SlateStarCodex and Caplan on ‘Why Are the Prices So D*mn High?’

SlateStarCodex, whose 2017 post on the cost disease was one of the motivations for our investigation, says Why Are the Prices so D*mn High (now available in print, ePub, and PDF) is “the best thing I’ve heard all year. It restores my faith in humanity.” I wouldn’t go that far.

SSC does have some lingering doubts and points to certain areas where the data isn’t clear and where we could have been clearer. I think this is inevitable. A lot has happened in the post World War II era. In dealing with very long run trends so much else is going on that answers will never be conclusive. It’s hard to see the signal in the noise. I think of the Baumol effect as something analogous to global warming. The tides come and go but the sea level is slowly rising.

In contrast, my friend Bryan Caplan is not happy. Bryan’s basic point is to argue, ‘look around at all the stupid ways in which the government prevents health care and education prices from falling. Of course, government is the explanation for higher prices.’ In point of fact, I agree with many of Bryan’s points. Bryan says, for example, that immigration would lower health care prices. Indeed it would. (Aside: it does seem odd for Bryan to argue that if K-12 education were privately funded schools would not continue their insane practice of requiring primary school teachers to have B.A.s when in fact, as Bryan knows, credentialism has occurred throughout the economy)

The problem with Bryan’s critiques is that they miss what we are trying to explain which is why some prices have risen while others have fallen. Immigration would indeed lower health care prices but it would also lower the price of automobiles leaving the net difference unexplained. Bryan, the armchair economist, has a simple syllogism, regulation increases prices, education is regulated, therefore regulation explains higher education prices. The problem is that most industries are regulated. Think about the regulations that govern the manufacture of automobiles. Why do all modern automobiles look the same? As Car and Driver puts it:

In our hyperregulated modern world, the government dictates nearly every aspect of car design, from the size and color of the exterior lighting elements to how sharp the creases stamped into sheet metal can be.

(See Jeffrey Tucker for more). And that’s just design regulation. There are also environmental regulations (e.g. ethanol, catalytic converters, CAFE etc.), engine regulations, made in America regulations, not to mention all the regulations on the inputs like steel and coal. The government even regulates how cars can be sold, preventing Tesla from selling direct to the public! When you put all these regulations together it’s not at all obvious that there is more regulation in education than in auto manufacturing. Indeed, since the major increase in regulation since the 1970s has been in environmental regulation, which impacts manufacturing more than services, it seems plausible that regulation has increased more for auto manufacturing.

As an empirical economist, I am interested in testable hypotheses. A testable hypothesis is that the industries with the biggest increases in regulation have seen the biggest increases in prices over time. Yet, when we test that hypothesis as best we can it appears to be false. Remember, this does not mean that regulation doesn’t increase prices! It can and probably does it’s just that regulation is not the explanation for the differences in prices we see across industries. (Note also that Bryan argues that you don’t need increasing regulation to explain increasing prices, which is true, but I still need a testable hypotheses not an unfalsifiable claim.)

So by all means let’s deregulate, but don’t expect 70+ year price trends to reverse until robots and AI start improving productivity in services faster than in manufacturing.

Let me close with this. What I found most convincing about the Baumol effect is consilience. Here, for example, are two figures which did not make the book. The first shows car prices versus car repair prices. The second shows shoe and clothing prices versus shoe repair, tailors, dry cleaners and hair styling. In both cases, the goods price is way down and the service price is up. The Baumol effect offers a unifying account of trends such as this across many different industries. Other theories tend to be ad hoc, false, or unfalsifiable.


Addendum: Other posts in this series.


A cynical person might suggest that the auto industry discovered the old fashioned IBM model of profits - don't worry about making a large profit on the initial sale, just get the customer dependent on what you provide.

Or to put it differently, you can't use Ford parts to repair a Toyota - and the auto manufacturers are very aware of that fact. The question that the car prices versus car repair prices graph does not answer is how much of the cost is due to the mechanic, and how much is due to the parts (or the more or less implied requirement that work be done at a dealer so as to keep a warranty valid).

Or to use an even more obvious analogy - the cost of a printer versus the cost of its ink would likely resemble that auto graph, starting at 100 somewhere in the 1970s till today. And considering the opposition to right to repair laws, one can even assume that the growing profit from repair played a major role in lowering the initial vehicle cost.

