Political Knowledge and Policy Representation in the States

Abstract

Political knowledge is central to the success of representative democracy. However, public policy has been shown to follow public opinion even despite low levels of political information in the electorate. Does this mean that political knowledge is irrelevant to policy representation? We consider whether knowledgeable electorates are better able to achieve representative policy outcomes. Using the heterogeneity in the responsiveness of government across the states, we consider how state political knowledge moderates the connection between citizen ideology and the policy outcomes of state government. Using national surveys and multilevel logit with post-stratification, we develop measures of collective political knowledge in the states. We test whether knowledgeable electorates are more likely to secure representative political outcomes than less politically informed constituencies. We find that as state political knowledge increases, so does the correspondence between the preferences of the public and the ideological tenor of state policy outcomes.

Keywords

political knowledge representation state politics

Political knowledge is at the heart of representative democracy. If elected officials are to vote in line with the preferences of their constituents, they need some knowledge of what the people in their district want. If citizens are to hold politicians accountable for their choices, they need information to evaluate how well that representative performs his or her job. Indeed, debates about the degree to which the U.S. system of government fits a model of representative democracy often hinge on questions of political knowledge and information. Citizens delegate political decision-making power to elected representatives—who may choose to follow the wishes of their constituents or not. For citizens who want to insure against the risk of an unresponsive representative, knowledge is a source of political power.

Citizens who hold greater stocks of information and political knowledge are better equipped to hold representatives accountable as well as generally take on the challenges of good citizenship. Political information helps citizens form opinions on what they believe in politics and what they think politicians should do (Krosnick & Milburn, 1990). Political knowledge helps people communicate their wishes to elected officials and identify unresponsive politicians (Delli Carpini & Keeter, 1996). Political knowledge is not only central to definitions of good citizenship (Berelson, 1952) but also has been demonstrated to promote civic competence in practice. When people are knowledgeable about politics, they make choices at the ballot box that better reflect their interests (Gelman & King, 1993; Lau, Andersen, & Redlawsk, 2008). Because knowledgeable citizens demonstrate more ideological constraint and attitude consistency, they arguably send sharper signals to politicians about their preferences (Delli Carpini & Keeter, 1996). Those who hold more political information are also better able to understand the political stories they see in the news and better prepared to translate their political attitudes into action (Graber, 1984; Popkin & Dimock, 1999).

Although political information helps citizens meet the demands of citizenship, we do not have decisive evidence that levels of political knowledge promote policy responsiveness. Although individual voters with political information are better equipped to hold politicians accountable, this does not guarantee that knowledgeable electorates collectively are more successful at obtaining representative policy outcomes from elites. In fact, a common refrain in political science is that ideological policy representation comes despite the knowledge levels of the public. Surveys demonstrate the profound limits to Americans’ knowledge of politics—about 20% can name the three branches of government, about half know what a filibuster is, and usually less than half can name their representative in Congress (Delli Carpini & Keeter, 1996). But even as people struggle with the facts of how government works, we also see that government responds to the demands of the public. Public preferences explain the policy outcomes of government at the national level as well as at the state and local level (Erikson, Wright, & McIver, 1993; Palus, 2010; Stimson, MacKuen, & Erikson, 1995).

This suggests that policy responsiveness does not require a knowledgeable public. Instead, citizens might obtain representative policy outcomes through strategies of low information rationality—using heuristics to approximate fully informed decision making (Lupia, 1994) or by relying on online processing, where people can make informed decisions even absent the ability to recall the specific evidence that informed their judgment (Lodge, Steenbergen, & Brau, 1995). Policy responsiveness could also reflect the wisdom of crowds, where the opinions of the collective are coherent and informative even if many of the underlying opinions of individuals are nonattitudes or noise (Converse, 1990). Or, policy responsiveness might be maintained through political institutions that act in the public interest—whether the collective representation offered by political parties or the scrutiny of journalists as a check on politicians (Hochschild, 2010).

We argue that the level of political knowledge held by the public promotes the quality of policy representation. Even though an informed public may not be a necessity to secure representative outcomes from government, we believe that it helps. To explore the consequences of political knowledge for representation, we focus our attention on variations in knowledge across the states, and how these differences affect the connection between citizen liberalism and state policy outputs. States are an important case to examine, given the substantial policy responsibilities held by the states as well as the trend toward increasing state responsibilities over time. Because both levels of state political knowledge and the quality of policy representation vary across the states (Lyons, Jaeger, & Wolak, 2013; Lax & Phillips, 2012; Pacheco, 2013), examining state policy outcomes provides an excellent opportunity to examine the quality of representation across contexts. We expect that when state electorates hold more political information, they obtain policy outcomes that are closer to their ideological leanings. Using national surveys and multilevel logit and Poisson models with post-stratification, we create multiple measures of state political knowledge. We then test the conditioning effects of aggregate knowledge on ideological representation using variations in public policy outputs across the 50 states. We find that collective political knowledge strengthens the connection between citizen ideology and state policy outcomes. Political knowledge promotes ideological policy representation.

Explaining Policy Outcomes in the States

Some states produce liberal policies whereas others have a conservative tilt. What explains these state-to-state variations in the ideological tenor of policy outcomes? It has a lot to do with the preferences of the state electorate. The liberalism of citizens in a state strongly predicts the ideological composition of public policy in the states (Erikson et al., 1993). Not only is there a connection between opinion and policy generally, but state public opinion is also influential on a range of specific policy outcomes in the states (i.e., Brace, Sims-Butler, Arceneaux, & Johnson, 2002; Camobreco & Barnello, 2008; Haider-Markel & Kaufman, 2006; Hill & Hinton-Andersson, 1995; Johnson, 2001; Johnson, Brace, & Arceneaux, 2005; Mooney & Lee, 2000; Norrander, 2000).

However, policy and public preferences are not perfectly congruent. Conservative electorates are more likely to get conservative policy than liberal policy, but the gap between preferences and policy is sometimes large and sometimes small. Elected officials are not all as equally quick in reacting to changes in public preferences, where members of the House are faster to respond to opinion than members of the Senate (Stimson et al., 1995). Some citizens wield greater influence on policy than others, as a product of their wealth or political resources (Bartels, 2008; Flavin, 2012; Gilens, 2012; Rigby & Wright, 2011). Some issue domains reveal a closer congruence between preferences and policy than others (Burstein, 2006; Shaffer & Weber, 1974; Sutton, 1973; Wlezien, 2004). We also see that some states produce policies that are more representative of citizen ideology than others (Lax & Phillips, 2012; Pacheco, 2013; Sutton, 1973; Treadway, 1985).

Why is the match between preferences and policy stronger in some states and weaker in others? The roots of this gap could be in institutions, where legislative rules and levels of professionalism can promote more representative policy outcomes (Burden, 2005; Lax & Phillips, 2012; Maestas, 2000). Others suggest that the accessibility of ballot initiatives promotes the correspondence of public opinion to policy, though evidence on this point tends to be quite mixed (Arceneaux, 2002; Camobreco, 1998; Gerber, 1996; Lascher, Hagen, & Rochlin, 1996; Lax & Phillips, 2012; Monogan, Gray, & Lowery, 2009). It might be due to patterns of interests and engagement, where some states have more active interest groups than others, though the moderating effects tend to be weak (Gray, Lowery, Fellowes, & McAtee, 2004; Schneider & Jacoby, 2006). The nature of representation in the states could also depend on the character of the policy agenda, as when issues are more salient, government responsiveness is more likely (Lax & Phillips, 2009a; Pacheco, 2013).

