Political Gender Gap and Social Dominance Orientation

Abstract

In the last 30 years, there was a rise in the political gender gap. The Center for the American Woman and Politics data shows that a larger proportion of women than men vote for the Democratic Party. Women tend to differ from men in several political issues endorsing more welfare policies and progressive policy changes and are less racist than men. Social dominance orientation (SDO) has been theorized to account for political gender differences. Men have higher average levels of SDO than women, everything else being equal. To test the gender invariance hypothesis, we believe it’s important to take into consideration people who not only identify with groups that hold specific hierarchy-enhancing or hierarchy-attenuating ideologies but who participate in them promoting their values and ideas. In this chapter, we describe the findings of research on gender differences in SDO of activists in political parties that range from extreme left-wing to extreme right-wing.

1. Introduction

The increase of participation of women in politics has revealed that women and men are political actors with distinct political preferences [1, 2]. In the last 30 years, there was a rise in the political gender gap. The Center for the American Woman and Politics [1] data show that a more substantial proportion of women than men vote for the Democratic Party. The last presidential elections (2016) revealed a sizable 11 percentage-point gender gap, 42 percent of women voted for Trump versus 53 percent of men [1]. This is not a new phenomenon according to the data of CAWP in the last two decades, since in the 1996 presidential election, women voters tend to prefer more a democratic candidate than men (the gender gap variance has varied in these years from a minimum of 7 points to a maximum of 11 points). The 2016 gender gap was one of the largest ones. Also, in Europe, most countries show either no gender gap or that women are more left-wing than men. Recent research [3] based on the analysis of the European Values Study/World Values Survey that combines data spanning from 1989 to 2014 reveals that there is a gender-generation gap. In the younger cohorts, women are more left-wing oriented than men. Researches on the gender difference in political issues point out that there is a wide difference in programs and issues that women and men support. Women, in general, are in favor of government spending on social welfare, education, and health. They are more likely to favor programs for medical care, schooling, and gun control.

On the contrary, they tend to oppose more military spending or the use of force to solve conflicts and are against capital punishment [4, 5, 6]. They support less discriminatory policies and have more positive attitudes toward homosexuals than men [7]. Women have lower levels of prejudice [8], authoritarianism [9, 10], and anti-egalitarianism, [11] are more worried about potential international conflicts [12], and, in general, hold less punitive attitudes [11]. Social dominance orientation (SDO) has been theorized to account for political gender differences [13, 14, 15].

2. Social dominance orientation

The social dominance theory (STD) aims to understand how group-based social hierarchy is formed and preserved [13]. According to Sidanius and Pratto [13], postindustrial societies tend to develop group-oriented social hierarchies that support long-term human survival. In these hierarchical societies, intergroup conflicts and oppressions contribute to maintaining the status quo of the social system. SDT suggests that an individual orientation called social dominance orientation [13, 16, 17] is a potential explanatory factor of sociopolitical sex differences. The SDO has been defined as a personal desire for group-based dominance, mirroring an individual’s support for group-based hierarchies [13]. People higher in SDO tend to support hierarchy-enhancing legitimizing myths such as prejudice, racism, sexism, militarism, support for the death penalty, and coercive social power across societies and contexts [16, 18, 19, 20, 21, 22]. Men tend to score higher than women in SDO [13]. Such differences may be, to a certain extent, determined by the desire of males to justify their dominant position in society. As Sidanius and Pratto [13] point out, our contemporary hierarchical system is mostly “andrarchical” since men tend still occupy most of the highest positions of political and economic power. Therefore, men should support social systems that maintain hierarchies since they tend to hold privileges due to occupying higher positions in society.

In our societies, women and men usually have different roles in the group-based hierarchy. Men tend to be more numerous in the police, military, lawyers, judges, and business executives areas, whereas women are in a more significant number in the teachers, social workers, and charity volunteer areas [23]. Overall, men are inclined to participate in institutions or hold roles that enhance hierarchy and females on the contrary to institutions that diminish hierarchy [24]. The SDT maintains that those that occupy positions in society that reinforce the existing group inequality or strengthen in-group status are more likely to be social dominance-oriented than out-groups are.

