Policy Research Working Paper 9822 Do Gender Norms Become Less Traditional with Displacement? The Case of Colombia Eliana Rubiano-Matulevich Gender Global Theme October 2021 Policy Research Working Paper 9822 Abstract Conflict-induced displacement is associated with loss of multilevel linear regression models show that gender norms human and physical capital and psychological trauma. condoning violence against women relax with displacement, Households and social structures that produce and repro- while those that limit women’s economic opportunities duce gender norms are disrupted, providing opportunities become more rigid. The findings also reveal a misalign- for change. This paper operationalizes a definition of gender ment between attitudes and behaviors in other domains. norms that brings together the behaviors and attitudes Displaced women have less rigid patriarchal attitudes, but of displaced and non-displaced women using household their ability to decide about contraception and their own survey data for Colombia. The results of a two-step estima- earnings decreases following displacement. tion involving kernel-based propensity score matching and This paper is a product of the Gender Global Theme. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at erubiano@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Do Gender Norms Become Less Traditional with Displacement? The Case of Colombia * Eliana Rubiano-Matulevich+ Keywords: gender norms, conflict, internal displacement. JEL codes: J16, D10, D74, O15, F22. The authors of this paper conducted their research under Gender Dimensions of Forced Displacement project. The project is co-led by Lucia Hanmer and Diana Arango under the guidance of Hana Brixi, Global Director, Gender Unit, The World Bank Group. * This work is part of the program ‘Building the Evidence on Protracted Forced Displacement: A Multi-Stakeholder Partnership’. The program is funded by UK aid from the United Kingdom's Foreign, Commonwealth and Development Office (FCDO), it is managed by the World Bank Group (WBG) and was established in partnership with the United Nations High Commissioner for Refugees (UNHCR). The scope of the program is to expand the global knowledge on forced displacement by funding quality research and disseminating results for the use of practitioners and policy makers. This work does not necessarily reflect the views of FCDO, the WBG or UNHCR. + World Bank Group, Washington DC, USA. The author would like to thank Lisa Andersson, Franziska Gassmann, Khalid Koser, Ortrun Merkle, Zina Nimeh, Isabel Ruiz, Melissa Siegel, and Jennifer Waidler for their thorough review of earlier versions of this paper. The author is grateful to Diana Arango, Gabriel Demombynes, Kathryn Falb, Lucia Hanmer, Michelle Hynes, Patricia Justino, and Jeni Klugman, for their insightful comments and suggestions. Ragui Assad, Daisy Demirag, Jocelyn Kelly, Caroline Krafft, Ezequiel Molina, Maria Beatriz Orlando, Julieth Santamaria, Sergio Rivera, and Migration Days workshop participants at Maastricht University provided valuable comments to earlier versions. 1. Introduction Women can be disproportionately affected by the negative effects of conflict-induced displacement. Access to essential services such as reproductive health care can be disrupted. Displacement can result in higher levels of gender-based violence (GBV) (Annan & Brier, 2010; Callaway & Martin, 2011; Cohen et al., 2013; Vu et al., 2014; Wirtz et al., 2014). However, it can also provide opportunities to challenge gender norms that limit women’s access to opportunities and ability to make decisions. Following displacement, women might adopt new roles that would not have been possible before (Justino et al., 2012; Meertens & Stoller, 2001; Pirtskhalava, 2015). For example, in the absence of men, displaced Nuer women in South Sudan took on male responsibilities as income providers and assumed roles perceived as male, including negotiating dowries. To date, however, partly due to data limitations, few quantitative studies take account of the changes in gender relations among women and men in situations of displacement. This paper aims to bridge this knowledge gap by building on the work by Heise and Cislaghi (2020), who propose a definition of gender norms that brings together two streams of theory and practice around gender equality. The first stream is the work on social norms, which emerged from studies in social psychology and evolved with behavioral economics (Bicchieri, 2005; Mackie et al., 2015). The second stream is the study of gender norms advanced by feminist scholars (Badgett & Folbre, 1999; Connell & Pearce, 2014; Connell & Pearse, 2015). The definition is operationalized by measuring behaviors and attitudes (Alesina et al., 2013; Harper et al., 2020). Following previous studies, survey clusters are used as a proxy for reference networks (Blakely, 2000; Kelly et al., 2018; Storey & Kaggwa, 2009; Uthman et al., 2011; Vyas & Heise, 2016). The analysis focuses on Colombia, a country with a long-standing conflict and home to the second largest internally displaced population in the world. It uses three rounds of Demographic and Health Surveys (DHS) data for the 2005-2015 period to examine the extent to which gender norms that limit women’s access to reproductive health, economic opportunities, and mobility, and norms that tolerate violence against women, and endorse patriarchy become less traditional among 2 women in situations of displacement. The empirical approach involves a two-step estimation using kernel-based propensity score matching (PSM) and multilevel linear regression models to estimate the effect of displacement behaviors and attitudes on the matched sample of women. The findings show mixed evidence regarding norm change. Gender norms that tolerate violence against women become less traditional with displacement, while those that limit women’s economic opportunities become more rigid. Results also reveal a misalignment between attitudes and behaviors in specific domains of gender norms. In particular, displacement reduces the likelihood of agreeing with patriarchal statements such as ‘families with men have less problems’ or ‘a good wife obeys her husband,’ but women’s ability to decide about contraception and their own earnings, proxies for behaviors, decreases with displacement. This is consistent with previous studies showing that displaced women in Colombia report increased controlling behaviors when they pursue employment while their husbands are unemployed. Furthermore, gender norm transgression might lead to a backlash of more restrictive gender norm attitudes. Displaced men’s controlling behaviors might be exacerbated by psychological trauma, stress and loss of financial stability (Hynes et al., 2016; Wirtz et al., 2014). The remainder of this paper is organized as follows. The next section presents a review of the literature. Section 3 describes the Colombian context. Section 4 presents the theoretical framework. Section 5 describes the data, followed by Section 6, which presents the empirical approach. Section 7 discusses the results before concluding in Section 8. 2. Literature Review Displacement disrupts social and community relations, alters the structure and size of households, and is associated with changes in gender roles (Gururaja, 2000; Ibáñez & Velásquez, 2009; Ibáñez & Vélez, 2008; Levine et al., 2019; Vélez & Bello, 2010). In some cases, displacement might also provide opportunities to renegotiate gender roles (Aysa-Lastra, 2011). Women often take on the role of providers and protectors of families when their male partners die, disappear, are recruited by armed groups, or migrate in search of economic opportunities (Ayssa & Massey, 2004; El- Bushra, 2000). For example, Meertens and Stoller (2001) and Meertens and Segura‐Escobar 3 (1996) document changes in gender roles among internally displaced populations (IDP) in Colombia. The authors argue that most displaced rural women were raised in a context of patriarchal traditions characterized by rigid gender roles around domestic chores and participation in agricultural activities close to the home. When fleeing to urban settings, displaced women often become the main breadwinners for the first time in their lives. Gender norms that assign women to the domestic sphere have significant economic consequences in the context of displacement. As women enter paid employment, these gender norms are reproduced by occupational segregation (Badgett & Folbre, 1999). Displaced women tend to be employed as domestic workers or they are often engaged in petty trade (Bouta et al., 2005). They also maintain their roles as primary caregivers, creating a double burden compounded by poor security, limited infrastructure, and gender norms (Culcasi, 2019; Petesch, 2017; Pirtskhalava, 2015). For example, Culcasi (2019) reports that Syrian refugee women in Jordan have become breadwinners for their households, but their responsibility as the caretakers for their families has not diminished. Similar dynamics have been reported for IDP widows in Nepal (Ramnarain, 2016), Chechen refugees in the Czech Republic (Szczepanikova, 2005), and IDP women in Darfur (De La Puente, 2011). Notwithstanding displaced women’s increasing participation in the labor market, the evidence on the effect of displacement on intra-household bargaining power is mixed. In a study on the impact of conflict on women’s activities, Justino et al. (2012) find that women in Colombia participate more actively in labor markets during and immediately after conflict. Moreover, greater engagement in paid work is accompanied by improvements in women’s economic empowerment within households. In another study for Colombia, Calderón et al. (2011) confirm that IDP women work more hours per week than non-IDP women in rural areas. However, their ability to participate in important household decisions remains unaltered. Further, IDP women often report increased domestic violence when they pursue employment or education while their husbands were unemployed (Wirtz et al., 2014). In Turkey, Gulesci (2018) finds that displaced men were more likely than their non-displaced counterparts to display controlling behaviors, either by limiting their wives’ movements or social interactions. At the same time, displaced women were more inclined to believe that domestic violence is acceptable, compared to non-displaced women. 4 Women’s ability to perform activities deemed unsuitable pre-displacement often depends on the contestation of gender norms at the community level. In the above mentioned study in Darfur, De La Puente (2011) shows that IDP women were involved in health-related activities at the community level, but they did not participate in decisions related to the infrastructure or management of the camp, which were perceived as male fields. In contrast, Ramnarain (2016) finds that in the aftermath of the conflict in Nepal, widows engaged in employment outside the home and some of them even crossed over into male-dominated fields, such as construction labor or transport. 