DEPARTMENT OF ECONOMICS
ISSN 1441-5429
DISCUSSION PAPER 45/16
The Economic Consequences of Social Network
Structure
Matthew O. Jackson,∗ Brian Rogers† and Yves Zenou‡§
Abstract:
We survey the literature on the economic consequences of the structure of social networks. We
develop a taxonomy of ‘macro’ and ‘micro’ characteristics of social inter-action networks and
discuss both the theoretical and empirical findings concerning the role of those characteristics
in determining learning, diffusion, decisions, and resulting behaviors. We also discuss the
challenges of accounting for the endogeneity of networks in assessing the relationship between
the patterns of interactions and behaviors.
Keywords: Social networks, social economics, homophily, diffusion, social learning,
contagion, centrality measures, endogeneity, network formation.
JEL Classification Codes: D85, C72, L14, Z13
∗
Department of Economics, Stanford University, the Santa Fe Institute, and CIFAR, e-mail:
jacksonm@stanford.edu, http://www.stanford.edu/∼jacksonm.
†Department of Economics, Washington University in St. Louis, email: brogers@wustl.edu,
http://pages.wustl.edu/brogers.
‡Department of Economics, Monash University, Stockholm University and IFN,
Email: yves.zenou@monash.edu, https://sites.google.com/site/yvesbzenou/.
§We thank the editor Steven Durlauf, two anonymous referees, and Ben Golub for very helpful comments.
Matthew Jackson gratefully acknowledges financial support from the NSF under grants SES-0961481 and SES-
1155302 and from grant FA9550-12-1-0411 from the AFOSR and DARPA, and ARO MURI award No.
W911NF-12-1-0509. Yves Zenou acknowledges financial support from the Swedish Research Council
(Vetenskapr˚adet) under grant 421–2010–1310, and from PA, and ARO MURI award No. W911NF-12-1-0509.
Yves Zenou acknowledges financial support from the Swedish Research Council (Vetenskapr˚adet) under grant
421–2010–1310, and from the French National Research Agency (ANR) under grant ANR-13-JSH1-0009-01.
© 2016 Matthew O. Jackson, Brian Rogers and Yves Zenou
All rights reserved. No part of this paper may be reproduced in any form, or stored in a retrieval system, without the prior
written permission of the author
monash.edu/ business-economics
ABN 12 377 614 012 CRICOS Provider No. 00008C
1 Introduction
Humans are inherently social beings. We rely on each other for sustenance, safety, gover-
nance, information, and companionship. Production, exchange and consumption of goods
and services largely take place in social settings where the patterns and nature of interactions
influence, and are influenced by, economic activity. This embeddedness of many economic
transactions means that abstracting from social structure comes with the risk of severely
misunderstanding behaviors and their causes.
1
In particular, designing many economic poli-
cies requires a deep understanding of social structure. Consider the following representative
examples:
• Criminality is often a social behavior and accounting for peer influences and networks
of interactions can lead to more effective policies aimed at reducing crime.
• Increasing the employment rate and wages of a disadvantaged group requires under-
standing that many jobs are obtained via social contacts and the underlying social
networks exhibit patterns that can result in persistent inequality and poverty traps.
• Improving the human capital investments of a given group must account for the fact
that one’s decisions regarding education and labor market participation are often heav-
ily influenced by decisions of family and friends, both through learning and comple-
mentarities.
• Integrating schools not just in terms of ethnic or racial composition, but in terms of
friendship formation and cross-group interactions, requires understanding when and
why students are compelled to seek friendships with others similar to themselves.
• Enhancing new technology adoption requires a proper understanding of how peoples’
opinions and beliefs are shaped by word-of-mouth communication.
• Sustaining informal risk-sharing and favor exchange depends on social norms and sanc-
tions, and social structure provides new insights into how communities overcome basic
incentive problems.
This is, of course, only a partial list of the many economic behaviors that are shaped
at a fundamental level by network patterns of interaction. For instance, beyond such “so-
cial” networks, other interactions, such as international trade and political alliances, have
inherent network structures that shape the impact of policies and help us understand con-
flict and other inefficiencies. Given the importance of social context and the emerging tools
1
See Granovetter (1985) for a seminal discussion.
2
researchers are currently developing to account for it, there has been a rapid growth of anal-
yses of economic behavior that consider social context, appearing in an array of applied and
theoretical literatures both within and outside of economics. We do not attempt to provide
a comprehensive survey of the economic literature on social networks.
2
Instead, we provide
a framework for understanding how networks of interactions shape behavior.
Most importantly, there are robust regularities in how network structure relates to be-
havior, involving the network-based notions of density and distribution of connections, seg-
regation patterns, and the positions of key nodes. Our narrative - using this framework -
pulls together major insights that have emerged from empirical and theoretical analyses of
how social structure relates to the behaviors and well-being of the people in a society.
We emphasize that the relationship between social structure and economic behavior is
not unidirectional, as the relationships that constitute a given network are endogenous and
determined partly by economic behaviors. In particular, the symbiotic relationship of social
context and behavior complicates empirical analysis, since the relationships among most of
the variables of interest are endogenous. It is thus essential for many economic questions
to understand how networks form, evolve, and interact with behaviors. From an empirical
perspective, these questions arise at a unique time in which large network data sets are
rapidly becoming available, along with the computing power to analyze them.
In Section 2, we first elaborate on a few specific examples in order to ground the discus-
sion and illustrate our major themes. In Section 3, we propose a classification of network
characteristics and discuss how they relate to behavior, and we use this classification as the
base for the remainder of the article. Sections 4 through 7 present detailed descriptions of
how specific network characteristics relate to economic behavior. In Section 8, we discuss
some challenges that arise with empirical analyses in networked settings, devoting particular
attention to endogeneity problems, which are ubiquitous in the study of social interactions.
