Legislative Party Polarization and Trust in State Legislatures

Abstract

We argue that citizens’ trust attitudes are inversely related to party polarization because polarization tends to encourage political conflict, which most people dislike. We further posit that partisans trust attitudes are driven by the ideological extremity of the opposing and their own parties for similar reasons. Using roll-call-based estimates of state legislative party polarization and public opinion data collected in 2008, we show strong evidence in favor of our theory: higher levels of party polarization within legislative chambers depresses citizens’ trust in their legislatures. Among partisans, we also find that trust attitudes respond to the ideological extremity of the opposing party but not to a citizen’s own party’s extremity. We further find that as citizens’ interest in politics increases, they react more strongly to polarization when forming their trust attitudes. Finally, partisans become less responsive to the ideological extremity of the opposing party as they become more politically interested.

Keywords

trust polarization state legislatures

Over the past several decades, two important changes have taken place in American politics: (1) Political elites within the two major parties have become increasingly ideologically polarized at both the state (Shor & McCarty, 2011) and national levels (Theriault, 2008), and (2) the degree to which citizens feel trust toward government and government institutions declined dramatically during the 1960s (Citrin, 1974; Miller, 1974) and remains low today. Political trust is an important concept because it provides elected officials with greater “leeway to govern effectively” (Hetherington, 1998, p. 803). This suggests that low levels of trust may be problematic for elected officials. The extant research is mostly silent on the relationship between party polarization and trust in government (but see Hetherington & Rudolph, 2015), and much of the research focuses solely on trust attitudes expressed toward the national government. Although U.S. citizens are represented by multiple governments, they tend to pay less attention to politics at the subnational level than they do to national-level politics. Whereas trust in national government institutions has fallen, there is considerably more heterogeneity in attitudes toward state government institutions (Kelleher & Wolak, 2007; Richardson, Konisky, & Milyo, 2012; Wolak & Palus, 2010). As such, we turn our attention toward this state-level variation in trust and address the following question: To what extent does legislative party polarization affect citizens’ trust attitudes toward state legislatures?

More polarized parties express very different policy positions and thus provide citizens with clearer choices on Election Day relative to less polarized parties. In this sense, they may produce some benefits for citizens. However, polarized parties also produce a potentially significant cost: Politics tends to become more contentious as the ideological gap between parties becomes wider, and citizens tend to dislike partisan conflict, even when it produces policies they prefer (Atkinson, 2017). We argue that this conflict—or at least citizens’ perceptions of conflict—should lead citizens to become less trusting of government institutions because they view those institutions as being comprised of politicians who are more interested in demonizing their opponents than they are with producing beneficial public policies. We further argue that, among partisans, the ideological extremity of the two major parties may individually influence trust attitudes. While increasing distance between the two parties may lead partisans to perceive of the opposing party as more threatening, the source of that distance can be the opposing party, their own party, or both. These perceptions may discourage citizens from ceding authority to government institutions. Finally, we argue that more politically interested citizens should be more likely to be aware of polarization and party extremity than less interested people.

We test our theory using public opinion data drawn from a module embedded in the 2008 Cooperative Congressional Election Study (see Richardson et al., 2012) and Shor and McCarty’s (2011) measures of state legislative party ideology. We find strong evidence that party polarization drives trust attitudes. As the parties in a state legislative chamber become more polarized, citizens are less likely to express trust in their state legislatures. We observe this relationship for both chambers and when we use an average measure of the level of polarization in both state legislative chambers. We also find that among partisans, these trust attitudes are driven in large part by the ideological extremity of the opposing party and are not affected by the extremity of the in-party. In addition, we find that the inverse relationship between state legislative polarization and trust in state legislatures becomes stronger as citizens’ levels of interest in politics increases. The inverse relationship between opposition party ideological extremity and trust weakens, however, as people become more politically interested.

These findings strongly suggest that elite party polarization can powerfully inform the way in which people assess government institutions. Given that we test our theory using the characteristics of state legislatures, that is, those of subnational institutions, they also provide evidence that at least some citizens are paying attention to the happenings of subnational politics. Our results are thus novel in the sense that we test our theory across a range of low information environments that vary in terms of their institutional, social, and geographic contexts.

Process, Performance, and Political Trust

Early research on political trust focused on citizens’ evaluative orientations toward the political system as a whole, rather than individual institutions within that system (Easton, 1965; Stokes, 1962). Miller (1974) argued that these evaluations are driven in large part by the normative expectations that citizens hold about how the government has and should operate, thus making these evaluations process-oriented. When the processes governing political outcomes become contorted or manipulated, trust in government ought to decline. This process-oriented perspective is supported by more recent work on the relationship between scandals and trust (Bowler & Karp, 2004; Chanley, Rudolph, & Rahn, 2000; Orren, 1997), and citizens’ level of involvement in the democratic process (Towfigh et al., 2016). Recent work on ideological polarization in legislatures has painted a picture of real distortion in governing processes (Lee, 2015). Polarized chambers are more likely to use uncivil discourse in their hearings and debates (Dodd & Schraufnagel, 2012, 2013), more likely to experience gridlock on a variety of issues (Kousser, 2010), and more likely to exhibit brinksmanship in negotiations between parties (Barber & McCarty, 2015). These failures of the legislative process have the potential to do considerable harm to trust in political institutions.

In contrast to this work, a number of studies have suggested that trust has a strong performance component (Craig, 1993; Erber & Lau, 1990). For example, when the state of the national economy is strong, citizens tend to express more trust in the national government (Chanley et al., 2000; Citrin & Green, 1986; Hetherington, 1998). Higher crime rates also lower trust, and when a sitting president experiences a boost in approval ratings, citizens express more trust (Hetherington, 1998; Mansbridge, 1997). Finally, a number of individual-level characteristics influence trust in governing institutions: (1) Perceptions of greater congruence between preferences and outcomes leads to higher trust (Flavin, 2013), (2) Higher levels of trust in the federal government encourage greater trust in institutions (Hetherington & Nugent, 2001), and (3) Higher levels of education appear to produce expressions of higher levels of trust (Cole, 1973). Beyond its effects on governing processes, polarization also seems to have a number of performance-based consequences. Recent work has suggested that polarization may be at the heart of the rise in income inequality in the United States (McCarty, Poole, & Rosenthal, 2006). Hetherington (2001) also suggests that polarization results in different partisan perceptions of the same performance outcome, making it challenging for a chamber controlled by a party to ever be trusted by the opposition. Polarization also exaggerates incongruence between elected members and their constituents, as the median member of any constituency is the minimal distance between members of a constituency and their representative (Downs, 1957; Hinich & Munger, 1997). Given that party polarization can influence both citizens’ perceptions of governing processes by encouraging gridlock, and alter citizens’ perceptions of their own congruence with government and the performance of that government, there is good reason to believe that party polarization is likely tied to trust attitudes toward legislatures.

