« Upcoming MSU symposium to examine North Carolina Racial Justice Act | Main | "State parole system called a 'sad mess'" »

February 19, 2011

Great new article "celebrating" a quarter century of federal sentencing under the SRA

Inspired in part by my recent (serious!?!) post suggesting we ought to consider how IBM's Watson computer could aid sentencing decision-making, a couple of helpful readers pointed out this new article by J.C. Oleson in the University of Richmond Law Review titled "Blowing Out All the Candles: A Few Thoughts on the Twenty-Fifth Birthday of the Sentencing Reform Act of 1984." 

There are many aspect of this piece that make in a must-read this long weekend (even though there is something a bit mathematically peculiar about celebrating the 25th birthday of the SRA in 2011).  In light of my Watson post, I especially liked the passage spotlighting that "sociologist Max Weber imagined a kind of sentencing computer that would collect relevant facts and dispense a just sentence," and that "Marvin Frankel, the patron saint of sentencing guidelines, acknowledged that computers could be useful in bringing parity and fairness to sentencing."  The piece also includes these concluding thoughts:

Instead of basing federal sentences on political intuitions, the Commission could provide sentencing judges with meaningful data about which available sentences are most effective in reducing recidivism.  Improvements in risk assessment and technology have made it possible for the Commission to provide judges with data that were scarcely imaginable twenty-five years ago.

Even five years ago, given the acrimonious climate between Congress and the courts, it was difficult to envision a system of this kind.  But much has changed.  Given the Feeney Amendment, Booker and its progeny, and a growing interest in evidence-based policy, an actuarial sentencing information system is not only intellectually conceivable, but socially and politically viable.

So that, Sara, is my wish for your birthday: an actuarial sentencing information system that allows federal judges to impose data-driven sentences that are effective, efficient, and fair.  It is something that, at twenty-five, you might become.

February 19, 2011 at 12:23 PM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Great new article "celebrating" a quarter century of federal sentencing under the SRA:


"Instead of basing federal sentences on political intuitions, the Commission could provide sentencing judges with meaningful data about which available sentences are most effective in reducing recidivism."

Reducing recidivism is only one aspect of sentencing, albeit an important one. Some crimes are so destructive that a low sentence is inappropriate even if it would ensure the offender never commits another crime. For instance, if statistics showed that six months in jail prevented most murderers from reoffending, no one would think that's an appropriate sentence. On the other hand, some offenses are so insignificant that it's not particularly offensive that the defendant might commit the crime again. I'm thinking possession of drugs, illegal entry by foreigners, and similar crimes. Most sentencing schemes, see 18 U.S.C. 3553 for example, take into consideration other factors like rehabilitation, retribution, and incapacitation.

One last factor of sentencing is purely political: legislators always want to create new crimes with harsh penalties just so they can pretend they are "doing something about crime." Which is why we see inane things like 10-year mandatory minimums for people who transport drugs, the tendency toward making every new crime a felony (in Texas, there are 11 different felony offenses related to oysters), and the like.

Anyway, we certainly do want judges to have recidivism data available, but at the same time, I would suspect that neither defendants nor prosecutors would want judges to consider the likelihood of recidivism to be the primary factor at sentencing.

Posted by: C.E. | Feb 19, 2011 3:05:56 PM

I urge every innocent defendant railroaded by the incompetent lawyers to demand a Daubert hearing on every aspect of sentencing. Prove that every element of a sentence has a goal, and that it has been shown to achieve that goal. Otherwise the sentence is anti-scientific garbage.

Posted by: Supremacy Claus | Feb 19, 2011 4:44:27 PM


I would accept most of your list of crimes that do not necessarily warrant long sentences, but illegal re-entry is a crime that pretty much demonstrates that the person just doesn't care about US law. In order to get an illegal re-entry conviction, the person must have either already voluntarily left or been deported, and then after having been ordered not to come back they do so anyway. In so doing they are likely to commit a rash of other crimes, identity fraud of various forms being the archetypal example.

When someone has so thoroughly proven themselves a scofflaw I see no reason to cut them any slack.

Posted by: Soronel Haetir | Feb 19, 2011 6:38:18 PM

An interesting study would be recidivism in relation to the number of felonies, as C.E. suggests.

Say in 1820 there were 10 felonies and in 2011 there are 1,000 (instead of the probable thousands).

Obviously, more recidivists would commit felonies in 2011 than they did in 1820, and so felony recidivism would appear to be much more serious (prevalent) now compared to 1820.

Posted by: George | Feb 19, 2011 7:54:24 PM

In Vietnam, with no enforcement of the law, 40% of soldiers indulged in opium and heroin. Upon return that fraction fell to 2%, the same as that of the aged matched male population that had not gone to Vietnam. The risks of crime suppress crime. The rate is lower in a small town, where secrets are impossible. The likelihood of punishment is more of a deterrent than the severity of punishment (see all the pickpocketing that took place at the executions of pickpockets).

That means one thing only. The population must be allowed to enforce the laws. These should be the minimum laws to maintain safety and civilization, so they are understood and respected. Excessive prohibitions bring contempt on the lawyer and on the law. The list of FBI Index felonies is enough.

The regular citizen must be trained in firing weapons. There should be a duty to kill all criminals at the scene. If someone makes no attempt and the omission is caught on videotape, it should be a summary offense with a $100 ticket sent in the mail to the failing citizen. The risk of crime will be high in the chance of detection, and in the severity of the outcome.

Crime will diminish through attrition, after all the criminals are killed. Funding should be generous for detection of the genetic predisposition to criminality, so all fetuses carrying it may be aborted. But the Vietnam studies shows normal people commit crimes when the risk drops low enough.

Posted by: Supremacy Claus | Feb 19, 2011 9:27:57 PM

Along with the mathematical oddity, there is the citation to "United States v. Blakely."

Kind of reminds me of John Belushi in Animal House. "Was it over when the Germans bombed Pearl Harbor?" One of the animals says, "The Germans?" but another says, "Forget it. He's rolling."

Posted by: Kent Scheidegger | Feb 21, 2011 12:21:13 PM

It all depends on the slant of how the program is written...How much weight is given to offenses 10 yrs ago and how much weight is given to the last 2 yrs beofre the instant offense and after arrest...There are lots of other factors that just can't be programmed in without undue influence by the current obscene guidelines....What the Feds want is the absolute max sentence, so I think we are still heading on the wrong trail...Need to get Holder on track, to brief his AUSA, the judges and the circuit courts have to want to play ball as well....

The fair sentenceing act, actually increased other types of sentences...One would think, is there ever a day when they stop jacking them upward, and start to go downward...This has to be achieved first, then continual improvement and after a time, you don't need a computer or rigid guidelines at all...Once the Iron curtain has been knocked down...

This kind of mindset needs to go away...
The ashcroft memorandum: prosecute to the highest level charge and ensure guideline sentences...Holder says, WIN....Mostly at any cost. These is my thoughts on it...Computers are only as good as the program behind them...With the mindset of the guidelines, forget the computer, just be more of the same...Computers are an awful view to take on sentenceing people, they are even more sterile than some of the AUSA and Federal judges...Hard to believe that one..

Posted by: Josh | Feb 22, 2011 4:00:44 PM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB