A Conversation about the Impact of Social Networks on Economic Outcomes


Even though we think of ourselves as independent actors, social networks shape our opportunities, power, opinions, and behaviors. Networks are helpful in understanding phenomena from persistent wage inequality to polarized beliefs, and from financial crises to patterns of war. Professor Matthew Jackson of Stanford University recently discussed his research on the impact of social networks on communities and economies, as well as the good and bad of how our networks are evolving, at the October 25, 2017, Federal Bank of Atlanta's Public Affairs Forum. He also discussed his work in this Economy Matters podcast episode with Raphael Bostic, president and CEO of the Atlanta Fed.

Raphael Bostic: Hello everyone, and welcome to another Federal Reserve Bank of Atlanta's Economy Matters podcast. I'm Raphael Bostic, President and CEO of the Federal Reserve Bank of Atlanta, and it's distinctly my pleasure to welcome you to another edition. Today we're going to talk about networks, and as you probably are aware you have networks—either professional or personal—and they are quite important to your lives. They're important to how we work. And we have a number of scholars who have started to look at the issue of networks, and I'm very pleased to have one of the leading ones with us today. We have Matthew JacksonOff-site link of Stanford University. He's a professor who works on this. Welcome, Matthew.

Matthew Jackson: Thank you, Raphael. It's a pleasure to be here.

portrait of Matthew Jackson

Matthew Jackson

Bostic: And it's really good to have you here. Your interests are in game theory and microeconomic theory—a lot of super technical stuff about how it all fits together, and how these structures operate. But there are also a bunch of really interesting angles and perspectives that are really relevant for how we live on a day-to-day basis. I'm really grateful that you're here so we can try to unpack a number of these things. As part of your work, you've recently finished drafting a book. It's called The Human NetworkOff-site link. I actually read the textbook, the more technical one with all the math. But this Human Network book is really...tell us a little bit about what the plan was with that.

Jackson: The idea is that, as you said, a lot of our behavior is impacted by networks. Our opinions, our opportunities, our outcomes really are influenced by people around us, and so understanding what that social structure looks like and how it impacts our behaviors is really important. In the book I try to look through what are the real drivers of behaviors in terms of what products we buy, whether or not we educate ourselves, which jobs we get, whether countries go to war with each other, how polarized is politics—these are questions where social structure is really important, and we're trying to understand what we know about that and what we can say about how things are changing.

Bostic: What can we say is one big discovery that you've had as part of your research, in terms of how networks function and operate?

Photo of Raphael W. Bostic

Raphael Bostic

Jackson: From an economist's perspective, one of the most important things is what are known as externalities—that what we do impacts other people. So whether or not I choose to become informed about something…you know, do I learn about vaccinations? Do I do my homework on that? That ends up impacting what my friends end up knowing, and what they believe. So networks are just littered with externalities. Pretty much everything we do ends up impacting other people through what we tell them, what we learn from them, and so fundamentally there is a disconnect there and things can get very inefficient. It might be that I end up not spending as much time learning about vaccinations as I should, and then my friends don't know as much as they should. And spending time, investing—that produces a social good which percolates through a network. And understanding exactly how that works is important.

Bostic: It's very interesting how we've started to see the—you were talking about vaccinations—how we've started to see the return of a number of infectious diseases that used to be extremely rare. It really does speak to the relevance of what you're talking about, in real time. So in terms of big discoveries, that's one thing. Has anything really surprised you in terms of the research, or the analysis, or how the field has evolved?

Jackson: I guess one of the things that's—maybe it's not as surprising when you look at it from a personal perspective, but it's pretty striking when you look at it from a scientific perspective—is how segregated our networks are. When you actually draw out networks, you can begin to see communities. You can see people who connect with each other more frequently, and others where there are separations. And those divides in networks are very prevalent. So no matter where you're looking at networks—if you're looking in high schools, you're looking in grade schools, you're looking among professional societies, whether you're looking at collaborations between different people—you see these kinds of divides emerge. And they have pretty profound implications, so it's difficult to get information to spread across these kinds of things. You can have different norms emerge and different cultures emerge across these divides, and so that's one of the things that I think I hadn't realized how prevalent it really was and how striking it is when you look into it.