While elite polarization certainly influences process and performance, both of which are key elements in structuring trust attitudes toward institutions, the origins of polarization also likely play a role in hampering or encouraging trust. Some research suggests that elite polarization can affect public opinion. More polarized parties should communicate increasingly divergent messages, thus sending strong signals that the parties prefer starkly different political outcomes (Hetherington, 2001). These messages should suggest to citizens which issue positions are congruent with which party, therefore offering them exemplars of what Democrats and Republicans should look like (Carmines & Stimson, 1989; Zaller, 1992). There is also evidence that elite polarization encourages citizens to rely more heavily on party cues and decreases the importance of substantive information during the opinion-formation process (Druckman, Peterson, & Slothuus, 2013). Partisan conflict, while not precisely the same concept as party polarization, is more prevalent when parties diverge from one another to a greater extent and also appears to affect public opinion. For example, Atkinson (2017) finds that partisan conflict drives down support for policy initiatives, and Ramirez (2009) observes that Congressional approval declines as conflict increases, while Hinchliffe and Lee (2015) suggest that party electoral competition in the United States is strongly related to legislative polarization in subsequent sessions. These findings are directly relevant for our own theoretical argument below.

Legislative Polarization and Trust Attitudes

Citizens’ political views are informed in part by their perceptions of elite-level party polarization (Jones, 2015; Theiss-Morse, Barton, & Wagner, 2015; Uslaner, 2015). Specifically, we focus on citizens’ evaluations of government institutions and argue that legislative polarization affects the degree to which citizens trust legislatures. More specifically, we posit that the likelihood that a citizen expresses trust in a legislative body should decline as that legislature becomes increasingly polarized. Three reasons underlie our line of thought.

First, as noted above, elite polarization encourages party conflict. Extant research suggests that citizens harbor a deep dislike of partisan conflict. Parties that are constantly at odds with one another may be perceived of as being more interested in maintaining ideological purity at the cost of failing to work on behalf of citizens when crafting legislation (Carson, Koger, Lebo, & Young, 2010; Ramirez, 2009). Thus, citizens may be increasingly reluctant to grant their legislatures trust and the leeway that is associated with it as legislative polarization increases.

Second, the news media tends to focus on political conflict and legislative maneuvering rather than on political compromise, thus largely failing to produce substantive coverage of public policy issues (Johnson & O’Grady, 2013; Morris & Clawson, 2005). These kinds of “strategic” frames increase cynicism among the public (Jackson, 2010). Some research also suggests that coverage of the Speaker of the House’s activities increases during times of high conflict and polarization (Johnson & O’Grady, 2013). Thus, the mainstream news media as a whole is likely to spend more time covering both the conflicts that are generated by legislative polarization and polarization itself as legislative polarization increases. Pundits are especially likely to spend time discussing polarization and blame it for the political ills of the day. Thus, citizens should respond to these negative messages about polarized legislatures by expressing lower levels of political trust.

Third, as parties become increasingly polarized, citizens should view extreme opposition parties as greater social-identity threats (Branscombe, Ellemers, Spears, & Doosje, 1999; Huddy, Mason, & Aarøe, 2015). More specifically, partisans should view the opposing party as more threatening for two reasons: (1) Their own elites are likely to put forth a greater effort to demonize the opposition as polarization increases, and (2) They are more likely to notice differences between the parties when polarization is greater. This phenomena is likely magnified by the fact that the degree to which partisans dislike those who identify with the opposing party has grown over time (Iyengar, Sood, & Lelkes, 2012). Independents, at least those who are not ideological extremists, should also find polarized parties more threatening because they both appear to hold dangerously extreme policy preferences (King, 1997). Thus, trust should decline as legislative polarization increases because citizens should view one or both parties as being increasingly threatening.

These three assumptions (that citizens dislike conflict, that the news media is more likely to report on conflict, and that opposing party extremity poses a social identity threat to citizens) lead us to draw what we call the “polarization hypothesis”: citizens are less likely to express trust in legislatures as legislative polarization increases. In addition, the third argument leads us to draw a second hypothesis, which we call the “extremism hypothesis”: partisans are less likely to express trust in legislatures as their own and the opposing party becomes increasingly ideologically extreme. It is important to note that while these two hypotheses are clearly related, there are important differences between them. Research on legislative polarization has consistently treated polarization as a relational concept measuring the distance between two political objects (Garand, 2010; McCarty et al., 2006; Shor & McCarty, 2011; Theriault, 2006). Polarization may increase in a legislature when a citizen’s own party becomes more extreme, when the opposing party becomes more extreme, or both. The social identity-based threat that people may feel likely exists regardless of which party is more ideologically extreme.

While our theory suggests that citizens’ trust attitudes should be informed by the ideological extremities of both parties, it is also possible that these attitudes are responsive only to the extremity of their own or the opposing parties.¹ Learning which of these possibilities is correct will add to our understanding of the origins of trust attitudes, so we seek to observe which one is best supported by the data.

That said, our expectations may only hold among citizens with higher levels of interest in politics. There is a large body of research suggesting that most citizens tend to be inattentive to politics (e.g., Campbell, Converse, Miller, & Stokes, 1960; Delli Carpini & Keeter, 1996). Those who are either not exposed to information or who do not process information about legislative polarization and party ideological extremity cannot use that information to inform their trust attitudes. Thus, in addition to the direct hypotheses we describe above, we expect to observe conditional relationships between citizens’ trust attitudes and their legislatures’ degrees of polarization and party extremities. In terms of the polarization hypothesis, we expect that more interested citizens will be less likely to express trust in their legislatures than will less interested citizens as legislative polarization increases. For the extremity hypothesis, we expect that more interested partisans will be less likely to express trust in their legislatures than will less interested citizens as their own and the opposing party becomes increasingly ideologically extreme.

Research Design

We test our theory at the subnational level because states offer a great deal more potential variance in legislative polarization than exists within single time points for trust attitudes in the U.S. Congress. In addition, state legislatures vary on a number of institutional dimensions, so this environment allows for a richer test of our theory. We use data drawn from the University of Missouri’s team content module embedded in the 2008 Cooperative Congressional Election Study (CCES; see Richardson et al., 2012) for individual-level public opinion data.

The CCES is a large-scale survey of a representative population of adult citizens in the United States. The University of Missouri team content module includes responses from 1,000 people. In addition to being asked a number of standard demographic and political questions, respondents were also asked to express the degree to which they trust their state legislatures. The question was worded as follows: “How much of the time do you think you can trust each branch of government? Please answer for each branch of the Federal or your state government listed below.” For this research, the key response was for a respondent’s state legislature. The possible responses were “hardly ever” (26%), “some of the time” (55%), “most of the time” (18%), and “just about always” (1%).

