Will rising powers undermine global norms? The case of gender-balanced decision-making

Abstract

Will rising world powers tilt global norms in less progressive directions? While there has been much theoretical speculation on this question, few scholars have explored it empirically. This article uses existing empirical evidence to test the extent to which countries compromise their own values to gain favor with influential countries. We analyze the gender of ambassadors sent to rising powers and established powers and suggest that this analysis generates insight into the likely effects of rising powers on global norms. We find that the probability that a country will send a woman ambassador to a rising power is significantly lower than the probability that it will send a woman ambassador to an established world power. We show that this difference increases when the sending country exhibits high levels of international economic dependence. Our findings imply that the rise of countries like China, Russia, and India may undermine the global gender-balanced decision-making norm and progressive global norms more broadly.

Keywords

Ambassadors gender-balanced decision-making global norms rising powers

Introduction

Over the past three decades, a diverse set of actors has driven the creation of global norms associated with domains like democracy, torture, gender, and the environment. Researchers have studied the evolution, consolidation, and effects of global norms; their work suggests that effects of norms are considerable (Checkel, 1998; Cortell and Davis, 2000; Hafner-Burton, 2013; Keck and Sikkink, 1998). As these studies have been published, global power balances have begun to shift. Rising powers like Brazil, Russia, India, China, and South Africa (the so-called BRICS) have developed rapidly. These large, high-growth countries are gaining economic and political clout in a globalizing world. To what extent might rising powers affect the global normative order and the behaviors of other states?

Some observers fear that rising powers will erode progressive norms (Castañeda, 2010; Chin and Thakur, 2010). Par Engstrom (2012: 1–3) has considered this question in the context of human rights:

[i]f political, economic and social power is important in understanding the development of international human rights regime, what are the implications as power shifts in the international system? … Countries such as Brazil, China, India, and South Africa are not likely to develop approaches to human rights in line with the “moral consensus” that many human rights advocates, and Western policy-makers, seek to advance.

Others contest the idea that rising powers will undermine liberal institutions and norms. Authors such as Breslin (2013), Ikenberry (2011), Kahler (2013), and Nel (2010) suggest that emerging powers are internally diverse and, in certain cases, disposed to maintain the normative and institutional structures that have facilitated their ascent.

To date, however, much of the thinking about the effects of rising powers has been speculative and theoretical. Our approach is different. We use existing empirical evidence to test the idea that rising powers will undermine global norms in the future. We do this by investigating the extent to which states already compromise their values to gain favor with influential countries. We focus on the global norm on gender-balanced decision-making (GBDM) (Krook and True, 2012). Over the course of the post-war era, there has been increased international attention to women’s participation in domestic and international political institutions. The United Nations (UN) World Conferences on Women culminated in the 1995 Beijing Platform for Action, which called for a minimum of 30% women in decision-making posts. The GBDM norm has both promoted and been shaped by the adoption of gender quotas and increases in women’s representation in legislatures, cabinets, and corporate governance boards. It is possible that the increasing prominence of countries that challenge or undermine the GBDM norm will weaken the norm over time and that levels of women’s representation across the world will decline. The stakes of such a development could be very high; indeed, recent research suggests that lower levels of women’s political representation are associated with higher levels of interstate conflict (Caprioli, 2000; Koch and Fulton, 2011), higher levels of intrastate civil conflict (Melander, 2005a), higher levels of human rights abuses (Melander, 2005b), and higher rates of rape and domestic violence (Pinker, 2011).

To explore the possible impact of rising powers on the GBDM norm (and on progressive global norms more broadly), we analyze the gender of ambassadors in country dyads. We use dyadic ambassador data to investigate the extent to which countries compromise their values to accommodate the values of rising powers. We work from the premise that countries use ambassadorial appointments in two ways: to present themselves to host states and to respond to hosts’ values. As bridges between domestic and international arenas, ambassadors are symbols of sending countries’ values and signals of their countries’ ability to comprehend hosts’ values.

Our empirical argument focuses on the way states behave toward established powers and rising powers. In general, we use G7 (Canada, France, Germany, the UK, the US, Italy, and Japan) countries to represent established powers and BRICS to represent rising powers. We also concentrate, within these categories, on “G5” countries (Canada, France, Germany, the UK, and the US), which have been GBDM leaders and/or advocates, and the “RIC” countries (Russia, India, and China), which have generally been GBDM laggards. The empirical argument has three components. First, we show that sending countries, all else equal, appoint more women to host countries that value GBDM and fewer women to host countries that do not value gender equality. Second, we show that these patterns are more pronounced when sending states are economically dependent. When a sending state is economically dependent, it is more likely to mirror a powerful host state’s values on GBDM. Third, we show that relative to their domestic approaches to GBDM, economically dependent countries send more women ambassadors to established powers than they do to rising powers. It is possible that as rising powers play larger geopolitical roles, their position may inch a bit closer to that of established powers. Taken together, however, our arguments, which are based on the past performance of established and rising powers, suggest that rising powers may threaten women’s participation in decision-making and undermine recent representational gains. More broadly, they suggest that rising powers may dislodge global norms and usher in a less liberal global order.

We develop our argument in six sections. The first discusses the GBDM norm. The second considers rising powers and canvasses opinions about the extent to which they might threaten global norms. The third section lays out hypotheses. The fourth discusses data, and the fifth presents empirical results. The final section provides conclusions and presents avenues for future research.

The GBDM norm

As Krook and True (2012) demonstrate, the global norm on GBDM has a long, continuously evolving history. Early in its post-war history, the UN and its Committee on the Status of Women discussed the rights of women to vote, contest, and hold political office. Support for the notion of women’s political inclusion was always far from universal; while 1952’s Convention on the Political Rights of Women passed by a vote of 46–0 with 11 abstentions, more than 40 countries that signed the Convention reserved the right to opt out of certain of its provisions. In the years after the Convention was signed, activists and certain member states used successive conferences to expand the GBDM’s remit, moving beyond the “mere” focus on voting and eligibility and progressively highlighting the importance of women’s actual representation in decision-making roles. The UN World Conferences on Women, which began in 1975, were critical promoters of the GBDM norm’s representational turn, as was the UN General Assembly’s adoption of the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) in 1979 (Jacob et al., 2014; Krook and True, 2012; Paxton et al., 2006).

It is important, in this context, to recognize two facts. First, the development of the GBDM norm has not been driven exclusively by Western states or actors. The chronology of seminars and conferences that shaped the norm — Bangkok (1957), Bogotá (1959), Addis Ababa (1960), Ulan Bator (1965), Helsinki (1967), Accra (1967), Mexico City (1975), Nairobi (1985), and Beijing (1995) — is dominated by Southern and non-Western locales. This is an important reminder that transnational activism and Southern input have been critical to the norm’s development. Actors from developing countries have adopted and championed the norm before and after Beijing, often integrating it into regional and national political orders. Latin Americans and Africans have played particularly critical roles in the formulation, negotiation, and implementation of the GBDM norm, and Latin American countries have been leaders in establishing quotas for GBDM (Baldez, 2004; Htun and Jones, 2002). In 2015, nine of the 12 countries in the world’s top 10 for women’s parliamentary representation (including ties) were in sub-Saharan Africa and Latin America, and the African Union’s (AU’s) constitutive act requires that at least 50% of AU commissioners be women. While the notion of GBDM is consistent with the Western liberal tradition, the GBDM norm has also been shaped by and influential within non-Western contexts.

Second, G7 states have often promoted the norm; Hillary Clinton, for example, gave a number of speeches during her tenure as secretary of state that championed the importance of bringing women into power. However, established powers’ domestic actions have not always aligned with their global rhetoric. The US, for example, has never ratified CEDAW, and the US and Japan are below global averages for women’s legislative representation.

The GBDM norm has continued to evolve in the late 20th and early 21st centuries. In 1995, the Beijing Platform for Action set a target of 30% women in decision-making posts, and the UN’s Women, Peace and Security agenda has stressed the importance of gender balance in conflict prevention, peacekeeping, peace building, and post-conflict reconstruction. UN institutions, in parallel to these processes, have stressed the importance of gender balance in the UN system (notwithstanding episodes of sexual abuse by UN peacekeepers and a failed recent attempt to select a female secretary-general).

