Understanding why people are Social Distancing (or not)

Policy interventions and mass public health messaging have succeeded in persuading large swaths of the US population to stay home, but not everyone is doing so. In our national survey using the Ipsos Knowledge Panel, we set out to ask a few simple questions: why are people social distancing, or not social distancing? What’s driving their actions? And how can we target them to change behavior?

hero icon

Key Findings



  • 87% of Americans understand how to properly social distance, but only 47% do it all of the time.

  • Two of the biggest influences on whether people social distance or not are 1) perceptions of the risk to them and their close friends and family (i.e. how much of an impact social distancing is thought to have on their health and friends’ and family’s health) and 2) community norms (how much they think others in their community are going to social distance). These factors outweigh demographics or political party alone.

  • We can divide the population into four types of ‘social distancers.’ For example, the segment that is social distancing at the lowest rate – only 26% consistency - tends to see low risk to themselves or their immediate circle from not social distancing, and does not perceive it to be the community norm.

  • It matters where people get their information and trusted information sources are underutilized, especially health care providers.

  • 14.6% of respondents thought that Coronavirus could have a severe impact on their physical health, while 68.1% thought that it could have a severe impact on the American economy.



The survey was fielded using Ipsos’s Knowledge Panel to poll 2,500 Americans on March 27 – 31, 2020. It was designed using the CUBES behavioral framework to understand the beliefs, risk perceptions, social norms, awareness, and personal experiences around Coronavirus and social distancing.

Knowledge of Social Distancing Isn’t Driving Action, Respondents Face Many Barriers


While a large majority understand how to properly socially distance, we see drop offs between intention (i.e. desire to socially distance), and action.

Some of these drop offs are driven by specific interpretations of what it means to social distance: certain respondents show extra-cautious patterns of response with 10.9% saying it isn’t okay to go grocery shopping. In contrast, socializing was the most common reason cited for making exceptions: 20.2% said it was okay to spend time with friends outdoors and 10.6% said it was okay to socialize with non-symptomatic friends indoors.

Social distancing poses significant financial barriers to many respondents: at the time of the survey only 38% said they could work from home and only 50% were confident they could pay their bills for the next two months. Some respondents also have less faith in their own ability to stop the spread of the virus through social distancing.

Distaning image 1

Driven by Perceptions of Risk and Community Norms, There Are 4 Distinct Groups of Social Distancers


To get people to start and/or adhere to social distancing, shifting beliefs around risk and community norms is key. Our machine learning models show that considering an individual’s perceptions more strongly predicts social distancing behaviors than just demographics alone. For example, people who believe it is the norm in the community to social distance are 2.31 times more likely to follow the practice than those who believe it is not the norm. Active cases, death rates, and state-level policies such as stay-at-home orders don’t seem to be significantly impacting a person’s risk perception around social distancing.

Using machine learning, we identified four different segments of the American public with different social distancing beliefs and behaviors. These segments differ both in the drivers behind their social distancing actions and their actual behaviors. The segment (Group 4 below) with the least adherence to consistent social distancing, with only 26% participation, has low perception of social distancing norms and low risk perception -- and is more likely to be male and Republican.

Distaning image 2

Demographics, Political Ideology, and Structural Inequalities Also Play a Role


51% of women are always social distancing compared to 42% of men. Generationally, Baby Boomers are social distancing at the highest rates, and the youngest generation, Generation Z, is practicing social distancing much more inconsistently or not at all.


Distaning image 3

While Republicans and Democrats have similar levels of knowledge around social distancing, we see Republicans have both lower intention to act on social distancing behaviors.


Distaning image 4


Current COVID case and fatality data (although early and incomplete) shows that Black Americans are disproportionately impacted by the pandemic. We find Black respondents are less likely to feel that their actions can make a difference in stopping the spread of the virus. 31% of Black respondents were never or only sometimes social distancing, followed by 29% of Hispanic respondents, and 27% of white respondents.

We Are Under Leveraging Some of the Most Trusted Sources for Sharing Information


Currently, healthcare providers are the most trusted sources of information, but some of the least heard. Respondents said social media sources and President Trump, in contrast, are the least trusted sources of information around COVID-19, and they're also among their least used sources of information on the crisis.

Distaning image 5

Our methodology


Poll information

The survey was fielded using Ipsos’s Knowledge Panel to poll 2,500 Americans on March 27 – 31, 2020. It was designed using the CUBES behavioral framework to understand the beliefs, risk perceptions, social norms, awareness, and personal experiences around Coronavirus and social distancing. This poll is based on a nationally representative probability sample of 2,500 general population adults age 18 or older. The study was conducted in English and Spanish. The data were weighted to adjust for gender by age, race, education, Census region, metropolitan status, and household income. The margin of sampling error is plus or minus 2.4 percentage points at the 95% confidence level, for results based on the entire sample of adults. Poll data can be found here

Predictive model information

We performed a series of multivariate logistic regressions to identify characteristics that were associated with a higher likelihood of always following three social distancing behaviors: avoiding groups of people, staying home as much as possible, with exception of solo outdoor activities, and maintaining physical distance from people, with the goal of staying at least 6 feet from others. Variables were tested for multicollinearity and no variables used in any regression were found to have a Variance Inflation Factor (VIF) greater than 2.5.

Segmentation methodology

A K-means clustering algorithm was used to identify clusters of individuals that differed on the following six variables: perception of community social norms regarding social distancing, risk perception around social distancing, perceived difficulty of social distancing, degree of information seeking about coronavirus, level of worry about coronavirus, and perceived self efficacy in stopping the spread of coronavirus. These variables were selected for segmentation based on their relations to social distancing behavior observed in the predictive models as well as possible actionability.

About us


Achieving health and social impact requires bold solutions. But all too often, well-meaning programs fail because they make assumptions without a full and deep understanding of why these problems exist and persist.

At Surgo Foundation, we bring together all the tools available from behavioral science, data science, and artificial intelligence to unlock solutions that will work. Surgo Foundation is working in areas such as coronavirus, maternal health, tuberculosis, and HIV in the US and around the globe.

For more information, please contact the Covid-19 team: covid19@surgofoundation.org


Copyright © 2020 Surgo Foundation


Design: Federica Fragapane. Development: Paolo Corti

Surgo foundation logo