In integrating women and gender into development, what is the significance of the following: the 1970 publication of Ester Boserup’s landmark writing-Women’s Role in Economic Developm

Engendering development and disasters Sarah Bradshaw Principal Lecturer, Development Studies, Department of Law and Politics, University of Middlesex, United Kingdom

Over the last two decades the different impacts of disasters on women and men have been acknowledged, leading to calls to integrate gender into disaster risk reduction and response. This paper explores how evolving understandings of ways of integrating gender into development have influenced this process, critically analysing contemporary initiatives to ‘engender’ development that see the inclusion of women for both efficiency and equality gains. It has been argued that this has resulted in a ‘feminisation of responsibility’ that can reinforce rather than challenge gender relations. The construction of women affected by disasters as both an at-risk group and as a means to reduce risk suggests similar processes of feminisation. The paper argues that if disaster risk reduction initiatives are to reduce women’s vulnerability, they need to focus explicitly on the root causes of this vulnerability and design programmes that specifically focus on reducing gender inequalities by challenging unequal gendered power relations.

Keywords: disasters, feminisation of responsibility, gender, men, poverty, women, vulnerability

Introduction While ‘disaster’ is a contested notion (Cardona, 2004; Oliver-Smith, 2002; Quarantelli, 1998), it is usually understood that one occurs when an individual or group is vul- nerable to the impact of a natural or human-made hazard, i.e. they are unable to ‘anticipate, cope with, resist and recover from’ an event (Blaikie et al., 1994). Being a woman does not in itself lead to greater vulnerability, but women may be more vulnerable to hazards than men, given the unequal gendered power relations that limit women’s access to and control over resources. Vulnerability to an event is not based on sex or biological differences between men and women, but rather due to how society constructs what it means to be a man or a woman—i.e. what roles they should play and how they should behave—and this also influences how risk is per- ceived and responded to, with the concept being understood differently by men and women (Gustafson, 1998). Their lack of information, education and engagement with preparedness activities means that women faced by a perceived risk often do not feel that they can act, do not know when to act or do not know how to act on warnings. Socially constructed roles and norms mean that women cannot leave their homes without male permission, for example, or their roles as carers for children and the elderly slow their escape, meaning more women than men may die in an event. While reliable fatalities data disaggregated by gender and generation is still largely missing (Mazurana et al., 2011), a study by Neumayer and Plümper (2007) concluded that in countries where a disaster has occurred and where the socioeconomic status of women

doi:10.1111/disa.12111

Disasters, 2014, 39(S1): S54−S75. © 2014 The Author(s). Disasters © Overseas Development Institute, 2014 Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

