Joshua Robison

Home » Research


My research combines disparate literatures from political science, psychology, and communication studies to understand political behavior, public opinion formation, and how citizens evaluate political representatives. My research focuses particularly on the concept of motivation. On the one hand, one goal of my research is to understand why citizens differ in their desire to pay attention to politics, vote, and otherwise enter the political fray. Here I focus on concepts such as political interest and civic duty and trace how aspects of citizens’ political and social environments interact with their predispositions to structure these motivations to engage with politics. On the other hand, I am deeply interested in how citizens understand the motives of others, and particularly political elites, and the subsequent effect beliefs on this front have on evaluations and preferences. Here I focus on questions such as: how does the news media frame the motives of elites and what influence do such messages have on political interest and trust, when and how do political elites frame perceptions of their trustworthiness through their own communications, and how do beliefs about the motives of elites interact with partisan reasoning processes to influence persuasion efforts?

In the sections below I detail some of the studies I have conducted to advance these broader interests.

Motivating Interest and Engagement

Perhaps the most important predictor of the degree to which citizens engage with politics is their level of interest in politics. Political interest denotes how motivated citizens are to pay attention to politics and it has long been tied to a higher likelihood of voting, volunteering, political discussion, and political knowledge. Unfortunately, political scientists lack knowledge concerning the origins of political interest.  I do so in two articles. First, I focus on how personal values and personal identity influence interest and engagement with politics. Values are consistently shown to affect political preferences, but have never been explored as drivers of interest . I show that the influence of value priorities on interest is substantially greater than that of commonly used demographic variables. In fact, I find that the explanatory power of values rivals party identification. In my second study, which is forthcoming at Political Studies, I explore the role that social motivations play in driving political interest. I show that highlighting the potential social gains of engagement—for example, respect, solidarity, and larger social networks—can dramatically change a person’s willingness to engage with politics. Importantly, I show that identifying a motive for engagement, in this case a social motive, can actually counteract the effects of low efficacy, which would otherwise detract from engagement.

Mass Communication and Democratic Citizenship

Political interest and political trust are two key dimensions of democratic citizenship. Citizens’ interest in politics fundamentally structures the extent of their engagement with public affairs. Meanwhile, political trust influences perceptions of governmental legitimacy, whether citizens abide by the law, and preferences concerning public policies that may be socially desirable but are individually costly. Ultimately, democracies rely on some degree of political interest and trust among the mass public to operate.

In a study currently under review, I explore the influence of horse-race media coverage and elite polarization on interest, trust, and learning. I also explore how media choice affects these relationships. I show that horse-race coverage—which frames politics as a zero-sum game played by strategically oriented players—leads to a more interested but less trusting populace even when participants were given a choice over what they read and could have avoided this content. Framing political issues through the lens of elite polarization, on the other hand, has a more consistently negative effect.

In ongoing research, I am exploring how different frames of the causes of elite polarization affect interest and trust. For instance, in a forthcoming study at Political Communication, my co-author (Kevin Mullinix) and I investigate how the news media frames elite partisan polarization. Existing studies of how the news frames the motives of elites suggest that these descriptions focus primarily on strategic motives such as the need to win voters and appeal to special interests. However, in a novel content analysis of news coverage of elite polarization, we show that news coverage of elite partisan polarization deviates from this pattern. Rather than frame polarization as a phenomenon rooted in the self-interest of elites, journalists instead frame it as driven by deep differences in underlying values and ideological beliefs. In a survey-experiment reported in the study, we show that this choice of causal frame has deep implications, with political Independents reacting particularly negatively, i.e. becoming much less interested in politics and trusting of government institutions, when they believe polarization is caused by strategic rather than values-based motives. This study is the first to consider how citizens think about the causes of this important trend in American politics.

I continue to explore the inter-relationship between communication and perceptions of trustworthiness in a recently completed study where I focus on the efforts of political candidates and elected officials to shape beliefs about their own trustworthiness. Here the focus is on a particular context—cases when politicians change their minds or ‘flip-flop’ on an issue. Existing work suggests policy switches will lead to negative evaluations, but this existing work fails to take into account the ‘policy dialogue’ surrounding the switch, which includes attempts by policy switchers to explain why they changed positions. Using two survey experiments, including one conducted on a large, nationally representative US sample, I show that such explanations mitigate the negative costs of repositioning and do so by positively influencing perceptions of the elite’s motives, thereby challenging the conclusions of past studies that repositioning is necessarily a gateway to electoral harm.

Additional Projects

What’s the Matter with “Following Politics”? 

As part of my dissertation research, I also explore how we measure political interest. In the course of my research, I found a previously un-explored pattern in survey results: the American National Election Studies (ANES) has consistently delivered drastically lower estimates of mass interest than have alternative studies utilizing the very same measure. Based on this discovery, the ANES conducted a survey experiment on my behalf to explore the issue in greater depth. The core finding is that the interest measure that the ANES has used for over forty years, and which various researchers have relied on, is subject to previously unknown biases.