We recoded this variable such that it equaled 0 if the respondent reported trusting their state legislature “hardly ever” and 1 if they gave any other response. Thus, we have a measure of whether or not a respondent reported trusting the state legislature at all. We did this because there is a clear increase in the level of legislative polarization among respondents who expressed any of the three higher levels of trust compared with those who chose the lowest level, but there are no clear differences among those three higher levels of trust. This binary indicator of trust serves as the dependent variable in the analyses testing the polarization and extremity hypotheses. Because the dependent variable is dichotomous, we use logistic regression to evaluate our hypotheses.²

To assess the possibility that citizens’ reported trust attitudes in state legislatures are essentially identical to their levels of trust in the federal government—a more salient political object relative to state legislatures—we include a placebo test later in the article. In this set of tests, the dependent variable is trust in the federal government. Respondents were asked the same question as above and could give the same set of responses. They could have reported that they trusted the federal government “hardly ever” (49%), “some of the time” (42%), “most of the time” (8%), or “just about always” (less than 1%). As was the case with our indicator of trust in state legislatures, we collapse this measure to a binary indicator for ease of comparison. This choice does not affect the substance of our findings.³

Our primary independent variable for testing the polarization hypothesis is the degree of legislative polarization in a state during 2008. These data are drawn from the American Legislatures data (Shor & McCarty, 2011) on state legislative ideology. We used these data to create chamber-level measures of party polarization. First, we calculate the median across the ideal points estimated for members of the Democratic and Republican Parties in each chamber. Then, we subtracted these medians from one another to calculate a measure of chamber polarization for which higher values indicate greater polarization. Thus, we have measures of legislative polarization in both the lower and upper chamber of each state.⁴ We use each as our primary independent variable of interest in separate models, but we do not include both in the same model due to collinearity, as the two indicators are highly correlated (.87) with one another. We do, however, include a third measure of legislative polarization that is the mean of the polarization in the lower and upper chambers. We expect to observe negative coefficients for these variables in our models.

We use a different, but related, concepts as our primary independent variables of interest in our tests of the extremity hypothesis: the ideological extremity of a given partisan’s own and opposing - that is the party with which respondent does not identify - parties.⁵ We use the absolute value of the median party ideal point within each chamber. This gives us a direct measure of the ideological extremity within a party by chamber. Higher values indicate greater ideological extremity.

To test the conditional portion of our theory, we must account for respondents’ levels of political interest. We do so with an indicator ranging from 0 to 2 for which higher values correspond with higher levels of expressed interest in politics. Respondents could say that they were 0 = “not much interested,” 1 = “somewhat interested,” or 2 = “very much interested” in politics and current events. This variable is used as a control when we report the direct tests of the polarization and extremity hypotheses. When we do the conditional test of our hypotheses, we interact the interest variable with our measures of state legislative party polarization and extremity. Thus, in the models including the interactions, our primary quantities of interest are the marginal effects of polarization or extremity on trust attitudes at each level of expressed interest in politics.

In addition to the covariates meant to test our primary hypotheses, we control for potential differences in trust driven by partisanship by including two dummy variables that are coded 1 if respondent identifies with either the Democratic or the Republican Parties and 0 if they do not.⁶ We also include dummy variables indicating whether or not respondents are women, African American, or Latino and include a variable capturing the age in years of respondents, which allows our statistical models to account for differences in levels of trust across age groups. We also account for differences in our respondents’ levels of education by including a measure ranging from 1 to 6 in which higher values correspond with higher levels of education attainment.

Finally, we included four additional control variables tapping into the political and economic contexts in which respondents found themselves in their states: (1) the state unemployment rate,⁷ (2) respondents’ perceptions of the strength of the national economy,⁸ (3) the number of legislative chambers controlled by the respondent’s party,⁹ and (4) the presence of divided government.¹⁰ The first controls for the likely dissatisfaction that citizens will hold toward government institutions when unemployment rates are higher, the second controls for the finding that more positive economic perceptions tends to lead citizens to express higher levels of trust, and the third controls for the possibility that citizens express lower levels of trust when there is more salient political conflict between legislatures and governors. We report the summary statistics for each of the variables used in our analyses in Table 1.

Table 1.

Summary Statistics.

	M	SD	Minimum	Maximum
Trusts the state legislature	0.74	0.44	0.00	1.00
Trusts the national government	0.51	0.50	0.00	1.00
Legislative polarization
Lower chamber	1.59	0.67	0.52	3.21
Upper chamber	1.58	0.54	0.48	2.79
Average	1.58	0.59	0.55	3.00
Opposition party extremity
Lower chamber	0.80	0.42	0.02	1.92
Upper chamber	0.79	0.40	0.01	1.79
Average	0.79	0.39	0.03	1.72
In-party extremity
Lower chamber	0.84	0.43	0.02	1.92
Upper chamber	0.82	0.39	0.01	1.79
Average	0.83	0.40	0.03	1.72
Level of political interest	1.60	0.61	0.00	2.00
No. of legislative chambers controlled by respondent’s party	0.94	0.90	0.00	2.00
Divided government	0.45	0.50	0.00	1.00
Respondent is a Democrat	0.46	0.50	0.00	1.00
Respondent is a Republican	0.43	0.50	0.00	1.00
Respondent is female	0.49	0.50	0.00	1.00
Respondent is African American	0.06	0.23	0.00	1.00
Respondent is Latino/a	0.06	0.23	0.00	1.00
Age in years	48.45	15.80	18.00	93.00
Level of education	3.45	1.50	1.00	6.00
State unemployment rate	5.78	1.06	3.00	8.30
Perceptions of the national economy	1.50	0.69	1.00	5.00

Results

Testing the Polarization Hypothesis

We evaluate the polarization hypothesis first using the results of the models reported in Table 2. Each model uses a different measure of legislative polarization as the primary independent variable of interest. Polarization in the lower chamber is shown in the first column, upper chamber polarization is used in the second column, and the average polarization of both chambers is utilized in the last column. Recall that we expect the coefficients for legislative polarization to have negative signs.

Table 2.

State Legislative Polarization and Trust in State Legislatures.

	Lower chambers	Upper chambers	Average
Legislative polarization	−0.332*	−0.423*	−0.406*
	(0.143)	(0.208)	(0.172)
Interest in politics	−0.069	−0.079	−0.074
	(0.206)	(0.204)	(0.204)
Number of legislative chambers controlled by respondent’s party	0.400*	0.401*	0.401*
	(0.143)	(0.143)	(0.142)
Divided government	0.023	0.025	0.028
	(0.254)	(0.240)	(0.247)
Respondent is a Democrat	0.663	0.677	0.673
	(0.464)	(0.464)	(0.464)
Respondent is a Republican	0.257	0.281	0.272
	(0.352)	(0.353)	(0.352)
Respondent is female	0.129	0.130	0.130
	(0.197)	(0.196)	(0.196)
Respondent is African American	−0.155	−0.172	−0.167
	(0.489)	(0.483)	(0.487)
Respondent is Latino/a	−0.145	−0.169	−0.152
	(0.414)	(0.411)	(0.413)
Age in years	−0.007	−0.007	−0.007
	(0.006)	(0.006)	(0.006)
Level of education	0.039	0.039	0.039
	(0.084)	(0.084)	(0.084)
State unemployment rate	−0.073	−0.075	−0.069
	(0.118)	(0.116)	(0.117)
Perceptions of the national economy	0.337†	0.336†	0.337†
	(0.189)	(0.190)	(0.190)
Intercept	1.166	1.314	1.251
	(0.832)	(0.810)	(0.822)
Random intercept
$σ_{S t a t e}^{2}$	0.181	0.142	0.161
BIC	964.329	963.869	963.852
N	835	835	835

Note. Cell entries are estimated coefficients generated using mixed-effects logistic regression along with probability weights. Standard errors are reported in parentheses. BIC = Bayesian information criterion.

p ≤ .05 (two tailed), †p ≤ .05 (one tailed).

