Introduction
For decades, America has engaged in a rigorous, and sometimes bitter, debate regarding the government’s role in providing health care to its citizens. In recent years, this divisive policy debate has emerged in several state legislatures. For instance, in 2017, the California legislature strongly considered enacting a state-sponsored single-payer health care system through S.B. 562 or the “Healthy California Act.” Footnote 1 If enacted, this would be the first single-payer system in the United States. During the consideration of S.B. 562, there was strong support for the policy as most polls showed that most Californians supported the establishment of a single-payer system (Reference MurphyMurphy 2017). Footnote 2 Additionally, the advocates of S.B. 562 argued that it would “eliminate the high premiums, deductibles, and other costs that force far too many to skip needed care” (Reference BurgerBurger 2018). While a significant portion of citizens living in California viewed S.B. 562 as an inspiration for the single-payer health care system in America, some politicians and journalists expressed deep concerns on this revolutionary bill. In particular, several individuals highlighted the large price tag for S.B. 562. The San Diego Union-Tribune Editorial Board (2017a) candidly stated “it was downright silly” to approve the bill without addressing the costs associated with it. A coalition of doctors and other medical professionals argued that S.B. 562 would “dismantle the healthcare marketplace and destabilize California’s economy” (Reference MasonMason 2018).
A few months after being introduced in the Senate, S.B. 562 sailed through the California chamber mostly along partisan-lines. Next, the bill proceeded to the State Assembly where it faced several political hurdles. In particular, Speaker Anthony Rendon (R-CA) publicly expressed heavy skepticism toward the bill. “S.B. 562 was sent to the Assembly woefully incomplete,” Speaker Rendon proclaimed, “Even senators who voted for S.B. 562 noted there are potentially fatal flaws in the bill including the fact it does not address many serious issues, such as financing, delivery of care, cost controls … ” Footnote 3 Thus, few were surprised when Speaker Rendon sent S.B. 562 to the Rules Committee—a committee filled with his political allies—and ordered the bill not to proceed. Footnote 4 By using this procedural tactic, Speaker Rendon was able to effectively kill the bill (Reference MurphyMurphy 2017). While several prominent newspapers and politicians hailed Speaker Rendon’s decision as “the right move” (e.g., Editorial Board 2017b), several progressives in California expressed deep frustration at their leader for using his power to block a key policy on their legislative agenda. For example, Deborah Burger, co-President of the California Nurses Association and ally of S.B. 562, decried, what she considered, Speaker Rendon for “petulantly refus[ing] to do his job” and for using his procedural powers to “block[] public hearings and any amendments on the bill.”
California’s S.B. 562 illustrates how a state Speaker’s institutionalized powers can have significant consequences on policy outcomes. However, the willingness of the California State Assembly members to endow their Speaker with these powerful procedural tools stands in stark contrast to other states, such as Alaska and North Dakota, whose leaders are much more limited in authority. For example, Speaker Bryce Edgmon (D-AK), the current leader of the Alaska House of Representatives, has little sway over the committee assignments, leadership structure, or even the legislative staff (Reference MooneyMooney 2013). This raises the following question: why are certain states’ Speakers more powerful than others?
One important institutional characteristic that might partially account for the amount of formal powers delegated to a state Speaker is legislative term limits. While other state legislative factors, such as professionalism (Reference ClucasClucas 2007) and the distribution of partisan preferences (Reference Aldrich, Rohde, Dodd and OppenheimerAldrich and Rohde 2001; Reference RichmanRichman 2010), are associated with a powerful chamber leader, the extent to which term limits are a factor remains uncertain (Reference MooneyMooney 2012). Building off Reference MooneyMooney’s (2012) collective action theoretical framework on Speaker power, I argue that legislative term limits might incentivize the members to delegate more tools to control the lawmaking process to their chamber’s leadership to overcome collective action problems for two reasons. First, one consequence of term limits is that a decrease in institutional familiarity and policy expertise tends to occur (Moncrief, Niemi, and Powell 2008) and this may pose a challenge to lawmaking. As a result, legislators might be willing to delegate more institutional tools to their leadership to overcome this obstacle to governing. Second, term limits tend to weaken relationships among the legislators serving within the chamber and enhance the influence of the Speaker (Reference Sarbaugh-Thompson, Thompson, Elder, Commins, Elling and StrateSarbaugh-Thompson et al. 2006). Consequently, the lack of personal relationships among the members may present a challenge to the chamber regarding passing important legislation and accomplishing their policy goals (Kirkland 2011). Thus, the chamber might be more willing to delegate power via tools to control the lawmaking process to their leadership. I test this theoretical expectation with Reference MooneyMooney’s (2013) measure of Speaker power across 49 states between 1981 and 2015.
This article contributes to our understanding of legislative leadership and term limits for several reasons. First, the power delegated to a state Speaker can have a wide range of consequences for public policy. If term limits tend to empower the Speaker, then this may ease their ability to affect legislation. This would provide additional insights regarding the politics surrounding California’s S.B. 562 and the debate about establishing a single-payer health care system. After the enactment of term limits, the California Speaker obtained more control over the appointment of key leadership officials and the committee system (Reference MooneyMooney 2013). During his tenure, Speaker Rendon filled the Rules Committee with his political allies to help further his political agenda (Reference WhiteWhite 2016). Eventually, this committee blocked the passage of S.B. 562 (Reference MurphyMurphy 2017). Thus, given the potential consequences a Speaker may have on policy, it is important to understand the factors that influence the chamber’s decision to delegate to their leadership. Second, most studies that analyze Speaker power rely on cross-sectional data which has a few limitations. Specifically, this type of data structure cannot analyze the changes in Speaker power across time. Instead, I employ a time-series-cross-sectional data set and find that term limits are associated with more authority being delegated to the Speaker.
