PSY22-6a

Psychological Science 21(5) 737 –744 © The Author(s) 2010 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0956797610368810 http://pss.sagepub.com

Anecdotal evidence from various domains of society suggests that power undermines people’s sense of morality, corrupting their thoughts and behavior. In the political domain, newspa- pers repeatedly report on government officials who have extramarital affairs despite decrying the breakdown of family values or who use public funds for private benefits despite condemning governmental waste. This pattern of sanctimony combined with lechery and gluttony has led some observers to suggest that double standards are the hallmark of politicians (Runciman, 2008). In the economic domain, captains of indus- try have recently asked the government for billions of dollars to protect their banks, industries, and companies from eco- nomic ruin, but at the same time have secured financial bonuses amounting to millions of dollars for themselves while continuing to clamor about the divinity of free-market capital- ism (Kanagaretnam, Lobo, & Mohammad, 2008).

In the research we report here, we aimed to test whether there is a direct, causal link between the experience of power and moral hypocrisy. By moral hypocrisy, we mean a failure to follow one’s own expressed moral rules and principles. We propose that power increases hypocrisy, so that the powerful show a greater discrepancy between what they practice and what they preach than the powerless do. Given that powerful individuals often make crucial decisions that have moral con- siderations, the question whether power increases moral hypocrisy is important. Nonetheless, the relationship between power and hypocrisy has not been tested empirically.

Hypocrisy

Although the terms immorality and moral hypocrisy are some- times used interchangeably, it is important to emphasize their conceptual difference. Someone who behaves in an immoral manner is not a hypocrite if he or she admits that this behavior is immoral and does not expect other people to behave any better. Hypocrites, as the opening examples highlight, are peo- ple who publicly uphold strict moral norms, expecting and demanding others to follow them, but who privately violate these espoused standards in their own behavior.

The phenomenon of hypocrisy has been approached in two ways in the research literature. Some authors have studied the discrepancy between what respondents think is normative and how they actually behave (Batson, Kobrynowicz, Dinnerstein, Kampf, & Wilson, 1997; Batson & Thompson, 2001; Batson, Thompson, Seuferling, Whitney, & Strongman, 1999; Stone, Wiegand, Cooper, & Aronson, 1997). Others have studied the discrepancy between what respondents believe other people should do in a given situation and what they actually would do themselves (see Valdesolo & DeSteno, 2007, 2008). In the

Corresponding Author: Joris Lammers, Tilburg Institute for Behavioral Economics Research (TIBER), Department of Social Psychology, Tilburg University, Warandelaan 2, 5037 AB Tilburg, The Netherlands E-mail: [email protected]

Power Increases Hypocrisy: Moralizing in Reasoning, Immorality in Behavior

Joris Lammers1, Diederik A. Stapel1, and Adam D. Galinsky2 1Tilburg University and 2Northwestern University

Abstract

In five studies, we explored whether power increases moral hypocrisy (i.e., imposing strict moral standards on other people but practicing less strict moral behavior oneself). In Experiment 1, compared with the powerless, the powerful condemned other people’s cheating more, but also cheated more themselves. In Experiments 2 through 4, the powerful were more strict in judging other people’s moral transgressions than in judging their own transgressions. A final study found that the effect of power on moral hypocrisy depends on the legitimacy of the power: When power was illegitimate, the moral-hypocrisy effect was reversed, with the illegitimately powerful becoming stricter in judging their own behavior than in judging other people’s behavior. This pattern, which might be dubbed hypercrisy, was also found among low-power participants in Experiments 3 and 4. We discuss how patterns of hypocrisy and hypercrisy among the powerful and powerless can help perpetuate social inequality.

Keywords

power, morality, hypocrisy

Received 6/15/09; Revision accepted 10/5/09

Research Article

738 Lammers et al.

studies reported here, we used both approaches, examining (a) the discrepancy between expressed standards and behavior and (b) the discrepancy between judged appropriateness and defense of a moral transgression when it is committed by one- self versus another person.

We posit that power increases hypocrisy because, com- pared with the powerless, the powerful are both (a) stricter in their moral judgments on how people should behave and (b) more lenient in following these moral norms themselves. The effect of power on the expression of moral standards goes in completely opposite directions depending on whether those standards refer to how other people should behave or how one personally does behave.

