Authors: Varun Goel*, UNC Chapel Hill
Topics: Medical and Health Geography
Keywords: spillover, health, interventions, water, disease, bangladesh
Session Type: Paper
Start / End Time: 8:40 AM / 9:55 AM
Room: Governors Square 16, Sheraton, Concourse Level
Presentation File: No File Uploaded
Public health interventions aimed at specific populations may impact not only the direct recipients but also other people living in close physical or social proximity to the direct recipients. These indirect effects, also referred to as spillover effects, are often controlled for at the study design phase of most health impact evaluation studies to enable unbiased measurement of the treatment effect. However, these spillovers may themselves be of substantive interest and critical to evaluating health impact; cost-effectiveness studies excluding spillovers may underestimate or overestimate health impact and lead to erroneous policy considerations. While there is significant literature on spillover (herd) effects in studies of vaccine efficacy, gaps remain in understanding the mechanism and measurement of spillover effects across other health interventions. In this paper, I evaluate deep tubewell interventions in rural Bangladesh installed as part of an arsenic mitigation program to provide safe drinking water. I use spatial difference-in-difference methods with propensity score matching to evaluate the direction and magnitude of spillover effects, and the various mechanisms through which those indirect effects may augment or undermine the overall impact of deep tubewells on provision of safe drinking water in rural Bangladesh. Using comparative models, I show that there may be both positive and negative spillover effects of deep tubewells, and that the magnitude and direction of those effects is spatially heterogenous, and modified by factors such as tubewell density, distance, and arrangement in relation to other water, sanitation and hygiene interventions.