Abstract
A large sample of 14,306 men and women was used to examine the relationship between social power and sexual arousal to consensual sadomasochism. Results showed that power increases the arousal to sadomasochism, after controlling for age and dominance. Furthermore, the effect of power on arousal by sadistic thoughts is stronger among women than among men, while the effect of power on arousal by masochistic thoughts is stronger among men than women. These findings refute common beliefs, reinforced through novels such as Fifty Shades of Grey, that the desire for sadomasochism reflects a desire to play out power dynamics in the bedroom. Instead, the effect of power is driven through a process of disinhibition that leads people to disregard sexual norms in general and disregard sexual norms associated with their gender in particular. These results add to an emerging literature that social power changes traditional gender patterns in sex.
With over 100 million copies sold, the popularity of the novel Fifty Shades of Grey (James, 2011) shows people’s fascination with sadomasochism and their desire to better understand why some people want to mix sex with domination and pain. Although only a few have ever engaged in sadomasochistic sex, many people experience arousal to such thoughts (Weinberg & Kamel, 1983). Despite this, there is little research into the antecedents of sadomasochistic thoughts. Given the central role of power in sadomasochism, one particularly interesting question is how social power steers people’s attraction to sadomasochism.
One idea on this relation—and perhaps one reason why sadomasochism is often strongly criticized—is that sadomasochism is a form of playing out power differences in sex, where the powerful adopt the controlling sadistic role and exert control over the powerless who adopt a more passive role. This idea may ultimately draw on Freudian notions that sadism is an aberration of the aggressive instinct to exert power over others, while masochism stems from low-power motives to provide pleasure to others and avoid guilt (Deutsch, 1930; Freud, 1905/1953). James’ (2011) recent novel, for example, intuitively follows this playing out hypothesis in its portrayal of a powerful millionaire who takes the sadistic role and an insecure student who adopts the masochistic role. Yet despite this having reached the status of cultural truism, no research has tested the truth of this link between power and sadomasochism.
Power and Disinhibition
Based on the recent social psychological research on power, we dispute this playing out hypothesis and present an alternative, disinhibition hypothesis. This hypothesis is based on a wealth of findings, showing that power leads to a state of behavioral disinhibition, meaning that the powerful are more likely to act on their impulses (Galinsky, Gruenfeld, & Magee, 2003; Keltner, Gruenfeld, & Anderson, 2003; Lammers, Galinsky, Gordijn, & Otten, 2008; Smith & Bargh, 2008; Whitson, Liljenquist, Galinsky, Magee, Gruenfeld, & Cadena, 2013). Such a state of disinhibition can help people ignore or overcome situational pressures and can make them more likely to act in line with their goals and desires (Galinsky et al., 2003; Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008; Guinote, 2007a; Hirsh, Galinsky, & Zhong, 2011; Slabu & Guinote, 2010).
This power-induced disinhibition can cause positive, virtuous effects. For example, if people with a strong moral identity feel powerful, then they are more likely to overcome inertia and act on their inclination to help others (DeCelles, DeRue, Margolis, & Ceranic, 2012). And if people are strongly committed to a personal relationship, power can increase their tendency to forgive (Karremans & Smith, 2010). But due to its general disinhibiting effect, feelings of power can also help people to disregard and violate situational norms and thus cause negative, corruptive effects. For example, power has been shown to cause people to steal, cheat, or make traffic violations (Lammers, Stapel, & Galinsky, 2010; Trautmann, Van de Kuilen, & Zeckhauser, 2013; Yap, Wazlawek, Lucas, Cuddy, & Carney, 2013; for an overview, see Lammers, Galinsky, Dubois, & Rucker, 2015). This link between power and disinhibited counternormative behavior may be even further catalyzed by the fact that power increases feelings of social distance, making the powerful even more likely to ignore and disregard social norms (Lammers, Galinsky, Gordijn, & Otten, 2012; Magee & Smith, 2013).
Sadomasochism has clear counternormative elements, as it mixes sex and love with violence and domination. Much of the literature treats people who engage in sadomasochism as in need of therapy or even as a potential danger to society (e.g., Langevin, 2003). Also, people are less willing to admit having any sadomasochistic fantasies, unless they are in a disinhibited and aroused state, in which case they more easily admit being excited by the thought of spanking or being spanked in a sexual context (Ariely & Loewenstein, 2006; Imhoff & Schmidt, 2014). Given the strong link between power and disinhibition, which leads people to engage in counternormative behavior, we predict that social power also increases arousal to sadomasochism.