Ooops - got to go. The motor warning light just turned red, and only the dealer has the technology to turn it off - should only cost a couple of hundred dollars.

The fly in this ointment is that cars have also grown more reliable and you don't need to take your car to a dealer to get it repaired.

I suspect cars are cheaper because of outsourcing. Can't outsource education and health care.

'and you don't need to take your car to a dealer to get it repaired'

One can assume the number of fender benders has remained more or less constant, though the regulatory framework changed - 'NHTSA amended the bumper standard in May 1982, halving the front and rear crash test speeds for 1983 and newer car bumpers from 5 miles per hour (8 km/h) to 2.5 miles per hour (4 km/h), and the corner crash test speeds from 3 miles per hour (5 km/h) to 1.5 miles per hour (2 km/h). In addition, the zero-damage Phase II requirement was rolled back to the damage allowances of Phase I. At the same time, a passenger car bumper height requirements of 16 to 20 inches (41–51 cm) was established for passenger cars.

NHTSA evaluated the results of its change in 1987, noting it resulted in lower weight and manufacturing costs, offset by higher repair costs.'

Further information from the article - 'As an example, in 1990 the Insurance Institute for Highway Safety conducted four crash tests on three different-year examples of the Plymouth Horizon. The results illustrated the effect of the changes to the US bumper regulations (repair costs quoted in 1990 United States dollars):

1983 Horizon with Phase-II 5-mph bumpers: $287
1983 Horizon with Phase-I 2.5-mph bumpers: $918
1990 Horizon: $1,476'

Regulatory changes and effects on costs go both ways, something ever so conveniently forgotten by people like Prof. Tabarrok. That is why the point about parts is relevant. In Europe, all cars use easily replaceable and standardized lightbulbs (still seems to be true, but LEDs are certain a new factor), and a set of headlight, brake light, and turn signal bulbs might cost 8 euros. There is no reason to buy a sealed unit, one of those neat auto repair cost tricks in the U.S., which has nothing to do with Baumol.

Clockwork, FYI, you can get an ODB 2 scanner on Amazon for $20 that will let you read the indicator codes and clear them. You can google the code to find out what it means in English. Great for saving the $200 bucks just to have them tell you the gas cap got left off.

True, but the reason leaving the gas cap off sets a code and requires service is because of government emissions requirements.

Similarly for the small build up of dust/debris on the Mass Air Flow sensor which will render the auto inoperable.

Many other codes are not required by law but are the progression to more industrial electronic controls on vehicles. Many autos are becoming physically standardized as customers value the improvements in software capability. Same parts, different cosmetics/amenities with just different microcontroller programming. Same with appliances.

The real cost increase in auto service is the "mechanic" now has to be a full electrical/mechanical industrial controls technician skilled in electronics diagnostics, or be a parts changer. But for the auto mechanic, each car represents a different industrial control system rolling in after being run through salt, snow, water, dust, debris, etc. The Rust Belt seems to have the most interesting auto-diag problems with the conditions impacting not just modules but also wiring.

A cynical person might also consider the insurance industry as a price driver of auto repair costs.

IOoops - got to go. The motor warning light just turned red, and only the dealer has the technology to turn it off - should only cost a couple of hundred dollars.

Not quite, the threat of right to repair laws got the industry to agree back in 2002 to make tools and training available to independent repair shops as they do to their dealer based service stations. Only Mass. passed a right to repair law but the fact is this is essentially a regulation. If the auto dealerships tried to cut off all shops outside of Mass, how could the knowledge be limited to just one state? It would also motivate more states and even the Federal gov't to pass right to repair laws. Hence the industry has 'voluntarily' agreed to not cut independent shops out. This is why you see plenty of independent auto repair shops.

But this also illustrates just how hard it is to analyze regulation. In essence this is a regulation even though it doesn't exist on the books. It carries some cost to dealerships and manufacturers who would love to get a monopoly on repairs. Yet how would you analyze regulation when you have to consider even regulations that don't pass as possible regulations.

To be honest, I was thinking about John Deere in terms of current right to repair laws.

