Computational Research for Equity in the Legal System Training Program (CRELS)

CRELS is currently accepting applications. Please apply here. Deadline to apply is April 29, 2024.

CRELS particularly welcomes applications from students whose identities are underrepresented in STEM and data science. We seek to cultivate multiple multidisciplinary, racially, dis(abled), and gender-diverse cohorts of future leaders in the AI//Science Technology and Society//Social Justice space.

Questions about CRELS? Please contact CRELS' executive Director, Dr. Harpreet Mangat: bids-crels@berkeley.edu.

The UC Berkeley Computational Research for Equity in the Legal System Training Program (CRELS) trains doctoral students representing a variety of degree programs and expertise areas in the social sciences, computer science and statistics.  

Launched in 2023 with a $3-million grant from the National Science Foundation (NSF), this five-year multidisciplinary training program in data science and social science disciplines fosters a new computational social science research community and leads the integration of research on the social implications of AI. CRELS is supported by a 3-million dollar, 5-year grant from NSF’s NRT Research Traineeship Program. The grant supports Ph.D. students focusing on fundamental, longstanding challenges related to inequality and its connection to criminal justice institutions in the United States. Fellows will be trained for careers at the intersection of the studies of inequality, criminal justice, data science, and the social implications of artificial intelligence (AI) and big data. They will generate new scientific knowledge and develop novel tools for large-scale data integration and analysis.

This program is a collaborative effort currently led by David J. Harding, professor of Sociology, outgoing Faculty Director of the Berkeley Social Science Data Laboratory (D-Lab) and a Berkeley Institute for Data Science (BIDS) Faculty Affiliate; Philip Stark, professor of Statistics in the UC Berkeley Department of Statistics; Aditya Parameswaran, associate professor in the UC Berkeley School of Information (I School) and Electrical Engineering and Computer Sciences (EECS); Stephen A. Small, professor in the UC Berkeley Department of African American Studies and Director of the Institute for the Study of Societal Issues (ISSI); Rebecca Wexler, assistant professor in the UC Berkeley School of Law; Erin Michelle Turner Kerrison, assistant professor in the UC Berkeley School of Social Welfare; Marion Fourcade, professor in the UC Berkeley Department of Sociology; Amy E. Lerman, professor in the UC Berkeley School of Public Policy and Director of the Possibility Lab; Avi Feller, associate professor in the UC Berkeley School of Public Policy and in the UC Berkeley Department of Statistics; Andrew Grillo-Hill, PhD, Senior Research Associate and Lead Evaluator with the Science and Engineering team at WestEd; Cathryn Carson, professor and chair in the UC Berkeley Department of History; and Tim Thomas, research director at the UC Berkeley Urban Displacement Project and a Berkeley Institute for Data Science (BIDS) Affiliate.

For full details about this program’s implementation and training faculty, see the "Overview" below. Please contact bids-crels@berkeley.edu with any questions.

___________________________

Overview

The Berkeley Computational Research for Equity in the Legal System Training Program (CRELS) is a new five-year multi-disciplinary training program in the social sciences, computer science and statistics. CRELS will leverage UC Berkeley’s faculty expertise in the law and social science of criminal justice systems (police, courts, prisons, jails, community corrections), data science, and the social implications of AI. Computational research on topics such as prosecutorial decision-making, police misconduct, and the links between evictions and contact with the criminal justice system will address fundamental, longstanding challenges related to inequality and its connection to criminal justice institutions in the United States. The program will benefit from Berkeley’s rich, multi-disciplinary intellectual environments at the Berkeley Institute for Data Science (BIDS), Berkeley D(ata)-Lab, and the Kavli Center for Ethics, Science, and the Public, as well as a longstanding track record of mentorship and training for first-generation and underrepresented PhD students at the Institute for the Study of Societal Issues (ISSI).

