Mr. Frank Bowman

Professor of Law
Indiana University School of Law
July 13, 2004

Blakely v. Washington and the Federal Sentencing System

I am grateful to the Committee for the opportunity to testify today regarding the U.S. Supreme Court's very recent decision in Blakely v. Washington and its effect on the federal sentencing system. My testimony will address four questions: (1) Does Blakely v. Washington apply to the Federal Sentencing Guidelines? (2) Should Congress take any immediate legislative action in response to the Blakely decision? (3) If Congress acts in the near term, what sensible options are available to it? (4) Regardless of whether Congress acts in the near term, should the Blakely decision act as a catalyst for a thorough re-evaluation of the state of federal sentencing?

I. The Effect of Blakely v. Washington on the Federal Sentencing System

A detailed analysis of the Supreme Court's opinion in Blakely v. Washington, __ U.S. __, 2004 WL 1402697 (June 24, 2004), is beyond the scope of this testimony. In summary, the case involved a challenge to the Washington state sentencing guidelines. In Washington, a defendant's conviction of a felony produces two immediate sentencing consequences -- first, the conviction makes the defendant legally subject to a sentence within the upper boundary set by the statutory maximum sentence for the crime of conviction, and second, the conviction places the defendant in a presumptive sentencing range set by the state sentencing guidelines. This range will be within the statutory minimum and maximum sentences. Under the Washington state sentencing guidelines, a judge is obliged (or at least entitled) to adjust this range upward, but not beyond the statutory maximum, upon a post-conviction judicial finding of additional facts. For example, Blakely was convicted of second degree kidnapping with a firearm, a crime that carried a statutory maximum sentence of ten years. The fact of conviction generated a "standard range" of 49-53 months; however, the judge found that Blakely had committed the crime with "deliberate cruelty," a statutorily enumerated factor that permits imposition of a sentence above the standard range, and imposed a sentence of 90 months. The Supreme Court found that imposition of the enhanced sentence violated the defendant's Sixth Amendment right to a trial by jury.

In reaching its result, the Court relied on a rule it first announced four years ago in Apprendi v. New Jersey, 530 U.S. 466, 490 (2000): "Other than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt." In the years since Apprendi, many observers (including myself) assumed that Apprendi's rule applied only if a post-conviction judicial finding of fact could raise the defendant's sentence higher than the maximum sentence allowable by statute for the underlying offense of conviction. For example, in Apprendi itself, the maximum statutory sentence for the crime of which Apprendi was convicted was ten years, but under New Jersey law the judge was allowed to raise that sentence to twenty years if, after the trial or plea, he found that the defendant's motive in committing the offense was racial animus. The Supreme Court held that increasing Apprendi's sentence beyond the ten-year statutory maximum based on a post-conviction judicial finding of fact was unconstitutional.

In Blakely, however, the Court found that the Sixth Amendment can be violated even by a sentence below what we have always before thought of as the statutory maximum. Henceforward, "the 'statutory maximum' for Apprendi purposes is the maximum sentence a judge may impose solely on the basis of the facts reflected in the jury verdict or admitted by the defendant." Blakely v. Washington, supra (Opinion of Justice Scalia; emphasis in original).

Accordingly, the Federal Sentencing Guidelines seem to fall within the Blakely rule. A defendant convicted of a federal offense is nominally subject to any sentence below the statutory maximum; however, the actual sentence which a judge may impose can only be ascertained after a series of post-conviction findings of fact. The maximum guideline sentence applicable to a defendant increases as the judge finds more facts triggering upward adjustments of the defendant's offense level. In their essentials, therefore, the Federal Sentencing Guidelines are indistinguishable from the Washington guidelines struck down by the Court. Thus, although the Court reserved ruling on the application of its opinion to the Guidelines, there seems little question that it does impact the Guidelines.

This basic assessment seems to be born out by the actions of federal district and appellate courts over the last two weeks. As related in illuminating detail in the written testimony of Judge Paul Cassell for the morning's hearing, and as documented on an up-to-the-minute basis on Professor Doug Berman's marvelous website, http://sentencing.typepad.com, it appears that every federal district judge who has ruled so far has held Blakely applicable to the federal sentencing guidelines, rendering them either facially unconstitutional or unconstitutional as heretofore applied. The consensus on the applicability of Blakely to federal sentencing was joined last Friday by the Seventh Circuit. United States v. Booker, __ F.3d __, No. 03-4225 (7th Cir. July 9, 2004), slip op. at 10.

Unfortunately, although there seems to be wide agreement that Blakely applies to the federal system, there is no agreement whatever on how to apply it. In the last two weeks, federal sentencings all over the country have stopped or been postponed while courts and litigants assess the situation. As judges begin to rule, they face three basic options: (a) find that Blakely does not apply to the federal sentencing guidelines and proceed as though nothing has happened; (b) find that the Sentencing Guidelines survive, but that each guideline factor which produces an increase in sentencing range above the base offense level triggered by conviction of the underlying offense is now an "element" that must be pled and proven to a jury or agreed to as part of the plea; or (c) find that the Guidelines are facially unconstitutional, in which case judges can sentence anywhere within the statutory minimum and minimum sentences of the crime(s) of conviction.

Consider these options and their practical consequences:

(a) Blakely does not apply to the Federal Sentencing Guidelines: So far, the only federal judge to articulate this position in writing has been Seventh Circuit Judge Frank Easterbrook, dissenting from Judge Richard Posner's opinion holding Blakely applicable to the guidelines. United States v. Booker, __ F.3d __, No. 03-4225 (7th Cir. July 9, 2004). Of the district court judges who have so far issued opinions addressing the Blakely's impact on the Guidelines, none has found that Blakely does not apply.

(b) Blakely transforms the Guidelines into a part of the federal criminal code: The second possibility is that the guidelines remain constitutional as a set of sentencing rules, but that the facts necessary to apply the rules must be found beyond a reasonable doubt by a jury or be agreed to by the defendant as a condition of his or her plea. In effect, all Guidelines rules whose application would increase a defendant's sentencing range would be treated as "elements" of a crime for purposes of indictment, trial, and plea.

