Stevenson and Cox on Eugenic Criminology and the Birth of Predictive Algorithms in Criminal Justice

Megan T. Stevenson, University of Virginia School of Law, and Robynn Cox, UC Riverside, University of Southern California, Schaeffer Center for Health Policy and Economics, have published Eugenic Criminology and the Birth of Predictive Algorithms in Criminal Justice. Here is the abstract.

This Article tells the story of the birth of predictive algorithms in criminal justice. Known as risk assessments, these tools are widely used today to make decisions about bail, sentencing, and parole. Their roots trace back to the 1920s, when statistical prediction tools were first proposed for use in criminal justice decision-making. In this Article, we show that risk assessment found its origins in the ideas of eugenic criminology: namely, that crime is mostly caused by an inferior subclass of humanity, tainted from birth. Risk assessment was conceptualized as a way of sorting between the "normals" who were amenable to reform and the "sub-normals" who, due to their inferior genes, were not. Such "born criminals" were seen as requiring indefinite confinement within isolated penal colonies in order to protect society from crime, prevent procreation, and provide care for those in need of paternalistic guidance. We tell this story in part because it is a fascinating piece of history, marked by bigotry, bravado, and an almost fanatical optimism about mankind's ability to engineer a perfect society. But we also tell it because the ideas and practices of eugenic criminology are not widely known. While "tainted origins" do not automatically condemn the ongoing use of risk assessment, understanding history can help identify ways that the past lives on in the present.

Download the article from SSRN at the link.

Klerman on Quantitative Legal History @USCGouldLaw

Daniel M. Klerman, USC Gould School of Law, is publishing Quantitative Legal History in the Oxford Handbook of Historical Legal Research (Oxford, forthcoming). Here is the abstract.

Legal historians seldom use statistics, but this is a missed opportunity. Quantitative methods are particularly helpful in understand core legal history issues, including the effect of legal change and the influence of multiple factors on legislation, judicial decisionmaking, and citizen behavior. Recent work by Gavin Wright, Paul Mahoney, and Michele Landis Dauber shows how tables, graphs, and regression analysis can be woven into persuasive historical narrative and analysis. Collaboration between legal historians and quantitative social scientists also provides an untapped avenue to enrich the field.

Download the essay from SSRN at the link. 

Colorado Prosecutors, Race, and the Death Penalty in the Early Part of the 21st Century

Meg Beardsley, Washington and Lee University School of Law, Sam Kamin and Justin F. Marceau, both of the University of Denver College of Law, and Scott Phillips, University of Denver, University of Denver, are publishing Disquieting Discretion: Race, Geography & the Colorado Death Penalty in the First Decade of the Twenty-First Century in volume 92 of the Denver University Law Review (2015). Here is the abstract.

This Article demonstrates through original statistical research that prosecutors in Colorado were more likely to seek the death penalty against minority defendants than against white defendants. Moreover, defendants in Colorado’s Eighteenth Judicial District were more likely to face a death prosecution than defendants elsewhere in the state. Our empirical analysis demonstrates that even when one controls for the differential rates at which different groups commit statutorily death-eligible murders, non-white defendants and defendants in the Eighteenth Judicial District were still more likely than others to face a death penalty prosecution. Even when the heinousness of the crime is accounted for, the race of the accused and the place of the crime are statistically significant predictors of whether prosecutors will seek the death penalty. We discuss the implications of this disparate impact on the constitutionality of Colorado’s death penalty regime, concluding that the Colorado statute does not meet the dictates of the Eighth Amendment to the Constitution.

Download the article from SSRN at the link.

U. S. Privacy Law After One Hundred Twenty Years

Robert Sprague, University of Wyoming, College of Business, Kevin Grauberger, University of Wyoming, and Nicole Barberis, Bloomberg LP, have published One Hundred Twenty Years of U.S. Privacy Law Scholarship: A Latent Semantic Analysis. Here is the abstract.

This paper reports results from a research project aimed at identifying fundamental privacy law principles derived from the writings of legal scholars and commentators using probabilistic topic modeling, which is comprised of a suite of algorithms that attempt to discover hidden thematic structures in large archives of documents. Topic modeling algorithms are statistical methods that analyze the words of texts to discover topics (themes) contained within, how those topics are connected to each other, and how they change over time. A latent Dirichlet allocation process, which identifies sets of terms that more tightly co-occur, is incorporated into the topic modeling analysis to identify words most closely associated with each identified topic. The latent Dirichlet allocation therefore provides insight into the context in which each identified topic occurs. Our analysis reveals that privacy law in the United States comports most closely with the Georgia Supreme Court’s 1905 description of privacy from the seminal case Pavesich v. New England Life Insurance Company: “the right of a person to be secure from invasion by the [government or] public into matters of a private nature.”

Download the paper from SSRN at the link.