'to make tools and training available to independent repair shops as they do to their dealer based service stations'

And this is a cost driver, compared to the past, where a mechanic (generally) did not need to buy high tech diagnostic equipment from each auto manufacturer. It is an interesting discussion (and certainly includes OEM parts), but there is a reason why though large repair shop chains will likely stay in business, a local mechanic has a lot of challenges compared to 30 years ago.

Which leads to the idea it would be interesting (if difficult) to compare the cost of body repairs/painting over that time span, as that is less impacted by the increasing level of technology found in current cars.

Another area would simply be tire replacement - fairly straightforward to compare, and something that still can be done by many mechanics using the same equipment they have owned for decades.

Put a piece of electrical tape over the spot on your console where the warning light/code shows. Also if you hear a strange noise turn the radio on.

The point about car design being ruined by government regs is spot on. Modern SUVs / CUVs / cars are marginally safer in a crash, but the lack of outward visibility contributes to poorer driving.

And drivers' skill and accountability, while never high, are approaching comic levels.

And yes, vehicle costs are slowly dropping. But an engineer in the auto biz made the case that they SHOULD be much less expensive. After 120 years of production, they SHOULD be commodities. A 2001 Camry could be made for $15K today - if regs were frozen at 2000 levels. Of course, this engineer admitted he probably wouldn't have a auto industry job if that were the case.

Not marginally safer in a crash. Vastly safer.

For occupants, definitely safer. For people on foot, in wheelchairs, on bicycle? Vastly more dangerous.

Objective evidence? Pedestrian deaths from motorvehicles has been falling dramatically until recently and the recent increase seems to be due to people looking at their phones while walking and not cars being invisible due to, what, regulations that automakers have to camouflage their cars?

Vastly safer in a crash?
Ok. I'll concede somewhat since I did use the phrase "in a crash".

I'll clarify:
Driving overall is marginally safer.
If you tend to crash a lot, you're vastly safer in one of today's tanks.
Driving a motorcycle overall is less safe.

Pretty much any modern vehicle is safer on a crash than an equivalent vehicle would have been 40 years ago. We've made huge strides in that area.

This thread doesn't stand up to the data. Since 1966 deaths from auto crashes has gone down both in absolute numbers, in terms of deaths per size of the population and in terms of deaths per miles driven. This would capture both deaths caused by cars crashing into each other and cars hitting people on foot/wheelchair/bike. I have no idea if drivers are comically bad when compared to drivers from decades ago (I think this would be hard to measure, for example, in earlier ages we had a lot more tolerance for DWI but then today we have more distractions in the car). Nonetheless, it would seem engineering has made cars safer faster than we can be poorer drivers, if indeed that is happening.

I would think we've taken lots of the gains in absolute safety in relatively faster and more distracted driving, too. Radar crash warning = video games/texting while driving. Better crash test results = going 85, not 75.

Cars are safer, roads are safer, there's less tolerance for drink driving, more restrictions on underage driving. This is one if the success stories of our time.

This is apples and oranges here. Cars are Goods, Education and Healthcare are services, you can't compare the effects of regulation on goods and services like that.

The largest cost in autos is labor, and that can be reduced by utilizing robots, manufacturing where labor is cheap etc etc.

Education and healthcare cannot realize those same efficiencies, as they are highly location specific, and cannot be easily replaced by robots.... yet.

That's his point.

But, as Arnold Kling pointed out, you also include a chart showing that barbers have had no Baumol-driven increase since 1960. Why not? Are barbers really less skilled than folks mending shoes and clothing? And why is skill relevant anyway -- that does not seem essential to the Baumol story at all. Why the dramatic increase in the wages of physicians and nurses (far above those of auto-mechanics)? Aren't barriers to entry a big part of the story? (Immigration would have no effect, BTW, unless those barriers to entry were removed -- immigrant MDs can't practice unless they repeat residency in the U.S. and those slots are limited, so more immigration would not result in greater supply). Lastly, nobody seems to be addressing the huge increase in higher-ed costs despite most instruction now (unlike in 1960) being done by low-paid, non-tenure track faculty (and why has their pay -- like barbers -- failed to show the Baumol effect)? Another mystery -- unlike doctors and nurses, salaries of K12 teachers have show little real increase over many decades. System costs have gone WAY up, but again, this seems driven entirely by hiring more teachers, administrators, and ancillary staff for a given number of students. How does Baumol account for not having no increase K12 teacher pay even as productivity increased in other sectors?

