Electoral Incentives and Legislative Organization

Abstract

We investigate the relationship between electoral institutions and committee autonomy in the context of U.S. state legislatures. The distributive theories of legislative organization suggest that electoral rules that make personal reputations more important motivate legislators to decentralize power and enhance committee autonomy to be able to target particularistic goods to their local constituencies. We argue that the distributive theories have direct implications for the relationship between candidate selection procedures and committee autonomy. The need to reach out to a large number of voters and to amass significant financial resources in states with more inclusive candidate selection procedures such as the open primary makes representatives more dependent on special interests, which is conducive to legislative particularism and committee autonomy. We take advantage of the great variation across the American states to investigate the effects of candidate selection procedures, a factor neglected in the previous literature. Examining 24 state legislatures from 1955 to 1995, we find that the inclusiveness of the selectorate, or the body electing candidates, has a significant effect on committee autonomy with more inclusive primary elections leading to more autonomous committee systems. By contrast, however, term limits were not a significant predictor of committee autonomy. This contributes to our understanding of how legislators amend institutional arrangements to achieve their electoral goals.

Keywords

legislative politics legislative committees comparative legislatures legislative professionalism time series quantitative methods methodology

Introduction

In democracies, legislators must negotiate, cooperate, and work together to establish institutions that further their individual and collective goals. While parties seek to advance their policy agendas through the strategic structuring of institutions, individual members have incentives to create institutions that advance their goal of reelection. Mayhew’s (1974) seminal work argues that the electoral connection compels legislators not only to cultivate their personal reputation with constituents but also to design institutions in ways that enhance their ability to deliver particularistic benefits to their constituents. Congressional scholars argue that individual members’ incentives to cultivate a personal vote explain why members choose to endow committees with strong property rights that enable members to create policy changes and redistributive benefits for which they can claim credit back in the district (Katz and Sala 1996; Shepsle and Weingast 1987). More broadly, this body of research maintains that electoral systems in which legislators must appeal to a much broader selectorate produce legislative institutions where power is decentralized and committees are more autonomous.

Although earlier research sought to understand the centrality of committees in the U.S. Congress, much less work has been devoted to other institutional settings where the party’s influence over elections, legislative careerism, and the electoral connection vary considerably. This research seeks to expand our understanding of how electoral institutions influence committee autonomy through a comparative institutional analysis of U.S. state legislatures. The significant variation in electoral rules across the U.S. states provides a unique opportunity to examine how and under what conditions electoral rules influence legislative organization while holding constant confounding factors including the number of effective parties, social cleavages, and broader patterns of politics. We take advantage of this opportunity, for the first time, testing systematically the implications of the distributive theory of legislative organization in the context of committee autonomy in U.S. state legislatures and subject them to numerous robustness checks.

Our findings demonstrate some support that electoral rules shape legislative organization. Specifically, we find strong evidence that inclusiveness of the selectorate, a factor neglected in the previous literature, is associated with greater committee autonomy. Moreover, this effect is a much more powerful determinant of committee autonomy than factors studied more extensively in the previous literature such as institutionalized committee property rights or codified jurisdictions, which are insignificant once we include the inclusiveness of the selectorate in the models. State electoral practices influence members’ strategies for achieving reelection. In more open primary systems, the broader selectorate compels legislators to cultivate a personal vote and curry favor through particularism, which is facilitated by decentralizing power and expanding committee autonomy. Although the inclusiveness of the selectorate has a strong effect on committee autonomy, we do not find evidence of a linkage between implementation of term limits and the level of autonomy granted to committees. These results have broad implications for our understanding of institutional change in legislatures.

The Personal Vote and Committee Power

The importance of the electoral connection is a mainstay of U.S. congressional research. Elections serve as the primary mechanism of citizens’ popular control of government. Reelection-seeking members primarily seek votes through the collective reputation of their party (i.e., the party brand name) or through their own personal reputation. The extent to which legislators rely upon their party’s collective reputation versus their own personal image is largely a function of the electoral rules. Indeed, legislators’ strategies to increase their likelihood of reelection have been the subject of much scholarly attention. Mayhew’s (1974) seminal work, asserting that members seek to advance their electoral goals through the provision of particularistic benefits to constituents, has inspired an enormous amount of research, ranging from studies of the effect of constituency service on representation (Fenno 1978; Johannes 1984) to how advertising, credit-claiming, and position taking affect electoral prospects (e.g., Fiorina 1977).

The general conclusion of this body of literature is that the desire for reelection under electoral rules that require legislators to appeal to a broader selectorate encourages legislators to provide particularistic or pork-barrel rewards for which they could claim credit. While indirect evidence of the electoral connection is abundant, direct tests of the effect of electoral connection are rare. Constrained by the lack of variation in national-level electoral rules, a small number of works have focused on the impact of significant exogenous changes in national electoral rules such as the rise of primaries or the shift to the Australian ballot on legislative behavior. Most of these studies find that candidate-centered electoral rules are conducive to credit-claiming and legislative particularism. Meinke (2008) found that the direct election of U.S. senators was associated with greater sponsorship of constituency-oriented legislation when state partisanship was strong, a result in line with analysis of the period leading up to the Progressive reforms (Carson and Jenkins 2011). Wittrock et al. (2008) found that the shift to the Australian ballot (especially its office bloc version) in the American states was associated with legislative demand for more valuable committee assignments and pork-barrel projects. However, Schiller (2006) failed to identify significant differences in electoral careers and bill sponsorship patterns in the periods before and after direct elections (Schiller 2006).

Congressional scholars also have used Mayhew’s (1974) simplifying assumption that members of Congress are “single-minded seekers of reelection” in theories of institutional choice. That members of Congress select rules and institutions to further their electoral interests has become a tenet of contemporary congressional scholarship. Some of the most influential theories in American politics suggest that stronger incentives for incumbents to build personal votes explain the emergence of strong committees in Congress (Katz and Sala 1996; Shepsle and Weingast 1987).¹ According to Shepsle and Weingast (1987), representatives choose to organize the legislature not around strong parties but around strong committees that allow them to affect policies important to their constituents and to target pork-barrel projects to their districts. Katz and Sala (1996) investigate the consequences of the shift of the Australian ballot for committee tenure patterns and find that it led to the emergence of committee property rights—the norm of reappointing representatives to their same committee at the start of each Congress. Squire et al. (2004) broaden this analysis to the U.S. state legislatures, which were exposed to the same electoral innovation at approximately the same time. However, in contrast to Katz and Sala (1996), they do not find evidence of the emergence of committee property rights in U.S. state legislatures following the introduction of the Australian ballot.

