Dave Donaldson (Professor of Economics at Massachusetts Institute of Technology, United States) is one of the two keynote speakers at the upcoming annual 13th Tinbergen Institute Conference: International Trade and Development on June 21-23, 2018 at Tinbergen Institute Amsterdam. In this column Donaldson writes about his work on industrial policy and its potential effects and what to expect at the conference. Registration for this conference is open until June 12, 2018.
My talk is about industrial policy and its potential effects. Perennial arguments for industrial policy rest on three beliefs. First, that production processes display external economies of scale—such that a nation’s productivity in a given sector is increasing in its scale in that sector, and in a way that is not fully internalized by the firms in that sector. Second, that such scale economies differ across sectors—such that any policy-induced expansion of scale in one sector does not just lead to an equal and opposite contraction of productivity in some other sector. And third, that countries produce highly substitutable and tradable goods—such that a country can simultaneously expand scale in one sector without driving down the price of its own output, and find useful foreign alternative versions of the goods in the sector that it chooses to shrink.
The problem is that it is notoriously difficult to credibly estimate aggregate economies of scale and hence put numbers on these three forces. In my recent work with Dominick Bartelme, Arnaud Costinot and Andres Rodriguez-Clare we have set out to make progress here and in that way arrive at a better understanding of when and where industrial policy might succeed.
We start by showing how international trade data can be used to circumvent two well-known obstacles to credible estimation of aggregate economies of scale: the difficulties of measuring aggregate productivity when products proliferate in their unobserved quality levels and their dauntingly complex patterns of substitutability; and the statistical bias caused when observed scale is codetermined by both supply and demand forces. Our estimates, based on this methodology, imply that external economies of scale do indeed exist (in all of our 15 traded manufacturing sectors), and do indeed differ across sectors.
But when we go on to ask what these estimates imply for the gains from industry policy, our results are sobering. Even when implemented perfectly, and in a way that is optimized for one country (to the partial detriment of other countries), the benefits of such policies appear likely to be quite small — they average just 0.3% of GDP. Given the many real-world challenges involved in implementing such idealized policies, the gains from industrial policy are likely to prove elusive.
Dave Donaldson is a Canadian economist and a professor of economics at Massachusetts Institute of Technology, United States. He was awarded the 2017 John Bates Clark Medal by the American Economics Association, recognizing his innovative and scholarly contributions in the field of international trade. Donaldson has conducted extensive research on international trade, with applications to the return on infrastructure investments, climate change, food security, and optimal policy design. These contributions have lead to publications in top journals such as Review of Economic Studies, American Economic Review, Quarterly Journal of Economics, and Journal of Political Economy.
Learn more about the 13th Tinbergen Institute Conference: International Trade and Development.