I'm still feeling quite unsatisfied with the Baumol story when salaries have exploded in certain service occupations (medicine) and have hardly budged in others (barbering and K12 teaching). If, in an alternate universe with the reverse pattern, where teacher and barber salaries had exploded and nurse salaries had stagnated, would the explanation still be Baumol?

As someone who thought and commented on this earlier:

1. It seems clear to me that cost disease can be happening, without it being the only thing happening.

2. If cost disease is about compensation for equivalent work/skill, barbers are going to be viewed in a different category, non-equivalent to doctors.

If cost disease is about compensation for equivalent work/skill, barbers are going to be viewed in a different category, non-equivalent to doctors.

But that's not what it's about. Baumol says all occupations that don't and can't really have productivity gains (doctors, barbers, tailors, teachers) should nevertheless see pay increases to keep pace with occupations that do have productivity gains (otherwise nobody would go to into the flat productivity occupations). The skill level shouldn't really matter.

But what we're actually seeing is a weird, inconsistent pattern where some lower skill occupations (tailors) show increases and others (barbers) do not. Auto mechanics show increases, but K12 teachers do not. Physicians do, but the PhD non-tenured faculty who do most of the teaching do not.

I guess I'm viewing Baumol through a behavioral economic lens. I see the *way* teacher's salaries rise is that they go to school with, rub shoulders with, live beside engineers, and so they and their entire value group views them as the same, to be paid the same. And so as engineering productivity rises, teachers salaries also.

Barbers, no, and that's partly a class issue, yes.

That’s a totally reasonable theory, but it’s absolutely not Baumol.

Baumol is solely a function of productivity rising in one sector affecting wages in the non productivity rising sector.

My guess is that Baumol is a factor but heavily outweighed by globalization of labor markets, barriers to entry, and subsidization of demand.

In 1970, the K-12 student/teacher ratio was 22.3. In 2015, it was 16.0.

That's a 28% cut in workload.

Also, the cost of gold-plated health insurance has exploded, while retirement benefits (pensions and health care) that were worth 15% of pay in 1980 are worth 40% of pay now just because long-term interest rates went from 10% to 4%.

In 1956, I sat in a NYC parochial school classroom with 50+ pupils.

This week's Barron's has a lengthy article on (heterodox?) MMT. Not sure I can wrap my small mind around all that mental masturbation. Sounds to me like a central planner's 'wet dream.' It said, since 1990, health care, housing, and education (I'm omitting bank services not charged to customers) worth/prices/costs have risen. Some common features cited in the article: lack of competition, opaque pricing, heavy subsidies, regulatory constraints on supply.

In the same Barron's issue, James Grant (one trick pony) has an article on interest rates.

Of course, when Sister Margaret Ignatius (who basically worked for room and board) sent home a note regarding my misbehavior, it would be double jeopardy squared. Today, some parents would excoriate the teachers. That isn't economics, though.

Where has gone two decades of greatly-increased public education spending? The short-answer, according to a Manhattan Institute report by Josh McGee: 'State and local governments have catastrophically mismanaged their teacher pension systems. The cash infusion to K-12 has been used largely to pay for irresponsible pension promises politicians made to teachers’ unions and justified to the public with shoddy accounting. . . . '

I think you can trace every problem in American back to too much money being given to working people.

Apples and oranges. Public schools have to cover special-needs children today, they didn't in 1970.

"In 1970, the K-12 student/teacher ratio was 22.3. In 2015, it was 16.0.
That's a 28% cut in workload."

But what's the median ratio for classroom teachers? I suspect the drop has to do with adding more special purpose instructors than in reducing the size of the typical classroom. I know my kids elementary classrooms were nowhere near a 16:1 ratio (and that in a high-spending, left-leaning district where the teachers' union basically runs the BOE)

But, as Arnold Kling pointed out, you also include a chart showing that barbers have had no Baumol-driven increase since 1960. Why not? Are barbers really less skilled than folks mending shoes and clothing?

Perhaps barbers have had productivity increases. I've noticed a lot of the hair cutting chains I go to seem pretty good at cutting down on wait times and keeping their barbers and hair people busy all day. With apps to check people in and offer sophisticated loyalty rewards they very well might have found ways to increase productivity.

