Blog Post

Using Social Safety Nets to Improve Food Consumption

Social protection programs – specifically social safety nets – can meaningfully increase poor populations’ food consumption and asset holdings, according to a new study published in World Development .

Researchers from IFPRI and Cornell University* compiled a database of social protection programs and food security and asset outcomes, covering studies from Latin America and the Caribbean, Africa south of the Sahara, Middle East and North Africa, South Asia, and East Asia and the Pacific. The database includes studies from 1994-2016 and focuses on social safety net programs, including cash transfers, public works programs, and food transfers or vouchers. Using this database, the authors pooled estimates of the impact of social protection on household food security and asset formation using meta-analysis techniques.

The analysis of food security outcomes focused on caloric intake and the value of food consumed (food expenditure), as these were the most commonly reported measures across the examined studies. Both of these measures were then further broken down into grains, fruits and vegetables, animal source foods, and other foods. For the analysis of asset formation, the authors broke down physical asset measures into: livestock, non-farm productive assets (bicycles, mobile phones, etc.), farm productive assets (excluding livestock), land, and savings.

The studies included in the database covered a range of impact estimate time periods, from one year to ten years. Generating a single estimate for each study-program was a complex process; the authors began by using the outcomes reported for the whole study sample when possible; when this was not possible, they averaged the impact estimates across sub-groups to create one study sample. When studies presented multiple time periods for an outcome, the authors chose the longest time period. The authors pooled estimates to create a mean summary impact and a 95 percent confidence interval.

The results show that social protection programs, specifically social safety nets, have a significant positive impact on poor populations’ food consumption. Overall, caloric intake by program beneficiaries rose by 8 percent; this overall impact is significant at the 1 percent level. Across the regions included in the database, South Asia saw the greatest increase in caloric intake: 12 percent. Latin America and the Caribbean and Africa south of the Sahara saw increased caloric intakes of 7 percent and 6 percent, respectively.

Household food expenditures also experienced large impacts as a result of participation in social safety net programs. Overall, the analysis found a 13 percent increase in the value of food consumption across all of the study regions; again, this finding is significant at the 1 percent level. As with caloric intake, South Asia saw the largest increase, with food expenditures rising by 19 percent. Latin America and the Caribbean experienced a 13 percent increase, and Africa south of the Sahara followed closely with a 12 percent increase.

The authors used the food expenditure measure to examine consumption of different food groups. Overall expenditures on grains increased by 13 percent as a result of participation in social safety net programs; in Latin America and the Caribbean and Africa south of the Sahara, expenditures on grain rose by 14 percent and 17 percent, respectively.

Expenditures on fruits and vegetables followed a similar pattern, with a seven percent overall increase (significant at the five percent level) and an 11 percent increase in Africa south of the Sahara. For fruits and vegetables expenditures, the type of social safety net program appears very significant. For example, in Uganda, participation in a food transfer program led to a 23 percent decrease in expenditures on produce, while a cash transfer program in Peru led to a 30 percent increase.

For animal source food expenditures, participation in social safety net programs showed an overall 19 percent increase; this is significantly higher than the increases seen for grains or fruits and vegetables. The largest increase was seen in Africa south of the Sahara, where expenditures on animal source foods rose by 32 percent.

In terms of the impact of social safety net programs on asset formation, the study finds that results for livestock ownership are somewhat varied. Of 15 studies, five saw a small positive impact (between one and six percent), one saw no impact, and one saw a negative impact on livestock ownership. However, more than 40 percent of the studies saw a large positive impact on livestock ownership as a result of social safety net programs; overall, the study found that participation in these programs increased the likelihood of livestock ownership by 14 percent.

The results for non-farm productive assets and farm productive assets are similarly varied. However, overall, participation in social safety net programs increases non-farm asset ownership by seven percent and farm asset ownership by 35 percent.
Results also show that social safety net programs can increase both the likelihood of poor households holding savings and the value of those savings. The likelihood of holding any type of savings rose by 49 percent as a result of participation in a social safety net program, and the value of savings rose by 61 percent.

The analysis found no impact on land ownership as a result of social safety net programs, but the authors caution that this could be due to the small number of studies that look at this relationship.

Overall, the analysis suggests that social safety net programs have an important role to play in reducing global hunger. The findings regarding asset ownership in particular point to the fact that these programs can help address the causes of poverty itself, leading to more long-term and sustainable poverty reduction.

*Melissa Hidrobo (IFPRI), John Hoddinott (Cornell University), Neha Kumar (IFPRI), and Meghan Olivier (IFPRI)

By: Sara Gustafson, IFPRI