State Policy Outcomes and State Legislative Approval

Abstract

Does the public’s approval of their state legislature reflect their satisfaction with the outputs of state government? Using survey responses from the 2014 Cooperative Congressional Election Study, we consider the roots of public approval of state legislatures. We find that people are more likely to voice approval of their state legislature when it produces policy outcomes that correspond with their interests. Liberals view their state legislature more positively when policy outputs are liberal, while conservatives evaluate their state legislature more favorably when policy outcomes are conservative. These effects are the most pronounced among those who are the most knowledgeable about state politics. Using panel data from 2012 to 2014, we also show that changes in state policy liberalism are associated with changes in state legislative approval. Even though we have reasons to be pessimistic about the quality of citizens’ assessments of state government, our results demonstrate that citizens evaluate their state legislatures based on the policy outcomes they provide.

Keywords

state legislative approval state politics public opinion

In a representative democracy, we delegate decisions to elected officials to act on our behalf. We hope that politicians will work to produce policies that reflect our preferences, but we risk receiving outcomes that are not well-aligned with our political desires. For scholars, the congruence between the choices made by political elites and the preferences of the public inform the performance of representative democracy (e.g., Erikson, MacKuen, and Stimson 2002; Erikson, Wright, and McIver 1993). Yet even as scholars are interested in the character of policy representation afforded by officeholders, does the same hold true for citizens?

Although we have often assumed that policy responsiveness is important to citizens, questions remain about whether people evaluate governmental institutions based on the policy outcomes that they provide. When asked about their expectations of officeholders, people say they want politicians to listen to their concerns and vote in line with the concerns of constituents (Grill 2007; Hibbing and Theiss-Morse 2002). At the national level, this seems to be confirmed. We see that members of Congress who vote in ways congruent with their constituents benefit at the ballot box (Ansolabehere and Jones 2010). As the distance between policy outcomes and the liberalism of the public grows, congressional approval falls (Ramirez 2013). At the state level, however, it is less evident that policy outcomes matter for how people assess state elected officials. At the dyadic level, scholars find only weak evidence that state legislators are punished or rewarded at the polls for their ideological congruence with their districts (Hogan 2008; Rogers 2017). We consider the public’s responsiveness to state policy outcomes, exploring whether people evaluate state legislatures based on the ideological tenor of the policy outcomes they provide.

Are people more likely to approve of their state legislature when it delivers outcomes that come closer to their ideological preferences? Or are state policy outcomes unimportant to how people evaluate the performance of their state legislature? Using responses from the Cooperative Congressional Election Study (CCES), we explore whether people report higher approval of their state legislatures when their preferences are better aligned with the policy outcomes of state government. We find that increasing levels of state policy liberalism are associated with greater state legislative approval among liberals and lower levels of approval among conservatives. We also find that those with greater knowledge of state politics are more responsive to the ideological outcomes of state government than those with lower levels of political knowledge. Using the 2012–2014 CCES panel survey, we show that changes in state policy liberalism are associated with changes in state legislative approval over time.

Investigating the consequences of state policy outcomes for public approval is important for understanding the character of representative democracy in the states. Over time, states have taken an increasingly active role in policymaking, and hold policy authority in many domains relevant to citizens’ lives—from education to environmental issues to infrastructure and transportation. The decision to pass one policy versus another has important consequences—and as such, it is valuable to know if people’s evaluations of the performance of their state legislatures are responsive to those policy choices.

This research also informs our understanding of the connections between public evaluations and policy outcomes in the states. Scholars have affirmed that public preferences inform the policy choices made by state elected officials (Caughey and Warshaw 2018; Erikson, Wright, and McIver 1993). Yet this connection is not a perfect one, as the relationship between policy preferences and political outcomes varies across states and over time (Caughey and Warshaw 2018; Lax and Phillips 2012; Pacheco 2013; Wright and Winburn 2002). In considering this heterogeneity in policy responsiveness, it seems that policy representation works differently at the state level than at the national level (Erikson, Wright, and McIver 1993). Although elections are important to helping citizens obtain congruent policy outcomes from national institutions (Erikson, MacKuen, and Stimson 2002), they seem less important to encouraging policy responsiveness in the states (Caughey and Warshaw 2018). State electoral turnover is not necessarily responsive to the policy choices made by state legislators (Hogan 2008; Rogers 2017).

Given these reservations about the strength of the tie between the will of the state electorates and the outputs of state government, it is important to consider whether people evaluate state legislatures based on the policy outcomes of the states. In finding that people are responsive to the ideological tenor of state policy outcomes, our results suggest that people are better able to hold state legislatures responsible for their policy outputs than has been previously believed. Even in a low-information domain like state politics where the dyadic ties between state representatives and their constituents are weak, state electorates respond to state outcomes in ways that have the potential to promote collective accountability.

Origins of State Legislative Approval

Why do people approve of their state legislatures? Broadly speaking, people are more likely to approve of their statehouse when they feel that state officeholders are responsive to their concerns (Patterson, Hedlund, and Boynton 1975; Patterson, Ripley, and Quinlan 1992). In considering the predictors of state legislative ratings, approval follows in part from the performance of state government. When economic optimism is high, unemployment rates are low, and the fiscal health of the state is strong, people report higher levels of trust and approval of their state legislature (Flavin 2013; Hamman 2006; Kelleher and Wolak 2007; Richardson, Konisky, and Milyo 2012).

