislature] repeats itself with legislation and new members.” A second trend among the term-limited states is that even though members note term limits increasing partisanship; few legislators cite the occurrence of conflict on committees. When conflict does occur it is on a select few committees, this lack of conflict can lead to a lack of deliberation or drive to get the best policy. As several legislators noted, there is little conflict and things pass through the committee somewhat easily. An Oklahoma legislator said that in committees, “most votes are unanimous but a few cause conflict and splits.” A Nebraska senator remarked, “it’s a little too easy to get things done in a unicameral, with only one house.” In some instances, because committee chairs have the most experience, members from outside the committee look to the opinion of the committee chair and how they vote on legislation.

186

J. BUTCHER

A third trend among the term-limited states is how the lack of deliberation leads to rushed or biased legislation. A senior legislator from Missouri warned, “people, are passing legislation and not saying anything because they will not be around for the outcome.” The same legislator stated, “you cannot depend on committee chairs because they too have their own agendas.” A fellow Missourian said, “I rarely see long-term decisionmaking focus.” A senator from Nebraska expressed their shock at “how little time there is to do the discernment you actually want to do.” A legislator from Oklahoma said that bills pass through committee “because of who the author is, not because of what the bill is.” Legislators attest that because they are short on time; they simply pass the legislation they are presented with. Another member from Missouri shared that some “unintended consequences (of term limits) are that members in their last term are running dicey legislation because there are no consequences [to them politically].” The shifting dynamics of legislative careers under term limits not only alter the ability of committees to do their work but the perceptions of committees are changed as well. Legislators are increasingly going outside of legislative committees for information. For example, a Missouri legislator said: “committees are important for anecdotal information... the important technical information comes from lobbyists and experts because they give reports on both sides and do their own research.” One member from Nebraska said that chairs are challenged for a lack of institutional knowledge and members must turn to other sources. A different Nebraska senator noted that it is actually staff who runs committees given that they have more experience. A long-serving member from Ohio, who served before and after term limits, noted some important changes that he has seen: “There use to be more confidence in committee work and you could rely on and feel comfortable with the process more so than you can today with term limits and inexperience on both sides of the aisle and in leadership.” In an attempt to compensate for the lack of expertise present in many state legislatures committees have become even more centralized under leadership but only when there is the ability to do so. In some legislatures, leadership has been able to usurp committee autonomy and micromanage committees. This was noted by a number of lawmakers from Ohio. This trend is more likely to occur in a state with a divided/ consecutive term limit, like Ohio, because members can gain more political experience. Several members noted that the control over committees does vary by the speaker but that under term limits committee chairs are no longer granted “free-reign” because there are no longer veteran

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

187

committee chairs. Even in the less experienced and less professionalized states a lot of what happens in reference to legislation first goes through leadership and members tend to only rely on the advice from the most experienced members.

Summary There are many existing theories about term limits and their influence on legislative committees. Kousser (2005, 117) argued that under term limits committees are “less independent, respected, and active.” Committees were thought to defer to the executive branch more frequently (Cain and Wright 2007; Powell et al. 2007). Committees are not as independent because they are not capable of developing the necessary expertise to assert themselves. Members no longer have the skills necessary to make policy, which has made the legislature a weaker branch of government (Sarbaugh-Thompson et al. 2010; Bowser and Moncrief 2007; Powell 2007; Cain 1996). In this chapter, I test theories about legislative committees in termlimited states in three different ways. First, there is an evaluation of policy deliberation by examining the number of bills sent directly to the floor and by assessing how long committees hold onto bills. Contrary to what many scholars anticipated, committees in the lower chamber actually deliberated legislation for longer in states with term limits. In the upper chamber, however, three of the four types of term limits pushed legislation through at an accelerated rate. In addition to this analysis, I find that it is more likely for a divided/consecutive legislature to add amendments to legislation once it reaches the floor, indicating a lack of deliberation within the committee. States with unified term limits see fewer amendments added to legislation and have shorter deliberation times. Second, I assess the gatekeeping role of committees and the number of bills that are killed in committees. Specifically, this section examines the passage rate of bills and whether or not bills have subsequently changed on the legislative floor. Gatekeeping is the most noticeable difference between states with and without term limits. Unified term-limit states struggle with gatekeeping; fewer bills die in committee and the governor becomes the ultimate gatekeeper. Last, I utilize legislator interviews to uncover the dynamics of committees and how they matter to legislators. Specifically, this analysis is centered on the control of the committee chair over the committee, how

