The Redistricting Cycle,Partisan Tides,and Party Strategy in State Legislative Elections

Abstract

State legislative elections are increasingly shaped by two factors that influence the prospects of winning a majority: the redistricting cycle and partisan tide elections. Winning control of the redistricting process offers the prospect of shaping elections for the next decade, making majority status significantly more valuable than it otherwise might be. Partisan tides, on the contrary, can dramatically alter perceptions of which seats are safe or vulnerable and of whether majority status is obtainable or not. In this article, I examine how the proximity of redistricting and the presence of partisan tides are reflected in the strategies of the party organizations that contest state legislative elections. Using party finance data from 29 states during the period from 1996 to 2010, I find that parties’ majority-seeking behavior is more intense in states with legislative redistricting, when redistricting is imminent, and when partisan tides favor the minority party.

Keywords

party organizations parties and interest groups parties and elections legislative elections legislative politics elections redistricting

In both 2000 and 2010, the national parties focused heavily on state legislative races because the results of those races had important implications for redistricting, and by extension, control of the U.S. House. Then-unprecedented financial resources were directed to this task in 2000 (Storey 2000), and by 2010, the importance of this priority was reflected in new organizations such as the Democratic Redistricting Trust and its Grand Old Party (GOP) equivalent, the Redistricting Majority Project (D’Aprile 2010).

But from the perspective of state parties, an infusion of national money was not necessary to inspire an intense interest in gaining or maintaining state legislative majorities for these pivotal periods. After all, the linkage between the redistricting process and legislative composition is far more direct in state legislature redistricting: the party which controls the process draws all the districts in the state, rather than just the districts for one state, as in congressional redistricting (Herron and Wiseman 2008). Thus, gaining control of redistricting can give a party a decade-long structural advantage in its effort to maintain majority status.

This intense focus on pre-redistricting elections can be seen as one implication of the “redistricting cycle” (Hetherington, Larson, and Globetti 2003), the patterns arising in the post-Baker v. Carr era in which redistricting occurs at regular intervals (Cox and Katz 2002). The power to control the redistricting process makes majority party status in a legislature more valuable at the start of the decade than it might otherwise be. Because contributors show a propensity to seek out the most powerful institutions (Salka 2009), it is unsurprising that previous studies have found some evidence that levels of party spending increase in elections surrounding redistricting (Moncrief 1998). However, such studies have not examined the distribution of contributions across elections. The implicit assumption of many party finance studies is that temporal factors such as the redistricting cycle do not matter, an assumption made explicitly in Clucas (1992), who, while recognizing the argument that optimal strategies could be different throughout the redistricting cycle, ultimately believes strategies will not vary because “control of the Assembly is likely to be decided over a series of elections” (p. 281).

But while 2000 and 2010 shared a proximity to the next round of redistricting, they differed in an important way: 2000 was a relatively neutral year marked by an extremely close presidential election, while 2010 was a year characterized by a decidedly pro-Republican partisan “tide” (Jacobson 2011). Partisan tides exist when two conditions hold. First, one party has a clear competitive advantage at the national level due to factors such as economic conditions or presidential popularity. Second, perceptions of this advantage are broadly perceived and acted upon at the elite level, producing partisan disparities in the emergence of quality challengers and the distribution of donations and resources (Basinger and Ensley 2007; Jacobson and Kernell 1983). As a result, partisan tides are “self-fulfilling” (Jacobson and Kernell 1982).

While such elections have historically been somewhat infrequent, the three elections between 2006 and 2010 broke the historical pattern rather dramatically insofar as all three featured strong partisan tides (Campbell 2011). Medvic (2010) argues that the nationalization of campaigns is, in part, responsible for the increasing prevalence of such elections, particularly in midterm elections. At the state legislative level, where elections are generically lower-information affairs, these tides may be particularly influential: between 2006 and 2010, chambers in 23 states experienced new legislative majorities, with chambers in 11 states experiencing two successive changes in this span.

Like control of redistricting, partisan tides reshape the calculus that political parties face in developing electoral strategies. First, a gerrymander might alter a district in a way that shifts it from safe to vulnerable (or vice versa); likewise, the presence of a partisan tide may change perceptions of a district such that a formerly uncompetitive district is now viewed as competitive. Second, both gerrymanders and partisan tides can change a party’s beliefs about whether obtaining enough seats to win majority status is realistic.

In this article, I examine how the proximity of redistricting and the presence of partisan tides are reflected in the strategies of the party organizations that contest state legislative elections. The seeking of majority party status in legislatures is generally thought to fuel different strategies by majority and minority parties: all else equal, minority parties will adopt an “offensive” strategy in which they are more supportive of incumbents, while majority parties will adopt a “defensive” strategy in which they are more supportive of challengers and open-seat candidates (Dwyre and Stonecash 1992; Herrnson 1996). These patterns have been viewed as universal, except in states with no interparty competition (Gierzynski 1992).

If parties adopt such macro-level strategies, it stands to reason that election-specific factors such as proximity to redistricting and partisan tides will help shape these strategies. I argue that parties will adopt majority-seeking strategies when redistricting is imminent and when the state legislature is the main arena for the redistricting process. I also contend that partisan tides will influence party strategies, intensifying majority-seeking when the tide is favorable and producing a countervailing force when tides are unfavorable.

To assess these claims, I analyze the contribution patterns of party organizations in the lower house chambers of 29 state legislatures during the period from 1996 to 2010. By including years throughout the redistricting cycle and election cycles with varying partisan tides, I offer a highly comprehensive look at the behavior of state party organizations.

The article is organized as follows. First, I discuss previous examinations of offensive and defensive strategies and theorize why conditions related to redistricting and partisan tides should influence patterns of party contributions. Second, I discuss some of the measurement issues that pertain to modeling offensive and defensive strategies, and develop a novel strategy for the measurement of state-level partisan tides. Third, I discuss the data and multilevel modeling strategy used in this study. Finally, I present results and discuss avenues for future research.

Offense Defense Theory (ODT)

In describing the contribution strategies of party organizations, early work (e.g., Jacobson 1985–1986) noted that party organizations face a collective action problem: convincing safe incumbents to forego financial support to build the size of the party. Doing so maximizes the “efficiency” of party campaign contributions (i.e., channeling money to candidates in competitive races). A number of national (Glasgow 2002), single-state (Stonecash and Keith 1996), and comparative state studies (Gierzynski 1992) of party finance have established that to varying degrees, party organizations have become capable of instituting such efficient strategies.