Most of the attention to representational inequalities in the realm of state politics has focused on the character of institutional environments, where some environments are more conducive to legislators acting on the wishes of their constituents. We argue that the attributes of the electorate also shape the quality of political responsiveness in the states, a line of inquiry that has received little attention in prior research. An important exception is given by Barrilleaux (2006), who shows that the diversity of ideological preferences in a state tempers the effects of citizen ideology on policy outcomes. We build on this by considering how the sophistication of the electorate facilitates a connection between public preferences and state policy.

The Importance of Knowledge for Representation

We are sometimes surprised by the correspondence of public preferences and policy outcomes in the states, given a sense that voters are not well enough informed about state matters to enforce accountability from state-elected officials (Treadway, 1985). To reconcile a sense of an uninformed state electorate against clear evidence of state policy responsiveness, Erikson et al. (1993) turn to arguments of preference aggregation. Perhaps, some citizens are not well informed about state matters, and maybe not all hold meaningful attitudes on all policy issues that state governments consider. But in the aggregate, when we sum together the orientations of the state electorate at large, those who answer survey questions with guesses or nonattitudes cancel each other out. The aggregate character of the policy mood of the state electorate then reflects the attitudes of those with substantive attitudes and preferences (Converse, 1990; Erikson, MacKuen, & Stimson, 2002; Page & Shapiro, 1994). Even if some have uninformed views or nonattitudes, the collective ideological preferences of the state electorate carry more meaning than attitudes of singular citizens.

We agree that this process of preference aggregation should help legislators act in line with the public will, even absent a fully informed public. However, such an account does not rule out information as a part of the representation story. We expect that although policy follows collective preferences on average, citizen ideology does not have equal influence in all environments. We believe that the resources of the citizenry of a state—as captured by their level of political knowledge about state politics—affect the responsiveness of state policy outputs to citizen ideology. When knowledge is high, we expect a closer match between the liberalism of preferences and policy. When the public’s knowledge is low, we expect a weaker connection.

A knowledgeable electorate can improve the quality of representation through several pathways. First, a knowledgeable electorate should be better able to communicate his or her preferences to politicians. One challenge for politicians in following the preferences of their constituents is in knowing what they want. After all, if many struggle to name the branches of state government or the name of the governor, how can they communicate their preferences on state policy proposals? As levels of political information climb in a state, such worries should decline. Voters who are informed about politics are better equipped to signal to elected officials what they want from state government. After all, those with more political knowledge are more likely to be involved in campaigns or community projects or otherwise contact elected officials (Verba, Schlozman, & Brady, 1995). When citizens are better informed about state politics, state officials should be better able to gauge the preferences and priorities of the public. Indeed, in states where citizens are better educated, campaign contributors appear to have less state legislative influence—presumably because citizens have greater voice (Powell, 2012).

Political knowledge should also help voters hold politicians accountable for the performance of state government. If citizens know little about what their state legislature tackles or how the governor votes, how can they hold state officials accountable for those choices? When the level of political knowledge in a state increases, voters should be better able to distinguish the competitors on the ballot and choose the candidates that most align with their interests. We know that elections are an important way that citizens direct the ideological tenor of the policies passed by government (Erikson et al., 1993), so greater stocks of knowledge can serve to improve the quality of decision making in state elections. A politically aware citizenry is better able to enforce accountability, recognizing when representatives are not voting as the district would prefer and voting them out of office.¹

Finally, although it is clear that collective opinions are more coherent and sophisticated than individual opinions of Americans, collective opinions are not free from distortions (Althaus, 2003; Bartels, 1996). People want different policies, value different considerations, and choose different candidates depending on their level of information (Bartels, 1996; Delli Carpini & Keeter, 1996; Sniderman, Glaser, & Griffin, 1991). If information is unequally distributed in the electorate, then the collective signals sent by voters might be unrepresentative of the opinions of the state at large (Converse, 1990). When political knowledge in a state is low, we should expect it to be unevenly distributed in the state electorate. If only a small share of the state electorate is politically informed, they could become the only constituency of concern to representatives. Rather than consider the preferences of all, representatives could instead respond only to those capable of monitoring their political choices. But when a state electorate is widely knowledgeable, policy distortions should be less likely. When many hold political information, then representatives must be concerned with the interests of the citizenry as a whole, rather than an informed few.

We remain agnostic about which of these mechanisms is most important, but suspect that all are important pathways by which knowledge operates to promote policy representation. Our focus instead is on demonstrating whether knowledge plays any role in moderating the relationship between citizen preferences and state policy outcomes. By considering the effects of knowledge on the responsiveness of state government generally, we can broadly inform the mechanisms of representation. If state policy representation is conditioned on the political sophistication of the electorate, it will suggest that citizen monitoring of government helps enforce legislator accountability. If differences in political knowledge across the states have no effect on how state policy outcomes are produced, it will indicate that policy representation arrives through other mechanisms. Perhaps the choices made by elites are more important than the capacities of the electorate, or the congruence of policy and preferences could merely reflect opinion sharing between legislators and the character and culture of the districts from which they are drawn.

Data and Measures

To consider how political knowledge affects the translation of public preferences into policy outcomes, we first need to measure the amount of political knowledge held by the constituency of a state. In particular, we need to identify measures of knowledge of state politics. If people are to hold the governor and the state legislature accountable for the performance of state government, they need some knowledge of state matters, apart from general knowledge of national politics. Because knowledge of state politics is the most useful kind of information electorates could have in securing representative state policy outcomes, a measure of state knowledge is preferable to some measure of general political knowledge. General political knowledge and domain-specific knowledge are distinctive in character (Gilens, 2001; Iyengar, 1990), and specifically, people’s knowledge of state government is distinct from general political knowledge in its origins and nature (Delli Carpini & Keeter, 1996). Although education and interest are associated with political knowledge generally, knowledge of state politics has less to do with dispositional political interest or demographic characteristics than knowledge of national politics (Lyons et al., 2013).² As such, a measure of the public’s awareness of state politics is better suited to test the specific effects of political knowledge than something such as a measure of educational attainment at the state level.

We expect state political knowledge to be the most important in translating citizen preferences into state policy outcomes. However, finding a good measure of people’s knowledge of state politics is tricky. Whereas many surveys ask people about their knowledge of national politics, very few surveys ask about people’s knowledge of state politics. From the limited existing options, we identify three surveys with questions about state political knowledge. From these surveys, we create two different kinds of measures of political knowledge—specific knowledge and partisan knowledge. Our first measure reflects correct identification of the name of the governor of the state. As a measure of political knowledge, identification of one’s governor seems a reasonable indicator of the public’s level of knowledge of state politics. A person who cannot name the governor in a survey might still be otherwise informed about state matters, but we believe that most people who are knowledgeable about state politics would be able to answer this question. Because it is not asked in the course of a campaign season, we expect it reflects people’s general awareness of state political matters, not simply the campaign efforts of a particular campaign season or slate of candidates. We also believe such a measure is preferable to items asking about factual knowledge of the design of state institutions or the rules of the game, as it should better capture contemporary knowledge of the performance of state government. Being able to name the governor is the kind of useful operational knowledge that helps people hold state-elected officials accountable (Lupia, 2015).