Engendering development and disasters S55

is low, more women than men die, or they die at a younger age. For those women who do survive there may be longer-term and more intangible ‘secondary’ impacts such as a rise in violence or greater insecurity in employment (Bradshaw and Fordham, 2013). Disasters, then, are gendered events. The last two decades have seen an increased interest in ensuring a gender perspec- tive in post-disaster response efforts, and more recently there have been initiatives to mainstream gender into disaster risk reduction (DRR) initiatives. The Hyogo Framework for Action (HFA), which provides the global framework for DRR, is a good example of how this has been approached. Its opening section states that a gender perspective should be ‘integrated into all disaster risk management policies, plans and decision-making processes, including those related to risk assessment, early warning, information management, and education and training’ (UNISDR, 2005, p. 4). However, while it has been suggested that this provides the ‘most explicit reference to gender of any other international policy frameworks for DRR’ (see UNISDR, 2009), it is not without limitations. Most importantly, its call to integrate gender into all areas of DRR did not result in gender being integrated even into the HFA itself, and in the remainder of the document gender and women are mentioned only twice: once when discussing early-warning systems and once when discussing the need to ensure equal access to appropriate training and educational opportuni- ties. This suggests a lack of real commitment to adopting a gender perspective by the international agencies responsible for DRR. That there is a lack of commitment to integrating gender into DRR was a view shared by participants at the International Conference on Gender and Disaster Risk Reduction in 2009, who noted that gender ‘remains a marginalized issue in the current national and international negotiations around DRR and climate change adaptation’ and that gender considerations have been ‘hardly applied as a fundamental principle in policy and framework development’ (cited in UNISDR, 2009, p. 7). A survey in 2010 by the UN International Strategy for Disaster Reduction also suggests that gender is not being prioritised by those working in DRR. The study reviewed over 14,000 diverse pieces of material and classified them into themes and issues. This was followed by a survey of 1,856 DRR professionals to ascer- tain broad acceptance of this categorisation. It also asked them to indicate if these themes and issues were areas of specialisation and which areas needed more expertise (UNISDR, 2011). Analysis of the results shows that the highest number of respond- ents stated that ‘disaster risk management’ was their area of ‘expertise’ (61 per cent), and this also scored highly as an area needing strengthening (36 per cent). Newer areas that are key to disaster ‘risk reduction’ scored lower on both expertise and as areas needing strengthening. For example, 42 per cent of respondents said they had expertise in risk identification and only 28 per cent thought it was an area that needed strengthening, while climate change scored lower still on expertise (20 per cent), but was more readily identified as an area that needed greater expertise (35 per cent). However, the theme that scored lowest on both expertise and recognition of the need to strengthen knowledge of this area was ‘gender’, with only 13 per cent suggesting

Sarah Bradshaw S56

they had expertise in the area and 13 per cent suggesting it was an area that needs more expertise. It seems, therefore, that there is still a long way to go to convince DRR profes- sionals of the importance of adopting a gender perspective. In contrast, development professionals seem to have been more open to incorporating gender into their pro- jects and policies. The ‘success’ of efforts to integrate gender into development is witnessed by the World Bank’s commitment to ‘engender’ development via the adop- tion of a gender mainstreaming strategy in 2001, while even the most male-dominated of all the development agencies, the International Monetary Fund (IMF), proclaimed in 2012 that empowering women is just ‘smart economics’. The evolution of how women and gender have been integrated into the develop- ment process is not only more ‘advanced’ than the same processes in disasters, but also much better documented (see El Bushra, 2000; Kabeer, 1994; Moser, 1993; Ostergaard, 1992; Saunders, 2002; Young, 1993). This paper seeks to begin to redress the balance, exploring how gender has been incorporated into disaster policy and practice to date. It draws on the processes to ‘engender’ development in order to better understand processes to integrate gender into disasters. It begins with a consideration of the early stages of integrating women into development, before exploring the advances made in promoting gender as an issue in disasters. It then examines how gender has been integrated into relief and reconstruction in practice, using the debates within gender and development to problematise this. The next section considers more recent advances in engendering development, examining the notion of gender mainstreaming and how this operates in development policy, using the World Bank as a case study. The final section extends these ideas of engendering development to the disasters field, highlighting lessons to be learned.

Integrating women and gender into development Initial theorising around development presented the process as gender neutral. While the two main theories of development—modernisation theory and dependency theory—make no explicit mention of women or gender, the inherent assumption was that policies and projects that help men would also help women. The turning point came in 1970 with the publication of Boserup’s landmark text Women’s Role in Economic Development, which effectively demonstrated that ‘development’—or, rather, processes of modernisation—can harm women. While critiqued in turn (see Saunders, 2002), her work was ground breaking at the time since it illustrated that policies were not gender neutral, but based on gendered assumptions. Boserup’s work made clear the negative impact development could have on women and instigated a movement to integrate women into development. In particular, it is said to have helped inspire the UN Decade for Women (1976–1985) and a series of UN World Conferences on Women starting with a meeting in Mexico in 1975, followed by meetings in Copenhagen (1980) and Nairobi (1985). The fourth and last full meet- ing was held in 1995 in Beijing and the resultant Beijing Platform for Action now acts

