Political philosopher Kevin Vallier has an interesting but, I think, ultimately confused post over at Bleeding Heart Libertarians suggesting that consequentialism and value subjectivism—two views frequently held by economists—are in tension. To summarize briefly, he argues:
- Value subjectivism posits that states of affairs have value if and only if people subjectively value those states of affairs.
- Reasons for action are provided by these subjective values.
- Therefore value subjectivism entails that reasons for action are (all) agent-relative.
- Consequentialism gives to all moral actors the same overriding aim: Maximize overall good (however defined).
- Therefore consequentialism entails at least one (overriding) agent-neutral reason for action.
- Therefore value subjectivism and consequentialism cannot both be correct.
There’s a simple mistake here, and a slightly more subtle one. The simple mistake is that economic theories of practical rationality are supposed to be descriptive and predictive: They give an account of our explanatory or motivating reasons. The form of value subjectivism involved here does not entail any deep metaphysical commitments about the nature of value: It requires only the near-tautological premise that if you want to explain how people behave, the only relevant “reasons” are the ones people take themselves to have. This is a familiar distinction that has nothing to do with a gap between “self interested” and moral reasons. If I smash my laptop in a moment of frustration, it is perfectly coherent to say, in a descriptive and explanatory mode, that my reason for acting as I did was provided by my desire to vent my annoyance, and also that I had no good normative reason to act as I did, given my considered goals and preferences. Whatever objective values or reasons for action there may be, they are only motivationally effective to the extent that the values are internalized subjectively, and the reasons for action apprehended. Again, this should not be a controversial metaphysical thesis; it’s a tautology. What you have reason to do depends (in part) on the facts; what is rational for you to do depends on your beliefs. Subjective values are the only ones with explanatory relevance; it doesn’t follow that they’re the only ones with normative relevance.
The slightly more subtle mistake is conflating a view about the structure of motivation with a substantive view about what makes states of affairs objectively valuable—and in particular, with the view that subjective mental states are uniquely capable of making states of affairs objectively good or bad. Everyone has a subjective reason to defect in a one-shot Prisoner’s Dilemma; everyone’s defecting makes the outcome worse, even though subjective levels of satisfaction are the only inputs into that outcome. Despoiling the environment for short-term benefit can make the outcome worse by creating huge amounts of subjective misery, even if no current agents care about the effects of their actions on future generations.
A sufficiently selfish population may have no motivating reason, based on their subjective pattern of concerns, to avert a future ecological catastrophe—and in this case, economic theory can predict that they will fail to do so. They will act morally badly, on one plausible view, because they fail to align their subjective preferences with the objective disvalue of avoidable suffering. That objective disvalue may be wholly dependent on subjective mental states: If our descendants turn out to prefer postapocalyptic hellholes, everything’s copacetic. You can coherently believe that there are objective (normative) reasons for action, but that only subjective mental states can be ultimate sources of such reasons.
I have some more general doubts about the usefulness of making the agent-relative/agent-neutral distinction so central—the real action, as it were, being in the specification of what counts as a reason—but that’s probably a knot to untangle some other time.
Update: Maybe the most concise way to put this is that Vallier isn’t really making an argument against consequentialism, but against externalism about reasons generally. Consequentialism is not unique in suggesting that, for instance, you should not murder an annoying rival when you are certain to get away with it, even if you have a strong present desire to do so. That it may be economically rational to act on this desire just reminds us that theories of economic rationality are not, and are not supposed to be, moral theories.