The news that Canada’s Prime Minister Justin Trudeau has named a gender-balanced cabinet — and one that is more representative along other dimensions, including religion and ability — is refreshing. (The cabinet of his predecessor, Steven Harper, included six women among its 26 members in 2006).

Even more encouraging than Trudeau’s leadership on this issue — and his lack of excuses — is his apparent understanding of the need to address inequality. When asked why he named a government comprised of 50 percent women, he said simply, “Because it’s 2015.” As succinct as his response was, it reflects much greater clarity about the problem of inequality — and its solution — than much public discussion on the issue.

Scholars who study politics and gender reject conventional wisdom that women’s marginalization from leadership is due to poorer performance or lesser qualifications (women now outnumber men in advanced degrees in the United States, as in conservative Islamic countries), arguing instead that it reflects unfair recruitment practices. In addition to gender-based biases, inequality can also arise from basic social practices, including the tendency to socialize with others of the same gender. Men simply know fewer women and often prefer to work with other males. This observation helps expose the fallacy that gender equality will be easily achieved without institutional solutions like quotas.

Unfortunately, debates about gender have also figured prominently in the U.S. presidential primary races, and in troubling ways. Despite the presence of female candidates Hillary Clinton and Carly Fiorina, public debate on and media coverage of the issue are often devoid of an understanding of the mechanisms that create and perpetuate inequality.

In a recent article in Governance — currently ungated — I write about issues of women’s access to government resources and the role that often overlooked structural factors play in contributing to this problem. Although my research takes place in two authoritarian countries in North Africa, it highlights why subtle social dynamics can perpetuate women and other minorities’ lack of access to political power, even when social stereotypes and sexism are diminishing in society in general.

Drawing on an original survey of 200 Moroccan and Algerian parliamentarians and field research conducted 2006-2007, I investigate the link between parliamentarian gender, quotas, and service provision to women. A form of constituency service, parliamentary clientelism refers to providing help with personal and community problems, including medical services, judicial or bureaucratic corruption, jobs, money or grants, or electricity. Due to high levels of corruption and economic difficulties in many non-democratic settings, elected officials often spend much of their time trying to help citizens with these issues. Yet I find that women are relatively excluded from these informal service networks with parliamentarians, due in part to structural factors. On average, 20-29 percent of requests that Moroccan and Algerian parliamentarians receive are from women. The effects of gender quotas are striking. While electing women increases service responsiveness to females, parliamentarians elected through quotas experience mandates to help women and are more responsive to females than members of either sex elected without quotas.

To explain gender gaps in clientelistic services — and to think about why progress toward gender parity in United States and many other relatively egalitarian societies has stalled — I extend Elin Bjarnegård’s (2013) theory of homosocial capital, which she developed to explain male dominance in Thai parliamentary politics. Homosociality refers to friendships, collaborations and other non-romantic relationships with others of the same gender. Bjarnegård argues that, as long as males are numerically dominant in positions of power, they will continue to enjoy advantages accumulating homosocial capital — predictable relationships with individuals who are similar and have resources needed to win elections.

According to Bjarnegård, homosocial capital is composed of instrumental and expressive resources. Expressive resources are dispositional similarities which facilitate close, predictable relationships and access to instrumental resources needed to succeed in campaigns.

Men and women often prefer friendship and work collaboration with others of the same gender, due to greater ease relating (e.g., similar behavior and interests, fewer conflicts and sexual harassment concerns, higher trust). Male dominance is reproduced because instrumental and expressive resources are seldom in opposition for men. Males have more instrumental resources and can rely on homosocial linkages. In contrast, women must foster heterosocial networks to obtain resources, while enjoying less expressive resources, such as trust and familiarity, in networks with men.

I extend Bjarnegård’s theory of homosocial capital accumulation by incorporating structural and normative aspects of patriarchy into the framework and identifying two gender gaps: in parliamentarians’ supply of and citizens’ demand for services. Patriarchal structures are formal laws and policies such as personal status (e.g., marriage, divorce, inheritance, mobility) and employment laws (e.g., wage discrimination, sex segregation) which give males more instrumental resources and reproduce stereotypes about women’s traits and abilities (i.e., what women are like and can do), including their ability to provide wasta (use of a connection to gets things one needs). Patriarchal norms and attitudes are traditional views of gender roles which identify public space as male, discourage mixing of unrelated males and females, and justify gender discrimination (i.e., what women should do).

