India has the largest population exposed to inland flooding globally, with poor individuals constituting the highest-risk group in the country. Early warning systems for floods can play a critical role in reducing the risks floods pose to both lives and livelihoods, yet many questions remain. How can early warning systems be designed to reach poor citizens in under-resourced settings? Which, if any, non-informational barriers limit the impact of these systems in poor communities?
In the summer of 2019, in partnership with India’s Central Water Commission, the Google Flood Forecasting Initiative launched its location-based early warning system for riverine floods in the Ganges-Brahmaputra river basin. Drawing on technological advancements, this system generates highly accurate, localized flood predications which Android smartphone users with location services enabled receive as notifications in affected areas. Despite the potential, in practice flood alerts often remain inaccessible to those who need it the most, owing to a range of last-mile implementation challenges in low-income settings related to communication infrastructure and technology, state capacity, and other factors such as social trust and low literacy levels. Since the product’s launch, Inclusion Economics have collaborated with Google on ways to improve the access of these alerts among vulnerable households in Bihar, one of India’s poorest states.
We are currently enrolling a panel of around 6,000 households across our sample communities to order to extend our impact evaluation to examine the longer-run effectiveness of community-based early warning systems in inducing avoidance actions and improving health and socioeconomic outcomes, over three flood seasons. Results from this work hold the promise of providing novel, causal estimates of the value of early warning systems and provide insights on overcoming last-mile challenges to improve the lives of vulnerable populations disproportionately impacted by climate change.
Requisite Skills and Qualifications:
The Tobin RA will help with literature reviews, support ongoing surveys and, depending on skill set, write code to clean survey data, scrape data and conduct initial analysis. Skill and experience with econometrics software such as R or STATA to run econometric analysis, as well as Python skills, is valuable. Successful RAs will be detail oriented and able to work independently.