If this approach is adopted, it has different effects on two classes of cases: (1) those who have already been convicted by trial or plea, and (2) those who have not yet been convicted. For those in the first category, treating the Guidelines as elements of the offense will, in many cases, produce sentencing windfalls. If guidelines enhancements are elements requiring a jury verdict or a defendant admission, judges cannot apply guidelines provisions -- such as those involving role in the offense, loss amount in fraud cases, drug amounts greater than those triggering mandatory minimums, vulnerable victim enhancements, and a host of others - that were not decided by a jury or admitted by the defendant himself as part of the plea. During the last week, several district court judges have used essentially this approach to reduce the sentences of convicted defendants whose cases were awaiting sentencing or pending appeal.

For cases in the second category, those where guilt has not yet been adjudicated, the problems are far more vexing. If Guidelines adjustments were henceforward to be treated as elements of a crime to be proven beyond a reasonable doubt at trial, a host of new rules and procedures would have to be devised. At this point, no one has fully mapped out all the modifications that would be required; however, the list would seem to include at least the following:

? The government would presumably have to include all guidelines elements in the indictment. However, this is not certain. Perhaps guidelines enhancements sought by the prosecution could be enumerated in separate sentencing informations; but if so, such a procedure would presumably have to be authorized by statute and might not pass constitutional muster.
? If guidelines elements were required to be stated in indictments, grand juries as well as trial juries would have to find guidelines facts, and thus grand jurors would have to be instructed on the meanings of an array of guidelines terms of art - "loss," reasonable foreseeability, sophisticated means, the differences between "brandishing" and "otherwise using" a weapon, etc.
? Grand juries have hitherto been prevented from considering sentencing factors, both because they have been legally irrelevant and because many such factors were thought prejudicial to the defendant. Several U.S. Attorney's Offices have begun considering whether it will be necessary to bifurcate grand jury indictments and presentations by presenting the "substantive" section of the indictment in one session, and then, after the grand jury has returned a true bill on the substantive offense, presenting the sentencing portion of the indictment with supporting evidence.
? Since guidelines enhancements would be elements for proof at trial, the Federal Rules of Criminal Procedure and local discovery rules and practices would have to be revised to provide discovery regarding those elements.
? New trial procedures would have to be devised. Either every trial would have to be bifurcated into a guilt phase and subsequent sentencing phase, or pre-Blakely offense elements and post-Blakely sentencing elements would all be tried to the same jury at the same time. There is now no provision in federal statutes or rules for bifurcated sentencing proceedings, except in capital cases, and there is at least some doubt that such bifurcated trials would even be legal in the absence of legislation authorizing them.
? If a unitary system of trial were adopted, the judge would be required to address motions to dismiss particular guidelines elements at the close of the government's case and of all the evidence, before sending to the jury all guidelines elements that survived the motions to dismiss.
? In either a unitary or bifurcated system, the judge would be obliged to instruct the jury on the cornucopia of guidelines terms and concepts, and the jury would have to produce detailed special verdicts.

The prospect of redesigning pleading rules, discovery and motions practice, evidentiary presentations, jury instructions, and jury deliberations to accommodate the manifold complexities of the Guidelines should give any practical lawyer pause. It is doubtful that judges alone could effect the transformation. Legislation and Sentencing Commission action would almost certainly be required to modify the Sentencing Reform Act, the Guidelines, and the Federal Rules of Criminal Procedure to accommodate the new constitutional model, a process that would take months or years to accomplish. In the interim, uncertainty would be endemic.

A second consequence of treating all Guidelines sentencing enhancements as elements would be to markedly alter the plea bargaining environment. This reading of Blakely would transform every possible combination of statutory elements and guidelines sentencing elements into a separate "crime" for Sixth Amendment purposes. This has two consequences for plea bargaining: (a) As a procedural matter, each Guidelines factor that generates an increase in sentencing range would have to be stipulated to as part of a plea agreement before a defendant could be subject to the enhancement. (b) More importantly, negotiation between the parties over sentencing facts would no longer be "fact bargaining," but would become charge bargaining. Because charge bargaining is the historical province of the executive branch, the government would legally free to negotiate every sentencing-enhancing fact, effectively dictating whatever sentence the government thought best within the broad limits set by the interaction of the evidence and the Guidelines. The government would no longer have any obligation to inform the court of all the relevant sentencing facts and the only power the court would have over the negotiated outcome would be the extraordinary (and extraordinarily rarely used) remedy of rejecting the plea altogether.

A plea bargaining system that operated in this way might benefit some defendants with particularly able counsel practicing in districts with particularly malleable prosecutors. On the other hand, making sentencing factor bargaining legitimate would dramatically increase the leverage of prosecutors over individual defendants and the sentencing process as a whole, leading to worse results for some individual defendants and a general systemic tilt in favor of prosecutorial power.

In any case, any benefit to defendants would inevitably be uneven, varying widely from district to district and case to case. To the extent that the Guidelines have made any gains in reducing unjustifiable disparity, a system in which all sentencing factors can be freely negotiated would surely destroy those gains. (Prevention of this outcome was, after all, the point of the Guidelines' "relevant conduct" rules, see U.S.S.G. §1B1.3.) It might be suggested that the Justice Department's own internal policies regarding charging and accepting pleas to only the most serious readily provable offense would protect against disparity; however, the experience of the last decade, during which variants of the same policy have always been in place, strongly suggests that local U.S. Attorney's Offices cannot be meaningfully restrained by Main Justice from adopting locally convenient plea bargaining practices. Once previously illegitimate "fact bargaining" becomes legally permissible charge bargaining, no amount of haranguing from Washington will prevent progressively increasing local divergence from national norms.

