Despite the recognised importance of gender equality for economic development and post-conflict reconstruction, the tendency to privilege a patriarchal state-centric understanding of the political has consistently caused gender to be silenced from the international relations literature. Challenging the false dichotomy between personal and political, this dissertation explores the gendered effects of imposed economic sanctions on women’s empowerment. The conceptualisation of empowerment as multidimensional, consisting of both bottom-up individual assertiveness and top-down structural inclusion informs the choice of mixed-methods employed. Anchoring its analysis in cross-national time-series data over a twenty-year period, this dissertation finds that the negative effects on governments’ respect for women’s economic and social rights are more pronounced than those on women’s political rights. Extensive and non-human rights sanctions are shown to be respectively worse than limited sanctions imposed for human rights promotion. Providing qualitative evidence from Iraqi women’s experiences sheds the light on more informal aspects, such as shifts in household dynamics and re-traditionalised gender roles. The contribution to the field is two-fold. On one hand, it represents the first effort to calculate the exact change in probabilities of women’s rights violations depending on the severity, objectives and types of sanctions imposed. On the other, it is the first piece to rely on a mixed-method approach to explore the effects of sanctions on women’s rights, allowing to tackle both the public and the private, the formal and the informal realm. We hope that the conclusions presented incentivise policymakers to incorporate women’s rights in their preliminary assessments of the target, privilege targeted over comprehensive sanctions and reduce their duration.
The assumption that women and men are under the same socio-political constraints essentially perpetuates an unfounded myth of gender neutrality (Caprioli, 2004a). As stated in the UN Human Development Report, “no society treats its women as well as its men” (1996, p. 32). In 2016, women represented only 22.8% of all national parliamentarians (UN Women, 2015) and a mere 39.3% of the total labour force worldwide (World Bank, 2016). More often than not, embedded hierarchical structures and ascribed gender roles still cast women to second-class citizens, disadvantaging them in the economic, political and social spheres (Caprioli, 2004a; Enloe, 2000; Tickner, 2001). The curtailed resources, opportunities and agency women enjoy under normal conditions and the different stamp tools of statecraft imprint on women and men make it difficult not to extrapolate the exacerbated gender discrimination women must be confronted to under economic strangulation.
The importance of gender equality for economic development and post-conflict reconstruction has not been understated (Caprioli, 2000). Yet, policymakers’ inclination to deploy economic sanctions is far from declining and no systematic effort has been made to remedy to the negative impact of economic sanctions on women’s livelihoods. The reasons gender has been ignored in the sanctions literature are very much the same reasons it has so long been overlooked in political science. The tendency of the discipline to privilege a top-down patriarchal state-centric understanding of the political has consistently bolstered the myth that gender issues are detached from world politics and perpetuated a false dichotomy between the private and the public. This dissertation seeks to unmask the gendered personal politics of economic sanctions by analysing the extent to which the imposition of economic sanctions impacts women’s empowerment.
The contentious nature of the terms ‘economic sanctions’ and ‘women’s empowerment’ deserves clarification. On one hand, definitions of economic sanctions have tended to be peculiar and muddy, to the extent that “that there is much to be said by avoiding [the term] altogether” (Baldwin, 1985, p.36). This dissertation will understand economic sanctions as a tool of statecraft “distinct from economic warfare and trade wars”, which works “to lower the aggregate economic welfare of a target state” (Pape, 1997, p.93) and whose primary yet not only goal is punitively or preventively coercing the target government into altering its political policies. This dissertation also understands economic sanctions as an umbrella term for comprehensive and targeted sanctions, with no restrictions as to the goal of the sanctions or the size of the sanctioning coalition. On the other, women’s empowerment has been difficult to define, so much that for some, the value of the concept lies in its fuzziness (Kabeer, 1999). This dissertation will borrow Kabeer’s definition of empowerment as “the process by which those who have been denied the ability to make strategic life choices acquire such an ability” (1999, p.435). Empowerment will also be understood as three-dimensional, shaped by interrelated resources, agency and achievements, and consisting of both bottom-up individual assertiveness and top-down structural inclusion in all three economic, political and social spheres.
This conceptualisation of women’s empowerment guided the choice of mixed research methods used. Whereas only a quantitative study can systematically examine the impact of economic sanctions on state-induced women’s rights legislations on a large-N sample, qualitative evidence is required to contextualise the findings and tackle on-the-ground unquantifiable norms and obstacles to women’s empowerment. Anchoring its analysis in both cross-national time series data covering a period of twenty years and a case study of Iraq, this dissertation argues that the burden of economic sanctions is overwhelmingly felt by women in all three economic, political and social spheres. Only a mixed-method approach, allowing for one to offset the drawbacks of the other, can paint an accurate picture of the effects of sanctions on women’s rights.
The contribution to the literature is two-fold. First, whereas earlier attempts (Drury and Peksen, 2014) are limited to evaluating whether economic sanctions yielded a significant impact on women’s rights, the details of the how much remain unexplored. This dissertation represents the first effort to estimate the exact probabilities of women’s rights violations depending on the severity, objectives and types of sanctions imposed. Relying on three-category ordinal instead of binary sanctions variables allows us to calculate the change in likelihood of economic, political and social rights abuses according to whether the sanctions are limited, extensive, multilateral and aimed at promoting human rights. Second, despite the consensus in the literature that women’s empowerment programmes are best monitored through mixed methods, no work has yet sought to blend quantitative and qualitative methods to analyse the impact of a nonviolent tool of statecraft on women’s rights.