This article proceeds as follows. First, I briefly highlight the literature on state Speaker power. From there, I develop the argument, based on Reference MooneyMooney’s (2012) collective action theoretical framework on legislative leadership power, that legislative term limits should be a key determinant in a state Speaker’s power via the delegation of institutional tools that control the lawmaking process. I then describe my data, research design, empirical results, and robustness checks. Finally, I review the potential implications of my findings for term limits and leadership power.
Previous Literature on Leadership Power
At the core of politics is power (Reference DahlDahl 1957), and state Speakers can use their power to impact policy. For instance, the Colorado media has observed how Speaker Crisanta Duran (D) of the Colorado House of Representatives has used the power delegated to her office to control the legislative agenda. In particular, Speaker Duran has almost complete authority over bill referral, and she frequently sends legislation that is antithetical to her party’s core goals to the Veterans Committee, which is commonly referred to as the “kill committee.” Speaker Duran unilaterally appointed the “kill committee’s” chair and decided which members would be assigned to it. These tools ease Speaker Duran’s ability to control legislation (Reference AguilarAguilar 2017). Or as one Colorado legislator candidly stated when describing the Veterans Committee, “They have five hard core Democrats there who always vote no on Republican bills, with a few exceptions, but generally once it goes to the kill committee, the fix is in.” Legislation that would undermine the Affordable Care Act or “Obamacare” and a bill that would loosen gun regulations are just a few examples of legislation that has been sent to the “kill committee” (KGNU 2016). This example demonstrates how the tools delegated to the Speaker can sometimes have a substantive impact on legislative outputs. As this section will demonstrate, legislative scholars have also observed the importance of state Speakers and this has resulted in a well-developed literature.
Early research on state legislative organization uses the institutional and political variation among the states’ legislatures as a way to test prominent legislative theories (Reference ClucasClucas 2001; 2009) For instance, Richard Reference ClucasClucas (2001) examines state legislatures and the amount of power delegated to House Speakers to provide evidence for David Reference MayhewMayhew’s (1974) electoral theory of legislative politics. Specifically, he finds that legislators in electorally competitive states are more likely to delegate legislative control to the chamber’s leadership, so the leaders can more easily gain access to staff and valuable committee assignments for their reelection prospects. In a later article, Reference ClucasClucas (2009) again uses state legislatures to demonstrate the robustness of a prominent legislative theory, conditional party government (CPG). CPG posits that when the majority party is ideologically homogeneous and ideologically distinct from the minority party, then more power is delegated to the chamber’s leadership (Reference Aldrich, Rohde, Dodd and OppenheimerAldrich and Rohde 2001). Clucas finds indirect evidence that if the criteria for CPG are satisfied, then the majority party leader in the lower chamber does in fact obtain more influence on public policy.
Within this line of inquiry, legislative scholars have also analyzed state legislatures not just for their insights into legislative theories, but to uncover the consequences of recent political reforms and political trends (Reference ClucasClucas 2007; Reference MooneyMooney 2012; Reference RichmanRichman 2010). For instance, in the 1960s, state legislative reformists fought for more professionalized chambers to enhance their policymaking capabilities and effectiveness (Reference SquireSquire 2007). Reference RichmanRichman (2010) finds that this policy change tends to weaken the office of the Speaker. He contends that the large legislative staff and high salaries, in professionalized legislatures, allow the legislators to more easily achieve their policy and reelection goals without the assistance from the Speaker. In another prominent example, Reference MooneyMooney (2012) explores how partisan polarization, a salient political phenomenon, influences control over the legislature. Mooney finds that the distribution of the parties’ preferences has little effect on the amount of power delegated to the leadership. Instead, he finds that collective action problems caused by internal dynamics play a more prominent role.
Germane to my research, state politics scholars have focused on a more recent political reform—term limits and their effect on political power. For instance, term limits are associated with governors who can easily expand their influence over policy (Reference Miller, Nicholson-Crotty and Nicholson-CrottyMiller, Nicholson-Crotty, and Nicholson-Crotty 2011) because the chamber is unable to defend their institutional interests. With respect to chamber leadership powers, conventional wisdom among legislative scholars suggests that they result in weaker leaders (Reference Bowser, Jones, Kurtz, Rhyme, Weberg, Farmer, Rausch and GreenBowser et al. 2003; Reference Carey, Niemi, Powell and MoncriefCarey et al. 2006; Reference MooneyMooney 2012). Reference BowserBowser (2005, 113) succinctly describes this perspective, “Perhaps the most noticeable changes in many term-limited legislatures have to do with leadership. Leaders rise to the top more quickly than before but stay for a briefer period and wield less influence than in the past.” Specifically, this literature argues that Speakers in term-limited states are less powerful because they lack the lawmaking experiences and do not have the time to cultivate the skills necessary to obtain institutional tools that can control the lawmaking process. Conversely, the conventional approach claims that leaders in nonterm-limited states have the time to develop their skills in acquiring power and the members are incentivized to invest in the internal power structure.
A few students of legislative politics have nevertheless suggested that term limits may sometimes empower the legislative branch and the chamber’s leadership. For instance, Reference Farmer, Little, Farmer, Mooney, Powell and GreenFarmer and Little (2007) uncover that term limits in Ohio are associated with a more powerful Speaker. They claim that the Speaker’s recruitment efforts and frequent fundraising for the members of his party result in a loyal following. In another prominent example, Reference MooneyMooney (2012) suggests that legislators in term-limited states might delegate to their leadership to overcome collective action problems associated with term limits. Footnote 5
While the current literature does offer a deeper understanding of the factors at play with legislative power, there are still more questions than answers regarding term limits and their implications. Specifically, the literature provides two different depictions of term limits and their influence on leadership power. I enter this debate by taking a different approach when exploring the effects of term limits on the legislative branch. Specifically, I analyze how this political reform influenced actual chamber rules. Most scholars have explored the impact of term limits on legislative power with legislative surveys. While this type of data has provided deeper insights into the politics of term limits, a limitation to this approach is that these scholars can only capture how legislators perceive term limits’ effect on their chamber leadership. No study has explored the consequences of term limits with actual tools the Speaker can use to control the chamber over an extended period of time or with multiple states. Importantly, I aim to uncover if term limits influence the actual legislative structure and organization within a chamber and not just the perception among legislators.
