Justice Alito's Dissent in Cunningham v. California: How Can Someone So Wrong, Be So Right?

By VIKRAM DAVID AMAR AND AARON RAPPAPORT
Friday, Feb. 16, 2007

A few weeks ago, the Supreme Court issued perhaps the biggest decision of the Term to date: Cunningham v. California. There, the Court held that California's current criminal sentencing system violates a defendant's Sixth Amendment rights.

As discussed in Part One of this series, the Court's ruling seemed to be a straightforward application of a rule developed in Apprendi v. New Jersey and several subsequent cases - namely, that a judge may not impose a sentence above the statutory "maximum" based on his or her own fact-finding. For purposes of this rule, the statutory maximum is defined as the highest sentence that the defendant could lawfully receive based solely on the jury's verdict or the defendant's guilty plea (that is, without any additional fact-finding by the judge).

California's system violated this rule, since it permitted judges to exceed the presumptive sentence set by the legislature based on their own finding of aggravating factors, rather than on account of facts found by a jury beyond a reasonable doubt.

One of the most interesting features of Cunningham was Justice Alito's intricate dissenting opinion, which was also joined by Justices Kennedy and Breyer. As one of Justice Alito's first opinions in this controversial area of criminal constitutional law - and one of his most prominent opinions to date more generally -- his writing deserves careful study.

Particularly interesting is how his dissent highlights and explores one of the great puzzles emerging from this line of cases: When, and why, does the exercise of judicial discretion at sentencing violate the Sixth Amendment?

The Puzzle of Judicial Discretion: When Is It Constitutional, and Why?

In Apprendi and later cases, the Supreme Court criticized so-called "determinate" sentencing schemes (described in detail in the earlier column in this series) that allow judges to use their discretion to find facts that then trigger enhanced sentences or sentence ranges.

The puzzle is that, at the same time, the Court has upheld a purely advisory federal guidelines scheme, a system that actually increases the discretion of judges over sentencing. Moreover, the Justices uniformly agree that "indeterminate" sentencing schemes, where trial courts have virtually unreviewable leeway to set sentences within a broad range, are constitutional. So which is it? Is judicial discretion good, or bad?

As Justice Alito recognizes in his Cunningham dissent, the "bright-line rule" set forth in Apprendi was intended to identify when judicial discretion is acceptable. However, Alito argues that the line is not quite as bright as the majority in Cunningham suggests, and certainly not dispositive in assessing whether the California sentencing system is constitutional. Alito advances two arguments in support of this claim:

First, Alito argues that the Apprendi rule is sharply limited. It applies, he says, only when judges increase sentences beyond the maximum based on specific "fact-finding." (Emphasis added.) According to Alito, the California system allows judges to exceed the maximum based on other criteria besides fact-finding, such as their own policy judgments about whether a higher sentence for particular categories of crimes is needed to serve a valid purpose of punishment (such as deterrence). This added judicial discretion, Alito implies, saves the California system, rendering it constitutional.

Alito's argument faces serious problems. Perhaps most importantly, his reading of state law is questionable. The Cunningham majority did not agree with his view that judges in California have unlimited discretion to increase sentences above the maximum. Rather, they held that California law permits enhanced sentences only after judicial fact-finding has occurred.

Moreover, even if Alito's interpretation of state law had been accepted by the majority, his argument still might falter. Notably, Alito never explained why the possibility that judges can sentence above the maximum based on their own policy judgments - as opposed to their own fact-findings -- makes the system constitutional (and fairer for the defendant). One might think, for example, that if the Court's concern in this line of cases is about the dangers of judicial discretion, Alito's reading of state law would not necessarily solve the constitutional problem.

Is the Supreme Court's Booker Precedent Incoherent?

Alito's second point deserves closer attention. He argues that "[e]ven if California did require that a sentencing court find some aggravating 'fact' before imposing the upper term sentence, that would not make this case constitutionally distinguishable" from what the Supreme Court upheld as constitutionally permissible in United States v. Booker.

In Booker, the Court struck down the federal sentencing guidelines system under the Apprendi rule. But at the same time, in its remedial order, the Court allowed the guidelines to remain as an "advisory" system of rules. Alito suggests that this remedial order effectively limits the Apprendi rule, since an advisory system necessarily allows judges to impose sentences above the "maximum" based on the judges' own fact-finding.

How so? Alito suggests that the key here is to realize that sentences imposed under the advisory system are still to be reviewed by appellate courts for "reasonableness." Alito argues that appellate reasonableness review means that for each offense, there is a sentence which "represents the [most] onerous sentence that [could] be regarded as reasonable in light of the bare statutory elements found by the jury."

On Alito's view, any higher sentence is not per se illegal, but it would have to be justified in order to be held reasonable on appeal. Booker's remedial order "necessarily anticipates that the imposition of sentences above this level may be conditioned upon findings of fact made by the judge and not the jury." As Alito puts the point, "if reasonableness review is more than just an empty exercise, there inevitably will be some sentences that, absent any judge-found aggravating fact, will be unreasonable."

Presto! Booker's remedial portion thus suggests that judges can, in fact, justify sentences above the "maximum" based on their own fact-finding!

