Are humans rational or rationalizing agents? How should human behavior be modeled?
Some of the most recent developments in economic theory have found their way into the current student curricula in institutions of economics and management. The area of Behavioral Economics, for instance, builds largely around the empirical evidence against the axioms and algorithms of game theory, and the idea of rationality in economics. Simply put, the content of Behavioral Economics is largely about how individuals frequently fail to fully exploit the benefits of sound reasoning. In other words, people do reason, but not well enough. It is also argued that this aspect of individuals must be taken into account in the understanding of policy implications. To provide a simple example, the mere announcement of price hikes in the near future is enough to generate some immediate economic (spending) activity by the consumers … one doesn’t have to actually raise prices in the future – the consumer frequently fails to question the actual viability of such hikes.
Most of the current literature in Behavioral Economics draws from the results of social (field and laboratory) experiments aimed at testing the predictions of microeconomics and game theory. As it turns out, the general consensus is that neither is a good enough predictor of human behavior (for further details I’d recommend Ariely’s Predictably Irrational and Smith’s Rationality in Economics). So, experiments in economics and psychology take pride in (often) refuting the predictions of microeconomics and game theory (that’s right! You get published for refuting existing theories). There is, however, an interesting point that is being missed out here. If, for instance, game theory was indeed a good predictor of human behavior, then (as paradoxical as it will sound), game theorists would soon enough run out of their jobs. Let me explain: if you and I (or any economic agent) really behaved as per the predictions of game theory, then the need to explicitly study game theory wouldn’t arise at the first place. If British Telecom knew exactly how to design their auctions, then they wouldn’t have hired Prof. Ken Binmore (a big name in game theory). Paradoxically enough, therefore, both of those refuted, and those who refute, are happy with the existing experimental results.
A direct interpretation of this is that, when you read statements like “agent A plays strategy X,” in standard texts of game theory, what is actually meant is “agent A, who is actually a game theorist himself plays strategy X.” Clearly, A isn’t just anybody … and to be fair, game theory, at least in the more advanced levels, requires deep thought – deep enough to prevent just anybody from acquiring the necessary skills. The same arguments hold for formal microeconomic theory.
In my opinion, we overestimate the ability of the human mind to be endlessly rational, and underestimate its ability to endlessly rationalize human action … hence the excessive academic interest in (and, sometimes even heavy reliance on) the former. The latter, on the other hand, is so important, that it has the potential to explain almost all human evil (let alone ‘selfishness’ – the common academic axiom). In a nutshell, humans are not (infinitely) rational, but (again infinitely) rationalizing agents. We’d jump the traffic signal, when we know that no one’s around … even if it is broad daylight. There is a direct satisfaction – we have saved time. The indirect satisfaction comes from our ability to rationalize our action (we didn’t cause an accident … and therefore didn’t hurt anybody … so we’re good souls) … and consequently hold a good opinion of ourselves (although the subconscious self may still suggest otherwise). If you still have a different opinion, I will take the liberty to leave you with a rather disturbing thought … that even terrorists can rationalize all their action … and they often do. You’re free to dismiss the opinion presented here (and the terrorists are free to reject yours).
We handpick our company, principles, teachings, and opinions, to comfort ourselves, whenever our actions are in conflict with our subconscious morality. On a lighter note, any economist, who, on occasion, is found to be stingy, can conveniently blame it on his conditioning in game theory and microeconomics that often assume intrinsic human selfishness.
Development economists frequently need to know how economic agents would react to policy nudges. Marketing professionals need to capitalize on advertisement strategies that trigger impulsive buying. Both economics and business studies aim at understanding human behavior with varied approaches such as microeconomics, game theory, psychology and behavioral studies. While the former areas witness the advantages of mathematical rigor (math is a neutral language, devoid of strong emotion, and involves precise and logical thinking), the latter focus on the role of emotions in the shaping of human behavior which are exceedingly complex for any axiomatic mathematical formalization (although, some of the recent developments in economics and game theory account for such emotions through what are called as models of bounded rationality). It is, therefore, difficult to conclusively reject the merits of any single approach in favor of another. One can’t write any subject off from graduate school coursework. I will only maintain that all the approaches are necessary for a fuller understanding of human action.
To add a final note, the limits of game theory in making predictions about human action has been well acknowledged in the newer approach called bounded rationality … there is a bound to how rational one can be, but (as I have argued before) there is no limit to how much one can rationalize. The behavioral economist often says to the game theorist “humans don’t act according to the dictates of game theory,” to which, the game theorist politely responds “which is why you need us.”