The results of all three models provide strong support for our theory. Note that all three of the coefficients for legislative polarization are negative as expected. They also all differ significantly from zero (p ≤ .05), thus allowing us to confidently reject the null hypothesis that legislative polarization and trust in state legislatures are statistically unrelated. But to what degree does legislative polarization shape trust attitudes?

To address this question, we turn to a series of predicted probabilities which we plot in Figure 1.¹¹ We generated these predicted probabilities by simulating across the observed values of our measures of lower chamber polarization, upper chamber polarization, and the average level of polarization across chambers. We held the control variables at their observed values for each observation in our data.

The first plot in Figure 1 shows the predicted probability of expressing trust in state legislatures across all values of legislative polarization in the lower chamber. When polarization is at its lowest, the predicted probability of expressing trust in the state legislature is .81. As polarization increases, the predicted probability of expressing trust declines. At the high point of legislative polarization, this predicted probability is .66, a total decrease of .15.

The predicted probability of expressing trust similarly declines as polarization increases in the upper chamber of a state’s legislature. The predicted probability here ranges from .86 when upper chamber polarization is at its nadir to .70 when polarization is at its peak, a decline of .16.

We next turn to the final plot, which presents the predicted probability of expressing trust in the state legislature across our averaged measure of chamber polarization. The pattern here is similar to the results of the previous analyses. The predicted probability of expressing trust in the state legislature is .85 when polarization is at its minimum. This probability declines as polarization increases. At the maximum of polarization, the predicted probability is .69, a total decline of .20.

To what extent are the effects of state legislative polarization on trust attitudes similar to those of some of the other covariates in our models? We compare the effects of legislative polarization with those of the number of chambers controlled by a respondent’s party and perceptions of the national economy indicators using the “divide by four rule” (see Gelman & Hill, 2006, p. 82), which allows researchers to divide coefficients estimated using logistic regression by four to calculate the maximum change in the predicted probability of a success given a one-unit increase in a given independent variable. We report these values in Table 3, and they suggest that one-unit increases across these variables lead to substantively similar, in terms of magnitude, effects on the predicted probability of trusting the state legislature. Do note, however, that the range of the variables is not identical; the polarization variables range from about 0.5 to around 3, the chamber control indicator ranges from 0 to 2, and the economic perceptions variable ranges from 1 to 5. Because of these observed ranges, it appears as if respondents’ economic perceptions have the largest possible effects on trust attitudes in terms of magnitude compared with state legislative polarization and partisan chamber control.

Table 3.

Maximum Changes in Predicted Probability of Trusting the State Legislature When a Covariate Increases by One Unit.

	Lower chambers	Upper chambers	Average
Legislative polarization	−0.08	−0.11	−0.10
Number of legislative chambers controlled by respondent’s party	0.10	0.10	0.10
Perceptions of the national economy	0.08	0.08	0.08

Note. Quantities estimated using the divide by four rule (Gelman & Hill, 2006) using the values reported in Table 2.

Next, we consider the possibility that respondents’ levels of interest in politics condition the effects of state legislative polarization on their trust attitudes. The results of these models are shown in Table 4. Due to the presence of an interaction between our measures of legislative polarization and respondents’ levels of political interest, the raw coefficients do not provide us with sufficient information to interpret the results of our conditional test of the polarization hypothesis. To do this, we must calculate the marginal effects of legislative polarization on trust attitudes at each level of the political interest variable (Brambor, Clark, & Golder, 2006). We do so and plot them in Figure 2. The points represent the marginal effects, while the vertical lines running through them show the 95% confidence intervals. If the confidence intervals overlap with the dashed line at 0 on the Y axis, then we fail to reject the null hypothesis of no relationship at p ≤ .05.

Table 4.

How Political Interest Conditions the Effects of Legislative Polarization on Trust in State Legislatures.

	Lower chamber	Upper chamber	Average
Legislative polarization	−0.032	−0.401	−0.208
	(0.333)	(0.520)	(0.429)
Interest in politics	0.279	−0.055	0.150
	(0.452)	(0.594)	(0.523)
Polarization × Interest	−0.200	−0.014	−0.131
	(0.198)	(0.326)	(0.259)
Number of legislative chambers controlled by respondent’s party	0.394*	0.401*	0.397*
	(0.144)	(0.147)	(0.145)
Divided government	0.026	0.026	0.030
	(0.255)	(0.241)	(0.249)
Respondent is a Democrat	0.622	0.674	0.647
	(0.481)	(0.478)	(0.482)
Respondent is a Republican	0.234	0.280	0.255
	(0.366)	(0.363)	(0.365)
Respondent is female	0.117	0.129	0.123
	(0.199)	(0.198)	(0.199)
Respondent is African American	−0.167	−0.173	−0.175
	(0.492)	(0.484)	(0.490)
Respondent is Latino/a	−0.148	−0.170	−0.157
	(0.417)	(0.414)	(0.416)
Age in years	−0.008	−0.007	−0.007
	(0.006)	(0.006)	(0.006)
Level of education	0.039	0.039	0.039
	(0.085)	(0.083)	(0.084)
State unemployment rate	−0.074	−0.075	−0.070
	(0.118)	(0.116)	(0.117)
Perceptions of the national economy	0.321†	0.335†	0.329†
	(0.192)	(0.190)	(0.192)
Constant	0.731	1.282	0.962
	(0.936)	(1.032)	(0.989)
Random intercept
$σ_{S t a t e}^{2}$	0.186	0.142	0.164
BIC	969.806	970.592	970.161
N	835	835	835

Note. Cell entries are estimated coefficients generated using multinomial logistic regression along with probability weights. The base category is the lowest level of trust in state legislatures. Standard errors clustered on states are reported in parentheses. BIC = Bayesian information criterion.

p ≤ .05 (two tailed), †p ≤ .05 (one tailed).

The marginal effects of legislative polarization on trust attitudes exhibit a consistent pattern across the three models: They are all negative and increase in magnitude—that is, move further away from zero—as a respondent’s level of political interest increases. In the lower chamber and average chamber models, the marginal effect only differs significantly from zero at the highest level of interest. None of the marginal effects from the upper chamber model differ significantly from zero, though we should also note that the marginal effects are almost identical across the values of political interest, so a case can be made that the interaction between upper chamber polarization and political interest is empirically unnecessary in this case.

