Pretrial release program uses discredited gauge of success

Posted Monday, Aug. 06, 2012  comments  Print Reprints
A

Have more to add? News tip? Tell us

Last of two parts

How many criminal defendants released from jail on cut-rate personal bonds skip out on court dates?

Leah Gamble, Pretrial Release Services manager at Dallas County, is an open book.

"Our no-show rate is high," Gamble said. Almost 30 percent of such Dallas defendants have missed a court date, she'll tell you.

Then, she takes a look at Tarrant County's rate of defendants who dodged court. Her eyes bug out.

It's just under 3 percent -- to her, that would mean that, out of 100 defendants, only three played hooky.

"How can that be?" she said. "That's the lowest of most of the places I've known."

According to Tarrant County Administrator G.K. Maenius, the low number is a sign that Tarrant County is working overtime to chase down suspects who abscond.

"It's pretty much that our flight risk is fairly minimal," Maenius said. "Even the ones who don't show up, it's not like they've skipped the country and gone to Mexico.

"They just decided not to show up and eventually they will be in front of that judge."

But that doesn't necessarily tell the whole story.

Tarrant is using a discredited method of calculating how often defendants out of jail on pretrial release fail to show up for court dates, a Star-Telegram examination found.

The method creates the illusion of success while misleading judges about the rate of no-shows, say state and national directors of similar pretrial programs.

"Politicians like to do it. It looks good; it looks like the program is more effective," said Dennis Potts, assistant director of the Harris County Pretrial Program.

"But just tell the truth. Somebody tell the truth; somebody just tell the truth."

It's impossible to determine the truth about Tarrant's pretrial release program.

The county does not reveal how many suspects' cases are resolved or how many defendants don't show up or otherwise fail to meet the terms of their release.

State law requires that pretrial programs provide a no-show rate for defendants.

The failure to appear, or FTA, rate reflects the competency of pretrial programs and gauges whether defendants are adjudicated.

But the law doesn't specify how the rate is to be calculated.

Harris, Travis and Dallas counties use what's seen as a best-practice method, recommended as far back as 1979. It is the same calculation the courts use to measure bond forfeitures of surety companies whose criminal defendants don't show up, Potts said.

If a defendant misses one court date, that counts as one failure to appear.

So, if 10 defendants miss a court date out of 100 defendants who have court appearances, that's a 10 percent failure to appear rate. (This method is known as the defendant-based rate.)

"It's the most conservative number and therefore the largest number when you're looking at a defendant-based rate. It's a lot more honest," Potts said.

By that calculation, Harris County says its failure to appear rate is just under 9 percent. Travis says its rate is around 11 percent.

Tarrant's rate is tiny, but that's because of the difference in the way it's calculated.

Michelle Brown, director of Tarrant County's Pretrial Services, divides by the total number of court appearances instead of the number of defendants, which makes for a smaller rate.

For example, if there are a combined 10,000 scheduled court appearances, and 100 of those were missed, her FTA rate is 1 percent.

The method is flawed, Potts said, with weaknesses that bail bondsmen, in particular, like to point out.

"It's the way the numbers work," Potts said. "If you divide any number by a larger number, you're going to get a tiny percentage. ... Obviously you have more appearances than you have people."

With its calculation, Tarrant is swinging open the door to criticism of pretrial release programs, he said.

Brown says the county's outdated computer system doesn't sort data on personal bonds, so it would be difficult to do it any other way.

"One of the things they're trying to do across the nation is have every program calculate it the same, but every program doesn't have the technology to do it the same," she said.

"That may happen, but at this time there are a lot of programs that use this."

Brown says she doesn't plan to change her method or provide any other numbers.

The public can't calculate a defendant-based rate for Tarrant because the county's annual reports don't disclose the total of defendants who leave the program either through success or failure. Nor do the reports include the disposition of those cases.

"This has been the way that we've done it," she said. "It works well for us."

A lack of data

Tarrant's failure to appear calculation is one example of how the pretrial release program lacks data to help gauge success.

County and district criminal court judges in Tarrant County say they depend on the pretrial staff to report violations of bond terms. Those could include when a defendant in their court commits a new offense while on pretrial release.

And lacking information about the total of defendants who exit pretrial, judges say they can't tell whether their decisions about pretrial release proved good or bad.

"Since the judges do not manage pretrial release, we do not have access to data that would allow us to analyze whether current procedures provide enough necessary information to make good decisions about who should receive a pretrial bond," judges said in a written statement to the Star-Telegram.

Judges do take some actions to address failure to appear by defendants released through the pretrial program.

Judges can revoke personal bonds, and between Jan. 1. 2011, and March 8 this year, they revoked about 80 personal bonds, records show.

They can also declare the bonds insufficient, which puts a hold on them for an indefinite period. Since 2009, hundreds of personal bonds have been declared insufficient, county records show.

Some personal bonds are, indeed, forfeited. Only those bonds allow the county the automatic ability to pursue collection of the unsecured bail amount. Over the past two decades, the county has been unable to collect hundreds of thousands of dollars from forfeited personal bonds under the pretrial program, county records show.

But judges stress that they could do a better job if they managed pretrial and had access to more data about cases.

Gamble doesn't let technology shortcomings stand in the way of her calculations. Although it is a barrier, she said, she tracks every defendant in the Dallas program from start to finish, by hand.

"We have terrible data management challenges," she said. But "the law says you have to account for individuals who missed court dates. It specifies that."

In Travis County, judges run the pretrial program.

Irma Guerrero, manager for the Travis County Pretrial Services Program, says her office has to cross-check its data with other court coordinators who enter data for individual criminal court judges.

"It becomes complicated to pull accurate data," she said.

But it's her priority, she said.

Years ago, when Travis County struggled with data collections, the pretrial staff met with judges, clerks and court coordinators to solve the problem, Guerrero said.

"We said, 'You know we have to pull data and we have to be on the same page,'" she said. "That was our priority, to make sure data gets entered.

"Even now, we're still trying to refine our reports, and make it as accurate as possible," she said.

Yamil Berard, 817-390-7705

Twitter: @yberard

Looking for comments?

We welcome your comments on this story, but please be civil. Do not use profanity, hate speech, threats, personal abuse or any device to draw undue attention. Our policy requires those wishing to post here to use their real identity.

Our commenting policy | Facebook commenting FAQ | Why Facebook?