Punishment Without Conviction: Controlling the Use of Unconvicted Conduct in Federal Sentencing

Transcript

1 Be rkele nal of C rimin al L aw y J our | Issue 2 Article 3 Volume 17 2012 nt W ithout C on viction: C on trol Puni g the shme lin ncon victed C onduc t in F ede ral Use of U in g Sentenc rald Leon Ge ard Chr istine D iet er Recomme nded Citation ten al Sen eder t W nconvicted Co se of U he U ntrolling t nviction: Co ithout Co , cing Gerald Leonard and Christine Die ter, Punishmen nduct in F erkeley 17 B . L. 260 (2012). rim J. C Link to publisher version (DOI) https://do i.org/10.15779/Z382D08 epository. It has been aw Journals and Related Materials at Berkeley Law Scholarship R n access by the L ree and ope This Article is brought to you for f accepted for inclusion in Be e epository. For mor rkeley Law Scholarship R riminal Law by an authorized administrator of Be rkeley Journal of C [email protected] information, please contact .

2 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted I F ALL 2012 SSUE 17.2 Punishment Without Conviction: Controlling the Use of Unconvicted Conduct in Federal Sentencing ** * Gerald Leonard and Christine Dieter A BSTRACT Federal sentencing law is widely applied to punish offenders not only for the offenses of which they have been convicted, but also, in the same proceeding, for offenses of which they have not been convicted. Unlike many scholars, we accept that federal courts can, in the right circumstances, legitimately enhance sentences for facts and conduct found at sentencing, even when those facts and conduct constitute uncharged offenses or even charges on which the defendant actually won an acquittal. But we argue that in identifiable cases, the use of such sentencing facts does cross the line from appropriate contextualization of the offense of conviction to punishment for a separate offense of which the defendant has never been convicted. We demonstrate that crossing this line contravenes the Sentencing Reform Act, the Federal Sentencing Guidelines, and the Constitution. We then offer a principle and a mode of analysis for ensuring that courts punish only for offenses of conviction, even as they do substantial fact- finding at sentencing. We examine cases of federal sentencing for second-degree murder to explain how this principle works and then explain the benefits and challenges of applying the principle more generally. * Professor of Law, Boston University School of Law. Many thanks to Ken Simons and Carissa Hessick for helpful comments. ** J.D., 2012, Boston University School of Law. 1 Published by Berkeley Law Scholarship Repository, 2012

3 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 261 P ITHOUT I NTRODUCTION Every person with a working understanding of federal sentencing law knows this apparently shocking fact: offenders have long been sentenced to prison for offenses of which they have never been ne convicted. Once a federal court has convicted a defendant of o offense then that conviction allows the court to sentence for any number of other offenses that were never even charged or that were charged — but resulted in acquittal. Although this practice has long been common, it only began to provoke sustained outrage among legal scholars after the 1 advent of the U.S. Sentencing Guidelines in the late 1980s. Before then, it had occurred quietly in trial courts, where no law governed the practice and no appellate review disciplined the outcomes. With the advent of the Guidelines, however, it became open and obvious that individuals say, drug trafficking — faced sentences — convicted of one offense say, a — enhanced in sometimes huge proportions by a separate offense — that had either gone uncharged or of which they had been murder acquitted. Sometimes, such “unconvicted conduct” can be and is used for appropriate purposes. But in some cases, courts simply sentence offenders for crimes of which they have never been convicted. The issue we raise here is complicated. Unlike many scholars, we believe that the use of unconvicted conduct at sentencing is often 2 At the same time, we share legitimate, as the Supreme Court has held. much of the outrage of scholars who object to this practice. It is with Elizabeth Lear’s 1993 tempting to agree, for example, condemnation of the use of “acquitted conduct” when Lear offers the 3 example of United States v. Rivera-Lopez In that case, the defendant was . convicted of a pair of drug charges but acquitted on a third count, 4 The sentencing judge possession of three kilograms of cocaine. nevertheless found that the defendant had possessed the three kilograms and that the possession was “relevant conduct,” to use the Guidelines 5 As Lear reports, “the addition of the relevant term, for this sentencing. conduct resulted in the identical punishment range[] which the defendant[] would have encountered had [she] been convicted on all 1 See, e.g. , Susan N. Herman, The Tail that Wagged the Dog: Bifurcated Fact-Finding Under the Federal Sentencing Guidelines and the Limits of Due Process , 66 S. EV AL L. R . . 289 (1992); C Is Conviction Irrelevant? , 40 UCLA Elizabeth T. Lear, L. R EV . 1179 (1993); Kevin R. Reitz, , 45 S Sentencing Facts, Travesties of Real-Offense Sentencing . 523, 573 (1993). R EV . L. TAN 2 See infra Part III.B. 3 United States v. Rivera-Lopez, 928 F.2d 372 (11th Cir. 1991). 4 Id. at 372. 5 See id. at 373. https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 2 DOI: https://doi.org/10.15779/Z382D08

4 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 262 B OURNAL OF 6 Since the defendant’s acquittal on a major charge had literally counts.” no effect on the ultimate sentence, it is hard to deny that the judge punished her for that very charge. To take an even starker example, consider the sentencing of Nelson Frias. Frias was convicted on two weapons charges — for which the Guidelines indicated a sentence of no more than eighteen months — 7 At sentencing, the and was simultaneously acquitted on a drug charge. trial judge disregarded the jury’s acquittal, found that Frias actually had committed the drug offense, and calculated a Guidelines sentence in the 8 In order to neighborhood of twenty years based on his drug quantity. — notwithstanding the acquittal — the fully sentence for the drug quantity trial judge took the unusual step of ordering the ten-year statutory maximums for the weapons convictions served consecutively, rather than 9 concurrently. Frias’ sentence jumped from twelve to eighteen months, 10 which would have applied without the drug facts, to twenty years. Even after the Second Circuit ordered the trial judge to consider a downward departure from the Guidelines sentence, Frias was still 11 sentenced to twelve years. Even in the age of the Guidelines, then, punishment for 12 It is done now with more unconvicted offenses continues unabated. 13 but there is still little transparency, and subject to the regularities of law, official attention to the risk that some defendants face punishment for unconvicted offenses. So we agree with Lear and others that there is a serious problem here, but we do not accept their analysis of the problem or their 6 note 1, at 1197. Lear, supra 7 United States v. Concepcion, 983 F.2d 369, 374-75 (2d Cir. 1992). 8 at 376, 385-86. Id. 9 Id. 10 Id. at 376, 389. 11 United States v. Frias, 39 F.3d 391, 392 (2d Cir. 1994). 12 See, e.g. 37-38 (11th Cir. 2010) (increasing , United States v. Smith, 370 F. App’x 29, sentence for illegal gun possession based on conduct underlying acquittal for carrying a firearm in furtherance of an assault on law enforcement officials); United States v. Ashqar, 582 F.3d 819, 823-25 (7th Cir. 2009) (applying terrorism enhancement based on acquitted conduct to increase sentence for obstruction of justice and criminal -70 (4th Cir. 2007) contempt);; United States v. Sampson, 245 F. App’x 263, 269 (increasing sentence for perjury and conspiracy to commit perjury based on pending unrelated murder charges). 13 The Guidelines were originally binding on the federal courts. After United States v. Booker, 543 U.S. 220 (2005), they are now formally advisory, but, in fact, they still exercise substantial control over federal sentencing, see Rita v. United States, 551 U.S. 338, 351 (2007), and, for our purposes, remain nearly as important as when they were See infra Part III. formally binding. 3 Published by Berkeley Law Scholarship Repository, 2012