In addition to our recall-based measure of specific state political knowledge, we also construct a set of measures that capture partisan knowledge of state politics. Using responses from the Cooperative Congressional Election Study (CCES), we create measures of people’s correct identification of the partisan composition of their state government. Using the 2006 CCES, we create a state-level measure of the share of respondents who correctly name the partisanship of the governor. Using the 2007 CCES, we create a duplicate measure of gubernatorial partisan knowledge as well as a scale of correct answers to questions about partisan control of the lower house of the state legislature, the state senate, and the governor.

These measures of partisan knowledge are a useful complement to our specific political knowledge measure. Political scientists have argued that tests of political recall can be too high of a bar for evaluating citizen competence in holding elected officials accountable. After all, people need not remember the name of the incumbent, so long as they are able to recognize his or her name on the ballot and vote accordingly (Mann & Wolfinger, 1980). Knowing the partisanship of the governor is arguably among the most useful things people might know about state politics, particularly if they want to be able to make informed choices at the ballot box. Although these two types of measures of political knowledge (specific knowledge and partisan knowledge) do not encompass the full range of knowledge that people might hold about state politics, we believe that they represent some of the most relevant kinds of political knowledge to helping citizens secure representative policy outcomes from state government. Also, using multiple measures of political knowledge, we can mitigate concerns that our results reflect some attribute of our measures, rather than the fundamental differences in the political knowledge of state electorates.

Specific Knowledge of State Politics

We start by considering the degree to which specific knowledge of state politics moderates the connection between citizen ideology and policy liberalism in the states. For our measure of specific knowledge of state politics, we rely on a February 2007 survey conducted by Pew Research Center for the People and the Press. In it, 1,502 respondents were asked to identify the governor of their state.³ The question is reasonably difficult for respondents, where 66% of the sample were able to correctly identify the governor. With 1,502 respondents in the survey, we should not trust a measure that simply disaggregates correct answers by state.⁴ So we instead create a measure of state-level political knowledge using multilevel regression and post-stratification (MRP) (Lax & Phillips, 2009b; Park, Gelman, & Bafumi, 2004, 2006). In this approach, a state-level measure is constructed by taking responses from a national survey, building a multilevel logit explanatory model, and weighting the predictions from this model by the population characteristics of the state. Such an approach has been shown to produce measures that have strong external validity, better than disaggregating national samples particularly when state samples are small (Buttice & Highton, 2013; Lax & Phillips, 2009b; Park et al., 2004). By using multilevel modeling with post-stratification, we can also create estimates of state political knowledge for Alaska, Hawaii, and Wyoming, even though no survey respondents reported residence in these states.

To construct our state-level measure, we follow the advice of Lax and Phillips (2013). We estimated a multilevel logit model of knowledge of the governor.⁵ For the individual-level model of political knowledge, we include random effects associated with age, education, and the interaction of race and gender.⁶ These demographic factors have all been shown to be important predictors of levels of political knowledge (Delli Carpini & Keeter, 1996). Among the state-level predictors, we include fixed effects of the percent college educated by state and per capita personal income at the state level.⁷ We take the predictions of state political knowledge for each demographic subset of the sample, and then weight these responses according to how these groups are represented within states, using post-stratification and population data for demographic groups by state.⁸ Table 1 shows our estimates of knowledge of state politics by state.⁹ We see that levels of state political knowledge are not uniform, but variable across the states. The lowest level is in Virginia, where the knowledge level is 35%, whereas the highest level is in California, where 88% answer the state political knowledge question correctly.¹⁰

Table 1.

Levels of Political Knowledge Across the States.

State name	2007 Pew	2006 CCES	2007 CCES
State name	% who recall name of governor	% correctly identifying party of governor	% correctly identifying party of governor	Items correct on partisan knowledge scale
Alabama	63.76	82.11	71.40	1.45
Alaska	68.44	90.40	81.47	1.78
Arizona	52.10	91.34	82.10	1.78
Arkansas	67.08	73.32	72.27	1.57
California	88.31	94.09	85.95	2.14
Colorado	50.54	84.68	80.70	2.00
Connecticut	80.51	85.28	86.96	1.89
Delaware	69.05	69.61	75.97	1.58
Florida	70.13	91.66	84.28	1.88
Georgia	64.94	84.97	75.59	1.73
Hawaii	67.23	87.85	83.83	1.88
Idaho	61.15	80.24	77.22	1.93
Illinois	70.38	87.46	85.92	1.96
Indiana	63.03	75.63	74.45	1.44
Iowa	60.89	78.23	79.41	1.84
Kansas	65.53	73.51	73.53	1.71
Kentucky	70.52	78.12	81.03	1.58
Louisiana	67.83	80.78	79.28	1.60
Maine	70.06	84.62	81.58	1.78
Maryland	60.95	87.73	82.81	2.06
Massachusetts	67.56	90.54	90.61	2.36
Michigan	80.43	90.79	89.50	1.69
Minnesota	82.52	88.75	85.31	2.02
Mississippi	57.32	72.45	76.24	1.38
Missouri	59.11	82.46	80.37	1.63
Montana	60.82	80.88	83.39	1.71
Nebraska	64.55	76.36	76.03	—
Nevada	63.56	72.66	78.64	1.39
New Hampshire	63.79	82.17	84.59	1.96
New Jersey	68.47	85.05	88.79	2.10
New Mexico	69.30	89.96	78.15	1.70
New York	65.96	81.71	80.50	1.63
North Carolina	45.49	68.30	69.31	1.57
North Dakota	69.40	78.93	78.13	1.64
Ohio	44.75	80.44	80.71	1.56
Oklahoma	66.99	76.29	75.64	1.42
Oregon	62.51	85.71	83.33	1.98
Pennsylvania	71.37	81.62	77.70	1.36
Rhode Island	65.96	73.53	78.24	1.98
South Carolina	58.81	77.79	72.53	1.70
South Dakota	71.31	77.87	78.34	1.72
Tennessee	68.25	77.38	70.17	1.42
Texas	60.07	90.85	84.11	1.92
Utah	67.89	85.53	77.76	2.08
Vermont	69.76	76.05	83.29	2.05
Virginia	35.25	82.35	84.48	1.65
Washington	64.01	88.47	85.53	2.07
West Virginia	66.10	75.45	70.96	1.64
Wisconsin	76.22	82.40	82.42	1.62
Wyoming	75.59	83.81	83.70	1.63

Note. CCES = Cooperative Congressional Election Study.

We expect that levels of political knowledge in the states moderate the translation of citizen ideology to the liberalism of policy outcomes in the states. As our measure of citizen ideology, we use measures created by Enns and Koch (2013) that aggregate survey responses on ideological identification across the states. Our measure reflects the percent liberal identifiers in a state in 2006 as a share of those who identify as liberal or conservative.¹¹ As our measure of policy outcomes, we rely on the index of state policy liberalism as created by Sorens, Muedini, and Ruger (2008) and updated through 2008 by Sorens and Ruger.¹² The measure reflects state policies on a wide array of ideological issues that face the states, from law enforcement to education to the environment to eminent domain reform. It is similar to measures of policy liberalism created by Erikson et al. (1993) and updated by Gray et al. (2004), though it is based on a far larger number of issues and more narrowly anchored in the specific policy outputs of state government (rather than general state policy outcomes). We use the first dimension of the measure, where higher scores indicate states with more liberal policies.¹³

Knowledge as a Moderator of Citizen Ideology’s Effect on State Policy Outcomes

We are interested in the degree to which the connection between the ideology of state residents and the ideological tenor of state policy outputs is conditioned by levels of state political knowledge. Using regression, we consider the effects of the liberalism of the state electorate as moderated by levels of state political knowledge. Results are shown in Table 2.¹⁴ We find that as the level of political knowledge in the state electorate increases, so does the correspondence between the preferences of the public and the policies of the state. We find a near significant interaction term, indicating that the effects of ideology on policy vary by knowledge level.¹⁵

Table 2.

