The data on how coronavirus is spreading at schools and colleges is inconsistent, erratic — and sometimes purposely kept out of the public’s reach.
Federal rules don’t specifically require tracking or reporting the numbers by school or college, despite pressure from President Donald Trump and Education Secretary Betsy DeVos to open schools and colleges for in-person classes. The result is a distorted picture of how and where the virus may be spreading, not just for parents, teachers, students and professors, but the cities and towns where campuses are located.
The plan in Texas was for the state to gather and share data, but the state teachers union said it is launching its own tracker because the Texas education and health departments won’t share any data for a few more weeks. Even then, it won’t break the information down by school. In Florida, the state department of health has told schools that information about cases is considered confidential and has pressured schools to quit public announcements about cases. Some districts are ignoring that command and sharing the information anyway. In Tennessee, state officials at first balked at sharing statewide data about cases in schools, citing student privacy. The state reversed that position and the new plan is to provide the information starting later this month.
Ad hoc trackers have cropped up to fill the national void. Kansas high school teacher Alisha Morris started her own tracker, which has been taken on by the National Education Association. The New York Times is surveying outbreaks on college campuses. Its latest numbers show at least 51,000 cases across hundreds of institutions and at least 60 deaths since the pandemic began.
The data about cases and testing in education settings is crucial to determining whether students are to blame if there is a sudden rise in the broader community. Researchers have found that kids infected with the virus are more likely than adults not to display symptoms and can still transmit the disease. If those children aren’t tested, they could spread the virus at home to older, more vulnerable relatives. If local officials don’t track what’s happening in schools, they may miss the seeding of a much more dangerous outbreak in the neighborhood.
Tracking the virus in a uniform and systematic way could also provide a national picture of the policies best suited to keep the virus out of schools and from spreading outside of schools.
“We’re going to have thousands and thousands of local experiences, which will not be comparable, and I’m afraid we’re going to have a lot of anecdotes and no useful data,” said Irwin Redlener, director of the Pandemic Resource and Response Initiative at Columbia University.
Some colleges and universities have their own public dashboards with detailed information about the number of tests given, the number who tested positive and how many are being quarantined. Others aren’t sharing anything, including with faculty who may have taught an in-person class with a student who has tested positive.
What data is accessible — sometimes crude tallies of total infections — isn’t the kind of stuff that would make it easy for public health officials to glean new strategies, said Tom Inglesby, director of Johns Hopkins University’s Center for Health Security.
“We’re not going to learn practical information that will help us make value decisions about asymptomatic screening on a large scale.” he said. “We don’t know what best practices are yet. … Why aren’t we doing everything we can to learn from what’s happening?”
Last month, a group of Senate Democrats pressed the Trump administration to start tracking outbreaks at colleges, warning that the lack of federal guidance on reporting coronavirus infections among students was "likely to create a patchwork of inconsistent information across states." They wanted an answer by Sept. 2 but haven’t received a reply.
The federal Education Department said it does not plan to track school or college cases or outbreaks.
To try to fill the information vacuum, AASA, The School Superintendents Association, created its own tracker to examine how things are evolving at schools based on their reopening plans. Just over half of school districts are taking a hybrid approach to reopen, just under a third are fully online and about 10 percent are open for fully in-person instruction, said Noelle Ellerson Ng, who leads the association’s legislative and advocacy efforts.
"I think the [Education] Department is uniquely positioned to absolutely play this role and we welcome the opportunity to have them work in this space collaborate, partner and lead," Ellerson Ng said. "Absent that work, we find ourselves in a position to fill that void and put data and science where it's not being provided by the federal government."
The Superintendents Association tracker will aim to monitor baseline information such as state, community type, poverty rate and opening status that would be updated every two weeks. The group said the tracker could be expanded to include which policies school districts used to reopen, such as using masks, visors, staffing patterns, school hours, and more.
“I don't know if it'll necessarily change decisions, but it will definitely allow people to be more confident in the decisions and plans that they put in place,” she said.
Trump and DeVos at times have downplayed the effects of the virus on children and young adults. In late July, Trump said children are less likely to get sick and die from Covid-19. “They don’t catch it easily,” he said. “They don’t bring it home easily. And if they do catch it, they get better fast.”
According to tracking by COVKID, run by epidemiologists who track cases in children and teens, nearly 618,000 children and teens have confirmed cases of Covid-19 as of Monday, though it estimates more than 3 million have been infected with the virus when accounting for undetected cases — the CDC has also said the true number of infections may be 10 times higher than the official tally. A total of 124 children and teens have died from the virus.
DeVos in July said that “There’s nothing in the data that suggests that kids being in school is in any way dangerous." She has been less forceful about students returning to in-person classes than her boss, however.
A drawback of the data that is gathered is it’s not necessarily coming from places where schools are reopening first or fast, noted Emily Oster, an economics professor at Brown University who offers data-based guidance to parents, including in her New York Times bestselling book “Cribsheet.” "One of the things that is sort of unfortunate is that many of the places that are opening are also places which, for whatever reason, are not in a situation where they collect a lot of data.”
“So you say: ‘Okay, let me pick the states I think would be most likely to be kind of systematic about collecting this.’ Those are not the places where schools are opening first. And so I think that that's, to some extent, influencing our ability to learn about this,” said Oster, who has risen in fame as parents seek advice informed by data during the pandemic.
The American Federation of Teachers, which has 1.7 million members, adopted a resolution this summer saying officials should only consider reopening schools where less than 5 percent of people tested for coronavirus turn up positive. The World Health Organization recommends hitting that 5 percent mark before for reopening the economy more generally. Some school systems hosting classes on campus have rates five times that.
The overall rate of positivity among Florida children is more than 14 percent. In Glades County, which abuts Lake Okeechobee in South Florida, the rate of positive tests among children is 27.8 percent. Soon after shifting classes from in-person to online, the University of North Carolina at Chapel Hill reported a positive rate of 31.3 percent.
School districts that are openly sharing cases based on their own tracking said they wouldn’t have it any other way.
Cherry Creek School District Superintendent Scott Siegfried created his own tracker, which he credits for keeping Covid-19 infection rates low.
His Colorado school district has 55,000 students and 9,000 employees. Elementary schoolchildren are in classrooms full-time, if their parents chose that option, while older students have in-person classes part time and work online from home the rest of the time. All parents were offered the option of working exclusively online, and that was the choice of about 11,000 students.
For the first weeks of school, the district largely was able to keep coronavirus cases low. More than 6,000 staff were tested, and nine tests came back positive.
Siegfried's tracker relies on a metric based on four data sets for its county: test positivity rate based on a 14-day average, daily hospitalizations, the 14-day incident rate per 100,000 residents and a three-day average of daily cases. Siegfried said he introduced this metric to the community as the way the school district would make decisions about whether to shut down if they do see a rise in cases.
“Whether we have school or not, or how long we’re in school is completely dependent upon our behavior — whether we wear masks, whether we limit parties, whether we limit gatherings, social distancing,” he said. “If it starts to spike in our community, it'll spike in schools and we'll close.”
The school district said the metrics are "based on science, facts and the best interests of our students and staff" and that "they are not based on politics, perception, popularity or what other school districts are doing.”
Still, seeing other school districts shutter due to case surges prompts concern.
“I'm worried every day and every night as I don't sleep — it's always a worry.” Siegfried said. “We will have infections just like it happens in our community.”
“But, if I were to use the data metric that we developed here in Cherry Creek and apply it to California, Arizona, Mississippi, Florida or Texas, we would absolutely not be in school,” he said. “We would be remote because those communities did not do as good a job as our state, county and community did.”
Jennifer Scholtes contributed to this report.