People’s evaluations of their state legislature also reflect the character of policy processes in the state. At the national level, mistrust of Congress is thought to be rooted in fears that legislators are out of touch with the demands of the public, focused on partisan squabbles and interest group demands rather than the concerns of the electorate (Hibbing and Theiss-Morse 1995, 2002). As state legislatures increasingly resemble Congress in terms of their legislative professionalism, with longer sessions and larger staffs, their approval falls. People offer warmer evaluations of citizen legislatures than professionalized ones (Kelleher and Wolak 2007; Richardson, Konisky, and Milyo 2012; Squire 1993).

People also lean on their partisanship when asked to offer their appraisals of their state legislature. People offer warmer ratings of their state legislature when the institution is controlled by members of their own party, and provide more negative assessments of state legislatures controlled by members of the rival party (Banda and Kirkland 2018; Flavin 2013; Richardson, Konisky, and Milyo 2012).¹

State Policy Outcomes and Constituent Response

To what extent do people consider state policy outcomes when considering their feelings about their state legislature? There are reasons to be pessimistic that citizens are responsive to the ideological tenor of state policy outcomes. Voters are often uninformed about matters of state politics, where nearly half of Americans are unable to name any branch of state government and even fewer can name their state legislators (Lyons, Jaeger, and Wolak 2013; Songer 1984). Media coverage of state politics is often low, reflecting the dwindling presence of reporters in state capitals as well as the general decline of newspaper reporting (Wilson 2014). If people are unaware of the debates being waged in their state legislature, then the outputs of state government may not have much of an imprint on how people evaluate state institutions. In considering the low levels of citizen knowledge of the specifics of state legislatures, Squire and Moncrief (2015, 210) pessimistically conclude that “although many people may express an opinion about their legislature’s job performance, there might not be much substance behind those evaluations.”

From the perspective of state legislators, public preferences are important to how votes are cast, but few legislators voice worries about being held accountable for their votes in the statehouse (Jewell 1982). Given the uncontroversial nature of many state legislative bills as well as their limited salience to the electorate, state legislators do not see policy issues as particularly important to their chances for reelection. Nearly all incumbents who seek reelection to the state legislature win it, and many state legislative elections go uncontested.

Studies of state legislative election results have also raised doubts about whether the ideological leanings of state legislators are related to voters’ choices at the ballot box. In exploring the effects of state legislator ideology on a lawmaker’s vote share and odds of reelection, Rogers (2017) finds only modest evidence of electoral punishment for state legislators who are ideologically incongruent with their districts. For state legislators in safe districts and in states with weak media presence in the state capitol, ideological incongruence may well go unpunished by voters. Likewise, Hogan (2008) finds only modest evidence that ideologically extreme state legislators are any more likely to face strong challengers at the next election as compared with state lawmakers more in step with their districts. That said, in looking at legislators’ roll call votes, Birkhead (2015) uncovers stronger evidence of a tie between state legislative ideology and the preferences of state electorates. He finds that ideologically extreme state legislators receive lower vote shares than moderate representatives, and are more likely to face opposition at the next election.

Rather than considering the dyadic ties between districts and their state legislators (Hogan 2008; Rogers 2017), we instead examine constituents’ reactions to collective policy outcomes in their state, and the degree to which people evaluate their state legislature in response to the ideological tenor of state policy outcomes. We argue that this kind of collective accountability is also important to state politics. When asking state legislators how they see their roles as representatives, many see themselves as representatives of state interests rather than district ones (Wahlke et al. 1962). We also know that even when dyadic ties are weak, citizens’ preferences can still be reflected in policy outcomes through collective representation (Weissberg 1978). Indeed, when asking citizens about what kind of representation they prefer, experimental studies have shown that citizens value collective representation even more than they value dyadic representation (Harden and Clark 2016).

At the national level, scholars have confirmed that people are responsive to the ideological tenor of collective policy outcomes, adjusting their demands of government in response to liberalism of policy outputs (Erikson, MacKuen, and Stimson 2002; Wlezien 1995). People also calibrate their attitudes toward governmental institutions based on the policies those institutions produce. Congressional approval responds to policy outcomes, where the public tempers their approval of Congress when they feel as if there is a discrepancy between their partisan policy preferences and Congress’s failure to meet them (Ramirez 2009). When the ideological mood of the country diverges from policy results, the public adjusts its evaluations of Congress accordingly (Ramirez 2013).

At the state level, scholars have shown that the ideological content of state policy outcomes informs both policy demands and appraisals of government. In considering thermostatic responsiveness to social policy, Pacheco (2013) finds that people’s policy preferences respond to the ideological tenor of previous state policy outcomes. Although scholars have not considered the effects of state policy liberalism for legislative approval, they have shown that the ideological leanings of the state parties are consequential for how people appraise government. Banda and Kirkland (2018) find that those in states with ideologically polarized parties are less likely to say that they trust their state legislature, while Wolak and McAtee (2013) find partial evidence that the liberalism of the state parties is related to people’s approval of the political parties in their state. Flavin (2013) also finds that ideological similarities between citizens and their state government can cultivate stronger feelings of trust in state government.