Figure 1 presents data on women’s legislative representation in the post-Beijing era (1995–2015), highlighting developments in the G5, the G7, and the BRICS.¹ The trend lines show that established powers have followed an upward trajectory and have generally outpaced emerging powers when it comes to integrating women into legislative office. Established powers are still far from gender equality; two G7 states (Japan and the US), for example, continue to fall below the global mean for women legislators in 2015. For rising powers, there is more variation. South Africa dramatically increased women’s legislative representation after 1994 and is now well above the global average (41.8% versus a global average of 22.6%). China (23.6%) is also currently above the global mean, but its legislature is weak, and women remain under-represented at the top echelons of party and state power (Zeng, 2014). Brazil, India, and Russia have experienced slight gains in women’s legislative representation in recent decades but remain below global averages.²

Figure 1.

Women’s legislative representation over time.

In the realm of women’s cabinet representation, G7 states generally outperform BRICS states, with the exception of South Africa. In 2015, the global average level of women’s cabinet representation was 18.7% (Inter-Parliamentary Union, 2015). Russia (6.5%), China (11.5%), and Brazil (15.4%) were below this level. India (22.2%) was slightly above it, and South Africa (41.7%) was significantly above it. All G7 states were above the global average, and France (50%) and Italy (43.8%) were substantially higher.

According to the Global Gender Gap Report 2014, which measures gender gaps in health, education, and economic participation and opportunity, of the BRICS states, only South Africa (ranked 18th) performed well. The four other rising powers fall in the bottom half of the table, with Brazil at 71, Russia at 75, China at 87, and India at 114. Among the G7 states, only Japan ranked in the bottom half of the countries (coming in at 104 out of 142 states).

While there is variation within established powers and within rising powers when it comes to current levels of women’s political representation, there are important contrasts between the two sets of powers in citizens’ diffuse attitudinal acceptance of GBDM. Examining public opinion on gender and leadership, for example, respondents in India (62%), Russia (57.3%), South Africa (51.6%), China (47.9%), and Brazil (28.4%) are more likely to agree with the statement, “On the whole, men make better political leaders than women do,” than respondents in the European Union (EU) (26%)³ and the US (19.4%) (World Values Survey, 2014).

To summarize, the development of the GBDM norm and the commitments of world leaders to women’s representation have raised international attention to issues of gender and leadership. The resulting environment has increased the salience of gender and encouraged chief executives to signal their commitment, via cabinet, ambassadorial, and other appointments, to GBDM. Of course, other considerations continue to affect executives’ appointment decisions. However, recent international developments have increased the weight of gender as a factor in their appointment calculus.

Rising powers and the future of the GBDM norm

To what extent might rising powers decrease the power of GBDM and other global norms? Thinkers have considered this question since at least 2001, when Jim O’Neill coined the acronym “BRIC” to describe a group of “large emerging economies with high growth potential” (Armijo and Roberts, 2014). By 2006, the BRIC grouping had become a real institution. BRIC states began organizing summits parallel to meetings at the UN and the international financial institutions (IFIs). In 2009, Russia hosted the first BRIC leader summit, and the BRICS have met annually since then. South Africa, whose population and gross domestic product (GDP) are smaller than those of the BRICs, provided the nascent group with an attractive presence in Africa and received an invitation to join the group in 2010 (Hervieu, 2010).

Since 2010, the BRICS have grown more economically prominent. According to Armijo and Roberts (2014), BRIC states will represent four of the top seven economies in the world by 2020, surpassing all European states except Germany. Russia’s 2015 BRICS presidency website boasted that the BRICS comprise 42% of the world’s population, 26% of its territory, and 27% of its GDP (BRICS, 2015).

Might these rising powers construct alternative global normative and institutional orders? There are certainly signs of coordinated action and influence. In 2011, for example, the BRICS served simultaneously on the UN Security Council and cooperated frequently:

The five found it valuable to interact and adopt joint positions that would not have been predicted from self-interest or differences in regime types and values … BRICS countries coordinated their positions, and were often obstructive on sanctions or the use of force. Susan Rice, then the US Permanent Representative to the UN, questioned the voting of the three democratic BRICS as not always “consistent with their own democratic values.” (Armijo and Roberts, 2014: 16)

Beyond 2011, rising powers have taken congruent positions on otherwise divisive international issues. All BRICS opposed North Atlantic Treaty Organization (NATO)-led support for opponents of Muammar al-Qaddafi in Libya, for example, and all have opposed Western efforts to pressure Bashar al-Assad out of the Syrian presidency. “BICS” states have refrained from criticizing Russian actions in Ukraine and have rejected calls to join the Western/Japanese/Australian sanctions regime (Keck, 2014). The New Development Bank, conceived as an alternative to the Western-dominated IFIs, has been the BRICS’s most significant institutional project. There are at least some signs, therefore, that rising powers are willing and able to confront the institutions and norms that existing powers have worked to establish.

In light of these developments, certain observers have implied that rising powers might undermine progressive global norms. Ikenberry (2011: 56), for example, references “anxious observers [who] argue that the world will not just look less American — it will also look less liberal.” Chin and Thakur (2010: 120) are concerned that rising powers will end up “exporting a variation of ‘the developmental state’ as a new model … recalibrating the multilateral order to set aside claims of universal civil and political rights.” Hurrell (2007: 291–292) posits that rising powers, while “sometimes professing a liking for multilateralism, have tended to resist the effective delegation of authority to international bodies.” Furthermore, Keukeleire and Hooijmaaijers (2014) note that, despite their many differences, the BRICS share a desire to reform the UN and the IFIs, to prioritize the G20 over the G7, and to defend state sovereignty as the supreme principle of international politics. Castañeda (2010: 112) encapsulates all of these concerns:

Constructing [a] web of [liberal] international norms has been slow and painful, with less overall progress … than some have wished for. Many countries of what used to be called the Third World have contributed to parts of the edifice: Mexico to disarmament and the law of the sea; Costa Rica to human rights; Chile to free trade. But now, the possible accession of Brazil, China, India, and South Africa to the inner sanctum of the world’s leading institutions threatens to undermine those institutions’ principles and practices.

Anxiety about the implications of rising powers for global norms and institutions is not universal. Indeed, some observers are optimistic. Bello (2014), Desai (2013), and Pieterse (2011), for example, express hope that rising powers construct a more just global normative order. Others (e.g. Culp, forthcoming) suggest that rising powers are already pulling their weight. Perhaps the predominant view, however, is that rising powers are internally diverse competitors that are unlikely to fundamentally challenge the status quo, chip away at entrenched norms, and/or generate powerful new norms or institutions (Ikenberry, 2011; Kahler, 2013; Nel, 2010; Wang and French, 2013; Wu, 2016). Breslin (2013: 629, emphasis in original) provides a useful summary of this line of argument:

While the BRICS states might agree on what they don’t like and what they are dissatisfied with, there doesn’t seem to be a shared vision of what a radically different world order might look like. Rather, the agenda seems to focus primarily on the redistribution of power within the existing order. We might suggest that an attempt to articulate a clear alternative by any member could split the BRICS.

This debate leads to important questions about the fate of the GBDM norm. All BRICS signed and ratified CEDAW, but they did so well before their years of rapid development culminated. As discussed earlier, with the exception of South Africa, no rising power has a progressive record on women’s political representation or gender-equality outcomes.

Why ambassadors? Hypotheses about the effects of rising powers on the GBDM norm

To move toward hypotheses, we return to the observation that the GBDM norm increases the significance of gender as a variable affecting heads of governments’ appointment decisions. It is unlikely, we suggest, that the GBDM norm would affect all executive appointments in similar ways. Ambassadors, for example, are far more exposed to the world polity (Meyer et al., 1997) than other executive appointees. Other things equal, chief executives who are inclined to signal their acceptance of the norm can do so more effectively with an ambassadorial appointment than with an appointment to domestically focused posts (Jacob et al., 2013). Ambassadors live abroad, engage daily with foreign interlocutors, and participate in deeper and more frequent global interactions than most of their counterparts in national ministries and agencies.

For these reasons, we suggest that ambassadorial appointments are an important means of measuring states’ commitments to the GBDM norm. Diplomatic appointments take place at the nexus of domestic and international arenas and have an important symbolic component: by appointing ambassadors with specific identities and experiences, executives send signals to other states. They use ambassadorial appointments to signal their own values and to communicate their appreciation of host countries’ values. Jonsson (2008: 35) encapsulates this symbolic function of diplomatic appointments: “the diplomat is seen as a representative in the same way that a flag represents a state.” Similarly, Nowotny (2011: 46) observes:

Like flags, diplomats can also be used for sending coded but well-understood messages. Embassies can be opened or closed. Diplomatic recognition can be granted to, or withheld from a state. Ambassadors can be sent, can be withdrawn or may be called home for consultations. Diplomats may, or may not mark with their presence an event in their host country. … It is useful to have such a symbolic language. It sends a message that is fairly clear, but avoids causing damages that other expressions of displeasure, such as economic sanctions, might entail.

