Sex Differences in Sport Remain When Accounting for Countries’ Gender Inequality

Abstract

The spectator lek hypothesis argues that sex differences in preferences for sport largely stem from evolved predispositions and thus should be universal or near universal, whereas socioconstructivist hypotheses argue that such sex differences are entirely socially constructed and thus should vary as a function of a society’s gender inequality. To test these competing hypotheses, cross-national nested data were acquired from the International Social Survey Program (s_s = 49,729, n_countries = 34). Hierarchical linear modeling was used to examine whether sex differences in sport are universal or near universal when controlling for countries’ gender inequality. Findings indicate that even when controlling for gender inequality, sex differences remain in reporting sport as one’s most common activity, in watching sport, and in attending sport events and for agreeing with the statement that one plays sport to compete against others. Although this study was limited by the homogeneity of the sampled countries and the use of self-report measures, these findings nonetheless support the spectator lek hypothesis. Future research should examine case studies (e.g., matrilineal societies) that can specifically test the assumptions of the spectator lek hypothesis.

Keywords

sex differences gender inequality index evolutionary psychology Bernoulli modeling sport psychology male spectator lek hypothesis

Sex differences in preferences for sport are well documented throughout Western societies (e.g., Apostolou, 2015; Deaner, Balish, & Lombardo, 2016; Deaner & Smith, 2013). What is less clear, however, is why these sex differences exist and why they are so pervasive. Answers to such questions have a variety of societal implications, ranging from the development of national-level athletics policies and laws (e.g., Title IX; see Brake, 2010) to grassroots sport programming.

Hypotheses focused on sex differences in sport have been guided by social constructivist hypotheses focused on socialization effects (e.g., parenting practices shaping children’s preferences; Fredricks & Eccles, 2005). Adaptationist theory, as an alternative to social constructivism, maintains that both evolved predispositions and sociocultural factors contribute to sex differences (Apostolou, 2015; Balish, Eys, & Schulte-Hostedde, 2013; Deaner, Balish, & Lombardo, 2016; Deaner & Smith, 2013; Lombardo, 2012). The key difference is that adaptationist hypotheses move beyond the centrality of socialization by incorporating three meta-theoretical assumptions (Buss, 1995; Geary, 2010). The first assumption presumes that the mind, as an evolved system, can be productively reverse engineered by generating ultimate-level (i.e., functional) theories that center on ancestrally recurrent selection pressures differentially encountered by males and females. In other words, the mind can be understood by testing for specific selection pressures that could have shaped sex differences in evolved predispositions. The second assumption contends that differential selection pressures—within the context of other evolutionary constraints—should have crafted gender-specific sociomotivational mechanisms that are finely tuned to address the hypothesized sex-specific adaptive challenges. The third assumption supposes that to investigate possible sex differences in sport in modern societies, it is necessary to understand how the modern sporting environment activates male and female sociomotivational mechanisms that evolved to motivate adaptive behavior in the ancestral past (Apostolou, 2015; Balish et al., 2013; Deaner et al., 2016; Lombardo, 2012).

The specific adaptationist hypothesis that we test here, the spectator lek hypothesis (Lombardo, 2012), posits that sport acts as context in which males aggregate and compete with one another. Unlike mating leks in some avian species (Lombardo, 2012), males in sport leks are hypothesized to compete for status and allies among same-sex peers. One prediction of the spectator lek hypothesis is that due to differences in obligatory parental investment, men should participate more than females do. It follows that sex differences in preferences for sport largely stem from evolved mechanisms and thus should be universal, although the magnitude of the sex difference may vary across cultures and contexts (Apostolou, 2015; Deaner et al., 2016; Deaner & Smith, 2013; see also Schmitt, 2014). By contrast, social constructivist theories hypothesize that sex differences in sport should vary strictly as a function of the equality of opportunities and socialization practices (Chalabaev, Sarrazin, Fontayne, Boiché, & Clément-Guillotin, 2013; Fredricks & Eccles, 2005; see also Wood & Eagly, 2012).

Cross-cultural data provide opportunities for testing theories of sex differences in sport, given that cross-cultural evidence directly bears on these competing predictions. For example, are sex differences in sport universal? Or does difference exist largely as a function of sociocultural factors? Two noteworthy studies have examined cross-societal variation of sex differences in sport participation. Deaner and Smith’s (2013) systematic examination of the Human Relations Area Files (HRAF) probability sample found that male overrepresentation in sport is universal (e.g., Brown, 1991; Norenzayan & Heine, 2005), supporting the adaptationist claim of a male predisposition for physical competition. Deaner and Smith (2013) found in all 50 societies with relevant data that male sport (i.e., sport played only by boys and/or men but not girls and/or women) occurred more often than female sport, supporting the idea that male overrepresentation in sport is universal (e.g., Brown, 1991; Norenzayan & Heine, 2005) and the adaptationist claim of a male predisposition for physical competition. However, this analysis also revealed that the sex difference in sport participation was significantly greater in patriarchal than nonpatriarchal societies, suggesting that both evolved predispositions and sociocultural factors (e.g., patriarchy) have meaningful influences on sex differences in sport participation.