188

J. BUTCHER

much conflict exists within committees, and how much legislators rely on committees for information. Overall, committee chairs have a great deal of control over their committees. Yet, committee conflict varies by state and by how much experience a lawmaker has. Of most interest is the variation in how much members rely on committees for information. While members across the states consistently rely on committees for at least minimal information, committees are not the primary place for legislators to gather information for legislation. As a result, members must look beyond committees and even beyond the walls of the legislature in order to gain information to make legislative decisions. One fear brought by term limits was that each legislator would propose more legislation in order to build up his or her legacy. As a result, committees would be overwhelmed and would ease up on their responsibilities to engage in deliberation and gatekeeping. This test finds mixed support for this claim; term-limited legislatures do not propose more legislation. In fact, there are no instances of a term-limited legislature introducing significantly more legislation, but there are multiple instances of fewer pieces of legislation in each session. What is of note is that term-limited legislatures are more likely to have a bill vetoed by the governor. While divided/consecutive states have, in many ways, been able to adapt to the changes brought by term limits they still differ from states without term limits when it comes to committee deliberation and gatekeeping. The curious component is how different the states with unified limits are from their non-term-limited counterparts. These unified states have far fewer pieces of legislation, fewer bills die in committee, and there are fewer legislative amendments to legislation. In sum, unified states have fewer legislative proposals and they spend less time deliberating. Given the insight from legislators above, the differences are likely attributed to the differences in overall legislative experience and the members who serve in leadership positions and as committee chairs. This examination of committees reveals an interesting story about term limits and un-covers some of the differences among the states. While there are fewer bills introduced in each legislative chamber with term limits, there are also fewer bills killed, committees spend less time deliberating, and in some states, more bills are amended on the floor. Even in the instances where there appears to be little difference between term limits and states without term limits, the interview responses should be cause for concern. Members are openly less willing to accept the information

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

189

produced by committees and point out the flaws in the limited time to deliberate, which results in pushing rushed legislation.

Bibliography Baker, Travis J., and David M. Hedge. 2013. “Term limits and legislativeexecutive conflict in the American States.” Legislative Studies Quarterly 38 (2): 237–258. Basehart, Hubert Harry. 1980. “The effect of membership stability on continuity and experience in US State legislative committees.” Legislative Studies Quarterly pp. 55–68. Berman, D. 2007a. “The effects of legislative term limits in Arizona: More churning, more chaos, and a diminished institutional role for legislators.” In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Berman, David. 2007b. “Legislative climate.” In Institutional change in American politics: The case of term limits, edited by Bruce Cain Karl Kurtz, and Richard G Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4, pp. 107–118. Bowser, Jennifer Drage, and Gary Moncrief. 2007. “Term limits in state legislatures.” In Institutional change in American politics: The case of term limits, edited by Bruce Cain Karl Kurtz, and Richard G Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4, pp. 10–21. Box-Steffensmeier, Janet M., and Bradford S. Jones. 2004. Event history modmodel: A guide for social scientists. Cambridge University Press. Cain, Bruce E. 1996. “The varying impact of legislative term limits.” Springer Chapter 1: 21–36. Bruce E. Cain, and Thad Kousser. 2004. Adapting to term limits: Recent experiences and new directions. Public Policy Institute of California. Cain, Bruce, and Gerald Wright. 2007. “Committees.” In Institutional Change in American Politics: The Case of Term Limits, edited by Bruce Cain Karl Kurtz, and Richard G. Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4, pp. 73–89. Cain, Bruce, Thad Kousser, and Karl T. Kurtz. 2007. “Institutional imbalance: The effect of six-year limits in California.” In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Carey, John M., Richard G. Niemi, and Lynda W. Powell. 2000. Term limits in state legislatures. University of Michigan Press. Francis, Wayne L. 1989. The Legislative committee game: A comparative analysis of fifty states. Ohio State University Press.

190

J. BUTCHER

Francis, Wayne L., and James W Riddlesperger. 1982. “US State legislative committees: Structure, procedural efficiency, and party control.” Legislative Studies Quarterly pp. 453– 471. Hamm, Keith E., and Gary Moncrief. 1982. “Effects of structural change in legislative committee systems on their performance in US States.” Legislative Studies Quarterly pp. 383–399. Hamm, Keith E., Ronald D. Hedlund, and Nancy Martorano. 2006. “Measuring state legislative committee power: Change and chamber differences in the 20th Century.” State Politics & Policy Quarterly 6 (1): 88–111. Hamm, Keith E., Ronald D. Hedlund, and Stephanie S. Post. 2011. “Committee specialization in U.S. State legislatures during the 20th Century: Do Legislatures tap the talents of their members?” State Politics & Policy Quarterly 11 (3): 299–324. Jewell, Malcolm E and Marcia Lynn Whicker. 1994. Legislative leadership in the American States. The University of Michigan Press. Kousser, Thad. 2005. Term limits and the dismantling of state legislative professionalism. Cambridge University Press. Kousser, Thad, and John Straayer. 2007. “Budgets and the policy process.” In Institutional change in American politics: The case of term limits, edited by Bruce Cain Karl Kurtz, and Richard G. Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4, pp. 148–164. Kurtz, Karl T., Bruce E. Cain, and Richard G. Niemi. 2007. Institutional change in American politics: The case of term limits. University of Michigan Press. Makse, Todd. 2017. “The retention of expertise and productivity in state legislative committees.” State Politics & Policy Quarterly 17 (4): 418–440. Powell, Lynda W., Richard G. Niemi, and Michael Smith. 2007. “Constituent attention and interest representation.” In Institutional change in American politics: The case of term limits, edited by Bruce Cain Karl Kurtz, and Richard G Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4, pp. 134– 147. Powell, Richard J. 2007. “Executive-legislative relations.” In Institutional change in American politics: The case of term limits, edited by Bruce Cain Karl Kurtz, and Richard G Niemi. Ann Arbor, MI: University of Michigan Press Chapter 4, pp. 134–147. Rosenthal, A. 2007. “Living with term limits.” In Legislating without experience: Case studies in state legislative term limits. Lexington Books. Rosenthal, Alan. 1974. Legislative performance in the states; Explorations of committee behavior. The Free Press. Rosenthal, Alan. 1990. Governors and legislatures: Contending powers. CQ Press.

