60 1 Figure 6. Random effect estimates: Do voters prefer political candidates of the same racial/ ethnic minority group? Both Black Both Asian Both Other Both Latinx 0.00 0.05 0.10 0.15 0.20 Overview of the results of four meta−analyses showing that all four categories do not differ from each other significantly. Figure 6: Random effects estimates: Do voters prefer political candidates of the same racial/ethnic minority group? While the majority of the studies from the unjust and useful schools draw samples from US-populations, researchers apply a shared identification perspective to a wider range of contexts. This includes Brazil, India, Jordan, Kenya, Lebanon and Uganda. Only recently, scholars apply the shared identification framework to the US (e.g. Lemi and Brown, 2019). For instance, Aguilar et al. (2015) study the effect of ballot length on Brazilian voters who share racial/ethnic characteristics with candidates up for election. When respondents evaluate a short ballot, with three candidates only, ‘white’ and ‘brown’ respondents do not necessarily favor racial/ethnically congruent candidates whereas ‘Black’ respondents do. As the ballot length increases, all groups tend to favor racial/ethnically congruent candidates. Kao and Benstead (2021) study the conditions under which voters prefer women candidates in Jordan. They find that, overall, women are disadvantaged at the ballot box. However, sharing the same ethnicity with a candidate closes the gender gap completely. This means that (mostly male) respondents who will otherwise not choose a woman candidate will do so if they share the same ethnic identification. The authors call this surprising in light of the patriarchal inclinations of tribalist societies and point to the importance of an intersectional understanding of the dynamics of voting behavior (pp. 31–32). Non-experimental research echoes the findings in this meta-analysis. English, Pearson, and Strolovitch (2018), use surveys to study attitudes towards Members of Congress and other representative bodies in the United States and find clear shared racial/ethnic identification effects as well. Indeed, Blacks commonly understood to ‘vote as a bloc’ (Tate, 2003: 64). This is not a new finding; nor is it confined to voting behavior as it influences attitude formation as well (Mansbridge and Tate, 1992). We find that comparisons between experimental and non-experimental data point to shared racial/ethnic identification in a similar fashion (Philpot and Walton, 2007; Stout and Le, 2017). The chance of descriptive representation increases the turnout rates of ‘co-
RkJQdWJsaXNoZXIy MjY0ODMw