Introduction

The ineffectiveness of gender mainstreaming and gender-responsive interventions has been debated for decades. Various approaches to integrating gender as a fundamental part of policy processes, such as Women in Development (WID) and Gender and Development (GAD), have been intensively discussed in the literature (Razavi and Miller 1995). Since the 1995 Fourth World Conference on Women in Beijing, gender mainstreaming has been widely adopted as the preferred approach to addressing gender inequality, with development stakeholders promising to place gender equality issues at the centre of policy decisions. The United Nations Economic and Social Affairs (2000, p. 2) has defined gender mainstreaming as “…the process of assessing the implications for women and men of any planned action, including legislation, policies or programmes, in all areas and at all levels […] The ultimate goal is to achieve gender equality”. The United Nations (2002, p. 2) also stresses that “the mainstreaming strategy must be adapted to the particular subject under discussion” and that “there is no set formula or blueprint that can be applied in every context”. In sum, gender mainstreaming is a strategic instrument for achieving gender equality as an ultimate development goal of human rights and social justice, but not a strategic objective in itself. The mainstreaming process, which starts with gender analysis models that are “not static and […] reflect the changing contexts in which they operate” (Bacchi and Eveline 2010, p. 68), aims to identify, articulate, and define strategic objectives based on country and sector-specific contexts. However, the inclusion of gender analysis into the policymaking process alone is not sufficient to validate the strategic objectives derived from it, as policymaking involves multiple political interests and epistemological perspectives of policymakers. Nevertheless, much attention is still given to the mainstreaming strategy itself in policy evaluation (Brouwers 2013).

The inadequacy of gender mainstreaming in the policy process has been demonstrated in numerous studies around the world. For instance, mainstreaming agendas tend to depoliticise and trivialise the problem of gender inequality (Bock 2015; Acosta et al. 2019), fail to articulate their objectives (Verloo 2005), lose hold of feminist political objectives (Milward et al. 2015), are ineffectively implemented in practice (Shortall and Marangudakis 2022; Devkota et al. 2022), and have an integrationist and technocratic approach (Allwood 2013). In addition, feminist scholars criticising the adoption of gender in development policies point to the tendency to focus on economic rationality, failing to tackle gendered power relations or structural inequality (Ferguson 2010; Chant 2012). As a result, development policies tend to contribute to entrenching and perpetuating gender inequality through gender interventions (Razavi 2007; Prügl and Joshi 2021). These critiques raise the question of whether the problem lies in the failure of mainstreaming itself or in the problematisation generated by the process of mainstreaming.

It is precisely this question that we seek to explore in this article. Rather than assessing the effectiveness of gender mainstreaming in policy, we direct our attention toward the problem definition that sets the policy agenda for gender equality. For this, we follow the Foucauldian understanding of policy as a discourse that produces effects as a body of knowledge (Bacchi and Eveline 2010; Prügl 2011). Rather than assuming the legitimacy of current gender narratives in agricultural and rural development (ARD) policy, this study questions how gender policies as discourses create specific problems, produce effects, and govern subjects.

We have chosen to focus on the contexts of Myanmar and Nepal, which are particularly interesting as they are at different stages of institutionalising gender in their policies. While the Nepalese government has made gradual progress in integrating gender into its sectoral policies since the 1980s, with a commitment to gender-responsive budgeting (FAO, 2019), gender mainstreaming in Myanmar remains nascent (Chakraborty 2016). Despite their different contexts and histories of gender integration, gender inequalities persist in Nepal and Myanmar, which rank 96th and 106th, respectively, in the 2022 Global Gender Gap Index (The World Economic Forum 2022).

Below, we discuss the value of applying Bacchi’s “What’s the Problem Represented to be?” (WPR) approach, drawn from Foucauldian influenced poststructural methods of policy analysis, to understanding gender discourses in ARD policies, and present our research questions. We then provide country-specific gender contexts for Myanmar and Nepal. This is followed by the methodology, and finally, the findings are presented and discussed with respect to the research questions.

Analytical framework: the WPR approach

According to Bacchi (2000), progressive social change is difficult to achieve given that the way problems are represented conceals power relations, diverting attention away from the structures that create inequality and holding individuals responsible for their own failures. Thus, she stresses the importance of analysing how policy proposals represent ‘problems’ and produce particular discourses and suggests analysing the definition and representation of the problem rather than policy tools or proposed solutions in order to understand policy outcomes. Bacchi perceives discourse as “the knowledges through which we are governed,” reflecting the Foucauldian view of policy as constitutive practices that shape our reality (2017, p. 21). In this vein, Bacchi and Goodwin argue (2016, p. 23) that the way in which issues are represented (problematised) in policies generates 'discourse’, which determines a set of interconnected effects including discursive effects (“what can be thought and said”), subjectification effects (“how they are produced as specific kinds of subjects”) and lived effects (“which discursive and subjectification effects translate into people’s lives form part of the analysis”).

Bacchi pays particular attention to how we are constituted as subjects through latent problematisations in policy texts (2009). For example, in policy discourse, certain groups—such as the disabled or otherwise disadvantaged—are assigned and defined positions that can disempower these ‘created’ groups (Bacchi 2000, p. 54). In this regard, Bacchi emphasises the importance of studying policies as gendering practices, which she describes as a process of constituting ‘women’ and ‘men,’ sometimes perpetuating and reinforcing traditional gender norms (Bacchi 2017). She also argues that by assigning people to positions, policies set groups of people against each other, such as a group responsible for the problem versus those who are not, in what Foucault calls ‘dividing practices’ (Bacchi 2009). Furthermore, analysing policies as gendering and dividing/differencing practices calls our attention to intersectionality, which explains the interaction of “a range of social dynamics—gendering, racializing, heteronorming, disabling and so on” (Bacchi 2017, p. 34). Such a perspective is useful because it shows how and where power is concentrated and exercised against others (ibid.), and how it creates multiple sites of marginalisation that government services tend to neglect.

