Colorado: Parole Board’s Track Record Difficult to Assess

The killing of Tom Clements, executive director of the Colorado Department of Corrections (CDOC), made national news. Investigation ultimately connected the high-profile killing to a prisoner who was erroneously released from prison years before he should have been released. While the early release was attributed to court error, it put the entirety of Colorado’s parole system under scrutiny.

What has emerged from this scrutiny is a picture of inept record-keeping, at least over a period of four years which ended in December, 2009, as a result of which it is now all but impossible for auditors looking at recidivism rates to assess the extent to which Colorado’s parole board was making good or bad decisions when it granted early release to prisoners it believed would not pose a threat to the public safety.

The problem began when, in December, 2005, the CDOC stopped releasing prisoners on weekends. Rather than delay scheduled weekend releases to the following Monday, the CDOC released prisoners on the preceding Friday instead. At the same time, it began classifying those early releases as “discretionary” paroles—a label that until then had been reserved for prisoners who had been granted early release by the parole board.

The difference between the two groups of discretionary parolees is stark. The “Friday parolees” were released only a day or two early, whereas those granted early parole by the parole board were released months or years before completing their full sentences.

Because the two groups were lumped together, researchers cannot now reliably assess the soundness of the parole board’s decision-making during that four-year period. A “Friday parolee” who reoffends is indistinguishable from a board-selected “discretionary parolee” who reoffends. Both may have been released “early,” but the former, unlike the latter, does not reflect an exercise of discretion by the parole board that the early release would not endanger public safety.

Said Maureen O’Keefe, CDOC’s research director, “Once I realized how serious [the record-keeping problem] was, [I realized] we couldn’t determine whether the parole board was making good decisions.”

Still, auditors were able to glean some important information from the parole data. According to O’Keefe, excluding the four years of tainted data, discretionary releases by the parole board consistently constituted roughly 40 percent of all paroles in every year except 2011. In that year, discretionary releases constituted only 25 percent of all parolees.

Another trend to emerge from the audit is that the state’s overall prison population has been in decline for three years, declining by about 325 in 2010, by about 250 in 2011, and then, astonishingly, by over 1,500 in 2012.

The decline in Colorado’s prison population reflects a number of factors: most significantly, deliberate changes in state law over the past four years have reduced both the number and length of prison sentences; additionally, five years of declining crime rates in Colorado have slowed the rate of new prison commitments; lastly, discretionary parole by the parole board necessarily reduces the prison population.

In September 2012, the parole board began using a data-driven instrument to inform its parole decision-making. From inputs such as the prisoner’s age, the charge(s) that sent him/her to prison, previous criminal record, and infractions committed and educational programs completed while incarcerated, the instrument produces an advisory recommendation to the parole board about whether or not the prisoner should be released. In each case, it provides an estimate of the prospective parolee’s risk of re-offending.

It is too early to evaluate whether, in terms of recidivism, the parole board’s use of this new tool effectuates better results.

See: www.coloradoan.com, April 21, 2013 article.

Subscribe today

Already a subscriber? Login