Moderating Effects of Knowledge on Liberal Policy Outcomes.

	Policy liberalism
Knowledge of governor’s name	−47.51	−58.20^†
Knowledge of governor’s name	(31.15)	(31.18)
State citizen ideology	−41.76	−71.40
State citizen ideology	(61.75)	(63.02)
State Knowledge × State Ideology	159.37^†	191.73*
State Knowledge × State Ideology	(90.16)	(90.44)
Legislative liberalism	—	0.04^†
Legislative liberalism		(0.02)
Constant	10.00	17.96
Constant	(21.13)	(21.25)
R ²	.73	.75
N	50	50

Note. OLS regression estimates. Standard errors in parentheses. OLS = ordinary least squares.

†

p < .10. *p < .05.

With interaction effects, it is important to go beyond looking at only the significance of the interaction term to also examine the marginal effects of ideology across the states. To get a better sense of the moderating effect of state knowledge, we graph the marginal effects of citizen ideology across the range of state political knowledge in Figure 1. At the lowest levels of state political knowledge, the effects of public liberalism on policy are statistically indistinguishable from zero. In other words, when state knowledge is low, public preferences are unconnected to the ideology of state policies. In states where more than 49% of the electorate is expected to identify their governor correctly, we observe a statistically significant effect of citizen ideology on state policy outcomes.¹⁶ As the level of state political knowledge increases, the relationship between public liberalism and policy increases in magnitude. For example, in a state such as Mississippi, where 57% are knowledgeable about state politics, we would expect a one standard deviation increase in public liberalism to increase the liberalism of state policy by 3.9 points (about three quarters of a standard deviation increase in policy liberalism on this scale). However, if the level of knowledge in the state was instead 77%, we would expect the same standard deviation increase in citizen liberalism to result in a 6.4 increase in policy liberalism—an effect that is almost twice as large. As the knowledge of the state electorate climbs, so does the ideological responsiveness of state government. When state political knowledge is at its highest level, liberal and conservative electorates obtain the most ideologically distinctive policy outcomes.

Figure 1.

Marginal effect of ideology on policy liberalism, across levels of specific political knowledge of name of the governor.

When people are knowledgeable about state politics, they might be more successful in choosing candidates who align with their policy interests. It might also be that knowledgeable citizens create pressures on state-elected officials even outside elections, and this strengthens the connection between citizen ideology and the tenor of state policy outcomes. If the knowledge still conditions the translation of public preferences into policy liberalism even after controlling for the relative liberalism of the state legislature, then it will indicate that knowledge’s influence is not through electoral choices alone. To test this, we next add a measure of the liberalism of the state legislature—as shown in column 2 of Table 1. As our measure of state government ideology, we use an updated version of Berry, Ringquist, Fording, and Hanson’s (1998) government ideology measure for 2007.¹⁷ In adding the ideological composition of state government to our model, we do not dilute the moderating effects of state knowledge in the translation of citizen ideology into state policy outcomes. The slope of the marginal effect for citizen ideology shown in Figure 1 only becomes steeper with the addition of this control. This suggests that aggregate levels of state political knowledge promote policy representation even outside the electoral connection.

In the online supplemental appendix, we also tested the degree to which the effects of state political knowledge are distinctive from the moderating effects of knowledge of national politics. We find that the moderating effects of knowledge of the governor are robust to the inclusion of the interaction of citizen ideology and measures of national political knowledge, as well as controls for the interaction of citizen liberalism and levels of educational attainment in the states.¹⁸ This suggests that the effects we observe are not simply a product of the differences in the levels of resources across the states, but instead a specific consequence of levels of state political knowledge.

Heuristics Knowledge and Ideological Policy Representation in the States

We have shown that greater stocks of state politics knowledge facilitate the translation of citizen ideology into state policy outcomes. In states where people are very likely to know the name of the governor, the effects of citizen ideology on policy are greater. Yet, people might not need to know the name of the governor to still send signals to state government about their preferences. As such, we next consider whether partisan knowledge of state politics works in similar ways as our measure of specific knowledge of state politics.

To measure partisan knowledge, we first consider people’s ability to correctly identify the party affiliation of the governor of their state. In the 2006 and 2007 Cooperative Congressional Election Study, respondents were asked, “Do you happen to remember the party affiliation of the Governor of your state?” and were given four response options, Democrat, Republican, Independent, and “don’t know.” We code answers as correct or incorrect based on the partisanship of the governor in office, counting both “don’t know” responses and wrong answers as incorrect. This question captures knowledge of a useful shortcut in state politics. If one knows the party of the governor, then they are better able to assign partisan blame when encountering news of state policy successes or failures, which could in turn shape later vote choices or party evaluations.

We take advantage of the very large sample size of the 2006 Cooperative Congressional Election Study to create a nationally disaggregated measure, pooled by state as the average rate of correct answers by state. Given a suitably large number of respondents, this approach has been shown to yield valid and reliable estimates of state attitudes (i.e., Brace et al., 2002; Carsey & Harden, 2010; Erikson et al., 1993; Jones & Norrander, 1996; Norrander, 2000).¹⁹ By testing our models with a second measure of state political knowledge that asks about a different aspect of knowledge and relies on a different method of aggregation, we can also improve our confidence that our prior results are not simply a function of our selected measure of knowledge or the multilevel regression with post-stratification approach used to create our prior measure.

We find that people’s knowledge of this partisan heuristic is higher than specific recall of the name of the governor, as 85% of the sample is able to correctly identify the party affiliation of the governor of their state (compared with the 66% that were able to recall the governor’s name).²⁰ Levels of partisan heuristic knowledge by state are shown in the second column of Table 1. The lowest level of correct knowledge of the partisan heuristic is in North Carolina, where 68% correctly answer this question, and highest in California, where 94% know the party of the governor in office.

We also create a scale of partisan knowledge of state politics using questions about partisan control of both the governor’s office and the state legislature, taking advantage of a set of questions included in the 2007 Cooperative Congressional Election Study. Survey respondents were asked to identify the party of the governor as well as partisan control of upper house and the lower house of the state legislature.²¹ We combine these items to form a scale of partisan knowledge of state government. In the case of the 2007 CCES, the sample size is about a quarter of the size of the sample of the 2006 CCES, with 10,000 respondents rather than the more than 36,000 respondents of the 2006 study. As such, we rely on an MRP approach rather than simple disaggregation of survey responses into state samples.²² We use the identical approach that we did to create our state-aggregated measure of recall of the governor’s name. We first create a measure of correct knowledge of the governor’s party in 2007, to allow for comparisons with our 2006 measure. Second, we produce a scale of partisan knowledge based on the number of correct answers to questions about knowledge of partisan control of the statehouse, the state senate, and the governor’s office.²³

People are less likely to know the party composition of their state legislature than they are the party of the governor. In the 2007 CCES, 79% correctly name the party of the governor, whereas 41% correctly name the party that controls the statehouse and 46% correctly name the party that controls the state senate.²⁴ Using a multilevel Poisson model with post-stratification, we generate scores of the predicted number of correct partisan knowledge items by state, as shown in Table 1.²⁵ It is a useful measure of political knowledge for our purposes in that it assesses people’s knowledge of state politics beyond knowledge of the governor, so that it should be less likely to be determined by any particular distinctive attributes of the state’s chief executive. It is also composed of knowledge items of greater difficulty, which should result in a scale that can more strongly differentiate states of high and low levels of knowledge.