By introducing the WPR approach into policy analysis, Bacchi allows us to interrogate how policies represent specific issues as problems and leave others unproblematised; how the way in which problems are represented directly affects people’s lives; who benefits from a policy and who does not; and what epistemological approaches are chosen to seek solutions. Bacchi suggests that the first step in policy analysis is to examine the proposal in order to identify what is problematised. Using the example of a policy proposing training programmes for women in order to increase their job opportunities, she explains that the offer of training as a solution assumes that women’s lack of training is ‘the problem’ (Bacchi 1999). According to Bacchi, this presumes that women need to change, silencing other potential problems, such as working conditions. Furthermore, designating women as the problem affects how women perceive themselves and how others perceive them. As a result, some women avoid training out of fear of stigmatisation, which impedes significant societal change (ibid.). Adopting Bacchi’s idea and basic questionsFootnote 1 (Bletsas and Beasley 2012, p. 21), we began our analysis by examining the gender proposals prescribed by ARD policies with the following research questions.

  1. (1)

    What is the ‘problem’ represented to be?

  2. (2)

    What are the assumptions underlying the problematisation?

  3. (3)

    What is left unproblematised or silenced in this representation?

Bacchi’s approach is useful for understanding whether and how gender equality is targeted in policy, as what is problematised represents how policymakers and stakeholders interpret and engage with gender inequality. It allows us to analyse the changes in discursive practices and how these affect policy priorities and governing technologies concerning gender equality. Furthermore, it is useful to understand how the changes (re)constitute rural women as political subjects and how their subjectivities change as a consequence, which Foucault termed ‘subjectification.’ It also provides a relevant space for examining transformational potential in rural communities in countries such as Nepal and Myanmar, where intersecting social markers significantly contribute to gender inequality.

Gender and ARD in Myanmar and Nepal

Myanmar

Myanmar is a south-eastern Asian country bordering India, Bangladesh, China, Laos, and Thailand, and was once the world's leading producer and a major exporter of rice (Shrestha et al. 2022). Although Myanmar has experienced rapid development during its brief period of democratic transition, agricultural and rural development has been slower compared to its neighbours. Seventy-two percent of the population lives in rural areas, where 85 percent of poverty is concentrated; 30 percent of the rural population is landless and has no source of income other than as agricultural labour (ADB 2017). Myanmar is the only country in Asia to have missed out on the Green Revolution. While its Asian neighbours have doubled their agricultural output over the past five decades, Myanmar’s land and labour productivity is the lowest in the region (World Bank 2016). In addition, tensions and conflicts between ethnic minorities and the Burman Union government have long been a source of disruption to agriculture. In comparison to the Burman Buddhist majority, ethnic minorities have received limited public support, including agricultural extension, healthcare, and education (World Bank 2018; Mukherjee 2019; Kipgen 2015). In addition to male migration to neighbouring countries, ethnic minority men have joined ethnic armed organisations, resulting in a heavier labour burden on women in agriculture and rural livelihoods (Hedström and Elisabeth 2020). While the country has undergone a gradual shift from traditional gender norms toward gender equality during the democratic transition, the status of Myanmar women has historically been low, and significant gender disparities exist. Myanmar`s constitution refers to women primarily as mothers, thereby imposing strong gender roles, and it portrays gender issues as existing only among ethnic minority groups (ADB et al. 2016). The gender equality section of the constitution reflects a patriarchal perspective, stating that ‘nothing in this section shall prevent the appointment of men to positions that are naturally suitable for men only’ (ibid, p. xviii). Supported by the Burman Buddhist concept of ‘Phon’, which refers to the innate superiority of men over women, male dominance is accepted in social, religious, economic, and political spheres (Nwe 2010). As a result, in the agricultural sector—where women’s contribution is significant—gender inequalities persist in access to land, extension services, wages, and knowledge (Hansen et al. 2022; UNCTAD 2021; Conner and San 2021).

Nepal

The South Asian nation of Nepal, bordering China and India, is rich in agricultural, ethnic, cultural, and linguistic diversity. Although the country is about to leave the Least Developed Country (LDC) category, multidimensional poverty persists, particularly in rural areas. Factors contributing to rural poverty are multifaceted, including limited access to basic infrastructure and services due to the remoteness and poor connectivity of hill and mountain areas (Sapkota 2018); inequalities due to social, cultural and religious factors that vary across regions (Gentle and Tek 2012; Dahal et al. 2022); and high dependence on agriculture for livelihoods (Shahi et al. 2022). Agriculture accounted for 21.3 percent of the country’s GDP in 2021 and 70 percent of rural livelihoods and employed 56.6% of the nation’s working-age population (World Bank 2023). Crops and livestock account for 75 percent of agricultural GDP. Most farms are small-scale (FAO 2019), with low use of inputs and low labour productivity. Within Nepal`s ethnic and culturally diverse population, ethnicity, caste, gender, religion, age, economic status, geography, and other social markers shape social relations (Spangler and Christie 2020), creating different forms of power dynamics and intersecting vulnerabilities. This is particularly true of gender relations and norms.

