Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age by Bernard E. Harcourt

Summary

In "Against Prediction," Bernard E. Harcourt challenges the widespread reliance on actuarial prediction instruments in policing, sentencing, and surveillance. He argues that predictive profiling, often justified as neutral and efficient, actually perpetuates social inequalities and produces counterproductive effects in the justice system. Harcourt weaves together empirical evidence, legal criticism, and philosophical analysis to question the legitimacy and efficacy of actuarial methods. The book is a call to reconsider our increasing dependence on data-driven criminal justice practices and to recognize their hidden societal costs.

Life-Changing Lessons

1. Actuarial prediction tools can unintentionally reinforce existing biases, leading to systemic discrimination, especially against already marginalized communities.

2. Increasing reliance on prediction in policing and punishment changes societal behavior, often creating feedback loops that exacerbate crime or inequality.

3. True justice requires continuously questioning and examining the tools and rationales used to manage populations, ensuring they align with broader ethical and democratic principles.

Publishing year and rating

The book was published in: 2007

AI Rating (from 0 to 100): 91

Practical Examples

1. Police Drug Courier Profiles

   Harcourt examines how law enforcement agencies use drug courier profiles to predict who might be carrying contraband. These profiles, often based on statistical correlations, can lead to racially disproportionate stops and searches. This practice is critiqued for reinforcing stereotypes and feeding a cycle of profiling that targets specific groups.

2. Sentencing Guidelines

   The book discusses the use of actuarial tools in sentencing, such as risk assessment instruments predicting recidivism. Harcourt argues these tools commodify risk and shift judicial discretion to algorithms, often masking deeply ingrained biases under a guise of objectivity.

3. Automated Highway Surveillance

   Harcourt addresses the deployment of automatic license plate readers and other forms of surveillance that use predictive algorithms to identify suspicious vehicles. While presented as efficient crime-fighting tools, they also foster a culture of suspicion and erode civil liberties.

4. Broken Windows Policing

   The author explores 'broken windows' theory policing, where minor infractions are targeted based on the prediction that they lead to more severe crimes. Harcourt critiques the empirical evidence behind this theory and highlights its role in disproportionately policing poor and minority communities.

5. Preventive Detention Practices

   The book looks at the use of statistical techniques in preventive detention, where individuals can be incarcerated based on predicted future danger rather than actual crimes committed. Harcourt discusses the ethical and legal dilemmas of punishing people for acts they haven't committed, and the risks this poses to fundamental rights.