One of the ideas floating around in the wake of the federal govt’s economic update is the notion that more should be done to promote research and development in Canada. The usual way of promoting this project is to play up the link between R&D, technical progress and productivity: if the payoff from subsidising  R&D would be an improvement in Canada’s anemic rate of productivity growth, then it might be an idea worth pursuing.

There are a couple of reasons to think that subsidising R&D is not the way to improve productivity growth. The first one is raised by EclectEcon: R&D doesn’t have to take place in Canada in order for us to benefit from it. If we can buy the new technology more cheaply than we can produce it domestically, then we’re better off trading for it – just as we trade for other goods and services.

Another is that it’s not clear that a subsidy for R&D would pass a cost-benefit analysis – and that it’s even possible to imagine that too much R&D is being done. Many R&D projects amount to what Philippe Aghion and Peter Howitt call a Schumpeterian process of ‘creative destruction’: displacing an existing technology with one that is (possibly only slightly) better. The private returns to R&D can be enormous: the prospect of displacing Microsoft must animate the imagination of many IT entrepreneurs. But society’s measure of the gain would be the difference between the old and new technologies; just replacing Bill Gates with another multi-billionaire wouldn’t affect society’s welfare very much. Same thing for medical research.

If too much of what the R&D sector does consists of ‘business stealing’, then it’s entirely possible that – from society’s point of view – some of the resources currently engaged in R&D should in fact be be diverted toward other sectors.