A sending state can demonstrate its commitment to gender-equality norms through the identity of the ambassadors it appoints.

We work from the assumption that this commitment is dyadic, rather than universal. The commitment shown by sending country S can vary with the identity and values of host country H. We expect that S will signal greater commitment to gender equality to an H that is more committed to gender equality, and, conversely, that S will be less likely to signal a commitment to gender equality to an H that values gender equality less:

H1: The more H’s behavior accords with the GBDM norm, the more likely S will be to appoint a woman ambassador to H.

We assume, in addition, that chief executives take some ambassadorial appointments more seriously than others. They are likely to think harder about appointments when they are in an economically weak position. The more economically dependent S is, the more likely it will be to choose an ambassador (e.g. send a signal) that will reflect H’s approach to the GBDM norm. This line of argument draws on Bush’s (2011) study of gender quotas, which finds that countries more dependent on foreign aid are more likely to respond to implicit pressures by external forces to demonstrate their commitment to gender norms. Leaders of dependent states seek to signal their compliance with these norms in order to attract aid or enhance their reputation. If a powerful country signals its commitment to GBDM, dependent countries will be more likely to dispatch women ambassadors to its capital. On the other hand, countries that depend on a powerful partner that flouts the GBDM norm will be less likely to post women to that partner:

H2: The more economic power H exercises over S, the more likely S will be to mimic H’s approach to the GBDM norm.

Of course, a thorough explanation of why and where women are appointed ambassadors would require attention to broader conditions in S. Levels of women’s education, recruitment procedures for diplomatic careers, the gendered division of professional labor, rules governing diplomats’ employment conditions, and other factors are all likely to influence how many women serve in a country’s ambassadorial corps (Bentley and Warner, 2000; Jordan, 2010; Rahman, 2011). We work from the assumption, though, that while domestic conditions in S influence how many women are included in S’s ambassadorial ranks, they cannot fully account for the countries to which specific ambassadors are posted. Since ambassadorial appointments are dyadic, we can compare a sending country’s appointments to different host countries while holding sending country characteristics constant. Differences in appointments should therefore be attributable to host country characteristics.

Evidence from case studies supports the idea that H’s gender climate affects S’s appointment calculus. Diplomats, the literature suggests, need to fit in with the local culture, and countries take local values into account when assessing the likelihood that potential ambassadors will fit in particular locales. Neumann (2012), for example, indicates that when Oslo’s diplomatic academy accepted its first female diplomat, Gudrun M., in 1938, diplomatic training normally commenced with a placement abroad. Searching for a placement, the Norwegian foreign minister asked Norway’s legation in Italy about assigning the female diplomat to Italy. The head objected, stating: “As seen from Italy, it would appear utterly strange for a lady to be registered (anmeldt) as an attaché. Amongst the hundred names in the Italian diplomatic list, there is no lady” (Neumann, 2012: 141). Accepting the argument that it would seem odd to have a woman diplomat in Rome, the foreign minister placed her in London instead.

Studying US women diplomats between 1945 and 2004, Beckett (2008: 29) finds that the “policy of a receiving country regarding women, to an extent, affects the sending government’s decision to appoint a female representative.” Noting that the first female ambassador from the US was sent to Denmark, Beckett (2008: 32) suggests: “perhaps the U.S. government had fewer qualms in assigning a female diplomat to a country whose own government already consisted of a number of women representatives.” Even as women serve as ambassadors in a more diverse range of countries, they are far less likely to be placed in Asia and the Middle East, regions with lower levels of women’s political representation (Inter-Parliamentary Union, 2015).⁴ We expect, therefore, that dependent countries will use gendered appointments strategically: relative to the extent to which its domestic behavior coincides with the GBDM norm, S will “over-project” to partners (e.g. established powers) that it sees as norm leaders and “under-project” to partners that it sees as norm laggards (e.g. rising powers):

H3: A dependent S will “over-appoint” women ambassadors to established powers and “under-appoint” women to rising powers.

Data

Our dependent variable is the percentage of ambassadors that a country hosts that are women. In general, our raw data come from the protocol departments of host countries’ ministries of foreign affairs; Table 1 provides more detail about data sources, and Table 2 summarizes dependent variable values. Data on ambassadors come from fall 2015. Protocol departments often indicate the sex of ambassadorial appointees (e.g. “Her Excellency”). In the event that an appointee’s sex is not clear, we determine it through a combination of internet searches and press scrutiny.

Table 1.

Data sources.

Host country	Source	Data as of	URL
Brazil	Ministry of External Relations	10/2015	http://www.itamaraty.gov.br/index.php?option=com_content&view=article&id=3663&Itemid=112&lang=pt-BR
China	Ministry of Foreign Affairs	11/2015	http://www.fmprc.gov.cn/mfa_eng/wjb_663304/zwig_665342
India	Ministry of External Affairs	10/2015	http://mea.gov.in/indian-missions-abroad.htm
Russia	Embassypages.com	11/2015	http://www.embassypages.com/russia
South Africa	Ministry of Foreign Affairs	10/2015	http://www.dfa.gov.za/foreign/sa_abroad/hom.htm
Canada	Department of Global Affairs	10/2015	http://www.international.gc.ca/cip-pic/mission_info-info_mission.aspx?lang=eng
France	Ministry of Foreign Affairs and International Development	10/2015	http://www.diplomatie.gouv.fr/fr/IMG/pdf/copie_de_ambassadeursdif_cle81d992.pdf
Germany	Ministry of Foreign Affairs	10/2015	http://www.auswaertiges-amt.de/cae/servlet/contentblob/332372/publicationFile/209421/DtAuslandsvertretungenListe.pdf
Italy	Ministry of Foreign Affairs and International Cooperation	10/2015	http://www.esteri.it/mae/it/ministero/servizi/italiani/rappresentanze
Japan	Ministry of Foreign Affairs	10/2015	http://www.mofa.go.jp/about/emb_cons/mofaserv.html
UK	Foreign and Commonwealth Office	10/2015	https://www.gov.uk/government/world
US	Department of State	8/2015	http://www.state.gov/documents/organization/245951.pdf

Table 2.

Women ambassadors hosted by selected countries.

Country	# of women ambassadors	# of male ambassadors	% women ambassadors
Germany	40	131	23.39
South Africa	31	117	20.95
France	37	142	20.67
Italy	33	131	20.12
Canada	31	132	19.02
UK	31	145	17.61
Japan	27	145	15.70
Brazil	19	105	15.32
US	28	159	14.97
China	19	143	11.73
India	15	134	10.07
Russia	11	136	7.48

Note: G7 countries in plain font; BRICS countries in italics.

We measure GBDM norm compliance in four ways. Our first and second measures are the percentage of women in the legislature and the percentage of women in the cabinet. Data for legislators come from the Inter-Parliamentary Union (2015); for ministers, we code the gender of all ministers included in the January 2010 CIA Chiefs of State and Cabinet Members report. Our third measure is the 2015 World Economic Forum’s Global Gender Gap Political Empowerment Sub-Index (World Economic Forum, 2015). This index combines three measures: ratio of females with seats in Parliament over the male value (weighted 0.310); ratio of females at the ministerial level over the male value (weighted 0.247); and ratio of the number of years of a female head of government over the past 50 years over the male value (weighted 0.443). This data source differs from the previous sources in that it includes data on the head of government, combines the three measures into a single index, and weights the three component measures in a novel way. Since this measure puts particular emphasis on the presence of a female head of government, countries with a female president or prime minister perform especially well. India, for example, was ranked ninth in 2013 and 15th in 2014, despite its low levels of women’s legislative representation.

Finally, we use the United Nations Development Programme’s 2014 Gender Inequality Index (GII) to measure gender equality more broadly. The GII indicates gender-based disadvantages in the areas of reproductive health, empowerment, and the labor market. Measures include maternal mortality ratio, adolescent fertility rate, share of parliamentary seats held by each sex, attainment of secondary and higher education levels, and labor market participation rate. GII scores range from 0 (perfect gender equality) to 1, where either gender does as poorly as possible in all areas measured in the index (UNDP, 2014b: 5). For our included countries, values range from 0.041 (Germany) to 0.563 (India) (UNDP, 2014a).