3. Colombian Context The number of IDPs in Colombia constitutes 12 percent of the 45.7 million people who had been forced to flee worldwide by the end of 2019, resulting in the second largest displacement in the world, behind only the Syrian Arab Republic (IDMC, 2020). Displacement in Colombia is directly linked to violence, but the underlying causes are as complex as the protracted conflict itself. Similar to other situations of conflict in the world, the attacks on civilians are the main triggers for displacement in Colombia (Bohra-Mishra & Massey, 2011; Czaika & Kis-Katos, 2009; Engel & Ibáñez, 2007; Lischer, 2007; Shultz et al., 2014). The violence is directly linked to land disputes and drug trafficking, but it is also used as a strategy of armed groups to exert greater power and destroy social networks (Ibáñez, 2008; Kay, 2001; Perez, 2002). Official figures indicate that the main groups responsible for the displacement are guerilla and paramilitary groups (Ibáñez, 2009). The nature of the conflict implies that displacement is not confined to specific areas of the country. Between 1997 and 2018, more than 6 million people were forced to flee their homes from 90 percent of the country’s 1,123 municipalities (Figure 1) and its incidence has been more intense in rural areas characterized by weak institutional presence (Angrist & Kugler, 2008; Ibáñez, 2009). Nonetheless, cities hosting large populations of IDPs, such as Medellín, Cali and Bogotá, have also experienced violence by parties of the armed conflict and criminal groups, which has led to the phenomenon of intra-urban displacement (Atehortúa et al., 2013; Jacobsen, 2011). 5 Figure 1. Number of IDP expulsions, 1997-2018 Source: Author based on Colombian Registry of Victims (RUV). The decades-long armed conflict in Colombia has affected men and women in a number of ways. Men have been more likely than women to be kidnapped, killed, injured, and forcibly recruited by armed actors (Oficina del Alto Comisionado para la Paz, 2020). Women and girls, on the other hand, are more likely to be victims of sexual violence and forced labor. They often assume the role of household heads and tend to be the caregivers for disabled family members (Bouvier, 2016). Rural women are particularly vulnerable, as they have limited access to land and other productive assets (World Bank, 2019). This situation is aggravated by the exposure to sexual violence in situations of displacement (ABColombia & US Office in Colombia, 2013). The Colombian state has established a solid normative framework for gender equality. On paper, the legislation recognizes women’s rights, penalizes GBV, and mandates a 30 percent quota for women in electoral lists. The peace accord between the government and the Revolutionary Armed Forces of Colombia (FARC for its Spanish acronym) is often referred to as a model for gender inclusion, as it recognizes gender inequities in multiple areas; guarantees the rights of rural women; and addresses the rights of the victims (PRIO Centre on Gender, Peace and Security, 2016; Ruiz- 6 Navarro, 2019). Furthermore, the Victims and Land Restitution Law, which enables victims of the conflict to receive assistance and reparation, established preferential treatment for IDP women and reparations for survivors of sexual violence (Bouvier, 2016; Valcarcel & Samudio, 2017). Notwithstanding, the legislative framework has not yet translated into conditions of gender equality. Gender discrimination prevents applicants from receiving property rights after their husbands have died or disappeared (Garcia-Godos & Wiig, 2014; Meertens, 2010). Access to justice also remains a challenge for victims of conflict-related sexual violence, despite an increase in the number of formal complaints (United Nations Security Council, 2020). One of the challenges to achieve gender equality in Colombia lies in deeply rooted gender norms. Women are expected to take on the bulk of domestic responsibilities, whereas men are seen as the head and main breadwinners for their families (Chant, 2002). Affordable, good-quality childcare services are lacking, and no legal provision exists for paid parental leave to be shared between both mother and father (World Bank, 2019, 2020). In fact, women do nearly four times as much as much unpaid domestic and care work as men do. 1 According to van der Gaag et al. (2019), alongside outliers such as Bangladesh and Algeria, Colombia is one of the countries where both the laws and gender norms around caregiving are relatively resistant to gender equality. These structural barriers are reflected in women’s lack of access to economic opportunities. 4. Theoretical Framework In general, norms specify rules, conventions and institutions that dictate what should or should not be done (Harper et al., 2020). Gender norms, in particular, are defined by Cislaghi and Heise (2020) as: Social norms 2 defining acceptable and appropriate actions for women and men in a given group or society. They are embedded in formal and informal institutions, nested in the mind, and produced and reproduced through social interaction. They play a role in shaping women and men’s (often 1 World Bank Gender Data Portal (national estimates). https://datatopics.worldbank.org/gender/. Accessed February 4, 2021. 2 Rule of behavior related to the differences in societal expectations for women and men. Individuals prefer to follow such rule if they believe that most people in their reference network conform to it and believe they should follow it (Bicchieri, 2005; Mackie et al., 2015). 7 unequal) access to resources and freedoms, thus affecting their voice, power and sense of self. 3 (p. 415). Following Cislaghi and Heise (2017) and using elements outlined in Marcus and Harper (2015) in relation to gender norm change, this section describes the main individual, social, material, and structural factors that could shift gender norms in situations of displacement. Depending on the context, they can either promote a positive change, that is, gender norms become less traditional and new practices emerge, or a negative change, which entails more discriminatory practices. Gender norms are learned early in life. They are adopted and endorsed by women and men through their behaviors and attitudes (Cislaghi & Heise, 2020; Harper et al., 2020; Lundgren et al., 2019). Gender norms are reinforced or contested in school, at the workplace, by the media, and other social institutions (Tenenbaum & Leaper, 2002). Aspirations and skills acquired over time contribute to the reproduction or change of gender norms. For instance, gender norms become manifest in educational materials that portray characters in stereotypical roles in the household and at the workplace (Blumberg, 2008; Islam & Asadullah, 2018; Mahmood & Kausar, 2019; Miroiu, 2004). Providing people with access to unbiased education materials and curriculum can thus contribute to norm change. For instance, a semester-long course on gender equity as part of preservice training in Turkey developed more favorable attitudes toward gender equality among aspiring teachers (Erden, 2009). Improved access to education can also foster more liberal attitudes and a break in the intergenerational transmission of gender norms (Marcus & Harper, 2015). Mass media can reflect and sustain gender norms over time, but it can also foster positive change. By moving from remote to more densely populated areas, displaced women and men may access both factual and overt messaging about gender equality. These messages can also be transmitted through social interaction and popular entertainment programs that present an alternative vision of gender relations. For example, evidence suggests that soap operas played an important role in the reduction of fertility rates in Brazil and in shifting gender norms around domestic violence in Nigeria (Banerjee et al., 2019; Ferrara et al., 2012). 3 There are multiple definitions of gender norms. For instance, Connell and Pearce (2014) define them as the beliefs and rules, in a given community or institution, about the proper behavior of men and women. See Cislaghi and Heise (2020) for a detailed review of concepts. 8 Migration can reduce exposure to the structures that tend to reinforce gender norms, such as traditional and religious leaders (Muñoz-Boudet et al., 2013). Behavioral studies suggest that what other people do has a greater influence than what they say, particularly for behaviors that are visible, such as child marriage (Palluck & Ball, 2010). For behaviors that are less visible, norms are more likely to be spread by people talking about endorsing them (Bursztyn et al., 2020). Gender norms are influenced by material conditions and the environment in which individuals are born and live. Gender norms underpin inheritance laws, ownership and control over assets, and intra-household dynamics (Agarwal, 1997; Connell & Pearce, 2014). For example, land privatization via government redistribution programs has often disadvantaged women by placing land in the hands of male relatives (Whitehead & Tsikata, 2003). On the other hand, the loss of assets by displaced families represents an overall loss of wealth for the household, but it might also ‘level the playing field’ for women. This could foster changes around decision making, resource allocation, and patriarchal notions around men as the main breadwinners. Policies, regulations, and institutional biases reinforce gender norms. In the labor market, for instance, gender norms influence recruitment, the work environment, wage differentials, and career progression. In Colombia, displaced men often face large spells of unemployment as their agricultural skills are less relevant in urban settings. Displaced women, on the other hand, are frequently employed as domestic workers who are poorly remunerated (Meertens & Stoller, 2001). These dynamics can challenge patriarchal gender norms whereby men are no longer the breadwinners and increase intra-household tensions (Calderón et al., 2011; Meertens & Segura‐ Escobar, 1996). On the other hand, the recognition of women’s economic contribution might call patriarchal norms into question (Gutmann & Viveros, 2005; Jensen, 2012). 5. Data and Descriptive Statistics 9 This paper uses data from the Colombian DHS for 2005, 2010, and 2015. 