The love of power is the demon of men. 

~ Nietzsche in “Daybreak”

The study of interpersonal power is popular again in social psychology. For example, Dacher Keltner and colleagues have developed a comprehensive theory of the motives underlying psychological states (and traits) of high (and low) power. According to this theory, high-power individuals — or individuals in states of high power — are primarily motivated to exploit opportunities to gather rewards. In contrast, low-power individuals — or individuals in states of low power — seek to avoid risk and harm. Adam Galinsky’s research program on the motivational, perceptual, and behavioral consequence of high (low) power (e.g., Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008) is broadly consistent with Keltner’s theory. Galinsky and colleagues find that high power makes people go and get, perceive the world egocentrically, and care less about their fellow beings. Galinsky et al. have developed several simple and elegant methods of inducing, if only temporarily, a psychological experience of having (vs. lacking) power over at least one other person. This psychological transformation can be achieved by (a) simply recalling an experience of high (low) power, (b) playing the role of a high (low) power person, or (c) being exposed to semantic or symbolic material suggesting high (low) power.

These researchers define interpersonal power as the relative advantage in the control over resources, or more broadly, the ability to administer rewards and punishments. Power relations are rarely entirely asymmetric such that one person has all the power and the other has none. It is sensible to assume that, while power differentials are common, they are not total. The question is: Can they be? In search of an answer, I consider game theory, and the concept of veto power in particular. As Steve Laffey, former mayor of Cranston, Rhode Island, put it during a visit to my classroom, “There’s tremendous power in not giving a crap.” How right was Mr. Laffey?

Let us begin with the Dictator Game. In this game, one player, the dictator, has some funds, say $10, at his disposal. He may give any portion of this amount to the other player (let’s call him the subject — this player has so little power that he does not even have a name in the literature). Many dictators give a little money, with the average transfer approaching 20 percent. The usual interpretation of this finding is that many dictators are not entirely selfish. They have social preferences; they care about the welfare of others, if only a little, and they may be troubled by inequality. At first glance, social preferences are not about power. The power differential is built into the structure of the game. It exists before the dictator makes his decision, and the moment he does, it is gone. Power in the dictator game is self-eliminating. Once it is exercised, it evaporates. Hence, the dictator game is marginally useful as a paradigm for the study of social power. Of course, the game can be played repeatedly under the assumption that power renews itself like an allostatic drive. At second glance, the Dictator Game confounds power over others with power over the self just as it confounds other-regarding preferences with self-regarding preferences. In this game, these preference are negatively correlated. The confound can be removed by giving the dictator an endowment of $10 to keep no matter what, and to allow him to give some fraction (from nothing to all) of another $10 to the subject. The subject is again at the dictator’s mercy, but he need not worry that the dictator’s largesse can only come at the expense of the dictator’s absolute wealth. Relative wealth remains an issue, however.

In the Dictator Game, the subject is helpless. He has no Laffey power. Not giving a crap can only be a psychological consolation; it cannot affect outcomes. In contrast, the Ultimatum Game empowers the subject. The ultimator (the “proposer”) makes an offer as to how divide a purse of say $10. If the ultimatee agrees, the money is divided as proposed. If he does not give a crap (no veto) no one gets anything. Knowing that most people are not rational in the sense that they prefer having something to having nothing, and knowing that they get outraged at grossly unequal offers, most proposers offer more (about 40 percent) than the minimum (about 30 percent) that most responders are willing to accept. In this game, both players have power. The proposer has the power to articulate a deal, and do it in a mildly self-serving way, and the responder has the opportunity to Laffey it. Which type of power do people prefer? Although the responder’s power is categorical and final, the proposer’s power leads to higher expected values in the game. It stands to reason that most interested individuals will prefer to be the proposer. This is one point against Laffey’s theory of power. If negotiations can lead to ultimatums, he who proposes first has an advantage. He forces the other to figure out if he really doesn’t give a crap.

Now consider the Game of Chicken. Here, each party has a choice between cooperation, C, and defection, D. The outcome depends on the combination of the two choices. If both cooperate, CC, they both get their second highest payoff, say 3. If both defect, DD, the result is catastrophic for both, say 1. Yet, each is tempted to defect if there is a chance that the other cooperates. The payoff for unilateral defection, DC, is best, say 4, whereas the payoff for unilateral cooperation, CD, is second worst, say 2. Notice that when both players act simultaneously, power is symmetrical. By moving from cooperation to defection, a player can reduce a cooperator’s payoff by 1 point and a defector’s payoff by 3 points. The effect of this move on own payoff depends on the other’s choice (+1 if the other cooperates; -1 if the other defects). Each player can thus hurt the other more than he can hurt himself. As no one really wants the mutual defection payoff, many cooperate in this game. Veto power remains a possibility and thus a threat.

Brilliant economist Thomas Schelling (1960) juiced up the Game of Chicken by introducing signaling — and thus a paradox of power. If a player manages to switch to defection and send a signal to the other that there is no way for him to go back — even if there were a change of mind (chicken!) – this player has power. The burden of avoiding mutual destruction now falls on the other player, who is still able to cooperate, and thereby getting 2 instead of 1, while allowing the first player to capture 4. With this maneouver, the Game of Chicken turns into an Ultimatum Game. The first player proposes a 4:2 split, which the second player can reject, resulting in 1:1. If we take these ordinal values as absolutes, the proposed split is large enough for most players to accept. The paradox of power is that the first player can gain power over the final outcome (and over the other player by forcing his hand) only by surrendering his own action power. This is strategic behavior at its best and most devious. As in the ultimatum game, there is power in speed. He who disables his own action power first wins. Consider Schelling’s illustration. Two reckless youths are racing towards each other. He who swerves first to avoid collision is the chicken, but he gets to live. The shrewd strategist will throw his steering wheel overboard, letting the other driver know that he better swerve. If the other does not, both die, but the shrewd strategist feels less guilt in the beyond.

Schellingian shenanigans are not possible in the Prisoner’s Dilemma. Here, the rank order of preferences is DC > CC > DD > CD. If you signal defection (or cooperation) to the other player before he does, he will defect if he is rational.[1] What’s good for the chicken is not good for the jailbird.

Now that the brush is cleared, I will soon take a look at what Brams’s Theory of Moves has to say about Laffey power. Stay tuned for the next essay and don’t touch that dial.

[1] Unconstitutionally unable to resist the clever pun — however groan-inducing it my be — I propose to call this idea "signal defection theory." 

Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A. (2008). Power reduces the press of the situation: Implications for creativity, conformity, and dissonance. Journal of Personality and Social Psychology, 95, 1450-1466.

Keltner, D., Gruenfeld, D. H., & Anderson, A. (2003). Power, approach, and inhibition. Psychological Review, 110, 265-284.

Schelling, T. C. (1960). The strategy of conflict. Cambridge, MA: Harvard University Press.

You are reading

One Among Many

Bayes Odyssey

Where will the stat wars lead us?

The Kingdom of Bayes

The priors are low.

Interest of Conflict

Can you assert yourself without losing social capital?