Identifying the role of expected fertility as a driver of gender discrimination is crucial to guide policy solutions, but is hindered by the fact that pregnancy discrimination is typically addressed by broader gender-discrimination policies. This was not the case in the U.S., where before the 1978 Pregnancy Discrimination Act (PDA) pregnancy constituted a legal exception to existing gender discrimination laws. Exploiting states’ staggered adoption of laws similar to the PDA, we use difference-in-differences estimates based on survey data to show that the PDA reduced employment and hiring of women of fertile age, while having no detectable effect on dismissals. Our estimates imply that nearly 30% of the gender employment gap at the time was due to potential fertility. A simple search-and-matching model explains these results: by raising firing costs for discriminating employers, the PDA discouraged hiring but was not enforced strongly enough to prevent dismissals of pregnant employees. Finally, we show that pre-existing equal pay laws limited wage adjustments, exacerbating the negative effect on employment.
While the gig economy has been growing rapidly, evidence on its implications for the labor force remains sparse. This paper shows that gig work can buffer income losses arising from unemployment shocks. Matching a large sample of administrative employment data with information on platform entry, we demonstrate that recently unemployed workers are more likely to take up gig work and less likely to receive unemployment benefits (UB) when a gig delivery platform is available to them. Our identification relies on a difference-in-differences strategy, exploiting the staggered market entry of gig platforms across German districts. The analysis suggests that the effects are permanent and heterogeneous across various socioeconomic dimensions. Moreover, we discuss substantial fiscal externalities that could arise as a consequence of workers’ substitution of unemployment insurance with income from gig work.