5 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 263 P ITHOUT prescriptions for reform. Scholars like Lear have objected to such sentences primarily on grounds that we find inadequate. In particular, Lear and others have insisted that conduct underlying an acquitted 14 charge should be entirely off-limits at sentencing (or very nearly so). But that position reflects a misunderstanding of the larger problem. The issue, after all, is not so much the use of acquitted conduct but more conduct. That is, we should be less focused broadly the use of unconvicted and more broadly concerned with the acquittals on judicial disregard of conduct that has not been proven at imposition of punishment for any trial (or by plea). Whether that conduct was “acquitted” or never charged at all, we should be very worried when it suddenly surfaces at sentencing, dramatically enhancing the defendant’s punishment. Moreover, once one sees that the problem is much bigger than acquitted conduct, and that the issue of unconvicted conduct arises whenever any fact-finding is done at sentencing, it becomes implausible to advocate a universal rule against the use of such conduct at sentencing. The Supreme Court has firmly rejected such a rule, and to eliminate the use of unconvicted conduct at sentencing would radically alter centuries of established practice, pushing sentencing closer to a straight-time system — one in which every offender convicted of the same offense serves exactly the same time. The value of some judicial fact-finding and discretion at Frias. In that case, it sentencing is readily illustrated by another look at probably made good sense to take account of Frias’ drug activities in sentencing him for his weapons offenses, whether the drug facts came in despite an acquittal on those charges (as actually happened) or despite charge them at all at the trial stage (also a the prosecution’s choice not to frequent occurrence). Once at sentencing, the judge found them true by a preponderance of the evidence. The judge then logically determined that a weapons offender is more culpable and more dangerous if he possesses the weapons to facilitate drug deals than if he possesses them, for example, for self- defense. Of course, Frias’ jury declined to find every element of the drug charge proven beyond a reasonable doubt; at sentencing, however, the judge remained free to find it more likely than not that Frias was sufficiently involved with the drugs that they should contextualize his weapons conviction. Thus, a modest enhancement for the drugs would have been legitimate and even appropriate. But the 14 Lear, supra note 1, at 1185; Carissa Byrne Hessick & F. Andrew Hessick, Recognizing Constitutional Rights at Sentencing , 99 C . L. R EV . 47, 58-60 (2011); Reitz, supra note 1 , ALIF Reforming the Federal Sentencing Guidelines’ Misguided Approach to at 547-52; David Yellen, , 58 S Real-Offense Sentencing . L. R EV . 267, 275 (2005). TAN https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 4 DOI: https://doi.org/10.15779/Z382D08

6 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY C RIMINAL L AW [Vol. 17:2 OURNAL OF 264 B J Frias’ enhancement indicates that he was punished much sheer size of less for the weapons than for the drug deal itself, and the same is true of many other federal sentencings. Using unconvicted conduct to contextualize the conviction is appropriate, but punishing an offender for the unconvicted conduct is not. Thus, the conundrum: How does one determine, as a general matter, when an enhancement has crossed the line from appropriately accounting for the convicted offense in its full context to inappropriately punishing for a separate offense, of which the defendant has never been convicted? After more than twenty years of Guidelines sentencing, the federal system still operates without meaningful rules to prevent punishment beyond the offense of conviction. This article will not offer a complete resolution of that conundrum, but it will offer an approach that can be readily applied in certain categories of cases and that might serve as a starting point for extending the discipline of law into the harder cases. We begin with the Supreme Court’s own affirmation that a federal sentence must indeed punish only for convicted offenses, a principle affirmed in the very case in which the Court approved the use of acquitted conduct at 15 sentencing. To justify the use of acquitted conduct, the Court pointed to the difference in standards of proof —“beyond a reasonable doubt” at the conviction stage and mere preponderance at the sentencing stage — and the fact that general verdicts do not represent particularized findings 16 Thus, an acquittal on one charge need not exclude of fact by the jury. the evidence underlying that charge from the sentencing proceeding on a 17 related charge. But the Court’s more important principle was that, while unconvicted conduct could legitimately be used to contextualize the offense of conviction, it could not be used to punish for an unconvicted 18 offense as such. In this article, we argue that, although it is challenging 15 United States v. Watts, 519 U.S. 148, 154-55 (1997). See 16 Id. at 155-56. 17 Id. at 155. For example, an acquittal on a charge of possessing cocaine with intent to distribute does not necessarily represent a finding that the defendant did not commit the offense but only that the jury had some modest doubt on the question of guilt. Or, even if the jury can be said to have affirmatively found the defendant innocent on at least one element of the offense — say, the intent to distribute — that does not mean that the jury found the defendant innocent of all elements of the offense, such as the simple possession of the drugs. In any case, the crucial point is that a general verdict leaves us ngs of fact to speculate what the jury’s precise findings were and whether particular findi at sentencing really are in conflict with the jury’s general verdict. 18 See id. at 154-55. 5 Published by Berkeley Law Scholarship Repository, 2012

7 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D [Vol. 17:2 UNISHMENT C ONVICTION ITHOUT P 265 W (how to figure out how to implement this principle in a case like Frias many years exactly are too many for the additional drugs proven to the judge at sentencing?), there are cases where its application is straightforward. Our prime example of such a case involves a conviction for second-degree murder. Under federal law, first-degree murder must take 19 The absence the form of either premeditated murder or felony murder. of premeditation or an underlying felony means that the offender has, at 20 But what if the sentencing worst, committed second-degree murder. judge finds premeditation where the jury did not? Can the judge enhance degree murderer’s sentence on the basis of a post -conviction a second- finding that the murder was, in fact, premeditated? We argue that a judge who does so has clearly crossed the line to punishing for an — no matter the size of the unconvicted first-degree murder enhancement. The combination of premeditation and unlawful killing is by statutory definition first-degree murder, not second-degree murder. This judge thus punishes the offender for a different offense from the offense of conviction. And that, we argue, is inconsistent with the Sentencing Reform Act of 1984, inconsistent with basic principles of the Guidelines, and inconsistent with the Constitution as the Supreme Court has read it. We begin, in Part II, by describing a couple of second-degree murder cases that came out differently in the Fourth and Tenth Circuits. Next, in Part III, we look at the Sentencing Reform Act of 1984, the Guidelines’ treatment of “relevant conduct” at sentencing, and th e constitutional limits on post-conviction fact-finding indicated by the case law of the Supreme Court. From these sources, we argue in Part IV that 19 18 U.S.C. § 1111(a) (2006). Murder is the unlawful killing of a human being with malice aforethought. Every murder perpetrated by poison, lying in wait, or any other kind of willful, deliberate, malicious, and premeditated killing; or committed in the perpetration of, or attempt to perpetrate, any arson, escape, murder, kidnapping, treason, espionage, sabotage, aggravated sexual abuse or sexual abuse, child abuse, burglary, or robbery; or perpetrated as part of a pattern or practice of assault or torture against a child or children; or perpetrated from a premeditated design unlawfully and maliciously to effect the death of any human being other than him who is killed, is murder in the first degree. Any other murder is murder in the second degree. Id. 20 Id. In contrast to first-degree murder, second-degree murder is simply defined as Id. “[a]ny other murder.” https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 6 DOI: https://doi.org/10.15779/Z382D08