Engendering development and disasters S57

as a framework for annual sessions to review progress. The movement to integrate women into development is generally seen to have two key ideological threads— Women in Development (WID) and Gender and Development (GAD). WID emerged in the 1970s and set out to deliberately target more development resources at women. It has its roots in liberal feminism and was and is the acceptable face of engendering development, and, it could be argued, also more recently disas- ters. Women’s subordination was seen to stem from their exclusion from the market sphere, and their limited access to and control over resources. The key to change was seen to be via laws and legislation around education and employment to effectively place women ‘in’ existing processes, i.e. the aim was to maximise women’s access to the ‘modern’ sector. While women’s significant productive contribution was made visible, their reproductive role was downplayed. Women’s ‘problem’ was diagnosed as insufficient participation in a benign development process because of an oversight by policymakers. As the name suggests, WID sought to integrate women into exist- ing development processes. However, this integration did not stretch to integration into development agencies, and instead WID led to the setting up of separate offices and officers (Rathgeber, 1990). For many, while WID interventions have made women more visible in the devel- opment process, the way in which this has taken place is problematic (Bandarage, 1984; Jaquette, 1982; Parpart, 1993). Critics questioned if you could—and, more importantly, should—just try and integrate women, or ‘add women and stir’. Those who questioned the dominant development discourse of modernisation also ques- tioned the benefits for women that being more integrated into this process would bring. There was also a suggestion that the focus on women constructed them as ‘the problem’, rather than locating this in the unequal gender roles and relations that form the basis of gender subordination. While most literature presents GAD as emerging as a counter-movement to WID, it is important to note that one other development occurred before the move from WID to GAD. Women And Development (WAD) saw one very small change lin- guistically, but one very large shift ideologically. It could be suggested that if WID was a reflection of modernisation theory, then WAD is a gendered dependency theory. WAD located gender struggle in the structure of capitalism, so—as with Marxist feminists—it privileged capitalism over patriarchy. For this reason the extent to which it actually had an explicit aim to ‘engender’ development or the extent to which it was ‘properly feminist’ was questioned (Saunders, 2002, p. 7), and while it is part of the WID–GAD progression, it is often overlooked. The more usual counterpoint to the WID school has been GAD. Here two changes may be seen to arise in response to the criticisms levelled at WID above—from ‘in’ to ‘and’, and ‘women’ to ‘gender’. The change to ‘gender’ reflects the fact that it is not women that are the problem, but the unequal power relations between men and women, while ‘and’ suggests the need to explore both gender and development. In other words, it is not sufficient to add women to existing processes of development, but it is also essential to problematise such development. The focus on gender suggests

Sarah Bradshaw S58

that rather than looking at women in isolation, there is a need to address the imbal- ance of power between women and men. The resultant projects are not just about improving women’s access to income generation or girls’ access to education, but are focused on challenging the social norms that keep girls from attending school and the structural inequalities that mean women earn less than men for comparable work, for example. This approach also gives special attention to the oppression of women in the family and highlights the need for policy initiatives to enter the ‘private’ sphere. While attractive in theory, GAD has its origins in academic thinking and may be seen to be a more theoretical understanding than the practically oriented focus of WID. Initially non-governmental organisations (NGOs) embraced GAD, but more recently institutions such as the World Bank have stated that they follow a GAD approach. However, while many say they have a GAD approach, what they often do in practice is WID (Rathgeber, 1990)—a criticism that may also be levelled at recent attempts to gender disasters (see below). Although WID and GAD are the two mainstream approaches to engendering development adopted by both international development agencies and NGOs, other approaches also exist. In particular those working with grass-roots women’s organi- sations and Third World feminists became disillusioned with both WID and GAD. They questioned the extent to which it was possible to generalise about women’s situation and position, suggesting such thinking had led to the construction of Third World women as a homogeneous group of ‘benighted, overburdened beasts, help- lessly entangled in the tentacles of regressive Third World patriarchy’ (Parpart, 1995, p. 254). In contrast, the Third World feminist tradition highlighted diversity and difference (see Mohanty, 1991), emphasising that inequalities of power exist on many levels, not just male/female, but rich/poor and First World/Third World. It suggested the need to recognise that women inhabit multiple sites of oppression and what is seen to be a key issue for a low-income woman may be as much class-based as gender based-discrimination. Just as there are multiple sites of oppression, there is also the need to address multiple inequalities of power. The ‘empowerment approach’ places a consideration of power relations as central (see Rowlands, 1997). The critique from women from the Global South that ‘Third World’ women were being constructed by WID/GAD as a homogeneous group of oppressed ‘others’ (Mohanty, 2003) promoted debate around both approaches, but they were also being questioned by another set of critics. By the end of the 1990s what had variously been defined as ‘men in crisis’ or ‘troubled masculinities’ were recognised (Chant, 2000, p. 7). This supposed crisis arose in part from the success of WID/GAD, which meant that girls were ‘over-achieving’ compared to boys in education, for example, but also processes of globalisation had seen the rise of women’s employment and men’s under- or unemployment and, related to this, women assuming decision-making roles in households. At the same time men had been generalised as lazy in the development discourse (Whitehead, 2000), sitting around talking while women work, and it was suggested that gender discourse had constructed men as ‘lurking’ in the background imagined as ‘powerful and oppositional figures’ (Cornwall, 1998, p. 46), and that