Launched by a multidisciplinary research team that includes Berkeley’s Division of Social Sciences, Social Science Matrix, D-Lab, College of Computing, Data Science, and Society, Berkeley Institute for Data Science, Institute for the Study of Societal Issues, Human Technology Futures group, Possibility Lab, Eviction Research Network, EPIC Data Lab, and Kavli Center for Ethics, Science, and the Public, CRELS is supported by a $3-million dollar, 5-year grant from NSF’s NRT Research Traineeship Program. This innovative program aligns with the NSF’s Big Ideas, including Harnessing the Data Revolution, Growing Convergence Research and Transforming Education and Career Pathways. It seeks to create a link between these ambitious goals and Berkeley’s faculty expertise in the social sciences, criminal legal systems, data science, and the ethics and social implications of AI.

Eligibility

Doctoral students who meet the following requirements are eligible to be CRELS trainees:

  • Enrolled in full-time, on-campus PhD programs in the Social Sciences Division, the Division of Computing, Data Science and Society, or in the professional schools at UC Berkeley. 
  • Have completed the first-year course requirements in their home departments; Trainees will be selected for the program each year during the spring semester, typically the spring of their first year of graduate studies.

Funding

Trainee Fellows will be funded for one year, typically in their third year of graduate school, but exceptionally prepared students may receive funding in their second year and students in later years will also be considered. Funding includes a 12-month stipend and in-state tuition and fees. The CRELS stipend is $34,000. Trainees’ home departments are responsible for the difference between the CRELS stipend and GSR salary minimums. Funding in non-funded years typically comes from faculty research grants, internal or external fellowships, or teaching assistantships. Only US Citizens and legal permanent residents are eligible for stipends, tuitions, and fees, due to NSF regulations.

Trainee Scholars are not eligible for stipends or tuition/fees but receive travel or research funding. Trainee Scholars need not be US Citizens or Legal Permanent Residents. 

Program Goals

CRELS emphasizes a team science approach to problem solving and prepares students to generate new scientific knowledge and develop novel tools for large-scale data integration and analysis.

Trainees can expect to acquire the following core competencies:

Core Competencies illustration

Note: The program will consider accepting other coursework in substitution for these courses, including coursework at other institutions.

Program Design

Key components of CRELS include:

  1. A flexible set of pathways for trainees to acquire essential competencies;
  2. Two available “tracks” through the program for trainees with either more or less background in mathematics and computation;
  3. Courses in data science, computing, applied statistics, Inequality and Criminal Legal Systems, Social Implications of AI and Big Data, and on Reproducibility and Collaborative Computational Research;
  4. Participation in a multi-disciplinary CRELS workshop that will meet weekly and will include discussion of work in progress, recently published research, and professional development.
  5. Professional development in team science, research ethics, science communication, and publishing.
  6. Mentoring from faculty from multiple disciplines.

The program accommodates the requirements of trainees’ home PhD programs while providing sufficient flexibility to explore specific interests through an individualized training plan. This built-in flexibility and careful sequencing of required elements ensures that trainees’ time to degree is not delayed.

Major Research Efforts and Projects

The generation and dissemination of knowledge in this training program will emerge from three collaborative, multidisciplinary research projects that aim to discover and analyze potential interventions that will reduce inequalities resulting from the criminal justice system. In so doing, these projects will provide opportunities for scientific advances across multiple disciplines and give trainees both project-based and team-based opportunities for scientific training and professional development. 

  1. California Law Enforcement Accountability Network (CLEAN): At UC Berkeley’s EPIC Data Lab, the CLEAN project aims to identify points of intervention in policing institutions, cultures, and practices to reduce police misconduct and excessive use of force, which fall disproportionately on racial and ethnic minorities and the poor. 
  2. Data-Informed Prosecutorial Decision-Making: Leveraging ongoing work at UC Berkeley’s Possibility Lab, we will assess the ways in which machine learning and other data science tools can effectively support efforts to reduce mass incarceration, increase racial equity, and improve efficiency in prosecutorial case processing.
  3. Eviction and Criminal Justice System Involvement: Prior research on eviction has documented its frequency, racial/ethnic inequalities, and its impacts on a family’s future housing options, health, economic well-being, and child outcomes. Most eviction data are buried within court documents, unavailable to researchers. The multi-institution research group Eviction Research Network (ERN) aims to fill these gaps by collecting and processing more complete eviction data using tools such as NLP to mine electronically scanned court records in underrepresented jurisdictions. We will then link these data to police records and consumer data to study the relationship of criminal justice contact and eviction.

