In his 1869 book on The Subjection of Women, the economist and political philosopher John Stuart Mill wrote that he had “repudiated the notion of its being yet certainly known that there is any natural difference at all in the average strength or direction of the mental capacities of the two sexes, much less what that difference is.” Whatever the differences, the political implications should still be what Mill thought.
An article in The Economist suggests that perhaps women make for better physicians than men. It cites a number of recent medical studies concluding that female doctors have better medical outcomes in terms of patient survival and hospital readmissions (“Do Women Make Better Doctors than Men? Research Suggests Yes,” August 7, 2024). The data comprised hundreds of thousands of medical records in Canada and the United States. The Economist notes that they were retrospective studies, which are less reliable than the controlled-randomized sort. It could be, for example, that women doctors covered by the studies were, for whatever reason, assigned the least serious cases, which would be controlled in a study that randomly assigned doctors to patients.
But why is it important to know whether women or men are better physicians? The question looks strange, except perhaps for hospitals, clinics, and medical groups who, if discrimination were legally allowed, would be interested in hiring the most efficient doctors—pushing the salaries of women doctors above their less efficient male counterparts’. Since sex discrimination is illegal in hiring (we would now say “gender discrimination,” which has the advantage of avoiding a culturally hated three-letter word, but I’ll stick with Mill’s terminology), there must be another reason why the question has become a research agenda.
In a free society, whether males or females are better doctors would have no philosophical or political implications, regardless of genetic or social causes. (Once the hypothesized difference in productivity has been priced into salaries, it would have little business interest either as the prices would provide sufficient information.) The question would be no more important than whether left-handed or right-handed doctors are better.
Now, it does seem rather obvious, doesn’t it, that women are genetically more empathetic and caring than men. The Economist suggests that information on the relative competence of men and women as physicians would help male doctors change what they don’t do right. But then, one would think that the same sort of studies on other social groups—say, white doctors vs. black doctors or left-handed doctors vs. right-handed ones—would also be useful. Why is that not the case? Certainly, such studies further reinforce the cage of group identities, but this should not be an objection for our group-loving intellectual establishment, except for the fact that some groups are more loved than others.
I agree, of course, that whatever research question somebody wants to investigate is his own business, although there is an issue as to whether the researcher should force others to finance his research. I have discussed this issue in a few previous EconLog posts–for example, about how spurious scholarly journals are helped by government financing of higher education. Freedom of research is the only way to know, as best as possible, that no important question has been neglected.
Given the zeitgeist of our time, we may wonder if the studies on the relative efficiency of male and female doctors would have been published if they had found that male doctors are better. Or perhaps such studies were buried by professional and academic journals? Imagine the headline in the press, “A Government-Subsidized Study Claims that Men Are Better Doctors than Women”! Mrs. Grundy (whose opinions have moved along with the times) would turn in her grave. In this area like in others, a free market in ideas is essential to the search for truth.
Let’s return to John Stuart Mill and examine how he justified the formal freedom of women to compete with men in all occupations, a more enlightened approach than the coercive one we are now used to. In The Subjection of Women, as I previously wrote on this blog, Mill argued that the emancipation of women would benefit everybody in society (or, should we say, would satisfy general rules beneficial to everybody) by allowing each person to contribute to the activities in which he or she perform best. Mill viewed discrimination against women as either harmful or superfluous. It was harmful if it prevented women from competing and proving themselves better than, or as good as, their male counterparts. It was superfluous if women could not or would not compete in certain jobs or tasks–garbage collectors, say. Mill saw no reason to prevent women, especially with discriminatory laws, from competing in any field of activity, but no reason for the government to help them either. What’s important is the formal liberty to compete, whatever the result is, whoever proves better at responding to individual preferences expressed on markets.
******************************
Those who have struggled with DALL-E will understand my frustration. The image was supposed to show male and female doctors on each side of a wall. But the bot did not understand. For more than an hour (for example), I tried to have him replace the woman standing up on the men’s side with a male doctor, or at least replace the head of the woman with the head of a man. I tried to teach the robot the secrets of life and the basics of anatomy. I finally gave up. Here is the image, in all its robotic imperfection.
READER COMMENTS
Richard W Fulmer
Aug 21 2024 at 10:53am
“Better” in this case has been defined to mean higher rates of patient survival and fewer hospital readmissions, which is entirely reasonable. Suppose for the sake of discussion, that further study reveals that women doctors make better diagnoses because they spend more time with their patients and more time listening.
Question: Will requiring that all doctors spend more time listening to their patients improve overall healthcare results? Possibly. But what if spending more time with each patient means that fewer patients are seen? Do better outcomes for those who are seen outweigh the cost to those who aren’t? Would the outcome for the entire population – both the treated and the untreated – be better?
