Communicating Covid-19 messages at scale via social media

Researchers applied insights from earlier Covid-19 messaging research to launch a large-scale Facebook messaging campaign and corresponding evaluation reaching more than 35 million people.
Tablet displaying video messages recorded by health care workers
Illustration: J-PAL

Since the start of the Covid-19 pandemic, public health officials have been urgently working to share potentially lifesaving information to citizens to prevent the further spread and surge of the virus. Understanding the effectiveness of these messaging campaigns at large-scale is critical for controlling both the ongoing Covid-19 crisis, as well as future pandemics. With this in mind, a team of interdisciplinary researchers leveraged findings from earlier Covid-19 messaging research to design and launch a large-scale video messaging campaign and corresponding evaluation, reaching more than 35 million Facebook users across 13 states in the lead-up to the 2020 Thanksgiving and Christmas holidays. The campaign not only led to a reduction in holiday travel, but also resulted in a subsequent decline in Covid-19 infections.

The Problem

Throughout the course of the Covid-19 pandemic, there has been a critical need to rapidly and effectively disseminate information at large scale in order to promote preventative health behaviors and curb the further spread of the virus.

In November 2020, Covid-19 cases were rapidly increasing in the United States. In response, the Centers for Disease Control and Prevention (CDC) issued public guidance encouraging people to stay home rather than travel for the upcoming holidays (Thanksgiving and Christmas) in order to prevent the spread and potential surge of the virus.

Prior research has suggested that nurses and physicians tend to be highly trusted within the United States and can effectively spread public health messages, including over social media. However, at the time that the CDC issued stay at home guidance, there was no evidence to support whether these types of communications campaigns were effective at scale. Understanding the effectiveness of such large-scale public health campaigns may be critical for controlling both the ongoing Covid-19 crisis, as well as future pandemics.

The Research

Two randomized evaluations conducted at the start of the Covid-19 pandemic in the United States found that video messages delivered by a diverse group of doctors can shape individuals’ knowledge and behaviors.

Beginning in May 2020, an interdisciplinary team of economic and public health researchers—including J-PAL Affiliates Marcella Alsan (Harvard), Abhijit Banerjee (MIT), Emily Breza (Harvard), Arun Chandrasekhar (Stanford), Esther Duflo (MIT), and Benjamin A. Olken (MIT)—partnered with the Center for Diversity and Inclusion at Massachusetts General Hospital (MGH) to evaluate the impact of video messages communicating Covid-19 information on knowledge and information-seeking behavior of Black and Latinx individuals.1

In total, 42 physicians of varying age, gender, race, and ethnicity recorded video messages that communicated a range of Covid-19 information. Participants saw a sequence of three videos, each showing a different physician of the same race or ethnicity. The videos covered, respectively, (1) major symptoms and asymptomatic transmission, (2) advice about social distancing and prevention, and (3) information about mask wearing.

Participants who viewed health messages had fewer knowledge gaps related to Covid-19 symptoms, preventive behaviors, and transmission. For Black participants only, viewing a message from a physician of the same race also increased their information-seeking behavior, though this same effect was not observed for other groups.

In a second study, from August to September 2020, researchers evaluated the effects of video messages tailored to Black communities on recipients’ Covid-19 knowledge and preventive health behaviors.2 Participants were randomly assigned to receive either a series of three Covid-19 videos recorded by physicians of varied age, gender, and race on symptoms, case numbers, and CDC social distancing guidelines or a series of three placebo videos that discussed generic health topics. In both treatment and comparison groups, participants were also randomly assigned to see one of two American Medical Association statements on either structural racism or drug price transparency.

Consistent with findings from the May 2020 study, receiving Covid-19 health messages from a diverse set of physicians increased Covid-19 knowledge and shifted information-seeking and self-protective behaviors, regardless of the viewers’ race, socioeconomic background, or political beliefs. However, messages that highlighted structural racism in health care or the racial disparity in mortality due to Covid-19 did not have an additional impact on participants’ knowledge or behavior.

Findings from these two studies suggest that physician messaging campaigns can be effective in conveying Covid-19 information to people of diverse backgrounds that can lead to changes in information-seeking and preventive health behaviors. The relatively low cost of these interventions also suggested a promising avenue for encouraging preventive behaviors at scale.

From Research to Action

Researchers applied insights from earlier Covid-19 messaging studies to launch a large-scale Facebook messaging campaign and corresponding evaluation reaching more than 35 million people.