8 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 266 B OURNAL OF a second-degree murder conviction cannot authorize a sentence based on a post-conviction finding of premeditation. We argue further, in Part V, — that recognition of the underlying principle that the sentence must be carefully tailored to punish only for the offense of conviction and not for — ty anything else has the potential to impart some sorely needed rationali to federal sentencing. T I. ILEMMA HE D 21 Consider two murder prosecutions: United States v. Kelly , a 1993 22 , United States v. Barber case from the Tenth Circuit, and a 1997 case from the Fourth Circuit. In the first, Jimmy Gene Kelly admitted to killing his victim, and the jury only considered whether the murder was 23 premeditated. Strong evidence pointed towards premeditation: a history of bad relations with the victim, allegations that the victim spread rumors about Kelly’s family, and Kelly’s statement to a friend eight days 24 On the day of the before the murder that he intended to kill the victim. 25 crime, Kelly invited the victim fishing but did not bring a fishing rod. Notwithstanding all of this evidence, the jury found that Kelly had not premeditated the murder and accordingly convicted him of second- 26 degree murder. The second-degree conviction pointed to a Guidelines base 27 which , when combined with Kelly’s offense level of thirty-three, 21 United States v. Kelly, 1 F.3d 1137 (10th Cir. 1993). 22 United States v. Barber, 119 F.3d 276 (4th Cir. 1997). 23 , 1 F.3d at 1138. Kelly 24 Id. at 1138-39. 25 Id. 26 Id. at 1138. As discussed earlier, premeditation marks the distinction between federal first-degree and second-degree murder. notes 19-20 and accompanying text. See supra 27 The United States Sentencing Guidelines require a sentencing judge to go through a number of technical steps to produce a final sentence. First, the judge must determine a base offense level for the particular offense, as specified in the Guidelines manual. Then the judge must add or subtract points to reflect the particular facts of the crime and the offender, ending up with a final offense level. At the same time, the judge must determine a criminal history score, again adding up points as dictated by the Guidelines to express the offender’s history of prior convictions and other criminal activity. The judge then refers to the Guidelines sentencing table, finding the offense level on one axis and the criminal history score on the other axis, locating the box in the table where those two scores meet, and finding there the narrow sentencing range within which the judge must presumptively choose a sentence. Even then, however, the judge is free to sentence outside that range if she or he can explain adequately what makes this case 7 Published by Berkeley Law Scholarship Repository, 2012

9 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 267 P ITHOUT criminal history score, required a prison sentence of 151-188 months. After having heard the evidence at trial, however, the judge declared that “the record was pretty clear to me as the judge listening to it that 28 premeditatio He increased the offense level to forty- n was present.” one and sentenced Kelly to 360 months, almost twice as long as the maximum sentence within the presumptive Guidelines range for second- 29 degree murder. 30 In Barber , the Fourth Circuit confronted a similar situation. Defendants Anthony Barber and David Hodge pled guilty to second- degree murder for shooting the leader of a drug distribution scheme in which Barber had become entangled and robbing the victim of the fifty 31 dollars he carried. In a statement made to the Government as a condition of his plea bargain, Barber admitted that the murder was 32 Both defendants received a base offense level of thirty — premeditated. the Guidelines level for second-degree murder minus a reduction for — which led to a sentencing acceptance of responsibility for their crimes 33 range of 97-121 months. However, the sentencing judge in this case also found premeditation, departed upwards to an offense level of thirty- seven, and sentenced each defendant to 210 months — as in Kel ly , almost 34 double the Guidelines’ maximum presumed sentence. In each case, the sentencing judge had to decide whether to increase the defendant’s sentence pursuant to a finding of premeditation. Could this finding appropriately contextualize the second-degree murders, thus punishing the offenders for a second-degree murder made all the worse because it was premeditated? Or would the judge’s reliance on a finding of premeditation amount to punishment for a distinct — — crime which does not exist in premeditated second-degree murder the statute books and of which neither had been convicted? The Fourth and Tenth Circuits reached contradictory conclusions: the Fourth Circuit found it obvious that premeditation makes a second-degree murder an unlike the run of the mill case. See the Guidelines Manual for details. U.S. G UIDELINES M ANUAL ch. 1, pt. A, introductory cmt. 4(h) (1987). S ENTENCING 28 Kelly , 1 F.3d at 1139. The judge also considered use of dangerous instrumentality, restraint of the victim, and extreme conduct as potential aggravating factors warranting Id. at 1138. sentencing departure from the Guidelines. 29 Id. 30 United States v. Barber, 119 F.3d 276 (4th Cir. 1997). 31 at 279. Id. 32 Id. 33 Id. 34 The conditions of Barber’s plea agreement prevented the judge from using Id. Barber’s statement regarding premeditation against him;; however, the judge did use Id. premeditation as one factor justifying the sentencing departure in Hodge’s case. https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 8 DOI: https://doi.org/10.15779/Z382D08

