Why America imprisons so many

Writers: 30 years ago, prosecutors took over from judges controlling sentencing They say this led to vast disparities in sentencing and many more people imprisoned Writers: It ignores that federal judges are experienced and know to exercise discretion Writers: Many prosecutors have little experience deciding who will go to prison and for how long

Editor's note: Mark W. Bennett is a U.S. District Court judge in the Northern District of Iowa. Mark Osler is a professor of law at the University of St. Thomas in Minneapolis and previously served as a federal prosecutor in Detroit. Bennett was the sentencing judge and Mark Osler the lead counsel for the defendant in the 2009 case of Spears v. United States, in which the Supreme Court supported Bennett's reasoning in holding that sentencing judges could "categorically reject" harsh sentencing guidelines for crack cocaine.

(CNN) -- Nearly 30 years ago, Congress embarked on a remarkable and ultimately tragic transformation of criminal law. Through the establishment of mandatory sentences and sentencing guidelines, discretion in sentencing was shifted from judges to prosecutors.

After the changes, prosecutors largely controlled sentencing because things like mandatory sentences and guideline ranges were determined by decisions they made.

This change ignored the fact that federal judges are chosen from the ranks of experienced members of the bar precisely because their long legal careers have shown the ability to exercise discretion.

Mark W. Bennett Mark Osler

It also ignored the contrasting truth that many federal prosecutors are young lawyers in their 20s and 30s who have little experience making decisions as weighty as determining who will be imprisoned and for how long.

The primary reason for the changes was well-intended, though: Members of Congress wanted more uniformity in sentencing. That is, they wanted a term of imprisonment to derive from the crime and the history of the criminal rather than the personality of the person wielding discretion.

After nearly 30 years, we know how Congress' experiment turned out, and the results are not good. Federal judges have been relatively lenient on low-level drug offenders when they have the discretion to go that way. Turning discretion over to prosecutors via mandatory sentences and guidelines not only resulted in a remarkable surge in incarceration, it does not seem to solve the problem of disparities.

Let's look at just one way that prosecutors exercise this discretion: the enhancement of narcotics sentences under 21 U.S.C. 851, or proceedings to establish prior convictions. These enhancements, at a minimum, double a drug defendant's mandatory minimum sentence and may raise the maximum possible sentence. They are based on criminal history and can apply when a drug defendant has a prior qualifying drug conviction, no matter how old. They were enacted as part of the War on Drugs. They often make a huge difference in the sentence that results.

The 851 ruling applies, though, only if a prosecutor decides it should, and therein lies the rub. Federal judges are sometimes willing to vary from the drug sentencing guidelines because they are often too harsh, particularly for low-level drug offenders. Application of 851 enhancements by prosecutors robs judges of this discretion. Once the discretion shifted to prosecutors, the prison population skyrocketed.

And our analysis of the way these enhancements have been used reveals a deeply disturbing dirty little secret of federal sentencing: the stunningly arbitrary application of these enhancements by prosecutors within the Department of Justice.

The numbers tell the story. Our home states are fairly typical in their wild disparities: A federal defendant in Iowa is more than 1,056% likely to receive a 851 enhancement than one in Minnesota.

Nor are these Midwestern neighbors an anomaly. In the Northern District of Florida, prosecutors apply the enhancement 87% of the time, but in the bordering Middle District of Georgia, they are used in just 2% of relevant cases.

There is also breathtaking disparity within federal district within the same state (PDF). For example, in Florida, prosecutors in the Northern District apply the enhancement 87% of the time, but in the Southern District, it is used only 14% of the time. In the Eastern District of Tennessee, offenders are 3,994% more likely to receive an enhancement than in the Western District of Tennessee. In the Eastern District of Pennsylvania, a defendant is 2,257% more likely to receive the enhancement than in the Middle District of Pennsylvania. The disparities are startling.

In August, Attorney General Eric Holder announced steps to establish more discipline within the Department of Justice in how this discretion is used. It is a promising step but only that: a step. It is unclear how firm the attorney general is willing to be in tracking and constraining the use of this kind of discretion by prosecutors in different areas.

The larger lesson, and the more important one, is that after nearly 30 years, we still have gross and tragic disparities in federal sentencing, with the added burden of too many people put in prison, caused by mandatory sentencing and harsh sentencing guidelines.

Tentative steps at reform will not be enough. It is time for a radical rethinking of the project as a whole and a recognition that this grand experiment in shifting discretion to prosecutors has failed.

Follow us on Twitter @CNNOpinion.

Join us on Facebook.com/CNNOpinion.