Uber is something of a gift to economist types, because they’re collecting and collating vast amounts of information which can be mined for insights. One such finding is that, even in the complete and total absence of any discrimination on the basis of gender, there's still a gender earnings gap.

Note that I said “earnings” there, not "pay." This is something vitally important to understand in the more general conversation (I would say series of whines, but that would be unkind, so I won’t) about the gender pay gap. We’re not, and we don’t, measure differences in pay — we measure differences in earnings. These are not the same thing.

A difference in pay, strictly defined, purely on the basis of gender is illegal and has been for some decades. This isn’t something we’d expect to be a feature of a society as litigious as modern America. And we don’t really see any large number of lawsuits alleging that legal principle is being broken.

What we do see is a great deal of political campaigning insisting that men and women, on average, earn different amounts. This has the advantage of actually being true (something quite useful in an allegation). It’s the next step in the logic which may or may not be correct – that this difference in earnings is a result of discrimination.

Which is where our Uber data comes in.

The platform does not distinguish between male or female drivers. Pay rates differ by not one whit. It would be one heck of a stretch to try and insist that customers were preferentially picking male drivers, leaving women with less work to do.

Yet we do still see a difference in average earnings. Some 7 percent in fact – very close to the more general gender gap once we’ve accounted for the usual things like levels of education and so on in the wider economy (that popular 77 cents number is before making all those sorts of corrections).

That is, in the absolute and complete absence of any discrimination, we still see a gender earnings gap. Thus we cannot assume, and shouldn’t, that any gap we can see is the result of discrimination.

A good analysis of the findings is here, the paper itself here. Here’s the punchline:

We find that the entire gender gap is caused by three factors: experience on the platform (learning-by-doing), preferences over where/when to work, and preferences for driving speed.

None of that is external discrimination by customers or the employer in any manner. Yet there's still that 7 percent gap in earnings, which gives us two important points in that ongoing conversation about gender equity.

If in the entire absence of any form of discrimination we’re seeing a gender earnings gap, then we cannot go around shouting that the existence of a gender earnings gap is due to discrimination. Which rather explodes the rhetoric of that 77-cent gap.

But much more importantly, we’ve got to decide what we’d like to do about this.

Do we simply accept that if the working choices of individuals lead to a gap in group earnings well, that’s that then? That’s the classically liberal answer of course, if free choice leads to differences then so be it.

Or, do we say that the existence of different group outcomes means something must be done? If, on average, women's approach to paid work is different than that of men, then we’ve got to overturn our society? No doubt we could find some progressives who would insist that we should, rather neatly showing the difference between a true liberal and a progressive.

My own two subsequent questions would be these: What can we do if zero discrimination still leads to this gap? And why would we do anything anyway?

Tim Worstall (@worstall) is a contributor to the Washington Examiner's Beltway Confidential blog. He is a senior fellow at the Adam Smith Institute.

If you would like to write an op-ed for the Washington Examiner, please read our guidelines on submissions here.