Tracing ethnic ties in Africa’s largest slum

On the ground and with satellite images, an MIT Sloan professor examines how Nairobi’s poorest pay rent and improve their homes

November 26, 2013

Tavneet Suri

Tavneet Suri

Nobody really knows how many people live in Kibera, Africa’s largest slum. There is an enormous discrepancy in estimates, which put the population somewhere between 170,000 and more than one million people.

Residents of the neighborhood often live without electricity or running water. The streets are wound with open sewers and other refuse. Homes are ramshackle and dilapidated. And because Kibera—located in Kenya’s capital city of Nairobi—was nationalized after Kenyan independence in 1963, the slum is government-owned.

So why is someone else collecting rent?

“Even in places where you don’t have formal institutions, informal institutions arise,” said Tavneet Suri, an associate professor in applied economics at MIT Sloan. “People raise a structure and say ‘I own this structure and therefore I have rights and I have enough rights that I can rent it.’ So these informal institutions arise and they have big consequences.”

   The Kibera slum in Nairobi, KenyaThe Kibera slum in Nairobi, Kenya

Throughout 2012, Suri surveyed residents, chiefs, and elders across Kibera’s roughly 3.75 square miles to determine how rents are set and how home improvements are made. The results reveal a socioeconomic system that finds favoritism along ethnic lines and sees roles for everyone from government-appointed administrators to street gangs.

Results of the survey are presented in “There Is No Free House: Ethnic Patronage and Property Rights in a Kenyan Slum [PDF],” a working paper by Suri, MIT Sloan Professor Thomas Stoker, and Benjamin Marx, a PhD student in the MIT Department of Economics.

The researchers found that when Kibera landlords share a tribe with the area chief, residents see higher rents by a measure of 6 to 11 percent. And when residents themselves share a tribe with their chief, rents will be lower by the same percentage.

They also found that areas of the slum with a high gang presence saw less egregious increases in rent. This may be because gang members, although often violent criminals, act as proxies for residents in disputes.

“It seems they protect people’s rights against landlords in some places,” Suri said. “They’re able to enforce some better behavior.”

Chiefs in Nairobi are government-appointed and rotate around the city, rarely returning to manage the same area twice. There are four chief positions in Kibera. On average, a chief oversees one part of Nairobi for 23 months before moving to the next, the researchers found. Their post may take them from the slums to a wealthy section of the city overnight. Although they have some governance duties, the chiefs in Kibera spend most of their time mediating disputes among parties, including landlord and tenants.

Landlords, meanwhile, are sometimes long-term residents of Kibera who do not have official property rights to the land there, but have been allowed by the government to build, buy, and rent properties. Suri’s survey data found that today about 55 percent of landlords do not live in Kibera themselves.

The researchers also found that residents are more likely to improve on their homes when they share a tribe with the area chief. Investments in home improvement often require permission from an elder or chief, Suri said. And residents with larger savings, roots in the neighborhood, and the ability to pay the increased rent associated with improvements are more likely to improve on their home, the researchers write in their article.

Suri said this type of research can help policymakers better understand Kibera and other similar neighborhoods in Africa and around the world.

With barriers on site, a new approach to research

“Every week in the field we had an incident,” Suri said of the nearly yearlong project. From stolen phones and other electronic devices to a research assistant hit by a bus (but not seriously injured) on a narrow slum entrance road, work in Kibera was not easy.

But Suri, who is Kenyan, has been managing research teams in Africa since 2007 and has been studying the continent for a decade. She works with a Kenyan research manager, Suleiman Asman, who has been overseeing all her work there for the last several years. For each survey, Asman supervises a project manager who in turn oversees a group of 20 to 25 people who conduct surveys and collect data. Back in Cambridge, Suri’s team is rounded out by staff at the Abdul Latif Jameel Poverty Action Lab, known as J-PAL, where Suri is the scientific director for Africa.

“I think what’s happened over time in the slum that’s amazing is that our survey staff became known by the elders,” Suri said. “So over time it’s gotten much safer for them because they’re known, the elders know them, and they’re in the community.”

Still, even with significant access to Kibera, its leaders, and its residents, Suri’s team confronts the high cost of research there, and seeks new approaches to their work that can save time and money. For the study on rents and home improvements, the team purchased high-resolution daytime satellite images and cataloged the luminosity of slum roofs, 96 percent of which are made of corrugated iron. The images helped the group measure home improvements across the densely-populated slum.

“The thing about slums everywhere in the world is the roofs are just made of flat metal,” Suri said. ‘They buy cheap metal sheeting and put it up. And this is true all over Africa, Asia, and Latin America. And the sheeting needs to be replaced, because how long is metal sheeting going to last? So you can look at the replacements over time, because you can measure how bright the sheeting is.”

Suri, whose work in sub-Saharan Africa covers issues in agriculture, household finance, and political participation, among other topics, turns next to a voter mobilization study—work done with Benjamin Marx and Vincent Pons, both PhD students in the MIT Department of Economics—in which her team sent more than a million text messages to voters in the March presidential election in Kenya. Results will be forthcoming, but like her other work, Suri hopes her findings will contribute to a body of research that can help direct policy in Africa.

“Ultimately, our goal is to improve welfare in some way in developing countries,” Suri said. “The idea is to create good, rigorous evidence on things that have a bearing on policy.”