What is behind the cost disease?

I don’t have a similar graph for subway workers, but come on. The overall pictures is that health care and education costs have managed to increase by ten times without a single cent of the gains going to teachers, doctors, or nurses. Indeed these professions seem to have lost ground salary-wise relative to others.

That is just one bit from a very excellent blog post by Scott Alexander.


Is some part of it the associates, assistants, techs, aides, and so forth?

These days, it seems there are doctors, and their assistants, and nurses, and the assistants to the nurses, techs, and aides to them. Everybody who has a certificate/license has some number of people helping them.

The places where employees still seem to get real money (for example software) there aren't typically so many layers of assistants - everybody who actually touches code has to be an A player.
But maybe that's just a weird industry.

Surely the increase in paraprofessionals is tied to considerations which the article mentioned: those extra people (i) complete government and insurance company paperwork and (ii) administer extra tests. (Or, in the hospital, walk people to the bathroom.)

The same holds true for education. Increased specialization, when in decades past the teacher wore multiple hats. Social worker, school nurse, special ed - when I was in grammar school in the 60's, the teacher did all that.

Our grammar school parking lot was sparsely populated. We had one teacher for each classroom, a principal, the principal's secretary, a librarian, music teacher, janitor. That's all. No cafeteria - lunches were made and packed by mom. Nowadays, my local K-5 school has an overflowing parking lot. At
least 60 cars on any given day. I doubt educational outcomes have improved proportionally.

How many costs are because of rules, regs and diktats from their state and the Feds? How much must now be put into the lawyers fees portion of the budget?

" What is behind the cost disease? "

.... if only we had legions of expert professional economists to definitively answer these types of questions for us common folks.

Are economists so under-paid and disrespected... that top-notch people must seek other professions? Why aren't economists seeing more financial rewards for their work?

I loved this piece, and the first possible explanation, that markets might just not be working.

I like economists, but this might be an answer they avoid.

This is where he lost me. How can rising costs in primary education, in which 90%+ of students go to government schools, be a market failure and not a government failure? Or take higher education, where a large proportion of tuition is paid for through government subsidized loans.... Or hospitals, the most regulated industry you could imagine paid for via a heavily regulated and government funded third party payer that was the result of employers reacting to government wage controls during WWII. Nope, government can't be the cause of rising costs there either. Must be something else. (And it doesn't mean something else can't be involved, too, but his dismissal of government failure as a cause when it is the most common thread between these industries is laughable.)


(how do economists miss this obvious point?)

As he says, private systems in parallel suffer the same "costs disease."

I know parents who pay $60K per year for private elementary school. Is it a good school? Probably, but maybe the reasons it is "worth $60K" are more bs that we like to admit.


That's a different beast. Apples to oranges. The $60,000 isn't necessary to get the best education, but it is necessary to signal status and keep the school full of other elites with high status and lots of money. I also know people who send their kids to private schools that spend was less per student than the public schools. Similar with higher ed though part of the issue there is that the state schools are trying to emulate the top private universities not realizing that they can't all be Harvards and shouldn't strive to be.

@anon +1

All we have to do is look at veterinary medicine, where government intervention is basically non-existent. Yet the cost of animal medicine has risen at the same, if not higher rate than human medicine.



Does this mean there's a market failure in animal care or, what I suspect is more likely, that people are willing and able to spend more on their pets?

When I was a kid, we took our dogs to the vet twice. Once to get spayed and neutered. Another time to put them down when they weren't doing well.

Now people get cancer treatment for their dogs. That's going to be costly and result in a dramatic increase in veterinary spending as more and more people prefer to do that over just putting them to sleep or letting them die naturally without intervention. This has nothing to do with market efficiency.

The issue with private schools is partly status signaling, and partly ensuring that little Biff and Muffy (yes, I'm being deliberately and sarcastically stereotypical) go to school with with other kids whose parents care about status signaling as well as quality education in general, at a school where any disruptive influences can be summarily kicked out. If the local public schools cannot provide, then there's a market for private schools that will. Even one disruptive child per classroom takes a significant toll on the other kids.


I'm not necessarily saying your explanation is wrong. You seem to be saying that human healthcare costs are rising in large part because of government interventions distorting the market. Whereas animal healthcare costs are rising because of cultural shifts in the relationship between humans and pets. These represent two exogenous trends, that just so happen to be producing similar trajectories in their respective cost curves.

It's certainly plausible, but strikes me as extremely coincidental. Occam's razor would suggest that two concurrent series that look very similar are most likely being driven by some common X-factor. Since government intervention doesn't apply to veterinary medicine, that can't explain the X-factor. That's not to say that government doesn't distort (human) healthcare costs, only that the large bulk of the long-term trend probably isn't attributable to a change in government distortion. It's perfectly plausible that government made healthcare 5 times more expensive in 1970, as well as 5 times more expensive in 2017. The net impact is even larger, because it's distorted from a higher base. But that doesn't really explain what's the percent increase from 1970 to 2017.

@Doug: "....look at veterinary medicine, where government intervention is basically non-existent. "

Incorrect. U.S. veterinary practice is heavily regulated -- though noticeably less than the staggering level of government intervention in human health care markets.

Try opening your own local veterinary clinic ... without any government licenses or approvals --- your "non-existent" regulators will swarm you like sharks to chum.


Fair enough point. I'll quickly admit I know little about the details of the veterinary business. I may be reasoning from a cartoon impression of how veterinary medicine works.

I think it is kind of self-evident, as we look around a mass consumption society, that people are herd animals, and will pay more If neighbors, or role models are. That is true of state college and private middle schools.



Surely not if you can get the same thing for less? How is it losing to the Jones's to get the same things as them for lower prices? Surely that's the definition of winning?

what markets?

"Is some part of it the associates, assistants, techs, aides, and so forth?"

That's the most obvious answers. Schools have a substantially larger staff than they did historically.

I don't see it quite that way, there are validation engineers, trainers, project managers, documentation writers and so on who don't strictly need to be there. There are also all the bangalore tech workers and junior engineers large orgs are eventually forced to accept. But there is competition, so the org as a whole has to be efficient, and I guess a certain number of Indian workers can reduce the number or salary of A players required. It also isn't new Fred Brooks proposed a surgical team type of thing with IBM back in the day.

Working in software, I see the exact same dynamics here, too, although in early stages.

In mid nineties, 90% of our software stack was developed in-house. Lots of (incorrect) specification and (outdated) documentation was required for anything to meet the definition of done.
It used to cost X amount of dollars and Y amount of man days to finish a certain project.

In 2017, 90% of our software stack is high quality, free of charge components. Specifications or documentation written only when necessary (rarely). Level of developers' craftsmanship is *much* higher than before. Integrated development environments are way more powerful.
We spend much less time waiting for computer, we can search for an answer in seconds and often find it.
Yet, it still costs *at least* X amount of dollars and Y amount of man days to finish that same project.

So where does all the spare time and money go?
* User Experience designs, reviews, discussions, the results of which are at best marginally better than what a developer would come up with. Often done in products that are intended to be used by few people anyway
* Useless certifications and trainings
* Management flying across the globe to attend face-to-face meetings that could perfectly well be done via a Skype call
* Developers worshiping their own test-driven religions
* IT/Network administrators that get in the way more than they help
* Consultants, who are often up to 5 times as costly as in-house employees at the same skill level. Management prefers them, as they can be fired at will
* Powerpointers of various kind

In general, the teams actually touching the code itself are shockingly small. A general rule of thumb might be - if a SW company of 1000 employees announces a "new flagship product that will conquer the market", expect something like 7-10 people genuinely working on the product.

"User Experience designs, reviews, discussions, the results of which are at best marginally better than what a developer would come up with"

I doubt understanding what customers want is going to give you only a 'marginally' better product.

What you refer to is called "requirements analysis". User experience design has very little to do with it. It focuses on making product prettier and easier/intuitive to use.

I saw a comment on Slate Star Codex that I thought was interesting, but didn't get any responses:

"To offer a counter-point: It may not be that things have gotten [relatively] more expensive, but that certain things – specifically those used to calculate CPI – have gotten [relatively] much cheaper.

