This is the third of a series of responses by Mario Rizzo and Glen Whitman, authors of Escaping Paternalism, for my Book Club on their treatise.
Once again, we’d like to thank Bryan for hosting this book club. We also appreciate the many insightful contributions in the comments section. In this post, we’ll discuss a handful of questions raised over the course of the book club but especially in Bryan’s last installment.
But first, a personal request. If you’ve read the book, we would truly appreciate your reviewing it on Amazon!
And now for the remaining topics…
In our earlier reply on this topic, we may have taken the question too literally. For many, “falsifiability” is simply a shorthand way of asking Is this claim scientific? or Is this claim supported by evidence? And we believe the answer is a clear yes for both questions. There are mountains of evidence for inclusive rationality. It takes the form of people engaging in all manner of self-regulatory behaviors, from mental budgeting to diet plans to self-reward-and-punishment schemes to environmental structuring to mutual support groups. It takes the form of people learning over time. It takes the form of people reducing their “biases” in response to higher costs and indulging them more in response to lower costs (a phenomenon that Bryan himself identified and named). It takes the form of behavioral economic research revealing tendencies that – shorn of unwarranted normative judgments – evince the existence of non-standard preferences. And so on.
Some might reply, “What you’re doing is seeking confirmation rather than falsification.” But these are in fact two sides of the same coin. Scientific hypotheses quite often generate predictions about the existence of certain phenomena. If a thorough search fails to turn up examples of such phenomena, then eventually that failure is deemed a falsification. For example, evolutionary biologists hypothesized – on the basis of primitive marsupial fossils in South America and contemporary marsupials in Australia – that transitionary marsupial fossils would be found in Antarctica (which at one point connected South America to Australia). This hypothesis led to explorations that did indeed find such fossils. But what if these explorations had turned up nothing? With enough failures, biologists would eventually have had to move on to other, better-supported hypotheses. That’s how falsification works in evolutionary biology: a search for confirmatory evidence that might not materialize.
The process is similar in the study of human decision-making. If the search for inclusively rational behaviors had turned up nothing, we would have had to reconsider our position. But it didn’t turn out that way.
Children and Opiate Addicts
Even if inclusive rationality is generally correct, there is still the question of its limits. Are there any exceptions? This, we think, is what Bryan really wants to know when he asks about falsifiability; hence his frequent references to children and opioid addicts. We are happy to take these as probable exceptions, at least in many instances. Why do we hesitate to go beyond “probable”? Because we are experts in neither child psychology nor opioid addiction. What we know of them from general awareness and personal observation tends to support that hypothesis, but we don’t want to oversell our understanding.
We also suspect that, in both cases, the story is more complex than “these people aren’t rational.” It seems likely that some child and addict behaviors can be explained as the result of peculiar but not irrational preferences. Other behaviors, especially for children, can be explained as mere ignorance about how the world works (kids don’t necessarily understand that throwing a tantrum won’t make the broken toy fix itself) or testing the limits of their power (maybe the tantrum will motivate Mom and Dad to go buy a new toy). None of this means genuine irrationality isn’t also involved – only that teasing out its role could be a difficult undertaking.
Could other exceptions be found? Sure. Some varieties of mental illness might qualify. There might be some specific behaviors even in “normal” (adult, non-addicted, mentally healthy) people that would qualify. But if our research has taught us anything, it’s that rationality is more complex than you think and often appears where you least expect it. Behavioral economists have found countless “gotcha” cases of alleged rationality – but then closer examination forces us to say, “Hmm, maybe not.” The inclusive rationality approach forces analysts to have some humility, and to examine more carefully rather than indulging the impulse to find fault.
Bryan’s overarching perspective comes out most clearly in Part 5 of the book club. Here, he makes clear his support for a notion of objective welfare.
We should emphasize that Bryan’s position was not our primary target in the book. Many behavioral paternalists (i.e., new-school paternalists) explicitly disavow the notion of objective welfare, even if they sometimes unwittingly slip into that frame of mind. Like most economists, behavioral paternalists embrace the subjectivism of preferences and personal values. We have met them on that playing field.
By contrast, although Bryan is not a paternalist, he shares the old-school paternalists’ fundamental value judgment: that some people’s preferences are just wrong, and they would be better off if corrected. This is a philosophical position we do not share, so perhaps we should simply agree to disagree.
