A Half-Finding about Physician Incomes

Bloomberg News carried this fascinating article on September 2:

  

“Female physicians in the U.S. continue to earn less than their male counterparts, with the pay gap widening during the past two decades to more than $50,000 annually in 2010, researchers said.”

The article suggests that because of limitations in the source data the researchers “couldn’t adjust for a physician’s specialty or practice type.”   Thus the study would seem fatally flawed.

 But then the researchers go on to conjecture:

". . .  or do female physicians have less opportunity to enter higher paying specialties despite having similar preferences as male physicians?”

I think that's a reasonable conjecture, yet I wonder.  For one thing, what incentives are there for hospitals and group practices to hire male physicians at substantially higher compensation than equally-qualified female physicians?  For another, doesn't managed care in fact provide pretty much the opposite incentive?

I also wonder because worry has been expressed for years about a growing shortage of physicians in the U.S.    This shortage will become much worse if predictions turn out to be true that the ACA substantially increases the number of patients (by increasing the number of insured people) while the number of physicians increases only nominally.   This implies a lot more physicians will be needed than are likely to be in practice over the next few years, so demand for physicians would soar.  Don’t these circumstances actually increase women physicians’ bargaining power?

So I wonder - what's the mechanism, exactly, under which equally-qualified female physicians are being denied opportunities that exist, and that the female physicians seek?  

As it stands, the article reports only a half-finding: it's certainly true that there's an overall gap in median incomes.  But the research does not explain why this might be so.  Clearly more research is needed.

Keep in mind that statistically, it’s treacherous to rely on an overall result, be it median or average.  My favorite example is that, on average, Americans have one testicle and one ovary.  While that's certainly true, don’t expect to meet any such American right soon.