The Roots of Citizens’ Knowledge of State Politics

Abstract

Do citizens have the information they need to hold state politicians accountable? We consider what people know about state politics and whether knowledge of state government is rooted in the same factors that explain knowledge of national matters. We argue that while knowledge of national politics is rooted within individual dispositions like education and political interest, knowledge of state politics depends on the political climate of the state. When state political environments provide more information and greater incentives to become informed, people are more knowledgeable about state matters. Even if citizens are not always well versed in matters of state politics, they appear to monitor the business of the states. Citizens are most knowledgeable about state politics in the cases where information is arguably most important—when state governments are underperforming, when political competition is high, and when the political parties in the state are ideologically divided.

Keywords

political knowledge state contexts government accountability

How much do people know about their state government? By conventional wisdom, a public that tends to be uninformed about national politics is even less informed about matters of state government. If true, then people’s ability to monitor state officials and communicate their preferences on subnational issues will be limited. People might not have the information needed to hold state politicians accountable, nor possess sufficient knowledge to make the right choices on state ballot initiatives. We examine why some people are more knowledgeable about state politics than others, and the degree to which knowledge of state politics is rooted in the same kinds of factors that promote knowledge of national matters.

Past studies suggest that the roots of political knowledge are in people’s dispositions, where knowledge is highest among those with the greatest education and political interest. But if knowledge is based in the traits of citizens, then low levels of knowledge will likely be static, unaffected by the business of state government. We argue that people’s ability and incentives to become informed about state matters will depend in part on the political context of their state. Given a public that tends not to find state politics as interesting as government at other levels (Jennings and Zeigler 1970) as well as the significant undercoverage of state politics in the news (Graber 1993; Layton and Walton 1998), we expect that people’s knowledge of their state government will be influenced by the nature of the state’s political climate. In a state marked by weak economic performance or mismanaged state government, citizens arguably have stronger incentives to be informed about the business of the state as well as greater opportunities to learn about the state through the news. If people’s knowledge of state politics is responsive to political conditions like these, it suggests that even in the face of low knowledge of state matters, citizens could learn enough about state politics to be able to avoid costly mistakes—a threshold of reasoned choice described by Lupia and McCubbins (1998).

Using a 2007 Pew survey, we consider the roots of different kinds of knowledge about state government. We compare the individual-level roots of political knowledge about national politics versus state politics, to see whether citizens acquire information differently about each level of government. We also explore whether state-to-state differences in political contexts translate into variations in knowledge across states. Using multilevel modeling, we look beyond campaign effects to explore how state political environments influence the knowledge people hold about politics. We consider explanations rooted in government performance, the contentiousness of partisan politics, political competition, and the accessibility of news from the state capital. We find that knowledge of state politics is greatest in the environments where political information is arguably most important—when state economic performance is poor, when state government is poorly managed, and when party distance and partisan competition is greatest. People might not pay attention to state politics all of the time, but it appears citizens can successfully monitor state government—as state political knowledge is highest in the places where state governments are underperforming.

Do People Learn about Their State in Different Ways than They Learn about National Politics?

How knowledgeable are people about politics in their state? Roeder (1994, 34) suggests that in the public’s eyes, state governments are the “invisible layer of government.” When asked whether their state has a constitution, only 54% answered correctly. When asked to name any branch of state government, 44% could not even name one.¹ When asked whether they approved or disapproved of the performance of the state legislature, 25% provided no answer at all (Patterson, Ripley, and Quilan 1992). Compared with knowledge of national government, fewer people know the name of the governor than the name of the president, and are less likely to know about their state legislators as they are their member of Congress (Delli Carpini and Keeter 1996; Songer 1984).² But people’s knowledge of national politics is fairly low as well (Delli Carpini and Keeter 1996). When asked in 1952 to name the three branches of the national government, 52% of respondents failed to name a single branch.³ While 56% of respondents in the 1994 American National Election Study knew that the president nominates judges to federal courts, 51% of respondents in a 2009 survey knew whether the judges on their state’s supreme court are elected or not.⁴

This raises the question of just how distinctive knowledge of state politics is. Are those who are uninformed about national government similarly ignorant about government at other levels? Or do people acquire knowledge about politics in their state via routes that do not entirely mirror the pathways to learning about national politics? Based on past research, we might expect that those knowledgeable about state politics are the ones who are also knowledgeable about national politics. Political knowledge is in large part a function of one’s ability to understand politics combined with a motivation to follow political events (Delli Carpini and Keeter 1996; Luskin 1990; Smith 1989). Those who have higher levels of education and cognitive capability are more likely to comprehend the news they encounter, and those who find politics interesting put more effort into keeping up with politics. Because such factors are rooted within individuals, it suggests that those who are knowledgeable about one domain of politics will be knowledgeable about others.⁵ If knowledge is rooted in people’s ability to understand politics, then the roots of state knowledge should vary little from what predicts knowledge about national government.