4 Surveys are representative of the female population ages 13-49 at the national, urban, and rural levels. They collect information on health outcomes and socio-economic characteristics. The most recent waves include attitudes towards gender equality, women’s role in society, GBV, and intra-household decision-making. The DHS employs two-stage sampling designs. Primary sampling units (PSUs) or clusters are sampled in the first stage, and households in the second stage. This design results in a multilevel data set. The sample consists of 37,211 households (41,344 women) in 2005, 51,447 households (53,521 women) in 2010, and 44,614 households (38,718 women) in 2015. Men were interviewed in the 2015 round, but they are not included in the analysis because of the limited set of questions on attitudes and behaviors in the questionnaire. The surveys allow for the direct identification of displaced household members. Questions ask whether the respondent lived in one or more places in the last 5 years, the date of migration, and the reason for migrating. 5 Displaced households are defined as those that had at least one member who was forced to flee due to conflict. This is a reasonable assumption for Colombia, where nearly 91 percent of IDPs migrate with all household members (Ibáñez, 2008). Furthermore, to facilitate the provision of reparations and other entitlements, as the IDP status is attached to a household, it is transmitted across generations (Sarzin, 2017; Shultz et al., 2014). IDPs were oversampled in the three survey waves and they represent 7 percent of the individuals who migrated internally each survey-year. 5.1. Measurement of gender norms Gender norms relate to multiple spheres of life. This study focuses on norms around reproductive health, economic opportunity, mobility, violence against women, and patriarchy. The analysis of gender norms in other spheres, although important, remains beyond the scope of this paper because of data limitations. For the purposes of operationalization, the analysis combines the definition proposed by Cislaghi and Heise (2020) with elements of the social norms theory (Bicchieri, 2005, 4 Earlier surveys are excluded from the study either because they do not sample IDPs or do not include the set of questions required to conduct the analysis of gender norms. 5 The module also asks people if they moved to another municipality in the same department, to another department, or within the same municipality. 10 2017). 6 Specifically, it pays attention to the role of a reference network, which refers to the group of people whose actions and beliefs individuals care about when they act. Depending on the context and the sphere of life that gender norms refer to, this group can be given by neighborhoods, villages, or people on the street. Hence, as people move from one place to another, they move across reference groups and may knowingly change their behavior to comply with the norms in place in the new setting (Choe et al., 2014). The analysis focuses on behaviors or actions and attitudes or empirical expectations (Alesina et al., 2013). 7 These two components are measured and analyzed separately, rather than combined in an index. Following previous studies, survey clusters are used as a proxy for reference networks (Storey & Kaggwa, 2009; Uthman et al., 2011; Vyas & Heise, 2016). The prevalence of behaviors and attitudes at the reference network level is inferred by aggregating (non-self) reported values across individuals in the same cluster, as it is reasonable to think that people residing in the same cluster might have direct contact with each other. This approach is consistent with the feminist literature which has theorized gender norms as having blurry boundaries, rather than focusing on a particular group with similar demographic characteristics (Oakley, 2015). Also, following gender norms theory, which focuses on the alignment between the norm and personal attitudes, the analysis in this paper assumes that gender norms become less (more) traditional when both attitudes and behaviors become less (more) rigid. The analysis examines 17 items (10 attitudes and 7 behaviors) classified into five domains. Reproductive health, is measured using one indicator for attitudes and one indicator for behaviors, as shown in Table 1. The question on attitudes refers to women’s approval of contraception to prevent pregnancy. The proxy for behavior measures the ability of women to decide upon the use of contraception. The norms that limit women’s access to economic opportunities, are measured by one indicator for attitudes and two indicators for behaviors. These items deal with the intersection of family and work as well as the ability of women to decide on their earnings. The norms that limit women’s mobility, are proxied by one indicator for attitudes and one indicator for 6 As articulated by Cislaghi and Heise (2020), operationalizing this definition using only quantitative measures also requires recognizing that they may fall short in capturing institutional aspects. 7 According to Harper et al. (2020), although gender norms are invisible, they are reflected in behaviors and attitudes. 11 behaviors, which capture women’s ability to make decisions around their own mobility. There are four indicators to measure tolerance towards violence against women and two proxies for violent behaviors. Patriarchal gender norms include six indicators measuring the disagreement with statements around male dominance within the household. Four of the indicators on attitudes are combined into a single measure that differentiates women who disagree with all statements compared to those who agree with at least one of them. All indicators are assigned a 1 for a less traditional attitude/behavior and 0 otherwise. Only the more recent waves include questions about attitudes towards gender equality. Table 1. Attitudes and behaviors indicators Attitude or behavior Sphere Component Attitude or behavior question nontraditional if response Surveys Abbreviation is Do you approve or disapprove that Reproductive Attitude couples use a method to prevent Approve All App contra health pregnancy? Main decision maker for the use of Sole decision maker for Behavior 2010, 2015 Use dec contra contraception contraception Women’s most important role is to Attitude Disagree 2015 D Wcare hh care for the household and to cook Respondent works & Main or shared decision on how to decides how to spend All Decide money Economic spend money money opportunities  Behavior Who cleans the house, prepare food, Respondent & partner, clean bathroom, wash clothes, buy partner more, partner 2015 Share chores food/supermarket, pay bills, take care alone, neither of sick (chores x=1-8) It is normal that men do not allow Attitude Disagree 2015 D rest mobility their wives to go out Mobility No one, decision not made, Who has final say on visits to family or Behavior respondent, or respondent 2010, 2015 Say visits relatives & partner When men are mad it is better not to Disagree 2015 D tempt men tempt them Attitude Women that stay in a relation after Violence against Disagree 2015 D Wbeat & stay being beaten is because they like it women Has friends who abuse their wives No 2015 No abu friend Would call the attention (or has done Behavior Yes 2015 Call abu friend it) of a friend who abuses a woman Men always have the last word on Disagree 2015 M last word household decisions Families with a man have less Disagree 2015 M less prob problems Attitude Men are head of households Disagree 2015 M heads Patriarchal A good wife always obeys her husband Disagree 2015 Wife obeys norms Disagreement with all patriarchal Disagree with all 2015 No patriarchy statements Who has final say on making large Respondent alone or Say imp Behavior household purchases and own health 2010, 2015 respondent & husband decisions care 12 5.2. Descriptive statistics Table 2 shows statistics comparing IDP and non-IDP women interviewed in the three rounds of DHS. On average, IDP women are younger and less educated than their non-IDP counterparts. They are less likely to be married but more they likely to be widowed, or not be in union. IDP households have one more member than non-IDP households; they have slightly more children and adult members, but fewer elderly. The average time in displacement is 2.5 years. Table 2. Descriptive statistics Observations Displaced Non-displaced Diff Individual characteristics Age 110,992 28.26 29.86 -1.61*** (0.23) (0.03) (0.26) Years of education 110,772 7.46 9.02 -1.57*** (0.08) (0.01) (0.09) Marital status Never married 110,977 0.33 0.39 -0.06*** (0.01) (0.00) (0.01) Married 110,977 0.14 0.19 -0.05*** (0.01) (0.00) (0.01) Cohabiting 110,977 0.36 0.29 0.07*** (0.01) (0.00) (0.01) Widowed 110,977 0.03 0.02 0.01*** (0.00) (0.00) (0.00) Not in union 110,977 0.14 0.11 0.03*** (0.01) (0.00) (0.01) Employed 110,992 0.46 0.51 -0.04*** (0.01) (0.00) (0.01) Household characteristics Size 110,992 5.91 4.88 1.03*** (0.06) (0.01) (0.05) Children (0-5) 110,992 0.81 0.54 0.27*** (0.02) (0.00) (0.02) Children (6-14) 110,992 1.43 0.96 0.47*** (0.03) (0.00) (0.02) Adults (15-64) 110,992 3.54 3.18 0.36*** (0.04) (0.00) (0.04) Elderly (65+) 110,992 0.13 0.20 -0.07*** (0.01) (0.00) (0.01) Female-headed 110,990 0.41 0.34 0.07*** (0.01) (0.00) (0.01) Years in displacement 1,874 2.50 (0.04) Source: Author based on DHS 2005/2010/2015. Note: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. 13 In terms of reproductive health, most women approve the use of contraceptives, but less than 1 in 5 women use any modern method and are the sole decision makers regarding contraceptive use (Figure 2). When it comes to economic opportunities, the proportion of women disapproving their relegation to the domestic sphere is lower than the share disagreeing with traditional attitudes in the other four dimensions. Also, few women decide on their earnings. Regarding mobility, IDP and non-IDP women are equally likely to endorse nontraditional attitudes and behaviors. When it comes to violence against women, few women disagree with the statements ‘when men are mad it is better not to tempt them’ or ‘women stay in abusive relations because they like it.’ In contrast, most of them reject the statement ‘men have the last word in household decisions’ and would call out a friend who abuses another woman. In terms of patriarchy, few women disagree with all four statements around men’s superiority in the household. In addition, IDP women are less likely than non-IDP women to have a say in important household decisions. Figure 2. Differences in attitudes and behaviors questions Source: Author based on DHS 2005/2010/2015. Note: The bars show the percentage of women with non-traditional responses. *** p<0.01, ** p<0.05, * p<0.1. 14 6. Empirical Strategy Following Ho et al. (2007), the empirical approach involves a two-step estimation. In the first stage, the analysis employs kernel-based propensity score matching (PSM) to pre-process the data and control group for the displaced (treatment) before applying the parametric analysis in the second stage. 