Social and economic networks

<p>Romesh Vaitilingam interviews Matthew Jackson for Vox<br />
<br />
March 2009<br />
<br />
Transcription of an VoxEU audio interview [http://www.voxeu.org/index.php?q=node/3250]</p>
<p>[music]</p>
<p><strong>Romesh Vaitilingam</strong>: Welcome to &quot;Vox Talks, &quot; a series of audio interviews with leading economists from around the world. My name is Romesh Vaitilingam, and todays interview is with Professor Matthew Jackson of Stanford University.</p>
<p>Matt and I met at the American Economics Association's annual meetings in San Francisco in January 2009, where we talked about his new book, &quot;Social and Economic Networks&quot;, published by Princeton University Press. We began by talking about the most modern form of social networking, the website Facebook.</p>
<p><strong>Matthew Jackson</strong>: One thing that comes to mind is that I had a conversation recently with somebody in the admissions office at Stanford. One thing that Facebook is doing is essentially changing the way that people interact, especially teens.</p>
<p>It used to be that Stanford organized a freshman orientation week where they would have all the students come in. It was a chance for the students to meet each other, because they were coming to a new place, they didn't know anybody, they wanted to make sure that the students would have a chance to make friends and feel a little more comfortable before school started.</p>
<p>This person indicated that that's become somewhat obsolete because of Facebook. Now, when the students come in, they've been able to make contact with other students who are prospective Stanford students by getting a list of people who've been admitted and then checking out their Facebook pages, finding other students who have similar interests to their own, and actually contacting those students ahead of time and making appointments to meet.</p>
<p>If one person likes karate or something, they can meet somebody else who does karate, or rock climbing or whatever it is, they have similar interests. So, it's facilitated people's meeting quite a bit.</p>
<p>There is a whole series of things that it does that has basically changed the way in which people can interact and the way they can meet each other, and that has impacts for the way that social networks evolve.</p>
<p><strong>Romesh</strong>: Are these outcomes all positive, do you think? We tend to look at people playing on the Internet.</p>
<p><strong>Matthew</strong>: Yeah.</p>
<p><strong>Romesh</strong>: We think how much time is being wasted. Are there some downsides?</p>
<p><strong>Matthew</strong>: Actually, one thing that one can be concerned about with something like Facebook is what one terms in social networks literature &quot;homophily.&quot; Homophily refers to the fact that people tend to associate with other people who are very similar to themselves.</p>
<p>I've been doing some research with a student at Stanford, Ben Gollug. We're looking at the impact of homophily. One thing that Facebook does is it allows you to meet people who are more similar to yourselves more easily. So, one concern that one might have is the fact that you can imagine that a network might begin to fragment, then.</p>
<p>It could be that people, as they more easily find others who are very similar, tend to only associate with people who have very similar interests. Then, the network begins to segment itself. You have less contact across different groups than you would if meetings were more by chance and friendships were formed serendipitously rather than with this advantage of being able to locate people with very similar interests.</p>
<p>Depending on what you're looking at, it has its upsides and downsides. You're able to find somebody who has very similar interests. That's nice, on the one hand. You can associate with them, but it also slows dissemination of certain kinds of information. It fragments the networks, so, people might not have access that they might have had otherwise, if they were forced to interact more with people who have more diverse interests. It's not clear what the ultimate outcome of this is.</p>
<p>On the one hand, it also facilitates... When you think about more generally the electronic, computer age, the way that we communicate and so forth, it allows you to maintain more relationships than you had before. It's easy to correspond with co authors around the world and do research that way. Business can be done with more associates. The ease of communication, in some sense, facilitates more relationships.</p>
<p>But, at the same time, it also facilitates fragmentation. There are dual, competing I guess you might say influences that are going on at the same time.</p>
<p><strong>Romesh</strong>: You have these other issues of people coming together for positive outcomes, but also some dangerous stratification. Really, are these functions of this technology? Or, are there broader issues that go on around these software issues and computer networks?</p>
<p><strong>Matthew</strong>: Yeah, I think these are more generally broad issues. They are important to understand socially in trying to understand why people interact the way they do, how social networks form and the technology. I think part of the reason that networks have become such a hot topic across a wide variety of disciplines right now is that it's becoming increasingly clear that a lot of human interaction occurs through networked communications systems and other kinds of systems.</p>
<p>That has just made it more evident than it was before. But, these have always been issues. I think that right now what we're seeing is a transformation of society in the way that we interact with each other.</p>
<p>When I think about the way that my daughters communicate with their friends, they're texting. They're in constant contact with each other, no matter where they are in the world. If they travel somewhere, or if they go on a trip with me, they can be in contact with their friends thousands of miles away, which was never something that was true when I was growing up.</p>
<p>I think that the rapid change of communication technology, together with the more evident networking aspects of things like Facebook, LinkedIn, and other computer based network sites have just made it really apparent. We're also going through a time of rapid change.</p>
<p><strong>Romesh</strong>: What does economics bring to understanding these social networks, their positive sides and their negative sides?</p>
<p><strong>Matthew</strong>: I guess there are sort of two aspects to it. One would be the toolbox that economics brings to the table, in terms of thinking about people's motivations and their intents, and modeling exactly how a network might form. What are people doing when they're forming their friendships? Why are they forming certain friendships? Why are they forming certain kinds of business associations? Are they doing it in an efficient manner?</p>
<p>Actually, a fascinating set of questions is arising right now in terms of the recent financial credit crunch. How does a network of business transactions that are going on in terms of lending back and forth between different banks how does that structure contribute to the acceleration of the credit crunch?</p>
