If I could speak to that, Professor Paul Gendreau of St. Thomas University prepared a meta-analysis of studies on the effect of prison terms on recidivism for Public Safety and Emergency Preparedness Canada. He examined every study in North America over the last fifty years that met the criteria of being proper research. What he found was that there was not one study—not one—showing that longer terms in prison reduced recidivism. Generally there was very little correlation at all, but the degree to which there was a correlation, longer terms led to higher rates of recidivism—essentially confirming the notion many of us who work in prisons have that they are in themselves criminogenic environments.
If you take people who typically get conditional sentences who aren't criminalized in their lives and for many it's their first offence, and you put them through a process that first addresses the issues they have—often there are addictions and so on, especially alcoholism—you have much better potential in the long term than putting them into an environment that's inherently criminalized.
Then you have to go through the very difficult process of reintegration after a prison term. After you've been in prison, you've lost your community roots, your job, and your place in society. Coming back is very difficult, and none of those things increase the likelihood of success.