There is a graph showing inflation by type of good. It shows that CPI is broken out into manufactured goods (that have cratered in price) and services, namely health and education, that have greatly exceeded CPI.
Another way of phrasing this is that male professions exist in the free market and have dramatically succeeded in making society better, so much better many of these men have lost their jobs because the sectors are so productive.
Meanwhile, the female sectors of health and education, which are largely paid for by government and feature a mix of public sector unions and professional guilds, have eaten up all surplus generated by the male sectors and returned…products whose value we can’t seem to measure (education hasn’t returned better outcomes, healthcare spends more without measurable outcomes). Everyone seems to think Amazon is a success and cars are better today, but nobody likes how our healthcare system works.
But you know what, these unproductive sectors have been very productive at creating “jobs”. It’s been a good deal to be in nursing the last few decades. And it’s been a good deal to be in teaching, at least in the blue states with the gold plated benefits/pensions and relatively high pay.
As for business and professional services it would be interesting to see a breakdown. The most female coded business professional, Human Resources, grew 64% over the last decade versus 14% for the overall economy. I suspect that trend goes back further. I don’t think any of us believe HR is adding business value, but we do know that it’s required because of government regulation and lawsuits. I have no doubt the ridiculous HR trainings I click through without reading had to be created by some women with a fake email job.
“Goods make society better and services don’t” is not the correct takeaway from the Baumol effect, which is descriptive and not about the value, in a “does this improve society sense,” of output. Rather, it’s that some sectors simply can’t be automated or otherwise increased in productivity as much as others because of the nature of the work.
As for HR, having a part of your company whose job is to handle payroll and benefits for you so you can
1. Get info when you need and it and
2. Otherwise think about it as little as possible
Is clearly value additive imo, at least to a point, and is well worth an annual hour-long sensitivity training.
I think your first point has some merit, but your second regarding HR is borderline satire in the modern business world. I know very few people who would identify HR’s central function as payroll processing and many who wish it was.
Again, you seem unwilling to address the fact that these sectors are not market based. Education is a public sector union. The ama and healthcare sector more broadly is a highly regulated guild and the main payer is the government.
In those few areas of medical care that aren’t a government regulated mess, like plastic surgery, costs have come way down.
Handling payroll and benefits clearly hasn’t gotten harder (hell my mom used to work at one of those places that needed punch cards to do it). The increase in HR is clearly related to regulations especially around civil rights/sensitivity.
It is true that they can’t be effectively automated *as well as* (note my use of the phrase “as much as”), say, manufacturing. It has to do with the nature of the work and its outputs much more than it has to do with the presence of unions or guilds. Think about what we’re discussing here: if my factory makes 50 widgets a day, it is quite likely that more efficient technology and production methods can allow me to make lots more widgets in the same amount of time. Better technology, management, and processes can probably produce *some* efficiencies in the business of treating patients or educating children but fundamentally you cannot produce exponentially greater numbers of healthy patients or educated children this way because the output (e.g., a successfully treated patient) is relatively unresponsive to improved technology compared to manufacturing outputs. To use Baumol’s own classic example, “output per man-hour of the violinist playing a Schubert quartet in a standard concert hall is relatively fixed."
While I’m sure that public sector unions, lawyers’ and doctors’ “guilds” and so on produce inefficiencies, I am skeptical that they’re the major driver of the Baumol effect (as was Baumol himself). You’ll recall that manufacturing work was quite heavily unionized once upon a time (albeit by private sector unions), but apparently this was not an obstacle to automating much of that sector. With healthcare, if credential gatekeepjng and public sector bloat were the major drivers of cost inflation we should expect to see the biggest cost inflation in the U.S. or perhaps Europe, but in fact healthcare cost growth rates are… lower in the U.S. and Europe than the global average and well behind the Asia-Pacific and MENA reasons as reported here: https://www.wtwco.com/en-us/news/2025/11/double-digit-healthcare-cost-increases-projected-to-persist-into-2026-and-beyond
1) K-12 educational expenditure is 3x what it was in the 1960s inflation adjusted. Much of the increase is in non-teaching personal. This isn't a case of the idea that teachers can't be automated (btw, they could be). It's that places like NYC are spending $600,000 per classroom and it's all getting eaten up by non-teaching personal.