<p>That points to a second side of issues which is there are a lot of applications that economics is interested in that have to do with networked interactions.</p>
<p>So on the one side, it's a toolbox of a particular way of modeling and analyzing things which is well suited to understanding certain kinds of questions about how certain works might come into being and evolve and the other being as a whole set of applications which are fairly important. Most markets are networked. There's very few markets.</p>
<p>We are standing here at the meetings where all kinds of people are interviewing, and a lot of the interviews happen because their advisor knew somebody at another school and said look you should be paying attention to this student. There's enormous importance of social contacts in economic settings. I think it's a two way street in terms of the interaction of networks and economics.</p>
<p><strong>Romesh</strong>: How do you go about collecting data on these kinds of issues and how these networks fit together? I mean Facebook, I guess you can join Facebook yourself and try to monitor there? What about other areas because some are much more informal and not knowing about technology and there's no public trace, perhaps connections within banks.</p>
<p><strong>Matthew</strong>: So actually let me give you an example. I am working on a project right now with Abhijit Banerjee at MIT and Esther Duflo, and what we are doing is mapping out social networks in rural India in advance of a new micro lending setting. So there is micro finance coming in. They are going to be lending in these villages and we wanted to understand how the lending patterns depended on the social networks that are in place in terms of who ends up borrowing, who they form groups with, are some people favored in terms of their access to credit compared to other people, and how does that depend on their social standing?</p>
<p>So what we have done we have gone into 77 different villages outside of Bangalore, the area in Southern South Central India and literally interviewed more or less somebody from almost every household. From these interviews we've mapped out who borrows from whom, who talks to whom, what are the religious interactions between different people, wwhat are the loans that have gone on, what are the relationships in terms of family relationships, marriage relationships and so forth. We have a whole set of different social network data and we put that together with financial data. That's a very labor intensive way of doing things. We are literally going in, not us but there is a group in India which is helping interview these people, and that's one way you do it.</p>
<p>The other way as you mentioned is something like Facebook where you got this information digitized and available. There, there's a big problem of how to make sense of the millions of bits of data that you have and making it useful. So on the one hand you've got something which is an old fashioned way of doing it where you are literally interviewing. You get exactly the information you want but it can be very expensive to gather. The other you can get millions of bits of data but the challenge there is sorting through it and making sense of it. And then how do we interpret the fact that somebody's Facebook page has a link to somebody else's Facebook page. That's a little more difficult to interpret than the fact that somebody loaned somebody else some money. So one we know something of what it means, the other we are not sure exactly what it means. So data collection is a very interesting and rapidly changing area.</p>
<p><strong>Romesh</strong>: Where are we in this whole research program on social networks? You say economics brings some very valuable tools to try to understand them. You have got all these interesting and innovative ways of collecting data. Where are we going in terms of the findings and perhaps the lessons we can draw from this for policy in various areas.</p>
<p><strong>Matthew</strong>: I think it's an exciting time because of the explosion in data, because of the interesting questions like the credit crunch that are arising. Right now I am writing a commentary on a paper by Fowler, Dawes, and Christakis that shows that social networks are influenced by genetics so that people's socialability and the way that they form relationships, the number of relationships that they form and so forth can be influenced by heredity. Just a lot of interesting new information that is coming to light right now which I think sets the table for a lot of questions to be answered.</p>
<p>I think part of what's happened in terms of the economic research is that economics has matured to a point right now where we know that we have basic models of markets that we understand, but we know they are missing a lot of features and the patterns of human interaction are one of the big features which have been missing. The other big feature is the psychology of human interaction.</p>
<p>I think that two big areas of growth in the next decade or two will be, one, behavioral economics which is already quite large and growing, and the other will be social economics where we are really trying to understand how the social structure affects economic interactions. I think that the again the two-way interaction between economics and social interaction should continue and be a very active area. Right now I think it's a growth area. There's a lot we don't understand and a lot of exciting questions. It appears that there is a lot of fruitful information to be gathered.</p>
<p><strong>Romesh</strong>: Final question Matt, how worried do you think we should be about this issue of stratification driven by technology and the power of networks. This seems to run very much counter to our desire to dealing with inequality in some way. If we are not worried about static inequality, we are worried about mobility and inability of future generations to get on.</p>
<p><strong>Matthew</strong>: Great. So I think the ultimate answer to that probably depends on the area of the application because are these competing factors? The ease of interaction means we can have more interaction as well as more stratified interactions. So ultimately how this plays out depends on the application in terms of whether or not we are seeing more segmentation and less communication across groups or ultimately whether there is more, and also what forms of communication are taking place and whether or not there are real incentives for overcoming some of the things that might be negative from some of this.</p>
<p>Ultimately, I think it will depend on the particular application whether we are thinking about economic interactions in terms of loans or whether we are thinking of just dissemination of information in public opinion. These can be very different in the way that they disseminate. So I think it will take time and part of this will be empirical question as sort of a horse race to some extent.</p>
<p><strong>Romesh</strong>: Matt Jackson, thank you very much.</p>
<p>[music]</p>