Why do education, health care, and child care cost so much in America?

Three possible causes.

"health insurance claim form" by franchiseopportunitiesphotos, CC BY-SA 2.0

In a recent post over at his new substack (which you should check out), Marc Andreessen rails against the exploding prices of big-ticket service industries like college, health care, and child care. He posts the following famous graph from the American Enterprise Institute:

Source: AEI

I added the little blue box around “Average Hourly Wages”. Everything above that blue box has gotten less affordable for the average American wage earner since the turn of the century. Everything below it has gotten more affordable (even housing, though obviously not in cities like San Francisco). You can see that the things that have gotten less affordable are all big-ticket services — things where you pay human beings a large amount to take care of you or your family members. The things that have gotten more affordable are all physical goods.

Now, I myself have posted versions of this graph a bunch of times in the past, and I absolutely share Marc’s frustration. In a post back in December, I argued that the rising cost of services was a major reason that the U.S. middle class still feels squeezed, despite the good economy.

But it’s actually not so easy to figure out why this is happening. Marc argues that services are simply simply over-regulated and over-subsidized, and yes, this is one plausible explanation. But the story is probably more nuanced. When Timothy B. Lee looked at service industries overall, he found that there are a lot that have actually gotten a bit more affordable:

Source: Timothy B. Lee

Why should haircuts, laundry, restaurants, and car repair get more affordable over time, while lawn care, day care, college, veterinary care, and health care get less affordable? It defies simple explanation.

So when there’s no simple explanation, we go with a complex one. In this case, there are at least three major factors that are probably pushing up the prices of health care, child care, and college. First, let’s talk about the most basic factor — the tendency of labor-intensive services to get more expensive as the economy grows.

Factor 1: Baumol’s cost disease

The very first thing to remember, when talking about service prices going up, is that service prices are supposed to go up as the economy grows. And they’re supposed to go up faster than the prices of physical goods. This is called “Baumol’s cost disease”, though it’s not really a disease.

Dietz Vollrath has a great post explaining Baumol’s cost disease in layperson’s terms. But let me give a more abbreviated one. There are more capital-intensive industries in the economy, like manufacturing, and more labor-intensive ones like health care. Technological progress tends to make us richer by giving us new physical tools — i.e., better capital. That naturally lowers the prices of stuff that it takes a lot of capital to create — like manufactured goods, but also stuff like transportation. And thus our society become richer.

But labor-intensive industries don’t experience the same amount of productivity growth from improved technology. William Baumol’s famous example was that no matter how rich society gets, it still takes a string quartet the same amount of time to do a live performance. But because society keeps getting richer, people are able to pay more and more money for these labor-intensive services. Someone who makes $100k a year and who likes string quartets will pay a lot more to see a performance than someone who makes $30k a year. So the price of string quartets keeps going up and up.

Service industries in general tend to be more labor-intensive than other industries (at least, until AI really gets going, I suppose!). So Baumol’s cost disease explains why services, even when they become more affordable over time, tend to see faster inflation than the overall price level of the economy at large.

Some scholars, like Eric Helland and Alex Tabarrok, argue that Baumol’s cost disease by itself can explain most or all of why big-ticket services have gone up in price. But I don’t think this is correct. For services like haircuts, laundry, restaurants, and car repair, yes, this looks like a reasonable explanation — these are very labor-intensive services, and their prices have gone up almost 1-for-1 with wages. But the prices of health care, child care, and college have gone up much more than wages. So this can’t be a situation like the string quartet — the increasing cost of an hour of labor simply can’t be the reason that it takes so many more hours of labor to afford health care, child care, and college.

(You might argue that some of these services, like medical care and college education, require specialized labor. But their prices have gone up much more than wages for college degree holders and advanced degree holders too.)

Something is just different about health care, child care, and college.

So we’ll have to look at additional factors. First, let’s talk about Marc’s theory.

Factor 2: “Cost disease socialism”

In fact, I first heard the idea that high service prices are caused by a combination of restricted supply and subsidized demand from…Marc Andreessen. But the best elaboration of the idea is in a 2021 report from the Niskanen Center called “Cost Disease Socialism”. They argue at length that the government’s efforts to make health care, housing, child care and college more affordable by throwing money at them end up backfiring, because the government is simultaneously restricting the supply of these things via regulation.

