We want to thank Bryan one more time for hosting this book club, which has been entertaining and enlightening for both of us. Although the discussion has naturally gravitated toward points of disagreement, in truth we and Bryan are largely on the same page. It’s worth enumerating some of our many points of agreement, most of which relate to the policy application of behavioral findings:
- The policy readiness of behavioral research has been greatly oversold. Much less is known about the “standard biases” than we have been led to believe.
- The importance of incentives and learning has particularly been neglected by behavioral economists looking to use policy to correct alleged irrationality.
- Behavioral paternalists have also severely neglected the role of self-regulation and mutual self-help. Bryan questions the extent and efficacy of these self-regulatory measures, but he surely agrees that they exist and are relevant to policy judgments.
- The policy-relevance of behavioral findings is hobbled by the problem of local knowledge. Both the biases themselves and their self-regulation are context-dependent and individualized, which creates big problems for “one-size-fits-all” policymaking.
- The policy-relevance of behavioral paternalism is further hobbled by the interaction of biases, which can offset each other. It’s a mistake to judge the “wrongness” of a single bias in isolation, and thus also to attempt to correct a single bias in isolation.
- Practically speaking, behavioral paternalism provides vague scientific cover for those who would propagate “the feel-good seat-of-the-pants paternalistic policies governments wished to adopt anyway” (Bryan’s words, which we quite like).
- The problem gets even worse if policymakers are afflicted by the same biases that regular people are accused of holding. For reasons including the perverse incentives of politics and Bryan’s own “rational irrationality,” we can expect people to indulge their biases more often in the political sphere, often with deleterious consequences.
- Slippery slope effects are real, both in general and in paternalist policymaking specifically, and the indulgence of biases in the political sphere further greases the slope.
Even outside applications to policy, we and Bryan agree on quite a lot. Specifically:
- Behavioral paternalists have been too quick to deem people irrational, using a strict application of the neoclassical model of complete and transitive preferences. This model allows no room for people to change their minds or discover their preferences over time.
- It is perfectly reasonable and rational to allow some inconsistencies to exist in your mind. As Bryan says, in a sentence we wish we’d written ourselves, “Trying to make all your beliefs consistent is as foolish as trying to keep your house perfectly clean.” We would only amend this to allow for inconsistency of preferences as well as beliefs.
- The evidence supporting the existence of many supposed biases is not as unambiguous as it seems, and we do not understand their operation as well as we’ve been led to believe. Some things that are called “biases” may be nothing of the kind. (Bryan doesn’t buy our reasoning in every specific case, but he seems to agree this is sometimes true.)
So where do we and Bryan differ? There are various small points of difference, but the most important and persistent is that (consistent with the literature) we adopt a subjective theory of value, whereas Bryan believes in a form of objective welfare. We have trodden this ground well in previous posts, so we’re inclined to let it lie here.
However, we do wish to respond to one specific point from Bryan’s most recent post. Once again, he invokes the case of children. But this time, his point is not that children might be inclusively irrational (a point we have conceded), but that, in trying to shape our own children’s behavior, we are revealing our own support for objective welfare! Needless to say, we disagree. The reasons are complicated, but the primary one is that most of the ways we control our children fall in category of “dealing with people who don’t actually understand the world yet.” In Bryan’s specific example, he imagines “a bright child [who] stubbornly insisted that he wanted to play with a loaded gun after you thoroughly warned him of the risks.” We would indeed intervene in this case. But we would do so in large part because we don’t actually believe they genuinely grasp the risks, regardless of how bright they are or how much they insist. Yes, this is a paternalist impulse – appropriately directed at a child rather than an adult – but it doesn’t prove the existence of objective welfare. (There is much more we could say about this example, and the justification for paternalism for children in general, but we need to end this somewhere.)
In any case, our disagreement with Bryan about objective welfare confirms the wisdom of our having given the book a nested “even if” structure. At each stage of the book, we essentially say, “Even if you don’t agree with the arguments so far, we’re now offering you an additional and independent argument for resisting paternalism.” Bryan finds the book’s early arguments only moderately persuasive, but the later arguments strongly persuasive. That works for us! (We suspect that different readers might be more persuaded by the earlier arguments.)
As a final observation, we would note that Bryan’s personal anti-paternalism relies to a great extent on his libertarian (and specifically Huemerian) political philosophy. That’s fine, and we have a good deal of sympathy for that point of view. But as we say in the book, we are not philosophers, and therefore we focus on our comparative advantage: conceptual and consequentialist problems with behavioral paternalism. We endeavor to offer argumentation for the anti-paternalist position that doesn’t require one to come from a libertarian philosophical perspective. In many respects, our book is an immanent critique of behavioral paternalism, and it’s one we hope will cause even non-libertarian supporters of that position to reconsider.