We consider how the policy outcomes in the states inform people’s evaluations of their state legislature. We expect that liberals will be more likely to approve of their state legislature as the liberalism of state policy increases, while conservatives will reveal higher approval as the conservatism of policy outcomes increases. Even if people find it difficult to punish or reward their state representatives based on their specific job performance, we believe that people are better equipped to provide feedback to state legislatures based on their collective policy outputs. Voters might not know how their representative in the state legislature is voting on legislation, but the general ideological tenor of state policy-making is more easily observed. Just as citizens form impressions about the ideological tenor of national political outcomes (Erikson, MacKuen, and Stimson 2002), we expect that they also take cues from their state environment about the ideological directions of state policy outcomes.

Even with declines in statehouse reporting, people still learn about the happenings of state government from the news (Johnson 2013). Although state representatives may not receive much individualized news attention, the policy debates of the state capitol are more likely to capture media coverage. The ideological signals from state legislatures have arguably only sharpened in a time of increased party polarization (Shor and McCarty 2011) and sometimes only limited electoral competition (Shufeldt and Flavin 2012). We see this in people’s knowledge of state politics, as many know the party balance in state government even if they do not know the partisanship of their state representative (Patterson, Hedlund, and Boynton 1975). Indeed, people’s ability to correctly identify partisan control of their state legislature is only modestly lower than their accuracy in identifying partisan control of Congress.² Even though national politics receives far more press attention than state politics, Americans nonetheless still express a strong interest in state politics. In a 2014 survey, Americans were nearly as likely to say they were interested in keeping up with news about state politics as they were to express an interest in following news about the federal government.³

We expect that state constituents are interested in state outcomes and able to draw on what news they see about state government to make inferences about the general ideological direction of policy outcomes in their state. Although they may not be well versed on the bills that come before the state legislature, we believe that they are able to gauge whether state policy outcomes are moving to the right or to the left. We further expect that the effects of state policy liberalism on state legislative approval will be greatest among those most knowledgeable about their state legislature.

Data and Measures

To investigate citizens’ responsiveness to the policy outcomes of state government, we rely on responses to the 2014 CCES. In the preelection wave of the survey, respondents were asked whether they approved of how their state legislature is doing their job, on a 4-point scale ranging from strongly disapprove to strongly approve.⁴ In 2014, few voiced enthusiastic support for their state legislature. Only 8 percent said they strongly approved of the job their state legislature was doing, and 39 percent said they somewhat approved.⁵ Disapproval is more common than approval, with 25 percent voicing strong disapproval and 29 percent saying they somewhat disapproved.

Because we are interested in variations in levels of approval of state legislatures across the states as well as how approval varies across contexts as a function of respondent ideology, we rely on multilevel modeling and a random intercept random coefficient specification. Given clustering in survey responses at the state level, standard estimation approaches run the risk of underestimated standard errors and false positives associated with the effects of level-2 variables. By using a multilevel modeling approach, we minimize those risks (Steenbergen and Jones 2002).⁶

We consider the imprint of ideology on state legislative approval using a 7-point measure of self-reported respondent ideology.⁷ Using a random coefficient specification, we allow the effects of ideology to vary across state contexts, and model that variation as a function of the liberalism of policy outcomes in the state. As our measure of the ideological tenor of state policy, we rely on state policy liberalism scores for 2014, as created by Caughey and Warshaw (2016). Using a measurement model and data on 148 policy outcomes across the states, they create annual scores of state policy liberalism. Higher scores on this measure indicate greater liberalism in state policy outcomes. If people’s approval of their state legislature follows from the policy outputs of state government, we should expect to see that liberals who live in states with more liberal policy outcomes will be more likely to approve of their state legislature than conservative respondents in that state. Conservative respondents should likewise offer warmer ratings of their state legislature when they reside in states with conservative policy outputs.

We include controls for party control of state government, to distinguish whether people are uniquely responsive to the ideological tenor of state policy outcomes apart from patterns of party control of state government. We include an item indicating whether the respondent lives in a state where their own party controls state government and a second reflecting those who reside in a state where the rival party controls state government.⁸ Because stronger state economies are traditionally associated with higher approval of state legislatures, we control for state unemployment rates.⁹ We also control for levels of state legislative professionalism, given that people tend to be less trusting and less approving of professionalized state legislatures (Kelleher and Wolak 2007; Richardson, Konisky, and Milyo 2012). We rely on scores created by Bowen and Greene (2014).¹⁰ Higher scores indicate professionalized state legislatures with longer sessions and higher salaries, while lower scores are tied to less professionalized citizen legislatures. We expect to see lower levels of state legislative approval in states with highly professionalized legislatures.

We also include a control for party competition in the states. At the national level, party conflict tends to be associated with lower levels of confidence in Congress (Durr, Gilmour, and Wolbrecht 1997; Ramirez 2009). But at the state level, measures of divided government and party conflict seem unrelated to people’s approval of state legislatures (Banda and Kirkland 2018; Hamman 2006; Kelleher and Wolak 2007). In fact, some have suggested that party competition might promote positivity toward state government, as people are more likely to feel that their views are represented in politics in states with higher levels of party competition (Flavin and Shufeldt 2016; Wolak 2018). We test this by including a control for the level of party competition in the states. We rely on an index of state party competition created by Hinchliffe and Lee (2016) constructed from data on the closeness of elections for state legislative seats and governor’s races. The measure is reverse-coded, where high scores on this measure indicate places where one party dominates in state elections and low scores represent states with strong two-party competition. At the individual level, we include controls for the respondents’ attention to politics as well as their knowledge of partisan control of the state legislature, to explore whether familiarity of state politics is associated with greater criticism (Hibbing and Theiss-Morse 1995).¹¹ We control for congressional approval to capture individual differences in negativity of affect toward government. Finally, we control for respondents’ educational attainment, age, race, and gender.¹²

Multilevel ordered logit results are shown in Table 1. We confirm a significant interaction effect associated with respondent ideology and the liberalism of state policy outcomes. In the appendix, we plot the marginal effects for state policy liberalism, which demonstrate that policy outcomes inform state legislative approval for all but political moderates. For liberals, increasing policy liberalism is associated with greater approval of state legislatures, while for conservatives, increasing liberalism is associated with declines in state legislative approval.