6

CHANGES TO INTERNAL DELIBERATION: COMMITTEES

191

Sarbaugh-Thompson, Marjorie, John Strate, Kelly LeRoux, Richard C. Elling, Lyke Thompson, and Charles D. Elder. 2010. “Legislators and administrators: Complex relationships complicated by term limits.” Legislative Studies Quarterly 35 (1): 57–89. Sarbaugh-Thompson, Marjorie, Lyke Thompson, Charles Elder, John Strate, and Richard Elling. 2004. The political and institutional effects of term limits. Springer. Schrag, Peter. 1996. “The populist road to hell: Term limits in California.” American Prospect 24: 24–30. Squire, Peverill, and Gary Moncrief. 2019. State legislatures today: Politics under the domes. Rowman & Littlefield.

CHAPTER 7

Changes to External Relations: State Agencies

Agencies sometimes need to be taken with a grain of salt. —Oklahoma Representative (Republican)

In 2018, several states, including Oklahoma, experienced massive teacher walkouts. Teachers left their classrooms to protest low pay and a lack of resources, with little government response. Fast forward one year, after the 2018 election, more than ten new teachers were elected to the legislature (Butcher and Kim 2021). Newly elected high school teacher John Waldron began to focus on legislation addressing education concerns, including a bill to amend school evaluations. In response, the State Superintendent of Public Instruction clarified that “they had invested a great deal of time in developing the system and didn’t want it changed.”1 To put it briefly, the administrator made it clear that changes to education policy were not part of the agenda. Legislative action and agency administration are closely intertwined in state government, but term-limited legislators are not around long enough to know how to counter the more experienced agency workers. Centered on the 1 Waldron, John. “I Was an Angry Teacher Fighting for Better Education Policy. Now,

I’m Shaping It.” Education Week April 11, 2019 (last accessed October 12, 2023), available at https://www.edweek.org/leadership/opinion-i-was-an-angry-teacher-fightingfor-better-education-policy-now-im-shaping-it/2019/04

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Butcher, Navigating Term Limits, https://doi.org/10.1007/978-3-031-39423-2_7

193

194

J. BUTCHER

implications of changing legislative careers, in this chapter, I explore to what extent the experience of term-limited legislators has altered the relationship with state agencies. The bureaucracy is an integral part of state government, but it receives far less attention than state legislators and governors. As Richard Elling (1992, 1) said, “modern government is bureaucratic government no less at the state than at any other level.” Thinking of the quote at the start of this chapter, have term limits diminished the long-standing reliance that lawmakers and administrators have on each other? This chapter focuses on the relationship between state legislatures and state agencies. Previous research has evaluated this relationship through the lens of legislative strength and the ability to dictate agency action. A subset of studies looks at the legislature’s capacity to conduct oversight of state agencies. Institutional factors that may weaken the legislature in comparison with state agencies receive less attention than other dynamics of state government. Using interviews, I examine the disparity between diminished legislative capacity and reliance on state agencies. Studies of congressional oversight have revealed the importance of principal-agent models, delegation, and latent oversight (McCubbins and Schwartz 1984; Ogul and Rockman 1990; Weingast and Moran 1983). However, the relationships that exist at the federal level differ from those that occur at the state level. In the states, “bureaucracy more closely resembles a collection of heterogeneous agencies than a monolithic institution,” (Brudney and Hebert 1987, 186). Broadly, agencies vary by type, structure, and funding (Brudney and Hebert 1987). And due to term limits the relationship between the legislature and state agencies has shifted. This shift now means that what we know of congressional oversight cannot be applied directly to the states. There is an unprecedented dynamic occurring in states with term limits; legislatures are becoming weaker while agencies are becoming more assertive. Currently, the literature focuses on a lack of legislative oversight. However, legislatures have increasingly relied on agencies for information and policy guidance. Because term limits remove legislators from office, decreasing legislative experience, the assumption is that legislators have become dependent on agency administrators. This assumption, however, has merited little attention and has yet to be explored. The following assessment gives insight into the changing relationship between legislatures and agencies. It begins with a look at current understandings of legislative agency relations and how they work together in

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

195

states without term limits. Following that, there is an evaluation of the decreasing levels of agency monitoring and increasing dependence on state agencies. This analysis reveals that while monitoring is not linked to legislative experience; reliance on state agencies is. Under term limits, newly elected legislators rely more on agencies for information, while their senior counterparts are not.