SDT claims that men and women should exhibit differences in SDO due to strategies that follow from evolutionary theory. Sidanius et al. [25] maintain that different psychological and behavioral predispositions between males and females in terms of sexual and reproductive behavior are the core of gender difference in society. From this perspective, sex differences in orientation toward group-based social inequality (SDO) are the effects of human reproductive strategies. Sidanius and Pratto [13] put forward that reproductive inequality implies economic inequality and economic inequality implies political inequality. Sidanius, Pratto, and Bobo [25] formulate the gender invariance hypothesis from a perspective of theoretical biocultural interaction: “Not only should men have a higher average level of SDO, but this higher average level of male SDO should also be found after cultural, situational and environmental factors are considered” (p. 1000) [25].

Two invariance hypotheses have been proposed. The “strong” version and the “soft” one. The strong version claims that SDO differences between men and women should not vary across cultural factors, situational factors, or both. There should, therefore, be no significant interaction between sex gender and cultural-situational factors. In other words, the strong version of the invariance hypothesis predicts that the difference in SDO between men and women should be essentially invariant across all major cultural, environmental, and situational factors such as country of national origin, ethnicity, education, income, age, political ideology, racism, religious beliefs, and gender role attitudes. However, since the claim of the biological roots of gender differences is less relevant in SDT today [18], a contextual variation is, to some extent, allowed in the soft version hypothesis [26]. The soft version of the invariance hypothesis asserts that men will always show higher levels of SDO than women, everything else being equal. It claims that although gender might interact with several cultural-situational factors, this interaction will always be ordinal and never disordinal. Whereas the male-female differences in SDO might show some significant variations across cultural factors, situational factors, or both, females should never have significantly higher SDO than males within the same sociocultural context [25]. Both socialization experiences and belonging to hierarchy-attenuating or hierarchy-enhancing settings can increase or diminish SDO [27]; however, women should never have a significantly higher SDO than men. For instance, different professional groups may vary in their levels of SDO [20]; however, within a specific professional group, men should report higher levels of SDO than women [28].

Quite a lot of studies have attempted to investigate, if and under which, circumstances the invariance hypothesis holds. Several studies conducted mainly by Sidanius and colleagues supported the validity of the invariance hypothesis, both with samples of students and adult residents of the United States and in many foreign countries [13, 14, 15, 24, 25, 29, 30].

For example, in their cross-cultural study on male-female difference in SDO that involved 10 countries (Australia, Canada, Israel, Mexico, Palestine, Republic of China, New Zealand, the former USSR, Sweden, and the United States), Sidanius and Pratto showed that males are significantly more social dominance-oriented than females in 39 of the 45 samples [13].

Also, Wilson and White [31] in their study based on students and adults revealed that males were more social dominant and politically conservative than women. Social dominance mediated the relationship between gender and conservatism.

Furthermore, studies confirm that even in countries that traditionally promote gender equality, the gender gap in social dominance orientation prevails [32, 33].

Contrary evidence emerged, however, in other studies. Research based on student and adult samples from Australia, the United States, Ireland, and Sweden did not confirm the main gender effect [34, 35, 36, 37, 38].

In Taiwan, females scored higher than males, but the difference was not significant [39], and in two samples in Israel and Australia, men did not score significantly higher than females [34, 39]. In Küpper and Zick’s [40] first study, women unexpectedly showed higher levels of SDO than men.

Some studies on the gender invariance hypothesis investigated whether group differences in SDO can be explained by group identification.

Wilson and Liu [41], following the social identity theory (SIT) perspective, predicted that males who identify strongly with gender group should exhibit higher SDO scores than low-identifying males and that females who identify strongly with their gender group should score lower than low-identifying females. Their findings showed that the gender-SDO relationship was moderated by the strength of gender in-group identification: increasing group identification was associated with decreasing SDO scores for females and increasing SDO scores for females. Sidanius and Pratto [28], however, criticized this study for not meeting the criteria, “all else being equal in principle.” They underlined that they should have compared men and women with similar levels of gender identification.

Also, Huanga and Liu [42] analyzed the controversy in the literature concerning whether group differences in SDO can be explained by group identification. They hypothesized that if SDO acts as a stable individual difference, it should maintain its relative relationship with gender (i.e., men should have higher SDO than women) even when the demographic group is saliently primed. Alternatively, from a situational priming perspective, one might expect gender differences in SDO to be significant only when gender is salient. Their first research involved 1605 adults in Taiwan, and they found that contrary to SDT’s invariance hypothesis, men were higher on SDO than women only when gender was salient.

