On Rationality (Economic Terminology, #1)

As an economist, I often find myself talking past people when trying to explain complicated economic theories.  Surprisingly, this is less because of the in-depth knowledge required, and far more because we aren’t using the same terminology.  Many words used in economic contexts have very different meanings than their common usage.  Utility and value, for one.  Margin, for another.  And perhaps the most common source of confusion is the concept of rationality.

In common usage, “rational” basically means “reasonable” or “logical.”  The dictionary definition, according to a quick google check, is “based on or in accordance with reason or logic.”  Essentially, in common usage a rational person is someone who thinks things through and comes to a reasonable or logical conclusion.  Seems simple enough, right?

But not so in economics.  Traditional economic theory rests on four basic assumptions–rationality, maximization, marginality, and perfect information.  And the first of those, rationality, is the single biggest source of confusion when I try to discuss economic theory with non-economists.

To an economist, “rational” does not in the slightest sense mean “reasonable” or “logical.”  A rational actor is merely one who has well-ordered and consistent preferences.  That’s it.  That’s the entirety of economic rationality.  An economically rational actor who happens to prefer apples to oranges, and oranges to bananas, will never choose bananas over apples when given a choice between the two.  Such preferences can be strong (i.e., always prefers X to Y) or weak (i.e., indifferent between X and Y), but they are always consistent.  And those preferences can be modeled as widely or narrowly as you choose.  It could just be their explicit choices among a basket of goods, or you could incorporate social and situational factors like altruism, familial bonds, and cultural values.  They can be context dependent–one might prefer X to Y in Context A, and Y to X in Context B, but then one will always prefer X to Y in Context A and Y to X in Context B. It doesn’t matter: what their preferences actually are is irrelevant, no matter how ridiculous or unreasonable they might seem from the outside, so long as they are well-ordered and consistent.

This isn’t to say preferences can’t change for a rational actor.  They can, over time.  But they’re consistent, at the time a decision is made, across all time horizons–if you give a rational actor the choice between apples and bananas, it doesn’t matter whether they will receive the fruit now or a day from now.  They will always choose apples, until their preferences change overall.

An irrational actor, then, is by definition anyone who does not have well ordered and consistent preferences.  If an actor prefers apples to bananas when faced with immediate reward, but bananas to apples when they won’t get the reward until tomorrow, they’re economically irrational.  And the problem is, of course, that most of us exhibit such irrational preferences all the time.  For proof, we don’t have to look any further than our alarm clocks.

A rational actor prefers to get up at 6:30 AM, so he sets his alarm for 6:30 AM, and wakes up when it goes off.  End of story.  An irrational actor, on the other hand, prefers to get up at 6:30 AM when he sets the alarm, but when it actually goes off, he hits the snooze button a few times and gets up 15 minutes later.  His preferences have flipped–what he preferred when he set the alarm and what he preferred when it came time to actually get up were very different, and not because his actual preferences have changed at all.  Rather, he will make the same decisions day after day after day, because his preferences aren’t consistent over different time horizons.  The existence of the snooze button is due to the fact human beings do not, in general, exhibit economically rational preferences.  We can model such behavior with fancy mathematical tricks like quasi-hyperbolic discounting, but they’re by definition irrational in economic terminology.

And that’s why behavioral economics is now a major field–at some point between Richard Thaler’s Ph.D research in the late 1970s and his tenure as the President of the American Economics Association a couple years ago, most economists began to realize the limitations of models based on the unrealistic assumption of economic rationality.  And so they began to start trying to model decision making more in keeping with how people actually act.  Thaler last year predicted that “behavioral economics” will cease to exist as a separate field within three decades, because virtually all economics is now moving towards a behavioral basis.

In future editions of this series, we’ll look at other commonly misunderstood economic terms, including the other three assumptions I mentioned: marginality, maximization, and perfect information.