Annual Kellogg event a homecoming for Prof. Matthew Jackson who tells how networks influence social mobility
4/27/2007 - “They say opposites attract, but when you hear the data that’s not really the case,” said Matthew O. Jackson, during his April 25 lecture at the Kellogg School on social networks and their economic implications.
It turns out that most people prefer the comfort of familiar opinion, interests and habits, seeking the company of those more like themselves. There are potential costs and benefits related to the strategy, according to Jackson, a Stanford economist and former Kellogg professor who delivered his paper, “Social Structure, Segregation and Economic Behavior,” to a capacity audience at the 25th annual Nancy L. Schwartz Memorial Lecture.
In many cases, he said, people form social networks based on similarities, creating homogeneous groups that often have implications for how information is shared. Such “homophily” — a tendency for people to interact more frequently with others like themselves — can result in segregated networks even when there are opportunities to interact with others from diverse backgrounds. This segregation can prove broadly influential, even life-defining.
“Social structure is very important in determining economic interactions,” said Jackson, a professor in the Kellogg School’s Managerial Economics and Decision Sciences Department from 1988 until 1997. His presentation showed how homophily can influence decisions related to education, employment and overall social mobility.
In networks that display considerable segregation, the degree of trust and familiarity among groups will cause information to converge (come to a common view) more slowly than in nonsegregated groups. If one of those groups contains members more likely to value education, for example, while another consists of members less likely to do so, each group’s membership will be more likely to model their peers’ decisions. In addition, because the groups are segregated from one another, there is less chance of “cross talk” between them, making it unlikely for divergent views to co-mingle. Even if cross talk does exist, it’s likely to be limited, meaning that “segregated networks are going to move much more slowly to convergence,” according to Jackson.
Whether convergence is fast or slow is typically trivial with respect to mundane decisions (e.g. seeing a new film or buying a new product), but the implications can be crucial when considering issues like education, a key factor in social mobility. In these decisions, starting from a segregated position can mean that it takes a generation for an “update period” — a step along the way toward convergence — to occur.
“Starting conditions can be hugely influential when starting from a segregated group,” said Jackson, who shared other insights from his research.
Larger groups tend to be more biased toward homophily, and these groups tend to form more overall friendships. Minorities, on the other hand, form relatively more friendships with other minority groups relative to the total population size. The benefits of homophily, noted Jackson, include easier communication with those similar to oneself (e.g. in terms of education, social class, cultural experience, etc.). But the information shared tends to lack diversity. The advantage of cross-group interaction includes an increase in diversity, but this comes at a cost: such communication tends to be more challenging precisely because of the differences among those interacting.
Jackson cited three other key points from his homophily studies: 1. Social immobility actually increases with social influence, since children tend to make similar decisions (including with respect to education) as their parents; 2. There are persistent differences and inequality associated with segregated societies; 3. Economic inefficiencies can result from the under-investment in education of some high-ability children and the over-investment by some low-ability children.
“Beyond the issue of equity, there are pure efficiency concerns,” said Jackson, the author of several books and more than 70 scientific papers. He has garnered wide praise for his research, including his seminal work on the role of social networks in job searches, wage inequality and social mobility.
He said that segregation also holds implications for how people form opinions and whether and how certain behaviors will spread, including new product adoption.
“People are more prone to adopt software, games or phone plans as more of their friends adopt the same product,” said Jackson. “In addition to this social effect, different groups tend to have different innate propensities for adoption of a given product. The level of integration/segregation of a society then affects the overall adoption rate of a product.”
Said Ehud Kalai, the James J. O’Connor Distinguished Professor of Decision and Game Sciences at Kellogg: “Matt Jackson is the true scholar we were looking for when we established this lecture series in 1983. He is both accomplished and remains tremendously prolific.”
Established in honor of the late Nancy L. Schwartz, who joined Kellogg in 1970 and was the first woman to be appointed to an endowed chair at the school, the lecture series attracts extraordinary scholars in the field of economics. Among its previous speakers, have been Nobel Prize winners Robert C. Merton, Daniel Kahneman, Joseph E. Stiglitz, Kenneth J. Arrow, and Robert J. Aumann. In all, nine Nobel laureates have delivered papers at the lecture, most of them having earned the honor after speaking at Kellogg.
“This shows we pick serious people doing good work who later end up winning the prize,” said Kalai.
Kellogg Professor Brian Rogers has collaborated recently with Jackson on social networks research and the two worked together at the California Institute of Technology, where Rogers earned his social sciences doctorate before arriving at Kellogg in 2006.
“Working with Matt is an enviable experience for many reasons, not the least of which is that one begins to see how his mind analyzes problems,” said Rogers. “He is perhaps the most lucid, insightful and diligent economist I have interacted with. He has a natural instinct to find important questions to work on, and demonstrates a keen ability to develop strong and elegant answers to those questions.”
For additional information on the Nancy Schwartz Memorial Lecture, visit kellogg.northwestern.edu/meds/Schwartz.