Importantly, we do not aim to suggest that everyone who experience power will experience arousal to sadomasochism. For many—if not most—people, the idea of mixing sex with domination and pain is simply unappealing. For them, power will not have any effect on sadomasochism. But for those people who for some reason feel attracted to sadomasochism, experiencing a sense of power adds strength to their dormant impulses. Therefore, we expect that overall power will increase arousal to sadomasochism.
Crucially, however, we note that the playing out hypothesis and the disinhibition hypothesis make opposite predictions on the nature of that relationship. Specifically, the playing out hypothesis expects power to increase arousal to sadism but to reduce arousal to masochism. After all, sadism is the high-power, controlling role and masochism the low-power, submissive role in sadomasochism. If sadomasochism is driven by a tendency to play out real-life power differences in the bedroom, then the powerful should feel less aroused at exercising the low-power masochistic role, because it is inconsistent with their real-life position of power. In contrast, the disinhibition hypothesis predicts power to increase arousal to both. After all, both sadism and masochism constitute a form of sexual disinhibition (Ariely & Loewenstein, 2006; Imhoff & Schmidt, 2014). If power has general disinhibiting effects, it should increase both.
Gender and Double Disinhibition
Second, although sadomasochism is seen as counternormative overall, we propose that sadism and masochism may be seen as distinctly counternormative for men and women. Specifically, gender norms associate men with dominant and active and women with submissive and passive behaviors (Eagly, 1987; Koenig & Eagly, 2014). Transgression of such gender norms typically leads to negative reactions. For example, women may experience backlash if they show agentic, leadership behavior (Rudman & Glick, 2001), while men may experience backlash for being modest (Moss-Racusin, Phelan, & Rudman, 2010).
These societal gender norms also dictate behavior within sadomasochism. This is demonstrated by a strong descriptive norm that men are more likely to report experiencing arousal to sadistic thoughts, while women are more likely to report experiencing arousal to masochistic thoughts (Hsu, Kling, Kessler, Knapke, Diefenbach, & Elias, 1994; Person, Terestman, Myers, Goldberg, & Salvadori, 1989). Therefore, although sadomasochism may be counternormative overall, female sadism and male masochism may be even more counternormative than opposite combinations. Not only do they break the norm against mixing sex and violence, but they also violate gender norms against female dominance and male submission.
Again, this prediction differs from what would be predicted based on the idea that sadomasochism is a form of playing out power differences. One reason why sadomasochism is often strongly criticized from a feminist perspective is that sadomasochism is a form of playing out genderized power dynamics in sex, by giving men the dominant, sadistic role and women the submissive, masochistic role (Dworkin, 2006; Linden, Pagano, Russell, & Star, 1982). James’ (2011) recent novel also follows this gender pattern. Yet if the alternative disinhibition hypothesis is correct, then individual differences in power should attenuate gender differences, in particular by drawing powerful men more strongly to masochism compared to powerful women’s draw to masochism.
Power and Dispositions
Third and finally, the disinhibition hypothesis predicts that chronic dispositional differences may interact with power and further strengthen its counternormative effect (Chen, Lee-Chai, & Bargh, 2001; Galinsky et al., 2008; Guinote, 2007b; Guinote, Weick, & Cai, 2012; Hirsh et al., 2011). Specifically, given the strong link between sadomasochism and dominance (Ernulf & Innala, 1995; Moser, 1988), dispositional differences in dominance motivation may affect these results, in the sense that the effects of power (and possibly the power–gender interaction) on sadomasochism may be particularly strong for those high in dispositional dominance motivation. The playing out hypothesis does not make any predictions on the effect of disposition.
Method
To test these hypotheses against each other, we used a large-scale Internet-based sample composed of people of all ages, in which we measure dominance motivation, power, and arousal by sadistic and masochistic thoughts. We use a correlational design because the topic of our research precludes an experimental or observational design. We report how we determined sample size, all data exclusions, and all measures (see also the Supplementary online materials [SOM] at http://spps.sagepub.com/supplemental).