In contrast to the distributive theories of congressional organization, the informational theory of legislative organization (Krehbiel 1990; 1991) argues that the role of committees is to provide specialization, reducing uncertainty about policy outcomes and making the legislative process more efficient. Under the informational theory, committee autonomy depends on the extent to which the preferences of committee median are aligned with the preferences of the chamber median. Partisan theories of legislative organization instead argue that the majority party leadership uses its institutional prerogatives to advance the party’s policy agenda (Cox and McCubbins 2005). The majority leadership controls access to committees, stacking them with members who will protect the party’s collective reputation and facilitate the passage of the party’s policy agenda. As the relative strength of the majority party increases, the party leadership becomes more willing to provide committees with more autonomy as this does not threaten the dominance of the majority party (Binder 1997; Cox and McCubbins 1993; Dion 1997), conditional on the ideological homogeneity of the majority party (Cooper and Brady 1981; Cox and McCubbins 1997). Martorano (2006) compares the explanatory power of the distributive, informational, and partisan theories in the context of state legislatures in 24 states from 1955 to 1995. She finds that committee strong property rights or codified committee jurisdictions are not positively correlated with committee system autonomy, which contradicts the implications of the distributive theory.

The broader implication of distributive theory is that candidate-centered electoral systems should result in legislatures with stronger committees. Due to the lack of variation in electoral systems, previous direct tests of the electoral connection in the context of institutional choice in the U.S. Congress have focused on discrete electoral reforms such as the shift to the Australian ballot. By contrast, comparativists interested in the electoral connection have taken advantage of the significant variation in electoral rules and candidate selection procedures across countries to test the empirical implications of theories of institutional choice in the U.S. Congress. For example, Martin (2011) investigates whether more candidate-centered electoral rules are associated with stronger committees in 39 democratic legislatures. He finds that more candidate-centered electoral rules are associated with stronger committees but only when legislators cultivate strong personal support through legislative particularism (pork) and not when they instead rely on extra-legislative constituency service.

Theoretical Expectations

The American states allow us to hold confounding factors such as social cleavages, party systems, and broader patterns of politics constant. While the electoral system for legislative elections is held constant across states, there is significant cross-state variation in candidate selection procedures and term limits. State legislatures offer contrasts to Congress on a couple of key dimensions, including term limits in many states and different kinds of candidate selection methods (e.g., open or closed primaries). We argue that these features of the electoral environment motivate legislators to cultivate a personal vote. The distributive theory of legislative organization would imply that such electoral incentives would also affect committee autonomy. Somewhat surprisingly, these implications of the distributive theory of institutional choice have not been tested in the context of U.S. state legislature. We contribute to the literature on legislative organization by testing systematically and rigorously the implications of the distributive theory of legislative organization.

Inclusiveness of the Selectorate and Legislative Particularism

We draw on recent work in comparative politics to investigate the implications of candidate selection procedures for committee autonomy. Rahat and Hazan (2001) argue that candidate selection methods affect the incentives that individual politicians face in important ways. They classify selectorates, the body that selects candidates, on a continuum according to their exclusiveness or inclusiveness. In the most inclusive selectorates, the entire electorate can vote for a candidate. At the other extreme are the most exclusive selectorates where a candidate is nominated by one party leader. While all types of American primaries are to some extent inclusive procedures located in the “electorate zone” (i.e., closer to the inclusive end of the continuum), there is variation in the degree of their inclusiveness: the “blanket primary” where every registered voter can vote for candidates from any party would be located toward the inclusive end of the electorate zone. By contrast, American “closed primaries,” which require voters to register their party affiliation before or on the day of the primaries, are located toward the exclusive end of the “electorate zone” (see Rahat and Hazan 2001, 301–302). Because enrollment procedures vary widely across states, classifying states by primary type is more complicated than the simple open versus closed dichotomy (Carr and Scott 1984; Finkel and Scarrow 1985; Jewell 1981). In a number of states, party enrollment is required at the time of registration (Arizona, California, Connecticut, Delaware, Florida, Kentucky, Maryland, Nebraska, Nevada, New Mexico, New York, North Carolina, Oregon, Oklahoma, Pennsylvania, South Dakota, and West Virginia). By contrast, in Colorado, Kansas, Maine, New Jersey, and Rhode Island, enrollment does not occur until the citizen votes at the primary. How inclusive/exclusive primaries are depends on the specific restrictions in state laws (Kolodny and Katz 1992; Ranney 1981).²

According to Rahat and Hazan (2001), more inclusive candidate selection procedures motivate individual candidates to appeal directly to the voters, circumventing party leaders. Assuming that party leaders care about maintaining seats in the legislature, inclusive candidate selection procedures may not only undermine party discipline as party leaders “allow” cross-pressured members to vote with their constituency rather than party (when the two diverge) but also contribute to legislative particularism. Where candidate selection procedures are more inclusive, candidates rely less upon being loyal team players, and instead they rely more upon currying favor with their inclusive selectorates. Enlarging the selectorate not only makes the individual support base of a candidate more important but also leads to a need to increase financial resources to be able to reach more voters. This may empower local interest groups and produce slates of candidates with ties to special interests (Rahat and Hazan 2001, 313–14). Rahat and Hazan (2001, 314) find that electoral reforms in Israel that enlarged the selectorate were associated with less cohesive legislative parties. Scholars are only beginning to address the effects of the inclusiveness of the selectorate. Here, we take advantage of the great institutional variation across the American states to test these novel theories and their implications for institutional choice. If in fact more inclusive selectorates are associated with stronger incentives for legislators to initiate and support distributive policies, the distributive theory of legislative organization would imply that more inclusive selectorates should be associated with more autonomous committees.

While in the U.S. system there is variation across states in primary procedures, all primary elections are generally inclusive procedures. Even in closed primaries, party elites often have little control over who runs under the party label. For instance, in many cases, Tea Party candidates have sought office despite the discouragement of national, state, and local Republican Party leaders and have appealed to voters directly and worked with interest groups, such as FreedomWorks. This suggests that variation in primary procedures and the inclusiveness of the selectorate, in particular, might not lead to significant differences in the political importance of interest group support and legislative particularism across states. Drawing on the previous literature on primary types, we propose an alternative causal mechanism, which suggests a positive relationship between the “openness” of primary types and the level of legislative particularism. A number of studies have examined the effects of primary type on candidate ideology (Brady et al. 2007; Gerber and Morton 1998; Jewell 1984; Westley et al. 2004). Most relevant to our analysis, Gerber and Morton (1998) focus on the effects of primary closedness on candidate ideology in the context of nonpresidential primaries. Gerber and Morton hypothesize that candidates nominated in closed primaries are expected to be more extreme than candidates in other primary systems as a result of two effects. According to theoretical models in the “party elites” literature (Aranson and Ordeshook 1972; Coleman 1971; 1972; Wittman 1977; 1983; 1991), candidates are chosen by policy-motivated as opposed to office-seeking party elites. Given the uncertainty about the election outcome, the expected utility of the party’s median voter may be maximized by a policy that diverges significantly from the general election median. As a result, under closed primaries, candidates’ positions in closed primaries will diverge from the general election median.³

