Nutrition-sensitive social protection (NSSP) refers to a subset of social protection poverty alleviation programs that include targeting approaches, interventions, and conditions to address the underlying determinants of malnutrition (Ruel 2013). Design features that can enhance nutritional impacts of social protection programs include targeting of transfers to nutritionally vulnerable populations, providing fortified or nutritious foods, behavior change communication (BCC) interventions, establishing linkages to health services, and including women’s empowerment activities (Alderman 2016, Olney 2021, Manley 2022).
Investment in social protection programs is growing (Olney 2021, FAO 2023), and the social protection sector is a key stakeholder in many national multisector nutrition strategies. Since 2020, the combined impacts of the COVID-19 pandemic, climatic crises, and protracted conflicts have contributed to global rises in food prices and 122 million more people facing hunger (FAO 2023). Social protection has been a critical policy instrument for mitigating these shocks and will continue to be important for advancing the world towards Sustainable Development Goal 2 to eliminate hunger and end all forms of malnutrition.
As NSSP programs scale-up, governments need to monitor who in their target populations is being reached. However, at present, there is no standard approach for defining and collecting NSSP coverage indicators. DataDENT has been systematically working to address this measurement gap.
We first introduced our work in a March 2021 blog post that highlighted the complexities of measuring NSSP program coverage and shared plans to develop a measurement framework and NSSP coverage indicators for household surveys. DataDENT used a similar approach in 2018-2020 to develop indicators for IYCF counseling coverage. In this post, we provide an update of the NSSP coverage measurement work.
What have we learned about measurement of NSSP coverage?
In 2021-2022, DataDENT carried out literature/desk reviews, secondary analyses, and a pilot phone survey. After a brief review of NSSP evaluation literature and World Bank’s ASPIRE social protection database, we identified a subset of five ASPIRE social assistance program types that have potential to be nutrition-sensitive because they provide a food or cash transfer andcan be adapted to include a nutrition goal:
- Unconditional cash transfers (poverty-targeted cash transfers, emergency cash support)
- Conditional cash transfers
- Food & in-kind transfers (food stamps/rations/vouchers, nutrition feeding programs, in-kind emergency support)
- School-feeding
- Public works, workfare direct job creation (cash-for-work, food for work/training/assets)
Next, we reviewed sixteen social protection programs in LMIC with nutrition-sensitive potential including, among others, the Integrated Nutrition and Social Cash Transfer (IN-SCT) in Ethiopia, the Transfer Modality Research Initiative (TMRI) in Bangladesh, Filets Sociaux Program in Mali, and school feeding programs in Malawi. These programs vary widely in the type of nutrition-relevant benefits provided.
Using these findings, we identified seven elements of NSSP programs that may be measured in large-scale surveys (Figure 1). Two of the elements – provider type (e.g., government vs. NGO) and modality (e.g., mobile money) – were considered lower priority for inclusion in national surveys.
Figure 1: Proposed elements for consideration in NSSP coverage measurement
We also reviewed household survey tools from national surveys in over 70 LMICs to identify questions related to the above elements. Many of the surveys were supported by global programs including DHS, MICS, LSMS, or PMA. We identified over 396 unique questions across surveys that measured one or more elements, but no single questionnaire included all elements. Social protection modules in World Bank and MICS surveys included a combination of questions referring to program names and questions referring to benefits received. DHS surveys often did not have standard social protection modules but did ask about a range of nutrition interventions. Using these existing datasets, we tried to operationalize indicators for NSSP coverage based on the questions identified for 11 large-scale population-based surveys (Peru 2012 DHS, Argentina 2019 MICS, Nigeria 2018 LSS, Ethiopia 2018 SES, Malawi 2019 HIS, Nepal 2019 MICS, Bangladesh 2016 HIES, Bangladesh 2019 MICS, India NFHS 2016, India NSS 2021, Nepal NDHS 2016).
Alongside this secondary analysis, we included pilot questions about receipt of NSSP programs in a 2021 telephone survey conducted in two states in India during the COVID-19 pandemic, in order to examine respondent comprehension of these questions.
Detailed results of Phase 1 activities will be available in forthcoming publications. Key takeaways included:
- None of the national survey tools reviewed collected all core NSSP elements.
- In World Bank LSMS & MICS surveys, it is not feasible to link social protection program modules to questions on health and nutrition interventions among nutritionally vulnerable populations.
- DHS surveys include few questions related to social protection and typically only about specific programs.
- We could not consistently operationalize coverage indicators across existing national survey datasets.
- Respondents in India did not recognize the formal names of social protection programs suggesting that questions should ask about benefits received.
Based on these findings, we have moved forward into Phase 2 to develop new indicators and questions that can be integrated into household surveys to better estimate NSSP coverage.
Building global consensus on NSSP definition to measure coverage
One major barrier to DataDENT advancing the NSSP measurement agenda has been a lack of global consensus around an operational definition of what makes a social protection program nutrition sensitive. In July 2023, DataDENT joined practitioners and researchers from Nutrition International, WFP, FAO, IDS and IFPRI who are working towards consensus around key NSSP program features. The preliminary list includes:
- program design includes a clear nutrition goal/objective.
- includes a conditional or unconditional cash, food, or in-kind transfer or school feeding transfer that targets the most nutritionally vulnerable populations (e.g., pregnant, and lactating women, children <2y, women of reproductive age 15-49 years, adolescents 10-19 years, school-age children 5-9 years).
- links to one or more nutrition or health actions (e.g., inclusion of fortified or biofortified food, delivery of nutrition interventions such as BCC or iron-folic acid supplementation, linkage to antenatal care)
- applies a gender lens to strengthen women’s empowerment.
This collaborative group will continue to refine this list and plans to publish NSSP programming principles and guidance that will include implications for coverage monitoring.
Next steps: Developing new indicators and household survey questions to measure NSSP coverage
DataDENT’s Phase 2 work on NSSP coverage will involve further testing and refining of new indicators and household survey questions, gathering input from the broader NSSP community, developing guidance, and advocating for indicator and question uptake.
We aim to address several measurement challenges:
- formulating questions so respondents will accurately distinguish between involvement in a program that is NSSP “by design” from other social protection programs available in country (e.g., asking about program name and/or benefits received, asking about linked actions, specifying targeted populations)
- operationalizing the indicator denominator (e.g., using ASPIRE social protection indicator approach with households below a poverty threshold and/or using households with nutritionally vulnerable populations per program design)
- how to integrate new questions into established survey questionnaires (e.g., aligning recall periods)
In Figure 2 below, we illustrate where our proposed NSSP coverage indicator fits along a theoretical cascade of coverage-related indicators that starts with identifying the target population for NSSP to several indicators of effective coverage that account for quality of implementation. (Tanahashi et al. 1978, Amouzou 2019, Marsh 2020).
Figure 2: NSSP coverage indicator cascade
In 2024, DataDENT will conduct formative research in Ethiopia and India to support survey question development. Cognitive interviews will be conducted to refine questions that align with respondents’ understanding of social protection programs and connected benefits. Then we will pilot the survey questions at scale and identify cost drivers for adapting and fielding questions. DataDENT will engage a wide range of stakeholders in the indicator and question development and review process including technical experts, household survey programs, program implementers, and other data users. We will widely disseminate the resulting indicator guidance and advocate for uptake of the indicators by global and national household survey programs.
Follow us on LinkedIn or X @data_dent for updates.