10 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 OURNAL OF 268 B especially heinous instance of that offense and considered it 35 the Tenth Circuit decided that premeditation was simply accordingly; 36 not an available finding in a sentencing for second-degree murder. These decisions highlight a diversity of sentencing practice and theory but not in — that, in effect, allows federal judges in most of the country to consider premeditation in sentencing for second- — the Tenth Circuit 37 degree murder. Although both courts relied heavily on the structure and purpose of the Guidelines, their decisions raise questions more fundamental than just the proper interpretation of the particular guideline for second- degree murder. Both decisions acknowledge judges’ wide -ranging authority to consider all relevant information at sentencing, including 38 information underlying acquitted offenses and uncharged conduct. But they force us to ask whether we can nonetheless draw a line between, on the one hand, those facts that may properly alter a defendant’s sentence by providing appropriate context and, on the other, those facts that must be excluded because they would lead improperly to punishment for some offense other than the offense of conviction. This problem appears not just in murder cases but is embedded deeply and widely in any system that rests on post-conviction fact-finding. 35 The Fourth Circuit affirmed the trial court’s decision to sentence Barber and Hodge for premeditated second-degree murder, but only by virtue of an even split in an en banc court. Barber , 119 F.3d at 284-85. Notwithstanding this even split, which rendered any opinion dictum, Judge Wilkins, who had only recently completed his term as Chair of the Sentencing Commission, wrote an opinion seeking to legitimate such uses of sentencing facts, even when they were apparently inconsistent with the offense at 287-90. That opinion remains the best statement of the law for the of conviction. Id. Fourth Circuit even though it technically has no authority. 36 United States v. Kelly, 1 F.3d 1137, 1140-41 (10th Cir. 1993). 37 The Tenth Circuit has remained steadfast in its position. See United States v. Hanson, 264 F.3d 988 (10th Cir. 2001) (affirming trial judge’s refusal to base departure in Kelly , 1 F.3d at 1140-41 second -degree murder case on premeditation and robbery); ditation in second-degree (reversing lower court’s departure decision based on preme cf. United States v. Wolfe, 435 F.3d 1289, 1304 (10th Cir. 2006) (“The district murder); court in this case, in departing upward from the involuntary-manslaughter range halfway to the range provided for second-degree murder, never explained why Wolfe’s conduct, which resulted in two involuntary-manslaughter convictions, should instead be degree murder involving malice aforethought.”). treated as more like second- 38 See infra Part III.A for a discussion of the Supreme Court’s decisio ns addressing the use of uncharged and acquitted conduct in sentencing. 9 Published by Berkeley Law Scholarship Repository, 2012

11 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 269 P ITHOUT : HE ODERN T AW OF F EDERAL S ENTENCING M S TATUTE , II. L , C ONSTITUTION G UIDELINES A. Federal Sentencing Guidelines: Grappling with Judicial Discretion In 1984, Congress passed the Sentencing Reform Act, which 39 created the United States Sentencing Commission. The Act charged the Commission with developing a system of mandatory sentencing 40 guidelines that would impose law on judges’ sentencing decisions. Earlier in the twentieth century, and stretching back into the nineteenth century, sentencing practice had often followed a rehabilitative model, focusing on the defendant’s individual crime and characteristics and 41 seeking to impose a sentence specifically addressing these particularities. Ideally, the individually tailored sentence would maximize the chances for successful “treatment” of the particular offender and for her or his ultimate return to society as a productive citizen. This model suggested to the federal courts that no concrete rules should constrain the trial y flexible discretion in tailoring a sentence for any court’s necessaril 42 particular offender. at the height of the vogue for rehabilitation (or In fact, in 1949 — the Supreme Court “reformation,” as it was then commonly called)— Williams v. New York that core trial rights, such as the famously held in right to cross-examine adverse witnesses and even to have full access to 43 the evidence used against oneself, simply did not apply at sentencing. That holding depended on the belief that the sentencing judge must oper “treatment” for the offender, a task that demanded determine the pr the judge’s possession of the fullest possible information about the offender, even if the latter’s rights were curtailed to facilitate the gathering of that information. The Court celebrated the success of 39 Sentencing Reform Act of 1984, Pub. L. No. 98-473, § 217, 98 Stat. 1987, 2017. 40 Rachel E. Barkow, Recharging the Jury: The Criminal Jury’s Constitutional Role in an Era of Mandator y Sentencing , 152 U. P . L. R . 33, 88-89 (2003) (describing the development of A EV the Sentencing Guidelines). 41 AVID J. R OTHMAN , C ONSCIENCE AND C ONVENIENCE 31-32, 60 (1980); Frank O. D Debacle: How the Supreme Court Has Mangled American Sentencing Law and How Bowman, III, , 77 U. C It Might Yet Be Mended HI L. R EV . 367 (2010). . 42 For a history of this theoretical approach to criminal law in the twentieth century, see Gerald Leonard, Towards a Legal History of American Criminal Theory: Culture and Doctrine from Blackstone to the Model Penal Code , 6 B . C RIM . L. R EV . 691, 803-826 (2003). UFF 43 Williams v. New York, 337 U.S. 241, 245-50 (1949). https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 10 DOI: https://doi.org/10.15779/Z382D08

12 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY C RIMINAL L AW [Vol. 17:2 OURNAL OF 270 B J modern rehabilitationism and, with utterly benevolent intentions, used that theory to render sentencing a sphere of nearly complete lawlessness: Modern changes in the treatment of offenders make it more necessary now than a century ago for observance of the distinctions in the evidential procedure in the trial and sentencing processes. . . . Under the practice of individualizing punishments, investigation techniques have been given an important role. Probation workers making reports of their investigations have not been trained to prosecute but to aid offenders. Their reports have been given a high value by conscientious judges who want to sentence persons on the best available information rather than on guesswork and inadequate information. . . . We must recognize that most of the information now relied upon by judges to guide them in the intelligent imposition of sentences would be unavailable if information were restricted to that given in open court by witnesses subject 44 to cross-examination. Consistent with this ruling, criminal sentences were rarely even subject to appellate review, since appellate courts were poorly equipped to second- guess judgments made in the trial courts about proper, individualized 45 Professor Kate Stith and Judge José Cabranes later “treatment.” described the extraordinary scope of sentencing discretion with some horror: What made sentencing authority truly extraordinary was not the broad discretion the judge exercised, but, rather, the fact that the decision was virtually unreviewable on appeal. The lack of appellate review meant that the or even the bizarre — — unreasonable or inexplicable decision at this stage was beyond correction. In addition, no common standards or principles were articulated to 46 guide the exercise of judgment in sentencing. 44 Id. at 247-50 (citations omitted). 45 See Dorszynski v. United States, 418 U.S. 424, 431, 443-44 (1974). 46 Kate Stith & José A. Cabranes, Judging Under the Federal Sentencing Guidelines , 91 Nw. U. L. Rev. 1247, 1251-52 (1997) (citation omitted). 11 Published by Berkeley Law Scholarship Repository, 2012

