Amal Boughnim (Bordeaux School of Economics)
5 November 2024 @ 12:30 - 13:30
- Past event
The Anatomy of Gender Gaps in Scientific Productivity Across the Career Cycle and Generations
Abstract: Whereas gender differences in science have long been reported, the literature remains inconclusive in many respects, and much remains to be learned about how (and why) these differences evolve across generations and disciplines and materialize at different career stages. We provide new evidence on these issues based on unique administrative data on more than 94 thousand professors and researchers employed at French higher education and research institutions, and their publication outputs recorded over nearly 60 years. This data allows us to properly control for compositional effects, avoid misrepresentations of the career relying only on publication data, properly take into consideration that women do (or did) more often change names, and properly control for numerous fixed effects.
We find that, although gender representation has improved across time and generations, the gender gap has not decreased in younger generations. It is even larger in fields characterized by a larger representation of women. Interestingly, the gender gap has a specific shape over the career cycle that essentially perpetuates across cohorts: it reaches a maximum of 35% of men’s average annual productivity at age 43 and then declines in the late career in all fields except life sciences. We further examine two potential factors of gender productivity gaps that are consistent with these stylized facts. Using an event study performed on a subset of professors, we show that mothers are more impacted than fathers by the birth of their first child in the long run. Taking advantage of the fact that there are two distinct careers for tenured academics in France—one for (full-time) researchers and one for professors (who also have teaching duties)—we show that there are widening differences in the gender gap among professors, relative to researchers, over the career cycle. We suggest that the two phenomena may be the consequence of a correlation between gender and attitudes in a multitasking context where career progression essentially depends on one task.