A major impetus for this increased efficiency was the changing functional roles of party organizations, and specifically the rise of legislative caucus campaign committees (hereafter, LCCs), which came into prominence in the 1980s and 1990s (Gierzynski 1992; Rosenthal 1995; Shea 1995). LCCs and other new party organizational structures helped solve this collective action problem by insulating decision-makers from the pressures of incumbents.

While electoral prospects are enhanced by targeting contributions to the most competitive races, such “seat maximization” strategies (Jacobson 1985–1986) are not the only decision rules that parties might follow. Party loyalty has often been seen as a potential predictor of party contributions (Leyden and Borrelli 1990), although others have failed to find such patterns (Schecter and Hedge 2001). Other studies have examined campaign professionalism (Francia et al. 2003; Herrnson 1989), seniority (Stonecash 1988; Stonecash and Keith 1996), and leadership positions (Thompson and Cassie 1992) as predictors of receiving a contribution.

Another possibility is that parties will not merely consider individual races independently but instead pursue a macro-level strategy where the goal is winning control of the legislative chamber. While this idea has its roots in the literature on Congress (e.g., Herrnson 1989), a most explicit elaboration of the logic can be found in Gierzynski’s (1992) study of state LCCs. He argues that a party’s status in the legislature will influence the way party organizations allocate money to candidates. Majority parties will seek to protect their majority by pursuing a “defensive” strategy of protecting seats they already hold, while minority parties will seek to obtain a legislative majority by pursuing an “offensive” strategy of targeting seats they do not hold. Similar ideas are referred to elsewhere as “party maximizing strategies” (Clucas 1992), “additive versus protectionist strategies” (Thompson, Cassie, and Jewell 1994), or described without a label (Stonecash 1988). Hereafter, I refer collectively to these ideas as Offense Defense Theory (ODT).

Findings consistent with ODT and its assumptions have often been found in studies of Congress and single-state legislatures. Heberlig and Larson (2012), for example, find strong evidence that congressional parties marshal resources toward the end of majority party control and do not merely follow seat maximization strategies. However, the applicability of these findings to the states is not straightforward: the value of majority party status differs in many ways, including in the area of redistricting control to which I will return shortly. Moreover, LaRaja, Orr, and Smith (2006) contend that state party finance is largely driven by the state electoral and legal environment and not by national conditions or laws.

At the state level, Thompson, Cassie, and Jewell (1994) find evidence consistent with ODT in the competitive states of New Jersey and Pennsylvania but find less evidence in the less competitive state of North Carolina. Gierzynski (1992) also offers mixed support for ODT, finding that in some states, majority parties supported incumbents and minority parties supported non-incumbents. However, in pooled models, this pattern was found for only one of the three years in question.¹ Gierzynski (1992) also offers one prerequisite condition for ODT to hold: he theorizes (p. 29) and finds (p. 107) that patterns consistent with ODT do not exist in states with little interparty competition. The subsequent literature, however, has largely failed to confirm the universality of these claims.

Longitudinal work, while comparatively rare in the party finance literature, has also offered insights into party contribution strategies. In addition to majority and minority party status, parties may also be influenced by electoral conditions in the present election (Gierzynski 1992). Parties facing strong electoral conditions will be more offensive, and those facing poor conditions will be more defensive. Other longitudinal studies have also examined parties’ capacity to adjust to district-level changes across elections (Stonecash and Keith 1996). However, the election year itself has not been treated as an important factor.

Incorporating the Redistricting Cycle

Because of the redistricting cycle, neglecting temporal factors in the study of party finance strategies is a serious oversight. Because obtaining or maintaining majority party status is the purported motivation for offensive and defensive strategies, it follows that changes in the value of majority party status should be associated with changes in the use of these strategies.

Controlling the redistricting process is certainly one valuable benefit of being the state legislative majority party, albeit a benefit that occurs just once per decade.² Losing a redistricting battle may have negative implications not just for the party in the legislature but for every dimension of party behavior in that state, including party recruitment, fund-raising, and mobilization efforts. And while there is still debate in the academic literature about parties’ ability to take advantage of the redistricting process (see McDonald 2004 for a review of this debate), the fact that parties believe such advantages exist is ultimately most persuasive.

As such, I argue that party organizations will pursue the most starkly offensive and defensive strategies in states where control of redistricting is at stake. Majority party organizations will be more protective of incumbents in such states, compared with states where redistricting is conducted by a partisan or non-partisan commission.³ Conversely, minority party organizations will be more aggressive in targeting incumbents in states with legislative redistricting than in states with non-legislative redistricting. These predictions are summarized in the Redistricting Control Hypothesis.

Redistricting Control Hypothesis: The magnitude of contribution biases, as predicted by ODT, will be greater in states with legislative control of redistricting.

Because redistricting occurs regularly following each Census, the salience of the redistricting process should be greatest as this period approaches. The belief that party competition intensifies as the redistricting cycle concludes is widely held (Moncrief 1998; Rosenthal 1995) if not systematically examined, so the contention that party strategy might vary over time is not without foundation. However, parties’ increased motivation to regain their majority does not equate to an increased ability to compete. Because the end of the redistricting cycle features entrenched incumbents and hesitant potential challengers (Hetherington, Larson, and Globetti 2003), extreme party contribution strategies may be necessary to achieve goals.

While Clucas (1992) argues that control of redistricting inspires party strategy throughout the decade, I argue that elections closest to redistricting are the only ones after which a party can reasonably conclude they will control redistricting with certainty. Thus, majority parties will be especially protective of incumbents as redistricting approaches, and minority parties will be especially aggressive in targeting majority incumbents. However, these patterns will only exist in states with legislative redistricting; in states with non-legislative redistricting, the imminence of redistricting will not influence party finance patterns. These predictions are summarized in the Redistricting Imminence Hypothesis.

Redistricting Imminence Hypothesis: In states with legislative redistricting, the magnitude of contribution biases, as predicted by ODT, will be greater when there are fewer remaining election cycles before the decennial Census.

Partisan Tides and Party Strategy

Because the presence of a partisan tide affects everything from party recruitment and fund-raising to incumbent behavior, parties should be motivated to alter contribution strategies as well: parties who anticipate a partisan tide may wish to focus their resources on either incumbents (if the tide is unfavorable) or challengers (if the tide is favorable).

Past studies (e.g., Gierzynski 1992; Krasno and Green 1988) have treated partisan tides as a national-level phenomenon measured invariantly across states. However, partisan tides may not be uniform across states: for example, the Republican tide of 2010 did not appear to extend to California (Dennis and Widestrom 2012). As such, any comparative analysis of party strategies in state legislative races requires accounting for the fact that elections for the state legislature may also be driven by state-level factors. State legislative races can be affected by coattails in gubernatorial races (Hogan 2005), perceptions of state party ideology, and which party controls the state legislature (Bishop and Hatch 2011). Moreover, the factors that affect approval of the state legislature differ across states (Richardson, Konisky, and Milyo 2012) and also differ from the factors that influence individuals’ attitudes toward Congress (Kelleher and Wolak 2007).

