If you take studies that compare the gross earnings of men and women and control for some of the obvious differences in decision-making, the gender gap tends to disappear. For example, I remember hearing of a study that looked at something like "single, childless people aged 25-34 with a college degree and living in an urban environment." In this narrow demographic slice, women made something like 2% more money than men.
And then if you look at the US census data on occupations by gender, you can easily notice a trend in which jobs have what gender ratios. Roughly speaking it's something like "if the pay and requirements are equivalent, which is a more preferable job" often lands on female dominated fields.
Education, training, and library occupations: 73.1% female
Healthcare support occupations: 86.5%
Law enforcement workers including supervisors: 19.7%
Office and administrative support occupations: 70.8%
Natural resources, construction, and maintenance occupations: 5.1%
Production, transportation, and material moving occupations: 22.8%
The one big outlier to this pattern that I noticed was "Computer, engineering, and science occupations" at 25.7%. This also is a considerable category outlier in terms of skill requirements and competitiveness. "Management occupations" seems like it should also have some of this effect but to a lesser degree, and sits at 40.3%.
Source: American Community Survey, table S2401
Another interesting tidbit is comparing S2401 with S2402 - comparing all workers with only full-time year-round ones. I haven't done a full look, but one thing that jumped out at me is that excluding part-time or part-year workers drops the overall workforce percentage from 47.5% to 43.1%. It might be confounded somewhat by school teachers being considered "part year" and mostly women, though.