The expectation of divergence between the closed primary election median and the general election median voter is consistent with the findings of a number of empirical studies, which suggest that voters that participate in closed primary elections tend to have strong partisan ties and extreme partisan preferences (Abramowitz, Rapoport, and Stone 1991; Beck and Sorauf 1992; Buel and Jackson 1991). Gerber and Morton (1998, 311) further hypothesize that in open primary elections, the likelihood of electing moderate candidates is higher compared to closed primary elections due to the greater likelihood of “sincere cross-over voting”—voters sincerely voting from their most preferred candidate in the other party. In closed primary elections, “sincere crossover voting” is substantially more costly because voters must decide to vote in a given party’s primary before the election and register as a member of that party.⁴ This implies that in more open primary procedures, candidates from both parties will converge to the general electorate median. We hypothesize that the inability of candidates in states with open primary procedures to distinguish themselves ideologically due to centripetal incentives may motivate them to rely more heavily on pork, particularistic policies, and constituency service to improve their electoral prospects. This tendency may be reinforced by greater competition in states with more open primaries where the competition from a wider range of groups increases competition (Hogan 2003, 109). This implies that more open and inclusive candidate selection primary procedures should be associated with greater committee autonomy.

Hypothesis 1: Committee autonomy is greater in states that employ more inclusive primary procedures.

Term Limits and Legislative Particularism

The assertion that legislators prefer strong committees that would enable them to target pork-barrel projects to their constituencies and to affect policies important to them rests on the assumption that legislators are in fact seeking reelection. If, however, the electoral connection is severed through the imposition of term limits, the incentives to build a strong personal vote and, by implication, to create strong committees would be reduced, regardless the specific character of the broader electoral system. Several studies have found that term limits reduce legislative particularism (Carey, Niemi, and Powell 1998; Herron and Shotts 2006). Despite the dearth of research at the state level, there are several studies that address this question through systematic analysis of retirees in Congress. The expectation is that because these subsets of members lack the incentives that reelection holds, their performance and behavior during their last term would reflect how legislators would behave when the electoral connection is severed (similar to those restricted from reelection due to term limits). Herrick, Moore, and Hibbing (1994) find that overall roll-call attendance for retirees does decline and attention to constituency service is depressed. Other studies also find evidence of shirking among legislators or a decrease roll-call participation rates (Lott 1990). Examining state legislators during 1999–2000, Wright (2007) identifies significant difference in roll-call participation rates for those retiring only in legislatures with very high workloads.

Comparative studies of term limits and legislative particularism motivate a reconsideration of the expectation that the imposition of term limits should be associated with weaker incentives to provide pork. Carey (1996) examines bill sponsorships patterns in Costa Rica and Venezuela and finds evidence of widespread legislative particularism among Costa Rican legislators who lack the motivation of reelection due to term limits as well as the electoral incentives that tend to produce particularistic behavior. Similarly, Uslaner (1985) finds that even in a country like Israel, where electoral rules are not candidate centered (closed party list in a single national district), there is considerable effort by members of the Knesset to cultivate a personal vote through constituency service and casework. Carey (1996) argues that members engage in legislative particularism even if they cannot reap the benefit of reelection because Costa Rican legislators are motivated by post-legislative career ambitions—legislators expect that deputies most successful in accruing pork are most likely to receive patronage appointments after their legislative careers. While the points Carey raises are difficult to test on a single case, we believe that institutional variation in term limits in the American state legislatures may help us understand why we might see particularism and, by implication, stronger committees, even absent the incentives to cultivate a personal vote or to seek reelection. Taking advantage of the institutional variation of term limits in the American states, we test these competing theories in the context of American state legislatures.

Hypothesis 2a: The autonomy of legislative committees is greater in states without term limitations.

Hypothesis 2b: The autonomy of legislative committees does not vary according to term limitations.

Data and Method

We constructed a dataset of electoral rules and other legislative features for 24 lower state legislative houses for the 1955–95 period. Data on electoral rules were collected by the authors. These 24 states possess variation on our main variables of interest: type of candidate selection procedure and term limits. Data on candidate selection procedures were gathered from both the Secretary of State Web sites for each state and the National Conference of State Legislatures (NCSL).⁵ We operationalize the degree of inclusiveness or exclusiveness of the selectorate in the following way. We create a contextual variable for the inclusiveness of the selectorate that ranges from 0 (most exclusive) to 4 (most inclusive). Although all of the states in our sample use a primary to nominate candidates, these primaries vary a great deal in terms of their inclusiveness of voters. We present information on type of primaries used in the states in our sample in Figure 1. Open primaries used in Connecticut, Delaware, Illinois, Michigan, and Minnesota are fairly inclusive procedures (“Inclusiveness” = 3). At the other end of the continuum are closed primaries used in Delaware, Florida, Kansas, New Jersey, Nevada, Ohio, Oklahoma, and Pennsylvania (“Inclusiveness” = 0). Partially open and partially closed primaries used in the rest of the states are procedures with intermediate levels of inclusiveness, with partially open primaries closer to the inclusive end of the continuum (“Inclusiveness” = 2 and 1, respectively).

Figure 1.

Primary type by state.

As Figure 1 demonstrates, states that used only one type of candidate selection method in the period under consideration, such as Michigan or Minnesota, are the exception rather than the rule. Ten of the 25 states (Arizona, California, Florida, Iowa, Illinois, Kentucky, Pennsylvania, Kansas, Kentucky, Virginia) used all types of primaries for at least one year. In many states, however, experimentation with different types of primaries was short-lived: only 2 of the 10 states (Illinois and Kentucky) used all types of primaries for three or more consecutive elections. Overall, there is significant variation in primary types within and across states. While the outcome variable committee autonomy does not vary with as much frequency over the period of study, in all states with the exception of Arizona, Connecticut, Kansas, Montana, and North Carolina, committee autonomy varies sufficiently over time, and in some states such as Delaware, Iowa, Illinois, Indiana, South Carolina, West Virginia, and Wyoming, committee autonomy varies significantly (Martorano 2006). We collected data on which states adopted legislative term limits. During the time frame of this study, six states had adopted term limits (see Table 1). Term Limits is a dummy variable coded 1 if term limits have been enacted and 0 otherwise.

Table 1.

Term Limits, 1955–95.

State	Term limits	Year enacted
Arizona	Consecutive ban	1992
California	Lifetime ban	1990
Connecticut	No term limits
Delaware	No term limits
Florida	Consecutive ban	1992
Illinois	No term limits
Indiana	No term limits
Iowa	No term limits
Kansas	No term limits
Kentucky	No term limits
Michigan	Lifetime ban	1992
Minnesota	No term limits
Montana	No term limits
North Carolina	No term limits
Ohio	Consecutive ban	1992
Pennsylvania	No term limits
Rhode Island	No term limits
South Carolina	No term limits
South Dakota	Consecutive ban	1992
Vermont	No term limits
Virginia	No term limits
Washington	No term limits
West Virginia	No term limits
Wyoming	No term limits

Control Variables

It is important to hold constant the confounding factors that could also affect committee autonomy. A major competing theory that provides an alternative explanation for committee autonomy is the informational theory of legislative organization. The informational theory implies that greater issue complexity is associated with greater committee autonomy. Policy Complexity is operationalized as follows:

Policy Complexity = \frac{Number of bills introduced × Average length per bill}{Legislative days in session} .

