Every state in the US is starting to relax stay-at-home guidelines and allowing businesses to reopen. But even before the vast majority of states relaxed measures meant to keep people physically apart and slow the spread of the coronavirus, people were starting to travel more and see each other more, according to research from Northeastern's Network Science Institute.
“You start to see a gradual gradient in early May, which you follow when the federal physical distancing guidelines are not renewed,” he says. Brennan Kleina fifth PhD student who helped analyze location data from millions of mobile devices.
Since the end of April, the average person in the United States had reduced their daily mobility by between 45 and 55 percent, the researchers found, and the number of people who came into contact with at least one other person outside their home had decreased by 65. and 75 percent. The recent rise is small, nowhere near pre-pandemic levels, but the team continues to update the data in public board to help people visualize what is happening in their own situations.
“The question is whether that curve will continue to rise,” he says Matteo Chinazzi, research associate at the Network Science Institute. “To see if these behaviors will actually return to normal or if this just means that there is a little more activity compared to what we had with the stay-at-home policies, but still a long way from what would be typical.”
This data also helps guide models that predict how the disease will spread, which many governors and public health officials use to guide their decisions.
“You can plug our numbers into the epidemic model of your choice to show some kind of reference to reality,” says Chinazzi. “You can look at the actual data and say, 'OK, for the city of Boston or for Seattle, I know this happened.'
Location data provided by Cuebiq, a company that collects GPS data from consenting users' mobile devices. This data is anonymized and excludes members of vulnerable populations, such as minors or incarcerated individuals, but still provides the ability to analyze large-scale mobility patterns.
“Mobility data is really powerful for some of these big insights about how people move, not only during normal periods, but also during epidemics or pandemics like COVID-19, and also how people respond on a global scale in containment strategies like social distancing or shelter in place,” says Brennan Lake, who runs Cuebiq's Data for good program.
The researchers tracked five metrics to understand how people move around the United States. At the individual level, they measured how far the average person traveled from their usual area, how many people had at least one contact per day, and how often two people were in the same location. At the regional level, the researchers measured the daily volume of travel and transit between major urban areas. Of these, only daily travel volume has not increased since the beginning of May.
But just because people are moving more doesn't necessarily mean they aren't safe.
For example, the researchers track what they call “co-location events,” meaning two mobile devices are within 60 feet of each other within a five-minute window. While this is a very good proxy for the number of contacts, it doesn't tell us what these people did.
“If I go for a walk with Brennan and we walk around the Chestnut Hill Reservoir for 30 minutes, we'd be in the same location six different times,” he says. Stephen McCabe, another of the PhD students who have devoted their time and energy to this project. “These are not necessarily going to be contacts in the sense of breaking social norms about staying six feet away from each other.”
Wearing masks, meeting outside with small numbers of people and keeping your distance can help minimize the risks of leaving your home, according to public health officials. If people travel a little further and see a little more people, these safeguards could make a difference.
“Not all mobility enhancements are created equal,” says McCabe. “And they're certainly not all bad, per se. Getting on with life under this, to use the cliché, 'new normal' is a learning process.”
For media inquiriescontact Jessica Hair at j.hair@northeastern.edu or 617-373-5718.