In recent years, rural women in Nepal have experienced dynamic transitions in their lives due to rapidly changing political, social and economic conditions, including demographic shifts associated with male migration and agrarian changes—for example, in terms of labour reorganisation, land use and management, and agricultural practices (ibid.). Such transitions have contributed to the shifting, contestation and renegotiation of women's responsibilities and roles as individuals and within households and communities, affecting and reshaping gender norms; power relations (Leder 2022); and women's identity, subjectivity and decision-making power in different and complex ways (Spangler and Christie 2020). At the same time, great disparities persist between women and men in educational level, economic activities, formal employment, income, access to assets, legal rights and mobility (The World Economic Forum 2022). In Nepal, women provide 55–85 percent of all agricultural labour (Dhital 2010); 80 percent are employed in agriculture; and 76.4 percent receive no monetary compensation (FAO 2019). Women own less than 0.1 hectares on average, while less than 10 percent of the land is registered in women’s names (Goodrich et al. 2021). Rural women’s undervalued or unrecognised contribution to agriculture has resulted in lower wages, limited ownership and access to assets and information, and limited control over the sale of agricultural products (ibid). Although Nepal became the first country in the world to recognise a third gender in 2011 (UNDP 2014), those who identify themselves as gender diverse experience discrimination and are rarely discussed in development policy. In both Nepal and Myanmar, intersectional marginalities of gender and other social markers, such as ethnicity, caste, and religion, are the key barriers to achieving gender equality and social justice. While scholarly attention to gender policy research is scarce in both countries, there is a particular lack of attention in the field of agricultural and rural development. A combined search of Scopus, Web of Science and Google Scholar yielded only three relevant studies for Nepal and one for Myanmar.Footnote 2

Methodology

In this study, we aim to examine how gender policies as discourses within the agricultural and rural development sectors of Myanmar and Nepal create specific problems, produce effects, and govern subjects. To do this, we use Bacchi’s WPR approach and critically analyse the representation of the problem, the assumptions underlying the problematisation, and what is silenced in policy documents.

Documents analysed and inclusion criteria

This research examined 87Footnote 3 documents by the Myanmar and Nepali national governmentsFootnote 4 and the top ten donors listed by the Organisation for Economic Co-operation and Development (OECD) for these nations, which contain agricultural and rural development policies, strategies, and implementation guidelines. These included documents in English and the countries’ official languages—Burmese and Nepali—published from 1990 to 2020 and publicly available online. For Myanmar, 32 documents were reviewed for analysis, 16 of which were in English and 16 in Burmese, including 24 documents published by the government and eight by bilateral and multilateral donors. For Nepal, 55 documents were reviewed—34 in English and 21 in Nepali, of which 48 were national policies and strategies and seven were donors’ strategies. Government documents were retrieved from government ministry websites, platforms for international development cooperation—the Myanmar Information Management Unit (http://themimu.info/), and the Aid Management Information System of the Nepalese Ministry of Finance (https://amis.mof.gov.np/). In addition, policy documents in Burmese and Nepali were collected by Burmese and Nepali-speaking researchers using the same criteria as those used for the English documents.

Data analysis

Analysis of the documents by the first author consisted of three stages (See Fig. 1. Document Selection Procedure). First, a deductive coding strategy was used to identify documents containing keywords related to gender and inclusion,Footnote 5 yielding 74 documents for text analysis. Of these 74 documents, 35 were identified that explicitly mention gender equality as one of the policy’s objectives, with proposals to achieve it. These documents were used to analyse the problem representation, assumptions, and silences within ARD policies. To ensure inter-coder reliability and validity of the analysis, the first author repeated the analysis for each country for the English document review using Atlas.ti. For the Burmese and Nepali language documents, a double coding strategy was used with independent coders following the same coding strategy and protocol as that carried out for the English documents and the data was compared and validated by the first author in collaboration with three Burmese and two Nepali speaking researchers to ensure accuracy and transparency.

Fig. 1
figure 1

Document selection procedure

Full size image

Findings

Problematisation of women’s lower participation

As Bacchi suggests, we examined policy proposals starting from what is represented as a problem, finding one prominent problematisation: the low level of women’s participation in agricultural and rural development activities, such as joining farmer and water or land user groups, and in community-level decision-making processes. We identified 35 policy documents that proposed women's participation as a key solution to gender inequalities and two documentsFootnote 6 that proposed women's participation without mentioning gender equality. While more than 40 percent of the documents mentioned gender equality, no substantive proposals were made other than promoting women’s participation. These documents indicate that in ARD policies, the two countries have different levels of commitment to gender equality and gender mainstreaming. While the Nepalese government has an overarching policy of mainstreaming at the national level, in Myanmar, there is little recognition of this issue in policymaking. Despite this difference, increasing women’s participation at the individual level, limited to development activities, was predominantly represented as a solution to gender inequality in both countries. Those policies that propose increasing women’s participation generally did not discuss how and why participation would promote development and achieve gender equality. Instead, gender inequality is implicitly or explicitly reduced and equated to women’s lower level of participation.

Whereas lower participation is one of the symptoms of gender inequality, it is termed and framed as a ‘women’s’ problem, instead of the exclusion of women being problematised. Many simply added ‘increased women’s participation’ as a general prescription without diagnosing specific problems or causes of inequality or elaborating on whether or how participation would provide a solution. When women’s participation was mentioned, it was usually in relation to cooperatives, producer groups, water or land user groups, and village development committees. Myanmar’s Agricultural Development Strategy and Investment Plan (ADSIP) document encourages women’s participation in farmer groups in order to facilitate their collective access to productive resources such as credit, savings, training, and inputs to “improve the performance” of women (2018, p. 27), while Nepal’s Agricultural Development Strategy (ADS) does the same “to achieve economies of scale in technology dissemination, marketing, finance, and logistics” (2015, p. 7). MoAD (2004) declares that the participation of women in all possible fields will be increased to 50 percent to improve production and productivity, without mentioning gender inequality as a goal. The policies proposed in these documents mandated participation by a minimum 30, 33, or 50 percent of women in ARD programmes, depending on the policy.

