According to the paper, inmates with short prison sentences may not concern themselves with rehabilitation because they understand that they will soon be released regardless of whether they are rehabilitated.
Conversely, inmates with long prison sentences may conclude that rehabilitation is just not worthwhile since their release dates are so far off.
“The article demonstrates that an increase in the length of a sentence raises the ‘option value’ of keeping an inmate in prison,” Bernhardt said. “In other words, the value of the information gleaned from observing an inmate’s behavior while in prison is greater for a prisoner with a longer sentence. This is because information only has value when it can be used to make parole decisions, and the benefits from such decisions can be exploited when an inmate has a longer prison sentence.”
Paradoxically, pushing the benefits of rehabilitation into the distant future can discourage rehabilitation by inmates.
“Most people in prison are there precisely because of their lack of impulse control, so very long sentences can cause impatient inmates to conclude that rehabilitation is just not worthwhile,” Bernhardt noted.
Bernhard and co-researchers Steeve Mongrain, of Simon Fraser University, and Joanne Roberts, of the University of Calgary, conclude that long sentences are likely to be desirable when information is easy to “disentangle” – that is, when non-rehabilitated inmates are far more likely to be involved in prison incidents.
“When you see an inmate do something to indicate that he is not rehabilitated, you can exploit that information, and keep him in until the end of his sentence,” Bernhardt said. “But if you continue to see that he is attempting to reform – by participating in drug plans, working responsibly, abstaining from fights, taking
courses – then you update your impression to conclude that he is reformed, and exploit this new information by reducing his sentence in a meaningful way.”
For inmates with longer sentences, they must serve more time with good behavior in order to be released early, but with more time remaining on their sentences.
“For example, suppose that a single prison fight indicates that someone is not reformed,” Bernhardt said. “Then an inmate with a five-year sentence and no fights might be released after two years, with three years of their original sentence left unserved. On the other hand, an inmate with a 12-year sentence might be released after seven years, with five years left unserved.”
The key points are that the inmate with the shorter sentence and good behavior not only serves less time, but is released with less time left on his original sentence, Bernhardt says.
“This leads the state to release an inmate with a shorter sentence before one with a longer sentence when it has essentially identical beliefs about their prospects for rehabilitation,” he said.
However, if prison incidents are poor indicators of rehabilitation, long sentences could reduce rehabilitative effort. To encourage rehabilitative efforts, sentences shouldn’t be too short, but they also can’t be too long, either, Bernhardt says.
“To the extent that inmates whom the state believes are less likely to be rehabilitated receive longer sentences, this effect is reinforced,” he said.
A practical implication of that research is that although an increase in sentences may reduce crime, it can also increase prison populations, which not only leads to greater expenditures by the state, but also the possibility of less rehabilitation and correspondingly higher rates of recidivism.
“If you lock people up forever, then almost by definition, they can’t offend in the population,” Bernhardt said. “But locking people up for very long periods is quite costly; California is a prime example of that. So if the sentence length destroys incentives to rehabilitate, then recidivism ultimately may be higher.”
The paper was published in the Journal of Law, Economics and Organization.
The article, “Rehabilitated or Not: An Informational Theory of Parole Decisions,” is available online.
Phil Ciciora, Business & Law Editor | 217-333-2177; email@example.com