Economic Reflections 2: Monetary Behavior Incentives

In ch. 2 of What Money Can’t Buy, Michael Sandel discussed the use of market incentives to change unhealthy behavior such as smoking. Discussing the conclusions from various experimental studies, he noted that “cash incentives seem to work better at getting people to show up for a specific event … than at changing long-term habits and behaviors;” the “most encouraging” study of paying people not to smoke resulted in over 90 percent smoking again within six months (p. 59.) I propose that an economic analysis can indicate why long term behaviors seldom change in such cases.

A competing good arises with the emergence of a non-smoking incentive (NSI;) one can choose either smoking or the NSI, but not both. For a given level of NSI, some number of smokers with a smoking demand DS will find the incentive demand DI greater, and they will choose the NSI and give up smoking.  Increasing the NSI supply (i.e., raising the payment) will increase DI, and more smokers will give up smoking; decreasing the NSI supply (i.e., lowering the payment) will decrease DI and fewer smokers will take it. It would seem that the more smokers accept the NSI, the more the smoking demand DS drops.

I suggest however that DS has (at least) two components, a psychological, or internal, demand DPS and an actual consumption, or external, demand DCS.  Without the NSI, these two demands will more or less converge. With the NSI, the two diverge; the difference between them is a non-monetary opportunity cost of the smoker to accept the NSI. With consumption of the NSI over time, if DPS drops sufficiently a smoker may not resume after the NSI stops. That over 90 percent of smokers resume within six months indicates that DPS does not permanently change, if it changes at all; that is, only DCS decreases, while DPS either does not change or approximately reverts to its former level.

In my view Sandel rightly raises moral questions about a perfectionist polity’s use of monetary incentives to change people’s behavior. Moral considerations aside, however, if my analysis holds, it would seem that a proper understanding of the internal costs to incentivize consumers, and the ability of the incentive to change those internal costs, can be an important step toward understanding whether or not incentives can have the desired outcome.

* This is adapted from a series of one page papers I wrote for an independent study in micro- and macroeconomics; the material included the excellent text The Economic Way of Thinking by Heyne, Boettke, and Prychitko, as well as several texts I brought in from communitarian, market anarchy, and experimental economic perspectives.