We expect, however, that people’s knowledge of state politics has different roots than knowledge of national politics. While people’s ability to learn should promote knowledge of all levels of government, we expect that knowledge of state matters will be particularly responsive to the incentives and opportunities to learn about state politics. State politics fails to attract people’s interest in the same way as politics at other levels (Jennings and Zeigler 1970).⁶ As Dahl (1967) argues, the states are unimportant to citizens—not as proximate as local governments and not as important as the national government. Not only are people less likely to follow state politics than national politics but information about state events is also harder to find. In the case of national politics, the opportunities to learn are plentiful—if interested in doing so, one can easily follow national events by watching the nightly news, a favored cable program, or listening to the news on the radio. But in the case of state politics, it can be much harder to stay informed. National news rarely covers state politics (Graber 1993), and even local news is biased in favoring of covering national campaigns over subnational races (Kaplan, Goldstein, and Hale 2005). Regional newspapers often offer poor coverage of state events, given the low levels of interest of their audience as well as the challenges of covering news from the capital (Graber 1993). As newspapers cut the size of their state capital press corps in the face of diminishing resources, fewer stories emerge from the statehouse (Dorroh 2009; Layton and Walton 1998).

Because news about national politics is easily accessible from a range of news sources, the most informed will be the ones who are the most able and inclined to keep up with it. But in the case of state politics, keeping informed is more difficult. If indeed people are less interested in state politics than they are in national politics, and if news about state government is difficult to obtain relative to other kinds of news, then it might be unreasonable to expect citizens to be continuously informed about matters of state governance. The better standard might be that of the monitorial citizen, one not actively patrolling for political information but attentive to political circumstances that demand increased attention as signaled by the media or other groups (Schudson 1998; Zaller 2003). If we instead take a standard of burglar alarms over police patrols, then we should look at whether citizens are more likely to be informed in the political environments where their political attention is most necessary—such as when state governments are underperforming or when partisan differences divide the state capital.

We expect that the variance in knowledge of national politics is mainly across individuals, whether the interested and able are the most informed. In the case of state politics, we expect knowledge varies not simply across individuals but across state contexts as well, less driven by interest and more responsive to the kinds of state environments that incentivize learning about state matters. In this way, knowledge of state politics will reflect not only just who you are but also where you are and the nature of the political context you inhabit. If true, this runs contrary to studies of political knowledge that describe it as unidimensional, by suggesting that some are knowledgeable about particular domains of politics but not others (Iyengar 1990) and where some environments facilitate political learning and others do not.⁷ The implications of this are favorable for the quality of policy representation in the states. If knowledge of state politics is rooted in the same factors that explain knowledge of national politics, then it suggests that matters like the performance of state government matter little for citizen monitoring of state business. But if knowledge at each level has distinctive roots, then it suggests that knowledge of politics is not strictly a trait of individuals, but dependent on the circumstances and responsive to political conditions.

The Individual-Level Roots for Knowledge in State and National Politics

To consider the distinctiveness of knowledge of state politics, we examine the roots of knowledge about each level of government and whether the knowledge that people hold about national politics is rooted in the same factors that explain knowledge about state politics. We rely on a February 2007 survey conducted by Pew Research Center for the People and the Press. In the survey, 1,502 adults from 47 states and the District of Columbia were interviewed about their levels of political awareness in a battery of items about politics and current events.⁸ The timing of the survey allows us to assess levels of political knowledge outside of campaign seasons, and the variety of knowledge items gives us the opportunity to consider the distinctive bases of national political knowledge versus state political knowledge.⁹

We consider three measures of political knowledge. We explore the general roots of knowledge about national politics using a count of the number of correct responses to 20 questions about current events and politics, which are listed in the appendix. Nine items involve identification of various political figures, four items ask about American involvement in Iraq, and seven items concern current events such as whether Congress passed minimum wage legislation and the ideology of the Chief Justice.¹⁰ In this sample, the average number of correct responses is 11. We assess knowledge about state government using an item that asks respondents to identify the governor of their state, coded 1 for correct answers and 0 for incorrect or “don’t know” responses. For sake of comparison, we also consider a dichotomous measure of knowledge of national politics, based on correct identification of the vice president. In this sample, respondents are as likely to identify the vice president as they are the governor of their state. Sixty-nine percent correctly identify Dick Cheney, and 68% correctly identify the governor.¹¹