"Clothing repair" on the other hand has not been the subject of any such focus. People seem to simply buy new cloths rather than use a tailor more than once on an outfit. But I could imagine a data driven productivity boost is possible among tailors that may consolidate the number in an area and use apps and online ordering to get each tailor to output more 'clothing repairs' per hour than before.

If nothing else, those two graphs strongly suggest why most things are not made to be repaired these days!

Costs must go up for GDP to go up because economic growth is not measured in megatons of trash, pollution, destruction.

If the price of products that lasted longer andservices cost less that reduced the need for more services, GDP would be plunging. A $3000 car that needs only solar energy that comes from a $2000 solar array and both last for 50 years, and everything else does the same, GDP per person will fall toward $1000. If health care came up with a couole of shots costing $10 that not only prevented infections and viruses, but also cancer, obesity, heart disease, etc, would mean only 1 doctor per 10,000 population to deal with accidents, especially if clothing were engineered to be protective at a cost of $25 so no one was injured in drive by shootings.

Economies are zero sum. Can't have higher GDP and lower costs. Cost is paying for GDP, either paying the workers, or paying to consume it. And you can't pay the consumers $100 while charging the workers $200 to consume it. Workers are consumers, consumers are workers.

Unless you redesign the economy to have government hand out cash thats used to buy stuff with none of that called cost, and government paying costs in ways that make lower costs the most highly rewarded.

Say, everyone gets $100,000 year, but if you deliver a Tesla Roadster to the government for $10 that government sells for $200,000 you are designated "ideal citizen" and recognized with a $10 billion annual award.

Government is the major buyer of medical services but not the major buyer of automobiles, big difference. Government tends to buy at a profitable price for the medical industry, and the medical industry thus always plans to grow. Government will always cause cycles as medical purchases grow large and we have to have a reset to re-assemble a yield curve that predicts prices better. It is the later, the continual resetting of prices to compensate for growing medical industry. That is your disease. Call it Baumol? It is a matter of choice, I prefer the Baumol to be the general mechanism of adjust the yield curve to make pricing more accurate.

What is the equilibrium, Tyler keeps asking. We have two ways to arrive at equilibrium, the Euler method and the Quantum mechanical method. The quantum mechanical version is the tree trunk concept, we will always go back to the tree trunk and reset roundness because government constantly grows the medical industry.

The Euler method says we have no quantum constraints. This we end up with a single medical industry, in LA occupying three square miles. If you are seriously ill fly to LA. Otherwise visit the local nurse.

Boundary conditions apply, that is why we have the abstract algebra tree, the tree trunk is our yield curve. This is the equilbrimum that applies, boundary conditions. The effect of boundary conditions is to bankrupt Congress before the huge LA medical center is even planned. In other words, we get more of an Occam's Razor effect, and decide to simply default on all the medical promises made that exceed boundary conditions..

Shoe repair people and tailors/dressmakers had some of the worst historical pay of any job. Every little girl learned to sew. Any man could fix shoes. These jobs are synonymous with abject poverty. Shoe repairers tended to own their own stores, but they were basically closets, often in basements of commercial buildings. Fast food workers make more in terms of purchasing power than either of these jobs did.

Barbers were originally overwhelmingly business owners of a middling class. Overhead was low as commercial rents were low. You’ll find the flat pay for barbers to be primarily due to increased real estate costs eroding wage increases.

K12 teachers DO make phenomenally more than they used to. It used to be a job where the pay wouldn’t pay for living alone—you could literally only survive as a boarder, and in many small towns, there were teacherages attached to jobs to partially offset the terrible wages. My great grandfather was a superintendent, and his wife worked as a court reporter to make ends meet.

Doctors make far more because of the educational opportunity cost as well as the cost of the education and the few people who are able to become doctors.

I find the argument that immigration would lower medical costs to be insane. Immigrants are overwhelmingly low skill. They also are from countries where behaviors leading to much higher rates of disease and injury are common. (Just buckling seatbelts and using car seats is something many don’t do.) They also abuse emergency rooms and can’t be convinced to use medical services appropriately. All of these things cause a huge burden on the hospitals where there are the most immigrants.

