Abstract
Psychology literature has long argued that both men and women perceive a woman competent in male tasks as less likeable and that this biased perception explains women's underrepresentation in the top positions of the labor market, sometimes called the competence-likeability tradeoff. However, almost all studies test this conjecture in hypothetical decisions and scenarios that do not necessarily translate into hiring and promotions in the real world. I provide evidence that this tradeoff does not exist in a real decision. Using a laboratory experiment where I exogenously vary the opponent's gender and relative performance in an IQ test -- which measures competence in male tasks -- I find that neither men nor women are less generous to a woman than to a man with a better IQ test performance. The results are not due to women's in-group preference, distaste against male losers, a biased belief that women are worse at the IQ test than men, experimental manipulation failure, or ex-post randomization failure. These findings suggest that the competence-likeability tradeoff does not explain the women's underrepresentation in the labor market.