None of the policies of the two countries seeks to comprehend what inhibits participation, whether participation would effectively enhance access to resources, or whether their productive performance (agricultural yield and income) depends on such resources. There is also very little discussion of women’s participation in the decision-making process in rural development in general. Where it is mentioned, it is limited to women’s mandated numerical representation in community-level decision-making and occasionally in government-appointed committees. Nepal’s policies promote women’s leadership more than Myanmar´s policies do. Interestingly, Myanmar documents do not specifically mention women as policy targets, rather referring to ‘men and women’ collectively and using non-committed language when mandating women’s participation—for example, “women may participate in land use decision-making processes”(MoECAF 2016, p. 36); “a minimum quota for women’s participation mandated when necessary” (MoALI 2018, p. 33); and”this can include quotas for including women” (MoNREC 2019, p. 104; italics added by the authors).

Overall, the policies reviewed problematised ‘women’ and their lower or non-participation, particularly in the agricultural domain, and thus propose increasing women’s participation as a solution to gender inequalities. These proposals also make a quantum jump from numerical participation to empowerment to gender equality without providing evidence or assessing barriers to women’s participation. Thus, several underlying assumptions are involved in representing women’s participation as a predominant problem, as will be discussed below.

Underlying assumptions

Gender equality as a tool for economic development

The most prominent underlying assumption is that gender equality is an instrument for achieving economic development. Policies consider agriculture primarily as an industry, and economic development as the principal goal of agricultural and rural development. Nepal’s key strategies for agricultural development are governance, higher productivity, profitable commercialisation, and increased competitiveness (MoALD 2015). Similarly, Myanmar’s ADSIP (2018) lists governance, productivity, market linkages, and competitiveness as its core strategic pillars. In addition, agriculture is presented as an economic activity, prioritising productivity and profitability to the exclusion of its cultural and social heritage values in agrarian societies. In this reduced understanding of agriculture, policies consider gender equality as a tool for economic development, constituting women as economic subjects and highlighting their productive functions in development. For example, Nepal’s ADS (p. 7) describes the objectives of engaging women as achieving “…economies of scale in technology dissemination, marketing, finance, and logistics” and to “improve the productivity and income.” Myanmar’s ADSIP also states that “socio-economic and farming systems research would particularly consider gender roles to optimise suitable technology packages” (p. 53). It calls for special attention to be given to women smallholders as they can contribute to increased economic efficiency and growth, as well as poverty reduction, thereby reducing inequality. USAID’s Integrated Country Strategy for Myanmar prioritises gender equality for “reaching a country’s full economic potential (p. 7)” and ensuring “equal access to credit and economic opportunity (p. 7).” These narratives describe the instrumental use of gender equality and further presume that increased individual competency automatically leads to economic empowerment, regardless of institutional and structural circumstances. According to this logic, women's economic empowerment is prioritised above other fundamental goals, such as achieving well-being and women’s rights.

Unwilling and less productive women

Another underlying assumption is that women are unwilling or unavailable to participate in development activities or are excluded from participation. However, it is unclear whether women are being encouraged or even mandated to participate because they are reluctant to do so, because they wish to participate but are unable, or if these proposals are a measure to prevent farmers’ groups and development projects from excluding women by setting a ratio of women to men participants. None of the documents reviewed explains why women are mandated or promoted to participate in development activities. The ‘unwilling women’ scenario problematises women’s choice not to engage in socioeconomic activities, assuming that women lack awareness or interest in such engagement. The ‘unavailable women’ and ‘excluded women’ scenarios problematise the environment that prevents women’s participation, presupposing that if women are given the opportunity to participate in development activities, they will do so. However, as Cornwall (2008) notes, there are a variety of factors why individuals do not partake in such activities. Rather than identifying and responding to these factors, policies expect women to increase their engagement by setting certain ratios and quotas. Furthermore, by failing to include men in the problem statement and setting men's productivity as a benchmark for that of women, the policies reviewed assumed that women have fewer skills, less knowledge, and lower productivity. Nepal’s ADS (2015) exemplifies this implicit logic by aiming to “double the agricultural productivity and incomes of small-scale food producers, in particular women… (p. 37).” Nepal´s National Cooperative Policy (2012) promises to provide vocational and skills training for women cooperative members. To empower them, the 2013 Irrigation Policy prioritises women’s skill development and access to resources and information. Myanmar’s ADSIP targets a 60 percent increase in productivity and income for women, compared to 50 percent for men. This implicitly naturalises men’s primary role as normative farmers in carrying out agricultural work, assuming their homogeneously and relatively higher productivity than that of women.