We close with a summary and some concluding remarks in Section 9.
2 Illustrative Examples
In order to ground our discussion, we start by expanding on some of the examples mentioned
in the Introduction in which network structures are of primary importance in determining
behavior. Each of the following four examples illustrates a theme that we elaborate upon
2
Some aspects of networks have been covered in previous surveys. See, in particular, Jackson (2003,
2004, 2005, 2011), Ioannides and Datcher-Loury (2004), Granovetter (2005), Jackson and Yariv (2011),
Jackson and Zenou (2015), as well as the books by Demange and Wooders (2003), Vega-Redondo (2007),
Goyal (2007), Jackson (2008a), Benhabib, Bisin and Jackson (2011), Jackson and Zenou (2013), and
Bramoull´e, Galeotti and Rogers (2016).
3
below.
First, many criminal behaviors do not occur in isolation, but rather take place in a
social context.
3
Indeed, criminals often have friends or acquaintances who have themselves
committed several offenses. These social ties among criminals can serve as a means whereby
individuals actively or passively influence one another to commit crimes. In fact, not only the
behavior of direct friends, but also that of the larger structure of an individual’s network,
predicts criminal behavior. Influence occurs through a number of channels, as criminal
behaviors involve many complementarities, including role models, learning, and increased
opportunities, which can lead individuals to undertake criminal acts. Moreover, some crimes
inherently involve team production (e.g., production and trafficking of illegal drugs and
goods) and require criminals to work with accomplices. These complementarities can then
feed back and affect the social network in which an individual resides, as they may constitute
relevant components of the decision to invest in relationships. This, in turn, can reinforce
behaviors and erode investments in more productive human capital and opportunities.
Second, we observe persistent inequality on a number of dimensions (e.g., wages, pro-
motions, health, etc.) between ethnicities, genders, and other social classes. Important
components of these differences relate to segregation patterns in interaction, as segregation
in network structures affects how information flows, what access individuals have to various
opportunities, and how decisions are made. In sufficiently segregated networks, different
behaviors, norms, and expectations can persist in different communities which, in turn, can
have consequences for human capital investments, career choice, and various other behaviors.
Once outcomes differ across communities, individuals have different investment incentives
since they have different opportunity costs of, and benefits from, education and other de-
cisions. The differences in costs and benefits stem from complementarities in behaviors, as
there are often advantages to choosing similar behaviors to our neighbors. For instance,
returns to education are higher if one has educated friends who can provide information
about the optimal pursuit of an education, and eventually can serve as contacts for access
to skilled jobs. So optimal behavior is likely to be different across communities even if un-
derlying preferences are not systematically different. The differences can be reinforced by
complementarities and thus become persistent. Hence, it is essential for economists to un-
derstand why networks often exhibit strong segregation patterns, why those structures seem
to be so persistent, and how those patterns affect behavior and outcomes. Recent studies
have made significant progress on each of these facets.
Third, one of the most extensively studied network phenomena is diffusion. The spreading
3
It is well-established that crime is, to some extent, a group phenomenon, with sources of crime and
delinquency that can be traced to the social networks of individuals (see e.g. Sutherland (1947), Sanercki
(2001), Warr (2002), Calv´o-Armengol and Zenou (2004), Patacchini and Zenou (2012)).
4
of ideas, information, behaviors, and diseases, are all network-based phenomena. A most
prominent application is from epidemiology: how does a contagious disease spread through
a population? Finer details of network structure have only recently been systematically
incorporated in answering this kind of question. Features such as segregation, network
density, the distribution of links, the joint characteristics of linked individuals, as well as
potential changes in the network arising from individuals’ reactions to the contagion, are all
important to understand. A second application of diffusion centers on technology adoption:
when should we expect a new technology be widely adopted? What do the dynamics of
market penetration look like, and what factors determine success or failure of adoption?
More generally, it is important to understand which aspects of network structure enhance
or impede diffusion. How do the answers to these questions depend on the nature of the
diffusion process?
Fourth, and finally, cooperative behavior prevails in some environments, and not others.
Particular manifestations of behaviors that require cooperation include informal risk sharing
and favor exchange, the provision of various (local) public goods, and economic exchange;
all of which matter greatly in the development of a society. These are all inherently network
phenomena, as people react to their neighbors, and what they hear about others’ behaviors.
Pro-social behavior is routinely observed, even in contexts with little in the way of formal
institutions to provide sanctions. This is true in both the developing world and the developed
world, as many interactions are more easily governed by social sanctions than relying on
costly formal contracting. The means of providing appropriate incentives often relies in
large part on social structure. For example, information about misbehavior can spread
quickly through an individual’s network, leading to negative reactions in future interactions
for those whose actions conflict with social norms. Social structure plays a prominent role
in determining the forms of cooperative behavior that can be maintained.
3 Classifying Network Characteristics
With these motivating examples in hand, we now offer a framework through which to un-
derstand how structural properties of a network impact the behaviors of the agents who
comprise the network.
4
The framework is based on the fundamental characteristics of net-
works. We focus on four major characteristics (that we define below for those new to the
subject): degree distributions, homophily patterns,
5
clustering, and the centrality of nodes.
Naturally, there are many other facets of such inherently complex structures that can also
be important. We focus on these four because they are particularly prominent, fundamental,
4
See Jackson (2014).
5
Homophily is the tendency of agents to associate with other agents who have similar characteristics.
5
