Louisiana prisons can't track inmates accurately, audit finds

The Louisiana Department of Corrections is not properly tracking inmate locations and release dates, in part because it uses "antiquated" data management software.

That is one of the findings of a recent report from the Louisiana Legislative Auditor. Among other findings:

The Louisiana Department of Corrections is not properly tracking inmate locations and release dates, in part because it uses "antiquated" data management software.
That is one of the findings of a recent report from the Louisiana Legislative Auditor. Among other findings:


•    The department may lose track of inmates for weeks at a time, especially if they're housed as state inmates in parish jails.
•    Data about individual inmates often is missing, incomplete or wrong.
•    Department staff sometimes calculate inmate release dates differently.

 

"Successful reform is dependent on having accurate and complete information on Louisiana's offender population," the audit report says. "DOC does not have adequate policies and procedures to manage offenders."
 

The audit report, released in October, predated the Nov. 1 implementation of the Justice Reinvestment Act, a sweeping criminal justice reform that aims to reduce the prison population by at least 10 percent and save more than $262 million in the next decade.
 

The reform, which the legislature enacted with bipartisan support, touches sentencing, early release and eligibility for probation and parole, among other things. But it has drawn opposition from sheriffs, including Caddo Parish Sheriff Steve Prator, and from state Attorney General Jeff Landry.
 

The Department of Corrections acknowledged many of the problems and promised solutions.
Among the steps it is taking: The department has hired a contractor to help determine whether a $3.6 million data management software, purchased in 2015 and abandoned the same year, can be salvaged, with an answer expected by the end of the year.

 

Data discrepancies
 

The legislature and governor enacted the state reforms in June, giving the Corrections Department the task of re-calculating inmate release dates and selecting offenders eligible for earlier release.
 

The department manages more than 35,000 adult offenders— more than half of whom are housed in parish jails. Its inmate database software dates to 1970 and last was updated in 1991, according to the audit.
 

Nineteen percent of offender files evaluated by the legislative auditor contained at least one data entry error, the audit noted. Of the files sampled by the legislative auditor, 270 were missing sentencing dates, 102 were missing sentence start dates and more than 1,000 were missing full-term sentence dates.
 

In some cases, local courts sent incorrect or incomplete data to the Corrections Department, according to the audit. 
 

"Errors in data fields limit DOC's ability to accurately identify offenders, assess their eligibility for participating in programs or earning credits for early release and calculating their release dates," the audit said.
 

The department's Secretary James LeBlanc, in a response filed with the audit, said the auditor found 26 errors in more than 10,000 sampled data points — an error rate of less than 1 percent.
LeBlanc said that manually entering data increases the potential for error, but that there is no way to automatically enter data.

 

"In our opinion, it is much more logical and efficient to enter as much data as is available into an incomplete...file than it is to simply stack offender paperwork awaiting a complete record and not have any record of the offender... at all," he said. 
 

Release dates
 

The 2017 audit noted the department's process for calculating offender release dates is "inconsistent" and that it "does not have any policies, procedures, manuals or agency-wide guidance that details how to calculate release dates."
 

Legislative auditor staff asked two different DOC employees to compute the release date for a hypothetical offender. The result differed by 186 days, according to the audit.
 

"DOC does not have a formalized, consistent method for calculating offender release dates," the audit said. "An offender could be held too long if the release date was miscalculated and not caught until shortly before release."
 

This finding provides context for the conflict between Gov. John Bel Edwards' office and Caddo Sheriff Prator in the weeks before the first early release of inmates under the prison reforms on Nov. 1.
 

Prator sparked both controversy and concerns about what exactly would happen on Nov. 1 initially said the reforms would result in nearly 200 violent felons being released back to the parish. The governor's office put the number at closer to 35 nonviolentoffenders.
 

Soon after, the governor's office said Prator had calculated release dates incorrectly. Prator acknowledged the error, although a spokeswoman said that the corrections department had provided numerous lists of inmates to be released on Nov. 1 and that the number and names of offenders slated for release were in almost constant flux.
 

Cindy Chadwick, the sheriff's spokeswoman, said more recently that the sheriff's office had problems with corrections department files even prior to the Justice Reinvestment Act.
 

"Caddo Correctional Center staff checks each inmate for holds, detainers or warrants... before their release," Chadwick said in an email. "We frequently find inmates who still have pending charges prior to their release."
 

LeBlanc, in response to the audit, said calculating release dates is "very complex" and involves up to 20 variables, including parole revocation recalculations, length of sentence, credit for time served, good time release provisions, good time credit lost due to behavior and credits earned for certified treatment or rehabilitation programs. 
 

Legislative action also presents complications, LeBlanc said.
 

"Each year, as legislation is passed, our entire staff must be retrained to learn how the new laws affect their work while still applying old statutes that remain applicable," he said. 
 

Time computation is an entry-level job, in a staff area that had 33 percent turnover last year, LeBlanc said. Supervisors check, on average, more than 1,300 release cases among the average 4,213 offender files reviewed each month.
 

While reviewing every time computation is "not a reasonable expectation," LeBlanc said, he pledged that supervisors will check all time computations by "less experienced" staff.
 

"Of the errors reported, only one impacted the actual time computation of the offender," LeBlanc said, "and it would have been discovered and corrected as part of the review at discharge."
 