Why the Powerful Are Strict in Moral Judgments About Other People Intuitively, one would suspect that the degree to which people judge others’ behavior depends exclusively on the degree to which that behavior is morally objectionable. Research has shown, however, that people’s inclination to judge also strongly depends on more peripheral factors, such as whether they feel entitled to judge other people (Yzerbyt, Schadron, Leyens, & Rocher, 1994). The experience of power might very well increase such feelings of entitlement: High-power roles are culturally associated with the right to judge other people in order to maintain moral norms (Foucault & Gordon, 1980). Individuals in high-power roles, such as judges, teachers, or police officers, are socially entitled to lay down rules and to demand and ensure that other people follow them. This hypothesis—that feelings of power make people more likely to judge others and be more strict in their moral judgments when doing so—is strengthened by our previous demonstra- tion (Lammers & Stapel, 2009) that the powerful tend to be more focused on rules and less willing to make exceptions to those rules than the powerless are. In addition, powerful people are more inclined to voice their opinion, telling other people what to do and what to refrain from doing (DePaulo & Friedman, 1998; Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008). Also, anger—an emotion associated with increased power (Keltner, Gruenfeld, & Anderson, 2003)—is associated with increased judging of other people (Chaurand & Brauer, 2008). All in all, the powerful appear to create strict and broad constructions of moral standards.

Why the Powerful Are Less Strict in Their Own Moral Behavior Despite the previous demonstrations that the powerful espouse exacting moral standards, other research has found that they are less strict than the powerless in the morality of their own behavior. According to Keltner, Gruenfeld, and Anderson’s (2003) approach/inhibition theory of power, the powerful are more focused on the potential rewards of any action than the powerless are and therefore tend to follow their self-interest

more (see also Galinsky, Gruenfeld, & Magee, 2003; Guinote, 2007; Smith & Bargh, 2008). Normally, social disapproval acts as a check against self-interest (Batson et al., 1997; Gil- bert & Jones, 1986), but feelings of power reduce sensitivity to social disapproval (Emerson, 1962; Thibaut & Kelley, 1959), thus reducing the grip of social norms and standards on power holders’ behavior (Galinsky et al., 2008). As a result, even very strong norms, such as those regulating sexual behavior or compassion, are often ignored by the powerful (Bargh, Ray- mond, Pryor, & Strack, 1995; Van Kleef et al., 2008).

Moral Hypocrisy and the Crucial Role of Legitimacy On the basis of these two separate literatures, we propose that power inspires hypocrisy: It makes people stricter in moral judgments of others but less strict in their own behavior. We tested this hypothesis in a series of experiments in which we manipulated power and then measured hypocrisy. Further- more, we propose that a crucial factor driving these two effects is that a position of power carries with it a sense of entitlement (De Waal, 1983; Overbeck, 2009; Weber, 1948). As a result, the powerful feel they are entitled to deviate a bit from the moral rules that they demand other people follow. If a position of power is denied a sense of entitlement—for example, because the power is seen as illegitimate—the effect of power on hypocrisy should be eliminated. Powerful people who feel that their position is illegitimate are less inclined to assertively take what they want (Lammers, Galinsky, Gordijn, & Otten, 2008) and at the same time are less inclined to judge others for doing so, compared with people who feel their power is deserved (Chaurand & Brauer, 2008). Therefore, in our final study, we independently manipulated power and its legitimacy to test whether legitimacy crucially moderates the effect of power on hypocrisy.

Experiment 1: Judging Versus Engaging in Cheating In our first experiment, we manipulated whether participants experienced a feeling of high or low power and then measured hypocrisy, which we defined in this case as the discrepancy between expressed standards (what respondents think is nor- mative) and actual behavior (following Batson et al., 1997, 1999). We asked half of the participants whether cheating was acceptable and gave the other half the actual opportunity to cheat. We predicted that, compared with the powerless, the powerful would be less accepting of cheating, but would actu- ally cheat more.

Method Participants and design. Participants were 61 Dutch stu- dents (47 female, 14 male; mean age = 19.3 years) who took part for course credit. They were seated in closed cubicles and

Power and Hypocrisy 739

were randomly assigned to one condition of a 2 (power: high vs. low) × 2 (dilemma: judgment vs. behavior) between- participants design.

Procedure. We first induced a sense of power (high vs. low), using an experiential power prime. Participants in the high- power condition were asked to recall an experience of high power, and those in the low-power condition were asked to recall an experience of low power (Galinsky et al., 2003).

