Savings revisited: a replication study of a savings intervention in Malawi
3ie Replication paper 18
Jesper Stage and Tharshini Thangavelu replicate the results of a randomized evaluation carried out by Lasse Brune and colleagues in 2016 in Malawi. The original study evaluated an experiment where tobacco farmers in Malawi were assisted in setting up savings accounts for part of their harvest revenue, which would be used to purchase inputs ahead of the next planting season. Most farmers did not actually make use of this offer. However, on average, those farmers who were offered help with setting up savings accounts tended to spend more on inputs in the next planting season, leading to higher incomes.
The replication authors conducted push-button and pure replications that replicated the original results, with only very minor discrepancies. They further examined the robustness of the results by conducting estimation analyses, with results that were also largely similar to those reported in the original study. In their theory of change analysis, the replication researchers focused on those farmers who actually accepted help in setting up bank accounts and then used them. These farmers subsequently had considerably higher input spending and higher incomes than their counterparts (i.e. those farmers who were not offered help with setting up accounts). Overall, the savings intervention appears to have been successful in improving livelihoods among the targeted farmers, and the results from the replication study support those from the original study.