Human Rights

Do rights at home boost rights abroad? Sexual equality and humanitarian foreign policy

Alison Brysk

Department of Political Science and Global & International Studies Program, University of California, Santa Barbara

Aashish Mehta

Global & International Studies Program, University of California, Santa Barbara

Abstract

Does women’s empowerment strengthen global good citizenship? We test theories of democratic foreign policy and feminist international relations that suggest that more deeply democratic countries with greater gender equity will be stronger international human rights promoters. First, the direct empowerment of women as policymakers and civil society constituencies may shift states’ incentives and ability to pursue international human rights initiatives. Second, greater sexual equality may lead to feminist socialization of the wider society to promote human rights values. We test these predictions by measuring the relationship between five different measures of sexual equality and a country’s propensity to support 30 international human rights outcomes, including legal commitments, humanitarian assis- tance, and sanctions, controlling for previously established contributing factors such as level of development and democratic regime type. We find that more sexually equal countries are more likely to support international com- mitments to constrain state violence against individuals, international measures to combat gender and sexual orienta- tion discrimination, and more and higher quality development assistance. However, sexual equality appears to yield less benefit for more costly human rights initiatives: yielding sovereignty to international legal institutions, promot- ing economic rights through concessionary trade policies, or adopting diplomatic sanctions against pariah states. These effects are stronger in democratic states, where citizen empowerment translates more readily into foreign policy, and are also found in a sample that excludes the Western powers.

Keywords

deep democracy, gender equity, international human rights, sexual equality

Introduction

Does women’s empowerment strengthen global good citi- zenship? This study will systematically test whether domes- tic sexual equality strengthens human rights foreign policy, above and beyond previously studied factors such as dem- ocratic regime type. Our findings are consistent with the- ories that predict that women’s empowerment deepens democracy, projects women’s preferences for peace and care globally, socializes makers of foreign policy to combat gender discrimination, and mobilizes transnational civil society to foster a greater supply of global governance in sev- eral key areas.

Why is it important to add women’s empowerment to the supply side of global governance? First, it is critical to understand the factors that lead a small number of ‘global Good Samaritans’ to contribute disproportionately to the world’s humanitarian collective goods. Leading propo- nents of peace, international law, humanitarian aid, and democratization – like Sweden and Costa Rica – are not just democracies, but relatively egalitarian social democra- cies, notable for the empowerment of women (Aggerstam,

Corresponding author: [email protected]

Journal of Peace Research 2014, Vol 51(1) 97–110 ª The Author(s) 2013 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0022343313500805 jpr.sagepub.com

by guest on August 4, 2015jpr.sagepub.comDownloaded from

2008; Brysk, 2009; Manners, 2008). At the same time, several lines of scholarship suggest that improving sexual equality brings a variety of benefits for a nation’s develop- ment and peace, so it makes sense to trace the possible consequences for global governance (Sen, 1999). On the negative side, a recent study shows that sexual inequity and women’s insecurity predicts international aggression (Hudson, Ballif-Spanvill & Caprioli, 2012). Positively, greater equity at home is linked to improved domestic human rights outcomes for all citizens (Melander, 2005), and local women’s consciousness, participation, or empowerment in receiving countries improves the suc- cess of international human rights initiatives (Gizelis, 2009). In a complementary vein, case study research sug- gests that at the global level, empowered women in foreign policy institutions and civil society have played an impor- tant role in promoting particular foreign policies such as international measures against violence against women, and gender-based asylum policies for women fleeing gendered forms of persecution (Alfredsson, 2009; Keck & Sikkink, 1998; Spees, 2008).

In order to extend these findings and map the broader influence of gender relations on foreign policy, we sys- tematically examine the relationship between women’s empowerment and a broad set of human rights foreign policy initiatives, including most outcomes considered in previous studies, and controlling for previously known factors that generally predict international human rights promotion. We test the statistical relationship between domestic sexual equality and a representative package of humanitarian commitments and outputs, including human rights treaties and conventions, participation in international human rights institutions, humanitarian economic assistance and complementary economic flows, votes on sanctions, and specific measures against discrimination against women and by sexual orientation, using five different measures of sexual equality and women’s empowerment. We find that, other things being equal, countries with more empowered women make stronger international commitments to restrain state violence against vulnerable individuals, to humani- tarian aid, and to global measures against sexual discrim- ination. Since there is wide debate about the ultimate humanitarian impact of particular global human rights policies on recipient societies, this study simply tests the influence of gender balance on the generation of forms of human rights policy previously studied, and does not predict that any treaties, aid packages, or sanctions adopted will necessarily produce an improvement in local human rights conditions on the ground. (This debate is discussed below.)

Our analysis of the international impact of a state’s gendered character contributes to both liberal democratic and constructivist approaches to foreign policy which pre- dict that states’ identities and domestic characteristics may influence their international niche and mode of oper- ation (Hopf, 2002; Smith, 2001). Our study also builds on the feminist international relations analysis that gen- dered power relations do not stop at the water’s edge, and that more gender balanced societies may speak ‘in a differ- ent voice’ in the global arena (Gilligan, 1982; Tickner, 2001). Finally, our findings shed light on the question ‘what do women want?’ in foreign policy, and support the prediction that empowered women will seek protection against violence, care for the vulnerable, and equal rights for all (Steans, 2006).

Feminist theory and foreign policy: Why gender matters for global citizenship

The central contention of feminist international relations analysis is that gendered consciousness, roles, and experi- ences structure world politics. We use this insight to extend three overlapping lines of international relations theory that each claim that some forms of domestic identity will influence states’ international behavior: con- structivist socialization of states, the Kantian liberal view of democratic international relations, and domestic politics analyses of foreign policy. Combining the predictions of these theories, women’s empowerment should change conventional masculinist socialization in state foreign policy (constructivist feminism), increase women’s civic participation to build transnational colla- boration and advocacy (transnational feminism), and foster foreign policymakers’ ability and incentives to enact women’s preferences as leaders and interest groups in deep democracies (liberal feminism). Feminist values are said to include cooperation, care, justice, and equity – and we will critically examine arguments that predict that states with more empowered women will pursue these values: from the gendered experience of insecurity to a hypothetical ‘ethic of care’.

Constructivist feminist approaches to international relations hold that visions of national interest, the ambit of foreign policy, and the nature of the global system are constructed by gender relations (Enloe, 2000; Pettman, 1996). Constructivists contend that foreign policy is a constructed response to perceptions of national interest, filtered through states’ roles and identities – not an inev- itable response to objective conditions as realists argue. Specifically, constructivist approaches stress that ‘foreign policy is what states make of it’ (Smith, 2001); states may

98 journal of PEACE RESEARCH 51(1)

by guest on August 4, 2015jpr.sagepub.comDownloaded from

choose to become global citizens as an alternative pathway to power and influence, animated by an alternative concep- tion of national interest based on identity. Moreover, global discourses on women’s human rights help to constitute the identities of states that position themselves as middle pow- ers, ‘moral powers’, or leaders of regions – promotion of women’s human rights becomes a ‘mark of civilization’ in international society (Towns, 2010). The mechanism of constructivist feminist socialization works in part by enlarging the issue agenda and stakeholder constituency across gender lines and coalition-building on global com- mons concerns like public health (Baer & Brysk, 2009; Weldon, 2006). Feminist agendas and interdependence arguments are picked up by both male and female national and global norm entrepreneurs, information elites, and the ‘common sense’ of global governance – such as the link between women’s education, population growth, and development (King & Mason, 2001; Sen, 1999).

On the transnational dimension of feminist interna- tional relations theory, a systemic transformation of femin- ist consciousness and collaborative modes of global politics should foster greater attention, resources, and receptivity to global problems that disproportionately affect women (Peterson & Runyan, 2010). Women’s feminist, abolition- ist, or pacifist movements have a long historical association with transnationalism, dating from at least the 19th cen- tury. In 1938, Virginia Woolf famously asserted, ‘As a woman, I have no country. As a woman, I want no country. As a woman my country is the world’ (Brown, 2003: 183). Women’s transnational advocacy movements have been influential in shifting global agendas and leading states’ pol- icies on issues such as female genital mutilation and sex traf- ficking that affect women as a class, but not the movements’ own members (Brysk, 2013; Brysk & Choi- Fitzpatrick, 2012). Moreover, systematic evidence from 70 countries over four decades shows that domestic and transnational feminist mobilization in civil society signifi- cantly improves both domestic and international policies on violence against women (Htun & Weldon, 2012).

