Notes from DRUID
| Dick Langlois |
I am in Copenhagen for the DRUID 25th Celebration Conference, which finished up yesterday. It’s called the 25th Celebration because it’s the 25th DRUID conference -– there have been generally two a year since the organization started in 1995. This conference represents a transition for DRUID, which has grown considerably over the years. One indication of transition is that the old scientific advisory board, of which I had been a member since 1996, has been dissolved and a new one reconstituted. The new board is made up of a number of smart and interesting people, but it tends more toward management and economic geography and away from the theory of the firm and industry as represented by the likes of Bo Carlson, Brian Loasby, and George Richardson (and me). It was perhaps fitting that Brian was asked to press the button that touched off the fireworks over the harbor after last night’s conference dinner.
Steven Klepper presented the first keynote, a further development of his long-term research program on industry structure and the birth and death of firms. (I missed the very beginning of the talk because I was having breakfast with Nicolai; fortunately, the paper is available here.) What Klepper does is essentially top-flight quantitative economic history. In this paper he takes on the conventional wisdom (A) that Silicon Valley is unique because of the rate of spinoffs it engendered and (B) that universities are crucial to the spinoff process. It turns out that the early auto industry in Detroit and the early tire industry in Akron had almost identical spinoff patterns, both sans university. (In fact, there were more spinoffs than in Silicon Valley.) In Klepper’s account -– notably different from most accounts –- clusters arise when new profit opportunities get seized by defection of key personnel rather than through internal diversification. In all cases, the cluster tend to consist of successful spinoffs from already successful firms. Genuine new entrants and spinoffs from less-successful firms seldom prosper. Defections have to do in large part with the dysfunctionality of the parent company, involving a problem either with expectations (as when the soon-to-be defectors couldn’t convince management of the value of their ideas) or of incentives (read: inadequate stock options). There is an interesting connection here with the Penrose/Chandler theory of the growth of the firm. Penrose seems to assume, and Chandler more than assumes, that firms always build internal capabilities and then use their excess resources to diversify internally into profitable related areas. Klepper shows that those opportunities often result in the formation of new firms.
Chuck Sabel gave the keynote before the conference dinner on Thursday. He basically strung together several pieces of his work, but the talk was engaging nonetheless. Part of the talk recycled a debate about the Vanishing Hand that appeared in Enterprise and Society in 2004. (See also my unpublished rejoinder.) I was amused to see my name on one of his slides. As Dan Raff and I were both in the audience, we got to respond during the question period, which we both did in a very cordial way considering the absurd caricature of our ideas he had presented. Much of the presentation outlined new work Sabel is doing with Ricardo Hausmann and Dani Rodrik on industrial policy for South Africa. The idea is that, after the alleged collapse of the so-called Washington Consensus, it is OK to advocate activist industrial policy again. He framed this in a way I liked –- that economic growth involves the removal of constraints. As I pointed out in my comments, it’s an idea that is essentially my own view of vertical integration -– firms integrate (or disintegrate) to remove the constraints to entrepreneurial profit imposed by the architecture of capabilities in which they find themselves. I also like the idea as a way to think about economic growth, though in my view, of course, the constraints that need to be removed are predominantly those imposed by government controls and poor institutions rather than those that would be removed by affirmative government provision of the many kinds of subsidies and industrial public goods Sabel et al. argue for.
One interesting feature of DRUID conferences are debates. Both Nicolai and I have been participants at various times. One of the debates this year was on the value of sophisticated game theory to managerial practice. Sid Winter and Dan Levinthal argued against game theory. Interestingly, they seemed to sway the audience, as the “after” vote was more in their favor than the “before” vote. Of course, it helps to be right. (Eventually a video will be available here, along with videos of past debates.)
Best bons mots of the conference. Sid Winter introduced the term vampirical theory: a theory that is repeatedly killed by evidence but always returns from the dead. And Brian Loasby noted that Merton and Scholes were the first people to win a Nobel Prize for alchemy, as their theory claimed to be able to transmute uncertainty into risk. As their own experience with Long Term Capital Management showed, of course, this form of alchemy was ultimately no better than its medieval counterpart.