Disclaimer: The views expressed within this Think Piece are those of the author(s)and do not necessarily represent the views of UNICEF. This Think Piece has not been edited to official publication standards and UNICEF accepts no responsibility for errors.
Motivation and framing
Social protection is a leading strategy for addressing poverty, vulnerability to shocks, and under investment in human capital at-scale in low- and middle-income countries (LMICs). Its popularity among governments is due in part to evidence showing that social protection is effective at bolstering household consumption and food security, and increasing investments in productive assets and education, among others.
Poverty, vulnerability and well-being have inherent gender dimensions, thus it is not surprising that gender considerations have historically shaped certain design features of social protection in LMICs. Since the late 1990s, with the emergence of conditional cash transfers (CCTs) as social welfare policies in Latin America, women were typically targeted instrumentally as a way to achieve main programme goals related to improving human capital of children. More recently, the narrative has expanded to acknowledge the intrinsic value of increasing gender equality and facilitating women’s empowerment. In 2016, the Sustainable Development Goals (SDGs) identified social protection policies as vital to achieving targets under Goal 5 (gender equity and empowerment of women and girls), in addition to calling for minimum social protection coverage, by sex and age, as part of Goal 1 (end poverty and inequality). In 2018, the Social Protection Inter-Agency Cooperation Board of the International Labour Organization (ILO) formed its first ever working group on gender in preparation for the 63rd Commission on the Status of Women (CSW), with a priority theme that included social protection systems. Through preparation for CSW, background papers have addressed the importance of social protection to address gender dimensions of wellbeing. Thus, social protection, a policy instrument which has been traditionally promoted for its role in reducing poverty and vulnerability, is now also being recognised for its potential contributions to gender equality.
Given this recent focus on gender, it is worth reflecting on the breadth of rigorous evidence available to guide programming to achieve these goals. The majority of papers commissioned in anticipation of CSW to inform whether social protection is achieving results for women were conceptual, or summarized ‘promising’ case studies, rather than providing a comprehensive and critical assessment of impacts. However, recent reviews of cross-country evidence broadly agree on a number of conclusions, particularly that, although there is promise for some forms of social protection to facilitate gender-equality and empowerment of women, this is far from automatic. For many domains of women’s wellbeing, evidence is still limited and additional research is needed. Further, since social protection broadly has poverty objectives, there has been less commitment to exploring both programme design and innovative research to explore the potential for closing gender gaps.
While an impressive number of research efforts are ongoing—and will contribute to knowledge in the coming years—based on the current evidence landscape, we present our view of five priority research gaps (beginning with highest importance) to advance understanding of how to leverage social protection for gender equality and the wellbeing of women and girls.
Gap 1: Evaluations able to unpack the contribution of design components across social protection instruments to improvements in gender equality
A recurring theme across reviews is the inability to make concrete design recommendations based on existing research. Understanding design options and their differential effects on men and women (and boys and girls) is essential, as most large-scale programmes are still not designed with gender in mind; gender is neither the core business of social protection, nor is it typically a priority for large-scale implementation. Therefore, to move from ‘gender blind’ programming towards ‘gender transformative’ approaches, a nuanced understanding of design options is needed. Yet, current guidance on design and implementation is largely built around case studies of successful programmes and policies, without developing an understanding of how outcomes would differ under alternative programme designs. For example, in a recent quantitative review of social safety nets in Africa, we considered a number of design features of hypothesized importance for gender equality, including: 1) gender-based targeting, 2) conditionalities and behavioural features (e.g. labeling, nudges, messaging), 3) payments and transfer mechanisms, 4) integrated approaches, and 5) gender-aware operational features. Across design features, there were few studies able to assess differential impacts on women’s wellbeing outcomes.
This lack of evidence across key programme features means that we often have little persuasive power to change programme design (often at a cost) along the continuum towards gender-transformative approaches, as we cannot foresee the expected differential effects. For example, although integrated and ‘plus’ approaches are becoming popular for their potential positive and synergistic impacts, there is very little evidence of their comparative effectiveness. Until more is known, incentives are low for large-scale social protection programmes to take on additional components, often already served by vertical programming, at the expense of expansion of the core coverage to more beneficiaries. As another example, we know that cash transfers reduce intimate partner violence, a severe form of gender discrimination, yet we do not know which design features will maximize reductions, as evaluations have not been designed to test these features. To fill these gaps, we need both interventions which experiment with gender transformative components, as well as high quality evaluations with intent to capture differential design effects. The challenge to generating this evidence is primarily a methodological (design) challenge. Often implementers are unable or unwilling to experiment for the purpose of research in a way that allows evaluation of design features. In addition, it can be more costly to undertake studies with the complexity of research designs that are required to evaluate programme features.