Well in Illinois teachers make a ton of money, with good medical benefits and pensions guarantee of 3% annual increase. The story of underpaid teachers is long past. Regarding shoe repair, just had sole replaced for $55. Not sure what the labor cost portion was, but the cost is about what an oil change for my 2013 Honda Accord. Good news is that now my shoes are good for another 2 years or more. Will be back for an oil change in about 6 months.

Regarding shoe repair, just had sole replaced for $55.

But that doesn't tell us much. Maybe the guy can repair a sole for $55 in ten minutes but in terms of income there's a big difference between him having two jobs per day or six jobs per hour for 8 full hours in the day!

K12 teachers DO make phenomenally more than they used to.

No, not compared to 50 years ago. Make sure you look at the 'constant dollars' side of the chart, which shows only a minor increase in the national average (from $52.8K to $56.3K).

Agree need to use constant dollars but also must account for benefits. Only familiar with Illinois where teachers health benefits are good to great and pensions are outstanding, which are not taxed at the state level.

But rich benefits are nothing new. I have an aunt was a teacher between roughly 1962-92. She's been retired since her early 50s and the state has been paying her a pension and providing health coverage for her and her husband for more than 25 years (with plenty more to go).

But the monetary value of those benefits has skyrocketed

Seriously, could someone explain for me how immigration would help lower medical costs? If that were the case shouldn't we already be swimming in BOGO's, twofers and the medical equivalent of those Bed Bath & Beyond coupons?

Have not been in a hospital or rehab facility recently, right? My wife has had two knee and hip replacement surgeries in recent years . I was really surprised how many of the nurses and lab technicians were immigrants, including many from Africa and Haiti.

I did see that, especially when my dad was still around, but aren't immigrants adding greatly to the patient load as well?

I think he means that immigration would decrease medical prices if you had immigration of medical practitioners at the same proportion as existed in the native population and those practitioners were allowed to practice once they came, i.e. if there were an increase in supply.

However this is not the case, so there is a constant supply and an increased demand.

It would be nice if you addressed The substance of Scot Alexander’s comments

"A testable hypothesis is that the industries with the biggest increases in regulation have seen the biggest increases in prices over time. Yet, when we test that hypothesis as best we can it appears to be false. Remember, this does not mean that regulation doesn’t increase prices! It can and probably does it’s just that regulation is not the explanation for the differences in prices we see across industries."

Or that your test isnt very good. Absence of evidence and all that.

I think that the Baumol effect is big but you also need an increase in demand to make it explain 95%+ of the increase in spending on healthcare and schooling. That is were Governments subsidizing demand contributes. Government does not subsidize demand for cars nearly as much as for schooling and healthcare.

If 95% of spending rise is dues to all the Baumol effect, we should ask what percent of the healthcare cost is direct labor? To look at a non the Baumol effect part of healthcare, drugs, drugs use very little direct labor, are drug prices are falling but very slowly.

So, I think it takes both rising demand and the Baumol effect to explain the increase in spedning.


If it is 95%+ Baumol effect, since a lot of schooling healthcare seems to have such low bang for the buck. how do we lower demand AND increase efficiency.

Direct instruction and tutoring seems to work in schooling but there is little move in that direction AND is it totally irrational for a person to increase his life time consumption of non- healthcare stuff by say 12% and only use the most effective healthcare? I don't think that is irrational at all, the UK provides something like that though the NHS, though it provides much more than the one third that would save use 12% of spending.

BTW there have also been a sharp rise in costs of road/subway construction which is mechanizing fast.

The big issue I have with all of this, is that there's absolutely no looking at the demand side of the equation. What does the effect of what the market is "willing" (sometimes forced, in the case of health care) to pay have on prices? The market I'd point out is cell phones, where the vast majority of focus is on the flagship high-end phones, with relatively little on mid-cost or low-cost phones.

People are willing to pay a high cost for stuff. That's simply the way it is, and for most status positioning goods, higher cost is seen as a positive, not a negative. There's a strong cultural element to this that's for the most part generally ignored, and I think if you're trying to reign in costs, that really does have to be part of the discussion.

You did a great job on the study, and an even better job explaining the Baumol effect and the findings. Excellent work!

It's obvious to me the Baumol effect is a big part of the story. I think there are two caveats.
1. The contribution of the Baumol effect varies by industry, even among services industries, from explaining most of the effect to almost none of the effect.
2. The overall economy wide contribution is no more than 50%. That is huge amount of explanitory power, but it's half or less than the whole story. Your book presents it at 100%, and I think that's clearly wrong.

