It was 1975 when the first UN Women’s Council declared the “decade for women,” and 1995 when gender mainstreaming became a global policy initiative, so why hasn’t the international development sector gained more knowledge about and better results in addressing persistent gender inequalities? The answer lies in the mundane and bureaucratic practices within the development project life cycle, and the United States Agency for International Development (USAID) provides instructive examples.

Development projects – projects in developing countries that seek to increase economic or social well-being of the poor – typically fail for one or more of these reasons: poor design, poor implementation, or poor assessment (the evaluation fails to capture the changes brought about by the project). Thus, if US development policy were to begin projects with a genuine incorporation of gender social norms in project design, it could emerge as a leader in addressing global gender inequalities.

Consider that USAID’s agricultural development initiatives recognize that the majority of smallholder farmers are women.

The organization cites that strategic investment in food security is crucial, noting the destabilizing riots in 30 countries during the 2008 spike in world food prices. And USAID’s interest in women in agricultural development is shared by the World Bank, the Bill and Melinda Gates Foundation, and the International Fund for Agricultural Development (IFAD), among other powerful partners: together, they argue that agricultural productivity will increase when women receive development project benefits equal to those their husbands, fathers, and/or brothers have received.

Barriers to participation include women’s opportunity cost. They make essential contributions to childcare, housekeeping (from preparing food to washing clothes), and farming, and so their time is valuable enough to the household to prevent their ready participation.

Societal barriers in many country contexts also mean financial disadvantages for women including land inheritance laws, restricted mobility, and low access to finances. As a result, organizations’ excitement around engaging women in development projects can be quickly dampened by the empirical realities of delivering development within an overarching gendered division of labor.

During recent research for the International Center for Wheat and Maize Improvement’s  (CIMMYT) Gender and Social Inclusion Research team, I highlighted three examples from project evaluation reports that reveal how and why gender social norms are key components in agricultural development project planning. In one example, despite the heavy targeting and training of women to engage in one project’s beekeeping initiative, the final evaluation found that men disproportionately retained leadership in the Beekeeper’s Association. A second project sought to assist smallholder farmers, yet invested in export-oriented value chain enhancements, when the majority of women were engaged in domestic value chains. Lastly, one project’s final evaluation found its female trainees, tasked with teaching their neighbors, were discounted by their communities as possible sources of knowledge. Better attention to social norms around gender in the project design and implementation phases might have helped all three programs flourish.

This state of affairs has given rise to a strong interest in building a knowledge base around what works for gender social norm change in agricultural development.

Organizations are reviewing project portfolios, looking for answers. My recent review for CIMMYT of agricultural development projects in Ethiopia revealed that, of 13 agricultural development projects, only 2 demonstrated successful engagement with gender norms. These were both USAID-funded and run by CARE Ethiopia, an organization with a long-term, organization-wide commitment to mainstreaming gender and promoting women’s empowerment. The evaluations from the other 11 projects did not include women and/or gender relations in the theory of change (a design issue), did not adequately capture women and/or gender in the project monitoring system (an implementation issue with negative impacts on available evaluation data), did not include a gender-specific question in the evaluation contract (an evaluation issue that could mean implementation was successful, but not captured), and/or presented qualitative research anecdotally, not systematically (an evaluation issue that undermines rigor).

An International Fund for Agricultural Development (IFAD) review of global projects suggests this is a broader phenomenon. In 2017, IFAD assessed projects from 2011-2015 against their internal 6-point Gender Equality and Women’s Empowerment (GEWE) Policy, selecting projects for their inclusion of gender in the design from “partial gender mainstreaming” to “gender transformative.”

Out of 171 GEWE-rated projects, only 17 evaluations showed an impact on gender equity or equality, including those with a “partial gender mainstreaming” rating. In response, IFAD loosened their original inclusion criteria of counterfactual evidence to grow their sample to 57 evaluations.

We see recurring themes between the CIMMYT and IFAD reviews. As in my review of the Ethiopian agriculture programs, the IFAD evaluation team noted limitations primarily about the lack of data for learning about gender and development in ways that could inform better interventions. Although these projects were selected for their inclusion of gender in the design, design, implementation, and evaluation issues all contribute to the lack of meaningful data about gender in projects. For instance, IFAD noted difficulty in assessing the rigor of evidence and social change, given that some was anecdotal evidence while simple beneficiary output numbers are presented as ‘robust evidence’. It was difficult for IFAD to determine whether perhaps their internal gender rating system was not accurate, implementation strayed from the project design, and/or the evaluations failed to capture meaningful changes. Because evaluation is the ‘last’ of the three project failure points, we can assume that successful design and implementation would yield adequate data in the evaluation phase; that is, evaluation failures start earlier in the process.

So while we can applaud a gender-mainstreaming mandate in international development since at least 1995, recent reviews of agricultural development programs demonstrate that women and girls are still not captured in project design, addressed in implementation, nor captured in monitoring and evaluation systems and reports.

Most projects lack a gendered design, compounded by implementation activities which aim to quickly intervene in the social worlds of dynamic people and communities, and the corresponding evaluations have little data to learn from, treat gender as an ancillary matter, and/or are not required to assess the gender-related aspects of development projects. The introduction of ‘gender and development’ was meant as a response to ‘women in development,’ but largely reinterpreted as performance on sex-disaggregated targets. Focusing on gender social norms highlights the relationships between and among women and men and advocates for gendered design.

If the development sector broadly, and USAID more specifically, is serious about ensuring that women benefit from development projects, they should intentionally invest in gender social norms at the highest level of project design. By placing gender social norms within the design, projects would promote women’s engagement and assist in garnering the support of their communities and households. Gender social norms, then, become an omnipresent feature of project implementation, mirroring how gender is experienced by individuals in everyday life. USAID and other donors could drive gender equality by contracting organizations with demonstrated proficiency in gendered design, such as CARE Ethiopia. With gendered project designs in place, social norms should have an overt measurement, including qualitative research, to assess progress. The data produced from these activities will be inherently gendered, provide project managers insight into differential outcomes, and provide data for evaluation. The resulting evidence base will become a firm foundation for US development institutions and partners to better understand and enact ‘what works for gender social norm change.’