This dissertation proceeds as follows. The next section will first outline the key debates critical to economic sanctions and women’s empowerment in the existing literature, thus highlighting the gap in knowledge this contribution seeks to address. Next, a theoretical framework is laid out, from which are derived a number of hypotheses about the expected impact of different types of sanctions on women’s economic, political and social rights, tailored around the severity, duration, objectives and senders of sanctions. After formulating a suitable methodology and discussing the variables used, the models are empirically tested using a large-N sample and the findings are interpreted. This dissertation finds that while economic sanctions have a detrimental impact on women’s empowerment at large, the effects on women’s economic and social rights are more pronounced than those on women’s political rights. The likelihood of women’s rights violations is found to be positively correlated with the severity of sanctions. Non-human rights sanctions are shown to be worse for women’s economic, political and social rights than those imposed for human rights promotion. While multilateral sanctions lead to a higher likelihood of limited women’s political and social rights than unilateral sanctions, the latter seem more detrimental for women’s economic rights. A quantitative analysis alone only tells half the story. Because a case study is needed to shed the light on more informal aspects of women’s empowerment, such as shifts in household dynamics and societal norms, this dissertation incorporates qualitative evidence from Iraq under sanctions. Since all intricate analyses leave much room for error, the following section discusses a number of limitations this dissertation runs into. Lastly, the conclusion summarises the findings, outlines policy implications and suggests areas for further research.
Whereas policymakers continue to praise economic sanctions as a better alternative to unnecessary war (Pape, 1997), debates around the effectiveness, opportunity cost and desirability of deploying sanctions continue to plague academic circles.
Most of the existing literature on economic sanctions focuses on their effectiveness. Arguing that the pessimism is unwarranted, a number of authors contend that conclusions about the ineffectiveness of sanctions are pre-determined by a selection bias (Olson, 1979). Building the argument exclusively on a combination of prominent disastrous cases (Blake and Kemm, 2006; Lektzian and Souva, 2007; Parker, 2000) leads to successes being underrated and failures overblown (Rogers, 1996). In an unprecedented move towards a systematic large-N study of economic sanctions, “the bedrock study on the effectiveness of economic sanctions” (Pape, 1997, p.91) by Hufbauer, Schott and Elliot (1990) found that more than one third of sanctions imposed since 1914 have been met with successes (Elliott, 1998).
Yet, and in opposition to policy makers’ intuition, the prevailing opinion in academic circles remains sceptical (Blanchard and Ripsman, 1999; Morgan and Schwebach, 1997). While Pape (1997; 1998) and Drury (1998) questioned the methodological and conceptual robustness of the HSE quantitative study, Haass (1998) and Hovi et al. (2005) pointed at sanctions fatigue, decreasing international compliance, and regime diversion counterstrategies as obstacles to their successful implementation.
Stemming from the premise that “sanctions are neither impotent nor are they a panacea” (Kirshner, 1997, p.32), a body of work has rejected the counterproductive binary debate, focusing instead on when and under what circumstances sanctions are likely to achieve intended aims. Whereas the instrumental school of thought ties severity of economic deprivation to ability to spur political change (Drury, 1998; Galtung, 1967; Kaempfer and Lowenberg, 1988), a number of scholars question that predictive link, stating instead that specific international conditions are needed to increase the likelihood of success (Blanchard and Ripsman, 1999; Kaempfer et al, 2003; Kirschner, 1997). Arguing that seeing international bargaining in isolation from domestic dynamics overlooks that “sanction episodes are examples of two-level games” (Morgan and Schwebach, 1995, p.4), yet another strand links success of sanctions to the target’s domestic institutional structure (Allen, 2005; 2008; Brooks, 2002; Morgan and Schwebach, 1997).
Dissenting from the orthodox dichotomous measure of success, authors argued that the reason behind this academic quagmire is scholars asking the wrong questions (Baldwin, 2000). In that sense, what seems to be a puzzling stance of policymakers “lie[s] less in the ignorance of government officials than in the naivete of the research” (Lindsay, 1986, p.153). Studies have also focused on the reason behind the lack of consensus. While some took issue with methodological problems (Baldwin, 2000; Drezner, 2003), others raised conceptual and measurement issues. Definitions of success abound, yet “the state of disagreement […] remains somewhat of an embarrassment” (Marinov, 2005, p.565). Whereas Barber recognises the diversity of possible benchmarks sanctions should be evaluated against (1979), Parker advocates a narrow conception of success, excluding “any symbolic, expressive or domestic political function” (2000, p.239).
While prior to the 1990s, studies about effectiveness represented the bulk of the literature, the widely reported negative repercussions of economic strangulation in Haiti and Iraq have led to a shift in academics’ attention. Acknowledging that the international affects the personal and that sanctions seem to bring “assured civilian pain versus doubtful political gain” (Weiss, 1999, p.501), a number of studies have analysed the humanitarian consequences of economic sanctions (Cortright and Lopez, 1997; Gordon, 1999; 2011; Haass, 1997).