Engendering development and disasters S59

through this ‘gender roles and relations’ had become shorthand for inherently oppo- sitional relations. In practical if not in ideological terms, the case had been presented that men needed to be explicitly included in development. From the start some questioned the exist- ence of this ‘crisis of masculinity’ (see Chant and Gutmann, 2000), not least since problems of ‘masculinity’ might be better read as problems emerging from processes of globalisation and related to a crisis in international capitalism rather than gender. The masculinity projects that sprang up led to more questioning, since many pro- jects focused on male subjectivities—the personal constructions and understandings of ‘maleness’ and the implications of this for relationships with others. The focus reinforced men as self-reflecting subjects and complex individuals, yet the ‘Third World woman’ remained positioned as an object of subordination, not least since women rarely have the luxury to reflect on their ‘femaleness’—a luxury even less accessible when men’s projects compete for limited funding from the gender funding pot. More importantly for some, masculinity projects seemed to forget the vested interest men have in resisting change and the inherent contradiction in the aim to ‘empower men to dis-empower themselves’ (Redman, 1994). Pearson (2000, p. 44) suggests the construction of men as being in crisis at that time might have little to do with any real crisis of masculinity, but rather may be seen ultimately as aiming to dilute the radical women-focused development agenda that had arisen. More generally, by the turn of the millennium the advances made in pro- moting women’s rights were under attack. The 1990s had witnessed the ‘rise of rights’ (Eyben, 2003), with many organisations and international development agencies adopt- ing some form of ‘rights-based approach’ to development (Molyneux and Lazar, 2003; Piron, 2005), and within this the concepts of reproductive health, reproductive rights and sexual rights had become popularised (Petchesky, 2000). While the rights-based approach has not been without its critics (see IDS, 2005; Molyneux and Cornwall, 2008; Tsikata, 2004), the potential of rights for increasing recognition of women’s demands as legitimate claims has made rights particularly attractive to women’s move- ments. Some of the most effective organising over the past two decades has been around rights-related claims (Antrobus, 2004; Ruppert, 2002), including perhaps the greatest achievement—recognition of the fact that women should live free from violence as a ‘right’. Yet despite the agreements made at international conferences on women’s rights and the advances made in including women in development practice, for many gender activists the Millennium Development Goals (MDGs) highlighted that a backlash had begun. Launched in 2000, the MDGs were to frame the development agenda until 2015. What was included and excluded has been important in ideological and funding terms. While the MDGs contain a goal focused on gender equality and women’s empower- ment, they make no mention of sexual and reproductive rights, while ending vio- lence against women is also missing. This led some gender activists to declare that the abbreviation MDG is better understood as ‘Most Distracting Gimmick’ (Antrobus, 2004). Since the MDGs were formulated there have been some more positive devel- opments, such as the formation of UN Women in 2010. This merged four key UN