Martorano (2006) measures average length per bill by counting the number of pages in the session laws of the previous session, multiplying by the number of lines per page, and dividing by the total number of enactments. To facilitate interpretation, the measure is divided by 100.

Another competing theory of legislative organization is the partisan theory. According to this theory, committees serve the needs of the majority party. This implies that the greater the relative size of the majority party, the greater committee autonomy.⁶ Conversely, when a majority party starts losing its legislative advantage, it tends to centralize power to restrict minority rights (Binder 1997; Dion 1997). We operationalize Majority Party as the percentage of seats held by the majority party.

We also created a variable for Reform Period, which is coded 1 for the years 1965–79. This variable accounts for three relevant historical events. The first event is the mandatory reapportionment of legislative districts in 1962 and 1964, which, in contrast to later reapportionments, took place after the “one-man, one-vote” decision (the Baker v. Carr in 1962 and Reynolds v. Sims 1964 decisions) and for the first time in many state legislatures. The resulting increase in chamber size and the influx of new members may have increased committee system autonomy. As a result of the mandatory reapportionment, in many state legislatures, there was a shift in power from legislators who represented rural districts to those that represented urban districts. Committee members and chairmen may have used the power of congressional committees to thwart policy changes sought by the growing percentage of metropolitan representatives, such as the reallocation of government benefits from rural to metropolitan voters. In addition, in many chambers, there was a shift in power (conceptualized as the percentage of members holding seats) from legislators who represented rural districts to those who represented urban districts, potentially affecting both member and legislative chamber goals. The second event the variable Reform Period takes into account is the publication of The Sometimes Governments (Citizens Conference on State Legislatures [CCSL] 1971), which stipulated that virtually every state legislature had to adopt structures and procedures that strengthened committee systems. The Watergate scandal (1972–73) led to a movement to pass government Sunshine Laws, which would contribute to greater openness of the committee stage of the legislative process. In addition, congressional reforms of the committee system (Rohde 1991) were followed by a push for reforms at the level of state legislatures that would empower all legislative subunits (see Martorano 2006, 214). Separately, we also take into account another factor that took place at the end of the reform period and into the 1980s, namely, the adoption of sunset legislation in many states. The adoption of sunset statutes led to a flood of government programs that required periodic renewals and significantly increased the workload for state legislatures. This resulted in adjustments of committee rules, which might have affected committee autonomy. Due to this overload, many state legislatures subsequently repealed the sunset legislation, which likely resulted in new adjustments to committee organization. Sunset Law is a binary variable coded 1 for the years when sunset legislation was in place.

We also control for legislative professionalism. Legislative professionalism might affect committee autonomy in two different ways. Several studies have found that legislative professionalism affects the procedures adopted by state legislative bodies (Freeman and Hedlund 1993; Squire 1993; 1997; Thompson 1986). More professional legislatures have been shown to have greater independence from other state government officials such as governors and, as a result, are more autonomous. We expect that professional legislatures will have more autonomous committee systems that would allow them to have a maximum impact on policy and to facilitate legislators’ career ambitions. More professional and higher-paying legislatures might also heighten legislators’ motivation to get reelected (Gamm and Kousser 2010, 156). The desire to be reelected is likely to motivate representatives to empower legislative committees, which would allow them to cater to the needs of local constituents and contribute to their reelection efforts. We use King’s (2000) measure of legislative professionalism. In addition, we include control variables for New Speaker and Tenure of the Previous Speaker. Changes in the level of committee autonomy are more likely following the election of a new speaker (Swift 1997), especially if the previous speaker held the position for a long time. New Speaker is a dummy variable that indicates whether the speaker is new for the current session. Tenure of the Previous Speaker counts the number of sessions the previous speaker held the position. To account for partisan differences in positions on committee autonomy, we include a dummy variable Democratic Control of Chamber coded 1 if the chamber is controlled by the Democratic Party.

Finally, Martin’s (2011) research alerts us to the importance of considering how legislators are able to cultivate a personal vote. In some areas of the world, legislators’ ability to provide fiscal particularism through earmarks is constrained when the executive has sole authority to introduce measures involving spending. Martin argues that, in cases where fiscal particularism is unavailable to legislators, they lack the incentives to structure the legislature around committees. In the 24 state legislatures that we examine from 1955 to 1995, both fiscal particularism and extra-legislative constituency service are available to members. Therefore, we do not need to include controls for fiscal particularism as Martin does.

Measuring Committee System Autonomy

We use Martorano’s (2006) operationalization of committee system autonomy. To measure committee system autonomy, Martorano examines the extent to which committees are able to receive, screen, shape, and affect the passage of legislation. If the committee system is truly autonomous, the rules would stipulate that all legislation has to be referred to a committee for consideration before final passage. This committee right, however, is meaningless, unless committees have the right to screen out legislation that is considered bad public policy or that goes against member, party, chamber, or constituency preferences. Committee autonomy is also greater if bills referred to the committees do not have to be considered and/or reported back to the floor, there are no deadlines for committee action, and it is not difficult or impossible to withdraw legislation from committee consideration.

Committee autonomy is also enhanced if committees are able to shape legislation by sponsoring bills, offering committee amendments before consideration on the floor, or offer substitute legislation. Committee autonomy is undermined if committee decisions can be easily canceled out easily during consideration on the legislative floor and final passage (for more details, see Martorano 2006). The dependent variable Committee System Autonomy is an additive index based on the aforementioned committee rights and procedures, with negative scores indicating low committee autonomy and positive scores indicating high committee autonomy. The individual scores for each component, which range from −1 to 1, were summed. Thus, the composite index ranges from −11 to 11. For the legislatures in our sample, committee system autonomy ranges from −5 to 5 and has a mean of −0.117 and a standard deviation of 2.26.

Empirical Analysis

We present in Table 2 the main results of different models using the time-series cross-sectional (TSCS) dataset. A Wooldridge test indicated that a correction for autocorrelation was necessary. We present fixed and random-effects models corrected for first-order serial correlation.⁷ The Breusch–Pagan test for random effects rejected the null hypothesis of no state random effects, which warrants the panel estimation adopted here. A Sargan–Hansen test of over-identifying restrictions suggests that a random-effects model is inappropriate.⁸ Explanatory variables that do not vary within states must be excluded in a fixed-effects framework; however, this is not an issue in our analysis. Our main interest is in the electoral system variables, which, like the rest of the explanatory variables, vary significantly over time.⁹

Table 2.

