William Baumol was an economist specializing in industrial organization and technological change - applied microeconomics. William Bowen was also an economist, but someone who gained more fame in university administration, serving as president of Princeton University, and writing many books - good ones - on the challenges facing American higher education.
In 1966 together they published Performing Arts: The Economic Dilemma for the Twentieth Century Fund, but if you want a pithy summary, you can go to this paper from 1965, in the American Economic Review. This is where they explain the concept we now know as cost disease, and its implications for the arts.
Most people in arts management are familiar with the capsule description: it takes a certain number of musicians a certain number of minutes to perform Beethoven’s seventh symphony, and this hasn’t changed since it was first performed in 1813. But obviously no orchestra today could get away with paying musicians the wages they earned in 1813, and so we face a problem.
In this post, I’ll try to do two things: explain how cost disease happens; and ask whether it is, in fact, a problem.
How does cost disease happen? I will break it down into steps…
Any country is, in aggregate, only as rich as the value of what it produces - goods, services, extraction of natural resources. Productivity is our word for the value of output per worker, and countries with high productivity will have higher incomes per worker.
For most of human history, productivity grew very slowly, and in turn average living standards grew very slowly. This changed dramatically with the industrial revolution, when productivity growth began rising very quickly (Brad DeLong says things really began to take off in the West in terms of ordinary living standards in the 1870s).
What causes productivity to increase? A few possibilities: workers can become better educated, and healthier. We can invest in more infrastructure and privately-owned capital goods like factories and office buildings and machines. We can have government policies that encourage workers and capital investments to go to their most productive uses (centrally-planned economies were terrible at this, over-investing in some things and under-investing in others. Tariffs also serve to direct resources away from more valuable, and towards less valuable, uses). And we can have technological change: new knowledge, often embodied in new machines.
Technological change ends up being the most important in the long run, because it is not subject to diminishing returns. A country can invest only so much in physical capital before it doesn’t quite know what to do with even more roads or factories. But you can always find ways to incorporate new technologies.
Some technologies are labor-saving: they allow us to get as much output, or even more output, than before, with fewer workers. I’ll give four examples of sectors that have had a lot of labor-saving technological change. First, telecommunications. If you watch old movies you know you used to need a telephone operator to make a call. In my lifetime you still needed one for (enormously expensive!) overseas calls. But we just don’t need operators anymore. Given the number of calls we make now, there are not enough workers in the world that could handle the volume as telephone operators. Second, transportation. When I was a boy I lived just a few blocks from the harbor in Vancouver, and I remember how many people had jobs loading and unloading ships. But if you go to a seaport now, even a really big one like Rotterdam, you simply don’t see that many workers. Standardized-container shipping greatly reduced the need for longshoremen. And those massive container ships that cross the oceans? They typically have a crew of just a couple of dozen. Third, manufacturing. Cars, steel, appliances, use a lot of robots on the assembly line, with far fewer people needed to do things like punch rivets. And fourth, agriculture. Food output per farm worker has grown fantastically in the past century - farms now represent a tiny proportion of the workforce, with output higher than ever.
But some sectors are not amenable to labor-saving technological change. If I visit a fourth-grade classroom today, the labor needs are much the same as when I went to school. You can’t through technology have one teacher handle five hundred ten-year olds at a time, nor can you cover the learning needs in half an hour per day. Some tasks require a human to be physically present for an irreducible amount of time. Getting a haircut. Or a massage. Or getting your annual physical or dental exam. Or most aspects of the justice system. Counselling and social welfare. Child care and elder care. It’s not that there is no technological change: we know more about how to teach, and have a wider range of things to teach about. There have been great advances in medical technology. But in the end you still need workers (when I worked for the government of the Canadian province of Saskatchewan, the provincial government budget had reached a point where half of all expenditure was on health care. But the bulk of this was spending on personnel - you can’t really reduce the number of nurses you need).
Across the economy, there has to be a kind of equilibrium in the labor market, between the effort required to get the necessary qualifications for a job, and the wages and quality-of-life from working in that job. If one type of job takes a long time in terms of getting qualified, but wages and conditions are really crummy compared to what you could get in a different sector, then people gravitate to where on the whole it is a better deal. In each sector, in the long run, wages have to be competitive with other sectors or you will find it really hard to attract any employees.
A recap: our economy gets richer through productivity growth, which mostly comes from technological change. Labor-saving technological change is quite uneven across sectors. But wages cannot get too far out of line across sectors. So…
Some sectors - telecommunications, manufacturing, shipping, agriculture - have very fast labor-saving technological change, but wages in those sectors rise in a way consistent with an economy-wide average. As a result, we expect costs, and in turn prices, of things from those sectors to, relatively, fall. Other sectors have little or no labor-saving technological change, but wages there have to keep up with the growing wage levels across the economy. As a result, we expect costs, and in turn prices, of things in these sectors to, relatively, rise. And that is what we have come to call cost disease - rising labor costs relative to labor-saving productivity gains in some sectors (although not all sectors).
Agatha Christie, recalling her youth, said “I couldn’t imagine being too poor for servants, or rich enough for a car.”
So let’s think through the implications.
It is not that sectors without much labor-saving technological change are “inefficient”, or that they lack incentives to get their acts together. It is just in the nature of some services that a human cannot be replaced by technology (even if they can be greatly aided by technology, they cannot be replaced by it). If you’ve been in higher education for a few decades you will remember how “MOOC’s” were going to revolutionize the sector. They didn’t.
