Maybe that's what it used to be based on but more recent things I've seen have been talking about the difference for the same work, education and experience, i.e.:
"Even when comparing the sexes with the same job title at the same company and using similar education and experience, the gender pay gap persists: Men earned 2.4 percent more than women on average, down slightly from last year"[1]
"Procurement Leaders recent research shows that female buyers are paid less than male buyers. That is, women are earning less for the same work." [2]
"After accounting for job, industry, education and experience, Blau and Kahn determined that 38 percent of the wage gap comes from factors “unexplained.”" [3]
The meta argument I am trying to make is that recent arguments about this have shifted to including accounting for jobs (and usually experience levels), since jecjec was claiming that it's still the antiquated version that does not account for any of those things. If you want to have a discussion about the quality of any of those articles or the studies they're based on, well, that's a different discussion entirely, I present these as evidence that the discussion has moved, not that any of them are correct.
Band 5 staff earn between £21k and £28k. Band 8d staff earn between £65k and £81k.
When we restrict ourselves to a single healthcare profession (in this example I used widwives) that has many more women than men at the entry level jobs we still see men being promoted above women.
As others have pointed out, 2.4% is within the margin of error.
The "accounting for job, industry" is very rough, so "social science" lumps together economists (66% male, $70k median income) and social workers (68% female, $40k median income).
"Hours worked" is also almost certainly not accounted for properly, as that is not a linear relationship of you get %x more for %x more hours. Instead, being willing to put in extra hours is seen as a token of submitting to the dominance hierarchy and key factor towards advancement. Saying something like "I structure my work so I get everything done in 40 hours" is not the correct answer when asked about your willingness to put in extra hours, because that's not what this is about. And men are generally more willing to do this than women, and men that are not are shut out of promotion just as much (or probably more) than women.
Also there is risk-taking. A study reported on in the Süddeutsche Zeitung[1][2](German) shows that even in controlled laboratory conditions, women chose low-risk strategies even when the "risk" is almost entirely theoretical and the advantages of a higher risk strategy clear. As the high-risk strategies on average lead to greater rewards (as in real life), the authors report there was a 23% gap in male/female earnings within that study.
Finally, it turns out that the lives of the top earners actually aren't all that great. They more or less suck. Women are more savvy about this, and have more alternatives, whereas men are much more motivated to stay in such negative environments for the status/money rewards[3].
When you put these factors together, they actually appear to over-explain the gender wage gap, and of course that isn't entirely unlikely[4]
"Even when comparing the sexes with the same job title at the same company and using similar education and experience, the gender pay gap persists: Men earned 2.4 percent more than women on average, down slightly from last year"[1]
"Procurement Leaders recent research shows that female buyers are paid less than male buyers. That is, women are earning less for the same work." [2]
"After accounting for job, industry, education and experience, Blau and Kahn determined that 38 percent of the wage gap comes from factors “unexplained.”" [3]
[1] http://www.cnbc.com/2016/12/05/men-still-earn-more-than-wome... [2] https://www.forbes.com/sites/jwebb/2016/03/31/women-are-stil... [3] https://www.washingtonpost.com/news/wonk/wp/2016/03/08/its-2...