The economic theory of why this would fail is very simple. If you increase demand for something (by throwing money at it), but you reduce the supply of that thing, you just end up pumping up the price:

This hurts the consumer as well as the taxpayer. The only thing it really accomplishes is to create more jobs. But they’re unnecessary, make-work jobs — the economy doesn’t actually produce more, so hiring more people is just a waste of society’s overall time and effort. To politicians who want to say they created a bunch of jobs, this might feel like a win, but really it’s just a way of having people dig ditches with spoons.

In fact, I do think that cost disease socialism is a big factor in some of America’s excess costs (especially housing or anything involving land use). But when it comes to the big-ticket services that have become so much less affordable, the story seems much more nuanced.

First let’s talk about college. We heavily subsidize higher education, via cheap loans, grants, and state funding for universities. But it’s not a particularly regulated industry. There was a big boom in for-profit universities in the 2000s:

Source: capsee

A lot of these for-profits went bust in the 2010s, but this wasn’t due to some sort of regulatory crackdown — they simply failed to provide value for money.

Meanwhile, no regulation is stopping public universities from increasing supply — as you can see from the chart, they increased enrollment by millions in the 2000s. Private nonprofits (including those fancy Ivy League schools everyone always talks about) didn’t increase enrollment as much, but this isn’t because some law is preventing them from doing so; instead, it seems to be some sort of play for exclusiveness and prestige.

In other words, for higher education, only the subsidy piece of the “cost disease socialism” story is present. (Of course, here there’s also the story that people viewed college as an increasingly valuable investment, so maybe we don’t need cost disease socialism as much for this one.)

Next, child care. This something we don’t really subsidize much — only a few hundred dollars a year for a toddler, compared with about $14,000 in the average OECD country. And yet despite the lack of demand subsidies, U.S. child care is far more expensive than in most other rich nations. It is highly regulated (though not clearly more so than in other countries). So again, only half of the “cost disease socialism” story is present here, though here it’s the regulatory half.

Finally, let’s talk about health care. Here, both pieces of the “cost disease socialism” story are strongly present. The U.S. heavily subsidizes health care, by having the government pay for some of it directly, by giving tax breaks to employers who provide health insurance, and so on. And of course health care is incredibly regulated. So I’d say that this is the service industry where Marc and Niskanen have the strongest case.

But it’s worth noting that pretty much every other rich country in the world regulates health care more than we do, including price controls. And all of them subsidize health care very heavily via universal health insurance systems. And all of them have much cheaper health care than the U.S., while providing similar quality. So in the clearest-cut case of cost disease socialism, the proven cheaper alternative seems to be…actual socialism.

Meanwhile, Marc’s assertion that the use of technology has been effectively outlawed in these expensive service industries doesn’t seem to fit the facts on the ground. We just went through a pandemic where practically every college student in the country was forced to go to school over the internet instead of in person. Some costs were saved, but outcomes dramatically worsened. Massive open online courses have been available for years, but their effectiveness is highly questionable and they haven’t replaced traditional education. Online coding bootcamps like Bloom Institute of Technology exist and are perfectly legal, but they’ve merely added to traditional higher education instead of supplanting it.

In other words, just because education hasn't yet been able to disrupt the traditional college model doesn't mean it hasn't been given the chance to try. Maybe with AI it will finally succeed. Meanwhile in child care I don’t know of any proposed technology that could reduce costs; no one is suggesting that kids be left by themselves and monitored via webcam. Maybe robot day care workers, but we're still a long way from that.

And to say that technology isn’t used in health care is a little odd, since much of the nation’s technology investment goes into that industry, including tens of billions of dollars of venture capital every year. Telemedicine, AI stuff, various digital health apps, and software to increase efficiency at hospitals are all very big deals. I’m willing to believe that regulation prevents all this tech from being used efficiently, but the tech is definitely there.

Finally, it’s hard to see big differences in regulation between the services that have become less affordable and those that have become more affordable. Restaurants are incredibly highly regulated, as is car repair. Cosmetology has one of the most onerous occupational licensing regimes around. Until the pandemic, restaurants didn’t see substantial labor-saving automation for decades, and to my knowledge the hair and beauty industry still isn’t teched up at all. Yet these services have become slowly more affordable over the last two decades.

So the “cost disease socialism” explanation for expensive services seems to be an important factor, but not the whole story — just like Baumol’s cost disease is a factor, but not the whole story. I think we need one more piece to complete the puzzle: luxury effects.

