William Burrell

Corrections Consultant

by Doron Taussig

Howard Cain, the man police say fired the gun that killed Sgt. Stephen Liczbinski, was once safely behind bars. He went into the prison system in 1997 to serve a nine- to 18-year sentence for robbery. Nine years later, he was granted parole at his first parole board hearing.

Obviously, releasing Cain was a mistake. But we got to wondering: Is there a better way to do this? Or is this parole thing essentially a crapshoot? We spent two minutes with William D. Burrell, formerly an associate professor of criminology at Temple and now an independent corrections consultant working with the Urban Institute on a report about parole.

City Paper: Is it just clear in retrospect that some people shouldn't be paroled, or is there a way to predict these things?

William Burrell: Part of the challenge is trying to predict future behavior. This is not an exact science. We have risk assessment instruments ... they do a pretty good job of saying whether a person is likely to offend or not. The ability of those instruments to predict violent behavior [is not as good].

Probably the best way to think about risk assessments is to look at how insurance companies set rates. Let's say you're a 17-year-old male. These people have a high rate of getting in accidents, so they're charged at a high rate. Yet there may be a 17-year-old who's never going to get in an accident his whole life.

CP: Is leniency the biggest problem with parole, or is it just cases like this one that get attention? Are there a lot of people who should get parole, but don't?

WB: The problem we face with this kind of incident is obviously it draws tremendous public and political attention to parole. But this is one case out of thousands [in Pennsylvania]. There's always the risk of making policy based on the individual case — the exception rather than the rule.