It’s undeniable that food and manufactured goods have become immensely cheaper thanks to technology/globalization. If you use them as your metric for price changes, then in comparison other prices will seem overvalued. If you instead calculated CPI based on older technology (say using the price of organic farms or handmade toys), you’d see much less of an increase. So the story might not be that health/schools/etc cost more, but that they cost about the same and that food and consumer goods cost much, much less."

I was wondering if anyone here might have thoughts on this?

I'm no expert on CPI, but is it undeniable that food and mfg goods have become immensely cheaper?

Here's what I could find on food

I don't know the data well enough to draw any conclusions, but it's not obvious to me. I don't know how to evaluate mfg goods, but it's not clear to me that things like cars and washing machines are cheaper.


I don't agree with the original assertion, but you're begging the question by normalizing to conventional CPI. The original assertion is that healthcare represents something closer to "true inflation". Because it's a sector where productivity changes are minimal, it's a much better proxy of actual monetary inflation, unaffected by gains from production efficiencies. If that's the case (which I'm not saying it is), then "true monetary inflation" has actually averaged 6-7% instead of 2%. If food prices have risen 3% a year in nominal terms, we'd say food costs have risen 1%/yr in real terms, assuming we're using 2% CPI. However if inflation has actually been running at 6%, then a 3% nominal rise, actually represents a 3%/yr decline in real terms. Which represents substantial productivity gains.

Again, I don't agree with this hypothesis (see my comment below). But simply looking at a nominal price series, or even "consensus" CPI, doesn't refute the original point. The assertion is that inflation is substantially under-estimated, so anything that grew less than the cost of healthcare would have represent real productivity gains in the framework.

I'm not sure about food and mfg goods specifically, but there must be something that is relatively cheaper.

The CPI is calculated as the average change in price levels. If healthcare and infrastructure costs are rising much faster than inflation, there has to be something out there that is getting relatively cheaper. We can't have a Lake Woebegone economy where all prices are rising faster than inflation. I think the original point is correct. If all we know is that some prices are rising much faster than average price level, it's impossible to say whether the explanation is (1) that there is some factor specific to the highly inflationary goods causing their prices to rise or (2) that there is some factor specific to the lower inflation goods dragging down the CPI. I suspect it's some combination of the two.

By the way, I think Scott was far too quick to dismiss the explanation that government involvement is the culprit. For one thing, he underestimates the extent to which pricing in the education and healthcare sectors are driven by government. Even though private colleges and hospitals aren't directly run by the government, the influence of the government in those sectors is huge. Areas of healthcare that look more like a normal private market (like Lasik and other elective procedures without third-party payors) have not seen similar cost increases. Increasing costs of veterinary care is an increasing counterpoint, but I think that's just a separate trend of people becoming much more pet-obsessed (I'd be curious to see if there is a similar trend in livestock veterinary costs). Also, another perfect example of an industry where costs are spiraling out of control is the military. Unfortunately, it's tough to measure the efficiency of military spending with something like CPI, but I think it's clear that the costs of the U.S. military have grown much faster than inflation and have also gotten more expensive relative to other countries.

I think this phenomenon is a combination of three factors: (1) technology and trade have massively reduced the prices of a lot of goods, which lowers the CPI and causes other goods to appear more expensive relative to inflation, (2) sectors with more government involvement have a tendency to be less efficient and to become more expensive over time and (3) there is something uniquely dysfunctional about American politics that causes anything our government touches to be far more expensive than other countries. Of those factors, #1 is a good thing, #2 is largely unavoidable but a good reason to support smaller government and #3 deserves a great deal more study (as I think it's tremendously important and I have no idea what's causing it).

Agricultural raw material prices have risen much less than the CPI.

But food at home and eating out prices have risen about the same as the all items CPI no matter how far back you go.

Our food purchases are much more processed foods and less plain raw materials than it was a generation or two ago. Moreover, the variety has massively improved.For example, we can now buy fresh vegetables -- imported from Latin America-- all year long.. This does not show up in the CPI or most measures of consumer well being.

That's a plausible hypothesis, but viewed through that frame of reference, median wages have also gone down tremendously. It's still the case that the median person spends at least six times more of their paycheck on healthcare. If healthcare is a closer metric to "true prices" then manufactured good, then that means median wages have fallen by around 80%. It also means that overall GDP has crashed since 1970, since the price deflator now averages 6-7%, and nominal GDP has only averaged around 4%. It would mean that the economy has literally been in recession 90% of the past 40 years.

The only reason we're not all starving in the street is from the miraculous gains in manufacturing productivity and automation. But again, in this framework, that process is largely exogenous to the terrible macroeconomic situation. If those gains slow down even a little, and the macro trends continue, we're probably facing imminent economic collapse. So maybe this is a plausible hypothesis, but it certainly comes with a whole lot of extreme implications. I think medicine/education specific cost disease is a much more likely explanation.

You have to be careful and consider total compensation: pension, health insurance, etc. are typically pretty good for these employees. Still, overall he's right and I suspect that every single thing superfluous to a genuinely good education: fancy IT, administrators, guidance counselors, etc. all push things upward.

The sectors that are most regulated, most government-controlled, the most Unionized have had their costs rise the most without any improvement in the product?

I may be naive, but I would have thought that more regulators will result in more regulation requiring even more regulators. Public choice theory. Union officials and government bureaucrats aren't there to reward teachers, nurses or customers. They are there to get big fat pensions.

Two words to spoil your party: veterinary care

Keep thinking :)

Last year, IIRC there was a post at MR referencing a paper from NBER -http://www.nber.org/papers/w22669 and a little flurry of articles on how pet care was becoming more expensive. These seemed to focus on the growth in total spending rather than the cost of specific procedures (I couldn't access the paper, so I could be wrong).

Pet insurance has not really caught on, so most pet owners are paying out of pocket, and how much a pet owner will spend is a function of his attachment to a pet and his income/wealth. Wealthier people are willing to spend for ultrasounds, MRIs and surgery for their pets. People of moderate means are much more likely to euthanize or let their pets die when the cost to keep them alive becomes high relatively to their income. I've seen a 17 year old Schnauzer getting chemotherapy in NJ and a 13 year old cat with diabetes left to let the disease takes its course in PA.

We have two dogs and recently one (a 10 year old lab) was diagnosed with a cancerous anal gland. We went back to our old vet in NJ and he did the surgery, but only after he confirmed by x-ray that the cancer had not spread throughout the body. He also cleaned our other dog's teeth (under anesthesia) and the total bill was $1,700. All follow up visits were at no additional charge. I think the charge was reasonable, but we could have decided not to proceed.

I am skeptical about excessive inflation in pet care - I think that owners are just willing to go further in keeping their pets healthy and alive.

I agree: i am always pleasantly surprised when i take my maine coon to the vet. even something like castration - which i thought to be a relatively complicated procedure - costs only 85 euro (in the netherlands). Meanwhile human health care (insurance premiums) rises at a healthy 5%-10% per year (w/o a corresponding expansion of coverage).

even something like castration – which i thought to be a relatively complicated procedure

I once had a conversation with a vet while he was doing one. Given the amount of time he spent making eye contact with me, it's pretty clear he could have done it with his eyes closed.

To the extent costs have grown, they have just leaked over from humans like the capex of an MRI machine. It may cost a few thousand, but they don't take 10k a year for doing nothing like insurance companies.

Couldn't you make the same argument for healthcare? A huge proportion of total spending comes during the last several years of life. Maybe in the past these procedures were limited to the wealthy living near top reaearch hospitals. We invented this industry of expensive old people care.

So the solution could be to cap medicare spending after a certain dollar amount or age, but then that sounds a lot like a "death panel".

So when I take my dog to the vet how is it government controlled and regulated? Standards, yes. But they only get paid when they provide a service that I decide is worth my money. If they don't provide timely, affordable and high quality service, they don't get money.

Supply is artificially limited. Naturally that will make prices artificially high even though, ceterus paribas, each transaction that occurs is mutually beneficial

Since January 2000 the CPI for veterinary services has increased from
100 to 224.1 while the CPI for physicians services rose from 100 to .160.2.

January 2000 is the first observation for veterinary services.