Factor 3: Luxury effects

When I look at Timothy B. Lee’s chart above, one service in particular jumps out at me: veterinary care. The cost of vet care has soared far above wages and most of the other services on the list, rivaled only by the cost of college (and since the cost of college on that chart doesn’t include financial aid, vet care is probably #1). Why?

The reason is very obvious: Pets are a luxury. The news is full of stories about how Americans are spending more and more on their pets as their incomes go up, with affluent Americans spending by far the most. In economics, a luxury is defined as something where spending increases more than 1-for-1 with income, in percentage terms.

Is it possible that big-ticket service items like health care, college, and child care are luxury items? Maybe. Helland and Tabarrok think so:

The evidence is that as income has increased, the share of spending that consumers allocate to sectors like education and healthcare has increased, so the expenditure-share effect will magnify the [cost increase from the] Baumol effect.

This is just a way of saying that as Americans’ wages have risen — due to increased productivity in industries like manufacturing — they have chosen to spend a greater percent of those wages on education and health care, pushing up the prices of those goods even faster than wages. This idea isn’t Baumol cost disease — it’s just a natural change in preferences as people get richer.

The problem with this explanation is that it’s kind of hard to prove, because we don’t actually know how demand will change as people get richer than they’ve ever gotten before. For example, in a very lengthy 2018 blog post, a blogger calling himself “Random Critical Analysis” asserted that America’s excessive health spending can be fully explained by the fact that health care is a luxury item. But to make his curve fit the U.S. data, he had to assume that the U.S. isn’t an extreme outlier on the spending curve — in other words, he had to assume his conclusion. (Note: I had written a much longer explanation of why that blog post assumes its conclusion, but I deleted it in order to spare you a long and boring tangent.)

In other words, maybe America spends so much on health care, college, and child care because we’re just so dang rich, and those things are just what rich people like to spend their money on. It’s hard to know.

For health care and college, the luxury story kind of makes sense. Rich people value their lives very highly and will spend a lot to prolong them. And college is just plain fun, especially if you’re a rich kid and don’t have to work while you’re in school.

As for child care, I can easily believe that nannies are a luxury. But for day care — which is extremely expensive in the U.S. — it’s harder to believe this story. The reason is that rich women are much more likely than middle-class women to be stay-at-home moms at every level of education. If staying home with the kids is a luxury good, it doesn’t make sense that dropping them off at day care would also be a luxury.

But I guess I could believe that Americans pay more for day care because they’re more complacent and don’t bother to price-shop as much. You’ll often hear rich people say that the best thing about being rich is not having to worry about how much you spend. Maybe poorer consumers in Japan or Spain might aggressively price-shop for a cheaper day care center, while Americans might just sigh and pony up a bit of extra cash. Complacency could be an expensive luxury. (In fact, I kind of suspect that the pervasive complacency of a wealthy society is responsible for quite a few cases of overpricing in the U.S., though I don’t have the data to prove or disprove this hypothesis yet.)

So luxury effects have to be on the list of potentially important factors in the mystery of high American service costs.

Putting it all together

There’s no silver-bullet answer for why big-ticket services have gotten so unaffordable in the U.S. But perhaps we’re starting to beat the question down under the weight of partial explanations.

For health care, perhaps it was a combination of:

• subsidies and regulations with little cost oversight from a central government body, and

• the fact that rising incomes make people want to spend much more on easing their pain and prolonging their life.

For college, perhaps it was a combination of:

• the fact that it was viewed as a crucial human capital investment in an era of increasing inequality,

• cheap loans and scholarship subsidies, and

• rich people willing to pay through the nose for their kids to have four years of fun

And for child care, perhaps it was a combination of:

• heavy regulation, and

• parents who were able to fork over cash instead of price-shopping.

This is a deeply unsatisfying patchwork of explanations. It smacks of overdetermination and just-so storytelling. But that doesn’t mean it’s wrong! Health care, child care, and college are complex, unique markets, and there’s no reason to expect that they’d all have one simple cause, or the same set of causes. In any case, what we really need is a lot more research to sort this out.

But if there is one unified take-home message from all of this, it’s that Americans have been too complacent about excessive costs in a number of critical areas. We simply pay far too much for health care, child care, construction, housing, and lots of other things as well. Even if the particular reasons all turn out to be different, we need a society-wide effort to drive down costs across numerous areas of our lives. Whatever the methods, abundance must be the goal.

---