Mark Granovetter: The Need to Take Social Processes as Seriously as the Economic or Psychological

Jim Coleman, after whom my Fellow position is named, might seem a major exception and yet the reality is that Jim’s greatest impact on policy came from his 1966 report on schools and inequality, which had a long, ponderous name that no one remembers. Everyone just called it “The Coleman Report;” but Jim wrote this actually when his academic career was just beginning and the deeper he got into theoretical argument the more his influence on policy receded. My own research has been on social networks and economic sociology, which might seem to be natural subjects to influence social policy. And yet there is no Council of Social Advisors to complement the Council of Economic Advisors—and by the way, that is complement with an “E”—and sociological advice has not really been sought at the highest levels of policy formation. Now why is that? It could be because the advice of professional economists has been so extraordinarily successful in preventing economic storms and calming economic waters, but there are a few inconvenient facts standing in the way of that conclusion. Actually, there is a very nice book by the historian Michael Bernstein at UCSD, which points out that actually the influence of economists on policy has ebbed and flowed quite a lot in the course of the twentieth century and twenty-first century and you cannot always assume that it is that important.

To come back to the sociology situation, my Italian colleague Carlo Trigilia asked a few years ago specifically why economic sociology, which would seem to have many ideas to offer for policy, has had so little influence. And his argument is that most economic sociologists favor meso- or macro- level analysis, talk about the way social networks function, and that is very complicated. And because it is so complicated it is really hard to draw policy recommendations from it, whereas economists study the incentives of individuals, we all know what that means, and recommends that policy shape those. So that is easy to understand, although I think in practice in fact it is usually a lot more complicated than it looks. I would like to argue though that in fact, contrary to what Carlo argued, that even though a lot of consequences of social network processes are hard to predict—and if you look at the recent literature on complex networks it really is quite technical and complex—but nevertheless I think there are still some policy recommendations that might flow from studies of social networks that are actually pretty easy to understand and would be clear improvements over present practice.

So let me talk a little bit about a whole class of policy ideas that sociologists could develop if they had the time and energy, and if policy were taken more seriously by our discipline. But that probably is not going to happen until sociological policy ideas are actively solicited by policy-makers. So as you see there is a vicious circle here because not many of us want to spend time spinning policy if there is not much interest in it, but there will not be much interest in it if we are not doing it. How to break out of that circle is an important topic, but that is a topic for another day. So let me suggest how we might go about this. The simplest way to talk about my idea is to say that it is extremely cost-effective to leverage the strength of existing social network processes.  Employers do this all the time – they do not know the theory but they know what works.  For example, when they hire through their own employees’ social networks because they know that the new recruits will be quickly socialized into the workplace and subjected to effective social control through their ties to current employees.  Employers do not pay to create the trust and the relationships that have these effects, nor could they if they wanted to.  So it seems to me that leveraging social networks that currently exist is the closest thing that we have to a free lunch in social life and, you know, cost matters a lot.  And there are many examples of this.  So for example, consumers buy used cars from relatives because they expect better disclosure and fairer prices, and they do not pay for those benefits, quite the contrary.  Those raising money for good causes follow the contours of their networks to spark contagion and competition among those who care about one another’s opinions, and this makes so-called affinity fundraising highly effective.  Unfortunately, parenthetically, I should say this cuts both ways, as we know from famous cases of affinity fraud, such as the network strategies of Bernie Madoff—but that was just in parentheses. All the bad things have to be in parentheses, you know.

There are some easy translations of this principle into public policy.  Consider public efforts to give the poor and/or the unemployed new skills to help them emerge from poverty.  This has been a big emphasis for many years.  But the vast majority of such programs in the United States are conceived as efforts to improve the attitudes, the human capital, and the search skills of those enrolled, and the reason we think of it that way, I think, has to do with the influence of psychology and economics on public policy.  And the fact is that most such programs have not been terribly successful and that the few that have worked well have been those where trainers, those helping these people, just happened to have good network connections to local employers which they cultivated and used to help them place them.  And there is some interesting research on this, not much, but there is a little bit, Edwin Melendez and the late Ben Harrison worked on this.  Now, the network principle here is really simple.  In normal labor markets, most people find jobs through personal contacts, and if you get skills and you get psychological counseling but you do not get connections to employers then you are still disadvantaged compared to those who are connected to the people in employing companies.  The solution obviously is to develop programs where some personnel are responsible for creating network connections to employers, gaining their trust, and brokering connections to their firms for clients they can vouch for because they actually know them.  But this does not happen very often and one of the reasons is that training has often been identified as the province of clinical psychologists, whose conception of professional identity precludes the dirty work of traveling to factories, which are not very pleasant environments compared to the nice places where they do this training.  So, the sociology of professions helps us see, I think, not only why the policy is misguided but also how vested interests create resistance to the kinds of changes we need.

I think there are many such network applications to policy, including the important task of finding budding terrorists, for example, which is very much of a network problem.  Those concerned with public health are already somewhat aware of the need to find and focus on key network nodes or hubs, as we say, which are critical in the spreading of a disease or a bad habit like smoking or drug use, and of the importance of concentrating your efforts on neutralizing their impact, because that is much more cost effective than targeting random users or random sexual partners.  But systematic ideas about social networks are very slow to enter the general policy discourse, and I think we need to make more strenuous efforts in this direction.  And I want to add that such initiatives are really politically neutral.  In fact, network policy principles should have a lot of appeal to people who want to reduce the role of government, because when you leverage the strength of existing social networks you reduce the need to construct large bureaucracies whose interface with natural processes is stilted and difficult and costly.  And in fact, if you look at the history, that has been one of the things that has hampered many well-meaning social programs.

So this is a very abbreviated appeal for a more serious look at how social policy could be improved and streamlined by taking social processes just as seriously as economic and psychological ones, and if my comments on this occasion move the dial even a little bit in that way, I will be very gratified.