Foels and Pappas [43] tested the invariance hypothesis by measuring the relationship between sex and SDO while controlling for the effects of gender socialization. They demonstrated that the sex difference in SDO is mediated by gender socialization.

Lee et al. [33] addressed the dispute between SDT and social identity theory (SIT) in a meta-analysis. Their research showed that in what has been predicted by SDO, gender differences on SDO were more substantial and more stable than differences between ethical and racial groups in the United States and worldwide.

Other studies on the gender invariance hypothesis explored the influence of various kinds of presumed hierarchy-enhancing or hierarchy-attenuating settings. Several studies have shown that university majors and career choices are associated with either hierarchy-enhancing (HE, e.g., racism) or hierarchy-attenuating (HA, e.g., human rights) legitimizing myths [44]. Dambrun et al. [45] examined the impact of HE vs. HA academic major on stereotyping. They found that students in psychology were less social dominance-oriented than students in law. Moreover, while males were more social dominance-oriented than females in law, no sex difference was found for psychology majors. Authors conclude that their results “can be taken to suggest that social-cultural variables may affect scores on SDO and modify gender differences on SDO” (p. 130). They also notice that female law students had higher SDO scores than male psychology students; this finding is in opposition to the strong version of the invariance hypothesis that men should always score higher than females in SDO.

Sidanius et al. [30] showed conflicting results. In their longitudinal study, they measured the SDO of men and women once a year for the 4- and half-year period. Their findings show that even after controlling for the characteristics of students’ academic majors (hierarchy-enhancing or hierarchy-attenuating), males showed significantly higher SDO scores than females did, across the entire college career.

Research, based in Sweden, on gender differences in SDO in social structures varying in equality enhancement and gender composition revealed a main effect of gender on SDO despite the degree of political equality or gender composition. There was an interaction effect only in associations where women were the majority of members [26].

Bathalka et al., [46] investigated the gender invariance hypothesis in similar cultural, ideological, and status contexts. Their findings revealed either no effect for gender or an interaction between gender and the relevant social context and only a small effect size of gender. Overall the authors underline that their results disconfirm the gender invariance assumption of SDT. In their second study, students were categorized according to disciplines HA or HE (literature, languages, psychology, social studies, and anthropology majors were grouped as HA and economics, law, and business as HE majors). Their research showed that whereas HE/HA predicted SDO, gender did not.

Reviewing the literature, we find that most, but not all, studies have found significant differences in favor of males in SDO. However, some studies we discussed showed that males’ SDO scores changed according to environmental and socialization variables such as being embedded in hierarchy-attenuating environments, not identifying strongly with their gender, or living in societies whose cultural values are more egalitarian and less competitive.

3. Aims

Most of the studies that found an invariant gender gap involved students or adolescents [40]. To investigate further the possible causes of the increasing political gender gap, we need to conduct studies with members with strong salient group identities, where the influence of adult socialization egalitarian myths may have the opportunity to influence SDO. On this line, studies on the ideological divide may be done involving militants and politicians. With dispositional features such as personality traits and value differences between conservatives and liberals, right-wing and left-wing voters may be present among ordinary citizens or college students [47] but are more pronounced in groups of party activists, extremists, and politicians. Already in the 1960s, McClosky et al. [48] showed the ideological conflicts were much higher between democratic and republican activists and party leaders than among party voters.

To further explore the gender identity hypothesis and the political gender gap, we need to compare people who not only identify with specific hierarchy-enhancing or hierarchy-attenuating ideologies but who participate actively and continuously to political parties or groups which uphold and promote those ideas. Activism in political parties is a matter of choice in modern society. As Huddy [49] underlines, people who choose to be activists in particular political groups already may hold some hierarchy-enhancing or hierarchy-attenuating ideals. However, their SDO may be heightened or lessened by their prolonged exposure and their internalization of legitimizing myths (i.e., according to “soft” gender invariance hypothesis).