Table 1.

Policy Congruence and State Legislative Approval.

	State legislative approval, 2014
State policy liberalism	−0.524* (0.062)
Ideology	−0.029 (0.028)
Policy Liberalism × Ideology	0.178* (0.019)
Own party control of state government	1.058* (0.064)
Opposing party control of state government	−0.435* (0.084)
Legislative professionalism	−0.081* (0.037)
State unemployment rate	−0.141* (0.065)
History of single party control of state politics	−2.190* (0.843)
Congressional approval	3.163* (0.134)
Follows politics	0.302* (0.068)
Knowledge of partisan control of state legislature	−0.067* (0.018)
Education	−0.124* (0.064)
Age	0.000 (0.001)
Nonwhite	0.094* (0.042)
Female	0.020 (0.033)
Cutpoint 1	−1.598* (0.425)
Cutpoint 2	0.164 (0.425)
Cutpoint 3	3.258* (0.423)
Variance components
Variance, intercept	0.424* (0.085)
Variance, ideology	0.037* (0.007)
Covariance, intercept, and ideology	−0.105* (0.024)
Deviance	91,572
N (number of states)	42,442 (48)

Source. Cooperative Congressional Election Study (2014).

Multilevel ordered logit estimates, standard errors in parentheses.

p < .05.

In Figure 1, we show predicted levels of state legislative approval across the range of state policy outcomes for those who identify as very liberal as well as those who identify as very conservative. To simplify the presentation, our plot sums the probability of strongly approving and somewhat approving of the state legislature to show overall probability of approving one’s state legislature across different state contexts.¹³ Consider the case of a state with particularly conservative state policy outcomes. In this context, a person who identifies as very conservative would have a 70 percent likelihood of approving of their state legislature, while a person who self-identifies as very liberal has only a 21 percent predicted probability of approving of their state legislature. As the liberalism of the state’s policy outcomes increases, the approval ratings of conservatives decline as the ratings of liberals increase. For a state at the highest level of policy liberalism, a strong conservative has only a 25 percent probability of approving of the state legislature, while a strong liberal has a 68 percent probability of approving of the legislature.¹⁴ For ideologues, their views of their state legislature reflect in part the ideological tenor of the policy outcomes that the state provides.

Figure 1.

Predicted approval of state legislatures, by state policy liberalism.

In Table 2, we report the predicted probabilities of state legislative approval at high and low levels of the control variables. We confirm that people are more likely to approve of their state legislature when state government is controlled by members of one’s own party, and less likely to approve when party control is held by the opposing party. Likewise, state economic health has the expected effect on state legislative approval, where moving from the highest level of state unemployment to the lowest increases the predicted probability of approving of the state legislature from 39 to 54 percent.

Table 2.

Predicted Probabilities of State Legislative Approval.

	Pr(Strongly approve)	Pr(Somewhat approve)	Pr(Somewhat disapprove)	Pr(Strongly disapprove)
Own party control of state government	.08	.55	.28	.10
Opposing party control of state government	.02	26	.40	.32
Divided government	.03	.35	.39	.23
Minimum, state unemployment rate	.05	.49	.32	.13
Maximum, state unemployment rate	.03	.36	.39	.22
Minimum, state legislative professionalism	.05	.46	.35	.15
Maximum, state legislative professionalism	.02	.30	.40	.27
Minimum, party competition in state elections	.02	.30	.40	.27
Maximum, party competition in state elections	.04	.45	.35	.16
Minimum, knowledge of state politics	.04	.42	.37	.18
Maximum, knowledge of state politics	.03	.39	.38	.20
Minimum, follows politics	.03	.35	.39	.22
Maximum, follows politics	.04	.42	.37	.18
Minimum, congressional approval	.02	.26	.40	.33
Maximum, congressional approval	.29	.61	.09	.02
Less than a high school degree	.04	.42	.37	.18
Graduate degree	.03	.39	.38	.20
Nonwhite	.04	.40	.38	.19
White	.04	.42	.37	.18

Source. Cooperative Congressional Election Study (2014).

Predicted probabilities based on results in Table 1, holding other variables at their means.

Consistent with prior studies, we confirm that people report higher approval of amateur state legislatures compared with professionalized ones. Interestingly, people report higher approval of state legislatures in states with a history of party competition for control for state government than in places dominated by one-party control of state government. This is contrary to national-level evidence that partisan conflict hurts congressional approval (Durr, Gilmour, and Wolbrecht 1997), and is consistent with state-level evidence that party competition is valued in the states as a signal of responsiveness to voters (Flavin and Shufeldt 2016; Wolak 2018). Those who express greater interest in politics are more likely to approve of the job their state legislature is doing, while greater education and higher political knowledge tend to be associated with more negative ratings of the state legislature. Generalized favorability toward government, as captured by our indicator of congressional approval, is associated with greater approval of the state legislature as well.

