It is extraordinarily difficult to determine the extent to which the gender wage gap reflects discriminatory behaviors by employers or differences in productive capacities between men and women. We note that where piece-rate work is performed, wages should in principle reflect productivity differences and that it is more difficult to discriminate on the basis of gender because one is paid for what one produces. With this as our point of departure, we compared men and women working in the same occupation in the same establishment, in three countries, the U.S., Norway, and Sweden. Our findings are easy to summarize. First, the gender wage gap is clearly smaller under piece- than under time-rate work. According to the argument put forth here, one third of the gap at the occupation-establishment level is due to discrimination, while two thirds of the gap is due to productivity differences.
The main finding is that the productivity differences between sexes in typically male-dominated blue-collar industries are very small, of 1-3%. Second, in age groups where women on average have extensive family obligations, the wage gap is higher than in other age groups. Third, under time-rate work, the wage gap is more or less independent of supposed productivity differences between men and women, while under piece-rate work. The wage gap mirrors quite closely assumed productivity differences, with women receiving a wage premium in female-advantageous settings and a penalty in male-advantageous settings. Fourth, in two of three countries women sort more often into piece-rate work than men.