A fascinating paper came out recently on SSRN, which should be of interest to anyone concerned with the economics of higher education.  Its purpose was to answer a most interesting question: is cost-inflation in higher education driven by internal factors, or external ones?

There are two leading theories about cost-inflation in higher education.  The first, proposed by William Baumol (whose new book I mentioned last week), argues that external factors are to blame.  Education, as a labour-intensive good, says Baumol, will always see its costs rise at a faster-than-average rate in an economy where constant productivity gains are the norm.  Howard Bowen, on the other hand, suggests that the cause is internal: because virtually anything can be defended on the grounds that it increases “quality”, universities’ main impulse is to raise as much money as possible, and then spend it on a more-or-less infinite list of priorities.  This is what’s known as the “revenue theory” of college cost inflation.

These two concepts are not mutually exclusive, but they do lead to very different policy remedies for the problem of cost-inflation. It would therefore be a good thing if we understood the extent to which the Baumol and Bowen effects contributed to the overall phenomenon; what’s cool about this new paper, Measuring Baumol and Bowen Effects in Public Research Universities, by Robert Martin and R. Carter Hill, is that it manages to use empirical data from 143 research universities to confront this very question.

Martin and Hill’s conclusion is that, for the most part, the Bowen effect prevails.  For every $1 that the Baumol effect has raised costs, the Bowen effect has raised costs by $2.  They also note – in a finding that engendered some amused/outraged comments from the American professoriate – that the most cost-effective ratio of tenure-track professors to professional administrators (no, I couldn’t find a good definition of what that term means) is 3:1.

I’m not familiar enough with the IPEDS database to critique the paper comprehensively, but it seems to me that the analysis ignores any kind of output related to research, which has been a significant driver of increased costs over the years. And the observation that 3:1 is the “most efficient” ratio is just odd: expenditures can still bring positive returns even if they are not maximally efficient (Economics 100: profit maximizes where Marginal Cost = Marginal Revenue, not where Actual Cost is lowest).

One should also be careful about assuming these results would hold in Canada.  My initial impression is that growth in administration has been more restrained here than in the US.  But we’ll never know for sure. Sadly, Canada has no IPEDS equivalent which would allow us to look at non-Academic staff numbers over time.