Endogenous Rules and Legal Decision Makers

Much of the current empirical legal research avoids discussions about the emergence of endogenous rules, as distinct from formal laws that dictate how legal actors are to conduct the business of the criminal justice system. However, heuristics and norms often evolve when law enforcement agencies face new institutional environments, such as the decriminalization of certain activities or a desire of the public to remove bail from the criminal justice system. Adjustments to these informal processes are often omitted from academic research, as this requires deep institutional knowledge of the legal system.

The Computational Justice Lab, through numerous partnerships with law enforcement agencies, district attorney’s offices, public defender’s offices, courts/judges, probation departments, and a variety of other agencies within the criminal justice system, is engaging in research to understand these institutional features and endogenous strategic behavior. The lab aims to integrate on-the-ground knowledge into our analyses to shed light on important aspects of the criminal justice system that are often missed in current research.

The Open Criminal Court & Jail Justice Initiative provides the best example of the interdisciplinary, group-based efforts of the Computational Justice Lab. This project is currently crawling over one-third of online jail rosters across the United States on a daily basis. We are also actively combining this content with criminal case content that is made publicly available in various jurisdictions throughout the country.

In combining these data, we are generating new insights regarding aspects of the criminal justice system. For example, what is the impact of pretrial on final disposition? How are jails handling overcrowding? Which individuals face the highest risk of injustice from the legal system?

In addition to developing these data for research purposes, we are also actively working with agencies that are committed to assisting high-risk inmates.

Open Criminal Court & Jail Justice Initiative

Open Civil Court Data Initiative

In conjunction with the Rand Corporation, the lab is compiling a comprehensive data set of civil court data. These data will represent the most comprehensive list of case content, with machine learning algorithms utilized to determine the content, case outcomes, and type of case. Running a continuous crawl of these sites, we continue to update these data, which are being utilized in a series of research projects, such as the prediction of compensatory and punitive damages, identification of culpability thresholds, and other pertinent aspects of civil trials.

Identifying Human Trafficking

Human trafficking poses one of the most egregious crimes against humanity. For the past six years members of the Computation Justice Lab have worked intensively with law enforcement agencies to identify instances of human trafficking, including both labor and sex trafficking. These ground truth are currently being utilized by US attorneys and private-public partnerships to identify instances of human trafficking and inform law enforcement agencies of such instances. A current project, in conjunction with Liberty Asia, involves the identification of labor trafficking that examines financial transaction data at the transaction level in southeast Asia.

Officer safety is one of the most important issues that law enforcement agencies face. By improving officer safety, law enforcement agencies can attract higher quality officers. The overwhelming majority of officers that are killed or injured on the job result from domestic violence calls for service. Although this is often known after the fact, determining this information prior to the arrival of an officer could offer opportunities for law enforcement to enhance their safety.

Working directly with administrative records from law enforcement agencies and established working relationships, we will develop algorithms that generate a risk score for law enforcement agents responding to calls for service. We will then utilize a randomized controlled trial to determine the effectiveness of the officer safety tool in reducing officer involved deaths and injuries.

Enhancing Officer Safety

Peer Effects in Law Enforcing

Peer effects in labor productivity, education and social interactions have been well-established. Police engage with numerous peers throughout their career, starting with their initial training. Working with a number of law enforcement agencies, this project aims to unearth the impact of law enforcement peer interactions on the quality of policing. Specifically, we examine how different peer interactions during the training process impacts downstream policing conduct, such as violent interactions that include officer involved shootings.