Moving to extend a feminist perspective to the liberal democratic approach to international relations, domestic politics arguments show that factors such as a state’s regime type, leadership beliefs, class structure, culture, and religion are projected in distinctive ways into the international arena (Beasley et al., 2013; Hartz, 1955; Hopf, 2002). The most well-known application of state type to global citizen- ship is the democratic peace theory, derived from Imma- nuel Kant’s vision of a liberal democratic international community, and bolstered by contemporary findings that mature democracies are less prone to conflict among their ranks (Ray, 2008). Doyle summarizes the mechanisms that

produce the observed liberal democratic peace as an inter- twined combination that links domestic and constructivist dimensions and parallels feminist arguments: institutional representation of citizens, ideological commitment to human rights, and transnational interdependence (Doyle, 2005; Maoz & Russett, 1993). More recently, analysts of regional integration and value promotion by ‘middle pow- ers’ have gone beyond democratic regime type to link more deeply democratic states that are more equitable and parti- cipatory to more globalist foreign policies (Cooper, Higgott & Nossal, 1993; Dunne, 2008; Kaldor, 2007).

How does feminism affect this deep democracy model of human rights foreign policy? Deepening democracy by empowering women may translate to foreign policy by cre- ating interest group pressures by women’s organizations, an attentive constituency for feminist issues, and transnational ties among women’s organizations that affect foreign policy bureaucracies and participate directly in international insti- tutions (Joachim, 2003). Foreign policy is a three-level game shaped by diplomats, executives, parliaments, domestic interest groups, public opinion, transnational advocacy networks, international institutions, and coali- tions (Evans, Jacobson & Putnam, 1993; Smith, Chatfield & Pagnucco, 1997). Changes in the power and participa- tion of women in society could operate to shift foreign pol- icy through all of these channels: with increasing sexual equality, women become foreign policy decisionmakers, lobby foreign policymakers as groups and via public opin- ion, and socialize foreign policymakers. Deeper democratic participation by women’s groups does seem to have played a key role in fostering international organizations’ attention to a range of women’s rights issues, including the designa- tion of rape as a war crime (Spees, 2008; Stienstra, 1994). Foreign policymakers in societies with empowered women are socialized to attend to ‘women’s issues’; for example, in a debate on the introduction of gender-based asylum in Canada, an initially unresponsive immigration minister was forced to defend himself – and eventually yield to women’s groups’ pressure – by reference to the norm, ‘let me assure you that the Minister does not condone discrimination against, or persecution of, women’ (Alfreds- son, 2009: 191–192). Conversely, Lynn Savery (2007) shows that less deeply democratic and more patriarchal electoral democracies are less likely to adopt international women’s rights norms.

Women’s empowerment at home may also have an impact on foreign policy via the inclusion of more women as foreign policymakers in diplomatic and representative institutions. Liberal feminists contend that giving women voice, vote, and representative leadership roles will change the global system (Tickner, 2001). In this ‘women in

Brysk & Mehta 99

by guest on August 4, 2015jpr.sagepub.comDownloaded from

politics’ argument, women as foreign policymakers are in a position to allocate resources to international issues that disproportionately affect women – and female executives and parliamentarians often state that they do so. Demo- cratic women’s leadership may also circulate to the inter- national level, and women foreign policymakers may be socialized by domestic women’s rights experience. Costa Rican jurist Sonia Picado, who became the first female judge on the Inter-American Human Rights Court, stated, ‘My experience with international human rights started with the fight for women’s rights here at home’ (Interview, 5 June 2003). Although isolated elevation of female elites to national leadership is not necessarily pro- gressive, feminist theory suggests that more systematic incorporation of a formerly excluded half of society broadly throughout decisionmaking institutions should improve the human rights agendas and grassroots access of a wide range of bodies relevant to foreign policy (Meyer & Prugl, 1999). For example, Japan’s Sadako Ogata, the first female head of the UN High Commission on Refu- gees, returned home to head the Japanese International Cooperation Agency, where she markedly increased the grassroots and humanitarian focus of aid.

If empowering women does make a difference, what foreign policy goals are more gender-balanced societies more likely to pursue – and why? Perhaps because the state is gendered and constructed by war, historically the empow- erment of women is associated with demilitarization of the state (Kaufman & Williams, 2007; Steans, 2006) and women participate more in peace movements and conflict resolution within the state (Goldstein, 2001). In democra- cies, there are observable differences in women’s and femin- ists’ individual attitudes, political mobilization, and voting behavior regarding peace, social welfare, and discrimination that may translate to the global level. In most OECD coun- tries, there is a consistent ‘gender gap’ in which women are less supportive of the use of military force and more suppor- tive of humanitarian policies (Conover & Sapiro, 1993; Togeby, 1994). As individuals, women worldwide are more tolerant of all forms of difference (Norris & Inglehart, 2003), and in most countries, women support gay rights and marriage equality more strongly. In many cases, these differences by gender identity are cross-cut or even oversha- dowed by gender ideology. Thus, in a situation of acute nationalist conflict in the Middle East, women are not more peace-promoting than men – but across the ‘most different cases’ of Israel, Egypt, Palestine, and Kuwait, more feminist men and women alike are more supportive of Middle East peace (Tessler & Warriner, 1997).

These preferences for peace, care, and equity may be explained by gendered experiences, gendered thinking,

or gender roles, and our study will attempt to contribute to assessing these overlapping explanations. Gendered experiences of vulnerability to violence, poverty, and dis- crimination give many women an interest and sensitivity to these issues – even if not directly affected, all women are vulnerable to rape and domestic violence, the economic and health burdens of childbearing and child-rearing, and social persecution for private behavior that challenges ascribed roles (Kaufman & Williams, 2007). If the pre- ference for peace is because women are the first and worst victims of state violence and state-sponsored patriarchal societies (Hudson, Ballif-Spanvill & Caprioli, 2012), we would expect empowered women to support limitations on state violence and antidiscri- mination measures.

Alternatively, some feminist theory argues that women’s distinctive political values come from gendered thinking. These feminists contend that women engage in more com- passionate moral reasoning in the public sphere (Cho- dorow, 1978; Gilligan, 1982), and thus follow an ‘ethic of care’ in public policy (Held, 2006), including foreign policy (Robinson, 2011).

Feminist theories also predict that gender roles and identities such as motherhood may influence foreign pol- icy goals. Since women are the primary parents in most societies, many women will be especially sensitive to issues that involve the nurturing and protection of chil- dren (Ruddick, 1989). Women mobilize as mothers’ movements to advocate for human rights and receive preferential response: from the Nobel Laureate Liberian women’s peace movement that ended that country’s decades-long civil war, to the human rights campaigns of Mothers of the Disappeared in Latin America, to food security campaigns and economic protests banging empty pots and pans worldwide (Peters & Wolper, 1995; Turpin & Lorentzen, 1996). If this gender role mobilization translates into foreign policy, we should see differentially greater support by countries with greater sexual equality for children’s rights and self-designated maternalist global commons issues.

In sum, if states can choose to be good global citizens, and citizens in more deeply democratic states have more latitude to influence these choices, improving sexual equality should move foreign policy in a more globalist direction. If women are more vulnerable to violence, poverty, and discrimination, then where women have more influence, they should encourage their societies to be better global citizens as promoters of peace, care, and tolerance. Women’s empowerment should make a difference in foreign policy through a combination of increasing female participation in government, greater

100 journal of PEACE RESEARCH 51(1)

by guest on August 4, 2015jpr.sagepub.comDownloaded from

leverage of the gender gap in public opinion on govern- ment, mobilization of women’s movements in civil soci- ety, and diffusion of feminist norms across national and international institutions. Together, these arguments imply that greater domestic sexual equality should lead to greater support for human rights treaties, international legal institutions, international sanctions against human rights violators, humanitarian aid, projects that improve the welfare of women and children, and measures to combat discrimination.

Putting sexual equality to the test: Methods, data, and testable hypotheses

We examine countries’ commitments or contributions to 30 international human rights initiatives as of 2009–11, regressing each of them separately on each of five differ- ent measures of sexual equality as recorded in 2007–08. We control for countries’ basic economic characteristics, and the averages over appropriate time frames of three key sociopolitical indices (described below). With the exception of a few outcomes with small sample sizes, we also control for an interaction of the sexual equality measure with a democracy index in order to see whether sexual equality had different effects on rights promotion in autocracies and democracies. The precise form of the regression varies, depending upon the characteristics of the dependent variable: we use ordinary least squares when the dependent variable is continuous, probit when it is binary, ordered probit when it is ordered categorical (e.g. vote against, abstain, vote in favor), and Tobit when it is censored (e.g. contributions to the UN population fund, which have a lower bound of zero). Online supple- ment A summarizes the dataset and sources. We have 195 countries1 in the dataset, although missing data for some variables reduce the effective sample sizes as indi- cated in tables throughout.