So, unless I'm missing something, Alex:

1) Does not address the substance of SSC's comments
2) Does not link the SSC review, only links the original cost disease post
3) Quotes Scott addressing the book's conclusions -- "I find their case pretty convincing. And I want to believe. If this is true, it’s the best thing I’ve heard all year. It restores my faith in humanity" -- as though it were a review of the book itself

I'm not 100% sure about my read on point 3, but still, ???

Even assuming that Baumol's cost disease is a significant contributor to price increases in education and health care services and that regulatory policies are irrelevant, Tabarrok's conclusion that:

" Whatever the explanation for this difference in productivity growth, however, it’s a difference, like the difficulty of domesticating huckleberries, that cannot be traced back to policy.
We should neither ignore this difference nor make too much of it" is a testable hypothesis and it has been disproven countless times.

In industries where armchair economists did not simply "close their eyes and think of England" deregulation achieved massive cost savings.

The funniest thing in the history of economics has to be Tabarrok's insistence that his little regulatory word count algorithm has proven "rising federal regulation cannot explain secular trends in economic dynamism." As if word counts would not be random. Does he even begin to consider that for example all the tiny detail federal regulations in the hyperregulated automobile industry actually undermines the premises of his methodology?

Definitely an Ig Nobel candidate.

To what extent can the Baumol effect explain the observation by Collison and Nielsen of decelerating scientific progress?

Why Are the Prices so D*mn High is indeed a great read, thanks!

One question bothers me, though: why was the price of physicians and nurses multiplied by three relative to barbers and bus drivers? It seems there haven't been productivity gains in these sectors, either. Does the Baumol effect only work for skilled labor?


Improvements in technologyl/drugs has also made time spent with a doctors more valuable to consumers but not barbers . With the increase in wage premium workers get for a college degree, higher education is also more valuable. If output could be measured and use to compare occupations instead of wages their would probably not be much difference.

Thanks Joan,

I think you’re onto something when you write “if output could be measured”. Indeed, it is the subjective value of the service provided which determines how much the market is ready to pay for the factors of production. So counting the number of haircuts and hours of lecture produced with one hour of labor is not a good metric. Instead of quantities, I think we should focus on the value of these services.

Cars today are complex. Ever looked under the hood of an Audi? That's right, you can't even see the engine. In the 1950s my teenage brother and his buddies regularly worked on their cars, from changing the oil and spark plugs to making adjustments to the engine to make it more powerful. Not today. Servicing an automobile today requires much more training than in the past; hence, the cost is greater, much greater.

On the other hand, appliances, televisions, etc. have gotten much cheaper, in large part because they are produced in places like China. Indeed, they are so cheap that it makes no sense to repair them when they break: just replace them. That means there are fewer in the business of repairing them, and their cost is much higher, especially in relation to the cost of what they are repairing.

As for shoes, I have mine repaired, even though it costs about as much as the cost of a new pair. Why? The styles of today's shoes look ridiculous, almost as ridiculous as the Pee Wee Herman suits. That means my shoes are 20-30 years old, some even older. That's another thing: today's shoes aren't made as well as shoes in the past. My father wore Bally shoes before Bally shoes were, well, Bally shoes. If he were alive today (he's been dead over 50 years) he'd be wearing the same Bally shoes. I remember when I went to work as a lawyer and purchased the clothes I needed to wear for my work. It took some time, but I eventually purchased a pair of Bally shoes. I still have them. They look almost new. I remember meeting with a client, my best client, the first time I wore those shoes. He kept looking at my shoes, probably because he was concerned that he was indirectly paying for them. Finally, he looked at me and asked: "Do you think the shoes make the man?"

The correct answer to that question is "Of course! We are nothing but our shoes."

Hi Alex,

The Baumol effect = dog bites man. You've rediscovered that water is wet.

Wages are obviously going to roughly the same across industries in a given country because the cost of living drives wages at the floor and the cost of education and experience (vs profit) drives the higher end.

The interesting issue has nothing to do with wages. It's the lack of productivity in some businesses and the abundance of it in others.