Pochemuchka, what was the point you wanted to make?

Seems suspicious. What is your source?

I found this which claims CPI increase by government is way overstated. Note differences in methodologies.


Veterniary care has followed a very different trend line. Total spending has increased because people are now electing to spend thousands on their pets' health, when that would have been unthinkable 20 years ago, but the cost for basic procedures has not increased. See things like LASIK costs for a better comparison.

I have heard, but do not know, that it is harder to get into veterinary school than into medical school. Is the one-off nature of veterinary care a byproduct of restricted supply of veterinarians?

I am very familiar with both large and small animal vets, my ex wife was an extensive user of both. Veterinary school is just as expensive as med school and yes it is more competitive. As such, the people that are admitted are almost exclusively very smart, driven, upper-middle class women. They go into veterinary because they can afford it and because they love animals. Once they are are vets, within 3 years the large animal vets become small animal vets, because large animals break outdoors in bad weather, which is a hard job. Five years after graduation the small animal vets have either given it up because being mom has started, or they have moved on to a specialty or multi vet practice that has office hours of 30 hours per week. The older vets worked 60+ hour weeks, the new ones half that, minus the mom attrition, so there is increasing demand over a smaller hour base, so they can charge more. There are always outliers, of course, ymmv. There is a lot of pressure to engage in "maximal intervention", which drives up costs. Many "old school" vets or those running economically-matched practices in low-income areas have run into discipline problems with vet boards for not using "enough" anesthesia and/or doing triage instead of intensive care.

That is the obvious answer.

That is not *the* obvious answer. It is merely *one of the* obvious answers.

Alexander does a great job listing the possible or probable causes of the "cost disease" with respect to the three areas he discussed (education, health care, and subway building). To me, the "most obvious answer" is "all the above". There are probably more (I would add, e.g., the fact that while pensions have not "obviously" gone up, the cost of paying for them, past, present and future, obviously have). It always amazes me how, with respect to issues like this, with multiple variables, people (I'm looking straight at you economists), want to choose just one as "the cause". It amazes me even more, given the alleged advanced state of the "science" and the amount of money we throw into *that*, that there is so much doubt about what has caused this tremendous increase in cost in two of the largest segments of public spending (health care and education). It is not that discovering the answers to those questions requires some great advancement of theory; this seems to me largely a matter of close analysis and arithmetic.

So, why the alleged uncertainty and the inclination to seek *the answer* as if it were only one thing? If I may give *the answer*, I think it boils down to the fact that the "science" has been infected by politics. It is also extremely unsexy for an academic to give twenty-five causes of a particular outcome. That doesn't fit nicely into the headline news and is a bit messy when trying to line up behind a particular political faction or interest group. For example, I'm sure that increased exposure to medical malpractice tort liability has contributed to the overall health care "cost disease". But, when pressed to do something about it, the normal answer from opponents (e.g., the trial attorney lobby and their political beneficiaries) is that "well, that's not *the* cause of the problem; it only accounts for x percent!! And, so it is, down the line of all those multiple contributors to the problem such that nothing at all is done.

Interesting comment by jhertzlinger regarding why US subway construction costs 10x as much as in Europe as it's difficult to blame regulation.

"One possible theory: The sort of people who keep costs down in the public sector in Europe are working for Silicon Valley in the US."

we have a tech sector as well you know...

American governments are happier to give money to large non-defence corporations (more so at the state level).

More broadly responding to SSC - is the issue More Knowledge? That is, are we now in a world where everybody knows about things that go wrong (the media and now of course the web) and the qualities of every feature of every competing institution, product, etc., are well known to both producers and consumers?

So if *everybody knows* that going to Tier A College adds 40% to expected life earnings, does this cause (a) everybody wants in to Tier A regardless of costs, and (b) every other college has to try to match every feature of Tier A regardless of relevence? If *everybody knows* (or thinks they do) that hospitals with tracheotomy kits in every room are better, does that force all hospitals to have tracheotomy kits in every room? (See current Seattle Times article about a tragic death apparently partly caused by not having the right airway restoration tools at hand...)

If every provider has to be 100% complete in 100% of all things all the time (whether relevent or not) that would tend to drive up system wide costs.

I think so, and in education the fallacy is compounded by an old data problem. Education differentials of all kind were higher when there was less education.

If that were true, shouldn't everyone would have learned that education is wasteful signalling by now and adjusted accordingly?

Besides that, I don't think it explains much about Healthcare cost inflation. My sense is that the industry has gotten significantly more complex over the last thirty years, for a variety of reasons, and that the average person hasn't kept up with this at all and the industry is basically opaque to them.

Finally since I'm on a roll -

What are the costs and compensations as fractions of GDP? As fractions of available income? Is there some effect where yes healthcare grows from 10% to 15% of GDP, but GDP grows by 4x, and so healthcare looks 6x more expensive in "real" terms but is really only 1.5x more expensive?

Should all of these discussions be driven solely by percentage of GDP, or percentage of median discretionary income, or other such measures?

Given that in the end money always goes to people (it's a human construction) is the real effect that you spend between 7% and 15% of your net income (money, life wealth, status points) on healthcare, and that's just the way it is? That economic balancing forces mean that large groups (like teachers) will never make more than say 1.2x the median wage, no matter what? But also that education will never cost less than 4% nor more than about 12% of GDP?

Taking health care as an example, or biomedical research, or chemical research - we can do any of these - what happens when
Person A decides to buy, Person B decides to charge, and Person C pays (grants, insurance, etc) is that A has no inherent limit.
B has no inherent limit. C has no power. Eventually, of course, this all loops around through society, but there is slack in so many
of the joints.

So, money tends to accumulate to B. But that makes B a very lucrative place to sit. So gatekeepers emerge, extracting some of
the value from B before that person attains the position (medical schools) or creating kinds of supplies that are 'necessary' or
providing 'regulatory support.' And charging what the market will bear. So eventually, the surplus is largely extracted out of B,
and B is still a fine thing (being a doctor, or research institution, or whatever the unit of selection is) but there are all these hangers-on.
If any of them manage to extract more value from the proposition, they get their own epiphytes. Finally, there is this great sucking
sound as the prices B has to charge keep going up and C has to pay more and more but can't effectively negotiate because B
has great evidence of how much everything is costing _him_.

Look at the granting institutions, or the insurance companies... they are stuck in this cycle. They need to extract more and more
from society to fund this tangle of suppliers.

The only way to break the cycle is to break the loop (make A pay for what they want). Unfortunately, that obviously breaks the whole
process and suddenly B can't afford anything, the suppliers go belly up, and B can't get what is needed to do the job to current
specifications and standards. If all the medical schools shut down... so the specifications would need to be rewritten to meet the
real, funded demands of people who can pay. This would take time with tons of frictions, and the regulations would be a mess,
and probably nobody would get much healthcare or education or research or whatever for a while, and nobody would want to
enter those professions and ...

so the kinetics of a solution are missing.

I endorse this comment.

I made basically this same comment at Scott Alexander's blog.

yes for healthcare, education.

what about vet bills?

1) wealthier people with fewer children can and do spend more on their pets, enabling longer lives
2) due to 1, new technologies, medications and procedures are developed
3) new vets work substantially fewer hours than previous vets
4) professional and social "minimal care" ratchet effect

i don't see why prices wouldn't go up if there is value. take college education or medical care. if people believe they are getting value for these services and don't perceive that there are alternatives the price will naturally go up as people will be willing to pay for them.

Pick some consumer product, like headphones or vodka. There is a price quality curve, but it levels out. $1000 headphones are not 10x better than $100 models.

If too many people buy the $1000 model, for small psychic or auditory benefits, you have a cost disease. People just don't care as much in consumer goods. It might even be an economic "benefit" that people buy too much crap at Costco. But is it?

This is not the reason. A Harvard degree costs 3x as much as a state U degree, but it's more than three times better. Similarly, state U costs about the same as State U inferior branch.

What it is is administrative costs.

