Dondena Interviews

In an engaging talk this spring, New York University’s Delia Baldassarri presented a study of how some Ugandan producer organizations were able to cooperate—and thus make more money selling better crops—more effectively than others. The works sheds light on the perennial topic of free-ridership, and discusses the influence of things like reciprocity and cooperation.

            This fits with her broader research on networks and how they can help explain our behaviour. And later this year, another project may also reveal what would happen in your neighborhood if you dropped a letter on your way to the post.

Give me a summary of the talk you’re giving today.

The talk is about a group of farmer cooperatives in Uganda and the basic question is, how is it possible that some of these farmer groups are capable of overcoming collective action problems and some are not. Collective action problems are problems in which individuals face a trade-off between their own individual interests and the interest of the collectivity. The classic example is going on strike. The benefit you derive from going on strike most of the time is that you get a higher paycheck or some other benefits, but if your coworkers go on strike and you go to work you make money and you get some benefit. There’s the same thing with taxes, taxes are good for everybody but it’s better to have other people pay taxes rather than me, and so on.

            So these types of collective action problems are pervasive in any type of society, and the question I’m trying to answer is, what are the conditions that allow people to overcome them? And in this presentation today I’m going to focus on how mechanisms of reciprocity and interpersonal relationships can allow them to overcome their own self-instincts and actually cooperate. The question is whether people decide to cooperate, instead of free-riding, in collective action situations because they are altruistic or because they feel some type of solidarity toward members of the group, or whether they actually find it convenient from an economic perspective. Namely, individuals may cooperate not because of altruism or solidarity, but because they understand that in the context of repeated interaction, it’s actually better for the individual to cooperate than defect. Because defection might work once, but once you defect people might be upset with you if you are in a context in which you are supposed to interact with people over time. Therefore, you might be subject to the threat of sanctioning—monetary or reputational. So the idea here is to pick up all the mechanisms that might lead to cooperation through interpersonal relationships.


And so what is the data like? Where does this come from?

This comes from the experiences of these farmer cooperatives, founded six years ago by a collective effort led by the Ugandan government and USAID, in which farmers were invited to form cooperatives in order to exploit economies of scale.

            Essentially, for poor farmers in African villages and, in general, farmers around the world, it’s much more convenient to sell their crops together rather than selling them individually because if you sell a large amount of crop you get a better price for your own production. The goal of this cooperative effort was to get these farmers to sell their coffee together, but also improve the quality of their product. Again, if you improve the quality you might also actually improve the price, the final benefit you get from your activity.

            Some of these farmers groups were extremely successful, others were not. In some places, all the farmers sold their product collectively, in other places just 10% or 15% do. So the question is why there are these huge differences in a context in which essentially the quality of the land, the level of education, farmers’ skills and capabilities are almost the same. There must be some individual and group-level dynamic going on that makes it possible for some villages to be very effective in these farmer group activities where other villages instead fail.

            What I find, essentially, is that the extent to which these farmers have developed relationships among themselves – social networks – helps them to achieve higher levels of cooperation. This is because interpersonal relationships makes it possible to influence each other, pass along information and sanction defectors. So the bottom line is that mechanisms of reciprocity and cooperation might emerge through social networks.


How does this fit in with the bigger picture of what you do in your research?

I always try to understand social problems using a network approach. Meaning, I believe individuals have values and interests, but in order to explain their behavior it is extremely important to also understand the context in which they are embedded, the people around them, and the group motivations, and norms, that can merge from these patterns of relationships.

            I tend to use a network approach to explain cooperation, but also to explain conflict, for instance. I do believe that when people form groups they also enter dynamics in which they differentiate themselves from other groups, so there is the possibility for cohesion within the group but also there is the possibility of conflict with people from different groups.

            For instance, in other areas of my research I study dynamics of political polarization in American public opinion. Over the last 30 years the division between Republicans and Democrats in the U.S. has grown immensely. This polarization is not only visible in Congress, public opinion, or in terms of party agendas. but also in citizens’ everyday experience: individuals often report that they live in very segregated networks.

            For instance, when I started working on this topic, I had lived in New York for three years and I realized I knew only one Republican. This is partly due to the fact that I do live in an environment in which people mostly tend to be Democrat. I am a sociologist, I’m in academia, I live in Manhattan, but this is not sufficient to explain the level of political homogeneity of my social network. The remaining part of the story is that we are very reluctant to talk about politics with people we don’t know whether we are going to agree with. Essentially, we talk politics only with people we anticipate we are going to agree with. And by doing this we reinforce this feeling that we live in very segregated networks and the people around us are either all Democrats or all Republicans. Instead, if we had the chance to actually take an ‘objective’ picture of our networks and really know the political preferences of every single person that we interact with, then we’d probably find out that our networks are not as segregated as we perceive them to be. So this is another example of how I use networks to understand how people form groups and interact in them.


Are there many people working in this field?

