Pretrial Algorithms

The Bail Project / Newsroom / Pretrial Algorithms

Pretrial algorithms, commonly known as “risk assessment tools,” have become one of the leading alternatives to cash bail despite their fundamental flaws and potential to codify racial disparities. Here, we collect articles about this alarming trend.

The Bail Project opposes the use of pretrial algorithms to help decide whether an individual is incarcerated pretrial. Together, with data scientists and legal scholars, we argue that pretrial algorithms, also known as risk assessment tools, disproportionately harm people of color, oversimplify complex individual and case outcomes, undermine standards of procedural justice, and rely on data that is inaccurate, outdated, and unreliable....

Poor people who receive welfare benefits from the state are increasingly at the mercy of cold, calculated algorithms that can cut off their assistance at any moment. Because being poor increases the chances of someone going to jail in America, people also face the prospect of judgement by pretrial algorithms, which purport to measure a person's likelihood of committing a new crime or skipping court while out on bond....

An op-ed from leading thinkers in statistics, data science and law explains the fundamental problems with pre-trial risk assessments: "Applying 'big data' forecasting to our existing criminal justice practices is not just inadequate — it also risks cementing the irrational fears and flawed logic of mass incarceration behind a veneer of scientific objectivity."...

A study released by the Center for Court Innovation last week offers further proof that pretrial risk-assessments tools⁠—which some states have turned to in place of cash bail⁠—assign higher risk scores to Black people compared to white, meaning the former are more likely to remain incarcerated where risk assessments are used. While others have made similar observations, the study adds value to the discussion because it suggests this kind of racism is intrinsic to the risk assessment model by design, and not particular to just one or a handful of assessments. ...

The New York Times reported that a glitch that prevented East Baton Rouge Parish's jury database from updating properly left over 150,000 people off the jury rolls: "Across the country, computer-reliant jury coordinators have for years confronted database problems that kept otherwise-eligible potential jurors from being called to the nation's courthouses...