Our outcome variables belong to the following five sets of measures of global citizenry. First, we considered whether countries have ratified ten different human rights treaties (conventions and protocols). In cases of near-universal ratification like the ICCPR, we chose the optional protocols to these treaties to test state

commitment – in this case, Optional Protocol 1 permits individual access to the treaty monitoring body, and Optional Protocol 2 commits the country to abolishing the death penalty. Second, we looked at whether they signed and ratified the Rome Statute on the Interna- tional Criminal Court (ICC), and whether they have bilateral immunity agreements (BIAs) shielding United States citizens from the court.2 Third, we examined their votes on UN resolutions on pariah states, extrajudicial executions, and the death penalty, including some selected because they involve a commitment to freedom from discrimination based on sexual orientation. Fourth, we examined states’ commitment to development, as proxied by both the quantum of aid they provide and the five most human rights-related components of the Com- mitment to Development Index. This last set of variables is measured only for a small group of developed countries, and so does not include a full set of controls or an inter- action between sexual equality and democracy. Finally, we examined funding commitments to the UN Population Fund, the international agency most exclusively dedicated to women’s reproductive rights.

Online supplement B examines the relationship between our outcome measures, with a view to establish- ing whether they do indeed capture some national proclivity for global citizenship. Unrelated outcomes would cast doubt on the argument that national charac- teristics such as sexual equality promote participation in rights initiatives. Our analysis shows that the levels of a country’s support for different initiatives are indeed related to each other, but also that these levels of support vary across outcomes and countries. Thus, we argue, national character seems to matter for human rights pol- icy, but our outcomes test it in distinctive ways.

We have experimented with seven different measures of sexual equality, but only discuss five in the article. The Gender Equity Index (GEI) combines data on female– male ratios of several variables along three dimensions: economic activity (labor-force participation and estimated income), empowerment (the percentages of women in tech- nical positions; management and government; parliament; and ministerial level positions), and education (literacy, plus primary, secondary, and tertiary enrollment). The Gender Inequality Index (GII) combines information on maternal mortality, adolescent fertility, secondary-school comple- tion, parliamentary representation, and labor-force1 We use present-day UN country definitions. When constructing

historical variables for countries that were reunited during the period that these scores were recorded (Germany and Yemen), we attribute to the combined country the average values of these variables for their constituent countries during the years they were divided. For those that separated (the former Soviet and Yugoslav republics), we use scores for the combined nation in the pre-separation years.

2 The Bush administration sought BIAs with other countries as a means of ensuring that US citizens were not handed over to the ICC without US government permission.

Brysk & Mehta 101

by guest on August 4, 2015jpr.sagepub.comDownloaded from

outcomes. It therefore includes the same dimensions as the GEI, but adds health outcomes to the mix. In keeping with a human development approach, gender-equitable health outcomes are likely to be strong signals of deep democracy because they reflect a democratization of the capabilities and the social status of women and girls. In consideration of more strategic ‘women in politics’ arguments, we also considered a direct measure of women’s political participa- tion in the most internationally influential representative institution: parliament (WIP is Women In Parliament).3

Motivated by Hudson, Ballif-Spanvill & Caprioli’s (2012) critique that the widely used measures discussed above do not incorporate information on the legal rights and protections of women, we also utilize two measures drawn from the database provided by WomenStats.org. The Physical Security of Women (PSOW) scale assigns countries to a four-point scale (1 is best, 4 worst) based on the security of women (in law, practice, custom, and statistics) against domestic violence, rape, marital rape, and murder. The Inequity in Family Law (IFL) scale translates information on the law and practice regarding age of marriage, polygyny, consent in marriage, abortion, divorce, marital rape, and inheritance to a five-point scale (0 is best, 4 worst). Being categorical, these two measures present some practical complications. Regression coeffi- cients are difficult to identify when all countries in one sexual equality category have the same discrete human rights outcome. Moreover, while interactions between dummies capturing these sexual equality measures and the polity index could certainly be included in the regres- sions, this would result in an unwieldy profusion of marginal effects, given the number of outcome variables analyzed. We therefore report on the basic results using these measures in the text below, exclude interactions between the polity and sexual equality variables, and confine the results table and detailed discussion to online supplement E.

Our control variables were chosen based upon prior research on determinants of human rights foreign policy and behavior (Apodaca & Stohl, 1999; Poe, Tate & Keith, 1999). Per capita GDP (corrected for purchasing power parity) and the Human Development Index (HDI) are indicators of economic development, which could per- mit countries to train their sights on international

concerns or alter their incentives to do so.4 In order to dis- tinguish wealth from true development and control for the presence of potentially exploitative economic structures, we also introduce a measure of economic dependence on resource rents, averaged over 1995–2007 to smooth out price volatility. Trade dependence (exports plus imports as a share of GDP) is intended to capture, inter alia, the commercial importance of international relations. A priori, this might induce countries to more actively sup- port human rights initiatives, if international conformity increases access to foreign markets, or conversely to become reticent to support international initiatives that could strain relations with trading partners. Social connec- tivity (KOF, also averaged over 1995–2007) serves as a proxy for how well-informed about international affairs and connected to global information flows a country’s cit- izens are likely to be, and as an indicator of social moder- nization alongside economic development.

The Polity IV measure captures prima facie democ- racy, while the political terror scale (based on Amnesty International reports) captures the effects of internal on external human rights outcomes. These corrections are central to our core contention that sexual equality influ- ences international support for a progressive agenda, even holding constant a country’s current political sys- tem. The timing of these political terror and democracy measurements could matter, especially because some of our outcome variables (e.g. the War Crimes Convention) have been in existence for decades. For these older out- comes, we control for a wide range of possible effects using the average value of these measures between the date the treaty was opened and 2008.5 When examining more recent human rights outcomes we correct for the average historical political terror and polity scores going back as far as they are available (1975 for the terror scale, 1948 for the polity measure), but also correct for the average values of these variables between 2002 and 2008 to allow more recent history to have different effects. Sexual equality is interacted with this more recent average polity measure to allow the influence of sexual equality to vary with recent democratic conditions. Our hypothesis regarding the effect of democracy was that oppressive regimes would be less inclined to raise stan- dards or expectations of human rights at the international

3 The article excludes our sixth and seventh measures of sexual equality: the Gender Empowerment Measure (GEM) and the population sex ratio. The sex ratio is seldom statistically significant, and both the coverage of and the results from using GEM are qualitatively similar to the GEI. The GEM also has a smaller sample size.

4 Our tables provide results controlling for per capita GDP. Replacing per capita GDP with the HDI does not significantly alter our conclusions. 5 For treaties that were opened for ratification prior to 1976, when countries were first measured on the political terror scale, we use the average value of the political terror score from 1976 to 2008.

102 journal of PEACE RESEARCH 51(1)

by guest on August 4, 2015jpr.sagepub.comDownloaded from

level, although of course the converse situation is possible, if oppressive regimes support international human rights to deflect attention from their own record. To ensure that regression coefficients are readily interpretable, we scale the sexual equality and polity measures so that they have mean 0 and standard deviation 1.6

This wide variety of dependent variables and sexual equality measures permits us to test a number of hypoth- eses. For convenience, let b refer to the coefficient on sexual equality and c to the coefficient on the interaction between sexual equality and polity.

H1: Other things equal, more sexually equal countries are more supportive of humanitarian foreign policy.

For the average, anocratic country, this involves a test of the size and significance of b. For an otherwise average country that is one standard deviation more (less) demo- cratic than the mean, the test involves bþc (b–c). The mag- nitudes of the effects of sexual equality in the probit models are captured by its marginal effects, conditional on the scaled polity variable taking on the values –1, 0, and þ1, and on all other variables being at their mean values. In the ordered probit we similarly present the marginal effects for the outcome most supportive of international human rights (e.g. in a ‘vote against, abstain, vote in favor’ type variable, we present the marginal effect of sexual equality on the probability that a country votes in favor).

H2: Sexual equality is more closely associated with support for international human rights in demo- cratic countries.

This is simply a test of the significance of c. How strongly democracy conditions the effectiveness of sexual equality can be discerned by comparing the magnitudes of the marginal effects for autocratic countries with those of democratic countries when c is statistically significant.7

H3: Sexual equality is more closely associated with international human rights commitments when

women occupy positions with the authority to make such commitments.