The literature on the humanitarian impact of economic sanctions is split between quantitative analyses and case studies. The former includes systematic studies on the impact of sanctions on public health (Allen and Lektzian, 2013), media openness (Peksen, 2010), property rights (Peksen, 2014), income inequality (Afesorgbor, and Mahadevan, 2016), GDP growth (Neuenkirch and Neumeier, 2015) poverty (Neuenkirch and Neumeier, 2016) and human rights (Drury and Li, 2006; Peksen, 2009; Peksen and Drury, 2010). The effects of different types of sanctions – relating to their goal, size of the sanctioning coalition and extensiveness – have also been explored (Wood, 2008; Von Soest and Wahman, 2015). In parallel, a number of case studies look at the impact of sanctions on medicine and healthcare, life expectancy, toddler mortality, income, food and clean water availability, education and literacy levels, providing evidence from Rhodesia (Galtung, 1967), South Africa (Doxey, 1972), Haiti (Gibbons and Garfield, 1999), Cuba (Garfield and Santana, 1997; Gordon, 2016), Burma (Seekins, 2005), Burundi (Hoskins and Nutt, 1997), former Yugoslavia (Devin and Dashti-Gibson, 1997),Iran and Syria (Moret, 2015), and Iraq (Ali and Shah, 2000; Al-Nasrawi, 2001; CESR, 1996; Gordon, 2009).
The negative humanitarian consequences of economic sanctions have not gone uncontested. While Pattison argues that sanctions are morally preferable to intervention or government apathy (2015), Lopez contends that the conclusions of exceptional cases should not overshadow cases where sanctions have been successful without inflicting terrible humanitarian damage (1999).
In spite of its comprehensiveness and heterogeneity, the literature on the consequences of sanctions remains inherently state-centric (Drury and Peksen, 2014; Buck et al, 1998). With the exception of a few case studies, most of the literature fails to disaggregate the effects of sanctions by social groups and very few efforts have been made to systematically examine the humanitarian consequences on the youth, the elderly or women in a large-N study.
The local impact of different international dynamics such as neoliberal globalisation or militarisation on women’s status has been highlighted in the literature. Whereas economic globalisation and membership in international organisations have largely been found to improve women’s conditions (Gray et al, 2006; Richards and Gelleny, 2007), military interventions and civil war are likely to exacerbate structural and cultural violence against women (Byrne, 1996; Caprioli and Douglass, 2008; Peksen, 2011), increase sex trafficking and systematic rape (Ashford, 2008; Hynes, 2004) and redefine gender identities (Cooke and Wilcott, 1993; Enloe, 2000; Hagemann et al, 2010; Sjoberg and Via, 2010; Tickner, 2001). Yet, feminist scholars have largely neglected the consequences of nonviolent foreign policy tools.
The silence about economic coercion in the feminist scholarship has been mirrored by a similar invisibility of women in the international relations literature on sanctions (Buck et al, 1998). Despite the dynamic gender literature in international relations and for all the efforts allocated to exploring the workings of economic sanctions, most “analyses tend to have been either gender-neutral or gender-blind” (ibid, p.69). The reasons gender has been ignored in the sanctions literature are very much the same reasons it has so long been overlooked in international relations. The tendency of political science to focus on aggregated units, privilege a top-down patriarchal understanding of the political and disregard what is seen as the private realm has consistently steered the focus away from gender roles (ibid; see also Enloe, 2000; Tickner, 2001).
It would be unfounded to assert that nothing has been written on the impact of sanctions on women’s rights. A few case studies dedicate their analyses to the gendered impacts of sanction episodes. Providing evidence from Burma, Haiti, Syria, Iran, Cuba and Iraq, authors pinpoint the rise in prostitution (Seekins, 2005), decline in women employment (Gibbons and Garfield, 1999), reduction of micro-credit schemes (Moret, 2015), decline in female literacy and income (Hoskins and Nutt, 1997), shifting constructions of femininity and an increase in domestic violence and state-induced societal conservatism (Al Ali, 2005; Al Ali and Pratt, 2009a; Al Jawaheri, 2008; Buck et al, 1998). Narrative analyses are valuable insofar as they privilege depth over breadth and document qualitative aspects of women’s disempowerment. Yet, individual case studies cannot confront the assumptions to a large N-test, thus necessarily impeding their analytical and replicability potential.
Drury and Peksen are the only ones to have conducted a systematic cross-national assessment of the correlation between economic sanctions and women’s rights (2014). However, their study leaves us with more questions than answers. First, by relying on binary sanctions variables, the authors do not include the severity of sanctions in their model. Second, they conclude that human rights sanctions do not have the same negative impact on women’s rights than non-human rights sanctions. Yet, they fail to estimate by how much the likelihood of women’s economic, political and social rights violations would increase under human rights sanctions, or if the target was under limited or comprehensive sanctions. Third, they include the objective and the size of the sanctioning coalition in their explanatory variables. However, they fail to anchor their rationale for doing so in a theoretical framework about the expected possible effects of different types of sanctions on different types of women’s rights. Lastly, without the insight of a case-study, both their hypotheses and findings remain locked in a high-tower of theoretical abstraction. Statistical methods are “simple windows on complex realities” (Kabeer, 1999, p.447). Their analysis might have explored top-down legislations promoting women’s rights, but more informal aspects of women’s empowerment, such as intrahousehold dynamics and entrenched norms, remain hidden.