Electoral Incentives and Committee Autonomy, 1955–95.

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Inclusiveness of the selectorate	0.34236*** (0.04197)	0.33414*** (0.04151)	0.32721*** (0.04184)	.31701*** (0.04173)	0.92345*** (0.11222)	0.77967*** (0.11596)
Term limits	0.19266 (0.27881)	0.02711 (0.27809)	0.02664 (0.27804)	0.02405* (0.27616)	0.16946 (0.51145)	0.15147 (0.48936)
Reform period	0.03841 (0.10213)	0.06676 (0.10305)	0.0666 (0.10312)	0.05307 (0.10255)	0.25362 (0.19354)	0.31052* (0.18701)
Democratic control of chamber	0.063 (0.07718)	0.13735* (0.08036)	0.13989* (0.08059)	0.12738 (0.08016)	0.10823 (0.13901)	0.024605 (0.13902)
Legislative professionalism	1.54374 (1.03101)	1.31637 (1.01304)	1.29811 (1.01392)	1.147693 (1.00906)	0.17351 (1.96082)	−1.47207 (1.9149)
New Speaker	0.01531 (0.05815)	0.02685 (0.05883)	0.02192 (0.06027)	0.015449 (0.05989)	−0.06625 (0.12485)	−0.10009 (0.12317)
Tenure of previous speaker	0.00285 (0.02631)	−0.0061 (0.02682)	−0.00635 (0.02709)	−0.01141 (0.02698)	0.05103 (0.06265)	0.02899 (0.06066)
Policy complexity		0.00008 (0.00002)	0.00008*** (0.00002)	0.00008*** (0.00002)	0.00011** (0.00005)	0.00014** (0.00005)
Majority party strength		−0.00619 (0.00408)	−0.00548 (0.00411)	−0.00428 (0.00411)	−0.00671 0.00751	−0.00136 (0.00751)
Property rights			−0.30444 (0.21208)	−0.35553* (0.21159)	0.45778 (0.43715)	−0.11245 (0.42748)
Codified committee jurisdictions			−0.07088 (0.35127)	−0.0522 (0.34888)	−6.14209 (4.28258)	−6.68821 (4.42288)
Sunset Law				−0.36534** (0.14287)		−0.96328*** (0.25779)
Constant	−1.27761*** (0.05824)	−1.12432 (0.07244)	−1.1579*** (0.07327)	−1.0334*** (0.073)	−1.95047*** (0.14349)	−1.46203*** (0.17671)
Number of observations	458	439	439	439	143	143
Number of states	24	24	24	24	8	8
Average number of obs. per state	19.1	18.3	18.3	18.3	17.9	17.9
F statistic	3.67***	4.07***	4.06***	8.11***	9.40***	10.69***
Within R²	0.1389	0.1770	0.1815	0.1946	0.4546	0.5105
Between R²	0.4994	0.3157	0.2655	0.2398	0.7542	0.5132
Overall R²	0.3834	0.3010	0.2726	0.2622	0.6286	0.5733

***

p<0.001,**p<0.01,*p<0.05

In models 5 and 6, we restrict the sample only to cases where there is a change to a primary type in one direction (from more closed to more open or vice versa) that is sustained over time.¹⁰ This restriction reduces the number of observations to 143. We find that, in line with our theoretical expectations, more inclusive candidate selection procedures are associated with greater committee autonomy. This is consistent with the claim that the effort to reach out to more voters in states with inclusive candidate selection procedures and to amass greater financial resources motivates legislators to delegate more autonomy to legislative committees, which would give them greater access to particularistic goods they could use to target to their local constituents. By contrast, the adoption of term limits does not have a significant effect on committee autonomy. Our theoretical priors about the effect of term limits on committee autonomy were mixed: while some theoretical arguments imply a positive correlation between the adoption of term limits and committee autonomy, others predict the opposite. The absence of a significant effect of term limits could be due to the absence of either of these effects or to these two competing effects canceling each other out. However, further research is necessary to distinguish between these two explanations. We find that Democratic Control of the Chamber is associated with greater committee autonomy. We also find that the adoption of sunset laws reduced committee autonomy—Sunset Law has a negative and significant coefficient in models 4 and 6. In the restricted sample, Inclusiveness of the Selectorate again has a positive and statistically significant coefficient, larger in magnitude. The other results are very similar to those from models 1 to 4, with the exception of the Reform variable, which has a positive and statistically significant coefficient in model 6.

We find that Policy Complexity has a positive and significant coefficient, which is consistent with the implications of the informational theory of legislative organization. Growing policy complexity and the need to maximize efficiency and reduce the uncertainty of the legislative process necessitates more autonomous committees. Previous research has found that strong committee property rights have a negative and significant effect on committee system autonomy, which goes against the implications of the distributive theory of legislative organization (Martorano 2006). We find that, similarly to her results, Codified Committee Jurisdictions does not have a significant coefficient in any of the models. However, in contrast to her findings, we find that Property Rights do not have a significant effect either. Our results show that once we take into account the inclusiveness of candidate selection procedures, committee property rights lose their explanatory power. While our findings cast doubt on the claim that the empirical pattern provides “damaging evidence for the distributive theory” (Martorano 2006, 221), we need to examine alternative explanations for the difference in our results. Given that Property Rights is an outcome variable, which is likely to be dependent on electoral incentives and, therefore, on electoral institutions, its inherent endogeneity may bias the coefficients. In other words, if Inclusiveness of the Selectorate affects committee system autonomy not only directly, but also indirectly through Committee Property Rights, Committee Property Rights would be an endogenous variable, which implies that our original model is inappropriate. The endogeneity of some of the explanatory variables violates the assumption of no correlation between the regressors and the error terms and renders ordinary least square (OLS) estimates biased and inconsistent.

To address this, we use a generalized method of moments (“difference GMM”) estimator suggested by Arellano and Bond (1991).¹¹ This estimator first differences the levels equations to remove the unit fixed effects and instruments the first-differenced predetermined and endogenous variables on lags of the levels variables sufficiently deep to be uncorrelated with the first-differenced error term. Because a large instrument set could overfit the predetermined and endogenous variables and bias parameter estimates toward their OLS parameters, we use only the second lags of the endogenous variables as instruments. We fit the model with robust standard errors consistent with panel-specific autocorrelation and heteroskedasticity in one-step estimation.¹² The GMM model assumes that the errors are not serially correlated, which can be tested by checking whether the first-differenced residuals exhibit second-order serial correlation. The results from the Arellano–Bond estimation are presented in Table 3. In models 7 and 9, Property Rights and Codified Committee Jurisdictions are treated as exogenous; in models 8 and 10, they are endogenized. According to the AR(2) test statistics presented in Table A1 of the online appendix, there is no serial correlation in levels, which is tested by checking for second-order correlation in first differences.