This will be rejected if the marginal effects of the WIP measure are statistically insignificant.

H4: Sexual equality is associated with outcomes that have no distinctive value for women as an overall interest group, suggesting a socialization effect.

This hypothesis will be rejected if sexual equality is insignif- icant for outcomes that have no distinctive value for women as an overall interest group. These outcomes include opposition to the death penalty and LGBT dis- crimination, both of which affect a very small percentage of women.

In keeping with our objective of shedding light on what types of international outcomes may be affected by domestic sexual equality, we split our outcome variables into three groups according to three areas of hypothesized preferences women may bring to the table: freedom from violence, care, and tolerance. We also note that some out- comes – sanctions against pariah states, pro-development trade policies, and support for the International Criminal Court – are overtly costly, because they may involve loss of aid or trade. These distinctions permit us to investigate two sets of follow-up questions:

Follow-up A: Does sexual equality influence interna- tional human rights outcomes involving (i) the right to freedom from state violence; (ii) an ethic of care; or (iii) discrimination?

Follow-up B: Do the effects of sexual equality extend to costly international outcomes?

We draw inferences conservatively. We utilize robust standard errors everywhere, and also run the regressions with incrementally larger sets of controls to examine the fragility of our results to the inclusion and exclusion of con- trols. We only report results from the model with the most complete set of controls. Coefficients on independent variables, if significant with full controls, were generally of the same sign when controls were excluded. Finally, the use of indices that combine sexual equality measured along several dimensions could obscure effects of sexual equality if only some included dimensions matter.

With respect to interpretation, we claim only to have established a relationship between sexual equality and global human rights policy, conditional on a rich set of control variables. While the best analyses of international commitments (e.g. Simmons, 2010) take into account information on the timing of commitments and allow

6 This scaling was performed using means and standard deviations calculated from the subset of countries reporting our three main continuous sexual equality indices (the GEI, GII, and WIP measures, defined below). 7 Note that c captures the effects of small changes in the polity score on the marginal effect of sexual equality. In practice, in a probit model there will be circumstances when c > 0 but large increases in sexual equality could nevertheless reduce the marginal effect of sexual equality. It is this non-linearity that motivated us to present the mar- ginal effects of sexual equality for autocratic, anocratic, and democratic countries separately (Ai & Norton, 2003).

Brysk & Mehta 103

by guest on August 4, 2015jpr.sagepub.comDownloaded from

for country fixed effects, gender equity measures have only become available from a large number of countries in recent years, precluding this type of analysis. Similarly, there appear to be no good instruments that could be used to cleanse the measured relationship of biases aris- ing from reverse causality.

We select our sample and dependent variables delib- erately, in an effort to distinguish the often- overlapping but analytically distinct phenomena of development, sexual equality, and global human rights participation. We include many variables, such as treaty ratification, which are available to less modernized states that are not aid providers, and we have run our analysis on the full sample of countries and on a subset of this dataset that excludes countries belonging to the Western Europe and Others Group.

Finally, we show only that sexual equality promotes humanitarian initiatives, without presuming that these initiatives always achieve their objectives. In consider- ation of the debate on the impact of treaties, we follow Beth Simmons’s (2010) findings that treaties are often effective, with the presence of sufficient enabling domes- tic factors that overlap with those we are considering, such as civil society mobilization. We do not take a posi- tion on whether aid is helpful for human development, given significant evidence that motivations for aid- giving are not entirely altruistic (Bueno de Mesquita & Smith, 2009; Younas, 2008) and frequent mismanage- ment of aid funds (Easterly & Pfutze, 2008). Rather we simply ask whether more sexually equal countries are apt to give more aid, and whether measures of aid quality designed to capture a more serious commitment to development are related to sexual equality. Our sanctions measure is UN resolutions, which are purely diplomatic and thus transcend the debate on possible humanitarian harm of economic sanctions.

Analysis and results

Tables I–III present the main regression results using the GEI and WIP measures, clustered around our three areas of human rights policy. We relegate results using the GII measures, which are broadly similar to those using the GEI, to online supplement C in order to keep the tables readable. However, results from these regressions are included in the tallies of significant effects below. We present marginal effects for probit and ordered probit regressions and regression coefficients for all other regres- sion types. We also calculate derivatives of the predicted outcome with respect to sexual equality for typical coun- tries with a government that is either autocratic (has a

polity score one standard deviation below the mean) or democratic (polity score one standard deviation above the mean). For a country with an average polity score, the marginal effects/coefficients provide the effect of sex- ual equality. The predicted outcome is the dependent variable itself for the Tobit regressions, the probability of a positive human rights outcome for probits, and the most positive human rights outcome for ordered probits.

We now turn to our hypothesis tests, beginning with the combined effects of sexual equality and democracy (H1 and H2), which are generally positive. We have 30 outcome variables, 23 of which provide sufficient observations to support estimation of the interaction between sexual equality and democracy. For each out- come, we have three regressions, each corresponding to a different sexual equality measure. Thus, we have 90 regressions from which we can test H1 and 69 regres- sions from which we can test H2. We find that for an average country, sexual equality enhances human rights in 42 out of 90 regressions. For countries with democ- racy measures one standard deviation above the mean, equality boosts human rights in 25 out of 69 regressions, while for those that are one standard deviation below the mean it does so for 23 out of 69 regressions. Moreover, the effects can be quite large. For example, a one stan- dard deviation increase in the GEI increases the prob- ability of ratifying the second optional protocol to the ICCPR by approximately 14% in a democracy, 23% in an anocracy, and 15% in an autocracy. The results on PSOW and especially on IFL provide further evi- dence that countries with higher sexual equality promote international human rights more assiduously. Thus, H1 is confirmed for democratic, anocratic, and autocratic countries alike.

The interaction between sexual equality and democ- racy is statistically significant in 17 of 69 regressions (H2), and is of the anticipated sign – democracy appears to bolster the effects of sexual equality in 16 of these cases. The derivatives of the outcome variables with respect to sexual equality are larger at a point estimate in democracies than in autocracies in 15 of these 16 cases. Thus, democracy often appears to facilitate the projection of sexual equality onto the world stage (H2).

Turning to the evidence on possible political pathway effects of sexual equality, we find that having women in parliament matters for an anocratic country for 12 out of 30 outcome variables. WIP help with five out of 23 outcomes in democracies, and seven out of 23 in auto- cracies. Thus, the hypothesis (H3) that sexual equality advances international human rights via political mechanisms cannot be rejected.

104 journal of PEACE RESEARCH 51(1)

by guest on August 4, 2015jpr.sagepub.comDownloaded from

T ab

le I.

F re

ed om

fr om

st at

e vi

ol en

ce ag

ai n

st in

di vi

du al

s

D er

iv at

iv e

w .r

.t .

se xu

al eq

ua lit

y at

: M

ar gi

na l

ef fe

ct s

A gr

ee m

en t

(d ep

en de

nt va

ri ab

le )

T yp

e of

re gr

es si

on

Se xu

al eq

ua lit

y m

ea su

re

P ol

it y ¼

–1 St

d. de

v.

P ol

it y ¼

þ 1

st d.

de v.

Se xu

al eq

ua lit

y

P ol

it y–

se xu

al eq

ua lit

y in

te ra

ct io

n G

D P

pe r

ca pi

ta R

es ou

rc e

re nt

s P

op ul

at io

n K

O F

in de

x T

ra de

:G D

P ra

ti o

P ol

it ic

al te

rr or

– cu

rr en

t P

ol it

y– cu

rr en

t

P ol

it ic

al te

rr or

– hi

st or

ic al

P ol

it y–

hi st

or ic

al Sa

m pl

e si

ze

(1 )

G en

oc id

e P

ro bi

t G

E I

0. 07

8* 0.

05 3

0. 06

8 –0

.0 21

–0 .0

05 –0

.0 01

0 0.

01 1*

* –0

.0 05

0. 19

3* *

–0 .0

42 12

7 W

IP 0.

02 3

0. 03

3 0.

02 8

0. 00

6 –0

.0 06

–0 .0

01 0

0. 01

2* *

0. 01

6 0.

21 8*

* 0.

01 7

14 2

(2 )

W ar

C ri

m es

P ro

bi t

G E

I .2

30 **

0. 09

3 0.

20 0*

* –0

.0 87

–0 .0

29 **

0. 00

4 –0

.0 01

0. 01

6* *

–0 .0

24 0.

