We examine the extent and channels through which cross-market differences in women’s relative labor market outcomes and non-labor market outcomes (e.g. marriage and fertility) are determined by differences across markets in sexism — defined as beliefs about the appropriate role women should play in society. Using data from the GSS to measure sexism, we document that the large and stable cross-state differences in women’s outcomes are strongly related to overall sexism in the population. To isolate the causal effect of residential sexism, we use a sample of internal migrants and an IV strategy that exploits plausibly exogenous variation in the prevailing sexism migrants are exposed to in their state of residence due to migration costs. Next, we attempt to distinguish the role of norms vs. discrimination in determining this observed relationship. First, consistent with the idea that norms matter, we find that conditional on state of residence fixed effects, women born in more sexist states have poorer relative labor market outcomes, are more likely to be married, and have children earlier. Second, to examine the role of discrimination, we find that as predicted by taste-based discrimination models, gender gaps in labor force participation and offer wages at the state-level are (a) more strongly related to male sexism as opposed to female sexism (b) are negatively related to the sexism of the median male, but not to any other percentile of the distribution of male or female sexist beliefs in a state. By contrast, women’s non-labor market outcomes are more strongly associated with mean female sexism, and do not vary systematically with the median or any other percentile of the female or male sexism distribution. Overall, these results suggest that while both prejudice-based discrimination and norms appear to separately matter for cross-state gender gaps in labor market outcomes, women’s non-labor market outcomes are influenced largely by norms.
Please sign up for meetings below: