Conventional wisdom suggests construction has done relatively well combating the spread of COVID-19 on U.S. jobsites.
While there have been outbreaks at high-profile projects such as Allegiant Stadium in Las Vegas and as many as 13 San Francisco Bay Area jobsites, the industry as a whole hasn’t drawn attention for the debilitating scourges that have made news in healthcare, nursing homes and the meatpacking industry.
“Our industry has not suffered the kind of widespread workplace outbreaks that have hobbled so many other sectors of the economy,” wrote Stephen Sandherr, CEO of the Associated General Contractors trade association in a letter to members earlier this month, encouraging firms to stay vigilant against COVID-19 amid a recent surge of more than 160,000 new daily cases across the country.
But now, academic studies, an increasing amount of local public health data and anecdotal media reports are calling into question construction’s clean bill of health on COVID-19 mitigation. Consider:
- Construction had 106 COVID-19 outbreaks in Washington state through Nov. 12, the third highest total for all settings, behind only restaurants and produce packing, according to the Washington State Department of Health.
- Construction and manufacturing had the third highest total of all outbreak settings tracked by the Michigan Department of Health and Human Services through Nov. 12, behind only long-term care facilities and schools.
- Construction sites account for the second highest number of COVID-19 clusters in Nashville, Tennessee, behind only long-term care facilities and ahead of bars, according to the Metro Public Health Department.
- Construction workers in Texas are five times more likely to be hospitalized with COVID-19 than workers in other industries, according to a study conducted by researchers at the University of Texas Austin COVID-19 Modeling Consortium.
- A CDC study that examined COVID-19 workplace outbreaks in Utah found construction to have the second highest number of cases among all industries studied.
These data points contradict the widely held belief that the mitigation steps of screening, personal protective equipment and social distancing practices on jobsites have been effective in curbing workplace spread of the novel coronavirus in the construction industry.
They also bring into question, given the lack of a national data clearinghouse of COVID-19 infections by occupation in the U.S., whether the construction industry, which has been deemed essential in most states and continued working, has done as well as it thinks it has in preventing workplace spread.
“Outbreaks in construction sites may be very common,” said David Bui, epidemic intelligence service officer at the Centers for Disease Control and co-author of the study that found construction sites in Utah had the highest number of workplace outbreaks behind manufacturing. “Capturing data on workplace outbreaks has been challenging for a lot of health departments, so national level data is not available.”
The collective numbers have also emerged at a time when several states are walking back re-openings and even shutting down certain business sectors again as cases soar. Three states — Virginia, Michigan and Oregon — have implemented emergency rules in response to the pandemic, including protocols for screening of employees and visitors, physical distancing measures and the use of masks and face shields.
The AGC’s Nov. 3 letter to its members urging continued diligence against COVID-19 was motivated by a concern of the initial worksite lockdowns of March returning to the sector.
“Unfortunately, the number of confirmed cases has begun to grow at an alarming rate in many parts of the country,” Sandherr wrote in the letter. “It is safe to assume that any number of state and local officials will come under significant pressure to impose comparable new lockdown measures here during the next several weeks and months. In other words, our industry once again finds itself a simple stroke of the pen away from being forced to stop work.”
No national data
Part of the problem, multiple experts contacted for this article said, is that there is no single source for tracking COVID-19 infections across the construction industry on a national level. While the Occupational Safety and Health Administration tracks positive cases in the workplace via its Form 300 Log of Work-Related Injuries and Illnesses, those are only tabulated annually.
Brian Turmail, vice president of public affairs and strategic initiatives at the AGC, said the association’s position that construction has curtailed widespread outbreaks thus far is grounded in an absence of media and company reports that have pinpointed them in other industries.
“It’s based on a lack of data,” Turmail said. “Many other sectors of the economy have had widespread shutdowns. Meanwhile, we haven’t really heard from members of any kind of significant shutdown because of the coronavirus or spread of coronavirus on the jobsite.”
Construction Dived found news reports of more than 20 outbreaks of COVID-19 on construction sites in the U.S. and Canada. See our tracker here.
But other industry watchers said there are almost certainly more cases of COVID-19 at construction sites than are being publicly tracked.
“When you add it all up, there are outbreaks occurring that are just not getting reported,” said Russell Carr, president of Austin, Texas-based health and safety consultancy Berg Compliance Solutions. Part of the problem, he said, is construction’s “macho” attitude of working through illness, and that workers aren’t financially motivated to honestly report potential exposure, or even positive tests, because they know they’ll miss work if they do.
“There’s every incentive to just ignore it and go to work, and no incentive to report it and stay home,” Carr said. “People are just continuing to go to work.”
Turmail said some contractors do offer paid sick leave for workers who test positive for COVID-19.
Inconsistent mitigation practices
Remy Pasco, a graduate research assistant and one of the authors of the University of Texas study, said part of the reason why construction workers are five times more likely to be hospitalized with COVID-19 is simply due to the fact that they have continued to go to jobsites during the pandemic, while those in other professions have been working remotely.
“I should note that the rest of the population we compared these workers to is everyone not in the construction industry,” Pasco said. “There are some people in the other industries that were considered essential workers, and some that were probably working from home as well.”
He said the study’s comparison group, which was drawn from anonymized data of more than 2,200 COVID-19 patients admitted to Austin-area hospitals through Aug. 20, included individuals in the healthcare, farming, education and legal sectors, as well as retired and unemployed workers.