Although exhaustively retracing definitions of women’s empowerment is beyond the scope, highlighting different conceptualisations of the term is necessary to understand the issues related to its measurement. Women’s empowerment has repeatedly been pushed to the top of development agendas and several entities have been tasked to design a methodology of empowerment (Alsop and Heinsohn, 2005; Giele and Smock, 1977). Yet it remains used interchangeably with words like power (Beegle et al, 2001), agency (Gage, 1995) or autonomy (Dyson and Moore, 1983), and no consensus has been reached with regards to its definition, let alone its operationalisation. Whereas Sen and Grown (1987) are amongst those who emphasise the individual as the vehicle of bottom-up empowerment, others see top-down “social inclusion in institutions as the key pathway to empowerment” (Malhotra et al, 2002, p.4; Bennett, 2002). While Giele and Smock (1977) propose a six-pillar framework including political involvement, work conditions, family legislations, education rights, health services and cultural image, perhaps the most influential conceptualisation of women’s empowerment lies in Kabeer’s three-folded framework (1999). The equal weight she puts on interrelated resources as pre-conditions, agency as ability and achievements as outcome bridges the gap between potential and “actualised choice” (ibid, p.443). However, the overarching wisdom is that women’s empowerment should be conceptualised as a process (Bennett, 2002; Beteta, 2006; Kabeer, 1999), occurring in a holistic bubble moulded by a range of spheres (Charmes and Wieringa, 2003). Empowerment should be measured at different units of analysis, and be about agency and structure, the formal and the informal sector, the public and the private.
Conventional feminist scholarship has typically been reluctant to use quantitative methodology in gender studies (Caprioli, 2004b; Drury and Peksen, 2014). First, both the possibility for indicators to capture less institutionalised aspects and the adequacy of specific gender indicators such as the Gender Development Index and the Gender Empowerment Measure have been debated (Moghadam and Senftova, 2005; Schuler, 2006). Because improvements in state-induced legislations on women’s rights do not consider on-the-ground impediments to their implementation, some scholars contend that empowerment can only be understood qualitatively (Sen, 1993). Second, the conventional feminists’ scepticism towards quantitative methods can be explained by their understanding of quantitative research as a positivist methodology inherently rooted in international relations and pushing a masculine agenda (Caprioli, 2004b).
This dissertation seeks to address two gaps, one in content and the other in methodology.
First, for all the literature economic sanctions have generated, there has been no comprehensive large-N study assessing the likelihood of women’s economic, political and social rights violations depending on severity, objective and sender of sanctions. The silence of the literature on the impact of economic sanctions on women’s empowerment should not detract from its relevance. The persistence of cultural and structural gender inequality makes women more susceptible to be disadvantaged by tools of international statecraft (Peksen, 2011; Caprioli and Douglass, 2008). Any theoretical framework which does not include the gender-specific impact of sanctions is bound to remain incomplete and lead to distorted assessments of the desirability of deploying economic sanctions (Ashford, 2008; Gray et al, 2006).
Second, with the literature on economic sanctions and women’s empowerment devoted entirely to either atheoretical and non-generalisable qualitative case studies or empirical testing not grounded enough in empirics and ethnographic examples (Dashti-Gibson et al, 1997, p.608), there has been no initiative to blend quantitative and qualitative approaches into one piece of work. Whereas a large-N study can identify trends and patterns not made explicit by narrative evidence (Apodaca, 1998), a case study can add “qualitative flesh on quantitative bones” (Tarrow, 2010, p.97). Empowerment in one sphere, namely legal and political rights, does not mean empowerment in another, namely societal norms and household bargaining power. Drawing on Caprioli’s defence of mixed methods research in gender studies, this dissertation argues that it is only through a combination of theory, details and validity that research is advanced (2004b). Both methods are tied by “complementary distinctiveness” (Lieberman, 2005, p.450), used to explain different things. The pronounced methodological dichotomy and Popper’s “myth of framework” are unfounded (Caprioli, 2004b, p.256). Given the holistic conception of women’s empowerment, relying solely on one approach only tells one part of the story. Only a mixed-method approach to studying the toll economic sanctions take on women’s empowerment can accumulate the benefits of both, offset the drawbacks, and paint a gender-sensitive picture of the impact of sanctions on both top-down given potential choice and bottom-up actualised achievements.
The assumption that women and men are both under the same socio-political constraints perpetuates an unfounded myth of gender neutrality (Caprioli, 2004a). The differential treatment, resources and agency of women and men in normal conditions makes it easy to extrapolate the exacerbated gender discrimination under economic strangulation (Buck et al, 1998). Theoretically, the brushstrokes painted by economic sanctions on the women’s empowerment canvas are not so clear. In times of economic hardships, the increasing difficulty to make ends meet might lead women to enter the workforce. With women holding more economic rights, both household bargaining dynamics and societal perceptions of gender relations can change, leading in the long-term to extended political and legal rights in the public sphere (Bertocchi, 2011; Doepke et al, 2012; Geddes and Lueck, 2002).
However, despite the potential opportunities economic sanctions present theoretically, the actual picture is bleaker. By pushing the target state into “coping” mode – “in which individuals and families engage in order to ensure their everyday survival” (Al Ali and Pratt, 2009b, p.22) – sanctions distort everyday economic, political and social patterns and exacerbate inflation, instability and violence. Notwithstanding the evident paradox, economic sanctions largely end up deteriorating women’s rights instead of promoting them (Buck et al, 1998; Peksen and Drury, 2014). The anticipated effects of sanctions on women’s economic, political and social rights cannot be understood in isolation. The three sections below, although distinct for the sake of clarity, should thus be seen as interrelated and mutually reinforcing.