In a second study, Apostolou (2015) abstracted data from the 2007 International Social Survey Program (ISSP), and included interviews from residents in 37 countries about their sport and leisure-time activities. Each ISSP member country interviewed residents regarding their sport and leisure time activities. Apostolou (2015) found that males reported higher levels of overall sport participation, team sport participation, and watching sport. Furthermore, it was found that males and females reported different reasons for sport participation: Males were more likely to agree that they participate in sport “to compete,” whereas females were more likely to agree that they participate in sport to “look good.”

These studies provide critical, preliminary insight into cross-societal variation in sex differences in sport participation. However, neither study explicitly addressed the potential explanatory power of gender inequality, which, according to socioconstructivist theories, is crucial (see, for example, Wood & Eagly, 2012). For example, while demonstration of sex differences in sport across a number of different cultures could be interpreted as powerful evidence for adaptationist predictions, this interpretation rests on the assumption that countries included in the analyses vary substantially in their sociocultural composition. In reality, most cross-national studies lack substantial cultural diversity (for a discussion, see Pollet, Tybur, Frankenhuis, & Rickard, 2014). Accordingly, it is plausible that the presence of sex differences across a relatively small set of countries simply stems from shared sociocultural influences within each country (Mace et al., 1994). For example, there may be many countries whose cultural values and institutions suppress women’s motivation and opportunities for leisure time, including sport participation. To account for such possibilities, cross-national studies can statistically account for, rather than assume, the role of countries’ gender inequality. That is, a more direct test of the universality of sex differences in sport participation would account for cultural factors known to affect such relationships, rather than assume their absence or irrelevance. For example, Lowen, Deaner, and Schmitt (2016) found that in the summer Olympics, countries with lower levels of gender inequality win more medals and send more participants, and these patterns hold for both men and women.

The aim of this study was to assess whether sex differences in sport and sport fandom remain when controlling for countries’ gender inequality. We hypothesize that sex differences in self-reported behaviors (i.e., sport participation and fandom) and motives (e.g., I play sport to compete) for sport will remain even when controlling for gender inequality. This prediction follows from the male spectator lek hypothesis (Lombardo, 2012), which predicts persistent sex differences because participation in sport should primarily—but not entirely—reflect evolved sex differences in sport motivation. To test this hypothesis, we analyzed the 2007 Leisure Time and Sport survey ISSP module, which includes data on sport behaviors from 34 countries and thus can be linked to other country-level data, such as the gender inequality index (GII) developed by the United Nations.

Method

Sample

Individual-level data were acquired from the ISSP, an ongoing multinational cooperation conducting surveys covering topics relevant to social sciences research. Between 2006 and 2008, the ISSP conducted a Leisure Time and Sport survey (Scholz & Heller, 2009) that gathered nationally representative data from 36 countries. However, due to the lack of data for separated countries, we collapsed East and West Germany into one Germany category, and Jewish and Arab Israel into one Israel category. Thus, this study examined 34 countries (Argentina, Australia, Austria, Belgium, Bulgaria, Chile, Croatia, Cyprus, Czech Republic, Dominican Republic, Finland, France, Germany, Great Britain, Hungary, Ireland, Israel, Japan, Latvia, Mexico, New Zealand, Norway, Philippines, Poland, Russia, Slovak Republic, Slovenia, South Africa, South Korea, Sweden, Switzerland, Taiwan, the United States, Uruguay). This survey accumulated 49,729 individual responses, with sample sizes across countries ranging from 906 to 2,907. Females were slightly overrepresented (55.1% self-reported as female) while ages ranged from 15 to 98 years, with more than half (56.7%) being between the ages of 20 and 50. The ISSP Leisure Time and Sport survey is publically available for research purposes (see Scholz & Heller, 2009).

Dependent Variables

This study examines seven dependent variables: (a) sport participation, (b) team sport participation, (c) watching sport, (d) watching team sport, (e) attending sporting events, (f) agreement that one plays sport to compete, and (g) agreement that one plays sport to look good. For sport participation and team sport participation, both variables were derived from the question “What is your most frequent activity?” (see Table 1 for coding structure). For the purposes of deciding which activities constitute sport, we defined sport participation as any physically competitive activity that commonly involves mutually agreed upon rules between two or more sides. Team sport is defined similarly but is limited to intergroup competition that involves three or more people per side (note that this definition excludes racket sport that involve two per side). The variables watching sport and watching team sport were constructed according to the same coding structure. The variable attending sport events was derived from the question “How often do you do each of the following activities in your free time? Attend sporting events as a spectator?” Response options included (a) daily, (b) several times a week, (c) several times a month, (d) several times a year of less, and (e) never. Those who responded that they attend several times per month or more were categorized as spectators, and those who attend sporting events less than several times a month were categorized as nonspectators. The variables concerning participant’s motives for sport were derived from the question “Please indicate how important the following reasons are for you to take part in sport or games: (a) to compete against others and (b) to look good.” Response options included (a) very important, (b) somewhat important, (c) not very important, and (d) not important. Those who responded that the reason is very important or somewhat important were categorized as in agreement, and those who reported that the reason is not very important or not important were categorized as not in agreement.