2) "is relatively unresponsive to improved technology"
at a fundamental level it is unresponsive because nothing can improve it. Education can't change IQ. Most education is a mixture of glorified babysitting and gatekeeping to extractive guilds.
Healthcare is similar. We plow resources into a mixture of trying to re-animate a corpse (end of life care), blatant scams (anything involving poor people with $0 cost shares), and Red Queen Race nonsense (ever heard of coding, an army of "nurses" goes to delirious old peoples homes and asked them to regurgitate codes for revenue).
Your second point here supports the Baumol effect argument. The outputs of education and healthcare are, for a number of reasons, simply relatively unresponsive to productivity enhancements compared to the outputs of, say, manufacturing. Typically productivity enhancements in a given sector result in both wage increases and employment decreases, but sectors that can’t reap the productivity enhancements are nonetheless incentivized to raise wages to compete for workers. Then they pass the higher input costs of labor along in the form of higher costs (whereas sectors like manufacturing can lower input costs by decreasing net employment even as wages are going up). This is true, of course, all around the world. I don’t deny administrative bloat also increases costs but the fact that these observed effects are not U.S.-or west-specific suggests against the common “it’s women being promoted into unproductive sectors for affirmative action reasons” type theories.
And every single advanced country features the government as primary payer in education and healthcare. And every single country sees high trend is these sectors, shocking!
And in any case where the government wasn’t the payer we didn’t see massive increases in cost. Plastic surgery and LASIK got better and cheaper, because they aren’t covered by the government.
Most private schools still operate for less than most public schools. In states that have implemented school vouchers, which often pay only a fraction of what the public schools spend per kid, we’ve seen swift disenrollment from public schools. Turns out the private educational market can turn out a competitive product at half or less the cost as soon as you end the monopoly.
So yes I think that in a world where the government wasn’t providing infinite backstop to Eds and Meds that these sectors would do more with less. It’s not an iron law that healthcare has to eat our entire economy. In Singapore where they enforce market discipline healthcare is 3% of gdp.
How does polling line up in this stuff again? Oh yeah women vote to the left of men by a wide margin (bigger government) with especially strong support for Eds and Meds subsidies. We gave women the vote and they voted for the government to fund their own jobs!
Now we’ve got out of control cost spirals in pointless red queen races that are eating our entire economy. But at least every women that can get an education degree (lowest average iq of ba grads) can get a sweet job where you get two overgenerous pensions by 65, summers off, and if there is a bad flu you can decide your just not going to go to work for two years (the men though can’t, they are essential workers).
I’m not sure whether this is intentional obfuscation but the analysis has a pretty gaping hole. Your labor participation rate by education level holds, if you don’t consider the proportion of each population who attained the relevant level. Of course, the number of women who have attained a college degree has increased more than double that of men over that ~30 year period. So women have shifted a significant portion of their population (between 20-30% from what I have seen) from the low labor participation pool to the high labor participation pool and that data tends to align perfectly with the industries recognized as increasingly bloated (healthcare and education, with HR another consideration). If your labor participation numbers are accurate, it’s not clear why this trend is not more clearly represented in your central analysis.
The fact that we can look back to the 1990s to see where women received help isnt surprising. I remember the bombshell report that dropped shocking many Americans: Affirmative Action workplace policies from the 1970s to the 1990s, which was designed to remedy discrimination against African Americans and racial minorities and promoted heavily in the media, primarily benefitted White women.
Guess who has been promoting the idea of girl boss culture since the 2000s? White women and the media.
Thank you for citing my Argument piece! Your essay is excellent. I also wrote about the Stepman essay in my newsletter this week: https://darbysaxbe.substack.com/p/in-defense-of-husbands
There is a graph showing inflation by type of good. It shows that CPI is broken out into manufactured goods (that have cratered in price) and services, namely health and education, that have greatly exceeded CPI.
https://www.visualcapitalist.com/inflation-chart-tracks-price-changes-us-goods-services/
Another way of phrasing this is that male professions exist in the free market and have dramatically succeeded in making society better, so much better many of these men have lost their jobs because the sectors are so productive.
