More thoughts on Israel Kirzner: Dynamics

A couple days ago Pete Boettke shared speculation from Thomson Reuters that Kirzner could win the Nobel this year along with Baumol. I have no thoughts on that - I don't know his work well enough to assess its merits. But that post and yesterday's events got me thinking more about Kirzner than I typically do, and one thing that I've been puzzling over (but didn't have well formulated enough to ask him yesterday afternoon) was what Kirzner would think of "dynamic equilibrium" and really the transition dynamics of that sort of model.


As I noted before, Kirzner's concerned with out-of-equilibrium behavior and if there is convergence to an equilibrium how that behavior gets us there. Contrary to Kirzner's own assertions this is not something that the mainstream is unconcerned with. One answer to this question that has been particularly amenable to modeling is the idea of transition dynamics and dynamic equilibrium. As far as Kirzner's reservations about equilibrium economics go it's something of a misnomer. On the stable arm nothing is stationary, people often aren't satisfied and they aren't finished doing things. But they are acting rationally, and when we identify dynamic equilibrium we are talking about a sequence of rational decisions that lead to a more stable equilibrium. The rapidity of convergence depends on various conditions.


Kirzner is very skeptical of the mainstream theoretical apparatus so I doubt he would think much of the point. But I do wonder what people who see things a little differently but who still appreciate what Kirzner has to say think about how transition dynamics in these models relate to his thinking about the market process. What are agents doing and thinking in dynamic equilibrium that's similar to what they're doing and thinking for Kirzner, etc.


As a side note, Vernon Smith made a vague reference to my line of thinking here in his speech (after Kirzner's). He started by talking about how he's been right a lot in his career, but he's also been wrong a lot and how he doesn't know much about the long-term future of things but he does know the next step and if he takes a series of good next steps he gets along pretty well (he broadened this point to people in general who get along pretty well in the world).


Too me this sounded in a very rough way like Smith was solving a Bellman equation. In this approach to dynamic programming you just worry about being optimal at every point in time. The future is nicely tucked away in the value function and you come to that when the future comes around. You don't waste time or energy worrying about planning and optimizing a long stream of decisions.