For half of the participants, we measured cheating behav- ior, using a paradigm adapted from Fischbacher and Heusi (2008). Participants were told that they would be compensated for time spent in this experiment by participating in a lottery for one of three prizes (€100, €50, and €25). Depending on the number of tickets they won, they would have more or less chance to win a prize. In the privacy of their cubicle, partici- pants used two differently colored 10-sided dice to determine their own number of lottery tickets. That is, they rolled each of the dice once and determined the resulting number themselves. Logically, the theoretical outcome lay between 0 and 99, and the mean outcome per condition should have been 50. But obviously, the procedure offered participants ample opportu- nity to cheat. Hence, actual cheating behavior (averaged within this condition) would be reflected in a significant upward mean deviation from 50.

The other half of the participants did not enter this lottery. Instead, we asked them whether it is morally acceptable for people to overreport their traveling expenses (on a 9-point scale from 1, fully acceptable, to 9, fully unacceptable).

Results and discussion

We first standardized the responses to both the morality judg- ment (9-point scale) and the cheating measure (a 100-point scale). Next, we performed a 2 (power: high vs. low) × 2 (dilemma: judgment vs. behavior) analysis of variance on par- ticipants’ decision and found the expected interaction effect, F(1, 57) = 7.33, p = .009, η

p 2 = .12, and no main effects (Fs <

.01; see Fig. 1). High-power participants were stricter in their judgment than low-power participants, finding it less accept- able to overreport traveling expenses, t(57) = 1.78, p = .08. However, they claimed a higher number of lottery tickets than low-power participants did, t(57) = 2.09, p = .04.

Experiments 2, 3, and 4: Judging One’s Own and Other People’s Moral Transgressions As we predicted, compared with the powerless, the powerful were less tolerant of cheating but more likely to cheat, demon- strating heightened hypocrisy. Yet a disadvantage of our approach in Experiment 1 is that judgment and behavior could not be compared directly, because they were measured on dif- ferent scales and involved different settings (lottery tickets and traveling expenses). We therefore could not quantify hypoc- risy precisely.

Consequently, in the next experiments, we decided to use a different approach. Following Valdesolo and DeSteno (2007, 2008), we measured hypocrisy as the discrepancy between the moral acceptability and appropriateness of one’s own moral

4

5

6

7

8

High Power Low Power

Condition Condition

Ju dg

m en

t

40

50

60

70

80

High Power Low Power

B eh

av io

r

Fig. 1. Results of Experiment 1: mean judgment score (on a scale from 1, fully acceptable, to 9, fully unacceptable) and number of lottery tickets participants reported winning (the measure of cheating behavior) as a function of condition. A significant upward deviation from 50 lottery tickets represents cheating behavior. Unstandardized means are shown. Error bars represent ±1 SE.

740 Lammers et al.

transgressions and other people’s moral transgressions. Because judgments of one’s own behavior and other people’s behavior can be measured with comparable scales, we could directly calculate hypocrisy as the absolute discrepancy between the two measures. We also aimed to generalize our findings by manipulating power with three different tech- niques and by measuring hypocrisy with three different moral dilemmas. Because the three experiments had similar meth- ods, we describe them together.

Method Participants and design. One hundred seventy-two Dutch students (129 women, 43 men; mean age = 20.5 years) took part in partial fulfillment of a course requirement. In each experiment, participants were randomly assigned to one con- dition of a 2 (power: low vs. high) × 2 (target: oneself vs. other people) between-participants design. In total, 42 students par- ticipated in Experiment 2, 88 students participated in Experi- ment 3, and 42 students participated in Experiment 4.

Procedure. In Experiment 2, we manipulated power positions by simulating a bureaucratic organization and randomly dis- tributing a high-power role (prime minister) and a low-power role (civil servant) in a bureaucracy simulation (adapted from Anderson & Berdahl, 2002). Prime ministers were told they could control and direct the civil servants. In Experiment 3, we primed the experience of power using the same recall manipu- lation as in Experiment 1. In Experiment 4, we primed the concept of power, using Chen, Lee-Chai, and Bargh’s (2001) word-search puzzle.

