Politics (theory): Lukacs, computationalist systems, consumer boycotts, Ulrich Beck
I've been meaning to write a follow up to this post for a while now. In the earlier post, I attempted to draw a distinction between positivist and computationalist systems. Specifically, I argued that while positivist systems have as their foundation immutable "givens" (or "facts")--and so are vulnerable to Lukacs' criticism that they exclude the possibility of acting on the system--computationalist systems replace the givens with inputs, over which something external to the (partial) system has control. This prevents the system from ever being truly deterministic.
This may have sounded like totally abstracted intellectual wanking, but I actually had something pretty specific in mind at the time: mathematical economics, especially competitive equilibrium models and consumer theory.
My economics is pretty rusty, since shortly after getting into it I got frustrated with it and moved to psychology. But here's how mathematical microeconomics works: you set up some model, preferably a general one. You count up a number of agents--consumers, producers, whatever--and define them in terms of a number of exogenous variables: their utility curves, the initial endowments of the consumers, the costs of production, etc. Sometimes you make some conditions about the nature of the exogenous variables: utility is a strictly increasing, quasiconcave functions of real-valued vectors that represent the amount of different commodities consumed, etc. Certain functions--which encode the assumptions of the model about, for example, the effect of consumer actions on commodity price--are posited. And then, with the assumption that each agent is acting in an instrumentally rational way, you can prove things about the resulting distribution of goods. It's probably Pareto optimal, since that's something economists consider a promising social welfare result.
Of course, this is recognizably exactly the sort of positivist system that Lukacs complains about. Here you've got your facts, in the form of the exogenous variables and assumptions of the model, and then the about market prices and the distribution of wealth and so on are rationally deducible and necessary results of the facts. There's no real room for agency here. I think that a long time ago I read something--by Peter Boettke, IIRC--saying that at one point, Chicago school economists were so sure that the world worked like this that they spent a lot of time trying to deduce what people's utility functions were based on the current market prices. Then as these models were shown to be more and more implausible, you get people like Joseph Stiglitz who rebuild the models to show that the market is failing because people are underinformed. So the only options for the economic agent are: rational informed behavior (which is predicted by the model), ignorance, or irrationality. And so the "natural laws" of the market get their dual nature as immutable and normative laws.
Lukacs' solution to this problem, I guess, would be to reject the individualist, instrumental rationality of the agents in these models and replace it with the totality-facing rationality of the class-conscious proletariat. But this idea hasn't worked out so well. (Maybe because class consciousness is a chimera?*)
But what if we start looking at the exogenous variables of the model not as "facts," but as "inputs"?
There are lots of directions to go at this point, but I want to choose one: utility functions. In all the economics I ever studied, the utility function representing the "preferences" of the consumers was always considered to be a bare fact posited (or generalized over) before the math was done. But if we hold the rest of the model still, we find that it defines a number of functions from consumer preferences to commodity distributions, market prices, and so on. Even varying just one set of preferences has an impact on the (computed) economy as a whole. If consumer A likes butter less, they will (being rational) be less willing to pay for butter, reducing the price of butter, reducing the production of butter, reducing the number of people employed in butter-making capacity, and so on.
What this sets up is the possibility of the consumer boycott as a means for social change.
There's this guy Ulrich Beck who I think I've brought up a lot, but I forget how much detail I've gone into about this (or anything else here, actually). Also, I've unfortunately lost everything I had written by him. But here was one of his ideas, as far as I can remember and at risk of repeating myself: In today's global power struggle, there are many forces at work. Some are states. One is the monolithic force that is Capital. Capital flies around the world looking for ways to reproduce itself, and it is winning the power struggle against states, because states need capital to hang out with it to flourish, but capital can bludgeon states into doing what it wants by using the threat on non-investment. So if some state doesn't do something that attracts capital, capital flows somewhere else, and that state is screwed.
So what can control capital, if not states? Beck's answer is: the consumers. For capital to grow itself, it needs to be invested in projects that result in consumers buying things. So just as capital can threaten states with non-investment, consumers can threaten capital with non-purchase.
This doesn't make any practical difference if preferences are just givens, since no consumer has any reason to rationally alter their buying habits.** The economy will keep chugging away deterministically until it destroys the world.
But if preferences are flexible, then that means that, at least in the system described so far, there is room for social change. As I said, this is the way contemporary consumer boycotts work. Word gets out that Nike has shitty factory conditions, and thousands of people stop buying Nike shoes. Capital shifts. People become vegetarians because they believe in animal rights or the environment or whatever. Capital shifts again. And so on. The RIAA pisses a lot of people off, and they stop buying their albums and start supporting independent artists. Capital shifts once again. And supervening on each change in the economy is a change in society that is brought about by the action (or inaction) of consumers.
All this, and the hope for more changes like it, all becomes theoretically possible under a computationalist understanding of the economic system. Moreover, we get the possibility of individual action that effects a change in the totality. (Lukacs thinks this is impossible.)
* I can conceive of an account of class-conscious rationality that is not bogus, but I think this rationality entail certain kinds of rationality of individual members of that class. Of course, Marxism says members of the proletariat aren't supposed to be individuals under capitalism, but it looks like the persistently individual, instrumental rationality of people of all classes flies in the face of that. So I think that kind of personal rationality needs to be accepted in our social theory; other prescriptions need to either be compatible or be ignored.
