While there are many theories about the predictors of vote choice, far less is known about the dynamic processes by which voters reach these choices. Studies of campaigns frequently assume choices develop through activation of partisan attachment. This notion conflicts with recent evidence that campaigns facilitate ideological or issue voting. The discrepancy persists because researchers rely on inadequate data for investigating how relationships between predispositions and individual-level dynamics of vote intention may themselves change over the course of a campaign. I develop a framework for identifying campaign activation that emphasizes the temporal dependence of these relationships. Applying this framework to a year-long eleven-wave panel survey from the 2008 election, I test for partisan activation against ideological activation. Results show that the campaign activates ideology rather than party identification, challenging the long-standing assumption of partisan activation.
“Campaigns, Political Sophistication, and the Dynamics of Vote Choice.”
It is well known that outcomes and processes of decision-making differ across levels of political sophistication. Far less is known about how this stratification interacts with presidential campaigns. Does the intense flow of information exacerbate or mitigate these differences? Previous research shows mixed results about whether voters at the top, the middle, or the bottom of this distribution are most responsive to campaign information. This is the result of insufficient data, such as repeated cross sections or short two-wave panels. With these data, researchers cannot directly observe individual learning and shifts in decision-making. I use an eleven-wave panel survey collected over the year leading up to the 2008 election to examine who learns and who changes the bases of their decision-making over the course of the campaign. I find that less sophisticated voters exhibit higher rates of learning over the course of a campaign. Given enough time, the flow of information in campaigns begins to close gaps in voter knowledge. Since general political sophistication is highly correlated with socio-economic status, campaigns help to close the gap of political advantage by especially informing the less sophisticated.
The most prominent theories of campaigns emphasizes their role in providing information that allows voters to connect their attitudes and vote preferences. Yet little is known about who acquires issue-specific information or when they do so. In this paper I consider how the salience of an issue in campaign discourse and the personal importance of this issue to a voter jointly produce heterogeneous reductions in the cost of acquiring information. Using rolling cross section and panel data from two recent presidential contests to examine respondent knowledge of candidates' Social Security positions, I demonstrate that learning among issue public members outpaces that of non-members when the campaign prominently features this issue. The learning difference vanishes when it does not. Furthermore, learning especially accrues to seniors at the middle-to-low stratum of general political knowledge, challenging the notion that highly sophisticated voters most reap the rewards of campaign information.
“Time of Vote Decision in the 2008 Presidential Election,” with D.S. Hillygus.
Time of decision is critical to understanding the potential impact of the campaign environment on voter decision making. If a respondent settles on a preferred candidate before the campaign begins, it is clear that the campaign has little direct impact on that decision. Yet, existing analyses do not explicitly test for campaign effects on the time of decision. In this paper, we examine the time of the vote decision, estimating a duration model to predict when respondents settled on a preferred candidate. While time of decision is dependent upon voter sophistication (with more sophisticated voters deciding earlier), the gap between more and less sophisticated voters disappears in battleground states.
“Citizen Perceptions of Government Service Quality: Evidence from Public Schools,” with M.M. Chingos and M.R. West.
Conventional models of democratic accountability hinge on citizens’ ability to evaluate government performance accurately, yet there is little evidence on the degree to which citizen perceptions of the quality of government services correspond to actual service quality. Using nationally representative survey data, we find that citizens’ perceptions of the quality of specific public schools reflect publicly available information about the level of student achievement in those schools. The relationship between actual and perceived school quality is two to three times stronger for parents of school-age children, who have the most contact with schools and arguably the strongest incentive to be informed. However, this relationship does not differ by homeowner status or by respondents’ race, ethnicity, income, or education. A regression discontinuity analysis of an oversample of Florida residents confirms that public accountability systems can have a causal effect on citizen perceptions of service quality.
Political scientists have long argued that the influence of self-interests pales in comparison to individuals’ long-standing political dispositions. However, the principal differences in public opinion on merit pay are defined by key stakeholders (in particular parents and teachers) and generally not by partisans or ideological groups. These two groups also differ with regard to their propensity to update their views when exposed to new information about the policy. Across two survey experiments, the differences in treatment effects between parents and teachers were consistently larger than those between Republicans and Democrats. None of these groups, however, reveal markedly different structures of opinion formation on education issues. How teachers, parents, Democrats, and Republicans think about education policies does not distinguish these groups nearly as much as what they think.
Does public information about school quality lead parents to sort their children out of schools with relatively poor performance? Use of this exit option in response to information about school quality has the potential to indirectly foster school responsiveness to quality concerns. To determine whether this information affects student exit, I use a regression discontinuity design to examine the effect of school grades on exit. Results indicate that parents do not seem to respond to information about school quality generally and, thus, cast doubt on the effectiveness of indirect accountability to promote educational improvement. However, there is limited evidence that particularly poor school performance accompanied by institutional mechanisms for school choice promote student sorting away from low-quality schools.