13 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT 271 C ONVICTION [Vol. 17:2 W P ITHOUT By the 1970s, academics and legislators increasingly voiced concern that this individually tailored model of sentencing allowed for 47 Similarly situated defendants committing too much judicial discretion. similar crimes often received wildly disparate sentences across, and even within, jurisdictions. Some critics worried that the disparities might, in 48 Others felt that pervasive judicial discretion part, be racially motivated. meant that too many offenders got off with lighter sentences than they 49 At the same time, the great confidence in rehabilitation that deserved. had informed the Williams opinion and dominated penology in the mid- 50 twentieth century began to fade rapidly. reflected these The Sentencing Reform Act of 1984 (“SRA”) shifts in attitude. The SRA turned federal sentencing on its head in important respects, replacing boundless discretion with precisely articulated mandatory guidelines for federal sentencing, seeking to ensure that similar offenders would receive similar sentences, and rejecting rehabilitation altogether as a basis for any sentence of incarceration. The Commission shall insure that the Thus, the SRA provided that, “ guidelines reflect the inappropriateness of imposing a sentence to a term of imprisonment for the purpose of rehabilitating the defendant or providing the defendant with needed educational or vocational training, 47 . F RIENDS S ERV . C OMM ., S TRUGGLE FOR J USTICE A OBERT O. M 143-44 (1971); R D ENTENCING : T HE D ECISION AS TO T YPE , L ENGTH , AND C ONDITIONS OF AWSON , S 3-4 (1969); M. D A LAN ENTENCE ERSHOWITZ , T WENTIETH C ENTURY F UND T ASK S F C S ENTENCING , F AIR AND C RIMINAL P UNISHMENT 4 (1976); ORCE ON ERTAIN M E. F RANKEL , C RIMINAL S ENTENCES 103-04 (1973); L ESLIE T. W ILKINS ET ARVIN TRUCTURING ., G UIDELINES : S ENTENCING J UDICIAL D ISCRETION , at xiii-xviii S AL , 50 J. C Legal Values and the Rehabilitative Ideal (1978); Francis Allen, . L. RIM C OLICE S CI . 226, 227-29 (1959); Stith & Cabranes, supra note 46, at RIMINOLOGY & P 374-75. 48 , 24 G OLDEN The Dilemma of Difference: Race as a Sentencing Factor Placido G. Gomez, G U. L. R . 357, 358-60 (1994) ATE (“In 1979, blacks accounted for 10.1% of this EV country’s adult male population, yet occupied 48% of the beds in our state prisons.”). 49 supra note 41, at 374. Bowman, 50 See, e.g. , F RANKEL , supra note 47 , at 92 (“It is not disagreeable for people ‘in the community and in the field of criminal justice,’ assuming they have no more pressing things to do, to devote their energies to the attempted proof or accomplishment of — and vicious — is to cage prisoners for universal redeemability. What is disagreeable indeterminate stretches while we set about their assured rehabilitation, not knowing what to do for them, o r really, whether we can do any useful thing for them.”);; ORVAL M ORRIS , T HE F UTURE OF I MPRISONMENT 17 (1974); D ERSHOWITZ , supra note N 47 , at 3 (“By failing to administer either equitable or sure punishment, the sentencing if anything permitting such wide latitude for the individual discretion of system — various authorities can be so signified — undermines the entire criminal justice structure .”);; See generally, A H IRSCH , D OING J USTICE chs. 2-4 (1976). NDREW VON https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 12 DOI: https://doi.org/10.15779/Z382D08

14 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 272 OURNAL OF B 51 Instead, it aimed to medical care, or other correctional treatment.” “provide certainty and fairness in meeting the purposes of sentencing, avoiding unwarranted sentencing disparities among defendants with similar records who have been found guilty of similar criminal 52 conduct.” The SRA abandoned rehabilitation and, in § 3553(a)(2), defined the purposes of sentencing as follows: “ (A) to reflect the seriousness of the offense, to promote respect for the law, and to provide just punishment for the offense; (B) to afford adequate deterrence to criminal conduct; (C) to protect the public from further crimes of the 53 defendant . . . .” Rehabilitation of a sort —that is, providing “the defendant with needed educational or vocational training, medical care, or other correctional remained one purpose of treatment in the most effective manner”— 54 sentencing, as reflected in subsection (D) of this provision; but, as indicated above, it could no longer justify incarceration, only certain 55 forms of probation and treatment outside of prison. Williams In short, the SRA rejected a fundamental premise of — that incarceration was meant to rehabilitate and that judges must therefore have access to unlimited information about the offender to effect that rehabilitation —and focused instead on providing “just 56 punishment,” “adequate deterrence,” and a measure of incapacitation. In doing so, the Guidelines sentencing sought to equalize sentences among “defendants with similar records who have been found guilty of 57 Rehabilitation and discretion were out, similar criminal conduct.” by a simple theory that being “found guilty” of particular replaced conduct should bring predictable sentencing consequences and that prior criminal records should aggravate those consequences in predictable ways. The resulting sentences would be presumptively adequate to punish justly, deter, and incapacitate. Moreover, by introducing 51 see Tapia v. United States, 131 S. Ct. 2382, 2393 (2011) 28 U.S.C. § 994(k) (2006); (“[A] court may not impose or lengthen a prison sentence to enable an offender to complete a treatme nt program or otherwise to promote rehabilitation.”). 52 28 U.S.C. § 991(b)(1)(B) (Supp. II 2008). 53 18 U.S.C. § 3553(a)(2) (Supp. IV 2010). 54 Id . 55 See Tapia , 131 S. Ct. at 2390. 56 § 3553(a)(2). 57 Id . 13 Published by Berkeley Law Scholarship Repository, 2012

15 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 273 P ITHOUT substantive appellate review of sentences, the SRA ensured the sentencing courts’ faithful implementation of the above theory and 58 purposes of punishment. In theory, the United States Sentencing Commission, which was created by the SRA and charged with writing the mandatory Guidelines, could have implemented the above purposes by way of a pure “charge offense” system, in which the offender is punished strictly on the basis of the offense of conviction, without any attention to the particulars of the offense conduct or its context, although with some attention to the offender’s criminal record. Such a system largely ignores the contexts of particular offenses the facts that might make one weapons offense, for — example, much more severe than another. The Commission rejected such a mechanical approach, but it also rejected a pure “real offense” system. Under a pure real offense system, a defendant receives punishment based on whatever conduct the sentencing court believed 59 actually occurred, regardless of the precise offense of conviction. Instead, the Sentencing Commission created a modified real offense system of sentencing, adopting a middle ground that reflected the 60 language of the SRA. This system instructs the judge to use a defendant’s conviction plus any number of contextual factors, specified by the Guidelines as relevant to the offense of conviction, to establish an 61 offense level from one to forty-three. The judge then cross-references the offense level with the defendant’s criminal history category (on a scale from one to six), which provides a presumptive sentencing range in 62 months. In the modified real offense system designed by the SRA, it was clear that judges could no longer rely on the extraordinarily wide range of information and discretion that had informed sentencings under the rehabilitative model. But there remained the question of the exact scope of information that would be available to the judge and its precise purposes. Before the enactment of the SRA, federal law provided that, No limitation shall be placed on the information concerning the “ 58 See Koon v. United States, 518 U.S. 81, 100 (1996). 59 note 40. supra Barkow, 60 See Stephen Breyer, The Federal Sentencing Guidelines and the Key Compromises upon Which They Rest , 17 H L. R EV . 1, 8-12 (1988). OFSTRA 61 U.S. S G UIDELINES M ANUAL ch. 1, pt. A, introductory cmt. 4(h) (1987). ENTENCING 62 See id. ch. 5, pt. A (sentencing table). The table lists offense levels on the y-axis and criminal histories on the x-axis. The point of intersection yields the presumptive sentencing range. The Commission structured the table to include overlap in the ranges as the offense level moves consecutively higher or lower. Id. ch. 1, pt. A, introductory cmt. 4(h). https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 14 DOI: https://doi.org/10.15779/Z382D08

