What Makes Voters Prefer Highly Educated Candidates? Unpacking Demand-Side Drivers of Unequal Descriptive Representation
(working paper with Andreas Videbæk Jensen)
Political candidates with lower levels of educational attainment are less likely to get elected. This is part of a larger pattern where some socio-demographic groups are substantially underrepresented in political office, which poses a challenge to democratic legitimacy and representation. We show that this electoral disadvantage occurs even though lower-educated voters -- just like higher-educated voters -- prefer same-group candidates, and we present innovative experimental evidence on the mechanisms underlying these preferences. Voters use education as a heuristic for whether a candidate's policies will favor their own group, and they infer candidate competence and warmth from education information. When information about these factors is provided, lower-educated voters no longer exhibit same-group preferences, whereas the initially large education premium among higher-educated voters is only halved. The findings advance our understanding of the cognitive mechanisms at play when voters process information about political candidates and challenge existing work stating that education-based electoral discrimination only favors highly educated candidates. Our findings suggest that providing detailed information about candidates reduces ingroup bias, and that lack of supply of lower-educated candidates may keep lower-educated voters from getting the office-holders they actually prefer.
Is Racial Discrimination Driven by Race or Proxies for Race? An Experimental Examination of Our Concept of Racial Discrimination
(with Andrew James Latham and Kasper Lippert-Rasmussen)
Discrimination against racial and ethnic outgroup members is a pervasive phenomenon. In the public and academic debate, such differential treatment is often viewed as a product of race. In this article, we ask a more fundamental question: Is racial discrimination rooted in race or is it rather based on (perceived) proxies for race? Theoretically, we distinguish between three key explanatory factors that either constitute or is perceived to correlate with race across different philosophical traditions: people’s ancestry, their skin color, and their education level. To test which of these three factors that underlie racial discrimination, we conducted a large, pre-registered augmented conjoint experiment fielded among more than 15,000 respondents in the US and Brazil. In the experiment, respondents were randomly assigned to evaluate 2 x 10 hypothetical political candidates, co-workers, or neighbors in which the amount and type of information available about the profiles varied between respondents. This design allows us to estimate three key quantities of interest: the total effects of the three explanatory variables, the eliminated (mediation) effects of the three explanatory variables, and the lexicographical order effects between the three explanatory variables. Data is currently being collected.
Early Experiences with Economic Hardship and Political Trust in Adulthood
(working paper with Francesco Colombo, Peter Thisted Dinesen, and Kim Mannemar Sønderskov)