Fairness in Criminal Justice Risk Assessments: The State of the Art

Objectives: Discussions of fairness in criminal justice risk assessments typically lack conceptual precision. Rhetoric too often substitutes for careful analysis. In this article, we seek to clarify the trade-offs between different kinds of fairness and between fairness and accuracy. Methods: We draw on the existing literatures in criminology, computer science, and statistics to provide an integrated examination of fairness and accuracy in criminal justice risk assessments. We also provide an empirical illustration using data from arraignments. Results: We show that there are at least six kinds of fairness, some of which are incompatible with one another and with accuracy. Conclusions: Except in trivial cases, it is impossible to maximize accuracy and fairness at the same time and impossible simultaneously to satisfy all kinds of fairness. In practice, a major complication is different base rates across different legally protected groups. There is a need to consider challenging trade-offs. These lessons apply to applications well beyond criminology where assessments of risk can be used by decision makers. Examples include mortgage lending, employment, college admissions, child welfare, and medical diagnoses.

[1]  Howard G. Borden.  Factors for Predicting Parole Success , 1928 .

[2]  H. H. Marks,et al.  Statistics of diabetes. , 1946, The New England journal of medicine.

[3]  M. Rosenblatt.  A CENTRAL LIMIT THEOREM AND A STRONG MIXING CONDITION. , 1956, Proceedings of the National Academy of Sciences of the United States of America.

[4]  H. White.  Using Least Squares to Approximate Unknown Regression Functions , 1980 .

[5]  D. Freedman.  Bootstrapping Regression Models , 1981 .

[6]  D. J. Harris,et al.  Research on Sentencing: The Search for Reform , 1984 .

[7]  R. Berk,et al.  REVIEW ESSAY: DANGEROUS PEOPLE Criminal Careers and “Career Criminals” edited by Alfred Blumstein, Jacqueline Cohen, Jeffrey A. Roth, and Christy A. Visher. Washington, D.C.: National Academy Press, 1986. , 1987 .

[8]  R. Tibshirani,et al.  Generalized Additive Models , 1991 .

[9]  Sentencing by Numbers , 1993 .

[10]  Umesh V. Vazirani,et al.  An Introduction to Computational Learning Theory , 1994 .

[11]  Jan de Leeuw,et al.  An Evaluation of California's Inmate Classification System Using a Generalized Regression Discontinuity Design , 1999 .

[12]  E. Silver,et al.  A Multiple Models Approach to Assessing Recidivism Risk , 2002 .

[13]  Stephen Demuth.  RACIAL AND ETHNIC DIFFERENCES IN PRETRIAL RELEASE DECISIONS AND OUTCOMES: A COMPARISON OF HISPANIC, BLACK, AND WHITE FELONY ARRESTEES* , 2003 .

[14]  Eric R. Ziegel,et al.  The Elements of Statistical Learning , 2003, Technometrics.

[15]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[16]  W. Wu,et al.  Nonlinear system theory: another look at dependence. , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[17]  B. Harcourt,et al.  Against Prediction: Profiling, Policing and Punishing in an Actuarial Age by B.E. Harcourt , 2008 .

[18]  Radford M. Neal.  Pattern Recognition and Machine Learning , 2007, Technometrics.

[19]  M. Sevo.  The Role of Race: Should Race Be a Factor in Health Care Screening and Treatment? , 2008 .

[20]  Franco Turini,et al.  Discrimination-aware data mining , 2008, KDD.

[21]  J. Maindonald.  Statistical Learning from a Regression Perspective , 2008 .

[22]  Shalabh.  Statistical Learning from a Regression Perspective , 2009 .

[23]  R. Berk.  The Role of Race in Forecasts of Violent Crime , 2009 .

[24]  Klaus Nordhausen,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition by Trevor Hastie, Robert Tibshirani, Jerome Friedman , 2009 .

[25]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[26]  Jun Sakuma,et al.  Fairness-aware Learning through Regularization Approach , 2011, 2011 IEEE 11th International Conference on Data Mining Workshops.

[27]  Yuan Y. Liu,et al.  A Comparison of Logistic Regression, Classification and Regression Tree, and Neural Networks Models in Predicting Violent Re-Offending , 2011 .

[28]  Toon Calders,et al.  Data preprocessing techniques for classification without discrimination , 2011, Knowledge and Information Systems.

[29]  Steven L. Chanenson,et al.  Reform in Motion: The Promise and Perils of Incorporating Risk Assessments and Cost-Benefit Analysis into Pennsylvania Sentencing , 2011 .

[30]  R. Berk.  Criminal Justice Forecasts of Risk: A Machine Learning Approach , 2012 .

[31]  Xiangliang Zhang,et al.  Decision Theory for Discrimination-Aware Classification , 2012, 2012 IEEE 12th International Conference on Data Mining.

[32]  Richard Berk,et al.  Criminal Justice Forecasts of Risk , 2012, SpringerBriefs in Computer Science.

[33]  Toniann Pitassi,et al.  Fairness through awareness , 2011, ITCS '12.

[34]  Richard A. Berk,et al.  Overview of: “Statistical Procedures for Forecasting Criminal Behavior: A Comparative Assessment” , 2013 .

[35]  Greg Ridgeway,et al.  Linking prediction and prevention , 2013 .