16 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 OURNAL OF 274 B background, character, and conduct of a person convicted of an offense which a court of the United States may receive and consider for the 63 And this statutory relic purpose of imposing an appropriate sentence.” of the rehabilitation era remains on the books today. At the same time, no longer applied. Even if judges the SRA made clear that Williams -like access to information about the might still have full Williams offender, they could no longer use all of that information just as they pleased. Rather, the SRA required the Commission to implement in guidelines that would compel judges to impose sentences only for certa enumerated purposes, namely “just punishment,” “adequate deterrence,” 64 and incapacitation. To this end, the Commission was empowered — or to exclude whole categories of information in some cases, required — from consideration at sentencing, categories that had seemed relevant for 65 After the SRA and continuing to the present rehabilitation purposes. day, judges no longer have the freedom to rely on any and all information about the offenders they sentence but instead must sentence only on the basis of information rationally related to the statutory purposes enumerated in § 3553(a)(2). In addition to these statutory constraints, the Guidelines themselves, which are administrative rules promulgated by the Commission, further constrain judges. The fundamental provisions of 66 In its the Guidelines include § 1B1.3, the “relevant conduct” guideline. most pertinent part, this guideline orders that offense levels should be determined by examining “all acts and omissions committed, aided, abetted, counseled, commanded, induced, procured, or willfully caused by the defendant . . . that occurred during the commission of the offense of conviction, in preparation for that offense, or in the course of attempting to avoid detection or responsibility for that offense,” as well 67 This provision as all harm caused or intended by the offender. mandates consideration of all conduct underlying the offense of conviction. It does not call for consideration of conduct apart from the offense of conviction, but it does require attention to every detail of the offender’s conduct that relates to the offense of conviction. 63 18 U.S.C. § 3661 (2006). 64 See 18 U.S.C. § 3553(a)(2) (Supp. IV 2010). 65 28 U.S.C. § 994(c)-(e) (2006) (authorizing or requiring the Commission to exclude See or limit the use of numerous categories of evidence including the general exclusion of “the education, vocational skills, employment record, family ties and responsibilities, and community ties of the defendant.”). 66 ENTENCING G UIDELINES M ANUAL § 1B1.3(a) (1987). U.S. S 67 Id. 15 Published by Berkeley Law Scholarship Repository, 2012

17 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 275 P ITHOUT After determining an offense level (and a criminal history category), a judge chooses a sentence within the Guidelines’ narrowed, 68 but still substantial, range. Here, within those narrower bounds, the In determining the Commission allows judges their old discretion: “ sentence to impose within the guideline range, or whether a departure from the guidelines is warranted, the court may consider, without limitation, any information concerning the background, character and 69 Even se prohibited by law.” conduct of the defendant, unless otherwi when the Guidelines were mandatory, the court could depart upward or — — downward from the range as indicated in the provision just quoted but “only when it found ‘an aggravating or mitigating circumstance . . . that was not adequately taken into consideration by the Sentencing 70 Commission.’” This safety valve was intended to operate only in rare cases. In developing the Guidelines, the Sentencing Commission analyzed comprehensive criminal sentencing data and attempted to account for most relevant distinctions in conduct and criminal history in 71 devising offense categories. Still, the Commission necessarily recognized that some factors could never be fully accounted for in advance, and it sought to keep enough judicial discretion within the 72 system to appropriately address “an unusual case.” Thus, through the Commission, the SRA propelled sentencing practice away from a model of unbounded judicial discretion, premised s on rehabilitation, toward a more structured system that impose discipline on judges and a measure of predictability on sentencing. It requires that sentences of incarceration rest not on a rationale of rehabilitation but on a theory that similar offenses committed in a similar fashion by offenders with similar criminal records should be punished — with similar terms of imprisonment. Both the language of the SRA focusing on “ found guilty defendants with similar records who have been 73 — and the language of the Commission in § 1B1.3 — of similar conduct” make clear specifying conduct relevant to “the offense of conviction”— 68 Guidelines sentencing ranges have a minimum and a maximum, which is about 25% higher than the minimum. See id. ch. 5, pt. A (sentencing table). 69 Id. § 1B1.4. 70 Id. ch. 1, pt. A, introductory cmt. 4(b) (citing 18 U.S.C. § 3553(b)). 71 Id. introductory cmt. 3 (“[The Commission has] analyzed data drawn from 10,000 presentence investigations, the differing elements of various crimes as distinguished in substantive criminal statutes, the United States Parole Commission’s guidelines and statistics, and data from other relevant sources in order to determine which distinctions guidelines practice.”). were important in pre- 72 Id. introductory cmt. 3-4(b). 73 28 U.S.C. § 991(b)(1)(B) (Supp. II 2008) (emphasis added). https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 16 DOI: https://doi.org/10.15779/Z382D08

18 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 276 B OURNAL OF that federal sentencing is no longer about the rehabilitative penology of era, which left all law and trial protections behind as it the Williams brought the offender’s entire life before the bar. Rather, the statute and the Guidelines tie sentencing to the offense of conviction (in all its particularity) and the need to pursue just punishment for that offense, as well as adequate deterrence and incapacitation. Constitutional Case Law B. Although recent case law has rendered the formerly mandatory 74 Guidelines merely advisory, the basic principles of the SRA and the Commission’s effort to implement them remain in effect. Moreover, as this section will show, the Supreme Court has embraced the common sense principle that sentencing should bear a close relationship to the particular conviction that authorizes and legitimates punishment in the first place. Though Williams might seem to contradict that principle, the 75 It may well remain good law case has an uncertain status these days. insofar as it holds that a legislature may frame sentencing as a proceeding aimed at rehabilitation and that, if a legislature does so, it may authorize a judge to consider unlimited information about the offender without basic trial protections. But the scope and vitality of Williams ’s authority are in doubt not only because rehabilitation is now foreign to federal sentencing, at least for sentences of imprisonment, but also because more recent double jeopardy cases indicate that the rehabilitative Williams may not have the constitutional status they once premises of Williams did. While held that judges might constitutionally rely on a very broad scope of information obtained by informal means, two modern Supreme Court cases suggest that a sentencing judge faces real constitutional constraints on the ways in which she or he can use that information. The first of these cases is United States v. Watts , which squarely 76 legitimated the use of acquitted and unconvicted conduct at sentencing. for weakening the important role of Many commentators condemn Watts 77 juries and convictions in determining the proper bases of sentencing. . Rather, even as the case But we see no important problems with Watts 74 United States v. Booker, 543 U.S. 220, 226 (2005). 75 Heinous, Atrocious, and Cruel: Apprendi , Indeterminate Sentencing, and the W. David Ball, 109 C Meaning of Punishment, L. R . . 893, 926-29 (2009) ; Hessick & Hessick, supra OLUM EV note 14, at 85. 76 United States v. Watts, 519 U.S. 148, 156-57 (1997). 77 See sources cited supra note 14. 17 Published by Berkeley Law Scholarship Repository, 2012