SDO would suggest that being committed activists in a hierarchy-attenuating political groups could make individuals identify with the ideals of the groups. Therefore, male members in these groups could have lower SDO scores than males adhering and participating in hierarchy-enhancing political groups. So, we should find the highest scores of SDO in males belonging to right-wing political groups and the lowest in males active in extreme left-wing groups. However, according to SDT, even while absolute levels of SDO may vary across situations, men should still have relatively higher SDO than women within each political group. The predisposition of males to be temperamentally inclined to dominate, even when exposed to substantial and long-term environmental attenuating pressures, will produce nonetheless residual gender differences even among males and females belonging to groups who uphold egalitarian group values.

To understand if the gender gap in politics could be systematically observed within and across the left-wing and right-wing split, we aimed to compare the accuracy in predicting SDO by gender in four well-defined samples of male and female activists belonging to hierarchy-enhancing political groups – center right and extreme right wing oriented groups - vs. belonging to hierarchy attenuating political groups - center left and extreme left wing oriented groups.

As predicted by SDT, we should find higher mean SDO scores among male activists of extreme right-wing, hierarchy-promoting political groups and lower mean SDO scores in males members of extreme left wing, hierarchy-attenuating political groups. However, according to the invariant gender hypothesis of SDT, the difference between males and females within each group should be invariant across groups: the gender divide should be stronger than the political divide.

On the other hand, as more sustainable within a SIT perspective and gender similarities hypothesis [50], we should find no gender differences both in the hierarchy-attenuating extreme left-wing political groups and hierarchy-enhancing extreme right-wing political groups. Both settings should reinforce both in male and female members the dominant legitimizing myths of the political group they have chosen, and the political divide should be stronger than the gender divide.

4. Method

4.1 Participants

There is a general consensus in literature that SDO is a stable individual difference [51] although can relatively vary across some conditions. According to person-environment fit theory, authors [27] posited that people select hierarchy-enhancing environments according to their SDO levels, as well as environments attract and socialize people according to how much in such places hierarchy-enhancing myths are supported. As a result, high-SDO people tend to fit better in hierarchy-enhancing environments and low-SDO people tend to fit better in hierarchy-attenuating environments (e.g., [18, 27, 52]). In the present chapter, we opted to study the gender invariance hypothesis in people belonging to political groups with a different support of hierarchy-enhancing legitimizing myths accordingly to the literature that outlines that the SDO level among participants (of both sex) of political groups mirrors the different extents to which parties support group dominance.

We included 626 subjects, 350 males and 276 females, who had been for at least 2 years activists in political parties or associations belonging to two well-differentiated groups of (1) extreme left-wing and (2) extreme right-wing and having also two additional groups of (3) center left-wing and (4) center right-wing parties. We decided to invest considerable effort to secure a relatively large sample of political activists of different political parties. We sent emails to the address of local political parties asking to give the questionnaire to the activists. The questionnaire was accompanied by a letter of the Sapienza University of Rome stationery presenting the aims of the scientific research and guaranteeing anonymity and privacy. After mailing questionnaires and letters, we tried to recontact the political parties to ascertain that they received the questionnaires and to solicit their responses. However, since the response rate was low, after the first contacts had been established, snowball sampling was used to recruit other political activists:

1. Extreme left activists were 126, 70 males and 56 females (mean age 26.21, SD 4.83; range 16–34). About 9% have a low level of education, 61% a high school diploma, and 30% a college degree. These extreme left-wing organizations define themselves as supporting socialist, feminist, and ecologist issues and fighting against all social inequalities based on race, ethnicity, gender, or other discriminating features. Activists spend considerable time in these groups in weekly meetings and organizing protest marches, boycotts, and sit-in in favor of oppressed groups.

2. Extreme Right activists were 181, 123 males and 58 females (mean age 24.01, SD 5.07; range 14–34). About 23% had a college degree, 57% had a high school diploma, and about 20% did not finish high school. These groups promote attitudes and belief systems such as nationalism, racism, classism, sexism, ethnocentrism, and political-economic conservatism. Extreme right activists also meet weekly and often engage in nontraditional political activities bordering on illegality such as unauthorized protest marches that sometimes end in violence (fights with police officers or leftwing extremists).