Linking Term Limits to Speaker Power
Term limits are widely supported by the American public (Reference KousserKousser 2005); thus, it is not surprising that several states have adopted them. Table 1 identifies the 21 states that have adopted, implemented, and, in some cases, repealed legislative term limits. Since the implementation of term limits, several political consequences have emerged. Term limits have been shown to influence political behavior (Reference Kuhlmann and LewisKuhlmann and Lewis 2017; Reference Masket and LewisMasket and Lewis 2007), institutional organization (Reference Nicholson-Crotty and MillerNicholson-Crotty and Miller 2012; Reference WrightWright 2007), and public policy (Reference KousserKousser 2005; Reference LewisLewis 2012). While previous research suggests that term limits can have consequences on policy outcomes and legislative behavior, it is still unclear to what extent term limits have any influence on chamber rules. This study focuses on the effect of term limits on one particular type of chamber rule—the number of tools or powers delegated to the Speaker. The power a Speaker wields can have far-reaching consequences on policy outputs and government performance (Reference FiorinaFiorina 1986; Reference Hasecke and MycoffHasecke and Mycoff 2007; Reference Lazarus and MonroeLazarus and Monroe 2007; Reference RohdeRohde 1991); therefore, it is important to uncover the factors that incentivize the members to delegate to their leadership. In the remaining portion of this section, I rely on Reference MooneyMooney’s (2012) collective action theoretical framework on Speaker power to outline why legislative term limits are likely to incentivize the lower chamber to delegate more formal institutional powers to the state Speaker.
Table 1. States with Term Limits.
Source. National Conference of State Legislatures.
Note. In Oklahoma, the limit is a total of 12 years in the legislature. The total time may be split between the two chambers or spent entirely in one.
a. In 2012, California changed its term limit law for the lower and upper from 6 and 8, respectively.
b. In 2014, Arkansas changed its term limit law for the lower and upper from 6 and 8, respectively.
Reference MooneyMooney (2012) suggests that a legislature may decide to delegate tools to the Speaker to overcome the collective action problems associated with the lawmaking process. In other words, the members establish a principal–agent relationship with their legislative leadership to help overcome the problems that burden the members serving within the chamber (Reference Bender, Glazer and HammondBender, Glazer, and Hammond 2001; Reference SmithSmith 2000). From the perspective of the members, “the costs of producing collective benefits, including public policy,” are far too high for one legislator to endure (Reference ClucasClucas 2001, 320). Therefore, the chamber’s leadership is delegated power, resources, and institutional tools. The amount of power delegated to the leadership is in response to the amount of problems facing the legislature. Specifically, states facing more obstacles tend to delegate more tools and resources to their leadership than states with fewer challenges. In turn, legislative leaders are expected to use their power to help the legislators achieve their political goals (Reference ClucasClucas 2001). In other words, Mooney suggests that the primary reason state legislators delegate authority to their Speaker is to address collective problems within the chamber (p. 567). Footnote 6
Why might legislative term limits create collective action problems that incentivize the members to cede authority to their leadership? I suggest there are two reasons. First, one of the consequences of term limits is that the number of inexperienced legislators serving in the chamber surges because of the high legislative turnover. For example, in 1998, after the implementation of term limits, 57% of Michigan’s House members were forced to retire (Reference Carey, Niemi and PowellCarey, Niemi, and Powell 2009). This dramatic increase in veteran legislator turnover tends to strip the chamber of its legislative knowledge and expertise (Reference Moncrief and ThompsonMoncrief and Thompson 2001). Not only does this frequent turnover tend to undermine the development of institutional familiarity, but also a “brain drain” of policy expertise may begin to emerge within the legislature (Reference Bowser, Moncrief, Kurtz, Cain and NiemiBowser and Moncrief 2007).
Inexperience in lawmaking, policymaking, and the chamber procedural process might stymie the members from achieving one of their primary goals as legislators, good public policy (Reference FennoFenno 1978). Footnote 7 In particular, the lawmaking process tends to be overwhelming for even the most experienced legislator (Reference Packenham, Kornberg and MusolfPackenham 1970; Reference Squire and HammSquire and Hamm 2005, 118). A bill sponsor usually must bargain and logroll with their chamber colleagues to increase the chances of their legislation succeeding through the several lawmaking stages (Reference Carrubba and VoldenCarrubba and Volden 2000). Moreover, a legislator needs to strategically and prudently manage relations with the public, committee staff, and executive branch while trying to ensure the passage of their policy agenda (Reference SinclairSinclair 1989). Consequently, a member with little legislative knowledge and unaware of the steps necessary to pass a bill, due to term limits, might struggle to accomplish most of their policy goals.
To overcome this inexperience hurdle caused by term limits, the chamber might concentrate power around the Speaker. When power is concentrated around the Speaker, he or she can more easily enforce and implement policy decisions that will help individual members achieve their political goals (Reference ClucasClucas 2001; Reference MooneyMooney 2012). While Speakers in term-limited states may not have much prior legislative experience themselves, they can still use their delegated powers to organize the chamber and to ensure the passage of important bills. In other words, a chamber can operate more effectively if power is concentrated among one inexperienced legislator than if it is shared among several green lawmakers. Furthermore, the Speaker can use his or her connections to the party’s apparatus for informational and policy expertise to help enact important legislation (Reference Squire and HammSquire and Hamm 2005).
