Taxi Bigotry Isn't Black and White, Cont'd

Editor’s Note: This article previously appeared in a different format as part of The Atlantic’s Notes section, retired in 2021.

Many readers over the weekend began to debate how race plays a role in cab drivers favoring certain passengers over others. The latest reader to write in:

I’m white. Some years ago, after dinner in Midtown, a black acquaintance with a train to catch at Penn Station asked me to hail him a cab, explaining that taxis won’t stop for him because they don’t want to go all the way up to Harlem or out to Brooklyn or the Bronx. They want to go where they have a good chance of picking up another fare in short order. They don’t want to drive empty back to Midtown. It was an economic decision based on experience-informed probabilities.

Another reader makes a similar point but goes into much greater depth—and controversy:

In his essay, Doug Glanville says this:

I also have come to understand that drivers have major concerns about safety or about the economics of getting return rides after dropping someone off. But this is where bias is circular—you have to make a lot of assumptions to draw these sorts of conclusions without engaging with the passenger on any level.

and this:

But I knew what was going on: The driver had concluded I was a threat, either because I was dangerous myself or because I would direct him to a bad neighborhood (or give him a low tip).

He’s hand-waving these motivations away, basically indicating that they hold no weight and no justification. And it’s certainly true that no justification would make racism good, that nothing would make these events fair to a black person simply trying to get a ride home. But in coming to the conclusion that these taxi drivers simply “learned” their racism on arriving here and assuming they pull their biases out of thin air, Glanville has avoided the hard, real conversation for the easy, comfortable and ultimately useless one.

About money: Taxi drivers don’t make a lot of it.

Those immigrant taxi drivers he’s talking about are immigrants because very few born-and-bred Americans will take a job so bad. About 30-40 percent of the Chicago taxi drivers he’s talking about make less than minimum wage, and $12-13 an hour is high wages, for a Chicago driver. [CB note: a bigger followup study counters those minimum wage figures and found that the average income of Chicago cabbies is $12/hour average … and one-sixth of them actually lose money every day.]

What hard data there is shows that black people tip less in almost all situations, even when adjusting for income. See this 2006 study from the Cornell University School of Hotel Administration. The same paper notes that in what taxi-specific data there is, black people tip less and are four times more likely not to tip at all. Taking fares from black people means, for taxi drivers, significantly less money. There’s no statistically plausible way it couldn’t.

That Cornell study isn’t the most relevant for this topic, since it addresses restaurant tipping, but this one is: a 2005 study of 1,000 cab rides in New Haven, Connecticut, published in the Yale Law Journal. (Update: The reader points out that “the Cornell study had that taxi data in it, but it was buried pretty far down in the study.”) One of the study’s authors from Yale, Ian Ayres, summarized the findings on the Freakonomics blog:

1. African-American cab drivers, on average, were tipped approximately one-third less than white cab drivers.

2. African-American and Hispanic passengers tipped approximately one-half the amount white passengers tipped.

African-American passengers also seemed to participate in the racial discrimination against African-American drivers. While African-American passengers generally tipped less, on average they also tipped black drivers approximately one-third less than they tipped white drivers.

Passenger discrimination against African-American drivers was not subtle: African-American drivers were 80 percent more likely to be stiffed than white drivers (28.3 percent vs. 15.7 percent).

Back to our reader, who ventures into fraught territory:

About danger: Did you know taxi driving is the occupation with the highest on-the-job murder rates in the U.S.? It’s over twice as bad as the next closest job: being a cop.

This is a map of the racial breakdown of Chicago. This is a map of crime density in Chicago. They are nearly identical. The black areas are the high-crime areas, more or less, and also the low-income areas. Taking more fares from black passengers means going to more dangerous places. There’s no statistically plausible way it couldn’t. It also very likely means less return fares, which means less money for increased danger.

We can and must talk about how there’s a lot of history behind why those maps look like that. We can and must talk about policies that feed that problem and prolong it. We have to talk about why a lot of black men and women are denied jobs and opportunity. We even have to go all the way down the line and talk about people who are racist but don’t have a lot of control or money—taxi drivers, for instance.

What we can’t do, though, is ignore what having that conversation means. The studies that exist suggest that to a taxi driver, that tip money and those return trips mean the difference between not making the rent and making a just-barely-living wage. The murder statistics suggest that a taxi driver is more than anyone else in America justified in worrying about driving into high-crime areas.

Does every black person live in a dangerous neighborhood and refuse to tip? No. But statistically, the likelihood is increased.

Glanville calls these men racist, and he’s right. Nothing can justify that and make racism positive. But every relevant piece of data we have shows that that racism might mean the difference between barely scraping by and getting evicted, of being able to do a dangerous job in slightly less dangerous places or not. There’s a difference between a rich man in a safe place being racist and what Glanville is talking about.

What do you think?