19 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 277 ITHOUT P legitimated broad fact-finding and consideration of acquitted conduct, it simultaneously embraced a premise that should be understood to discipline fact-finding in fundamental respects. Watts was tried for possession of cocaine base with intent to 78 distribute and for gun possession in connection with a drug offense. The jury convicted him of drug possession but acquitted him on the gun charge. Nevertheless, the sentencing judge found by a preponderance of the evidence that Watts had possessed the gun in connection with the 79 drug offense and raised his sentence under the Guidelines accordingly. The Court of Appeals for the Ninth Circuit vacated the sentence, holding that a judge could not find facts, even by a preponderance of the evidence, that were “necessarily rejected” by the jury in reaching a verdict 80 of not guilty. The Supreme Court reinstated the sentence, emphasizing that, “[n]either the broad language of § 3661 [admitting sentencing facts suggests any basis for the with ‘no limitation’] nor our holding in Williams courts to invent a blanket prohibition against considering certain types of 81 evidence at sentencing.” The Watts holding rested on several grounds, among them that courts had long taken into account acquitted conduct along with many other categories of facts at sentencing; that § 3661 continued to say that “no limitation” should be placed on the information available to the judge at sentencing; and that § 1B1.3 of the Guidelines indicated implicitly in its text and explicitly in its authoritative commentary that tted and unconvicted conduct should come in as “relevant conduct” acqui 82 in appropriate cases. Still, none of these justifications would do the job if, as the Ninth Circuit had suggested, sentencing on the basis of acquitted conduct Clause. The Ninth Circuit’s position made violated the Double Jeopardy a good deal of common sense, since Watts had been tried and acquitted once on the gun charge and then at sentencing found himself in danger of punishment for that offense all over again. The Supreme Court, however, escaped that argument by holding that an offender is not “punished” for a particular sentencing fact that enhances the sentence. Rather, the offender is still punished only for the offense of conviction; the sentencing fact merely contextualizes that offense. Thus, the Court’s reasoning suggested one fundamental limit on the facts that a sentencing 78 Watts , 519 U.S. at 149-50. 79 Id. 80 Id. at 150. 81 Id. at 152. 82 Id. at 151-53. https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 18 DOI: https://doi.org/10.15779/Z382D08

20 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 278 OURNAL OF B court might consider. Those facts must serve only to contextualize the As we offense of which the defendant has actually been convicted: “ explained in Witte , . . . sentencing enhancements do not punish a defendant for crimes of which he was not convicted, but rather increase his sentence because of the manner in which he committed the crime of 83 conviction.” , the case on which Watts primarily relied, In Witte v. United States Witte was initially convicted of marijuana possession with intent to 84 distribute. A year prior to the marijuana incident, however, he had also 85 attempted to import cocaine. The sentencing judge determined that the cocaine attempt constituted part of the same criminal conspiracy and 86 accordingly enhanced Witte’s sentence on the marijuana charge. When the cocaine incident subsequently formed the basis of a separate indictment on charges of attempting to import cocaine, Witte argued that reusing conduct already considered at an earlier sentencing violated his 87 rights under the Double Jeopardy Clause. The Supreme Court, however, disagreed. According to the Court, Witte’s sentence at the first trial punished him only for his conviction for marijuana possession. The attempt to import cocaine contextualized the marijuana offense and enhanced its severity, justifying an increased sentence, but the increase 88 did not constitute separate punishment for a separate offense. To the extent that the Guidelines aggravate punishment for related conduct outside the elements of the crime on the theory that such conduct bears on the “character of the offense,” the offender is still punished present offense was carried out in only for the fact that the a manner that warrants increased punishment, not for a offense (which that related conduct may or may different 89 not constitute). This distinction between punishment for a sentencing fact and as contextualized by a sentencing punishment for the offense of conviction fact was fundamental. 83 Id. at 154. 84 Witte v. United States, 515 U.S. 389, 392-93 (1995). 85 at 391-92. Id. 86 at 393-95. Id. 87 Id. at 394-95. 88 at 402-03. Id. 89 Id. 19 Published by Berkeley Law Scholarship Repository, 2012

21 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 279 P ITHOUT Because consideration of relevant conduct in determining a defendant’s sentence within the legislatively authorized punishment range does not constitute punishment for that conduct, the instant prosecution does not violate the Double Jeopardy Clause’s prohibition against the imposition of multiple punishments for the same 90 offense. The clear implication of this holding is that if “relevant conduct” was used to punish for the sentencing fact itself and not just to contextualize the offense of conviction, then such use would violate the 91 Double Jeopardy Clause. Thus, the Court held both that the Guidelines do indeed use relevant conduct only for contextualization, as the language of the SRA and § 1B1.3 both suggest, and that it would be that is, to convert modified unconstitutional for them to do otherwise — real offense sentencing into pure real offense sentencing, rendering the offense of conviction a mere pretext for punishing the offender for any and every bad act the judge might identify at sentencing. Watts and Witte were, in part, Double We should note that, since Jeopardy cases, their constitutional holdings arguably apply only in cases 90 at 406. “[O]ur precedents . . . make clear that a defendant in that situation is Id. punished, for double jeopardy purposes, only for the offense of which the defendant is convicted.” at 397. Id. 91 Core Eighth Amendment cases are to the same effect, notwithstanding a recent -strikes laws in Ewing v. California , 538 argument that the Court’s legitimation of three U.S. 11 (2003) constituted an admission that courts could punish not only for the offense of conviction but also for prior offenses. See Carissa Byrne Hessick & F. , 97 C Andrew Hessick, Double Jeopardy as a Limit on Punishment R EV L. 45, 73- 75 ORNELL . was exactly that the offense of conviction might (2011). In fact, the reasoning of Ewing not evaded by be appropriately contextualized by the fact of the offender’s prior — — record. Where that prior record is severe, the meaning of the current offense is not just that the defendant committed, for example, a larceny, but that that latest larceny, when contextualized by the offender’s previous convictions, sta nds as evidence of the defendant’s incorrigibility. Such incorrigibility is rational justification for imposing an Ewing especially harsh sentence for this latest offense. It is true that there is language in that, when taken out of context, sounds like an endorsement of punishment beyond the offense of conviction See Ewing , 538 U.S. at 29 (“In weighing the gravity of Ewing's . offense, we must place on the scales not only his current felony, but also his long t makes clear that it does not mean by this history of felony recidivism.”). But the Cour that Ewing may be punished again for his prior offenses as such. Rather, it cites Witte for the proposition that “In repeatedly upholding such recidivism statutes, we have for the rejected double jeopardy challenges because the enhanced punishment imposed for the latest crime , which is considered to be an . . . [is] ‘a stiffened penalty later offense Id. at 25-26 (emphasis added). aggravated offense because a repetitive one.’” https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 20 DOI: https://doi.org/10.15779/Z382D08

