The principle -- generally accepted by economists -- is simple enough. Suppose you're a company that manufactures things (or, these days, contracts to have them manufactured). As time goes on, the manufacturing process gets more efficient. Productivity rises. So you spend less money to make more widgets.
This happens more or less through the entire economy. So we all (very generally speaking) get richer. (Obviously, I'm leaving out such factors as glaring income inequality, which normally I care a lot about.) Because we're richer, we can have things we didn't have before. Computers. iPhones. More sophisticated cars. More varied clothes and food. We take these things for granted. They're part of our lives. We expect to be paid enough so we can buy them. Which, if we work for a company that shows increased productivity, isn't hard for our employers to do.
But some big players in our economy get left out of this. These are institutions (very typically nonprofits) that don't show productivity gains. Orchestras, for instance. It takes just as many musicians to play a symphony now as it did 50 years ago. Or hospitals. Or universities.
Orchestras, in fact, are less productive than they were, because (see above) they need larger staffs, for marketing and development. And so orchestras fall behind the rest of the economy. Their costs keep rising, just everybody else's do. Just like General Electric, or Ralph Lauren, they have to pay higher salaries than they used to, so their musicians -- and the people on their staff -- can buy computers, and nicely varied food.A nice summary of the notion, as articulated by its main proponent William Baumol in a 1966 book and subsequent papers, can be found here.
There's something seductive about this idea: a nice graph pops into one's head in which the cost line mercilessly rises past the income line, something like those graphs of Medicare spending if we don't get our shit together fast.
But some things also seem fishy. Like, just how long does this cost-death take, anyhow? Private symphonies and opera companies in some semblance of their modern administrative forms have been around for at least 150 years. Compared to other horror stories of obsolescence by productivity (banister carvers and carriage whip makers anyone?), that seems like a pretty balmy fate. Also, the crux of the cost disease argument seems to come back to the idea that industries which cannot improve the unit productivity past some point are doomed--once you can't reduce staff or increase product volume any further, rising wage costs mean you're dead in the water. Yet we seem to be surrounded by business models that would also fail this test, like any independent urban restaurant that is predicated upon the labor that goes into food production and service, rather than the simple distribution of increasingly cheaper foodstuffs to a larger audience.
Tyler Cowen touches on many of these issues in a 1996 critique of the "cost disease" notion here, focusing on a couple of key points:
1) In economist speak, the "cost disease" hypothesis posits that substitution effects swamp income effects, i.e., the desire to switch away from the low-productivity symphony sector to a higher productivity substitute dominates decisions. But a scenario where increased incomes cause individuals to consume more of the symphony sector, even if it is more expensive, is no less plausible.
2) Why should we assume that the symphony orchestra and other performing arts organizations are fundamentally incapable of increasing their productivity? Baumol's example which Sandow repeats--it still takes the same 50 dudes to play symphony x it did in 1780--seems far too narrow a way to conceive of the musical services a modern arts organization is capable of rendering with the technology available to distribute their product. Recordings, and recent developments in live broadcast, are the obvious innovations, of course. But consider as well the ways that modern travel and communications allow the most modest string quartet to tour the globe. Even the ubiquity of large modern concert halls represent advances in distribution.
3) Finally, Cohen argues that the "cost disease" hypothesis ignores improvements in product diversity and quality and illuminates nothing more than rising nominal costs. Today's orchestra has a few hundred years of extra repertoire to offer relative to that 1780 orchestra, and large improvements in the quality of performance that create real additional value for orchestra consumers--value that can't be easily compared to the 'value' gleaned from cutting the timpani section.
Obviously, performing arts organizations are having a hard time of it right now, and I'm not saying that any of these things are a silver bullet or "prove" that everything is alright. But the cost disease idea and its predictions of inescapable economic annihilation for the performing arts seem just a bit too convenient for those who indulge in classical music pessimism. Blaming the current troubles on theories about the economic exceptionalism of arts organizations rather than understanding them in the context of the larger economy seems counterproductive.
Update: Matthew Guerrieri has a great post on the "cost disease" here (and kindly links to the above)...