When Harvard Business School professor Vincent Pons went to Kenya to conduct research in advance of the 2013 national elections, he discovered surprising lessons about how the ethnic makeup of teams affects the work they do—now published in a new working paper, “Diversity and Team Performance in a Kenyan Organization.”
For years, Pons had been researching electoral politics in France. “We found that if you provide information to voters one-on-one it increased participation,” says Pons, an assistant professor of business administration in the Business, Government, and International Economy unit.
“The working atmosphere was more cohesive among co-ethnic teams”
Pons wanted to see if the same kind of door-to-door outreach could work in Kenya, which had been wracked by ethnic violence in the 2007 election. Specifically, he wanted to see if the ethnic makeup of the teams doing the knocking influenced voters’ decision to register.
Pons and co-researchers Benjamin Marx and Tavneet Suri—both of the Massachusetts Institute of Technology—focused on Kibera, the largest slum in Africa, where 30 ethnic groups live side by side, and where Nairobi’s electoral violence was concentrated.
Working with Kenya’s electoral commission, Pons and his colleagues tracked the efforts of teams of canvassers as they braved the slum’s twisting streets, knocking on doors and trying to convince occupants to register to vote.
For each of 15,000 interactions, they recorded the ethnicity of the teammates and the citizens, whether the team was able to find the household, and the length of the conversation. Surprisingly, they found that ethnicity of the teammates compared to the voters had little bearing on how many households they visited.
“We were afraid teams would spend less effort finding households of different ethnicities, and they would spend all of their time trying to register their own people in order to gain power,” says Pons. “The fact that we didn’t see that impact was quite a striking result.”
Co-ethnic teams more productive
As Pons and his co-researchers analyzed the data, however, they found a disturbing result about the ethnic makeup within teams. Those canvassing teams made up of two people of different ethnicities completed fewer visits—16.2 percent less—than two people of the same ethnicity. Even when they did complete visits, they tended to be 48 percent shorter.
“The working atmosphere was more cohesive among co-ethnic teams,” Pons says. “In multi-ethnic teams, at the end of the project, teammates were more likely to complain about their partners and to report that they performed better than the other worker.”
The researchers also found that homogenous teams were more likely to split up in an effort to cover more ground, helping them gain time in the field.
Surprisingly, however, when Pons and his colleagues examined the difference in ethnicity between team members and their superiors, they found the opposite effect. When ethnicity was different, it raised the fraction of completed visits by 14 percent and led to meetings that were 69 percent longer.
“If you have a manager that is the same ethnicity as you, the guy is going to give you some more slack and be less on your back,” Pons surmises. “He might let you finish work early, or provide less monitoring.”
The researchers looked at homogeneity along other demographic statistics, including age and gender, and found much smaller effects—perhaps due to how crucial ethnicity is in a context where people are voting on ethnic lines.
Other research has found that teams made up of people from diverse cultural viewpoints show more creativity and versatility in unfamiliar situations. “Indeed, ethnic diversity may be very useful in completing a task that requires many different competencies,” says Pons. At the same time, the findings of Pons and his fellow researchers offer a note of caution in how teams are put together.
“Ethnic homogeneity matters a lot, and it may affect productivity in different ways, whether it’s on a horizontal or vertical dimension,” he says. “If you are going to bring together heterogeneous teams, you want to make sure that it happens in a way that people collaborate, and there is a lot of conversation and opportunities for them to adjust as they go.”
An invitation for further research
The findings in this study, Pons notes, may be particularly strong because of the salience of ethnicity in a country like Kenya that has experienced significant recent racial tension—and may not necessarily be as strong in a country like the United States or France. Even so, these trends are worth keeping in mind in any context in which race is a factor.
Pons would like to conduct a similar experiment in which researchers intervene in the middle to offer space for teammates to have conversation and air differences in order to see if that could improve performance and mitigate some of the negative effects of ethnic diversity while maintaining the positive aspects.
“The study is an invitation to do future work that would try and identify effective ways to ensure that negative impact is not there.”