Which Option is Best for Me? A Values-Based Proposal for Behavioral Economists

In this post, C. Tyler DesRoches discusses his recent article in Journal of Applied Philosophy on the normative foundations of behavioral economics.

Emmett is hungry. He only has enough money to purchase either a slice of cake or a piece of fruit. What’s the best option for Emmett? You might think that fruit is his best option. After all, that’s the healthiest option. In a recent article, I defend one way to make sense of this view, by proposing a values-based account of ‘true preferences.’ Let me explain.

You might think that whatever Emmett chooses is his best option. On this canonical view, held by many neo-classical economists, Emmett’s preference for cake or fruit is revealed through his choice and satisfying this preference makes Emmett better-off. One reason to doubt this story is that behavioral economists have consistently shown that people do not possess the kind of revealed preferences assumed by neo-classical economists, which entail that Emmett always chooses his best option. If behavioral economists are correct, then Emmett is a fallible creature who, like the rest of us, makes mistakes. Maybe he really prefers fruit over cake but chooses cake because he suffers from weakness of will, lack of self-control, or some other cause.

Here’s the rub. If we reject the traditional story about revealed preferences, then we still need some other way for deciding what makes Emmett better-off. Perhaps what makes him better-off is not the satisfaction of his revealed preferences but the satisfaction of his true preferences. True preferences have been defined variously among behavioral economists and philosophers of economics, but they’re typically viewed as a person’s ‘real’ or ‘genuine’ preferences. These preferences make room for the idea that Emmett truly prefers some option but fails to choose it. While the notion ‘true preference’ might sound plausible to some, others remain skeptical. I do think there’s a plausible story to tell about true preferences, however. It goes something like this.

Suppose ex hypothesi that Emmett is generally concerned about his health and appearance, and that he deeply regrets his occasional sugar binges. Health and appearance matter more to him than the pleasure that would be yielded by consuming unhealthy options. Emmett is not simply concerned with making one healthy choice between cake and fruit, but with making many successive healthy choices such that his choices jointly constitute a healthy pattern of consumption.

What does it mean for Emmett to be ‘generally concerned’ with his own health? One way to make this idea clear and distinct is to suggest that Emmett is committed to the value of his own health. Emmett’s value commitment is not a fleeting phenomenon but possesses diachronic stability across disparate choice contexts. This commitment, for Emmett, is different from any mere whim or fancy, each of which involves fleeting pro-attitudes. Upon reflection, Emmett approves of his attitude towards his health and is aware of no circumstances such that his attitude would become unstable in response to reflecting on them. Given this stability, Emmett does not deliberate about the significance of his own health in every choice situation. He accepts it as a matter of course. His health is a relatively fixed point in his deliberations and serves as the basis for planning and actions.

Now, it is one thing for Emmett to report that he’s committed to his own health and intends to follow through on a healthy pattern of consumption, but it is quite another for him to act on this value commitment over time. What should Emmett do? Ideally, he would be resolute. He should choose an acceptable plan at the outset, commit to it, and act on it. When he acts on a plan that is grounded in his value commitment, then Emmett’s actions – his choices over time – express his values-based preferences. These preferences are normative, affective, and relatively stable across choice contexts. Other things being equal, the satisfaction values-based preferences is strong evidence that Emmett has been made better-off.

You might think that my caricature of Emmett casts him as an impossible stoic. After all, it seems that he must choose healthy options, always and everywhere! In fact, there are at least three circumstances under which Emmett is justified in choosing unhealthy options. First, he can choose some unhealthy options while remaining committed to his own health. No healthy pattern of consumption informed by the best nutrition science requires a person to avoid every unhealthy option on every occasion. Second, Emmett might decide to reevaluate and abandon his value commitment. If so, then following through on his healthy pattern of consumption could be bad for Emmett. Third, he might find himself in the thralls of a particular choice situation (say, at his best friend’s wedding) where unhealthy options overwhelm Emmett’s ordinary evaluations. In this kind of situation, Emmett might judge that the benefits of choosing the unhealthy option outweighs the costs of sacrificing his values-based preferences. So long as this kind of circumstance doesn’t arise too often and Emmett returns to his healthy pattern of consumption, then Emmett can choose some unhealthy options in this context, too.

Values-based preferences redirect our attention from individual consumption choices to whole patterns of consumption. What matters for a person such as Emmett is that he chooses over time in a manner that expresses his value commitment. While I don’t pretend that values-based preferences apply to every choice situation, they do provide us with a rigorous framework for analyzing the welfare effects for a particular class of choice situations – self-acknowledged self-control problems – ubiquitous to behavioral economics. At any rate, I think this is a promising way to make sense of choosing one’s best option, especially when the main concern is making successive choices over time.