Doris Layton MacKenzie, Ph.D., is Professor in the Department of Criminology and Criminal Justice at the University of Maryland and Director of the Evaluation Research Group. She has an extensive publication record on such topics as examining what works to reduce crime in the community, inmate adjustment to prison, the impact of intermediate sanctions on recidivism, long-term offenders, methods of predicting prison populations, self-report criminal activities of probationers and boot camp prisons.
You’ve spent 20 years studying juvenile and adult correctional boot camps. How has your thinking evolved about them?
I started in the 1980s with a study of a boot camp in Los Angeles. At that point I was very skeptical about whether they worked. As I started interviewing staff and inmates, I found that there were some very strong relationships developing between staff and inmates. My research showed that boot camps didn’t have an impact on recidivism, but there did seem to be a positive atmosphere in the program. I found the same thing when I looked at boot camps nationally. Also, some states used them to meet other goals, such as reducing prison overcrowding in New York. There were a lot of negatives as well, including the lack of impact on recidivism and problems with staff training that have led to injuries and deaths in some camps. So on balance, I think maybe it’s better not to have these programs.
How has that message been received?
People are usually upset to hear it. A common thing I hear is, “my boot camp is different.” Part of the problem is that the people I speak with, including politicians, aren’t trained in social sciences. They’ve learned a case study approach in law that looks at past precedent. They tend to ask about particular programs they’re familiar with, while I talk about program results overall.
Has there been an impact on boot camps?
Boot camps aren’t as popular as they were ten years ago. There aren’t many new programs opening. I think it’s because there’s a lot of good research out there that shows that getting tough is not enough, so the field has moved away from these types of programs.
Ten years ago you helped write a high-profile report, submitted to Congress, about what works and what doesn’t in criminal justice policy. What has the impact of that report been?
I think it’s been very well received. The biggest impact is that practitioners think more about research now. Our challenge in writing the report was where to draw the line to say something is effective. It’s very difficult, because if we only used random assignment studies, we wouldn’t have very many to choose from. So we drew the line relatively low in terms of science. The lead author on the study, Larry Sherman, really pushed us to draw more conclusions from the research so that our findings would be useful to policymakers. As a result, people really liked it because we were clear about our findings, which researchers sometimes have a hard time doing. Another advantage of the report is that we showed exactly how we reached a decision about whether something worked or didn’t. That gave our readers a way of finding out exactly why we made the decisions we made.
Are there other important rules for researchers to follow?
It’s very important to be clear on what measures you’re using as examples of effectiveness. There are measures other than recidivism that are important, such as reducing prison overcrowding. Another challenge is making sure to use legitimate criteria for effectiveness. A lot of drug treatment literature only looks at people who complete programs and leaves out people who drop out. If we only looked at completers in boot camps, they would be phenomenally successful!
Do criminal justice programs do a good job of learning lessons from past failures?
I have some concerns about that. For example, I think reentry programs are making an error when they talk about giving offenders services like jobs and housing without first targeting something within the person. Just giving people services doesn’t seem to be justified by the research. There needs to be a change model in place.