Alternatively, legislators in nonterm-limited states can develop their policy expertise and acquire their legislative knowledge over time (Reference Sarbaugh-Thompson, Thompson, Elder, Commins, Elling and StrateSarbaugh-Thompson et al. 2006). Experienced legislators are aware of the potential pitfalls ahead of their legislative proposals and they can plan accordingly to ensure their success. This allows the legislators to easily achieve their policy goals on their own and without the aid of the Speaker (Reference LewisLewis 2012). Therefore, nonterm-limited legislators may have few reasons to delegate legislative tools to their Speaker.
The second reason that legislative term limits may create collective action problems, thus, resulting in the chamber delegating more power to the Speaker, is that term limits are associated with weaker legislative relationships between the members (Reference Cain and LevinCain and Levin 1999). Reference Sarbaugh-Thompson, Thompson, Elder, Commins, Elling and StrateSarbaugh-Thompson et al. (2006) provide evidence that suggests legislators serving in term-limited states have fewer relationships with other members and struggle to reach consensus on important policy issues than legislators without term limits. Sarbaugh-Thompson and her co-authors argue this is due to the term-limited members not having the time to develop a rapport or personal relationship with their colleagues. Without these personal relationships, it may be harder for the members to enact their policy agenda. In other words, for most members, getting their policy agenda through the legislature will require at least a simple majority in both chambers, which may be difficult to do if the relationships between the legislators have not been developed (Reference FennoFenno 1978; Reference KingdonKingdon 1973; Reference KirklandKirkland 2011). To overcome this collective action problem, the legislature may delegate more power to their chamber’s leadership to enhance the efficiency of the lawmaking process. This comports with Sarbaugh-Thompson et al.’s finding that term-limited members are usually influenced more by their Speaker. If the Speaker has more sway on the members’ decisions, then perhaps the chamber might be more willing to delegate tools to control the lawmaking process to their leadership.
Conversely, legislators in nonterm-limited states tend to have stronger working relationships with their fellow legislators (Reference Cain and LevinCain and Levin 1999). This should ease their ability to build consensus and enact important legislation (Reference Berman, Farmer, Mooney, Powell and GreenBerman 2007; Kirkland 2011), thus achieving their policy goals. If the members’ policy goals are met, then they might have less need to delegate institutional tools to their leadership.
To summarize, relying on Reference MooneyMooney’s (2012) collective action theoretical framework on Speaker power, I suggest that legislative term limits may incentivize the members to delegate more authority to their leadership to overcome collective action problems for two reasons. First, term limits increase legislative turnover and tend to result in a loss of legislative and procedural knowledge in the chamber. This loss of expertise might pose a challenge for legislators trying to pass their legislative agenda. Thus, legislators might be forced to delegate institutional tools to their chamber’s leadership to achieve their policy goals. Second, legislative term limits tend to weaken relationships among legislators within the chamber. Limited personal relationships can pose a challenge to enacting laws and for the members to achieve their policy goals. Given this, the members might be more willing to delegate power to their leadership. Therefore, my theoretical expectation is that legislative term limits should be associated with more tools delegated to the state Speaker. It should be noted that I am arguing that term limits results in the members delegating more tools to their Speaker and not necessarily that the chamber’s leader is effectively exercising this new power. Footnote 8
Data and Research Design
Dependent Variable
How should we measure state Speaker power? Both Reference BattistaBattista (2011) and Reference MooneyMooney (2012) note that the two primary instruments researchers typically used to gauge a Speaker’s power are legislative surveys and indices comprised of rules and formal institutionalized tools delegated to the Speaker. A major limitation with the legislative survey approach is that it cannot easily address factors that vary over time and the response rate for contemporary surveys is exceptionally low. Furthermore, the wording of the questions in legislative surveys can pose an additional challenge to comparing the effect of power across time (Reference BattistaBattista 2011). Therefore, and in an effort to avoid similar issues, I examine the effect of legislative term limits on state Speakers’ formal powers with the delegation of lawmaking tools and chamber rules with an additive index. This approach is commonly utilized by several legislative scholars (Reference Bagashka and ClarkBagashka and Clark 2014; Reference ClucasClucas 2001; Reference MartoranoMartorano 2004; Reference Martorano2006; Reference RichmanRichman 2010).
Reference ClucasClucas (2001) has created a Speaker power index based on several sub-indices that captures the chamber rules and various tools delegated to the Speaker. While his groundbreaking state Speaker power index is the foundation for numerous studies (Reference Miller, Nicholson-Crotty and Nicholson-CrottyMiller, Nicholson-Crotty, and Nicholson-Crotty 2011; Reference MooneyMooney 2012; Reference RichmanRichman 2010), one major limitation is that the measure only includes states legislatures in 1995. Moreover, it is not possible to expand the index with the original sub-indices into other time periods because of data availability (Reference MooneyMooney 2013). Therefore, I cannot examine if legislative term limits have any effect on Speaker power with this specific measure.
Reference MooneyMooney (2013) has recently constructed a similar index of Speaker power for 49 states and across multiple decades. To demonstrate his index’s validity, Mooney conducts various tests that show his index is strongly correlated with previous measures of Speaker power. For this analysis, I use Mooney’s index of Speaker power because it covers the time period before and after the implementation of legislative term limits and his index includes almost every lower chamber in the United States.
Reference MooneyMooney’s (2013) additive index is composed of five indicators with data collected from The Book of the States. The five indicators Mooney uses in his Speaker’s power index are the following: committee chair appointment power, committee assignment power, control over chamber’s leadership structure, procedural power, and control over legislative resources.
Committee chair appointment power is the Speaker’s ability to select committee chairs unilaterally. Committee chairs can use their gatekeeping powers to block legislation that could potentially upset their constituents and hurt their party’s reputation with the electorate. In addition, Speakers who can unilaterally appoint committee chairs can have more influence on policy, and they can threaten to withhold committee chair positions from legislators who rebel against the chamber’s leadership.
