If you had to guess, how often would you say a car smashes into a building in the United States? I’m not talking about the time you scratched the paint off your rear fender on a tight turn out of an alley. I mean solid crashes—like the time a New York City bus spent four days inside a Brooklyn townhouse last year. Crashes that do enough damage to get reported.
The best guess for the past few years has been 60 times a day—a number that a security consultant named Rob Reiter derived from a database he has built of 36,000 car crashes. It’s a number that’s been picked up by news reports and trade organizations, credited to Reiter’s operation, the Storefront Safety Council.
Last month, Reiter had his data reviewed by a risk management company, CHC Global, an arm of Lloyd’s of London, the insurance market. Their conclusion: The data was sound and there was nothing like it out there. But also: Reiter’s database was undercounting real-world collisions with buildings.
So Reiter’s new projection is that there are at least 100 American drivers veering their cars into buildings each day, or 36,500 a year, and that’s the low bound. The revised figure corresponds to 16,000 Americans injured in car-on-building crashes each year, he believes, and more than 2,500 people killed.
Putting a number on this phenomenon would be helpful to street designers, retailers, shopping mall owners, and so on. But figuring out how much property damage cars are causing each year—and how many human casualties result from those crashes—is a surprisingly tough problem. Even with the CHC Global analysis in hand, I can’t really say if the Storefront Safety Council numbers are on the money. A few experts I spoke to expressed some skepticism. Jonathan Stiles, a postdoc at the Collaborative Sciences Center for Road Safety at Florida Atlantic University, said he would not trust the numbers without a fuller account of the methodology than the Storefront Safety Council makes public.
Norman Garrick, a professor emeritus at the University of Connecticut, expressed a similar concern. He noted that Reiter’s estimate that 8 percent of building crashes result in a death is 10 times the National Highway Safety Transportation Administration’s own figure, and that that estimated death toll might be biased upward by the tendency of fatal crashes to make the news.
On the other hand, Reiter’s collision total lines up almost exactly with the only other source for this particular statistic I could find: a 1973 report prepared for the Department of Housing and Urban Development that estimated 20,000 to 40,000 car-on-building collisions a year. Doubling those numbers to account for the national growth in vehicle miles traveled puts HUD’s low bound in line with Reiter’s.
That’s not very reassuring, if you’re a building or a person inside one: 1973 was when the per capita car crash death rate hit its all-time high in this country. It was the drunk-driving era. Since then we’ve spent decades building to give cars the run of the built environment. Automobiles themselves now have driver-friendly features like sensors and cameras. Yet, to go by Reiter’s figures, we seem to have scarcely gotten any better at not running into buildings.
In any case, the fact that Reiter’s homespun approach has delivered the nation’s pre-eminent statistic on cars hitting buildings is a symbol of how the externalities of a car-dependent society go unseen and unstudied.
“We’re losing hundreds of people a year in what everything thinks of as a series of one-off unfortunate events,” Reiter said, “and the reality is, it’s organized mayhem”
For the past eight years, Reiter has spent 12 hours a week logging entries into his crash database. Articles and TV clips. Academic research. Fire and police reports. Litigation. Twelve Google Alerts, seven alerts on other online services, a couple dozen sleuths who send him a note every time someone’s driven their car into the supermarket again.
I first met Reiter five years ago, when I wrote for Slate about the possibility that the country’s counter-terrorism fixation might for once produce a more friendly, open urban environment by prodding officials to create more car-free spaces, in the vein of Wall Street in New York or Pennsylvania Avenue just north of the White House. At the time, Reiter worked for a Los Angeles-based bollard company, Calpipe. He invited me to sit on a panel at a security conference.
“I’m a security guy, a bollard and a barrier guy,” Reiter told me the other day. In other words, he is not exactly a disinterested party when it comes to raising awareness about how often cars crash into buildings. The organization he co-founded, the Storefront Safety Council, is the keeper of the car crash data—and an awfully good advertisement for bollards. “It’s the dumbest thing in the world,” he said of the humble bollard. “And it works really, really well.”
No disrespect to Rob, but my first thought was: How could no one else be keeping track of this? One answer might be found in an Insurance Institute for Highway Safety report on fatalities in car crashes with fixed objects: In 2020, only 210 people died in fatal collisions with buildings, according to federal data. More than half of fatal collisions are with trees, utility poles, and traffic barriers.
But people killed in car-on-building collisions may also be missing from the government’s pedestrian fatalities count, suggested the planner and writer Angie Schmitt, the author of Right of Way. “Official pedestrian fatalities leave out a few thousand fatalities that happen on private property,” she said.
Indeed, because so many of these collisions occur at relatively low speeds, in parking lots or on city streets, even if the death toll runs into the thousands, as Reiter alleges, it’s just a small fraction of total roadway fatalities. There is also lots of building damage, though the American Property Casualty Insurance Association didn’t have any data either. Nor did the National Safety Council. Advocates for Highway and Auto Safety was also at a loss.
The CHC Global audit suggests the Storefront Safety Council might be undercounting crashes by quite a bit. For example, the SSC counted two incidents when a car crashed into a Las Vegas bus shelter in 2020. But the risk consultants got data from Las Vegas showing that it happened 27 times that year. Storefront Safety Council batting average: One in 13.5.
Or consider a Texas Traffic Institute study, which Reiter worked on, which used corporate data that showed the nation’s 7-11 stores were absorbing more than one car crash a day. The SSC only records a fraction of those—leading the auditors to conclude that the SSC might only by counting one in every 91 crashes!
So: There’s some uncertainty in the projections. Maybe the broader lesson is simply to marvel at the fact that we don’t even know how many times a day a car smashes into a building in the United States. It’s just one of those things about car culture that goes uncounted, unthought of, until you’re the one waiting for the register when an SUV comes careening through the front window.