Table 1.

Coding for Dependent Variables Representing Types of Most Frequent Activities.

Type of activity	Sport	Team sport	Male distribution (%)	Female distribution (%)	Percentage male (%)
Team sport	1	1	0.1	0	54
American football	1	1	0.1	0	83
Baseball/softball	1	1	1	0.3	79
Basketball	1	1	5	0.9	84
Cricket	1	1	0.5	0.1	87
Ice hockey	1	1	0.3	0.1	84
Field hockey	1	1	0.1	0.1	46
Football, soccer	1	1	15.6	1.1	93
Handball	1	1	0.4	0.2	63
Netball	1	1	0.01	0.8	3
Polo, water polo	1	1	0.1	0.01	100
Rugby	1	1	0.7	0.1	84
Volleyball	1	1	1.3	2.2	36
Other team sport	1	1	1.2	0.3	82
Racket sport	1	0	0.01	0	10
Badminton	1	0	0.8	0.8	48
Squash	1	0	0.4	0.2	67
Table tennis	1	0	0.8	0.2	75
Tennis	1	0	1.9	1.3	57
Other racket or bat sport	1	0	0.1	0.1	61
Athletics and fitness sport	0	0	0.1	0	67
Athletics (e.g., javelin, sprinting)	1	0	1.2	0.8	59
Body training (e.g., body building)	0	0	1.7	1.2	57
Fitness (e.g., aerobics, exercise)	0	0	7	16	29
Jogging	0	0	6.8	6.2	51
Walking (e.g., Nordic walking)	0	0	25.3	45	35
Other fitness sport	0	0	0.7	1.9	25
Other sport	0	0	0	0	40
Adrenaline sport	0	0	0.1	0	71
Billiards/pool/snooker	0	0	0.6	0.1	87
Biathlon/triathlon	1	0	0.1	0	73
Bowling/curling/bocce	1	0	1.1	0.6	63
Boat sport (e.g., sailing, yachting)	0	0	0.5	0.1	76
Bullfight	0	0	0.01	0	100
Cockfighting	0	0	0.01	0	50
Cycling/mountain biking	0	0	8.7	5.8	59
Dancing (e.g., ballroom, ballet)	0	0	0.4	2.2	16
Darts	0	0	0.1	0.1	63
Fencing	1	0	0.01	0	33
Fishing/hunting	0	0	2.5	0.1	94
Golf/mini golf	0	0	2.7	0.9	74
Horse riding/horse racing	0	0	0.3	0.7	29
Ice skating	0	0	0.1	0.1	37
Skateboarding/roller-skating	0	0	0.2	0.3	38
Martial arts (e.g., boxing, judo)	1	0	1.4	0.5	72
Motor sport (e.g., motor racing)	0	0	0.6	0.1	83
Rodeo	0	0	0.01	0	10
Shooting (e.g., pistols, archery)	0	0	0.4	0.1	86
Swimming/diving/snorkeling	0	0	3.2	5.1	37
Surfing/waterskiing	0	0	0.3	0	98
Snow sport (e.g., skiing)	0	0	1.4	0.7	65
Other sport	0	0	2.3	2.5	46

Control and Independent Variables

To estimate countries’ gender inequality, we used the United Nations’ GII (measured in 2008, United Nations Development Programme, 2011). The index represents a measure of women’s advancement in a country, and replaces gender development index and gender empowerment measures as it more directly measures sex differences in social achievement and empowerment within a country. The GII is based on three measures: (a) reproductive health, (b) parliamentary representation and higher education attainment, and (c) the labor force participation rate. The GII is a continuous measure, with scores ranging from 0 to 100, with increasing scores representing more gender inequality (Permanyer, 2013).

Individual-level independent variables included age and education. For age, participants simply reported their age in years (15 to 98 years of age), which we then categorized into four groups (15-25, 26-35, 36-50, >50). For level of education, participants were asked their highest level of education. Those who reported receiving more than secondary-level education (e.g., high school) were categorized as highly educated (1), and those with secondary-level education or less were categorized as less educated (0).

Multilevel Modeling

Multilevel modeling approaches are powerful inferential techniques for examining sex differences across contexts because they correctly specify multiple levels at which meaningful variation can occur and simultaneously control for confounding variables at multiple scales of analysis (Snijders & Bosker, 1999; see also Pollet et al., 2014). Accounting for the hierarchical data more accurately models the natural nestedness of individuals within groups (e.g., families) and of groups within larger geopolitical entities (e.g., countries). Multilevel models can incorporate variance structures at different levels of analysis and reduce the burden, and potential bias, associated with the creation of multiple models for each level in the hierarchy. In this study, hierarchical linear modeling (HLM) was used as it accounts for the country-level clustering (HLM 6; Raudenbush, Bryk, & Congdon, 2004). Specifically, Bernoulli modeling is well suited for studying sex differences as it allows for the estimation of the relative probability of event occurrence (e.g., participate or do not participate in sport) by calculating multivariate odds ratios. Multivariate odds ratio compares the likelihood of event occurrence among different levels of a sociodemographic category, all the while accounting for other individual-level and/or country-level variables. In other words, Bernoulli modeling offers a way to examine the relative odds of males being more likely to report sport behavior than females while accounting for other factors such as countries’ gender inequality.