To consider the moderating effects of partisan knowledge on the connection between state preferences and policy, we rely on the same interactive specification.²⁶ If we observe a significant interaction effect associated with ideology and partisan knowledge, it will mean that both partisan knowledge as well as specific political knowledge promote ideologically representative outcomes in the states. Results of our regression models are shown in Table 3. We find that the effects of ideology on policy outputs vary across the states, depending on the share of the state electorate that possesses knowledge of party control of state government. We find significant interaction effects associated with the liberalism of the state electorate and each measure of state partisan knowledge.

Table 3.

Moderating Effects of Partisan State Knowledge on Liberal Policy Outcomes.

	State policy liberalism
	Knowledge of governor’s party, 2006 CCES		Knowledge of governor’s party, 2007 CCES		Knowledge of partisan control of state government, 2007 CCES
Partisan knowledge	−90.72*	−83.24*	−129.75*	−125.73*	−18.61^†	−16.57
Partisan knowledge	(35.62)	(37.13)	(42.83)	(43.39)	(9.91)	(10.26)
State citizen ideology	−161.25^†	−149.65^†	−282.91*	−276.82*	−37.04	−31.66
State citizen ideology	(83.26)	(85.05)	(107.86)	(108.72)	(51.64)	(52.28)
Partisan knowledge × Ideology	277.95*	259.27*	437.17*	425.51*	61.10*	55.81^†
Partisan knowledge × Ideology	(101.06)	(104.50)	(132.13)	(133.74)	(28.14)	(29.01)
Legislative liberalism	—	0.02	—	0.01	—	0.02
Legislative liberalism		(0.02)		(0.02)		(0.02)
Constant	51.73^†	46.03	81.42*	78.55*	9.41	6.29
Constant	(29.07)	(30.17)	(34.49)	(34.89)	(17.66)	(18.15)
R ²	.75	.75	.79	.79	.77	.77
n	50	50	50	50	49	49

Note. OLS regression estimates. Standard errors in parentheses. CCES = Cooperative Congressional Election Study; OLS = ordinary least squares.

†

p < .10. *p < .05.

The marginal effects of ideology across levels of partisan knowledge are shown in Figure 2. In the results from the 2006 CCES, we find that in states where less than 69% of the electorate is expected to know the party of the governor, citizen ideology does not have a significant effect on liberalism of state policy outputs.²⁷ As knowledge increases, so does the magnitude of the marginal effect of state ideology. Comparing the effects of citizen ideology on policy liberalism in the states for states at the 10th percentile of political knowledge with those at the 90th percentile, the effects of ideology are nearly double in size. The same pattern is seen in the case of the gubernatorial knowledge measure from the 2007 CCES. For states where less than 72% of the electorate is expected to know the governor’s partisanship, we find insignificant marginal effects associated with citizen ideology.²⁸ As partisan knowledge of the governor increases, so do the effects of citizen ideology on the liberalism of state policy outcomes.

Figure 2.

Marginal effect of ideology on state policy liberalism, across levels of partisan state political knowledge (2006 and 2007 CCES).

We also find that our scale of partisan knowledge moderates ideology’s influence on policy outcomes. Here, the marginal effects of ideology on policy liberalism are significant across the range of the knowledge scale in the states, but increasing in magnitude as partisan knowledge increases. Citizen ideology is a significant predictor of the liberalism of the state’s policy outcomes in both the state with the highest partisan knowledge and the lowest, but the magnitude of ideology’s effect on policy is twice as large in the state with the highest level of partisan political knowledge.

In the other models of Table 3, we include controls for the ideology of state government. If the effects of political knowledge diminished once we controlled for the conservatism of state legislators, it would suggest that part of knowledge’s influence was through the election of likeminded legislators. The general robustness of knowledge’s influence in the face of these controls suggests instead that state political knowledge plays an important role in encouraging ideological policy representation even outside of election seasons.

State Political Knowledge and Policy Congruence

So far, our analyses have focused on the ideological responsiveness of state government, and the degree to which state political knowledge facilitates the translation of public preferences into policy outcomes. Lax and Phillips (2012) propose a second metric of the quality of policy representation in the states: whether policy outcomes are congruent with the majority will of the state electorate. If most policies align with the ideological tilt of the majority, it suggests greater policy congruence. To test the robustness of our results, we next consider the effects of state political knowledge for the congruence of policy outcomes in the states, using the replication data set generously shared by Lax and Phillips (2012). To construct their measure of policy congruence, Lax and Phillips consider 39 policies in the states, related to issues of law enforcement, abortion, education, health care, and several other issues. Using multilevel modeling with post-stratification, they create measures of issue-specific state preferences in each domain. When a majority of the electorate prefers a conservative outcome and the state policy is conservative, the case is coded as congruent. Mismatches between the majority preference of the electorate and the direction of the existing policy are considered incongruent. A key virtue of this data set is that we are able to test the effects of political knowledge while controlling for prior explanations for policy congruence in the states. To interact various other institutional environments with citizen ideology would quickly consume our limited degrees of freedom. Because this data set uses congruence as its dependent variable rather than liberalism, we have greater leverage to explore the effects of state political knowledge on policy congruence relative factors such as issue salience and presence of ballot initiatives.

We next consider the effects of state political knowledge in explaining the policy congruence of state government across issue domains. We rely on multilevel logit models, specified as cross-classified multilevel models with random effects associated with the policy domain and the state. We control for the size of the opinion majority, the conservatism of the opinion majority, and the salience of the issue nationally, as Lax and Phillips (2012) find all three to be positively related to policy congruence in the states.²⁹ We also include a control for the level of interest group opposition in the states, given past studies of the role interest groups play in potentially inhibiting policy congruence (Gray et al., 2004; Lax & Phillips, 2012; Monogan et al., 2009; Schneider & Jacoby, 2006). Likewise, we include a control for whether the state allows for ballot initiatives and implements legislative term limits, given previous literature describing their potential importance to the quality of policy responsiveness in the states (Arceneaux, 2002; Camobreco, 1998; Lascher et al., 1996; Lax & Phillips, 2012; Monogan et al., 2009).

We also use measures of the liberalism of state government and the professionalism of the state legislature to conduct a partial test of the routes of political knowledge’s influence. As discussed earlier, policy representation can occur through elections, where voters select candidates who align with their interests and vote out incongruent representatives. Policy representation can also be promoted outside of campaign seasons through rational anticipation, where politicians use available information to forecast future electoral consequences for their policy choices. When legislators choose to follow public demands due to the anticipated reactions of voters at the next election, we observe greater policy representativeness. If knowledge’s main influence is through elections, where knowledgeable electorates are better able to select candidates who align with their interests and vote out incongruent representatives, then the effects of political knowledge should diminish in the presence of a control for the liberalism of state government. If the most knowledgeable state electorates are particularly adept at signaling their preferences to government outside of campaign seasons, then we should expect to see greater effects for knowledge in places that are most likely to be attuned to the needs of their constituents—states with high levels of legislative professionalism.³⁰

As such, we test three model specifications: a set of models that include our main measures of knowledge alongside explanations for policy congruence common to past studies, models that add controls for legislative liberalism, and a set of models where we add an interaction between state political knowledge with the professionalism of the state legislature. Results are shown in Table 4. Considering our measure of specific state knowledge, the share of the state predicted to be able to name the governor, we find no effects for knowledge on congruence. In the case of partisan knowledge, however, we find that policy congruence climbs with levels of state political knowledge. The greater the share of the state that can identify the governor’s party, the more likely it is that policy outcomes match the majority preference of the state electorate. On average, Lax and Phillips (2012) find 48% of policies are congruent with opinion majorities in the states. Our model predicts that in a state with the lowest level of knowledge of this partisan heuristic, the level of congruence will drop to 39%. At the highest level of state heuristic knowledge, our model predicts that policy outcomes are congruent with the majority preferences of the electorate 57% of the time. Likewise, we also find that states with higher average scores on the partisan knowledge scale also have significantly greater levels of policy congruence than states with low levels of state partisan knowledge. Knowledge of state politics plays a role in promoting the ideological congruence of state policy outcomes with the majority will of the state electorate.