However, Bryan also suggests that we, too, have accepted some notion of objective welfare. In response to a passage in the book where we concede – for the sake of argument – that we could indulge our intuition in certain very extreme cases that people are acting irrationally, Bryan responds: “I agree that ‘intuition’ (or just ‘common-sense’) says this. The reason, though, is that ‘well-considered well-being’ is a thinly-veiled version of objective well-being. The morbidly obese are plainly acting in accordance with their own preferences, but they are acting contrary to their own long-run happiness.”
On this point, we strongly beg to differ. Well-considered well-being is not just thinly veiled objective well-being. We dispute the notion that anyone who thinks carefully enough will choose Bryan’s personal values! On this front, the behavioral paternalists are right: the appropriate standard of well-being is the one you would impose on yourself. If the morbidly obese person looks at his life and genuinely concludes, “You know, all thing considered, this is the life for me,” we, as economists, have no objective basis for saying otherwise. If there are legitimate grounds for deeming this person irrational and possibly in need of help, it’s because his behavior is making him worse off from his own perspective.
We would also observe that the appeal to “common sense” is potentially tyrannical – not in Bryan’s hands, but in the hands of those who don’t share his libertarian value commitments. Common sense is often shorthand for “what we happen to like.” Laden with social desirability bias, common sense can become a cudgel for imposing one’s own values on others. Which is not to say we should never apply common sense; again, we might be willing to indulge that intuition in some extreme cases. But it is playing with fire, so to speak.
To be very clear, we don’t dispute the existence of objective standards. In principle, we can objectively define the choices that will maximize health, or lifespan, or long-term financial wealth. But how should those things be weighed against other values, such as spontaneity, indulgence, and hedonic pleasure? That is a matter of personal preferences and values. Most importantly, tradeoffs among competing values – some or all of which may have a claim to objective – are what is important in any policy evaluation. We cannot say, for example, that people are undersaving if we do not know the appropriate tradeoff between present and future “objective goods” at the margin.
Overapplication of “Irrationality”
A recurring theme of the book is that the term “irrational” has been applied far too broadly. “Irrational” is not a synonym for “wrong,” “undesirable,” “ill-informed,” or any number of other negative qualities. “Rationality” and “irrationality” relate specifically to the suitability of means to ends – that is, how well people match their choices to their goals, whatever those goals might be. At least, that is how we use them in the book. The need to distinguish irrationality from other negative things comes up in Bryan’s fourth book-club post:
RW embrace glorious candor. If we fully embrace their “inclusive rationality,” RW freely admit that we can no longer impugn paternalists as irrational. I’d add, however, that RW expose their opponents’ irrationality so convincingly that we should reject inclusive rationality. Many beliefs are arrant folly, and relabeling this arrant folly as “inclusive rationality” is bending over backwards for no good reason.
We would not want to be accused of supporting “arrant folly”! But it’s worthwhile to recognize that arrant folly can nevertheless be rational. We lay out the reasons why in the chapters on political economy and slippery slopes, but here’s the short version: Even fully rational policymakers can support bad policies. This can happen for numerous reasons. They may lack good information. They may not have absorbed the wisdom of our book! They may care only about what appeals to a rationally ignorant public. They may expect never to bear the long-run consequences of their policies. They may expect to be rewarded for the up-front benefits or illusory benefits of their policies. They may expect to receive greater status and larger budgets in administering such policies. They may enjoy exercising control over others. They may belong to (or take contributions from) interest groups that benefit from paternalistic policies.
Furthermore, as we emphasize in the book, even “irrational” contributors to bad policy – such as action bias, overconfidence, and confirmation bias – are not clearly irrational. They are tendencies with low costs and sometimes notable benefits to policymakers. This is Bryan’s own “rational irrationality” in action: the political arena does little to discourage, and much to encourage, the indulgence of our biases. And we don’t have to concede they are genuine biases to make this point! Some biases are just nonstandard (but totally relatable) preferences, such as “I enjoy feeling like I’m right, and I dislike admitting I’m wrong.” If we conclude that so-called biases like these are in fact inclusively rational, that doesn’t excuse the resulting bad policy. See above: even fully rational policymakers can support bad policies.
Finally, policymakers may not share our values. They may implicitly or explicitly believe in a notion of objective welfare (as Bryan does) and weight it heavily enough to outweigh pragmatic concerns. They may not share Bryan’s libertarian value commitments, which do a lot of the work in helping him reach anti-paternalistic conclusions despite his belief in objective welfare. We disagree with those who arrive at paternalism in this way, but we wouldn’t call them irrational. In the book, although some of our arguments may invoke libertarian values, for the most part we endeavor to provide reasons to resist paternalism that could persuade people who don’t share our ideological priors.