Next, we turn to the substantive effects of lower and average chamber state legislative polarization on trust in the state legislature as conditioned by political interest, which we present in Figure 3.¹² In the lower chamber model, the predicted probability of expressing trust in the state legislature when polarization is at its lowest value is about .77 and .82 among respondents with the lowest and highest levels of political interest, respectively. When polarization is at its highest value, the predicted probabilities of expressing trust for the same kinds of people are .76 and .63. In the average polarization model, these values range from .80 to .72 for the least interested and .82 to .63 for the most interested as polarization moves from its minimum to its maximum. Thus, it appears as if citizens’ trust in state legislatures declines as the legislative parties become more polarized and that these effects are strongest among those who are most interested in politics.

Placebo Test

While the above results seem to strongly suggest that legislative polarization affects the levels of trust that citizens hold in their state legislatures, some readers may be concerned that citizens are too inattentive to subnational politics for the results to be credible. Given the paucity of attention citizens devote to state government, it is possible that our results are simply an artifact of some relationship between polarization and citizens’ general attitudes toward all governments. Our theoretical expectations are very specific: We expect that the level of polarization in a legislature informs trust in that chamber. Thus, if our theory has merit, increases in state legislative party polarization ought to depress trust attitudes toward state legislatures (which we have already seen) but not toward other government bodies. As such, a placebo test is in order. To perform such a test, we observe how legislative party polarization influences trust in the national government. If our theory is correct, we will observe essentially no relationship between these two variables. We report the results of these tests in Table 5.

Table 5.

State Legislative Polarization and Trust in the Federal Government.

	Lower chambers	Upper chambers	Average
Legislative polarization	0.213	0.259	0.256
	(0.137)	(0.184)	(0.163)
Level of political interest	−0.292*	−0.285*	−0.288*
	(0.143)	(0.143)	(0.144)
Number of legislative chambers controlled by respondent’s party	−0.024	−0.029	−0.027
	(0.096)	(0.097)	(0.097)
Divided government	−0.200	−0.197	−0.200
	(0.186)	(0.184)	(0.185)
Respondent is a Democrat	0.863*	0.843*	0.853*
	(0.331)	(0.335)	(0.333)
Respondent is a Republican	0.623*	0.610*	0.615*
	(0.299)	(0.302)	(0.301)
Respondent is female	0.309†	0.310†	0.311†
	(0.162)	(0.163)	(0.163)
Respondent is African American	0.187	0.190	0.191
	(0.324)	(0.323)	(0.323)
Respondent is Latino/a	−0.339	−0.318	−0.332
	(0.343)	(0.348)	(0.344)
Age in years	−0.010*	−0.011*	−0.010*
	(0.005)	(0.005)	(0.005)
Level of education	0.054	0.052	0.054
	(0.069)	(0.069)	(0.069)
State unemployment rate	−0.185*	−0.188*	−0.189*
	(0.084)	(0.084)	(0.084)
Perceptions of the national economy	0.362*	0.362*	0.363*
	(0.125)	(0.125)	(0.126)
Intercept	0.395	0.387	0.426
	(0.693)	(0.698)	(0.698)
Random intercept
$σ_{S t a t e}^{2}$	0.015	0.021	0.016
BIC	1,221.425	1,221.530	1,221.224
N	865	865	865

p ≤ .05 (two tailed), †p ≤ .05 (one tailed).

In the previous section, the estimated coefficients for the legislative polarization variables were observed to be negative. In this analysis, all three are positive. In addition, none of these coefficients differs significantly from zero at a traditional level (p ≤ .05). Thus, these results provide no evidence that state legislative polarization affects citizens’ expressed levels of trust in the federal government. This suggests the systematic relationships we observe in prior analyses are not a function of general attitudes toward governments, and are instead a function of specific evaluations of specific institutions. Furthermore, note that the correlation between our measures of trust in state legislatures and trust in the federal government is about .32. While the two variables are positively correlated with one another, there is a substantial degree of independent variance in both measures.

We run one more set of placebo tests in Table 6. The specification is the same as in the first set of tests, but this time we interact political interest with our measures of polarization. We plot the marginal effects of interest in Figure 4. All of these marginal effects are positive, and only two—for people with the lowest and middle levels of political interest in the lower chamber polarization model—differ significantly from zero. Thus, we once again find no evidence that increases in state legislative polarization drives down trust in the federal government.

Table 6.

How Political Interest Conditions the Effects of State Legislative Polarization on Trust in the Federal Government.

	Lower chambers	Upper chambers	Average
Legislative polarization	0.606*	0.261	0.511
	(0.246)	(0.474)	(0.339)
Level of political interest	0.146	−0.283	−0.008
	(0.250)	(0.398)	(0.297)
Polarization × Interest	−0.262*	−0.002	−0.170
	(0.112)	(0.219)	(0.140)
Number of legislative chambers controlled by respondent’s party	−0.031	−0.029	−0.031
	(0.098)	(0.100)	(0.099)
Divided government	−0.207	−0.197	−0.202
	(0.187)	(0.184)	(0.186)
Respondent is a Democrat	0.813*	0.842*	0.820*
	(0.342)	(0.336)	(0.341)
Respondent is a Republican	0.591†	0.610*	0.590†
	(0.307)	(0.305)	(0.307)
Respondent is female	0.293†	0.310†	0.301†
	(0.162)	(0.162)	(0.161)
Respondent is African American	0.182	0.190	0.184
	(0.327)	(0.321)	(0.325)
Respondent is Latino/a	−0.351	−0.318	−0.343
	(0.351)	(0.354)	(0.350)
Age in years	−0.011*	−0.011*	−0.011*
	(0.005)	(0.005)	(0.005)
Level of education	0.055	0.052	0.054
	(0.069)	(0.069)	(0.069)
State unemployment rate	−0.193*	−0.188*	−0.194*
	(0.083)	(0.084)	(0.084)
Perceptions of the national economy	0.345*	0.362*	0.353*
	(0.123)	(0.123)	(0.123)
Intercept	−0.078	0.384	0.059
	(0.714)	(0.944)	(0.789)
Random intercept
$σ_{S t a t e}^{2}$	0.027	0.021	0.021
BIC	1,225.425	1,228.293	1,227.078
N	865	865	865

Note. Cell entries are estimated coefficients generated using mixed effects logistic regression along with probability weights. Standard errors are reported in parentheses. BIC = Bayesian information criterion.

p ≤ .05 (two tailed), †p ≤ .05 (one tailed).

Testing the Opposition Extremity Hypothesis

We report the results of our tests of the extremity hypothesis in Table 7.¹³ Once again, the results focusing on the lower chamber are in the first column while those of the upper chamber are reported in the second column. The last column shows the results when we average our extremity measures across the two chambers. Recall that these models omit independents, who do not have clear in and out parties. Thus, the following results hold only for partisans.

Table 7.

Party Extremity and Trust in State Legislatures.