Gap 2: Region-specific evidence and evidence synthesis
Past evidence reviews from LMICs, particularly on social safety nets and cash transfers, tend to be regionally skewed towards Latin America, where modern cash transfer programmes first gained prominence. However, as more evidence is generated, this is changing. Because programme design, poverty dynamics, and gender norms underlying potential for impact vary by region, regional-specific learning is needed. For example, Africa is likely to have both higher levels and depth of poverty, and less access to services and infrastructure as compared to Latin America. In addition, populations in Africa are more likely to live in areas which are drought or conflict prone, with the accompanying breakdown of routine services and governance structures. Due to these factors, social protection in Africa has traditionally had greater focus on resilience and shock-responsivity, and fewer punitive co-responsibilities related to service provision (which require intensive monitoring). From a gender perspective, Africa is unique in a number of important ways related to social norms and demographics. For example, there are higher percentages of HIV-affected households (including orphans and vulnerable children), higher fertility rates, and early marriage transitions, including into polygamous marriages in Africa. Due to the wide diversity of programme typologies and objectives, women have not necessarily been targeted instrumentally as recipients of social protection and safety nets in Africa, as they have in Latin America, however coverage by sex may vary by programme type.
These unique regional considerations translate into opportunities as well as potential restrictions on how social protection can be leveraged for gender equality, which global generalizations overlook. For example, Bastagli and colleagues review cash transfer impacts from LMICs and conclude there is strong evidence on women’s empowerment and decision-making increases specifically. However, we review similar indicators focusing on new evidence from Africa and find relatively weak evidence for changes in women’s empowerment and decision-making. South Asia, Southeast Asia and the Middle East have comparable idiosyncrasies with respect to gender norms, poverty, and programme design, which make them distinct. As social protection evolves to become more complex, so will our need for regional evidence and evidence synthesis reflecting gender realities and norms to make informed programme/policy recommendations.
At the national level, it is important to understand the social protection system within the country and how social protection is being and can be leveraged for greater impact on women’s wellbeing outcomes. Evaluating social protection systems, however, is challenging within a quantitative impact evaluation framework, and thus the majority of evaluation evidence available is focused on programmes (e.g. social safety nets, including cash transfers), rather than on effects of social protection system integration per se. Thus, while more comprehensive knowledge on functioning of national systems is relevant and necessary for understanding holistic impacts across populations and lifecycle stages, this type of knowledge needs to be built using institutional or political economy analysis or a case study approach.
Gap 3: Better indicators and intra-household gender-analysis (what gets measured gets done?)
A common theme among gender researchers and advocates across sectors is the need for better indicators, measurement and analysis by gender—research on social protection is no different. There is currently little understanding of coverage of social protection by sex in LMICs, and the majority of evaluations conduct minimal gender-disaggregated analysis. For example, many indicators are still collected at the household level (e.g. poverty, food security, productivity), despite well documented intra-household disparities. Further, analysis is often done comparing male- versus female-headed households, rather than contrasting men and women within the same household. Some of these gaps are simply a result of practicalities. Evaluations often need to prioritize key outcomes (domains) of interest in line with programme objectives to manage workload; questionnaires easily become overloaded and lengthy, overburdening participants. It is significantly more costly to interview two adults within a household, particularly if same sex enumerators are needed. Another challenge revolves around measurement, particularly of domains which are less traditionally measured in quantitative surveys, including ‘empowerment’, ‘autonomy’, ‘confidence’, and ‘gender norms’. For example, a new generation of impact evaluations specifically designed around gender issues are underway, which will expand the evidence base and provide some of the first rigorous evidence measuring women’s empowerment using holistic scales, like the Women’s Empowerment in Agriculture Index (WEAI). However, validation and local adaptation of rigorous measures to capture these complex concepts is needed to be able to measure power dynamics among diverse populations and settings. While qualitative work, as used to complement the WEAI impact evaluations, can help fill gaps in quantitative evaluations, mixed method work is still an exception, rather than the norm, and multiple types of evidence are needed to inform and influence diverse stakeholders. Further, low quality research, including poor measurement and biased analysis can result in misleading conclusions and sub-optimal policy decisions. Without proper measurement, indicators and analysis of important gender-related concepts within evaluations, and a budget to accommodate them, we will be unable fulfill the potential of evidence for change.