From a methodological standpoint, accounting for partisan tides is essential in accurately identifying the use or non-use of offensive and defensive strategies. When a minority party is facing a favorable partisan tide, the inclination will be to support challengers rather than incumbents, particularly because selection effects will produce an unusually strong set of candidates for one party and an unusually weak set of candidates for the other party (Krasno and Green 1988). Thus, acting consistently with ODT and responding to partisan tides can be observationally equivalent. Without controlling for partisan tides in an analysis, we cannot be sure whether a party’s contribution biases reflect partisan tides or the desire to obtain or retain majority status. Likewise, the lack of an apparent contribution bias could be the result of countervailing influences: being pushed in one direction by the partisan tide and in the other direction by the desire to obtain or retain majority status.

Methodological concerns aside, there is a theoretical linkage between the redistricting cycle and partisan tides as explanations for party contribution strategies. Being responsive to partisan tides and being responsive to the redistricting cycle share a common theme: the consideration of macro-level factors in addition to candidate and district-level factors; in other words, both represent a rejection of the “seat maximization” strategy to take advantage of favorable or valuable conditions.

Partisan Tides Hypothesis: Contribution biases, as predicted by ODT, will be amplified when partisan tides favor the minority party and weakened when partisan tides favor the majority party.

Data and Method

To assess the above hypotheses, I examine the allocation of party money to lower chamber state legislative candidates⁴ in contested races⁵ during the period from 1996 to 2010, excluding the 2001–2002 elections.⁶ The party finance data in this study were compiled from The Institute on Money in State Politics. A party organization was included in the analysis if it met three criteria. First, the organization must have contributed money⁷ to candidates in at least three elections during the period. Second, the organization’s contributions must have constituted at least 1% of all spending by candidates of that party in a given year.⁸ Third, I only include money from two types of party organizations: legislative campaign committees (LCCs) and state party organizations (SPOs).⁹ In total, 29 states were included, with a total of 323 party-level observations and nearly 20,000 candidate-level observations. A full list of cases included in the analyses is given in Table 1.

Table 1.

List of Party Organizations in Study.

	Democrats	Republicans
Alaska	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
California	98, 00, 04, 06, 08, 10	98, 00, 04, 06, 08, 10
Colorado	98, 00, 04, 06, 08, 10	98, 00, 04
Connecticut	00, 04, 06	98, 00, 04, 06
Florida	98, 00, 04, 06, 08, 10	98, 00, 04, 06, 08, 10
Iowa	98, 00, 04, 06, 08, 10	98, 00, 04, 06, 08, 10
Idaho	96, 98, 00, 04, 06, 10	96, 98, 00, 04, 06, 08, 10
Illinois	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Indiana	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Kentucky	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Michigan	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Minnesota	96, 98, 00, 04	96, 98, 00, 04
Missouri	98, 00, 04, 06, 08, 10	98, 00, 04, 10
North Carolina	06, 08, 10	06, 08, 10
New Mexico	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 10
Nevada	96, 98, 00, 04, 06, 08	96, 98, 00, 04, 06, 08, 10
New York	98, 00, 04, 06, 08	98, 00, 04, 08
Ohio	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Oklahoma	00, 04, 08	00, 04, 06, 08
Oregon	96, 98, 00, 04, 06, 08	96, 98, 00, 04, 06
Pennsylvania	98, 00, 04, 06, 08, 10	98, 00, 04, 06, 08, 10
South Carolina	96, 98, 00, 06, 08, 10	96, 98, 00, 06, 08, 10
Tennessee	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Texas	98, 00, 04, 06, 08, 10	98, 00, 04, 06, 08, 10
Utah	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Virginia	99, 03, 05, 07, 09	99, 03, 05, 07, 09
Washington	96, 98, 00, 04, 06, 08, 10	96, 98, 00, 04, 06, 08, 10
Wisconsin	98, 00, 06	98, 00, 08, 10
Wyoming	96, 06, 08, 10	96, 98, 00, 04, 06, 08, 10

To test the above hypotheses, I use a multilevel Tobit model. The Tobit model is appropriate because party contributions to candidates are strictly non-negative and censored at zero (Greene 2003).¹⁰ The Tobit model assumes that contributions are a response to a latent variable y*, which might be roughly interpreted here as a candidate’s general deservingness of receiving money. Candidates at or below a certain threshold of y* = τ receive zero dollars, while candidates for whom y* > τ receive money according to a continuous linear function of individual-level attributes and group-level factors.

I adopt a multilevel modeling framework because while party organizations tend to respond to similar factors in determining the allocation of party money to candidates (e.g., district competitiveness, quality of opponent, incumbency status), the weight an organization gives to macro-level variables may vary across states. The mean contribution also varies greatly across organizations. Thus, there is simply too much heterogeneity in the data to estimate a pooled model across all candidates in all states. Multilevel models can account both for heterogeneity in coefficients across organizations, and for the fact that the error structure is likely to be correlated within groups and to vary across groups.¹¹ Specifically, the model to be estimated is as follows:

y_{i j} = {\begin{matrix} y *_{i j} \equiv γ_{00} + \sum_{h = 1}^{p} γ_{h 0} x_{h i j} + \sum_{k = 1}^{q} γ_{0 k} z_{k j} + \sum_{k = 1}^{q} \sum_{h = 1}^{p} γ_{h k} z_{k j} x_{h i j} + U_{0 j} \sum_{h = 1}^{p} U_{h j} x_{h i j} + R_{i j} & i f y *_{i j} > τ \\ 0 & i f y *_{i j} \leq τ \end{matrix}} .

the x_hijs represent candidate-level factors such as incumbency status, district competitiveness, and the party that held the seat, subscripted for the candidate i and party organization j. The z_kjs represent organizational-level (and state level) factors such as organization type, legislative competitiveness, and the type of redistricting process, subscripted for the organization j. The coefficients γ_hk represent cross-level interactions between a district-level variable h and an organizational-level variable k, which are necessary to test the conditional hypotheses. The coefficient γ₀₀ represents the intercept when all variables are set at their grand means, the coefficients γ_h0 represent the main effects of the candidate-level variables, while the coefficients γ_0k represent the main effects of the organization and state-level variables. Finally, the error term U_0j is a random intercept term that allows mean contributions to vary across organizations, and the error term U_hj captures the random slopes of the macro-level two variables, whose effects are allowed to vary across organizations. The error term R_ij is an independent and identically distributed residual term that varies at level 1.