On average, economic sanctions cost the target state 3 percent of its GDP annually (Hufbauer et al, 2007). Inflation skyrockets, unemployment soars and shadow illicit activities expand (Galtung, 1967; Weiss et al, 1997). Theoretically, a government under sanctions might respond by intervening in markets and redistributing assets to appease citizens and reduce their incentives to revolt (Gutmann et al, 2016). However, the “logic of siege” dictates that more often, the leadership, the military and the political elite end up monopolising the scare resources, while redirecting the costs onto citizens (Gordon, 1999, p.125). The result is that those who are most on the margins of decision-making and least able to weather the consequences are those who bear the biggest burden (ibid; Cortright and Lopez, 1999, Gutmann et al, 2016). More specifically, due to both ingrained societal discrimination against women in the workforce and a recession-induced widening supply-demand gap in the labour market, women are often the first ones to be laid off, and those who keep their jobs are more likely to be demoted than men (Drury and Peksen, 2014). Moreover, the public sector, comprising the majority of employed women, is the one where salaries typically drop the most. While women make up 80 percent of employees in export-oriented industries, those sectors are the ones most adversely impacted and the first forced to shut down. Lastly, the government’s inability to afford childcare benefits and policies supportive to working women makes it cheaper for many to leave their full-time job (Drury and Peksen, 2014).
H1: The imposition of economic sanctions is likely to have an adverse impact on women’s economic rights.
Following from this logic, and assuming that the objective for which sanctions are imposed and their severity are independent, this dissertation also hypothesises that sanctions with the goal of promoting human rights and those with non-human rights related goals are likely to have the same negative impact on women’s economic rights.
H2: Human rights sanctions and non-human rights sanctions are likely to have similar negative effects on women’s economic rights.
With regards to the potential difference in impact depending on the size of the sanctioning coalition, whether multilateral or unilateral sanctions are expected to have a higher effect on women’s economic rights has not been looked at. On one hand, insofar as they affect a greater share of the international market, narrow the scope of alternative suppliers and reduce possibilities to use counterstrategies, there is support for the hypothesis that multilateral sanctions are more detrimental to women’s rights than unilateral ones (Bapat and Morgan, 2009; Martin, 1992). On the other, Kaempfer et al argue that success “is negatively correlated with the size of the sanctioner” (2003, p.519). Because they rely on a larger sanctioning coalition, the consequences of multilateral sanctions can be hampered by the lack of cooperation and individual temptations to free-ride among the sanctioning members (Kaempfer and Lowenberg, 1999; Miers and Morgan, 2002). In contrast, unilateral sanctions severing a strategic relationship between countries with initially dense trade linkages can inflict stronger damage than symbolic ones imposed by an international organisation with limited economic links to the target (ibid). According to Bapat and Morgan, the differential impacts of one or the other can also depend on the number of issues and actors involved (2009). This leads to two competing hypotheses.
H3a: Multilateral sanctions are likely to have a worse impact than unilateral sanctions on women’s economic rights.
H3b: Unilateral sanctions are likely to have a worse impact than multilateral sanctions on women’s economic rights.
Following from the instrumental theory of sanctions, according to which the costs of sanctions on the target’s economy are correlated with their duration (Bolks and Al Sowayel, 2000; Dashti-Gibson et al, 1997), we expect the asphyxiating effect of sanctions on women’s economic rights to worsen the longer economic sanctions are in place.
H4: Economic sanctions are likely to have a negative impact on women’s economic rights over time, albeit with a diminishing marginal rate of effect.
It is unclear which way economic sanctions steer women’s political rights. Since many sanctions are imposed to promote a democratic transition, it can go either way (Gutmann and al, 2016). Although the impact of sanctions on women’s political rights has not been explored at length, their impact on political rights at large has. While some argued that sanctions can weaken the coercive apparatus of the target and international pressure is likely to inflict accountability costs forcing the regime to concede political liberties (Galtung, 1967; Kirshner, 1997; Mack and Khan 2000), others have found no significant negative relationship on political rights (Gutmann et al, 2016). Others yet made a distinction depending on regime type, arguing that violations of political rights were more likely in democracies than in autocracies, because the latter offer less avenues for revolt than the former (Allen, 2008). However, the conventional wisdom remains that economic sanctions are counterproductive, insofar as they increase incentives to breach political rights (Oechslin, 2014). Instead of leading to an equal redistribution of the remaining funds, sanctions usually prompt the leadership to stiffen its grip on power and deflect the costs onto citizens, thereby creating more socio-economic grievances while solidifying its coercive apparatus (Peksen and Drury, 2009). Following from Gurr’s “relative deprivation” theory (1970), collective motives increase the incentives to dissent, either with violent or non-violent means. To crush resisters’ perception of regime fragility and reduce others’ incentives to join, the regime is likely to crack down on protesters (Drury and Li, 2006). Framing the sanctions as a foreign threat to the nationalist core makes it easier for the state to justify a state of emergency and paint protesters as traitors and top-down repression as a survival necessity (Peksen, 2009; Peksen and Drury, 2010).