Alito's argument here is not free from problems. At most, his argument simply highlights a possible inconsistency between the two parts of the Court's Booker opinion: the substantive portion that strikes down the binding guidelines (for violating the Apprendi rule), and the remedial portion that upholds the advisory guidelines (even though they appear to transgress the same rule). Some dissenters to the remedial portion of Booker themselves highlighted this same tension. But this potential inconsistency does not mean that the Apprendi rule should be discarded; it could be taken to mean simply that the Court was wrong in its remedial order to validate the advisory sentencing guidelines.

In any event, Alito is far too quick to assume an inconsistency exists between the substantive and remedial portions of the Booker opinion. A reconciliation might be possible - and this potential reconciliation, as we will explain, raises further questions about Alito's arguments in support of the California sentencing scheme.

The Error of Justice Alito's Ways

Justice Alito's claim that the two parts of the Booker opinion are inconsistent ignores some potentially salient distinctions between the binding guideline regime struck down in Booker's substantive opinion and the advisory system validated in the Court's remedial order:

One distinction concerns the way that a verdict is translated into a specific sentencing outcome. Under binding guidelines, the verdict generate a determinative sentencing range based on formal, explicit, publicly-articulated sentencing rules prescribed in advance by a legislative body. As a result, in a binding guideline system, it is obvious when judicial fact-finding leads to an increase in the sentence beyond the maximum determined solely by the verdict.

In contrast, under an advisory guideline system, the sentencing range associated with a verdict remains obscured (at least, until the appellate courts develop formal sentencing dispositions that define categories of reasonableness for different offenses). Under an advisory system, therefore, the effect of judicial fact-finding is much less overt, and possibly less troubling for that reason.

According to this distinction, the Apprendi rule should properly apply only to maximum sentences defined as the highest sentence that can be imposed without further judicial fact- finding conducted pursuant to explicit, formal and publicly ascertainable sentencing rules.

Such a reformulation (or clarification) of the Apprendi rule would explain why the binding federal guideline system was unconstitutional: because it allowed judicial fact-finding to exceed the maximum, under this definition.

It would also explain why the advisory system is constitutional. Since that system does not at this time contain any formal sentencing rules to establish a maximum sentence, the judicial fact- finding does not transgress the constitutional rule. Of course, if the appellate courts, in the future, formalize explicit rules to determine the maximum "reasonable" sentence for an offense, the advisory system might ultimately face similar problems. But that seems a long time away.

The problem with this reconceptualization of the Apprendi rationale is that the reason we would want to make a distinction between formal, overt and publicly-ascertainable judge-found facts, on the one hand, and informal, under-the-radar judicial fact-finding, on the other, is far from obvious. While the effect of judicial discretion is no doubt obscured in an advisory system, it continues to operate sub silentio. If judicial discretion is problematic, the exercise of that discretion perhaps ought to be problematic whether its effect is clearly visible or not.

Public ignorance of what the courts are doing shouldn't make judicial fact-finding constitutionally acceptable, should it?

Is the Problem Too Much Legislative Prescription? A Distinction With Strong Explanatory Power

But perhaps there is another distinction that could serve to reconcile the two parts of the Booker opinion. Maybe the problem with the prior, binding guidelines was not their public openness, but rather the institutional role played by the legislature or U.S. Sentencing Commission in formulating the standards.

In the previous, binding system, these two institutions determined the maximum sentence that could be imposed based solely on the verdict, and determined which facts should be taken into account to deviate from this maximum.

By contrast, in the current advisory guideline system, the judiciary makes these ultimate determinations. As a result, one could reconcile the two parts of Booker by clarifying that the maximum sentence, for Apprendi purposes, means the most severe sentence that the legislature or commission determines can be imposed based on the fact of conviction (or based on facts stipulated by the defendant in a plea agreement).

This reformulation would explain why (as the Court held) a binding guidelines system violates the Constitution, but an advisory guidelines system does not: A binding guideline system (such as the prior federal sentencing system) would violate Apprendi because - and to the extent that -- it allows the judiciary to increase the sentence beyond the maximum sentence established by the legislature or Commission, pursuant to facts the legislature or Commission has prescribed as important.

By contrast, an advisory guideline system survives constitutional review, because the legislature and the commission haven't determined the maximum sentence that can be imposed based solely on the verdict. As a result, the advisory system doesn't have a maximum sentence at all (short of the ultimate statutory maximum). Nor has the legislature tried to identify all the additional facts a judge must weigh in ultimately rendering a sentence.

Why the Distinction Is Not Only Explanatory, But Logically Defensible

This approach does a good job of explaining why binding guidelines are unconstitutional and advisory guidelines are lawful, and so it offers a plausible way to reconcile the two parts of the Booker opinion. But does this distinction make any sense?

Perhaps it does. The distinction is premised on the view that the defendant's liberty interests are threatened when a legislature (or Commission ) attempts to determine the effect of a specific fact on the defendant's ultimate sentence. The danger, as recent history suggests, is that the legislature will opt for long sentences based upon facts of questionable reliability.