Based on prior findings that physician-recorded video messages could effectively shift recipients’ knowledge and behaviors (regardless of race, education, or political leanings), an expanded interdisciplinary research team launched a large-scale intervention and randomized evaluation to test whether short messages distributed as part of a social media ad campaign could impact both preventive health behaviors and Covid-19 infection rates at the population level.3

Across 13 states, researchers randomly assigned counties to receive a Covid-19 Facebook ad campaign at varying intensity levels. In total, more than 35 million Facebook users received a 20-second video message recorded by a nurse or physician encouraging them to stay at home rather than travel in the lead-up to Thanksgiving and Christmas in order to prevent the spread of the virus. Data from the research team’s marketing partner, AdGlow, suggests that overall engagement with the video ads was high, with over 12 percent of targeted users watching at least three seconds of a video.4

In high-intensity counties (e.g. counties where 75 percent of zip codes received the video ad campaign), the average distance traveled decreased by nearly one percentage point in the three days leading up to each holiday, though not during other periods, suggesting the campaign was effective in deterring holiday-specific travel. Moreover, the campaign led to a subsequent decrease in Covid-19 infections at the population level—cases declined in treated relative to comparison zip codes by 3.5 percent in the two-week period starting five days after holiday travel. Researchers did not observe differences in impacts between Republican and Democratic counties or between urban versus rural counties, which suggests that the campaign did not lead to negative or unintended consequences in spite of a polarized political environment (a potential concern prior to running the ad campaign)

Combined, findings from these three US-based Covid-19 messaging campaigns suggest that healthcare professionals can be effective in sharing lifesaving messages with a diverse range of recipients and these messages can be effective when communicated at scale via social media campaigns. This work has important implications for preventing the spread of future pandemics as well as direct applications for disseminating additional Covid-19 related health information. Additional research is now ongoing to assess how similar messaging campaigns delivered via social media may be able to encourage Covid-19 vaccination in the United States.

References

Alsan, Marcella, Fatima Cody Standford, Abhijit Banerjee, Emily Breza, Arun G. Chandrasekhar, Sarah Eichmeyer, Paul Goldsmith-Pinkham, Lucy Ogbu-Nwobodo, Benjamin A. Olken, Carlos Torres, Anirudh Sankar, Pierre-Luc Vautrey, and Estehr Duflo. 2021. “Comparison of Knowledge and Information-Seeking Behavior After COVID-19 Public Health Messages and Messages Tailored for Black and Latinx Communities: A Randomized Controlled Trial.” Ann Intern Med. 174, no. 4 (April): 484-492. https://doi.org/10.7326/M20-6141.

Breza, Emily, Fatima Cody Stanford, Marcella Alsan, Burak Alsan, Abhijit Banerjee, Arun G. Chandrasekhar, Sarah Eichmeyer, Traci Glushko, Paul Goldsmith-Pinkham, Kelly Holland , Emily Hoppe, Mohit Karnani, Sarah Liegl, Tristan Loisel, Lucy Ogbu-Nwobodo, Benjamin A. Olken, Carlos Torres, Pierre-Luc Vautrey, Erica T. Warner , Susan Wootton, and Esther Duflo. 2021. "Effects of a Large-Scale Social Media Advertising Campaign on Holiday Travel and COVID-19 Infections: A Cluster Randomized Controlled Trial." Nature Medicine 27 (August): 1622-1628. https://doi.org/10.1038/s41591-021-01487-3.

Torres, Carlos, Lucy Ogbu-Nwobodo, Marcella Alsan, Fatima Cody Stanford, Abhijit Banerjee, Emily Breza, Arun G. Chandrasekhar, Sarah Eichmeyer, Mohit Karnani, Tristan Loisel, Paul Goldsmith-Pinkham, Benjamin A. Olken, Pierre-Luc Vautrey, Erica Warner, and Esther Duflo. 2021. “Effect of Physician-delivered COVID-119 Public Health messages and Messages Acknowledging Racial Inequity on Black and White Adults’ Knowledge, Beliefs, and Practices Related to COVID-19: A Randomized Clinical Trial.” JAMA Network Open 4, no. 7 (July). https://doi.org/10.1001/jamanetworkopen.2021.17115.

1.
The full research team included: Marcella Alsan, Abhijit Banerjee, Emily Breza, Arun G. Chandrasekhar, Fatima Cody Stanford, Esther Duflo, Sarah Eichmeyer, Paul Goldsmith-Pinkham, Lucy Ogbu-Nwobodo, Benjamin A. Olken, Anirudh Sankar, Carlos Torres, and Pierre-Luc Vautrey.
2.
For this second study, the full research team included: Marcella Alsan, Abhijit Banerjee, Emily Breza, Arun G. Chandrasekhar, Fatima Cody Stanford, Esther Duflo, Sarah Eichmeyer, Paul Goldsmith-Pinkham, Mohit Karnani, Tristan Loisel, Lucy Ogbu-Nwobodo, Benjamin A. Olken, Carlos Torres, Pierre-Luc Vautrey, and Erica Warner.
3.
The research team for the scaled up evaluation included: Burak Alsan, Marcella Alsan, Abhijit Banerjee, Emily Breza, Arun G. Chandrasekhar, Fatima Cody Stanford, Esther Duflo, Sarah Eichmeyer, Traci Glushko, Paul Goldsmith-Pinkham, Kelly Holland, Emily Hoppe, Mohit Karnani, Sarah Leigel, Tristan Loisel, Lucy Ogbu-Nwobodo, Benjamin A. Olken, Carlos Torres, Pierre-Luc Vautrey, Erica Warner, and Susan Wootton.
4.
This figure is well above industry standards for Facebook ads (1-2 percent average engagement) and Facebook video posts (6 percent average engagement).