This book is for anyone concerned about human and fiscal costs of jails and prisons. It shows how properly-developed algorithms can force the adoption of more legally sophisticated bail and sentencing practices that reduce incarceration, minimize racially biased decision-making, and maximize the use of resources, without sacrificing public safety.
Inhaltsverzeichnis
Preface; 1. Rationale: what risk algorithms can do for the criminal justice system; 2. Fit: why and when data about groups are relevant to individuals; 3. Validity: figuring out when risk algorithms are sufficiently accurate; 4. Fairness: avoiding unjust algorithms egalitarian injustice; 5. Structure: limiting retributivism and individual prevention; 6. Moving forward: the need for experimentation.