Richard W Fulmer
Aug 21 2024 at 12:10pm
I think that my comment is relevant only in the absence of a free market in healthcare. With government in (partial or complete) control of prices and supply, monitoring collective outcomes makes sense.
In a free market, doctors with low rates of patient survival and more hospital readmissions will lose patients as their professional reputations decline and the number of their lawsuits rise. Privately run clinics and hospitals will tend to get rid of them if for no other reason that they threaten their bottom lines.
If, on average, women (or black, brown, religious, atheistic, tall, or so endlessly on) doctors have better results – and, therefore, bring in more customers – they will be in greater demand leading others like them to enter the profession.
steve
Aug 21 2024 at 7:24pm
Not really. Patients have few ways to tell who is really the better doctor. Plus, we have found that when we make that info available it is largely ignored. A few infovores or the type of person who reads econ blogs might use the info but that is a small minority. (In my state we publish outcomes for some surgeries. These are mostly ignored and a number of studies have found the same thing.) Also, better when applied to individual patients is hard to determine. Some patients have better outcomes with a doc who isn’t as bright but is nice and good at encouraging compliance.
The numbers are a little easier to track for surgery but a bit more nebulous for primary care and medicine. Unless they are truly awful it takes quite a while to build up a meaningful data base.
I will say that the article hits some points that resonate with me as a long time department chair. The guys are more likely to cowboy it and not follow guidelines. Occasionally that is merited and the pt does well but in general following the guidelines gives better outcomes. You need to know when to detour. Women are more likely to seek and accept ideas from others. Males practice with more ego. So I think my very best 3 or 4 docs are male but the women screw up less often. Certainly seems like groups that have a mix of male and female do better than groups that are all one or the other.
Steve
Peter
Aug 22 2024 at 3:37am
As a manager, I think that is a common observation. I know generally when I want something done and successfully so, men will rarely fail though sometimes you have to lean hard on them to motivate their ego whereas if I just want some routine mundane task accomplished good enough women tend to be better as men get bored and can’t be bothered to even half ass it, i.e. they would rather fail than be mediocre and well often mediocre is what you want. BTW I lump openly gays in the women category on that though I find lesbians stay in the woman category, even dykes.
As you said it’s good to have a mixed team though if I had my druthers I’d only hirer straight men and bull dykes as you get all the mixed team goodness without all the sexual undertone drama that tend to make mixed teams more trouble then their worth.
steve
Aug 22 2024 at 12:35pm
The sexual undertone stuff is pretty muted with my people. The issue that kills me is pregnancies. It screws up work and projects when someone disappears 3 months at a time and then when they come back they all need to pump. We have a busy ICU and then I need to find someone to cover every day while they need to go pump.
Steve
Jose Pablo
Aug 21 2024 at 3:17pm
Given the zeitgeist of our time, we may wonder if the studies on the relative efficiency of male and female doctors would have been published if they had found that male doctors are better
In this regard, Roland Fryer is still my “favorite” example of how the “zeitgeist of our time” responds to unwanted academic research findings.
Jose Pablo
Aug 21 2024 at 3:28pm
The question looks strange, except perhaps for hospitals, clinics, and medical groups
The question (well, the answer) is irrelevant even to these groups (if discrimination were legally allowed).
The conclusion helps very little in deciding in a real situation, which specific individual A, male or B, female, is going to make a better physician, except in the absence of any other piece of information about both specific candidates. A very unlikely situation. And in this case (absence of any other information) the information only slightly improves the chances of hiring the best physician compared with tossing a coin to decide which candidate to hire.
Even in a world where females are, on average, better physicians. Candidate A, male, can be a far better physician than Candidate B, a female. Hospitals don’t hire “averages” they hire specific individuals.
The study is a total waste of research talent and time. Well, maybe only time.
Pierre Lemieux
Aug 21 2024 at 3:52pm
Jose: Except if all individuals are identical, the result of course depends on which individual you select. Yet, whatever the average, the selection also depends on the variance within a group. If you have one urn with 99 black balls and 1 white, and a second urn with 99 white balls and 1 black, the probability is 0.99 that the random selection of one ball in the first urn will yield black ball, but only 0.01 that you will have the same result when closing your eyes and selecting one ball from the second urn. Sorry to inflict elementary probability theory on you, but randomness cannot be totally eliminated from choice. And the chooser is not willing to pay an unlimited price to diminish the uncertainty.
In the case we are discussing, of course, the proportion of black balls and white balls is presumably quite close in the two urns–which, I suppose, is your argument.
Richard W Fulmer
Aug 21 2024 at 4:08pm
Perhaps a better use of research money would be to identify the “best” (however defined) doctors, regardless of their demographic category, and determine what they are doing that “bad” doctors aren’t.