Participants and Design
Readers of Quest and Glamour, a popular science and a general lifestyle magazine in the Netherlands, were asked to voluntarily complete a short online questionnaire, by means of an ad in a newsletter. The study was conducted at the website of Quest, on a site that regularly features small tests for voluntary participation (at http://www.quest.nl/test). As sample size, we a priori used all respondents who completed the items within a 3-week time span, excluding those participants with incomplete responses. In total, 19,408 respondents started the survey and 14,508 respondents (74.8%) finished all items. In addition, we dropped the 202 respondents who indicated an age lower than 16. 1 We chose 16 as the cutoff point, as this is the age of consent in the Netherlands (Wetboek van Strafrecht, 2014). This left 14,306 participants. The sample contained 9,016 (63.0%) women and 5,290 (37.0%) men. Participants ranged in age from 16 to 105 years. The mean age was 33.2 years (SD = 13.2 years) and the median age was 28.5 years (25th percentile: 22.9 years and 75th percentile: 42.1 years). Of all respondents, 17.2% had obtained a master’s, 37.0% a bachelor’s, 26.1% polytechnic, 18.8% secondary, and 0.9% primary education. A comparison with the general population of the Netherlands (Centraal Bureau voor de Statistiek, 2015) showed that our sample contained more young people, in particular people in their 20s, and less middle aged and older people, and that our sample was more highly educated than the general population. We did not measure ethnicity or socioeconomic status.
Measures
Power and dominance
Following earlier work, we looked at power as respondents’ position in the formal hierarchy of their primary job (Lammers, Stapel, et al., 2010; Lammers, Stoker, & Stapel, 2010b; Lammers, Stoker, Jordan, Pollmann, & Stapel, 2011). Participants indicated whether they were unemployed and/or studying (n = 4,696, 32.8%, coded as 1), employed in a nonmanagement function (n = 3,502, 24.5%, coded as 2), in lower level management (n = 1,574, 11.0%, coded as 3), in middle-level management (n = 3,156, 22.1%, coded as 4), or in top-level management (n = 1,378, 9.6%, coded as 5). Although this formal hierarchy scale is strictly defined an ordinal measure, earlier work has shown that it can be safely analyzed as a continuous scale, given that it strongly correlates with the subjective feelings of power and control over others (Lammers, Stoker, et al., 2010). The unemployed and students were categorized as the group with the least power, given that objectively they have little to no power, defined as either control over resources or the ability to give orders to other people, and given that they instead depend strongly on others. Furthermore, both students and unemployed have been shown to experience many of the psychological characteristics of powerlessness, such as anxiety, lower self-esteem, and lower sense of control (Andrews & Wilding, 2004; Bayram & Bilgel, 2008; Dekker & Schaufeli, 1995; Donovan & Oddy, 1982). To measure the desire to be in a position of dominance, participants completed the Dominance subscale of the Achievement Motivation Scale, on 5-point scales (Cassidy & Lynn, 1989; α = .82).
Sexual interest
We administered the sadism and masochism fantasy subscales of the Violent Sexual Interest Questionnaire (VSIQ; Larue et al., 2014). Five items measured sexual arousal to consensual sadism (example item, “It sexually arouses me to fantasize about torturing a consenting person”; α = .87) and 5 to consensual masochism (example, “It sexually arouses me to fantasize about being tortured by a person on my own demand”; α = .87). All items were completed on 5-point scales. We left out the nonconsensual sadism VSIQ items (sexual arousal about inflicting pain against a person’s will). We also measured sexual arousal to general eroticism (5 items; α = .78) and two other measures of disinhibition. These were also completed on 5-point scales. Given that they are not central to our predictions, we discuss these in the SOM.
Results
Data Analytic Strategy
We first standardized all independent variables. Next, we separately analyzed arousal by sadistic and masochistic thoughts using linear regression analysis. In each case, we first analyzed for the whole sample linear effects of power, dominance, gender, and their interactions, controlling for age. 2 Next, we interpreted interactions with gender by running separate linear regression analyses for both genders. In text, we only discuss effects that bear on our hypotheses (i.e., the main effects of gender and power and any interaction with power). All other effects are listed in Tables 1 and 2.
Regression Model on Arousal to Sadistic Thoughts.a
aValues indicate Bs and asterisks indicate associated ps.
Regression Model on Arousal to Masochistic Thoughts.a
aValues indicate Bs and asterisks indicate associated ps.
Sadism
Overall, men were more strongly aroused by sadistic thoughts (M = 2.38, SD = 1.06) than women were (M = 1.93, SD = 0.80), t(14,304) = 28.78, p < .0001, d = .48, 95% confidence interval [CI] = [0.42, 0.48]. A linear regression analysis of the effect of power, dominance, gender, and their interactions, controlling for age, on arousal to sadistic thoughts (Table 1, first column) showed that social power was positively related to arousal by sadistic thoughts, B = .026, SE = 0.010, 95% CI = [0.006, 0.045], t(14,297) = 2.56, p = .011. Crucially, and consistent with the disinhibition hypothesis, this effect was moderated by gender, B = −.039, SE = 0.016, 95% CI = [−0.071, −0.007], t(14,297) = −2.42, p = .016. We also found a three-way interaction between power, dominance motivation, and gender, B = .070, SE = 0.016, 95% CI = [0.039, 0.102], t(14,297) = 4.34, p < .0001.