Acknowledging that dependence can take economic, military, and other forms, we nonetheless focus here on economic dependence. We use World Bank categories to construct dummies for two types of highly dependent states — heavily indebted poor countries (HIPC)⁵ and countries in fragile and conflict-affected situations (FCS)⁶ — and two types of economically powerful countries — advanced industrial liberal countries of the Organisation for Economic Co-operation and Development (OECD)⁷ and non-OECD high-income countries (NOC).⁸

Empirical analysis

We present three empirical subsections, corresponding to our three hypotheses. First, we show that established powers host more women ambassadors than rising powers and that rising powers with low GBDM norm compliance host particularly few women ambassadors. Second, we show that countries that are economically dependent drive the observed differences between women ambassadors hosted by established powers and rising powers. For countries that are less economically dependent, there is less of a difference in the gender profiles of the ambassadors sent to established and rising powers. Third, we show that even after accounting for countries’ approaches to the GBDM norm, dependent countries send more women ambassadors to established powers and fewer women to rising powers. This argument emphasizes the role of dependency in determining whether the GBDM values that a country projects toward established and rising powers exceed or undershoot its domestic GBDM values.

Which countries host more women ambassadors?

We begin by asking whether established powers or rising powers host more women ambassadors and whether the percentage of women ambassadors that a country hosts relates to its domestic GBDM attainment. Figure 2 presents data on the percentage of women ambassadors hosted by G7 and BRICS countries on the vertical axis. The subfigures plot these data against the four selected indicators for domestic GBDM norm compliance.

Figure 2.

Percentage of hosted ambassadors that are women, by host GBDM compliance indicator.

Several points emerge from Figure 2. First, there is considerable variation in the percentage of women ambassadors hosted by the 12 countries, and there is a clear G7–BRICS dichotomy. Six of the top seven are established powers, and four of the bottom five are rising powers. South Africa is the only clear exception to the G7–BRICS dichotomy. Russia, India, and China make up the bottom three.

Second, there is a positive correlation in all subfigures between a country’s GBDM compliance and the proportion of women ambassadors it hosts. The subfigures also plot the linear regression fit through the data points, along with the 90% confidence interval. Despite the small number of data points, the bivariate relationships are statistically significant.⁹ Countries that perform well on GBDM indicators tend to host a larger proportion of women ambassadors, and countries that perform poorly on these measures tend to host a small proportion of women ambassadors.

Third, the countries toward the bottom left (relatively low on both GBDM compliance and hosting women ambassadors) are drawn from the BRICS — and, specifically, RIC — while the countries toward the top right are drawn from the G7. This finding is consistent with the argument that, other things equal, an S is more likely to send a woman ambassador to an H if S perceives H to be more committed to the GBDM norm. This preliminary evidence supports H1.

We also engage a comparison between established and rising powers after accounting for sender country characteristics. Conceptually, the gender of the ambassador that S sends to H should be a function of factors specific to S (for instance, the nature of the pool of ambassadors available to dispatch) and factors specific to H (for instance, H’s GBDM compliance). The use of statistical controls can potentially filter out some of the S-specific factors, but, as is well known, the use of statistical controls is inferior to the use of design controls. We take advantage of a simple design control to filter out S-specific effects: we use the difference in the gender of the ambassador sent by sender S to H₁ and H₂. S-specific factors will arguably be canceled out through this procedure.¹⁰

Table 3 presents results from tests of statistical significance of the difference in proportion of women ambassadors for all pairs of host countries among the G7 and BRICS. The table reports coefficients and standard errors as a matrix; the coefficient in each cell (X, Y) shows the estimated average difference in women ambassadors placed by sending countries to host X and host Y. A positive coefficient implies that, after accounting for sending country fixed effects, X is likely to host more women ambassadors relative to Y — and the size of the coefficient estimates the average difference in proportions.

Table 3.

Pair-wise differences in women ambassadors hosted by G7 and BRICS countries.

	FRA	DEU	GBR	CAN	USA	ITA	JPN	RUS	CHN	IND	BRA	ZAF
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
FRA		−0.02(0.04)[167]	0.03(0.04)[169]	0.02(0.04)[160]	0.05(0.04)[173]	0.01(0.04)[160]	0.06(0.04)[163]	0.12^*** (0.04)[143]	0.08^** (0.04)[157]	0.10^** (0.04)[144]	0.08(0.05)[121]	0.01(0.04)[143]
DEU	0.02(0.04)[167]		0.07(0.04)[161]	0.03(0.04)[152]	0.10^** (0.04)[166]	0.05(0.05)[156]	0.08^* (0.05)[159]	0.13^*** (0.04)[145]	0.11^*** (0.04)[153]	0.14^*** (0.04)[144]	0.07(0.05)[121]	0.03(0.05)[139]
GBR	−0.03(0.04)[169]	−0.07(0.04)[161]		−0.01(0.04)[158]	0.03(0.04)[172]	−0.06(0.04)[157]	0.02(0.04)[159]	0.04(0.04)[139]	0.01(0.04)[151]	0.06(0.04)[141]	−0.03(0.04)[119]	−0.05(0.04)[139]
CAN	−0.02(0.04)[160]	−0.03(0.04)[152]	0.01(0.04)[158]		0.04(0.03)[161]	−0.03(0.05)[151]	0.03(0.04)[152]	0.07^* (0.04)[134]	0.05(0.05)[143]	0.05(0.04)[134]	0.01(0.04)[117]	−0.05(0.05)[133]
USA	−0.05(0.04)[173]	−0.10^** (0.04)[166]	−0.03(0.04)[172]	−0.04(0.03)[161]		−0.06(0.04)[158]	−0.01(0.04)[167]	0.04(0.04)[139]	0.01(0.04)[155]	0.03(0.04)[142]	−0.05(0.04)[118]	−0.09^** (0.05)[142]
ITA	−0.01(0.04)[160]	−0.05(0.05)[156]	0.06(0.04)[157]	0.03(0.05)[151]	0.06(0.04)[158]		0.05(0.04)[152]	0.15^*** (0.04)[138]	0.07^* (0.04)[148]	0.09^** (0.04)[141]	0.07(0.05)[120]	−0.01(0.05)[138]
JPN	−0.06(0.04)[163]	−0.08^* (0.05)[159]	−0.02(0.04)[159]	−0.03(0.04)[152]	0.01(0.04)[167]	−0.05(0.04)[152]		0.06(0.04)[138]	0.02(0.04)[151]	0.03(0.04)[143]	−0.03(0.04)[121]	−0.09^** (0.04)[137]
RUS	−0.12^*** (0.04)[143]	−0.13^*** (0.04)[145]	−0.04(0.04)[139]	−0.07^* (0.04)[134]	−0.04(0.04)[139]	−0.15^*** (0.04)[138]	−0.06(0.04)[138]		−0.06^* (0.03)[137]	−0.02(0.04)[134]	−0.05(0.04)[111]	−0.10^** (0.04)[125]
CHN	−0.08^** (0.04)[157]	−0.11^*** (0.04)[153]	−0.01(0.04)[151]	−0.05(0.05)[143]	−0.01(0.04)[155]	−0.07^* (0.04)[148]	−0.02(0.04)[151]	0.06^* (0.03)[137]		0.02(0.03)[140]	−0.06(0.04)[116]	−0.10^** (0.04)[136]
IND	−0.10^** (0.04)[144]	−0.14^*** (0.04)[144]	−0.06(0.04)[141]	−0.05(0.04)[134]	−0.03(0.04)[142]	−0.09^** (0.04)[141]	−0.03(0.04)[143]	0.02(0.04)[134]	−0.02(0.03)[140]		−0.05(0.03)[113]	−0.09^** (0.04)[129]
BRA	−0.08(0.05)[121]	−0.07(0.05)[121]	0.03(0.04)[119]	−0.01(0.04)[117]	0.05(0.04)[118]	−0.07(0.05)[120]	0.03(0.04)[121]	0.05(0.04)[111]	0.06(0.04)[116]	0.05(0.03)[113]		−0.09^* (0.05)[115]
ZAF	−0.01(0.04)[143]	−0.03(0.05)[139]	0.05(0.04)[139]	0.05(0.05)[133]	0.09^** (0.05)[142]	0.01(0.05)[138]	0.09^** (0.04)[137]	0.10^** (0.04)[125]	0.10^** (0.04)[136]	0.09^** (0.04)[129]	0.09^* (0.05)[115]

Notes: Standard errors in parentheses; number of observations in square brackets.

