Work-release programs that allow Louisiana convicts to hold private-sector jobs as they near the end of their prison terms are beneficial to both the state and the inmates, but the Department of Public Safety and Corrections needs to do a better job of overseeing the programs, according to a new report by the state’s legislative auditor.

Such programs have been controversial, mainly because some contracts to run the programs have gone to political insiders and escapes of participants are common.

Also, while the programs are aimed at helping inmates save some money before they’re released, operators are allowed to keep the lion’s share of the inmates’ wages. The performance audit released Monday said that 55 percent of inmates leave their work-release program with less than $1,000 in the bank.

The report also confirmed that escapes are frequent — at least 254 over the three-year period reviewed — leading the auditors to recommend more active oversight by the department and by work-release operators themselves, including possible use of electronic monitoring.

Two of the programs that had the most reported escapes, including a controversial program based in Covington with ties to St. Tammany Parish Sheriff Jack Strain, have since been closed, and the number of escapes overall has fallen in each of the past few years. But the program with the most documented escapes, run by the West Baton Rouge Parish Sheriff’s Office, remains open.

Despite the various red flags, the audit found that Louisiana’s work-release system has a range of positive impacts overall. For starters, the programs save taxpayers roughly $12 million annually in incarceration costs because the state pays a lower daily rate to house offenders in such programs than it spends on those in traditional jails.

The programs also have been shown to reduce recidivism rates slightly, according to corrections officials, thus saving the state money and improving public safety. According to state data, 39.6 percent of inmates who go through work-release programs will be arrested again within five years, compared with 41.3 percent of inmates released from state prisons and 44.9 percent of those released from local prisons.

Inmates benefit from work-release not only because they are less likely to commit new crimes, the report said, but also because they tend to acquire at least some savings.

Employers who hire work-release participants also are winners: Not only do they get tax breaks for hiring offenders, but they don’t have to offer them as many benefits as regular employees, the report notes. Moreover, offenders are typically punctual and reliable.

Work-release also has been a major boon for those who run the programs — a group that includes many Louisiana sheriffs and a handful of private contractors. The report says operators took in approximately $55 million in revenue in the most recent year studied, the bulk of it coming from the 64 percent share of inmate wages that the overseers are allowed to keep. Operators also made $4.1 million from selling snacks and other items to inmates.

That commissary spending — an average of more than $1,300 per year per inmate, according to the report — is one reason so many inmates accumulate so little money in the bank. Auditors recommended putting caps on how much inmates can spend.

Of 38 work-release programs in the state, about two-thirds are run by local sheriffs, with the balance being run by private firms. Those firms handled almost half of the roughly 3,000 offenders in the program at any given time, and they took in about half of the $55 million the programs brought in.

Easily the biggest of the firms is Louisiana Workforce LLC. It is owned by Paul Perkins, a former business associate of Corrections Department Secretary Jimmy LeBlanc and former Angola Warden Burl Cain — a fact the audit does not highlight. The firm houses nearly a third of the state’s work-release inmates, according to the report.

One major question the report did not take on was whether work-release tends to work better under public or private management. Some, including LeBlanc, have argued that private operators are more apt to find inmates high-paying jobs because it increases their own profits. The sheriffs have no such incentive.

“In government, you’re gonna get the same pay no matter what,” LeBlanc said in a 2014 interview. “Why do I need to find (better) jobs if I’m gonna get paid the same amount of money? So I think privatization, given the right operator, is not a bad thing.”

But data collected by the auditors suggest that inmates in private work-release programs don’t actually make any more money than those who work under sheriffs. In fact, on balance, the sheriff-run programs collected about 5 percent more from inmate earnings than did the private ones, according to the audit.

Another finding of the audit was that work-release programs offer very little in the way of rehabilitation programming compared with state prisons or local jails. State prisons on average offer 61 rehabilitative programs, compared with seven in a typical local jail and three in a work-release program.

That’s not necessarily surprising: There is no financial incentive for work-release operators to provide such courses, and the state doesn’t require them to do so, the report said.

In fact, while the report does not mention it, offering such programming can actually hurt an operator’s bottom line because inmates who complete the programs become eligible for early release.

The report says the state should require more rehabilitation programming, and that recommendation was one of only a few in the report that LeBlanc endorsed in his response.

Because work-release is generally beneficial, the report says the state should maximize its use, starting with finding more inmates to participate. If all available slots in the programs were filled, the audit says, the state would save another $7 million.

However, LeBlanc disputed auditors’ methodology in computing that figure, saying there are fewer empty beds in work-release programs than the report found.

The audit said inmates in work-release are transferred frequently and correctional officials don’t always know where they are, another finding disputed by LeBlanc.

The audit also said the programs did a poor job of collecting court-ordered restitution from inmates, including more than $5 million owed to crime victims.

LeBlanc responded that the department can collect restitution only when it’s listed in a “specific judgment detailing how much is owed and to whom.” Auditors said they knew of the restitution only because it was detailed in such judgments.