How the Law School’s Coase-Sandor Institute Supports Empirical Scholarship

When the Law School’s scholars tap into massive data sets to answer pressing legal questions, they often draw support from the Coase-Sandor Institute of Law and Economics, an academic hub and research lab staffed by a small team of analysts trained to clean, organize, and synthesize data.

It’s a behind-the-scenes ingredient designed for Law School faculty whose cutting-edge scholarship includes time-consuming empirical work—and it is a resource John Rappaport, Professor of Law and Ludwig and Hilde Wolf Research Scholar, calls “significant.”

“It’s invaluable to have researchers with this specialized training—there are a lot of papers that wouldn’t be written without their help,” said Rappaport, a leading scholar of criminal procedure and policing who has used large data sets to examine issues related to police behavior. “There’s a lot of early-stage work with empirical projects, and it can create a real bottleneck. The work [of Coase-Sandor researchers] enables the faculty to focus on the higher-level stuff.”

Founded in 2011, the Coase-Sandor Institute puts the Law School’s signature interdisciplinary approach into practice, drawing on the analytic tools of economics to examine legal problems and interrogate how laws and institutions influence people’s behaviors and decisions. The Institute—which bears the name of principal donor Richard Sandor, the Aaron Director Lecturer in Law and Economics and the chairman and CEO of the American Financial Exchange, and his mentor, the late Nobel Laureate Ronald Coase, a longtime Law School professor—facilitates global scholarly collaboration, hosts lectures, and for many years held an annual Summer Institute for international scholars of law and economics. Increasingly, the Coase-Sandor Institute—and Law School scholars—have embraced an expanded and creative vision of the potential of law and economics, applying the discipline’s traditional methods to a host of new subjects, including policing, health care, and education.

The Coase-Sandor data lab, mostly staffed by college graduates with training in economics and quantitative methods and even a few graduate students working toward PhDs in economics or other social sciences, was created as legal scholars became increasingly interested in the power of data to examine complex questions about law and society. Faculty who pursue this work are experts in a wide variety of fields who all “benefit from getting technical help with the magnitude of data they are analyzing,” said Omri Ben-Shahar, the Leo and Eileen Herzel Professor of Law and the Kearney Director of the Coase-Sandor Institute.

The Institute’s researchers have contributed to a wide variety of scholarship, from Professor Anup Malani’s projects involving health insurance experiments in India to Rappaport’s work on policing to other scholars’ research on global bilateral labor agreements, federal bail reform, international organizations, comparative antitrust law, and more.

Sometimes work with one data set can lead to multiple projects. For instance, in 2016, Rappaport and two colleagues, Dhammika Dharmapala, the Paul and Theo Leffmann Professor of Commercial Law, and Richard McAdams, the Bernard D. Meltzer Professor of Law, turned to the Institute with an interest in measuring how—and whether—collective bargaining agreements affect the behavior of police officers. A 2003 Florida Supreme Court decision gave police officers who worked in sheriffs’ offices the right to enter into collective bargaining agreements, which meant the scholars could compare those officers with those in municipal departments unaffected by the court’s decision. And since Florida is unique among states for the sheer volume of police data it makes available to the public, Dharmapala, McAdams, and Rappaport saw an opportunity.

One of the Coase-Sandor researchers, Morgen Miller, worked with them to identify key information from various sources, including the FBI’s Uniform Crime Reports and the Florida Department of Law Enforcement’s database on police misconduct. The assembly and integration of these disparate sources of information into a single database allowed the faculty to test the impact of collective bargaining rights on police behavior. Ultimately, the scholars found that extending collective bargaining rights to sheriff’s deputies in Florida led to a 40 percent increase in violent misconduct.

From this endeavor came a spinoff project about police hiring and certification policies led by Rappaport, who focused on “wandering officers”—members of law enforcement who had been fired for misconduct and were later hired by another police agency in the state, often because information about the officer’s history of misconduct wasn’t shared efficiently with the new agency.

“That data gives a sample that makes it possible to potentially identify trends in general about policing,” said Ben-Shahar, the Institute’s director.

Rappaport also turned to Coase-Sandor for research support on a paper that used police liability insurance claims data to examine whether police behavior is getting worse (he concluded that it isn’t, but that society has become less tolerant of policing harms), and he is working with Coase-Sandor on forthcoming projects as well.

The Institute, he said, provides a level of expertise beyond what one typically could expect from law student research assistants. Data analysis takes time—and it takes time to acquire and clean data, too. Large-scale, nationwide data about police misconduct are not readily available, as there is no federal database that requires the systematic collection or publication of data about police misconduct. Data collection and publication practices also vary significantly between states. So researchers must familiarize themselves with different systems and practices, and sometimes they must merge data from multiple sources.

“This is a law school, so most [student research assistants] are training to be lawyers and they don’t have the skill set to support a high level of quantitative empirical research,” Rappaport said. “Even if they did, they don’t have the time because they’re law students!”

Many other parts of the legal system similarly lack centralized databases, and the assembly of information from many sources to test legal questions requires ingenuity from faculty and quantitative skills from research professionals. Miller, for example, holds a master’s degree and is a doctoral candidate in economics at the University of Michigan who has completed all but her dissertation.

“One of the things about this job that’s so interesting and exciting is that we’re learning about so many different topics. At any one time, I could be working on projects ranging from policing to comparative constitutions,” Miller said.