Bostic: Here at the Federal Reserve Bank of Atlanta, we're doing a lot to try to think about wage inequality, differential access to jobs and capital, and all those sorts of things. And what you're talking about with this issue of segregated networks really speaks perhaps as an underlying driver of why we see a lack of social mobility.

We know that Atlanta—this metro area—compared to others in the country, is almost dead last in terms of the ability of people from lower income levels to graduate up to higher income levels. And so maybe there's something that you can tell us to help us figure out, how do we structure a response that really gives people a full opportunity to have access to participate in the American economy? Are there things that you've seen that we can use to really help break some of this separation that exists in our networks?

Jackson: Sure. I think you hit the nail on the head in terms of, if you really want to understand inequality, immobility is a big part of the answer. And immobility comes because people are born into certain circumstances, and those circumstances put them in a network of information and peers that makes it difficult for them to learn what they need to learn and to access the jobs they need to get, and it feeds back on itself. I think there are two aspects to it. One is the feedback effect, so that if my friends aren't educating themselves, it's not in my interest to do it for me, and understanding that feedback is very important in overcoming...

Bostic: It kind of reinforces itself.

Jackson: Exactly. And that means that if you want a program that's going to lift people up, thinking on an individual level doesn't work. You have to think on group levels, and you have to move multiple people at the same time. And that's a better way of getting them to move up. The other thing is that a lot of it is basic information. It's remarkable how ignorant people can be about what it takes to move up, in terms of getting better jobs, or how valuable is an education. If you're looking at somebody who's a teenager, they're not thinking deeply about the future of their life two or three decades down the road. And so it has to be that there's people around them helping push them to do the things they need to do to educate themselves, and to help give them opportunities.

Bostic: Listening to you reminds me of my own life, actually. I grew up in a suburban community, and I was the first one who really had an opportunity to go to some of the best universities, from my family. We didn't know the importance of the PSAT in terms of getting access to scholarships and stuff, and it was a neighbor who had two older children who had gone through the process, that told us, “You'd better get him to show up for this test, because it's really important.” That was actually super critical, and it's all about this network thing.

Jackson: Exactly.

Bostic: My family, we were one of the few African-Americans in our community, and we had the good fortune to have befriended others from different backgrounds and different experiences that we could leverage. So it's actually very interesting on that.

I wanted to actually turn a little bit, and ask your perspective on networks today. So, networks have been around for a long time, right? This is something that exists for people in general, in fact. And in your book, the textbook version, you started by talking about the Medici family in Florence and how they orchestrated the creation of a network that put them central to everything that happened in Florence. Today, do you feel like networks are more important, or have things changed? I'm thinking about the rise of technology, which allows us to have much more distributed networks that engage like the…do you have thoughts on that?

Jackson: Yes, definitely, and actually a lot of what I'll talk about this evening in the forum relates to trends. I would put it as two basic trends that are going on that are technologically driven in networks. One is that we're seeing denser networks—we're able to connect with more people and to maintain relationships at greater distances. When you think about something like Facebook, you can keep in touch on a daily basis with friends who are living thousands of miles away, and family. And just the extent of our reach is much higher, and when you look at travel, people are moving around the globe more. That's shrinking our world and giving us a denser, richer network in some ways.

But at the same time, I think there's a second aspect which is more subtle and more fascinating, which is that the search technology that's available today makes it easier for us to find individuals that we wouldn't have found otherwise. And that actually leads us to sometimes connect with people who are more similar to ourselves, and it can create echo chambers, it can lead to people forming tight groups, sometimes at great distances that have very like-minded composition. And I think that increased segregation in some ways is something that's coming along with the technology, which isn't as obvious but is equally consequential.

