I share the skepticism as to whether self-control problems can be overcome with economic incentives. Yes, some people commit crimes even though they are threatened with punishments, and it may be that their discount rates are too high; the future penalty does not influence them as much as we – or even their future selves – would like. On the other hand, criminal behavior might better be traced to the uncertainty of being caught. In contrast, a voluntary or involuntary system of economic gains or losses with regard to weight can be made virtually certain. If contracts of the sort I have described for the workplace are entered into, participants will have agreed to be weighed on a monthly basis, and there will be no uncertainty about consequences.
The idea is therefore closer to a tax on cigarettes. We could reward non-smokers instead of taxing cigarettes, but the reward scheme is much more susceptible to false claims. How do we know whether someone has really not smoked during the last month? Cigarette taxes are, by the way, quite effective in reducing smoking. And a much higher tax than presently levied would do even more to control the development of new smokers. But taxation is much more difficult for obesity. Even if we could and would tax pounds or body mass indices, that is taxation after the fact. The self-control element suggests that we want to tax people in earlier time periods, as they contemplate an activity (like buying cigarettes), and not later, as they observe the outcomes (like developing lung cancer or gaining weight). We could tax fast food, particular fatty inputs, and so forth, but each of these will be grossly imprecise tools. Moreover, the self-control (or social utility) issues are different from those associated with smoking. A large percentage of people who smoked wished they did not, but a much smaller percentage of people who consumed a bag of potato chips yesterday wished they had not. My point is simply that just as smoking can be affected by economic incentives, so too can obesity rates. But taxes may be well suited in one area, while contracts (with more modest government involvement) do better in the other.
When I agree that we should be skeptical about economic incentives for this problem of self-control, I mean that we have little experience with the matter. Most illegal drugs are very expensive, and that high price seems to do better discouraging new users than addicted ones. If the per-pack tax on cigarettes were much higher than it is at present, a greater fraction of new smokers would undoubtedly be discouraged. Similarly, if our tax system made saving for retirement (arguably another self-control topic) yet more attractive, there would almost surely be more savings, and I do not mean simply a shift from one kind of savings to another. But if gaining weight meant that I would lose $20,000 a year rather than the $2,000 I used in my illustration in the earlier blog post, I would probably not enter into the agreement in the first place. I think I am too risk averse for that bet. That is a “problem” with an opt-in system in the interest of self-control.