After the manipulation of power, participants were pre- sented with a moral dilemma that was supposedly unrelated to the experiment and was presented as part of unrelated research on moral dilemmas. Specifically, the dilemma in Experiment 2 was about breaking traffic-related rules and norms: Is some- one who is late for an appointment allowed to break the speed limit to make the appointment on time, given that there is little traffic on the road? In Experiment 3, we asked participants about tax declarations: Is it acceptable to omit from one’s tax declaration additional wages that one earned in one’s spare time? In Experiment 4, the dilemma was about keeping a bike that was stolen and abandoned by the thief: If someone who needs a bike but has no money to buy one finds such an aban- doned bike, is it acceptable for him or her to take and keep the bike, rather than bring it to the police? We picked these dilem- mas because they all involve serious immoral behaviors (tech- nically, legal offenses or even felonies), but behaviors that many people occasionally engage in.

In all three experiments, participants in the others’-trans- gression condition were asked to rate how acceptable it would be for other people—specifically, for Renee, Kim, or Chris (all gender neutral names)—to engage in the described behavior.

Participants in the own-transgression condition were asked to rate how acceptable it would be if they themselves engaged in the described behavior. In Experiments 2 and 3, participants answered a single question on a 9-point scale from 1 (com- pletely unacceptable) to 9 (fully acceptable). In Experiment 4, they provided ratings on four 9-point scales (acceptable, admissible, o.k., and appropriate), and these ratings were averaged (α = .92).

We used a mood measure (“Currently I feel happy/cheerful/ satisfied/sad/unhappy/dejected”; αs > .85) and a power-manipu- lation check (“I feel influential/independent/leading/dependent/ unimportant/subordinate”; αs > .88) that have been used exten- sively in previous research (e.g., Lammers & Stapel, 2009).1 There were no mood effects (Fs < 1, ps > .5). The manipulation checks showed that participants felt more powerful in the high- than in the low-power conditions (ps < .001, η

p 2s > .55). In Exper-

iment 2, we also administered four items (α = .88) to check that low-power participants and high-power participants identified equally with their respective roles and found that they did (no dif- ference, p = .42). In Experiment 4, we asked participants what they thought the aim of the research was and determined that none were aware that the puzzle manipulated feelings of power.

Results In all three experiments, we found the same pattern of interac- tion between power and target. The interaction was marginally significant in Experiment 2, F(1, 38) = 3.53, p = .068, η

p 2 =

.09, and significant in both Experiment 3, F(1, 68) = 9.05, p =

.004, η p 2 = .12, and Experiment 4, F(1, 70) = 8.16, p = .007,

η p 2 = .18. None of the main effects in any experiment was sig-

nificant (Fs < .3, n.s.). Table 1 shows the mean ratings in all conditions, along with

the results of t tests on the level of hypocrisy (the discrepancy between ratings in the own-transgression and the others’- transgression conditions) within the high-power and low- power conditions. High-power participants showed significant moral hypocrisy, regardless of whether the behavior in ques- tion was speeding (p = .02), tax dodging (p = .03), or keeping a stolen bike (p = .06). In none of the low-power conditions did we find signs of hypocrisy. In fact, in two experiments, we found an unexpected significant effect in the opposite direc- tion (p = .03 and p = .04), indicating that low-power partici- pants were more lenient in their moral judgment of others’ transgressions than of their own transgressions. We return to this finding in the General Discussion.

An additional analysis, using the combined data of Experi- ments 2 through 4 (which employed the same design and the same 9-point measure of moral acceptability) and adding experiment as an additional between-participants factor, con- firmed that participants in the high-power conditions overall showed significant levels of moral hypocrisy, F(1, 74) = 13.14, p < .001, η

p 2 = .15. Low-power participants did not show

Power and Hypocrisy 741

moral hypocrisy; in fact, they showed the opposite effect, being more lenient in judging others’ transgressions than their own transgressions, F(1, 86) = 5.99, p = .02, η

p 2 = .08. Overall,

high-power participants were stricter than low-power partici- pants in judging other people’s behavior, F(1, 81) = 16.81, p < .001, η

p 2 = .17, but more lenient in judging their own behavior,

F(1, 85) = 6.24, p = .01, η p 2 = .07.

Experiment 5: The Moderating Effect of Illegitimacy The previous studies consistently found that high-power par- ticipants demonstrated moral hypocrisy, whereas low-power participants did not. We proposed earlier that one of the rea- sons why power increases hypocrisy is that the powerful feel entitled both to take more than other people and to prescribe how other people should behave. Accordingly, when power is disconnected from such entitlement—for example, when the powerful feel that their position is illegitimate—the power- induced hypocrisy effect should be eliminated.

