In 1966, William Baumol and William Bowen looked at the origins of rising salaries for live performances (music, theater, dance), and noted that an underlying issue was that such performances could not easily be made more efficient - productivity could not be increased (Baumol and Bowen, Wikopedia). The oft-quoted ( and quite convincing) formulation of this concept is that a Beethoven quartet must be performed by exactly the same number of musicians today as was required in the 19th century, and that the quartet requires roughly the same amount of time to perform. No increase in productivity over two centuries! This inability to increase productivity should, according to simple economic arguments, lead to flat incomes - rising income usually is a result of increased productivity. However, despite this, salaries in the performing arts had risen over time. Baumol and Bowen concluded that this occurred because it was necessary to keep salaries on a par with those in industries that were seeing productivity increases in order to keep workers in the performing arts. This "push" of salaries in industries without productivity increases is called Baumol's cost disease.
Very often, when someone poses the troubling question, " why is college so expensive?", the response is simply "Baumol's cost disease", said with an authority that suggests that should settle the discussion. In fact, used in this way, Baumol's cost disease is like the magician's gesture that is designed to get the audience's attention away from place where something important is happening. In reality, customer's don't care about the cost of making a product, they are concerned about the price they must pay, and that difference leads one down a potentially fruitful path of reflection about both Baumol's formulation, and critical issues in higher education.
When increases in productivity are difficult or impossible to achieve, the price of the product often can be controlled by spreading the costs over more consumers without raising costs proportionally. In the case of Baumol's Beethoven quartet, by increasing the size of the performance hall, more customers can share the costs of the concert, thus decreasing the cost/attendee (although net costs increase because the larger hall likely costs more to rent). This trick is used in higher education, of course - we move the lecture to a bigger hall. However, technology also enables the concert performance to be "repurposed" for a variety of different, expanded audiences. For example, the the concert can be broadcast over radio or tv, carried over the internet to be shown "live" in movie houses, and recorded for later play on CDs or mp3. Each of these repurposings has additional costs associated with it, of course, and so the final costs for each includes the cost of both the original performance and the intervening technology and personnel. Each aspect of the repurposing is worth doing financially (there may be other reasons, such as reputationally) only if it helps to increase overall net revenues.
It is worth noting that Baumol and Bowen's cost analysis focused only on the original performance, not its repurposing. Most of the repurposed performances have shown enormous productivity increases over the past decades because of their large technology component. Thus only the original component - the actual performance in a concert hall- is trapped in the land of no productivity increases.
Each of these repurposings is different from the original- hearing the quartet played in a small hall with 60 other people is very different from playing an mp3 of the concert in your car as you drive to work. But "different" does not necessarily mean inferior. Clayton Christensen argues that quality must be defined in the context of the job a customer is buying a product to do - if the product does that job very well, it is of high quality from the perspective of that customer. When I am driving to work, an mp3 recording that enables me to hear my favorite Beethoven quartet in the car is absolutely solving the problem that I want solved at that time, while having a ticket in my pocket to go to a live performance of the same quartet that evening is not solving any problem I have during my drive. Different contexts, different judgments of quality.
Higher education's first responses to cost containment was to model on Beethoven's quartets - we moved to bigger lecture halls, then recorded lectures and broadcast them, and more recently, put those recorded lectures on the internet. We in higher education have generally postulated that the purchasers of these repurposed lectures have been mislead or are ignorant because these repurposed lectures are inferior to the originals. In fact, although some customers certainly have been mislead, most are buying these products because they are the best available products to do the job the customer wanted and needed done - a job that for one reason or another precluded working with the time and location limitations of traditional campus courses.
However, the key reason that Baumol's cost disease in not the ultimate answer to the price of college question is that higher education now is developing new approaches to teaching and learning that numerous studies indicate should be more learning-effective and cost-effective than the traditional approaches. New online programs are not simply ineffective online versions of the original ineffective in-person lectures, but are carefully developed using the results of the latest in learning research. These new approaches can be used to supplement and improve the traditional "quartet" approach, or to replace it completely. In either case, opportunities to create economies of scale will occur, with resulting productivity increases.
All in all, "Baumol's cost disease" is an excuse for not looking at alternative, potentially much more effective, approaches to education that would lower price and improve learning. "Baumol's cost disease" is only the answer to, "What will happen if we maintain the status quo?"