The influence of lobbyists on politics in the United States is a given. We know they seek to sway politicians with donations of money. How do they decide which politicians to approach and how much to donate? Could it be that politicians’ social connections have something to do with the amount lobbyists are willing to give them? Marco Battaglini, Edward H. Meyer Professor of Economics, set out to find out. Collaborating with departmental colleague Eleonora Patacchini, he wrote a mathematical model formalizing the lobbying decisions of special interest groups.
“There’s a lot of literature about interest groups that promise donations in exchange for favors and to what extent that can affect politicians,” Battaglini says. “But we added an additional dimension to the problem: the fact that politicians are socially connected and look to each other for their decisions.”
Social Networks and Lobbyists’ Contributions
The researchers’ model predicted that the contributions of interest groups to politicians should be proportional to a politician’s centrality in the social networks of Congress. “If I’m socially central in Congress, then I can attract more money because if you convince me to support your cause, you will convince many more legislators, as well, through my social connections,” Battaglini explains.
To test the model, Battaglini and Patacchini collected data about where and when individual legislators in the United States Congress attended university. If two legislators went to the same school during the same time window, the researchers assumed a social connection between them. The network they constructed bore out the prediction; the centrality of the legislators in the social web correlated with the amount of money they received from interest groups.
Estimating Social Connections
The researchers are now expanding their research with additional projects, seeking to pinpoint the way social networks are constructed. “The problem is that you don’t know for sure the connections between individuals in any group,” Battaglini says. “We used alumni connection as a proxy for a social network, but we wanted to do better.” He and Patacchini are collaborating with Sida Peng, PhD’17 Economics and currently research economist at Microsoft, to determine a way to estimate social networks, using observable outcomes—such as whether a legislator’s proposals pass at the committee level. In another related project, they took a more theoretical approach and worked out a model that looks at how fundamentals such as alma mater, age, gender, and other variables decide social connections.
“Our starting motivation for these projects was to try to understand how legislators make decisions and whether connections are a factor when they vote.”
“Our starting motivation for these projects was to try to understand how legislators make decisions and whether connections are a factor when they vote,” Battaglini explains. “But once you’ve addressed these abstract general questions, they can have implications in other contexts. If you know how to estimate social connections using observables, you can use that to predict social connections in many different situations. For instance, banks interact by lending money to each other and so forth. If we have a banking crisis, these connections may actually help propagate shocks, so knowing about them may help predict how the shock on one bank may affect another bank.”
Policymakers and Protesters
In another project, Battaglini modeled the effect of petitions and public protest on the decision making of policymakers. In a situation where everyone would agree on the same policy if they knew the true state of the world, he showed that the larger the degree of conflict between the policymaker and the protestors, the less the protests will sway the policymaker’s decision. At the same time, the precision of each individual protestor’s information is crucial to the impact the protest will have on the policymaker. “If I have a lot of people with a vague idea of what’s happening, even if that knowledge is based on real information, it would still be useless,” Battaglini explains. “It cannot be aggregated. It cannot be put together to influence the policymaker. Even though, if the policymaker could see the information directly, it might have an effect on them.”
Battaglini’s model also suggests that the protest of a smaller group of very knowledgeable people may be more influential than the protest of a large group. The employee protests at Google are an example of this sort of highly informed group influencing decisionmakers, Battaglini notes. “Recently, the employees said they were against the company selling artificial intelligence technology to the government to be used in weapons,” he says. “In this case, the protest worked, and Google ultimately decided to end the contract with the government.”
Economics and Politics
Continuing his focus on the intersection of economics and politics, Battaglini collaborated with Stephen Coate, Economics, to model how negotiations at the political level affect public debt policy. “Traditionally, people studied debt, assuming there is a benevolent ruler making the decision—someone seeking the optimal debt level that will smooth out the cost of taxation over time,” Battaglini says. “But we wrote a theory where policies are decided as the outcomes of negotiations between different interests at the political level.”
The researchers showed that in a simple model with no growth but with shocks to the system—such as a war or an economic boom—public distortions and bargaining cause debt to increase over time. While the amount of debt fluctuates, it never disappears entirely. This is in contrast to the benevolent ruler, who would gradually accumulate assets until debt is eliminated and consumption becomes completely insulated from shocks. The researchers’ findings mirror real life. “If you look at democracies, what you see is a lot of debt,” Battaglini says.
Studies like this offer insights that can be useful to policymakers, Battaglini points out. “Understanding why debt increases over time in a systematic way is important if you want to design institutions that avoid the cost of having excessive debt,” he says. “More generally, it’s very difficult to make a prediction about public debt if you don’t have a theory about behavior in the future, because it’s not just about the shocks you predict may or may not arrive, but it’s about how you expect the political system to react to them.”
Battaglini is aware that his research interests don’t fit what most people expect from an economist. “People often narrowly define economics as markets and prices,” he says. “Most of my projects are about social interactions and how resources are allocated outside of markets. I’m fascinated by these social interactions and what determines them and how they depend on the institutions where these interactions take place.”