That Harvard degree is more than 3x better only because too many people have convinced themselves that Harvard and other "prestigious" institutions have magic pixie dust that graduates get sprinkled on them that makes them better than an equivalent student graduating from State U. This becomes a positive feedback loop where students think they must go to the most prestigious college available so that they can get their share of the magic pixie dust. In reality, smart and motivated students will do equally well wherever they go as long as certain minimum standards are met, which are almost certainly met by all flagship public state universities. The Ivy League is a shuck.

Is Harvard normally considered part of the "cost disease" though?
I am not sure luxury products are. They have to become mass luxury to move the curve on average cost.

Chefs buying Cuisinarts is one thing, all of the middle class buying one and sticking in the closet is another.

Put more simply, if it is a feature of "consumer capitalism" that people buy a lot of crap they don't need, is it surprising that they buy more crap than they need?

A rational person might spend less and meditate more, to achieve the same net happiness.

Maybe you should comment less and meditate more, eh?

That is worth thinking about.

a) There's a price signal from the public option; people will expect to spend more to get something better from the private option.

b) Even if it's feasible to deliver better education at a fraction of the cost, signalling and competition for talent won't make that common. It's possible to offer fantastic shoes for $10, but good luck finding that in the market.

Founder of Red Bull, explaining why he charges $3 for a 10-cent drink: "How will they know it's a premium product if you don't charge a premium price?"

Do any professors teach from old versions of textbooks? For instance could a professor teach from a used version of Mankiw's Principles, and save money for students that way? For instance a fifth edition of Mankiws Principles costs $10, whereas the latest edition (7th) costs $287 new. Nothing has changed in the principles of economics from edition five to seven.

This gets tricky. You can't guarantee that enough used copies of old editions will be available. Interestingly, the rise of the used markets may increase the price of new books as the publisher sells fewer new textbooks and has to charge more per book to recover fixed costs. It also encourages the publishers to come up with new editions more often to make the used books seem out of date. I'm hoping we can get a more robust and efficient e-book market going.

"Do any professors teach from old versions of textbooks?"

Yes, for exactly the reason you describe.

I studied Business Administration 101 in Italy on photocopies of microfilmed copies of a 1929 textbook, the venerable Zappa. It was actually the prescribed book, not a student's workaround.

Higher ed: Some combination of administrative bloat, fancy facilities, and increased personnel costs not captured by raw salaries. One possible contributor Alexander doesn't mention -- a decline in average teaching loads for faculty at research universities. Which doesn't translate into faculty free time, it translates into a lot of hours producing and publishing research that is little read and cited. We waste a lot of money on ZMP research papers.

K12: Massive expansion of non-classroom positions. Smaller class sizes. Earlier retirements. Far more expensive pensions and medical benefits (including retirement health benefits). The cost of pensions has exploded due to a combination of A) greater generosity, B) earlier retirement, and C) very low investment returns (in response to such results, CalPERS is bravely dropping it's return expectations from 7.5% to....7% -- and that only gradually).

I would exclude housing from his analysis, however. In the country as a whole, when you control for household size and amenities, prices have not increased beyond inflation. Alexander asks:

Or, once again, just ask yourself: do you think most poor and middle class people would rather:

1. Rent a modern house/apartment
2. Rent the sort of house/apartment their parents had, for half the cost

People might say #2, but their revealed preferences are very much #1. Not far from where Alexander lives and works is this place. It's at least as nice now as it was then and is not 'crime-y'. During first decades after it was built, it was full of families with kids. Now there are very few -- families simply won't live in 2-3 bedroom single-bath homes where kids have to share bedrooms.

A couple of things strike me here:

1. How much of this is due to a change in composition? Healthcare and education are both becoming much more metropolitan, cities with good schools and hospitals appear to grow more than those without. It is my impression that an ever larger fraction of the healthcare and education dollars are being spent in expensive locations like NYC, Boston, DC, California, and Seattle while less are being spent in lower cost places like Toledo, Lansing, or Iowa City. Logistics are cheaper, but compliance costs are higher in cities and any growth is having to pay higher costs due to a lack of open land for development.

2. How much inferior are the inputs today? For instance, in 1970 about 10% of children lived with a single parent. Today it is around 30%. I know many admirable single parents, but it is virtually impossible for one parent to have as much time for early reading, homework help, and the rest as two parents. In theory they might be able to contract out such efforts, but most single people, not surprisingly, have lower incomes than those who are married. Evidence also suggests that children of single parent families are more likely to have discipline issues. We tripled the proportion of these students so ceteris parabis would should be expecting a decline in educational performance.

But there are many other broad social issues that may be decreasing the non-formal educational support of today's students. For instance, in the past it was quite likely that child with an interest in math would know an engineer through church, the neighborhood, scouting, or some other social interaction of their parents. That is no longer the case, we have sorted very heavily so that it is obnoxiously common for doctors' families to only socially interact with other highly educated families. The general Bowling Alone phenomena diminishes other supports for students as well.

On the medical side, I can unequivocally say that, outside of smoking, patients start less healthy today. The average weight of patients coming through the ED has continued to increase, as has the pediatric patient weight. This is correlated to a huge number of disease states, it is astonishing that we have the same life expectancy today as back when we had a leaner population. Yes some of this is offset by smoking reduction, but we have only cut the smoking rate in half and the top age brackets smoked a huge number of pack-days before they quit. And obesity is far from the only health determinant that has decreased.

Far more patients live alone than ever before. This means that hard to visualize melanoma is caught later. Depression is more frequent. Stress levels are reported to be lower. We do not know all the mechanisms, but married people have substantially better objective health measures (e.g. lower rates of cardiovascular disease) and lower all cause mortality risk. Again, we also have the social side of things to consider. Patients who attend church/synagogue/mosque regularly are significantly less likely to die from all causes; while not viable policy, the naive math says that compulsory church attendance should do more for life expectancy than curing cancer. People report fewer friends and doing less activities with them. And of course there are the vices, today patients have done more drugs throughout their lives (drug use at 20 shows up in health bills at 50), drink more (a risk factor for pretty much everything), and have had more STIs.

3. We expect more professionals who have "authority" over other people. Back in the day, I am told, that if drunk patient swore at an attending three times the unofficial rule was to have a resident practice an intubation. Teachers 40 years ago were allowed to belittle students in class for lapses in student conduct. Today, we all have to play nicer. This takes time and effort. Doctors get whole courses on how to be more polite and respectful, we spend countless man-hours being measured on how well we follow the material, and if we slip too low on our numbers we get to go burn more resources to try to improve them. Ultimately, having better behaved healthcare workers means having more of them - having a shoulder to cry on takes time and money. In education, I suspect similar dynamics are in play, convincing a student to turn in his homework through soft persuasion may be better, but it takes an awful lot more time. This is made worse cost-wise by large organizations getting dedicated staff to make people feel happier while interacting with the organization. Once you have a referral service, it becomes exceedingly easy to just refer everyone if it is vastly more expensive.

I suspect that people just do not realize how incredibly different modern society is from those of days gone by nor how we have shifted a large amount of work from informal sectors to formal ones like education and healthcare. There are other dynamics in play but healthcare and education are no longer the same services they used to be. We have different inputs, we have different compositions of inputs, and we have different outputs. I wonder if society has just gotten more expensive at providing human intervention products (like healthcare, education, legal services, bureaucracy) because society has optimized for producing all the other products.

Very insightful. Not only do the inputs changes, but also how we see or treat things we used to ignore or not see.

Can't find any evidence online but i would doubt that 1.is the reason. In my state, it's the marginal State U branches, located in marginal areas and marketed to marginal students who would not have attended college in the past, which have grown the fastest.

There are a few factors at play here, but in general the mechanism behind dramatically rising prices in most of those sectors is a phenomenon you could call "Neo-Ricardianism" (credit: Arnold Kling). That is, given the current organization and structure of the economy, the following sectors enjoy a special ability to extract rents from the surplus income of workers, because acquiring their particular good, service, or credential, is something akin to a "license to live and work". Consider these sectors:

1. Real Estate (especially near the center of an industry's centralized hub, where most of the income in that sector is earned, e.g. Wall Street, Silicon Valley, DC Metro). That seems like the old Ricardianism, but note now that it's not the agricultural production of the land itself, but the geographically-tight web of high-productivity economic relationships that happen to enjoy economies of agglomeration in a particular spot. That spot may be arbitrary and due to historical contingencies, but it will enjoy huge amounts of inertia due to the challenge of social re-coordination.
2. Health Care.
3. Education.
4. Government

To oversimplify, one increasingly has no choice but to pay these pipers what they want.