Table 4.

Policy Congruence and Levels of State Political Knowledge.

	Policy congruence
	Knowledge as recall of governor’s name			Knowledge of governor’s party, 2006			Knowledge of partisan control of state government, 2007
Political knowledge	0.00	0.00	−0.02	2.80*	2.69*	−0.97	0.77*	0.68*	−0.38
Political knowledge	(0.01)	(0.01)	(0.01)	(0.99)	(0.98)	(1.86)	(0.28)	(0.28)	(0.49)
Size of opinion majority	0.13*	0.11*	0.11*	0.13*	0.11*	0.11*	0.13*	0.11*	0.11*
Size of opinion majority	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
Issue salience	0.56^†	0.49^†	0.49^†	0.56^†	0.49^†	0.50^†	0.58*	0.51^†	0.52^†
Issue salience	(0.29)	(0.27)	(0.27)	(0.29)	(0.27)	(0.27)	(0.30)	(0.28)	(0.28)
Conservative opinion majority	0.86*	0.86*	0.94*	0.91*	0.92*	0.99*	1.00*	1.00*	1.07*
Conservative opinion majority	(0.23)	(0.22)	(0.22)	(0.22)	(0.22)	(0.22)	(0.23)	(0.23)	(0.23)
Government ideological opposition	—	−0.54*	−0.54*	—	−0.53*	−0.53*	—	−0.52*	−0.54*
Government ideological opposition		(0.12)	(0.12)		(0.12)	(0.12)		(0.12)	(0.12)
Professionalism	—	—	−2.96	—	—	−11.19	—	—	−5.64
Professionalism			(3.64)			(7.63)			(3.47)
Knowledge × Professionalism	—	—	0.07	—	—	14.67^†	—	—	3.96*
Knowledge × Professionalism			(0.05)			(8.87)			(1.89)
Interest group opposition	−0.81*	−0.78*	−0.77*	−0.81*	−0.78*	−0.77*	−0.80*	−0.77*	−0.76*
Interest group opposition	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.13)	(0.13)	(0.13)
Term limits	0.48*	0.55*	0.46*	0.51*	0.58*	0.50*	0.65*	0.70*	0.61*
Term limits	(0.18)	(0.18)	(0.16)	(0.17)	(0.17)	(0.16)	(0.18)	(0.18)	(0.17)
Citizen initiative	−0.18	−0.24	−0.21	−0.27^†	−0.32*	−0.32*	−0.27^†	−0.32*	−0.33*
Citizen initiative	(0.17)	(0.17)	(0.15)	(0.16)	(0.16)	(0.15)	(0.16)	(0.16)	(0.15)
Constant	−12.39*	−10.81*	−9.83*	−14.41*	−12.81*	−10.02*	−13.83*	−12.18*	−10.66*
Constant	(2.01)	(1.94)	(2.07)	(2.15)	(2.08)	(2.45)	(2.08)	(2.01)	(2.13)
Variance components
Intercept, states	0.06	0.05	0.01	0.03	0.03	0.00	0.03	0.03	0.00
Intercept, states	(0.04)	(0.04)	(0.03)	(0.04)	(0.04)	(0.00)	(0.04)	(0.04)	(0.00)
Intercept, policy issue	2.70*	2.32*	2.36*	2.73*	2.36*	2.37*	2.83*	2.45*	2.49*
Intercept, policy issue	(0.73)	(0.63)	(0.64)	(0.73)	(0.64)	(0.64)	(0.76)	(0.67)	(0.68)
n	1,950	1,950	1,950	1,950	1,950	1,950	1,911	1,911	1,911

Note. Multilevel logit estimates. Standard errors in parentheses.

†

p < .10. *p < .05.

Considering our partial test of the routes of knowledge’s influence, we find that the effects of state political knowledge are positive and significant even after controlling for the liberalism of state government. This suggests again that knowledgeable electorates secure stronger policy representation through mechanisms other than election turnover alone.³¹ We also find evidence in line with arguments of rational anticipation. When considering the two measures of partisan knowledge, we find that their effects on promoting policy congruence are strengthened in states with higher levels of legislative professionalism (as indicated by the positive and significant interaction effects associated with partisan knowledge and professionalism).³² Although this is at best an indirect test of the rational anticipation mechanism, these results are at least suggestive that state political knowledge promotes policy representation through the improved communication between legislators and constituents in the states. When citizens have the resource of knowledge and legislators have institutional resources, representative policy outcomes are particularly likely.

Discussion

We have demonstrated that the knowledge held by the public alters the imprint of citizen ideology on the outcomes of state government. When state electorates are more knowledgeable about politics in their state, they are more likely to obtain policy outcomes that mirror the preferences of the state electorate. We should acknowledge that we consider only one type of representation here—ideological representation. It is possible that voters in states with lower levels of knowledge demand and receive different kinds of representation from elected officials. Furthermore, we expect that the substantive ideological representation that we focus on here is more likely to be conditioned on levels of state political knowledge than other forms of representation like descriptive representation or symbolic representation. We are also limited to using questions about two kinds of political knowledge, specific knowledge of the governor and knowledge of partisan control of government, as these are the best available data we could find. By finding similar results across different measures of state political knowledge, it provides reassurances that our results are not solely a function of the measures chosen. However, we cannot know with confidence that we would find the same patterns with measures of political knowledge rooted in other details of state governments.

A concern might be raised that two of our three measures of state political knowledge are capturing something about the traits of the governor, and that this is what drives the results we observe. For instance, perhaps long-serving governors are both better known and more adept at responding to constituent preferences. To explore this, we also considered whether state political knowledge still moderates the effects of public liberalism even when controlling for the direct or opinion-moderating effects of state attributes such as the length of the governor’s tenure in office or the amount of per capita campaign spending in the governor’s race in 2006. The effects we observe in our main models are robust to these controls.³³ Although such factors can contribute to the knowledge people hold, the influence of political knowledge on policy representation extends beyond their imprint.

Although our measures of state political knowledge are not perfect, we also have no strong reasons to think that our selected measures of state political knowledge are fundamentally different in nature than other measures of contemporary knowledge of state politics. Whatever the limits of our measures, using three measures of knowledge that are constructed in different ways and at different time points both inside and outside of a campaign season, we gain some confidence that our results are not an artifact of the particular idiosyncrasies of one type of measure of state political knowledge.³⁴

Conclusion

The connection between public preferences and citizen ideology is stronger in some states and weaker in others. Prior studies describe the heterogeneity in the responsiveness of state government as a function of institutional design or the choices made by political elites. In this article, we demonstrate that citizens also play an important role in securing representative outcomes from government. When state electorates are more knowledgeable about state politics, public preferences have a greater influence on state policy outcomes. Others show that political knowledge is a useful tool for individual voters when it comes to accomplishing the tasks of citizenship. We demonstrate here that collective political knowledge also promotes the performance of representative democracy.