To counteract that danger, the Apprendi rule says that any time the legislature (or Commission) gives a specific fact a weight, that fact must be submitted to the jury and proved beyond a reasonable doubt.

By contrast, judicial discretion over sentencing does not seem to pose a similar danger, perhaps because of an assumption that the judiciary will in general be more sympathetic to the liberty interests of the defendant. Thus, the rule suggests a certain way to understand the Apprendi line of cases. These cases represent a distrust of legislative (and Commission) power in the sentencing realm, not a distrust of judicial discretion per se.

Though this might be a profitable way to harmonize the two parts of the Booker opinion, the approach does not necessarily support California's sentencing scheme reviewed in Cunningham. Rather, the reformulated Apprendi rule suggests that Cunningham may have, in fact, been correctly decided.

After all, under California law, the middle term was deemed by the legislature to be the presumptively reasonable sentence. As such, it represents the "maximum" sentence under the reformulated rule. By allowing judges to sentence above the legislatively-identified maximum pursuant to their own fact-finding - some of which facts, at least, were themselves listed by the legislature as relevant -- the system might still implicate at least to some extent the constitutional rule. (Since only some of the aggravating facts on which a judge could rely were legislatively-identified, however, an argument can be made that the California scheme should have been less troubling to the Court than the laws struck down in the earlier cases in this line.)

Maybe Justice Alito Is Right, As to His Bottom Line, After All

Whether our various (re)formulations of the rule in Apprendi justify the Court's actions in this line of cases might depend on what the federal and state legislatures choose to do after their chosen determinate sentencing scheme is called into constitutional question.

To the extent that legislatures respond by giving more power to juries, or by moving towards indeterminate sentencing regimes that give judges more discretion, perhaps the end results will redound to the benefit of criminal defendants whose interests are supposedly being vindicated by the Sixth Amendment doctrine.

But to the extent that the legislatures respond to the judicial invalidation of sentencing schemes by imposing tougher sentences on offenders, or by creating legislative presumptions in favor of harsher sentences (with judges being empowered to find facts only to reduce, rather than increase, the punishments), the Court's rulings in the Apprendi line of cases might backfire.

These possibilities are underscored by two potential responses to the Cunningham decision itself - one hypothetical, and the other currently under legislative consideration:

First, consider a hypothetical. What if, as our colleague Michael Dorf has suggested, California were to keep its scheme of allowing three possible sentencing options - say, 16, 12 or 6 years -- for a given offense. Instead of establishing the middle term as the presumptive sentence, suppose the legislature set the presumption at 16 years, rather than 12. Under this scheme, unless and until the trial judge finds one or more specified facts of mitigation (and perhaps mitigation could include a finding of no legislatively identified aggravating facts), then the judge must choose 16 years. But if such mitigation (or lack of aggravation) is found by the court, the judge could reduce the sentence to 12 or 6 years.

Such a system would be similar to the one struck down by the Court in Cunningham, but also presumably immune from the Court's rationale for invalidation. Such a response by the legislature would presumably be worse for defendants, and yet it would seem to pass constitutional muster.

Or consider a proposal that is in fact somewhat likely to pass the California legislature soon: Senate Bill 40 (SB40) would amend the sentencing rules to give the court complete discretion to choose which of the three sentencing points it prefers. That change would avoid any plausible constitutional concerns by eliminating the legislative presumption in favor of any of the three terms at all. In effect, it would result in a discretionary sentencing scheme with three possible sentencing options.

In all likelihood, this change, too, would seem to make it easier (that is, easier than under the law that the Court struck down in Cunningham) for a judge to impose the higher term. Before, the sentencing judge needed to identify some fact to justify the higher term. Under SB40, it's not clear the court needs to do so.

This simply highlights the potential unwisdom of striking down the California system on the basis that it violates the defendant's Sixth Amendment rights. How can the defendant's Sixth Amendment rights be promoted by a piece of legislation that makes it easier to sentence to a higher term?

Thus, even under our various theoretical clarifications or reconceptualizations of the Apprendi line, the Court might not be reaching results that vindicate the values about which it cares, and, at a minimum, has a lot more explaining to do.

Justice Alito in dissent is certainly justified in his seeming frustration with the Court's failure (to date) to make clear to the world what this doctrinal line is really all about.


Vikram David Amar is a professor of law at the University of California, Hastings College of Law in San Francisco. He is a 1988 graduate of the Yale Law School, and a former clerk to Justice Harry Blackmun. He is a co-author, along with William Cohen and Jonathan Varat, of a major constitutional law casebook, and a co-author of several volumes of the Wright & Miller treatise on federal practice and procedure. Before teaching, Professor Amar spent a few years at the firm of Gibson, Dunn & Crutcher. Aaron Rappaport is a professor of law at the University of California, Hastings College of the Law in San Francisco. He is a 1991 graduate of Yale Law School, and a former clerk to now-Justice Stephen Breyer on the U.S. Court of Appeals. Prior to teaching, Professor Rappaport served as counsel to the Senate Judiciary Committee and to the Federal Communications Commission, and was an Assistant Director on the National Economic Council. Between 1996 and 2005, Professor Rappaport was also an editor of the Federal Sentencing Reporter, a leading journal on sentencing law.