Method Participants and design. One hundred five Dutch students (82 women, 23 men; mean age = 19.8 years) participated in the experiment in partial fulfillment of a course require- ment. Participants were randomly assigned to one condition of a 2 (power: high vs. low) × 2 (legitimacy: legitimate vs. illegitimate) × 2 (target: oneself vs. other people) between- participants design.

Procedure. We manipulated power and its legitimacy using an adapted experiential power prime. Half of the participants were asked to write about an experience in which they were in

a position of high power, and half were asked to write about an experience in which they were in a position of low power. Within each of these conditions, half of the participants were instructed to write about an experience in which their high or low power was legitimate, meaning that they were entitled to that powerful or powerless position, and the other half were instructed to write about an experience in which their high or low power was illegitimate, meaning that they personally believed they were not entitled to that position (see Lammers et al., 2008, for details).

We next used the same bike dilemma as in Experiment 4, but with only a single item to measure moral acceptability. Finally, participants completed a measure of mood (α = .89) and a power-manipulation check (α = .97) identical to those in Experiments 2, 3, and 4. We also used a legitimacy-manipulation check (α = .97). Our manipulation checks showed that we suc- cessfully induced feelings of high and low power, F(1, 97) = 232.61, p < .001, η

p 2 = .71, that were seen as either legitimate or

not, F(1, 97) = 59.05, p < .001, η p 2 = .38.

Results We found the predicted three-way interaction of power, legiti- macy, and target, F(1, 97) = 6.42, p = .01, η

p 2 = .06 (see Table

2). Under conditions of legitimacy, we replicated the finding that the powerful show moral hypocrisy, manifested in a dis- crepancy between judgments of their own and other people’s transgressions (p = .06). However, illegitimate power did not cause moral hypocrisy. In fact, we again found the reverse (p = .01). Participants in neither low-power condition showed moral hypocrisy.

Further analyses showed that the three-way interaction was fully mediated by the experienced feelings of power and legiti- macy (measured by the manipulation checks). Adding the

Table 1. Mean Acceptability Judgments and Differences Between Conditions in Experiments 2 Through 4

Target

Experiment and power condition Own transgression Others’

transgression Difference t test Hypocrisy/hypercrisy

Experiment 2: power roles, speeding High power 7.63 (0.92) 6.33 (1.43) 1.30 t(38) = 2.40, p = .02 Hypocrisy Low power 7.23 (1.24) 7.33 (0.87) –0.10 t(38) = –0.20, n.s. Neither Experiment 3: experiential priming, tax dodging

High power 7.57 (1.08) 6.63 (1.64) 0.94 t(84) = 2.21, p = .03 Hypocrisy Low power 6.84 (1.55) 7.70 (1.02) –0.86 t(84) = –2.20, p = .03 Hypercrisy Experiment 4: concept priming, keeping a stolen bike

High power 4.43 (1.79) 2.93 (1.64) 1.50 t(38) = 1.94, p = .06 Hypocrisy Low power 3.16 (1.93) 4.70 (1.51) –1.54 t(38) = –2.10, p = .04 Hypercrisy

Note: Standard deviations are given in parentheses. Significantly higher approval of the behavior in the own-transgression condition than in the others’-transgression condition is indicative of hypocrisy, and the reverse pattern is indicative of hypercrisy.

742 Lammers et al.

interaction of measured feelings of power, measured feelings of legitimacy, and target (which was highly significant, β = –0.50, p = .04) reduced the three-way interaction to nonsig- nificance, β = –0.22, p = .41 (see Table 3). Finally, although we found that positive mood was associated with increased moral leniency (p = .02), controlling for mood did not affect the results.

General Discussion Across five experiments, irrespective of how power was manipulated or hypocrisy was measured, we found strong evi- dence that the powerful are more likely to engage in moral hypocrisy than are people who lack power. In Experiment 1, we measured the discrepancy between moral judgments and actual immoral behavior and found that, compared with low- power participants, high-power participants engaged in more

immoral behavior but found such behavior less acceptable. In Experiments 2 through 5, we measured the discrepancy between the acceptability of one’s own moral transgressions and those committed by other people. The method we used in Experiment 1 had the advantage that actual behavior was mea- sured, but it did not allow us to compute an absolute degree of hypocrisy (a discrepancy). Across Experiments 2 through 5, the powerful judged their own moral transgressions as more acceptable than other people’s, but low-power participants did not.

