Dallas Police Chief Eddie García has repeatedly complained about a revolving door that lets violent suspects return to the streets to commit other crimes while they await trial. So García commissioned a criminologist to conduct a snapshot of those whom Dallas police officers have arrested to determine the likelihood that they will be rearrested or commit a violent crime while out of jail.

The sample included 464 people, including all 109 murder suspects, who were arrested in 2021. They were tracked through May 15, 2022.

Despite representing just a smattering of the thousands of people arrested during the year, findings are alarming. García presented the survey to the city’s Public Safety Committee last week.

Of the murder suspects, 37, or 33.9%, were released on bail or on their own recognizance. And of those released, five, or 13.5%, were rearrested within an average of 175 days and two were rearrested in connection with a violent crime.

Other stats are disturbing, as well. According to this sampling, a person arrested in the robbery of a business or of an individual or who had been charged with a weapons violation is more likely to be rearrested and more likely to commit a violent crime while awaiting trial than a person charged with murder.

The survey found that 80% of those arrested the robbery of a business and released pending a court date were rearrested within an average of 256 days. Two are accused of committing a violent crime while out of jail.

And 39.1% of those accused of robbing an individual and out of jail while awaiting a court date were arrested again within 120 days. Four are accused of committing violent crimes.

A similar pattern emerged with those charged with weapons law violations. Roughly, 39.1% were rearrested within 134 days, three in connection with a violent crime.

You don’t have to be a criminologist to see a pattern: If a person has been arrested in a robbery or for a weapons violation, the risk of being rearrested increases. Judges, who make the call on who is released before trial, need to know this. So, too, do state legislators who need to give bail policies a detailed review next session.

Violent reoffending is a common theme in Texas.

The Texas Constitution gives a person charged with a crime the right to be released pending trial if they can make bond assigned by the judge or magistrate, except in the case of capital murder. So while various counties use risk assessment questions to help determine the threat that a suspect poses to a community, the ability to make the bond amount is the overriding factor in whether a person stays in jail or gets out while awaiting trial.

The Damon Allen Act, also known as SB 6, passed during the second special session in 2021 with the idea of requiring better processes for judges and magistrates to assess risk.

Backed by Gov. Greg Abbott and named for a law enforcement officer who was brutally murdered while on duty, the measure restricted judges from approving personal bonds for defendants charged with certain violent crimes and directed judges and magistrates to look at a defendant’s criminal history prior to setting bail.

The law also called for a uniform court management system operated through the Office of Court Administration to alert judges to reasons that the accused should not be released before bail is set on a new charge.

García is not alone in the concern that Texas is getting it wrong on bail by letting out violent people who can post a few thousand dollars bond, while retaining poor suspects who are accused of nonviolent crimes and have no violent history.

As a result, Texas’ cash bail system isn’t a good measure of a person’s risk to the community, let alone the proper way to determine who stays in jail and who gets to go home.

For years now, Texas Supreme Court Chief Justice Nathan Hecht has expressed concern that the percentage of people in Texas jails has skyrocketed in the past quarter century from nonviolent offenders who are unlikely to reoffend or flee but who can’t come up with bond money.

In Dallas County, for example, a risk assessment tool available to help judges isn’t a foolproof way to weed out those who pose no risk as they await trial from those who are likely to commit more crimes if released.

“It is simply a tool in the tool box [to help set bail conditions] and was never meant to be a predictor of future behavior,” said Gordon Hikel, Dallas County assistant county administrator.

However, it is clear to us that judges need better information to make decisions about bail, or they need to make better decisions more consistently about how they determine risk before setting bail.

And the Legislature should consider changes in the law to better allow judges to hold high-risk, potentially violent defendants without bail. Pretrial supervision has to improve and should include an assessment of whether ankle monitors, a form of GPS tracking used in pretrial, probation and parole proceedings, are effective or offer a false sense of supervision and risk control.

Counties are busting budgets to house people who can’t make cash bail, and communities and police chiefs are rightly asking that dangerous criminals not return to the streets to offend again.

Hecht supports a change in the Texas Constitution to permit judges to deny bail to certain high-risk suspects and to create a fair, complete review of risk factors. For example, several states permit denial of bail for certain offenses, and the federal criminal justice system heavily relies on a determination of risk in detention or release decisions.

Jail is supposed to protect society from dangerous individuals. The criminal justice system in Texas must make sure that those in jail are the most dangerous, not just the poorest.