Shocks, migration, and food security in the age of COVID-19: reflections on using SLRC panel survey data

In 2019, Mercy Corps teamed up with a group of graduate students from Georgetown University’s McCourt School of Public Policy to explore the conditions under which households decide to migrate internally or externally, the effect of migration on food security, and the impact of international aid on migration outcomes and wellbeing. By the time their report was finalised in May 2020, COVID-19 had dramatically altered the context in which their research is interpreted.


Migration has long been thought of as a transformative force for improving individual and household welfare and promoting broader economic development. Yet increasingly restrictive migration policies and the detrimental impacts of COVID-19 on mobility have elevated the need to better understand the relationship between shocks, migration and welfare, and the implications of curbing mobility on households that rely on migration as a coping strategy.

While the literature indicates that migration can have significant development impacts (for example, through remittances), and can serve as a coping strategy for communities affected by shocks, there is often only limited understanding and attention paid to migration in local development programmes and policies. Therefore, the results of our study – a micro-level analysis of Wave 1 and 2 survey data from the Democratic Republic of Congo (DRC), Pakistan, and Sri Lanka – has important implications for the work that Mercy Corps and others are doing in contexts similar to those surveyed by the Secure Livelihoods Research Consortium (SLRC).

Our study found that households with a migrant differ in significant ways from households without a migrant, and that returns to migration were highly dependent upon the context. This reflects some of the lessons gleaned from SLRC research: specifically, that migration is not always the ‘game changer’ that it promises to be. Through our of experience using the SLRC data, we have generated several key lessons and reflections which may be of use to other practitioners working in similar fields.

Reflections on using the SLRC survey data

The panel structure of the dataset enabled fixed effects analyses

The first two waves of data collected by SLRC provided us with a six-year window of insights into migration and its effects in the DRC, Pakistan and Sri Lanka. Because the same households were interviewed in each wave – and the surveyors accounted for attrition in determining the original sample size – we could control for time-invariant characteristics in our analyses to better isolate the specific effects of different shocks or household characteristics on migration. For example, when examining the DRC, the fixed effects analysis revealed that crop/livestock disease had the strongest effect on migration in comparison to other potentially predictive variables such as exposure to conflict, the quality of local governance, or additional household characteristics which were dropped from the model.

Comprehensive modules allowed for thorough understanding of migrant and non-migrant households

A critical component of the analysis was differentiating characteristics of households that had at least one member who migrated either within or outside of the country for livelihood purposes from households without any migrants. We conducted extensive analyses of the variables in each module, their correlation with the migration-related variables in the dataset, and any statistically significant differences in those characteristics among migrant and non-migrant households. This allowed us to identify the most significant factors influencing the decision to migrate and the results of migration.

The SLRC survey was not designed to measure the causes or effects of migration, which limited the scope of the analysis and results

While the survey does include questions about migration, displacement, and remittances, it remains mostly focused on livelihoods, and thus lacks key information such as where people migrated to/from, how long they stayed, and whether they returned to their original household. This level of detail is critical in order to fully understand the drivers of and returns to migration. For example, the key migration question in the survey – “Did any of these activities involve migration inside/outside your country in the past three years?” – refers only to migration for livelihood-related activities, which limits the implications of the analysis. For the returns to migration, the remittances variable may be a more inclusive and accurate measure of the extent of households who are affected by migration (given that it may also include those households benefiting from remittances from individuals who migrated for reasons that were not livelihoods-related, or perhaps who migrated longer than three years ago), but it still leaves a significant proportion of the analysis to conjecture.


Our research findings raise important considerations for development and humanitarian programmes , as well as broader policies. In the context of the ongoing locust outbreak in East Africa, for example, understanding the likely effect of the shock on migration should be carefully factored into preparedness and response efforts. Further, while it is clear from the analysis that migration can have positive impacts on sending families’ wellbeing, this is not always the case. The international donor community should increasingly focus on fostering the development gains associated with more informed and productive migration, in addition to efforts that curb irregular migration.

Because migration can sometimes be the best option for many households faced with hardship, the limitations imposed by the COVID-19 pandemic pose a significant threat to remittances, food security, and long-term development. The closure of international borders and restrictions on internal movement will likely have profound income effects on the poor, particularly migrant workers and their families, thus raising concerns about food insecurity. The fallout of this is already apparent in key countries that Mercy Corps works in, including Nepal (one of the SLRC focus countries), where migrant workers have been unable to return to overseas jobs, minimising household income while also increasing household consumption needs. Exploring ways to ease these restrictions or provide supplementary safety nets to migrant-sending households must be prioritised and coordinated by practitioners and policy-makers, to mitigate the impacts of COVID-19 on the most vulnerable households. While this study begins to draw out some of these important implications for programming and broader policy, migration and mobility must remain at the forefront of conversations about development and humanitarian relief at this critical time.