Extending the effects on women’s empowerment, it is not difficult to see women’s political rights becoming scapegoats for state-led repression. As issues of economic stability and national survival take priority over women empowerment policies, women’s rights are the first ones to be neglected and the last ones to be brought back to the agenda, Moreover, integration in the international system has been considered crucial in improving governments’ compliance with international human rights norms and advances in gender equality have often been driven by transnational feminist movements. In the long-term, economic isolation is likely to expose sanctioned countries less to empowerment advances and lower their motivation to comply with international norms (Peksen, 2009).
H5a: Economic sanctions are likely to have a negative effect on women’s political rights.
H5b: Economic sanctions are likely to have a positive effect on women’s political rights.
Whether the impact on women’s political rights differs significantly depending on the goal of the sanctions imposed has not been explored either. While one study found that the impact of human-rights sanctions and non-democratic sanctions was not significantly different (Peksen and Drury, 2010), others argued that sanctions aimed at promoting better human rights were more detrimental than non-human rights sanctions (Peksen, 2009). Holding their severity constant, sanctions imposed with the goal to punish human rights violations might offer the target state the incentive to promote democracy (Von Soest and Wahman, 2015), of which women’s empowerment is a good proxy. Alternatively, if the targeted leadership frames the sanctions as an existential threat to its sovereignty, the human-rights sanctions might lead it to enhance its repressive apparatus to avoid appearing conciliatory (Peksen, 2009). The human rights sanctions regime might also counterproductively help shape a backlash against those trying to push for women’s political rights, as gender equality promotion can be equated with imperialism and foreign stooges.
H6a: Human rights and non-human rights sanctions are likely to have a similar impact on women’s political rights.
H6b: Human rights sanctions are likely to have a positive impact on women’s political rights.
H6c: Human rights sanctions are likely to have worse effects on women’s political rights than non-human rights sanctions.
With regards to the size of the sanctioning coalition, it is more likely that multilateral sanctions would have a worse effect on women’s political rights than unilateral ones. Following from mass mobilisation theory, the awareness that a larger number of states are opposing the regime would enhance the people’s perception of regime fragility, thus increasing incentives and lowering perceived costs to join a resistance movement (Nepstad, 2013). To avoid appearing weak, the regime would likely backfire with repression. Because they are seen as more discretionary, expendable and volatile, women’s political rights would be the first ones to be suppressed and the last ones on the government’s priority list.
H7: Multilateral sanctions are likely to have worse effects on women’s political rights than unilateral sanctions.
Insofar as sticks without carrots can only be efficient for a certain period, an authoritarian regime deprived from its financial resources and unable to economically co-opt its citizens into subservience will be forced to resort to political concessions (Gause III and Yom, 2012; Wood, 2008). Therefore, it is expected that over time, the political leadership will adapt and women’s political rights might not be negatively affected by the duration of economic sanctions. Moreover, the longer economic sanctions are in place, the more governments might feel compelled to enact top-down women’s rights political reforms, albeit more cosmetic than transformative.
H8: Economic sanctions are likely to have no or a positive impact on women’s political rights over time.
Deteriorating economic conditions confounded with political domestic instabilities create an environment conducive to increased societal turmoil and gendered violence. On one hand, the ‘criminalisation’ of the economy (Peterson, 2009) leads to an increase in the normalisation of gender-based crimes, honour killings and systematic gang rape (Caprioli and Douglass, 2008), feeding into pervading public insecurity and deterring women’s access to education or freedom of movement. On the other, as the strain on the target state increases, women’s rights are likely to become relegated as optional, making the authorities less likely to guarantee women’s social rights protection or punish those who violate them (Drury and Peksen, 2014). As nationalist power struggles ensue, citizenship is often redefined along socio-economic and political lines, constructed around hierarchical patriarchal structures and sustaining norms permissive of gender-based discrimination (Al Ali and Pratt, 2009b, Kamp, 2009), thus reducing opportunities for advocating for women’s empowerment. On a more individual level, soaring unemployment and inflation are likely to strengthen male-held perceptions of impotent manhood, for which men will compensate with increased domestic aggression and household bargaining power. That, combined with the awareness of increased gender-based violence in the streets and the strengthened ideas of men as protectors, is likely to lead men to tighten their grip over women in the day-to-day life (Caprioli, 2004a; Caprioli and Douglass, 2008). The reduction in the number of women in the workforce will contribute to a re-traditionalisation of gender roles.
H9: Economic sanctions are likely to have a negative impact on women’s social rights.
Whether human rights and non-human rights sanctions yield similar effects on women’s social rights is thorny. While strategically working towards women’s social empowerment to get rid of human-rights sanctions is a possibility, this dissertation hypothesises that the deteriorating economic conditions and shifting governmental priorities will likely offset the authorities’ incentives to advance women’s empowerment.
H10: Human-rights and non-human rights sanctions are likely to have the similar negative effects on women’s social rights.
The potentially different impact of multilateral and unilateral sanctions has not been verified either. Here, the same reasoning related to their impact on economic and political rights applies. Essentially, whether multilateral sanctions will yield worse effects than unilateral sanctions will depend on their cost to the target (measured in terms of sanctioning coalition cohesiveness, number of members and initial trade flows) and the regime response to resistance movements (Bapat and Morgan, 2009; Kaempfer and Lowenberg, 1999). However, state-led attitudes to women’s empowerment are no guarantee of altered societal attitudes. The adverse economic effects of sanctions might lead to a deterioration of women’s social rights, regardless of the type of sanctions imposed.