Political Knowledge and the Effects of State Policy Liberalism

We next consider the possibility of heterogeneity in the effects of ideology and state policy outcomes, conditional on people’s levels of political sophistication. We have proposed that citizens monitor the outcomes of state government and use the information they gather about policy outcomes to inform their evaluations of state legislative performance. If this is the case, then we should expect to see greater responsiveness to state policy liberalism among those with higher levels of knowledge of state politics. To test this, we interact state policy liberalism and respondent ideology with our measure of knowledge of partisan control of the state legislature. In Table 3, we confirm a significant three-way interaction effect, indicating that the relationship between ideology and state liberalism varies with respondents’ knowledge of state politics.

Table 3.

Political Knowledge, Policy Congruence, and State Legislative Approval.

	State legislative approval, 2014
State policy liberalism	−0.053 (0.072)
Ideology	0.070* (0.028)
State political knowledge	0.116 (0.075)
Policy Liberalism × Ideology	0.024 (0.023)
Policy Liberalism × State Political Knowledge	−0.400* (0.047)
Ideology × State Political Knowledge	−0.096* (0.025)
Policy Liberalism × Ideology × Political Knowledge	0.144* (0.017)
Own party control of state government	1.059* (0.060)
Opposing party control of state government	−0.334* (0.078)
Legislative professionalism	−0.068* (0.023)
State unemployment rate	−0.188* (0.045)
History of single party control of state politics	−2.249* (0.794)
Follows politics	0.327* (0.065)
Education	−0.130* (0.067)
Congressional approval	3.138* (0.112)
Age	0.000 (0.001)
Nonwhite	0.088* (0.044)
Female	0.035 (0.034)
Cutpoint 1	−1.664* (0.339)
Cutpoint 2	0.137 (0.342)
Cutpoint 3	3.278* (0.336)
Variance components
Variance, intercept	0.305* (0.092)
Variance, ideology	0.044* (0.009)
Variance, state political knowledge	0.009* (0.005)
Covariance, intercept, and ideology	−0.086* (0.031)
Covariance, intercept, and political knowledge	−0.012 (0.020)
Covariance, ideology, and political knowledge	−0.001 (0.006)
Deviance	90,289
N (number of states)	42,442 (48)

Source. Cooperative Congressional Election Study (2014).

Multilevel ordered logit estimates, standard errors in parentheses.

p < .05.

In Figure 2, we plot the marginal effects of state policy liberalism over the range of respondent ideology for those at high, low, and median levels of state political knowledge. We find the strongest marginal effects for state policy liberalism among those who are most knowledgeable about their state legislature. Among those at the median level of political knowledge, we find a significant marginal effect for state policy liberalism over most of the range of respondent ideology, but the effect is smaller in magnitude than is observed among those with higher state political knowledge. Among those who are at the lowest level of political knowledge, we fail to find a significant marginal effect for state liberalism over the range of respondent ideology. The effects of state policy liberalism on state legislative approval are concentrated among those who are more knowledgeable about politics in their state.

Figure 2.

Marginal effect of state policy liberalism of state legislative approval, by level of state political knowledge and ideological identification.

Changes in Policy Liberalism and Changes in State Legislative Approval

We find that liberals offer warmer ratings of their state legislature when they reside in states with liberal policy outcomes and that conservatives report higher approval when residing in states with conservative policy outcomes, and that the strength of this association increases with the sophistication of state constituents. Because our results are robust to controls for patterns of party control of state government, this suggests that the effects of state policy reflect more than just the congruence between conservative state governments and their conservative-leaning electorates. But to further demonstrate that the public is responsive to the ideological content of state policy outcomes, we next look at the dynamics of state legislative approval. Using panel data, we consider how people’s approval of their state legislature varies as a function of the changing liberalism of policy outcomes in the states.

We rely on the subset of respondents that participated in both the 2012 and 2014 CCES as participants in the panel survey portion of the CCES.¹⁵ As our dependent variable, we consider changes in people’s approval of their state legislature from 2012 to 2014. The measure ranges from −3 to 3. Higher scores indicate those who offered higher levels of approval in 2014, while negative scores reflect those whose opinions of their state legislature declined over the period. To explore how people respond to policy changes in their state, we use a differenced measure of changes in state policy outcomes from 2012 to 2014, again using the annual estimates of state policy liberalism from Caughey and Warshaw (2016). By looking at only the changes in policy liberalism, we essentially remove the distinctive histories of the states that have lead them to have a slate of liberal or conservative policy traditions and focus instead on the contemporary political climate of the states. It represents a tough test of our theory, as changes in policy liberalism tend to be modest over this time period.

We interact state changes in policy liberalism with the respondents’ self-reported ideology. We expect that conservatives will increase their support of their state legislature when policy outcomes move to the right, while liberals will be more likely to approve of their state legislature when policy outcomes take a left turn between 2012 and 2014.¹⁶ We control for shifts in partisan control of state government, as well as changes in state unemployment rates from 2012 to 2014.¹⁷ We include measures of changes in political attention, knowledge, and congressional approval, and the same slate of demographic controls as in Table 1.