	Lower chambers	Upper chambers	Average
Opposition party extremism	−0.791*	−0.819*	−0.903*
	(0.244)	(0.319)	(0.261)
In-party extremism	−0.218	−0.139	−0.210
	(0.294)	(0.368)	(0.327)
Level of political interest	−0.062	−0.076	−0.074
	(0.239)	(0.241)	(0.238)
Number of legislative chambers controlled by respondent’s party	0.341*	0.335*	0.328*
	(0.135)	(0.131)	(0.135)
Divided government	0.157	0.141	0.154
	(0.341)	(0.339)	(0.336)
Respondent is a Democrat	0.236	0.238	0.232
	(0.257)	(0.253)	(0.255)
Respondent is female	0.271	0.283	0.277
	(0.220)	(0.224)	(0.222)
Respondent is African American	−0.047	−0.060	−0.061
	(0.476)	(0.473)	(0.473)
Respondent is Latino/a	0.100	0.075	0.086
	(0.484)	(0.492)	(0.487)
Age in years	−0.009	−0.009	−0.009
	(0.006)	(0.006)	(0.006)
Level of education	0.063	0.061	0.062
	(0.084)	(0.085)	(0.084)
State unemployment rate	−0.210	−0.248	−0.220
	(0.176)	(0.183)	(0.180)
Perceptions of the national economy	0.186	0.201	0.195
	(0.186)	(0.185)	(0.186)
Intercept	2.782*	2.997*	2.942*
	(1.077)	(1.064)	(1.074)
Random intercept
$σ_{S t a t e}^{2}$	0.363	0.387	0.352
BIC	794.451	793.901	793.557
N	745	745	745

p ≤ .05 (two tailed), †p ≤ .05 (one tailed).

As expected, the estimated coefficients for our opposition party extremism variables are negative across all three models. The standard errors are also small enough relative to the coefficients that we are able to confidently (p ≤ .05) reject the null hypothesis that opposition party extremism is unrelated to these trust attitudes. This implies that the probability of expressing trust in the state legislature declines as a respondent’s opposing party becomes increasingly ideologically extreme. In other words, the probability that a Republican (Democrat) will express trust in their state legislature declines as the Democratic (Republican) Party becomes more liberal (conservative).

While the coefficients provide us with evidence favoring our theory, they do not allow us to directly observe the substantive impact of opposition party extremism on trust in state legislatures. To observe these substantive effects, we present a series of predicted probabilities generated from these models in Figure 5. As expected, all three plots show that the predicted probability of trusting the state legislature declines as opposition party extremity increases.

The first panel shows the effects of lower chamber opposition party extremity on trust attitudes. At the lowest level of opposition extremity, the predicted probability of expressing trust in the state legislature is .86. At the highest level of opposition party extremity, the predicted probability is .62, a total decline of .24. Upper chamber opposition party extremity exhibits a similar effect on trust as shown in the second panel; the predicted probability decreases from a starting point of .87 to .64 as opposition extremism increases from its minimum to its maximum values, a total decrease of .23. Finally, the last panel shows the substantive effects of our average chamber extremism measure on trust in the state legislature. The predicted probability of expressing trust among partisans who live in states with the least ideologically extreme opposition parties is .88. This value declines to .63 among those living in states with the most extreme opposition parties, a total change in the predicted probability of .25.¹⁴

Returning momentarily to the results presented in Table 7, we further note that, as expected, partisans’ trust attitudes do not appear to be responsive to their own party’s ideological extremity within—or averaged across—the chambers. The coefficients generated for the in-party extremism measure are negative across the three models, but none differ significantly (p ≤ .05) from zero. Thus, it appears that whereas opposition party extremism exhibits a substantively meaningful effect of partisans’ expressions of trust in their state legislatures, the ideological leanings of partisans’ own parties do not inform these same trust attitudes. While each party’s level of polarization tends to move with the opposition meaning both parties are typically polarized to similar degrees (Lebo, McGlynn, & Koger, 2007), it is important to note that our results suggest that citizens’ trust in state legislatures are driven more by fear of the opposing party than by discontent with the extremity of their own parties representatives. Thus, in the event of asymmetric elite polarization, we might expect asymmetric changes in the level of trust in state legislatures.

Our final analyses are presented in Table 8 and show the results of the conditional tests of the extremism hypothesis. Due to the presence of interactions between the opposition and in-party indicators and the interest in politics variable, we once again calculate and plot quantities of interest: the marginal effects of a one-unit increase in party extremism on trust attitudes across the values of interest in politics. We present these quantities in Figure 6.

Table 8.

How Political Interest Conditions the Effects of Party Extremity on Trust in State Legislatures.

	Lower chamber	Upper chamber	Average
Opposition party extremism	−1.409†	−1.715*	−1.769*
	(0.848)	(0.777)	(0.754)
Opposition party extremism × Interest	0.369	0.542	0.513
	(0.453)	(0.423)	(0.412)
In-party extremism	0.803	1.047	0.970
	(0.533)	(0.798)	(0.620)
In-party extremism × Interest	−0.632*	−0.689†	−0.705*
	(0.261)	(0.390)	(0.291)
Divided government	0.127	0.088	0.110
	(0.357)	(0.355)	(0.354)
Number of legislative chambers controlled by respondent’s party	0.333*	0.323*	0.316*
	(0.134)	(0.136)	(0.137)
Respondent is a Democrat	0.222	0.219	0.214
	(0.260)	(0.256)	(0.258)
Respondent is female	0.305	0.323	0.319
	(0.225)	(0.236)	(0.230)
Respondent is African American	−0.016	−0.031	−0.030
	(0.471)	(0.474)	(0.471)
Respondent is Latino/a	0.067	0.015	0.035
	(0.486)	(0.497)	(0.493)
Age in years	−0.010†	−0.009	−0.009
	(0.006)	(0.006)	(0.006)
Level of education	0.079	0.074	0.078
	(0.086)	(0.088)	(0.087)
Level of political interest	0.200	0.028	0.096
	(0.499)	(0.572)	(0.535)
State unemployment rate	−0.232	−0.276	−0.245
	(0.184)	(0.191)	(0.189)
Perceptions of the national economy	0.178	0.201	0.191
	(0.187)	(0.188)	(0.189)
Intercept	2.513†	2.975*	2.827†
	(1.463)	(1.459)	(1.486)
Random Intercept
$σ_{S t a t e}^{2}$	0.446	0.489	0.449
BIC	804.418	803.298	803.064
N	745	745	745

p ≤ .05 (two tailed). †p ≤ .05 (one tailed).

The marginal effects for opposition parties are shown in the left column and those for the in-party are shown in the left column. We first note that none of the marginal effects for the in-party’s ideological extremism differ significantly (p ≤ .05) from zero, a finding that mirrors what we observed in the unconditional test of the extremism hypothesis. Next, all of the marginal effects for the opposition party are negative as expected, indicating that increases in the opposition party’s level of ideological extremism on average correspond with decreases in the probability that a citizen expresses trust in their state legislature. Recall, however, that our expectation is that citizens with higher levels of political interest would respond more strongly to increasingly extreme opposition parties. That does not appear to be the case; the marginal effects of opposition party extremism decrease in magnitude as political interest increases. Still, all of the marginal effects are negative and all but one—that of those with the highest level of interest in politics in the upper chamber model—differ significantly from zero.