04 –0

.1 48

12 7

W IP

0. 05

6 –0

.0 80

–0 .0

11 –0

.0 74

–0 .0

27 **

0. 00

3 0

0. 01

5* *

0. 07

4 0.

04 8

0. 00

7 14

2

(3 )

C A

T P

ro bi

t G

E I

0. 09

3 0.

00 5

0. 03

0. 02

8 0.

00 1

–0 .0

02 0

0. 00

1 –0

.0 53

0. 05

8 0.

05 4

12 7

W IP

–0 .0

22 0.

07 0

0. 03

5 0.

05 –0

.0 01

–0 .0

01 0

0. 00

3 –0

.0 82

0. 04

1 0.

05 4

14 2

(4 )

C A

T P

ro to

co l

P ro

bi t

G E

I 0.

05 0

0. 03

0 –0

.0 27

–0 .0

64 –0

.0 24

** 0.

00 4

–0 .0

01 0.

01 3*

* –0

.1 38

–0 .0

71 0.

18 3*

* 11

8 W

IP –0

.0 32

0. 74

8 0.

01 6

0. 06

3 –0

.0 28

** 0.

00 3

0 0.

01 1*

* –0

.0 56

–0 .0

72 0.

22 6*

* 13

2

(5 )

IC C

P R

P ro

to co

l 1

P ro

bi t

G E

I 0.

00 9

.1 38

* 0.

14 2*

* 0.

13 2

–0 .0

21 **

0. 00

5 –0

.0 04

** 0.

00 2

–0 .2

02 0.

02 4

0. 28

9* *

12 7

W IP

.1 48

** .1

33 **

0. 16

0* *

0. 22

0* *

–0 .0

22 **

0. 00

7 –0

.0 04

** 0.

00 3

–0 .2

13 0.

00 7

0. 39

5* *

14 2

(6 )

IC C

P R

P ro

to co

l 2

P ro

bi t

G E

I .1

67 **

.1 39

* 0.

22 5*

* –0

.0 37

–0 .0

08 –0

.0 05

–0 .0

02 0.

01 3*

* –0

.0 7

0. 03

9 0.

10 8

12 6

W IP

.1 33

** .1

60 *

0. 20

8* *

0. 00

2 –0

.0 19

** 0.

00 3

–0 .0

01 0.

01 8*

* –0

.0 41

0. 05

2 0.

23 9*

* 14

1

(7 )

IC C

O rd

er ed

P ro

bi t

G E

I –0

.0 46

.1 86

** 0.

53 3*

* 0.

75 3*

* –0

.0 03

0. 00

6 –0

.0 04

–0 .0

30 **

–0 .0

61 –0

.0 91

1. 41

6* *

–0 .7

53 –0

.3 49

11 8

W IP

–0 .0

24 0.

01 0

0. 28

7 0.

36 1*

* –0

.0 07

0 –0

.0 04

–0 .0

07 –0

.3 01

0. 12

6 1.

05 2*

* –0

.6 87

** –0

.1 83

13 2

(8 )

B IA

P ro

bi t

G E

I 0.

08 1

–0 .0

09 0.

07 5

–0 .0

89 0.

00 5

0 0

0. 01

7 –0

.3 26

–0 .2

99 **

0. 25

6 0.

17 7

–0 .1

01 10

0 W

IP 0.

02 8

0. 08

1 0.

09 8

0. 04

3 –0

.0 01

–0 .0

01 0

0. 01

8* *

–0 .4

14 –0

.3 12

** 0.

27 5

0. 17

8 –0

.0 53

11 2

(9 )

IC C

P lu

s O

rd er

ed P

ro bi

t G

E I

–0 .0

24 .1

97 **

0. 30

7 0.

46 7*

* 0.

03 6

–0 .0

03 –0

.0 02

–0 .0

21 0.

01 5

–0 .4

46 1.

00 9*

* –0

.1 13

–0 .0

67 10

0

W IP

–0 .0

47 .1

84 **

0. 22

7 0.

46 2*

* 0.

01 1

–0 .0

07 –0

.0 03

0. 00

2 –0

.4 54

–0 .2

93 0.

92 3*

* –0

.1 54

–0 .0

57 11

2

(1 0)

M ya

n m

ar O

rd er

ed P

ro bi

t G

E I

–0 .0

19 .1

38 *

0. 22

3 0.

3 0.

00 4

–0 .0

04 –0

.0 07

** 0.

01 8

–1 .1

44 **

0. 24

–0 .5

17 0.

74 6*

* –0

.0 14

11 7

W IP

0. 01

6 0.

03 8

0. 09

8 0.

03 4

0. 01

1 –0

.0 02

–0 .0

06 0.

01 4

–0 .4

56 0.

20 9

–0 .5

34 **

0. 72

7* *

–0 .0

71 12

9

(1 1)

N .

K or

ea O

rd er

ed P

ro bi

t G

E I

–0 .0

71 .1

54 *

0. 20

7 0.

42 8*

* 0.

00 5

–0 .0

20 **

–0 .0

02 0.

00 8

–0 .5

43 –0

.2 02

–0 .2

79 0.

65 9*

* –0

.2 35

11 7

W IP

–0 .0

04 0.

07 1

0. 13

8 0.

15 2

0. 00

7 –0

.0 15

–0 .0

01 0.

00 9

–0 .2

33 –0

.1 31

–0 .3

87 0.

66 5*

* –0

.2 33

12 9

(1 2)

D ea

th P

en al

ty O

rd er

ed P

ro bi

t G

E I

0. 16

7* *

0. 16

9* *

0. 62

3* *

0. 11

7 –0

.0 43

** 0.

01 6

–0 .0

13 **

0. 02

7 –0

.5 73

0. 11

1 0.

08 2

0. 49

1* *

–0 .1

09 11

7

W IP

0. 11

3* 0.

06 4

0. 30

6* *

–0 .0

3 –0

.0 34

0. 01

1 –0

.0 10

** 0.

03 0*

* –0

.2 39

0. 11

1 0.

03 0.

48 9*

* –0

.0 21

12 9

** p

< .0

1, *

p <

.0 5.

M ar

gi n

al ef

fe ct

s ar

e ca

lc u

la te

d at

th e

m ea

n le

ve lo

f al

lo th

er va

ri ab

le s

ap pe

ar in

g in

th e

re gr

es si

on .D

er iv

at iv

es w

it h

re sp

ec t

to se

xu al

eq u

al it

y ar

e ca

lc u

la te

d po

st -e

st im

at io

n fo

r sp

ec if

ie d

po lit

y le

ve ls

w it

h al

l ot

h er

va ri

ab le

s h

el d

fi xe

d at

re gr

es si

on -s

am pl

e m

ea n

s. F

or or

de re

d pr

ob it

re gr

es si

on s,

th e

de ri

va ti

ve s

sh ow

n co

rr es

po n

d to

th e

ch an

ge in

th e

pr ob

ab ili

ty of

th e

m os

t h

u m

an -r

ig h

ts su

pp or

ti ve

ou tc

om e.

105

by guest on August 4, 2015jpr.sagepub.comDownloaded from

T ab

le II

. E

th ic

of ca

re

D er

va ti

ve w

.r .t

. se

xu al

eq ua

lit y

at :

M ar

gi na

l ef

fe ct

s

A gr

ee m

en t

(d ep

en de

nt va

ri ab

le )

T yp

e of

re gr

es si

on

Se xu

al eq

ua lit

y m

ea su

re P

ol it

y ¼

–1 st

d. de

v.

P ol

it y ¼

þ 1

st d.

de v.

Se xu

al eq

ua lit

y

P ol

it y–

se xu

al eq

ua lit

y in

te ra

ct io

n G

D P

pe r

ca pi

ta R

es ou

rc e

re nt

s P

op ul

at io

n K

O F

in de

x

T ra

de :

G D

P ra

ti o

P ol

it y

cu rr

en t

P ol

it ic

al te

rr or

– hi

st or

ic al

P ol

it y–

hi st

or ic

al Sa

m pl

e si

ze

O D

A to

G N

I R

at io

O L

S G

E I

0. 17

1* *

0. 01

5* *

0. 00

6 0.

39 7

27 W

IP 0.

15 2*

* 0.

01 2*

* 0.

00 9

–0 .1

31 27

C om

m it

m en

t to

D ev

el op

m en

t In

de x

– O

ve ra

ll O

L S

G E

I 1.

14 5*

* 0.

00 8

0 0.

99 5

22 W

IP 0.

70 1*

* 0.

02 5

0. 01

1 2.

88 5*

* 22

– A

id O

L S

G E

I 3.

25 5*

* 0.

