Would you make a promise on someone else’s behalf?
Posted Aug 26, 2019
"Everyone is a millionaire where promises are concerned." —Ovid
"Mexico will pay for the wall." —A lesser-known American poet
Much derided by economists, making promises is an essential ingredient of social life. Societies in which promises are kept are richer and happier than societies in which they are broken or not even made. Promise-making and promise-keeping are interlinked with issues of reciprocity, trust, and cooperation.
I will focus here on promises of positive transactions, setting aside threats, which are a special class of promise. That a promise will be kept cannot be a foregone conclusion. If it were, no promise would even be necessary. The existence of a promise entails the possibility of betrayal. This is what makes it interesting.
A special situation arises when one party, the Prometeur, or "P," makes a promise to a third party, the Bénéficiaire, or "B," on behalf of a second party, the Nettoyeur, or "N." P expects N to clean up her promise to B. Perhaps you think something like this should never happen. But we know it does, if only from time to time.
A parent might promise a reward to a child that she expects the co-parent to deliver, or a priest may promise the penitent a reward that only the big man upstairs can deliver (see, for example, Philippians 4:19). A mundane example might involve a job candidate who is told that she will receive a secondary appointment in a sister department as well—before consulting with that department. Of course, I am only generating these colorful examples from my vivid imagination.
We do know that second-party promises occur. How may we represent the strategic context? As I have done in many previous essays (e.g., Krueger, 2014), I take my cue from The Theory of Moves (Brams, 2011), generate plausible preferences rankings, and then evaluate best strategies. The matrix below shows my attempt for second-party promise-making.
P has a choice between making a promise to B now (rapide) or wait and consult with N first (lente). N has a choice between honoring the promise (accepter) or rejecting it (rejeter). B has interests, of course, but no strategic influence. Hence, we ignore B and explore the interaction as a game between P and N.
P’s primary interest is for N to go along—r/a and l/a respectively yield the highest payoffs of 4 and 3. P’s secondary interest is to choose rapide over lente because no matter how N acts, rapide yields a higher payoff. Hence, choosing rapide is P’s dominating strategy.
[Note: For P, rapide is the dominating strategy. This entails that it is the strategy that maximizes the minimum payoff (maximin) and thereby satisfies the motive of loss aversion. The inverse is not true. There are games, such as chicken or the volunteer’s dilemma, in which maximin is not dominating (Krueger, 2019).]
N’s first interest, when reacting to B’s choice, is to play tit for tat (Nowak & Sigmund, 1992). If P rashly makes a promise over N’s head, N will not honor it (out of a sense of honor). If P is willing to wait, N, in turn, will cooperate. N’s secondary interest is for P to cooperate and consult with N first. The sum of N’s payoffs associated with P playing lente (4 + 2) is greater than the sum associated with P playing rapide (3 + 1).
If P follows the simplest game-theoretic rule, she will play the dominating strategy, make a rapide promise, and see it fail. The result is a Nash equilibrium, but a tragic one. With a bit of social intelligence and a tolerance for a slight inequality favoring the other party, P will show restraint and play lente. This way, everyone, including B, will get what they want, except that P will not have the dubious and short-lived satisfaction of having successfully rushed things and having dominated N. A little restraint makes for a good game.
The promise game is, like the prisoner’s dilemma and other faces games, such as chicken or stag hunt, a non-cooperative, non-zero-sum game (Krueger, Evans, & Heck, 2017). In these games, the exercise of mutual trust, if it can be achieved, outperforms cold game-theoretic calculation. The tragedy is that trust is fragile (Evans & Krueger, 2009).
Let’s bring B back into the game. Why should B believe P’s claim that N will deliver the goods (sweets, eternal life, a secondary appointment)? Either N will provide the goods anyway, in which case P only delivers a prediction and not a promise, or P can make N deliver the goods—against N’s own best interest.
When there is an actual promise and not a mere prediction, P will seem either overconfident or all-powerful. The overconfident are foolish, and the all-powerful are frightening. Is it wise to be dependent on either?
What is more, a secondhand promise has few teeth, because P does not stand to lose much in case N does not deliver. N’s delivery, or lack thereof, will happen much later [in the Philippian case only in eschatological time], at which point P has long pocketed whatever she gets from B in return for the promise. If P and N live long enough to experience N’s rejection of the deal, then P can wash her hands and blame N.
You made me promises promises
Knowing I'd believe
You knew you'd never keep
Brams, S. J. (2011). Game theory and the humanities. Cambridge, MA: MIT Press.
Evans, A. M., & Krueger, J. I. (2009). The psychology (and economics) of trust. Social and Personality Psychology Compass: Intrapersonal Processes, 3, 1003-1017.
Krueger, J. I. (2014). A general power game. Psychology Today Online. https://www.psychologytoday.com/us/blog/one-among-many/201402/general-power-game
Krueger, J. I. (2019). The vexing volunteer’s dilemma. Current Directions in Psychological Science, 28, 53-58.
Krueger, J. I., Evans, A. M., & Heck, P. R. (2017). Let me help you help me: Trust between profit and prosociality. In P. A. M. Van Lange, B. Rockenbach, & T. Yamagishi (Eds.). Social dilemmas: New perspectives on trust (pp. 121-138). New York, NY: Oxford University Press.
Nowak, M. A., & Sigmund, K. (1992). Tit for tat in heterogeneous populations. Nature, 355, 250-253.