PRESS Experimental Research Workshop

Date & Time Dec 12 2017 4:30 PM - 6:00 PM
Speaker(s)
Brandon de la Cuesta and Chaya Crowder
Audience Restricted to Princeton University

What Can the State Do? Evaluating the role of voter priors on state capacity with a two-stage conjoint experiment

Brandon de la Cuesta, PhD Candidate, Politics

A major explanation for the low accountability observed in developing countries has been that citizens lack the necessary information to make inferences over candidate quality. Yet, in recent years, many studies have found that providing information on incumbent performance has little effect on voting behavior or other forms of accountability. I argue that this is due to the effect of citizens' priors on state capacity, which mediates---and in some cases severely diminishes---the value of information about incumbent performance. Where the state is strong, promises made by candidates in pre-election campaigns are feasible a priori. As a result, political competition revolves around the substance of those promises, rather than the probability with which they are fulfilled. Where the state is weak, however, it is more difficult for voters to use post-election outcomes as a basis for inference over incumbent quality. If an incumbent fails to deliver on his promises, is it because the state lacked the necessary resources or technical expertise, because the candidate lacked the requisite skills to implement his policies, or because he diverted resources for personal gain? To establish how voters understand poor performance in weak-state environments, I propose a two-stage conjoint experiment. For each of two hypothetical incumbents, subjects will hear a vignette about their pre-election platforms, which includes both promises and other common strategies such as vote-buying. After being asked to make reputational evaluations of each incumbent, subjects will then be presented with a vignette on post-election performance and asked to evaluate incumbent type as well as quality. Finally, to reduce potential social desirability bias, I propose a novel variation on the standard choice task in which existing data from a previous project is used to predict which among the two incumbents is more likely to be re-elected.

Whose Women's March: A Study of Instagram's Effect on Political Solidarity

Chaya Crowder, PhD candidate, Politics

The election and presidency of Donald Trump have stimulated broad-based resistance among members of the many marginalized groups who feel threatened by him, his positions, and the policies of his administration. One of the most notable mobilizations was the Women’s March the day after Trump’s inauguration. Political scientists have estimated that 4,157,877 people attended Women’s March protests across the nation, making the Women’s March was the largest recorded protest in American history. The organizers of the march sought to be inclusive of a diverse range of women and set out with intentionally intersectional goals. Consequently, the Women’s March provides an interesting test case for intersectional movement politics.  Instagram played a central role in how the organizers and participants of the march disseminated their message online. There were more Instagram posts using the hashtag #WomensMarch than posts using the hashtag #Election2016. This project uses a survey experiment to explore the impact of exposure to certain Instagram images and messaging on political behavior. Recent research has shown that the race of the messenger, the level of social endorsement (how many likes does a post have), as well as the content of can impact how messaging is perceived by the receiver.  This project explores the implications of previous research on a new social networking platform in the context of Anti-Trump mobilizations.