The analytical strategy involved constructing a series of Bernoulli models for each of the six dependent variables. First, unconditional models were calculated to assess if significant cross-country variation in the dependent variable existed (i.e., whether HLM was necessary). Second, the dependent variable (e.g., general sport participation, 0 = nonparticipant and 1 = participant) was regressed onto age (0 = >25, 1 = 15-25), education (0 = secondary education or lesser and 1 = more than secondary), and sex (0 = female and 1 = male) at Level 1 using a random intercept and random slope for sex.

To visualize how sex differences in the dependent variables might vary as a function of countries’ gender inequality, we conducted multiple logistic regressions to produce odds ratios for each dependent variable within each country. These odds ratios were then graphed according to the GII value of their corresponding country.

Results

The initial unconditional models that were constructed to test if HLM was necessary revealed that sex differences varied across countries in all but one (i.e., reason to play sport: to look good). Consequently, Bernoulli models were only constructed for the remaining five outcome variables. Overall, the results suggest that males are more likely to report that sport is their most common activity and to agree that one plays sports to compete against others, even when accounting for countries’ gender inequality. The following sections and accompanying tables describe each dependent variable in detail.

In Model 1, which examined general sport, those who were highly educated, younger, and male were more likely to report a sport as their most frequent physical activity (see Table 2). In Model 2, which examined team sport, those who were highly educated, younger, and male were more likely than females to report a team sport as their most frequent physical activity (see Table 2). In Model 3, which examined watching sport, those who were highly educated were as likely to report watching sport as those not highly educated, younger, and males. In Model 4—which examined attendance at sport events—those who were highly educated, younger, and male were more likely to attend sporting events (see Table 2). In Model 5, which examined agreement with a reason for sport—to compete—those who were highly educated, younger, and males were more likely to agree that sport is for competing against others (see Table 2).

Table 2.

Hierarchical Bernoulli Modeling for Sport (Models 1 and 2) and Team Sport Participation (Models 3 and 4).

	Model 1: Sport	Model 2: Team sport	Model 3: Watch sport	Model 4: Attend sport	Model 5: Compete
Level 2 variables	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)
Intercept	−2.60*** (.17)	−3.67*** (.16)	−0.22 (.19)	−3.75*** (.70)	−1.77*** (.17)
GII	1.56** (.91)	2.98** (.93)	2.43** (.60)	3.43*** (.23)	3.69*** (.83)
Level 1 variables	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]
Sex (male)	4.88*** [3.13, 5.34]	6.10*** [5.19, 7.19]	3.18*** [1.69, 3.17]	2.15*** [1.67, 2.78]	1.98*** [1.80, 2.18]
Education level	1.16** [1.04, 1.31]	0.97 [0.86, 1.09]	1.03 [0.96, 1.11]	0.84** [0.74, 0.95]	0.77*** [0.71, 0.85]
Age
15-30	6.56*** [4.88, 8.82]	19.07*** [15.94, 23.00]	1.47*** [1.27, 1.71]	2.09*** [1.74, 2.52]	1.93*** [1.66, 2.24]
31-45	3.15*** [2.42, 4.11]	8.54*** [7.53, 9.68]	1.28*** [1.26, 1.71]	1.88*** [1.62, 2.17]	1.31*** [1.17, 1.48]
46-60	1.66*** [1.30, 2.12]	3.42*** [2.94, 3.97]	1.14** [1.12, 1.47]	1.42*** [1.25, 1.61]	1.09 [1.00, 1.21]

Note. GII = gender inequality index; OR = odds ratio; CI = confidence interval.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

Figure 1 depicts how sex differences vary according to the GII of the corresponding country. This figure demonstrates that although there are several abnormally high sex differences in some countries with greater gender inequality, on the whole, sex differences do not seem to increase as GII increases. Indeed, across the three most gender-equal countries in the sample (Sweden, Norway, Switzerland), males were, on average, 11.3 times more likely to report a team sport as their most common physical activity, while in the three more gender-unequal countries (Philippines, South Africa, Dominican Republic), males were only 3.67 times more likely.

Figure 1.

Countries’ odds ratios for males, relative to females, for all dependent sport, team sport, watching, attending, and reason: to play sport to compete.

Discussion

Overall, even when accounting for the moderating effect of gender inequality, substantial and consistent sex differences in sport persist. Most notably, males were nearly 5 times as likely to report a sport as their most frequent physical activity and 6 times as likely for reporting a team sport. Furthermore, males were more than 3 times as likely to report watching sport on television, and twice as likely to report attending sporting events and to agree that they play sport to compete against others. These findings support the spectator lek hypothesis, and strengthen the claim that sex differences in sport participation, fandom, and reasons for playing sport may stem, in part, from universal, evolved predispositions (Deaner et al., 2016; Deaner & Smith, 2013; Lombardo, 2012).