22 8

–0 .0

76 –4

.8 67

22 W

IP 2.

43 9*

* 0.

27 7*

* –0

.0 75

0. 09

5 22

– T

ra de

O L

S G

E I

1. 55

9* *

–0 .0

85 –0

.2 15

1. 28

1 22

W IP

0. 5

–0 .0

62 –0

.1 67

4. 27

2 22

– In

ve st

m en

t O

L S

G E

I 0.

50 1

–0 .0

53 0.

15 3

–2 .8

62 22

W IP

0. 28

7 –0

.0 45

0. 16

–2 .0

17 22

– M

ig ra

ti on

O L

S G

E I

1. 15

5 0.

06 8

–0 .0

19 5.

87 7

22 W

IP 0.

75 6

0. 08

5 –0

.0 11

7. 73

9 22

– E

n vi

ro n

m en

t O

L S

G E

I 0.

88 6

–0 .0

13 –0

.1 38

0. 75

7 22

W IP

0. 68

8* *

0 –0

.1 4

2. 08

5 22

C R

C P

ro to

co l

1 P

ro bi

t G

E I

0. 09

5 0.

17 5*

0. 12

3* *

0. 04

1 –0

.0 01

0. 00

3 0

–0 .0

03 –0

.0 45

–0 .0

39 0.

00 7

12 2

W IP

0. 05

4 0.

12 0.

08 3*

* 0.

03 3

–0 .0

01 0

0 0.

00 2

–0 .0

77 –0

.0 32

0. 00

9 13

6

C R

C P

ro to

co l

2 P

ro bi

t G

E I

0. 02

4 0.

06 2

0. 04

0. 01

6 0

–0 .0

02 0

0. 00

1 –0

.0 99

–0 .0

11 –0

.0 1

12 2

W IP

–0 .0

04 0.

07 1

0. 02

8 0.

03 2

0 –0

.0 04

0 0.

00 2

–0 .1

64 **

–0 .0

37 –0

.0 29

13 6

D is

ab ili

ti es

P ro

bi t

G E

I 0.

00 6

0. 15

4 0.

08 7

0. 08

1 –0

.0 04

–0 .0

02 0

0. 00

1 –0

.2 37

–0 .0

32 0.

00 8

11 7

W IP

–0 .0

49 0.

09 4

0. 02

7 0.

08 1

–0 .0

05 –0

.0 07

0 0.

00 4

–0 .2

61 **

–0 .0

59 –0

.0 08

12 9

** p

< .0

1, *

p <

.0 5.

M ar

gi n

al ef

fe ct

s ar

e sh

ow n

fo r

pr ob

it re

gr es

si on

s, re

gr es

si on

co ef

fi ci

en ts

fo r

O L

S re

gr es

si on

s. M

ar gi

n al

ef fe

ct s

ar e

ca lc

u la

te d

at th

e m

ea n

le ve

l of

al l

ot h

er va

ri ab

le s

ap pe

ar in

g in

th e

re gr

es si

on .

D er

iv at

iv es

w it

h re

sp ec

t to

se xu

al eq

u al

it y

ar e

ca lc

u la

te d

po st

-e st

im at

io n

fo r

sp ec

if ie

d po

lit y

le ve

ls w

it h

al l

ot h

er va

ri ab

le s

h el

d fi

xe d

at re

gr es

si on

-s am

pl e

m ea

n s.

106

by guest on August 4, 2015jpr.sagepub.comDownloaded from

T ab

le II

I. G

en de

r eq

u it

y an

d to

le ra

n ce

D er

iv at

iv e

w .r

.t .

se xu

al eq

ua lit

y at

: C

oe ff

ic ie

nt s/M

ar gi

na l

E ff

ec ts

A gr

ee m

en t

(d ep

en de

nt va

ri ab

le )

T yp

e of

re gr

es si

on

Se xu

al eq

ua lit

y m

ea su

re

P ol

it y ¼

–1 st

d. de

v.

P ol

it y ¼

þ 1

st d.

de v.

Se xu

al eq

ua lit

y

P ol

it y–

se xu

al eq

ua lit

y in

te ra

ct io

n

G D

P pe

r ca

pi ta

R es

ou rc

e in

te ns

it y

P op

ul at

io n

K O

F in

de x

T ra

de :

G D

P ra

ti o

P ol

it ic

al te

rr or

– cu

rr en

t P

ol it

y– cu

rr en

t P

ol it

ic al

te rr

or –

hi st

or ic

al P

ol it

y– hi

st or

ic al

Sa m

pl e

si ze

(1 )

C E

D A

W P

ro to

co l

1 P

ro bi

t G

E I

0. 22

0* *

0. 03

0. 16

7* *

–0 .1

29 –0

.0 13

0. 00

6 –0

.0 01

0. 00

9 –0

.3 16

–0 .0

13 0.

10 3

12 2

W IP

0. 14

8* *

0. 00

3 0.

09 6

–0 .0

91 –0

.0 1

0. 01

0* *

–0 .0

01 0.

01 0*

* –0

.3 13

** –0

.0 04

0. 21

6* *

13 6

(2 )

Ir an

O rd

er ed

P ro

bi t

G E

I –0

.0 04

0. 23

9* *

0. 52

5* *

0. 55

0* *

–0 .0

1 –0

.0 07

–0 .0

05 0.

00 7

–0 .7

02 0.

07 4

–0 .3

61 0.

76 9*

* 0.

03 3

11 7

W IP

0. 04

2 0.

15 7*

* 0.

34 5*

* 0.

17 1

–0 .0

04 –0

.0 01

–0 .0

01 0.

01 0.

02 6

0. 06

2 –0

.3 33

0. 69

9* *

0. 12

7 12

9

(3 )

U N

P F

C on

tr ib

u ti

on s/

G D

P T

ob it

G E

I –0

.0 52

2. 10

2* *

1. 03

4* *

1. 08

6* *

0. 04

4 –0

.0 11

0 –0

.0 24

–0 .0

17 –0

.1 54

0. 59

8 11

7 W

IP 0.

04 5

2. 61

0* *

1. 32

5* *

1. 28

0* *

0. 02

3 –0

.0 14

0. 00

1 –0

.0 16

0. 39

–0 .2

29 0.

65 6*

* 12

9

(4 )

IC C

P R

–S ex

u al

or ie

n ta

ti on

O rd

er ed

P ro

bi t

G E

I 0.

11 0*

0. 17

1* *

–0 .5

90 **

–0 .1

47 –0

.0 15

0. 01

2 0.

00 2

–0 .0

12 –0

.1 21

0. 02

8 0.

00 8

–0 .5

20 **

–0 .1

24 11

7 W

IP 0.

01 8

–0 .0

05 –0

.0 27

0. 04

2 –0

.0 09

–0 .0

04 0.

00 1

–0 .0

15 0.

09 6

0. 05

6 0.

10 8

–0 .4

73 **

–0 .4

4 12

9

(5 )

IC E

SR –S

ex u

al or

ie n

ta ti

on O

rd er

ed P

ro bi

t G

E I

0. 12

1* *

0. 16

4* *

–0 .6

39 **

–0 .1

07 –0

.0 19

0. 00

5 0.

00 4

–0 .0

06 –0

.0 09

0. 20

8 –0

.3 04

–0 .4

85 **

–0 .4

14 11

7 W

IP 0.

06 8*

–0 .0

26 –0

.1 13

0. 20

8 –0

.0 23

–0 .0

05 0.

00 1

–0 .0

09 0.

03 0.

19 8

–0 .2

27 –0

.5 77

** –0

.6 38

** 12

9

(6 )

E xt

ra -j

u di

ci al

ex ec

u ti

on O

rd er

ed P

ro bi

t G

E I

0. 22

6* *

0. 19

0* *

0. 81

1* *

0. 03

3 –0

.0 32

–0 .0

04 –0

.0 08

** 0.

01 9

–0 .8

51 0.

05 7

–0 .3

87 0.

34 8

–0 .3

81 11

7 W

IP 0.

12 3*

0. 09

3 0.

37 8*

* 0.

02 4

–0 .0

36 0.

00 1

0 0.

02 3

–0 .1

09 0.

03 8

–0 .3

18 0.

48 5*

* –0

.0 63

12 9

(7 )

A n

ti L

G B

T O

rd er

ed P

ro bi

t G

E I

0. 00

3 0.