Results of the random intercept random coefficient multilevel regression model are shown in Table 4.¹⁸ We find that changes in state policy liberalism are associated with changes in public approval of state legislatures.¹⁹ In states where policy outcomes shifted to the right, strong conservatives are more likely to approve of their state legislature and strong liberals offer more negative evaluations of their state legislature. In Figure 3, we show the marginal effects of changes in state policy liberalism across the range of respondent ideology. For those who say they are liberal or very liberal, greater policy liberalism is significantly related to higher levels of approval, as indicated by the significant positive marginal effects. For moderates and left-leaning centrists, changes in state policy liberalism are unrelated to changes in state legislative approval. For those who identify as conservatives, changes in policy liberalism have a significant negative marginal effect, one that increases in size with the strength of conservative identification.

Table 4.

The Effects of Changes in State Policy Liberalism on Changes in State Legislative Approval.

	Change in state legislative approval, 2012 to 2014
Change in state policy liberalism over period	−0.343* (0.079)
Ideology	0.023 (0.019)
Ideology × Change in State Policy Liberalism	0.106* (0.030)
Change in own party control of state government	0.122 (0.082)
Change in opposing party control of state government	−0.216* (0.088)
Change in state unemployment rate over period	−0.127* (0.047)
Change in congressional approval	0.464* (0.056)
Change in attention to politics	−0.010 (0.121)
Change in state political knowledge	0.018 (0.019)
Education	0.060 (0.054)
Age	0.002 (0.001)
Nonwhite	0.015 (0.051)
Female	0.008 (0.028)
Constant	−0.420* (0.140)
Variance components
Variance, intercept	0.106* (0.041)
Variance, ideology	0.014* (0.006)
Covariance, intercept, and ideology	−0.034* (0.013)
Deviance	13,927
N (number of states)	7,898 (49)

Source. Cooperative Congressional Election Study (2012–2014) panel.

Multilevel regression estimates, standard errors in parentheses.

p < .05.

Figure 3.

Marginal effects of changes in state policy liberalism on changes in state legislative approval.

To demonstrate the size of these effects, consider two states in the sample: Mississippi and New Jersey. Each saw a similar magnitude shift in policy liberalism from 2012 to 2014—but in opposing directions, as state policy outcomes moved to the right in Mississippi and to the left in New Jersey.²⁰ How do these policy changes connect to people’s evaluations of their state legislature? The effects depend on the ideological leanings of the state constituent. When New Jersey moves left on state policy from 2012 to 2014, strong liberals in the state offer a warmer evaluation of their state legislature, with a predicted level of approval that is 0.08 points greater, all else equal. For strong conservatives in the same state, public approval of the state legislature is predicted to be 0.09 points lower. The converse holds in the case of Mississippi’s move toward more conservative public policy outcomes from 2012 to 2014. In a case where state policy outcomes shifted to the right, strong conservatives in Mississippi report warmer evaluations of their state legislature, with a predicted level of approval that is 0.09 points greater, all other variables at their means. Strong liberals in the state seem to express their displeasure at this right turn, where predicted state legislative approval is 0.07 points lower, all else equal. Changes in state policy outcomes are associated with changes in how people evaluate their state legislature.

It is important to note that the magnitude of these effects are not large. Given a standard deviation’s increase in state policy liberalism, we would expect that a strong liberal’s approval to be 0.12 points greater, while a strong conservative’s approval would be 0.14 points lower, all else equal. This represents about a fifth of a standard deviation change on a measure that runs from −3 to 3. This small effect size reflects in part the relative stability of people’s levels of state legislative approval over this period, as about 60 percent of the sample offers the same rating of their state legislature in 2012 as in 2014.²¹ Even so, the effects of state policy liberalism remain comparable in magnitude to the effects of changing state economic conditions. Over this period, a standard deviation’s increase in state unemployment rates would lead to an expected drop in state legislative approval by 0.09 points. Compared with shifts in party control of state government, the effects of shifts in state policy liberalism are smaller in magnitude. For those in a state that shifted from a time of divided party control in the 2012 survey to unified out-party control in the 2014 survey wave, we would expect a drop in state legislative approval of 0.22 points. In sum, we find that changes in economic conditions, partisan control of state government, and the ideological tenor of state policy outcomes all inform changes in public approval of state legislatures.

Conclusion

We demonstrate that the ideological tenor of state policy outcomes informs people’s evaluations of their state legislature. Liberals are more likely to approve of their state legislature when state governments deliver liberal policy outcomes, while conservatives offer warmer ratings of state legislatures in the face of conservative policy outcomes. As state policy outcomes shift to the right or left, people seem to update their ratings of the state legislature, giving more favorable ratings to state legislative chambers in states where policy outcomes have shifted in the direction of their own ideological leanings.

These findings are important evidence of the responsiveness of state electorates to the policy actions of state governments. We have reason to be pessimistic about state electorate’s attentiveness to state political outcomes. Because states can be low-information environments, we may worry about how much voters know about state government, and the degree to which state legislative elections ultimately serve as checks on the choices made by state representatives. Even so, our results indicate that citizens react to the policy outcomes that their state governments produce. This is consistent with other evidence that suggests that citizens use cues and heuristics to hold officeholders responsible in other low-information domains. Even as we know that many voters might not be able to recall the specific votes cast by their member of Congress, they are able to hold their representative accountable for casting out-of-step votes (Ansolabehere and Jones 2010). And even as many may struggle to name the specific bills that populate the agenda at the state legislature, our evidence suggests that citizens are responsive to the content of state policy outcomes overall. Although voters might not always use their state legislators’ votes to inform their choices on Election Day (Hogan 2008; Rogers 2017), their feelings about the state legislature are responsive to the ideological tenor of policy outcomes in the aggregate. Because people’s approval of their state legislature is associated with their vote choices in state legislative elections (Rogers 2016), these results suggest a possible pathway of how state legislatures may be held accountable for state policy outcomes.