Erikson et al. (1993) show that public preferences in the states drive public policy outcomes—a remarkable result given expectations of a public perceived as uninterested and uninformed about matters of state politics. We find here that the ideology–policy connection holds for almost all the states, where the effects of public liberalism fail to predict outcomes only for the handful of states that have particularly low levels of citizen knowledge of state politics. Importantly, as the knowledge of the state electorate increases, the effects of ideology on state policy outcomes moves from mild to very strong. The greater the share of the state electorate that is politically informed, the greater the level of ideological responsiveness of state government. When knowledge of state politics is low, uninformed electorates are in peril of getting unrepresentative outcomes.

Our findings highlight a way that policy representation in the states is limited, in that low levels of state political knowledge are associated with a disconnect between citizen ideology and policy outcomes. When voters are uninformed, they are less likely to receive representative outcomes. However, these results also provide some normative reassurance about how representative policy outcomes can be achieved through the collective knowledge of a state electorate. Even if an individual voter is not tuned into the business of state politics, he or she might still see representative policy outcomes as a result of the attentiveness of his or her fellow citizens of the state. In this way, political knowledge in the aggregate is a collective good with potential dividends for all in the state, where citizens can get ideological representation even if not personally informed about politics. However, as a collective good, it also implies that a person might see little incentive to become more informed about state politics, so long as others appear to be monitoring the business of government in the state capital.

By demonstrating the role that collective political knowledge plays in securing representative outcomes from government, we confirm that information is a central mechanism to political responsiveness. Prior research shows that political knowledge can affect which candidates are elected and which policies are passed—as fully informed voters think about policy issues in different ways and choose different candidates than the uninformed (Althaus, 2003; Bartels, 1996; Gilens, 2001; Sniderman et al., 1991). We demonstrate that levels of political knowledge not only alter outcomes but also promote representative policy outcomes. The greater the level of collective state knowledge, the greater the likelihood that state policy outcomes mirror the preferences of the state electorate at large. A politically knowledgeable person is argued to be a better citizen—better able to make informed political choices, more tolerant, more participatory, and more ideologically consistent (Delli Carpini & Keeter, 1996). We show that in the aggregate, a politically knowledgeable electorate also helps lead to normatively appealing outcomes—policies that match the public will. When state electorates are more politically knowledgeable, they are more likely to secure policy representation.

Footnotes

Acknowledgements

The authors thank John McIver and Julie Pacheco for their helpful comments on an earlier iteration of the article. They thank Justin Phillips and Jeffrey Lax for generously sharing their replication data on policy congruence in the states.

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

William P. Jaeger is the vice president of the Colorado Children’s Campaign, a research and policy organization. He completed his Master’s degree in Political Science at the University of Colorado at Boulder and also holds Master’s degrees in Education and in Public Policy.

Jeffrey Lyons is an assistant professor of political science at Boise State University. His research focuses on public opinion, political behavior, and state and local politics.

Jennifer Wolak is an associate professor of political science at the University of Colorado at Boulder. Her research interests include public opinion, political psychology, and state politics.

References

Althaus

S. L.

(2003). Collective preferences in democratic politics: Opinion surveys and the will of the people. New York, NY: Cambridge University Press.

Arceneaux

(2002). Direct democracy and the link between public opinion and state abortion policy. State Politics & Policy Quarterly, 2, 372-387.

Barrilleaux

(2006). Ideological cleavage, political competition, and policy making in the American states. In Cohen

J. E.

(Ed.), Public opinion in state politics (pp. 121-141). Palo Alto, CA: Stanford University Press.

Bartels

L. M.

(1996). Uninformed votes: Information effects in Presidential elections. American Journal of Political Science, 40, 194-230.

Bartels

L. M.

(2008). Unequal democracy: The political economy of the New Gilded Age. Princeton, NJ: Princeton University Press.

Berelson

(1952). Democratic theory and public opinion. Public Opinion Quarterly, 16, 313-330.

Berry

W. D.

Fording

R. C.

Ringquist

E. J.

Hanson

R. L.

Klarner

(2010). Measuring citizen and government ideology in the American states: A re-appraisal. State Politics and Policy Quarterly, 10, 117-135.

Berry

W. D.

Ringquist

E. J.

Fording

R. C.

Hanson

R. L.

(1998). Measuring citizen and government ideology in the American states, 1960-1993. American Journal of Political Science, 42, 327-348.

Brace

Sims-Butler

Arceneaux

Johnson

(2002). Public opinion in the American states: New perspectives using national survey data. American Journal of Political Science, 46(1), 173-189.

10.

Burden

B. C.

(2005). Institutions and policy representation in the states. State Politics and Policy Quarterly, 5, 373-393.

11.

Burstein

(2006). Why estimates of the impact of public opinion on public policy are too high: Empirical and theoretical implications. Social Forces, 84, 2273-2289.

12.

Buttice

M. K.

Highton

(2013). How does multilevel regression and poststratification perform with conventional national surveys? Political Analysis, 21, 449-467.

13.

Camobreco

J. F.

(1998). Preferences, fiscal policies, and the initiative process. Journal of Politics, 60, 819-829.

14.

Camobreco

J. F.

Barnello

M. A.

(2008). Democratic responsiveness and policy shock: The case of state abortion policy. State Politics & Policy Quarterly, 8, 48-65.

15.

Carsey

T. M.

Harden

J. J.

(2010). New measures of partisanship, ideology, and policy mood in the American states. State Politics and Policy Quarterly, 10, 136-156.

16.

Converse

P. E.

(1990). Popular representation and the distribution of information. In Ferejohn

J. A.

Kuklinski

J. H.

(Eds.), Information and democratic processes (pp. 369-388). Chicago: University of Illinois Press.

17.

Delli Carpini

M. X.

Keeter

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

18.

Enns

P. K.

Koch

(2013). Public opinion in the U.S. states: 1956 to 2010. State Politics and Policy Quarterly, 13, 349-372.

19.

Erikson

R. S.

MacKuen

M. B.

Stimson

J. A.

(2002). The macro polity. New York, NY: Cambridge University Press.

20.

Erikson

R. S.

Wright

G. C.

McIver

J. P.

(1993). Statehouse democracy: Public opinion and policy in the American states. New York, NY: Cambridge University Press.

21.

Flavin

(2012). Income inequality and policy representation in the American states. American Politics Research, 40, 29-59.

22.

Flavin

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

23.

Gelman

King

(1993). Why are American Presidential election campaign polls so variable when votes are so predictable? British Journal of Political Science, 23, 409-451.

24.

Gerber

E. R.

(1996). Legislative response to the threat of the popular initiative. American Journal of Political Science, 40, 99-128.

25.

Gilens

(2001). Political ignorance and collective policy preferences. American Political Science Review, 95, 379-396.

26.

Gilens

(2012). Affluence and influence: Economic inequality and political power in America. Princeton, NJ: Princeton University Press.

27.

Graber

(1984). Processing the news: How people tame the information tide. New York, NY: Longman.

28.

Gray

Lowery

Fellowes

McAtee

(2004). Public opinion, public policy, and organized interests in the American states. Political Research Quarterly, 57, 411-420.