Jose Pablo
Aug 24 2024 at 3:48pm
The best use would be, I think, to develop an algorithm in which you plug some independent variables and that gives you back the “expected overall quality” of an individual doctor (in terms, for instance, of patient survival and hospital readmissions).
The “useful” research will then identify a limited number of variables to be included in an optimized practical version of the algorithm, taking into account how “significant” the variables are for the prediction and the difficulty of obtaining reliable values for them (avoiding “garbage in garbage out” phenomena).
I doubt the male-female variable would make the cut into the optimized algorithm. And this even though it is pretty easy to get reliable values for this variable … well at least it used to be in the past.
Mactoul
Aug 21 2024 at 10:24pm
That women make better or worse physicians is very weak research since goodness of a physician is not well-defined quantity.
Stronger is the research finding that while men and women have roughly equal average IQ, the men IQ has greater variance than women’s.
Thus, more men than women are to be found at high IQ level. Thus more male geniuses than female.
Which in turn argues against the perennial liberal canard that it was the old patriarchy that didn’t allow full development of female achievement. Eg Mills himself.
Darwinism is no friend to Enlightenment dogma. Natural selection has no tendency to produce equality irrespective of what Adam Smith wrote.
Correlations have been produced by economists themselves between national IQ and per capita GDP.
There is Murray’s The Bell Curve that enraged all liberals. But never refuted.
Pierre Lemieux
Aug 21 2024 at 11:27pm
Mactoul: Definition of the studies I discuss (according to The Economist):
Ahmed Fares
Aug 22 2024 at 3:08am
That’s good for the patients. For the doctors, not so much.
Suicide risk for female doctors 76% higher than general population
Pierre Lemieux
Aug 22 2024 at 11:08am
Ahmed: Very interesting. Three rapid and non-expert reflections. First, it is easy to understand that people in constant contact with death, disease, and decline would be more affected by a conscience of the absurdity of things–especially women, who are more empathetic and caring. This would be amplified by the fact that scientific men and women are presumably less religious than the general population. Second, as The Guardian‘s story mentions, there might be other factors in the environment of women doctors’ work environment that cause this difference in suicide rates between female and male doctors. (Is the sexual harassment factor, mentioned by The Guardian, an easy exaggeration?) Third, I have read (but haven’t checked) that dentists also have an unusually high rate of suicide: if that is true, it would deflate part of what I just said.
I wonder if Steve has further comments bearing on that.
steve
Aug 22 2024 at 2:59pm
There was a lot of heterogeneity in the study and it is different than older studies that didnt show such large sex difference. I think it is likely true as the study looks fairly well done but it doesnt tell us if it is just a sex difference or if women choose practice specialties where suicide rates tend to run higher.
My impression is that women doctors more commonly marry other doctors now or other high achievers. It’s generally a job with longer hours and a fair bit of stress. What I have seen over and over is that it is still almost always the job of the woman to be the main care provider for children and aging parents. This is now markedly different than it was 30-40 years ago but still mostly true. That dual burden and needing to be perfect at both pushes a lot of women over the edge.
As an aside, I find a lot of these young docs are horrible, IMHO, at managing money. If you have two married docs their income is very high, yet they set up budgets where they are living paycheck to paycheck. Then when someone gets sick it stresses them even more.
Steve
nobody.really
Aug 23 2024 at 12:06pm
This blog recently hosted a discussion about making visiting a doctor more like visiting a veterinarian. I don’t recall the discussion observing that veterinarians have a higher suicide rate. Specifically, male vets kill themselves at 1.8 times the rate of the general population; female vets kill themselves at 3.4-5.0 times the rate of the general population.
For whatever that’s worth.
BC
Aug 22 2024 at 4:41am
Given that doctors make up a small fraction of the general population, I would guess that the apparent outperformance of women over men doctors reflects some sort of selection effect rather than general gender differences between women and men. Perhaps, women face a higher bar in getting through medical school or getting hired. Alternatively, perhaps men that would have made good doctors self-select into other professions, e.g., computer scientists or hedge fund managers, more often than women. (And, yes, I would say the same thing if the “less loved” group was found to outperform. Given the specially selected population, look for selection effects, e.g., DEI policies.)
Re: the DALL-E problems. I wonder whether some AI “safety” feature has inadvertently made it difficult to replace the female doctor with a male one, similar to the way the Google AI generated an image of Black and Asian Nazi soldiers. The concept of “safety” has certainly evolved…
David Seltzer
Aug 23 2024 at 3:14pm
Pierre: If women are better physicians than men, do they earn more or less? If women are better, is the demand for their services greater? If so, wouldn’t their salaries be higher than their male counterpart? In terms of human capital, is the economic value of a woman’s medical skills and experience greater? An idea for further research.
Comments are closed.