Across all five experiments, only the powerful showed hypocrisy. We found this pattern irrespective of whether the behavior in question was mildly inappropriate (cheating to obtain extra lottery tickets) or very inappropriate (a legal offense). Our final study demonstrated the crucial role of enti- tlement: Only when power is experienced as legitimate is moral hypocrisy a likely result. If power is not experienced as legitimate, then the moral-hypocrisy effect disappears.

Hypercrisy In Experiments 3 and 4, we unexpectedly found that low- power participants were in fact stricter in judging their own transgressions than in judging those of other people. This effect could be termed hypercrisy (from Greek hyper–, mean- ing “too much,” and kritein, meaning “being critical,” i.e., of oneself). That is, participants were overly critical—rather than insufficiently (hypo-) critical—of themselves.2 Given that moral hypocrisy is typically portrayed as a fundamental and deep-seated phenomenon (Valdesolo & DeSteno, 2007, 2008), this effect is surprising. We do note, however, that the hypoc- risy effect (among high-power participants) appears to be markedly stronger than the hypercrisy effect (among low- power participants). After all, the size of the latter effect was statistically significant only in Experiments 3 and 4. Also, in the overall analysis, the hypocrisy effect size (among high- power participants) was much larger than the hypercrisy effect size (among low-power participants). This might have been caused by the fact that by default people tend toward hypoc- risy, because of self-interest and motivated reasoning (Batson

Table 2. Mean Acceptability Judgments and Differences Between Conditions in Experiment 5

Target

Power and legitimacy condition Own transgression Others’ transgression Difference t test Hypocrisy/hypercrisy

Legitimate power High power 6.90 (1.85) 5.09 (1.64) 1.81 t(97) = 1.87, p = .06 Hypocrisy Low power 4.38 (2.22) 4.69 (2.50) –0.31 t(97) = –0.35, p = .72 Neither Illegitimate power High power 3.87 (2.13) 6.00 (1.77) –2.13 t(97) = –2.64, p = .01 Hypercrisy Low power 4.29 (2.81) 5.07 (2.37) –0.79 t(97) = –0.94, p = .35 Neither

Note: Standard deviations are given in parentheses. Significantly higher approval of the behavior in the own-transgression condition than in the others’- transgression condition is indicative of hypocrisy, and the reverse pattern is indicative of hypercrisy.

Table 3. Mediation Model for Experiment 5

Predictor Step 1 (β) Step 2: mediation

(β)

Power 0.41 0.20 Legitimacy 0.29 0.05 Target (own vs. other behavior) 0.21 0.31 Power × Legitimacy 0.23 0.13 Power × Target –0.06 0.01 Legitimacy × Target –0.42 –0.41 Power × Target × Legitimacy –0.56** –0.22 Experienced power — 0.20 Experienced legitimacy — 0.31 Experienced Power × Experienced

Legitimacy — 0.16

Experienced Power × Experienced Legitimacy × Target

— –0.50*

Note: The left column shows the standardized regression coefficients for the experimental conditions. The right column shows the coefficients for the mediators (experienced power and experienced legitimacy) together with the direct effects for the experimental conditions when the mediators were included in the model. *p ≤ .05. **p ≤ .01.

Power and Hypocrisy 743

& Thompson, 2001). This baseline tendency must first be overcome before the experience of decreased power can make people less critical of others and more critical of themselves. In contrast, the effect of elevated power builds on an already existing difference between how people judge themselves and others. Hypercrisy is therefore probably a less common effect than hypocrisy. Yet when it occurs, it occurs mainly among the powerless or among the powerful who clearly lack legitimacy.

Social inequality As a rule, human societies are unequal (Leavitt, 2005; Magee & Galinsky, 2008). Even egalitarian democracies are made up of a large group of powerless have-nots and a small group of powerful haves (Mills, 1956). A question that lies at the heart of the social sciences is how this status quo is defended and how the powerless come to accept their disadvantaged posi- tion. The typical answer is that the state and its rules, regula- tions, and monopoly on violence coerce the powerless to do so (Weber, 1948). But this cannot be the whole answer: Social order rests on more than fear of the law’s reach (Arendt, 1951).

One explanation is that there exists a general acceptance of inequality of social groups, not only among the powerful, but also among the powerless, in order to produce social harmony (Sidanius, Levin, Federico, & Pratto, 2001). Working-class people embrace ideologies that seem fair but actually rein- force the status quo (Gramsci, 1992). Members of low-status groups show out-group favoritism and hold negative stereo- types about their own group’s abilities (Jost & Banaji, 1994). All these and other findings show that the protection of social inequality is not something necessarily imposed by one group and resisted by the other. Rather, the stability of the system comes from within, in the sense that even the victims of that system contribute to its acceptance (Foucault, 1979/1991; Jost, Banaji, & Nosek, 2004).