Committee assignment power measures the ability of state leaders to decide committee assignments. Legislative committees are the primary vehicles in which bills are created; therefore, control over this aspect of the legislature gives a Speaker more power over policy outcomes. Moreover, legislators prefer to sit on committees that might benefit their constituents. Therefore, Speakers can use this tool to reward their allies in the chamber.
The control over the chamber’s leadership structure indicator accounts for the Speaker’s ability to decide which members can join the House’s leadership. Speakers who can unilaterally structure the chamber’s leadership can more easily influence the legislative agenda, and they can use these leadership positions as an incentive for legislators to fall in line.
Procedural power measures how much influence the Speaker has in bill referral. Leaders who can unilaterally decide which committee can hear a bill can also decide the fate of legislation. For instance, a common tactic Speakers use to block legislation that is antithetical to his or her policy agenda is to refer the bill to a committee that he or she knows will not offer a report (e.g., the Colorado example discussed earlier in the article).
The control over legislative resources indicator accounts for the independence of professional committee staff from the Speaker. Committee staff that is insulated from the Speaker’s influence is less likely to follow his or her policy preferences.
These sub-indices are equally weighted and summed together. Higher scores indicate that a Speaker has more legislative tools at his or her disposal to control the lawmaking process. For this analysis, I extend Mooney’s index measure to 2015. Specifically, I examine all states calculated in Mooney’s index between 1981 and 2015. Mooney does not calculate a Speaker power score for Nebraska because it does not have a lower chamber. Therefore, Nebraska is excluded from my analysis. For more details on Mooney’s coding of Speaker power, see Reference MooneyMooney (2013).
The approach I adopt to measure Speaker power is not without its shortcomings. Footnote 9 In particular, there are three potential issues with this approach. First, the legislative procedures in Reference MooneyMooney’s (2013) Speaker power index may not contain every tool that is delegated to the chamber’s leadership. Second, a given legislative tool may not have the same impact in every state. Third, interpreting state procedural rules is no easy task and is a subjective process. Thus, it is most certainly possible that The Book of the States might have misinterpreted a few state rules. I am conscious of these limitations, and I recognize that Speakers may use other institutional tools not included within Mooney’s index to influence the lawmaking process and the precise impact of a particular tool could vary by state. Unfortunately, I cannot adopt alternative methods (e.g., legislative surveys) to gauge Speaker power because of the low-response rate. However, by adopting this index approach, I am able to examine the changes in Speaker power before and after the implementation of legislative term limits. It is also important to remember that the Speaker power index I utilize is highly correlated with other well-established scholarly measures (Reference ClucasClucas 2001; Reference Miller, Nicholson-Crotty and Nicholson-CrottyMiller, Nicholson-Crotty, and Nicholson-Crotty 2011). Thus, despite its limitations, this study offers an important substantive contribution to the effects of legislative term limits.
My unit of analysis is the amount of power a Speaker has in the legislative session following the most recent legislative election for the lower chamber. I use this unit of analysis because legislative procedures and rules are relatively stable between elections. Furthermore, chamber rules usually change only after the election of new members (Reference Squire and HammSquire and Hamm 2005). This unit of analysis is also consistent with Reference MooneyMooney’s (2013) original coding and collection of the Speaker power index. Footnote 10
In Figure 1, I show the Speaker power across the states in 1981 (Top Panel) and 2015 (Bottom Panel). Darker shaded states indicate a Speaker has more tools to control the chamber. Footnote 11 Conversely, lighter shaded states indicate a Speaker possess fewer institutional tools to control the chamber. Footnote 12 As suggested by Figure 1, there is a wide variation of lower chamber rules among the U.S. states and it does change across history. Kentucky’s Speaker has historically been the weakest chamber leader in the United States with a score of zero between the 1980s and the early 2000s. During this time period, the Kentucky Speaker had no control over the distribution of committee assignments, bill referrals, legislative staff, and the appointment of committee chairs or leadership positions. Alternatively, Oklahoma’s Speaker, in 2001, is the most powerful leader in my data set with a score of 4.28. In 2001, the Oklahoma Speaker has unilateral control over committee assignments, the appointment of committee chairs, and bill referrals. In addition, the Oklahoma Speaker has significant influence over the legislative staff, and he can appoint over a majority of the chamber’s leadership positions. The average Speaker power score in my data set is 2.69 and the standard deviation is 0.80. Footnote 13
Figure 1. Speaker power by state in 1981 (top) and 2015 (bottom).
Note. The top figure is the Speaker score for each state in 1981. The bottom figure is the Speaker score for each state in 2015. Darker shaded states indicate that the Speaker is delegated more institutional tools. States that are white indicate an institutionally weak Speaker. Nebraska is filled with the mean value of the data set and is included for visual purposes only. Nebraska is not included in any of the data analyses.
It should also be noted that my data only includes the amount of power delegated to leaders in lower chambers and I do this for three reasons. First, Reference Hamm, Hedlund and MartoranoHamm, Hedlund, and Martorano (2001) find that differences between the upper and lower chambers are less distinct in state legislatures. Second, Reference MooneyMooney’s (2013) measure of Speaker power only examines lower chambers. Finally, I focus only on lower chambers to remain consistent with previous research (see, Reference Bagashka and ClarkBagashka and Clark 2014; Reference ClucasClucas 2001; Reference Clucas2007; Reference Clucas2009; Reference MartoranoMartorano 2004; 2006; Reference MooneyMooney 2012; Reference RichmanRichman 2010).
Key Explanatory Variables
My variable of interest is a measure of legislative term limits. In theory, the effects of term limits could vary depending upon the specific restrictions within a state. For example, it is possible that Michigan’s strict term limit law of only allowing House members to serve a maximum of six years could have a different effect than Nevada’s more lenient limit of 12 years. Therefore, I employ Reference Baker and HedgeBaker and Hedge’s (2013) continuous term-limitedness scores. Footnote 14 Like Reference Sarbaugh-ThompsonSarbaugh-Thompson’s (2010) term limit measure, Baker and Hedge’s continuous scores take into account three different components of a state term limit law: the maximum length of service allowed, presence of lifetime or consecutive service limits, and whether legislators are allowed to “recycle” between the chambers. Footnote 15 All other observations are coded as “zero.” Higher scores indicate a state has a more stringent term limit law.
