Are Female Candidates More Effective Populists? Causal Evidence From Bolivia

• In Event: Poster Session: Comparative Politics of Developing Countries II, African Politics & Middle East and North Africa Politics
  In Poster Session: Poster Session: Comparative Politics of Developing Countries II

Abstract

Conventional wisdom and recent scholarship agree that populist appeals to "the people" and against "an establishment" are a defining feature of political messaging in contemporary politics. Surprisingly, we still lack experimental evidence on whether populist appeals can be utilized with equal effectiveness by male and female candidates alike. If female candidates are seen as more credible when criticizing political elites characterized as failing to represent "the people", populist messaging could be a potent strategy for women to obtain political office. It is precisely their persistent political underrepresentation particularly in executive office that should enable women to call the "political establishment" out more persuasively. This perspective is consonant with recent findings that female candidates can more credibly criticize corruption among political elites (Barnes and Beaulieu 2014, 2019; Murray 2010) and are seen as more likely to bring about political change (Brown et al. 2011; Morgan and Buice 2013). To test whether female candidates are more effective populists, I conducted an original candidate choice survey experiment with a representative sample of 1346 respondents in left-wing populist Bolivia. In the vignette, I randomize the following characteristics of the hypothetical candidate: Gender (female or male), ethnicity (Aymara, Quechua or non-indigenous), age, marital status, the number and age of children, the political appeal (none vs. “people-centric” vs. “anti-establishment” vs. both “people-centric” and “anti-establishment”), the policy appeal, and leadership style (“makes important decisions on her/his own” vs. “likes to work with others when making important decisions”). Respondents observed two of such candidate profiles at a time – four times in a row – and were asked each time to choose which candidate they would prefer in the election. The average marginal component effects, conditional average marginal components effects and the average component interaction effects were estimated. In what appears to be only the fourth candidate choice experiment on gender in the developing world (Clayton et al. 2019), I do not find evidence that either the fully populist appeal or its components of people-centrism and anti-establishment critique interact with candidate gender in determining respondent candidate choice.

Author

• Julian Michel, UCLA