Exclusion of women from service provision occurs through mutually-reinforcing structural and dispositional factors operating on the supply and demand-side of exchanges. On the instrumental, supply-side, patriarchal structures limit female deputies’ instrumental resources needed to provide services. Because female leaders enjoy fewer patronage networks, women are less likely to be seen as good wastas and are less able to run and succeed in elections (see Table 1).

In Tunisia, 19.1 percent saw male deputies as better service providers, compared to 6.9 percent who saw women as better (73.9 percent believed there was no difference), according to a 2014 Transitional Governance Project survey of 1200 Tunisians conducted by United Nations Democracy Fund and Centre d’Études Maghrébines à Tunis, in collaboration with Lindsay Benstead, Ellen Lust, Dhafer Malouche and JMW Consulting.

In Jordan, 45.4 percent saw males as better, compared to 13.5 percent females (41.1 percent, no difference), according to a 2014 Program on Governance and Local Development poll conducted among 1,488 Jordanians by Lindsay Benstead, Kristen Kao and Ellen Lust.

In Libya, 28.2 percent believed a man would be more effective, compared to 10.9 percent women (60.9 percent, no difference), according to a 2013 National Democratic Institute poll of 1200 Libyans conducted by Diwan Research and the Transitional Governance Project, in collaboration with Lindsay Benstead, Ellen Lust, and JMW Consulting.

Table 1. Mechanisms of exclusion from service provision

On the instrumental, demand side, patriarchal structures exclude women from public space (e.g., employment, politics), giving them fewer resources and influence. Women are less likely to be civically engaged or lead civil society groups, Sufi brotherhoods, and tribes — groups which are critical for winning semi-competitive elections.

The Transitional Governance Project found voting in first post-uprisings parliamentary elections was 77 percent for Egyptian men, 58 percent for women; 75 percent for Tunisian men, 65 percent for women; and, 84 percent for Libyan men, 59 percent for women. As a result, serving women is a less effective electoral strategy than serving men.

On the expressive, supply side, patriarchal norms limit female deputies’ interactions with male elites. Female parliamentarians provide fewer services to constituents, who are more likely to be female.

On the demand side, patriarchal gender norms limit interaction and networks with male deputies. Afrobarometer data suggest women are less likely to contact members: 11.7 percent of Moroccan men, compared to 3.5 percent of women (p<.000) and 2.2 percent of Algerian men, compared to 1.7 percent of Algerian women (p<.533) had contacted a member in the previous year. According to the Arab Barometer, 56.6 percent of men and 43.4 percent of women used wasta during the last five years.

Where patriarchal structures and norms are robust and services distributed via informal networks, males have substantial advantages accumulating homosocial capital, succeeding in elections, and accessing services through clientelistic networks. Quotas of more than 50 percent of seats would be needed to achieve parity — 50 percent of requests provided to females — as long as women as less mobilized electorally and have fewer resources.

The results also show that economic and political context shapes the size of the gender gap. Algeria’s military-backed regime and oil economy excludes women from political and economic networks to a greater extent than Morocco’s oil-poor economy and monarchical regime. Quota design also affects the size of mandates: reserved seat quotas, implemented in Morocco, led to a larger impact of quotas on service provision to women than party list quotas, implemented in Algeria.

This study provides compelling evidence across two quota designs — reserved seat and party candidate — that quotas create mandates to serve women and increase gender equity in access to services. The findings are applicable to authoritarian regimes and patriarchal societies worldwide, such as Pakistan, Mali and Yemen, but illustrate that quota design, oil and regime type affect the size of gender gaps and mandate effects.

The study also has implications for the United States and other Western countries, where men still hold most leadership positions, including 95 percent of posts as CEO in S &P 500 companies and more than 80 percent of seats in Congress. Even as social attitudes become more accepting of gender equality, the preponderance of males in positions of power creates structural conditions that perpetuate women’s exclusion. Since, as a starting point, males enjoy more resources — and because men and women tend to develop more effective networks with others of the same gender — without quotas, whether legally enforced or voluntary, the lack of women in leadership will remain a stubborn and difficult problem to change.

Lindsay J. Benstead is an assistant professor of political science at the Mark O. Hatfield School of Government, Portland State University and contributing scholar in the Women’s Rights in the Middle East Program, James A. Baker III Institute for Public Policy, Rice University.