I focus on implemented term limits rather than adopted term limits because previous research suggests that implemented term limits are more likely to influence legislative organization (see Reference Bagashka and ClarkBagashka and Clark 2014; Reference KousserKousser 2005; Reference Miller, Nicholson-Crotty and Nicholson-CrottyMiller, Nicholson-Crotty, and Nicholson-Crotty 2018). More importantly, I focus on implemented term limits because my theory suggests that it is the loss of institutional familiarity, due to term limits, rather than the anticipation of term limits that drives the chamber’s decision to delegate more power to their Speaker. Footnote 16
Additional Explanatory Variables
I also control for other variables that could potentially influence the amount of power delegated to a state Speaker. As previously mentioned, Reference MooneyMooney (2012) finds that a prominent reason Speakers can easily acquire more power is to overcome collective action problems. Therefore, I control for the chamber size or the number of legislators serving in the chamber. Footnote 17 Specifically, legislators may empower their state Speaker with more tools to overcome problems associated with having a large number of members in the chamber. Rank and file members may also delegate more power to their Speaker to help address the dire economic conditions facing their constituents. The Unemployment rate variable is measured as the percentage of state residents who are unemployed at the start of the legislative session. Footnote 18 Also, states with larger populations may have more complex problems on their agenda (Reference RichmanRichman 2010). Therefore, I control for a state’s population in my model.
Reference Moncrief, Niemi and PowellMoncrief, Niemi, and Powell (2008) suggest that an increase in chamber turnover might result in more power being delegated to the Speaker. Specifically, a sudden rush of new and inexperienced members might make lawmaking more difficult for the chamber. Thus, with an increase in chamber turnover, there might be an increase in legislative power delegated to the Speaker. My measure of legislative turnover is derived from The Book of the States, and higher turnover indicates more freshmen serving within the chamber.
Reference ClucasClucas (2001) finds that party competition can influence the number of tools delegated to the state Speaker. I use the folded Ranney index to account for the level of partisan competition within a state. Footnote 19 This measure incorporates the partisan distribution of legislative seats and the incumbent governor’s average vote share in the most recent election. The folded Ranney index also captures the duration of partisan control for each state. Higher scores indicate increasing party competition within a state. I also include an indicator that control for the party system within a state (Reference Battista and RichmanBattista and Richman 2011). Reference Hedlund, Coombs, Martorano and HammHedlund et al. (2009) find that majority party size influences the organization within a legislative chamber. I measure this variable as the percentage of seats held by the majority party at the start of the session.
There are several other state institutional features that could influence the members’ decision to delegate power to their Speaker. For instance, I include Reference SquireSquire’s (2007) measure of legislative professionalism as suggested by Reference ClucasClucas (2007) and Reference RichmanRichman (2010). Squire’s index combines each legislature’s amount of salary, legislative staff, and length of session. More professionalized legislatures are associated with more powerful Speakers. The higher salaries, associated with professionalized legislatures, tend to increase the value of the seats to the members. Majority party members in professionalized legislatures might try to retain their valuable seats by relying heavily on their leadership and delegate tools to their Speaker in an effort to protect their party brand and to help their reelection efforts.
Citizen initiatives are associated with states facing several difficulties implementing policy (Reference LewisLewis 2012). Legislators may be forced to delegate more power to the leadership to overcome this institutional hurdle. The citizen initiative variable accounts for whether a state allows either constitutional or statutory citizen initiatives. For this variable, I use the continuous measure developed by Reference Bowler and DonovanBowler and Donovan (2004).
I also control for the state’s ideology because left-leaning states have been shown to expand the role of state government (Reference Alt and LowryAlt and Lowry 1994). To implement their policy agendas, liberal states may need to delegate more tasks to their leadership. I control for the preferences of the state with Reference Berry, Fording, Ringquist, Hanson and KlarnerBerry et al.’s (2010) state ideology scores. Higher scores indicate a state is more liberal.
Finally, the amount of power delegated to a Speaker may change over time. Therefore, I include a linear time trend. I also include a random effect for each state. I have also estimated a model with year and state fixed effects and the results are similar to those presented here (Supplemental Appendix A).
Findings
Given that the dependent variable is continuous, I estimate the coefficients using linear regression with state random effects. As a robustness test, I have estimated coefficients for a Tobit model and the results are consistent to those presented here (Supplemental Appendix B). The estimates are presented in Table 2. As expected, the legislative term limit variable is positively signed and statistically significant. This indicates that states with implemented legislative term limits are associated with a Speaker with more tools to help him or her to control the legislative process.
Table 2. State House Speaker Power—Testing Term Limits Hypothesis, 1981–2015.
Note. The models are estimated using linear regression with random effects for each state. The second column reflects the percentage change in the predicted Speaker power when shifting from the minimum value to the maximum value for the corresponding variable of interest. AIC = Akaike information criterion.
* p ≤ .05 (all one-tailed tests).
Substantively, the results from Table 2 can also be translated into percentage changes in the predicted Speaker power for each significant independent variable. To calculate these changes, I vary each corresponding variable from its minimum to maximum value, holding all other variables constant. Footnote 20 The last column in Table 2 presents the changes in Speaker power. According to the data, the states with the most stringent term limit laws tend to delegate around 11% more tools to their Speaker than states without term limits, all else being equal. Footnote 21 The legislatures with the most partisan competition in the United States tend to delegate to their Speaker 6% more authority than the least competitive legislatures. Chamber size also appears to have an impact on Speaker power. The largest chambers tend to delegate 46% more tools to their Speaker compared to the smallest chambers. Finally, the results suggest that the average amount of power delegated to a Speaker declined around 10% from 1981 to 2015. The remaining explanatory variables are not statistically significant at conventional levels.
