The social distancing of America

The social distancing of America

Location data from millions of smartphones shows how Americans’ travel habits changed in the early weeks of the coronavirus pandemic.

Week of March 23-27

Change from February, median weekday travel distance

Source: The Kochava Collective. Based on data from more than 7 million distinct devices between February and March 2020. Note: Counties with values calculated from fewer than 500 devices have been omitted.

The first week of March was the last normal workweek for most parts of the country. (Note: California data is not available under the provisions of the California Consumer Privacy Act.)

By the second week, travel was down in many places, especially in Seattle, the site of the first major outbreak of the coronavirus in the United States.

By the week of March 16-20, many Americans were no longer commuting, especially in urban cores like New York, Minneapolis, Chicago, Detroit and Denver.

And by the end of March, most of the country was staying close to home.

Large sections of public life came to a standstill over the month of March as the novel coronavirus - which causes the potentially lethal respiratory illness COVID-19 - spread across the United States. State and local governments shut down schools, ordered restaurants and other businesses to close and issued stay-at-home orders for millions. But the public policy response hasn’t been uniform, and not everybody has stayed close to home.

Using data shared with Reuters by the Kochava Collective, an Idaho-based company that helps businesses evaluate the effectiveness of online ads, we’re able to see which counties in the United States saw people (and their phones) travel longer or shorter distances in the first few weeks of March than they did in February.

Location data from smartphones has been used by governments and public health officials to fight the outbreaks in South Korea and Singapore. By tracing the movement and contacts of people who have been infected, officials have been able to inform people who may have been exposed. The strategy faces regulatory hurdles in the United States and Europe, where privacy restrictions prevent sharing of personal data.

Anonymized smartphone data in the United States shows some interesting trends. People in larger cities and urban corridors were more likely to change their travel habits, especially in early March. By the end of the month, most U.S. residents were traveling dramatically less than they did in February, but social and demographic differences were strong predictors of how much that changed.

Change in median distance traveled, March 23-27

Circles are sized by county population

People in lower-income counties were more likely to continue traveling. For many, social distancing is a privilege, especially for blue-collar workers or those in the service sector who are less likely to be able to work from home.

Behavior in urban and rural counties also contrasted sharply. The most populated counties (largest circles) decreased travel the most dramatically.

Residents of counties that voted for Trump in 2016, which also tend to be less urban, were less likely to change their travel habits.

Counties with more young people were more likely to see a reduction in travel. They also tend to be more concentrated in urban areas.

People age 65 and older are among those at highest risk from COVID-19. But in many counties with higher concentrations of seniors, people were slower to change their travel habits.

Editing by Jonathan Oatis.

Sources: The Kochava Collective; MIT Elections Lab. Median household income is from the 2012-2016 ACS 5-Year estimates and is shown in 2016 inflation-adjusted dollars. Age data is from the 2012-2016 5-year estimates. Share of rural population is from the 2010 Census.