Wednesday, December 16, 2009

Prosecutors' performance hard to measure: Excellent reporting out of Waco shows gaps in DA data

The McLennan County District Attorney is facing his first serious electoral challenge in a generation next year, and the Waco Tribune-Herald offered up excellent coverage this week that are must-reads for anyone concerned with how prosecutors operate or what issues voters should care about in elections:
(UPDATE: Links now dead: See excerpts at the Stand Down Project.) I may need to change up my Google News alert keywords because if it weren't for Doug Berman pointing them out at Sentencing Law & Policy, I wouldn't have seen these stories. Both are must-reads; the first one in particular, even for readers with no stakes in McLennan County justice.

The stories were inspired because a local officers' union endorsed incumbent John Segrest's opponent, claiming the sitting DA refused to file too many cases brought to him by local police. Cindy Culp at the Trib analyzed data from the McLennan DA's office and counties with comparable populations and discovered that most DAs do not wind up pursuing around half the cases brought to them by police, both in Waco and other jurisdictions that produced viable data. Here are some topline numbers:

By these data, the McLennan DA actually does a terrific job at clearing out cases that don't need to be prosecuted early in the process. They're prosecuting cases brought to them at about the same rate as other counties overall, but in Waco prosecutors are much more likely to refuse the case up front instead of filing a bad case and dismissing it months down the line after the county has paid for jail costs, indigent defense, etc.. I'd made this same point the other day regarding jail overcrowding in Nueces County:
A big key to processing cases more rapidly is for prosecutors to assess cases brought by officers as early in the process as possible, not after a defendant has been sitting in jail for hours, days or (worst case) even weeks. After all, why jail people who you're not going to charge? Harris and El Paso's DA's have implemented "direct filing" systems that get prosecutors involved much earlier in the process. In Harris County, police must get every arrests pre-approved by prosecutors who are on call 24 hours a day, 365 days per year to evaluate cases. (See Murray Newman's description of Harris' system here and here.) Getting prosecutors involved earlier is one of the smartest approaches I've heard to adding a little giddyup to the system. It solves or prevents a lot of other problems.
Culp's data add even more oomph to that critique - if up to half of cases brought to prosecutors are refused or dismissed, then delays in getting to that result take up a lot of extra time and resources in the system, creating problems for the jail, the courts and the county coffers.

There may be other reasons to critique Mr. Segrest's performance, but by these data he's doing a remarkably efficient job compared with his peers on identifying which cases need to be prosecuted.

Culp's report also points out the frustrating lack of data avaialable for evaluating prosecutors' performance, and for my money state legislators would be well-served to require formal reporting on some of the outcome measures where she sought information but couldn't consistently find it across jurisdictions. That opacity is a big reason why, as one source she quoted pointed out, incumbent American prosecutors are about as likely to lose reelection fights as were officials in Communist Russia.

Bully for Culp for putting this series together: One of the best pieces of reporting I've seen on a complex, nuanced subject.

5 comments:

Anonymous said...

What a terrible way to evaluate prosecutors. One criminal episode with one defendant may generate multiple charges. If the prosecutor accepts some charges, and refuses others, then how can you say that's a good performance measure (which happens frequently).

Gritsforbreakfast said...

It's just one data point, and the article IMO did a good job of coaxing out the nuance you describe. Certainly, though, if you're not going to wind up prosecuting half the cases, it's better to get rid of them early in the process before the county incurs incarceration costs, indigent atty fees, etc..

I'd encourage you to read both pieces if you haven't. She did a better job than me of splitting some of those hairs and evaluating what metrics are and aren't useful.

Brody said...

As a prosecutor in a small office, I'm always looking for ways to judge my performance. This is one that I hadn't particularly considered (I've mostly been looking at our clearance rate and time-to-clearance), but will now be added to my year end review process.

Any other metrics that those in this community would like to see their local DA/CA generate? In a perfect world, what would you want to know?

Anonymous said...

Whatever the problem is, Reyna ain't the one to fix it.

The Team said...

Hey Grits, recently while on the topic of plea bargaining, a former "career prosecutor" [his words] from Harris County commented on a post at Simple Justice re: "Plea Bargaining 201". He'd shared that he had witnessed three types of cases go to trial during the early 80’s,90's & early 00s. The very solid, very strong & very close.

He then claims that the remaining 95% of cases that were too weak to take to trial were "plead out". His personal experience could single-handedly explain one of the reasons why we’ve been enjoying overcrowding for decades at taxpayers expense. What a joke and it is on us. Unless another ADA from the same county and era sees it different?

No matter how one looks at it, by his measurement we see that the police obviously have arrest quotas that result in "round-ups" and by the ADAs refusal to simply release, they have clogged up the system. Meanwhile the 95% are constantly being placed on probation or they are in the county jail waiting to go to state jail or prison and no one knows how many are truly Not Guilty. (Not – in no way; to no degree).

We thank you for bringing us this very useful post.