8 The treatment group includes all women living in a household where at least one member was displaced due to violence, while the control group is defined as individuals who were not displaced. Voluntary migrants are excluded from the analysis. The main assumption is the exogeneity of the treatment, that is, armed groups attack civilians, seize their property and force them to flee. Therefore, displacement is not a voluntary decision to improve economic conditions (Ceriani & Verme, 2018; Ruiz & Vargas-Silva, 2015). In Colombia, in nearly 9 of every 10 cases, displacement is a reaction to being a victim of violent attacks (Ibáñez & Vélez, 2008). PSM estimates the propensity to be displaced based on the observable characteristics of interviewed women and their households ( ) = Pr( ∈ | = ). It employs a logistic regression of the indicator that takes value 1 for the displaced, and 0 for observations in the non-displaced sample, over the set of common variables. Predictors include age group, years of education, marital status, geographic area, and exposure to massacres at the municipality or department level with a two-year lag. Massacres are events in which four or more people are killed in a crime committed in the same place, at the same time, by the same perpetrators and the victims are defenseless people. 9 Because the surveys do not include detailed information on the municipality of origin, the analysis follows Calderón et al. (2011) and constructs a dummy variable equivalent to 1 if the household migrated within the same municipality and there were massacres in the two years prior to the date of migration; if the household migrated within the same department and there were any massacres in the department in the two years prior to the survey; or if it moved to another department where there were massacres in any other department two years 8 Kernel matching is a non-parametric matching estimator that uses weighted averages of all individuals in the control group to construct the counterfactual. One of the advantages of this approach is the lower variance that is achieved, because more information is used. A potential limitation is that observations used are bad matches. The common support condition is imposed to minimize this issue. 9 Victims belonging to the Public Force are not taken into account nor are cases in which the Public Force is fulfilling its Constitutional duty. 15 prior to their migration; and 0 otherwise. For non-displaced households, the variable assigns a 1 to households that live in a municipality where there were massacres with a two-year lag. The matching is done for each survey wave separately. The balancing property is fulfilled, that is, the mean propensity score is the same for individuals in the treatment and control groups. There is also a high degree of overlap between the two distributions, indicating that the common support assumption is satisfied (see balance tests in the Appendix). In the second stage, the analysis employs a multilevel linear regression model to estimate the effect of displacement on behaviors and attitudes on the matched sample. The model is given by equation (1): = + + + + + + + + (1) denotes behaviors or attitudes (measured separately) for individual i in cluster j in municipality l in department k at time y. The estimation controls for survey-year ( ) and department fixed effects ( ). contains municipality characteristics including the share of public expenditure allocated to the social sector and the value of royalties. denotes individual variables not included in the matching because they could have been affected by the treatment, such as household size, number of children under 5, sex of the head of household, employment status, and a wealth index. is equivalent to 1 if the woman is displaced and 0 for non-IDP women with similar characteristics (control group); is the coefficient of interest. is the (non-self) average of behaviors or attitudes aggregated at the cluster level, which serves as a proxy for the reference network. The multilevel model contains a fixed effects component, which consists of level-1 coefficients and a random effects component denoted that indicates variability across clusters. The estimation sequentially adds blocks of potential confounding variables to adjust for the characteristics of the women who comprise the reference network. To examine whether a multilevel model is appropriate, intraclass correlations were computed from the empty model with only the random error allowed to be free. The intraclass correlation captures 16 the proportion of variance that lies between level-2 units, which ranges from 0.15 for attitudes and behaviors around contraception to 0.9 for variables around violence against women. These amounts of variation are moderate to high, as Snijders and Bosker (2012) note that intraclass correlations with values between 0.05 and 0.2 are common. 7. Results Table 3 shows the effect of displacement on attitudes and behaviors around reproductive health. Columns (1)-(5) indicate that displacement does not alter attitudes towards the use of contraception. This is explained by the fact that most women in the sample agree with the use of contraception. When it comes to behaviors, columns (6)-(10) show that displacement reduces women’s ability to use and decide on contraceptive use. These patterns might be explained by the lack of access to sexual and reproductive health information and services, as well as different attitudes around the ideal family size (Harper et al., 2020). Male partners or female relatives might also exert control over the use of contraception (Solomon et al., 2019). Women’s lack of decision-making power around contraception could also be explained by less ‘gender-equitable’ practices within the household, which tend to the transmitted through the family and passed down to the next generation (Fernández & Fogli, 2009). Given the approach to determine whether gender norms become less traditional (or not) adopted in this paper, these findings do not provide evidence to suggest that displacement relaxes gender norms around reproductive health. 17 Table 3. Effect of displacement on gender norms around reproductive health Attitudes Behaviors Approve contrac Use & decide contra (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Displaced -0.01 -0.01 -0.01 -0.00 0.00 -0.05*** -0.03* -0.03* -0.03* -0.05** (0.01) (0.01) (0.01) (0.01) (0.01) (0.02) (0.02) (0.02) (0.02) (0.03) Household characteristics Children under 5 -0.00 -0.00 0.00 -0.00 -0.04** -0.04* -0.04** -0.02 (0.01) (0.01) (0.01) (0.01) (0.02) (0.02) (0.02) (0.02) Female headed 0.01** 0.01** 0.01** 0.01 0.10*** 0.11*** 0.11*** 0.10*** (0.01) (0.01) (0.01) (0.01) (0.03) (0.03) (0.02) (0.03) Wealth quintile Poor 0.02** 0.02 0.02 0.00 0.05* 0.05* 0.04 0.00 (0.01) (0.01) (0.01) (0.01) (0.03) (0.03) (0.03) (0.03) Middle 0.03*** 0.02** 0.02* 0.01 0.02 0.02 -0.01 -0.00 (0.01) (0.01) (0.01) (0.01) (0.03) (0.03) (0.03) (0.03) Rich 0.03*** 0.03*** 0.02** 0.01 0.01 0.01 -0.03 -0.04 (0.01) (0.01) (0.01) (0.01) (0.03) (0.03) (0.03) (0.04) Richest 0.03*** 0.03*** 0.03** 0.01 -0.00 0.01 -0.04 -0.07* (0.01) (0.01) (0.01) (0.01) (0.03) (0.03) (0.04) (0.04) Context Reference network 0.16** 0.090.13* -0.00 -0.03 -0.03 (0.07) (0.07) (0.06) (0.04) (0.04) (0.04) Royalties (ln) -0.00 0.00 (0.00) (0.00) Social investment (% exp) -0.00 -0.00 (0.00) (0.00) Constant 0.98*** 0.96*** 0.80*** 0.88*** 0.91*** 0.23*** 0.21*** 0.21*** 0.21*** 0.50** (0.00) (0.01) (0.06) (0.06) (0.07) (0.01) (0.02) (0.02) (0.04) (0.22) Observations 18,850 18,875 18,840 18,840 11,866 18,850 18,875 18,271 18,271 11,516 Number of groups 4,749 4,749 4,741 4,741 3,665 4,749 4,749 4,621 4,621 3,508 Department FE No No No Yes Yes No No No Yes Yes Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Source: Author based on DHS 2005/2010/2015. Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Reference group for wealth quintile is poorest. Turning to economic opportunities, Columns (2)-(5) in Table 4 show that displacement is associated with more traditional attitudes around women in the domestic sphere. Depending on the specification, displacement reduces the probability of disagreeing with the statement ‘a woman’s main role is family caregiving and cooking’ by 6-8 percentage points. This might be explained by the fact that social expectations that assign women to the domestic sphere can make them ‘protective’ of the unpaid care space, attaching value to their leadership of it and being resistant to 18 others’ involvement (van der Gaag et al., 2019). In terms of behaviors, columns (6)-(10) show that displacement reduces women’s ability to decide on the money they earn. In terms of the distribution of household chores, there are no significant differences between IDP and non-IDP women, except when including the full set of controls. In this specification, women tend to experience a slight redistribution of unpaid domestic work following displacement. Reference networks are strongly correlated with less traditional behaviors around domestic chores but do not influence women’s decision-making power over money. Overall, these estimates indicate that gender norms that limit women’s economic opportunities become more rigid with displacement. This finding is consistent with studies of Syrian refugees in Jordan (Culcasi, 2019), IDP widows in Nepal, Chechen refugees in the Czech Republic, and IDP women in Darfur (De La Puente, 2011; Ramnarain, 2016; Szczepanikova, 2005), which reveal that women can work for pay outside the home, but there is an expectation that they will be the main caregivers in the household. In the case of Colombia, Calderon et al. (2011) show that IDP women work more hours than non-IDP women in rural areas, but greater engagement in the labor market does not translate into improved bargaining power. Indeed, gender norms about paid and unpaid work are intertwined, but they can also move in different directions. According to Harper et al. (2020), this is the case when norms stretch to encompass women doing paid work, without any corresponding shifts in male responsibilities. 