The dependent variable, per capita party contribution, is constructed so contributions are comparable across states. The contribution made to a candidate is divided by the average district population in the state. Across all observations, the mean contribution is $0.21 per constituent (SD = $0.68), while the average non-zero contribution is $0.42 per constituent.

Independent Variables

Next, I turn to the key predictor variables required to test the hypotheses advanced above. The central claim of ODT is that both majority and minority parties will focus their resources on seats currently held by the majority party. Therefore, I create a dummy variable, majority-held seat, indicating whether a member of the current legislative majority party holds the seat,¹² and I predict a positive coefficient on this variable.

To assess the Redistricting Control Hypothesis, I create a dummy variable legislative control, which takes on a value of “1” if the legislature is the primary venue for state legislative redistricting and a “0” otherwise (based on McDonald 2004).¹³ The Redistricting Control Hypothesis asserts that parties will adopt more aggressively offensive and defensive strategies under this condition. To assess this hypothesis, I include an interaction between majority-held seat and legislative control; the coefficient on this term should be positive.

The Redistricting Imminence Hypothesis, on the contrary, asserts that more aggressive strategies will only be used when redistricting is imminent and that as redistricting becomes more distant, the strategies of parties in states with legislative and non-legislative redistricting institutions will become more similar. To assess this hypothesis, I first create a measure, years until redistricting, which counts the number of years between the current election and the next round of redistricting. Thus, the 2000 and 2010 elections (1999 and 2009 in Virginia) receive a “0,” the 1998/2008 elections receive a “2,” the 1996/2006 elections receive a “4,” and the 2004 elections receive a “6.” Because the hypothesis asserts that the imminence of redistricting will only influence contribution patterns in states with legislative control of redistricting, I create a three-way interaction between majority-held seat, legislative control, and years until redistricting. For proper specification, I also include all constituent terms (Brambor, Clark, and Golder 2005). The expectation is that there will be a negative coefficient on the triple interaction.

To assess the Partisan Tides Hypothesis, I include a variable, partisan tide, which estimates the advantage, expressed as a vote percentage, that a candidate of the state’s majority party would have had in a given state and year. Full details of the construction of this measure can be found in Online Appendix A. Because the hypothesis predicts that parties will contribute more money in seats they do not hold when the partisan tide favors them, the expectation is that majority parties give less money in majority-held seats when the tide favors the majority party, and more money when the tide is unfavorable, with the reverse being true for minority parties. To test this hypothesis, I create an interaction between majority-held seat and partisan tide, with the expectation that the sign on that coefficient will be negative.

I also include a number of control variables. The most important of these is legislative competition. As previously noted, Gierzynski (1992) asserts that ODT will only hold in states where there is enough competitiveness that the minority party might plausibly win a majority in the current election. Assessing competitiveness in a state legislature requires attention to two separate meanings of competitiveness: the difference between the seat shares of the two parties and the amount of district-level competition (Barrilleaux, Holbrook, and Langer 2002).

To capture both of these meanings of competitiveness, I create a measure majority party dominance, defined as the percentage shift in aggregate vote share that would be required for the minority party to gain majority party status. Specifically, I use JudgeIt, Version 1.3.4 (Gelman, King, and Thomas 2008), to simulate the aggregate vote share at which the minority party would be expected to win 50% of the seats in the legislature.¹⁴ Because competition may be a prerequisite condition for ODT to hold, higher values of majority party dominance may make parties less likely to follow ODT. Thus, I expect an interaction between majority-held seat and majority party dominance to take a negative sign.¹⁵

Finally, I include a number of district-level predictors of contributions. I include two dummy variables indicating whether the candidate was an incumbent or a challenger, with the expectation being that both will receive smaller contributions relative to candidates in open-seat races.¹⁶ The previous margin of victory serves as one measure of district competitiveness, with the expectation that more money will be given in districts with smaller previous margins. Likewise, I include a measure of the normal vote margin, using district-level presidential returns from the 2000 election (Wright 2004); this measure provides a comparable metric because data from the same presidential race are available for all districts, both pre-redistricting and post-redistricting. The expectation is that districts that are more competitive in national elections will attract more party contributions, all things equal. Finally, I calculate the difference in the percentage of non-party money raised by the two major party candidates in the race. Although this non-party money difference is unlikely to be completely exogenous to party money, this third measure of competitiveness has the advantage of being specific to the year and race in question.¹⁷ Summary statistics and sources for all data can be found in Table 2.

Table 2.

Summary Statistics and Data Sources.

Variable	Description	M	SD
Dependent variable
Party money^a	Per capita party money given to candidate	0.21	0.68
Contest-level variables
Majority party-held seat	Dummy = 1 if seat is currently held by the majority party in the lower chamber	0.58	0.49
Previous margin of victory	Difference in the two-party Democratic and Republican vote in prior legislative election	0.44	0.36
Normal vote margin^b	Difference in two-party Democratic and Republican vote in 2000 presidential election, adjusted by state	0.10	0.08
Non-party money difference^a	Difference between the percentage of non-party money raised by two major candidates	0.59	0.33
Incumbent candidate	Dummy = 1 if candidate is incumbent	0.38	0.49
Challenger candidate	Dummy = 1 if candidate is challenger	0.39	0.49
Organization-level variables
Control of redistricting^c	Dummy = 1 if legislature controls redistricting in state	0.71	0.46
Years until redistricting	Number of year between current election and next redistricting (i.e., 2001 or 2011)	2.48	2.14
Partisan tide^d	Estimated state-level partisan tide in favor of legislative majority party	−0.001	0.024
Chamber competitiveness^e	Estimated vote swing required to change legislative control of chamber	0.065	0.053
Total party spending^a	Organization spending as a percentage of all spending by its party’s candidates	0.126	0.094

The National Institute on Money in State Politics.

Based on Wright (2004).

Based on McDonald (2004).

Calculated by author. See Online Appendix A.

Based on Carsey, Berry, Niemi, Powell, and Snyder (2009).

Results

In Tables 3 and 4, I present the results of a series of multilevel Tobit models that predict the per capita party contribution made to each legislative candidate. I begin by assessing the effects of legislative control of redistricting, proximity to the next round of redistricting, and the presence of partisan tides. Next, I explore differences in patterns across states with and without legislative redistricting and differences in behavior among majority parties and minority parties. Finally, I consider whether these patterns extend only to states with competitive legislatures.

Table 3.

Multilevel Tobit Models of Per Capita Party Spending.