[36]  Elective Race: Recognizing Race Discrimination in the Era of Racial Self-Identification , 2013 .

[37]  Sonja B. Starr.  Evidence-Based Sentencing and the Scientific Rationalization of Discrimination , 2013 .

[38]  William L. Oliver,et al.  The Emergence of Machine Learning Techniques in Criminology , 2013 .

[39]  Richard A. Berk,et al.  Statistical Procedures for Forecasting Criminal Behavior , 2013 .

[40]  William Rhodes.  Machine Learning Approaches as a Tool for Effective Offender Risk Prediction , 2013 .

[41]  Josep Domingo-Ferrer,et al.  A Methodology for Direct and Indirect Discrimination Prevention in Data Mining , 2013, IEEE Transactions on Knowledge and Data Engineering.

[42]  A. Buja,et al.  Models as Approximations, Part I: A Conspiracy of Nonlinearity and Random Regressors in Linear Regression , 2014, 1404.1578.

[43]  Legal and Ethical Issues in the Prediction of Recidivism , 2014 .

[44]  Andrew Guthrie Ferguson,et al.  Big Data and Predictive Reasonable Suspicion , 2014 .

[45]  Carlos Eduardo Scheidegger,et al.  Certifying and Removing Disparate Impact , 2014, KDD.

[46]  M. Phil,et al.  A METHODOLOGY FOR DIRECT AND INDIRECT DISCRIMINATION PREVENTION IN DATA MINING , 2015 .

[47]  Melissa Hamilton,et al.  Risk-Needs Assessment: Constitutional and Ethical Challenges , 2015 .

[48]  Richard A. Berk,et al.  Machine Learning Forecasts of Risk to Inform Sentencing Decisions , 2015 .

[49]  Andreas Buja,et al.  Models as Approximations - A Conspiracy of Random Regressors and Model Deviations Against Classical Inference in Regression , 2015 .

[50]  R. Berk,et al.  Forecasting Domestic Violence: A Machine Learning Approach to Help Inform Arraignment Decisions , 2016 .

[51]  Andrew D. Selbst,et al.  Big Data's Disparate Impact , 2016 .

[52]  R. Berk.  A Primer on Fairness in Criminal Justice Risk Assessments , 2016 .

[53]  Marijn Janssen,et al.  The challenges and limits of big data algorithms in technocratic governance , 2016, Gov. Inf. Q..

[54]  Indre Zliobaite,et al.  Using sensitive personal data may be necessary for avoiding discrimination in data-driven decision models , 2016, Artificial Intelligence and Law.

[55]  K. Crawford.  Artificial Intelligence's White Guy Problem , 2016 .

[56]  COMPAS Risk Scales : Demonstrating Accuracy Equity and Predictive Parity Performance of the COMPAS Risk Scales in Broward County , 2016 .

[57]  Aaron Roth,et al.  Fairness in Learning: Classic and Contextual Bandits , 2016, NIPS.

[58]  L. Kewley,et al.  ZFIRE: THE KINEMATICS OF STAR-FORMING GALAXIES AS A FUNCTION OF ENVIRONMENT AT z ∼ 2 , 2016, 1606.00528.

[59]  Nathan Srebro,et al.  Equality of Opportunity in Supervised Learning , 2016, NIPS.

[60]  J. Reidenberg,et al.  Accountable Algorithms , 2016 .

[61]  Dean P. Foster,et al.  Impartial Predictive Modeling: Ensuring Fairness in Arbitrary Models , 2016 .

[62]  S. Jonathan.  Actuarial Justice: The Emerging New Criminal Law , 2017 .

[63]  Richard A. Berk,et al.  An impact assessment of machine learning risk forecasts on parole board decisions and recidivism , 2017, Journal of Experimental Criminology.

[64]  Alexandra Chouldechova,et al.  Fair prediction with disparate impact: A study of bias in recidivism prediction instruments , 2016, Big Data.

[65]  Avi Feller,et al.  Algorithmic Decision Making and the Cost of Fairness , 2017, KDD.

[66]  Kush R. Varshney,et al.  Optimized Data Pre-Processing for Discrimination Prevention , 2017, ArXiv.

[67]  Seth Neel,et al.  A Convex Framework for Fair Regression , 2017, ArXiv.

[68]  Jon M. Kleinberg,et al.  Inherent Trade-Offs in the Fair Determination of Risk Scores , 2016, ITCS.

[69]  Jon M. Kleinberg,et al.  On Fairness and Calibration , 2017, NIPS.

[70]  Seth Neel,et al.  Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness , 2017, ICML.

[71]  John Langford,et al.  A Reductions Approach to Fair Classification , 2018, ICML.

[72]  Temple,et al.  PP , 2018, Catalysis from A to Z.

[73]  Jure Leskovec,et al.  Human Decisions and Machine Predictions , 2017, The quarterly journal of economics.

[74]  A. Buja,et al.  Working with Misspecified Regression Models , 2018 .

[75]  Weijie J. Su,et al.  Assumption Lean Regression , 2018, The American Statistician.

[76]  Kristian Lum,et al.  An algorithm for removing sensitive information: Application to race-independent recidivism prediction , 2017, The Annals of Applied Statistics.

[77]  David M. W. Powers,et al.  Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation , 2011, ArXiv.

[78]  Statistical Learning from a Regression Perspective -2/E , 2020 .