The central theme of the project is to understand the role of data in the design of unbiased policies with regard to race and other factors for emergency and police priority dispatch systems, policing, justice systems, correction facilities and beyond. Towards that, the project aims to create a publicly available comprehensive “data hub” to foster the role of data in policy design, as well as develop analytic methods to evaluate biases using the data.
Researchers are building datasets from various law enforcement-related sources, like body camera images, cell phone mobility data, and social media posts. With interfaces that support users from different programming backgrounds, this initiative will benefit law enforcement researchers across the US.
Linking law enforcement data, which is collected through heterogeneous administrative processes, into a stitched data set will provide a more systematic characterization and insightful understanding of law enforcement.
A large proportion of police-citizen interactions are initiated by 911, and thus pairing 911 call data with police stop data provides a step toward a more systematic causal framework for estimating racial bias.
Applying a systematic framework to causally understand the effect of race on policing, the policing team looks at linked data quantitatively to estimate how different races benefit or suffer differently from the same policy interventions.
Predictive policing systems use narrowly scoped data and narrowly defined objectives that lead to 'hotspot' policing — disproportionate policing of small areas. What impact does this have on communities beyond how it effects crime? We examine how algorithms can lead to changes in police practices and policies.
We are very interested in connecting with MIT undergraduate students and stakeholders interested in the Policing vertical team and in future projects. Please email us at firstname.lastname@example.org. If you would like to be a sponsor and support our work, please reach out to email@example.com.
When it comes to racial profiling, data both hurts and helps. S. Craig Watkins speaks with show hosts Liberty and Scott about the damage policing data can do to communities and how data can also be used to solve the problem.
In this TEDxMIT talk, MIT Visiting Professor S. Craig Watkins addresses one of the fundamental challenges in the AI Ethics debate: computational models that discriminate against marginalized populations.
MLK Visiting Professor S. Craig Watkins looks beyond algorithm bias to an AI future where models more effectively deal with systemic inequality.
Devavrat Shah, Professor at MIT EECS, leads the Policing vertical team that consists of Fotini Christia (Ford International Professor in the Social Sciences at MIT), Jessy Han (MIT PhD Student, IDSS Social & Engineering Systems), Chris Hays (MIT PhD Student, IDSS Social & Engineering Systems), Andrew Miller (Assistant Professor of PoliSci at United States Naval Academy), Manish Raghavan (Drew Houston (2005) Career Development Professor at MIT Sloan and MIT EECS), Kristianny Ruelas-Vargas (Wellesley UROP), Craig Watkins (Ernest A. Sharpe Centennial Professor at the University of Texas Austin), and Chris Winship (Diker-Tishman Professor of Sociology at Harvard).