Legislative Oversight One of a legislature’s key duties is to oversee administrative agencies. As described by, Ogul and Rockman (1990, 7), “the purpose of oversight is to ensure that those to whom authority is delegated remain responsive.” Specifically, the legislature conducts oversight to ensure positive outcomes for their constituents, have a say in policy implementation, and monitor the state executive’s role. However, each of these roles requires someone with some level of government and policy expertise. Agency oversight is essential for the separation of powers. Much of the research evaluating legislative oversight has been centered on legislatures that can conduct oversight, namely more professional legislatures. There has been limited research evaluating the relationship between oversight and term limits. Current research concludes that oversight will decrease under term limits. Oversight is a tool to monitor agency activity, allowing the legislature to evaluate policy implementation and serve as a check on the executive branch. Oversight sometimes occurs when an agency has gone too far or caused a problem—i.e., fire alarm oversight (Baranowski 2001; McCubbins and Schwartz 1984). But as Baranowski argues, “effective oversight of state agencies requires the institutional capacity to oversee” (2001, 149). Such resources are not a given in all states. A legislature needs the capacity to monitor agencies and enforce any necessary sanctions to have oversight (Woods and Baranowski 2006). As previously noted, professionalization worked to increase legislative capacity (Rosenthal 1996, 1990). Professional legislatures are better able to increase oversight and regulation (Boehmke and Shipan 2015). Increased resources give the legislature greater influence (Woods and Baranowski 2006). While there is a relationship between professionalization and agencies, it is not a high indicator of policy influence (Baranowski 2001). However, the capacity that comes with professionalization is helpful for actively monitoring agencies (Baranowski 2001).

196

J. BUTCHER

In states without term limits, there is little motivation to conduct oversight unless it directly affects a member’s constituency and, as a result, their election prospects. Yet professional legislatures have a greater capacity to oversee because members have more resources to monitor an agency. This conflict closely aligns with McCubbins and Schwartz’s take on different types of oversight (1984). Legislators are likelier to wait for a fire alarm to be pulled because that is when oversight becomes electorally beneficial. The added variable of term limits and shortened tenure complicates current understandings of agency relations. Namely, there are concerns over whether or not members can gain enough experience and whether agency relations are worth the time when there is so little of it. Put another way, do inexperienced legislators conduct routine oversight? The answer to these questions remains unclear because the long-term effects of term limits and state agencies are unexplored. Term Limits and Oversight According to a study by Marjorie Sarbaugh-Thompson and colleagues (2010), before term limits, only some legislators viewed oversight as one of their job responsibilities. Studies in other states revealed that oversight was neither a top priority nor a common practice (Woods and Baranowski 2006). Regardless of the institution they served in, members with greater political ambition were less willing to conduct oversight because it was not electorally beneficial (Rosenthal 1981; Woods and Baranowski 2006). The changes brought by term limits can alter the relationship between the legislature and state agencies (Krause and Woods 2014). The worry was that term limits would result in a substantial cut to legislative capacity, leaving legislatures less able to compete with other governmental actors (Rosenthal 1990). When legislatures do not conduct oversight, they become a weaker actor, allowing agencies to take on a greater role (Powell 2007). Shortly after term limits had passed, there were already noticeable changes to oversight. One early study warned, “oversight is fatally weakened by term limits,” (Capell 1996, 86). Others observed a decline in oversight correlated with the implementation of term limits (Berman et al. 2004; Farmer 2007; Kurtz et al. 2007; Moen et al. 2005; SarbaughThompson et al. 2010). Given the hands-off approach by the legislature, implementation was left to bureaucrats. In Michigan, there was a decreased interest in oversight in the legislature (Sarbaugh-Thompson