Sarah Bradshaw S60

initiatives promoting the empowerment of women and seeks to hold the UN system accountable for its own commitments on gender equality. UN Women also seek to influence inter-governmental bodies in their formulation of policies, global standards and norms, including, as the MDGs come to an end, the ongoing process to formu- late a post-2015 development agenda and a new set of ‘Sustainable Development Goals’ (SDGs). The illustrative goals presented to the UN by the High Level Panel in June 2013 suggests advances from the MDGs, with violence against women proposed as a specific target and the explicit mention of sexual and reproductive health and rights. However, it is the Open Working Group of the General Assembly that has the ultimate responsibility for agreeing on a proposal on the SDGs, and a communiqué in February 2014 at the conclusion of the ‘stocktaking phase’ of the process suggests that the inclusion of sexual rights in any new set of goals is far from agreed. While the ‘exclusion’ of gendered rights from initiatives such as the MDGs has been a focus for gender activists of feminists over the decades, more recently the inclusion of gender has been problematised, and nowhere more so than in the poli- cies and projects of the World Bank and its plan to engender development through gender mainstreaming. Before considering the contemporary development land- scape further, an examination of processes to date to integrate gender into disasters is first needed.

The road to gendering disasters While Boserup’s work has been identified as the moment when a movement toward integrating gender into development emerged, no such defined movement to integrate gender into disasters in global agencies and agendas or official disaster discourse exists. Outside of the formal processes such as UN initiatives, perhaps the most coherent ‘movement’ in the gender and disasters field is the Gender and Disasters Network (GDN). This being said, many people working in the field of gender and disasters remain unaware of this organisation and its work. Conceptualised in 1997 as a re- sponse to a gender gap in disaster analysis and practice, it is presented as unique in its ‘strategy of engendering Disaster Risk Reduction by taking advantage of the virtual space created by the World Wide Web’ and it has become a ‘legitimate and respected voice for gender and disasters issues’ (Sanz et al., 2009, p. 15). Just as research was a key factor in initiating the WID movement and as the basis of the GAD approach, the importance of promoting and disseminating quality research has not been lost on the founders and members of the GDN, and it has become a repository for gender and disaster knowledge. Tierney (2012, p. 245) suggests that the field of disaster research has been charac- terised by a series of ‘critical disjunctures’, with discontinuities in research and the systematic neglect of some topics, and even collective resistance to the introduction of new ideas. In the field of disaster studies gender might be seen to be one such topic, and it has suffered from ‘periodic lapses of attention’ (Anderson, 2012). In 1982 Rivers published a key paper exploring gender differences and discrimination in disasters,