There are no magical solutions. Consider child care. It requires a certain number of carers per child per hour and you can’t get around that. It is expensive and will only get more expensive as wages rise. We can think of policy proposals: get the public sector to subsidize it? Sure, we could do that, but we would need to recognize this is just shifting the rising cost (perhaps quite justifiably) from the parent to the taxpayer. Encourage parents to just stay home and look after their own kids? Again, that doesn’t evade the problem: the cost of staying at home is the foregone wage of the stay-at-home parent, and that too is rising over time. Cost disease isn’t eliminated.
Many things the government pays for are in the cost disease world: education, health care, social services, the justice system. That means the cost of government is going to rise relative to other things over time. If we want consistency in the quality of public services, tax rates will need to rise over time. You cannot have tax rates from days of yore to fund contemporary governments at contemporary wages. It’s not government inefficiency - you can’t just issue blanket statements about how many public sector jobs you would eliminate.1
“Cost disease” is a funny name for this, because it is a shift in relative costs and prices that only happens because as a society we are getting richer. It’s not that we cannot afford teachers or nurses any more - we are richer than ever before. It’s just that their costs are going up compared to the prices of cars and microwave ovens.
Well, this is eventually supposed to get around to the arts.
The live performing arts are, as the Beethoven example I began with illustrates, a cost disease sector.
We cannot make a blanket statement about “The Arts” because the making and distribution of recorded arts has really fallen in cost. I’d say I first used my own money to buy a record in 1970.2 Since then, the cost of acquiring recorded music has gone down a lot (and the quality of recordings has increased at the same time). Further, the cost of good stereo equipment has drastically fallen - I now have what would be considered a pretty ordinary set-up these days, and its fidelity would have been far out of my dreams when I was a teenager. And if you want to record your own music, again, the technology available to you has gone far up in quality and far down in price - our high-school basement band collectively could never have managed to acquire a decent 8-track recording device, and now you can get something serviceable for $300.
Even for things on stage, at least some costs have fallen - transportation, for example. But personnel costs have risen in real terms.
Is there a way around this? Not really. Suppose a theatre company says, “right, from now on let’s only do plays with one or at most two actors.” That will cut costs for now, but as soon as you do that, you will see over time the wage demands of those one or two actors rising. Cost disease will always be with us.
The problem for the arts manager is to recognize this situation, and try to ensure that willingness to pay for tickets rises along with costs. Remember, the economy is richer - people are paying DoorDash to bring Big Macs to their apartments. Find efficiencies where you can, but plan ahead for rising labor costs.
What about arts policy? Sometimes people cite cost disease as a reason to have public funding for the arts. I am not quite convinced. There are a lot of cost disease sectors in the commercial world where we do not call for government funding: people can pay for their own haircuts, cab rides, and restaurant meals. We might say if we have a different reason for public funding of the arts - various concepts of its social benefits - then we might think that amount needs to rise over time as costs rise.
In other words, cost disease is a thing to consider in how we think about the amount of public funding for the arts if we have already decided, for some other reason, that public funding for the arts is a good thing.
Coda:
When I was in high school I played in a rock and roll band, and our goal was to get good enough to get gigs playing high school dances. Our own school (a very ordinary, middle-middle-class public school) had four or five dances with live bands each year, and they were really good, professional cover bands. During my time in high school we had Heart play at our school, twice (this was before their first album, but they were a force when it came to Deep Purple / Led Zeppelin covers), Sweeney Todd when Nick Gilder was their lead singer, and, in what represented the pinnacle of our musical career, we opened for Trooper at our own school.3
That’s a fuzzy picture of me at seventeen.
What high school today could afford such an expense? It wasn’t long after my day that disk jockeys took over: there was just no way that the meager ticket revenues from high school kids could pay for a serious band with serious gear. Cost disease.
Our band never did get to headline a high school dance - I really wanted to be a rock star, but it wasn’t in the cards.4 So I had to settle for economist.
cf “we will eliminate waste, fraud, and mismanagement.” lol.
Cosmo’s Factory.
Non-Canadians do not know who Trooper is. Every Canadian knows who Trooper is. This is just one aspect of the vast difference in cultural heritage between Canada and the US.
Another kid at our school at the time had more success in this regard…
Like you, I am not convinced, but for slightly different reasons. B & B asserted that theater and music (we'll stick with those) couldn't take advantage of technology to increase productivity, but it seemed to me that this was sleight of hand--movies, TV, and music recordings ARE the technological versions of theater and music. Yes, we still need the same number of actors or musicians, but it is the audience that is "more efficient"--you can only fit so many into the Broadway theater where "Wicked" is playing, but way more can see it when the movie is distributed across the nation. To assert that it isn't "the same experience" as live theater and music performances is true, but neither is a ride in a car "the same experience" as a ride on a horse, yet they are both means of transportation. The question that we have to face is whether theater is the horse-and-buggy of the arts, and if not, how to emphasize what makes it worthwhile. (By the way, the thing B & B didn't address is the ballooning personnel associated with the growth of institutions -- the salaries of artists are nothing ompared to the salaries of people "needed" to run an institution the size of, say, the Guthrie.)
I really enjoyed that, thank-you. It was well written, clear and I loved the way you presented the information. I learned a lot from it and will read it again. Well done!