Cell phone location data maps where people are social distancing — and not

Unacast grades states on social distancing

How well are people adhering to recommendations on social distancing and staying at home to help curb the spread of the novel coronavirus? Cellular device location data is providing some insights on that.

Norwegian company Unacast, which is backed by Telia and other investors, last week launched a map in response to the COVID-19 pandemic which analyzes smartphone location data from the United States, on a state-level scale, and gives each state a letter grade on how well its citizens are doing on social distancing.

According to Unacast, the top five areas which are doing the best at social distancing are the District of Columbia and the states of Alaska, Nevada, New Jersey and Rhode Island.

The bottom five states in terms of social distancing were Oregon, New Mexico, Idaho, Montana and Wyoming. Wyoming was the only state to earn an F on Unacast’s “Social Distancing Scoreboard,” which you can find here. 

Unacast’s CEO and co-founder Thomas Walle said in a blog post that his company believes that the scoreboard data, as well as other COVID-19-related data sets that it is working on, “will not only help make sense of what’s happening now, but unearth trends that will help project scenarios in the short- and mid-term future.”

Walle outlined how Unacast developed the data scoreboard. First, he said, the company started by looking at behavior that indicated “staying at home” compared to “not at home”; changes in the amount of time spent at home and not-at-home; changes in the dispersion of “activity clusters” of people gathering together; and change in the average distance traveled. Change in average distance traveled ultimately ended up being chosen as the best starting point, he added, because it correlates well with the number of confirmed cases (the more cases are confirmed, the greater the decrease in the average distance traveled on the county level, according to Walle); it doesn’t require “strong assumptions,” such as an assumed home location; and works independently of supply-side fluctuations.

“We get the most signals when people are moving so the metric is unaffected by changes in the ping frequency due to inactivity when people dwell at home,” Walle wrote. It also, he went on, captures how people adapt their everyday behavior, such as working at home, canceling travel and avoiding non-essential trips outside the home.

Valuable as the data may be, it still raises issues around individual users’ privacy. The Federal Communications Commission proposed $200 million in fines to the four national U.S. carriers earlier this month. Each of the carriers was accused of disclosing their customers’ location information, without their content, to a third party who wasn’t authorized to receive it and who sometimes was selling information on the location of identified individual’s phones information to non-law-enforcement or to individuals, and dragging their feet on shutting down that access once the problem became clear.

Unacast addressed those privacy concerns by saying that the scoreboard and COVID-19 tools is it working on “do not identify any individual person, device, or household.”

In order to calculate the underlying social indexing score, Walle said, “we combine tens of millions of anonymous mobile phones and their interactions with each other each day – and then extrapolate the results to the population level.” Since the company operates both in Europe and the U.S., it says it has been operating within the GDPR privacy law’s guidelines since May 2018 and also has adopted California’s CCPA guidelines for all of its U.S. operations.

Leave a Reply