Among the individual-level roots of knowledge, we consider first the effects of one’s ability to comprehend the news on knowledge of each level. Education helps provide people with the means to interpret political events and understand the news, facilitating political learning (Delli Carpini and Keeter 1996). We expect that increasing years of schooling (measured as a seven-point scale of educational attainment) will be as important to explaining knowledge about state politics as it is in explaining knowledge of national politics. We also consider the effects of people’s motivation to attend to political matters, as those who find politics interesting invest more effort into following the news and derive more satisfaction from acquiring political knowledge (Luskin 1990). We expect that while people’s interest in following politics will be important for knowledge of both levels of government, it will be more important to explaining knowledge of national politics. Given the relative difficulty of obtaining information about state government, someone who actively follows current events will make substantial information gains over one who avoids the news when it comes to national politics but maybe only limited gains relative to the inattentive in the domain of state politics. A citizen who pays attention to current events everyday learns much about national politics that an inattentive citizen misses, while learning only a modest amount more about state matters than the citizen not tuned into political news. We test this using a measure of how much the respondent reports that he or she enjoys keeping up with the news—a lot, some, not much, or not at all as well as a measure of strength of partisanship (a folded five-point measure of party identification).

Media coverage explains political knowledge generally (Barabas and Jerit 2009) but is seen as particularly important to explaining public knowledge of state politics (Delli Carpini, Keeter, and Kennamer 1994). To inform whether knowledge of state affairs is conditioned by the accessibility of information, we consider the effects of using locally produced news relative to reliance on national news sources. We consider whether those who report reading a daily newspaper or say they watch local television news in the evening are more knowledgeable about state politics than those who do not.¹² We also make a scale of the number of national news sources that a respondent reports watching or listening to regularly, from a list that includes national nightly network news, CNN, Fox News, NewsHour with Jim Lehrer, morning national news, The O’Reilly Factor, The Colbert Report or The Daily Show, National Public Radio, Rush Limbaugh’s radio show, and news magazines. We expect that while general levels of news consumption will be important to explaining knowledge at each level, those who follow local television news and read newspapers will be more knowledgeable about state politics than those who solely rely on nationally produced news sources.

We control for the respondent’s gender, race, and age, as is common in models of political knowledge. Women tend to be less politically knowledgeable than men (Delli Carpini and Keeter 1996; 2000). Some studies suggest that the gender gap in political knowledge closes in the domains of subnational politics, as women may be more engaged in the issues tackled by governments in their communities (Verba, Burns, and Schlozman 1997). In the case of knowledge of national politics, nonwhites tend to be less informed than whites and young people tend to be less politically informed than older people, but it is not clear that these differences hold across all domains of knowledge (Delli Carpini and Keeter 1996). We test whether there are demographic differences in knowledge of state politics or if these factors matter more in national domains rather than subnational.

We also expect that levels of knowledge of state politics will be particularly responsive to the character of the political environment. We consider four types of state-level explanations that might explain state-to-state variation in knowledge of state government: the performance of state government, the contentiousness of state politics, the level of partisan competition, and the accessibility of information about news of the governor. We cannot be certain with this data about which mechanisms connect state contexts to citizen learning. It could be citizen driven, where people pay more attention when states are failing to perform well, but it could also be facilitated by elites who make news about state politics easier to obtain, such as when media coverage climbs when times are tough or when contentious politics makes partisan news prominent. We consider the correlations between knowledge and different state contexts as a first cut to consider how political environments promote learning.

First, we expect that poor performance of state government will be correlated with greater citizens’ knowledge. If citizens are not continually patrolling for state political news, but instead attuned to the kinds of events that demand their attention, then the failures of government should serve as just such a fire alarm. When state government underperforms, it could motivate citizen attention and increase the accessibility of state news, whether from political opponents of the administration in office, increased media coverage, or in discussions with others in the state. To test this, we first consider the effects of state economic performance, using a measure of the state unemployment rate during the survey period. We expect that poor economic times will be associated with greater knowledge of state politics. While some states are well managed with efficient and well-functioning bureaucracies, other states struggle to meet the demands of managing state responsibilities. To see whether poor management of state government prompts greater knowledge of state matters, we rely on scores from the Government Performance Project on the quality of state management practices (Barrett and Greene 2005). States that score high on this measure do well at matters like workforce planning, managing state information, and monitoring infrastructure and budgets. We also consider the fiscal performance of the state, by considering how well states do at managing their finances. We include a measure of state fiscal health, reflecting the difference between total state revenues and state spending, divided by total state revenues (Berry and Berry 1990). When states spend beyond their means, or fail to have enough funds to meet state expenditures due to struggling economic times, we expect that public scrutiny of state government will be higher. As a result, those living in states with poor fiscal health will be more knowledgeable about state politics than those living in states with excellent fiscal health.