G7: Group of 7; BRICS: Brazil, Russia, India, China, South Africa; FRA: France; DEU: Germany; GBR: United Kingdom; CAN: Canada; USA: United States of America; ITA: Italy; JPN: Japan; RUS: Russia; CHN: China; IND: India; BRA: Brazil; ZAF: South Africa.

Consider first the pair-wise differences among G7 hosts. There are 21 distinct host pairs. In 19 of these cases, the difference in proportion does not attain statistical significance even at the 10% level.¹¹ Consider next the pair-wise differences among BRICS hosts. As noted earlier, South Africa is an exception, hosting a significantly greater proportion of women ambassadors than each of the other rising powers. However, if we bracket the South African case, then most other pair-wise differences among the BRIC countries are not statistically significant even at the 10% level.¹²

Among pairs between established and rising powers, there are 10 cases of statistically significant differences, in all of which the G7 country has the higher proportion of women ambassadors. Further, if we focus solely on RIC, then the same 10 cases continue to hold, and they account for about half of the pairs.

In sum, the results continue to hold even after accounting for sender characteristics through a fixed-effects design. Within the G7 and BRIC countries separately, pair-wise differences are mostly not statistically significant, and in G7–RIC pairs, about half of the differences are statistically significant.

Which countries send relatively more women ambassadors to established versus rising powers?

H2 suggests that the more economically dependent an S is on an H, the more likely it is that S’s ambassador will reflect H’s compliance with the GBDM norm. Among host countries who comply more with the GBDM norm (e.g. G7 countries), we would expect more women ambassadors being sent from more dependent countries than from less dependent countries. In addition, among host countries that comply less with the GBDM (e.g. rising powers, especially RIC), we would expect fewer women ambassadors being sent from more dependent countries than others.

H2 predicts that HIPCs will dispatch more women ambassadors to established powers than they will to rising powers. The descriptive statistics in Table 4 preliminarily test this expectation. Among the 29 HIPC senders for which we have data for ambassadors to G7 and BRICS countries (Panel A), 18.9% of the ambassadors sent to all the G7 hosts are women; the figure is only 10.8% for ambassadors sent to the BRICS. The difference (8.1 percentage points) is statistically significant at the 5% level. Among the 32 HIPC senders for which we have data for ambassadors to G5 and RIC countries (Panel B), 21.3% of the ambassadors sent to G5 hosts are women; the figure is only 5.7% for ambassadors sent to RIC hosts. The difference (15.6 percentage points) is statistically significant at the 1% level. The average G7 country hosts 8.1/10.8 = 75% more women ambassadors from HIPC senders relative to the average BRICS country, and the average G5 country hosts 15.5/5.7 = 272% more women ambassadors from HIPC senders relative to the average RIC country.

Table 4.

Women ambassadors hosted by G7 and BRICS countries, by sending country classification.

	HIPC	FCS	OECD	NOC	Rest
Panel A
G7	18.9	17.5	16.2	20.6	13.2
BRICS	10.8	6.7	11.8	13.8	6.7
G7–BRICS	8.1**(3.4)	10.8**(4.9)	4.4(3.7)	6.8(6.4)	6.5***(2.1)
Observations	29	18	31	20	72
Panel B
G5	21.3	18.2	16.9	18.9	12.6
RIC	5.7	5.3	10.4	17.5	8.6
G5–RIC	15.5***(3.8)	12.9**(5.5)	6.5(4.2)	1.4(6.8)	3.9(2.9)
Observations	32	19	32	19	81

Notes: Standard errors in parentheses. For G7 and BRICS rows, averages are for countries that send ambassadors to at least three G7 countries and at least two BRICS countries. For the G5 and RIC rows, averages are for countries that send ambassadors to at least two G5 countries and at least one RIC country.

Among non-HIPC senders, H2 predicts no such distinctions between women ambassadors sent to established and rising powers, irrespective of the level of domestic GBDM attainment of such non-HIPC senders. Consider particularly less dependent senders, as proxied by the World Bank’s “high income” classification. Whether a sending country has more or less domestic GBDM compliance, we expect little difference in women ambassadors sent out to established- and rising-power hosts. We use the aforementioned OECD and NOC categories to distinguish, broadly, between high-income countries with relatively high and relatively low GBDM compliance, respectively. The averages in Table 4 are consistent with our expectations: even at the 10% significance level, there is no statistical difference in women ambassadors posted to established and emerging powers, either by OECD senders or by NOC senders.

Finally, as an alternative to the HIPC classification (which focuses on economic dependence) and as a robustness check, we use the dichotomous FCS variable, which is an imperfect proxy for political dependence. Table 4 confirms that for FCS senders, there is a large and statistically significant difference in ambassadors sent to established powers relative to rising powers.

We can also test H2 using inferential statistics. Table 5 presents the results of a linear regression model including all 164 sending countries for which we have ambassador data. The dependent variable is the difference in the percentage of women ambassadors hosted by G5 and RIC countries. The independent variables of interest are dichotomous variables for HIPC, OECD, and NOC (with the rest of the world as comparator).¹³ Column (1) presents results without control variables, and other columns report results with the progressive addition of statistical controls.¹⁴ The dependent variable is the difference in women ambassadors that a country sends to the two groups (G5 and RIC); as discussed earlier, this fact mimics a fixed-effects approach. This fact is partly confirmed by the estimates in Table 5: most control variables fail to reach statistical significance. Across all columns, the key set of results — estimated coefficients for HIPC, OECD, and NOC (relative to the rest of the world) — are consistent with Table 3’s descriptive results. Compared to the rest of the world, HIPC countries consistently send a larger percentage of women ambassadors to G5 countries than they do to RICs. The difference is sizeable, with coefficients ranging from 11.2 to 14.9. Across columns, in addition, the coefficients for OECD and NOC countries have smaller sizes and do not attain statistical significance, even at the 10% level.

Table 5.

Difference in women ambassadors hosted by G5 and RIC countries.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
HIPC	11.59^**	11.18^**	11.70^**	11.38^**	12.63^**	12.41^**	14.91^**
	(4.73)	(4.81)	(5.28)	(5.27)	(5.59)	(5.87)	(6.13)
OECD	2.53	1.31	4.84	7.35	1.61	4.59	4.21
	(5.08)	(5.28)	(6.93)	(7.54)	(6.63)	(6.94)	(7.02)
NOC	−2.53	−2.40	−2.49	−2.85	−4.72	−2.75	−1.95
	(7.29)	(7.77)	(8.50)	(8.72)	(8.20)	(7.78)	(8.02)
Women legislators		0.13	0.06	0.08	0.03	0.01	−0.01
		(0.18)	(0.21)	(0.21)	(0.23)	(0.23)	(0.24)
Democracy stock				−0.01	−0.01	−0.01	−0.01
				(0.01)	(0.01)	(0.01)	(0.01)
Parliamentary system					8.58	9.93	10.55^*
					(5.75)	(6.02)	(6.11)
Legal quota						5.80	6.08
						(4.96)	(5.08)
Party quota						−7.02	−7.29
						(5.62)	(5.69)
Government fractionalization							4.32
Government fractionalization							(9.51)
Colonizer dummies	n	n	y	y	y	y	y
Constant	3.93	1.85	−0.04	−0.58	0.77	1.34	−0.11
	(2.88)	(4.01)	(6.47)	(6.51)	(6.02)	(6.10)	(6.09)
Observations	164	162	156	156	153	153	150

Notes: Standard errors in parentheses. Sending countries that send ambassadors to at least two G5 countries and at least one RIC country are included.

We now unpack the aggregate host categories used so far and make comparisons across each individual pair among the 12 host countries. For HIPC senders, Table 6 presents results from tests of statistical significance of the difference in proportion of women ambassadors for all pairs of host countries among the 12 G7 and BRICS countries. This is analogous to Table 3, where we presented similar pair-wise comparisons without homing in on HIPC senders.

Table 6.

Pair-wise differences in women ambassadors sent by HIPC countries and hosted by G7 and BRICS countries.

