Thursday, March 27, 2008

Paralogisms of Cognitive Microeconomics

I've been having a lot of fun recently with Kant's Critique. It was meant to apply to every aspect of human reason. In this strategy below, I'm arguing that microeconomics is no exception.


Cognitive biases and limitations are a function of individual consumer demand. In order for microeconomic models to play an adequate role in consumer choice theory, these limitations of human reason must be avoided. But they cannot. Interpersonal utility comparisons, and thus social choice theory, are impossible.

A “paralogism” is a logical argument which is not formally invalid, but rather, so ridden with unintentional, paradoxical or ambiguous reasoning—usually involving some sort of equivocation—such that it cannot be accepted in its entirety.

1st Paralogism: All consumer preferences are based on the presupposition that the consumer says to him or herself “I demand...." The individual is the subject and goods and services are the predicates. But economists often confuse the subjectivity of demand with a permanent, interpersonal (and often objective) indifference curve.

2nd Paralogism: Cognitive biases are internalized into indifference curves and thus become functions of the model, with the disclaimer "ceteris paribus". The isolation provided by the model does not lead to isolation of real-world cognitive biases. Cognitive biases cannot be avoided. Yet models which assume bounded rationality avoid cognitive biases.

All attempts to bound rationality, along with interpersonal utility comparisons, are normative economic models, not positive economic models. Positive economics is impossible. This is not an error of microeconomic theory; it is an error of public choice theory. The inherent judgmental and heuristic bias of comparing and abstracting choice models is inescapable for microeconomic theory.

2 comments:

Muser said...

Nice critique. Yes, if one's "self-interest" is socially constructed, then is it self-interest?

Acumensch said...

Hmmm yes, "self interest" might be thought of as a cognitive bias, you're right. That's part of the bounded rationality assumption as well.

But the policy implications of what I was suggesting are that there is no place for authoritarian policymaking. Anything which assumes it can make comparisons between people about their happiness, their indifference, or utility, are going to be subjective. There is no intersubjective truth that economic abstraction can obtain when it uses "utility" and "bounded rationality" as variables in the model.

See my point?

So, antitrust policy, taxation, distribution, price controls, all these things would be relatively pointless for microeconomics to talk about, and hence, pointless for policymakers to talk about as well.

Maybe that's a bit more "hardline" than people would have expected. But I hope elaborating on what I think this argument means in terms of policy helps to clarify my overall argument a bit more.