E is for Economics of the Arts
What is it good for?
Thirty years ago I went from being a kind of bog standard faculty member in an Economics Department to becoming my University’s Dean of the Faculty of Fine Arts (it’s called Media, Art, and Performance now - I was not around for the name change, though I was Dean through the harrowing process of constructing and moving into the new building space you see on the home page). It’s a funny story how it happened.
Anyway, when I was appointed, there were two kinds of reactions. A minority view, held by Board of Trustees sorts, was “this is great - somebody who will be able to make an economic case for the arts!”. The majority view was “are you joking? This will be a disaster for us; all economists care about is money.” I set about my task of confounding both groups, and switched the focus of my research from labor economics and tax policy (much as I was enjoying the study of the coordination of tax policy between states with no border controls but different value-added tax rates) to the economics of the arts, and that’s where I remained.
The way I approach studying arts markets and arts policy through economics is to see economics as a quite useful way to organize thinking, to get a picture of why things work the way they do, but to recognize its limits as a tool, and know that there are kinds of questions where trying to answer through the tools of economics is just not that helpful, and sometimes can take you in the wrong direction entirely. (My working paper - free download here - gives a more detailed look at why I think this).
Economics is useful for understanding why markets seem to work the way they do: Why, even in the age of the internets and apps, is it still the case there remain intermediaries between artists and audiences? Why are so many contracts set up as option contracts? Why do we observe such skewed earnings in the arts, with a handful of very rich and a multitude of barely getting by? Why do the live performing arts keep getting relatively more expensive? Economics is really good at this sort of thing: it’s an older book, pre-internet in fact, but Richard Caves’s Creative Industries remains highly recommended on all these questions.
But arts policy? I started as an enthusiast, but over time questioned more and more whether economics was a useful way to think about government support of the arts.
Let me amend that: economics is very useful in knowing how not to think about government support for the arts. Here are two wrong ways to think about economics and the benefits of government funding for the arts. I have written and taught about these errors for years, but in the end, on retirement, I guess at failed at reaching these “learning outcomes”.
The first error regards the purchase of complementary goods and services.
My antagonist in this battle is the incorrigible lobby organization Americans for the Arts. So let’s turn to their report Arts and Economic Prosperity (6th edition). Here is one extract:
What continues to set AEP6 apart from other national studies is its analysis of the event-related spending by arts and culture audiences. When people attend a cultural event, they often make an outing of it—dining at a restaurant, paying for parking or public transportation, enjoying dessert after the show, and returning home to pay for child or pet care. AEP6 shows that the typical attendee spends $38.46 per person per event, in addition to the cost of event admission. A ZIP code analysis of each of the 224,677 survey respondents shows that a third of attendees (30.1%) traveled from outside the county in which the event took place. Their event-related spending was more than twice that of their local counterparts ($60.57 vs. $29.77).
I am not questioning the actual figures: if they say a typical arts attendee pays $38.46 in addition to the cost of event admission, so be it. But that doesn’t answer the key question: who cares? If I purchase a car, I will also have a lot of other expenses involved in using the car: fuel, insurance, license and registration, routine maintenance and replacement parts, and, if I am unlucky, towing, and work getting done at the body shop. These are not things that make the auto industry great. They are costs I must bear if I am to use a car. That an increase in the demand for macaroni will lead to an increase in purchases of cheese is not something that leads me to want to prepare a report on Pasta and Economic Prosperity. It is not a measure of decreased prosperity if people are able to walk to shows rather than drive, yet that is an implication of this report.
Virtually any good or service I can think of has other things we purchase along with it, either as a necessary cost (gas for your car, a sitter for your kids), or as a nice thing to go along with the purchase (getting a sunroof installed in your car, going for a drink after the show). But that’s not a reason to advocate for public funding for the arts, or cars, or spaghetti.
The second error is the (ab)use of the Keynesian spending multiplier. On this topic I turn to … well, it’s Americans for the Arts again. From the same report:
AEP6 measures the economic impact of the arts using a methodology that enables economists to track how many times a dollar is respent within the local economy, and then to measure the economic impact generated by each round of spending (i.e., the direct, indirect, and induced economic impacts). Think of this as tracking a supply chain.
Consider this example:
A theater company purchases a five-gallon bucket of paint from its local hardware store for $100—a very simple transaction at the outset but one that initiates a complex sequence of income and spending by both individuals and other businesses.
Following the initial purchase, the hardware store may use a portion of the $100 to pay the sales clerk who sold the bucket of paint. The sales clerk then respends some of the money for groceries; the grocery store uses some of the money to pay its cashier; the cashier then spends some of the money for rent; and so on.
The hardware store also uses some of the $100 to purchase goods and services from other businesses, such as the local utility company, and then to buy a new bucket of paint from the paint factory to restock its shelf. Those businesses, in turn, respend the money they earned from the hardware store to pay employees and buy goods and services, and so on.
Some of these expenditures are local and some are outside the region. The local ones continue the local economic impact cycle. Eventually, the last of the $100 is spent outside of the community at which point it no longer has a local economic impact. It is considered to have “leaked” out of the community.
The total economic impact describes this full economic effect, starting with the theater’s initial paint purchase and ending when the last of the $100 leaks out of the community. It is composed of the direct economic impact (the effect of the initial paint purchase by the theater), as well as the indirect and induced economic impacts, which are the effects of the subsequent rounds of spending by businesses and individuals, respectively.
This comes from the analysis of shifts in aggregate demand (ignoring aggregate supply constraints) in a simple Keynesian model, where an exogenous increase in investment or government spending will generate a new equilibrium level of aggregate demand.
