New Journal Article on the Value of Commitment

A Simple Game of Strategy. To find out more about commitment, consider the following simple game, which takes place over three days: On day 1, a teacher tells you what he will do on day 3 and how committed he is to that action. On day 2, you work and show the teacher what you did. On day 3, the teacher takes an action having evaluated the work you completed on day 2. The more the teacher is committed to his announcement on day 1, the less his action on day 3 can depend on your action on day 2. For concreteness, imagine the teacher informs you on day 1: “In two days, I want to hire you as my assistant for the next year.” On day 2, you decide how hard you work and show your work to the teacher. On day 3, the teacher tells you if he will hire you or not.

What does “perfect commitment” mean? On day 1, the teacher could—if he wanted to—perfectly commit to his later action, by giving you an employment contract right then and there. But given this absolute promise, how hard would you work on day 2? The teacher will hire you, no matter what happens on that day. Since the teacher cannot change his mind on day 3, he might as well not inspect your work at all, and therefore you might decide not to complete any work and to take day 2 off instead.

What does “no commitment” mean? If the teacher does not want to commit on day 1, he could say “I might hire you in two days, and I have been known to tell that to people and then forget about making an offer.” How hard should you work now? After all, the teacher will inspect your output on day 2. If all the teacher wanted is to make you work hard on day 2, then his chances might be better with no commitment than under perfect commitment.

What is “intermediate commitment”? The teacher could try to make it harder for himself to get out of his prospective offer to you by committing himself somewhat. For example, he could say on day 1: “I plan to hire you for a year on the day after tomorrow. Here is a cheque for your first x days of work.” Wow! If x = 0, then we are back to no commitment. However, if x = 365, then the commitment is perfect. By choosing an x in between, the teacher can effectively vary his level of commitment by giving you some of the one-year job you so desperately want already on day 1.

What commitment level is best? The commitment level that is really best for the teacher depends on his and your preferences. But chances are that some intermediate x would give you the best incentives to work hard on day 2, and that is the strategic purpose the teacher’s decision accomplishes. By making it difficult, but not too difficult to deviate from his announced action, the teacher can give you incentives. In fact, the less the teacher commits, the more you have to take into account how your action on day 2 would determine his reaction on day 3. This is called the “strategic effect” of your action, as opposed to its “direct effect.” The direct effect of your work is how much utility it gives him. For instance, if you paint a nice picture, then he might hang it up in his office. But if he already perfectly committed to hiring you the next day, then painting that beautiful picture will not have any strategic effect, which (as any cold-blooded economist might predict) should make you think twice about putting in a lot of work to paint it.

Intermediate commitment in real life. Commitment in real life can be achieved by making it difficult to change the effects of actions. For example, regulators can commit to laws by increasing the voting threshold required to repeal them. Firms can make commitments by investing in research initiatives that are difficult to stop. Individuals can make commitments to themselves, for example, to quit smoking, by asking others to keep a sum of money paid to them now if they have failed at the attempt as determined three months from now. There are many examples. The concept of commitment is related to self-control (which is required or implied when making a commitment) and trust afforded by others (which is implied when your future action is as good as cast in stone). In real life, neither ‘perfect commitment’ nor ‘no commitment’ are actually possible. Every action you take is at least partially irreversible because changing it carries at least some small cost, be it just the disappointment or the decrease in self-confidence it causes. And virtually any announced action can be changed, as the perfect doomsday device does not exist. As a result, all practical commitments are intermediate: they can be changed, at a cost.

Reference
Weber, T.A. (2014) “A Continuum of Commitment,” Economics Letters, in press.
http://dx.doi.org/10.1016/j.econlet.2014.04.013