Yana Gallen, at Northwestern University, has written a paper summing up a study of gender gap in Danish employment. They found women are paid 16% less and are 12% less productive, but have unable to pin down the other 4% other than bias.
One important point, at least to me, was childless women were equally productive with men. I doubt that productivity loss stemmed from showing cow orkers pictures of their kids. Even in a family friendly utopia like Denmark, it's more likely exhaustion and stress, ongoing and accumulative. Days off and holidays bring no respite, just additional pressures, expected duties, and self-recrimination for not living up to the June Cleaver model.
Quote:
Abstract: Using Danish matched employer-employee data, this paper estimates the relative productivity of men and women and finds that the gender “productivity gap” is 12 percent–seventy five percent of the 16 percent residual pay gap can be accounted for by productivity differences between men and women. I measure the productivity gap by estimating the efficiency units lost in a firm-level production function if a laborer is female, holding other explanatory covariates such as age, education, experience, and hours worked constant.
To study the mechanisms behind the 4 percent gap in pay that is unexplained by productivity, I use data on parenthood and age. Mothers are paid much lower wages than men, but their estimated productivity gap completely explains their pay gap. In contrast, women without children are estimated to be as productive as men but they are not compensated at the same rate as men.
The decoupling of pay and productivity for women without children happens during their prime-child bearing years. I provide estimates of the productivity gap in the cross-section and estimates that account for endogenous sorting of women into less productive firms using a control-function approach inspired by Olley-Pakes.
This paper also provides estimates of the gender productivity gap across industries and occupations. Though the results do vary across industries and occupations, the overall estimate of the productivity gap is fairly robust to the specification of the production function.
|
The paper can be downloaded as a pdf at the link above.