Representativeness and democratic decision-making

The third underlying assumption of the problem representation is that women who participate in development activities will be empowered and be able to actively shape agendas for the benefit of women at large, contribute to gender equality, and have the same bargaining and voting power as men or members of higher social status in general. This further assumes that equal opportunities to participate will lead to equal power distribution and developmental benefits. Thus, this assumption allows for a quantum jump from numerical participation to increased production, to decision-making power and empowerment, and finally to contribution to gender equality. Nepal’s National Cooperative Policy (2012) demonstrates this presumption by claiming that, “As women's participation in the cooperatives is over 40 per cent, it can easily be assumed that the cooperative has now been a contributor to empowerment (p. 156)”, thereby equating participation with empowerment without providing any evidence of empowerment beyond a headcount. Moreover, none of the policies analysed discussed democratic participation. Rather, they use rhetorical words such as ‘meaningful’ and ‘full’ participation. We found no discussion of measures to facilitate and ensure more democratic participation. This assumption disregards power imbalances that underlie decision-making processes and the ways in which power is exercised within farmer or water and land user groups and community development committees. Thus, policymakers emphasise numerical participation but fail to propose measures to facilitate fair, democratic participation. Moreover, establishing such quantifiable outputs risks misuse of policy measures. For instance, the Nepalese government provides incentives for cooperatives and rural enterprises to reduce or waive taxes if women comprise 30 percent of their membership. Nonetheless, there have been reports of privileged leaders exploiting such government subsidies by mobilising disadvantaged groups to join fake cooperatives without genuine participation or engagement (Hairong and Yiyuan 2013; Khadse 2016). Leder et al. (2017, p. 244) demonstrate that when numerical participation is executed without considering power dynamics, participation may create “double barriers” for women, providing an example of Dalit women who participated in a group in which the other half were higher-caste men, resulting in them being unable to express their views and influence decision-making.

Legitimising exclusion of non-participants

Problematising non-participants without proposing how to include them also implicitly legitimises exclusion from policy support of people who do not wish to, do not agree with, or cannot participate in the interventions offered. Policies in Myanmar and Nepal rely on farmers and user groups to channel public support, such as training, information, and inputs, leaving limited space for the provision of alternative services to non-participants. Focusing on the provision of extension services only through this participation mechanism attributes individual choice and responsibility for the consequences of not being reached by government programmes. Such ‘inclusion through participation’ may lead to a system of exclusion that discursively segregates rural populations into women and men, participants and non-participants, contributors and non-contributors, and the empowered and vulnerable. Behind these underlying assumptions, we also identified unproblematised gender inequality issues as follows.

Unproblematised, invisiblised and silenced issues

Intersectional marginalities

The policies examined in this study have completely silenced the intersectional marginality of rural society, posing women as a homogeneous group regardless of age, physical and mental ability, ethnicity, religion, social class, and position in the household, thus silencing the intersectional power inequalities and disadvantages. Our co-occurrence tables (in Atlas.ti) revealed that when ‘women’ are presented, they are usually paired with other minorities such as youth, elders, ethnic minorities, and the lowest caste. Often, they are awkwardly added to at the end of a problem or proposal statement, for example, “…including women, the aged, ethnic minorities, and other groups… (MoAD 2016, p. 63)”, “…excluded groups such as women, youth, and other disadvantaged groups… (MoALD 2015, p. 73)”, “…vulnerable groups, such as women, children, and the elderly (MoLFRD 2015, p. 39), and “…particularly women and youth (MoPF 2018, p. 45)”. Policies briefly propose the inclusion of women and other marginalised groups, but hardly any positive action is mentioned for any groups other than women, and they furthermore neglect intersectional discrimination that women may face. While Nepal has ostensibly and habitually added marginalised ethnic, religious and caste groups to policy texts, Myanmar's policies rarely discuss the disproportionate intersectional gender discrimination faced by ethnic and religious minorities. Furthermore, gender is described as a binary category consisting exclusively of ‘men and women,’ each treated as a homogeneous group, thus heteronormativising gender relations. Aside from USAID’s recognition of LGBT people as being traditionally marginalised in its Integrated Country Strategy for Myanmar (2018, p.2), there is hardly any discussion of a third gender or a gender-diverse population. Although a “third gender” is officially recognised by the government in Nepal, none of the ARD policies reviewed recognised them.

Social reproduction and its depletion effects

The ARD policies of both nations unproblematised and naturalised the responsibility of social reproduction being imposed upon women. Few policy documents acknowledge women’s contribution or the multiple burdens that result from their productive and reproductive labour, and those that do fail to discuss viable solutions for gender inequality. For example, Nepal´s Renewable Energy Subsidy Policy (2016) suggests using renewable energy to reduce the ‘drudgery of rural women’ so that they may increase “their productive time” and “improve the family’s health and educational status”. Similarly, Myanmar’s Climate Change Strategy (2019) recognises women’s disproportionate workload due to “responsibilities inside and outside the household (p. 103)” It also raises the issue that the agricultural sector fails to recognise women as formal farmers and fishers, although they perform the majority of agricultural work. However, the proposed solutions include providing childcare services for women so that they can attend meetings. Thus, both policies further perpetuate the assumption that women are responsible for childcare and family well-being. Such proposals describe the responsibility for social reproduction as exclusively women’s, while none of these policies attends to the fact that such overburdening can lead to the depletion of their own health and well-being (Rai et al. 2014).

Social provisions and legal protection

Legal enforcement of land and labour rights and social policies, which are a prerequisite for improving women’s rights and well-being, are rarely discussed. Despite the heavy concentration on women’s economic empowerment, the policies fail to provide viable solutions for inclusion of women in formal employment or legal enforcement to guarantee women equal wages or adequate payment for their agricultural labour. Rather, policies address their contribution to the informal economy as subsistent producers and micro-entrepreneurs. While Nepal’s ADS (2015) “strongly recommends for the equal wage for male and female labors for the same work” (p. 61), Myanmar ADSIP (2018) aims to reduce the gender wage gap (p. 11) without legal enforcement. In Myanmar, 43 percent of women are agricultural wage workers, with a gender pay gap of 29 percent (Lambrecht et al. 2021). Meanwhile, 80 percent of Nepalese women are employed in the agricultural sector, earning 33 percent less than men (UNESCO 2015). While both governments recognised gender inequality in land rights, only Nepal laid out its plan to achieve 50 percent women’s or joint land ownership by 2030. MoECAF (2016) states equal land tenure and management rights for women and men, without providing a clear strategy for redressing the current disparity. Given that most rural women rely on subsistent farming and agricultural labour, lack of clear legal and social protection for women may exacerbate existing inequalities among women, especially those in more marginalised social positions.