Does Gender Inequality Affect Females’ Opportunities for Sport?

At the sociocontextual level, there are a number of barriers that influence sex differences in sport participation, such as differential access to public recreational areas and sport programs. For example, the enactment of Title IX, a U.S. federal law that required government-funded educational institutions to provide equal educational and sporting resources to males and females, substantially increased female sport participation within high schools and colleges (National Federation of State High School Associations, 2010; National Collegiate Athletic Association [NCAA] Research, 2010). Nonetheless, sex differences in sports behavior within the contemporary United States are still substantial (Deaner et al., 2012).

Another potential mediating mechanism that concerns opportunities is population-level changes in females’ reproductive goals, which may enable females more time to invest in sport. Increased gender equality may increase females’ autonomy and education, which may shift females’ reproductive goals such that females have fewer children but invest relatively more resources in those children (Subbarao & Raney, 1995). This shift may allow females to engage in sport more because they have delayed marriage and the age of first birth. For example, female education is negatively associated with number of offspring and positively associated with better health outcomes (Carneiro, Meghir, & Parey, 2013). Moreover, Carneiro and colleagues (2013) found that mothers with higher levels of education invested more in their children’s leisure time.

Does Gender Inequality Affect Females’ Interest in Sport?

A critical question is whether empowering women will eventually extinguish sex differences in sport interest. There is empirical evidence to doubt such claims. First, even in societies with relatively high gender equality, females are statistically underrepresented in sport (e.g., Apostolou, 2015; Deaner et al., 2012) and report less interest than males (Evans, Schweingruber, & Stevenson, 2002; Jacobs, Lanza, Osgood, Eccles, & Wigfield, 2002). Second, if females were equally interested in sport but lacked opportunities, then females should be equally less likely to engage in noncompetitive physical activities. However, gender inequality seems to fully moderate sex differences in physical activity (Van Tuyckom, Van de Velde, & Bracke, 2013) but does not fully or even substantially moderate sex differences in sport, as this study demonstrated. Finally, longitudinal evidence from youth in the United States suggests that relative sex differences in the perceived “coolness” of sport have apparently remained constant, despite the enactment of policies that empower female sport. For example, Buchanan, Blankenbaker, and Cotton’s (1976) early finding that young boys perceive sport as more important for popularity than do females has been replicated twice over the past several decades (e.g., Chase & Dummer, 1992; Chase & Machida, 2011) despite significant cultural change (e.g., converging gender roles) and the enactment of Title IX (Brake, 2010).

Future Research and Practical Applications

One avenue of future research that may help explain cross-country variation in sex differences in sport involvement is country-level averages of 2D:4D. Such an analysis should reveal whether 2D:4D (a proxy for prenatal testosterone exposure) drives the relationship between gender inequality and sport. Indeed, a recent study by Manning, Fink, and Trivers (2014) found that a country’s residual differences between the average males and females’ ratio of the length of the second digit and fourth digit (which is a reliable marker of sex differences in prenatal exposure to testosterone; Lutchmaya, Baron-Cohen, Raggatt, Knickmeyer, & Manning, 2004) is correlated with measures of gender inequality at the country level. Given that testosterone exposure increases competitive social orientations (Eisenegger, Haushofer, & Fehr, 2011) and prenatal testosterone exposure is associated with interest (Berenbaum, 1999; Frisén et al., 2009) and achievement in sport (Tester & Campbell, 2007), country-level sex differences in prenatal testosterone exposure could contribute to both countries’ sex differences in sport and countries’ gender inequality. From this view, countries’ average sex differences in prenatal testosterone exposure could be a common causal variable that explains why country-level gender inequality could be correlated but not be causally related with sex differences in sport participation.

Another potential driver of this relationship is country-level variation in polygyny (Ember, Ember, & Low, 2007; Murdock & White, 1969). Due to greater variance in the acquisition of mating partners within polygamous societies, males may show greater motivation to engage in competitions that afford social status (such as sport) and thus acquire mates. Supporting this, Ember (1974) demonstrated that polygyny may be due to an imbalanced sex ratio stemming from high male mortality in warfare. A more recent analysis by Kruger (2010) found that polygyny—which is theorized to intensify male mating competition—is associated with increased male mortality. Another possible avenue is a country’s historical parasite prevalence (Murray & Schaller, 2010), which may predict preferences for certain types of sport, such as team sport. In particular, the threat of pathogen transmission via intergroup contact may have selected for a psychology of out-group aversion or hostility, and thus regional pathogen prevalence may predict increased preferences for intergroup competition among team sports. Indeed, pathogen prevalence is associated with preferences for group-based morals (e.g., authority/respect, in-group/loyalty; van Leeuwen, Park, Koenig, & Graham, 2012; Winegard & Deaner, 2010). Similarly, disease avoidance variables (e.g., disgust avoidance) are associated with in-group favoritism (Navarrete & Fessler, 2006) and social dominance orientation (Hodson & Costello, 2007).