16 0*

* –0

.4 71

** –0

.4 54

** –0

.0 31

0. 06

6* *

0. 00

8 –0

.0 05

0. 34

9 –0

.2 3

0. 21

4 –0

.7 20

** –0

.3 62

11 7

W IP

0. 02

0. 01

1 –0

.0 71

0. 03

1 –0

.0 21

0. 01

6 0

–0 .0

07 0.

05 1

–0 .1

03 0.

41 2

–0 .5

52 **

–0 .5

73 **

12 9

(8 )

P ro

L G

B T

O rd

er ed

P ro

bi t

G E

I 0.

13 1*

0. 22

5* *

0. 76

9* *

0. 30

2 0.

02 1

–0 .0

13 –0

.0 07

** –0

.0 06

–0 .2

73 –0

.0 33

0. 26

0. 59

5* *

0. 22

5 11

7 W

IP 0.

15 0*

* 0.

03 5

0. 37

1* *

–0 .2

48 0.

02 0.

00 6

–0 .0

01 0.

00 4

–0 .1

07 0.

00 3

0. 05

2 0.

68 6*

* 0.

35 6

12 9

** p

< .0

1, *

p <

.0 5.

M ar

gi n

al ef

fe ct

s ar

e sh

ow n

fo r

pr ob

it re

gr es

si on

s, re

gr es

si on

co ef

fi ci

en ts

fo r

T ob

it .

M ar

gi n

al ef

fe ct

s ar

e ca

lc u

la te

d at

th e

m ea

n le

ve l

of al

l ot

h er

va ri

ab le

s ap

pe ar

in g

in th

e re

gr es

si on

. D

er iv

at iv

es w

it h

re sp

ec t

to se

xu al

eq u

al it

y ar

e ca

lc u

la te

d po

st -e

st im

at io

n fo

r sp

ec if

ie d

po lit

y le

ve ls

w it

h al

l ot

h er

va ri

ab le

s h

el d

fi xe

d at

re gr

es si

on -s

am pl

e m

ea n

s. F

or or

de re

d pr

ob it

re gr

es si

on s,

th e

de ri

- va

ti ve

s sh

ow n

co rr

es po

n d

to th

e ch

an ge

in th

e pr

ob ab

ili ty

of th

e m

os t

h u

m an

-r ig

h ts

su pp

or ti

ve ou

tc om

e.

107

by guest on August 4, 2015jpr.sagepub.comDownloaded from

This said, there is also substantial evidence for the hypothesis that sexual equality is helpful for outcomes with little value that is specific to women as an interest group, suggesting a values socialization effect (H4). Restricting attention to the six outcomes that fit this description: oppo- sition to the death penalty (Table I) and five measures involving gay rights (items 4–8 in Table III), we find that out of the 18 resulting regressions, sexual equality is signif- icant in 13 regressions in autocracies, 10 in anocracies, and 12 in democracies. All five gay rights initiatives received sig- nificantly greater support for countries with high PSOW and low IFL (online supplement E). This suggests that sex- ual equality operates through both political mechanisms and socialization.

Follow-up A asks which types of outcomes are influ- enced by sexual equality. Comparing the numbers of statistically significant (and ‘correctly’ signed) results across the three tables suggests that sexual equality may matter most for tolerance, which is associated more with gendered experience than gendered thinking, but still plays an important role across the board. This impression is strongly confirmed by the results on PSOW (online supplement E). However, we are reluctant to draw conclusions from differ- ences that are this subtle, given that the regressions included in each table are variable in terms of their sample sizes and dependent variable types, and that the number of dependent variables in each group, while large for a study like ours, is rather small for drawing inferences.

Finally, we look at politically costly international com- mitments (Follow-up B). Out of the 21 regressions run on our seven clearly politically costly outcomes (three measures of support for the ICC, pro-development trade policies, and sanctions against Iran, Myanmar, and North Korea), sexual equality only matters for anocracies in four. It matters for only one out of 18 regressions that include an interaction between polity and sexual equality in auto- cracies, but matters for seven in democracies. On the other hand, polity scores are strongly supportive of human rights initiatives in all these cases. This suggests that sexual equality alone is not especially effective in pushing support over the top for more costly and contentious commit- ments, unless it is combined with democracy. Results using the PSOW and IFL indexes similarly show very lit- tle effect of sexual equality on support for these particular commitments.

The control variables generally have the effects we anticipated. Per capita GDP enters most regressions negatively when it is significant, perhaps reflecting a lower willingness by the powerful to cede sovereignty. The exception here is aid (Table II) – richer countries give more and nominally better aid. In addition, more

populous countries are less likely to support international human rights initiatives. Resource intensity plays a negligible role. The KOF social globalization index pro- motes support for freedom from state violence, but not much else. A history of political terror is positively related with support for antidiscrimination measures and those promoting freedom from violence. Trade depen- dence is associated with reduced support for interna- tional initiatives, perhaps because such initiatives carry greater commercial costs for more trade dependent coun- tries. Consistent with previous literature, democracy, in a country of average sexual equality, seems to be helpful for human rights promotion.

Regression results obtained using a subset of countries that excludes the members of the Western European and Others Group are included as an online supplement, part D. The general relationship between sexual equality and support for human rights survives the exclusion of this core group of liberal democratic Western powers, which suggests the effects are not proxies for modernization or Western val- ues. We also note that widespread domestic women’s empowerment began in the 1970s with the second wave feminist movement and post-industrial modernization, but many aspects of humanitarian foreign policy did not expand greatly until the 1990s. The spread of women’s empower- ment and humanitarian policy are therefore unlikely to be reflections of the same phenomenon. Rather, we posit that women’s empowerment appears to be an enabling condition for global good citizenship.

Conclusion

In sum, states with empowered women do make better glo- bal citizens in certain critical areas. More sex-equitable countries are more likely to support international commit- ments against state violence against individuals, even when most of the victims are male. States with more empowered women also provide more and at least nominally higher quality development assistance and more reliably defend children’s rights to not be sexually exploited, demonstrat- ing an ethic of care. In addition, a higher level of sexual equality raises countries’ support for international antidis- crimination measures on both sex and sexual orientation. Democracy appears to be an enabling factor. However, our results suggest that sexual equality does not have much leverage on international positions that are politically costly unless it is accompanied by overt democracy. The pattern of our findings also seems to demonstrate that sexual equal- ity plays a role that goes beyond mere participation of women in politics and suggests that we pay greater atten- tion to socialization.

108 journal of PEACE RESEARCH 51(1)

by guest on August 4, 2015jpr.sagepub.comDownloaded from

What does this mean for human rights promotion pol- icy? Empowerment of women worldwide has the potential to strengthen the level and quality of global human rights policy. Thus, human rights promotion should include tar- geted efforts to improve gender equity in lagging developed democracies, such as Japan, which may improve their glo- bal citizenship. Conversely, foundering global humanitar- ian efforts should reach beyond the customary developed democracies to engage and empower the most gender- friendly emerging nations, such as recently female-headed Brazil, Chile, and Argentina. Finally, we see that strength- ening sexual equality may bolster the protection of all peo- ple who are vulnerable to state violence, extreme poverty, child abuse, and discrimination, as feminist socialization improves states’ efforts to protect all citizens. As the United Nations 1993 Vienna Conference on Human Rights con- cluded, human rights are interdependent and indivisible – and women’s rights are human rights.

Replication data The online appendix and replication data for the empiri- cal analysis in this article can be found at http:// www.prio.no/jpr/datasets.

Acknowledgements We thank Natasha Bennett, Quinn McCreight, and Aisa Villanueva for excellent research assistance, Aditee Maskey, Wayne Sandholtz, Alejandro Anaya, and Rachel Cicchowski for very productive discussions, and four anon- ymous referees for many useful suggestions. All errors are our own.

References Aggerstam, Lisbeth (2008) Ethical power Europe? Interna-

tional Affairs 84(1): 1–11. Ai, Chunrong & Edward C Norton (2003) Interaction terms in

logit and probit models. Economics Letters 80(1): 123–129. Alfredsson, LS (2009) Creating Human Rights: How Nonciti-

zens Made Sex Persecution Matter to the World. Philadelphia, PA: Pennsylvania University Press.

Apodaca, Clair & Michael Stohl (1999) United States human rights policy and foreign assistance. International Studies Quarterly 43(1): 185–198.

Baer, Madeline & Alison Brysk (2009) New rights for private wrongs: Female genital mutilation and global framing dialogues. In: Clifford Bob (ed.) The International Struggle for New Human Rights. Philadelphia, PA: University of Pennsylvania Press, 93–107.

Beasley, Ryan K; Juliet Kaarbo, Jeffrey S Lantis & Michael T Snarr, eds (2013) Foreign Policy in Comparative Perspective:

Domestic and International Influences on State Behavior. Washington, DC: CQ.