Meanwhile, the female sectors of health and education, which are largely paid for by government and feature a mix of public sector unions and professional guilds, have eaten up all surplus generated by the male sectors and returned…products whose value we can’t seem to measure (education hasn’t returned better outcomes, healthcare spends more without measurable outcomes). Everyone seems to think Amazon is a success and cars are better today, but nobody likes how our healthcare system works.
But you know what, these unproductive sectors have been very productive at creating “jobs”. It’s been a good deal to be in nursing the last few decades. And it’s been a good deal to be in teaching, at least in the blue states with the gold plated benefits/pensions and relatively high pay.
As for business and professional services it would be interesting to see a breakdown. The most female coded business professional, Human Resources, grew 64% over the last decade versus 14% for the overall economy. I suspect that trend goes back further. I don’t think any of us believe HR is adding business value, but we do know that it’s required because of government regulation and lawsuits. I have no doubt the ridiculous HR trainings I click through without reading had to be created by some women with a fake email job.
“Goods make society better and services don’t” is not the correct takeaway from the Baumol effect, which is descriptive and not about the value, in a “does this improve society sense,” of output. Rather, it’s that some sectors simply can’t be automated or otherwise increased in productivity as much as others because of the nature of the work.
As for HR, having a part of your company whose job is to handle payroll and benefits for you so you can
1. Get info when you need and it and
2. Otherwise think about it as little as possible
Is clearly value additive imo, at least to a point, and is well worth an annual hour-long sensitivity training.
I think your first point has some merit, but your second regarding HR is borderline satire in the modern business world. I know very few people who would identify HR’s central function as payroll processing and many who wish it was.
“Can’t be automated”
Is this true? Or is it illegal to automate them.
Again, you seem unwilling to address the fact that these sectors are not market based. Education is a public sector union. The ama and healthcare sector more broadly is a highly regulated guild and the main payer is the government.
In those few areas of medical care that aren’t a government regulated mess, like plastic surgery, costs have come way down.
Handling payroll and benefits clearly hasn’t gotten harder (hell my mom used to work at one of those places that needed punch cards to do it). The increase in HR is clearly related to regulations especially around civil rights/sensitivity.
It is true that they can’t be effectively automated *as well as* (note my use of the phrase “as much as”), say, manufacturing. It has to do with the nature of the work and its outputs much more than it has to do with the presence of unions or guilds. Think about what we’re discussing here: if my factory makes 50 widgets a day, it is quite likely that more efficient technology and production methods can allow me to make lots more widgets in the same amount of time. Better technology, management, and processes can probably produce *some* efficiencies in the business of treating patients or educating children but fundamentally you cannot produce exponentially greater numbers of healthy patients or educated children this way because the output (e.g., a successfully treated patient) is relatively unresponsive to improved technology compared to manufacturing outputs. To use Baumol’s own classic example, “output per man-hour of the violinist playing a Schubert quartet in a standard concert hall is relatively fixed."
While I’m sure that public sector unions, lawyers’ and doctors’ “guilds” and so on produce inefficiencies, I am skeptical that they’re the major driver of the Baumol effect (as was Baumol himself). You’ll recall that manufacturing work was quite heavily unionized once upon a time (albeit by private sector unions), but apparently this was not an obstacle to automating much of that sector. With healthcare, if credential gatekeepjng and public sector bloat were the major drivers of cost inflation we should expect to see the biggest cost inflation in the U.S. or perhaps Europe, but in fact healthcare cost growth rates are… lower in the U.S. and Europe than the global average and well behind the Asia-Pacific and MENA reasons as reported here: https://www.wtwco.com/en-us/news/2025/11/double-digit-healthcare-cost-increases-projected-to-persist-into-2026-and-beyond
1) K-12 educational expenditure is 3x what it was in the 1960s inflation adjusted. Much of the increase is in non-teaching personal. This isn't a case of the idea that teachers can't be automated (btw, they could be). It's that places like NYC are spending $600,000 per classroom and it's all getting eaten up by non-teaching personal.
