Another Uber driver arrested for sexual assault. That was one of the headlines when a Boston woman reported her driver "indecently touched her several times" last month, according to the Boston Police Department.

Such incidents seem frighteningly common now. In the past year alone, there have been several high-profile reports of drivers attacking passengers of ridesharing services like Uber. In the United States, there were assaults reported in Boston, Washington, D.C., Chicago, Oklahoma, Los Angeles, and Orlando.

But how dangerous is Uber compared with a taxi or limousine?

"There's no way to search for that," said Neva Coakley, a spokeswoman for the Boston Police Department. "We wouldn't be able to speak to that because we don't have data to support it. We don't distinguish between what type of suspects they are."

In other words, Boston doesn't track assaults by where they happen—in a taxi, in an Uber, or in someone's home—so there's no data to compare reports against Uber drivers versus taxi drivers or limo drivers. That's true in other cities, too. We asked police departments in five cities—Boston, San Francisco, Chicago, New York, and Washington, D.C.—for information about assaults against passengers of taxis or Uber cars. None of them tracked violent crimes at that level. This is meaningful because it underscores how the narrative about ridesharing and public safety is largely anecdotal. It raises another question, too: If Uber is potentially unsafe for passengers, what about taxis?

“We don’t keep specific data with relation to passengers being assaulted,” said Lee Jones, a spokesman for the New York Police Department.

“We don’t track offenses by whether they occurred in a taxi cab,” said Gwendolyn Crump of the Metropolitan Police Department in Washington, D.C.  

“As far as tracking the offenses in taxi cabs specifically, I don't have a way to look this up,” said Grace Gatpandan, a public information officer with the San Francisco Police Department in the city where Uber is headquartered. Police officers in Chicago did not respond to multiple interview requests.

Allan Fromberg, a spokesman for the New York City Taxi and Limousine Commission, said some Uber complaints are invisible to law enforcement. “Apparently, people often complain to Uber through driver ratings, which only Uber sees.” (A spokesman for Uber says it’s the other way around: Passengers often call the police before they complain to Uber, and the company finds out about alleged crimes through law enforcement.) One website that tallies Uber incidents is run by the Taxicab, Limousine & Paratransit Association, a trade association representing taxi and limo companies—a group that has a stake in making Uber look bad. “Unfortunately we don't have hard numbers, but anecdotally,” said Dave Sutton, a spokesperson for the association, “we see this as an outrageous number of incidents.”

* * *

Uber’s recent track record has been troubling. In addition to the recent assault reported in Boston, an off-duty Uber driver is being investigated in connection with a sexual assault near Los Angeles earlier this month, according to the Los Angeles Times. Uber insists it meets the highest standards for safety, and says its background-check process has been in place since April. That process applies to UberX rideshares, but not to black cars and SUVs—those are managed through partnerships with existing local limo companies. "Unlike the taxi industry, our background checking process and standards are consistent across the United States and often more rigorous than what is required to become a taxi driver," the Uber spokesman Taylor Bennett told us. Uber uses a private company called HireEase to conduct background checks.

"Ours cover courthouse records, county, state, and federal records," Bennett said. "We cover the gamut in terms of what we look at."

There have been several reports about loopholes that have led to Uber approving drivers with criminal histories including felony convictions. And the Fair Credit Reporting Act limits the amount of information HireEase is able to uncover. (“The [act] does not limit the time for reporting convictions,” the company explained on its website, “However, adverse matters that did not result in a conviction are only reportable for seven years under the FCRA.”)

So while taxi companies check a prospective driver's fingerprint records against a database that theoretically (more on that in a minute) includes a person's complete criminal history in the United States, Uber background checks use a database that can only go back seven years for some information. Late last year, San Francisco District Attorney George Gascon called Uber's background checks "completely worthless," according to the Los Angeles Times. And several Uber and UberX drivers in the Washington, D.C., area said Uber's background checks were hardly rigorous.

"Not really," said an Uber black car driver in Washington, D.C., who requested anonymity because he feared retribution. "Look, the real background check is getting fingerprinted and that takes how many weeks? This one is just online. A real background check is not like this. A real background check, the government approves you and checks everything. But Uber checks kind of. They're kind of background checks."

Uber, of course, disagrees with this assessment. Fingerprint scans, the kind used by many taxi services, Uber’s Bennett told The Atlantic, offer just a "snapshot in time." The FBI’s process takes up to 16 weeks, and involves checking an individual’s records related to arrests, federal employment, naturalization, or military service, according to information provided by the agency. “Well it goes as far back as the records we have, and we've been doing this since 1924,” FBI spokesman Stephen Fischer Jr., said. “The big key is it is a voluntary process, so law enforcement agencies are not required to provide us their arrest data and criminal history information. We rely on the agencies to provide us the most accurate and up-to-date information, as we are just the repository.”