3. Center left-wing activists were 111, 50 males and 61 females (mean age 26.71, SD 4.86; range 16–35). They are members of the center left-wing moderate parties. About 7% have a junior high diploma, 67% a high school diploma, and 26% a college degree. They engage in more traditional legal-political activities such as signing petitions, political campaigns, raising funds, and getting people to the voting polls.

4. Center right-wing activists were 208, 107 males and 101 females (mean age 27.20, SD 4.76; range 14–35). They belong to center right-wing parties. About 68% have a high school diploma, 24% a college degree, and 8% a junior high education. They also engage in more traditional party activities, like organizing fundraising events, helping party candidates, distributing documents, and getting voters to the poll.

4.2 Instruments

All subjects filled a questionnaire which contained:

1. A section in which subjects recorded age, sex, and educational level. Furthermore, to confirm and control the distinctive SS belonging to the selected groups, we measured their political orientation by means of a single item (a 10-point scale), where point 1 meant extreme right-wing orientation and point 10 meant extreme left-wing orientation.

2. SDO scale used was an Italian adaptation of the SDO scale [53].

5. Statistical analyses and results

We first performed an analysis of variance to ascertain the political orientation as firmly acting differences within groups in SDO males and females scores.

Results are shown in Table 1.

Gender	Political orientation	Means	Gender		Political orientation		Gender × political orientation
			F	P	F	p	F	p
Males	Extreme right-wing	3.74	19.85	<0.001	192.00	<0.001	3.16	<0.05
	Center right	3.27
	Center left	1.84
	Extreme left-wing	1.73
	Total	2.93
Females	Extreme right-wing	3.12
	Center right	3.03
	Center left	1.79
	Extreme left-wing	1.46
	Total	2.46

Table 1.

Four political groups, male and female—SDO means.

To deepen the test of the invariance hypothesis, we applied a multiple regression model involving social dominance orientation and gender across the four groups considered. A multigroup path analysis was performed to test the hypothesis of the influences of gender on social dominance, this way independently of political orientation. A dummy variable was created with 1 corresponding to males. Thus, in our regression model, the weight is the average difference between males and females on SDO.

Figure 1.

Conceptual model of the regression of gender on social dominance orientation.

SDO was treated as an exogenous latent variable with three indices. The latter was formed summing up groups of items of the scale. The figure below illustrates the conceptual model (Figure 1).

Several competing models were tested.

Hypotheses were as follows in ascending order of constraints:

• H1: both factor loadings and regression weight are different for each of the four political orientation groups.

• H0A: regression weight is invariant for each of the political groups.

• H0B: factor loadings of social dominance are invariant for each of the four political groups.

• H0C: both regression weight and factor loading are invariant for each of the political groups.

Results showed in Table 2 ascertain from these results that hypotheses H1 and H0A showed a relevant good fit.

Model	Chi	df	P
H1	13.86	8	0.09
H0A	18.37	11	0.07
H0B	23.86	14	0.05
H0C	29.76	17	0.03

Table 2.

Results of path analysis of the model tested.

In order to choose the best model, D2 statistic was calculated [54] as follows:

$$\text{D2}=\mathrm{Chi}$$ ^$2$ $$\mathrm{H}0\mathrm{A}-\mathrm{Chi}$$ ^$2$ $$\text{H1}$$ E1

$$\mathrm{df}=\mathrm{dfH}0\mathrm{A}-\text{dfH1}$$ E2

So,

$$\text{D2}=18.37\mathrm{–}13.86$$ E3

$$\mathrm{df}=11\mathrm{–}8=3$$ E4

$$\mathrm{p}=0.211.$$ E5

Since the reduction in Chi² is not statistically significant, it is possible to choose the model of the invariance of regression weight among the political groups. Focusing on this model, fit indexes were RMSEA = 0.07; NFI = 0.98; and CFI = 0.99.

Loading differences among the groups did not seem due to a different factorial structure.

Concerning the main hypothesis, these results showed that males express higher SDO than females invariantly of their right-wing or left-wing political orientation (beta = 0.21; R2 = 0.04).