More and more, one has to live near the center of a few hubs, because these are where the jobs are increasingly concentrated. The explanation is due to technological progress and I'll defer it for now. Central real estate is quasi-zero-sum - supply inelastic for both physical and regulatory reasons - and so can be bid up to the edge of one's income surplus.

If one is stuck in a particular small set of urban areas in order to have any shot at affluence, then one can't live there without paying the government whatever it wants to extract in the form of taxes, up to the point of totally predatory "stationary bandits" who deliver very little value to taxpayers in exchange for their revenues. And if one if a captive audience for whatever that government provides (or allows for) in terms of mass-transit opportunities, or else suffer extra additional hours of crushing commute, then there is created a terrific opportunity for additional rent-extraction that creates an almost irresistible political temptation and which will surely come into existence if one's local institutions are of low integrity and poor quality.

If one can't get easily get a good job without a "probable good worker certificate" credential issued by an higher educational institution, and furthermore there is a zero-sum status competition for prestige among those institutions then, again, they can charge up to the limit of one's expected future surplus in exchange for that credential.

And that medical care is like a license to live a life normal and healthy enough to earn a living is pretty obvious. Here, the opportunity for rent-extraction would not exist in a normal, competitive market, but the regulatory interventions in the sector are so intensive, with restrictions in supply and many subtle transfers and cross-subsidies, that the market behaves more like a Ricardian sector than a normal service industry.

Note that the identity of the individuals who are the prime recipients and beneficiaries of these rents is not always obvious, and certainly not always the nominal owners of property.

"More and more, one has to live near the center of a few hubs, because these are where the jobs are increasingly concentrated. "

That's really not true. Central cities have accounted for an ever-decreasing share of the population of their metro areas over the last half century and jobs have moved out as well. Less than 1/3rd of those who live in the Chicago metro area now reside in the city. Detroit is even more extreme -- in 1950 half of metro area residents were in the city. Now only about 1/7th are.

Also, the top metro areas from the middle of the 20th century (New York, Chicago, LA, Philadelphia, Detroit, Boston) have grown more slowly than the up-and-comers (Miami, Houston, Atlanta, Phoenix, Dallas, Denver, Seattle). So it's just not the case that jobs and population are becoming increasingly concentrated in city centers or in only a few metro areas .

"Less than 1/3rd of those who live in the Chicago metro area now reside in the city."

Focus on work, not bedrooms+commuting. Especially where most of the highest incomes are being earned. The dramatic Chicago skyline is as good a visual demonstration as any of the typical power law in the density of labor productivity.

Another relevant Klingism: These are all sectors in which government restricts supply and subsidizes demand.

Here's the post on the subject that *isn't* self-serving blind spot bullshit: The cost disease.

Very good.

Agree everyone should read this link. It is better than any comments or responses written on Scott's site or the comments here.

A few added comments before people follow the link. The overall inflation rate, adjusted for quality is pretty much zero. That means that these areas -- education, health care and government infrastructure and left-coast housing are the anomalies to be explained. Absent these increase areas, cost of living is surely dropping. Within these areas (dismissing housing per square footage which is not even getting more expensive except in a few screwed up cities) we see not just the cost disease (lack of productivity gains) but rent seeking, bureaucracy, bloat and exactly what one would expect in a market where the people using the service are not the ones paying directly for it, and/or there are negative dynamics (Scott refers to this elsewhere as Moloch or vicious cycles).

In education we are seeing something where the dynamic attractor is not educational outcomes but signaling impacts that you attended school with the elite (kids able to afford or whose parents can afford to send them to school with other elites). This, when accompanied by other people paying is guaranteed to end badly. The attractor is toward exorbitant prices. Health care has a similarly dysfunctional dynamic as the link explains.

Let me be succinct. Markets work when there is constructive competition to supply better products for less. When the dynamic is off, as it is in government contracting, schools and health care you get a different attractor and Scott has already written extensively on it, and its name is MOLOCH, aka the vicious cycle.

Well that was something different and refreshing. I like the idea of the "Trumpenreich".

It is a waste. No real understanding and no solution, just rant.

+1. Excessive rant mode is on at the link.

from the link:

Restore marriage 1.0.

why didn't anyone else think of this

I'll only weigh in on the cost of college as that's what I have first hand experience with. Twelve years ago one of my daughters started off at University of Rochester with a $4K merit scholarship (since we could pay the tuition, it was a modest but welcome stipend). Near the end of her freshman year we get a letter from the President saying that because of the need to keep and attract outstanding faculty tuition was increasing by 6% and there was a smaller increase in room and board. No offer was made to increase the scholarship money by a like amount. We received similar letters each of the next two years (maybe they had a form letter that was automatically sent each April). IIRC, by her senior year the tuition had risen to about $41K. The $16K in scholarship money that she received covered all of the increases in tuition as well as room and board.

The university graciously sent us a financial report every year which I read with a great deal of interest. Rochester has a hefty endowment and over the years received significant funding from the old Rochester industries that alas are pretty much gone now (Xerox, Bausch and Lomb and Kodak). the enterprising researchers in the medical school developed several childhood vaccines that resulted in hefty royalty payments from several pharma companies. Now this endowment is not in the same league as several of the Ivy schools but is still quite substantial for a school with an undergrad enrollment of 4000 students and maybe 3000 students in the grad and professional schools. Was any of that money used for scholarships? Not really apparent to me. It's strange to see University endowment scorecards being reported as though there is a contest for which advisor is doing the best. I think Yale is sitting on something like $38B right now and yet fundraising letters still go out to alumni as though the university is in a state of perilous poverty. Strange behavior.

Indeed. Conan O'Brien once said that Harvard's fund-raising slogan was something like "We don't need your money, but we want it." Having large and growing endowments seems to have become a goal of college administrators, having nothing to do with the other things colleges do.

One wag described Harvard, given its vast endowment, as a "hedge fund with a university attached."

"...Yale is sitting on something like $38B right now and yet fundraising letters still go out to alumni as though the university is in a state of perilous poverty."

And those alumni contributors get a charitable contribution tax deduction so that, in effect, all other taxpayers are contributing to a university that would not accept their kids.

As a country's per capita gdp grows beyond the need to feed and clothe itself, what do you expect that output to be spent on? IE what goods and services get consumed by the extra per capita wealth? Looks like education, health care and housing are certainly being consumed more. As they get bid up the marginal returns to the incremental expenditures goes way down. So as we add more extra-curriculars to schools, and more technogy to health care, and stainless steel appliances to kitchens, the cost goes up out of proportion to measurable impeovment. The question is will there be compelling alternative goods and services as society gets wealthier to displace spending on these items? Another obvious policy question would be if these goods and services were not taxpayer funded would spending go down and savings or alternative consumption go up?

What about just plain leisure? If the economy produces everything we need, with increasingly less work, why aren't cutting to six hour workdays, three day weekends, 12 weeks of vacation, and five maternity leave? Doesn't that seem a little bit better than health technology that doesn't add value, school administrators for programs that no one uses, or "prestige brand" kitchen appliances? I think 90% of people would accept that tradeoff.

There's something seriously socially broken, if we're all working to produce crap that no one needs, just because we've "got to make something". This may sound socialist, but it's not. It's a serious coordination problem. The vast majority of people would readily accept 1970 schooling, 1970 medicine (with modern cheap technology like ACE inhibitors), 1970 housing at 1970 prices, which they could afford by working three days a week. The problem is that those 1970 era products and prices have completely disappeared from the market. That's not just a lack of progress, but a serious regression.

The secondary problem is that there are few professional careers that let somebody work 25 hours a week for 5/8 of the full-time pay. Why? Mostly because of signaling externalities. We'd all like to work less, but want to impress employers as hard-workers. The employee that actually asks for 25 hours is dis-proportionality likely to be a slacker, who's not getting anything done even during his reduced hours.

Scott makes some valid and interesting points, particularly about risk tolerance. Sometimes he is trapped in erroneous comparisons.

