Women and the opportunity costs of employment
Just as the Confederations Cup was about to start, another relevant meeting took place in Saint Petersburg: the Second World Congress on Comparative Economics. The timing of the conference could not be more appropriate, as the year also marked the 100 anniversary of the Russian revolution. As in the last comparative economic meetings, GRAPE was there. This time represented by Lucas, who discussed our research on female employment rates in transition economies.
In this research, we addressed the evolution of female employment rate since the beginning of transition. For this, we compile over 1400 individual level databases, both from transition and advanced economies. Our findings can be split in three parts:
- Female employment rate fell in transition economies, whereas it increased in advanced economies. The change in employment was not driven by the increase in unemployment (women in transition appear to be less sensitive to business cycle fluctuations than women in advanced economies); nor is a by-product of too-high employment rates prior to transition. It appears that the main driver of the fall was connected to frictions in school-to-work transitions.
- The failure to recover observed since the 2000's might be related to the lack of any significant effect of opportunity costs of working. Whereas in advanced economies developments such as the increasing importance of skill work, or the higher education obtained by women helped to reduce employment gaps, the same did not occur in transition countries. For almost all proxies of opportunity costs we failed to find a significant relation in transition countries. The sole exception being female participation rate, which suggests the existence of a “good-will effect”
- Lastly, we explore if this was a pure transition phenomena; particularly, we provide some test for decreasing marginal effects of opportunity costs. If that is the case, the effect should be larger when the gender employment gap is larger. Unconditional quantile regressions indicate that this is the case.
It goes without saying that we are thankful for the comments we received, particularly from …. We will try to improve the paper to meet your suggestions.
The slides of the presentation are available below.