There are two issues here. One is that the increase in aggregate demand from an exogenous shock to spending only generates this “economic impact” in a depression economy, not in what we have now, where we are near full employment. It assumes that all of that increase in demand for goods and services can simply be met by otherwise unemployed resources, and thus with no consequent changes in prices.
The second issue is the same as with first error I documented: who cares? Every bit of new expenditure has the same effects according to this model. And so even if I thought this example from our local theatre company was very persuasive, I would find that I could also measure, by the same devices, the economic impact of cranberry farming in Massachusetts, the automobile sector, the personal fitness sector, or a NASCAR race in North Carolina. (I won’t list any more, but I will leave to the reader to pick any sector they like, do a search for “economic impact”, and something will come up).
Both kinds of errors I give here have a common element, which is that it applies a set of calculations to the arts as if it were any old sector - and indeed, as I point out, you could do it for any old sector - and treat it as something meaningful for thinking about the importance of the arts, and, ultimately, the public interest in the arts. But if the arts are just another economic sector like all the others, why have publicly-funded arts support? The numbers that Americans for the Arts present have no implications for public policy. They are just … numbers. If they came up with different numbers, it wouldn’t change anything.
The whole reason we have arts policy, public arts funding and agencies, is because we don’t think the arts are like cranberries or cars. Otherwise, why bother?
So, let’s turn to economic analysis applied properly.
The reason one would think that special, favorable treatment of some sector is warranted, whether through subsidizing consumers, or producers (which are administratively different but have the same goal), or through a tax exemption (say, for example, not subjecting it to general sales taxes), is that just leaving things to ordinary markets of buyers and sellers for the good would, in terms of social welfare, underproduce the good. We don’t normally worry about this. There are buyers and sellers of haircuts, and staplers, and tacos, and coffee mugs, and various prices for them and quantities purchased, and whatever those numbers are they are not a policy concern - nobody thinks people are buying too few haircuts (save for concerns about members of one’s immediate family). Ordinary markets for ordinary goods work really well at allocating resources.
But sometimes ordinary markets do not do so well, and this is where there are spillovers, also known as externalities. Suppose that when I make a purchase, someone else benefits from it (1) even though they are not involved in the selling of the good, and (2) the benefit is not one that occurs through the market, i.e., it is a benefit external to the price system. So, to take an example famous to economics students (at least to mine), apple growers benefit from the bees that pollinate their blossoms, but that is not an externality, since there are well-established markets where apples are grown for beekeepers, at a price, to come and place their hives in your orchard. Likewise, the late-night diner that gets an increase in business when there is a popular film at the cinema next door is not experiencing an externality - it is just the ordinary functioning of markets, where going to a movie and getting a slice of pie afterwards are complementary goods.
Some obvious positive externalities: schools - I benefit from children getting well cared for and educated even if I do not have children of my own; public health - I benefit from people not going around carrying infectious diseases. Even if I don’t care much about the welfare of others, and am rather self-centered in that regard, I would still be willing to have taxes collected that are used to pay for universal schooling and health clinics and vaccinations and such, since those things make my life more pleasant.
So: do the arts generate positive externalities? If I go to the opera, do people, or at least some people, who have no intention of attending the opera, benefit from my attendance? If yes, this would make a case for some sort of subsidy that leads to an increase in opera attendance.
But there are a few complications.
First, the external benefits have to exist not just in theory, or in the eyes of an advocate for greater government support for the arts, but in the mind of the person supposedly getting the external benefit. Economics, famously, does not question peoples’ tastes and preferences. And so if non-opera goers are unanimous in thinking “I couldn’t care less whether Rushton goes to the opera”1 then there is no externality, and I ought to be responsible for the full cost of my ticket (this issue will come up again in this series).
Second, if there are grounds to think there are positive externalities, then the policy-maker has to figure out which policies will best do the job of increasing arts production and consumption. In the US we subsidize nonprofit arts organizations with grants from federal, state and local arm’s-length funders, but does this actually have much effect? One of the first economists to think hard about this was Alan Peacock - I think this from 1969 was one of the first attempts to formalize analysis of government subsidy of the arts on externality grounds, and I highly recommend his book on arts funding and his practical work in an arts funding agency - and he makes the point that arts subsidies have to do something, they have to make something happen that otherwise would not have happened (I’ll come back to this at letter G).
Third, externalities should not be confused with so-called instrumental benefits from the arts. Suppose Rushton goes to the opera, but the experience of the performance is only a part of the benefit he receives, and he also benefits from getting out and about, seeing some familiar faces, demonstrating what a cultured person he is, and so on. Those are all personal benefits to him, and so do not count as externalities, which are about benefits to others not in attendance.
Fourth, what exactly do we think these externalities are? Why should someone care if Rushton goes to the opera? In textbooks on cultural economics there is a set of ideas usually tossed around in this chapter: non-attendees might still highly value the art form, though they are only able to access it through recordings, but are glad that someone is supporting the performing organizations; or, non-attendees might have no interest in the art itself, but think it makes their community look good in the eyes of others, that there is a local or national pride in having the art. Remember “economic impact” is not an externality, since it just represents very ordinary market activity. But this all seems pretty vague - when I used to teach this, I see now that I must have done an awful lot of hand-waving arguments.
What then is the most persuasive argument that positive externalities from the arts exist? The strongest case to my mind comes in the letter F…
We will not speak of people who say “I could care less” when they mean that, in fact, they could not care less.