2) "is relatively unresponsive to improved technology"
at a fundamental level it is unresponsive because nothing can improve it. Education can't change IQ. Most education is a mixture of glorified babysitting and gatekeeping to extractive guilds.
Healthcare is similar. We plow resources into a mixture of trying to re-animate a corpse (end of life care), blatant scams (anything involving poor people with $0 cost shares), and Red Queen Race nonsense (ever heard of coding, an army of "nurses" goes to delirious old peoples homes and asked them to regurgitate codes for revenue).
3) Nurses make bank:
https://www.wsj.com/economy/jobs/nursing-jobs-pay-prosperity-b2769391
Teachers make bank:
https://www.edweek.org/leadership/the-zero-sum-game-of-school-budgets-and-teacher-pensions/2024/10
Your second point here supports the Baumol effect argument. The outputs of education and healthcare are, for a number of reasons, simply relatively unresponsive to productivity enhancements compared to the outputs of, say, manufacturing. Typically productivity enhancements in a given sector result in both wage increases and employment decreases, but sectors that can’t reap the productivity enhancements are nonetheless incentivized to raise wages to compete for workers. Then they pass the higher input costs of labor along in the form of higher costs (whereas sectors like manufacturing can lower input costs by decreasing net employment even as wages are going up). This is true, of course, all around the world. I don’t deny administrative bloat also increases costs but the fact that these observed effects are not U.S.-or west-specific suggests against the common “it’s women being promoted into unproductive sectors for affirmative action reasons” type theories.
Who are they passing the higher costs on to?
Oh yeah, the government.
And every single advanced country features the government as primary payer in education and healthcare. And every single country sees high trend is these sectors, shocking!
And in any case where the government wasn’t the payer we didn’t see massive increases in cost. Plastic surgery and LASIK got better and cheaper, because they aren’t covered by the government.
Most private schools still operate for less than most public schools. In states that have implemented school vouchers, which often pay only a fraction of what the public schools spend per kid, we’ve seen swift disenrollment from public schools. Turns out the private educational market can turn out a competitive product at half or less the cost as soon as you end the monopoly.
So yes I think that in a world where the government wasn’t providing infinite backstop to Eds and Meds that these sectors would do more with less. It’s not an iron law that healthcare has to eat our entire economy. In Singapore where they enforce market discipline healthcare is 3% of gdp.
How does polling line up in this stuff again? Oh yeah women vote to the left of men by a wide margin (bigger government) with especially strong support for Eds and Meds subsidies. We gave women the vote and they voted for the government to fund their own jobs!
Now we’ve got out of control cost spirals in pointless red queen races that are eating our entire economy. But at least every women that can get an education degree (lowest average iq of ba grads) can get a sweet job where you get two overgenerous pensions by 65, summers off, and if there is a bad flu you can decide your just not going to go to work for two years (the men though can’t, they are essential workers).
I’m not sure whether this is intentional obfuscation but the analysis has a pretty gaping hole. Your labor participation rate by education level holds, if you don’t consider the proportion of each population who attained the relevant level. Of course, the number of women who have attained a college degree has increased more than double that of men over that ~30 year period. So women have shifted a significant portion of their population (between 20-30% from what I have seen) from the low labor participation pool to the high labor participation pool and that data tends to align perfectly with the industries recognized as increasingly bloated (healthcare and education, with HR another consideration). If your labor participation numbers are accurate, it’s not clear why this trend is not more clearly represented in your central analysis.
The fact that we can look back to the 1990s to see where women received help isnt surprising. I remember the bombshell report that dropped shocking many Americans: Affirmative Action workplace policies from the 1970s to the 1990s, which was designed to remedy discrimination against African Americans and racial minorities and promoted heavily in the media, primarily benefitted White women.
Guess who has been promoting the idea of girl boss culture since the 2000s? White women and the media.