Table 3.

Electoral Incentives and Committee Autonomy, 1955–95.

	Model 7 Arellano–Bond	Model 8	Model 9	Model 10
Inclusiveness of the selectorate	0.52143*** (0.13694)	0.52491*** (0.13488)	0.52764*** (0.13774)	0.52743*** (0.13466)
Term limits	0.20685 (0.22326)	0.32625 (0.34697)	0.20978 (0.21942)	0.32899 (0.33243)
Reform period	−0.12483 (0.08346)	−0.11715 (0.12152)	−0.10518 (0.08131)	−0.11536 (0.12075)
Democratic control of chamber	0.23461 (0.14303)	0.25025* (0.13673)	0.22906 (0.14083)	0.25691* (0.13743)
Legislative professionalism	1.09398 (1.23551)	0.65953 (1.28616)	0.92032 (1.19971)	0.69942 (1.19688)
New speaker	−0.04659 (0.07647)	−0.05939 (0.07776)	−0.04811 (0.07484)	−0.06538 (0.07871)
Tenure of previous speaker	0.03309 (0.04535)	0.03013 (0.0472)	0.03419 (0.04273)	0.03111 (0.04463)
Policy complexity	0.00011 (0.00005)	0.00013** (0.00006)	0.00011** (0.00005)	0.00013** (0.00006)
Majority party strength	−0.00351 (0.00529)	−0.00365 (0.00509)	−0.00406 (0.00543)	−0.00407 (0.00509)
Property rights	−0.80169 (0.75345)	−0.59616 (0.62383)	−0.52418 (0.67272)	−0.45182 (0.59898)
Codified committee jurisdictions	−0.26163 (0.56235)	−0.58401 (0.54813)	−0.3145 (0.57884)	−0.69708 (0.58225)
Sunset law			−0.04445 (0.19363)	−0.04995 (0.17815)
Number of observations	392	392	392	392
Number of states	22	22	22	22
Average number of obs. per state	17.82	17.82	17.82	17.82
χ²	29.02**	54.43***	33.25***	109.31***
Arellano–Bond AR(2) test	−1.45	−1.38	−1.47	−1.41
p Value, AR(2) test	0.147	0.168	0.141	0.157

***

p<0.001,**p<0.01,*p<0.05

The results from the Arellano-Bond estimation are very similar to those from the fixed-effects models presented in Table 2. Our main variable of interest Inclusiveness of the Selectorate again has a positive and significant coefficient in models 7–10. The coefficient of Term Limits is statistically insignificant. The coefficient of Sunset Law loses its statistical significance in the Arellano–Bond models. Policy Complexity is significant in models 8–10. Similar to our previous findings, unlike Martorano (2006), we find that Property Rights has an insignificant effect on committee system autonomy. Our results suggest that there is no support for one of the implications of one of the empirical implications of the distributive theory, namely, that the desire for reelection is associated with the emergence of strong committee property rights in the context of U.S. state legislatures. This finding is consistent with Squire et al.’s (2004) finding that the shift to the Australian ballot in the context of state legislatures did not lead to the strong property rights in U.S. state legislatures, in contrast to the results of the same change in electoral rules in the U.S. Congress. Our results, however, suggest that electoral incentives affect committee system autonomy, albeit through a different causal mechanism. In particular, the need to gain the support of a large number of local constituents in states with more inclusive primary procedures such as the open primary incentivizes legislators to amass significant financial resources and cater to special interests to that end. This is consistent with the direct implications of the distributive theory, according to which, electoral incentives motivate legislators to support strong committees that would give them greater access to particularistic goods they could use to buy the support of local constituents.

Robustness Checks

We subject our results to a number of robustness checks. If legislative positions are not very desirable, candidates may not care about the personal vote. To address this, we control for voluntary turnover measured by the mean years of service for members from Squire (1988). Those results are presented in Table A1 (models 13–14) of the online appendix. Our results are robust to the inclusion of this variable. In models 15–16, we take into account the interaction between electoral rules and district competitiveness. Hogan (2003) finds that competitiveness varies greatly as a function of state- and district-level factors. Running in a competitive district under open primary rules and then in the general election may motivate a representative to seek more of a personal vote compared to a noncompetitive district. Using the Carsey et al. (2012) dataset, we include in the analysis two measures of district competitiveness: District Margin of Victory and Number of Candidates per District. The results are presented in Table A2 of the online appendix (models 15–16). Those variables do not achieve statistical significance; they do not affect our results in any important ways, substantively or significantly. We have less confidence in these results. Adding these variables restricts the number of states and years for which data are available and reduces the number of observations to 32.¹³

Our measure of the “openness” or inclusiveness of primary procedures assumes that the relationship between inclusiveness and committee autonomy is linear. However, our theoretical argument about the effect of inclusiveness on candidate ideology allows the possibility of a nonlinear relationship between primary type and candidate ideology.¹⁴ To investigate the possibility of a nonlinear relationship between “openness” and inclusiveness of the selectorate, we replace the variable Inclusiveness of the Selectorate with dummy variables for open and semi-closed primaries (closed primary is the baseline category; see models 17 and 18 in Table A3 of the online appendix). We do not find evidence of a nonlinear relationship. In all models, Open Primary has a positive and statistically significant effect on committee autonomy. Semi-open primary is statistically indistinguishable from the reference category. The absence of a nonlinear relationship is consistent with the prevalence of sincere crossover voting in open primaries; however, additional research is necessary to fully investigate this. In models 19 and 20 (Table A4 of the online appendix), we investigate the possibility that the inclusiveness of the selectorate explains the decision to codify property rights and committee jurisdictions or that inclusiveness of the selectorate, property rights, and codified committee jurisdictions could be influenced by the same factors as the dependent variable committee autonomy. In model 19, we endogenize Inclusiveness of the Selectorate. In model 20, we endogenize Inclusiveness of the Selectorate and Codified Committee Jurisdictions and Property Rights. This model specification does not affect the significance of our results in any important ways, substantively or statistically.

Conclusion

Previous congressional studies maintain a causal connection between electoral rules, the personal vote, and legislative organization. Comparative studies examining these claims in more diverse institutional settings have produced mixed evidence concerning the effect of electoral rules on legislative organization, however. We examine the effect of electoral rules on committee system autonomy in the U.S. states. We believe the states are especially well suited due to their significant variation in the inclusiveness of the selectorate, term limits, and institutional arrangements, yet confounding factors like the effective number of parties, constitutional design, and broader patterns of politics are held constant. We apply in a novel way insights from the recent comparative literature to investigate the effects of electoral incentives on legislative organization and examine the empirical support for the direct implications of the distributive theory of legislative organization. We subject our findings to a battery of robustness checks to consider alternative explanations and a variety of models that address the inherent endogeneity of some of the explanatory variables.