When statehouses are divided by partisan battles, the business of state government becomes more newsworthy, and news coverage arguably is more likely. Partisan conflict also suggests higher stakes in the policy battle and perhaps also a greater need to monitor and attend to the political disagreements at hand. We investigate if knowledge is higher under divided government, coded 1 for cases where the state senate, statehouse, and governor’s office are not all controlled by the same political party. We also consider the effects of partisan division within statehouses, to see whether knowledge of state politics climbs as the political parties in the state legislature become more ideologically distinctive. As a measure of the amount of elite polarization in the states, we use the difference between the median legislative ideology of the Republican Party and Democratic Party in the state (Shor and McCarty 2010).¹³

Some states have high levels of party competition, where most elections are contested and candidates from either party have a viable chance of winning. In such contexts, political parties are likely to be more active and campaign efforts are likely to be more prevalent. For voters, there are stronger incentives to pay attention to campaigns as well as greater ability to encounter campaign information. As a result, we anticipate that citizens in states with a tradition of competitive and contested state elections will have higher rates of state-level political knowledge. To test this, we rely on a measure of electoral competition for state legislative seats (Holbrook and Van Dunk 1993). States that score high on this measure feature a high number of contested races and low margins separating the winning and losing candidates. Partisan divisions among the electorate could also result in higher levels of political knowledge. Discussing politics with people who hold opposing views can introduce people to new arguments and evidence, and the expectation of discussion with other knowledgeable individuals can motivate greater information gathering for use in political talk. When the partisan composition of a state electorate is heterogeneous and populated by even shares of Democrats and Republicans, then the opportunities to learn from partisan discussion are arguably greater. To assess this, we include a Herfindahl index of partisan concentration, where higher values indicate diverse partisan environments where there are equal numbers of self-identified Democrats, Republicans, and Independents, and low values indicate states where many share the same party identification.¹⁴

We also consider the consequences of the accessibility of news about the governor. We argued that the media likely play a particularly important role in educating (or failing to educate) citizens on matters of state government. We test this by considering how readily available information about the governor is at the time of the survey. If governors rarely make the news or if news from the capital is scarce in state newspapers, then we expect familiarity with the governor to be low. We assess this with a count of the number of news stories that mention the governor in the highest circulation paper in the state during the month of the survey.¹⁵ We also control for whether the governor is newly elected to the office, given that the survey is conducted in February 2007, only a month or two after the inauguration of a number of politicians elected in the 2006 elections.

The Individual-Level Roots of Knowledge of State Government versus National Government

We begin by considering the differences in the individual-level roots of political knowledge at the national and state level in Table 1, comparing the direction, size, and significance of individual-level factors in explaining knowledge at each level. We present two kinds of models. For the Level-1 model, we present a seemingly unrelated biprobit model with clustered standard errors, which allows us to compare the coefficients between the national and state identification models. For the Level-2 model, we use multilevel modeling and random-intercept logit models, as relying on clustered standard errors when Level-2 factors are included can lead to too-small standard errors for state-level variables. Our discussion of the effects of the individual-level roots of political knowledge focuses on the seemingly unrelated biprobit results, but we present multilevel logit models with individual-level predictors in Table 1 as well, so we can speak to how much additional variance is explained through the inclusion of the state-level factors in later models.¹⁶ Predicted effects from the Level-1 models are shown in Table 2.

Table 1.

Comparing the Roots of Knowledge in National and State Government.

	Multilevel models			Seemingly unrelated biprobit
	Correct answers, national knowledge	Correctly identifies vice president	Correctly identifies governor	Correctly identifies vice president	Correctly identifies governor	χ² (equality of coefficients)
Means
Education	0.070* (0.005)	0.317* (0.045)	0.251* (0.043)	0.185* (0.030)	0.145*(0.028)	1.06
Motive
Enjoys following politics	0.516* (0.035)	2.070* (0.262)	1.246* (0.255)	1.259* (0.168)	0.638* (0.165)	8.63*
Strength of partisanship	0.050* (0.026)	0.609* (0.208)	0.341† (0.205)	0.348* (0.147)	0.222* (0.114)	0.72
Opportunity
Reads a newspaper	0.020 (0.018)	0.201 (0.151)	0.362* (0.145)	0.125 (0.079)	0.202* (0.086)	0.45
Watches local television news	−0.007 (0.020)	−0.062 (0.171)	0.181 (0.161)	−0.059 (0.098)	0.082 (0.099)	1.62
Consumption of national news	0.014* (0.005)	0.023 (0.041)	−0.018 (0.039)	0.013 (0.021)	−0.013 (0.023)	0.77
Demographics
Female	−0.154* (0.017)	−0.503* (0.145)	−0.153 (0.138)	−0.286* (0.088)	−0.063 (0.092)	4.11*
Nonwhite	−0.121* (0.022)	−0.970* (0.168)	−0.591* (0.169)	−0.578* (0.093)	−0.284* (0.127)	2.89†
Age	0.003* (0.001)	−0.011* (0.004)	−0.001 (0.004)	−0.007* (0.002)	0.000 (0.003)	4.56*
Constant	1.613* (0.045)	−1.313* (0.348)	−1.387* (0.357)	−0.730* (0.252)	−0.798* (0.273)
Level-2 variance	0.000 (0.000)	0.000 (0.000)	0.484* (0.156)	—	—
Rho	—	—	—	0.410* (0.059)
n (Level 2)	1,263 (48)	1,303 (48)	1,302 (47)	1,302
Deviance	6,588	1,229	1,390	2,626
AIC	6,610	1,251	1,412	2,668

Note: AIC = Akaike information criterion. First column, multilevel Poisson estimates. Second and third columns, multilevel logit estimates. Other columns, seemingly unrelated bivariate probit with clustered standard errors. Standard errors in parentheses.

p < .05. †p < .10.