		Model 1	Model 2	Model 3
	Expected sign	All states	Legislative redistricting states	Non-legislative redistricting states
Constant		0.88 (0.07)***	0.83 (0.08)***	0.98 (0.15)***
Contest-level variables
Majority party-held seat	(+)	0.13 (0.03)***	0.14 (0.03)***	0.10 (0.05)**
Previous margin of victory	(−)	−0.33 (0.05)***	−0.28 (0.05)***	−0.44 (0.12)***
Normal vote margin	(−)	−2.44 (0.25)***	−2.50 (0.30)***	−2.34 (0.46)***
Non-party money difference	(−)	−1.33 (0.11)***	−1.30 (0.12)***	−1.38 (0.23)***
Incumbent candidate	(−)	−0.17 (0.03)***	−0.13 (0.04)***	−0.25 (0.06)***
Challenger candidate	(−)	−0.23 (0.04)***	−0.22 (0.04)***	−0.29 (0.07)***
Organization-level variables
Control of redistricting		−0.05 (0.06)	—	—
Years until redistricting		−0.004 (0.012)	−0.01 (0.01)	0.01 (0.02)
Partisan tide		−2.37 (1.39) *	−0.84 (1.58)	−6.65 (2.69)**
Majority party dominance		−0.47 (0.49)	−0.56 (0.57)	−0.50 (0.86)
Cross-level interactions
Maj. held × Control	(+)	0.05 (0.05)	—	—
Maj. held × Years until		−0.030 (0.012)**	−0.041 (0.014)***	0.004 (0.018)
Years until × Control		−0.02 (0.03)	—	—
Maj. held × Control × Years until	(−)	−0.045 (0.023)**	—	—
Maj. held × Partisan tide	(−)	−4.33 (1.21)***	−3.13 (1.43)**	−8.20 (2.26)***
Maj. held × Maj. dominance	(−)	−0.46 (0.42)	−0.91 (0.51) *	0.60 (0.66)
Model statistics
N (candidates/parties)		19,813/323	13,941/229	5,872/94
σ (random intercept)		0.30 (0.03)***	0.30 (0.04)***	0.30 (0.05)***
σ (random slope)		0.32 (0.04)***	0.35 (0.05)***	0.24 (0.05)***
σ (residual)		0.93 (0.07)***	0.88 (0.08)***	1.03 (0.15)***

Note. Robust standard errors in parentheses. Interaction and constituent terms mean centered.

p <.10. **p <.05. ***p <.01.

Table 4.

Interpretation of Interactive Term Marginal Effects.

		Expected party money per 1,000 constituents
Redistricting and party giving	Marginal effect (SE)	Majority-held seat	Minority-held seat	Percentage difference (%)
Legislative redistricting states
No elections before redistricting	0.245 (0.052)***	$679	$606	12.0
One election before redistricting	0.159 (0.035)***	$661	$614	7.7
Two elections before redistricting	0.074 (0.037)**	$644	$622	3.5
Three elections before redistricting	−0.012 (0.056)	$627	$631	−0.1
Non-legislative redistricting states	0.093 (0.046)**	$668	$640	4.5%

Partisan tides and party giving	Marginal effect (SE)	Majority-held seat	Minority-held seat	Percentage difference (%)
Pro-majority tide	0.022 (0.041)	$627	$617	1.6
Neutral	0.126 (0.028)***	$657	$616	6.7
Pro-minority tide	0.229 (0.041)***	$689	$615	12.0

Note. Calculations are based on the change in the truncated outcome, ΔE(y | y > τ, x) / Δk. Other variables held as their respective means (for continuous variables) or modal values (for discrete variables). Because the marginal effect of majority-held seat does not vary across years in non-legislative redistricting states, the estimate at the mean value of years until redistricting is presented.

p < .05. ***p <.01.

The first column of Table 3 presents the main model, with all party organizations and states included. The results indicate support for the idea that redistricting shapes party strategies, and, in particular, for the Redistricting Imminence Hypothesis. When redistricting is imminent, the difference between states with legislative and non-legislative redistricting is significant; however, the coefficient on the two-way interaction between majority-held seat and control of redistricting is not significant, indicating that the two types of states are more similar when years until redistricting is at its mean value. The results also indicate support for the Partisan Tides Hypothesis: candidates in majority-held seats receive significantly less money when the partisan tide favors the majority party.

Given the complexity of the interactions, these findings are re-expressed in Table 4 by calculating the marginal effects (and their standard errors) and the substantive effect sizes for different values of the relevant variables. Specifically, I calculate the marginal effect of being a candidate in a majority-held seat on the size of the predicted party contribution. Turning first to states with legislative control of redistricting, parties exhibit a contribution bias toward majority-held seats in the elections immediately prior to redistricting, as well the two elections prior to that. The size of that bias is considerable, increasing from 3.5% in the year ending in “6” to 12.0% in the year ending in “0.”

The impact of partisan tides can have similarly large effects on party contribution biases. In a state with a partisan tide one standard deviation more favorable to the minority party than the mean, both parties target majority-held seats aggressively, once again favoring such candidates by approximately 12.0% relative to otherwise identical candidates in minority-held seats. On the contrary, when the partisan tide is one standard deviation more favorable to the majority party than the mean, the contribution bias toward majority parties is no longer significantly different from zero. In this case, the propensity toward majority-seeking, as expressed in ODT, is counterbalanced by the effect of the partisan tide. With neither party believing that majority status is likely to change in this election, the motivation for a contribution bias disappears.

Finally, it is important to observe that parties act consistently with ODT even in states without legislative redistricting. Although the magnitude of contribution biases are smaller than in legislative redistricting states where redistricting is imminent, the marginal effect is significantly different from zero and implies a bias of around 4.5%. That parties would value majority status for reasons unrelated to redistricting is unsurprising, but given inconsistent findings regarding ODT in the previous literature, its applicability to these states is worth noting.

The other variables in the model largely behave as expected. All three measures of political competitiveness indicate that more money is spent in more competitive districts, and open-seat candidates generically receive more money than challengers and incumbents. The effects of chamber-level competitiveness are in the expected direction, but neither the main effect nor the interaction with majority-held seats is statistically significant. We can neither be confident that less money is spent in less competitive chambers nor that the lack of chamber-level competitiveness produces patterns less consistent with ODT.

It is possible, however, that there are additional nuances in the relationships between institutions, party organizations, electoral context, and time. Because previous studies have found inconsistencies in party strategies across place and time, I explore these relationships in subsets of the data, divided by redistricting control, majority status, and chamber competitiveness. Because these examinations are not the main subject of the article, this strategy seems preferable to introducing models with more triple-interaction terms and possibly quadruple-interaction terms. Because these split sample models do not constitute proper significance testing and as “the difference between significant and not significant is not significant” (Gelman and Stern 2006), these results should be interpreted with appropriate caution.