Our results have practical implications. First, this study calls into question the validity of the social constructivist thesis that sex differences are entirely attributable to sociocultural factors. If, on average, males are more predisposed to have interest in sport and team sport, then it will be worthwhile to reexamine sport policies that assume the opposite (Deaner et al., 2012).

Limitations

This study assumes that the sex difference in sport participation reflects a sex difference in preferences, but it may (partly) reflect sex differences in opportunities or external incentives. More direct measures of preferences and more comprehensive designs are needed to test the link between preferences for sport and actual sport participation. Nonetheless, this limitation would not explain why males report far more interest in sport (see further Deaner et al., 2016). Another potential limitation is that relying only on individuals’ most frequent physical activity ignores the totality of participants’ sport participation profile, which when accounted for may alter the sex differences.

This study was also limited as it relied on only 34 countries, many of which can be considered WEIRD (i.e., Western, educated, industrialized, rich, and democratic; Henrich, Heine, & Norenzayan, 2010). Henrich and colleagues (2010) have noted that compared with the rest of the world, people from WEIRD societies are outliers in many aspects of psychology. To further examine the universality of sex differences across different physical activities, it is necessary to study a greater number of more culturally diverse societies. Tests of universality would also benefit from case study methodologies that examine those cultures that have the strongest chances of falsifying theories that predict universality. For example, girls within matrilineal societies of the Khasi people in northeast India seem to exhibit more competitive behavior than boys in a simple risk-reward-based ball-tossing game (Gneezy, Leonard, & List, 2009), although there is no indication that Khasi girls show substantial sport interest (Deaner & Smith, 2013).

Another limitation of this study was the potential nonindependence of countries (Borgerhoff Mulder, 2001; Chick, 2000; Nunn, 2011). In particular, it would be desirable to test whether the moderating effect of countries’ gender inequality on sex differences in sport became stronger (or weaker) once the shared history (e.g., religious, colonial) of some countries was accounted for. This issue could be addressed in future studies with phylogenetic comparative methods.

Conclusion

This study highlights the importance of empirically accounting for the role of cultural variation—in this case, gender inequality—when examining evolved universal design. This study demonstrated that even when controlling for gender inequality, sex differences in sport remain. Males were nearly 5 times as likely to report a sport as their most frequent physical activity and 6 times as likely for reporting a team sport. Furthermore, males were more than 3 times as likely to report watching sport on television, and twice as likely to report attending sporting events and to agree that they play sport to compete against others. Nonetheless, future research should test for the moderating role of gender inequality in larger and more diverse samples. This finding thus supports the existence of a male predisposition for sports interest, a phenomenon predicted by the spectator lek hypothesis.

Footnotes

Acknowledgements

The authors would like to thank Mark Eys and Laurene Rehman for helpful comments and criticisms of earlier drafts.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The first author received scholarship funding from the Nova Scotia Helath Research Foundation.

Author Biographies

Shea M. Balish received an interdisciplinary PhD in sport psychology from Dalhousie University, an MSc degree in developmental sport psychology from Queen’s University. He currently holds a Banting Postdoctoral Fellowship at the University of Toronto, studying basic models of motivation in sport and exercise settings. He uses an interdisciplinary approach that focuses on multiple analytic levels (e.g., biological, individual, team, community, country) in relation to the determinants and outcomes of sport participation. His research is informed by evolutionary theory, and centers on how specific affective systems regulate status striving and friendship alliances, and how these systems can help explain youth sport participation, including youth sport attrition (e.g., dropout) and gender differences therein. He often tests theories that make claims of universality or cross-cultural variability and thus studies diverse societies.

Robert O. Deaner is an associate professor in the Department of Psychology at Grand Valley State University. His research focuses on sex and gender differences in sports motivation and cross-societal variation in girls and women’s sports. He is considered a leading expert in the evolutionary informed study of motivation for sport, especially sex differences therein.

Daniel Rainham is the Elizabeth May Chair in Sustainability and Environmental Health and an assistant professor in the Environmental Science Program. He is also a research associate with the Atlantic Health Promotion Research Centre and holds various cross-appointments at Dalhousie University. His research focuses on environmental patterns and changes as determinants of healthy behavior, including measures of social class.

Chris Blanchard is a professor at Dalhousie University with cross-appointments in psychology and health and human performance. He currently holds a Tier II Canada Research Chair in Cardiovascular Disease and Physical Activity, and was previously the University of Ottawa Research Chair in Population Health where he received an Ontario Career Scientist Award in Gender and Health.

References

Apostolou

(2015). The athlete and the spectator inside the man: A cross-cultural investigation of the evolutionary origins of athletic behavior. Cross-Cultural Research, 49, 151-173. doi:10.1177/1069397114536516

Balish

S. M.

Eys

M. A.

Schulte-Hostedde

A. I.

(2013). Evolutionary sport and exercise psychology: Integrating proximate and ultimate explanations. Psychology of Sport and Exercise, 14, 413-422. doi:10.1016/j.psychsport.2012.12.006

Berenbaum

S. A.

(1999). Effects of early androgens on sex-typed activities and interests in adolescents with congenital adrenal hyperplasia. Hormones and Behavior, 35, 102-110. doi:10.1006/hbeh.1998.1503

Borgerhoff Mulder

. (2001). Using phylogenetically based comparative methods in anthropology: More questions than answers. Evolutionary Anthropology, 10, 99-111.