Table 2.

Predicted Probability of Correct Identification of Officials at State and National Level.

	Multilevel models			Seemingly unrelated biprobit
	Predicted correct, national knowledge	Pr(naming vice president)	Pr(naming governor)	Pr(naming vice president)	Pr(naming governor)
Lowest level of education	8.84	0.53	0.52	0.54	0.52
Highest level of education	13.44	0.88	0.83	0.89	0.82
Does not enjoy keeping up with the news	7.82	0.45	0.52	0.44	0.54
Enjoys keeping up with the news	13.11	0.86	0.79	0.87	0.77
Independent	10.99	0.69	0.67	0.69	0.65
Strong partisan	11.55	0.81	0.74	0.80	0.73
Does not regularly read a newspaper	11.29	0.77	0.69	0.76	0.67
Regularly reads a daily newspaper	11.52	0.80	0.76	0.79	0.74
Does not regularly watch local television news	11.48	0.79	0.70	0.79	0.70
Regularly watches local television news	11.40	0.78	0.74	0.78	0.72
Lowest level of use of national news	10.97	0.77	0.74	0.77	0.73
Highest level of use of national news	12.67	0.81	0.70	0.81	0.68
Male	12.35	0.83	0.75	0.82	0.73
Female	10.59	0.74	0.72	0.74	0.70
White	11.71	0.82	0.75	0.81	0.74
Nonwhite	10.38	0.63	0.63	0.62	0.63
Age 18	10.47	0.84	0.74	0.84	0.71
Age 92	12.79	0.70	0.72	0.68	0.72

Note: Significant differences in boldface. The predicted counts of correct answers and predicted probabilities of a correct response were calculated using the coefficients from Table 1. The change in probabilities reflects a movement from the minimum value of the independent variable to the maximum, holding all other variables at their means.

We find that people’s ability to learn about politics, as modeled by education, is as important to learning about state politics as it is national politics. Moving from the lowest level of education to the highest increases the likelihood of identifying the vice president by 35%, and increases the likelihood of correctly identifying the governor by 30%. Being motivated to follow politics, however, is more important to explaining variations in knowledge of national matters than it is to understanding why some are more knowledgeable of state politics, as indicated by the significant test statistic associated with testing the equality of these coefficients across models. For a person who really enjoys keeping up with the news, the probability of knowing the vice president is 87%, 43 points higher than the likelihood of correct identification for one who has no interest in following the news. The increase in the case of state knowledge is lower but still substantial, such that one who enjoys following the news has a 77% probability of correctly identifying the governor, 23 points higher than one who has no interest in following the news. Partisans are more likely to correctly identify the vice president and their governor than independents are.

The contributions of media choices to knowledge are modest compared with the effects of means and motive. Newspaper readers have a 74% probability of knowing the governor of their state, whereas those who do not read a daily newspaper have a 67% likelihood of correctly identifying the governor. We find no effect of newspaper reading on knowledge of the vice president. Although television watching has sometimes been found to be negatively correlated with levels of political information (Delli Carpini and Keeter 1996) and is negatively correlated with national knowledge here, we find that local television news consumption is positively associated with knowledge of the governor. However, the effect is not statistically significant. It suggests that people’s media choices lead to differential knowledge about state government versus national government, but the evidence is not conclusive.

We find a number of demographic differences in knowledge about state matters versus national matters. Women are less knowledgeable about national politics than men, where women have a predicted likelihood of correctly identifying the vice president nearly 10 points lower than found for men. However, we find no evidence of a gender gap in knowledge of state politics. Women are as likely to correctly identify the governor of their state as men are. In the case of race, nonwhites are less likely to correctly identify the governor and vice president than white respondents.¹⁷ While increasing age is associated with greater knowledge of national politics overall, it is negatively associated with knowledge of the vice president and unrelated to knowledge of the governor. For all three demographic variables, we can reject the equality of coefficients across the state and national knowledge equations—demographic factors have different effects on knowledge of politics at each level. This could be an artifact of the sample, but it suggests that knowledge in the domain of state politics has roots that are distinctive from what explains knowledge in other domains.