** This isn't quite true. Even holding one's preferences and instrumental rationality constant, ones buying habits could change as a result of new information about the world or (if the model we're using allows for such things) new information about ones own preferences. I'm pretty sure these cases don't break the argument I'm trying to make though. Call me on it if they do.
This may have sounded like totally abstracted intellectual wanking, but I actually had something pretty specific in mind at the time: mathematical economics, especially competitive equilibrium models and consumer theory.
My economics is pretty rusty, since shortly after getting into it I got frustrated with it and moved to psychology. But here's how mathematical microeconomics works: you set up some model, preferably a general one. You count up a number of agents--consumers, producers, whatever--and define them in terms of a number of exogenous variables: their utility curves, the initial endowments of the consumers, the costs of production, etc. Sometimes you make some conditions about the nature of the exogenous variables: utility is a strictly increasing, quasiconcave functions of real-valued vectors that represent the amount of different commodities consumed, etc. Certain functions--which encode the assumptions of the model about, for example, the effect of consumer actions on commodity price--are posited. And then, with the assumption that each agent is acting in an instrumentally rational way, you can prove things about the resulting distribution of goods. It's probably Pareto optimal, since that's something economists consider a promising social welfare result.
Of course, this is recognizably exactly the sort of positivist system that Lukacs complains about. Here you've got your facts, in the form of the exogenous variables and assumptions of the model, and then the about market prices and the distribution of wealth and so on are rationally deducible and necessary results of the facts. There's no real room for agency here. I think that a long time ago I read something--by Peter Boettke, IIRC--saying that at one point, Chicago school economists were so sure that the world worked like this that they spent a lot of time trying to deduce what people's utility functions were based on the current market prices. Then as these models were shown to be more and more implausible, you get people like Joseph Stiglitz who rebuild the models to show that the market is failing because people are underinformed. So the only options for the economic agent are: rational informed behavior (which is predicted by the model), ignorance, or irrationality. And so the "natural laws" of the market get their dual nature as immutable and normative laws.
Lukacs' solution to this problem, I guess, would be to reject the individualist, instrumental rationality of the agents in these models and replace it with the totality-facing rationality of the class-conscious proletariat. But this idea hasn't worked out so well. (Maybe because class consciousness is a chimera?*)
But what if we start looking at the exogenous variables of the model not as "facts," but as "inputs"?
There are lots of directions to go at this point, but I want to choose one: utility functions. In all the economics I ever studied, the utility function representing the "preferences" of the consumers was always considered to be a bare fact posited (or generalized over) before the math was done. But if we hold the rest of the model still, we find that it defines a number of functions from consumer preferences to commodity distributions, market prices, and so on. Even varying just one set of preferences has an impact on the (computed) economy as a whole. If consumer A likes butter less, they will (being rational) be less willing to pay for butter, reducing the price of butter, reducing the production of butter, reducing the number of people employed in butter-making capacity, and so on.
What this sets up is the possibility of the consumer boycott as a means for social change.
There's this guy Ulrich Beck who I think I've brought up a lot, but I forget how much detail I've gone into about this (or anything else here, actually). Also, I've unfortunately lost everything I had written by him. But here was one of his ideas, as far as I can remember and at risk of repeating myself: In today's global power struggle, there are many forces at work. Some are states. One is the monolithic force that is Capital. Capital flies around the world looking for ways to reproduce itself, and it is winning the power struggle against states, because states need capital to hang out with it to flourish, but capital can bludgeon states into doing what it wants by using the threat on non-investment. So if some state doesn't do something that attracts capital, capital flows somewhere else, and that state is screwed.
So what can control capital, if not states? Beck's answer is: the consumers. For capital to grow itself, it needs to be invested in projects that result in consumers buying things. So just as capital can threaten states with non-investment, consumers can threaten capital with non-purchase.
This doesn't make any practical difference if preferences are just givens, since no consumer has any reason to rationally alter their buying habits.** The economy will keep chugging away deterministically until it destroys the world.
But if preferences are flexible, then that means that, at least in the system described so far, there is room for social change. As I said, this is the way contemporary consumer boycotts work. Word gets out that Nike has shitty factory conditions, and thousands of people stop buying Nike shoes. Capital shifts. People become vegetarians because they believe in animal rights or the environment or whatever. Capital shifts again. And so on. The RIAA pisses a lot of people off, and they stop buying their albums and start supporting independent artists. Capital shifts once again. And supervening on each change in the economy is a change in society that is brought about by the action (or inaction) of consumers.
All this, and the hope for more changes like it, all becomes theoretically possible under a computationalist understanding of the economic system. Moreover, we get the possibility of individual action that effects a change in the totality. (Lukacs thinks this is impossible.)
* I can conceive of an account of class-conscious rationality that is not bogus, but I think this rationality entail certain kinds of rationality of individual members of that class. Of course, Marxism says members of the proletariat aren't supposed to be individuals under capitalism, but it looks like the persistently individual, instrumental rationality of people of all classes flies in the face of that. So I think that kind of personal rationality needs to be accepted in our social theory; other prescriptions need to either be compatible or be ignored.
** This isn't quite true. Even holding one's preferences and instrumental rationality constant, ones buying habits could change as a result of new information about the world or (if the model we're using allows for such things) new information about ones own preferences. I'm pretty sure these cases don't break the argument I'm trying to make though. Call me on it if they do.