19 Table 4. Effect of displacement on gender norms around economic opportunities Attitudes Behaviors D Wcare hh Works & dec money Share chores (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) Displaced -0.04 -0.08*** -0.07*** -0.06** -0.07** -0.09*** -0.08** -0.07** -0.08** -0.07* 0.03 0.04 0.03 0.03 0.06** (0.03) (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.03) (0.03) (0.04) (0.02) (0.02) (0.02) (0.02) (0.03) Household characteristics Children under 5 0.04* 0.03 0.04 0.07** 0.04* -0.08** -0.09*** -0.08** -0.03 -0.03 -0.03 -0.06** (0.03) (0.02) (0.02) (0.03) (0.03) (0.03) (0.03) (0.04) (0.02) (0.02) (0.02) (0.03) Female headed -0.03 -0.03 -0.03 -0.03 -0.03 0.11*** 0.11*** 0.03 -0.00 0.00 0.01 -0.01 (0.02) (0.02) (0.02) (0.03) (0.02) (0.04) (0.04) (0.04) (0.03) (0.03) (0.03) (0.03) Wealth quintile Poor -0.20*** -0.16*** -0.15*** -0.12** -0.20*** 0.07 0.08* 0.12** 0.04 0.04 0.05** 0.02 (0.04) (0.04) (0.04) (0.05) (0.04) (0.05) (0.05) (0.05) (0.02) (0.02) (0.02) (0.03) Middle -0.29*** -0.22*** -0.24*** -0.22*** -0.29*** 0.15*** 0.18*** 0.30*** 0.10*** 0.09*** 0.09*** 0.06* (0.04) (0.04) (0.04) (0.05) (0.04) (0.06) (0.06) (0.06) (0.04) (0.03) (0.03) (0.03) Rich -0.38*** -0.30*** -0.33*** -0.26*** -0.38*** 0.17*** 0.20*** 0.26*** 0.09*** 0.08** 0.07* 0.03 (0.04) (0.04) (0.04) (0.06) (0.04) (0.06) (0.06) (0.07) (0.03) (0.03) (0.04) (0.03) Richest -0.44*** -0.34*** -0.35*** -0.29*** -0.44*** 0.12** 0.14** 0.22*** 0.09*** 0.07** 0.08** 0.02 (0.04) (0.04) (0.05) (0.07) (0.04) (0.06) (0.06) (0.08) (0.03) (0.03) (0.04) (0.04) Context Reference network 0.33*** 0.25*** 0.28*** -0.03 -0.06 -0.00 0.15* 0.17** 0.23** (0.06) (0.06) (0.07) (0.08) (0.08) (0.11) (0.08) (0.07) (0.10) Royalties (ln) -0.01* 0.01*** (0.00) (0.00) Social investment (% exp) 0.01 0.00 (0.00) (0.00) Constant 0.43*** 0.72*** 0.53*** 0.50*** 0.01 0.38*** 0.72*** 0.30*** 0.34*** -0.13 0.08*** 0.04* 0.03 0.02 -0.36 (0.00) (0.03) (0.05) (0.06) (0.38) (0.01) (0.03) (0.05) (0.07) (0.58) (0.01) (0.02) (0.02) (0.04) (0.25) Observations 15,799 15,798 15,755 15,755 9,170 15,799 15,790 15,790 15,790 9,170 9,090 9,089 8,852 8,852 5,017 Number of groups 4,213 4,213 4,170 4,170 2,363 4,213 4,205 4,205 4,205 2,363 3,783 3,783 3,546 3,546 2,017 Department FE No No No Yes Yes No No No Yes Yes No No No Yes Yes Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Reference group for wealth quintile is poorest. 20 Table 5 shows that displacement does not alter the attitudes of women towards wives’ ability to go out without telling their husbands. However, across most specifications (columns [6]-[8]), displacement increases the likelihood that women participate in the final say about visits to relatives and friends. Knowing someone who has greater decision-making power around mobility is also associated with less traditional behaviors. These results, however, do not provide enough evidence to say that displacement is associated with less traditional gender norms around mobility. In terms of the effect of displacement on gender norms around violence against women, columns (1) and (5) in Table 6 show that displacement reduces the likelihood of disagreeing with the statement ‘it is better not to tempt men when they are mad,’ but increases the probability of supporting the statement ‘women stay in abusive relations because they like it.’ In terms of behaviors, on the other hand, the effect of displacement is significant and stable across specifications for one of the two proxies. IDP women are more likely than their non-IDP counterparts to state that they would call out a friend who abuses a woman. These findings might reflect the fact that the acceptability of violence spans a continuum. Some women believe that violence is justified under certain circumstances, but they do not accept it completely (Harper et al., 2020). Furthermore, attitudes towards violence against women might be slow to change because of sticky norms that reflect patriarchy (Harper et al., 2020), but results in Table 6 provide some evidence to suggest that some attitudes and behaviors around violence against women appear to change with displacement. Strong legislative frameworks that support the rights of displaced women and condemn different forms GBV can also contest traditional gender norms. Some of these laws shape values and norms, which in turn, can influence individual attitudes and behaviors (Klugman, 2017; Nadler, 2017). For example, in a study of 12 Sub-Saharan African countries Maswikwa et al. (2015) found that the prevalence of child marriage was 40 percent lower in countries with laws against this practice compared with countries with no legislation. 21 Table 5. Effect of displacement and gender norms around women’s mobility Attitudes Behaviors D rest mobility Say visits (1) (2) (3) (4) (5) (6) (7) (8) (9) Displaced 0.05* 0.04* 0.03 0.04 0.05 0.06** 0.06** 0.06*** 0.04 (0.03) (0.02) (0.02) (0.03) (0.03) (0.03) (0.02) (0.02) (0.03) Household characteristics Children under 5 -0.00 -0.00 0.00 -0.05* 0.06*** 0.05** 0.03 (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.03) Female headed -0.02 -0.00 -0.02 -0.03 -0.04* -0.04* -0.06** (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.03) Wealth quintile Poor 0.05 0.04 0.02 0.03 0.06* 0.07** 0.08** (0.03) (0.03) (0.03) (0.04) (0.03) (0.03) (0.04) Middle 0.06* 0.04 0.03 0.04 0.04 0.06* 0.05 (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.05) Rich 0.11*** 0.09*** 0.09** 0.07 0.05 0.08** 0.06 (0.04) (0.04) (0.04) (0.05) (0.04) (0.04) (0.05) Richest 0.13*** 0.10*** 0.10*** 0.09** 0.01 0.04 0.03 (0.03) (0.03) (0.04) (0.04) (0.04) (0.04) (0.05) Context Reference network 0.20*** 0.14** 0.06 0.27*** 0.38*** (0.07) (0.07) (0.09) (0.08) (0.10) Royalties (ln) -0.00 (0.00) Social investment (% exp) 0.00 (0.00) Constant 0.72*** 0.67*** 0.54*** 0.62*** 1.21*** 0.69*** 0.63*** 0.42*** 0.32 (0.01) (0.03) (0.06) (0.06) (0.36) (0.01) (0.03) (0.06) (0.41) Observations 15,462 15,508 15,464 15,464 9,008 15,462 15,508 15,500 9,032 R-squared 4,195 4,196 4,152 4,152 2,358 4,195 4,196 4,188 2,382 Department FE No No No Yes Yes No No Yes Yes Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Source: Author based on DHS 2000/2005/2010/2015. Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Reference group for wealth quintile is poorest. 22 Table 6. Effect of displacement on gender norms around violence against women Attitudes Behaviors D tempt men D W beaten & stay No abu friend Call abu friend (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) Displaced -0.03* -0.01 -0.01 -0.00 -0.03* 0.05** 0.06** 0.06** 0.05* 0.03 -0.04 -0.04 -0.03 -0.03 -0.02 0.05*** 0.05*** 0.05*** 0.05*** 0.06*** (0.01) (0.01) (0.02) (0.01) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.02) (0.02) (0.02) (0.02) (0.02) Household characteristics Size Children under 5 -0.00 -0.00 -0.00 -0.00 -0.02 -0.02 -0.01 0.04 0.02 0.03 0.03 0.05 -0.02 -0.02 0.01 -0.03 (0.01) (0.02) (0.01) (0.02) (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.02) (0.03) (0.02) (0.02) (0.06) (0.02) Female headed -0.02* -0.02 -0.01 -0.01 -0.04* -0.04* -0.04* -0.03 -0.08*** -0.07*** -0.08*** -0.06* 0.01 0.01 0.01 0.00 (0.01) (0.01) (0.01) (0.01) (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.02) (0.03) (0.02) (0.02) (0.06) (0.02) Wealth quintile Poor 0.02 0.01 0.00 -0.03 0.01 0.01 0.00 0.03 -0.09** -0.07* -0.05 -0.06 0.03 0.03 0.01 -0.00 (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.03) (0.04) (0.04) (0.04) (0.04) (0.04) (0.03) (0.03) (0.06) (0.03) Middle 0.02 0.01 -0.00 -0.02 0.01 0.01 0.00 0.03 -0.08** -0.06 -0.07* -0.07 0.04 0.04 0.01 0.06* (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.04) (0.04) (0.04) (0.04) (0.04) (0.05) (0.03) (0.03) (0.06) (0.03) Rich 0.11*** 0.10*** 0.09*** 0.05 0.04 0.05 0.03 0.12** -0.12*** -0.10** -0.12*** -0.12** 0.07*** 0.07*** 0.01 0.07** (0.02) (0.03) (0.03) (0.03) (0.04) (0.04) (0.05) (0.05) (0.04) (0.04) (0.04) (0.06) (0.02) (0.02) (0.06) (0.03) Richest 0.13*** 0.11*** 0.11*** 0.04 0.10** 0.10** 0.08* 0.17*** -0.08* -0.06 -0.10** -0.13** 0.05 0.04 0.01 0.02 (0.02) (0.03) (0.03) (0.03) (0.05) (0.05) (0.05) (0.07) (0.04) (0.04) (0.04) (0.06) (0.03) (0.03) (0.06) (0.04) Context Reference network 0.10 0.02 0.15* 0.06 0.02 0.01 0.18*** 0.10 0.07 0.07 0.01 0.06 (0.07) (0.06) (0.08) (0.07) (0.07) (0.09) (0.06) (0.06) (0.08) (0.06) (0.06) (0.10) Royalties (ln) 0.00 0.00 0.00 (0.00) (0.00) (0.00) Social investment (% exp) -0.00 0.00 0.00 (0.00) (0.01) (0.00) Constant 0.11*** 0.08*** 0.07*** 0.08*** 0.33 0.28*** 0.28*** 0.26*** 0.29*** 0.11 0.55*** 0.65*** 0.53*** 0.66*** 0.29 0.86*** 0.83*** 0.77*** 0.84*** 0.56** (0.01) (0.02) (0.02) (0.03) (0.27) (0.01) (0.03) (0.04) (0.05) (0.38) (0.01) (0.03) (0.05) (0.06) (0.49) (0.01) (0.02) (0.06) (0.06) (0.28) Observations 15,509 15,508 15,464 15,464 9,008 15,509 15,508 15,464 15,464 9,008 15,494 15,493 15,449 15,449 9,001 15,413 15,412 15,368 15,368 8,945 R-squared 4,196 4,196 4,152 4,152 2,358 4,196 4,196 4,152 4,152 2,358 4,195 4,195 4,151 4,151 2,358 4,191 4,191 4,147 4,147 2,355 Department FE No No No Yes Yes No No No Yes Yes No No No Yes Yes No No No Yes Yes Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Source: Author based on DHS 2000/2005/2010/2015. Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Reference group for wealth quintile is poorest. 23 Finally, Table 7 shows the effect of displacement on patriarchal norms. IDP women are significantly less likely than non-IDP women with similar characteristics to disagree with individual statements such as ‘families with men have less problems’ and ‘a good wife obeys her husband,’ but there is no significant difference when asked about men as heads of household and men’s last word in household decisions. Also, displaced women are significantly more likely than non-displaced women to disagree with all patriarchal statements at the same time. Both groups, however, are relatively small in the sample. In terms of behaviors, columns (18)-(22) indicate that displacement does not alter women’s say in important household decisions. While these findings do not suggest changes in patriarchal gender norms, they reveal important changes around women’s attitudes which might indicate slow shifts in intra-household dynamics. Sensitivity analysis This section presents sensitivity analysis focused on the effect of conflict-induced displacement on gender norms among young women (ages 13-24). Table 8 shows the results for the preferred model specification, which includes the whole set of controls, as well as year- and department fixed effects. Overall, there is no evidence to suggest that displacement alters gender norms among young women, except for those around economic opportunity. Consistent with the results for the full sample, displaced young women are significantly less likely than their non-IDP counterparts to reject statements that relegate women to the domestic sphere and they tend to have less decision- making power over their earnings (when they work). Displacement also causes a strong rejection of violent behaviors towards women among this age group, but it does not alter attitudes. 