Overall, the roots of people’s knowledge of state politics are similar to the factors that explain people’s knowledge of national politics—those with greater ability and desire to follow politics are more informed about government at each level. However, knowledge of state politics is distinctive from national knowledge in that it is less responsive to people’s motivation to pursue the news and less reflective of demographic differences.¹⁸ What is also distinctive about knowledge of state politics is that it is partly a function of the state in which one resides. With multilevel modeling, we can partition the variance in knowledge between Level 1 (individual factors) and Level 2 (state factors). For state contexts to be relevant to explaining levels of knowledge, some of the variation in political information levels should be attributable to the state level. As shown in the multilevel modeling results in Table 1, in the case of national knowledge, we find no significant variation in knowledge attributable to state contexts. The intraclass correlation is less than .01, meaning that not even 1% of the variance in knowledge of the vice president is due to state environments. But in the case of knowledge of the governor, we do find significant variance associated with the state level. The intraclass correlation for the multilevel model of gubernatorial knowledge in Table 1 is .13, meaning that controlling for individual-level explanations, 13% of the variance in knowledge of state politics is due to state-to-state variations.¹⁹ While knowledge about national politics is essentially a product of individual experiences rather than the political environment, what people know about state politics depends in part on the state in which they live.

The State-Level Roots of Knowledge of State Politics

In Table 3, we add state-level explanations to our multilevel logit model of correct identification of the governor. We also present a multilevel model of knowledge of the vice president for sake of comparison, to show that national knowledge is not sensitive to variations in state contexts.²⁰ We find that people’s knowledge of state politics is highest in the situations when monitoring is most important—when state government is performing suboptimally. As seen in Figure 1, when unemployment is at its highest level, state residents have an 82% probability of correctly identifying the governor. When unemployment is at its lowest level, the probability of correct identification is significantly lower—64%. Governments that are mismanaged are associated with higher levels of public scrutiny as well. For the state that gets the lowest score on government management, the probability of state residents identifying the governor is 83%, compared with a 62% probability for those in the state with the best managed state government. The state of the state’s budget is related to knowledge, where those living in states with the highest level of deficit spending have an 84% probability of being able to name the governor, 23 points higher than what is found for those living in the state with the best fiscal health.

Table 3.

The Individual and Contextual Roots of State Political Knowledge.

			Pr(identify governor) when x is at
	Correctly identifies vice president	Correctly identifies governor	Minimum	Maximum
Performance of state government
State unemployment rate	−0.071 (0.097)	0.224† (0.125)	0.64	0.82
Quality of government management	0.100† (0.053)	−0.176* (0.076)	0.83	0.62
State fiscal health	−1.272 (1.962)	−5.568* (2.505)	0.84	0.61
Contentious politics in the states
Party polarization of state legislature	0.225 (0.167)	0.520* (0.240)	0.64	0.84
Divided government	−0.268 (0.169)	−0.097 (0.217)	0.75	0.73
Party competition in the states
Electoral competition	0.007 (0.008)	0.023* (0.010)	0.59	0.81
Partisan diversity of the electorate	−10.285 (6.509)	−11.383 (8.020)	0.86	0.71
Attributes of the governor
Media coverage of the governor	0.001 (0.002)	0.005* (0.003)	0.65	0.86
Newly elected governor	−0.188 (0.191)	−0.583* (0.258)	0.76	0.64
Means
Education	0.332* (0.047)	0.248* (0.044)	0.53	0.83
Motive
Enjoys following politics	2.066* (0.271)	1.220* (0.258)	0.53	0.79
Strength of partisanship	0.642* (0.213)	0.400* (0.206)	0.67	0.75
Opportunity
Reads a newspaper	0.200 (0.156)	0.371* (0.147)	0.69	0.76
Watches local television news	−0.065 (0.175)	0.229 (0.162)	0.70	0.75
Consumption of national news	0.015 (0.042)	−0.033 (0.040)	0.75	0.69
Demographics
Female	−0.518* (0.149)	−0.219 (0.140)	0.76	0.71
Nonwhite	−0.959* (0.175)	−0.586* (0.172)	0.76	0.64
Age	−0.011* (0.004)	−0.002 (0.004)	0.75	0.72
Constant	4.630 (4.184)	5.091 (5.129)
Level-2 variance	0.000 (0.000)	0.180* (0.096)
n (Level 2)	1,258 (45)	1,258 (45)
Block F test of Level-2 variables	8.37	25.94*
Deviance	1,173	1,327
AIC	1,213	1,367

Note: AIC = Akaike information criterion. Multilevel logit estimates. Standard errors in parentheses. Significant differences in boldface. The third and fourth columns represent the predicted probability of correctly identifying the governor when the independent variable is at its minimum and when it is at its maximum, holding all other variables at their means.

p < .05. †p < .10.

Figure 1.

Effects of state-level factors on knowledge of the governor.

