The political consequences of resource scarcity: Targeted spending in a water-stressed democracy. A replication study of Mahadevan and Shenoy (Journal of Public Economics, 2023)

Mahadevan and Shenoy (2023) assesses the use of state influence on funding for welfare programs in West Bengal for political targeting during periods of economic distress. Using a multidimensional regression discontinuity design, the authors find a misallocation of funding for a make-work program to support incumbent majority parties in electoral jurisdictions facing water-stress. This clientelism increased voting shares from farming communities using vote-buying strategies during a period of high unemployment (the dry season). We successfully computationally reproduce their results, and note some potential revisions to the replication packet to improve future replication. Further, we test the robustness replicability of the results through re-analyses modifying the definition of labor allocation, the definition of water-stress, as well as sub-analysis by voter turnout and voter population. We find that extreme water-stress jurisdictions are heavy recipients of reallocated labor, and find that labor is reallocated from part-time to full-time employment through the welfare program. Electoral victories from the ruling party successful implement vote-buying strategies in water-stressed electorates with high voter-turnout and large constituencies. This replication provides support for the internal validity of Mahadevan and Shenoy (2023)’s results and sheds a deeper light into the reallocation of welfare programs during periods of economic disaster.