Jeff Jacoby

The Justice Department's earlier recidivism study, though organized and presented differently, came to similar findings. It found that 67 percent of former inmates released from prisons in just 15 states had been rearrested for at least one serious new crime within three years. Those included, the bureau noted, "2,900 new homicides, 2,400 new kidnappings, 2,400 rapes, 3,200 other sexual assaults, 21,200 robberies, 54,600 assaults, and nearly 13,900 other violent crimes."

In April 2011, meanwhile, the Pew Center on the States issued its own report on recidivism. Its conclusion: "More than four out of 10 adult American offenders … return to prison within three years of their release."

Such recidivism rates are terrible. It is heartbreaking and alarming that so many criminal offenders emerge from prison only too ready to offend again. Too many inmates come out hardened and more antisocial than they went in. For decades, scholars, policy makers, social workers, and public-safety experts have searched for the holy grail of rehabilitation and effective sentencing that would give us a more humane corrections network — one less congested, less expensive, less unfair, and less of a revolving door for the addicted and the unstable.

The problem with holy grails is that they are more easily sought than found. Bill Keller of The New York Times recently described a range of seemingly promising strategies for addressing what has become "the hottest subject in criminal justice," the US system of mass incarceration. Among them: Easing mandatory-minimum sentences and three-strikes laws. Diverting nonviolent drug offenders to specialized courts focused on treatment. Counseling for inmates about to be paroled. Repeal of rules that bar felons from getting many kinds of occupational licences.

But would they work? Prison reform and rehabilitation programs have been earnestly advanced for decades, but that holy grail remains elusive — and recidivism remains sky-high. American sociologist Robert Martinson made waves 40 years ago with an influential essay that concluded: "With few and isolated exceptions, the rehabilitative efforts that have been reported so far have had no appreciable effect on recidivism." Decades later, reformers are still trying to figure out what works.

Prison is awful, there's no question about it. It doesn't turn criminals into model citizens. It can't be expected to cure dysfunction whose roots go back to a broken home or a lousy school.

But one thing we know prison can do: It can isolate criminals from society, and thereby make society safer. In the 1980s we began locking up more convicts for longer terms. Now we have the largest prison population on earth — and crime rates at 30-year lows. When it comes to crime and punishment, there's always a trade-off. At least until we find that holy grail.

Jeff Jacoby

Jeff Jacoby is an Op-Ed writer for the Boston Globe, a radio political commentator, and a contributing columnist for