Prator, in an October interview, said his questioning of the correction's department lists resulted in stopping the early release of at least 10 offenders because of other charges or warrants from other jurisdictions.
 

Chadwick said the sheriff is open to improving a system that Edwards has repeatedly said is not working.
 

"Regardless of normal operations or Justice Reinvestment, improvements in calculations of release dates and criteria can always be made and are welcomed," she said. 
 

Inaccurate offender locations
 

Another result of the data discrepancies is that the corrections department does not always know where its offenders are, especially when state offenders are housed in parish facilities, according to the audit.
 

Policy requires the department to approve transfers before they occur. It requires parish jails to report transfers to the state.
 

But because there is no formal time frame for that communication to occur, some parish jails report the transfer "the same day" and some "do not notify DOC at all," the audit noted.
 

The audit found that, on average, it took 22 days following a transfer from a parish jail to another facility for the state to update its accounts.
 

"This means that for 22 days, DOC did not know the location of a given offender," the audit said.
In one case highlighted in the audit, an offender was transferred 22 times in four years and often spent less than a month in a local jail before another transfer.

 

Eleven percent of files sampled by the legislative auditor found offenders were in different facilities than recorded in the inmate database. One case cited in the audit involved a offender convicted of second-degree murder who the database showed was housed in the Evangeline Parish Jail for 166 days — when he was in a facility 100 miles away.
 

"With the approximately 3,500 offender transfers per month at the local level, we must acknowledge the margin for human error," LeBlanc said in response to the audit.
 

He added that the existing practice "has not hampered our ability to meet our mission and has not cause an increased risk to public safety."
 

LeBlanc said offenders are accounted for at the local level, per agreement with sheriffs. He also said the DOC updates offender files as soon as the department receives notification but is "dependent" on the 104 local jail facilities for timely communication.
 

Ken Pastorick, spokesman for the Department of Corrections, said that in response to the audit, the department will require parish authorities to inform the department 24 hours before any transfer. The new requirement goes into effect Monday, Nov. 13.
 

Stakeholder reaction
 

Louisiana Attorney General Jeff Landry said the audit report reiterates concerns he has "voiced for months."
 

"The Governor rushed his dangerous reform package, putting our State's public safety in jeopardy," Landry wrote in an email. “The Governor is releasing criminals onto our streets before they are properly vetted and in spite of fears by their victims."
 

Landry also voiced concern that the governor's cost-saving claims are false and that expanding probation and parole eligibility without proper funding will create further risk.
 

"By putting the cart before the horse, the Governor continues to enact policies that don't make our State safer or benefit our hard-working taxpayers,” Landry said.
 

Shauna Sanford, a spokeswoman for the governor, said Edwards has spoken with LeBlanc and has received assurance "that the department has already completed several initiatives in the report and is taking the findings very seriously." 
 

"While careful consideration is being given to each one, it is important to note that the agency has not been hampered in its ability to meet is mission," Sanford said in an email.
 

Overall implementation of the Justice Reinvestment Act has been "smooth," Sanford added.
"This is something we have prepared for over the last several months," she said."We anticipate the releasing offenders will work with their probation and parole supervisors to transition successfully back to the community... (but) as with offenders, there is always the potential for them to reoffend."

 

Sanford said the offenders released on Nov. 1 would have been released within coming months anyway. She said that, because of the reforms, inmates received 50 to 100 hours of reentry programming.
 

"What was in place has not worked and led to fewer re-entry programs, more recidivism and more victims of crime," she said. "The (Justice Reinvestment Initiative) is based on decades of evidence-based research and nearly a year of study and input from a wide range of Louisiana businesses and a broad spectrum of advocates, community and political leaders the governor believes that this is an opportunity to effect real change for our citizens and communities."   
 

Recommendations, responses
 

In his response to the audit and 189 pages of supplemental documentation, Corrections Secretary James LeBlanc noted that the department has either addressed the audit recommendations or is taking steps to do so.
 

Include a timeframe for reporting offender transfers: The Department of Corrections agreed to revise its Basic Jail Guidelines.
 

Consider expanding billing audits to cover more than 7 of 114 facilities: As of August 2017, the DOC will complete a quarterly payment audit for at least 50 percent of its facilities.
 

Include local facilities when reviewing offender files for quality assurance: Six pre-class groups will review an average of at least 25 cases each month.
 

Establish procedures for using and monitoring reports to detect errors where edits can't be entered: The DOC will explore additional edits under the guidance of probation and parole staff
 

Develop formal policies and procedures for calculating offender release dates: Each specialist has a guide for time computations but the DOC will train all new hires and hold yearly in-service reviews following legislative session.
 

Establish procedures for monitoring, revoking and changing employee access to offender data: The DOC will develop a written policy and the Office of Adult Services and Staff will manage access.
 

Establish policies and procedures for making and monitoring changes to the CAJUN system: A written policy will be created for changes currently made using a help ticket database.
 

Develop a formal plan about whether the $3.6 million Offender Management System can be saved: The DOC is contracting with a vendor to see if the OMS can be re-instated.

© Gannett Co., Inc. 2017. All Rights Reserved


JOIN THE CONVERSATION

To find out more about Facebook commenting please read the
Conversation Guidelines and FAQs

Leave a Comment