Our findings support these notions that the powerless col- laborate in reproducing social inequality. The powerful impose more normative restraints on other people, but believe that they themselves can act with less restraint. The less powerful are less inclined to impose norms on other people, but more rigidly follow these norms themselves. This means that people with power take what they want not only because they can do so without punishment, but also because they intuitively feel they are entitled to do so. Conversely, people who lack power fail to get what they need not only because they are not allowed to take it, but also because they intuitively feel they are not entitled to it. Further, we demonstrated these effects even after minimal lexical priming of which participants were unaware. This suggests that these inequality-reinforcing processes may be at least partially automatic.

Our last experiment, however, found that the spiral of inequality can be broken, if the illegitimacy of the power dis- tribution is revealed. One way to undermine the legitimacy of

authority is open revolt, but a more subtle way in which the pow- erless might curb self-enrichment by the powerful is by tainting their reputation, for example, by gossiping (Keltner, Van Kleef, Chen, & Kraus, 2008). If the powerful sense that their unre- strained self-enrichment leads to gossiping, derision, and the undermining of their reputation as conscientious leaders, then they may be inspired to bring their behavior back to their espoused standards. If they fail to do so, they may quickly lose their authority, their reputation, and—eventually—their power.

Declaration of Conflicting Interests

The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.

Notes

1. We did not measure feelings of power in Experiment 4 because the semantic power prime we used in that experiment operates outside conscious awareness (Chen et al., 2001). 2. The term hypercrisy does not generate any hits in psychological databases.

References

Anderson, C., & Berdahl, J.L. (2002). The experience of power: Exam- ining the effects of power on approach and inhibition tendencies. Journal of Personality and Social Psychology, 83, 1362–1377.

Arendt, H. (1951). The origins of totalitarianism. New York: Har- court, Brace and World.

Bargh, J.A., Raymond, P., Pryor, J.B., & Strack, F. (1995). Attractive- ness of the underling: An automatic power → sex association and its consequences for sexual harassment and aggression. Journal of Personality and Social Psychology, 68, 768–781.

Batson, C.D., Kobrynowicz, D., Dinnerstein, J.L., Kampf, H.C., & Wilson, A.D. (1997). In a very different voice: Unmasking moral hypocrisy. Journal of Personality and Social Psychology, 72, 1335–1348.

Batson, C.D., & Thompson, E.R. (2001). Why don’t moral people act morally? Motivational considerations. Current Directions in Psychological Science, 10, 54–57.

Batson, C.D., Thompson, E.R., Seuferling, G., Whitney, H., & Strongman, J. (1999). Moral hypocrisy: Appearing moral to one- self without being so. Journal of Personality and Social Psychol- ogy, 77, 525–537.

Chaurand, N., & Brauer, M. (2008). What determines social control? People’s reactions to counternormative behaviors in urban envi- ronments. Journal of Applied Social Psychology, 38, 1689–1715.

Chen, S., Lee-Chai, A.Y., & Bargh, J.A. (2001). Relationship orienta- tion as moderator of the effects of social power. Journal of Per- sonality and Social Psychology, 80, 183–187.

De Waal, F. (1983). Chimpanzee politics: Power and sex among apes. London: Jonathan Cape.

DePaulo, B.M., & Friedman, H.S. (1998). Nonverbal communica- tion. In D. Gilbert, S.T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed., Vol. 2, pp. 3–40). New York: McGraw-Hill.

744 Lammers et al.

Emerson, R.M. (1962). Power-dependence relations. American Soci- ological Review, 27, 31–41.

Fischbacher, U., & Heusi, F. (2008). Lies in disguise: An experimen- tal study on cheating. Unpublished manuscript, Thurgau Institute of Economics, Kreuzlinger, Switzerland.

Foucault, M. (1991). Discipline and punish: The birth of the prison. New York: Vintage Books. (Original work published 1979)

Foucault, M., & Gordon, C. (1980). Power/knowledge. New York: Pantheon Books.

Galinsky, A.D., Gruenfeld, D.H, & Magee, J.C. (2003). From power to action. Journal of Personality and Social Psychology, 85, 453–466.

