Esther Duflo on encouraging social distancing: evidence from several randomized controlled trials

Video and Slides

On Thursday, February 11, Esther Duflo joined Markus’ Academy for a lecture on encouraging social distancing : evidence from several randomized controlled trials. Duflo is a Professor of Poverty Alleviation and Development Economics at MIT and a 2019 recipient of the Nobel Prize in Economics.

Watch the full presentation below and download the slides here. You can also watch all Markus’ Academy webinars on the Markus’ Academy YouTube channel.

Executive Summary

• Before a vaccine was available, behavior changes were the only way to protect oneself and others against COVID. Communication to induce behavioral changes was paramount. There has been a bombardment of information and dis-information. The question therefore arises: does communicating on the value of protection make a difference? Furthermore, does the information need to be tailored to particular groups to be effective?
  
• The first study was conducted in India during their May nationwide lockdown. While there was much public communication at the time, behavioral compliance was far from universal. We wanted to see the effect of having a well-respected public figure, Prof. Abhijit Banerjee, co-winner of the Nobel Prize with Esther Dufo, send a message to the public.
  
• Study found that the intervention changed behavior of people in the treatment villages. The content of the message wasn’t as important as receiving the message from a trusted source. There was also a spillover effect on non-geosubscribers, which suggests that the information spread. It didn’t affect what the people knew but affected how they acted on what they knew.

• The next two studies were aimed at understanding who is trusted in the US and why. The US COVID crisis has been described as a “syndemic”, where the health pandemic is exacerbated by the systemic racism epidemic. Doctors have tried reaching out to people directly to encourage social distancing and mask wearing — previous studies documented that doctors are among the most trusted in public opinion. Duflo developed public messages about COVID with doctors at MGH, targeting a study group of low-income people-of-color. The second study was conducted in September during a time of COVID fatigue in the US. 

• Study found that overall, intervention affected people’s knowledge about COVID. However, there wasn’t an overall impact on demand for links for more knowledge. There was an increase in effect when the doctor was African American and the recipient was African American, though. Other interventions, such as Dr. Birx videos, social norm messages, and acknowledgement of inequality, did not affect behavior. Limitations include no hard measure of behavior and the changing environment in the US around COVID and politics.

• The second study focused on systemic-racism acknowledgement and impact of videos from diverse doctors. Found that in September, African Americans were twice as likely to wear a mask outside and followed similar COVID precautions. The intervention had a large impact on knowledge levels, increased demand for more information, increased willingness to pay for a mask, and increased the number of people who practiced all safety measures. Variations in the message framing, including the ethnicity of the doctor, did not have a significant effect. Acknowledging the disparate racial burden did increase willingness to donate to COVID efforts. Effect on knowledge was largest for white Republicans while the effect on behavior was the same regardless of ethnicity and political affiliation.

• Study found that COVID messaging on social media by doctors had a significant effect on holiday distance travelled but no effect on staying home. In conclusion, an information campaign with actionable information can still get people to change their behavior, despite ambient noise from social media and even on highly politicized topics.