We find support that the inclusiveness of the selectorate has a significant effect on committee autonomy. In states where primary elections are more inclusive (i.e., not restricted by declared party membership), committee autonomy is much greater. This finding is robust across our different models. Electoral rules that shape legislators’ need to gain the support of a large number of local constituents (i.e., states with more inclusive primary procedures, such as the open primary) compel legislators to decentralize power and enhance committee autonomy to provide themselves with greater latitude to supply targeted, particularistic goods to constituents. This enhances legislators’ ability to cultivate a personal vote. Our research failed to show that the implementation of term limits in state legislatures had a significant effect on committee autonomy.

While this research assumes that committee autonomy enhances members’ ability to cultivate a personal vote, members may have different strategies for cultivating the personal vote. As Martin (2011) argues, some members may seek to supply pork to constituents (i.e., fiscal particularism), while others engage in extra-legislative constituency service to enhance their personal reputation with voters. When legislators rely upon fiscal particularism, we might expect legislators to expand committee autonomy; however, reliance on extra-legislative constituency service does not necessary imply that legislators will expand committee autonomy. Although in the U.S. states, fiscal particularism and extra-legislative constituent service are possible, future research may seek to examine the connection between committee autonomy and the types of activities legislators engage in to cultivate a personal vote. This would further elucidate the relationship between electoral incentives, institutional structure, and legislative behavior.

Footnotes

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.

Notes

Author Biographies

Tanya Bagashka is an Associate Professor in the Department of Political Science at the University of Houston. She received her PhD from the University of Rochester in 2008. She studies comparative institutions, corruption, and post-communist politics.

Jennifer Hayes Clark is an Associate Professor in the Department of Political Science at the University of Houston. She received her PhD from Indiana University in 2007. Her research examines state legislatures, political parties, and representation.

References

Abramowitz

Alan

Rapoport

Ronald

Stone

Walter

. 1991. “Up Close and Personal: The 1998 Iowa Caucuses and Presidential Politics.” In Nominating the President, eds. Buel

Emmett H.

Jr. Sigelman

Lee

(pp. 42-71). Knoxville: University of Tennessee Press.

Aranson

Peter

Ordeshook

Peter

. 1972. “Spatial Strategies for Sequential Elections.” In Probability Models for Startegic Elections, eds. Niemi

Richard G.

Weisberg

Herbert F.

(pp. 298-331). Columbus: Charles E. Merrill.

Arellano

Manuel

. 1993. “On the Testing of Correlated Effects with Panel Data.” Journal of Econometrics 59:87–97.

Arellano

Manuel

Bond

Stephen

. 1991. “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.” Review of Economic Studies 58:277–97.

Beck

Allen

Sorauf

Frank J.

1992. Party Politics in America. New York: Harper Collins.

Besley

Timothy

Case

Anne

. 2003. “Political Institutions and Policy Choices: Evidence from the United States.” Journal of Economic Literature 41:7–73.

Binder

Sarah

. 1997. Minority Rights, Majority Rule, Partisanship, and the Development of Congress. Cambridge, UK: Cambridge University Press.

Brady

David

Hahrie

Han

Pope

Jeremy

. 2007. “Primary Elections and Candidate Ideology: Out of Step with the Primary Electorate?” Legislative Studies Quarterly 32:79–106.

Buell

Emmett

Jackson

John

. 1991. “The National Conventions: Diminished but Still Important in a Primary Dominated Process.” In Nominating the President, eds. Buel

Emmett

Sigelman

Lee

(pp. 230-51). Knoxville: University of Tennessee Press.

10.

Carey

John

. 1996. Term Limits and Legislative Representation. Cambridge, UK: Cambridge University Press.

11.

Carey

John

Niemi

Richard

Powell

Linda

. 1998. “The Effects of Term Limits on State Legislatures.” Legislative Studies Quarterly 23:271–300.

12.

Carr

Craig

Scott

Gary

. 1984. “The Logic of State Primary Classification Schemes.” American Politics Quarterly 12:465–76.

13.

Carsey

Thomas

Klarner

Carl

Berry

William

Jewell

Malcolm

Niemi

Richard

Powell

Lynda

Snyder

James

. 2012. State Legislative Election Returns (1967-2010). ICPSR 34297. Ann Arbor: Interuniversity Consortium for Political and Social Research.

14.

Carson

Jamie L.

Jenkins

Jeffery A.

2011. “Examining the Electoral Connection across Time.” Annual Review of Political Science 14:25–46. http://www.annualreviews.org/doi/abs/10.1146/annurev-polisci-030310-221852 (accessed May 20, 2014).

15.

Citizens’ Conference on State Legislatures. 1971. The Sometimes Governments: A Critical Study of the 50 American Legislatures. Kansas City, MO: The Citizens’ Conference on State Legislatures.

16.

Coleman

James

. 1971. “Internal Processes Governing Party Positions in Elections.” Public Choice 11:35–60.

17.

Coleman

James

. 1972. “The Positions of Political Parties in Elections.” In Probability Models of Collective Decision Making, eds. Niemi

Richard G.

Weisberg

Herbert F.

Columbus: Charles E. Merrill, pp. 332-357.

18.

Cooper

Joseph

Brady

David

. 1981. “Institutional Context and Leadership Style: The House from Cannon to Rayburn.” American Political Science Review 75:411–25.

19.

Cox

Gary

McCubbins

Mathew D.

1997. “Toward a Theory of Legislative Rules Changes: Assessing Schickler and Rich’s Evidence.” American Journal of Political Science 41:1376–86.

20.

Cox

Gary

McCubbins

Matthew D.

1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press.

21.

Cox

Gary

McCubbins

Matthew D.

2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. Cambridge, UK: Cambridge University Press.

22.

Dion

Douglas

. 1997. Turning the Legislative Thumbscrew: Minority Rights and Procedural Change in Legislative Politics. Ann Arbor: University of Michigan Press.

23.

Fenno

Richard

. 1978. Home Style: House Members in Their Districts. Boston: Little, Brown.

24.

Finkel

Steven

Scarrow

Howard

. 1985. “Party Identification and Party Enrollment: The Difference and Consequence.” Journal of Politics 47:620–42.

25.

Fiorina

Morris P.

1977. Congress-Keystone of the Washington Establishment. New Haven: Yale University Press.

26.

Fiorina

Morris P.

1987. “Alternative Rationales for Restrictive Procedures.” Journal of Law, Economics, & Organization 3:337–45.

27.

Freeman

Patricia

Hedlund

Ronald

. 1993. “The Functions of Committee Change in State Legislatures.” Political Research Quarterly 46:911–29.

28.

Gamm

Gerald

Kousser

Thad

. 2010. “Broad Bills or Particularistic Policy? Historical Patterns in American State Legislatures.” American Journal of Political Science 104(1): 151–70.

29.