	FRA	DEU	GBR	CAN	USA	ITA	JPN	RUS	CHN	IND	BRA	ZAF
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
FRA		0.03(0.10)[35]	0.03(0.11)[30]	0.00(0.12)[30]	0.06(0.09)[36]	−0.07(0.11)[30]	0.03(0.10)[31]	0.19*(0.10)[26]	0.20**(0.07)[30]	0.04 (0.12)[23]	0.00(0.15)[20]	−0.07(0.10)[29]
DEU	−0.03(0.10)[35]		0.00(0.10)[29]	−0.07(0.13)[28]	0.06(0.09)[34]	0.00(0.11)[29]	−0.03(0.10)[30]	0.08(0.09)[26]	0.18**(0.07)[28]	0.05(0.12)[22]	−0.05(0.12)[19]	−0.04(0.11)[27]
GBR	−0.03(0.11)[30]	0.00(0.10)[29]		0.04(0.13)[26]	0.07(0.10)[30]	−0.12(0.12)[26]	−0.04(0.13)[26]	0.14(0.13)[21]	0.22**(0.09)[23]	0.21(0.12)[19]	−0.06(0.10)[17]	−0.09(0.14)[23]
CAN	0.00(0.12)[30]	0.07(0.13)[28]	−0.04(0.13)[26]		0.03(0.10)[30]	−0.12(0.15)[25]	0.00(0.13)[27]	0.22**(0.09)[23]	0.24**(0.09)[25]	0.10(0.12)[20]	−0.05(0.14)[19]	0.00(0.14)[23]
USA	−0.06(0.09)[36]	−0.06(0.09)[34]	−0.07(0.10)[30]	−0.03(0.10)[30]		−0.10(0.12)[30]	−0.07(0.10)[30]	0.12(0.09)[25]	0.21**(0.08)[29]	0.04(0.10)[23]	−0.16(0.12)[19]	−0.07(0.11)[29]
ITA	0.07(0.11)[30]	0.00(0.11)[29]	0.12(0.12)[26]	0.12(0.15)[25]	0.10(0.12)[30]		0.07(0.13)[27]	0.32**(0.12)[22]	0.32***(0.10)[25]	0.14(0.14)[22]	0.16(0.16)[19]	0.00(0.16)[26]
JPN	−0.03(0.10)[31]	0.03(0.10)[30]	0.04(0.13)[26]	0.00(0.13)[27]	0.07(0.10)[30]	−0.07(0.13)[27]		0.13(0.09)[24]	0.22**(0.08)[27]	0.05(0.08)[21]	−0.15(0.13)[20]	0.00(0.10)[24]
RUS	−0.19*(0.10)[26]	−0.08(0.09)[26]	−0.14(0.13)[21]	−0.22**(0.09)[23]	−0.12(0.09)[25]	−0.32**(0.12)[22]	−0.13(0.09)[24]		0.04(0.04)[23]	−0.22**(0.10)[18]	−0.25**(0.11)[16]	−0.15(0.11)[20]
CHN	−0.20**(0.07)[30]	−0.18**(0.07)[28]	−0.22**(0.09)[23]	−0.24**(0.09)[25]	−0.21**(0.08)[29]	−0.32***(0.10)[25]	−0.22**(0.08)[27]	−0.04(0.04)[23]		−0.19**(0.09)[21]	−0.29**(0.11)[17]	−0.20**(0.08)[25]
IND	−0.04(0.12)[23]	−0.05(0.12)[22]	−0.21(0.12)[19]	−0.10(0.12)[20]	−0.04(0.10)[23]	−0.14(0.14)[22]	−0.05(0.08)[21]	0.22**(0.10)[18]	0.19**(0.09)[21]		−0.20*(0.11)[15]	−0.05(0.14)[19]
BRA	0.00(0.15)[20]	0.05(0.12)[19]	0.06(0.10)[17]	0.05(0.14)[19]	0.16(0.12)[19]	−0.16(0.16)[19]	0.15(0.13)[20]	0.25**(0.11)[16]	0.29**(0.11)[17]	0.20*(0.11)[15]		−0.13(0.18)[16]
ZAF	0.07(0.10)[29]	0.04(0.11)[27]	0.09(0.14)[23]	0.00(0.14)[23]	0.07(0.11)[29]	0.00(0.16)[26]	0.00(0.10)[24]	0.15(0.11)[20]	0.20**(0.08)[25]	0.05(0.14)[19]	0.13(0.18)[16]

Notes: Standard errors in parentheses; number of observations in square brackets.

HIPC: Heavily indebted poor countries; G7: Group of 7; BRICS: Brazil, Russia, India, China, South Africa; FRA: France; DEU: Germany; GBR: United Kingdom; CAN: Canada; USA: United States of America; ITA: Italy; JPN: Japan; RUS: Russia; CHN: China; IND: India; BRA: Brazil; ZAF: South Africa.

Table 6 shows that for HIPC senders, among all 21 distinct G7 host pairs, there is not a single instance of a statistically significant (at the 10% level) difference in hosting women ambassadors. In sharp contrast, there are important differences among BRICS host pairs hosting HIPC ambassadors: the difference is statistically significant in a majority of pairs (six out of 10). Specifically, South Africa and Brazil host more women ambassadors from HIPC countries when paired with each of the RIC countries. In addition, even within the RIC grouping, India is a bit closer to South Africa and Brazil than it is to Russia and China. Finally, there are important differences in G7–BRICS pairs hosting HIPC ambassadors: there are 10 cases of statistically significant differences, in all of which the G7 country has the higher proportion of women ambassadors. However, all 10 cases are due to Russia and China: HIPC senders do not significantly distinguish between G7 countries, on the one hand, and South Africa, Brazil, and India, on the other. Rather, the key distinction that HIPC senders make is between G7 countries, on the one hand, and Russia and China, on the other.¹⁵

Dependence and the difference between domestic GBDM compliance and projected GBDM values

So far, the analysis has shown that established powers host more women ambassadors than emerging powers and that this result is driven by dependence. The notion that dependent countries strategically dispatch ambassadors to match the host’s GBDM values does not take into account the extent to which such strategizing requires dependent sending countries to diverge from their own GBDM values. Since domestic GBDM values can influence the extent of strategic value projection, we now incorporate domestic GBDM values into our empirical approach. Our argument, as encapsulated in H3, is that relative to what their domestic GBDM values would predict, dependent countries will send out more women ambassadors to established powers and fewer women ambassadors to rising powers. We test this notion in two ways.

Our first approach is direct. We begin by proxying a country’s domestic GBDM value via its percentage of women legislators. If a country were to externally project its domestic GBDM value, we would expect the difference between the percentage of women ambassadors it dispatches and its percentage of women legislators to be close to zero. H3 predicts that dependent countries will “over-project” their domestic GBDM value to established powers and “under-project” their domestic GBDM value to rising powers. Table 7 reports linear regression results. For the bottom half of the panel, the dependent variable is the difference between the percentage of women ambassadors sent by a country to RIC countries and its percentage of women legislators. We have two key independent variables: dichotomous variables for HIPC and high-income countries (the latter including both OECD and NOC countries), with the rest of the world as comparator. Columns (2)–(6) report results after introducing controls similar to those used in Table 5.¹⁶ In all columns, the HIPC coefficient is negative and significant, while the high-income country coefficient is insignificant. Thus, relative to their women legislator levels, HIPC countries do indeed send fewer women ambassadors to RIC countries. For high-income countries, on the other hand, there is no significant gap between the two.¹⁷

Table 7.

Women ambassadors hosted by G5 and RIC countries, relative to domestic GBDM compliance.

	Sending country type	(1)	(2)	(3)	(4)	(5)	(6)
G5 hosts	HIPC	5.66	3.76	3.75	3.04	2.77	4.20
		(4.19)	(4.87)	(4.89)	(4.81)	(4.74)	(5.11)
	OECD and NOC	−0.08	3.36	3.51	3.02	2.65	2.71
		(3.42)	(3.79)	(4.06)	(4.26)	(4.78)	(4.84)
	Observations	162	156	156	153	153	150
RIC hosts	HIPC	−5.88*(3.35)	−8.64**(3.71)	−8.64**(3.71)	−10.25***(3.79)	−10.34**(4.07)	−11.31***(4.28)
	OECD and NOC	−0.63	1.59	1.48	4.80	2.16	1.93
		(4.19)	(4.79)	(4.72)	(4.32)	(4.39)	(4.41)
	Observations	162	156	156	153	153	150

Notes: Standard errors in parentheses. In the top half of the table, the dependent variable is the difference between the percentage of women ambassadors sent to G5 countries and the sending country’s percentage of women legislators. In the bottom half of the table, the dependent variable is the difference between the percentage of women ambassadors sent to RIC countries and the sending country’s percentage of women legislators. Estimated coefficients for control variables not reported; models (2)–(6) correspond with models (3)–(7) in Table 5. Models incorporate countries that send ambassadors to at least two G5 countries and at least one RIC country.

