Predictive Sentencing addresses the role of risk assessment in contemporary sentencing practices. Predictive sentencing has become so deeply ingrained in Western criminal justice decision-making that despite early ethical discussions about selective incapacitation, it currently attracts little critique. Nor has it been subjected to a thorough normative and empirical scrutiny. This is problematic since much current policy and practice concerning risk predictions is inconsistent with mainstream theories of punishment. Moreover, predictive sentencing exacerbates discrimination and disparity in sentencing. Although structured risk assessments may have replaced 'gut feelings', and have now been systematically implemented in Western justice systems, the fundamental issues and questions that surround the use of risk assessment instruments at sentencing remain unresolved. This volume critically evaluates these issues and will be of great interest to scholars of criminal justice and criminology.

From random security checks at airports to the use of risk assessment in sentencing, actuarial methods are being used more than ever to determine whom law enforcement officials target and punish. And with the exception of racial profiling on our highways and streets, most people favor these methods because they believe they’re a more cost-effective way to fight crime. In Against Prediction, Bernard E. Harcourt challenges this growing reliance on actuarial methods. These prediction tools, he demonstrates, may in fact increase the overall amount of crime in society, depending on the relative responsiveness of the profiled populations to heightened security. They may also aggravate the difficulties that minorities already have obtaining work, education, and a better quality of life—thus perpetuating the pattern of criminal behavior. Ultimately, Harcourt shows how the perceived success of actuarial methods has begun to distort our very conception of just punishment and to obscure alternate visions of social order. In place of the actuarial, he proposes instead a turn to randomization in punishment and policing. The presumption, Harcourt concludes, should be against prediction.

'Is it fair for a judge to increase a defendant's prison time on the basis of an algorithmic score that predicts the likelihood that he will commit future crimes? Many states now say yes, even when the algorithms they use for this purpose have a high error rate, a secret design, and a demonstratable racial bias. The former federal judge Katherine Forrest, in her short but incisive When Machines Can Be Judge, Jury, and Executioner, says this is both unfair and irrational ...' See full reviewJed S RakoffUnited States District Judge for the Southern District of New YorkNew York Review of Books This book explores justice in the age of artificial intelligence. It argues that current AI tools used in connection with liberty decisions are based on utilitarian frameworks of justice and inconsistent with individual fairness reflected in the US Constitution and Declaration of Independence. It uses AI risk assessment tools and lethal autonomous weapons as examples of how AI influences liberty decisions. The algorithmic design of AI risk assessment tools can and does embed human biases. Designers and users of these AI tools have allowed some degree of compromise to exist between accuracy and individual fairness.Written by a former federal judge who lectures widely and frequently on AI and the justice system, this book is the first comprehensive presentation of the theoretical framework of AI tools in the criminal justice system and lethal autonomous weapons utilized in decision-making. The book then provides a comprehensive explanation as to why, tracing the evolution of the debate regarding racial and other biases embedded in such tools. No other book delves as comprehensively into the theory and practice of AI risk assessment tools.

All modern sentencing systems, in the US and beyond, consider the offender's prior record to be an important determinant of the form and severity of punishment for subsequent offences. Repeat offenders receive harsher punishments than first offenders, and offenders with longer criminal records are punished more severely than those with shorter records. Yet the vast literature on sentencing policy, law, and practice has generally overlooked the issue of prior convictions, even though this is the most important sentencing factor after the seriousness of the crime. In Paying for the Past, Richard S. Frase and Julian V. Roberts provide a critical and systematic examination of current prior record enhancements under sentencing guidelines across the US. Drawing on empirical data and analyses of guidelines from a number of jurisdictions, they illustrate different approaches to prior record enhancements and the differing outcomes of those approaches. Roberts and Frase demonstrate that most prior record enhancements generate a range of adverse outcomes at sentencing. Further, the pervasive justifications for prior record enhancement, such as the repeat offender's assumed higher risk of reoffending or greater culpability, are uncertain and have rarely been subjected to critical appraisal. The punitive sentencing premiums for repeat offenders prescribed by US guidelines cannot be justified on grounds of prevention or retribution. Shining a light on a neglected but critically important topic, Paying for the Past examines the costs of prior record enhancements for repeat offenders and offers model guidelines to help reduce racial disparities and reallocate criminal justice resources for jurisdictions who use sentence enhancements.