Economic rationality

Economists aspire to contribute expertise that is “value-neutral” or “scientific.” In the spirit of the Ruminathans, I take them at their word. Intentions aside, however, I agree with Mary L. Hirshfeld, whose Aquinas and the Market I am reading now, that orthodox economists’ assumptions about what it means to “choose rationally” actually contain deep ideological commitments, whether or not economists acknowledge them. In a way that complements Hirshfeld’s analysis, this post exposes those hidden commitments.

Classic economic arguments – e.g., rent controls reduce overall welfare – take the following form: The economist draws a bunch of supply and demand curves on a Cartesian grid and then shows that “the area between these curves is smaller than the area between those curves.” Q.E.D. The argument constitutes a “proof” because it can be expressed in mathematical terms familiar to anyone who has studied calculus.  

Anyone who has studied calculus will also remember that areas-under-curves arguments relied on a bunch of important assumptions about the shape of those curves, for example that they be “continuous.” The economist’s curves describe patterns of human behavior, and to get the curves to behave in a way that permits calculus-type arguments, economists make certain assumptions about human nature. Specifically, they make assumptions about the structure of human preferences: that we follow a certain logic when we make choices. Although adopted for technical reasons – to make the math work – they have come to define an economic idea of rationality.

At first blush they appear innocuous. Take “transitivity”: If I prefer X to Y, and Y to Z, it stands to reason that I also prefer X to Z. Simple enough. Transitivity is a requirement for rational decision-making, otherwise we’d get stuck in loops.

And admittedly, sometimes we do. We don’t always choose rationally in economists’ terms. Behavioral economists have created an entire sub-field to catalog the ways in which actual behavior systematically deviates from this standard of rationality. But they leave intact the definition of perfect rationality free from biases and limitations.

Another rationality assumption is “completeness.” For any pair of goods X and Y, the rational person either prefers X to Y, or Y to X, or is indifferent between the two. Either you prefer oranges to apples, or you prefer apples to oranges. Or you just don’t give a damn. As with transitivity, completeness superficially seems like a reasonable axiom. It does not force you to commit to a preference: You can remain indifferent. What you cannot do, however – at least not if you want economists to think you are rational – is withhold judgment entirely.

Economic rationality, characterized by transitivity and completeness of a preferences, is supposedly value-neutral because it does not prescribe a “correct” ranking of preferences. And contrary to some naïve critiques of orthodox economics, it does not imply that your narrow self-interest determines your preferences. It is entirely consistent with orthodox economics to say that “I prefer X to Y” where “X” is “peace on earth” and “Y” is “my full belly.” Economics simply describes how you make choices. With the appropriate preferences, it’s perfectly rational to go on hunger strike until all swords are beaten into plowshares.

There is a critique of this view that rests on the idea that some goods are incommensurable: They do not lend themselves to preference ordering. In other words, in addition to the options “I prefer X to Y,” “Y to X,” and “I’m indifferent between X and Y” there is the option: “WTF? X and Y aren’t even in the same universe!”

Although I agree with that view, I think there is a deeper problem with completeness as an axiom of rationality. In my view, a lot of rational and morally fraught reflection has to take place to even distinguish X and Y in such a way that a preference ordering becomes meaningful. The rational choice economist takes that prior reflection for granted, thereby smuggling in ideological commitments.

Consider any pop song. The following is a grammatically sound sentence, quite possibly something you’ve heard or said yourself: “I like the chorus better than the verse.” On the one hand, this sentence is intelligible. On the other, though, it’s beside the point: The chorus is the part of the song that the composer has designed to attract your main attention, and she has crafted the verse to contrast with the chorus in order to achieve that effect. A typical pop song will also contain a bridge that offers another contrast, its singular appearance resetting the mind’s expectations so that the final chorus can be both familiar and a yet new climax. You can clearly distinguish verse, bridge, and chorus from each other, but it is the integrated whole that you appreciate.

Just so with a well-told story, where the stage must be set for the denouement to matter; or a fine meal where the dessert – cloying on its own – rounds out the dining experience.

Clearly, we can draw distinctions – verses from choruses, desserts from entrées, etc. – where the things distinguished are not meaningfully ordered in terms of preference. Instead, it is meaningful to say that one thing serves another, or that they serve each other in different ways: The verse serves to highlight the chorus, but the chorus in turn gives the verse its purpose.

“That’s all well and good for distinctions within pop songs,” you might say, “but economists are talking about choices between pop songs, among which you favor some over others and spend more or less money on to download.” But an individual song may have been crafted by the composer as part of a cycle, or a concept album, in such a way that its deepest, truest impact unfolds in a specific context, and ripped out of that context it might lose meaning or even see it destroyed. We could at least have an argument about whether that’s the case. You may think Bohemian Rhapsody stands on its own, and I may feel it is the keystone in A Night at the Opera. And that argument is the point: Whether we see a given thing as part of an integrated whole or as an integrated whole in its own right is the output of a deliberative process, a process that precedes the establishment of a preference order.

The completeness assumption sneaks in ideological commitments in two ways. First, it effectively endorses the status quo agreement on what goods stand alone, as reflected in how they are packaged and affixed with a price. Rational choice theory does not dictate a correct ordering of that list, true, but it does stack the deck in favor of the dominant theory of what should be on that list. We largely agree that neither verses nor choruses make up unique goods, but individual songs do. The ringtone phase in the music industry is an interesting case study in that regard: It was an experiment in changing the list of goods, and its partial success and ultimate failure demonstrates quite clearly how deliberation about what constitutes a “good” is an ongoing social process.

The second way rational choice theory sneaks in an ideological commitment is by disarming challenges to the status quo. In a sense, there are two ways to interpret a choice. In the rational choice model, my choice reveals my underlying ordering of goods. When I buy lamb in the presence of lower-priced beef, I am expressing an underlying preference for lamb. On rational choice theory, I am rational insofar as my total choices reveal an underlying structure of coherent preferences, having the properties of transitivity, completeness, and in some models at least, “independence of irrelevant alternatives.” My actual preference structure may deviate from coherence so defined, but behavioral economists will no doubt have some kind of label to affix to my imperfect rationality.

Some of the time, we make lamb versus beef choices in a way that expresses an underlying preference. But I would argue that we also make many choices as value experiments to learn what a thing truly is and in what relation it stands to other things. And in many more cases, we deliberately avoid a choice – a choice to not choose – because we are withholding judgment or because we disagree with the framing of the choice. These two approaches – deliberate experimentation, deliberate withholding of judgment – are rational strategies for exploring and revising the list of goods. From the outside, however, the resulting choices may appear incoherent in the rational choice economist’s terms, even as they follow a clear internal logic.

And that is the problem: The deliberative process of defining society’s goods is necessarily a social process involving negotiation. But the rational choice economists have staked out the commanding heights of “rationality” as their own. Those whose internal logic of deliberation results in an incoherent pattern of preferences of the status quo list of goods are disarmed in that negotiation, disqualifying them from participating in their definition of an adult conversation.

When I was a child, I compiled ordered lists: favorite bands, favorite albums, favorite songs on the album, etc. To me, part of putting away childish things has been to give up the lists and the orderings and instead to recognize a thing’s value in how it stands in relation to other things, how it serves them or is served by them. Deliberating with others on the proper relationship between things, between choruses and verses, between dinner and dessert, is precisely what rationality is all about.