The power shield: Powerful roles eliminate gender disparities in political elections.

Are women less likely to win elections than men? Past analyses of U.S. elections have found little evidence of gender bias, leading some scholars to declare: "When women run, women win." However, across many professional domains, women face disparate outcomes in achieving leadership positions. The current research resolves this puzzle through a novel theoretical perspective and methodological advances. Theoretically, we propose that power frees women from restrictive gender norms, reducing gender bias. Thus, gender bias likely exists in politics but is more pronounced for lower-power candidates and less pronounced for higher-power candidates. Because incumbent candidates have more power and challenger candidates less power, we predicted incumbent women would be shielded from gender bias and achieve electoral parity with incumbent men. Conversely, we predicted challenger women would face particularly strong gender bias and disparate outcomes. Methodologically, we resolve prior scope-of-analysis limitations by analyzing every governor and U.S. senator election since women's suffrage (1920). Further, we developed a novel bootstrapping method that resolves regression assumption violations inherent in statistical analyses of candidate-level measures. Analyses revealed 2 important findings. First, our comprehensive dataset revealed that, contrary to past research, women were less likely to win elections than men overall. Second, we found evidence for a power shield effect: Male challengers were three times more likely to win than female challengers, men were 25% more likely than women to win open-seat races, but female incumbents fared just as well as male incumbents. These results suggest that some gender differences may be power differences in disguise. (PsycInfo Database Record (c) 2021 APA, all rights reserved).