Last, we turn to the substantive effects of these quantities in Figure 7, in which we plot the predicted probability of expressing trust in state legislatures across the level of opposition party extremity for each value of individual-level political interest. In the lower chamber, respondents with the lowest and highest levels of political interest have a predicted probability of expressing trust in the state legislature of .91 and .85 when extremism is at its minimum. These predicted probabilities shift to .52 and .65 at extremism’s maximum. We observe similar patterns in our upper chamber—.93 to .49 for the least and .85 to .67 for the most interested—and average chamber—.93 to .50 among those with the lowest and .86 to .66 among those with the highest levels of political interest—models. In sum, while these results show that the way that partisans’ trust attitudes respond—or do not respond—to the level of ideological extremity of their state legislative parties is conditioned by how interested they are in politics, the patterns on the whole are substantively congruent with direct tests of the extremism hypothesis.

Conclusion

In this research, we presented a theory about how legislative polarization affects citizens’ trust in their state legislatures. Our analyses produced evidence that strongly favors our argument; as polarization increases, the likelihood that citizens express trusting attitudes toward their legislatures appears to decline. In addition, we also found evidence supporting our argument that these trust attitudes among partisans are strongly informed by the ideological extremity of the opposing party. We further found more nuanced evidence suggesting that the trust attitudes of people with higher levels of political interest were more responsive to state legislative party polarization and less responsive to opposition party extremity relative to people with lower levels of political interest. We failed to find evidence, however, that the trust attitudes of partisans are influenced by the extremity of their own parties. It is possible that the latter nonfinding is driven by biased information processing by partisans or some other process. Future research might examine this and other possibilities.

Citizens appear to be more willing to cede greater authority to their state legislatures when those bodies are less polarized relative to when they are more polarized. This suggests that less polarized legislatures may be more empowered to act on behalf of their residents than their more polarized cousins. Polarization, then, drives trust attitudes down and may make it more difficult for legislatures to produce policy changes. It may also constrain the ability of legislators to act on behalf of their constituents in situations in which policy changes may benefit citizens but in ways that may not be obvious to those constituents who are lacking in political sophistication.

In addition, these results suggest that citizens are threatened by increasing ideological extremity among their out-parties. This implication still speaks to our argument about polarization, but it appears as if ideological extremity only directly matters to partisans in terms of the opposing party, not the extremity of the in-party. Increasingly distinct parties may encourage the formation of cognitive images of the opposing party as increasingly dangerous. It may be sensible from the perspective to partisans, then, to avoid ceding authority to government and government institutions when they fear the consequences of the opposing party taking control. They may be comfortable with granting greater leeway to the government when it is controlled by their party’s representatives, but they may perceive of unacceptable levels of danger in situations in which the government is dominated by the opposing party.

Our research is important for two reasons. First, it provides evidence of which political reformers should take heed. If they wish to produce institutions that people trust, one way to achieve that goal is to change the electoral rules in a way that encourages the election of less ideologically extreme legislators. Drawing state legislative districts that are more electorally competitive and that contain a partisan and ideological mix of citizens may help to reduce legislative polarization in the aggregate.

Second, Tip O’Neill and others may not have been entirely correct when they claimed that all politics is local, but our results imply that citizens are attentive enough to subnational politics that their attitudes about state politics are shaped at least in part by state-level conditions. Thus, citizens may be more attuned to subnational politics than some prior public opinion research suggests.

Our research is limited in that we only have individual-level public opinion data from a single year. Unfortunately, there is little available data on citizens’ levels of trust in subnational institutions. But given the variance in levels of legislative polarization across the states in these data, not to mention the different constellations of rules and institutions within each of the chambers, we are confident that these results are meaningful. Future work might expand the analysis to additional years, examine how trust in the U.S. Congress is shaped in response to levels of Congressional polarization both cross-sectionally and over time, explore why more interested people exhibited weaker responses to opposition party extremity, and observe how trust attitudes toward state courts vary in response to polarization and more generally the ideological make-up of state courts of last resort.

Footnotes

Acknowledgements

The authors would like to thank Tom Carsey, Shanna Pearson-Merkowitz, Emily Schilling, James Endersby, Bill Eubank, the anonymous reviewers, and the editor for their helpful feedback. We would also like to thank Boris Shor and Nolan McCarty for sharing their state legislative ideal point data and Lilliard Richardson Jr., David Konisky, and Jeffrey Milyo for allowing us access to their 2008 CCES survey 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.

Supplementary Material

Supplementary material is available for this article online.

Notes

Author Biographies

Kevin K. Banda is an assistant professor in the Department of Political Science at Texas Tech University. His research focuses on polarization, campaigns, elections, public opinion, political communication, and state politics.

Justin H. Kirkland is an associate professor in the Department of Political Science at the University of Houston. His research focuses on representation, legislative behavior, state politics, and methods of social network analysis.

References

Atkinson

M. L.

(2017). Combative politics: The media and public perceptions of lawmaking. Chicago, IL: The University of Chicago Press.

Barber

McCarty

(2015). Causes and consequences of polarization. In Mansbridge

Martin

C. J.

(Eds.), Solutions to political polarization in america (pp. 19-53).Washington, DC: American Political Science Association.

Bischof

Zurich

(2017). New graphics schemes for stata: Plotplain & plottig. The Stata Journal, 17(3), 748-759.

Bowler

Karp

J. A.

(2004). Politicians, scandals, and trust in government. Political Behavior, 26, 271-287.

Brambor

Clark

W. R.

Golder

(2006). Understanding interaction models: Improving empirical analyses. Political Analysis, 14, 63-82.

Branscombe

N. R.

Ellemers

Spears

Doosje

(1999). The context and content of social identity threat. In Ellemers

Spears

Doosje

(Eds.), Social identity: Context, commitment, content (pp. 35-58). Oxford, England: Blackwell.

Campbell

Converse

P. E.

Miller

W. E.

Stokes

D. E.

(1960). The American voter. New York, NY: John Wiley.

Carmines

E. G.

Stimson

J. A.

(1989). Issue evolution: Race and the transformation of American politics. Princeton, NJ: Princeton University Press.

Carson

J. L.

Koger

Lebo

M. J.

Young

(2010). The electoral costs of party loyalty in congress. American Journal of Political Science, 54, 598-616.

10.

Chanley

V. A.

Rudolph

T. J.

Rahn

W. M.

(2000). The origins and consequences of public trust in government: A time series analysis. Public Opinion Quarterly, 64, 239-256.

11.

Citrin

(1974). Comment: The political relevance of trust in government. American Political Science Review, 68, 973-988.

12.

Citrin

Green

D. P.

(1986). Presidential leadership and the resurgence of trust in government. British Journal of Political Science, 16, 431-453.

13.

Cole

R. L.

(1973). Towards a model of political trust: A causal analysis. American Journal of Political Science, 17, 809-817.

14.

Craig

S. C.

(1993). The malevolent leaders: Popular discontent in America. Boulder, CO: Westview Press.

15.

Delli Carpini

M. X.

Keeter

. (1996). What Americans know about politics and why it matters. New Haven, CT: Yale University Press.