For example he asks about whether people would chose the 70s-80s level of service for that era's price + a rebate of the cash. In the case of housing---the clear answer is "No"---people are building bigger houses. Home prices from repeat sales (which to some extent correct for this) show price appreciation.

Malpractice lawsuits was a big issue in the 70s and 80s. I remember as a teenager in the late 70s reading articles on the malpractice lawsuit crisis.

He acknowledges that ACE inhibitors cost only about $20/a month. Those are subject to FDA regulation. I have other non-prescription medicines that I pay about $10/month. Those are also subject to FDA regulation. His comparison of the regulated drug vs the unregulated supplement while citing correct numbers, must be missing an issue other than regulation, such as an intellectual property or orphan drug law. It could be that the supplement dosage is not well controlled. Or finally merely it is expensive to prove that the supplement does anything and the way we pay for it is by letting the manufacturer get a monopoly and charge a lot for the treatment.

You're looking at housing through a very narrow lens. Let's just consider a very basic demand case: a decent sized home, big enough for a family of four, in a good job market, with a not too long commute, in a safe neighborhood and decent school district. In 1970 that produce existed in a way that was affordable to a median single-earner household. In 2017 that product requires at least a two-income household, and even then is probably only relatively affordable at 75th percentile or higher.

You're right for the very high end of the market, people buying new housing stock, the construction component has improved. High-end 2017 homes are nicer and bigger than high-end 1970 homes. But I'd argue even the location component has gone backwards. The income required to live in an inner-ring suburb, let alone the urban core, of a high-end job market like NYC or the Bay Area is all but unattainable to anyone outside the top decile. And let's look at the bottom end of the market: the lower quintile is literally living in shipping containers. It's undeniable that this is a major regression since 1970.

You're missing something. The cost changes in the housing markets in NYC & SF are very different than the much smaller changes in places like Atlanta, Phoenix, Dallas and Houston. Look at http://www.economist.com/blogs/graphicdetail/2016/08/daily-chart-20, set the chart to prices in real terms and compare individual cities.

I have a single-income household of six with a reasonable (30m) commute and a 3500+ sq. ft house in one of the above metro areas. I agree you can't do that in SF or NYC, but that's because there is something special going on in a few locations, not because there is something inherent in the housing market in the USA increasing the costs. What is different about the increasing price areas vs. the others I leave as an exercise for the reader. :)

Some observations on colleges and college costs: 1) It is essential to account for the cost reductions that some students receive in scholarships in doing the analysis.

2) Because of the increase in the premium paid for educated labor, many students who now get sent to college in prior decades would have dropped out due to intellectual or emotional issues. We currently demand that colleges serve those students. There is a cost to this. There is also a benefit that is not being captured.

How many of those marginal college graduates are in careers that are any better than their non-college educated forefathers? We have a substantial proportion of college graduates doing jobs that thirty years ago didn't require a college degree. (see below link). It'd been thing if all those marginal graduates were becoming engineers or doctors. That'd represent a genuine gain, and maybe the increased cost to push more people through is worth it. But the vast majority are becoming retail managers, police officers, office assistants (i.e. secretaries) and the like. People used to be able to do those things without even having to go to college.

Now we intellectually torture people who aren't cognitively equipped for tertiary education, debase the standards of higher education, and burn a whole lot money doing so? Why, because employers demand much more college degrees? And why? Not because they actually need college skills, but because the system has pushed so many people through college, that it's a really bad sign not to go to college. It's like a circular loop of absurdity.


"Not because they actually need college skills, but because the system has pushed so many people through college, that it’s a really bad sign not to go to college."

Why shouldn't employers select for college grads when they get their educational assets for free? Businesses have to pay for their real estate, production machinery, accounting, advertising, etc. but they don't have to pay for the education that they're requiring of their employees and, as the graph indicates, those employees haven't increased their income over their uneducated predecessors. Employers are getting purportedly better employees for the same price as the uneducated dolts that once manned the enterprise.

Can any of these things fail and go out of business?

Yes? Costs will be constrained by the ability to pay of it's clients and competitive pressures.

No? Costs will be constrained only by the ability to pay of the tax payers, or the ability to borrow by the taxing authority.

In response to demands for lower costs, what happens in the organization if participants offer less service, complain to the clients, increase the costs to the clients by delays or other increased costs? Does the funding get increased, or is there a mass cleanout of personnel starting from the top?

The opening graph in the blog post you linked is bs.

Let's consider a similar graph: per capita labor income over time versus per capita hours of Labor supply over time. Why are wages going up so much when firms aren't getting any any extra labor?

Of course the problem is we need to account for the value of the labor, which is increasing (both because human capital is increasing and the value of output is increasing -- that is, the value of marginal product, which equals the wage of workers in competitive equilibrium, is increasing).

The same thing applies to education production. Human capital is more valuable now than ever. Education both produces human capital and is human capital intensive. It should be no surprise that the price of education is increasing.

Sure, I agree that a variety of regulatory problems exacerbate high prices in education (and healthcare and real estate and pharmaceutical development). But can we please avoid tendentious arguments? Or at least recognize them for what they are?

Stagnant results show that there is no change in output of human capital, though.

*value* of marginal product. Human capital is worth more now than in the past. I presume the author showed sat scores as a proxy for human capital. So increments in sat scores are worth more (there are so many things wrong with using sat scores, including extreme selection bias (which varies over time). And the fact that sats are normalized test scores. But ignore that for now). Since the marginal product of educational inputs is their value-added to sat scores (again, qualms with sat scores), the value of the marginal product of educational inputs increases with the price of human capital.

The point is that even if schools are teaching you the same amount, what they are teaching you is worth more.

This would induce higher production of education, But increased production of human capital is at least partly offset by the fact that education is a human capital intensive production technology - it takes educated people to make educated people. So it is unambiguous that education will get more expensive. Increasing the value of education pushes up both the supply curve and the demand curve.

More to the point, people look at that graph of sat scores and act like education is broken -- that those tends can be explained only by market or government failures. This conclusion does not follow.

Actually the "stagnant results" show us nothing and may not be stagnant. The tests, populations taking the tests, and the curricula have changed over the decades. While I have not delved into the details of how CATO constructed the graph; I am skeptical that they have accounted for any of this.

Some examples of changes that have major impact: 1) computer programming is certainly taught more now than in the 70s. Is that tested for? If so, then the how has it been normalized? and 2) Students who used to be viewed as disabled are more likely to be mainstreamed into schools.

It is utter nonsense to plot test scores and costs on the same graph like that. There is no reason that one would believe that doubling the per-pupil spending would double test scores anymore than doubling the amount spent in treating a disease will double a patients lifespan.

"Human capital is more valuble tab ever. "

2 objections:

1. The classic diamonds vs water problem. Supply and demand, not intrinsic value, determine prices.

2. Education isn't more valuble. Ever heard of credential inflation?

Maybe it is not cost that has changed, but the allocation of costs to a reference price that is never paid.

Take healthcare. If a hospital has a high proportion of uncompensated care, costs are allocated to paying patients. You could say that costs are increasing for paying patients, or you could say that, because of the lack of public funding for medical care for low income patients, that the costs have been shifted to insureds and patients to are able to pay.

Same with school. If states reduce their contributions to state universities, costs are shifted to paying students in the form of tuition increases; if colleges recruit students who do not have the ability to pay, those costs are shifted to students who can pay.

Maybe it is not a cost disease so much as a cost shifting disease.

Cross subsidies and other cost shifting games create a lot more space for waste and rent-seeking than everyone paying their own bills.

Not necessarily. Take insurance. If a large part of the market becomes uninsured, shifting costs to the insured, some insureds will cease buying insurance because their premiums cover the costs of the uninsured., whereas if there is a tax on the insured and uninsured to fund the cost of the uninsured who are mandated to pay if the can afford it and not if they can't, then insured market stay healthy with people continuing to buy insurance to cover their own healthcare.

Uncompensated care is a very low percentage (well under 5% of total spend). It would have been just as much prior to WWII. There wouldn't be any if hospitals charged what people could afford. Public funding is a majority of medical spending and hospitals very much like it that way due to the complete lack of accountability.

