What preferences do you want?
In traditional neoclassical economics, that’s a nonsense question, because preferences are taken as idiosyncratic, fixed, and exogenous. Of course, that’s only a modeling assumption; no one really claims that’s descriptively accurate.
But using that modeling assumption means that we miss a lot of richness in the process of deciding what we want. Sometimes we work on what we want to want.
There is a famous paper by two Chicago economics Nobelists, Gary Becker and George Stigler, entitled “De Gustibus non est Disputandum,” often translated “There’s no accounting for/arguing over tastes.” Their point was that taste is obviously important, but that it can’t be used directly as an analytical tool.
The logic was that reasoning from changes in quantity demanded to preference changes should be the last resort, not the first. Economists should focus on changes in parameters such as price, the price of other products, income, and regulatory constraints or changes in transaction costs. The discipline imposed by this approach forces the economist to look at market institutions and market parameters.
The Counterrevolution of Preference
There were dissenters from this position, of course. Douglass North, one of my dissertation advisers, encouraged economists to take a closer look at preferences, precisely because he thought that economists would be able to use an economic approach to understand what was going on. People, or at least most people, want to be seen, and even to see themselves, as moral. But this preference, like all economic preferences, might very well be something that can be modeled in terms of trade-offs.
Suppose you found a wallet, one that contained a recent ID and some money. Would you contact the person and try to return the wallet? Suppose you were reasonably certain that no one had seen you pick up the wallet; in one sense, the “rational” thing to do would be to take the money and throw away the wallet. Here are the alternative outcomes from which to choose:
Return the wallet: transaction costs of looking up the person, finding them, and returning the wallet. Assume that’s going to cost you an hour, and maybe more, of valuable time.
Keep the money, and throw the wallet away. You get an immediate net benefit of however much money is in the wallet.
An economist might say that we can’t say anything about where the underlying preferences come from, but we could say this:
The more money in the wallet, the higher the “price” of returning it, so the person is more likely to keep it as the amount of money increases.
The more difficult it would likely be to find the person, the greater the “cost” of returning it, so the person is more likely to keep it as the size of the city increases
The wealthier the person, the more likely that person is to return the wallet, if the preference to perceive oneself is a normal good, or luxury good, meaning a preference for thinking of oneself as honest increases with income.
I may have this wrong, of course. You could make other arguments for the signs of the effects. But the point is that it should be easy to perform experiments that look at the effects of behavior, at the margin, of acting on a preference to be “honest.”
Hero or Idiot?
Once I bought some things at one of those big box home stores, call it DIY Depot. One of the things was a HEPA room air filter, because my younger son was having trouble with allergies. I wasn’t paying much attention, and paid with a credit card. When I got to the car, I looked at the receipt, and the charge had been less than the price on the air filter machine alone. The cashier hadn’t charged me.
I immediately went back in, with the air filter machine in its box, and showed the cashier the receipt. I said, “I’m sorry, I didn’t notice, but you didn’t charge me for this!”
If I expected her to be amused, or happy, I was dead wrong. She turned red, and was almost shouting. “Why would you come back? Are you trying to get me in trouble?” I said no, I just wanted to do the right thing. At first, she refused to ring up the machine, and then did it angrily. “Do you think DIY Depot even cares? It’s just a big corporation. You aren’t a hero, you’re an idiot!” Chagrined, I paid for the machine and went out to my car.
It’s possible the cashier thought her boss would see this, or find out somehow, and she’d lose her job. But I don’t think that was it. I think the obvious explanation is the correct one: she was working for just over minimum wage, and the cost of that filter, which was clearly a medical device and not something frivolous, was more than $100. She was probably right that DIY Depot threw away 10 times as much merchandise every day, so my coming back in and paying wasn’t impressing anyone.
Really, I think she was angry that I was wealthy enough that going back and paying would even occur to me. I had gotten lucky; no one was going to catch me, and in fact no one would even know. To her, insisting on paying was just showing off.
Suppose that I had been much poorer, and that the device cost $10,000. Suppose further, though it’s implausible for something that valuable, that I could be certain that no one would know, or think worse of me. Would I have taken the machine back inside then? I guess I’d like to think so, but the point is that I may well have a reservation price where I’d take the windfall and leave.
Different people have different reservation “honesty prices,” and those prices are likely influenced, but not completely determined by, their income. On the other hand, there are some very poor people who would likely always do “the right thing,” and some wealthy people would sell out their honesty for just a few dollars, because they don’t value that self-perception very much.
The Preference Store
All of this brings me back to where I started. What preferences do you want to have? The question of meta-preference is also an economic problem, though the question of “supply” is rather murky. Imagine, though, that you went to the mall one day, and you noticed a store, between the Spencer’s Gifts and the Mr. Dunderbak’s, a store advertising preferences.
You walk in, remembering that your spouse had told you to pick up some preferences, because you had run out. You look around. Up front, on the aisle end caps, are some really cheap preferences: “Potato chips are preferred to fresh homemade bread is preferred to caviar.” The price of this preference is really low, because a person with this preference can be very happy eating cheap junk food.
In the liquor section, the cheap preferences are “Wine coolers are preferred to California red blends are preferred to French champagne.” A person with these preferences would never buy champagne, but it’s not because of the cost. They actually like carbonated wine coolers better, and achieve a higher level of satisfaction buying a cheap product.
The point is that you should want to prefer cheap stuff. It is irrational for a poor person to cultivate a taste for expensive things. You have to be able — to use Doug North’s phrase — to “afford your preferences.” Of course, if you are wealthy, and a meta-preference for expensive preferences is a “normal good,” then you might choose a different preference, one which implies you like excellent red wine, and can distinguish not just grape varieties but different vintages.
Thus, the question “What preferences would a rational person choose?” is not nonsense.
An interesting side point is the question of interpersonal utility comparisons, but not for different people. The question is, can one person compare the levels of happiness s/he feels as a result of cultivating a taste? It drives my friend Geoff Brennan of Australian National University crazy that I drink wine that comes in a box. He has wine tastings, from his extensive (and expensive!) wine cellar, and the best I can manage is to distinguish if the wines are red or white, and that’s only if I can see them. But I love wine, and often drink it.
Would I be happier if I “bought” the more expensive preference of cultivating a discerning taste? It’s true I would understand wine better, and appreciate distinctions that now escape me. And I could “afford my preference” if I did decide to cultivate a taste for better wine.
So, that let’s us hone the question a bit: would a rational person cultivate a taste for expensive rather than cheap wine? The answer would depend on the person’s meta-preference for having such a taste, the person’s income (a person is less likely to enjoy the cultivated taste, given the opportunity cost of other products forgone), and the price of expensive (and high quality) wine available to that person.
Economics of Preferences
There have been several calls to consider an economic inquiry into preferences and the ways that preferences interact with constraints. A famous one was by Denzau and North (1994), who were interested in the way that preferences might be shared, and might even be objectively “incorrect,” in the sense that people might hold, especially in a political context, beliefs about causal mechanisms that are empirically false. The problem is that it may be hard to update a preference, because it is more than an empirical belief.
Some of the best work on this question has been done by Bryan Caplan, both on political preferences, and — in some ways even more interestingly — on economic or social preferences. Earlier, Geoff Brennan and Loren Lomasky suggested that people might vote for things other than their objectively political ideal — in effect, going to the preference store (my words) and picking the preference they want to express — rather than choosing the vote to express based on their most desired outcome. But Caplan extended this insight in some interesting and important ways.
I think there are a lot of interesting questions here. Given the new ability to do randomized control trials online, I hope that some of the young, or not so young, scholars who read this will consider working on this question.