Non-Becker Rationality and Kahneman Irrationality

This article provides two Ideological Turing exercises: One argument in defense of hard-line Chicago rationality and one in defense of the existence of economic irrationality.

I don’t think Gary Becker or any of the big Chicago names (eg Friedman) ever firmly stated that individuals are rational (quick proof and another one), but let’s suppose that he did. He could make the following argument: For any given set of action there is some objective function which could have been maximized to render such action as rational.

Let’s call the above statement Non-Beckerian Objectivity. It’s non-Beckerian because Becker never said that and I don’t think he would. On the other hand we have Beckerian rationality. It is called such because Becker did actually say it. It’s defined as “maximization of a well-ordered function, such as a profit or utility function.”

Given Non-Beckerian objectivity, Beckerian rationality becomes tautological and unfalsifiable. We can call the combination Non-Beckerian Rationality. Again, I think this is out of keeping with the Chicago tradition which is oriented around empiricism and falsifiability, a la Friedman for example. Other Friedman statements:

  1. People may not be rational, but they act as if they are rational (in the long run at the market level).
    1. Becker’s correlate: They also act as if they are irrational. Because market efficiency is invariant to rationality.
  2. Most good models involve “wildly inaccurate descriptive representations of reality”
    1. Classic Chicago/neoclassical price theoretic assumption include that individuals are perfectly rational, perfectly informed, know their own preferences, know the present discounted value of the entire future, live forever, can rapidly make complex calculations, and can consume infinitesimal quantities.

These are consistent with the Chicago view as being between rationality-agnostic and friendly toward some degree of individual irrationality, leaning toward the latter.

I’ve made it clear that I disagree with the idea of hard-lined, tautological rationality as representative of Chicago, but for the Turing exercise I gave a coherent argument to that effect vis a vis Non-Beckerian Rationality. On the other side of things, I’m not a fan of hard-lined irrationality proponents as well. Like Becker is often straw-manned as a hard-lined rationalist, I think Kahneman is often straw-manned as a hard-lined irrationalist.

So let me give a Turing exercise and give two arguments in favor of the idea that Kahneman is a hard-lined irrationalist. Based on Thinking, Fast and Slow we can make a Two Systems Argument and a Two Selves Argument.

The two systems argument:

  1. A Kahneman agent has two systems with different objective functions
  2. If the agent fails to maximize either function they are irrational
  3. These systems present divergent objectives
  4. At any point in time the agent is maximizing on only a single system
  5. Therefore, at any point in time the agent is irrational

The two selves argument:

  1. A Kahneman agent has two selves with different happiness maximizing functions
  2. If the agent fails to maximize either function they are irrational
  3. These systems present divergent objectives
  4. At any point in time the agent is maximizing on only a single system
  5. Therefore, at any point in time the agent is irrational

In truth I think Kahneman would agree with Becker and Friedman in saying something like most people tend to act rationally most of the time, subject to certain constraints and biases. Perhaps he would go further and agree with his co-laureate Vernon Smith’s concept of ecological rationality and go so far as to say people are perfectly rational subject to constraints and biases. Clearly Kahneman acknowledges cognitive bias and constraints, but I think it is a bad interpretation of his work to say that he is a hard-lined irrationalist.

  •  
  •  
  •  
  •  
Tagged with: , , , , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *

*