Why is it Singapore, South Korea, and Israel get better results with 1/4th the spend as a %of GDP? It can't because of less cost shifting, can it? US hospitals are scum who scam the public via 3rd party payment.

Dan, the thing I don't understand is that insurance companies would seem to have significant incentive to control costs in the US system, but it just doesn't seem to work. And we've moved to high deductibles without much impact, and the insureds do feel that. Granted, it's very difficult to shop as a consumer, and it's difficult to find "discount" providers, but I don't see the system as without incentives - it's just not very responsive to those incentives.


I just looked up Singapore, and you do know, of course, that the government subsidizes healthcare through means testing, for example, if you are unable to pay for hospital costs, so that disproves, rather than supports, your point. Excellent paper at Columbia University and even Wikipedia. I did not look up the other countries but will if you want to challenge me.

Great article. An important point to remember is that your costs are someone else's income. Lowering costs, or even containing them, will be fought tooth and nail by the recipients of those costs.

I find it interesting that in education one of the key measures of quality is students per teacher. We almost always view a low student/ teacher ratio as good news and a sign of better education.

But that means that in education we view lower productivity as an improvement.

What other fields do this?

Alexander writes off the "Baumol effect" because workers wages in these industries have not increased. But I believe Baumols argument is that In competing for workers the rates of wage increase could simply be the same as more productive industries- but the cost of the service in those relatively low productivity industries would have to rise. The workers don't capture it.

These expenses didn't even exist just a short time ago. They're probably considered absolute necessities now:

Campus security

Since 2002 the University has invested nearly $16 million in campus security systems and infrastructure. Many resources are in place to monitor, promote, and ensure safety on the Twin Cities campus:

TXT-U, the University of Minnesota emergency notification text messaging system.
2,500 security cameras on the Twin Cities campus, monitored 24 hours a day.
Secured buildings: 215 buildings with U Card access.
Code Blue Emergency Telephones: 21 phones provide instant two-way communication with University Police with the push of a button. Located throughout the Minneapolis and St. Paul campuses.
267 emergency intercom buttons, located in the stairways of most parking structures on the Minneapolis and St. Paul campuses. Callers are immediately connected to the University's Police Security Emergency Command Center (PSECC).
The U’s Department of Safety includes the University of Minnesota Police Department (UMPD) and the Public Safety Emergency Communications Center (PSECC ) with professionals working in law enforcement and emergency preparedness.
Many safety resources for students such as the Gopher Chauffeur free transportation service and 612-624-WALK—the free 24-hour campus safety escort service—can be found on the What You Can Do page.

There's a lot of things like this.

My own alma mater has spent 100s of millions in parking garages in the last 20 years. Likely to be in the billions after debt service.

In the old days, professors and students would live close to campus or take the bus. But campus is harder to walk to and easier to drive to now.

That fits with with Alexander's comments about lower institutional risk profiles. (Same as doctors ordering unneeded tests and evaluations just to be safe from suit).

Colleges are positional goods, so like Picasso paintings, the cost of a limited number of Havard degrees will go up as people get richer. You can still get a cheap education, but not at a positional good college.

Health care is about the difference between 98% and 99%, the delta between 98% and 99% costs just as much as the first 98% because of diminishing returns. Such an improvement is almost unmeasurable statistically, but if it's your life and you have the money, won't you pay what it costs even if the chance is small it will do you good? Afterall you can't take it with you. Same as with vet costs, people are as sentimental about their pets as people now. So many more elective treatments instead of putting the animal to sleep. So these costs are rising with GDP. Absolutely you can get 1970's health care at 1970's costs - don't take insurance, pay cash, and decline any complex treatment. Same with 1970 vet costs - basically don't bother taking your cat to the vet, just get a new one when the first one dies.

On school costs, this is partly the state ownership aspect causing no-one to bother tacking waste (like too many admins) or innovate. But it is mostly about services, schools nowadays do a lot more things with their kids than in the past because people want it. Like music, dance, theatre etc etc.

"You can still get a cheap education, but not at a positional good college."

Can you? Haven't costs at directional-state-U increased at about the same rate (albeit from a lower level)?

Yes - in fact you can get a great education nowadays for almost nothing (Khan Academy etc). It's like buying a reproduction print of a famous painting, what people say they appreciate is the actual visual image, but what is valuable is the positional good of having the original which no-one else can possess.

I'll be the first person to mention this: Scott Alexander in his opening paragraph leaves a sarcastic reply about how TC did not properly define "cost disease", instead TC described general inflation, yet Alexander fails in his entire verbose post to properly define cost disease. Amazing. Go to Wikipedia and look up "cost disease". Here it is: "Baumol's cost disease (also known as the Baumol Effect) is a phenomenon described by William J. Baumol and William G. Bowen in the 1960s. It involves a rise of salaries in jobs that have experienced no increase of labor productivity in response to rising salaries in other jobs which did experience such labor productivity growth"

That's cost disease, not the unclear 'what-if' speculations of Alexander*. And TC says the piece is 'excellent'? Maybe in a Straussian fashion (where excellent means non-excellent).

*just one example of many, Alexander says: "So, imagine you’re a poor person. White, minority, whatever. Which would you prefer? Sending your child to a 2016 school? Or sending your child to a 1975 school, and getting a check for $5,000 every year?" First, a lot of people would prefer the $5k check (including myself). But in any event this is not cost disease. This is talking about paying $5k extra for the tiny advances since 1975 in education.

He references the Baumol Effect. He also points out that it doesn't seem to be valid.

@chuck martel - OK thanks, but Alexander's denial of the Baumol effect is based on an N=2 sample, which seems bogus. For example, he assumes teachers did have an increase in productivity, when in fact due to the Baby Boomers they did (class sizes got larger, enabling the same teacher to teach many more students, an example of a productivity gain, no matter what you may think of big classes)...so I don't think Alexander rebuts the Baumol effect.

Ray, respectfully to your point on education, I think Alexander's point is we're spending a whole lot more money on education but test scores (his measure of productivity) have hardly moved at all. I guess we can argue that test scores aren't a good measure of output, but that just shifts the case to what exactly are we producing more of for all that extra money if students aren't scoring better, and is it worth it, and if not why do costs continue to increase despite our concerns about whether it's worth it?

Maybe doctors, teachers, etc. have not seen any gains because we keep pumping out too many of them.

Note that in other industries (tech, energy, finance) the gains go to fewer.

Regulate industries don't have this dynamic.

If you are going to nitpick about the use and definition of terminology, to wit, "cost disease", you need go no further to the headline to this blog post "What is behind the cost disease"? and the text that follows:

"I don’t have a similar graph for subway workers, but come on. The overall pictures is that health care and education costs have managed to increase by ten times without a single cent of the gains going to teachers, doctors, or nurses. Indeed these professions seem to have lost ground salary-wise relative to others."

If you strictly define "cost disease" as "a rise in salaries...etc" as endorsed above, and you accept Alexander's finding that in those stated professions there has been no rise in salaries then, by definition, no "cost disease" is present (nor really a question of being "valid"). Alexander seems to fully understand the academic definition and use it properly. On the other hand, although I might be inclined to cut Cowen some slack in writing the headline on the basis of, perhaps 1) He's using a definition that is not from the textbook (in which case, however, Alexander's comment was completely justified); or 2) he's being facetious, having been caught in a terminological blunder one would not expect from an academic economist by making (intentionally?) the same blunder again here. Also, this, from the very first para of that Alexander piece: "Cowen assumes his readers already understand that cost disease exists". Well, that unsubstantiated assumption seems to persist.
Thus, your nitpicking would rather be directed at Cowen.

As usual, you've managed to get everything backward.

In response to Ray at comment 69.

Rather than speculate, we could compare budgets from today and 1960 for hospitals, schools and infrastructure projects and identify where the cost increases are coming from.

Has this been done?

What about the argument stated by James Howard Kunstler and others, that it is all caused by the sharply reduced ROI from oil acquisition. It used to be 100 to 1, and now can be as low as 5 to 1. There is the idea it must be at least 30 to 1 to maintain our standard of living, but since it is far below that, our standard of living keeps going down, or "our costs keep going up."