29.

Haider-Markel

D. P.

Kaufman

M. S.

(2006). Public opinion and policymaking in the culture wars: Is there a connection between opinion and state policy on gay and lesbian issues? In Cohen

J. E.

(Ed.), Public opinion in state politics (pp. 163-182). Stanford, CA: Stanford University Press.

30.

Hill

K. Q.

Hinton-Andersson

(1995). Pathways of representation: A causal analysis of public opinion-policy linkages. American Journal of Political Science, 39, 924-935.

31.

Hochschild

J. L.

(2010). If democracies need informed voters, how can they thrive while expanding enfranchisement? Election Law Journal, 9, 111-123.

32.

Iyengar

(1990). Shortcuts to political knowledge: The role of selective attention and accessibility. In Ferejohn

J. A.

Kuklinski

J. H.

(Eds.), Information and democratic processes (pp. 160-185). Urbana: University of Illinois Press.

33.

Johnson

(2001). The impact of social diversity and racial attitudes on social welfare policy. State Politics and Policy Quarterly, 1, 27-49.

34.

Johnson

Brace

Arceneaux

(2005). Public opinion and dynamic representation in the American states: The case of environmental attitudes. Social Science Quarterly, 86, 87-108.

35.

Jones

B. S.

Norrander

(1996). The reliability of aggregated public opinion measures. American Journal of Political Science, 40, 295-309.

36.

Krosnick

J. A.

Milburn

M. A.

(1990). Psychological determinants of political opinionation. Social Cognition, 8, 49-72.

37.

Lascher

E. L.

Jr. Hagen

M. G.

Rochlin

S. A.

(1996). Gun behind the door? Ballot initiatives, state policies and public opinion. Journal of Politics, 58, 760-775.

38.

Lau

R. R.

Andersen

D. J.

Redlawsk

D. P.

(2008). An exploration of correct voting in recent U.S. presidential elections. American Journal of Political Science, 52, 395-411.

39.

Lax

J. R.

Phillips

J. H.

(2009a). Gay rights in the states: Public opinion and policy responsiveness. American Political Science Review, 103, 367-386.

40.

Lax

J. R.

Phillips

J. H.

(2009b). How should we estimate public opinion in the states? American Journal of Political Science, 53, 107-121.

41.

Lax

J. R.

Phillips

J. H.

(2012). The democratic deficit in the states. American Journal of Political Science, 56, 148-166.

42.

Lax

J. R.

Phillips

J. H.

(2013). How should we estimate sub-national opinion using MRP? Preliminary findings and recommendations (Working paper, April 10, 2013). Retrieved from http://www.columbia.edu/~jrl2124/mrp2.pdf

43.

Lodge

Steenbergen

M. R.

Brau

(1995). The responsive voter: Campaign information and the dynamics of candidate evaluation. American Political Science Review, 89, 309-326.

44.

Lupia

(1994). Shortcuts versus encyclopedias: Information and voting behavior in California insurance reform elections. American Political Science Review, 88, 63-76.

45.

Lupia

(2015). Uninformed: Why people seem to know so little about politics and what we can do about it. Oxford, UK: Oxford University Press.

46.

Lyons

Jaeger

W. P.

Wolak

(2013). The roots of citizens’ knowledge of state politics. State Politics and Policy Quarterly, 13, 183-202.

47.

Maestas

(2000). Professional legislatures and ambitious politicians: Policy responsiveness of state institutions. Legislative Studies Quarterly, 25, 663-690.

48.

Mann

T. E.

Wolfinger

R. E.

(1980). Candidates and parties in Congressional elections. American Political Science Review, 74, 617-632.

49.

Monogan

Gray

Lowery

(2009). Public opinion, organized interests, and policy congruence in initiative and non-initiative U.S. states. State Politics & Policy Quarterly, 9, 304-324.

50.

Mooney

C. Z.

Lee

(2000). The influence of values on consensus and contentious morality policy: US death penalty reform, 1956-82. Journal of Politics, 62, 223-239.

51.

Norrander

(2000). The multi-layered impact of public opinion on capital punishment implementation in the American states. Political Research Quarterly, 53, 771-794.

52.

Pacheco

(2011). Using national surveys to measure dynamic U.S. state public opinion: A guideline for scholars and an application. State Politics & Policy Quarterly, 11, 415-439.

53.

Pacheco

(2013). The thermostatic model of responsiveness in the American states. State Politics & Policy Quarterly, 13, 306-332.

54.

Page

Shapiro

(1994). The rational public. Chicago, IL: University of Chicago Press.

55.

Palus

C. K.

(2010). Responsiveness in American local governments. State and Local Government Review, 42, 133-150.

56.

Park

D. K.

Gelman

Bafumi

(2004). Bayesian multilevel estimation with poststratification: State-level estimates from national polls. Political Analysis, 12, 375-385.

57.

Park

D. K.

Gelman

Bafumi

(2006). State-level opinions from national surveys: Poststratification using multilevel logistic regression. In Cohen

J. E.

(Ed.), Public opinion in state politics (pp. 209-228). Palo Alto, CA: Stanford University Press.

58.

Popkin

S. L.

Dimock

M. A.

(1999). Political knowledge and citizen competence. In Elkin

S. L.

Soltan

K. E.

(Eds.), Citizen competence and democratic institutions (pp. 117-146). University Park: Pennsylvania State University Press.

59.

Powell

L. W.

(2012). The influence of campaign contributions in state legislatures: The effects of institutions and politics. Ann Arbor: University of Michigan Press.

60.

Rigby

Wright

G. C.

(2011). Whose statehouse democracy? Policy responsiveness to poor versus rich constituents in poor versus rich states. In Enns

Wlezien

(Eds.), Who gets represented? (pp. 189-222). New York, NY: Russell Sage.

61.

Schneider

S. K.

Jacoby

W. G.

(2006). Citizen influences on state policy priorities: The interplay of public opinion and interest groups. In Cohen

J. E.

(Ed.), Public opinion in state politics. Palo Alto, CA: Stanford University Press.

62.

Shaffer

W. R.

Weber

R. E.

(1974). Policy responsiveness in the American states. Beverly Hills, CA: Sage.

63.

Sniderman

P. M.

Glaser

J. M.

Griffin

(1991). Information and electoral choice. In Sniderman

P. M.

Brody

R. A.

Tetlock

P. E.

(Eds.), Reasoning and choice: Explorations in political psychology (pp. 164-178). New York, NY: Cambridge University Press.

64.

Sorens

Muedini

Ruger

W. P.

(2008). U.S. state and local public policies in 2006: A new database. State Politics & Policy Quarterly, 8, 309-326.

65.

Stimson

J. A.

(2004). Tides of consent: How public opinion shapes American politics. New York, NY: Cambridge University Press.

66.

Stimson

J. A.

MacKuen

M. B.

Erikson

R. S.

(1995). Dynamic representation. American Political Science Review, 89, 543-565.

67.

Sutton

R. L.

(1973). The states and the people: Measuring and accounting for “state representativeness.” Polity, 5, 451-476.

68.

Treadway

J. M.

(1985). Public policymaking in the states. New York, NY: Praeger.

69.

Verba

Schlozman

K. L.

Brady

H. E.

(1995). Voice and equality: Civic voluntarism in American politics. Cambridge, MA: Harvard University Press.

70.

Wlezien

(2004). Patterns of representation: Dynamics of public preferences and policy. Journal of Politics, 66, 1-24.

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

0.25 MB