While people are especially likely to be knowledgeable about state politics when things are going poorly in the state, political competition in the state is associated with higher knowledge of state politics. While people are no more knowledgeable in states with divided government than in states with unified government, the ideological distance between the political parties is associated with state-level political knowledge. In the state where the political parties in the state legislature are most ideologically distinctive, residents have an 84% likelihood of giving a correct answer. When the parties are most ideologically similar, the probability of a correct answer is 64%. Electoral competition in the states is associated with greater knowledge of state politics as well, where moving from the lowest level of competition to the highest increases the likelihood of correctly answering the state politics knowledge question from 59% to 81%. However, the diversity of partisan preferences in the electorate is unrelated to knowledge of state politics. Overall, we find that people are more likely to be informed about state politics in states where the interparty electoral competition is high and the ideological distance between the parties is greatest.

Finally, knowledge of the governor is associated with the accessibility of news about the gubernatorial office. For a resident of the state where the governor is most often mentioned in the news, he or she has an 86% probability of naming the governor, compared with a probability of correct identification of 65% for one in a state where the governor is scarcely mentioned in the press, all else equal.²¹ This suggests that the newsworthiness of governors and the quality of newspapers’ monitoring of state events can influence citizens’ knowledge of state matters.

Adding state-level predictors increases the explanatory power of the model of gubernatorial knowledge. While we explain 14% of the variance in knowledge of the governor in the model with only individual-level explanations, we explain 21% of the variance in knowledge with the model that adds state explanations. Of the remaining unexplained variance, most is at the individual level (74%) rather than the state level (4%). While a block F test of the Level-2 variables shows that state factors explain a significant amount of variance in knowledge of the governor, they do not explain significant variation in knowledge of the vice president.²² Knowledge of national politics rests in people’s ability to comprehend political news and motivation to pursue it, while knowledge of state politics is rooted within individuals as well as the nature of the political contexts people inhabit.

Conclusion

Those with greater motivation to follow politics and ability to comprehend political news are more informed about government at both the state and national level, but we find that knowledge of state politics is less dependent on individual traits like political interest, gender, race, and age. This means that while knowledge of different levels of government shares common roots, being informed about national politics does not guarantee that an individual will be knowledgeable about state matters. Unlike the case of national political knowledge, we find heterogeneity in knowledge of state politics across individuals as well as heterogeneity across states, where some environments are more likely to promote citizens’ knowledge of state politics than others.

This means that people’s knowledge about state politics rests in part on the state in which they live; it is not exclusively a function of individual traits. State political knowledge is not static and constant—it is responsive to the political debates and partisan character of state government. Given that overall levels of political knowledge in the electorate have increased little over the years despite gains in education and information availability (Delli Carpini and Keeter 1991), it is tempting to think that the variance in political knowledge is mostly due to individual differences. Our findings, however, suggest that knowledge is not entirely trait based but strategically responsive to the political environment. Just as people’s knowledge of party control of Congress depends on the amount of party division and turnover in government (Bennett and Bennett 1993) and political events like September 11, 2001, promote greater learning about national politics (Prior 2002), we find that knowledge of state politics varies across state contexts, as a function of political events and political performance. People might not pay attention to politics all of the time, but when circumstances demand greater scrutiny, we find higher levels of knowledge about state politics.

The democratic demands on citizens vary from state to state. We find that the state environments most correlated with high knowledge of state politics are arguably the contexts where people have the greatest need to be informed—when state economic performance is weak, when management of state government is poor, when competition is high, when the distance between the two political parties is greatest, and when newspapers actively cover news from the governor’s office. With only a single question about state political knowledge in a single survey, we cannot know the extent to which people’s knowledge of state affairs is continually responsive to state climates, and more work is needed to know how robust these findings are to other times and other measures of state political knowledge. In this survey, however, knowledge of state politics is highest in the state environments where such information is arguably the most valuable.

Given lower levels of citizens’ interest in state politics as well as the challenges of finding news from the state capital in local news coverage, we might worry about citizens’ ability to stay informed about matters of state politics. Because knowledge of state politics is responsive to state-level influences, it suggests that even in the face of low information, monitorial citizens can have what they need to hold state officials accountable for their performance in office. People may not always follow the business of state politics but they appear to pay particular attention in instances where the state government is not performing up to expectations. In a scarce information environment like what is found for state politics, we find that people are able to become more knowledgeable under conditions where a lack of information could lead to costly mistakes.

Footnotes

Appendix

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Jeffrey Lyons is a PhD candidate in the Department of Political Science at the University of Colorado, Boulder. His research focuses on American political behavior as well as state and local politics.

William P. Jaeger is a PhD candidate in the Department of Political Science at the University of Colorado, Boulder. His research focuses on American political behavior and campaigns and elections.

Jennifer Wolak is an associate professor of political science at the University of Colorado.

References

Barabas

Jason

Jerit

Jennifer

. 2009. “Estimating the Causal Effects of Media Coverage on Policy-Specific Knowledge.” American Journal of Political Science 53(1): 73–89.

