Precinct or Prejudice? Understanding Racial Disparities in New York City's Stop-and-Frisk Policy

By Sharad Goel, Justin M. Rao, Ravi Shroff. 2016

252 citations recorded by the source corpus.

Subfields: Political Science and International Relations, Sociology and Political Science, AI Ethics, Fairness & Society, Social Sciences & Humanities

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Political Science and International Relations

In Sociology and Political Science

In AI Ethics, Fairness & Society

In Social Sciences & Humanities

Citations in the corpus

Distill-and-Compare: Auditing Black-Box Models Using Transparent Model Distillation (2017, background): “[18] proposed a simple risk scoring model for weapon possession: S = 3× 1PS + 1× 1AS + 1× 1Bulдe , where S is the risk score, PS denotes primary stop circumstance is presence of suspicious object, AS denotes secondary stop circumstance is sight or sound of criminal activity, and Bulдe denotes bulge ”
Stable and Fair Classification (2019, background): “, against African-Americans while predicting future criminals [31, 5, 8], granting loans [22], or NYPD stop-and-frisk [35].”
Algorithmic Decision Making and the Cost of Fairness (2017, background): “Experts increasingly rely on algorithmic decision aids in diverse settings, including law enforcement, education, employment, and medicine [5, 8, 19, 20, 25].”
Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment (2016, methodology): “Finally, we experiment with synthetic as well as real world datasets and show that our methodology can be eﬀectively used to avoid disparate mistreatment.”
Learning Certifiably Optimal Rule Lists (2017, methodology): “values, T2f0;1;2;3;4;5g. To employ this heuristic model in the eld, \...ocers simply need to add at most three small, positive integers ...and check whether the sum exceeds a xed threshold..." (Goel et al., 2016). Figure 8 directly compares various models learned by CORELS to the heuristic mode”
Learning Optimized Risk Scores (2016, methodology): “In the simplest case, this involves scaling and rounding the coefficients from a logistic regression model (Goel et al., 2016) or a linear probability model (US Department of Justice, 2005).”
WILDS: A Benchmark of in-the-Wild Distribution Shifts (2020, neutral): “, 2020), location and race shifts in the New York stop-question-and-frisk dataset (Goel et al., 2016), and category and time shifts in the Amazon and Yelp review datasets (Ni et al.”
Simple Rules for Complex Decisions (2017, methodology): “g, unit weighting) lead to accurate predictions [4, 9, 12, 14]; in particular, our strategy incorporates feature selection and more general integer weights to generate a richer family of simple rules.”
The Problem of Infra-marginality in Outcome Tests for Discrimination (2016, methodology): “Though originally proposed in the context of lending decisions, outcome tests have gained popularity in a variety of domains, particularly policing (Goel et al., 2016, 2017; Ayres, 2002; Knowles et al., 2001).”
Residual Unfairness in Fair Machine Learning from Prejudiced Data (2018, methodology): “D.) Goel et al. (2016) considered learning predictors from this data as a way to assess the implied decision thresholds used for determining pedestrian stops related to the criteria of “reasonable suspicion.””
Aggressive Policing and the Educational Performance of Minority Youth (2018, background): “African-American boys experience disproportionate contact with police (Geller 2018; Goel, Rao, and Shroff 2016).”
Learning Fair Classifiers: A Regularization-Inspired Approach (2017, background): “In settings such as loan approvals [19], policing [7], targeted advertisement [20], or criminal risk assessments [12], algorithmic fairness has to be taken into account in order to ensure the absence of discrimination.”
Beyond Distributive Fairness in Algorithmic Decision Making: Feature Selection for Procedurally Fair Learning (2018, methodology): “In a recent study, Goel, Rao, and Shroff (2015) designed a learning task based on appropriate features present in the datasets to predict whether a person is carrying an illegal weapon or not.”
Robustness to Spurious Correlations via Human Annotations (2020, background): “Previous work on this dataset has shown that racial minorities are stopped more frequently for less serious observations (Goel et al., 2016; Gelmand et al., 2007).”
Does Predictive Policing Lead to Biased Arrests? Results From a Randomized Controlled Trial (2018, background): “…et al. 2016; Horrace and Rohlin 2016), pedestrian stops (Harris 1994; Gelman, Fagan, and Kiss 2007; Fagan et al. 2010; Rios 2011; Fagan et al. 2015; Goel, Rao, and Shroff 2016; Stuart 2016), traffic tickets (Dunn 2009), drug enforcement and arrests (Black and Reiss Jr. 1970; Smith, Visher, and…”