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

197

et al. 2004). There were also fears that term limits would hinder the legislature’s ability to monitor budget offices (Jewell and Whicker 1994). In an interview about term limits, one legislator remarked, “term limits won’t increase turnover by much, but they will hurt legislative capacity” (Carey et al. 2000, 93). It appeared, early on, that the most significant impact of term limits would be that legislators would be unable to compete with other governmental actors. As states transitioned to term limits, it became clear that there was a lack of legislative oversight. In a study by Richard Powell (2007, 143), “observers stated that committee oversight of executive agencies is much less rigorous under term limits.” Another study evaluating legislative monitoring before and after the implementation of term limits found that the amount of time dedicated to oversight had decreased after implementation (Sarbaugh-Thompson et al. 2010). Under term limits, members lack experience and are less equipped to process information; as a result, they no longer hold the expertise necessary to constrain agencies (Krause and Woods 2014). Some states began to notice the decreased ability of the legislature to monitor agencies. States began to create other entities or committees dedicated to oversight in order to combat this problem. For example, Maine created an Office of Program Evaluation and Government Accountability (Powell 2007, 145). At the time, there were several reasons for the lack of oversight. First, the legislature found it difficult to conduct oversight because there was a loss of expertise with the increased turnover brought by term limits (Powell 2007; Sarbaugh-Thompson et al. 2010). Second, the more politically ambitious legislators who served under term limits tended to find oversight to be a poor use of their time (Ogul and Rockman 1990; Sarbaugh-Thompson et al. 2010). Members have little incentive to conduct regular oversight of state agencies because they receive no credit for doing so (Elling 1992; Powell 2007). Instead, legislators will focus on more electorally beneficial issues, such as policy creation or bringing dollars back to the district. Third, many new legislators did not know what it meant to monitor an agency. They did not understand the practice until they were in their final term in office (Sarbaugh-Thompson 2010). To summarize, the initial turnover brought about by term limits took away a great deal of expertise from the institution, resulting in new legislators who do not know what to do regarding oversight and are more focused on their political careers.

198

J. BUTCHER

The oversight that did occur has been linked to committee work. In other words, members on certain committees were more likely to conduct oversight. Serving on the Appropriations Committee, particularly, has been tied to practicing agency oversight (Sarbaugh-Thompson et al. 2010). Similarly, some committee members were more willing to challenge agency practices created before they were elected to office (Clucas 2003). Term limits cut off the long-term relationship between legislators and agency administrators, so new members are more willing to challenge old norms (Sarbaugh-Thompson et al. 2010). As a consequence of decreased oversight, agencies were better able to thwart legislative efforts to monitor their activities (Sarbaugh-Thompson et al. 2004). Additionally, the lack of oversight increased deference to the executive (Baker and Hedge 2013; Cain and Wright 2007; Kousser and Straayer 2007; Little and Farmer 2007; Sarbaugh-Thompson et al. 2010). During the transitional phase, studies affirmed that legislators were not actively monitoring state agencies. Instead, it was a job relegated to a few specific committees. The lack of oversight led to the creation of new entities to assist the legislature in monitoring. The lack of expertise and lack of oversight were theorized to lead to legislative dependence. The drastic changes brought by term limits were not only going to limit oversight, but these changes would lead to legislators becoming dependent on agencies for information. A Growing Dependency A long-standing theory is that term limits, by inducing turnover, would result in a lack of oversight. This lack of oversight would free agencies from legislative monitoring and sanctions. The implementation of term limits increased the legislature’s dependence on state agencies for information. With increased turnover and shorter terms, members do not have the time for information gathering. Instead of gaining institutional knowledge, legislators have come to rely heavily on state agencies and administrators for information and policy guidance. Term limits diminish the expertise that is required for policy-making, and in doing so, increase the power of bureaucrats (Bowser and Moncrief 2007; Kousser and Straayer 2007). Legislators now turn to the bureaucracy for information, while their predecessors were able to be their own experts. Newer legislators, in particular, are more willing to look to the bureaucracy as a primary information source (Cain and Wright

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

199

2007). Bureaucrats are more likely to work with these new legislators because they can exert greater influence on them (Cohen and Spitzer 1996). These changes are not just at the individual level but affect the institution (Cain and Wright 2007). The transitional phase of term limits saw a growing dependence on state agencies for information. Newer legislative members viewed agency administrators as experts to whom they frequently deferred (Powell 2007). There was increased reliance on staff and bureaucrats in termlimited states because staff and bureaucrats served longer than the members (Carey et al. 2000). As a result, there is greater bureaucratic influence on legislation (Nicholson-Crotty and Miller 2011). Overall, term limits increased the perceived influence of bureaucracy among legislators (Nicholson-Crotty and Miller 2011). There are consequences to the changing relationship between legislators and bureaucrats. Legislators have less knowledge, and bureaucrats have less incentive to be honest with them (Sarbaugh-Thompson et al. 2004). There is also the potential for the executive branch to go around the legislature; bureaucrats can avoid responding to the legislature altogether (Sarbaugh-Thompson et al. 2004). The inability of the legislature to monitor and enforce bureaucratic action shifts power to the executive branch because it is better positioned to monitor state agencies. As a result of increasing agency autonomy, the balance of power in the states is flowing in favor of state executives.