The top half of Table 7 reports results for the other dependent variable: the difference between the percentage of women ambassadors sent by a country to G5 countries and its percentage of women legislators. Although none of the coefficients attain statistical significance at the 10% level, the HIPC coefficients are higher than the coefficients for high-income countries, which offers (weak) support for H3.

Our second test of H3 is indirect. For each S, we calculate the percentage of women ambassadors dispatched by S. Then, for all sending countries, we regress this percentage against an indicator of S’s domestic GBDM compliance. This technique estimates the average global relationship between women ambassadors dispatched and domestic GBDM compliance. If S’s actual percentage of women ambassadors is more (less) than what is “predicted” by the average relationship, then S’s projected GBDM is greater (lesser) than its domestic GBDM compliance. We estimate a bivariate linear model of the percentage of women ambassadors sent to G5 (RIC) countries against the indicator of domestic GBDM compliance and use the estimated coefficient to construct a variable for the “predicted” percentage of women ambassadors. We then take the difference between the predicted variable and the actual percentage of women ambassadors. This calculation measures the degree to which S’s projected GBDM varies from its domestic GBDM compliance.

Table 8 reports linear regression results. Consider the bottom half, where the dependent variable is the difference between the percentage of women ambassadors that S sends to RIC countries and S’s predicted percentage of women ambassadors based on its percentage of women legislators. As before, the two key dependent variables are for HIPC and high-income countries, with the rest of the world as comparator. Columns (2)–(6) report results after integrating controls similar to those in Table 4.¹⁸ In all columns, the HIPC coefficient is negative, although it reaches significance at the 10% level in only one model, and the size of the high-income coefficient is smaller. The top half of Table 8 reports results for the other dependent variable: the difference between the percentage of women ambassadors sent by S to G5 countries and S’s predicted percentage of women ambassadors based on women legislators. The HIPC coefficients are positive and higher than the coefficients for high-income countries, although they attain significance at the 10% level in only one case.¹⁹

Table 8.

Difference in women ambassadors hosted by G5 and RIC countries using predicted values after accounting for senders’ domestic GBDM values.

	Sending country type	(1)	(2)	(3)	(4)	(5)	(6)
G5 hosts	HIPC	7.66*	6.72	6.78	7.05	6.93	8.22
		(3.95)	(4.67)	(4.70)	(4.76)	(4.74)	(5.12)
	OECD and NOC	3.10	5.05	3.96	3.12	2.74	2.84
		(3.14)	(3.58)	(3.66)	(3.77)	(4.22)	(4.2)
	Observations	162	156	156	153	153	150
RIC hosts	HIPC	−3.40(3.24)	−4.97(3.27)	−4.89(3.21)	−5.28(3.24)	−5.19(3.53)	−6.32*(3.72)
	OECD and NOC	3.31	3.68	2.04	4.92	2.27	2.09
		(3.75)	(4.45)	(4.35)	(3.97)	(3.97)	(4.01)
	Observations	162	156	156	153	153	150

Notes: Standard errors in parentheses. In the top half of the table, the dependent variable is the difference between the actual percentage of women ambassadors sent to G5 hosts and the predicted percentage of women ambassadors. In the bottom half, the dependent variable is the difference between the actual percentage of women ambassadors sent to RIC hosts and the predicted percentage of women ambassadors. Columns (2)–(6) report coefficients for (a) HIPC and (b) OECD and NOC after controlling for different vectors (coefficients of control variables not shown) analogous to those in Table 5.

Conclusions

The results weakly support the contention that, relative to what is predicted by their domestic GBDM compliance, HIPC countries send fewer women ambassadors to RIC than non-HIPC countries and more women ambassadors to G5 than non-HIPC countries. The relatively weak support for H3 suggests that multiple factors are likely to influence countries’ decisions about where to send specific ambassadors. In addition to the gender context of the host state, leaders also take into account the economic and military prestige of a particular posting. Examining ambassadorial appointments in the 50 states with the highest-ranked GDPs, for example, Towns and Niklasson (2016: 18) find that female ambassadors are less likely than male ambassadors to be sent to countries with high economic and military status.

While there has been much contemplation about the implications of rising powers for the contours of international politics, there has been little empirical scrutiny into the ways that rising powers might reshape the global order. Here, we have pursued one avenue of empirical analysis. In recent decades, the global GBDM norm has affected state behavior and helped to increase the representation of women in high-level decision-making positions. Overall, our analysis suggests that rising powers may weaken this global norm and decrease the likelihood that other states will act in accordance with the norm’s standards of appropriate behavior.

We recognize that countries (like the RICs) whose behavior has not historically accorded with normative standards can change. It is possible, in coming years, that their behaviors may come more closely to approximate normative standards. Assume, however, that such change does not transpire. In such a case, we would expect the increasing geopolitical importance of emerging powers to blunt international pressures to comply with the GBDM norm, a development that might be called “normative slackening.” In an environment of normative slackening, we would expect countries to face less international pressure to appoint women ambassadors and, consistent with research on the domestic institutional effects of global gender norms (Bush, 2011; Jacob et al., 2014), to de-emphasize issues of gender parity in legislatures and cabinets. The broader systemic effects of such normative slackening could be significant given the demonstrated associations between women’s political representation, on the one hand, and interstate peace, intrastate stability, human rights protections, and sexual violence, on the other.

Our analysis largely confirms the three hypotheses. It suggests that countries tailor their behavior to the normative environments of their powerful partners. We have shown that countries send fewer women ambassadors to powers that flout the GBDM norm than they do to powers that act in accord with the GBDM norm. In addition, we have demonstrated that dependence amplifies this effect: dependent countries are less likely to challenge the normative environment of powerful states. It is possible that rising powers may become more GBDM-compliant in the future. If such a development does not occur, though, and if GBDM-compliant states lose power to GBDM-non-compliant states, we expect the importance of GBDM as a factor in executives’ decision-making calculus to wane. In such a world, the power of the GBDM norm would diminish. This finding has important implications for the fate of other progressive norms — for example, norms regarding democracy, human rights, minority rights, lesbian, gay, bisexual, and transgender (LGBT) rights, and religious liberty — that established powers tend to recognize and emerging powers tend to flout.

On the relatively narrow (but insufficiently studied) question of the factors that affect ambassadorial appointments, we have argued that gender considerations matter. Specifically, while domestic factors affect the gender composition of a country’s pool of ambassadorial candidates, host-country characteristics help to determine the posts to which ambassadorial appointees are sent. Future research might scrutinize the relative importance of the host-country gender climate, when compared to other host-country conditions such as economic and military power, as a determinant of ambassadorial appointments.

Footnotes

Acknowledgements

The authors would like to thank Kerry Crawford, Jonathan Keller, Dolashree Mysoor, EJIR’s anonymous reviewers, and participants at the following venues for helpful feedback on previous drafts: a seminar at Azim Premji University’s School of Policy and Governance (Bengaluru, India; September 2014); the annual conference of the American Political Science Association (San Francisco, USA; September 2015); and the International Symposium on Development and Governance in the BRICS (Fudan University, Shanghai, China; September 2016).

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Author biographies

Suraj Jacob is Associate Professor in the School for Policy and Governance at Azim Premji University in Bangalore, Karnataka, India. His recent work has appeared in venues such as The Journal of Legislative Studies, Politics & Gender, and Studies in Indian Politics. His research interests include policy regimes, comparative public policy, gender and politics, and the political economy of development.

Melinda Adams is Professor of Political Science at James Madison University in Harrisonburg, Virginia, USA. Her recent work has appeared in venues such as Governance, International Feminist Journal of Politics, and Politics & Gender. Her research interests include gender and politics, women’s political representation, and the politics of development in West Africa.

John A. Scherpereel is Professor of Political Science and Director of the MA program in Political Science with a concentration in European Union Policy Studies at James Madison University in Harrisonburg, Virginia, USA. His recent work has appeared in venues such as European Politics and Society, Journal of Common Market Studies, and West European Politics. His research interests include gender and politics, territorial governance, political communication, and representation.