Ironically, if Blakely were ultimately determined to require (or at least permit) the Guidelines to be transformed into a set of "elements" to be proven to a jury or negotiated by the parties, the effect would be to markedly reduce judicial control over the entire federal sentencing process. Not only would district court judges be stripped of the power to determine sentencing facts and apply the Guidelines to their findings, but appellate courts would be stripped of any power of review. Neither jury findings of fact nor the terms of a negotiated plea are subject to appellate review in any but the rarest instances. Thus, the interpretation of Blakely discussed here would have the perverse effect of exacerbating one of the central judicial complaints about the current federal sentencing system - the increase of prosecutorial control over sentencing outcomes at the expense of the judiciary.

(c) Blakely renders the Federal Sentencing Guidelines facially unconstitutional: The third reading of Blakely open to judges is that it renders the Federal Sentencing Guidelines in their present form facially unconstitutional, at least within the current framework of procedural rules governing criminal trials, sentencings, and appeals. Several district court judges have issued rulings to this effect, including an elegant and persuasive opinion by Judge Paul Cassell who is appearing before you today. I think Judge Cassell is right and that the Supreme Court is likely to agree.

Blakely appears to require this result. Blakely finds it unconstitutional for the maximum sentence to which a defendant is exposed based purely on the facts found by a jury or admitted in a plea agreement to be increased based on post-conviction judicial findings of fact. The linchpin of the entire federal sentencing guidelines system is precisely such post-conviction judicial findings. The Guidelines model has three basic components: (1) post-conviction findings of fact by district court judges; (2) application of Guidelines rules to those findings by district court judges; and (3) appellate review of the actions of the district court. Both the Guidelines themselves and important components of statutes enabling and governing the Guidelines were written to effectuate this model. Although it is intellectually possible to isolate the Guidelines rules from the web of trial court decisions and appellate review procedures within which the rules were designed to operate, doing so does such violence to the language, legislative history, and fundamental conception of the Guidelines structure that one could save them only by transforming them by judicial fiat into something that neither the Sentencing Commission nor Congress ever contemplated that they would become. It is certainly true that when construing statutes facing constitutional objections that courts will attempt to save so much of the statute as can be saved consistent with the constitution. On the other hand, if the reading of a statute required to render it constitutional transforms the statute into something entirely at odds with its original design and conception, courts may properly strike down the statute in its entirety.

Not only does the reasoning and language of Blakely seem to require invalidation of the Guidelines, but the real world effects of the alternative Guidelines-as-elements interpretation outlined in the previous section will give thoughtful judges reason to shy away from it. Not only would such a system be remarkably ungainly, but far more importantly, it would, as noted, exacerbate those features of the current system that federal judges find most galling. If the only options facing the Court were (a) preserving a simulacrum of the Guidelines system that would make the features judges now find most objectionable even worse, or (b) striking the system down in its entirety and starting anew, it is hard to imagine that a majority of the justices would not strike down the system given a plausible constitutional argument for doing so.

Thus, while the Supreme Court could adopt a saving interpretation of the Guidelines which transformed them into elements of a new set of guidelines crimes, the Court could, without any violence to ordinary principles of constitutional adjudication, just as easily find the whole structure invalid.

If the Court finds the Guidelines as a whole unconstitutional, a different form of uncertainty would result. Presumably, district judges would then be free to sentence defendants anywhere within the statutory minimum and maximum sentences applicable to the offense(s) of conviction. Some judges might view the Sentencing Guidelines as being advisory and might continue to make factual findings called for by the Guidelines in order to aid them in using the Guidelines to guide their sentencing discretion. Other judges might conclude that the Guidelines, being unconstitutional, were of no further use to the judiciary.

Some have characterized this eventuality as a return to the legal regime that predated the enactment of the Sentencing Reform Act of 1984. However, it would differ in at least one critical respect. Prior to the SRA and the Guidelines, federal parole authorities exercised substantial authority over actual release dates of convicted prisoners. Their parole guidelines and discretionary judgments often had the salutary effect of evening out inter-judge sentencing disparities and reducing unreasonably high judicial sentences to more reasonable levels. This backstop on judicial sentencing authority would no longer be in place if the Guidelines were simply cast aside.

II. Should Congress Take Action in the Short Term?

Although, as discussed below, I have advanced a proposal for a legislative response to Blakely, there exists a real and substantial question about whether Congress should take any action in the near term.

1. The argument for legislating

The argument for taking some immediate action is straightforward. The Supreme Court has issued a decision that almost certainly renders the Federal Sentencing Guidelines unconstitutional, at least as now written and applied. However, the Court neither addressed the federal system in particular, nor gave any general guidance about how its opinion should be applied in the courts. Consequently, the federal criminal justice system is in a state of turmoil from which it is unlikely to emerge even after the Supreme Court adjudicates the applicability of Blakely to the Guidelines. Consequently, it would be desirable for Congress to pass legislation that would restore order to the federal sentencing process, at least while more permanent and fundamental reforms are considered.

2. The arguments against legislating

The arguments against taking prompt legislative action sound five basic themes, which I have labeled (a bit facetiously) as follows:

a. "Wait for the Supreme Court."

The first instinct of most of us before seeking legislation in response to Blakely is to look to the Supreme Court for additional guidance on two critical points: Are the Sentencing Guidelines constitutional under Blakely? If not, how can we fix the system to make it constitutional? The difficulty is that the Court will not tell us the answer to the first question for months, perhaps many months, even though the answer seems reasonably clear. The bigger difficulty is that if the Court does the expected and declares the Guidelines unconstitutional facially or as applied, the Court will probably not tell us very much about how to fix the problem. In which case, we will be in very much the same fix we are now.

b. "Don't worry. We've got everything under control."