H11a: Multilateral sanctions are likely to have a worse impact than unilateral sanctions on women’s social rights.
H11b: Unilateral sanctions are likely to have a worse impact than multilateral sanctions on women’s social rights.
H11c: Multilateral and unilateral sanctions are likely to have the same negative impact on women’s social rights.
Women’s economic rights are negatively correlated with the duration of sanctions and economic deterioration is associated with a re-traditionalisation of gender norms, which are not easy to shift once embedded.
H12: Economic sanctions are likely to have a negative impact on women’s social rights over time, albeit with a diminishing marginal rate of effect.
Lastly, with the exception of Carneiro and Apolinario (2016) who found no significant difference between the impact of comprehensive and smart sanctions, embargoes targeting the entire economy are generally found to have greater corrosive effects than targeted sanctions, consisting of trade sanctions and elites’ asset freezes (Peksen and Drury, 2010). Following from the instrumental perspective, this dissertation hypothesises that the impact of economic sanctions is tied to their severity and costs.
H13: Comprehensive sanctions are likely to have worse effects on women’s economic, political and social rights than limited sanctions.
A point of caution is necessary. Although this section has separated sanctions’ expected effects on women’s rights by type and goal, the impact of each might be influenced by the correlation between severity, objective and size of the coalition. For example, the fact that multilateral sanctions tend to be extensive or that unilateral sanctions are more likely to be non-human rights sanctions (Peksen, 2009; 2010; Wood, 2008) can alter the expected consequences of each.
To empirically test the hypotheses outlined, this dissertation relies on cross-sectional, time-series data, spanning 145 countries and twenty years, where the unit of analysis is country/year. The dataset used has been obtained by merging the dataset used by Peksen (2009) – itself a combination of Hufbauer, Schott and Elliott’s (1990) and the Threat of Implementation of Economic Sanctions datasets (Morgan et al, 2006) – with the Cingranelli and Richards (CIRI) Human Rights database (2006). The time span (1981-2000) has been determined by the accessibility of the data, as the overlap period of the two datasets. Instead of restricting the sample only to countries having been under economic sanctions, the analysis includes a global sample to decrease the likelihood of selection bias (Levy, 2008) . Since their observations only begin in 1990, the states emerging from the dissolution of Yugoslavia and the Soviet Union have also been incorporated (Apodaca, 1998). This section will first discuss the dependent, independent and control variables used in the models and then clarify a few methodological questions. The summary statistics of the variables presented are shown in Table 1 below.
This study uses the Women’s Economic Rights, Women’s Political Rights and Women’s Social Rights variables from the CIRI Human Rights Database (2006). Beside assimilating all three dimensions, these indicators paint a top-down picture of women’s rights, by quantifying governments’ degree of tolerance for women’s rights abuses. Women’s Economic Rights accounts for “equal pay for equal work, free choice of profession or employment without the need to obtain a husband’s or male relative’s consent, […] equality in hiring and promotion practices, job security […], non-discrimination by employers, the right to be free from sexual harassment in the workplace, the right to work at night […] and in occupations classified as dangerous, […] in the military and the police force” (Cingranelli and Richards, 2010, p.423-424).
Women’s Political Rights includes include “the right to vote, the right to run for political office, the right to hold elected and appointed government positions, the right to join political parties, and the right to petition government officials” (ibid, p.423). Women’s Social Rights measures “equal inheritance; the right to enter into marriage on a basis of equality with men, the right to travel abroad, the right to obtain a passport, to confer citizenship to children or a husband; to initiate a divorce; the right to own […] property brought into marriage; the right to participate in social, cultural, and community activities; the right to an education, the freedom to choose a residence/domicile, freedom from female genital mutilation of children and adults without their consent, and freedom from forced sterilisation” (ibid, p.424).
All three are four-category ordinal variables, ranging from 0 to 3, where 0 indicates the lowest (severely curtailed by law and practice) and 3 the highest (guaranteed and enforced by law) level of respect (Cingranelli and Richards, 2013).
The decision to use the CIRI indicators owes to more theoretical and methodological considerations and lies in their comprehensiveness, operationalisation and reliability. First, measuring women’s empowerment through top-down changes in state legislations has found supporters in the feminist literature. While Moghadam and Senftova (2005) argue that “empowerment comes about through legal reform and public policy changes” (p.391), others contend that empowerment is partly reflected through “the removal of institutional barriers” (Malhotra et al, 2002, p.4) and an “adequate legal and regulatory framework” (Beteta, 2006, p.234). Although structural gender-sensitive policies do not guarantee assertiveness at the individual level (Kabeer, 1999), legally transforming the rules is necessary to both bolster improvements over time and tilt women’s cost-benefit calculations (Malhotra et al, 2002). Second, women’s empowerment is a process, best grasped when analysed through time-series data (Bennett, 2002; Malhotra et al, 2002). The availability of the CIRI indicators in country-year format fits this purpose and allows to follow the evolution of women’s rights across different points in time. Third, the indicators have been developed with a mixed-method approach, embedding qualitative content analysis in the coding mechanism (Cingranelli and Richards, 2010). Fourth, the operationalisation of the indicators as four-category ordinal variables is in accordance with the nature of the variables: neither binary nor perfect, but leaving room for potential measurement errors (ibid). Fifth, the reliability of the indicators is verified and echoed in their close-to-perfect “Krippendorff’s r-bar intercoder reliability” (ibid, p.409).