Trainees will typically begin the program in the summer following their first year in graduate school with workshops and bootcamps to prepare them for fall coursework and to remediate any lack of experience with coding or statistics. Students who may need extra preparation and support for success in CRELS will be counseled by the faculty on necessary BIDS or D-Lab trainings. A week prior to fall instruction, trainees will attend a program orientation that covers program requirements and structure, advice on connecting with faculty mentors, training requirements, review of individual development plans, and an introduction to key program leadership and faculty. The program orientation and subsequent group meetings will serve to form close bonds within and across cohorts. Initially, each trainee will be matched with one mentor from their own discipline and another from a different discipline, based on their research interests and developmental goals, although they will be allowed to change mentors as the program progresses.

Program coursework will occur primarily in the second year of graduate school, with some courses possibly completed in the third year, depending on requirements of the home department. We also anticipate that some trainees will begin taking program courses in their first year of graduate school, in anticipation of application to the program. Given the wide range of backgrounds and prior experiences, each trainee's coursework will need some degree of tailoring in consultation with program faculty. The coursework will play a critical role in developing core competencies that will prepare trainees for effective engagement during their 3rd year in at least one of the research projects described below. 

Most trainees will take advantage of program funding in their third year. The CRELS workshops and most professional development activities will be available to all interested doctoral students on campus. Professional Development, mentoring, and other support services provided by ISSI will continue as students transition to dissertation research in their fourth and fifth years. A central goal of the fourth year in CRELS will be ensuring all trainees submit at least one paper for publication by the end of the year.

This diagram further illustrates the program design for the typical trainee:

Student Progression through Program illustration

Program Leadership and Affiliates

CRELS faculty and trainees will collaborate across disciplines to simultaneously address social-science and policy questions regarding inequality and criminal justice institutions, the development of tools and methods for leveraging newly available data from the criminal justice system, and ethical and social implications of big data and AI in the context of criminal justice.

Training Faculty

NRT Mentor Affiliations

Application Requirements

CRELS is currently accepting applications. Please apply here. Deadline to apply is April 29, 2024.

CRELS particularly welcomes applications from students whose identities are underrepresented in STEM and data science. We seek to cultivate multiple multidisciplinary, racially, dis(abled), and gender-diverse cohorts of future leaders in the AI//Science Technology and Society//Social Justice space.

Questions about CRELS? Please contact CRELS' executive Director, Dr. Harpreet Mangat: bids-crels@berkeley.edu.

Topics

Training topic page, Fellowship Programs topic page
Computational Research for Equity in the Legal System banner

Support CRELS

Please support CRELS here. Your support will help us continue our mission and goals. Our goal is to train PhD students in the social sciences, professional schools, statistics, and computer science in multidisciplinary research on the criminal legal system that leverages new data science methods and take into account the ethical and social implications of big data and artificial intelligence.

  • The research training program stands on a three-legged stool: criminal legal systems, data science, social and ethical implications of Artificial Intelligence (AI)

  • A larger goal is to build a multidisciplinary community at Berkeley and beyond that leverages the intellectual frameworks and tools of various disciplines to advance what we’re calling computational social science

  • Contribute to diversification of academia more generally and data science in particular by creating an intellectual community where all students can thrive and learn from one another

  • Why focus on criminal legal systems:

    • Prior research: a major driver of inequality in the US

    • Public records from the criminal legal system are increasingly available – need to process them into large-scale structured data

    • A topic that attracts a diverse range of students and motivates them to invest in the skills necessary to address topics like prosecutorial decision making, police misconduct, and the connections between eviction and criminal legal system contact

    • Criminal legal system is a site where AI is increasingly being used and misused

Donors who give an or more will receive annual updates from the CRELS program as well as invitations to special CRELS events and insider communications. Please contact CRELS executive director, Dr. Harpreet Mangat (bids-crels@berkeley.edu) if you have any questions.