Ductor, Goyal and Prummer (Review of Economics and Statistics, 2018) document gender differences in productivity in economics across 30 years. They argue that these differences can be explained by differences in the co-authorship networks of men and women. Our goal is to test whether differences in the co-authorship network of men and women can be explained by men and women having different expectations regarding the performance of others. In this paper, we examine gender differences in taskperformance and expectations in four different tasks (emotion-recognition, verbal ability, mathematical ability, and mental rotation). According to previous literature, these tasks differ in the (ex-ante) expectation that women will perform better than men. We find support for the hypothesis that women perform better than men when eliciting the beliefs of participants in the task and external observers. The use of monetary incentives to elicit expectations does not matter for the accuracy of the reported beliefs (i.e., we do not find support for a social desirability bias or demand-effect). These findings highlight that differences in networks could be explained by homophile or discrimination, rather than expecting different performance from men and women.