Friday, February 10, 2006

Improper access to Texas criminal records common, unchecked

Here's a new report (pdf) from the state auditor on the Texas Criminal Justice Information System, and the media summary. Of entities using the system's web interface to access criminal records from the Texas Department of Public Safety, 26 percent had unauthorized users registered, typically employees who no longer worked for the entity but retained their passcodes and access. Also, many improperly shared usernames and passwords among employees. In the management response, DPS declared that:
We have recognized that the ever increasing use of the criminal history data for licensing, employment, volunteerism, and other “non-criminal justice” purposes naturally creates a corresponding responsibility for controls over those entities. Our limited resources have prevented an adequate response to this rising need.
With one out of 20 Texans in prison, on probation or parole, this improper access opens up that information about an awful lot of people. What's more, regular readers know I think Texas restricts way too many jobs based on criminal history. Every time more employers are granted access to these records for purposes of employment screening, we reduce the chance that people who've been in trouble with the law in the past can find good jobs and make a life for themselves. So expanding access to the system harms prospects for employing ex-cons as well as risking the information being used inappropriately. Obviously state law requires DPS to share this information with a lot more folks than the agency can adequately manage.

The probation tracking system at the Texas Department of Crimnal Justice has improved, the report declared. Eleven percent of probationers' records were incorrectly tracked, compared to 46 percent when the system was last checked in 2001. That said, "improvement" is relative. With more than 605,000 probationers' flagged, according to the report, an 11 percent error rate would mean errors exist in records for more than 66,500 people.


Gritsforbreakfast said...

Of course, it would be far too drastic to say "employers should not engage in pre-employment screening." What I am saying is that state law restricts too many jobs (many hundreds)from felons, and makes information beyond conviction info available to so many people the state can't keep track of them. It's gotten far out of hand and ironically makes everyone less safe, not more so.

That's because desparate people still have to feed, clothe and shelter themselves and will do it by robbing and stealing if they can't make an honest living. So we should be careful only to exclude the most dangerous offenders from certain jobs, not all offenders from all jobs, which your comments seem to support.

Most ex-offenders want jobs and normal lives when they get out, but the system isn't designed to help them (check the links for some of my previous writing on the topic). I think we need to roll back most legal restrictions on ex-offenders employment, and stop using information beyond public data about convictions (as the report indicates, a lot of its flawed anyway) for the wide array of non-criminal justice uses cited in the report. Best,

Gritsforbreakfast said...

One more thing - I wasn't clear in the blog post and only implied it in my response above, but the DPS database includes LOTS of data besides public information about convictions that most employers use for background checks. Historically up to half that data had errors - now they're saying only 11% of records do, but that means every time it's used there's an 11% chance it's wrong. I don't think that additional data is especially probative compared to conviction data, and the high error rate makes it improper to use, IMO, for employment screening. Best,

Anonymous said...

How many of these are really "crimes", as opposed to being nothing more than contractual in nature? It is my understanding that the vast majority of the "criminal" matters are nothing more than violations of rules set forth in a contract between the individual, as a "person", and the state, as a "person".

While I have no problem with identifying those who have committed real crimes, I have a huge problem with labeling those who have managed to get themselves convicted of violating some rule, regulation, or statute which has nothing to do with what we would all agree is an actual "Bad Thing"...