On Monday the Census Bureau released new 2013 data showing median earnings by detailed occupation for full-time, year-round workers. The numbers also include the ratio of women’s earnings to men’s earnings by occupation, in cases where enough data were available for a reliable comparison.

Overall, the median earnings for full-time, year-round female workers were 78.8 percent of what their male counterparts earned. Of course, the mix of occupations that the sexes sort into varies significantly, and this compositional difference is often cited as a way to explain away the gender pay gap.

But, as research from Harvard’s Claudia Goldin has shown, more of the gender wage gap is explained by differences in pay within occupations than by differences between occupations. And the new Census data reveal just how pervasive this within-occupation gender wage discrepancy is. Of the 342 occupations for which enough data points were available, only nine showed women out-earning men in 2013. They are:

Occupational category % of occupation that is female Median earnings, men Median earnings, women Women’s earnings as a % of men’s Margin of error on earnings ratio
Producers and directors 37.3 $62,368 $66,226 106.2 10.3
Cleaners of vehicles and equipment 13.5 $23,605 $24,793 105.0 9.6
Wholesale and retail buyers, except farm products 49.2 $41,619 $42,990 103.3 5.9
Transportation security screeners 35.9 $40,732 $41,751 102.5 4.4
Social and human service assistants 78.8 $34,967 $35,766 102.3 11.6
Special education teachers 84.6 $46,932 $47,378 101.0 5.5
Transportation, storage and distribution managers 18.2 $52,017 $52,259 100.5 5.5
Dishwashers 15.5 $17,302 $17,332 100.2 7.4
Counselors 69.6 $42,299 $42,369 100.2 2.2

Some of these occupations are female-dominated (social and human service assistants, special ed teachers, counselors), but most are not. Some are low-income, and some are not. Interestingly, the occupational group in which women out-earn men the most is “producers and directors,” a field in which there’s been a lot of hand-wringing lately about bias against women.

Note that the margin of error is pretty big for all of these categories, which means it’s still possible that the typical woman in the field is in fact earning less than the typical man (and that likewise, some occupations in which women are reported as earning slightly less than men may actually have the median earnings for women outpacing those of male counterparts).

These new Census numbers don’t adjust for job tenure, experience, hours (beyond being full-time), education or other variables that could potentially affect earnings. When economists have tried controlling for these kinds of factors before, the wage gap has narrowed but did not disappear entirely. We don’t really know for sure what explains the persistent disparity in earnings between the sexes. Discrimination is one possible explanation, as is any difference in the way men vs. women negotiate and ask for raises. Some of Goldin’s more recent research has emphasized how much value the two sexes place on temporal flexibility (shorter or more flexible hours), and how willing employers are to accommodate those preferences.