Using social networks to spread change

Editor’s note: Nick Cooney is author of the new book Change of Heart: what Psychology Can Teach Us About Spreading Social Change. Below is an introduction he wrote for Socialbrite and a chapter from his book:

The non-profit world is filled with competing theories about how to create change, with many of those theories revolving around one key question: what motivates people to change their behavior? Whether you’re trying to increase donations, recruit new volunteers, or get members of the public to make a specific change in their lifestyle like eating healthy, whether you will succeed or fail depends in large part on your understanding of human psychology. As a non-profit director myself, I decided that I wanted facts and not just theories on how to more effectively influence others. I spent a year combing through the scientific record to learn proven techniques for how non-profits can more effectively persuade the public and succeed in their mission. The result: my new book, Change Of Heart: What Psychology Can Teach Us About Spreading Social Change. The following excerpt shares research on the role that social networks (and social networking sites) can play in helping non-profits succeed.

Excerpt: ‘Using social networks to spread change’

In taking a hypothetical look at the spread of vegetarianism, we made an assumption that the average person has about ten friends, each of those people has about ten friends, and so on. But in the real world, the number of friends a person has varies significantly from individual to individual. People who are particularly social might have many dozens of friends. Someone who is shy and reserved may have only one or two. Malcolm Gladwell uses the term connectors to describe people who have large numbers of friends and acquaintances. Other authors call these people hubs. Either way, these highly connected individuals sit at the center of vast social circles.

It’s because of connectors that the “Six Degrees of Kevin Bacon” game actually works. For those not familiar, this is a pop culture game where you try to connect actor Kevin Bacon to any other actor or actress based on the movies they’ve been in. For example if Kevin Bacon was in a movie with Danny DeVito, and Danny DeVito was in a movie with Jim Carrey, then Bacon would be two degrees away from Jim Carrey. In fact, Bacon can be connected to almost any other American actor or actress in only 3 to 4 links. This reason for this is connectors—the small number of actors who have been in a large number of movies (Barabasi).

Researcher Stanely Milgram wanted to find out how many links it would take to go from one randomly selected person in the United States to another. In other words, he wanted to find out how connected we are. To do so, he mailed out letters to randomly selected individuals in Nebraska and Kansas. In the letters he described the experiment and instructed the recipients that their goal was to get those same letters to a target person in Boston. The target’s name and information were given, but recipients were instructed to forward the letter to that person only if they knew him personally. Otherwise, they were instructed to forward the letter to whatever friend of theirs they thought most likely to know the target person in Boston. And so the letter was passed on from one friend to another, getting closer and closer to the target person. Want to take a guess as to how many times each letter was forwarded on before reaching its intended recipient?

On average, the letter reached the target person after being forwarded on only six times. So there are about six degrees of separation between any two randomly selected people in the United States. It is a small world after all—thanks in large part to connectors. Of the letters that reached the target person at his home, two thirds were sent to him by one man, a prominent clothing merchant. Of the letters that reached the target person at his office, half were sent by two other men. Connectors such as these individuals make the world a much smaller place (Travers and Milgram 1969).

One of the fundamental properties of the human social network is that most people have a small number of links (that is, friendships) but a small number of people have a large number of links. In fact about twenty percent of people hold about eighty percent of the links (Barabasi and Laslo), and research has found that genetics play a significant role in how connected a person is (Christakis and Fowler Connected 233). This 80/20 principle of uneven distribution holds roughly true for other issues as well, including the distribution of population among cities in the U.S., the size of businesses, the length of rivers (Rank–size distribution 2010) and the distribution of wealth (as of 2007, twenty percent of the U.S. population controlled ninety-three percent of the financial wealth) (Domhoff 2010). Networks like the human social network, in which links are distributed unevenly, are called scale–free models.

We’ve discussed how traits like obesity, recycling and caring about world poverty can spread through social networks, so that when one person changes their behavior they make others more likely to do the same. We’ve also discussed how some people—connectors—are linked to many more people than others. Putting these two facts together, we can see that connectors are much more influential than the average person. When they change a behavior, it will have much more of a ripple effect because they are within three degrees of separation of many more people than the average individual. Connectors are often the ones that help push trends forward, and they can make or break the success of a new idea.

Connectors who bridge different groups of people also play a vital role in the spread of ideas. Think of a randomly selected friend of yours. Now think of another randomly selected friend. Do those two friends know one another? Chances are they do. Most of us have social networks filled with friends who know one another and who share similar attitudes. An analysis of social networks on Facebook found that both altruistic and non–altruistic people had similar numbers of friends, but that they were embedded in separate networks of like–minded people (Christakis and Fowler Connected 300). The altruists stuck mostly with other altruists, and the non–altruists stuck mostly with other non–altruists. Tightly clustered groups are good for enforcing group behavior; if a lot of your friends are commenting to one another that a new health care bill is good, chances are you will agree with them. What tightly clustered groups are not good at is adopting new ideas. This is where connectors come in.

