A bail-bonds business near Brooklyn’s jail and courthouse complex in New York. (Kathy Willens/Associated Press)

Jonathan Lippman is of counsel at Latham & Watkins LLP and former chief judge of the state of New York.

Jonathan Lippman is of counsel at Latham & Watkins LLP and former chief judge of the state of New York.

From coast to coast, officials in the criminal-justice system are finally concluding that the nation’s cash bail system isn’t working.

California's chief justice, Tani Cantil-Sakauye, endorsed a judicial commission last month that found cash bail "unnecessarily compromises victim and public safety" and "exacerbates socioeconomic disparities and racial bias." On the other coast, a New York City commission I led came to a similar conclusion, determining that our jails are filled with people unable to post even modest bail amounts even though they pose little risk to public safety.

But is not enough to say that cash bail is unjust and contributes to the overuse of incarceration, especially when there are powerful forces — such as the bail bonds lobby — that are gathering to protect it. We must offer an affirmative vision as an alternative.

That is why I was pleased to see California judges, like our commission, call for the adoption of evidence-based assessment tools to make better decisions about who poses a risk to community safety and who can be safely released while their cases are pending.

Risk-assessment tools, which use scientific algorithms to make a prediction about the risk posed by each defendant, are not without their critics. Former attorney general Eric H. Holder Jr. and many public defenders across the country have argued that risk-assessment instruments effectively compound the racial bias that already exists in the criminal-justice system.

Their argument is simple: Police are more likely to arrest people of color. Prosecutors are more likely to charge people of color. And judges are more likely to sentence people of color to jail. All risk-assessment instruments rely on prior criminal conduct to one degree or another to predict behavior. Therefore, racial bias is effectively hard-wired into the risk-assessment process.

We need to wrestle with the concerns that Holder and others have raised, but we should not be deterred by them.

Well-designed risk-assessment tools draw on large data sets — aggregating and examining the behavior of thousands of defendants to predict how similar defendants will behave. Research has shown that science-based risk assessment is more accurate than the subjective analysis of the typical front-line criminal-justice practitioner. Risk-assessment tools allow prosecutors, probation officers and judges to make more informed and nuanced decisions about defendants.

No matter how well made a risk-assessment tool might be, it can be no better than the people who use it. As the Center for Court Innovation and others have argued, we need to introduce risk-assessment tools with care, making sure that the people who use them are qualified and trained and that we focus on results to make sure everyone is treated with fairness and consistency. We also need to make the algorithm for determining risk transparent so that we are not relying on prepackaged protocols that undermine or constrain professional discretion.

Fortunately, there are already many risk-assessment tools that can make the justice system fairer and more effective. Virginia's risk-assessment instrument led to the release of more than 550 nonviolent felons over a four-year period who would have otherwise been incarcerated. Contrary to conventional wisdom, the crime rate didn't budge. In Oregon, an actuarial risk-assessment tool effectively eliminated a persistent pattern of disparity in which judges were more likely to place African American young people in secure detention.

In my prior role as chief judge of the state of New York, I saw firsthand how the punitive lawmaking of the past generation — particularly mandatory minimum sentences — effectively moved discretion away from front-line judges. Our history tells us that we shouldn’t be limiting judges’ options; we should be expanding them. We need to give judges the tools they need to make the best decisions possible — and that means providing them with good risk-assessment tools.

Over the past few years, criminal-justice reform has become a bipartisan issue, with everyone from Sen. Bernie Sanders (I-Vt.) to the Koch brothers offering their endorsement. If we play our cards right, this opportunity might mean the end of mass incarceration in the United States. We should be using every possible tool to achieve this goal, including cutting-edge social science.