Information provision over the phone saves lives

A randomized controlled trial to contain COVID-19 in rural Bangladesh at the pandemic’s onset

Abstract

Lack of information about COVID-19 and its spread may have been a significant driver of excess mortality at the onset of the pandemic. In April and May 2020, we implemented a randomized controlled trial with more than 3,000 households in 150 villages in rural Bangladesh in order to study the behavioral and health effects of information provision about COVID-19. Our one-to-one information campaign via phone stressed the importance of social distancing and hygiene measures, and illustrated the consequences of an exponential spread of COVID-19 within a village. One half of the treated households were additionally offered an unconditional cash transfer (corresponding to 2-3 days’ income) to compensate for possible costs of adherence to social distancing rules. We find that information provision improves knowledge about COVID-19 and induces significant behavioral changes. Most importantly, information provision yields considerably better health outcomes in treated villages, with spillovers from treated to untreated households there. Our intervention reduces the number of reported deaths by about 50% in treated villages, suggesting that all over the country an estimated 5,500 lives could have been saved in the first year of the pandemic through timely one-to-one information provision. Cash transfers have hardly any additional effect on these outcomes.