Barrett

Katherine

Greene

Richard

. 2005. “Grading the States ‘05: A Management Report Card.” Governing 18: 24–95.

Bennett

Stephen E.

Bennett

Linda L. M.

1993. “Out of Sight, Out of Mind: Americans’ Knowledge of Party Control of the House of Representatives, 1960–1984.” Political Research Quarterly 46(1): 67–80.

Berry

Frances Stokes

Berry

William D.

1990. “State Lottery Adoptions As Policy Innovations: An Event History Analysis.” American Political Science Review 84: 395–415.

Dahl

Robert A

. 1967. “The City in the Future of Democracy.” American Political Science Review 61(4): 953–70.

Delli Carpini

Michael X.

Keeter

Scott

. 1991. “Stability and Change in the US Public’s Knowledge of Politics.” Public Opinion Quarterly 55: 583–612.

Delli Carpini

Michael X.

Keeter

Scott

. 1996. What Americans Know about Politics and Why It Matters. New Haven, CT: Yale University Press.

Delli Carpini

Michael X.

Keeter

Scott

. 2000. “Gender and Political Knowledge.” In Gender and American Politics: Women, Men, and the Political Process, eds. Tolleson-Rinehart

Josephson

J. J.

Armonk, NY: M.E. Sharpe.

Delli Carpini

Michael X.

Keeter

Scott

Kennamer

J. David

. 1994. “Effects of the News Media Environment on Citizen Knowledge of State Politics and Government.” Journalism Quarterly 71(2): 443–56.

10.

Dorroh

Jennifer

. 2009. “Statehouse Exodus.” American Journalism Review 31: 20–35.

11.

Erikson

Robert S.

Wright

Gerald C.

McIver

John P.

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

12.

Graber

Doris A.

1993. “Swiss Cheese Journalism.” State Government News 36: 19–21.

13.

Holbrook

Thomas M.

Van Dunk

Emily

. 1993. “Electoral Competition in the American States.” American Political Science Review 87: 955–62.

14.

Hox

Joop

. 2010. Multilevel Analysis: Techniques and Applications. New York: Routledge.

15.

Iyengar

Shanto

. 1990. “Shortcuts to Political Knowledge: The Role of Selective Attention and Accessibility.” In Information and Democratic Processes, eds. Ferejohn

J. A.

Kuklinski

J. H.

Urbana: University of Illinois Press, 160–85.

16.

Jennings

M. Kent

Zeigler

Harmon

. 1970. “The Salience of American State Politics.” American Political Science Review 64: 523–35.

17.

Kaplan

Martin

Goldstein

Ken

Hale

Matthew

. 2005. “Local News Coverage of the 2004 Campaigns: An Analysis of Nightly Broadcasts in 11 Markets.” Report of the Lear Center Local News Archive, USC Annenberg School, Los Angeles and University of Wisconsin, Madison.

18.

Layton

Charles

Walton

Mary

. 1998. “Missing the Story at the Statehouse.” American Journalism Review 20: 1–38.

19.

Lupia

Arthur

McCubbins

Mathew D.

1998. The Democratic Dilemma: Can Citizens Learn What They Need to Know. New York: Cambridge University Press.

20.

Luskin

Robert C.

1990. “Explaining Political Sophistication.” Political Behavior 12(4): 331–61.

21.

Patterson

Samuel C.

Ripley

Randall B.

Quinlan

Stephen V.

1992. “Citizens’ Orientations toward Legislatures: Congress and the State Legislature.” Western Political Quarterly 45(2): 315–38.

22.

Prior

Markus

. 2002. “Political Knowledge after September 11.” PS: Political Science & Politics 35: 523–30.

23.

Roeder

Phillip W.

1994. Public Opinion and Policy Leadership in the American States. Tuscaloosa, AL: The University of Alabama Press.

24.

Schudson

Michael

. 1998. The Good Citizen: A History of American Civic Life. New York: Free Press.

25.

Shor

Boris

McCarty

Nolan

. 2010. “The Ideological Mapping of American Legislatures.” Working paper. Retrieved from http://ideas.repec.org/p/ecl/prirpe/8-11-2010.html

26.

Smith

Eric R. A. N.

1989. The Unchanging American Voter. Berkeley: University of California Press.

27.

Songer

Donald R.

1984. “Government Closest to the People: Constituent Knowledge in State and National Politics.” Polity 17(2): 387–95.

28.

Verba

Sidney

Burns

Nancy

Schlozman

Kay Lehman

. 1997. “Knowing and Caring about Politics: Gender and Political Engagement.” Journal of Politics 59: 1051–72.

29.

Zaller

John

. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.

30.

Zaller

John

. 2003. “A New Standard of News Quality: Burglar Alarms for the Monitorial Citizen.” Political Communication 20(2): 109–30.