Your host spent his lunch hour today attending Glen Whitman's talk today, titled "Against the New Paternalism: Internalities and the Economics of Self-Control". Alas, the amusing word "internalities" never made it into the actual talk, but this should not at all be a distraction from what was a highly enlightening subject, in more ways than one. Of course, in all that follows, errors of interpretation are mine alone, and should not be attributed to Prof. Whitman. Ditto for quotes, which should be regarded as paraphrases.
For those finding the word "internality" opaque, it runs like so: An economic externality occurs when Fat Albert, not being forced by anyone to buy two tickets on a plane, crushes you into half of the seat you paid for. From your perspective, it's an externally-imposed cost, hence an externality. An internality occurs when Fat Albert, reaching for the concession-supply-size box of Twinkies that he bought at Costco, fails to consider the longer-term possibility of having a heart attack as perhaps outweighing the shorter-term pleasure of the next marginal Twinkie. At least that's the theory; the intent of the lecture was to argue that policies attempting to control internalities -- such as "fat taxes" -- are bad ideas.
Prof. Whitman began with a five minute talk about dieting ("en-light-ening"? Get it? [ugh -- ed.]), which was slightly hard to believe insofar as it doesn't look at all like he needs to. Of course, given the fact that SC had never seen him before (he looks remarkably like his brother, though), and has no reference to judge by, the alternate explanation is that he simply found a method that works. But I digress.
The point of his discussion of dieting techniques was that different things turn out to work for different people, and what works for any one individual often is highly idiosyncratic. The nascent SC diet has shown promising results, but features incredibly high start-up costs and an obsessive commitment to record-keeping. Prof. Whitman disclosed his own technique (if he wants seconds, he has to wait a half-hour), as well as his brother's (which your host didn't ask for permission to repeat, so he won't, but it probably would work well for an audiophile like SC, too). None of these three things are identical, though, and that turns out to be important.
Following the diet issue was a discussion of things we now know about preferences via the school of behavioral economics. The most fascinating result from your host's perspective was the discussion of "hyperbolic preferences", the idea that the discount rates we assign to expectations of future rewards are themselves variable and time-dependent. For example, if you tell a person that they can wait 100 days and receive 10 dollars, or 101 days and receive 15 dollars, they'll pick the $15. But if you come back to them on day 100 and say you can have 10 today, or wait a day for 15, they (shockingly to SC) pick the $10. From this, SC concludes that economists should be exempt from human-subjects boards, so that they can slap their idiot test subjects for blowing this no-brainer. More seriously, though, the point is that we do not generally license people to perform such slapping, and that creating a government mechanism for performing it (considering the fat tax as a metaphorical slap to Fat Albert for Twinkie consumption) will run into hard issues, which we'll get to at the end of this post.
The other result of interest is the "peak-end" weighting of preferences. Apparently, a study involving colonoscopies found that people gven an extra period of time at the end of the procedure with the tube in their rectum (which hurts, but not as much as the peak pain level, when the tube is being moved around -- or so I'm told, never having had this procedure myself) reported less pain than people who had the tube removed immediately. SC was so disturbed at the thought this could be true, he had to find the reference as soon as he got home -- here it is. The "peak-end" idea says that we weight our experiences according to an average of the peak of pain/pleasure, and the same measurement at the end of the experience. So a lower average pain produces a better overall experience, even though objectively more pain is being inflicted on patients getting the lower average, with the same peaks.
The point of these discussions is to show that our preferences are hardly all that obvious. As Prof. Whitman put it during a Q/A period after the lecture, "rationality isn't an assumption, it's a skill". Choices that would appear not to be grounded in rationality often have more complex causes, and can turn out to be superior when unintuitive facts about our preferences get taken into account.
This brings us back to the internality problem. Any solution to the diet problem that doesn't account for the surprisingly complicated personal dynamics of every individual is going to go wrong somewhere. Prof. Whitman discussed several kinds of internal bargaining that one can conduct with oneself -- rewards for good behavior, rigorously permitted exceptions (say, only eating Twinkies on weekends), or setting up a separate bank account to limit spending on indulgences (i.e, I can eat all of them I want, but when I've eaten $200 worth of Twinkies, I'm done -- for the week). In the paper linked above, Prof. Whitman casts these agreements as being forged between two parties only you really ever have access to -- your present self, and your future self. As soon as one attempts to construct a regulatory scheme that imposes present-day costs, this interferes with the bargaining conducted between your present and future selves, and has a variety of potentially ruinous effects.
For example, one might not develop the willpower and self-control beneficial to other areas of life if the only reason you don't buy Twinkies is that they cost $20 each. Prof. Whitman allowed that it might turn out (we have no idea at this point) that willpower in one domain doesn't transfer to others, so one might be a glutton with food, but a wise investor, or even a miser in other areas. But in that case, another ill effect is that some people who already exhibit preferences in their lives that go in the direction the policy is trying to steer people will be unfairly punished. Say Uncle Scrooge never spent a dime on anything, followed all the desirable public policy prescriptions -- saved enough for retirement, exercised regularly, didn't take out an adjustable-rate interest-only loan on his house, generally didn't make himself in any way a nuisance to society -- but allowed himself one Twinkie a month. Then that Twinkie goes up to 20x its original cost. Why should Uncle Scrooge suddenly pay for Fat Albert's inability to exercise self-control?
Naturally, this has all sorts of applications outside of dieting -- we could have the same argument about any of the other areas where Uncle Scrooge looks like a model citizen. SC found at least as much value in the discussion over how to recognize personal preferences and deal with them as he did in learning about the economics of self-control. But he'll stop talking about it now, because his future self just ordered him to take a break and hit the gym.
Recent Comments