At least some folks have been heard to suggest that Blakely just isn't that big a practical problem. That the Department of Justice and the courts will, in short order, work out procedures that bring the existing sentencing regime into conformity with Blakely and everything will soon be just fine. While I yield to no one in my respect for the problem-solving capabilities of the Justice Department and the federal bench and bar, this diagnosis seems unduly sanguine. Judge Cassell and others are surely right that the federal system is not in "chaos," at least if chaos means either utter paralysis or the absence of any thoughtful problem-solving, but even a cursory reading of Judge Cassell's testimony or Professor Berman's website shows a justice system in profound disarray. Any system in which judges are routinely issuing two, or even three, alternative sentences in a single case has big troubles. And the issues so far encountered have been mostly confined to dealing with cases where defendants have already been found guilty. The really tough problems will start to unfold as the system tries to make Blakely and the Guidelines co-exist in grand jury, discovery, trial, and appellate practice.

Whatever the official organs may say, the front line actors I have talked to recognize that no one has any clear idea how to proceed, and that even intelligent efforts to craft constitutionally acceptable procedures are quite likely to prove either constitutionally unacceptable or practically unworkable in the end.

c. "Let 100 flowers bloom."

The third school of thought is more realistic in that it does not pretend that achieving post-Blakely stability will necessarily be easy or quick. Rather, folks in this school tend to think the current sentencing system is badly flawed and that permitting a period of legal uncertainty would allow the collective talents of judges and lawyers to fashion new ad hoc sentencing models that would advance the general project of sentencing reform. The idea here is that sentencing policymakers would profit from studying the solutions arrived at in the field as they strive to apply Blakely and replace the current Guidelines model with something better.

I confess to some sympathy with this approach. I agree that the current federal sentencing system needs significant reform. As an academic, I think it would be intellectually interesting to see what judges, prosecutors, and defense counsel would do given the chance. I don't doubt that policymakers and their advisors would surely find valuable the many interesting solutions to the variety of problems created by Blakely that would doubtless emerge. The difficulty is that those solutions would inevitably vary from judge to judge, district to district, circuit to circuit, and week to week. Only some of the solutions would ultimately survive constitutional challenge and all the cases applying the ones that didn't would have to be re-sentenced. Therefore, as a practical lawyer, I recoil from consciously turning the federal criminal system into a social science experiment.

d. "Don't make a bad situation worse."

Some are concerned that legislating hastily risks creating even more problems than letting the system try to figure things out on an ad hoc basis. I agree entirely. Even uncertainty would be better than legislating a solution that would make the current situation worse. As I will discuss below, I believe that several of the proposed legislative solutions, notably the "Kansas plan" endorsed by the Federal Defenders, have that effect. However, at least two of the proposed legislative alternatives - making the guidelines advisory for a set period, and the proposal outlined below to raise the top of guideline ranges to the statutory maximum.

e. "Don't waste the moment."

A good deal of the opposition to any early legislation stems from a widespread feeling that the current sentencing system is badly flawed, but astoundingly resistant to change, and that the Blakely opinion represents a long-sought opportunity to fundamentally restructure the system. Those who seek such reform worry that any "temporary" legislative solution to the Blakely problem would, if constitutionally sound, tend to become permanent, thus delaying much-needed reform for years. I share this desire for reform, and the worry that the legislation I have suggested below will at least delay it. However, two considerations give me some assurance that in advancing this proposal I am not doing a disservice. First, I think the just and equitable operation of the federal criminal justice system requires stability and predictability, which it will not attain for at least several years without some legislative action. Second, I think those worried that an opportunity may be lost overestimate the transience of the political moment. I venture to predict that Blakely is the beginning of the end of the guidelines in their current form and that, if Congress, the Sentencing Commission, the Justice Department, and others do not move reasonably expeditiously in a cooperative and professional way to make necessary changes, subsequent judicial decisions will force change upon us. My objective is to help, in a small way, to create an orderly process for desirable change.

Therefore, taking everything into consideration and being acutely conscious that I may be dead wrong, I conclude that some legislative action is probably desirable. Absent congressional action, in the near term, the federal courts will continue in turmoil as judges try to negotiate the labyrinth created by Blakely. In the longer term, absent congressional action, either the Guidelines will be transformed by judicial decisions into an unwieldy annex to the criminal code, augmenting the power of prosecutors and decreasing the authority of judges, or more likely the whole structure will be invalidated as unconstitutional and the process of creating a federal sentencing system would have to begin anew. Such a process carries great risks for all those interested in federal sentencing. For Congress and the Sentencing Commission, seventeen years of work would be nullified. For prosecutors, the Guidelines have been a boon; acceding by inaction to the collapse of the current structure with no guarantee of what might replace it would present, at the least, a tremendous gamble. Even those who have no investment in the Guidelines and every interest in radical reform should be very concerned that any replacement could be even more punitive and more restrictive of judicial discretion than the Guidelines themselves.

III. What Legislative Options Does Congress Have?

If Congress wishes to act in the near term, it should only seriously consider legislation that meets four basic criteria:

? First, it has to be simple. Simple to write into statute, simple to understand.
? Second, it has to have easily predictable consequences for the sentencing system. If we can't really predict exactly how something will work in practice, we should not rush to enact it into law.
? Third, it has to solve, or at least greatly ameliorate, the short-term litigation crisis.
? Fourth, and closely related to the third point, it has to be easy to implement. Most importantly, implementing it cannot require immediate massive rethinking of interlocking sets of sentencing, trial, evidence, and appellate rules.

If we can't come up with an interim solution that meets all these criteria pretty well, we're probably better off doing nothing.

Thus far, three basic legislative options have been suggested:

A. Make the Guidelines advisory for a time certain

This option will be ably discussed by Ron Weich. I will not expand upon it here other than to say that it seems to meet all four criteria listed above. It is simple. It's effect on the sentencing system (if not on individual cases) is easily predictable. It would solve the short-term litigation crisis by removing the taint of uncertainty that now attends every sentencing proceeding. Finally, and critically, it would be easy to implement. Because guidelines enhancements would no longer have a necessary effect on a defendant's sentence, Blakely would not be implicated and there would be no need to redesign the entire plea, trial, and sentencing process to ram the round peg of the Guidelines into the square hole of Blakely. The objections to this proposal will doubtless be articulated by others. For myself, I think there is a great deal to be said for a period of advisory guidelines. Indeed, if politically possible, it seems to me a desirable and perhaps the best interim solution.