6. Discussion

Recent years have seen a rise in the political gender gap; women in Western European countries have gradually preferred more leftist parties [55]. In the United States, the share of women who identify with or lean toward the Democratic Party has increased. The last data from the Pew Research Center [56] show that among registered voters, 56% of women affiliate or favor the Democratic Party compared to 44% of men. Several authors have hypothesized that the difference between women and men on political attitudes and political party identification can be attributed to differences in SDO [5]. To investigate if the gender gap in politics goes deeper than traditional left-wing and right-wing division, we analyzed gender differences in SDO in activists of HE and HA political parties. According to the gender invariance hypothesis, all else been equal, men should still have significantly higher SDO than women. The predisposition of males to be temperamentally inclined to dominate will produce nonetheless residual gender differences even among males and females belonging to groups who uphold egalitarian group values.

To analyze the gender invariance hypothesis in a political context, we based our study on political activists. Political activists, in fact, not only identify with groups that hold specific hierarchy-enhancing or hierarchy-attenuating ideologies but actively participate, promoting their values and ideas within the group. Their active commitment in specific HE or HA groups should encourage in fact, even more, the internalization of legitimizing myths that may influence their levels of SDO. Our research was based on four samples of male and female activists belonging to hierarchy enhancing – right-wing – vs. hierarchy attenuating – left-wing - political groups. According with SIT and SDT, the SDO scores should be higher in right-wing groups than left-wing groups and highest in extreme rightwing groups and lowest in extreme left groups. However, according to SDT, even while absolute levels of SDO may vary across situations, men should still have significantly higher SDO than women. The SIT theory, on the contrary, would predict that groups on both sides of the political divide should attract males and females who, for the left, are strong egalitarian and do not favor the oppression of one group over another and, for the right males and females, who hold equally strong opposite views. Gender differences should be insignificant since egalitarian adult socialization experiences should promote egalitarian attitudes in both men and women activists and vice versa.

On the whole, our results sustain more the validity of the soft than the strong version of the invariance hypothesis [25]. We observed general differences in SDO score across political groups (e.g., SDO score for right-wing were higher than left-wing); therefore, males of all the activists’ groups showed a higher social dominance than females invariantly belonging to left- or right-wing political groups. The soft version implies indeed that SDO difference between men and women should be essentially constant across cultural and situational factors, everything else being equal. Our study confirmed that SDO differentiated men and women invariantly across cultural and situational factors such as political activism practiced both in moderate and extremist political groups. Male right-wing extremists had the highest SDO scores and female left-wing extremists the lowest. Still, the gender differences persisted in all groups, giving strong support for the temperamental differences in dominance predisposition.

Our data did show also that extreme right-wing women presented significantly higher scores of SDO than men belonging to left-wing parties. This result can be congruent with both SIT and SDT, which emphasizes the importance of group identification and the soft version of the SDO gender hypothesis that recognizes the influence of context.

7. Conclusions

In this study, the recent increase of the gender gap in political elections was addressed by analyzing gender differences in SDO of political activists. Our findings confirm the soft version of gender invariance hypothesis. However, our study has several limits; we still do not know if people who chose these different groups had originally higher or lower SDO or if the experience of belonging to different groups changes the SDO. Furthermore, we did not explore whether changes in SDO occurred in people who were active participants in different political groups. Future studies should ascertain whether SDO scores diminish after being an active member of the political moderate leftist group and if these changes occur in less time or more often when participating in extremist groups with even stronger egalitarian ideals. Moreover, to understand more the influence of group identification maintained by the SIT, the number of years of group involvement should be taken into consideration in future studies. Future research could also explore if new media like partisan Internet sites and social networks can enhance or reduce SDO.

Our results, with their present limits, also show that SDO is strongest in males and females in far right-wing groups, and these findings could help explain why these groups are now rising in recent elections in Europe and several other countries worldwide. One fundamental belief of SDT theory is that certain groups of people are entitled to rule over other groups [13]. The SDO scale elicits agreement-disagreement with statements such as “it’s a good thing that some groups are at the top, and others at the bottom” or “Some groups of people are simply inferior to other groups.” “America first,” which Trump launched, had a very precise meaning, to favor American-born citizens over immigrants. This slogan has been echoed by right-wing nationalist parties in Italy (“Italy first”) and Hungary, Poland, and Turkey. In all these countries, authoritarian leaders have emerged and are sustained by male and female supporters, who perceive themselves similar to their leaders, and in fact, both authoritarian leaders and followers probably share very high SDO. As a matter of fact, recent studies have shown that SDO is related to support for radical right parties [57].