The Crucial Criminal Cases that Will Start the New Supreme Court Term:
Testing the Federal Sentencing Guidelines' Constitutionality

On October 4, the first day of its new term, the Supreme Court will hear argument in two cases -- United States v. Booker and United States v. Fanfan - that raise an extraordinarily important question about crime and punishment in the U.S.

The cases involve the controversial Federal Sentencing Guidelines - a set of rules (the word "guidelines" is a bit misleading) that govern every federal criminal case. They ask the Court to decide whether the Guidelines are unconstitutional to the extent they allow a judge to enhance a convicted defendant's sentence based on his - not a jury's -- fact findings.

To understand why these cases are so crucial, some background is in order - about both the Guidelines themselves, and the Court's related decision last Term in Blakely v. Washington.

I will argue, in this column, that, while the guidelines may well be unconstitutional, they have some real value and should be modified rather than gutted altogether.

The Reason the Federal Sentencing Guidelines Were Adopted

Before the federal guidelines, federal judges wielded incredible power at sentencing. They could give a defendant any sentence the relevant statute allowed - and the range was often as wide as 10 or 15 years.

Tremendous inequities resulted: Two defendants with similar criminal records (or lack thereof) could commit two basically indistinguishable crimes - yet nonetheless received two wildly different sentences from different judges. A lenient judge might give a first time drug offender probation; a harsher judge in the same courthouse might "send a message" by sentencing the same kind of defendant to many years in jail.

How should this problem be solved? In 1987, Congress offered an answer - adopting a comprehensive federal sentencing guideline system intended to make sure that like cases, and like defendants, were treated alike - and thus equitably.

How the Federal Sentencing Guidelines Work

Under the Guidelines, federal judges no longer typically pick a sentence from within a ten-to-fifteen year range. Instead, they typically pick a sentence within a range that is roughly six months long.

Obviously, the result is to drastically compress the range within which the judge's discretion is exercised. Now the difference between a harsh and lenient judge is more like to be five months than five years - and like defendants ought to more or less be treated alike.

Since the statutory maximums and minimums haven't changed, how do judges figure out what the six-month range should be? The answer is that they look to two key factors: the severity of the offense, and the criminal history of the defendant. Once they've calculated the number assigned to each factor, they look up the two numbers on a grid - and voila! There is the relevant, roughly six-month range of sentences from which they will choose.

(In relatively rare cases, the judge can depart from this short guideline range. But the judge must do so because he or she determines that the case at issue raises a circumstance not contemplated by the guidelines themselves. Thus, for instance, a very novel type of computer hacking might strike a judge as beyond the consideration of the Guidelines' drafters.)

How Do Judges Figure Out Criminal History and Offense Level?

How does the judge figure out a given convicted defendant's criminal history and offense level, before using the grid to determine the sentencing range? The answer is: The judge looks at facts.

Those facts are presented at a sentencing hearing. Prior to the hearing, the judge reviews a PSR (pre-sentencing report). In the PSR, the prosecution -- with help from the probation office -- presents evidence.

The defense also has the opportunity to counter this evidence if it wishes. However, the defendant may already have conceded some of the facts in his guilty plea. In addition, some of the facts may be impossible to controvert given the jury's trial verdict.

For instance, to give a simple example, in a "felon with a firearm" case, a guilty verdict would preclude the defendant from challenging, at sentencing, whether in fact he had a firearm: The jury already said he did.

What kind of other evidence - besides the facts conceded in the plea, or implied by the verdict - might the judge consult?

To determine criminal history, it will be evidence as to the number and seriousness of the defendant's prior crimes, if any.

So suppose a drug dealer is indicted and convicted on only two deals. The prosecution can still offer evidence at the sentencing hearing about what it believes to be the full scope of the defendant's prior drug trafficking. So a two-deal case could lead, in theory, to a hundred-deal sentence. And to many observers, that has seemed very troubling: If a defendant was convicted of only two deals, shouldn't he be sentenced based on those deals alone?

To determine offense level, the evidence will be of the kind the Guidelines deem relevant to a given type of offense. For a drug crime, the Guidelines ask: What was the quantity of drugs? For a property crime, the Guidelines may ask: Was a gun used? For an economic crime such as fraud, the Guidelines ask: What were the damages caused?

The same problem occurs here too: The sentencing proof can far exceed the trial proof. At trial, a prosecutor might prove to the jury only a fraction of the damages caused by the defendant's fraudulent scheme. But at sentencing, the prosecutor may trot out - for the first time, for the judge alone -- evidence of a much bigger range of damages. And again, this has troubled many observers: Shouldn't the jury - which is hearing evidence about the scheme - also hear evidence of its impact?

The Blockbuster Decision in Blakely

During the last Supreme Court Term, in the case of Blakely v. Washington, the Court finally got around to addressing these issues - which had been lingering for about sixteen years, since the Guidelines were passed in 1987. (Why didn't the Court take a case on this issue earlier? It's hard to say.)

The Court's 5-4 decision in Blakely shook the foundations of state - and by implication, federal - sentencing. In particular, it held the State of Washington's criminal sentencing system unconstitutional - but the ramifications of its opinion were much larger.

In Blakely, the Court had to decide what the Sixth Amendment's guarantee of a jury trial really means. Does it literally mean simply that a jury trial must precede a conviction - with any sentence to follow fair game? The Court said no: It means much more.

According to Blakely, the Sixth Amendment means that juries -- rather than judges -- must decide any matter that a defendant did not concede (in his plea or otherwise), and that could lengthen his sentence beyond the maximum set out in state sentencing guidelines.

Does Blakely mean the federal sentencing guidelines, too, are doomed? Lower federal courts around the country have split over this issue.

In my view, the most likely outcome is that the Court (again by a sharply divided vote) will say yes - and it will therefore strike down at least part of the federal Guidelines, as contrary to the Sixth Amendment. If it issues a broad holding, it could gut the Guidelines - and effectively prevent further guidelines from being imposed. And in my view, that will be a big mistake.

The Bedrock Principle Of Sentencing: Discretion Will Remain

Here's the key insight that ought to govern any discussion of Blakely, the Guidelines, and this term's two follow-up cases: The guidelines did not take discretion out of sentencing. Instead, it simply transferred the discretion from judges to prosecutors.

At the risk of some oversimplification, here's why: Roughly 95% of all federal criminal cases are resolved by guilty plea - with no jury ever hearing any evidence at all.

In those cases, the defendant's sentence is basically determined by the prosecutor. The prosecutor chooses the federal crime(s) on which to indict. She chooses which facts (among many) to present to the judge - at the plea stage, and then at the sentencing stage. By pleading guilty, the defendant gives up the right to try to controvert those facts - and the prosecutor, by and large, shapes the terms of the plea.

The result of all this is that the prosecutor, in effect, has the power both to determine a defendant's sentence, and to undercut the supposed sentencing uniformity the guidelines were meant to achieve.

If she wants to go easy on a given defendant, she can indict for the most minor possible crime, present a skimpy criminal history, and downplay the offense. (Obviously, the prosecutor cannot ethically present false evidence, but she does have substantial discretion after complying with that ethics rules.)

Conversely, if she wants to go after a given defendant, she will simply max out the crimes she charges him with, charge him with as many crimes as possible, and at sentencing offer a kitchen sink's worth of evidence showing how shady his criminal history is, and also how dire the offense was.

Before the Guidelines, different judges gave very different sentences for similar defendants' similar crimes. Now, under the Guidelines, different prosecutors make decisions that predictably lead to very different sentences for similar defendants' similar crimes.