B. Raise the top of Guideline ranges to statutory maxima

Assuming that one is unwilling to confer increased sentencing discretion on judges, even for a short and experimental period, I believe that the Guidelines structure can be preserved essentially unchanged with a simple modification - amend the sentencing ranges on the Chapter 5 Sentencing Table to increase the top of each guideline range to the statutory maximum of the offense(s) of conviction.

As written, Blakely necessarily affects only cases in which post-conviction judicial findings of fact mandate or authorize an increase in the maximum of the otherwise applicable sentencing range. To the extent that Blakely itself may be ambiguous on the point, the Supreme Court expressly held in McMillan v. Pennsylvania, 477 U.S. 79, 89-90 (1986), and reaffirmed in Harris v. United States, 536 U.S. 545, 122 S.Ct. 2406 (June 24, 2002), that a post-conviction judicial finding of fact could raise the minimum sentence, so long as that minimum was itself within the legislatively authorized statutory maximum. It bears emphasis that Harris was decided only two years ago, and was decided after Apprendi and on the very same day as Ring v. Arizona, 536 U.S. 584 (June 24, 2002), the case whose reading of Apprendi Justice Scalia found so important in his Blakely opinion. Thus, the change I suggest would render the federal sentencing guidelines entirely constitutional under Blakely and Harris.

This proposal could not be effected without an amendment of the Sentencing Reform Act because it would fall afoul of the so-called "25% rule," 28 U.S.C. § 994(b)(2), which mandates that the top of any guideline range be no more than six months or 25% greater than its bottom. The ranges produced by this proposal would ordinarily violate that provision.

I have attached to this testimony as an appendix an annotated version of statutory language that would effectuate this proposal. Detailed discussion of the proposal is contained in the annotations. In addition to raising the top of the guideline range, this proposal contains the following major features:

1. The proposed bill would direct the Sentencing Commission to enact a policy statement providing non-binding guidance to district courts that would read roughly as follows:

(A) the typical case is a combination of offense and offender characteristics adequately represented by a sentencing range the maximum of which should not exceed the minimum by the greater of 25 percent or 6 months;
(B) a sentence that exceeds the minimum of the guideline range by the greater of 25 percent or 6 months should be based on one or more factors thatC
(I) are legally permissible;
(II) advance the statutory purposes of sentencing as set forth in 18 U.S.C. ' 3553(a)(2); and
(III) result in a reasonable sentence.

2. The bill would make any sentence above the guideline minimum appealable on an abuse of discretion standard. The fact that a judge imposed a sentence higher than that suggested by the policy statement for a typical case would be a factor in the determination of whether the judge had abused his or her discretion.

3. The bill would sunset after approximately 18 months.

The practical effect of such a bill would be to preserve current federal practice almost unchanged. Guidelines factors would not be elements. They could still constitutionally be determined by post-conviction judicial findings of fact. No modifications of pleading or trial practice would be required. The only theoretical difference would be that judges could sentence defendants above the top of the current guideline ranges without the formality of an upward departure. However, for the reasons discussed in more detail in the attached annotated version of the proposed bill, the likelihood that judges would use their newly granted discretion to increase the sentences of very many defendants above now-prevailing levels seems, at best, remote.

C. The "Kansas Plan"

The Federal Public Defenders and several other groups have expressed interest in attempting to amend federal law to convert Guideline factors into offense elements that must be pled and proven at trial, or admitted in a plea agreement. This suggestion is modeled on the approach taken by the State of Kansas when its court found state guidelines unconstitutional under Apprendi v. New Jersey. The same approach has sometimes been referred to as "Blakely-izing" the Guidelines.

The most important thing to realize about any effort to "Blakely-ize" the Guidelines is that the Guidelines as now written cannot, as a practical matter, be administered through the vehicle of jury trials and detailed plea agreements. If commanded by statute to treat guidelines factors as elements, judges will work together with lawyers in their districts and will cobble together various arrangements and accommodations in an effort to make the system work. And in due course, sentencings will proceed again. But the effort to make the system work will inevitably require that some sections of the guidelines be ignored or tortured into unrecognizable shapes. Rules will have to be amended, some probably by Congress. Procedures will have to be altered. Some of the resulting arrangements may be better than the Guidelines system we have, some may be worse. But such arrangements will differ markedly from district to district and they will bear only a cousinly resemblance to the Guidelines system now in place.

Perhaps the best way to summarize my views regarding the Kansas plan as interim legislation is that it fails every one of the conditions set out above for desirable near-term legislation.
? It is not simple. Although the defenders have put forward simple-seeming legislative language amending the SRA to implement the scheme, passing those few lines would in short order require revisiting a plethora of other statutes and court rules.
? It is not easy to predict the consequences for the sentencing system or even how it would work in practice.
? It would not solve the short-term litigation crisis, but would surely make it worse.
? Finally, it would not be easy to implement and would indeed require immediate massive rethinking of interlocking sets of sentencing, trial, evidence, and appellate rules.

IV. Conclusion

For reasons I hope to explain in more detail in my oral testimony, I do not see the problem created by the Blakely decision as a short-term glitch for a smooth-running federal criminal sentencing system. Rather, in my view at least, Blakely was decided because the federal sentencing system suffers from serious, long-term, structural problems. The question is not whether reform is needed, but how reform is to be accomplished. My preference is to enable the justice system to work smoothly while reform is crafted. It is to that end that I have offered my proposal and these remarks.

Appendix

[Annotated version of proposed legislation, with comments interlineated in boldface type. F. Bowman]
A BILL

To amend title 18 and title 28, United States Code, with regard to sentencing and for other purposes.

Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled,

SECTION 1. AMENDMENTS TO TITLES 18 AND 28, UNITED STATES CODE.
(a) AMENDMENT TO SECTION 3551.CSection 3551 of title 18, United States Code, is amended byC
(1) redesignating subsection (a) as subsection (a)(1); and
(2) inserting after subsection (a)(1), as redesignated by this amendment, the following:
"(2) Upon a finding of guiltC
(A) the statutory maximum sentence that may be imposed upon a defendant is (i) the maximum penalty provided by the statute defining the offense of conviction, including any applicable statutory enhancement; or (ii) if the statute defining the offense of conviction does not set forth a maximum statutory penalty, the maximum penalty authorized by this chapter for the grade of the offense;
(B) in a case involving multiple counts of conviction, the statutory maximum sentence that may be imposed upon a defendant is the sum of the maximum penalties provided by (i) the statute or statutes defining the offenses of conviction, or (ii) if one or more of the statutes defining an offense of conviction does not set forth a maximum statutory penalty, the maximum penalty authorized by this chapter for the grade of the offense; and
(C) the minimum sentence that may be imposed upon a defendant, if any, is the minimum sentence provided by the statute defining an offense of conviction, except as otherwise provided in this chapter."
[The original version of this provision was suggested by the U.S. Sentencing Commission. The basic idea behind it is that Blakely places great emphasis on the concept of "statutory maximum sentence," but the Sentencing Reform Act can be read to imply by its silence on the point that guidelines maxima have become the new de facto statutory maxima. This amendment would express in no uncertain terms that the statutory maximum sentence under federal law is the maximum sentence provided by statute. It might be suggested that such a provision would be unnecessary under the amended version of the guidelines because the guidelines maxima will be equal to the statutory maxima. However, what we want to stave off is a judicial finding that the new, much softer, 25% suggested ceiling is some form of "statutory maximum." With the addition of this provision, the Court would have to hold that the phrase "statutory maximum sentence" does not mean what Congress has expressly said it means. Indeed, if Congress enacts nothing else in the near term, it might consider enacting this provision. The mere presence of this language in federal statutory law would make it more difficult for a Court applying Blakely to find even the existing guidelines unconstitutional.
I have revised the original Commission language to include the phrase "statutory maximum sentence," and thus to make the effect of the amendment crystal clear.
I have also included language designed to deal with the problem of multiple count convictions. Commission drafters did not think the multiple count provision was necessary because of the existence of case law on consecutive sentences. I do not think it absolutely critical, just clarifying.]
(b) AMENDMENT TO SECTION 3553.CSection 3553 of title 18, United States Code, is amended B
(1) in subsection (a), by striking "The court shall " and inserting "Consistent with any applicable statutory maximum or statutory minimum sentence authorized under section 3551 of this subchapter, the court shall".
(2) in subsection (b)(1), by striking "an aggravating or" and inserting "a";
(3) in subsection (b)(2)(A), by striking clause (i) and by redesignating clauses (ii) and (iii) as clauses (i) and (ii), respectively; and
(4) in subsection(c), by striking "if the sentenceC" and all that follows through "is not of the kind," and inserting " if the sentence is not of the kind,".