Uber’s checks are at least rigorous enough that not everyone passes, Bennett said. Ten percent of Boston taxi drivers who took Uber’s background check failed, he said. And some Philadelphia UberX drivers who passed the city’s background test ended up failing Uber’s, he said.

In reality, no background check is perfect. That's according to Michael Fertik, author of The Reputation Economy, which explores the growing influence of a person's online reputation. "If someone has no DUI but every photograph of him is drinking, and then that guy gets into an accident, God forbid, then the media would excoriate Uber for having hired him," Fertik said. "At the same time, before the accident, you can't not hire him because he's photographed drinking."

"If someone has a history of disparaging women, given Uber's current PR challenges around women's safety, you have to ask, 'How is this person disparaging women? Does it feel real? Does it feel humorous? Do we care about that distinction?’"

Without a perfect background check, there’s no way to be fully confident about a hire. Every hire ends up being something of a risk analysis. And it’s not just the hiring process that’s imperfect, Fertik says. Humans happen to be flawed, too.

"There's a difference between making an offhand remark and assaulting someone,” he said. “At some point, people have to be able to live their lives. You can't expect everyone to be Mitt Fucking Romney. Not everyone is a robot. It's not like everyone is going to pass the ultimate background check. It doesn't mean they get to be president, but they might be able to become an Uber driver.”

* * *

What may be more revealing about the safety of Uber versus traditional taxicabs is what happens once an incident has occurred. After an assault against a passenger is reported, Uber says it immediately removes a driver’s access to the app while the investigation is underway. Uber also points out that as soon as a driver is suspended, that driver can no longer access the app to pick up a passenger. And without access to the Uber app, drivers don’t have access to passengers. "If there are instances of serious behavior that don't meet our quality standards then we absolutely will deactivate riders and drivers," Bennett said. The company doesn’t have numbers on how many drivers are suspended for good, versus suspended and reinstated, he said.

In cities like Washington, D.C., and New York, taxi companies and the commissions that oversee fleets describe a similar process for suspending drivers accused of assault. In Washington, drivers are suspended as soon as criminal charges are filed, said a spokesman for the District of Columbia Taxicab Commission, the city agency that oversees the local taxi industry.

Taxi drivers have been in the headlines just like Uber has. In the past year, there have been assaults against taxi passengers reported in Seattle, Washington, D.C., Portland, Fort Lauderdale, and elsewhere. In 2012, a rash of incidents in Washington—seven assaults over the course of a few weeks—prompted the District's taxicab commissioner to issue a warning to female passengers. At the time, the commissioner promised panic buttons would be installed by the end of that year. Now, three years later, the target date for installation is June of 2015. (Uber says it will add a “panic button” to its app for Chicago passengers later this year, according to the Chicago Sun-Times.)

The taxi commission in Washington acknowledged there are still concerns about assault—both from passengers and from drivers—though the scope of the problem is not clear. “It's not as if there’s a standard procedure where we get all incident reports,” said Neville Waters, a spokesman for the commission. “A lot of it depends on the commander of the particular ward. That also doesn't necessarily mean that all assaults that are occurring in a cab are being captured or that we're even getting that information.”

The questions about Uber’s safety are part of a much larger debate over regulation of what’s come to be known as the sharing economy. And the narrative about Uber rides being inherently more dangerous than taxi transportation is, not surprisingly, being pushed by those most threatened by Uber's success. Calling Uber a "success"—the company is valued at $40 billion—seems like something of an understatement. Uber is trouncing the competition. There were 2 million fewer taxi rides in Washington, D.C., last year compared with the year before, Waters told me—that’s a 9 percent drop in taxi business in a single year. And taxi business declined a stunning 65 percent in a two-year period after Uber showed up in San Francisco. All this is to say that the reputation Uber has as a danger to its passengers is developing against the backdrop of a business fight that Uber appears to be winning swiftly and handily.

Concerns over Uber safety are troubling—and the company hasn’t always seemed to take its riders’ concerns seriously. But there’s little to suggest that the newest form of ridesharing is significantly riskier than the old one. As for the relative safety of taxis versus ridesharing, Waters preferred not to speculate. “I know very little about the company’s operations,” Waters said. “They're pretty secretive and they certainly don't share data with us. I think it would be unfair to base an opinion... I've never used the service. I don't really know.”