It is important to highlight that the implementation of term limits did not radically change a legislature’s rules and procedures as indicated by the moderately sized substantive effects. For instance, after the implementation of term limits, it would be highly unlikely that a Speaker with little or no institutional tools would be delegated complete unilateral control over the entire lawmaking process. The moderate change in Speaker power is not surprising given that previous research shows that state legislative rules and procedures tend to change incrementally over time (Reference KingKing 2000; Reference MartoranoMartorano 2004; Reference Squire and HammSquire and Hamm 2005). However, it should be noted that the effect of legislative term limits on the chamber’s rules and procedure is similar to that of other prominent institutional features as demonstrated by Table 2. Furthermore, these changes to a chamber’s procedures can sometimes have a dramatic effect on policy outcomes as illustrated by California’s S.B. 562 and the decision not to adopt a single-payer health care system.
Robustness Check: Alternative Models
It is important that my findings remain robust to various model specifications and alternative measurements. Furthermore, it is vital that I provide other empirical tests that can help to further establish the link between implemented term limits and the rules within the lower chamber. In this section, I present five robustness tests with the results displayed in Table 3. For the most part, I use the same variables and methods as described in the prior section.
Table 3. Alternative Models.
Note. Models 2, 3, 5, and 6 are estimated using linear regression with a random effect for each state. Model 4 includes a lagged dependent variable and robust standard errors. Model 2 is estimated using Reference Sarbaugh-ThompsonSarbaugh-Thompson (2010) measure of term limits. Model 3 is estimated using the traditional dichotomous measure of term limits. Model 4 includes a lagged dependent variable. Model 5 includes only states that implemented term limits. Model 6 is a placebo test of all states between 1981 and 1990. AIC = Akaike information criterion.
* p ≤ .05 (all one-tailed tests).
First, Reference Sarbaugh-ThompsonSarbaugh-Thompson (2010) has created alternative measurements of legislative term limits. I employ her measure that calculates the ratio of the turnover rate within a legislature due to the implementation of term limits to the turnover rate from the prior decade. This measure is weighted relative to the size the legislature. Higher values indicate a more stringent term limit law. The results are presented in Model 2. It appears that the term limit variable is still statistically significant, and the coefficient is positively signed. Second, I employ the traditional measure of term limits—a dichotomous variable. The results are reported in the Model 3. Footnote 22 With this measure, I, again, find that implemented legislative term limits correspond to an increase in Speaker power.
Third, there are two reasons I should consider estimating a model with a lagged dependent variable. First, it is strongly possible that past tools delegated to the Speaker could influence the current level of power that is delegated to the Speaker. Second, it is possible that term limits were implemented in states with already powerful Speakers. In other words, I need to ensure that my results capture the change in Speaker power after the implementation of term limits. Including a lagged dependent variable allows me to estimate the change in my dependent variable (Reference MarkusMarkus 1979). Thus, as a robustness check, I have estimated a model with a lagged dependent variable, and it is labeled as Speaker Powert- 1. The results are displayed in Model 4. Even with the inclusion of a lagged dependent variable, the legislative term limits coefficient is positively signed and statistically significant.
Fourth, one concern with my analysis is that, for some reason, Speakers in term-limited chambers could have been powerful before the implementation of term limits and the amount of power delegated to the Speaker did not change after this new policy was carried out. In other words, the implementation of term limits, in itself, did not affect the amount of power delegated to the chamber’s leadership. Therefore, as a robustness check, I have estimated a model that only includes the states that have implemented legislative term limits. This model more directly tests the relationship between term limits and Speaker power. In addition, this analysis compares the amount of power a Speaker has over the lawmaking process before and after the implementation of term limits within a state. I use the Reference Baker and HedgeBaker and Hedge (2013) measure of term limits for this analysis. The results are reported in Model 5. Again, the legislative term limits variable is positively signed and statistically significant. This indicates that chambers with term limits saw an increase in the delegation of resources to the Speaker after the policy went into effect.
Finally, I adopt an empirical test that is similar to a placebo test (see Reference Miller, Nicholson-Crotty and Nicholson-CrottyMiller, Nicholson-Crotty, and Nicholson-Crotty 2018, 582). This test explores whether an unobserved state factor which happens to be associated with the implementation of term limits may be driving my empirical findings. For this test, the period of study is restricted to all state legislative sessions that occurred before 1990 because the first state to adopt term limits was in 1990. Furthermore, I create a new term limits indicator: states that eventually adopted and implemented term limits are coded as one and those that never did are coded as zero. If this new term limit variable is positively signed and statistically significant, then this would suggest that an unmeasured state characteristic that is correlated with the implementation of term limits could be causing the chamber to delegate more power to their Speaker, rather than the term limits themselves. The results are displayed in Model 6. The term limit variable is not statistically significant. This placebo test suggests that it is probably not very likely that some other unmeasured state characteristic that is associated with the implementation of term limits is influencing my results.
Conclusion and Implications
In 2016, the Missouri legislature considered a highly divisive bill, S.B. 656, which sought to loosen gun regulations. In particular, the legislation would allow most Missourians to carry a concealed gun without a permit on training. The bill also expanded legal protections for Missouri citizens who use deadly force to defend themselves. The bill was highly controversial because several Democratic politicians argued that it would result in a sharp rise in gun violence (Reference O’NeilO’Neil 2016). Therefore, the Republican-controlled legislature had to employ several tactics to ensure the bill’s survival. The Speaker of the Missouri House of Representatives, Todd Richardson (R), had several tools at his disposal to aid the passage of the measure and this was noticed by several opponents of the bill. For instance, one Missouri legislator claimed that the chamber’s leadership denied the bill’s opponents from speaking against the legislation on the floor and the Speaker used the legislative staff to their political advantage (Reference NewmanNewman 2016). In addition, the Missouri Speaker was allowed to unilaterally appoint members to the committee who crafted the final draft of the legislation—a power obtained by the Speaker after the implementation of term limits (Reference MooneyMooney 2013). Eventually, the bill was signed into law along party-lines and over the governor’s veto. This example highlights how the power a Speaker wields can have significant consequences on policy and legislative behavior. Furthermore, the findings from my study suggest that legislative term limits might partially account for why the Missouri legislators have delegated so much lawmaking authority to their Speaker.