Implications for Bureaucratic Relations In this section, I take an empirical approach to understand the relationship between state legislatures and agencies as they exist today. Legislative interviews allow an in-depth look at how current lawmakers interact with state agencies. The following analysis has two parts: agency oversight and agency dependence. First, legislators were asked how much time they spent monitoring agencies. Second, legislators revealed how much they rely on outside sources, particularly agencies, for information. The interviews reveal that while legislators spend little time monitoring state agencies, most rely on state agencies for information. There are three questions I use to gauge the amount of time spent monitoring and the extent to which legislators rely on agencies. To capture agency monitoring, legislators were asked, “How much time

200

J. BUTCHER

do you spend monitoring state agencies?” Responses were scaled: 1none, 2-little, 3-some, 4-a lot, and 5-most. There are two questions to better understand legislators’ reliance on agencies for information. First, members were asked, “Do you rely on outside sources for information, like lobbyists and agencies?” The responses to this were 1-yes and 0-no. Second, members were asked how they formed their decisions and where they got their information. The specific question about agencies is, “How much do you rely on state agency workers?” These answers are coded using the same five-point scale as above.2 Additionally, there are openended responses, which provide insight into lawmakers’ perceptions of agency monitoring and reliance on agency expertise. There is a consistent trend in the responses across the four states; legislators are skeptical of agencies but see them as valuable resources. Many legislators noted that agency workers have their own agenda and goals to pursue, so they have a healthy amount of skepticism. The remainder of this section looks at the differences in legislative agency monitoring and legislative dependence on agency expertise across the four states. First, there is an independent descriptive assessment of monitoring and reliance. Following that, there is a quantitative examination of monitoring and reliance to compare the two outcomes. While there are slight differences among the states, there are trends that are likely a result of the imposed term limits. Namely, there is less oversight and agency monitoring simply because legislators feel increased pressure for time. Additionally, while legislators are skeptical of agency intentions, they are more likely to rely on those agencies for information and expertise because they are pressed for time. Agency Monitoring When states implemented term limits, the leading concern about state agencies was that legislators would no longer conduct oversight. The theories as to why this change would happen centered around the diminished capacity to conduct oversight, related to policy expertise, and the reduced incentive to conduct oversight because of the shorter terms in office. Term-limited legislators were asked how much time they spend monitoring state agencies to understand whether these fears came to 2 In the cases where members answered in between numbers (i.e., 3.5), answers were rounded down because of their uncertainty that it was any higher than the half number.

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

201

Table 7.1 How much time legislators spend on agency monitoring None Missouri Nebraska Ohio Oklahoma Total (%)

3 1 3 3 10(11)

Little 15 7 11 13 46(51)

Some 10 5 5 5 25(27)

A Lot 2 1 4 1 8(9)

Most 2 0 0 0 2(2)

Total 32 14 23 22 91

Note Percentages are rounded to the nearest whole number

fruition. Table 7.1 contains the breakdown of responses by state.3 The answers show trends consistent across all four states. These responses are not necessarily representative of all term-limited states, but they present a sample of all four types of term limits. In Table 7.1, there are few members who respond that they spend either a lot or most of their time monitoring state agencies. In fact, the two categories make up less than 10% of the total responses. Nearly 30% of members report spending some time monitoring agencies. Most members report spending little time monitoring state agencies; just over 50% do so. After that, 10% of members report spending no time at all on monitoring. These trends are consistent among the states. In each state, most people report spending little time, followed by some time, and the third most common category is none. The lowest recorded responses in each state are in the most category, followed by a lot. When legislators were asked about how much time they spend monitoring state agencies, there were many different responses, most indicating a lack of oversight. The specific question asked was, “How much time do you spend monitoring state agencies?” While answers frequently demonstrated a lack of monitoring, legislators gave consistent reasons for this neglect. Some individuals said they simply did not have the time to monitor agencies. Recognizing the importance of monitoring, many members said they wished they could conduct more oversight. Another reason mentioned is that monitoring is viewed as a committee-specific task. One Oklahoma legislator said simply, “we don’t have time.” In response to agency monitoring activity, a high-ranking party leader in 3 Those who responded with “don’t know” are excluded from the analysis.

202

J. BUTCHER

Missouri said there is “not enough, I wish we could do more, but there’s only so much time in the day.” A member from Oklahoma expressed concerns about monitoring saying, “I need to do more (... I am) already seeing some problems.” On a more positive note, a Missouri senator said, “That’s something we should do a lot more of... not always a bad thing, also identifying the good things they are doing and what can be improved.” A senior Missouri legislator said there is “not enough” agency monitoring, while another said they “wish it was more.” A Nebraska state senator said, “needs to be more of that,” another senator said they are not able to monitor as much as they would like. Legislators in Ohio complained that they have very little time to conduct oversight because they are “always campaigning.” Other responses pushed the burden of conducting oversight onto the appropriations and budget committees. Those not serving on these committees did not view oversight as one of their jobs. One Missouri legislator said that oversight is vested in the budget process, stating it is a flaw in the design of agency monitoring. An Oklahoma legislator not serving on the budget committee responded, “None, it’s not my job.” That same legislator continued, saying, “I don’t have time to sit here and micromanage them—not a personal duty.” A Missouri representative said, “[monitoring is] what we need more of... some state agencies aren’t good at telling us the truth—biggest control is through the budget.” Coincidentally, their colleague reported that they spend a “tremendous” amount of time monitoring due to their role on the budget committee. Overall, the statements offered by legislators, and the initial scaled responses, indicate that there is a lack of oversight occurring in these specific state legislatures. Yet it is not conclusive that this is unique to term-limited states. Recall state legislators without term limits and even before term limits had little incentive to conduct oversight. The question then becomes, is the lack of oversight exacerbated in term-limited states, or are these legislators behaving rationally and seeking out opportunities with considerable electoral benefits? The answer to this question lies in how much experience members have, knowing that the intention of term limits is to cut experience. If agency monitoring is tied to legislative expertise, that would indicate a difference in term-limited states—new members simply do not know how to monitor.