24 Table 7. Effect of displacement on patriarchal gender norms Attitudes M last word M less prob M heads Wife obeys (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) Displaced 0.00 0.03 0.03 0.03 0.01 0.06** 0.09*** 0.08*** 0.09*** 0.05* -0.02 0.00 -0.00 -0.00 0.05 0.01 0.06** 0.06** 0.05** 0.10*** (0.02) (0.02) (0.02) (0.02) (0.03) (0.03) (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.02) (0.02) (0.03) (0.03) Household characteristics Children under 5 -0.04 -0.03 -0.03 -0.05* -0.03 -0.03 -0.03 -0.06** -0.02 -0.02 -0.02 -0.04 0.03 0.03 0.03 -0.00 (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.02) (0.03) Female headed 0.05** 0.05** 0.05** 0.04 0.10*** 0.10*** 0.10*** 0.11*** 0.16*** 0.15*** 0.14*** 0.13*** 0.03 0.03 0.03 0.02 (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.02) (0.03) (0.03) (0.03) (0.02) (0.03) (0.02) (0.02) (0.02) (0.03) Wealth quintile Poor 0.15*** 0.12*** 0.12*** 0.14*** 0.14*** 0.13*** 0.12*** 0.10** 0.14*** 0.11*** 0.12*** 0.08* 0.16*** 0.14*** 0.15*** 0.18*** (0.04) (0.04) (0.03) (0.04) (0.04) (0.04) (0.03) (0.04) (0.04) (0.04) (0.04) (0.05) (0.03) (0.03) (0.04) (0.04) Middle 0.21*** 0.16*** 0.17*** 0.16*** 0.18*** 0.16*** 0.17*** 0.12** 0.18*** 0.13*** 0.15*** 0.12** 0.28*** 0.24*** 0.27*** 0.28*** (0.04) (0.04) (0.04) (0.05) (0.04) (0.04) (0.04) (0.05) (0.04) (0.04) (0.04) (0.05) (0.04) (0.04) (0.04) (0.05) Rich 0.30*** 0.23*** 0.24*** 0.26*** 0.31*** 0.29*** 0.29*** 0.23*** 0.26*** 0.20*** 0.22*** 0.17*** 0.37*** 0.33*** 0.36*** 0.35*** (0.03) (0.04) (0.04) (0.05) (0.04) (0.04) (0.04) (0.05) (0.04) (0.04) (0.05) (0.06) (0.04) (0.04) (0.05) (0.06) Richest 0.34*** 0.26*** 0.27*** 0.30*** 0.33*** 0.30*** 0.30*** 0.24*** 0.33*** 0.25*** 0.26*** 0.27*** 0.45*** 0.39*** 0.42*** 0.45*** (0.03) (0.04) (0.04) (0.05) (0.04) (0.04) (0.04) (0.05) (0.04) (0.04) (0.04) (0.06) (0.03) (0.04) (0.05) (0.06) Context Reference network 0.27*** 0.17*** 0.16** 0.14** 0.08 0.07 0.27*** 0.14** 0.14* 0.17*** 0.10 0.03 (0.06) (0.06) (0.08) (0.06) (0.06) (0.08) (0.06) (0.06) (0.08) (0.06) (0.07) (0.08) Royalties (ln) 0.00 0.01* 0.00 0.01** (0.00) (0.00) (0.00) (0.00) Social investment (% exp) 0.00 -0.00 0.00 -0.00 (0.00) (0.00) (0.01) (0.00) Constant 0.71*** 0.52*** 0.36*** 0.43*** 0.11 0.45*** 0.24*** 0.19*** 0.21*** 0.53 0.52*** 0.45*** 0.18*** 0.31*** 0.13 0.19*** 0.14*** 0.18*** 0.33 (0.01) (0.03) (0.05) (0.06) (0.42) (0.01) (0.03) (0.04) (0.05) (0.44) (0.01) (0.01) (0.04) (0.05) (0.46) (0.03) (0.03) (0.05) (0.40) Observations 15,426 15,425 15,383 15,383 8,969 15,426 15,425 15,383 15,383 8,969 15,426 15,426 15,383 15,383 8,969 15,425 15,381 15,381 8,968 R-squared 4,184 4,184 4,142 4,142 2,355 4,184 4,184 4,142 4,142 2,355 4,184 4,184 4,142 4,142 2,355 4,184 4,140 4,140 2,354 Department FE No No No Yes Yes No No No Yes Yes No No No Yes Yes No No No Yes Yes Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 25 Table 7 (continued) Attitudes Behaviors No patriarchy Say imp decisions (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) Displaced 0.02 0.05** 0.04** 0.04** 0.05** 0.01 -0.00 -0.00 0.00 -0.02 (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03) Household characteristics Children under 5 0.01 0.01 0.01 -0.03 0.03 0.03 0.03 -0.00 (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03) Female headed 0.02 0.02 0.02 0.01 0.03 0.03 0.03 0.01 (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03) Wealth quintile Poor 0.06*** 0.04** 0.05** 0.03 0.10*** 0.10*** 0.08** 0.09* (0.02) (0.02) (0.02) (0.03) (0.04) (0.04) (0.04) (0.04) Middle 0.11*** 0.09*** 0.10*** 0.09** 0.13*** 0.13*** 0.10** 0.09* (0.02) (0.02) (0.03) (0.04) (0.03) (0.03) (0.04) (0.05) Rich 0.18*** 0.16*** 0.17*** 0.10** 0.09** 0.09** 0.05 0.01 (0.03) (0.03) (0.04) (0.04) (0.04) (0.04) (0.04) (0.06) Richest 0.25*** 0.21*** 0.22*** 0.18*** -0.03 -0.03 -0.07 -0.07 (0.03) (0.03) (0.04) (0.05) (0.05) (0.05) (0.05) (0.07) Context Reference network 0.17** 0.10 0.16 0.02 -0.05 0.04 (0.07) (0.07) (0.10) (0.15) (0.15) (0.20) Royalties (ln) 0.01*** 0.01 (0.00) (0.00) Social investment (% exp) -0.00 -0.00 (0.00) (0.00) Constant 0.13*** 0.01 -0.00 -0.00 0.30 0.60*** 0.51*** 0.50*** 0.52*** 0.72* (0.01) (0.02) (0.02) (0.03) (0.29) (0.01) (0.03) (0.05) (0.06) (0.43) Observations 15,426 15,425 15,381 15,381 8,968 15,426 15,425 15,425 15,425 8,996 R-squared 4,184 4,184 4,140 4,140 2,354 4,184 4,184 4,184 4,184 2,382 Department FE No No No Yes Yes No No No Yes Yes Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Source: Author based on DHS 2000/2005/2010/2015. Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Reference group for wealth quintile is poorest. 26 Table 8. Effect of displacement on gender norms among young women (ages 13-24) Reproductive health Economic opportunity Mobility Attitudes Behaviors Attitudes Behaviors Attitudes Behaviors Approve Use & decide D Wcare Works & dec Share D rest Say visits contrac contra hh money chores mobility (1) (2) (3) (4) (5) (6) (7) Displaced 0.01 -0.05 -0.08* -0.15* 0.04 0.03 0.08 (0.01) (0.05) (0.05) (0.08) (0.08) (0.05) (0.05) Household characteristics Children under 5 -0.02 0.08* 0.13*** 0.04 -0.88*** -0.11*** 0.16*** (0.02) (0.05) (0.04) (0.07) (0.09) (0.04) (0.05) Female headed 0.02* 0.05 -0.05 0.00 0.14** -0.07 -0.16*** (0.01) (0.05) (0.04) (0.06) (0.07) (0.04) (0.05) Wealth quintile Poor 0.01 0.05 -0.16** 0.06 0.07 0.08 0.16** (0.02) (0.06) (0.07) (0.06) (0.08) (0.07) (0.07) Middle 0.02* -0.03 -0.27*** 0.24*** 0.13 0.08 0.09 (0.01) (0.06) (0.08) (0.09) (0.09) (0.08) (0.08) Rich 0.02 0.04 -0.28*** 0.34** 0.00 -0.00 0.12 (0.01) (0.08) (0.09) (0.15) (0.12) (0.09) (0.09) Richest 0.02 0.15 -0.33*** 0.13 0.21 0.06 0.12 (0.01) (0.14) (0.09) (0.14) (0.19) (0.08) (0.09) Context Reference network 0.18 -0.02 0.32*** 0.09 -0.17 0.30** 0.54*** (0.11) (0.07) (0.11) (0.14) (0.10) (0.15) (0.17) Royalties (ln) 0.00 0.00 -0.01 0.00 -0.00 -0.00 -0.00 (0.00) (0.00) (0.01) (0.01) (0.01) (0.01) (0.01) Social investment (% exp) -0.00 -0.01 -0.00 0.02** -0.01* -0.00 -0.00 (0.00) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Constant 0.84*** 0.68 0.56 -1.44** 2.16*** 0.61 -0.02 (0.12) (0.49) (0.62) (0.72) (0.63) (0.54) (0.89) Observations 1,292 1,267 3,426 605 465 3,389 3,396 Number of groups 1,019 1,001 1,624 506 391 1,615 1,622 Source: Author based on DHS 2000/2005/2010/2015. Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Ref. group for wealth is poorest. 27 Table 8 (cont’d) Violence against women Patriarchy Attitudes Behaviors Attitudes Behaviors D tempt D W beaten & No abu Call abu M last Wife No Say imp M less prob M heads men stay friend friend word obeys patriarchy decisions (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) Displaced -0.05 0.03 -0.05 0.10*** -0.01 0.02 0.07 0.12*** 0.04 -0.02 (0.03) (0.05) (0.05) (0.03) (0.05) (0.05) (0.04) (0.04) (0.03) (0.04) Household characteristics Children under 5 0.01 0.09** -0.04 -0.06*** -0.07* -0.14*** -0.13*** -0.11*** -0.10*** 0.10** (0.03) (0.04) (0.05) (0.02) (0.04) (0.04) (0.04) (0.04) (0.03) (0.05) Female headed 0.01 0.01 -0.00 -0.00 0.07* 0.10** 0.11*** 0.00 0.02 -0.11** (0.02) (0.04) (0.04) (0.02) (0.04) (0.05) (0.04) (0.04) (0.03) (0.05) Wealth quintile Poor 0.03 0.09** -0.05 -0.03 0.09 -0.01 -0.05 0.19*** 0.02 0.13** (0.04) (0.05) (0.07) (0.03) (0.06) (0.06) (0.06) (0.07) (0.04) (0.06) Middle 0.03 0.08 0.01 0.04 0.08 0.07 0.06 0.23*** 0.04 0.13* (0.05) (0.06) (0.08) (0.03) (0.08) (0.08) (0.08) (0.07) (0.05) (0.07) Rich 0.08 0.12* -0.13 0.04 0.22*** 0.13 0.09 0.43*** 0.09 0.04 (0.06) (0.06) (0.08) (0.03) (0.07) (0.08) (0.09) (0.08) (0.07) (0.08) Richest 0.14* 0.30*** -0.10 0.02 0.20** 0.23*** 0.21*** 0.41*** 0.24*** 0.00 (0.08) (0.10) (0.09) (0.03) (0.08) (0.08) (0.08) (0.08) (0.06) (0.08) Context Reference network 0.31** -0.07 0.26** 0.28** 0.30*** -0.00 0.14 0.11 0.20 0.45* (0.15) (0.11) (0.12) (0.12) (0.11) (0.11) (0.11) (0.12) (0.14) (0.26) Royalties (ln) 0.00** -0.01* -0.00 0.00 0.00 0.00 0.01* 0.01** 0.01*** 0.01 (0.00) (0.01) (0.01) (0.00) (0.01) (0.01) (0.00) (0.00) (0.00) (0.00) Social investment (% exp) -0.01 0.01** -0.00 0.01** 0.00 -0.00 0.00 0.01 -0.01 -0.01 (0.00) (0.01) (0.01) (0.00) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Constant 0.56 -0.75 0.80 0.10 0.32 0.79 0.37 -0.37 0.54 0.48 (0.42) (0.51) (0.89) (0.29) (0.59) (0.55) (0.62) (0.66) (0.51) (0.61) Observations 3,389 3,389 3,387 3,361 3,377 3,377 3,377 3,377 3,377 3,385 Number of groups 1,615 1,615 1,614 1,610 1,612 1,612 1,612 1,612 1,612 1,620 Source: Author based on DHS 2000/2005/2010/2015. Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Ref. group for wealth is poorest. 28 8. Conclusions The findings in this paper show mixed evidence regarding norm change. Specifically, gender norms that tolerate violence against women become less traditional with displacement, while those that limit women’s economic opportunities become more rigid. These patterns are not necessarily surprising, as the evidence in this area is mixed. For example, Culcasi’s study (2019) of Syrian refugees in Jordan reveals that women can work for pay outside the home, but they continue to be the main caregivers in the household. Similar experiences have also been reported for IDP widows in Nepal, Chechen refugees in the Czech Republic and IDP women in Darfur (De La Puente, 2011; Ramnarain, 2016; Szczepanikova, 2005). The study also revealed a misalignment between attitudes and behaviors in specific domains of gender norms. For example, displacement is associated with less traditional patriarchal attitudes such as ‘families with men have less problems’ or ‘a good wife obeys her husband,’ but women’s ability to decide about contraception and earnings decreases following displacement. These findings shed light on the complexity of gender norm change, which does not operate in a vacuum. Moreover, change can be contradictory and improvements in one area do not imply that all others will automatically follow. The findings of this study have implications for policy making. Greater access to reproductive health for displaced women could be ensured through universal health coverage schemes, for example, by including contraception in basic packages. Social assistance programs for IDPs could also address barriers to contraception by providing information to increase uptake and men’s support for modern methods (Khan et al., 2016). While the expansion of reproductive health services might not guarantee a shift in gender norms, it might provide opportunities for women to overcome some of the challenges imposed by those norms (Malhotra et al., 2019). In terms of economic opportunities, the findings show that, in situations of displacement, paid work does not necessarily translate into increased decision-making power. Hence, providing access to economic opportunities is not a guarantee that gender gaps will be reduced if men have full control of the gains, as determined by patriarchal norms. Economic empowerment programs 29 for displaced women, in particular, should have built-in guidelines for the protection of women and should engage men in promoting more gender-equitable