The network approach has been extremely successful in the last 20 years. I’d say in many areas of society and social sciences in general, the concept of networks has been used a lot. In terms of really thinking about how interactions affect collective outcomes or interpersonal relationships we still have a lot of work to do.

            The thing is, several scholars tend to rely on very superficial measures of social networks. For instance, consider the concept of social capital, according to which individuals and groups may benefit from their social networks. Usually people measure it in terms of number of ties, or associational memberships, or trust, but they don’t really try to think about network properties or how, for instance, being imbedded in a very close-knit pattern of relationships actually isolates individuals from other types of relationships instead of getting them embedded in broader networks.

            So essentially what I would say is that yes, the concept of networks has been very successful. People use it all the time. However, they are probably less aware of the trade-offs that we always experience by using our social networks.


What kinds of things would you like to see people doing more of, in this space and even in the social sciences more broadly?

I’d like people first to become more familiar with the mathematical and graph theory basis upon which social network research is based. I’d also like them to engage more seriously in theory construction, in terms of thinking about how networks consequences can be quite complex. For instance, people believe that dense social relationships are good, both for the individual, than the collectivity. But consider this. If I form close ties with individuals in my neighborhood (the West Village), that means that I will strengthen my relationship with other upper middle class whites, thus leading to benefits of various sort. However, this would also further , exclude from our network of relationship, and thus from those benefits, people of different ethnoracial and economic background. This may exacerbate inequality, social exclusion, and lead to political conflict.            This is something that has been mostly ignored by many social scientists, for instances Putnam, who uses the concept of networks as this kind of magic wand that can solve all types of collective problems without realizing that by forming networks, we also exclude people, we also capture advantages and resources in society.


It’s like trading one problem for another.

Exactly. There is this intuition that people might benefit from networks, this idea of social capital, but there are also people that actually lose by not being embedded in these networks. We should be able to consider both sides in our research. And scientists are really behind on this. They don’t see that anytime there is someone who gains from relationships, there is probably someone who loses, either a social group or simply individuals. This problem is shared by economists too: they do tend to use networks in a macro context, doing macro analysis, they like the concept of social capital because it sounds like monetary capital. And they use it in the same way, but they don’t get to see specifics. So I’d love to see the concept of networks used more cooperatively.


So what is the next terrific academic thing you have coming up on your calendar that you’re excited about.

I have two things. In Italy, I’m conducting a “Lost Letter” experiment, which is an experiment I’ve conducted all over the country, in around 200 different cities going from big cities to small towns. The experiment is very simple. A bunch of letters, between 20 and 40 in each city, with an address on it and a stamp are lost in public spaces. The goal is to see how many of them get picked up by passersby and sent to the recipient. the rate of return is considered to be a measure of solidarity towards some unknown other.

            First of all, I am interested to see whether there are regional differences. There is some research, and plenty of stereotypes about differences between the north and south in terms of trust and solidarity. Second, scholars tend to believe that in small towns there is much more cohesion, and therefore people are nicer to each other. I don’t believe that, and I won’t be surprised if we find higher levels of returns in big cities rather than in small towns.

            Finally, in this experiment I consider 4 types of recipients, and randomized their distribution. One is an ordinary Italian citizen with a autochthonous name. The second is an immigrant, Mohammed Hassan, the third is a university professor ,and the fourth is a member of parliament. So I expect to see difference in their rates of returns.


Oh, that’s fascinating! So, would you mail the letter?

I would! I definitely would!


I never would, do you know why? I would think to myself, What if this letter was lost on purpose? What if the person sending this letter decided at the last minute, “You know what? I don’t want to mail this letter.” Maybe it was an old lover, or who knows what, or their brother, and the letter says something awful, and the person said, “I don’t want to send this letter.” I’d probably throw it away.

I have never thought of this!


So what’s the other thing—I can’t imagine anything more exciting than this, by the way...

The other project is to study dynamics of solidarity and cooperation between different ethnic groups. Some of the literature coming out right now, and the current debate on affirmative action in the US, argue that people who live in ethnically diverse neighborhoods, or even school environments, , have lower levels of trust. However, the support for this is very weak empirical evidence, which is based on silly attitudinal questions like “Do you think that most people can be trusted or not?”

            In contrast, there are empirical results from other studies in which people are as likely to perform CPR on someone who is having a heart attack in heterogeneous neighborhoods as well as homogeneous neighborhoods. Similarly, in a Lost Letter experiment, done recently in Chicago, the ethnic diversity of a neighborhood is not a predictor of the rate of letter return. The main predictor are instead economic conditions and residential stability.

            So the idea is to actually study whether intergroup relationships actually lead to lower levels of trust and cooperation in urban setting. Instead of relying on attitudinal measures of trust, I plan to actually have people engage in monitored exchanges, behavioral games, or other types of behavioural tests like the Lost Letter experiment. And hopefully I will find some other, more nuanced explanations of the relationship between intergroup contact and cooperation than the ones that have been advanced so far..


Learn more about Delia Baldassarri here. Learn more about the Dondena Seminar Series here.
















Last updated 10 December 2016 - 05:39:06