Though some connectors sit at the center of a dense cluster of friends, others are linked to a number of different clusters. For example a group of environmental activists in a small town might include one activist who is also a Republican committeeperson and a prominent member of a local church. That activist is particularly important because they can help transmit the behaviors of the environmental activists to other social circles via links of close friendship.

Some activists, particularly those who perceive themselves as having very progressive viewpoints, only want to collaborate with people who have virtually identical beliefs as them. Some activists will even refuse to work with anyone who doesn’t agree with them on an entire list of social issues—feminism, gay rights, animal rights, capitalism, etc. This attitude will create a tiny group with very similar views, but won’t help spread any of those views to the wider public. Instead of rejecting those who inhabit different social circles or have different viewpoints on other issues, we should embrace these connectors as particularly important for spreading our message to new audiences. A college social justice organization might be taken aback when a hard–partying fraternity brother shows up for a meeting, and be tempted to not take him seriously. But that person could help spread new behaviors to those in his own social circle, and might also provide the social justice organization access to a new pool of volunteers.

‘Activists should focus on changing the behaviors of connectors’

In trying to create social change, activists should focus on changing the behaviors of connectors. Because connectors are linked to so many people, or provide bridges between very different groups of people, getting a connector to make a change will have a much large ripple effect than getting an average person to make a change. For example, getting five connectors to spay or neuter their companion animal will lead to many more animals ultimately being spayed or neutered (as the effect filters to their many friends, to their friends’ friends, and to their friends’ friends’ friends) than getting five average people to do the same. Connectors spread new behaviors far (reaching a large number of people) and wide (reaching new social clusters), speeding up the spread of social change. But how can we activists find these powerful connectors?

The rise of online social networking sites like Facebook, Twitter and MySpace means we now have access to detailed maps of who is linked to whom and who the major connectors are—those with the largest number of friends or followers. A small portion of these people may be friend collectors who don’t personally know as many people as their profile suggests. But for the most part, those who have more online friends really are more connected to others. This doesn’t mean that someone with one thousand Facebook friends has one thousand meaningful links through which new behaviors can be spread. As was discussed earlier, changes spread only through deep social connections. But users with a large number of friends on Facebook usually have a larger number of meaningful online connections than users with a small number of friends. Research published in The Economist found that Facebook users with one hundred fifty friends maintained a relationship with nineteen to twenty-two other users on average, whereas Facebook users with five hundred friends maintained a relationship with thirty-nine to forty-seven other users (the range in each answer represents the difference between men and women, with women being more communicative) (The Economist 2009; Byron et al.). If you’re going to be promoting your cause online, finding the major connectors and focusing some of your outreach efforts on them can create a major ripple effect.

For example, imagine that a college environmental club has started a Facebook group called “We Want Ohio State to Switch to Wind Power!” In trying to get as many group members as possible (which would show student support for the initiative), naturally those in the environmental club will first approach their friends about joining the Facebook group. But the outreach needn’t end there. By browsing through friends’ profiles and the profiles of friends of friends, club members should be able to identify 20 of the most well–connected students at the school. Club members can then befriend those individuals, and after chatting them up for a few days can tell the connectors about the wind power campaign and ask them to invite their many friends to join the group.

Mapping out networks of our friends and our friends’ friends on these sites was made a little easier by the development of applications that automatically create such maps. The Twitter Mention Map Application (Mention Map) shows which other users a tweeter is talking about. The Facebook Social Graph application (Social Graph) plots a graph of all of your Facebook friends and their connections to one another. One major limitation of this and similar applications for Facebook and MySpace is that those currently in existence only allow you to look at users with one degree of separation from you. So you can find which of your friends are the most well–connected to other friends of yours, but you can’t see which of your friends are the most well–connected to the wider world. Furthermore, you can only look up your own social network map, and not anyone else’s. Perhaps more advanced applications will be developed in the future to allow mapping to at least two degrees of separation.

Despite their limitations, the graphing applications currently available can still be useful in helping increase your influence as an activist. For example, let’s say you’re holding a major conference and you want as many people from your social network as possible to come out. Naturally, you’re going to invite all of your Facebook friends to come to the event and post about it repeatedly on your wall. However after mapping which of your friends are the biggest connectors in your social circle, you can contact the five or ten most connected individuals and ask them to also post about the event to their wall. That way many of the people in your social circle will be hearing about the event not just from you, but from several other of their Facebook friends—making it more likely they’ll attend.

There are doubtlessly other ways to locate and use connectors to promote your activist cause. Think about the people you know and pick out those couple friends who seem to know everyone. Make a mental (or written) note of who these people are, and be sure to put special emphasis on talking to them when promoting an event, a behavior change, a fundraising drive or anything else.

Moving on from our immediate friends, there are other very visible connectors in society. Politicians, prominent businesspeople and civic leaders are major connectors, interacting with and befriending a wide variety of people. Whatever behavior an activist organization is promoting, it would make sense to put special emphasis on getting campaign materials in the hands of connectors like these. If an average person decides to go vegetarian, several dozen of that person’s friends will find out and maybe think about doing the same themselves. If a city council member goes vegetarian the many hundreds of people that the council member interacts with will find out about it, creating much larger ripples.