In private businesses, there are mechanisms to enforce productivity gains: selection and profit. In most public businesses and some insulated private businesses, these mechanisms aren't active.

Your comparison of the cost of goods vs services (car repair vs shoe repair) misses the point.

How much revenue do car and shoe repair generate today vs. 80 years ago? the answer is that they're a much smaller portion of the economy because the quality and price changes have dramatically reduced demand for car and (especially) shoe repair.

"it’s not at all obvious that there is more regulation in education than in auto manufacturing."

There is no law requiring you to buy a car, and there are no tax supported free cars being handed out to try to monopolize the market.

Although maybe President AOC will start handing out Teslas to everyone as part of the Green New Deal...

There is a vast difference between regulation of manufacturers and regulation of services. Two examples.

As per the Paris Accord, refrigerants which have high greenhouse gas potential are to be phased out. I'm not sure what the domestic market is doing, but commercial appliances are using propane as refrigerant, a natural refrigerant that is encouraged by regulation.

For manufacturers, there are some costs, and some savings. The refrigerant is substantially cheaper than 134A, the HFC that was commonly used. Some devices and practices were changed; the controls are outside the cabinet, and the wiring connectors are designed to not spark. Probably a slight decrease in cost on balance.

The reworking of the lines, the supply chain adjustments of components that are suitable was a cost. The handling of the flammable refrigerant requires some equipment and procedures. These costs can be spread over the production output however. On net prices are flat or slightly down.

But on my end, servicing this stuff the story is very different. The refrigerant is more expensive for me to purchase; the wholesalers sell 14 oz disposable cans of refrigerant grade propane for about $1 per oz here in Canada. About the same as 134A. The Montreal protocol made illegal disposable refrigerant containers, and here we go again. I can buy anywhere from 30lbs to 500lbs containers of 134A, and can handle it with normal procedures, but large quantities of propane require both transportation procedures and storage facilities that are very expensive. The wholesalers can handle the 14oz cans without any additional expenses, so that is what is available. No cost savings, slightly more.

Since the refrigerant charge if released into the cabinet would create an explosive atmosphere (these are in commercial kitchens with open flame very close) if there is a leak in one of these things I have to cover my risk. We remove the appliance to a well ventilated spot, put a pressure test overnight in the unit to assure that it is tight, then charge it to get it working. I estimate another hour or two of labor charge.

Those extra costs are per invoice. They aren't spread over the production of a factory. So the regulatory costs either are taken from the wages of my workers or passed on to the customer.

Another example. Almost everywhere by regulation any job that entails risk requires a safety meeting daily. I'd be very optimistic to suggest 15 minutes, and say you have a crew of 6. That is an hour and a half of time. (it likely is a half hour or more). If this is a construction crew where the additional labor costs are applied to a house or commercial property, the costs would be reflected in the construction costs or taken out of the wages of the workers.

Apply this same thing to a factory crew per shift. That shift will produce multiple items that the cost can be applied to.

Same with any non scalable endeavor. Medical, teaching, auto repair, dentistry, etc. You can't make a doctor 10 times more productive. I can't say to my crew that they will now handle 20-30 service calls a day each instead of the 2-3. Travel time would be longer than their workday.

So to say that regulations on car manufacturing compares to regulation on teaching is frankly ridiculous.

And the previous article said that these costs are one time. Again, someone should get out a bit more. The propane change is one of many many changes that are occurring, on top of the health and safety regulations that are adding costs substantially. Regulatory requirements are constantly changing, very seldom making things cheaper.

Industries where there is competition or for whatever reason can't pass on the costs deal with it by moving offshore, lowering the wages paid. What I face is the increased costs get close to the replacement costs of the equipment, so the factory gets the work and I don't. Industries that can pass on the costs do, and we see the increases.

"You can't make a doctor 10 times more productive."

Replace the doctor with a computer for diagnostics

Other leading Baumol experts declined comment, reportedly saying they were unable to find the "comments" link tucked up in that weird spot near those two charts.

Rather than comparing auto and health care costs, it would be more interesting to see a comparison between the cost of "traditional" health care and cosmetic, dental, and veterinary health services.

Per Megan MacArdle a while back dental and vet services have also been increasing in price as a rate faster than inflation too. Something I can easily credit just based on my own experiences.

Comments for this post are closed