Models 2 and 3 in Table 3 compare the results of two multilevel Tobit models specified as above: one that includes only states with legislative redistricting and one that includes only states with non-legislative redistricting. The results are broadly similar with a couple exceptions. First, as predicted by the Redistricting Imminence Hypothesis and supported by the triple interaction in Model 1, time matters in states with legislative redistricting and does not matter in other states. The second, more novel difference concerns the interaction between majority-held seat and partisan tide. The coefficient on this interaction, although significant in both models, is well over twice as large in the states without legislative redistricting. This implies that parties in such states may react more aggressively to partisan tides because they do not have to concern themselves as much with the importance of majority party status for redistricting purposes.

Turning to Models 4 and 5 in Table 5, I compare the contribution strategies across majority and minority parties. Two patterns of note emerge here: first, the negative coefficient on the triple-interaction is nearly twice as large among majority parties, indicating that they are likely to adopt a defensive strategy only when redistricting is very close. Minority parties, on the contrary, may adopt an offensive strategy in earlier years; this makes sense if we believe that winning a majority may be a multi-election plan for many minority parties. Second, the impact of partisan tides on contribution strategies is seemingly greater for majority parties than for minority parties. This could reflect the asymmetrical nature of competition for majority status: by definition, majority parties must only defend their vulnerable seats, while minority parties must retain their vulnerable seats and win majority-held seats to achieve a majority.

Table 5.

Additional Multilevel Tobit Models of Per Capita Party Spending.

	Model 4	Model 5	Model 6	Model 7
	Majority parties only	Minority parties only	Competitive states only	Uncompetitive states only
Constant	1.00 (0.12)***	0.75 (0.09)***	0.83 (0.18)***	0.74 (0.08)***
Contest-level variables
Majority party-held seat	0.11 (0.06)*	0.06 (0.05)	0.14 (0.14)	0.16 (0.05)***
Previous margin of victory	−0.38 (0.08)***	−0.26 (0.05)***	−0.46 (0.10)***	−0.21 (0.04)***
Normal vote margin	−2.74 (0.33)***	−2.14 (0.37)***	−3.18 (0.47)***	−1.66 (0.22)***
Non-party money difference	−1.32 (0.16)***	−1.30 (0.14)***	−1.68 (0.18)***	−0.92 (0.09)***
Incumbent candidate	−0.16 (0.06)***	−0.20 (0.05)***	−0.20 (0.06)***	−0.12 (0.03)***
Challenger candidate	−0.27 (0.07)***	−0.18 (0.04)***	−0.29 (0.06)***	−0.17 (0.03)***
Organization-level variables
Control of redistricting	−0.06 (0.08)	−0.06 (0.08)	0.00 (0.08)	0.02 (0.05)
Years until redistricting	−0.005 (0.016)	−0.004 (0.017)	−0.03 (0.03)	0.004 (0.01)
Partisan tide	−2.40 (1.89)	−2.22 (2.00)	−1.85 (2.78)	−2.12 (1.07)**
Majority party dominance	0.17 (0.66)	−1.21 (0.63)*	−5.02 (3.12)	−1.75 (0.59)***
Cross-level interactions
Maj. held × Control	0.11 (0.08)	−0.003 (0.015)	0.00 (0.09)	0.08 (0.06)
Maj. held × Years until	−0.024 (0.017)	−0.033 (0.015)**	−0.056 (0.018)***	−0.01 (0.01)
Years Until × Control	−0.05 (0.03)	0.003 (0.036)	0.03 (0.05)	−0.03 (0.02)
Maj. × Control × Years	−0.057 (0.031)*	−0.032 (0.031)	−0.072 (0.039)*	−0.02 (0.02)
Maj. Held × Partisan tide	−6.47 (1.82)***	−2.59 (1.47)*	−5.89 (1.81)***	−0.75 (0.73)
Maj. Held × Dominance	−0.29 (0.58)	−0.80 (0.58)	0.30 (2.60)	−1.83 (0.72)**
Model statistics
N (candidates/parties)	9,627/159	10,186/164	9,992/142	9,821/181
σ (random intercept)	0.30 (0.04)***	0.30 (0.05)***	0.40 (0.05)***	0.19 (0.03)***
σ (random slope)	0.35 (0.06)***	0.27 (0.04)***	0.26 (0.05)***	0.33 (0.05)***
σ (residual)	0.97 (0.11)***	0.86 (0.08)***	1.14 (0.11)***	0.64 (0.06)***

Note. Robust standard errors in parentheses. Interaction and constituent terms mean centered.

p <.10. **p <.05. ***p <.01.

Models 6 and 7 in Table 5 compare the contribution strategies in competitive and uncompetitive chambers with competitive chambers defined as ones in which the estimated voted swing required to change chamber control is 5 percentage points or less. As predicted by much of the previous literature, many of the patterns disappear or weaken significantly in the less competitive states. The relationships between time, redistricting control, and contribution strategy disappear almost entirely in the uncompetitive states, while remaining robust (and with larger coefficients) in the competitive states. The effect of partisan tides on contribution strategy also mostly disappears in the uncompetitive states. Last, even among uncompetitive states, the bias toward giving in majority-held seats decreases as the chamber becomes less competitive: we see hints of offensive and defensive strategies in remotely competitive states like Illinois and Nevada but none whatsoever in truly one-party states like Idaho and New York. Taken together, this indicates that parties in uncompetitive states may respond more strongly to district-level factors than to macro-level considerations such as majority control and the redistricting cycle.

Discussion

In this study, I show that the strategic behavior of party organizations is influenced by two important contextual factors: redistricting and partisan tides. These organizations, particularly since the rise of legislative campaign committees, are highly influential in state legislative elections not only as direct financial contributors but also as providers of party services (Francia et al. 2003). If party strategies are increasingly directed not toward seat maximization but rather toward attaining control of redistricting and responding to partisan tides, competition in state legislative elections may be fundamentally altered.

In examining party finance patterns, I began with the oft-encountered supposition that majority parties will tend to protect incumbents while minority parties will tend to support challengers. Based on the findings in this study, ODT appears to describe the behavior of political parties in states where the redistricting process is directly linked to the outcome of legislative elections. It can be inferred that parties are, indeed, cognizant of the value of controlling the redistricting process and that this recognition does, indeed, influence decision-making. To a lesser extent, ODT describes the patterns of other party organizations as well, but the accuracy of this conventional wisdom appears to vary by somewhat majority status and has considerable less applicability in uncompetitive states where majority status is unobtainable.