relationships (Heilman & Barker, 2018; van der Gaag et al., 2019). Given the long-term nature of displacement in Colombia, it is important to build capacity for both displaced women and men to access economic opportunities, which can eventually replace social assistance. Program interventions should identify occupations and sectors where they could work given their skills and include support services, such as flexible working hours and childcare to address constraints related to domestic responsibilities. Finally, the risk of sexual and GBV in situations of displacement has devastating impacts on the individual and on the household. This is a complex issue and requires a battery of interventions for prevention and response. Interventions that have worked in non-displaced settings such as strategies for addressing norms that condone violence against women, designing effective facilities and services for survivors, and engaging men and boys in prevention and response could be piloted in situations of displacement (see Jewkes et al., [2015] and Ellsberg et al., [2015]). This study has limitations. Large-scale household surveys rarely include questions to identify and measure a norm. Moreover, the choice of proxy indicators for attitudes and behaviors is driven by data availability and some of them could arguably be mapped onto multiple spheres or domains of gender norms. Given data limitations, mixed methods studies can offer more insights as the economics literature on the gender-differentiated effects of conflict-induced displacement evolves. The limited time span covered by the migration question might pose a challenge to capturing changes in gender norms. Nonetheless, the theoretical framework argues that displacement can accelerate change by, for instance, opening up economic opportunities for women in urban areas and reducing exposure to traditional structures that reinforce gender norms (Cislaghi & Heise, 2020; Harper et al., 2020; Marcus & Harper, 2015; Muñoz Boudet et al., 2013). The lack of data for men is another limitation. Gender norms are produced and reproduced by women and men; hence, the analysis omits part of the story. On the other hand, focusing on women’s views and their own decision-making power within the household provides an overview of intra-household dynamics that could hint at men’s and other household members’ gender norms. 30 References ABColombia, & US Office in Colombia. (2013). Women, Conflict-related Sexual Violence and the Peace Process. Agarwal, B. (1997). “‘Bargaining’” and Gender Relations: Within and Beyond the Household. Feminist Economics, 3(1), 1–51. https://doi.org/10.1080/135457097338799 Alesina, A., Giuliano, P., & Nunn, N. (2013). On the Origins of Gender Roles: Women and the Plough*. The Quarterly Journal of Economics, 128(2), 469–530. Angrist, J. D., & Kugler, A. D. (2008). Rural Windfall or a New Resource Curse? Coca, Income, and Civil Conflict in Colombia. Review of Economics and Statistics, 90(2), 191–215. Annan, J., & Brier, M. (2010). The risk of return: Intimate partner violence in northern Uganda’s armed conflict. Social Science & Medicine (1982), 70(1), 152–159. Atehortúa, C., Salcedo, J., & Vidal, R. (2013). Desplazados internos fuera de los campos. El papel de las autoridades locales en Colombia. Estudio comparado en Bogotá D.C. y Cali (Project on Internal Displacement). Brookings-LSE. Aysa-Lastra, M. (2011). Integration of Internally Displaced Persons in Urban Labour Markets: A Case Study of the IDP Population in Soacha, Colombia. Journal of Refugee Studies, 24(2), 277–303. Ayssa, M., & Massey, D. S. (2004). Wives Left Behind: The Labor Market Behavior of Women in Migrant Communities. In D. S. Massey & J. Durand (Eds.), Crossing the border: Research from the Mexican Migration Project. Russell Sage Foundation. Badgett, M. V. L., & Folbre, N. (1999). Assigning care: Gender norms and economic outcomes. International Labour Review, 138(3), 311–326. Banerjee, A., La Ferrara, E., & Orozco, V. (2019). Entertainment, Education, and Attitudes Toward Domestic Violence. AEA Papers and Proceedings, 109, 133–137. Bicchieri, C. (2005). The grammar of society: The nature and dynamics of social norms. Cambridge University Press. Bicchieri, C. (2017). Norms in the Wild: How to Diagnose, Measure, and Change Social Norms. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780190622046.001.0001 Blakely, T. A. (2000). Ecological effects in multi-level studies. Journal of Epidemiology & Community Health, 54(5), 367–374. 31 Blumberg, R. L. (2008). The invisible obstacle to educational equality: Gender bias in textbooks. PROSPECTS, 38(3), 345–361. Bohra-Mishra, P., & Massey, D. S. (2011). Individual Decisions to Migrate During Civil Conflict. Demography, 48(2), 401–424. Bouta, T., Frerks, G., & Bannon, I. (2005). Gender, conflict, and development. World Bank. Bouvier, V. (2016). Gender and The Role of Women in Colombia’s Peace Process [Background Paper prepared for the United Nations Global Study on 15 Years of Implementation of UN Security Council Resolution 1325 (2000)]. Bursztyn, L., González, A. L., & Yanagizawa-Drott, D. (2020). Misperceived Social Norms: Women Working Outside the Home in Saudi Arabia. American Economic Review, 110(10), 2997–3029. Calderón, V., Gafaro, M., & Ibáñez, A. M. (2011). Forced Migration, Female Labour Force Participation, and Intra-Household Bargaining: Does Conflict Empower Women? SSRN Electronic Journal. Callaway, A., & Martin, S. (2011). Internal Displacement and Internal Trafficking. In K. Koser & S. Martin (Eds.), The migration-displacement nexus: Patterns, processes, and policies. Berghahn Books. Centro de Memoria Histórica (Colombia) (Ed.). (2012). El Placer: Mujeres, coca y guerra en el bajo Putumayo (Primera edición en Colombia). Taurus. Ceriani, L., & Verme, P. (2018). Risk Preferences and the Decision to Flee Conflict (WPS 8376; Policy Research Working Paper). The World Bank Group. Chant, S. (2002). Researching Gender, Families and Households in Latin America: From the 20th into the 21st Century. Bulletin of Latin American Research, 21(4), 545–575. Choe, M. K., Bumpass, L. L., Tsuya, N. O., & Rindfuss, R. R. (2014). Nontraditional Family- Related Attitudes in Japan: Macro and Micro Determinants. Population and Development Review, 40(2), 241–271. Cislaghi, B., & Heise, L. (2020). Gender norms and social norms: Differences, similarities and why they matter in prevention science. Sociology of Health & Illness, 42(2), 407–422. Cohen, D., Hoover, A., & Wood, E. (2013). Wartime Sexual Violence: Misconceptions, Implications, and Ways Forward. United States Institute of Peace. 32 Connell, R., & Pearce, R. (2014). Gender norms and stereotypes: A survey of concepts, research and issues about change. UN Women. Connell, R., & Pearse, R. (2015). Gender Norms: Are They the Enemy of Women’s Rights?. UNRISD. Culcasi, K. (2019). “We are women and men now”: Intimate spaces and coping labour for Syrian women refugees in Jordan. Transactions of the Institute of British Geographers, tran.12292. Czaika, M., & Kis-Katos, K. (2009). Civil Conflict and Displacement: Village-Level Determinants of Forced Migration in Aceh. Journal of Peace Research, 46(3), 399–418. De La Puente, D. (2011). Women’s leadership in camps for internally displaced people in Darfur, western Sudan. Community Development Journal, 46(3), 365–377. Defensoría del Pueblo. (2019). Informe defensorial: Violencias basadas en género y discriminación. El-Bushra, J. (2000). Gender and forced migration. Forced Migration Review, 9, 4–7. Ellsberg, M., Arango, D. J., Morton, M., Gennari, F., Kiplesund, S., Contreras, M., & Watts, C. (2015). Prevention of violence against women and girls: What does the evidence say? The Lancet, 385(9977), 1555–1566. Engel, S., & Ibáñez, A. M. (2007). Displacement Due to Violence in Colombia: A Household‐ Level Analysis. Economic Development and Cultural Change, 55(2), 335–365. Erden, F. T. (2009). A course on gender equity in education: Does it affect gender role attitudes of preservice teachers? Teaching and Teacher Education, 25(3), 409–414. Fernández, R., & Fogli, A. (2009). Culture: An Empirical Investigation of Beliefs, Work, and Fertility. American Economic Journal: Macroeconomics, 1(1), 146–177. Ferrara, E. L., Chong, A., & Duryea, S. (2012). Soap Operas and Fertility: Evidence from Brazil. American Economic Journal: Applied Economics, 4(4), 1–31. Garcia-Godos, J., & Wiig, H. (2014). The Colombian Land Restitution Programme: Process, results and challenges, with special emphasis on women (NIBR-rapport 2014: 14). Norwegian Institute for Urban and Regional Research. Georgetown Institute for Women, Peace and Security (GIWPS). (2017). Inclusive Justice: How Women Shape Transitional Justice in Tunisia and Colombia. 33 Grabska, K. (2013). The Return of Displaced Nuer in Southern Sudan: Women Becoming Men?: The Return of Displaced Nuer in Southern Sudan. Development and Change, 44(5), 1135–1157. Gulesci, S. (2018). Forced Migration and Attitudes Towards Domestic Violence. In S. Anderson, L. A. Beaman, & J. P. Platteau (Eds.), Towards gender equity in development (First edition, pp. 97–116). Oxford University Press. Gururaja, S. (2000). Gender dimensions of displacement. Forced Migration Review, 9, 13–16. Gutmann, M. C., & Viveros, M. (2005). Masculinities in Latin America. SSRN Electronic Journal. Harper, C., Marcus, R., George, R., D’Angelo, S., & Samman, E. (2020). Gender, power and progress: How norms change. ALIGN/ODI. www.alignplatform. org/gender-power- progress Heilman, B., & Barker, G. (2018). Masculine Norms and Violence: Making the Connections. Promundo-US. Ho, D. E., Imai, K., King, G., & Stuart, E. A. (2007). Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference. Political Analysis, 15(3), 199–236. Hynes, M. E., Sterk, C. E., Hennink, M., Patel, S., DePadilla, L., & Yount, K. M. (2016). Exploring gender norms, agency and intimate partner violence among displaced Colombian women: A qualitative assessment. Global Public Health, 11(1–2), 17–33. Ibáñez, A. M. (2008). El desplazamiento forzoso en Colombia: Un camino sin retorno a la pobreza (1. ed). Universidad de los Andes, Facultad de Economía. Ibáñez, A. M. (2009). Forced displacement in Colombia: Magnitude and causes. The Economics of Peace and Security Journal, 4(1), 7. Ibáñez, A. M., & Velásquez, A. (2009). Identifying Victims of Civil Conflicts: An Evaluation of Forced Displaced Households in Colombia. Journal of Peace Research, 46(3), 431–451. Ibáñez, A. M., & Vélez, C. E. (2008). Civil Conflict and Forced Migration: The Micro Determinants and Welfare Losses of Displacement in Colombia. World Development, 36(4), 659–676. Internal Displacement Monitoring Centre (IDMC). (2020). Global Report on Internal Displacement 2019. Internal Displacement Monitoring Centre. 