Brake

D. L.

(2010). Getting in the game: Title IX and the women’s sports revolution. New York: New York University Press.

Brown

D. E.

(1991). Human universals. New York, NY: McGraw-Hill.

Buchanan

H. T.

Blankenbaker

Cotton

(1976). Academic and athletic ability as popularity factors in elementary school children. The Research Quarterly, 47, 320-325.

Buss

D. M.

(1995). Psychological sex differences: Origins through sexual selection. American Psychologist, 50, 164-168.

Carneiro

Meghir

Parey

(2013). Maternal education, home environments, and the development of children and adolescents. Journal of the European Economic Association, 11, 123-160. doi:10.1111/j.1542-4774.2012.01096.x

10.

Chalabaev

Sarrazin

Fontayne

Boiché

Clément-Guillotin

(2013). The influence of sex stereotypes and gender roles on participation and performance in sport and exercise: Review and future directions. Psychology of Sport and Exercise, 14, 136-144.

11.

Chase

M. A.

Dummer

G. M.

(1992). The role of sport as a social-status determinant for children. Research Quarterly for Exercise and Sport, 63, 418-424.

12.

Chase

M. A.

Machida

(2011). The role of sport as a social status determinant for children: Thirty years later. Research Quarterly for Exercise and Sport, 82, 731-739.

13.

Chick

(2000). Editorial: Opportunities for cross-cultural comparative research on leisure. Leisure Sciences, 22, 79-91.

14.

Deaner

R. O.

Balish

S. M.

Lombardo

(2016). Sex differences in sport interest and motivation: An evolutionary perspective. Evolutionary Behavioral Sciences, 10, 73-97.

15.

Deaner

R. O.

Geary

D. C.

Puts

D. A.

Ham

S. A.

Kruger

Fles

. . .Grandis

(2012). A sex difference in the predisposition for physical competition: Males play sport much more than females even in the contemporary U.S. PLoS ONE, 7(11), e49168. doi:10.1371/journal.pone.0049168

16.

Deaner

R. O.

Smith

B. A.

(2013). Sex differences in sport across 50 societies. Cross-Cultural Research, 47, 268-309. doi:10.1177/1069397112463687

17.

Eisenegger

Haushofer

Fehr

(2011). The role of testosterone in social interaction. Trends in Cognitive Sciences, 15, 263-271. doi:10.1016/j.tics.2011.04.008

18.

Ember

(1974). Warfare, sex ratio, and polygyny. Ethnology, 13, 197-206.

19.

Ember

C. R.

Low

B. S.

(2007). Comparing explanations of polygyny. Cross-Cultural Research, 41, 428-440.

20.

Evans

E. M.

Schweingruber

Stevenson

H. W.

(2002). Gender differences in interest and knowledge acquisition: The United States, Taiwan, and Japan. Sex Roles, 47, 153-167. doi:10.1023/A:1021047122532

21.

Fredricks

Eccles

(2005). Family socialization, gender, and sport motivation and involvement. Journal of Sport & Exercise Psychology, 27, 3-31.

22.

Frisén

Nordenström

Falhammar

Filipsson

Holmdahl

Janson

P. O.

. . . Nordenskjold

(2009). Gender role behavior, sexuality, and psychosocial adaptation in women with congenital adrenal hyperplasia due to CYP21A2 deficiency. Journal of Clinical Endocrinology & Metabolism, 94, 3432-3439. doi:10.1210/jc.2009-0636

23.

Geary

D. C.

(2010). Male, female: The evolution of human sex differences (2nd ed.). Washington, DC: American Psychological Association.

24.

Gneezy

Leonard

K. L.

List

J. A.

(2009). Gender differences in competition: Evidence from a matrilineal and a patriarchal society. Econometrica, 77, 1637-1664. doi:10.3982/ECTA6690

25.

Henrich

Heine

S. J.

Norenzayan

(2010). The weirdest people in the world? Behavioral & Brain Sciences, 33, 61-83. doi:10.1017/S0140525X0999152X

26.

Hodson

Costello

(2007). Interpersonal disgust, ideological orientations, and dehumanization as predictors of intergroup attitudes. Psychological Science, 18, 691-698. doi:10.1111/j.1467-9280.2007.01962.x

27.

Jacobs

J. E.

Lanza

Osgood

D. W.

Eccles

J. S.

Wigfield

(2002). Changes in children’s self-competence and values: Gender and domain differences across grades one through twelve. Child Development, 73, 509-527. doi:10.1111/1467-8624.00421

28.

Kruger

D. J.

(2010). Socio-demographic factors intensifying male mating competition exacerbate male mortality rates. Evolutionary Psychology, 8(2). doi:10.1177/147470491000800205

29.

Lombardo

(2012). On the evolution of sport. Evolutionary Psychology, 10, 1-28.

30.