Preventing a behavior from spreading

Some activist organizations aren’t trying to spread a particular behavior, they’re trying to prevent a behavior from spreading or reduce its prevalence in society. Examples include efforts to stop the spread of AIDS, to reduce the number of smokers and to reduce the number of obese Americans. Here too, targeting the connectors in a network will lead to much greater success than targeting the general population. AIDS prevention efforts should focus first on providing condoms and doing regular HIV testing of the biggest connectors: prostitutes and others with large numbers of sexual partners. By working with the most–connected one percent of the population, a non–profit can do as much to stop the spread of AIDS as they would have by working with a much larger percentage of the general population (Barabasi 139–142).

Christakis and Fowler conducted some theoretical research on how to deal with the spread of new diseases when widespread immunization takes a lot of time and a lot of money, and when there may not be enough supply to go around. For example the 2009 swine flu pandemic went on for months before vaccines were made available to the general population. Until there was enough supply to vaccinate everyone, what should federal health agencies have done? Reserved vaccines for those most susceptible to the illness, such as young children? Reserved them for those who had money to pay for them? Or distributed them on a first–come, first–served basis until they ran out?

Christakis and Fowler looked at what would happen if instead of immunizing everyone against a disease, the government focused on immunizing those who were most connected. The pair found that immunizing the thirty percent of the population that is most connected would do just as much to prevent the spread of the disease as immunizing ninety-nine percent of the population at random (Christakis and Fowler Connected 133).

Researcher David Bahr and colleagues wondered if they could create a model to replicate the way that obesity spread through the Framingham Heart Study participants. They created a computer simulation with several hundred thousand fictitious people, inserted a small percentage of obese individuals in to the mix, and let the simulation begin. Slowly, obesity began to spread through the simulated population in a way similar to how it had happened among the Framingham participants. Now that Bahr and his team had created a good model for the spread of obesity, they turned to their main question: how could the spread of obesity be stopped, or reversed?

One solution that worked extremely well in the simulated model was to have people diet not with friends but with friends of friends. That way the social norm of dieting was spread further through the social network and its influence was wider–reaching. It also put particular social pressure on the mutual friend (the person who knows both dieters) to make a similar change. When Bahr and his colleagues plugged this technique into their virtual society, the obesity epidemic began to reverse itself. In fact they found that it could be reversed quickly with only one percent of the population initially going on a diet, as long as the dieters were placed in the right spots throughout the social network (Bahr et al. 2009). Of course, such precision is not possible in the real world. But the fact remains that focusing on friends of friends can have more of an impact at spreading change than focusing on friends, and that it doesn’t take a large percentage of people to initiate sweeping social change.

What appears to be most effective stands in contrast to the initial impulse most of us have upon changing our own behavior, which is to talk about the change with those close to us. When a person learns about the environmental impact of plastic bottles, they’ll typically want to let their co–workers, friends and family members know about what they’ve learned and encourage them to stop buying bottled water. The same is true for a person who has just donated to the victims of a natural disaster or bought a hybrid car. Yet the research indicates that instead of focusing on these people, if we really want to spread that behavior we should focus on their friends.

Online social networks make connecting with friends of friends extremely easy. For example, if you post the following update on your Facebook page or Twitter feed, a couple hundred people might see it: “Undercover investigation: male chicks are ground up alive by egg industry! Video at http://tinyurl.com/mzamns. Please don’t eat eggs!” On the other hand, if you ask ten friends to post it on their Facebook pages or Twitter feeds then several thousand people (friends of friends) will see the posting, and your ten friends will be very likely to watch the video as well. If both you and a friends’ friend stop eating eggs, not only will that behavior ripple out across a much wider set of networks, but your mutual friend becomes more likely to stop eating eggs as well.

Alternatively, activists could contact friends of friends directly via social networking websites. “Hi, my name is Nick, we’re both friends of Michael Brown. I wanted to let you know the non–profit I work for is having a walkathon to help animals next month; if you’re interested in taking part here’s the website…” Not only do you reach a wider audience, but you also make it more likely that your immediate friends will come out to join the walkathon as well. Even if you don’t use social networking sites you can put this technique to use the old–fashioned way by asking your friends to put you in touch with friends of theirs who might be interested in your cause.

We might also set up situations where we introduce to one another two friends of ours who are making a similar change but who don’t know one another. For example an environmental activist might have two acquaintances who are both in the process of trying to reduce their home energy use. By introducing these two acquaintances to one another the activist provides them additional social support, which makes them more likely to succeed and helps the ripple effect of their changes get spread over a wide social network. It’s also possible that the two budding environmentalists have other mutual friends who could become influenced to reduce energy usage as well.

In working for social change, we as activists need to recognize that connections matter when it comes to spreading new behaviors through society. By focusing outreach efforts on connectors, befriending new connectors and involving friends of friends in our efforts, we can create large ripple effects that create more change more quickly.