Planned Economies

18 Apr 2024 22:51

A topic in economics which gets most of its interest from arguments about socialism. I am devoting this notebook to, mostly, the theory of planned economies, and especially the limits on the feasibility of such planning, rather than what actual putatively-socialist states did (or are doing).

In a cards-on-the-table spirit, I should say that I'm a market socialist, because I think the obstacles in the way of planning a good economy are severe. Fundamentally, I think the difficulties are (1) scaling up and (2) objectives. As to (1), in every way of formalizing the planning problem I know of, the problem gets more difficult as the number of variables to be planned grows --- and the computational complexity is supra-linear, it grows faster than the number of variables. This means that even if planning is perfectly feasible for a small entity, it will become much harder for an assemblage of many entities of that size. (I am even tempted to say that the supra-linear increase in complexity will often mean that there is diminishing returns to increasing the scope of planning, and so a natural scale for planning. But only "tempted", because that just looks at the cost of planning, not its benefits.) As to (2), firms in a monetary economy are handed their objective function to maximize, which is profit. (Non-profits with a clear mission are handed the objective function of minimizing cost.) If we're going to plan a whole economy, though, we don't have that convenience, and need to come up with an objective function, which is a hard, hard political problem. But I have written far too much about this elsewhere.

I also want to distinguish very sharply here between "planned economies" (which are not a great idea for complex and prosperous societies) and planning-in-economies (which is essential and inevitable). Under capitalism, firms draw up and (try to) implement plans all the time --- not perfectly successfully, but what endeavor is completely successful? Also under capitalism, and indeed under pretty much any imaginable rich economy, there are going to be large-scale, long-lasting infrastructure projects which simply cannot be left to market outcomes, but will have to be done by the state, or at least catalyzed by the state. (Think: roads, dams, power grids, schools.) There will be choices involved in whether, where, when and how to carry out those projects, and it sounds reasonable enough that those choices should be informed by some kind of study of their likely consequences, i.e., they should be planned. Moreover, once those choices have been made, precisely because infrastructure is long-lasting, they will shape future economic activity. Since this kind of planning can't be avoided, obviously it's possible --- but it not the same kind of thing as planning a whole economy! (Again, the issue of scale is crucial. And "thinking through the likely consequences" is going to be political.)

See also: Control Theory, Especially Distributed and Decentralized Control; Duality between Knowledge Centralization and Market Completeness; Otto Neurath; Project Cybersyn; The Socialist Calculation Debate