Lowen

Deaner

R. O.

Schmitt

(2016). Guys and gals going for gold the role of women’s empowerment in Olympic success. Journal of Sport Economics, 17, 260-285. doi:10.1177/1527002514531791

31.

Lutchmaya

Baron-Cohen

Raggatt

Knickmeyer

Manning

J. T.

(2004). 2nd to 4th digit ratios, fetal testosterone and estradiol. Early Human Development, 77, 23-28. doi:10.1016/j.earlhumdev.2003.12.002

32.

Mace

Pagel

Bowen

J. R.

Gupta

B. K. D.

Otterbein

K. F.

Ridley

. . . Voland

(1994). The comparative method in anthropology [and comments and reply]. Current Anthropology, 35, 549-564.

33.

Manning

J. T.

Fink

Trivers

(2014). Digit ratio (2D:4D) and gender inequalities across nations. Evolutionary Psychology, 12, 757-768.

34.

Murdock

G. P.

White

D. R.

(1969). Standard cross-cultural sample. Ethnology, 8, 329-369.

35.

Murray

D. R.

Schaller

(2010). Historical prevalence of infectious diseases within 230 geopolitical regions: A tool for investigating origins of culture. Journal of Cross-Cultural Psychology, 41, 99-108. doi:10.1177/0022022109349510

36.

National Federation of State High School Associations. (2010). 2009e10 high school athletics participation survey. Indianapolis, IN: Author.

37.

Navarrete

C. D.

Fessler

D. M.

(2006). Disease avoidance and ethnocentrism: The effects of disease vulnerability and disgust sensitivity on intergroup attitudes. Evolution & Human Behavior, 27, 270-282. doi:10.1016/j.evolhumbehav.2005.12.00

38.

National Collegiate Athletic Association (NCAA) Research. (2010). NCAA sport sponsorship and participation rates report, 1981e82 d 2008e09. Indianapolis, IN: Erin Irick.

39.

Norenzayan

Heine

S. J.

(2005). Psychological universals: What are they and how can we know? Psychological Bulletin, 131, 763-784. doi:10.1037/0033-2909.131.5.763

40.

Nunn

C. L.

(2011). The comparative approach in evolutionary anthropology and biology. Chicago, IL: The University of Chicago Press.

41.

Permanyer

(2013). A critical assessment of the UNDP’s gender inequality index. Feminist Economics, 19(2), 1-32.

42.

Pollet

T. V.

Tybur

J. M.

Frankenhuis

W. E.

Rickard

I. J.

(2014). What can cross-cultural correlations teach us about human nature? Human Nature, 25, 410-429. doi:10.1007/s12110-014-9206-3

43.

Raudenbush

S. W.

Bryk

A. S.

Congdon

(2004). HLM 6 for Windows [Computer software]. Lincolnwood, IL: Scientific Software International.

44.

Schmitt

D. P.

(2014). The evolution of culturally-variable sex differences: Men and women are not always different, but when they are . . . it appears not to result from patriarchy or sex role socialization. In Weekes-Shackelford

V. A.

Shackelford

T. K.

(Eds.), The evolution of sexuality (pp. 221-256). New York, NY: Springer.

45.

Scholz

Heller

(2009). ISSP study monitoring 2007 (German Social Sciences Infrastructure Services Technical Reports, 2009/5). Mannheim, Germany: German Social Sciences Infrastructure Services. Retrieved from http://www.gesis.org/fileadmin/upload/forschung/publikationen/gesis_reihen/gesis_methodenberichte/2009/TechnicalReport_09-5.pdf

46.

Snijders

T. A. B.

Bosker

R. J.

(1999). Multilevel analysis: An introduction to basic and advanced multilevel modeling. London, England: SAGE.

47.

Subbarao

Raney

(1995). Social gains from female education: A cross-national study. Economic Development and Cultural Change, 44, 105-128.

48.

Tester

Campbell

(2007). Sporting achievement: What is the contribution of digit ratio? Journal of Personality, 75, 663-678. doi:10.1111/j.1467-6494.2007.00452.x

49.

United Nations Development Programme. (2011). The 2011 human development report—sustainability of equity: A better future for all. Retrieved from http://hdr.undp.org/en/content/human-development-report-2011

50.

van Leeuwen

Park

J. H.

Koenig

B. L.

Graham

. (2012). Regional variation in pathogen prevalence predicts endorsement of group-focused moral concerns. Evolution & Human Behavior, 33, 429-437. doi:10.1016/j.evolhumbehav.2011.12.005

51.

Van Tuyckom

Van de Velde

Bracke

. (2013). Does country-context matter? A cross-national analysis of gender and leisure time physical inactivity in Europe. The European Journal of Public Health, 23, 452-457.

52.

Winegard

Deaner

R. O.

(2010). The evolutionary significance of Red Sox nation: Sport fandom as a by-product of coalitional psychology. Evolutionary Psychology, 8, 432-446.

53.

Wood

Eagly

A. H.

(2012). Biological construction of sex differences and similarities in behavior. In Zanna

(Ed.), Advances in experimental social psychology (Vol. 46, pp. 55-124). San Diego, CA: Academic Press.