Preferences for flexibility and gender labor market inequality

We use a search model of the labor market in which jobs are characterized by flexibility (such as the possibility of working from home, or discretion in choosing work-hours) to estimate the distribution of preferences over flexibility. In an hedonic wage model, a job amenity is estimated to convey positive utility only if the conditional mean wage of jobs with the amenity is lower than the conditional mean wages of jobs with-out the amenity. However, in an environment with on-the-job searching and wage posting, firms may gain positive prot by oering both a higher wage and the job amenity because doing so will reduce worker turnover. The observed wage distribution may then exhibit a positive correlation between wages and the job amenity even if workers are willing to pay for it. Our model parameter identification relies on differences in the predicted wage distributions over flexible and non-flexible jobs. We compare our estimates with the results obtainable from hedonic wage regressions that assume that wage differences arise from pure compensating differentials. To measure flexibility, we use detailed information about daily activities from the American Time-Use Survey, which we match with worker and job information contained in the Current Population Survey. We use the model estimates to perform counterfactual experiments, showing that flexibility has a substantial impact on the wage distribution but a negligible impact on the unemployment rate. This implies that subsidizing flexibility could reduce the gender wage gap without signicantly affecting overall welfare.