Gap 4: Cost-effectiveness, value for money and benchmarking
A cross-cutting theme critical to ensuring sustainability, expansion, programme design and implementation is value for money (including cost-efficiency and cost-effectiveness). While not generally considered a methodology critical for gender analysis, a better understanding of value for money ultimately tells us how well programme costs are converting to intended programme outcomes and impacts. To make comparisons across programme designs, and across outcomes, measuring cost-effectiveness will be essential in weighing trade-offs for more inclusive impacts. In addition, benchmarking to other economic programmes and policies is needed to fully understand the trade-offs between different implementation strategies a government may pursue to fulfill poverty objectives. Traditionally, programmes like social protection have faced methodological challenges in calculating comprehensive benefits in relation to costs due to the need to standardize and aggregate the latter across diverse outcome groups (e.g. health, education, food security). However, benchmarking techniques whereby different types of programmes (e.g. cash transfer vs. graduation programme) are contrasted within the same budget envelope offer one alternative to address this challenge, though issues with aggregation of benefits across a range of outcomes still exist. Without cost effectiveness measures, gender transformative programmes are likely always to appear more costly to implement and therefore less sustainable and more difficult to scale in the long run. In addition, value for money analysis may help ensure sustainability of gender-aware programmes within the national political and fiscal discourse. For example, the Mexican government’s recent decision to end the PROSPERA programme (formerly, Opportunidades and Progresa) was due in part to a loss of political support for continued spending on the programme. Although PROSPERA and its antecedents are among the most studied programmes in the world, better evidence of its cost-effectiveness may have strengthened the political case to keep this and similar programmes in Latin America. In sum, a better understanding of value for money would help identify the most efficient avenues for investment, which both are efficient at meeting gender objectives, as well as other programme priorities.
Gap 5: Future-looking policies and implementation features
The social protection field has advanced significantly over the past decade and is expected to be one of the key responses to poverty, vulnerability and exclusion during the SDG era. For research to stay relevant, it must always be forward-looking to anticipate future questions stakeholders may have. It is therefore useful to explicitly consider how future-looking policies and implementation features will interact with gender dynamics at the outset. These include, but are not limited to, the increased use of social protection in crisis settings, movement towards universal basic income (UBI), mobile technology and blockchain, micro-insurance, as well as effects on biological outcomes such as cognitive functioning, and macro-level processes (e.g. migration, urbanization, environmental and planetary health). For example, at least one small-scale NGO-led cash transfer programme run in Niger showed that transfers via mobile money conferred additional benefits to women—via time saving and ability to retain control of transfers—however we have little information if these benefits are consistently observed in larger-scale programming. Innovative programming and research based on technological advancements and in the most challenging acute disaster settings are increasingly being tested. Understanding gender implications should be part of these first effects to inform potential scale-up or expansion of programming.
Despite high-level commitments made by global stakeholders to advancing gender equity and equality through social protection, and the important role of this shared objective, there remains significant evidence gaps in understanding what this means in practice. We have proposed five priority evidence gaps to advance this understanding, related to: 1) programme design, 2) regional knowledge, 3) measurement and analysis, 4) cost effectiveness, and 5) future-looking issues. These priorities do not operate in isolation; each interacts with and should be considered alongside the others. We acknowledge that they are not exhaustive and are influenced by the type of research and viewpoints we bring to the intersection of social protection and gender. For example, understanding long-term, lifecycle, and intergenerational effects are certainly of importance, despite not being included as a top priority here. In many ways, setting out a research agenda for leveraging social protection for gender equality is similar to one which simply aims to make systems and programmes work better overall—no programme can fulfill objectives of poverty, inequality, and vulnerability while leaving half the population behind. We look forward to helping close these gaps.
Acknowledgements: This think piece draws heavily on concepts in a related chapter: Peterman, A., Kumar, N., Pereira, A. and Gilligan D. “Towards Gender Equality: A critical assessment of evidence on Social Safety Nets in Africa.” prepared as a chapter for the Annual Trends and Outlook Report (ATOR), “Gender Parity in Rural Africa: From Commitments to Outcomes” Quisumbing, A., Meinzen-Dick, R. and Njuki, J. (Eds). The chapter was undertaken with funding support from the CGIAR Research Program on Policies, Institutions, and Markets (PIM) led by the International Food Policy Research Institute (IFPRI).