Training against racism not enough to address issue in US companies

On May 29, Starbucks will close 8,000 locations to administer racial bias training for 175,000 of its employees. The move is a response to national outrage over the arrests of two black patrons while they were simply waiting for a meeting to begin at a Philadelphia coffee shop.
 

But racial bias training for employees is not enough to address the epidemic of discrimination by American companies.

Over the past two years, we have investigated discrimination in customer service by conducting large-scale field experiments in the hospitality industry. We have repeatedly found that front-line workers exhibit racial bias in the quality of customer service they provide. 
 

In one experiment, we emailed approximately 6,000 hotels across the United States from 12 fictitious email accounts. We varied the names of the senders to signal different attributes, such as race and gender, to the recipients. Names carry a lot of information; for instance, from our names — Alexandra Fe  ldberg and Tami Kim — you might infer that one of us is a white female and the other is an Asian female (and you would be correct). Similarly, in our studies we used names with which we know many Americans have strong race and gender associations. 
 

The inquiry these fictitious people made was simple: They asked for local restaurant recommendations. We tracked whether hotel employees responded and also analysed the content of the emails from those who did respond.

Across the range of responses, racial discrimination was clear. Overall, hotel employees were significantly more likely to respond to inquiries from people who had typically white names than from those who had typically black and Asian names. But racial bias did not end there. Discrimination also happened in many subtle ways.
 

Hotel employees provided 20 percent more restaurant recommendations to white than to black or Asian people. Employees’ politeness also varied by race. When responding to white people, employees were more likely to address them by name and to end their emails with a complimentary close (“Best,” “Sincerely”) than they were when responding to black or Asian people. And employees were more likely to go “above and beyond” in their service: They were three times as likely to provide extra information — even when the initial inquiry was just about restaurants — to white than to black or Asian people. The manager at Starbucks called the police with the claim that the two men were “refusing to make a purchase.” The manager did not believe that they were customers.
 

Our research suggests that this belief — that a nonwhite person cannot or will not be a legitimate customer — can indeed worsen discrimination in service delivery. In a follow-up study, we emailed nearly 2,000 hotels and made a similar inquiry about local restaurants. But this time, in addition to varying the race of the enquirers, we varied whether they made clear their intention to stay at the hotel. Indeed, making customer status explicit helped reduce discrimination against minority people. These results are in line with the recent incident at Starbucks: We see that service workers do not apply the rule that every person is a potential customer deserving the same level of service.
 

Achieving consistency in customer service is no easy feat. Good customer service could mean something different to every patron and every employee. As the sociologist Philip Selznick wrote, employees are “wholes,” not just “roles.” This applies to customer service. People bring their whole selves to work — including their unique experiences, backgrounds and world views.

While well-designed racial bias training is certainly a positive step forward and an important public statement, research on the long-term effectiveness of such training is mixed at best. Instead of relying primarily on trainings to remedy bias, if they truly want to transform the way they serve customers, companies need to make structural changes. 
 

© 2018 The New York Times News Service