Future work should assess whether the financing activities of parties are related to the institutional and electoral value of majority party status. Although redistricting is an important power the majority party possesses once a decade, majority party legislators in every session may have policy, power, and fund-raising advantages. While some work (e.g., Kim and Phillips 2009) has attempted to quantify the value of majority party status, their conclusions are limited to a small number of states and a narrow definition of value.

Future comparative analysis of organizational strategies should also assess whether parties adjust their strategies within the election cycle. Is there evidence that parties shift strategies within campaigns in response to, for example, changes in the presidential race or the popularity of the governor? While a few previous studies (e.g., Box-Steffensmeier, Radcliffe, and Bartels 2005) have examined such patterns, these studies have been at the congressional level and have not directly addressed ODT.

Aside from its empirical consequences, the influence of the redistricting cycle also suggests an emerging pattern in which some elections are inherently more consequential than other, not due to current issues or policy agendas but merely due to the year in which they occur. This is especially troubling if the substance of electoral campaigns and party priorities are less responsive to important state matters during some election years because state budget crises and other pressing issues do not predictably occur with the regularity that the redistricting cycle imposes on the electoral process. Because such a pattern is certainly anomalous and perhaps troubling in a democracy, the full implications of the redistricting cycle demand attention from normative scholars as well.

Footnotes

Acknowledgements

The author thanks Paul Beck, Jan Box-Steffensmeier, Justin Buchler, Delia Dumitrescu, Keith Hamm, Robin Kolodny, Andrea McAtee, Kira Sanbonmatsu, Craig Volden, Alan Wiseman, and Ron Weber for helpful comments on previous drafts of this paper, and thanks Jon Winburn and Jerry Wright for sharing data.

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 Biography

Todd Makse is Assistant Professor of Political Science at Susquehanna University. His research interests include elections and voting behavior, state politics, legislative politics, and policy diffusion.

References

Barrilleaux

Charles

Holbrook

Thomas

Langer

Laura

. 2002. “Electoral Competition, Legislative Balance, and American State Welfare Policy.” American Journal of Political Science 46:415–27.

Basinger

Scott J.

Ensley

Michael J.

2007. “Candidates, Campaigns, or Partisan Conditions? Reevaluating Strategic-Politicians Theory.” Legislative Studies Quarterly 32:361–94.

Bishop

Bradford H.

Hatch

Rebecca S.

2011. “State Partisanship, Perception of State Parties, and Voting in State-Level Elections.” Working Paper.

Box-Steffensmeier

Janet M.

Radcliffe

Peter M.

Bartels

Brandon L.

2005. “The Incidence and Timing of PAC Contributions to Incumbent U.S. House Members.” Legislative Studies Quarterly 30:549–79.

Brambor

Thomas

Clark

William

Golder

Matt

. 2005. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 14:63–82.

Campbell

James E.

2011. “The Midterm Landslide of 2010: A Triple Wave Election.” The Forum 8(4): Article 3.

Carey

John M.

Niemi

Richard G.

Powell

Lynda W.

2000. “Incumbency and the Probability of Reelection in State Legislative Elections.” Journal of Politics 62:671–700.

Carsey

Thomas

Berry

William D.

Niemi

Richard G.

Powell

Lynda W.

Snyder

James M.

2009. “State Legislative Election Returns, 1967–2003.” Release Version 5.

Clarke

Phillipa

. 2008. “When can group level clustering be ignored? Multilevel models versus single-level models with sparse data.” Journal of Epidemiology and Community Health 62: 752–58.

10.

Clucas

Richard A.

1992. “Legislative Leadership and Campaign Support in California.” Legislative Studies Quarterly 17:265–83.

11.

Clucas

Richard A.

2001. “Principal-Agent Theory and the Power of State House Speakers.” Legislative Studies Quarterly 26(2): 319–338.

12.

Cox

Gary W.

Katz

Jonathan N.

2002. Elbridge Gerry’s Salamander: The Electoral Consequences of the Reapportionment Revolution. Cambridge, UK: Cambridge University Press.

13.

D’Aprile

Shane

. 2010. “GOP Could Dominate State Redistricting.” The Hill. http://thehill.com/blogs/ballot-box/house-races/112505-gop-could-dominate-state-redistricting-after-election (Accessed April 4).

14.

Dennis

Christopher

Widestrom

Amy

. 2012. “A Deep Blue Hole? California, the Tea Party, and the 2010 Midterm Elections.” In Key States, High Stakes: Sarah Palin, the Tea Party and the 2010 Midterm Election, ed. Bullock

Charles S.

III (pp. 51–64). Lanham: Rowman & Littlefield.

15.

Dwyre

Diana

Stonecash

Jeffrey M.

1992. “Where’s the Party? Changing State Party Organizations.” American Politics Quarterly 20:326–44.

16.

Francia

Peter L.

Herrnson

Paul S.

Frendreis

John

Gitelson

Alan

. 2003. “The Battle for the Legislature: Party Campaigning in State House and State Senate Elections.” In The State of the Parties: The Changing Role of Contemporary American Parties, eds. Green

John

Farmer

Rick

(pp. 171–89). Lanham: Rowman & Littlefield.

17.

Gelman

Andrew

King

Gary

Thomas

Andrew

. 2008. “JUDGEIT II: A Program for Evaluating Electoral Systems and Redistricting Plans.” Accessed May 2014 from http://cran.r-project.org

18.

Gelman

Andrew

Stern

Hal

. 2006. “The Difference between ‘Significant’ and ‘Not Significant’ Is Not Itself Statistically Significant.” The American Statistician 60:328–31.

19.

Gierzynski

Anthony

. 1992. Legislative Party Campaign Committees in the American States. Lexington: The University Press of Kentucky.

20.

Gierzynski

Anthony

Breaux

David

. 1998. “Candidate Revenues and Expenditures in State Legislative Primaries.” In Campaign Finance in State Legislative Elections, eds. Thompson

Joel A.

Moncrief

Gary F.

(pp. 80-98). Washington, DC: Congressional Quarterly Press.

21.

Gierzynski

Anthony

Jewell

Malcolm

. 1992. “Legislative Caucus and Leadership Campaign Committees.” In Changing Patterns in State Legislative Careers, eds. Moncrief

Gary F.

Thompson

Joel A.

, (pp. 107–26). Ann Arbor: University of Michigan Press.

22.

Glasgow

Garrett

. 2002. “The Efficiency of Congressional Campaign Committee Contributions in House Elections.” Party Politics 8:657–72.

23.

Greene

William

. 2003. Econometric Analysis. Upper Saddle River, NJ: Pearson Education.

