The utility of doubt

Malcolm Gladwell gives us an engaging and sympathetic canter through the life and work of A. O. Hirschman here.

The doubts of Hirschman are clearly more appealing than the certainties of modern economists, based on more maths and less experience, but there are limits to this approach as economics.


The examples of Hirschman's thought that Gladwell gives are all well-expressed examples of good common sense: sometimes people make the best of a bad job; sometimes we wouldn't start something if we knew how hard it would really be; if you want someone's phone number then try looking in the phone book. That is all well and good, but the best (or at least the most exciting) parts of economics come from applying a clear-headed and rigorous approach to human behaviour and reaching counter-intuitive conclusions. This is an approach that can be taken too far (Noah Smith has made the point), but it is the approach by which economics really adds value (to use a horrible phrase). We don't need economists to tell us that phone books have phone numbers in them; we need them to tell us that Governments can (or cannot) borrow unimaginably huge sums of money safely, or that increasing the minimum wage does (or does not) help people on low incomes, and so on. People's common sense can go either way on these sorts of question, and we need some ordered way of investigating them.

So, to return to Gladwell's Hischman, we learn that "Digging through the Hoosac turned out to be a nightmare. The project cost more than ten times the budgeted estimate. If the people involved had known the true nature of the challenges they faced, they would never have funded the Troy-Greenfield railroad. But, had they not, the factories of northwestern Massachusetts wouldn’t have been able to ship their goods so easily to the expanding West, the cost of freight would have remained stubbornly high, and the state of Massachusetts would have been immeasurably poorer." Immeasurably? Really? Why not try to measure the improvement and see whether the tunnel really was worth digging in the first place? Maybe that money would have been better used on something else. (Closer to home, was the Channel Tunnel worth the cost? This concludes that "Property investors in the south of England are known to make regular use of the tunnels which has helped to boost what was already a very lucrative second home market in France for British property investors". That is not the greatest public benefit imaginable.)

Again, "The worst thing that ever happened to incompetent public-school districts was the growth of private schools: they siphoned off the kind of parents who would otherwise have agitated more strongly for reform." Clearly it is possible that that is true, and we may have Hirschman to thank for pointing out that Friedman's free-market alternative ignores that possibility as well as devaluing the political process. But the really interesting result - the sort of thing economists should be bringing to the party - would be to prove the causation and quantify the effects. Would state school pupils really benefit from the abolition of independent schools? And, if so, by how much? 

None of that is an attack of Hirschman, of course, but rather on Gladwell's Hirschman, namely the idea that economics would do better to come up with well-expressed doubts than stolidly-expressed theories.