In this article, I explore the relationship between legislative terms limits and the number of tools delegated to the state Speaker. Specifically, building off Reference MooneyMooney’s (2012) theoretical framework, I suggest that legislative term limits should incentivize the chamber to delegate more tools to their leadership. Footnote 23 When examining the chamber’s rules in 49 states between 1981 and 2015 with Reference MooneyMooney’s (2013) Speaker power index, I do indeed find that the implementation of term limits is associated with an increase in the Speaker’s power. Importantly, these findings speak to the debate concerning term limits and their effect on power within the legislative branch and its leadership (Reference BowserBowser 2005; Reference MooneyMooney 2012). While conventional wisdom suggests that legislative term limits weaken and destabilize the legislative branch and its leadership, I find that they can sometimes empower the chamber’s leadership in terms of institutional tools delegated to the Speaker. The Speaker can use these tools to help overcome the collective action problems associated with legislative term limits (Reference MooneyMooney 2012).
Before concluding, a few caveats are in order. First, I am not asserting that term limits strengthen every aspect of the legislative branch and its leadership. Indeed, other legislative power structures, such as the committee system (Reference Cain, Wright, Kurtz, Cain and NiemiCain and Wright 2007) and the parties’ reelection apparatuses (Reference Apollonio and La RajaApollonio and La Raja 2006), are most likely weakened by the presence of legislative term limits. Rather, I claim these findings challenge the conventionally held view that the legislative term limits weaken most facets of the legislative branch and its leadership. My results cast doubt on this view by suggesting that at least the Speaker can acquire more lawmaking tools or pass chamber rules that will help him or her carry out their branch’s policy agenda.
Second, it should be noted that there are several dimensions of leadership power (Reference BattistaBattista 2011; Reference Battista and RichmanBattista and Richman 2011; Reference MooneyMooney 2012). In particular, Speakers might be able to use informal tools, knowledge, and their own personal skills to influence the lawmaking process. It is important to recognize that my results do not speak to this particular dimension of power. However, future scholars should more closely examine the impact of term limits on this particular issue. In addition, it is important to recognize that these findings are not in direct opposition to previous studies which conclude that term limits reduce the influence of the Speaker (e.g., Reference Carey, Niemi, Powell and MoncriefCarey et al. 2006; Reference Miller, Nicholson-Crotty and Nicholson-CrottyMiller, Nicholson-Crotty, and Nicholson-Crotty 2011). These findings only suggest that Speakers are delegated more formal authority and tools by the members. It is certainly possible that a Speaker who is delegated the tools to more easily craft legislation does not have the personal characteristics necessary to shape public policy.
Third, it is important to recognize that the Mooney index that I employ may not perfectly capture a Speaker’s power. For instance, the Mooney measure does not account for the Speaker’s ability to influence the floor agenda—a significant limitation. In addition, some of the Mooney index scores deviate from some scholarly accounts of state Speakers. For example, Reference Farmer, Little, Farmer, Mooney, Powell and GreenFarmer and Little (2007) argue that the Ohio State Speaker has a great deal of influence in the lawmaking process. However, the Mooney index would suggest that the Ohio Speaker has less power than the average state Speaker. I encourage future scholars to develop additional measures of Speaker power and explore its relationship with term limits. Footnote 24
Fourth, interpreting chamber rules and procedures is a complicated and subjective task. It is certainly possible that The Book of the States, the main source utilized in this analysis, may have misinterpreted a few chamber rules due to the complexity associated with the legislative process. Footnote 25 If the chamber rules are indeed systematically misinterpreted, then this might call into question the relationship between legislative term limits and Speaker power. Therefore, I encourage future scholars to examine alternative sources that document legislative procedures used in the states to verify the accuracy of this Speaker power measure.
I hope that my findings stimulate additional research on the effects of term limits on other aspects of legislative organization, moving this research in new directions. For example, it is unclear how term limits may impact committee power (see, Reference Bagashka and ClarkBagashka and Clark 2014). My findings combined with prior research (Reference Carey, Niemi, Powell and MoncriefCarey et al. 2006) would suggest that term limits should result in a less powerful committee system. The literature suggests that powerful Speakers tend to undermine the autonomy and independence of the committee system so they can enact a partisan agenda (Reference Aldrich, Rohde, Dodd and OppenheimerAldrich and Rohde 2001). However, no study, as of yet, has analyzed chamber rules concerning committee autonomy within all U.S. states. In addition, future research should explore the role fundraising has in influencing the level of power delegated to a Speaker. It is possible that Speakers use their fundraising connections as a mechanism to acquire more power. Rather than a collective action–based theoretical framework, this possible scenario would suggest a more entrepreneurial context. Thus, there is still a great deal to learn about the implications of term limits on legislative organization.
Author’s Note
Laine P. Shay is now affiliated with Texas A&M-Corpus Christi, Corpus Christi, TX,USA.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD
Laine P. Shay https://orcid.org/0000-0002-3480-6729
Supplemental Material
Supplemental material for this article is available online.
Author Biography
Laine P. Shay is an assistant professor in the Department of Social Sciences at Texas A&M-Corpus Christi. His research focuses on American Institutions, Legislative Politics, and State Politics.
