Not too long ago, Bill Easterly and Justin Lin squared off at an event at the World Bank over the wisdom of industrial policy in developing countries. While I am sympathetic to Bill's position, judging by the mood of the crowd in the room, I would have to call the debate a tie.
Bill has returned for round two with a new working paper that formalizes some of the arguments he made during that debate. In the Power of Exports, the New York University Professor (and now avid and entertaining Twitterer) points out that tiny Fiji dominates the market for exports to the U.S. in the category of "Women's, girl's suits, of cotton, not knit." More generally, the market for exports tends to be dominated in each country by a few big hits (and what's more, to a particular importing country), whether women's cotton suits or "ceramic bathroom kitchen sanitary items not porcelain." Mathematically, this is described by a power law, where the likelihood of observing a particular value decreases exponentially with the size of that value.
From this observation, Bill concludes that industrial policy—in other words, attempts to discover big hits—is a bad idea. These big hits are so rare that it is much better to leave the discovery process to many decentralized entrepreneurs than a single government agency with limited foresight.
Perhaps I missed something, but I am still left with the question I had at the end of my last post on the topic. Bill argues that the process is akin to finding a single needle in a haystack, but perhaps there are many needles in there? Let's go back to Bill's favorite example of Fiji. Currently, that country dominates the market for women's cotton suits. Does it really make sense that ex ante Fiji couldn't have also achieved a big hit in men's suits? Or cotton blend suits? Or any comparable garment that requires similar inputs of capital and labor and similar distribution networks?
Update: I shouldn't limit this question to Bill Easterly. He was joined on this paper by co-authors Ariell Reshef and Julia Schwenkenberg. Apologies for the omission.
Join the Conversation