To decompose these gender interaction effects on sadistic fantasies, we ran separate linear regression analyses for women (Table 1, second column) and men (third column). Power increased arousal to sadistic thoughts among women, B = .032, SE = 0.009, 95% CI = [0.014, 0.050], t(9,011) = 3.49, p = .0005 but not among men, B = −.024, SE = 0.016, 95% CI = [−0.055, 0.008], t(5,285) = −1.45, p = .146. This is consistent with the disinhibition hypothesis, given that male-associated sadism is more counternormative for women than for men.
Among men, we found an interaction between power and dominance motivation, B = .067, SE = 0.015, 95% CI = [0.039, 0.096], t(5,285) = 4.60, p < .0001, showing that power increased arousal by sadism among men high in dominance motivation (+1 SD), B cond = .039, SE = 0.014, t(5,285) = 2.82, p = .005, but decreased arousal among men low in dominance motivation (−1 SD), B cond = −.057, SE = 0.017, t(5,285) = −3.42 p = .0006. This finding, that power increases the effects of disposition on sexual arousal, also fits the disinhibition hypothesis.
Masochism
Overall, women were more strongly aroused by masochistic thoughts (M = 2.22, SD = 1.00) than men were (M = 2.01, SD = 1.01), t(14,304) = 12.16, p < .0001, d = .21, 95% CI = [0.17, 0.24]. A linear regression analysis of the effect of power, dominance, gender, and their interactions, controlling for age, on arousal to sadistic thoughts (Table 2, first column), showed a positive effect of power on arousal by masochistic fantasies, B = .025, SE = 0.011, 95% CI = [0.003, 0.047], t(14,297) = 2.22, p = .027. In other words, consistent with the disinhibition hypothesis power increases both forms of sadomasochism—not just sadism but also masochism. Furthermore, this effect was moderated by gender, B = .049, SE = 0.018, 95% CI = [0.013, 0.085], t(14,297) = 2.67, p = .008.
Decomposing these gender interaction effects by running separate linear regression analyses for women and men (Table 2, second and third column, respectively), we found among women a small positive effect of power on arousal by masochistic thoughts, B = .033, SE = 0.011, 95% CI = [0.011, 0.056], t(9,011) = 2.90, p = .004. Importantly, however, that same effect of power on arousal by masochistic thoughts was 81.8% stronger among men, B = .061, SE = 0.016, 95% CI = [0.030, 0.091], t(5,285) = 3.85, p = .0001. In other words, consistent with the disinhibition hypothesis, power particularly strongly increases arousal to female-associated masochism among men, while this effect is weaker among women.
Discussion
In literature, sadomasochism is portrayed as a form of playing out power dynamics. But how veridical is this cultural depiction? Our results show it is at least partially wrong. Social power disinhibits and leads people to ignore social norms. Therefore, it increases arousal to any form of sadomasochistic thought, including masochism. Furthermore, power increases arousal to sadism particularly strongly among women and increases arousal to masochism particularly strongly among men—that is, power particularly strongly increases arousal to the form of sadomasochism that is normally reserved for the opposite gender. Both findings strongly fit with the idea that the effects of power on sadomasochism are due to general processes of disinhibition associated with power.
Our results contribute to an emerging literature showing that social power can lead people to ignore social norms (for an overview, see Lammers et al., 2015). Our results suggest that this is the case for even the most intimate aspects of people’s cognition, such as their arousal to specific sexual thoughts. Social power, even if measured in a completely different domain as people’s relative position in their hierarchy at work, is associated with increased arousal to sexual cues, including those associated with sadomasochism. In doing so, power has the potential to attenuate strong genderized sexual norms and well-entrenched gender differences (see also Kunstman & Maner, 2011; Lammers et al., 2011).
Our findings are striking, given that sadomasochism is sometimes criticized from a radical feminist perspective as a continuation of patriarchal power relations. Sadomasochism that places men in the sadistic and women in the masochistic role is seen as an expression of hatred against women (Dworkin, 2006; Linden et al., 1982). On the one hand, given that in our sample men were on average more aroused by sadism and less aroused by masochism than women, our results partially fit with those ideas. Yet on the other hand, our finding that among men power increases arousal mainly to masochism and not to sadism, and our finding that power also increases arousal to sadomasochism among women suggest that the relation between power and sadomasochism is not solely one of oppression but also one of liberation—freeing people from genderized sexual norms. All in all, our results paint a more nuanced picture of the relation between power, dominance, and sadomasochism.