(c) AMENDMENT TO SECTION 3742.C Section 3742 of title 18, United States Code is amended B
(1) by striking subsection (a)(3) and substituting a new subsection (a)(3) as follows:
A(3) is greater than the sentence specified as the minimum of the applicable guideline range to the extent that the sentence includes a greater fine or term of imprisonment, probation, or supervised release than the minimum established by the applicable guideline range, or includes a more limiting condition of probation or supervised release under section 3563(b)(6) than the minimum established in the guideline range.@
(2) in subsection (e), by redesignating subsection (e)(4) as subsection (e)(5) and inserting a new subsection (e)(4), as follows:
A(4) if within the applicable guideline range, adequately reflects the statutory purposes of sentencing as set forth in 18 U.S.C. ' 3553(a)(2), is otherwise consistent with the guidelines and policy statements set forth in the applicable Guidelines Manual, and is a reasonable sentence.@
(3) in the paragraph following current section (e)(4), redesignated herein as (e)(5), by adding at the end the following C AWith respect to determinations under subsection (4), the court of appeals shall review the district court=s imposition of a sentence within the applicable guideline range on an abuse of discretion standard.@
(4) in subsection (f)(2), following Aand is plainly unreasonable,@ by inserting "or a sentence imposed within the applicable guideline range constituted an abuse of the district court=s discretion,@.
[The foregoing amendments to 18 USC 3742 create a right of appeal on an abuse of discretion standard for sentences between the guidelines minimum and the guideline/statutory maximum. There are three categories of concern about such a right of appeal.
FIRST, will any constraint on judicial discretion within the new guideline range violate Blakely? I believe that if one were to specify, by statute, a numerical or percentage range (such as six months or 25% above the minimum) that triggered a right of appeal which was not available below that threshold, such a provision would run considerable risk of violating Blakely. In effect, the existence of such a numerical trigger of appellate review would create a presumptive sentencing range immune to appellate scrutiny not dissimilar from current guideline ranges. I have consulted widely on this point with academics, practitioners, and others. Some feel that a numerical appellate trigger of abuse of discretion review would not necessarily violate Blakely, but the majority seem to feel that such a trigger is a feature to be avoided, if possible.
The language suggested here creates no statutorily-endorsed range free from appellate inspection. Instead, it grants district courts discretion to sentence anywhere within the newly expanded range, while granting a right of appeal of any sentence within that range.
SECOND, will an abuse of discretion standard of review be sufficient to ensure (a) that district courts will not begin routinely sentencing above the current guidelines ranges, and (b) that defendants who are sentenced substantially above the current guideline ranges have some meaningful appellate recourse? I think the answer to both parts of the question is yes.
Evidence of judicial behavior over the last decade demonstrates that judges almost never sentence higher than the existing guideline ranges. In 2002, the upward departure rate was 0.8%. US Sentencing Commission, 2002 Sourcebook of Federal Sentencing Statistics, tbl. 27 (2004). And even among defendants sentenced within existing guideline ranges, 60% were sentenced at the bottom of the existing guideline range and an additional 15% were sentenced in the lower half of the range. US Sentencing Commission, 2002 Sourcebook of Federal Sentencing Statistics, tbl. 29 (2004). Only 10% of those sentenced within the existing guideline range (or roughly 6% of all defendants) were sentenced at the top of existing ranges, thus suggesting at least the possibility that, unconstrained by the guidelines, the judge might have gone still higher. Id.
It is certainly possible that some judges in some classes of cases may be disposed to impose longer sentences than they did before. However, the discretion to do so accorded the district court under the proposed law is not without limits. The Guideline policy statements recommended to the Commission in Section 2 below will, if adopted (and they surely would be), express the view of both Congress and the Commission that sentences within the 6-month or 25% range will ordinarily be about right. Because that view will be a part of the Guidelines, a judge's decision to sentence above the suggested range will become one of the factors to be taken into account by a court of appeals reviewing the district court's exercise of its discretion under newly-amended Section 2742(e)(4) above. That is, under the new law, an appellate court determining if a district court abused its discretion will be required to consider whether a sentence within the newly expanded guideline range "adequately reflects the statutory purposes of sentencing as set forth in 18 U.S.C. ' 3553(a)(2), is otherwise consistent with the guidelines and policy statements set forth in the applicable Guidelines Manual, and is a reasonable sentence.@ In short, if a judge imposes a sentence more than 25% above the minimum, that would be one factor suggesting an unreasonable exercise of sentencing discretion.
In my opinion, the strong signal regarding congressional intent embodied in the proposed legislation and resulting policy statement will have a significant limiting effect on judicial disposition to sentence higher than the 25% range.
THIRD, will a right of appeal of all sentences above the guidelines minimum generate a flood of frivolous litigation? I think the answer is no, for several reasons: (a) Under current practice, roughly 35% of all cases are sentenced below the guideline minimum as a result of downward departures. US Sentencing Commission, 2002 Sourcebook of Federal Sentencing Statistics, tbl. 27 (2004). As noted above, of those cases sentenced within the existing guideline range, roughly 60% are sentenced at the guideline minimum. US Sentencing Commission, 2002 Sourcebook of Federal Sentencing Statistics, tbl. 29 (2004). This means that roughly 75% of all cases are now sentenced at or below the guideline minimum. Thus, assuming that current trends continue, at most 25% of current defendants would even be theoretically eligible to appeal under the new law. (b) Once the new right of appeal went into effect, it would create an incentive for prosecutors to agree to sentences at the low end of the range and for judges to sentence at the low end to avoid appeals. I strongly suspect that the effect of this legislation would be to reduce still further the percentage of cases sentenced above the minimum of the guideline range. (c) 96% of all cases are now resolved by plea. An increasingly common feature of such plea agreements is appeal waivers. I have no reason to doubt that U.S. Attorney's Offices will amend their plea waiver language to embrace the new right of appeal. DOJ might attempt to require waiver of any appeal for any sentence in the newly expanded range. However, given the extreme risks that would pose for defendants (who could be sentenced to astronomically high sentences with no recourse) and given the amendment to the guideline policy statements creating a "typical case" zone of 25% or 6 months, the more likely approach will be to require defendants to waive an appeal for any sentence within that zone. (d) In any case, challenges to a district court's sentencing discretion within the newly expanded range will very rarely be successful and only if the sentence is markedly higher than the now-existing ranges. In those cases where challenges are raised to sentences within the newly expanded range, appellate courts will deal with them summarily in the overwhelming majority of cases and give serious attention only to the outliers -- which is, after all, exactly what we want them to do.
(d) AMENDMENTS TO TITLE 28, UNITED STATES CODE.CSection 994 is amendedB (1) by striking subsection (b)(2) and by redesignating subsection (b)(1) as subsection (b); and
(2) in subsection (w)(3), after Athese documents,@ by inserting Aincluding an analysis of district courts= sentences in criminal cases within a guideline range,@.