A barrel of West TX Intermediate Crude is $53.86

Not $140

New pundits, please!

Wow, that makes me nostalgic for the peak oil wars of the mid 2000's. Latest BP Statistical review suggest world oil reserves now at highest ever level, but those people just won't quit.

@carlospin The oil price is still relatively low. It's just the oil drillers are not making any money. It's called low ROI. https://www.nytimes.com/interactive/2016/business/energy-environment/oil-prices.html

None of the cost disease industries are "winner take all"

Isn't the increase in per capita spending in health care and education a function of adding a lot more students and patients to the pool of participants. Fixed costs in health care (for hospitals, LTFs, etc.) and in education (for the campus) are high, so adding to the pool of participants necessarily greatly increases costs by a lot. Or stated another way, the per capita cost of health care and education is low when a large segment of the population doesn't participate in either.

Wouldn't that only be relevant if schools and hospitals were operating at lower capacity than before?

If you can find "Peter Van Doren discusses health care" on CSPAN, he goes through the whole thing. For healthcare, 50% of resources are used by 5% of patients. The rest is just arguing about who pays for this. Healthy American tend to overuse the system as well.

Education is pure careerism. Employers want credentials in lieu of actually measuring performance. Actual measurement would either cost too much or simply be impossible.

If America can't do healthcare or education right anymore, and some other first world nations can, the benefits to lowering transaction costs and buying these things from those countries must be huge.

Telepresence for medicine? Online curricula plus some dose of telepresence for primary ed, and pry pure online for high school and higher? It's not functional enough yet, but the robots are getting more agile and we're getting VR already so maybe it is solvable.

Or, one step earlier, flying out for any not-super urgent serious hospital treatment? Sending your kids to a college abroad? The stupid signaling rat race argument gets in the way, but maybe Harvard can agree Sorbonne is no worse at teaching and quietly offer the students an option of learning there while getting Harvard diploma and paying a fraction of tuition? Don't see this happening, but maybe some solutions exist?

Ultimately, neither market nor government would work if people don't have any choice. Maybe providing this choice, if only at the margin, to the few % of the population ready to fly out for hospital and optimize internationally for college, is a force needed to slowly push those systems to a better place?..

There is rampant and spreading ... faux-careerism. Engineers are encouraged to become "six-sigma black belts", doctors are drained off into administration, teachers have onerous continuing ed requirements. One makes ones bones in education by positing theories. These theories collect the worst of our social sciences.

Part of it is subsidy - if you subsidize something, it costs more and there is more of it. The rest is credentialism - one does not advance one's career by doing a good job, but by the paper chase. In cases, this can make sense but many credential programs live in the world they wish that was, rather than the world we're in .

Alexander lost me with his first graph. How do we start out in 1970 with a 0% "total cost"? And I note both the dotted line (not explained) between 1970 & 1980 and the notion that this cost should include EVERY cent spent on a child in a given year - which is what is says. I further note that test scores are graphed on the same vertical axis, which is risible. My preemptive conclusion is that there is an axe being ground there, and the source isn't reliable - it's clear that the graph isn't justifiable. I find this entire discussion stupid. (although I should be more PC) The notion that education in 1970 meant the same thing as it does in 2017 is a clear indication of impaired thinking. For instance, let's compare the time-on-task of adult-child interactions in 1970 vs 2017. Is the number of working mothers significant? Is the reduction in the number of children per household significant? Is the educational attainment of a teacher with, say, 5 years experience in 1970 higher or lower than that today? Relative to what? What has been the impact of the dramatic shift from single earner households to two earner households on the market for teachers? Does economic science really apply over multi-generational time periods? and on, and on.

Grinnell College (located in Grinnell, Iowa) has properly spent more than 1 billion USD on campus construction, reconstuction, rearrangement etc. over the past 20 years. If instead they had spent like half of that (which would still have been more than enough - most of it is soulless extravaganza) and reinvested the rest, their humongous endowment would probably be like 750 million USD higher now.

Currently, it's 1.649 billion USD - it could have been 2.4 billion. With 1,705 cheap-to-educate students (its a liberal arts college - so no expensive graduate research labs etc., plus an extremely cheap location), they could probably just charge room, board plus some sort of infrastructure fee, and go tuition free.

Never gonna happen.

And I am not counting all the other areas of "fat" where millions of dollars disappear each year.

Thoroughly depressing

He was wrong about veterinary costs and regulation. In the 60's, any drug could be used on animals and you could even buy them off the shelf at places like dairy supply houses. Nowadays, you have to hire a vet and get it from FDA approved facilities with a script (prescription) and only use it on the animal species that it is approved for applied by the Vet. For almost all species outside of cattle, pigs, and chickens, almost nothing is approved so you need someone or group with an "Investigational permit" and be part of a drug study at your expense.

I was raising aquatic animals where nothing is approved (wrong species) and to treat a sick fish legally is almost impossible under today's FDA. They even went so far as ruling that Ice, Salt, Vinegar, garlic etc. are legally DRUGs that they must legally regulate (these materials impact the health of fish, therefore are drugs), but to stop a year-long fight they created a new class of drugs called "low regulatory priority drugs" keeping the option open to screw anyone who opposes them in the animal husbandry business.

The FDA is leading the way in Vet costs.

Does Alex have a dog? ;)

That basically isn't true. The FDA is evil and destructive, but the regulations are more permissive than that.

"The FDA is evil and destructive"

& you're as crazy as a shit house rat.

I investigated purchasing canine levothyroxine for myself. Turns out I can't buy it without a prescription from a vet. I wonder if some of the extension of FDA to pet medicines is to prevent this activity. (Would be interesting if pharmaceutical industry supports it). I was able to buy fish antibiotic without a prescription fwiw.

I guess I'll add that the difficulty of an individual trying to take control of some of their own medical situation is difficult in the present system. Outlets outside of the system-as-designed are systematically shut off, but I suspect its not just the FDA. It seems the industry itself has no interest in providing a cheaper alternative. There are quite a few things people seem to have to go through a doctor to get that I can't understand the need for, especially in a world of high deductibles where we're trying to encourage cost consciousness on part of the patient. But the system seems to keep trying to thwart those efforts. I'm trying to find cheaper ways but barriers are in place.

Could you write more about this post?

While I am inclined to agree with much of it, it does strike me as a complicated problem to study/unravel - e.g.:
The comment that cell phones declining in price dramatically is who things should be it. Now, is it the case that some (or even much) of that price decline comes because of fundamental science research etc. at universities which is partly funded by that increased college costs? Heck, it could be even more direct - colleges are taking that money and basically starting a hedge fund that invests or outright buys other companies. i.e., my college cost has gone up not because I'm funding my own education but because I'm funding other developments that may benefit me in other ways.

I don't necessarily believe this argument myself. But it seems like there are mechanisms/channels to be considered wrt how you attribute progress in one part of society to costs in another.

The revealed truth about cost disease is that, at the margins, there is a stronger constituency for increasing costs than decreasing them.

People complain about college costs, but consider a 5% increase in the budget vs a 5% decrease.

5% increase: students and parents complain, but few are willing to leave over a matter of 5%.

5% decrease: crisis and uproar. Faculty and staff are unwilling to consider a hiring or wage freeze. Administration is unwilling to forego building projects. Students actually will make decisions based on luxury dorms and climbing walls.

Similarly, in medicine I don't think you will find many places that are willing to admit to doing 5% less than average in return for 5% lower costs. There might be liability issues involved with doing 5% less than the "standard of care."

In practice, people seem to get much more worked up about 5% worth of bells and whistles than 5% of costs, even when they don't translate directly into outcomes.

I don't see how you can fight cost disease without building a constituency for lower costs.

Good points. One thing I don't understand, however, is why do the health insurance companies not have incentive to drive lower costs? Or if they have incentive, why is it not sufficient? Most actual consumers are fairly clueless about costs and coverages, but are sensitive to the premiums (probably not as sensitive in the US as they could be because it comes out of the paycheck pre-tax; would be much more sensitive if they personally wrote the $1000/month check) . But the insurer should be in a position to make good decisions on this.

Comments for this post are closed