Brown, Heloise (2003) ‘The Truest Form of Patriotism’: Pacifist Feminism in Britain, 1870–1902. Manchester: Manchester University Press.

Brysk, Alison (2009) Global Good Samaritans: Human Rights as Foreign Policy. New York: Oxford University Press.

Brysk, Alison (2013) Changing hearts and minds: Sexual pol- itics and human rights. In: Thomas Risse, Stephen C Ropp & Kathryn Sikkink (eds) The Persistent Power of Human Rights: From Commitment to Compliance. Cambridge: Cambridge University Press.

Brysk, Alison & Austin Choi-Fitzpatrick, eds (2012) From Human Trafficking to Human Rights: Reframing Contempo- rary Slavery. Philadelphia, PA: University of Pennsylvania Press.

Bueno de Mesquita, Bruce & Alastair Smith (2009) A polit- ical economy of aid. International Organization 63(2): 309–340.

Chodorow, Nancy (1978) The Reproduction of Mothering: Psychoanalysis and the Sociology of Gender. Berkeley, CA: University of California Press.

Conover, Pamela Johnston & Virginia Sapiro (1993) Gender, feminist consciousness, and war. American Journal of Polit- ical Science 37(4): 1079–1099.

Cooper, Andrew F; Richard A Higgott & Kim Richard Nossal (1993) Relocating Middle Powers: Australia and Canada in a Changing World Order. Vancouver: University of British Columbia Press.

Doyle, Michael W (2005) Three pillars of the liberal peace. American Political Science Review 99(3): 463–466.

Dunne, Timothy (2008) Good citizen Europe. International Affairs 84(1): 1–11.

Easterly, William & Tobias Pfutze (2008) Where does the money go? Best and worst practices in foreign aid. Journal of Economic Perspectives 22(2): 29–52.

Enloe, Cynthia (2000) Bananas, Beaches, and Bases: Making Feminist Sense of International Politics. Berkeley, CA: University of California Press.

Evans, Peter B; Harold Karan Jacobson & Robert D Putnam (1993) Double-Edged Diplomacy: International Bargaining and Domestic Politics. Berkeley, CA: University of California Press.

Gilligan, Carol (1982) In a Different Voice. Cambridge, MA: Harvard University Press.

Gizelis, Theodora-Ismene (2009) Gender empowerment and United Nations peace-building. Journal of Peace Research 46(4): 505–523.

Goldstein, Joshua (2001) Women and War. New York: Cam- bridge University Press.

Hartz, Louis (1955) The Liberal Tradition in America. New York: Harcourt, Brace & World.

Held, Virginia (2006) The Ethics of Care: Personal, Political, and Global. New York: Oxford University Press.

Hopf, Ted (2002) Social Construction of International Politics: Iden- tities and Foreign Policies. Ithaca, NY: Cornell University Press.

Brysk & Mehta 109

by guest on August 4, 2015jpr.sagepub.comDownloaded from

Htun, Mala & S Laurel Weldon (2012) The civic origins of progressive policy change: Combating violence against women in global perspective, 1975–2005. American Polit- ical Science Review 106(3): 548–569.

Hudson, Valerie M; Bonnie Ballif-Spanvill & Mary Caprioli (2012) Sex and World Peace. New York: Columbia University Press.

Joachim, Jutta (2003) Framing issues and seizing opportuni- ties: The UN, NGOs, and women’s rights. International Studies Quarterly 47(2): 247–274.

Kaldor, Mary (2007) Deepening democracy. In: David Held & David Mepham (eds) Progressive Foreign Policy. Cam- bridge: Polity, 36–54.

Kaufman, Joyce P & Kristen P Williams (2007) Women, the State, and War: A Comparative Perspective on Citizenship and Nationalism. Lanham, MD: Rowman & Littlefield.

Keck, Margaret & Kathryn Sikkink (1998) Activists Beyond Borders. Ithaca, NY: Cornell University Press.

King, Elizabeth M & Andrew D Mason (2001) Engendering Development: Through Gender Equality in Rights, Resources, and Voice. Washington, DC: World Bank.

Manners, Ian (2008) Norm power Europe. International Affairs 84(1): 1–11.

Maoz, Zeev & Bruce Russett (1993) Normative and structural causes of democratic peace, 1946–1986. American Political Science Review 87(3): 624–638.

Melander, Erik (2005) Political gender equality and state human rights abuse. Journal of Peace Research 42(2): 149–166.

Meyer, Mary & Elizabeth Prugl, eds (1999) Gender Politics in Global Governance. Lanham, MD: Rowman & Littlefield.

Norris, Pippa & Ronald Inglehart (2003) Rising Tide: Gender Equality and Cultural Change Around the World. Cam- bridge: Cambridge University Press.

Peters, Julie & Andrea Wolper, eds (1995) Women’s Rights, Human Rights. New York: Routledge.

Peterson, Spike V & Anne Sisson Runyan (2010) Global Gen- der Issues in the New Millenium. Boulder, CO: Westfield.

Pettman, Jan Jindy (1996) Worlding Women: A Feminist Inter- national Politics. New York: Routledge.

Poe, Steven C; C Neal Tate & Linda Camp Keith (1999) Repression of the human right to personal integrity revisited: A global cross-national study covering the years 1976–1993. International Studies Quarterly 43(2): 291–313.

Ray, James Lee (2008) Democracy and International Conflict: An Evaluation of the Democratic Peace Proposition. Colum- bia, SC: University of South Carolina Press.

Robinson, Fiona (2011) The Ethics of Care: A Feminist Approach to Human Security. Philadelphia, PA: Temple University Press.

Ruddick, Sara (1989) Maternal Thinking. Boston: Beacon. Savery, Lynn (2007) Engendering the State: The International

Diffusion of Women’s Human Rights. Abingdon & New York: Routledge.

Sen, Amartya (1999) Development as Freedom. Oxford: Oxford University Press.

Simmons, Beth (2010) Mobilizing for Human Rights: Interna- tional Law in Domestic Politics. New York: Cambridge Uni- versity Press.

Smith, Jackie; Charles Chatfield & Ron Pagnucco, eds (1997) Transnational Social Movements and Global Poli- tics: Solidarity Beyond the State. Syracuse, NY: Syracuse University Press.

Smith, Steve (2001) Foreign policy is what states make of it. In: Vendulka Kubálková (ed.) Foreign Policy in a Con- structed World. Armonk, NY: ME Sharpe.

Spees, Pam (2008) Women’s advocacy in the creation of the international criminal court: Changing the landscapes of justice and power. In: Karen Anderson & Mary Hawkes- worth (eds) War and Terror: Feminist Perspectives. Chicago, IL: University of Chicago Press.

Steans, Jill (2006) Gender and International Relations: Issues, Debates and Future Directions. Boston, MA: Polity.

Stienstra, Deborah (1994) Women’s Movement and Interna- tional Organizations. Basingstoke: Macmillan.

Tessler, Mark & Ina Warriner (1997) Gender, feminism, and attitudes toward international conflict: Exploring relation- ships with survey data from the Middle East. World Politics 49(2): 250–281.

Tickner, J Ann, (2001) Gendering World Politics. New York: Columbia University Press.

Togeby, Lise (1994) The gender gap in foreign policy atti- tudes. Journal of Peace Research 31(4): 375–392.

Towns, Ann E (2010) Women and States: Norms and Hier- archies in International Society. Cambridge: Cambridge University Press.

Turpin, Jennifer & Lois Ann Lorentzen, eds (1996) The Gen- dered New World Order. New York: Routledge.

Weldon, S Laurel, (2006) Inclusion, solidarity, and social movements: The global movement against gender violence. Perspectives on Politics 4(1): 55–74.

Younas, Javed (2008) Motivation for bilateral aid allocation: Altruism or trade benefits. European Journal of Political Economy 24(3): 661–674.

ALISON BRYSK, b. 1960, PhD in Political Science (Stanford University, 1990); Mellichamp Professor in Global Governance, University of California, Santa Barbara (2010– ); main interest: human rights; most recent book: Speaking Rights to Power (Oxford University Press, 2013).

AASHISH MEHTA, b. 1974, PhD in Agricultural and Applied Economics (University of Wisconsin-Madison, 2004); Assistant Professor, Global Studies, University of California, Santa Barbara (2007– ); main interest: political economy of development.

110 journal of PEACE RESEARCH 51(1)

by guest on August 4, 2015jpr.sagepub.comDownloaded from