Not much, you might think, when you hear about some of the politically correct idiocies they’re generating these days. Supporters of universities often argue they produce positive externalities that warrant public support, much as the positive externalities put off by anti-malaria spray warrant public support. These days, however, it sometimes seems the externalities are mainly toxic. 

Another argument university presidents seldom fail to make when seeking higher budgets is that universities bring big economic payoffs, a claim made much more often than the number of peer-reviewed studies supporting it warrants. As a result, a new working paper from the National Bureau of Economic Research, written by Anna Valero and John Van Reenen, both at the London School of Economics, comes to a neatly-phrased conclusion that seems bound to be widely quoted: Double the number of universities in a region and you increase per capita GDP by four per cent. (The paper is “The Economic Impact of Universities: Evidence from Across the Globe.”) 

The data and econometrics necessary to come to such a conclusion, to say nothing of the complex and multifold forces and feedbacks linking universities and GDP, are together so daunting that many of us simply won’t believe the conclusion is meaningful. Double the number of universities (of whatever quality, size or focus?) anywhere (city, town or desolate tundra?) and you’ll get four per cent more per capita GDP, no matter what, no matter how financed, sounds more than implausible, even if the authors seem to have added all the right controls and raised all the appropriate self-doubts.

Not being an econometrician myself, what I usually find most interesting in such studies is the description of the raw data and some of the “what ifs” that go along with them. Valero and Van Reenen use a data set that includes, before they pare it down a bit, 16,326 universities worldwide, with universities being defined as “higher education institutions that offer at least a three- or four-year professional diploma or a post-graduate degree.” Seven countries (the United States, Brazil, Philippines, Mexico, Japan, Russia and India, in descending order) account for more than half the universities in the world. The U.S. has 13 per cent on its own. Interestingly, in the period the study focuses on, post-1955, “there has been very little exit from the university sector.” That would be a story in itself. In how many other sectors is “very little exit” also the case over the last 60 years? 

One wrinkle in the study of particular interest to Canadians: Though a region’s GDP rises if it increases its stock of universities, it also rises, usually even more, if neighbouring regions increase their stock. If there are these spillovers across regions, that suggests there may be a need for a federal role in university education (if, that is, there’s need for government at all—see below). Unfortunately, the interregional spillovers only show up when regions are close together, less than 200 kilometres apart. In a country as vast as ours, that may make them irrelevant.

The most interesting what-if is an example for the United Kingdom, where in 2010 there were 171 universities across 10 regions. If one university were added to each region, the model predicts an increase in U.K. GDP of £11.3 billion per year. With universities spending £160 million per year on average the cost of 10 more is £1.6 billion per year, which looks like a pretty good benefit/cost ratio. That dripping you hear is university presidents salivating over the prospect of bigger budgets. 

But, careful! Is an increase in GDP a net benefit to society? People give up leisure to get it. It’s not all gravy. Even more importantly, does it necessarily follow from the benefits being greater than the costs that governments need to subsidize this process?

If I learn something at university that makes potential employers willing to pay me more, that does increase the value I add via economic activity, which means by definition it increases GDP. But the employers do pay me for that skill, and that makes me willing to pay the university to acquire it. And the university is more than willing to sell me my acquisition of it. So the whole process may take place—extra universities will increase GDP—without any minister having to do anything.

Just because something is good, even good for GDP, doesn’t mean it requires subsidy.