Our results push back against popular assumptions that people are singularly partisan in how they think about their state legislature. Although we confirm that people like state legislatures more when the state government is controlled by members of one’s own party, we also show that people are responsive to more than just shared partisanship. For state legislatures, this means that the policy outcomes they produce are consequential for their standing in the eyes of the public. When state legislatures work to move policy outcomes to the left or to the right, citizens respond to these changes. In this way, members of state legislative bodies have influence over how the institution is perceived. In passing legislation that appeals to as many state constituents as possible, state legislatures can increase their popularity in the electorate.

Our results also help explain overall levels of public approval of state legislatures. Surveys have regularly shown that more people disapprove of their state legislature than approve of it (Rosenthal 2008; Squire and Moncrief 2015). This might reflect people’s general cynicism toward legislatures (Hibbing and Theiss-Morse 1995). But if approval of state legislatures is responsive to the ideological content of state policy outcomes, then it is perhaps less surprising to find high levels of pessimism about state institutions. In pushing policy outcomes in a liberal direction, state legislatures will not only secure support among liberals in the electorate but also inspire the ire of conservative constituents. Ideological policy change in either direction will tend to leave opposing segments of the electorate unhappy. Given this, states will find it challenging to secure broad support among both liberal and conservative ideologues in the electorate. Even as states are colored red or blue on electoral maps, state electorates remain ideologically heterogeneous. In Gallup surveys, even the reddest states on the map have less than a majority of self-identified conservatives (Newport 2015). As levels of ideological polarization in the electorate climb (Shor and McCarty 2011), it will likely become even more difficult for state legislatures to boost their standing in the public though policy changes.

Supplemental Material

Supplemental_Appendix_(1) – Supplemental material for State Policy Outcomes and State Legislative Approval

Supplemental material, Supplemental_Appendix_(1) for State Policy Outcomes and State Legislative Approval by Stefani Langehennig, Joseph Zamadics and Jennifer Wolak in Political Research Quarterly

Footnotes

Appendix

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

Supplemental Material

Replication code and data available at: . Supplemental materials for this article are available with the manuscript on the Political Research Quarterly (PRQ) website.

ORCID iD

Jennifer Wolak

References

Ansolabehere

Stephen

Jones

Philip Edward

. 2010. “Constituents’ Responses to Congressional Roll-Call Voting.” American Journal of Political Science 54:583–97.

Ansolabehere

Stephen

Schaﬀner

Brian F.

2014. Cooperative Congressional Election Study, 2014: Common Content. Cambridge, MA: Harvard University.

Ansolabehere

Stephen

Schaﬀner

Brian

. 2014. Cooperative Congressional Election Study: 2010-2014 Panel Study. Amherst, MA: University of Massachusetts, Amherst.

Banda

Kevin K.

Kirkland

Justin H.

2018. “Legislative Party Polarization and Trust in State Legislatures.” American Politics Research 46:596–628.

Birkhead

Nathaniel A.

2015. “The Role of Ideology in State Legislative Elections.” Legislative Studies Quarterly 40:55–82.

Bowen

Daniel C.

Greene

Zachary

. 2014. “Should We Measure Professionalism with an Index? A Note on Theory and Practice in State Legislative Professionalism Research.” State Politics & Policy Quarterly 14:277–96.

Caughey

Devin

Warshaw

Christopher

. 2016. “The Dynamics of State Policy Liberalism, 1936–2014.” American Journal of Political Science 60:899–913.

Caughey

Devin

Warshaw

Christopher

. 2018. “Policy Preferences and Policy Change: Dynamic Responsiveness in the American States, 1936–2014.” American Political Science Review 112:249–66.

Durr

Robert H.

Gilmour

John B.

Wolbrecht

Christina

. 1997. “Explaining Congressional Approval.” American Journal of Political Science 41:175–207.

10.

Erikson

Robert S.

MacKuen

Michael B.

Stimson

James A.

2002. The Macro Polity. New York: Cambridge University Press.

11.

Erikson

Robert S.

Wright

Gerald C.

McIver

John P.

1993. Statehouse Democracy: Public Opinion and Policy in the American States. New York: Cambridge University Press.

12.

Flavin

Patrick

. 2013. “Policy Representation and Evaluations of State Government.” State and Local Government Review 45:139–52.

13.

Flavin

Patrick

Shufeldt

Gregory

. 2016. “Party Competition and Citizens’ Political Attitudes in the American States.” Electoral Studies 44:235–44.

14.

Grill

Christopher J.

2007. The Public Side of Representation: A Study of Citizens’ Views about Representatives and the Representative Process. Albany: State University of New York Press.

15.

Hamman

John A.

2006. “Public Opinion in the States: Determinants of Legislative Job Performance.” In Public Opinion in State Politics, edited by Cohen

Jeffrey E.

, 79–101. Palo Alto: Stanford University Press.

16.

Harden

Jeffrey J.

Clark

Christopher J.

2016. “A Legislature or a Legislator Like Me? Citizen Demand for Collective and Dyadic Political Representation.” American Politics Research 44:247–71.

17.

Hibbing

John R.

