Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

How Much Does Transit Spread COVID-19? The Jury Is Still Out.

We assume that squishing people together on subways and buses, along with urban density in general, accounts for much of the virus's spread. But when you look at the evidence, it's a blurred picture.

New York City's subway system, carrying millions of passengers daily, seems like the ideal way to spread COVID-19. But the evidence isn't so clear. (Shutterstock)
On March 11, just as New York City was beginning to shut down because of the COVID-19 pandemic, a friend and I rode the subway from our Brooklyn homes over to Manhattan's Lower East Side to see the classic 1982 film Blade Runner. We knew it was a bit foolhardy. But in a mostly empty theater we watched Harrison Ford battle replicants. Afterward on the subway platform, we said goodbye.

That was the last time I saw my friend, and the last time I took the subway. Sheltering in has reduced my need to take a bus or train, but it's also because I'm scared. I imagine sitting close to others with the air circulating the coronavirus. And I know that thousands of the Metropolitan Transportation Authority's more than 70,000 workers have contracted the disease and that more than 130 have died.

This combination of reduced need and increased fear is why the region's famously busy subway, bus and commuter-rail system is carrying a small fraction of the millions it used to daily. Chicago, Washington, D.C., and other transit-dependent cities have seen similar drops as millions stay home or turn to other means of getting around. You might need to get on a waiting list to buy a bicycle.

In an effort to make their systems safer, transit authorities are scrubbing and cleaning as never before. New York City, for the first time in the subway's history, has suspended 24-hour service so workers can deep clean the trains in the wee hours. Around the world, technology is being brought to bear: Shanghai is trying out ultraviolet light to kill the virus in buses and trains. Hong Kong is using a disinfection robot.

It may be taken as a given that squishing people together on a train or bus is sure to spread COVID-19. But what if this isn't as true as it might appear? What if New York's status as the nation's densest and most transit-dependent city is not the primary reason its inhabitants have the highest infection rate of any big city in the country and a death toll of at least 21,000? The answer is important, because most cities, even suburban ones, have some type of mass transit.

I am not a neutral observer. I like big cities. I like not owning a car, and I like taking buses and trains. So take what I say here with a grain of salt, but hear me out.

When I look around the world, I see dense cities with large mass transit systems, such as Hong Kong, Seoul, Shanghai and Tokyo, that have far lower COVID-19 infection and death rates. So clearly, density and crowded trains and buses don't automatically produce a pandemic.

And when you look at New York City more closely, the evidence blurs. Staten Island, the most suburban part of the city with the least transit use, has one of the city's highest rates of confirmed cases, as do less-dense parts of Queens. There are high infection rates in more suburban areas of the metropolitan region, such as Nassau County on Long Island, and upstate in the rural area of Ulster County.

The East Side of Manhattan has the most crowded subway in the city, the Lexington Avenue Line. Pre-social distancing, it carried more than a million passengers a day, more than the entire systems of Boston, Chicago or Washington. Yet the districts along the line generally have lower COVID-19 case and death rates than the citywide average. Could this be because the area's wealthier inhabitants left town, perhaps carrying the infection with them? Hard to say given what we know and don't know at this point.

There are also big transit hubs where multiple lines come together, such as the Fulton Street complex in Lower Manhattan and Atlantic Center in Brooklyn. One would think this might correlate with infection rates, but no such evidence leaps out at you when you look at infection rates by ZIP codes.

On the website Market Urbanism, the economist Salim Furth studied the evidence and concluded, surprisingly, that car use in New York City correlates more strongly with infection than subway use. "To the small extent that transportation options matter," Furth writes, "automobiles appear to be more dangerous disease vectors than subways." While no one can say exactly why this might be so, Furth suggests two possibilities: Those dependent on subways may have reduced their travel more than the automobile-dependent, and drivers' greater mobility gives them the means to carry the virus beyond their immediate neighborhoods.

Still, all things equal, riding on a bus or train must have some positive correlation with spreading COVID-19 as well as other diseases. It stands to reason. But perhaps the risk is actually relatively low and countered by wearing a mask and gloves and regular cleaning of the cars.

What is needed are some good studies on how much mass transit and density spread disease and, if they do, how it can be countered. Perhaps a good place to start is studying why all those MTA workers got sick. Was it due to their direct duties, such as driving trains or being in ticket booths? Or could other factors have had big impacts, such as daily worker meetings or where the workers live?

Getting clearer answers is important, because history is driven by perception, and if people think riding a bus or train will infect them, they will avoid doing so regardless of the real risk factor.

My own theory about why New York City has such high infection rates is that it's the airports, which bring in more international passengers than arrive in any other U.S. city. It now appears that the coronavirus began spreading from Wuhan, China, to Europe in December, much earlier than had been previously thought. It's quite possible that infected people from Europe and Asia began streaming into the region's three international airports weeks or even months before the state's first official cases were diagnosed on March 1. The disease would have had much more time to spread unhindered than in cities with lower international traffic.

My logical mind tells me all this. But for the time being I'm still avoiding the subway and buses. I'll find other ways of getting around until better data confirms that they're safe, or at least as safe as they were before the arrival of the coronavirus. Maybe I'll get on a waiting list to buy a bike.

Governing's opinion columns reflect the views of their authors and not necessarily those of Governing's editors or management.

An urban affairs and infrastructure columnist for Governing. He can be reached at or on Twitter at @Amcities.
Special Projects