Are Women Penalized for Coauthoring in Academic Economics? - Evidence from the Lab
Year: 2017
Type: dissertation
Abstract: The large discrepancies in the fraction of male and female tenured academics have been subjected to a vast amount of research.Among the explanations of the male dominance are differences in preferences between men and women, structural conditions that favor men and mere discrimination.Within the field of economics, Sarsons (2017) assigns the disparity in promotions to tenured positions, to differences in recognition for joint work.Her findings suggest that male economists experience an equal raise in probability of promotion of writing an extra paper, independently of whether they collaborate on publications.Women attain the same increase in probability when publishing individually, but are substantially less credited for joint work.Similar analyses in sociology does not reveal the same coauthor penalty for women.This leads her to believe that the alphabetical order of coauthors in academic economics and ordering per contribution in sociology constitute the difference.This thesis explores the notion of coauthor penalty by two means.I first relate coauthor penalty to existing economic theories of discrimination.Coauthor penalty fits well with the concept of statistical discrimination.That is, discriminatory behavior based on the use of heuristics under limited information.I find that coauthor penalty is largely in line with Phelps' (1972) seminal model of statistical discrimination.Furthermore, Sarsons' (2017) results simile those of empirical literature claiming to identify other cases of statistical discrimination.ordered alphabetically (alphabetical treatment).The rest observed tables where each of the four subjects' pairs were listed according to who did best (first-author treatment).The best one was placed first.We asked them to choose two team members who would earn them money in the subsequent quiz.In part 3, all the participants were shown similar tables as in the previous part, but could observe the four subjects' individual score in each quiz from part 1.We asked them to pick one team member for the following quiz.In the last part of the experiment they answered two questions: who did worst and who did best in the quiz in part 3? They were allowed to choose from three candidates.I find that subjects consistently use performance variables when assessing candidates in part 2 and 3. Furthermore, there is no evidence of coauthor penalty for females.When information on performance is absent, the subjects make decisions independent of gender.Moreover, I find no evidence suggesting that females are less likely to be chosen relative to males in the alphabetical compared to the first-author treatment, controlling for available signals on performance.In both treatments, subjects seem to choose according to the score of the candidates and, in the first-author treatment, the number of times a candidate is listed first.
Source: NVA (Nasjonalt vitenarkiv)
Author: Vegard Sjurseike Wiborg
Institution: University of Oslo
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
Funder: Universitetet i Oslo
OpenAlex ID: W2751986960