Theiss-Morse

Elizabeth

. 1995. Congress as Public Enemy: Public Attitudes toward American Political Institutions. Cambridge: Cambridge University Press.

18.

Hibbing

John R.

Theiss-Morse

Elizabeth

. 2002. Stealth Democracy: Americans’ Beliefs about How Government Should Work. New York: Cambridge University Press.

19.

Hinchliffe

Kelsey L.

Lee

Frances E.

2016. “Party Competition and Conflict in State Legislatures.” State Politics & Policy Quarterly 16 (2): 172–97.

20.

Hogan

Robert E.

2008. “Policy Responsiveness and Incumbent Reelection in State Legislatures.” American Journal of Political Science 52:858–73.

21.

Jewell

Malcolm E.

1982. Representation in State Legislatures. Lexington: University of Kentucky Press.

22.

Johnson

Martin

. 2013. “Media Politics in the States.” In Guide to State Politics and Policy, edited by Niemi

Richard G.

Dyck

Joshua J.

, 151–60. Washington, DC: CQ Press.

23.

Kelleher

Christine

Wolak

Jennifer

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

24.

Lax

Jeffrey R.

Phillips

Justin H.

2012. “The Democratic Deficit in the States.” American Journal of Political Science 56:148–66.

25.

Lyons

Jeffrey

Jaeger

William P.

Wolak

Jennifer

. 2013. “The Roots of Citizens’ Knowledge of State Politics.” State Politics & Policy Quarterly 13:183–202.

26.

Newport

Frank

. 2015. “Mississippi, Alabama, and Louisiana Most Conservative States.” Gallup News, February 6. news.gallup.com/poll/181505/mississippi-alabama-louisiana-conservative-states.aspx.

27.

Pacheco

Julianna

. 2013. “The Thermostatic Model of Responsiveness in the American States.” State Politics & Policy Quarterly 13:306–32.

28.

Patterson

Samuel C.

Hedlund

Ronald D.

Boynton

G. Robert

. 1975. Representatives and the Represented: Bases of American Support for the American Legislatures. New York: Wiley.

29.

Patterson

Samuel C.

Ripley

Randall B.

Quinlan

Stephen

. 1992. “Citizens’ Orientations toward Legislatures: Congress and State Legislatures.” Western Political Quarterly 45:315–38.

30.

Ramirez

Mark D.

2009. “The Dynamics of Partisan Conflict on Congressional Approval.” American Journal of Political Science 53:681–94.

31.

Ramirez

Mark D.

2013. “The Policy Origins of Congressional Approval.” The Journal of Politics 75:198–209.

32.

Richardson

Lilliard E.

Konisky

David M.

Milyo

Jeffrey

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

33.

Rogers

Steven

. 2016. “National Forces in State Legislative Elections.” The Annals of the American Academy of Political and Social Science 667:207–25.

34.

Rogers

Steven

. 2017. “Electoral Accountability for State Legislative Roll Calls and Ideological Representation.” American Political Science Review 111:555–71.

35.

Rosenthal

Alan

. 2008. Engines of Democracy: Politics and Policymaking in State Legislatures. Washington, DC: CQ Press.

36.

Shor

Boris

McCarty

Nolan

. 2011. “The Ideological Mapping of American Legislatures.” American Political Science Review 105:530–51.

37.

Shufeldt

Gregory

Flavin

Patrick

. 2012. “Two Distinct Concepts: Party Competition in Government and Electoral Competition in the American States.” State Politics & Policy Quarterly 12:330–42.

38.

Songer

Donald R.

1984. “Government Closest to the People: Constituent Knowledge in State and National Politics.” Polity 17:387–95.

39.

Squire

Peverill

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

40.

Squire

Peverill

Moncrief

Gary

. 2015. State Legislatures Today: Politics under the Domes. 2nd ed. Lanham: Rowman & Littlefield.

41.

Steenbergen

Marco R.

Jones

Bradford S.

2002. “Modeling Multilevel Data Structures.” American Journal of Political Science 46:218–37.

42.

Stimson

James A.

MacKuen

Michael B.

Erikson

Robert S.

1995. “Dynamic Representation.” American Political Science Review 89 (3): 543–65.

43.

Wahlke

John C.

Eulau

Heinz

Buchanan

William

Ferguson

LeRoy C.

1962. The Legislative System. New York: Wiley.

44.

Weissberg

Robert

. 1978. “Collective vs. Dyadic Representation in Congress.” American Political Science Review 72:535–47.

45.

Wilson

Reid

. 2014. “The Precipitous Decline of State Political Coverage.” The Washington Post, July 14. www.washingtonpost.com/blogs/govbeat/wp/2014/07/10/the-precipitous-decline-of-state-political-coverage/.

46.

Wlezien

Christopher

. 1995. “The Public as Thermostat: Dynamics of Preferences for Spending.” American Journal of Political Science 39:981–1000.

47.

Wolak

Jennifer

. 2018. “Feelings of Political Efficacy in the Fifty States.” Political Behavior 40:763–84.

48.

Wolak

Jennifer

McAtee

Andrea

. 2013. “Parties in the Public Eye.” State Politics & Policy Quarterly 13:398–421.

49.

Wright

Gerald C.

Winburn

Jonathan

. 2002. “Patterns of Constituency-Legislator Policy Congruence in the States.” Presented at the State Politics and Policy Conference, Milwaukee, May 24-25, 2002.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.86 MB