Gerber

Elizabeth

Morton

Rebecca

. 1998. “Primary Election Systems and Representation.” Journal of Law Economics, & Organization 14:304–24.

30.

Heckelman

Jacc

. 2004. “A Spatial Model of U.S. Senate Elections.” Public Choice 118:87–103.

31.

Herrick

Rebekah

Moore

Michael K.

Hibbing

John R.

1994. “Unfastening the Electoral Connection: The Behavior of U.S. Representatives When Reelection Is No Longer a Factor.” Journal of Politics 56:214–27.

32.

Herron

Michael

Shotts

Kenneth

. 2006. “Term Limits and Pork.” Legislative Studies Quarterly 31:383–403.

33.

Hogan

Robert

. 2003. “Sources of Competition in State Legislative Primary Elections.” Legislative Studies Quarterly 28:103–26.

34.

Jewell

Malcolm

. 1981. “The State of Legislative Research.” Legislative Studies Quarterly 6:1–25.

35.

Jewell

Malcolm

. 1984. Parties and Primaries: Nominating State Governors. New York: Praeger.

36.

Johannes

John

. 1984. To Serve the People: Congress and Constituency Service. Lincoln: University of Nebraska Press.

37.

Kanthak

Kristen

Morton

Rebecca

. 2003. “Primaries and Turnout.” Working Paper. http://www.nyu.edu/gsas/dept/politics/faculty/morton/KanthakMort.pdf .

38.

Katz

Jonathan

Sala

Brian

. 1996. “Careerism, Committee Assignments, and the Electoral Connection.” American Political Science Review 90:21–33.

39.

Keith

Bruce

Magleby

David

Nelson

Candice

Orr

Elizabeth

Westlye

Mark

Wolfinger

Raymond

. 1992. The Myth of the Independent Voter. Berkeley: University of California Press.

40.

Kenney

Patrick

. 1986. “Explaining Primary Turnout: The Senatorial Case.” Legislative Studies Quarterly 11:65–73.

41.

King

James

. 2000. “Changes in Professionalism in U.S. State Legislatures.” Legislative Studies Quarterly 25:327–43.

42.

Kolodny

Robin

Katz

Richard

. 1992. “The United States.” In Party Organization: A Data Handbook, eds. Katz

Richard

Mair

Peter

. London: SAGE, pp. 871-930

43.

Krehbiel

Keith

. 1990. “Are Committees Composed of Preference Outliers?” American Political Science Review 84:149–63.

44.

Krehbiel

Keith

. 1991. Information and Legislative Organization. Ann Arbor: University of Michigan Press.

45.

Lott

John R.

Jr.

1990. “Attendance Rates, Political Shirking, and the Effect of Post-elective Office Employment.” Economic Inquiry 28:133–50.

46.

Martin

Shane

. 2011. “Electoral Institutions, the Personal Vote and Legislative Organization.” Legislative Studies Quarterly 36:339–61.

47.

Martorano

Nancy

. 2006. “Balancing Power: Committee System Autonomy and Legislative Organization.” Legislative Studies Quarterly 31:205–34.

48.

Mayhew

David

. 1974. Congress: The Electoral Connection. New Haven: Yale University Press.

49.

Meinke

Scott

. 2008. “Institutional Change and the Electoral Connection in the Senate: Revisiting the Effects of Direct Election.” Political Research Quarterly 61:445–57.

50.

Rahat

Gideon

Hazan

Reuven

. 2001. “Candidate Selection Methods.” Party Politics 7(3): 297–322.

51.

Ranney

Austin

. 1981. “Candidate Selection.” In Democracy at the Polls, eds. Penniman

Howard

Butler

David

Ranney

Austin

. Washington, DC: American Enterprise Institute, pp. 75-106.

52.

Rohde

David

. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press.

53.

Schiller

Wendy

. 2006. “Building Careers and Courting Constituents: U.S. Senate Representation, 1889-1924.” Studies in American Political Development 20:185–97.

54.

Shepsle

Kenneth

. 1978. The Giant Jigsaw Puzzle. Chicago: University of Chicago Press.

55.

Shepsle

Kenneth

Weingast

Barry

. 1987. “The Institutional Foundations of Committee Power.” American Political Science Review 81:85–104.

56.

Shepsle

Kenneth

Weingast

Barry

. 1994. “Positive Theories of Congressional Institutions.” Legislative Studies Quarterly 19:149–79.

57.

Squire

Peverill

. 1988. “Career Opportunities and Membership Stability in Legislatures.” Legislative Studies Quarterly 13:65–82.

58.

Squire

Peverill

. 1993. “Professionalization and Public Opinion of State Legislatures.” Journal of Politics 55:479–91.

59.

Squire

Peverill

. 1997. “Another Look at Legislative Professionalization and Divided Government in the States.” Legislative Studies Quarterly 22:417–32.

60.

Squire

Peverill

Hamm

Keith

Hedlund

Ronald

Moncrief

Garry

. 2004. “Electoral Reforms, Membership Stability and the Existence of Committee Property Rights in American State Legislatures.” British Journal of Political Science 35:169–81.

61.

Swift

Elaine

. 1997. The Making of an American Senate: Reconstitutive Change in Congress, 1787-1841. Ann Arbor: University of Michigan Press.

62.

Thompson

Joel

. 1986. “State Legislative Reform: Another Look, One More Time, Again.” Polity 19:27–41.

63.

Uslaner

Eric

. 1985. “Casework and Institutional Design: Redeeming Promises in the Promised Land.” Legislative Studies Quarterly 10(1): 35–52.

64.

Weingast

Barry

Marshall

William

. 1988. “The Industrial Organization of Congress.” Journal of Political Economy 96:132–63.

65.

Westley

Christopher

Calcagno

Peter

Ault

Richard

. 2004. “Primary Election Systems and Candidate Deviation.” Eastern Economic Journal 30:365–76.

66.

Wittman

Donald

. 1977. “Candidates with Policy Preferences: A Dynamic Model.” Journal of Economic Theory 14:180–89.

67.

Wittman

Donald

. 1983. “Candidate Motivations: A Synthesis of Alternatives.” American Political Science Review 77:142–57.

68.

Wittman

Donald

. 1991. “Spatial Strategies When Candidates Have Policy Preferences.” In Advances in the Spatial Theory of Voting, eds. Enelow

James M.

Hinich

Melvin J.

Cambridge, UK: Cambridge University Press, pp. 66-98.

69.

Wittrock

Stephen

Nemeth

Howard

DiSarro

Brian

Squire

Peverill

. 2008. “The Impact of the Australian Ballot on Member Behavior in the U.S. House of Representatives.” Political Research Quarterly 61:434–44.

70.

Wooldridge

Jeffrey

. 2002. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press.

71.

Wright

Gerald C.

2007. “Do Term Limits Affect Legislative Roll Call Voting? Representation, Polarization and Participation.” State Politics and Policy Quarterly 7:256–80.