In 1964, The New York Times ran a front page story that has had a profound effect on the field of psychology. Thirty seven people, the story explained, witnessed the murder of Kitty Genovese, but not one of them called the police. A 38th rang only after Genovese, who was stabbed to death, had already passed away.

The tale, a depressing if flawed nod to the loneliness of urban life—the events, it turns out, didn't unfold exactly as the Times recalled—gave rise to the study of a deeply disturbing phenomenon known as the bystander effect. The crux is that groups can be painfully inefficient organisms. When the responsibility to act doesn't rest on the shoulders of any one individual, each person waits for someone else to assume the role of hero—whether out of shyness, deference, or a simple human desire to conform—ensuring that no hero emerges.

People, psychologists have observed time and again ever since, are only less likely to help someone in need when others are around. The counterintuitive behavioral kink is so prevalent that it has even been found to manifest itself in children.

In many ways, a new study, published in the American Journal of Public Health, further confirms the phenomenon. Using data from the National Emergency Management Information System (NEMSIS), which tracks the details associated with millions of emergency calls from around the United States, researchers at Cornell University found that only 1 in 39 people (or roughly 2.57 percent of people) who suffered a medical emergency in public received any help beyond a call for an ambulance. And that number is probably kinder than it seems, since it doesn't discount support by family members and friends.

But the study also reveals a nefarious side to our collective tendency toward inaction. While the phenomenon is a troubling one for society as a whole, it appears to have a disproportionate effect on some. Black patients, the NEMSIS data show, are significantly less likely than their white counterparts to receive help of any form before paramedics arrive.

"There is a huge racial disparity in terms of who gets help during public emergencies," said Erin York Cornwell, an assistant professor of sociology at Cornell University, and the study's lead author. "It's really very concerning. It's what's most troubling."

The chart below, plucked from the study, shows just how great that disparity is. The likelihood of someone receiving support from a bystander is so minuscule, the researchers had to zoom into the very bottom spectrum of probabilities (that's what second chart, magnified on the top right, is). And when they zoomed in, it was clear that black people received help far less often, even after adjusting for population (that's what the low, mid, and high density indicators mean). Overall, white people were more than twice as likely to receive help than black patients (the probabilities were 4.5 percent and 1.8 percent, respectively).

There are several reasons why people are reluctant to lend a hand in these situations. The bystander effect certainly appears to take its toll. There are few places better fit to show the phenomenon, after all, than out in the streets when unexpected emergencies bubble up, confronting crowds of complete strangers with the chance to help. It's no coincidence that the rate of support was highest in low density (i.e. less populate) areas.

There is also possibility that helping might undermine the efforts of paramedics, especially if one misdiagnoses the problem. The risk of helping, in other words, might appear to outweigh the risk of not helping.

But even still, Cornwell says she expected the rates at which people receive this kind of support to be much higher.

"In many cases, it's as simple as giving someone a blanket, a bottle of water, some juice, or a little food, " she said. "We're talking about things like holding someone's head, keeping someone still, or helping someone take medication—things that require no training at all."

As for the racial disparity, there are structural reasons—or really, shortcomings—that might help explain why it exists. The researchers adjusted for socioeconomic differences in the observed data, but say it's possible that residential segregation plays a role, because it "concentrates black patients in neighborhoods lacking services and institutions." Without them, Cornwell says, residents are much less likely to be trained in various forms of basic medical support, like CPR.

She also points to sociological research, which suggests poorer communities have weaker social bonds and less communal trust. That, she says, could discourage support and concern for others.

Socioeconomic circumstances, however, likely only tell part of the story. While the researchers don't know the race of the bystanders who did and did not help those suffering from medical emergencies, they believe a certain level of racial bias may be at play as well.

"The difference is just too significant," said Cornwell.

The data do show at least one anomaly, which could support that hypothesis. For whites, the bystander effect appears to have an enormous impact. Whites patients in the less populated areas are much more likely to receive support than those in more populated ones. For black patients, the difference isn't nearly as significant, suggesting that it could have more to do with the individual than with the environment.

Racial bias, at least the implicit variety, remains surprisingly common in the United States, even among those who believe themselves immune. There are signs that it exists in our court system, where white federal judges are roughly four times as likely as their black counterparts to dismiss race discrimination cases outright and half as likely to rule in favor of people alleging racial harassment at work. It might affect our doctors, who, according to a recent study, routinely under-treat black patients for pain. And its role in policing has been a topic of much scrutiny, underscored, at least in part, by The Washington Post's Pulitzer prize winning investigation, which showed law enforcement fatally shoot unarmed black men at disproportionately high rates.

"I think this echoes what we've seen in terms of bias against blacks in public settings," said Cornwell. "Not to push this too far, but the evidence is stacking up that in public spaces African Americans are unfairly perceived as more dangerous or threatening."