ENLARGE

Photo: McKay Savage

Members of a rural Indian self-help group can access microfinance loans, hold their own savings and start businesses. According to the photographer, “Their newfound confidence and authority shows clearly in their faces and the many questions they ask and how they now drive the change agenda in their villages.”

Spreading the word

It’s the oldest information-diffusion process in the world: word of mouth. In a major new study of that process, researchers tracked communication through social networks as a microfinance company began marketing loans in 43 villages in southern India.

FYI, these are old-fashioned, voice-actuated social networks.

“I have a basic interest in understanding how social networks influence human behavior,” says study co-author Matthew Jackson, a professor of economics at Stanford University. Advance knowledge that the Bharatha Swamukti Samsthe bank would enter these communities “represented a unique opportunity” to gather data on social networks, and then test how communication in the networks affected the decision to borrow from the bank.

Small “microfinance” loans are made to members of groups of borrowers, which guarantee each individual loan. Loans are typically used to start a business by buying a cow, say, or a sewing machine. The loans are usually made to women; in this case, women aged 18 to 57.

Before the bank entered the villages, a small army of research assistants asked about demographics, income and social relationships:

“Where do you borrow money?”

“Who do you borrow kerosene from?”

ENLARGE

Photo: Peter Haden

A whiteboard lists statistics for a branch of the Mahasemam Trust, a microfinance institution in Chennai, India.

“What temple, church or mosque do you attend?”

After the bank came to town, the researchers returned to identify the new microfinance borrowers.

It’s who you know

As expected, the new borrowers, who were presumably enthusiastic about the bank, were more likely to pass along information about the bank than people who knew about the bank but did not take out a loan.

Still, those non-borrowers accounted for about one-third of overall information transmission. Jackson found that surprising, since he expected the borrowers to be even more dominant in spreading the word about the bank and the new loans.

The case may be unusual, as Jackson acknowledges, since microfinance has received so much positive coverage in the Indian media. “The main thing they had to do was announce that it was available in the village. If this was another product that you had never heard about, maybe you would listen more to the advice of a friend.”

Who’s got clout?

The Why Files. Data from Arbitron Inc. and Edison Research

People aged 18 to 24 are the most likely to have a profile on a social network. On average, they also sport the highest average number of Facebook friends: 510.

If you want to spread the word about something, should you try to reach the person with more friends, or the person with friends who are more important? Here, quality beat quantity. “On a superficial level, it’s obvious that if you want to reach more people by word of mouth, you want to hit someone who is well connected, but it turned out that having more connections isn’t a good predictor of whether you spread information well in the village,” Jackson says. “You need to be well connected to key people in the village, and they need to have good connections to other key people.”

As Jackson talked about “friends,” we had a FB moment: Could the results from rural villages in Southwest India pertain to people who have hundreds of “friends” on the king of the social networks? Can we extrapolate from villagers to screen slaves?

To some extent, Jackson thinks we can. “There is a substantial amount of word-of-mouth in any diffusion process in the developed and developing world.” He hopes a new definition of centrality proposed in the new study, which accounts for quality and quantity of friends, will be useful in future work on networks.

Photo: Gates Foundation

A child receives oral polio vaccine from a house-to-house vaccination team in a village in Bihar, India, in 2010. Public-health campaigns often need to spread the word in places without much electronic communication.

Sifting the network

But who, we wondered, would want to spend so much time and money identifying the biggest grapes on the grapevine? Aside from researchers, nobody would, as Jackson recognizes. “If you are going to bang on everybody’s door, and interview them, you might as well tell them about your product.”

Instead, researchers are using their findings to streamline the process of finding key network nodes. “It looks like people have an ability to tell you who are the right people,” Jackson says. Ideally, it should be possible to determine the minimum number of people needed to find the key people.

We always knew that opinion leaders affect attitudes about new ideas and products, so what is new? First, spreading basic information about the presence of a new financial option seemed to be more important than opinions conveyed along with the information.

Second, the really influential person “is not necessarily the person with the most connections, or who is most central in the network,” Jackson says. “It’s somebody who is well connected in the sense that they are well connected to people who are well connected.”