Buenas ideas sobre política industrial

 Ayer escuché este episodio del Columbia Energy Exchange, y me pareció buenísimo (bueno, casi todo, había trozos donde divagaban un poco). Pero en general una aproximación muy buena a la política industrial, con algunos trozos buenísimos:

I think good industrial policy builds markets not monopolies, and it avoids single point dependencies. And you do that by having milestones and clawbacks and sunset clauses, so you protect taxpayers. And then the last question that I think we have to ask every time, and I don’t just mean privately within government, I mean publicly with democratic accountability and transparency, what’s the exit strategy? I don’t think the goal of any of these interventions should be permanent dependence on public support. There may be certain examples of where that’s necessary, but in general, the government’s role I think should shrink as private capital scales up and markets mature. And the common thread for all those questions is discipline. That’s the most successful examples of industrial policy like the Reconstruction Finance Corporation or DARPA, or, to be fair, Operation Warp Speed. They were all mission-driven. They were all time-bound and they were all grounded in a theory of market failure.

 I think we need to simulate a multi-player, multi-stage game theory of if we use our controls to maintain or expand as largely as possible and our rivals do the same, are we net better off at equilibrium 5, 10, 20, 50 steps down the line? You have to pass a certain threshold of efficacy before you get into this kind of game because otherwise you end up in an escalatory tit for tat and a generation later you ask the question, why the hell are we in this to begin with? 

Let’s institutionalize the practice of industrial policy. What is the strategic objective? What’s the market failure? Why is your intervention solving that failure? How do you sustain competition? What’s your exit plan? I think the public deserves to know and to grapple with the trade-offs, obviously.