Men’s and intergenerational engagement

None of the policies reviewed discusses the need to engage men or intergenerational actors such as in-laws—who are key decision-makers in patrilineal households- in addressing gender equality. While women are encouraged to increase their labour input as producers and active agents fighting poverty, men’s role in sharing responsibility for social reproduction is nowhere to be found in gender equality narratives. Nepal’s Fifteenth Plan (2020) briefly mentions, “Active participation of men will be ensured for the achievement of substantive gender equality (p. 264)” without further elaborating on measures to assure such commitment. Among the donors’ documents, the DFID’s Operational Plan for Myanmar (2014) specifies engaging men and boys, yet its discussion is limited to preventing sexual violence in the context of political conflict. Numerous studies have reported that men’s proactive engagement in gender interventions and sensitising them to gender equality expedite the construction of an environment conducive to women’s political, economic, and social participation (Esplen et al. 2006). Feminist scholars stress the importance of engaging men and boys in balancing the inter-relational and intra-household power relations that constitute gender practices at the individual, household, and community levels. Furthermore, in patriarchal families, powerholders are usually husbands or in-laws who promote or limit women’s agency and participation. Thus, women’s socioeconomic opportunities depend on the family’s labour and emotional support (Leder et al. 2017). However, policies do not address the micro-level power relations that facilitate or restrict women’s participation. In the gendering process of ARD policy, men tend to be alienated from discussing gender equality and are constructed as bystanders, which may inhibit the realisation of gender inequality.

Institutionalising gender knowledge

The participation discourse is silent regarding the institutionalisation of gender knowledge. Lack of institutional gender knowledge and expertise within the government poses a great risk of producing and reproducing limited gender discourses and inappropriate proposals for gender equality, which may lead to unexpected negative consequences. While the Nepalese government is strongly committed to mainstreaming gender equality and social inclusion (GESI) in all ministries, neither country is fully committed to institutionalising gender knowledge within the government. Nepal’s ADS (2015) highlights the integration of the GESI strategy into its operational mechanism while paying little attention to fostering gender expertise. The only relevant proposal is to appoint at least two full-time professional staff to integrate GESI into ADS, and one gender staff at each district office—covering 1.7 to 6.1 million people depending on the district (NSO, 2021)—to design, implement and monitor agricultural programmes. It mentions fostering and recruiting more female junior field extension officers and notes that additional expertise required for the GESI integration will be outsourced to specialised UN agencies such as UN Women. This implies a lack of strategies to foster and maintain gender expertise in agriculture at the central level, apart from using such expertise to develop a GESI strategy. None of the Myanmar policies reviewed mentioned a plan to foster gender experts within the government system. Its ARD policies discuss some workshops for government staff, sparsely throwing in gender terms such as “gender-sensitive,” “gender equitable civil society representation”, and “gender-equitable and participatory planning.”

Discussion

This study aims to understand why social progress towards gender equality is slow by examining what aspects of gender inequality are problematised in the ARD policies of Myanmar and Nepal. Our findings show that women’s participation is represented as the predominant problem and proposal in the studied policies without evidence, constituting the central discourse of gender equality. By equating the lack of women's participation with gender inequality, women are placed at the centre of this problematisation and blamed for persistent social injustice and slow development outcomes. In this manner, the discourse of women’s participation turns gender inequality into a women’s agenda (Bacchi 2017), normalising gender as a heteronormative concept of women and men and homogenising women's rural life experiences, which demonstrates a simplistic and rigid understanding of gender. As a result, differences among women and intersecting marginalities or privileges became invisible. Such discourse can also be interpreted as a dividing practice (Bacchi, ibid), creating a binary opposition—such as those who participate versus those who don’t, and women versus other stakeholders (power holders) who are not responsible for gender equality such as men, privileged castes, major ethnic groups, and the governments—while generating a critical gap in policymaking and implementation. Through the process of gendering and othering, women have been regarded as the sole party responsible for gender equality and have been instrumentalised, suggesting that decades of feminist critiques on technical, instrumental participation (Cleaver 1999; Shrestha and Clement 2019, Paul 2019) have not yet been adequately considered by ARD policymakers.

Instrumental subjectification of rural women

The problematisation of women’s participation makes women the explicit target of ARD interventions, producing specific subjects (Bacchi and Goodwin 2016) in the form of participating or non-participating women, which constitutes gendered development subjectivity (Moor et al. 2017). Under this frame, women who participate are constituted as more productive, empowered, responsible, educated, and modern subjects. At the same time, they are seen as less skilled or knowledgeable, in need of fixing and empowering through intervention activities. They are regarded as less productive in agriculture than men and are treated as an under-utilised labour force whose participation could further increase productivity, failing to recognise their actual significant contributions to production and reproduction. The argument that women are less productive is unsubstantiated and derived from limited research methodologies, as productivity cannot be calculated individually in family farming systems where household members share agricultural tasks (Doss 2018). Therefore, policy narratives that promote women's increased participation in agriculture due to a perceived lack of agricultural skills and an assumed correlation with lower productivity may be subject to inherent biases. In this discourse, participating women are constructed as subjects who should and can produce as much as men, and are therefore expected to do more and have higher productivity or output than non-participants. Such subjectivities thus normalise the image of less-productive women and idealise women participating in development interventions.

