> I think the people who wrote freakonomics did adjust for gender pay gaps between men/women who have the same job and found that the gap is much smaller[0], an

> I think the people who wrote freakonomics did adjust for gender pay gaps between men/women who have the same job and found that the gap is much smaller[0], and I understand how that is a problem if a man gets paid more than a woman for doing the same job. Even here, "women make less than men within a given field, ergo discrimination" is the wrong takeaway. It's very likely that just like Goldin observed that different fields make different tradeoffs between flexibility and pay, so do different job postings within a field (however, her study isn't equipped to measure this). This means that within, say, software engineering, there are some postings that pay very well but require long, inflexible hours while others pay less well but are more accommodating. In all likelihood, women disproportionately self-select into these more accommodating jobs within the field, just like women disproportionately self-select into the more accommodating fields.