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

203

Agency Reliance While scholarship has paid a great deal of attention to agency monitoring and the changes brought by term limits, there has been less attention to how much legislators have come to rely on agencies for information. Legislators are serving with less experience and less policy expertise, which is by design. Term limits are intended to bring in new, inexperienced members. The cost of this change remains overlooked: not only do termlimited members not monitor agencies, but legislators go to state agencies for information. With limited knowledge, lawmakers go to the entity that has to actually implement policy, the state agency. State agencies gain informational power over the legislature because legislators have become dependent on that knowledge. I use two different questions related to information. The first question asked legislators if they use outside sources for information. While the second asked how much they rely specifically on state agency workers for forming their opinions or making decisions. Again, these results are consistent among the four states. Table 7.2 contains the breakdown of responses to whether or not members go to outside sources to gain information. Specifically, the question asked if they go to external sources such as lobbyists and state agencies. Most members confirmed that they do go to outside sources for information. In fact, 91% of members interviewed said they go outside the legislature to get information from lobbyists and state agencies. Looking at the number of members who were willing to go to outside sources for information and who have to rely on that information is an essential first step to understanding agency dependency. Looking specifically at state agencies, Table 7.3 contains the responses from members on how much they rely on state agency workers. There is slightly more Table 7.2 Do legislators rely on outside sources for information?

Yes Missouri Nebraska Ohio Oklahoma Total (%)

No 31 13 18 22 84(91)

Total 3 1 3 1 8(9)

Note Percentages are rounded to the nearest whole number

34 14 21 23 92

204

J. BUTCHER

Table 7.3 How much legislators rely on state agency workers None Missouri Nebraska Ohio Oklahoma Total (%)

2 2 0 1 5(5)

Little 4 4 8 2 18(19)

Some 19 7 7 13 46(49)

A Lot 12 0 8 6 26(28)

Most 0 0 0 1 1(1)

Total 37 13 23 23 93

Note Percentages are rounded to the nearest whole number

variation here, but the distribution is mainly consistent among the states. From the states, most members rely on state agency workers for information and forming their opinion either some or a lot. While most members do not say they rely on state agencies the most, they depend on them greatly. Less than 25% of members say they rely on agencies none to little. Overall, there is more reliance on agencies for information than there is not. Interviews with state legislators revealed support for this suspected changing relationship between state agencies and legislatures. First, legislators said they spend little to no time monitoring state agencies. Few members prioritize it and spend a lot or most of their time on monitoring. Legislators, however, rely on outside sources for information, and they rely heavily on state agencies. More than 90% of respondents look to external sources for information. Specifically, over 75% of respondents rely on state agency workers some to a lot. State legislators were also asked about whether they gather information from state agencies or if they rely on agencies for information. The responses from most states indicate that legislators tend to rely on state agencies for both general information and technical expertise. However, one Missouri representative noted that working with agencies varies by gubernatorial administration. Unified/Lifetime An Oklahoma representative stated they go to agencies sometimes for consultation, especially regarding the influence of policies. One legislator said they would go to secretaries of agencies for information. Another member said they “absolutely” go to agencies, “how else will

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

205

you know?” Similar to other state legislators, Oklahomans go to agencies for information expertise because agency workers are the ones who retain institutional knowledge. Several members said they do not go to agencies; instead, agencies go to them. Similarly, another member said agencies come to initiate a conversation, but they rely on agency heads more so than other actors (i.e., lobbyists). Unified/Consecutive In Nebraska, state senators had little to say about their relationship with state agencies. When they did speak about agencies and the information they supply, their comments tended to be positive and indicated a good working relationship between the two actors. A senator from Nebraska shared that they go to agencies for information, especially when there are questions about policy mechanisms. They depend on agencies regarding what the agencies do and how they do it. Legislators stated various reasons why they would go to agencies: to get current information, to see how bills influence agencies, and to get important facts for bill writing. One senator said, “you have to (listen to agencies) simply because they’ve been working on it for years.” The positive feelings toward agency information may result from the limited terms legislators can serve. Divided/Lifetime Several Missouri legislators said they look to state workers because they have the knowledge, information, and specialized expertise. One senator agreed that “bureaucrats have tremendous knowledge,” they continued, “the knowledge of bureaucrats is much greater because every eight years legislators have to reeducate their workforce.” Another member said bureaucrats are helpful when they choose to give information. One representative spoke against agencies, saying, “they don’t like to share a lot, they stone wall,” they continued, “because of term limits they wait us out.” Divided/Consecutive A first-term representative from Ohio said you have to get information everywhere. They said, “I try to be a sponge: go, first and foremost, to my experience, then to experts and people on the ground level.” The same member commented, “if you need numbers or stats you have to go to them.” A different member pointed out that “It’s difficult to figure out the nuts and bolts” of policy and legislation. A representative in