34 Islam, K. Md. M., & Asadullah, M. N. (2018). Gender stereotypes and education: A comparative content analysis of Malaysian, Indonesian, Pakistani and Bangladeshi school textbooks. PLOS ONE, 13(1), e0190807. Jacobsen, K. (2011). Profiling urban IDPs: How IDPs differ from their non-IDP neighbours in three cities. In K. Koser & S. Martin (Eds.), The migration-displacement nexus: Patterns, processes, and policies. Berghahn Books. Jensen, R. (2012). Do Labor Market Opportunities Affect Young Women’s Work and Family Decisions? Experimental Evidence from India *. The Quarterly Journal of Economics, 127(2), 753–792. Jewkes, R., McLean Hilker, L., Khan, S., Fulu, E., Busiello, F., & Fraser, E. (2015). What Works to Prevent Violence Against Women and Girls—Evidence Reviews. [Paper 3: Response Mechanisms to Prevent Violence Against Women and Girls]. What Works to Prevent Violence. Justino, P., Cardona, I., Mitchell, R., & Müller, C. (2012). Women Working for Recovery: The Impact of Female Employment on Family and Community Welfare after Conflict. UN Women. Kay, C. (2001). Reflections on rural violence in Latin America. Third World Quarterly, 22(5), 741–775. Kelly, J. T. D., Colantuoni, E., Robinson, C., & Decker, M. R. (2018). From the battlefield to the bedroom: A multilevel analysis of the links between political conflict and intimate partner violence in Liberia. BMJ Global Health, 3(2), e000668. Klugman, J. (2017). Gender based violence and the law [Background Paper prepared for the 2017 World Development Report Governance and the Law]. Kroc Institute. (2020). Towards implementation of women’s rights in the Colombian Final Peace Accord: Progress, opportunities and challenges (Barometer Initiative, Peace Accords Matrix). University of Notre Dame. Levine, S., Ullah, Z., Ullah Khan, S., Ahmad, M., Ghulam, S., Iqbal, R., & Uddin, S. (2019). The impact of displacement on gender roles and relations: The case of IDPs from FATA, Pakistan. Overseas Development Institute. Lischer, S. K. (2007). Causes and Consequences of Conflict-Induced Displacement. Civil Wars, 9(2), 142–155. 35 Lundgren, R., Burgess, S., Chantelois, H., Oregede, S., Kerner, B., & Kågesten, A. E. (2019). Processing gender: Lived experiences of reproducing and transforming gender norms over the life course of young people in Northern Uganda. Culture, Health & Sexuality, 21(4), 387–403. Mackie, G., Moneti, F., Shakya, H., & Denny, E. (2015). What are Social Norms? How are They Measured? UNICEF / University of California, San Diego, Center on Global Justice. Mahmood, T., & Kausar, G. (2019). Female Teachers’ Perceptions of Gender Bias in Pakistani English Textbooks. Asian Women, 35(4), 109–126. Marcus, R., & Harper, C. (2015). How do gender norms change. Overseas Development Institute. Maswikwa, Richter, Kaufman, & Nandi. (2015). Minimum Marriage Age Laws and the Prevalence Of Child Marriage and Adolescent Birth: Evidence from Sub-Saharan Africa. International Perspectives on Sexual and Reproductive Health, 41(2), 58. Meertens, D. (2010). Forced displacement and women’s security in Colombia. Disasters, 34, S147–S164. Meertens, D., & Segura‐Escobar, N. (1996). Uprooted Lives: Gender, Violence and Displacement in Colombia. Singapore Journal of Tropical Geography, 17(2), 165–178. Meertens, D., & Stoller, R. (2001). Facing Destruction, Rebuilding Life: Gender and the Internally Displaced in Colombia. Latin American Perspectives, 28(1), 132–148. Miroiu, M. (2004). All in One: Fairness, Neutrality and Conservatism – a Case Study of Romania. PROSPECTS, 34(1), 85–100. Muñoz-Boudet, A. M., Petesch, P., Turk, C., & Thumala, A. (2013). On norms and agency: Conversations about gender equality with women and men in 20 countries. World Bank. Nadler, J. (2017). Expressive Law, Social Norms, and Social Groups. Law & Social Inquiry, 42(01), 60–75. Oakley, A. (2015). Sex, gender and society. Ashgate Publishing. Oficina del Alto Comisionado para la Paz. (2020). Estadísticas de Asistencia Integral a las Víctimas de MAP y MUSE. Oficina Del Alto Comisionado Para La Paz. Palluck, E., & Ball, L. (2010). Social norms marketing aimed at gender based violence: A literature review. International Rescue Committee. 36 Perez, L. E. (2002). Desplazamiento forzado en Colombia 1995-1999: Una aproximación empírica a las relaciones entre desplazamiento, conflicto armado y desarrollo. In El desplazamiento forzado en Colombia: Compromisos desde la universidad. Petesch, P. (2017). The Oxford Handbook of Gender and Conflict. In F. Ní Aoláin, N. Cahn, D. F. Haynes, & N. Valji (Eds.), Agency and Gender Norms in War Economies. Oxford University Press. Pirtskhalava, E. (2015). Displacement Effects on Gender Roles, and Family Structure: Muslim Meskhetians in the USA. Sociology and Anthropology, 11. PRIO Centre on Gender, Peace and Security. (2016). Gender and Inclusion in the Colombian Peace Process (Issue 4). Peace Research Institute. https://gps.prio.org/utility/DownloadFile.ashx?id=3&type=publicationfile Ramnarain, S. (2016). Unpacking Widow Headship and Agency in Post-Conflict Nepal. Feminist Economics, 22(1), 80–105. Ruiz, I., & Vargas-Silva, C. (2015). The Labor Market Impacts of Forced Migration. American Economic Review, 105(5), 581–586. Ruiz-Navarro, C. (2019). A feminist peace in Colombia? (Feminist Foreign Policy). Sarzin, Z. (2017). Stocktaking of Global Forced Displacement Data. The World Bank. Shultz, J. M., Ceballos, Á. M. G., Espinel, Z., Oliveros, S. R., Fonseca, M. F., & Florez, L. J. H. (2014). Internal displacement in Colombia: Fifteen distinguishing features. Disaster Health, 2(1), 13–24. Snijders, T. A. B., & Bosker, R. J. (2012). Multilevel analysis: An introduction to basic and advanced multilevel modeling (2nd ed). Sage. Solomon, T., Nigatu, M., Gebrehiwot, T. T., & Getachew, B. (2019). Unmet need for family planning and associated factors among currently married reproductive age women in Tiro Afeta District, South West Ethiopia, 2017: Cross-sectional study. BMC Women’s Health, 19(1), 170. Storey, D., & Kaggwa, E. B. (2009). The Influence of Changes in Fertility Related Norms on Contraceptive Use in Egypt, 1995–2005. Population Review, 48(1). Szczepanikova, A. (2005). Gender Relations in a Refugee Camp: A Case of Chechens Seeking Asylum in the Czech Republic. Journal of Refugee Studies, 18(3), 281–298. 37 Tenenbaum, H. R., & Leaper, C. (2002). Are parents’ gender schemas related to their children’s gender-related cognitions? A meta-analysis. Developmental Psychology, 38(4), 615–630. United Nations Security Council. (2020). Report of the Secretary-General to the Security Council (S/2020/487). Office of the Sexual Representative of the Secretary-General on Sexual Violence in Conflict. Uthman, O. A., Moradi, T., & Lawoko, S. (2011). Are Individual and Community Acceptance and Witnessing of Intimate Partner Violence Related to Its Occurrence? Multilevel Structural Equation Model. PLoS ONE, 6(12), e27738. Valcarcel, A., & Samudio, V. (2017). Colombia: Durable solutions for the forcibly displaced. Forced Migration Review, 56 (Latin America and the Caribbean: building on a tradition of protection). van der Gaag, N., Heilman, B., Gupta, T., Nembhard, C., & Barker, G. (2019). State of the Worlds Fathers: Unlocking the Power of Men’s Care. Promundo. Vélez, O. R., & Bello, A. H. (2010). Family Breakdown in Colombia. Forced Migration Review, 34, 69–70. Vu, A., Adam, A., Wirtz, A., Pham, K., Rubenstein, L., Glass, N., Beyrer, C., & Singh, S. (2014). The Prevalence of Sexual Violence among Female Refugees in Complex Humanitarian Emergencies: A Systematic Review and Meta-analysis. PLoS Currents. Vyas, S., & Heise, L. (2016). How do area-level socioeconomic status and gender norms affect partner violence against women? Evidence from Tanzania. International Journal of Public Health, 61(8), 971–980. Whitehead, A., & Tsikata, D. (2003). Policy Discourses on Women’s Land Rights in Sub- Saharan Africa: The Implications of the Re-turn to the Customary. Journal of Agrarian Change, 3(1–2), 67–112. Wirtz, A. L., Pham, K., Glass, N., Loochkartt, S., Kidane, T., Cuspoca, D., Rubenstein, L. S., Singh, S., & Vu, A. (2014). Gender-based violence in conflict and displacement: Qualitative findings from displaced women in Colombia. Conflict and Health, 8(1), 10. World Bank. (2019). Colombia Gender Assessment. The World Bank Group. World Bank. (2020). Women, Business and the Law 2020. World Bank Group. 38 Appendix Table A1. Balance test, 2005 Unmatched Matched (common support) Variable Treated Control % bias P-value Treated Control % bias P-value Age (base: age 35-49) Age: 13-17 0.12 0.11 3.20 0.14 0.12 0.12 0 1 Age: 18-24 0.21 0.20 3.60 0.09 0.21 0.21 0 1 Age: 25-34 0.27 0.25 4.00 0.06 0.27 0.27 0 1 Education Years of education 7.02 8.66 -42.60 0.00 7.04 7.04 0 1 Marital status (base: single) Married 0.52 0.50 3.70 0.09 0.52 0.52 0 1 Widow 0.03 0.02 7.50 0.00 0.02 0.02 0 1 Separated or divorced 0.14 0.11 7.20 0.00 0.14 0.14 0 1 Geographic area Urban 0.79 0.75 9.80 0.00 0.79 0.79 0 1 Massacres in t-1 or t-2 (origin) 0.66 0.30 78.40 0.00 0.66 0.66 0 1 Survey year Year (2010) 0.48 0.39 18.00 0.00 0.48 0.48 0 1 Year (2015) 0.22 0.30 -18.80 0.00 0.22 0.22 0 1 Note: Figures are in percentages (%) or as indicated. The p-values are for the two-sample t-test with equal variances (Ho: Difference in means =0 and Ha: Difference in means ≠0). Variables capturing household size and composition are omitted from the matching because they are likely to be affected treatment. Table A2. Balance test, 2010 Unmatched Matched (common support) Variable Treated Control % bias P-value Treated Control % bias P-value Age (base: age 35-49) Age: 13-17 0.12 0.11 5.20 0.09 0.12 0.12 0 1 Age: 18-24 0.22 0.19 5.90 0.06 0.21 0.21 0 1 Age: 25-34 0.27 0.25 3.00 0.34 0.27 0.27 0 1 Education Years of education 6.57 8.61 -53.60 0.00 6.59 6.59 0 1 Marital status (base: single) Married 0.51 0.51 2.00 0.53 0.52 0.52 0 1 Widow 0.02 0.02 4.40 0.13 0.02 0.02 0 1 Separated or divorced 0.13 0.11 5.50 0.07 0.13 0.13 0 1 Geographic area Urban 0.75 0.73 4.10 0.20 0.75 0.75 0 1 Massacres in t-1 or t-2 (origin) 0.84 0.29 130.90 0.00 0.83 0.83 0 1 39 Table A3. Balance test, 2015 Unmatched Matched (common support) Variable Treated Control % bias P-value Treated Control % bias P-value Age (base: age 35-49) Age: 13-17 0.12 0.11 3.00 0.51 0.11 0.11 0 1 Age: 18-24 0.22 0.20 4.70 0.30 0.22 0.22 0 1 Age: 25-34 0.27 0.26 2.10 0.64 0.27 0.27 0 1 Education Years of education 8.21 9.18 -26.00 0.00 8.24 8.24 0 1 Marital status (base: single) Married 0.52 0.50 3.60 0.44 0.52 0.52 0 1 Widow 0.02 0.01 2.60 0.55 0.01 0.01 0 1 Separated or divorced 0.16 0.11 14.80 0.00 0.16 0.16 0 1 Geographic area Urban 0.82 0.75 18.00 0.00 0.82 0.82 0 1 Massacres in t-1 or t-2 (origin) 0.84 0.22 159.70 0.00 0.84 0.84 0 1 40