Abstract
The exponential growth of available information about routine police activities offers new opportunities to improve the fairness and effectiveness of police practices. We illustrate the point by showing how a particular kind of calculation made possible by modern, large-scale datasets - determining the likelihood that stopping and frisking a particular pedestrian will result in the discovery of contraband or other evidence of criminal activity - could be used to reduce the racially disparate impact of pedestrian searches and to increase their effectiveness. For tools of this kind to achieve their full potential in improving policing, though, the legal system will need to adapt. One important change would be to understand police tactics such as investigatory stops of pedestrians or motorists as programs, not as isolated occurrences. Beyond that, the judiciary will need to grow more comfortable with statistical proof of discriminatory policing, and the police will need to be more receptive to the assistance that algorithms can provide in reducing bias.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 181-232 |
| Number of pages | 52 |
| Journal | New Criminal Law Review |
| Volume | 20 |
| Issue number | 2 |
| DOIs | |
| State | Published - Mar 1 2017 |
Fingerprint
Keywords
- Big data
- Discrimination
- Police
- Statistical proof
- Stop-and-frisk
ASJC Scopus subject areas
- Law
Cite this
Combatting police discrimination in the age of big data. / Goel, Sharad; Perelman, Maya; Shroff, Ravi; Sklansky, David Alan.
In: New Criminal Law Review, Vol. 20, No. 2, 01.03.2017, p. 181-232.Research output: Contribution to journal › Review article
}
TY - JOUR
T1 - Combatting police discrimination in the age of big data
AU - Goel, Sharad
AU - Perelman, Maya
AU - Shroff, Ravi
AU - Sklansky, David Alan
PY - 2017/3/1
Y1 - 2017/3/1
N2 - The exponential growth of available information about routine police activities offers new opportunities to improve the fairness and effectiveness of police practices. We illustrate the point by showing how a particular kind of calculation made possible by modern, large-scale datasets - determining the likelihood that stopping and frisking a particular pedestrian will result in the discovery of contraband or other evidence of criminal activity - could be used to reduce the racially disparate impact of pedestrian searches and to increase their effectiveness. For tools of this kind to achieve their full potential in improving policing, though, the legal system will need to adapt. One important change would be to understand police tactics such as investigatory stops of pedestrians or motorists as programs, not as isolated occurrences. Beyond that, the judiciary will need to grow more comfortable with statistical proof of discriminatory policing, and the police will need to be more receptive to the assistance that algorithms can provide in reducing bias.
AB - The exponential growth of available information about routine police activities offers new opportunities to improve the fairness and effectiveness of police practices. We illustrate the point by showing how a particular kind of calculation made possible by modern, large-scale datasets - determining the likelihood that stopping and frisking a particular pedestrian will result in the discovery of contraband or other evidence of criminal activity - could be used to reduce the racially disparate impact of pedestrian searches and to increase their effectiveness. For tools of this kind to achieve their full potential in improving policing, though, the legal system will need to adapt. One important change would be to understand police tactics such as investigatory stops of pedestrians or motorists as programs, not as isolated occurrences. Beyond that, the judiciary will need to grow more comfortable with statistical proof of discriminatory policing, and the police will need to be more receptive to the assistance that algorithms can provide in reducing bias.
KW - Big data
KW - Discrimination
KW - Police
KW - Statistical proof
KW - Stop-and-frisk
UR - http://www.scopus.com/inward/record.url?scp=85025471535&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85025471535&partnerID=8YFLogxK
U2 - 10.1525/nclr.2017.20.2.181
DO - 10.1525/nclr.2017.20.2.181
M3 - Review article
VL - 20
SP - 181
EP - 232
JO - New Criminal Law Review
JF - New Criminal Law Review
SN - 1933-4192
IS - 2
ER -