Galinsky, A.D., Magee, J.C., Gruenfeld, D.H, Whitson, J., & Liljen- quist, K.A. (2008). Social power reduces the strength of the situ- ation: Implications for creativity, conformity, and dissonance. Journal of Personality and Social Psychology, 95, 1450–1466.

Gilbert, D.T., & Jones, E.E. (1986). Exemplification: The self-pre- sentation of moral character. Journal of Personality, 54, 593–615.

Gramsci, A. (1992). Prison notebooks (J.A. Buttigieg, Trans.). New York: Columbia University Press.

Guinote, A. (2007). Power and goal pursuit. Personality and Social Psychology Bulletin, 33, 1076–1087.

Jost, J.T., & Banaji, M.R. (1994). The role of stereotyping in system- justification and the production of false consciousness. British Journal of Social Psychology, 33, 1–27.

Jost, J.T., Banaji, M.R., & Nosek, B.A. (2004). A decade of system justi- fication theory: Accumulated evidence of conscious and unconscious bolstering of the status-quo. Political Psychology, 25, 881–919.

Kanagaretnam, K., Lobo, G.J., & Mohammad, E. (2008). Determi- nants and consequences of large CEO pay. International Journal of Accounting and Finance, 1, 61–82.

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

Keltner, D., Van Kleef, G.A., Chen, S., & Kraus, M.W. (2008). A recip- rocal influence model of social power: Emerging principles and lines of inquiry. In M.P. Zanna (Ed.), Advances in experimental social psychology (Vol. 40, pp. 151–192). London: Academic Press.

Lammers, J., Galinsky, A.D., Gordijn, E.H., & Otten, S. (2008). Ille- gitimacy moderates the effects of power on approach. Psycho- logical Science, 19, 558–564.

Lammers, J., & Stapel, D.A. (2009). How power influences moral think- ing. Journal of Personality and Social Psychology, 97, 279–289.

Leavitt, H.J. (2005). Top down: Why hierarchies are here to stay and how to manage them more effectively. Boston: Harvard Business School Press.

Magee, J.C., & Galinsky, A.D. (2008). Social hierarchy: The self- reinforcing nature of power and status. Academy of Management Annals, 2, 351–398.

Mills, C.W. (1956). The power elite. Oxford, England: Oxford Uni- versity Press.

Overbeck, J.R. (2009). Concepts and historical perspectives on power. In A. Guinote & T. Vescio (Eds.), The social psychology of power (pp. 11–66). New York: Guilford Press.

Runciman, D. (2008). Political hypocrisy: The mask of power, from Hobbes to Orwell and beyond. Princeton, NJ: Princeton University Press.

Sidanius, J., Levin, S., Federico, C., & Pratto, F. (2001). Legitimiz- ing ideologies: The social dominance approach. In J.T. Jost & B. Major (Eds.), The psychology of legitimacy: Emerging per- spectives on ideology, justice, and intergroup relations (pp. 307– 331). New York: Cambridge University Press.

Smith, P.K., & Bargh, J.A. (2008). Nonconscious effects of power on basic approach and avoidance tendencies. Social Cognition, 26, 1–24.

Stone, J., Wiegand, A.W., Cooper, J., & Aronson, E. (1997). When exemplification fails: Hypocrisy and the motive for self-integrity. Journal of Personality and Social Psychology, 72, 54–65.

Thibaut, J.W., & Kelley, H.H. (1959). The social psychology of groups. New York: Wiley & Sons.

Valdesolo, P., & DeSteno, D. (2007). Moral hypocrisy: Social groups and the flexibility of virtue. Psychological Science, 18, 689–690.

Valdesolo, P., & DeSteno, D. (2008). The duality of virtue: Decon- structing the moral hypocrite. Journal of Experimental Social Psychology, 44, 1334–1338.

Van Kleef, G.A., Oveis, C., Van der Löwe, I., LuoKogan, A., Goetz, J., & Keltner, D. (2008). Power, distress, and compassion: Turn- ing a blind eye to the suffering of others. Psychological Science, 19, 1315–1322.

Weber, M. (1948). Class, status and party. In H. Gerth & C.W. Mills (Eds.), From Max Weber: Essays in sociology (pp. 180–195). New York: Routledge.

Yzerbyt, V., Schadron, G., Leyens, J.-P., & Rocher, S. (1994). Social judgeability: The impact of meta-informational cues on the use of stereotypes. Journal of Personality and Social Psychology, 66, 48–55.