16.

Dodd

L. C.

Schraufnagel

(2012). Congress and the polarity paradox: Party polarization, member incivility and enactment of landmark legislation, 1891-1994. Congress & the Presidency, 39, 109-132.

17.

Dodd

L. C.

Schraufnagel

(2013). Taking incivility seriously. In Frisch

Kelly

(Eds.), Politics to the extreme (pp. 71-91). New York, NY: Palgrave Macmillan.

18.

Downs

(1957). An economic theory of democracy. New York, NY: Harper & Row.

19.

Druckman

J. N.

Peterson

Slothuus

(2013). How elite polarization affects public opinion formation. American Political Science Review, 107, 57-79.

20.

Easton

(1965). A systems analysis of political life. New York, NY: John Wiley.

21.

Erber

Lau

(1990). Political cynicism revisited: An information-processing reconciliation of policy-based and incumbency-based interpretations of changes in trust in government. American Journal of Political Science, 34, 236-253.

22.

Flavin

(2013). Policy representation and evaluations of state government. State and Local Government Review, 45, 139-152.

23.

Garand

J. C.

(2010). Income inequality, party polarization, and roll-call voting in the U.S. Senate. The Journal of Politics, 72, 1109-1128.

24.

Gelman

Hill

(2006). Data analysis using regression and multilevel/hierarchical models. Cambridge, UK: Cambridge University Press.

25.

Hetherington

M. J.

(1998). The political relevance of political trust. American Political Science Review, 92, 791-808.

26.

Hetherington

M. J.

(2001). Resurgent mass partisanship: The role of elite polarization. American Political Science Review, 95, 619-631.

27.

Hetherington

M. J.

Nugent

J. D.

(2001). Explaining public support for devolution: The role of political trust. In Hibbing

J. R.

Theiss-Morse

(Eds.), What is it about government that Americans dislike? (pp. 134-151). Cambridge, UK: Cambridge University Press.

28.

Hetherington

M. J.

Rudolph

T. J.

(2015). Why Washington won’t work: Polarization, political trust, and the governing crisis. Chicago, IL: The University of Chicago Press.

29.

Hinchliffe

K. L.

Lee

F. E.

(2015). Party competition and conflict in state legislatures. State Politics & Policy Quarterly, 16, 172-197.

30.

Hinich

M. J.

Munger

M. C.

(1997). Analytical politics. Cambridge, UK: Cambridge University Press.

31.

Huddy

Mason

Aarøe

(2015). Expressive partisanship: Campaign involvement, political emotion, and partisan identity. American Political Science Review, 109, 1-17.

32.

Iyengar

Sood

Lelkes

(2012). Affect, not ideology: A social identity perspective on polarization. Public Opinion Quarterly, 76, 405-431.

33.

Jackson

(2010). Strategic media, cynical public? Examining the contingent effects of strategic news frames on political cynicism in the United Kingdom. The International Journal of Press/Politics, 16, 75-101.

34.

Johnson

O’Grady

(2013). Speakers and the spotlight explaining media coverage of leadership in the house of representatives. American Politics Research, 41, 498-532.

35.

Jones

D. R.

(2015). Declining trust in congress: Effects of polarization and consequences for democracy. The Forum, 13, 375-394.

36.

Keele

(2005, August). The authorities really do matter: Party control and trust in government. The Journal of Politics, 67, 873-886.

37.

Kelleher

C. A.

Wolak

(2007). Explaining public confidence in the branches of state government. Political Research Quarterly, 60, 707-721.

38.

King

D. C.

(1997). The polarization of American parties and mistrust of government. In Nye

J. S.

Zelikow

P. D.

King

D. C.

(Eds.), Why people don’t trust government (pp. 155-178). Cambrige, MA: Harvard University Press.

39.

Kousser

T. B.

(2010). Does partisan polarization lead to policy gridlock in California? California Journal of Politics and Policy, 2, 1-23.

40.

Lebo

M. J.

McGlynn

A. J.

Koger

(2007). Strategic party government: Party influence in congress, 1789–2000. American Journal of Political Science, 51, 464-481.

41.

Lee

F. E.

(2015). How party polarization affects governance. Annual Review of Political Science, 18, 261-282.

42.

Mansbridge

J. M.

(1997). Social and cultural causes of dissatisfaction with the U.S. government. In Nye

J. S.

Zelikow

P. D.

King

D. C.

(Eds.), Why people don’t trust government (pp. 133-154). Cambridge, MA: Harvard University Press.

43.

McCarty

Poole

K. T.

Rosenthal

(2006). Polarized America: The dance of ideology and unequal riches. Cambridge, MA: MIT Press.

44.

Miller

A. H.

(1974). Political issues and trust in government: 1964-1970. American Political Science Review, 68, 951-972.

45.

Morris

J. S.

Clawson

R. A.

(2005). Media coverage of congress in the 1990s: Scandals, personalities, and the prevalence of policy and process. Political Communication, 22, 297-313.

46.

Orren

(1997). Fall from grace: The public’s loss of faith in government. In Nye

J. S.

Zelikow

P. D.

King

D. C.

(Eds.), Why people don’t trust government (pp. 77-108). Cambridge, MA: Harvard University Press.

47.

Ramirez

M. D.

(2009). The dynamics of partisan conflict on congressional approval. American Journal of Political Science, 53, 681-694.

48.

Richardson

L. E.

Jr. Konisky

D. M.

Milyo

(2012). Public approval of state legislatures. Legislative Studies Quarterly, 37, 99-116.

49.

Shor

McCarty

(2011). The ideological mapping of American legislatures. American Political Science Review, 105, 530-551.

50.

Stokes

D. E.

(1962). Popular evaluations of government: An empirical assessment. In Cleveland

Lasswell

H. D.

(Eds.), Ethics and bigness: Scientific, academic, religious, political, and military (pp. 61-72). New York, NY: Harper.

51.

Theiss-Morse

Barton

D.-G.

Wagner

M. W.

(2015). Political trust in polarized times. In Bornstein

B. H.

Tomkins

A. J.

(Eds.), Motivating cooperation and compliance with authority (pp. 167-190). New York, NY: Springer.

52.

Theriault

S. M.

(2006). Party polarization in the US congress member replacement and member adaptation. Party Politics, 12, 483-503.

53.

Theriault

S. M.

(2008). Party polarization in congress. Cambridge, UK: Cambridge University Press.

54.

Towfigh

E. V.

Goerg

S. J.

Glöckner

Leifeld

Llorente-Saguer

Bade

Kurschilgen

(2016). Do direct-democratic procedures lead to higher acceptance than political representation? Public Choice, 167, 47-65.

55.

Uslaner

E. M.

(2015). Congressional polarization and political trust. The Forum, 13, 361-373.

56.

Wolak

Palus

C. K.

(2010). The dynamics of public confidence in U.S. state and local government. State Politics & Policy Quarterly, 10, 421-445.

57.

Zaller

J. R.

(1992). The nature and origins of mass opinion. New York, NY: Cambridge University Press.

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

0.57 MB