Overall, Nepali women work 4.8 times more hours than men, including approximately three more hours of unpaid care and housework (Budlender and Moussie 2013). At the same time, 41 percent of women suffer from anaemia (Morrison et al. 2021) and show high rates of uterine prolapse due to heavy workloads in both productive and reproductive roles (Bodner-Adler et al. 2007). Myanmar shows a similar pattern, with women working 4–5 times more on reproductive tasks or twice as much as men in terms of average total hours worked (ADB et al. 2016). Nevertheless, policy proposals emphasise women's roles and responsibilities while remaining silent regarding women’s increased workload through participation (Lyon et al. 2017; Margolies et al. 2023) as well as their disproportionate burden of social reproduction and the potential physical, mental, and emotional depletion caused by the additional participation on top of their existing productive and social reproductive work (Rai et al., 2014). By failing to offer alternatives based on a more comprehensive gender equality discourse, including improved social infrastructure; legal enforcement for equal rights; policies to recognise, reduce, and redistribute unpaid care work (Ferrant et al. 2014): and compensation for women's multiple burdens and opportunity costs, the proposals individualise and isolate the problem of gender inequality and fail to adequately address it.

Furthermore, reliance on women’s participation assumes that all women have an equal voice, bargaining power and capacity to engage in democratic decision-making processes and to represent the diverse realities faced by different women. This also reflects, as Cleaver (2012) notes, the dominant assumption that participation itself is empowering for women. These assumptions expect immediate empowering effects from participation, ignoring that physical presence does not always translate into voice, bargaining power and equitable redistribution of development outcomes. As Cornwall argues, “participation is about power and control” (2008, p. 271) and participation without consideration of social power dynamics can exacerbate inequalities and disempower women (Cornwall ibid., Leder et al. 2017, Paul 2019). When policies neglect power dynamics in participation processes and fail to address such power imbalances, participating women are at further risk of marginalisation and disempowerment within the group as a result of their ostensible participation. For instance, when women are physically present at a meeting but face continuous public critique and are neglected in the decision-making process, this sense of powerlessness may be internalised, resulting in gendered subjectivity (Leder 2022). Shrestha and Clement (2019) exemplify how a 33 percent quota for women in a water user group became instrumentalised and tokenistic; women`s seats were mostly filled by higher caste women, who inadequately represent lower caster women's interests, and de-facto decision-making was carried out informally by a few men outside of the official process (Panta and Resurrección 2014). Not only gender but also ethnicity, caste, and religion greatly contribute to the intersections of marginalisation in Nepal and Myanmar. Therefore, even within women's groups, women from higher castes and majority ethnic, religious, or political groups tend to have more say and bargaining power, raising questions of representativeness in the absence of measures to facilitate a democratic decision-making process. This implies that gendered subjectivity created through participation can reinforce and perpetuate harmful gender norms that women have internalised through generations of multiple hierarchies. In this sense, as Paul (2019) argues, the discourse of ‘participation’ operates to (re)produce particular meanings of ‘women’.

On the other hand, non-participating women are labelled as non-compliant. They fall into this group because they are unable, reluctant, or even resistant to participate, and are blamed for their ignorance and unwillingness to participate in development interventions. They are subjectified as traditional, passive, disempowered, uncooperative, and irresponsible women because participation is associated with social responsibility (Cleaver 1999, p. 605). However, framing non-participants in this way, while failing to offer alternative ways of benefiting from government or development assistance, can also be interpreted as exclusion, which resonates with Cleaver’s (ibid) criticism of ‘the exclusionary nature of participation’. Although the participation approach aims to include all target women, it effectively limits the provision of public resources to those who participate under the project’s term and fails to include those who, for whatever reason, do not or cannot participate. This risks penalising women who exercise their power as autonomous individuals to disagree with and/or resist particular government intervention by restricting their access to productive resources, services, and information. The exclusion of these women is automatically legitimised by their choice not to participate in such interventions, placing them outside decision-making and public support.

Furthermore, mandating women’s participation in terms of numbers or ratios based on men’s participation is problematic. Such proposals are made without any attempt to understand the conditions leading to women's lower or non-participation and the means to address barriers that might exist. Women cannot or do not participate because of time constraints due to livelihood activities and the disproportionally allocated social reproductive responsibilities on them, as well as lack of physical and mental well-being, lack of social networks, restrictions on their movement due to gender norms, and many other reasons (Agarwal 2001; Leder 2022). However, the discourse is fixated on mandating women's participation without problematising the lack or absence of social infrastructure and policies that might facilitate women's active engagement in socio-economic activities or offering more comprehensive proposals to address gender inequality.

The participation discourse reduces women’s roles to being—or failing to be—agents of development depending on their level of participation. In doing so, the discourse denies any other role or subjectivity of women as possible or relevant and thus plays a role in shaping and reproducing gendered lives (Miller 2010). Through such discursive practices, women are (re)constituted as particular development agents under the gaze of the policy, internalising such values and policing themselves to perform the new development subject produced by the participation discourse. In this manner, gender interventions exacerbate gender inequalities and foster gendered development subjectivity, as indicated by Molyneux (2006) and Moor et al. (2017).

Alienating men and distancing the governments

With the predominant focus on women and their participation, only women are called upon as subjects responsible for gender inequality, while all other key stakeholders are alienated, such as men, people with higher social positions within patrilineal households (in-laws), members of the upper class or ethnic majority, and the government. Feminist scholars highlight the need for men’s engagement in gender equality at the household level, as women experience gender inequalities most profoundly and personally within the family (Chant and Brickell 2013, p. 88). Cornwall (2000, p. 19) argues that misrepresenting and failing to engage men in gender discourse leads to missing opportunities for transformative development. Others emphasise an understanding of relational and intersectional gender power relations rather than focusing on empowerment at the individual level (Leder 2022), which requires awareness and support from all stakeholders at the household, community and societal levels. By failing to see gender as a shifting and relational concept, and engage men and other agents of social change, policies continue to fail to (re)negotiate the ever-changing gendered power relations. The exclusion of these key actors effectively prevents them from being part of progressive social change by failing to address intra-household power dynamics and inequitable power distribution in society.

