http://www.uncoverage.net/2011/05/1500-prisoners-released-in-california-with-no-supervision-450-very-dangerous/1500 Prisoners Released in California with NO Supervision, 450 “Very Dangerous”
Posted on May 26, 2011 by Jane Jamison| 1 Comment
California, once again, the poster for the utter failure of progressivism. Liberals spend billions and years spinning the latest social craze, but they can’t provide the basics of government: Good schools, infrastructure, and public safety. California is becoming a very dangerous place, thanks to elected liberals.
California is already reeling under the mandate of this week’s Supreme Court ruling which is ordering the release of at 33,000 to 45,000 inmates due to “inhumane” prison overcrowding. As we reported this week, Governor Brown is promising the public will be protected, and that most of the prisoners will be housed instead in country prisons. How can he make that promise when the state is running a $26 billion deficit and part of his budget “cuts” involve transferring inmates to local jurisdictions, many of which have overcrowded jails and tight, over-drawn budgets?
Even without turning loose 46,000 prisoners in the state, California has demonstrated over and again that it cannot perform the basic services required by government.
The state of California quickly paid child-kidnap victim Jaycee Dugard a $20 million settlement last year, without a trial. It was shown that local police and parole officials were abysmally negligent in their duties by ignoring that a convicted sex offender was keeping Jaycee in a shack in his backyard for 18 years.
Today, we learn that the state has mistakenly freed hundreds of dangerous felons, thanks to a flawed computer program put together by one of our “esteemed” UC schools.
“More than 450 violent criminals were placed on unsupervised parole in California thanks to a flawed automatic risk assessment computer program developed by UC Irvine researchers, according to a special report by the state Office of the Inspector General.
The risk assessment was supposed to help a new state program select parolees to go without supervision after their release from prison. The program was part of a broad California Department of Corrections and Rehabilitation (CDCR) effort to reduce the number of parolees who are returned to prison on technical parole violations, which in turn would reduce prison overcrowding.
The Inspector General, however, found that over a seven-month period in 2010, the computer program was wrong nearly 24 percent of the time. About 15 percent of the more than 10,000 inmates placed on unsupervised parole from January to July 2010 should not have received that privilege, the report concludes.
In fact, “Over 450 of these ineligible offenders carry a high risk for violence, and some of these ineligible offenders may have already been discharged from non-revocable (unsupervised) parole after completing 12 months of parole, thereby precluding CDCR from taking action to correct the parolee’s inappropriate placement on non-revocable parole,” the report says.
The Office of the Inspector General says the error rate is caused by the computer assessment not taking into account information about prior supervised parole violations, and does not consider certain data about juvenile offenses.
UCI officials said the report is deeply flawed.
“I am deeply disappointed that this revised version of the report contains many of the same inaccuracies and poorly crafted science as did the earlier version, even though I brought these errors to the attention of your office,” wrote Susan Turner, director of UCI’s Center for Evidence-Based Corrections, in a letter to the Inspector General’s office (which you can read in its entirety here). We get into UCI’s objections in greater detail below.
INSPECTOR GENERAL SAYS….
The report found that the primary database the prison system uses to assess inmates contains incomplete data, and that some assessments that should have been scored manually were instead scored by machine.
The computer assessment program uses a algorithm developed by UCI to produce a risk assessment score. As the report states, “The algorithm searches the electronic criminal history data of each offender for convictions, which are identified as ‘score-able events.’ This data is then fed into a (risk-assessment) calculator. The resulting score for each offender is made available to staff through a … link on (the prison system’s) computer network.”
The report says about 96 percent of inmates scores can be calculated using the computer system, but the rest must be figured manually, because they include criminal history data from other jurisdictions outside of California.
In her letter, Turner said that the Inspector General’s analysis was inaccurate in two major respects.
“First, investigators essentially compared ‘apples’ to ‘oranges’ in their exercise when they compared their own manual scoring of hard copy rap sheets with the automated CDCR computer algorithm used for the vast majority of offenders under the supervision of CDCR,” she wrote. “The discrepancies identified by OIG were largely caused by OIG staff using ‘old’ scores generated by an out-of-date scoring algorithm to compare with their hand scored calculations.
“Second, and perhaps more disturbing, the OIG manual coding considered factors that are not programmed into the current automated version of the tool. Based on these ‘extra’ factors, OIG claims that its coding is not only different from, but also more accurate than, the CDCR tool. While OIG’s manual coding is different from CDCR’s automated coding (because OIG’s coding includes additional factors), there is no evidence that OIG’s manual coding is more accurate than CDCR’s automated coding. The additional factors that OIG inserted into its own manual coding have never been tested to verify their accuracy – either by the OIG or by CDCR. The conclusion in the report – that OIG’s manual coding is more accurate than CDCR’s automated coding — is nonsensical because it amounts to claiming accuracy for a procedure which was never tested for accuracy.”
The Inspector General’s second finding — that criminal history records typically are incomplete — “is a limitation of which researchers and criminal justice professionals are well aware of,” Turner wrote. “However, ‘rap sheets’ are the standard used by the justice system and there currently exists no alternative, as the report itself notes.”
And the Inspector General’s third finding – that juvenile data was handled inconsistently – “relates only to the miniscule number of juvenile records (much less than 1 percent) that are contained in the records. I have discussed this issue repeatedly with investigators from OIG. I have explained that the tool counts adult convictions on adult charges in adult courts. If a juvenile has an adult conviction on an adult charge, it is counted. Otherwise, it is not counted. This is a simple and concise rule that is consistent with Penal Code 3000.03. Therefore, OIG’s report is incorrect.”
The state corrections department continues to refine and test the tool to increase its accuracy and make the most appropriate decisions about offenders, Turner wrote.
“We agree that this tool, like any tool, can be improved. An evidence-based dialogue that accurately reflects the tool and possible ways it, and its use, might be improved, could benefit the CDCR, our State and our citizens. However, poor methods, suppositions and bad logic masquerading as evidence that this tool is flawed helps no one and only wastes scarce resources,” she wrote.
California Watch reports that the correction’s departments “top brass” were not pleased with the report, saying it “goes beyond its relevancy; in fact we must also take issue with its facts and central conclusion.””