Cook County judges routinely make bail decisions for crime suspects contrary to what the judges’ new risk-assessment system calls for, according to a review of more than 1,500 cases this year obtained by the Chicago Sun-Times.

A review by the Cook County sheriff’s office showed judges’ bail decisions differed from the guidelines about 85 percent of the time. The review tracked cases in 30 weekday sessions — between Feb. 10 and March 29 — in Central Bond Court at the main Leighton Criminal Courthouse at 26th and California.

The sheriff’s study found bail decisions were “inconsistent,” even when defendants’ backgrounds and the charges they faced were factored in.

The 90-page study, the results of which Chief Cook County Judge Timothy Evans disputes, also found that the amount and conditions of bail varied widely depending on which judge was presiding on a given day.

In a speech last month in Chicago, Illinois Supreme Court Justice Anne Burke, who pushed for the new assessment system, said: “One of the key problems with our bond court judges is not just their unwillingness to apply the risk assessments when making their decisions but stems from the fact that they are not being sufficiently trained and supervised and are not being held accountable by the system.”

The system was launched in July 2015 to consider two key factors in addition to the 36 bail factors set out in Illinois law:

• The threat that a defendant might pose to the community.

• And the likelihood of showing up for subsequent court dates.

Those assessments are done by Evans’ pretrial service division. It assigns a number according to the risk level that corresponds to a recommended bail.

One aim was to identify more nonviolent offenders who qualify for release pending trial.

Pat Milhizer, director of communications, Office of the Chief Judge. | Rich Hein/Sun-Times

Pat Milhizer, director of communications, Office of the Chief Judge. | Rich Hein/Sun-Times

Pat Milhizer, a courts spokesman, says that’s exactly what’s happening. In the three months before the assessments began, 52 percent of defendants charged with nonviolent, non-weapons felony charges were released on bail or electronic monitoring, Milhizer says. From January through May, that rose to 67 percent, he says.

Mike Carroll, director of information services for Evans, says the sheriff’s sampling was too small to draw conclusions — 7 percent of cases that came through Central Bond Court this year through May 31.

Milhizer says judges and pretrial services officers have received extensive training on the new system but that it doesn’t replace sound judgment. He points to a case in which, based on the assessment system, the recommendation was to release a man charged with gun possession even though he was accused of pointing the gun at someone and pulling the trigger twice, without it firing. Despite the recommendation, the judge set a significant bail, Milhizer says.

Cara Smith. | AP file photo

Cara Smith. | AP file photo

Cara Smith, Sheriff Tom Dart’s policy chief, said, “It is unfortunate that the chief judge’s office would disregard the objective findings of this report, which could be tremendously helpful in improving the bond court process and improving public safety.”

The sheriff’s office did bail comparisons of defendants charged with similar crimes. For instance, two drug defendants had risk levels that called for release from jail. A 28-year-old white man suspected of cocaine possession was freed on a no-cash bond while a 33-year-old black man was ordered held on $50,000 cash bail for possession of two Viagra pills.

The study also examined how judges’ decisions varied with the risk assessments for certain crimes. In gun-possession cases, the assessments called for defendants to be jailed without bail in about 5 percent of cases but judges set bail for 98 percent of those suspects and freed the rest on electronic monitoring, according to the study.

Most of those gun-possession cases involved “D bonds,” meaning the defendant was required to post cash to be released. About 20 percent of them were able to post bail and were freed pending trial.

Gun possession was the charge for which most defendants were released on bail, according to the sheriff’s study.

In comparison with gun offenders, only 2 percent of those charged with shoplifting were able to post bail, the study found.

In her June 23 speech, Burke said, “Judges — apparently distrustful of the information provided to them by pretrial services — refuse to allow eligible individuals to be released on their own recognizance and, instead, continue to require large cash bonds, even for relatively minor, nonviolent crimes. But even low bonds are too large if you are poor.”