Bostic: That's very interesting, because it speaks to the rise of Breitbart in news bays, or Black Lives Matter being able to perpetuate as far as possible—or even the white supremacist movement, or sometimes even a book club, that can persist, and it can have profound impacts on our social discourse and how we talk to each other. It occurs to me that there is potentially an intersection between fake news and the ability to propagate it quickly, and by tapping into particular networks as opposed to others. And so understanding how all this works seems quite central to understanding how we are or aren't getting along today.

Jackson: Right, exactly. And I think the technology has accelerated a lot of this. It's possible now to create a website and to make it look very real and to post whatever one wants on it, and to form groups around very narrow subject matter and very specific things, and attract interest in that. It becomes more and more difficult as an individual to tell what's fact from fiction when you're being inundated from all different directions, and we don't necessarily have personal experiences with a lot of the things we're trying to judge and make sense of.

So I think along with this sort of smaller world comes a lot of noise, and an increase in noise and an increase in special interests and very specific groups forming. And it's not obvious how that all plays out.

Bostic: This is an economics podcast, so I want to make sure we talk a little bit about economics. We've talked about applying this network theory approach to a lot of different aspects of the field—microeconomics, labor markets, political economy, urban growth, and how cities persist and grow. Do you feel like there's one part of the field where this is particularly appropriate to apply?

Jackson: I think understanding development is one area where networks—social networks in particular—have become very popular. And the idea there is that we're looking at trying to understand how people become informed about the many things around them and how much they're influenced by their neighbors. So, do they use the right fertilizers? Do they take up loans when they should take up loans? Are they educating their kids? Are they investing in better health?

So there's a whole series of things which are heavily dependent on people around them. And I think development is probably the first area for this to happen, just because it's an easy area to study this, because you're looking at communities that are very well-defined and very insular, and it's easy to see what's happening. So if you go into—I've been working in India, as I talked about earlier today—you go into a small village there, it's pretty well-defined. But the same thing is happening in the developed world. If you go into the suburbs of Atlanta, or the inner city in Atlanta, or in San Francisco and so forth, there are still communities of people that are all dependent on each other for these kinds of information. And understanding exactly how that affects people's decisions and investments is one area that's really been where networks have been growing quite dramatically, in terms of the economic research.

Bostic: I was talking to someone about tonight—I was really excited about having you here and learning more about networks—and they asked a question about, are social networks the same as economic networks? Is there a distinction that we should have in our head as we think about this? How do you think about that?

Jackson: There are very big differences between some types of networks. When we think about some economic networks, we could be thinking about the financial network, for instance. So we're thinking about which bank has which counterparties, and those counterparties have investments that affect the bank's indirect exposure to risk. That's a very economic network, and it's one that still allows us to take basic understandings of things like contagion of disease, which is a very social and more human network, and apply those same tools and understand how a shock to one bank can propagate to another.

So to some extent, the applications can be different in terms of the precise actors that are involved and what kinds of relationships they're holding. But some of the basic principles and some of the underlying theory and models are quite portable, and so some of the insights that we get about who's central, who's influential, who do we have to worry about—those concepts can be the same even though the specific applications are different.

Bostic: It occurs to me that this is something that could be really relevant for us as we think about contagion in financial markets and runs on banks and all those sorts of things, to the extent to which networks either amplify or dampen an impulse towards a particular type of behavior outcome. We should definitely be thinking more about that.

Next, I wanted to turn—I'm a psychologist by training, I did psych before I did econ—and so we talked social psych when I was a freshman in college, and all the ways that people interact and how the fact that we're interacting affects the outcome and how the dynamic plays out, and I feel like these networks are very much like that. So economists are starting to really appreciate much more the soft parts of economics—how people interact, emotions, connectivity, and all that kind of stuff. But it is something that has had a history in other fields. So are you working with sociologists and psychologists and maybe political scientists, in trying to think about ways that we might develop our models and our theories to be more practical?

