DSC180A-Fair-Policing

Is predictive policing fair?

Schedule

Week	Topic
1	Introduction
2	Background: Police, Traffic, and Discrimination
3	Traffic Stops Data
4	Causal Inference
5	Causal Inference II
6	Veil of Darkness
7	Improvements on Veil of Darkness I
8	Improvements on Veil of Darkness II
9	Impacts and Ethics
10	Work on Proposals

Introduction

Tags: Causal Methods, Discrimination, Data Journalism, Policy
Data: Administrative data, Geographical Data
Methods: Causal Inference, Natural Experiments.

The most common point of contact between police and residents are through traffic stops. While traffic stops help keep the roads safe, police also have a host of other reasons behind them, and these stops can have wildly different effects on the drivers being pulled over. This project will initially look at how various notions of fairness play out in traffic stops in San Diego in 2014-2015. In particular, the project approaches the question “does being identified as a certain race cause a person to get pulled over more often?”

Background

The current state of law enforcement in the United States is the result of a long history, heavily influenced by the legacy of slavery, the rise of the automobile, and the war on crime. Some of this history is covered in the following references:

Suspect Citizens: What 20 Million Traffic Stops Tell Us About Policing and Race by Frank R. Baumgartner, Derek A. Epp, Kelsey Shoub
Policing the Open Road: How Cars Transformed American Freedom by Sarah A. Seo

Data

Open data on traffic stops exist in cities and states throughout the United States.

The project will primarily use San Diego traffic stops data.
Stanford open policing has cleaned traffic stops data collected from across the country. This is a good reference source, but use the original source when possible.

Possible Projects