Furthermore, our results are novel by showing that they also liberate men from genderized sexual norms. Existing research has focused only on flirting and infidelity—forms of sexual behavior where gender norms are stricter for women than for men (Kunstman & Maner, 2011; Lammers et al., 2011; Lammers & Maner, 2015). In that research, power can liberate women from sexual norms. The current research is the first to show that power can also liberate men from restraining sexual norms, by making them particularly more likely to show masochism—a behavior that is stereotypically more strongly associated with women.
To be fair, we do note that the effects of power in our sample were small, compared to the effects of gender, age, or dominance motivation. Overall, we found that men were more aroused by sadism and women more by masochism. This is consistent with some findings in literature (Hsu et al., 1994; Larue et al., 2014; Person et al., 1989) and inconsistent with other findings (Dawson, Bannerman, & Lalumiere, 2014; Donnelly & Fraser, 1998). The large sample size of our study, compared to those of studies in existing literature, offers relatively robust support for the idea that men are more attracted to sadism and women more to masochism. This may reflect chronic power differences between men and women. Nonetheless, by showing that differences in power do not further strengthen but rather attenuate this gender difference, these results paint a more subtle and nuanced picture of the effect of power in sadomasochism.
One interesting observation was that we found that among women, power increased arousal to both sadism and masochism to an equal extent, whereas for men, power only increased arousal for masochism but not sadism. Furthermore, among men, differences in dominance motivation had a stronger positive effect on arousal to sadism and a stronger negative effect on arousal to masochism than among women. In other words, arousal to sadomasochism may depend more strongly on power and less strongly on personality for women than for men. This observation fits Baumeister’s (2000) erotic plasticity hypothesis, which holds that the female sex drive responds more strongly to sociocultural variables, while the male sex drive responds more strongly to dispositional variables.
In recent discussions around the depathologization of consensual sadomasochism, it has been stressed that many practicing individuals are not hampered by the practice in their career or pursuit of other life goals (Richters, De Visser, Rissel, Grulich, & Smith, 2008). Our findings fit with that idea. Although our research does not directly test any positive or negative effect of sadomasochism on people’s performance and well-being, either at work or elsewhere, our finding that people are most likely to engage in consensual sadomasochism at the highest levels of management suggests that sadomasochism does not necessarily hinder a successful career.
Limitations
One limitation of our findings is the correlational nature of the data, which prevents strong claims about causality. We do note that existing experimental evidence on power and disinhibition in sexual behavior (although not related to sadomasochism) points in similar directions (e.g., Kunstman & Maner, 2011). Another limitation of our findings is that our sample, although very large, contained more women, was younger, and was better educated than the general population of the Netherlands. Future research may want to extend our findings using a sample that is representative on these characteristics.
A further limitation to our findings is the fact that we relied exclusively on self-reports. Given the link between power and disinhibition on which our conceptual model is based, an alternative explanation for our results could therefore be that power merely leads to a more disinhibited and therefore freer and less socially desirable expression of feelings (e.g., Galinsky et al., 2008). On the other hand, the use of an Internet-based method of assessment reduces social desirability concerns overall, which lessens this concern at least partially (Whisman & Snyder, 2007). Furthermore, we note that the large sample and the use of well-validated measures speak to the strength of our findings. Finally, we note that, given the topic of our research obtaining better data by using experimental designs or observational data is simply not possible. As Maslow (1942) already wrote, in circumstances such as these, where it is not possible to obtain better data, one knows that the data are shaky, and the reader is warned of this, using only self-report data is nonetheless defensible.
Conclusion
Power, measured as one’s position in the hierarchy at work, affects people’s arousal to sadomasochistic thoughts. After controlling for gender, age, and dominance motivation, power is associated with increased arousal to both sadistic and masochistic thoughts. Yet the strength of these relations differs between men and women. By increasing disinhibition, power leads people to ignore genderized sex norms, making powerful women more aroused by male-associated sadism and powerful men more aroused by female-associated masochism. These results add to an emerging literature that social power is associated with an increased disregard for sexual norms and thus changes traditional gender patterns in sex.
Footnotes
Author Contributions
Joris Lammers and Roland Imhoff developed the study concept and design. Data collection was performed by J. Lammers. J. Lammers and R. Imhoff performed the data analysis and interpretation. J. Lammers drafted the manuscript and R. Imhoff provided critical revisions. All authors approved the final version of the manuscript for submission.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.