206

J. BUTCHER

their second term candidly stated, “saying I’m not influenced would be deceiving, but I do not reach out to them.” For Ohio legislators, it is clear that if they want detailed information on a particular policy; they have to go to the source, which is the relevant agency. It is clear that term-limited legislators are more than willing to go to outside sources for information, with an overwhelming 90% of lawmakers stating that they rely on external sources. Additionally, over 75% of members rely on information provided by state agencies, either some or a lot. This is likely a direct effect of having legislative term limits. Term limits have unintentionally given informational power over state agencies as members have become more reliant on agency workers for their policy expertise and knowledge of policy implementation. Legislators and Agencies While it is evident that there is a lack of monitoring in the states, as well as an increased reliance on state agencies as an information source, it is unclear if these are direct effects of term limits. The descriptive information above illustrates that there is more to the relationship between states and agencies than previously thought. This section centers on the differences in agency monitoring, reliance on agencies, and how states with term limits have interacted with agencies. Early scholars who feared a lack of oversight in term-limited legislatures failed to account for the pre-existing lack of oversight. Members who desire to be reelected take action to get reelected, and oversight is not shown to be an electorally beneficial act. Therefore, a member’s level of experience should not be indicative of their incentive to conduct oversight. In other words, the motivation to conduct oversight by monitoring state agencies is not present, with or without term limits. With the focus on monitoring, there was little attention paid to how a lack of experience plays out in reality. Monitoring may not be rooted in the level of legislative experience but in how much a member relies on a state agency. With increased turnover and shorter lengths of service, members need to seek information from someplace. With few experienced members, state agencies seem the likely option. State agencies help to develop policy and are the entities that implement the policy; for a new legislator, they constitute a primary source of information. Agencies are not only not being monitored, but they are the ones who hold informational power.

7

CHANGES TO EXTERNAL RELATIONS: STATE AGENCIES

207

Using the same interview questions above, I use ordered logistic regression to understand the relationship between experience and monitoring and reliance. Keep in mind these two different outcomes are in opposite directions. Monitoring implies that a legislator has enough time and knowledge to conduct oversight. Reliance implies that members lack knowledge and thus turn to agencies to fill in that information gap. Research has long suggested that inexperience would lead legislators to monitor less while increasing their reliance on agencies for information. The key independent variable for this analysis is experience. The experience variable captures the number of years that each member has served. The minimum number is zero; these are the members who just began serving in the legislature. The most years served in this sample is 19. Recall legislative careers in Ohio are similar to those in non-term-limited states because members can return and thus are able to serve for extended periods. Theoretically, experience is the critical difference between termlimited and non-limited states. The lack of experience is the reason for the lack of oversight and for increased informational reliance. If experience is significant, it can be assumed that experience is a key factor for monitoring or reliance. If experience is insignificant, then it indicates there is no difference based on experience in states without term limits. Both dependent variables, monitoring and reliance, are measured on a five-point scale, “none,” “little,” “some,” “a lot,” or “most” of time spent on monitoring/level of reliance of agencies. The model contains an indicator for three states, Ohio, Missouri, and Oklahoma, with the reference point being Nebraska. Additionally, the model includes a variable for prior legislative experience and indicators for White legislators, Republicans, and males. Table 7.4 contains the results of the ordered logit for both monitoring and reliance. This analysis allows for a comparison of the states while presenting an assessment of the importance of legislative experience. Agency monitoring is not significantly different for the states of Ohio and Missouri, but Oklahoma has significantly less monitoring than Nebraska. Based on this analysis, it appears that the level of monitoring does not vary by the level of experience. Additionally, being Republican or male do not have a statistically significant relationship with monitoring. However, White lawmakers are less likely to spend time monitoring state agencies. Broadly, the amount of time that a legislator spends on monitoring is not dependent on whether or not he or she has experience or what party they are from.

208

J. BUTCHER

Turning to how much legislators rely on state agencies. If term limits are going to do anything, it is more likely that they will give greater informational power to the bureaucracy because there will be a constant influx of new members. On top of that, there will be fewer senior members to help guide all of the new members. As a result, legislators will rely on the bureaucracy because of their inexperience. Table 7.4 demonstrates just how significant that relationship is. Each state is significantly more likely to rely on its administrative agencies for information than Nebraska. Legislators with more experience are significantly less likely to rely on state bureaucrats (p