The truth is that there is no fair and effective way to eliminate discretion -- and, therefore, some disparity -- from the criminal sentencing process. (For an object lesson as to why killing off discretion is dangerous, it's useful to look to Congress's "mandatory minimum" sentences for various drug offenses - which have resulted in Draconian sentences for countless two-bit drug runners. This result is morally dubious, hugely expensive to taxpayers, and not particularly effective at stopping drug trafficking. In contrast, had judges actually chosen sentences for these hapless drug mules, they would probably not have thrown the book at them, and might have considered alternatives like job training and addiction treatment.)

So the real question, then -- one made much more urgent by the prospect of the Supreme Court scuttling the current guideline system - is how to channel and allocate sentencing discretion, recognizing that there is no way to eliminate it.

Keeping that in mind, let's ask: Should we have Guidelines at all? And if so, what should they look like: In whom should they vest discretion, and how much?

Guidelines for the Guidelines: How They Could Be Improved

First, we should indeed have some kind of Federal Sentencing Guidelines. Ten or fifteen year disparities in sentences in similar cases were simply too great. We should not just turn back the clock, and re-create the often unfair regime that gave rise to the guidelines in the first place. Albeit imperfect, the sentencing guidelines have been quite useful in providing objective criteria for sentencing and a benchmark for what a fair sentence should look like.

Second, if we have to choose between discretion for prosecutors, and discretion for federal judges, it's probably better to give discretion to the latter.

Federal judges' independence is protected by the Constitution - which stipulates, in Article III, that they cannot be removed, nor can their salaries be reduced, and that they sit for life, not a finite term of years.

In contrast, federal prosecutors may be bucking for their own career objectives, and are not insulated and independent in the same way federal judges are. Young prosecutors may want to impress their seniors; older prosecutors may seek Administration positions. Federal judges may seek elevation to a higher court - but mostly, they just want to keep on being federal judges.

Federal judges also have a special kind of blessing: They have been appointed and confirmed by the Senate. In addition, they have a special kind of openness: They make their rulings in open court. In contrast, line prosecutors such as Assistant U.S. Attorneys were simply hired - not confirmed; have no special claim to constitutional authority; and make their decisions in the privacy of the U.S. Attorney's office, without public scrutiny.

For all these reasons, it makes sense to give Article III judges relatively more discretion in sentencing, and federal prosecutors relatively less. (That, in turn, would require something that is sorely needed anyway: more judgeships to address the federal courts' often-crushing current workloads.)

What a New Set of Guidelines Should Look Like

So what should a new set of Guidelines look like? For one thing, it could offer judges wider ranges from which to choose - if ten years' discretion is too much, six months' is also too little. What about splitting the difference - with judge choosing with a three or four year range?

A new set of Guidelines could also allow judges to depart downward more often in first-offender cases, especially when the crime is nonviolent - and to consider drug treatment and job training as sentencing alternatives in such cases. A drug mule who's served a jail time can be a hardened criminal; a drug mule who's had job training can turn his life around.

Would a wider sentencing range simply reintroduce pre-Guidelines unfairness? Obviously, the wider the range, the worse the risk is that this could occur. But there are remedies short of very narrow sentencing ranges.

Chief Judges of particular districts and circuits could set up systems to try to ensure some sentencing regularity among different judges. Conferences on sentencing held across circuits could also perform this function.

Moreover, additional guidelines meant to narrow the range - perhaps to the original six months or so -- could become hortatory rather than mandatory, a recommendation rather than a requirement.

Finally, if the Court does insist - in this Term's two crucial cases -- on junking the mandatory Guidelines altogether, they could replace with another version the entirety of which is simply suggestive.

Would making the Guidelines "merely" suggestive repose too much trust in judges? There are two answers to this:

First, if the Court junks the current, mandatory Guidelines, there is little realistic alternative except suggestive guidelines - so we may have to trust judges whether we like it or not.

Second, no matter what the Court does, as I argued above the reality of discretion will remain. Criminal sentencing is one area where society probably has little choice better than trusting judges to do their jobs well. Fortunately, most will.