He also said the researchers saw evidence of inconsistent mitigation practices from jobsite to jobsite, especially between large sites to small ones.
“I don’t want to generalize and say that all construction work sites are not doing enough, because we know that’s not true,” Pasco said. “We have seen exemplary work sites, where people are wearing masks, they’re washing their hands and they’re trying to socially distance as much as possible. But then you go one block further and some smaller sites just don’t have the capability of having the resources to handle it.”
The study, published in the American Medical Association’s peer-reviewed JAMA Network Open journal Oct. 29, set out to test a model that was developed to help the city of Austin determine what would happen if construction sites stayed open during the pandemic. With a high degree of accuracy, it proved the model’s predictions that construction workers are four to five times more likely to be hospitalized with COVID-19.
“It’s not something we suspect anymore,” Pasco said. “It’s something we’ve seen and established.”
But Pasco also acknowledged the study didn’t look at causality, and instead just observed data collected from COVID-19 patients who worked in numerous industries, including construction. In other words, the study didn’t analyze or conduct contact tracing to determine how those patients contracted the virus.
Turmail pointed to a later report suggesting that the COVID-19 outbreaks among construction workers in Austin were due to socializing away from work, and not necessarily their actions on jobsites.
“Really the challenge here is that there is sort of the unique demographics of the construction industry, particularly in Austin, and kind of living conditions and social settings are more of a factor than the lack of jobsite safety precautions,” Turmail said. “It doesn’t change the fact that a higher percentage of people with coronavirus obviously still poses a higher risk for other workers, and that we’ve got to protect against that.”
Executives at large construction companies, including New York-based Plaza Construction, Minneapolis-based Mortenson Construction and San Jose, California-based subcontractor Rosendin Electric said an increasing challenge across the industry has been policing workers’ behavior away from the job.
“When we’ve had COVID infections on our projects and we’ve investigated those through contract tracing, the vast majority of those have been from outside of work activities,” said Derek Cunz, senior vice president at Mortenson, which was part of a joint venture with McCarthy to build Allegiant Stadium in Las Vegas, where at least 31 workers tested positive for COVID-19 in May. “We’ve been reminding people and communicating more now about making safe choices outside of work.”
Recent research from the National Institute for Occupational Safety and Health suggests that construction workers have a higher prevalence for participating in the types of social and other behaviors that put them at risk for COVID-19 infection. The study found that six health risk behaviors were more common in the construction industry: smoking, smokeless tobacco use, binge drinking, no leisure-time physical activity, not always using a seatbelt and getting less than seven hours of sleep a day.
Several sources contacted for this article said that if COVID-19 numbers were significantly higher on construction sites than commonly thought, the public would have heard about that by now, even in the absence of a national data reporting framework.
Take, for instance, the experience of Dr. Dan Carlin, CEO of New Hampshire-based JobSiteCare, which provides telemedicine services to remote construction sites. Since the onset of the pandemic, he has coordinated daily COVID-19 screenings of 10,500 workers for Florida-based Moss Construction.
In that timeframe, he’s found just 171 confirmed cases of COVID-19 among those construction workers, or less than 1.6% of the entire workforce. And due to those daily screenings, which include a mandatory temperature check, Moss hasn’t had any shutdowns during that time.
But while Carlin pointed to screening, social distancing and mask wearing as the holy grail of mitigation measures that has helped produce those results, as a doctor, he also knows many of construction’s underlying characteristics are on his side, too.
Those include the industry’s practice of donning personal protective equipment as a job requirement in normal times, as well as many work sites essentially being located outdoors, which has proven to be one of the strongest deterrents to COVID-19 spread.
“When this party started back in February, we thought it was a surface virus,” said Carlin, a former medical officer in the U.S. Navy who also oversaw the WNBA’s season bubble, which had zero COVID-19 positives, this year. “But what we’ve learned is the way the virus really knocks us out is when it gets in the air in a closed environment.”
Carlin said that if he was forced to choose between handwashing and wearing a mask to mitigate COVID-19 spread, he’d choose a mask, hands down.
“Honestly, that’s the most important thing, is you’ve got to limit atmospheric exposure,” Carlin said. “And the way to do that is work outside.”
Lack of comprehensive data
Brian Knudsen, vice president at Portland, Oregon-based Andersen Construction, learned that the hard way when a pre-construction meeting held indoors in October led to an outbreak at the company. In all, 16 cases have been linked to the meeting, including 13 employees and 3 close contacts, according to the Jackson County Health Department.
Knudsen said people were wearing masks during the meeting, but that sometimes workers take their masks off indoors, and instead choose to maintain social distance.
“We’ve learned that this is a very contagious illness,” Knudsen said.
Like others in the industry, he sees increased prevalence of workers’ activities off the jobsite contributing to increased risk on the job now.
“It feels like it’s going to kind of flip around, where before maybe work was bringing it home, now maybe it’s going to be more home bringing it to work,” Knudsen said.
The lack of comprehensive data on the number of outbreaks in construction nationally leaves the industry to practice a guessing game of where the numbers truly are, related to other sectors.
“So, we’ve seen widespread outbreaks in other industries, but we haven’t seen them in construction,” said Jack Dennerlein, a professor at Northeastern University in Boston who studies occupational health and safety in construction. “But we don’t have a good surveillance system, either.
“So what we might be seeing with these outbreaks in other industries may just be the tip of the iceberg, poking above the fray of our poor data system. Below the water, there might be some stuff going on in construction, and we just haven’t been able to measure it.”