Analyzing the risk

Risk assessments are supposed to make the criminal justice system better by predicting which defendants are likely to commit new crimes. Defendant scores are given to judges during criminal sentencing in nine states, and there’s a push to mandate their use in federal prisons. But the risk assessments aren’t accurate, only somewhat more reliable than a coin flip. Black defendants are falsely flagged as future criminals at a high rate while white defendants regularly get mislabeled as low risk.

In Detroit, Project Greenlight uses video surveillance technology to try to solve and deter street crime. Instead of the police requesting private security video after a crime occurs, businesses pay to install the Greenlight system, which then streams video live to police analysts. The police claim it has lowered crime, but researches have found no evidence that it affects violent crime rates. Critics see the system as a pay-to-play system in which businesses buy better police protection. And they say that the cameras, and the use of facial recognition software, bias enforcement against people of color.

Researchers trained Tucson police officers in procedural justice, which involves treating people with dignity and respect by giving them a voice and showing neutrality, to help them de-escalate situations with fewer arrests and less violence while building trust in the community.

Brazil’s Federal Police created a database of gold samples from different regions of the country to help confirm the origins of gold suspected to be illegally traded. Each sample is analyzed to understand its molecular composition, atomic structure, and morphological features. That information can be cross-referenced for matches with suspicious gold during investigations.

In 2018, Massachusetts legislators passed a law allowing people with certain lower-level criminal records to seal the records or expunge convictions altogether. The law was meant to remove the burden that a criminal record imposes on people seeking jobs or renting apartments, particularly when the conduct is no longer illegal, such as marijuana possession. Thousands have managed to get their records sealed. But relatively few have sought or won expungement, thanks to an overly restrictive and complicated process that hasn't been explained well to those who could benefit.

When prosecutors in Boston and Baltimore stopped prosecuting certain non-violent, minor offenses, crime of all sorts, from minor to violent, went down. Getting prosecuted and jailed can in itself increase the probability that people will get in more trouble. When their mistakes are not compounded by an arrest record that limits their job and housing prospects, they are apt to stay out of trouble. The prosecutors and research found that simply excluding cases of non-violent, minor drug possession, disorderly conduct, shoplifting from the system does not encourage more crime; probably the opposite.

By turning over its criminal justice data to Measures for Justice, a nonprofit developer that turns raw data into publicly available reports, the Yolo County district attorney has a much better grasp on the work that it has been doing. Better data mean better-informed decisions about criminal justice reforms. The investment in the new system is prompting policy changes because of racial disparities showing up in the numbers. And that is prompting many other DAs to clamor for the same kind of system.

Starting in the early 1990s, thousands of American towns and cities passed "crime-free housing" ordinances that purport to reduce crime in rental housing. The laws encourage or require landlords to evict tenants based on calls to police, or the criminal records of tenants or their houseguests. The laws' lack of due process has turned them into a tool promoting segregation and retaliation against people of color. Frequently enacted in places that have begun to diversify racially, they give police too much power to declare certain residences a nuisance. Legal challenges have begun to curb some abuses.

Nearly three decades after Washington voters made their state the first to enact a three-strikes law, imposing life imprisonment for repeated, serious offenses, some prosecutors have found ways to avoid the law's effects that are seen as unduly harsh or racially biased. Some have interpreted a law authorizing resentencing to apply to three-strikes cases. Others have pushed the governor to grant clemency more often. This new willingness to question the law's effects is not universal among prosecutors, and the state Supreme Court soon will weigh in on the issue.

The spread of COVID-19 in jails prompted many releases from custody and a surge in donations to bail funds that pay for people's release. But those fixes have done little to address the underlying challenges of detaining millions of people before trial, either because they cannot afford cash bail or because risk-assessment tools deem them a threat to public safety or unlikely to return to court. In two South Florida jails, the struggles over containing the virus, providing due process to criminal defendants, and ensuring public safety have brought the debate into sharper focus.