24.

Heberlig

Eric S.

Larson

Bruce A.

2012. Congressional Parties, Institutional Ambition, and the Financing of Majority Control. Ann Arbor: University of Michigan Press.

25.

Herrnson

Paul S.

1989. “National Party Decision Making, Strategies, and Resource Distribution in Congressional Elections.” Western Political Quarterly 42:301–23.

26.

Herrnson

Paul S.

1996. “Party Strategy and Campaign Activities in the 1992 Congressional Elections.” In The State of the Parties, eds. Green

John C.

Shea

Daniel M.

(pp. 83–106). Lanham: Rowman & Littlefield.

27.

Herron

Michael C.

Wiseman

Alan E.

2008. “Gerrymanders and Theories of Law Making: A Study of Legislative Redistricting in Illinois.” Journal of Politics 70:151–67.

28.

Hetherington

Marc

Larson

Bruce

Globetti

Suzanne

. 2003. “The Redistricting Cycle and Strategic Candidate Decisions in U.S. House Races.” Journal of Politics 65:1221–34.

29.

Hogan

Robert E.

2005. “Gubernatorial Coattail Effects in State Legislative Elections.” Political Research Quarterly 58:587–97.

30.

Jacobson

Gary C.

1985-1986. “Party Organization and Distribution of Campaign Resources: Republicans and Democrats in 1982.” Political Science Quarterly 100:603–25.

31.

Jacobson

Gary C.

2011. “The Republican Resurgence in 2010.” Political Science Quarterly 126:27–52.

32.

Jacobson

Gary C.

Kernell

Samuel

. 1982. “Strategy and Choice in the 1982 Congressional Elections.” PS: Political Science and Politics 15:423–30.

33.

Jacobson

Gary C.

Kernell

Samuel

. 1983. Strategy and Choice in Congressional Elections. New Haven: Yale University Press.

34.

Kelleher

Christine A.

Wolak

Jennifer

. 2007. “Explaining Public Confidence in the Branches of State Government.” Political Research Quarterly 60:707–21.

35.

Kim

Henry A.

Phillips

Justin H.

2009. “Dividing the Spoils of Power: How Are the Benefits of Majority Party Status Distributed in State Legislatures?” State Politics and Policy Quarterly 9:1–23.

36.

Kousser

Thad

. 2005. Term Limits and the Dismantling of State Legislative Professionalism. Cambridge, UK: Cambridge University Press.

37.

Krasno

Jonathan S.

Green

Donald P.

1988. “Preempting Quality Challengers in House Elections.” Journal of Politics 50:20–36.

38.

LaRaja

Raymond

Orr

Susan E.

Smith

Daniel A.

2006. “Surviving BCRA: State Party Finance in 2004” in Green

John

Coffey

Daniel

, eds. The State of the Parties (5th Edition, pp. 113–34). Lanham, MD: Rowman and Littlefield.

39.

Leyden

Kevin M.

Borrelli

Stephen A.

1990. “Party Contributions and Party Unity: Can Loyalty Be Bought?” Western Political Quarterly 43:343–65.

40.

Maas

Cora J.

Hox

Joop J.

2005. “Sufficient Sample Sizes for Multilevel Modeling.|| Methodology” European Journal of Research Methods for the Behavioral and Social Sciences 1: 85–91

41.

Makse

Todd

. 2012. “Strategic Constituency Manipulation in State Legislative Redistricting.” Legislative Studies Quarterly 37:425–50.

42.

McDonald

Michael P.

2004. “A Comparative Analysis of Redistricting Institutions in the United States, 2001-2002.” State Politics and Policy Quarterly 4:371–95.

43.

Medvic

Stephen K.

2010. Campaigns and Elections: Players and Processes. Boston: Wadsworth.

44.

Moncrief

Gary F.

1998. “Candidate Spending in State Legislative Races.” In Campaign Finance in State Legislative Elections, eds. Thompson

Joel A.

Moncrief

Gary F.

(pp. 37–58). Washington, DC: Congressional Quarterly Press.

45.

Richardson

Lilliard E.

Jr. Konisky

David M.

Milyo

Jeffrey

. 2012. “Public Approval of State Legislatures.” Legislative Studies Quarterly 37:99–116.

46.

Rosenthal

Cindy S.

1995. “New Party or Campaign Bank Account? Explaining the Rise of State Legislative Campaign Committees.” Legislative Studies Quarterly 20:249–68.

47.

Salka

William M.

2009. Reforming State Legislative Elections: Creating a New Dynamic. Boulder: Lynne Rienner.

48.

Schecter

David L.

Hedge

David M.

2001. “Dancing with the One Who Brought You: The Allocation and Impact of Party Giving to State Legislators.” Legislative Studies Quarterly 26:437–56.

49.

Shea

Daniel M.

1995. Transforming Democracy: Legislative Campaign Committees and Political Parties. Albany: State University of New York Press.

50.

Sigelman

Lee

Zeng

Langche

. 1999. “Analyzing Censored and Sample-Selected Data with Tobit and Heckit Models.” Political Analysis 8:167–82.

51.

Stonecash

Jeffrey M.

1988. “Working at the Margin: Campaign Finance and Party Strategy in New York Assembly Elections.” Legislative Studies Quarterly 13:477–93.

52.

Stonecash

Jeffrey M.

Keith

Sara E.

1996. “Maintaining a Political Party: Providing and Withdrawing Party Campaign Funds.” Party Politics 2:313–28.

53.

Storey

Tim

. 2000. “Election 2000: The States are Crucial.” State Legislatures, September.

54.

Thompson

Joel A.

Cassie

William

. 1992. “Party and PAC Contributions to North Carolina Legislative Candidates.” Legislative Studies Quarterly 17:409–16.

55.

Thompson

Joel A.

Cassie

William

Jewell

Malcolm E.

1994. “A Sacred Cow or Just a Lot of Bull? Party and PAC Money in State Legislative Elections.” Political Research Quarterly 47:223–37.

56.

Winburn

Jonathan

. 2008. The Realities of Redistricting: Following the Rules and Limiting Gerrymandering in State Legislative Redistricting. Lanham: Lexington Books.

57.

Wiseman

Alan

. 2005. “Partisan Strategy and Support in State Legislative Elections: The Case of Illinois.” American Politics Research 33:376–403.

58.

Witko

Christopher

. 2005. “Measuring the Stringency of State Campaign Finance Regulation.” State Politics and Policy Quarterly 5:295–310.

59.

Wright

Gerald

. 2004. “Representation in America’s Legislatures.” Indiana University, Bloomington and National Science Foundation, Arlington, VA.