References

Armijo

Roberts

(2014) The emerging powers and global governance: Why the BRICS matter. In: Looney

(ed.) Handbook of Emerging Economies. New York, NY: Routledge, pp. 503–524.

Baldez

(2004) Elected bodies: The adoption of gender quota laws for legislative candidates in Mexico. Legislative Studies Quarterly 29(2): 231–258.

Beck

Clarke

Groff

et al . (2001) New tools in comparative political economy: The database of political institutions. World Bank Economic Review 15(1): 165–176.

Beckett

(2008) The role of US women diplomats between 1945 and 2004. Electronic Theses, Treatises and Dissertations, Paper 1178.

Bello

(2014) The BRICS: Challengers to the global status quo. Foreign Policy in Focus. Available at: http://fpif.org/brics-challengers-global-status-quo (accessed 11 July 2016).

Bentley

Warner

(eds) (2000) Frontline Diplomacy: The U.S. Foreign Affairs Oral History Collection. Arlington, VA: Association for Diplomatic Studies and Training.

Breslin

(2013) China and the global order: Signaling threat or friendship. International Affairs 89(3): 615–634.

BRICS (Brazil, Russia, India, China, and South Africa) (2015) BRICS in numbers. Available at: http://en.brics2015.ru/infographics/20150301/24260.html (accessed 14 July 2016).

Bush

(2011) International politics and the spread of quotas for women in legislatures. International Organization 65(1): 103–137.

10.

Caprioli

(2000) Gendered conflict. Journal of Peace Research 37(1): 53–68.

11.

Castañeda

(2010) Not ready for prime time: Why including emerging powers at the helm would hurt global governance. Foreign Affairs 89(5): 109–122.

12.

Checkel

(1998) The constructive turn in International Relations theory. World Politics 50(2): 324–348.

13.

Chin

Thakur

(2010) Will China change the rules of global order? The Washington Quarterly 33(4): 119–138.

14.

Cortell

Davis

(2000) Understanding the domestic impact of international norms: A research agenda. International Studies Review 2(1): 65–87.

15.

Culp

(forthcoming) How irresponsible are rising powers? Third World Quarterly.

16.

Desai

(2013) The BRICS are building a challenge to Western economic supremacy. The Guardian, 2 April. Available at: http://www.theguardian.com/commentisfree/2013/apr/02/brics-challenge-western-supremacy (accessed 11 July 2016).

17.

Engstrom

(2012) The future of human rights in a global order of change and continuity. Available at: http://ssrn.com/abstract=2078948 (accessed 14 July 2016).

18.

Gerring

Thacker

Alfaro

(2012) Democracy and human development. Journal of Politics 74(1): 1–17.

19.

Hafner-Burton

(2013) Making Human Rights a Reality. Princeton, NJ: Princeton UP.

20.

Hervieu

(2010) South Africa gains entry to BRIC club. The Guardian, 19 April. Available at: http://www.theguardian.com/world/2011/apr/19/south-africa-joins-bric-club (accessed 14 July 2016).

21.

Htun

Jones

(2002) Engendering the right to participate in decision-making: Electoral quotas and women’s leadership in Latin America. In: Craske

Molyneux

(eds) Gender and the Politics of Rights and Democracy in Latin America. New York, NY: Palgrave, pp. 32–56.

22.

Hurrell

(2007) On Global Order: Power, Values, and the Constitution of International Society. Oxford: Oxford UP.

23.

Ikenberry

(2011) The future of the liberal world order. Foreign Affairs 90(3): 56–68.

24.

Inter-Parliamentary Union (2015) Women in national parliaments. Available at: http://www.ipu.org/wmn-e/classif.htm (accessed 18 August 2015).

25.

Jacob

Adams

Scherpereel

(2013) External dependence and the feminization of foreign affairs. Annual meeting of the Midwest Political Science Association, Chicago, IL, USA.

26.

Jacob

Scherpereel

Adams

(2014) Gender norms and women’s political representation: A global analysis of cabinets, 1979–2009. Governance 27(2): 321–345.

27.

Jonsson

(2008) Global governance: Challenges to diplomatic communication, representation, and recognition. In: Cooper

Hocking

Maley

(eds) Global Governance and Diplomacy: Worlds Apart? New York, NY: Palgrave Macmillan, pp. 29–38.

28.

Jordan

(2010) “Hillary effect” cited for increase in female ambassadors to U.S. The Washington Post, 11 January. Available at: http://www.washingtonpost.com/wp-dyn/content/article/2010/01/10/AR2010011002731.html (accessed 15 July 2016).

29.

Kahler

(2013) Rising powers and global governance: Negotiating change in a resilient status quo. International Affairs 89(3): 711–729.

30.

Keck

Sikkink

(1998) Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca, NY: Cornell UP.

31.

Keck

(2014) Why did BRICS back Russia on Crimea? The Diplomat, 31 March. Available at: http://thediplomat.com/2014/03/why-did-brics-back-russia-on-crimea/ (accessed 15 July 2016).

32.

Keukeleire

Hooijmaaijers

(2014) The BRICS and other emerging power alliances and multilateral organizations in the Asia-Pacific and the global South: Challenges for the European Union and its view on multilateralism. Journal of Common Market Studies 52(3): 582–599.

33.

Koch

Fulton

(2011) In the defense of women: Gender, office holding, and national security policy in established democracies. Journal of Politics 73(1): 1–16.

34.

Krook

True

(2012) Rethinking the life cycles of international norms: The United Nations and the global promotion of gender equality. European Journal of International Relations 18(1): 103–127.

35.

Melander

(2005a) Gender equality and intrastate armed conflict. International Studies Quarterly 49(4): 695–714.

36.

Melander

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

37.

Meyer

Boli

Thomas

et al . (1997) World society and the nation-state. American Journal of Sociology 103(1): 144–181.

38.

Nel

(2010) Redistribution and recognition: What emerging regional powers want. Review of International Studies 36(4): 951–974.

39.

Neumann

(2012) At Home with the Diplomats. Ithaca, NY: Cornell UP.

40.

Nowotny

(2011) Diplomacy and Global Governance: The Diplomatic Service in an Age of Worldwide Interdependence. New Brunswick, NJ: Transaction.

41.

Paxton

Green

Hughes

(2003) Women in parliament 1945–2003: Cross-national dataset. ICPSR 24340-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research, 2008-12-22. Available at: http://doi.org/10.3886/ICPSR24340.v1 (accessed 13 December 2016).

42.

Paxton

Hughes

Green

(2006) The international women’s movement and women’s political representation, 1893–2003. American Sociological Review 71(6): 898–920.

43.

Pieterse

(2011) Global rebalancing: Crisis and the East–South turn. Development and Change 42(1): 22–48.

44.

Pinker

(2011) The Better Angels of Our Nature: Why Violence Has Declined. New York, NY: Viking.

45.

Quota Project (no date) Global database of quotas for women. Available at: www.quotaproject.org (accessed 19 July 2016).

46.

Rahman

(2011) Women in diplomacy: An assessment of British female ambassadors in overcoming gender hierarchy, 1990–2010. Available at: http://www.unc.edu/depts/diplomat/item/2011/0104/comm/rahman_women.html (accessed 15 July 2016).

47.

Towns

Niklasson

(2016) Gender, international status, and ambassador appointments. Foreign Policy Analysis. Available at: https://dx-doi-org.web.bisu.edu.cn/10.1093/fpa/orw039

48.

UNDP (United Nations Development Programme) ( 2014a) Gender inequality index. Available at: http://hdr.undp.org/en/composite/GII (accessed 19 July 2016).

49.

UNDP (2014b) Technical notes. Available at: http://hdr.undp.org/sites/default/files/hdr14_technical_notes.pdf (accessed 13 December 2016).

50.

Wang

French

(2013) Middle range powers in global governance. Third World Quarterly 34(6): 985–999.

51.

World Economic Forum (2015) The global gender gap report. Available at: www3.weforum.org/docs/GGGR2015/cover.pdf (accessed 19 July 2016).

52.

World Values Survey ( 2014) Wave 6 2010–2014, official aggregate v.20150418. Available at: http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp (accessed 15 July 2016).

53.

(2016) China and global governance: Four pathways to cross a threshold. Annual meeting of the International Studies Association, Atlanta, GA, USA.

54.

Zeng

(2014) Women’s political participation in China: Improved or not? Journal of International Women’s Studies 15(1): 136–150.