SEC. 2. DIRECTIVE TO THE UNITED STATES SENTENCING COMMISSION.
(a) IN GENERAL.CThe United States Sentencing Commission shallC
(1) amend the Federal sentencing guidelines to provide that the sentencing ranges established in Chapter Five, Part A consist ofC
(A) the minimum sentence provided for each category of offense and offender characteristics set forth in the United States Sentencing Commission, Guidelines Manual, effective November 5, 2003; and
(B) a maximum sentence equal to the maximum term of imprisonment authorized by the statutory offense of conviction, or in a case involving multiple counts of conviction, the sum of the maximum terms of imprisonment authorized by the statutory offenses of conviction.
(2) consider promulgating a policy statement providing guidance to the court regarding imposition of a particular sentence within the guideline sentencing range providing that:
(A) the typical case is a combination of offense and offender characteristics adequately represented by a sentencing range the maximum of which should not exceed the minimum by the greater of 25 percent or 6 months;
(B) a sentence that exceeds the minimum of the guideline range by the greater of 25 percent or 6 months should be based on one or more factors thatC
(I) are legally permissible;
(II) advance the statutory purposes of sentencing as set forth in 18 U.S.C. ' 3553(a)(2); and
(III) result in a reasonable sentence.
(b) EMERGENCY AMENDMENT AUTHORITY.CThe United States Sentencing Commission shall promulgate amendments pursuant to subsection (a) not later than the effective date of this Act in accordance with the procedure set forth in section 21(a) of the Sentencing Act of 1987 (Public Law 100B182), as though the authority under that Act had not expired.
(c) RECOMMENDATIONS. The United States Sentencing Commission shall, as soon as reasonably possible, make recommendations to the Congress regarding additional amendments to the relevant titles of the United States Code necessitated by enactment of this Act. In accordance with 28 U.S.C. § 994(o), the United States Sentencing Commission shall also study, review and analyze whether any guideline amendments or statutory changes are necessary or desirable in light of Blakely v. Washington, __ U.S. __ , Supreme Court Docket No. 02-1632 (June 24, 2004), or any subsequent related Supreme Court decision and report its recommendations as soon as reasonably possible to Congress. It is the sense of Congress that the United States Sentencing Commission would be materially aided in carrying out the recommendations of this subsection by convening an ad hoc advisory group consisting of individuals and representatives of institutions and professional groups concerned with and knowledgeable about federal sentencing to work with the Commission in formulating appropriate recommendations.
[The first two sentences in this section are drawn from Sentencing Commission language and language suggested by congressional staff. I think there are two phases here. First, the Commission is going to have a prompt 14-day turnaround for the immediate guideline language; nonetheless, there may be statutory wrinkles that we haven't thought of. The Commission's first job should be to study that point and see if there are any other conforming statutory changes that need to be made to make this Act consistent with other statutes. Second, after that job is done, the Commission should turn its attention to the bigger issues posed by Blakely and get hopping on a broad-based study of the implications of Blakely and related cases. Finally, I am not sure that a suggestion to form an advisory group is necessary. I suspect the Commission would do it anyway; still, such a suggestion might serve to ensure that the Commission's study and deliberations are not kept in-house, but reach out to a wide circle of interested persons and institutions.]
SEC. 3. EFFECTIVE DATE.CThis Act shall take effect 14 days after the date of the enactment of this Act and shall apply to all federal sentencings occurring on or after its effective date.
[The original Sentencing Commission draft proposed an effective date of 90 days post-enactment. This seems too long. A primary reason for legislating now is to provide immediate relief and certainty to a system in chaos. Waiting 90 days to some extent vitiates the point of the bill. A 14-day period would be too short if the Commission were being asked to do anything very complicated in the interim, but all they are directed to do before the effective date is (a) knock the tops off the guideline ranges, and (b) promulgate a policy statement whose text is set forth in the Act, and which in any case the Commission itself drafted. The text of the guideline amendments necessary to accomplish (a) will take a little thought, but since the Commission knows this bill might be coming, it is not unreasonable to expect them to have a drafting solution in place immediately.
That said, it does make sense to set the effective date off some short period in the future in order for the Commission to pass its guideline amendments and policy statements and get them into effect on the same day as the statutory provisions of the bill. You don't want a hiatus where the statutory changes have become effective but the guidelines change have not.
Finally, the bill should expressly apply to sentencings on or after its effective date. There will be some sentencings to which the bill may not be applicable for ex post facto reasons, but because the bill essentially reenacts the existing Guidelines, there will be very few defendants who will be able to claim that they are adversely affected by application of the new rules. ]
SEC. 4 SUNSET.C Effective on and after January 1, 2006, the provisions of this Act and the amendments to statutes made pursuant to this Act shall no longer have the force of law unless reenacted by Congress, and the amendments to the Federal Sentencing Guidelines and policy statements made pursuant to section 2(a) of this Act shall no longer have the force of law unless repromulgated by the U.S. Sentencing Commission and approved by Congress pursuant to law.
[The sunset language suggested by congressional staff would sunset the bill if the Supreme Court finds the original guideline constitutional. The following language was suggested: "This Act shall cease to have effect if the United States Supreme Court determines that the United States Sentencing Guidelines are not subject to the jury and evidentiary proof requirements set forth in Blakely v. Washington." This approach has two effects that may be thought objectionable.
First, as we discussed in Washington on Wednesday, it is very difficult to craft language that would really be self-executing. By which I mean that the Supreme Court in addressing the application of Blakely to the Guidelines could do all sorts of things not easily captured in the suggested language, or indeed in any statutory language. For example, assume that the Court finds that Blakely applies to "upward departures" under the guidelines (which is what the upward adjustments in the Washington scheme were called), but not to determinations of guidelines factors generating ranges? Although such an approach would be illogical, it's precisely the sort of face-saving compromise the Court might reach for if one of the Blakely majority recoils from wrecking the guidelines but wants to seem consistent with Blakely. If that were to be the Court's conclusion, the Court would have found that the jury and proof requirements applied to one part of the Guidelines, but not to the rest of it. Would the bill sunset or not? Alternatively, suppose that the Court were to impose SOME new jury and proof requirements on the Guidelines, but slightly different ones than those "set forth in Blakely v. Washington." Would the bill sunset or not? Or suppose you get a 4-1-4 decision with four justices applying Blakely to the Guidelines and one concurring in the result on ambiguous grounds. Does the bill sunset or not? I do not think we should introduce this sort of uncertainty into an already complicated situation. Hence, any sunset provision should involve a date certain, rather than a difficult-to-define contingency.
Second, it has been my impression that a number of participants in this process are desirous of ensuring that any legislation sunset in order that an interim solution to an immediate problem does not become the permanent condition of the sentencing system without further careful deliberation and legislative action. I agree with this approach. I think the approach I am suggesting serves the purpose of restoring stability to a system in crisis. However, many of the underlying conditions that created the crisis will persist for the foreseeable future. I believe a sunset provision is necessary to focus the attention of Congress, the Commission, and other sentencing actors on the need for prompt and careful re-evaluation of the federal sentencing system in light of Blakely and other factors.
I have selected a roughly 18-month sunset period because at least that much time will certainly be required for the Court to make a decision, for the rest of us to absorb it, and for the Commission and others to consult meaningfully on what changes to the existing system would be desirable. January 1, 2006 also seems a reasonably felicitous point in the political cycle because it would require action in late 2005, the calendar year prior to the next election.]
SEC. 5 SEVERABILITY.--- If any provision of this Act, an amendment made by this Act, or the application of such provision or amendment to any person or circumstance is held to be unconstitutional, the remainder of this Act, the amendments made by this Act, and the application of the provisions of such to any person or circumstance shall not be affected thereby.