What your cellphone reveals about Weld County’s compliance with stay-at-home orders
By Kelly Ragan
A couple companies are tracking your movement via your cellphone for pandemic research, and they say Weld County isn’t doing very well with stay-at-home orders.
According to Unacast, a New York-based company that tracks our cellphones to collect mobility data, Weld earned a D on its Social Distancing Scoreboard as of Wednesday.
Unacast’s data allows folks to compare travel now to what activity in the community was like prior to COVID-19, according to its website.
According to a report by Colorado Politics, Unacast originally sought to determine the average time people spent in their homes due to the pandemic, but the company pivoted. Because cellphone signals are most reliable when they’re moving, the company is now focusing on changes to the average distance people travel.
So, the company created what it calls a Social Distancing Scoreboard. It reports two percentage ranges for each county and state: average mobility and number of non-essential visits. The average mobility figures are based on distance travelled, according to the company, and non-essential visits include trips for “non-grocery retail goods and services.” (The company acknowledges this is an imperfect definition of non-essential, and it promises to improve this metric.)
In Weld, the scorecard shows folks are traveling less than before COVID-19; average mobility shrank in the county by about 24-40%, and non-essential visits in Weld declined by 55-60%. The scorecard also knocked the county for having 537 confirmed cases of the virus.
As of Wednesday night, Weld reported at least 607 positive cases and 35 deaths. Keep in mind the situation is rapidly changing and it’s common for different systems to update numbers at different times.
Larimer County earned a B, with Unacast showing average mobility shrank by 40-55% and non-essential visits declined by 70%. The report showed 161 confirmed cases of COVID-19.
Adams County earned a C. With Unacast reporting 388 confirmed cases, the cellphone data showed average mobility declined 40-55% and non-essential visits decreased by 60-65%.
Statewide, Coloradoans have cut back the distance they travel each day by about 37% since late February, the data show. That’s earned the state a B-, in line with the country’s collective score.
You can see the county and state scores for yourself here.
“The Social Distancing Scoreboard and other tools being developed for the Covid-19 Toolkit do not identify any individual person, device or household,” wrote Thomas Walle, CEO and co-founder, in a statement on the company’s website.
Google is also tracking your phone for its own COVID-19 Community Mobility Report. In the report, Google stipulates “location accuracy and understanding of categorized places varies from region to region,” so they don’t recommend comparing rural to urban areas with the data.
According to Google’s report, Weld’s compliance with stay-at-home orders looks like this:
Retail and recreation sites visits (think restaurants, libraries and movie theaters) are down 51% from the baseline.
Grocery store and pharmacy visits are down 27%
Park visits are down 12%
Transit station (think bus stops) use is down 60%
Workplace visits are down 40%
Residential visits are up 13%
Check out the data for yourself here.
Jennifer Finch, spokesperson for Weld County, wrote in an email that county officials don’t know how many people are following orders.
“The fact is, we really have no idea how many people are or are not abiding by the stay-at-home order,” Finch wrote. “We are a county that is 4,000 square miles in size, home to 300,000-plus residents and 32 municipalities.”
She also wrote the data provided by the platforms don’t tell the whole story.
“For example, someone living in a rural community has to travel farther to go to a grocery store – this cellphone tracking may see that trip as non-essential, however it is very essential for that family,” Finch wrote. “I have to come to work every day – that doesn’t mean I’m not abiding by the stay-at-home order.”
But according to a blog post put out by Google on April 3, Google isn’t measuring essential versus non-essential trips; it’s measuring how current travel compares to “trends over several weeks, with the most recent information representing 48-to-72 hours prior.”
And it’s likely Unacast’s system would consider a long trip to the grocery store from a rural area “essential” travel, since the data would show the cellphone traveling to and from a grocery store, pharmacy or pet-supply store; in theory, that the phone made the trip from a rural area over a long distance wouldn’t matter. Such trips could, however, decrease Weld’s score in the average mobility metric.
Still, the important aspect of the data is it compares travel before and during social distancing measures. So epidemiologists, elected officials and even Joe Schmoe on the internet (it’s me, I’m Joe Schmoe) can use the data to see if we’re staying home more than we were a couple weeks ago.