The researchers, economists Redzo Mujcic and Paul Frijters, performed their experiment in Brisbane in the Australian state of Queensland, a region they compared to the American South in its historical opposition to racial equality. They enlisted men and women of different races and wearing different attire to board buses and try to scan a fare card with no money on it. When the machine beeped to let the driver know that the card was empty, the experimenter told the driver that he or she was out of cash but was trying to get to some stop about a mile away. Because the trips were short and the buses run regularly, the experimenters were able to try several times an hour and gather data on 1,552 separate requests for a free trip. They made notes of the race of the driver and whether they were allowed to ride for free, and Mujcic and Frijters compiled the results, which were also noted by The New York Times yesterday..
They found that about two out of three requests were granted -- slightly more in wet weather, at night and when the bus was mostly empty, as you might expect. Yet Indians and blacks were less likely to get a free ride than whites and other Asians. This was generally true no matter what the driver's race was, although black drivers more readily allowed the passengers to remain on board and their biases were less pronounced, and Indian drivers let Indian passengers ride free as often as whites passengers, as shown in this chart from the paper.
Blacks and Indians were better off depending on how they dressed. Drivers granted only 36 percent of requests from blacks and 51 percent from Indians dressed casually, compared to 72 percent from whites. When blacks and Indians wore business attire and carried a briefcase, their rates of success jumped to 67 percent and 83 percent respectively. Wearing what looked like an army uniform had an even larger effect on how drivers perceived them. Blacks in uniform got free rides in uniform 77 percent of the time, and Indians 93 percent.
Regardless of attire, though, drivers preferred whites. They succeeded 93 percent of the time dressed for business and no less than 97 percent of the time in uniform, as shown in the chart below.
A free ride on a bus is a small thing, but it's worth considering the implications of these findings for a couple of perennial issues in the United States.
In the debate over affirmative action, the Supreme Court has insisted that colleges use racially neutral indicators such as family income in order to create a diverse student body. Advocates of diversity are split on whether this approach will work. This study, at least, suggests that even wealthy people of color are likely to confront bias, whether in college admissions or later on.
The study is also a reminder that these biases might be largely unconscious and held by generally well-intentioned people. Mujcic and Frijters surveyed drivers separately to ask whether they'd let their experimenters ride for free, showing them pictures of the group. When asked, the drivers were more likely to say they'd give black and Indians a free ride than to say they'd let whites ride free.
It's possible this discrepancy reflects simple dishonesty on the part of drivers, but it could also be that the drivers simply don't think of themselves as racist and behave differently in practice than they would when asked to think about the situation in advance. Extensive research has shown that most whites share an unconscious automatic bias against blacks, even those who believe strongly in racial equality. This psychological fact is a major problem for law enforcement agencies, which are increasingly working to train their officers to find their way around these biases to prevent hasty decisions on the beat.