Engendering development and disasters S61

and since then the production of gendered disasters knowledge has made large advances. The end of the 1990s saw one of the main disasters journals, the International Socio- logical Association’s International Journal of Mass Emergencies and Disasters, publish its first special edition dedicated to women and disasters, and in the interim decade a number of edited texts have focused on gender and disasters, starting with The Gendered Terrain of Disasters (Enarson and Morrow, 1998). In 2009 Fordham edited a special edition of the journal Regional Development Dialogue dedicated to ‘Gender and Disas- ter Management’, and in the same year an edited collection on Women, Gender and Disasters (Enarson and Chakrabarti, 2009) was published. These texts include con- tributions from a wide range of scholars and practitioners from a variety of countries, including scholars from the Third World, demonstrating how the field has expanded. While texts have considered disasters and development (see, for example, Pelling, 2003), it was not until 2013 that the first text dedicated to better understanding gen- der and disasters in the developing-world context was published (Bradshaw, 2013). Towards the end of the 20th century events such as Hurricane Mitch in Central America saw advances in terms of understanding how gender roles, relations, and identities are constructed and reconstructed in the context of disasters (see, for exam- ple, Bradshaw, 2001; Cupples, 2007). High-profile events during the early part of the new millennium, including the 2004 Indian Ocean tsunami, meant an upsurge in case studies and survivor narratives (see, for example, Oxfam International, 2009) and accounts from field workers that focused on both women and men (see, for exam- ple, Clarke and Murray, 2010). Research after Hurricane Katrina allowed new voices to be heard (see, for example, David and Enarson, 2012) and brought new directions in gendered research (see, for example, D’Ooge, 2008; Fothergill and Peek, 2008; Gault et al., 2005), and advances in the theoretical discussions on the meaning of disasters (see, for example, Brunsma, Overfelt and Picou, 2007). The GDN Gender and Disasters Sourcebook (2006) initiative has sought to disseminate existing studies such as these and to publicise good practice by making resources available via the internet to academics and practitioners. Although such virtual forums are important, a number of conferences have also brought together those working in the field, illustrating how thinking around gender in the disaster context has evolved. The 1998 ‘Women and Disaster’ conference in Vancouver was followed in 2000 by the meeting in Miami with the theme ‘Reaching Women and Children in Disasters’. In 2004 about a hundred women and men met to discuss ‘Gender Equality and Disaster Risk Reduction’. The change in titles from ‘Reaching Women and Children’ at the 2000 conference to ‘Gender Equality’ at the 2004 conference parallels the shift in language witnessed in the gender and develop- ment field. For the first time, during the 2004 meeting a breakaway discussion was held largely consisting of men. It identified the current practice of focusing discus- sions on identifying and mitigating women’s vulnerabilities as ‘limiting in the long run’ and called for engaging with ‘men and boys in equal measure’ (Mishra, 2009, p. 35). This move toward reasserting masculinities also demonstrates parallels with the changing and contested gender and development discourse.

Sarah Bradshaw S62

The 2004 conference led to the Honolulu Call to Action (see Anderson, 2009), which calls for disaster risk, gender, social equity, and environmental issues to be considered in an integrated way and highlights the gaps in the MDGs in terms of DRR and gender. The GDN has since gone on to seek to influence the interna- tional policy dialogue, with a delegation being invited to make an oral and written statement at the 2007 ‘Global Platform for Disaster Risk Reduction’ in Geneva, for example, and through involvement in the discussions around the development of the second phase of the HFA post-2015. Some progress has therefore been made since the late1990s, when Fordham (1998) noted that the incorporation of a gender focus into disaster work had still not advanced much further than revealing the situation of women, and that—if addressed at all— gender had been integrated into disaster practice as a demographic variable or per- sonality trait and not as the basis for a complex and dynamic set of social relations (Enarson, 2000). Yet as late as Hurricane Katrina in 2005 the ‘not noticing’ (Saeger, 2012) of the gendered dimension of the disaster by the media and expert responders alike was highlighted, and despite the advances made by those working in the field, it seems that gender is yet to be fully mainstreamed in humanitarian relief, integrated into research and field projects undertaken by the major disaster centres, and included in disaster training courses (Enarson and Meyreles, 2004). The gendering of the dominant disasters discourse has advanced slowly. Where the most rapid advances have been made is not in the integration of women into DRR, but the inclusion of women in post-disaster reconstruction activities.

Women in development and disasters in practice In the contemporary post-disaster context, studies suggest that women are included in the reconstruction projects, and are targeted as beneficiaries and participants. For example, as far back as 1998 an evaluation of projects post-Hurricane Mitch financed with funds through the Disasters Emergency Committee (ECA, 2000) concluded that the projects ‘tended to favour women and children in the distribution of products and services’ and that women were included as:

• beneficiaries of the self-construction of houses with the title issued in the woman’s name;

• participants in construction projects (roads, houses, bridges), which helped break certain stereotypes regarding their capacity to work; and

• beneficiaries of productive programmes in order to reduce their economic vulner- ability. These included chicken rearing, agricultural project