Jackson: Yes, definitely. Actually, networks is very exciting because it's one of the most interdisciplinary areas of research in terms of it having applications in lots of different fields. And in particular, I've been working recently with some psychologists—for instance, Sylvia Morelli and Jamil Zaki—and we looked at what kinds of personalities fill certain roles in a network. If you look at, for instance, a person who's very, very central and influential in a network of support and intimate trust, those tend to be empathetic people. If you look at people who are central in a much more open social network, where people go to parties and hang out together—for instance, in a university—then those people tend to be more fun-loving and happy people, as opposed to empathetic. So you can see different personality traits filling different roles in networks.

Bostic: I think some of the empathetic people might say they're happy, too. [laughter]

Jackson: They're not mutually exclusive.

Bostic: Exactly. That's very interesting. You know, being part of a large organization here, I do think a lot about what kinds of networks we're fostering institutionally, to allow us to be a more effective organization. Have you done much work in terms of organizational structure and how catalyzing or shepherding networks can impact organizational performance?

Jackson: It's actually something that…you could think of social engineering, to some extent, and certainly within a company or an organization it becomes very important who's communicating with whom. A lot of companies try to structure their teams and their internal workings in special ways by having certain powers and jurisdictions given to different entities within the company. But that doesn't necessarily correspond to what ends up happening internally—people tend to go to people that they think can get things done.

So the internal workings can be very different from the external, or from the designed workings—that's something that is being studied mostly in management science and in other areas. It's not something that I've looked at personally much, but understanding the dynamics of networks means that we can figure out how teams work, and try to do better in getting more productive and more creative groups together.

Bostic: So as a last question, I'm curious as to how did you come to be a network expert? What brought you to see networks as an interesting thing worth studying? Tell us a little bit about that path.

Jackson: Sure. My original training was in more microeconomic theory and game theory. And I was having a conversation one day at lunch with a friend of mine, Asher Wolinsky, and we started talking about power. And we thought about what, really, is power? What does it mean? Where does it come from? And it became clear that power comes from being able to mobilize other people and from having specific connections, and we became very interested in who becomes powerful, and how does that work. And then we started reading the sociology literature, which is quite extensive on the subject. And our perspective was a bit different as economists, because we think of a lot of relationships as ones that we choose. So we started thinking about it from a more strategic perspective in terms of, for instance—as you mentioned earlier—Cosimo de' Medici: how were the relationships engineered? Who did he marry his daughters to? How did that work, and who becomes powerful and how?

And so then it just sort of opened a door, to me. And I realized that economics had been very anonymous before, and a lot of the models are ones where you're not talking about the specific structure of who interacts with whom, you're just—it's a market, and markets have prices, and…

Bostic: And you have representative agents, right?

Jackson: Exactly, exactly. And so the idea here was to take that structure much more seriously, and then once you start reading and learning more about this then it became pretty obvious that this was “first order” in terms of having impacts in our lives.

Bostic: It got under your skin and into your blood [laughter], and we are the beneficiaries of it. I've been talking with Matthew Jackson. It's been a pleasure, I've really enjoyed the conversation, learned some things. He's a professor at Stanford University. He's an expert in networks. And if you are so moved, I would encourage you to go find his book—actually…it's not done yet, is it?

Jackson: It'll be out soon—hopefully early 2018.

Bostic: Well, that's good. Keep an eye out for it. It's called The Human Network, and it can give you a lot of insights into how networks work. Matthew, thank you for joining us today.

Jackson: Thank you, Raphael—my pleasure.

Bostic: And that brings us to the end of another Economy Matters podcast episode. I want to thank you for listening once again, and ask that you please visit the Atlanta Fed's web page—it's frbatlanta.org. On the web page you can find economic and banking information that might be useful for you, as well as materials on related topics that we are working on here at the Bank. So thanks again, and I look forward to having you join us for the next edition.