This study has been published by The International Journal of Public Opinion Research and can be found here

Question Format and Political Knowledge

The implications of seemingly benign survey measurement issues also animate a second project that I have worked on. In this project, I explore the implications of different ways of measuring political knowledge. Political knowledge is one of the central variables of political behavior and public opinion research, helping to explain persuasion, participation, and attitudes. As I demonstrate in this paper, however, these conclusions are intimately connected with the format in which knowledge questions are asked; simply changing from an open-ended format to a multiple-choice substantially effects estimates of how knowledgeable the mass public is, who is knowledgeable, and the connection between knowledge and participation, racial stereotype holding, and evaluations of President Obama. These results suggest that previous research on political knowledge may be both under-estimating mass knowledge and over-estimating the influence of knowledge as an independent variable.

This study has also been published by The International Journal of Public Opinion Research and can be found here.

Political Interest and Partisan Selective Exposure

A considerable amount of research documents an ideological or partisan bias in media exposure: liberals and Democrats are more likely to be exposed to liberal-leaning media while conservatives and Republicans are more likely to be exposed to conservative-leaning media. These “unconditional” models, conducted in the mid-2000’s find a sizable pattern of selective exposure, but almost no research investigates the scope conditions of such selectivity. Using Pew Research Center data from 1996 to 2012, my co-author (Thomas Leeper) and I document that exposure to ideological or partisan media is heavily conditioned by time, audience size, and individuals’ interest in national politics. Selective exposure seems to be limited to extreme ideologues viewing a handful of specific television programs. This research continues my substantive focus on the relationship between political motivation and elite communication in affecting mass-level outcomes.

The Causes and Consequences of Issue Public Membership

While the focus of my dissertation is on a citizen’s general interest in politics, I am currently engaged in an ongoing project with James Druckman and Thomas Leeper that explores the causes and consequences of citizens’ issue-specific engagement.  Even generally disinterested citizens may indicate that a particular issue is of great importance to them, with potentially significant consequences for political evaluations and engagement. While issue importance has long been discussed, empirical attention to its role in political evaluations has been limited. In this on-going project we use previously unexplored data from the 2008 American National Election Studies Panel Survey to delve into the role that issue importance plays in structuring vote choice as well as the social origins of issue involvement. In an article currently under review, we show that individuals in politically diverse social networks do not report weaker attitudes (i.e. attitudes characterized by lower levels of subjective importance or extremity) contrary to expectations generated by a prior study on the subject. Meanwhile, in an article currently under construction, we find little evidence for the common claim that proximity voting is positively influenced by issue importance (Krosnick, 1988). Instead, we find evidence of proximity voting across a variety of issues regardless of how subjectively important the issue is to the voter in question. This deviation from prior studies appears to stem from changes in the political environment, most notably the growth of elite polarization. Highly polarized parties enable individuals who do not care about a particular issue to nevertheless correctly place the parties and thus understand the stakes of the contest without the added issue-specific knowledge importance facilitates. Both studies contribute to broader questions concerning preference formation and the nature of voting and also build on my research interests concerning the influence of social context on political engagement and public opinion.

The Relationship between Social Class and Partisan Preferences

As a project member of the Perceptions of Social Stratification and Voting Project at Aarhus, have co-designed a large mail survey of Danish citizens with Rune Stubager to examine the relationship between perceptions of social class and partisan evaluations. While social class cleavages have long played a crucial role in explaining party politics in Western industrialized nations, their continued influence has been called into question. However, we challenge this perspective with new data and show that Danish citizens—individuals one might expect to be particularly unlikely to still use class as a political heuristic—in fact possess clear stereotypes of social classes that they use to make sense of politics. Individuals use their class stereotypes, we show, to understand the classes they agree with on politics and, by extension, which parties they should support. We will be designing and fielding another national survey this Fall to further examine this relationship with a particular aim of untangling the causal relationship between social class and vote choices via the use of survey experiments. This project aligns with my deeper interests in understanding the relationship between political motivation and communication in affecting citizen engagement and preferences.

Public Opinion Among College Athletes (w James Druckman, Samara Klar, and Mauro Gilli)

Finally, along with James Druckman, Samara Klar, and Mauro Gilli, I have published three articles exploring the attitudes and behaviors of college student-athletes concerning a variety of policy issues, including recreational and performance enhancing drug use, Title IX, and the payment and unionization of athletes. Each topic has generated a great deal of public attention, but the views of the athletes who are most intimately affected by them have not been explored. This is unfortunate as the nature of student athlete attitudes and behaviors may greatly influence the types of policy response that are both appropriate and likely to occur. Our three studies speak both to fundamental questions in the literature on preference formation as well as to the broader public policy debates centering on these topics.

The first article concerns the use of banned performance enhancing drugs and binge drinking by student athletes. We use an innovative method (a list-experiment) to estimate drug use and find that a substantial number of student-athletes have knowingly taken performance enhancing drugs, many more than report doing so using a more traditional self-report. This paper has been published in Social Science Quarterly and can be found here.

A second paper concerns unionization and the payment of athletes; a contentious issue that has received a great deal of public attention recently. This article focuses on the role of social context in shaping these preferences. We find that athlete opinion shifts dramatically toward favoring unionization when student-athletes finish their collegiate careers due to the change in their social environment from one dominated by college administrators and coaches who are opposed to unionization to a more diverse environment. This study has recently been published in PLOS One and can be found here.

In a final study, we explore knowledge of and opinion toward Title IX among student-athletes. Our study is the first such exploration of student athlete opinion on the issue. We find that substantial portions of those affected by this policy do not fully understand the boundaries of the law, which consequently has implications for its potential application and revision. This article has been published at The Sports Journal and can be found here.