The "Electability" Trap: Bias in Beliefs on the Quality of Minority Candidates

• In Event: Understanding Vote Choice

Abstract

Scholarship argues that negative social expectations harm underrepresented candidates (Alexander et al. 1993; 1991). For instance, respondents perceive of men as more "capable" candidates, and regard Blacks as "poorer leaders" (Bos 2011; Holman et al. 2011; Huddy et al. 1993; Tesler 2013). However, there exists a puzzle as to precisely how stereotypes inhibit underrepresented candidates’ electoral chances. Despite negative stereotypes, women garner comparable and oftentimes improved, levels of vote support as compared to men (Lawless et al. 2008; Sanbonmatsu 2002). Analysis similarly finds contradictory evidence that Blacks face reduced vote support, as compared to whites (Doherty et al. 2019; Highton 2004).

To address this puzzle, the focus of contemporary scholarship centers upon "electability"—the perception that a candidate profile’s strengths ensure victory, given preferences of the electorate and the characteristics of previously elected individuals (Bateson 2020; Karl et al. 2014; Loepp et al. 2020). Thus, the modern zeitgeist on underrepresentation addresses the ways respondents evaluate candidate profiles according to their personal criteria for candidates’ ideal positions, personality, competency, and convictions.

However, there are two shortcomings with the electability framework, one theoretical and one methodological. Regarding the first, conceptualization of electability relies on the assumption that the criteria for judgment prevail regardless of candidate identity—this mistakenly implies voters experience equal perceived utility from same attribute, when in fact a respondent’s value of a profile trait varies by candidate profile (Moskowitz et al. 1996). Regarding the second, analyses of viability often rely upon Choice-based Conjoint design, which focuses on animus and aversion in post-hoc decision-making, overlooking the role of implicit social expectations in shaping ex-ante standards. This project addresses both concerns, offering two central contributions: (1) a theory of Experience and Qualification Bias in respondents’ criteria for judging candidate profiles, and (2) an inventive methodological framework, Menu-based Choice design, to estimate the effects of identity on these criteria.

Theoretical Contribution: Therefore, in addition to the myriad challenges faced by underrepresented
candidates, this project introduces Expectation and Qualification Bias—the asymmetrical standards regarding which profiles are sufficient for victory and support—as an obstacle in the primary and election stage. This improves upon scholarship by clarifying harmful considerations that impact underrepresented candidates before voters reach for the ballot box. Discrimination might present itself as differing evaluations of identity groups and candidate traits, given the importance of their credentials to voter perceptions of viability (Funk 1999; Hellweg et al. 1989). For example, respondents differentially rate the qualifications of underrepresented job candidates, with women and racial minorities evaluated on multiple criteria and faced with tougher requirements (Bertrand et al. 2016; Moscatelli et al. 2020). Findings of inconsistent evaluative standards scaffold scholarship that examines variation in support for candidates based on their race/ethnicity (Caruso et al. 2009), party affiliation (Rahn 1993), and gender (Hassell et al. 2019; Hoyt et al. 2009).

Methodological Contribution: This paper builds upon prior scholarship, by moving beyond Choice-based Conjoint analysis, which captures the marginal effect of profile attributes on choice, to a Menu-based Choice inquiry that estimates the marginal probability a voter requires an indispensable profile attribute given a candidate’s identity. This model addresses the presence of high utility attributes such as higher education compensating for social expectations or discrimination (Fogliatto et al. 2012). I posit that conflicting findings of discriminatory vote-choice are artifacts of the structure of an experimental arrangement that focuses on product rather than process in decision-making.

Broader Impacts and Findings: I find that female and racial minority candidate profiles correlated with a higher probability of preferences for messages depicting high work-ethic and competency. In addition, these underrepresented candidate profiles corresponded to a higher likelihood of recommendations to pursue more rigorous occupations, schooling, and greater years of experience. This skew in demands extended to attributes such as age, marital status, and family size—each impacting women and women of color as increased child-care responsibilities and economic downturns represent significant factors in women’s labor force participation (Higgins-Dunn 2019; Parker et al. 2020; Rugaber et al. 2021). As we seek to broaden diversity in democracy, awareness of these asymmetrical standards may break the cycle of imploring and expecting women and racial minority candidates to collect strong signals in ways not expected of their counterparts.

Author

• Jennifer E. Cryer, University of Southern California