Confirming the theoretical expectations, the variables measuring women’s economic, political and social rights are not entirely statistically independent. Generating the correlation matrix (shown in table 2 below) reveals that of the combinations possible, only economic and social rights are strongly correlated. However, since those two variables are neither explanatory nor included in the same model, multicollinearity is not a problem.
The economic sanctions variables have been extricated from the dataset used by Peksen (2009), itself merging data from both the TIES (Morgan et al, 2006) and the HSE databases (1990) and only covers imposed sanctions, not threats thereof. In order to better account for the harshness of sanctions and unlike other studies treating economic sanctions as binary variables, all the sanctions variables are three-category ordinal variables, where for every year, 0 indicates no sanctions, 1 signifies limited sanctions (such as partial export restrictions or asset freezes) and 2 refers to extensive sanctions (Peksen, 2009).
Economic Sanctions indicates the absence of presence of partial or comprehensive sanctions. The models also include variables accounting for the goal, senders and duration of sanctions. While Human Rights Sanctions takes the value of 1 if the country in question is under limited human rights sanctions and the value of 2 if under extensive human rights sanctions, and 0 otherwise, Non-Human Rights Sanctions is coded 1 or 2 if limited or extensive sanctions have a non-human rights related goal. Similarly, Unilateral Sanctions and Multilateral Sanctions are coded 1 or 2 if the country is facing limited or extensive sanctions imposed by individual countries or by the UN or other intergovernmental organisation, respectively. Although binary pairs, generating the correlation coefficients showed the Human Rights Sanctions and the Non-Human Rights Sanctions variables on one hand, and the Multilateral Sanctions and Unilateral Sanctions variables on the other not to be significantly correlated.
Lastly, to account for the long-term effects of economic sanctions, the variable Sanction Duration indicates the number of years a country has been under sanctions. The variable is logged, to control for the curvilinear relationship between the duration of sanctions and the dependent variables, namely that the cumulative negative effect of sanctions will have a diminishing rate of effect and that the slope will be less steep over time (Lewis-Beck et al, 2004).
Drawing on the literature to find strong theoretical justification, the models also incorporate control variables. First, the models include a War dummy variable. The war-gender nexus has been debated in the literature. Militarised conflict has been found to affect women and men differently, and either exacerbate gender-based discrimination and systematic violence (Ashford, 2008; Byrne, 1996; Enloe, 2000), or on the contrary, open opportunities for redefining gender relations and yielding greater opportunities for women. The War variable was created by drawing on the Civil War (0-1) and Interstate War (0-1) indicators created by the Armed Conflict Database of the Peace Research Institute of Oslo (Strand et al, 2005). Assuming that civil war and interstate war yield similar effects, the two binary variables were then aggregated into one dummy War, coded 1 if the country was facing civil and/or interstate war in a given year, and 0 if facing neither.
Second, the literature on the effect of sanctions depending on regime type is far from straightforward. While some argue that economic strangulation can precipitate the collapse of authoritarian regimes (Gasiorowski, 1995) and that violations of political rights are more likely under sanctions in more democratic states (Allen, 2008), others contend that the lack of accountability makes authoritarian states more likely to crack down on protestors (Wood, 2008). Following from Allen and Lektzian (2013), the binary Democracy variable was obtained by simplifying the index of the PolityIV dataset (Marshall and Gurr, 2014). Democracy was coded 1 when polity scores were 6 or greater, and 0 if the polity scores ranged between – 10 and 5.
Third, because a higher level of economic development is expected to either yield greater educational and professional opportunities, leading to greater autonomy and bargaining power (Doepke et al, 2012; Gray et al, 2006; Richards and Gelleny, 2007), or reinforce a patriarchal system confining women to devalorised and low-wage ‘feminised’ jobs (Sassen, 1996), the natural log of GDP per capita (in 1995 constant US dollars) is included.
Fourth, to control for regional differences, the models include six regional dummy variables, North America, Latin America, Europe, Africa, Asia and Oceania (reference category).
First, since the three dependent variables are ordinal variables with more than two categories, whose values follow a logical sequential order, all the models are the product of ordered logistic regression . Second, each model has included the Huber-White robust standard errors to correct for the models’ heteroskedasticity. Third, year dummies were included to control for year effects and prevent causal relationships to be mistaken for aggregate trends. Fourth, a one-year lag of all time-variant variables was included. While lagging the women’s rights variables allows to correct for autocorrelation , a one-year lag of the independent variables prevents simultaneous causality by ensuring that the response variables temporally follow the explanatory variables.
 The results remained mostly similar when the sample was restricted to countries having faced economic sanctions.
 Although not shown due to space restrictions, not aggregating civil war and interstate war variables into one variable yielded similar results.
 Theoretically, using ordered probit regression is also suggested. The models using ordered probit regression are not displayed but yielded similar conclusions.
 It has been suggested that four-category ordinal variables do not enclose enough information to be lagged. Although not shown, not lagging the dependent variables yielded similar results (with the exception of a more significant negative effect of human rights sanctions on women’s economic rights)