The limited problem definition and silenced alternative proposals in policy texts manifest the failure to institutionalise gender knowledge and expertise within the government institution. Our analysis resonates with the critique by Devkota et al. (2022) that gender mainstreaming has largely failed in Nepal, with responsibilities falling on local and regional staff, and that there is a need to strengthen institutional capacity within the Ministry of Agriculture and Livestock Development on both local and national levels. We argue that the lack of institutionalised gender knowledge and expertise has contributed to the reduced discourse focusing on women's participation to address gender injustice (and thus to a discursive enclosure or limited discourse). This risks limiting the discursive capacity of government—and in turn, its capacity for identifying, defining, and solving problems (Andrews 2019)—as well as the discursive strategies that forge alliances to mobilise resources for gender equality (Cornwall et al. 2007). This discursive enclosure is implicated in agricultural development interventions and influences public discourses and practices on gender (in)equality. Consequently, it produces “subjectification and lived effects” as Bacchi and Goodwin argue (2016). Our findings show that women’s individual capacities and knowledge are emphasised as tools for gender justice, while government capacity and institutional gender knowledge—crucial to achieving transformative social change—are not problematised.

This raises doubts about governments’ political will to achieve gender equality; rather, they may individualise and isolate social exclusion issues, delegating the responsibility of participation to women, community-based organisations, and field officers in order to compensate for the lack of institutionalised gender knowledge and expertise. As observed by Mukhopadhyay and Prügl (2019), ARD policies relied upon and utilised women’s participation as a political technology to engage women to consent to its interventions. Echoing a critique by Gupta and Sharma (2006), this can be interpreted as the government distancing itself from or even abandoning its responsibility to mediate gendered power in institutional and social spheres, as well as its social welfare duties.

Conclusion

This paper examines which gender inequalities are problematised in the ARD policies in Myanmar and Nepal, and how they are problematised, as well as the impact of this problem representation on discursive practices and rural women’s subjectivities. Our study shows that in these countries, a lower level of women’s participation in development activities is represented as the predominant problem in ARD policies. This problematisation assumes gender equality as a tool for economic development, women’s un(willingness) to participate in development activities, their lower productivity, democratic decision-making in the participation process, and the legitimacy of excluding non-participants. With such inherent assumptions, increasing women’s participation in development interventions is proposed as a panacea for gender inequality, placing women at the centre of the problem and effectively blaming them for persistent social injustice and slow development outcomes. Despite decades of critiques of technical and instrumental participation, women’s participation is still the central discourse of gender equality, echoing and returning to the 1980s Women-in-Development (WID) approach (Tavenner and Crane 2022), which emphasised women’s productive contribution to development (Razavi and Miller 1995).

In the present study, policies’ heavy focus on women's participation silences other factors contributing to gender inequality, such as intersectional marginalities, social reproduction and its depletion effects, social provisions and legal protections, men’s and intergenerational engagement in addressing gender inequality, and the government’s institutional knowledge of gender. By unproblematising and silencing these factors, gender injustice has been reduced to women’s lower participation in development interventions, failing to address the power relations, structures, and institutions. This is problematic not only because the reduced approach precludes a greater diversity of discourses that are fundamental to transformative social change, but also because it has produced gendered development subjectivities that consequently impact women’s lived experiences within the narrowly defined rationality of economic development.

Using Bacchi’s approach, we have teased out how the problem is defined and represented; what other discourses are unaccounted for; how these said or unsaid discourses exercise power to determine intervention tools; and the ways such approaches engage rural women’s identities, subjectivities, and positions. We argue that analysing policy as discourse is critical because it reveals the ways in which discursive practices shape policies, and how gender inequalities are understood, constructed, and narrated as discourses shift, and consequently constitute reality. The effectiveness of gender policies should therefore be assessed in terms of what gender analysis underpins the strategy, what discourse is produced from it, what is problematised as a result of the analysis and how it is problematised, what is proposed as a solution, what the discourse allows to be thought and said, how it constitutes and positions political subjects, and how it affects the real lives of subjects. For these reasons, we argue for more attention to analysing how problems are represented as discourses, beyond examining the success of a policy or a mainstreaming strategy as a tool for gender equality.

Regardless of differences in the gender integration progress between Myanmar and Nepal, the central gender discourse is surprisingly alike: women’s participation. We interpreted that such a reduced problem definition and representation delays social progress towards gender equality, as it prohibits discursive opportunities for diverse accounts of—and alternative proposals for addressing—gender equality. Nevertheless, the examples of Myanmar and Nepal cannot be used to generalise that the limited discourse embedded in national ARD policies slows progress toward gender equality in other countries. Indeed, in order to examine the full potential of a country's gender equality strategy, there is a need for a more comprehensive analysis of the policies of other sectors which might support gender proposals included in ARD policies. However, our study suggests that the strategy of gender mainstreaming may be insufficient to achieve gender equality and that there is an urgent need to scrutinise the problems produced by gender analysis in policymaking as well as in the mainstreaming process.

We contribute to the literature on gender and development and to the study of gender policy by demonstrating how gender discourse engages in discursive practices, subjectifying and (re)positioning rural women, and we call for working towards a deeper understanding of problem representation by interrogating gender policy as discourse. In particular, more empirical evidence is needed on how such discursive practices influence the subjectification of policy subjects and their real lives in the process of implementing gender policies, in order to contribute to developing more accountable policy proposals for transformative social change.