22 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 OURNAL OF 280 B where the challenged sentencing fact has already been the subject of an ) or a conviction ( Witte ). But the Watts adjudication, whether an acquittal ( premise of the Court’s affirmances in those cases is that the sentencing regime created by the SRA and the Guidelines provides only for punishment of the offense of conviction, not punishment of unconvicted offenses. The constitutional argument for limiting sentences to offenses of conviction also gains strength from consideration of the jury right under the Sixth Amendment. It should seem obvious that, under the Constitution, the State cannot punish someone for an unconvicted offense just because that person has been convicted of another offense. After all, what is the purpose of the Sixth Amendment jury right, if not to protect persons from punishment until they have been duly convicted of the conduct for which they are being punished? This obvious argument seems to us the correct argument. But considering the Court’s recent abundance of case law on the jury right, it does seem necessary to elaborate. In recent years, the Supreme Court has sought to give meaningful , the Court struck down content to the jury right. In Apprendi v. New Jersey a law that would have bypassed the jury by increasing the statutory maximum sentence for the defendant’s crime based on a mere judicial 92 finding that racial bias motivated the commission of the offense. In making its decision, the Court emphasized its concern for protecting the constitutional power of the jury to find those facts essential to punishment. In the earliest days of the criminal justice system, the jury verdict often definitively established a defendant’s punishment, 93 Similarly, American eliminating all judicial discretion in the sentencing. juries have always had the authority to issue unreviewable acquittals, 94 “stand[ing] between the individual and the power of the government,” thus excluding judicial discretion from the process. And while judges gradually gained general authority to contextualize the offense by finding additional facts at sentencing, the jury’s verdict has a lways set the boundaries for the exercise of that authority. The Court’s holding in reaffirmed that judges can find facts that contextualize an Apprendi offense and enhance punishment but that the jury must first convict the 92 Apprendi v. New Jersey, 530 U.S. 466, 491 (2000). 93 See, e.g. , Barkow, supra note 40, at 78-79; Chris Kemmitt, Function over Form: Reviving the Sentencing Body , 40 U. Criminal Jury’s Historical Role as a M ICH . J.L. R EFORM 93, 98- 99 (2006). 94 United States v. Booker, 543 U.S. 220, 237 (2005). 21 Published by Berkeley Law Scholarship Repository, 2012

23 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 281 P ITHOUT 95 defendant. That conviction must set a hard statutory maximum to 96 control the judge. Mandatory sentencing guidelines presented a different but related challenge to jury power, since judges would routinely find facts (“relevant conduct”) at sentencing that resulted in higher Guideline s sentences than the jury conviction alone would have authorized. In the cases after , the Court ruled that it was unconstitutional for judges to find Apprendi facts that raised a defendant’s sentence above the mandatory range 97 and United States Blakely v. Washington established by the Guidelines. In 98 , v. Booker the Court held that a defendant has a Sixth Amendment right to have a jury do that sort of fact- finding: “Any fact (other than a prior conviction) which is necessary to support a sentence exceeding the maximum authorized by the facts established by a plea of guilty or a jury verdict must be admitted by the defendant or proved to a jury beyond a 99 reasonable doubt.” This holding preserved the jury right not by limiting judges’ traditional fact -finding discretion, but rather by ensuring that judges exercise their discretion only within the bounds set by the jury’s conviction. Since these bounds might still be quite broad in terms of years of imprisonment, one might doubt that the Court has really preserved much in the way of jury power. After all, legislatures remain free to create sentencing regimes with broader or narrower sentencing ranges, as they case, the prefer. And, in an odd twist in the famously strange Booker Supreme Court itself restored very wide sentencing ranges to the federal judiciary simply by reading the mandatory quality of the Guidelines out 100 Although judges must still consult the Guidelines and of the statute. make an initial determination of the appropriate Guidelines sentence, the judge has full authority to sentence the defendant anywhere within the 101 The statutory range, subject to reasonableness review upon appeal. Court thus reintroduced a breadth of judicial discretion approaching that of the pre-Guidelines era, eroding the value of the vaunted jury right. A substantial commitment to the jury power remains in the Blakely Court’s case law, however. In , Justice Scalia acknowledged that 95 Apprendi , 530 U.S. at 490 (“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.”). 96 Id. 97 Blakely v. Washington, 542 U.S. 296 (2004). 98 Booker , 543 U.S. 220. 99 Id. at 244. 100 Id. at 259. 101 Id. at 259-61. https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 22 DOI: https://doi.org/10.15779/Z382D08

24 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY C RIMINAL L AW [Vol. 17:2 OURNAL OF 282 B J the Court’s injection of a strengthened jury right into determinate sentencing schemes might lead legislatures to revert to indeterminate 102 But he sentencing schemes that would increase judicial discretion. explained why even such a response would still leave Sixth Amendment principles intact: [T]he Sixth Amendment by its terms is not a limitation on judicial power, but a reservation of jury power. It limits judicial power only to the extent that the claimed judicial power infringes on the province of the jury. Indeterminate sentencing does not do so. It increases judicial discretion, to be sure, but not at the expense of the jury’s traditional function of finding the facts essential to lawful imposition of the penalty. Of course indeterminate schemes involve judicial factfinding, in that a judge (like a parole board) may implicitly rule on those facts he deems important to the exercise of his sentencing discretion. But the facts do not pertain to whether the right to a lesser sentence — and that defendant has a legal makes all the difference insofar as judicial impingement 103 upon the traditional role of the jury is concerned. The advent of advisory Sentencing Guidelines necessarily increased judicial discretion at sentencing. But it did not affect the principle that every jury verdict must retain some substantial meaning, including at least that it limits the sentence to the legal maximum for the offense of conviction . In fact, Justice Scalia’s majority opinion in Blakely suggested that a substantial jury right must include at least some version of the principle we argue for in this article. Addressing critics of the decision, he argued that their logic would hold “ that a judge Apprendi could sentence a man for committing murder even if the jury convicted — or of him only of illegally possessing the firearm used to commit it 104 But no ” making an illegal lane change while fleeing the death scene. one, he claimed, perhaps with some naïveté, would actually advocate such an “absurd result”: “ The jury could not function as circuitbreaker in the State’s machinery of justice if it were relegated to making a determination that the defendant at some point did something wrong, a 102 Blakely , 542 U.S. at 308-09. 103 Id. 104 Id. at 306. 23 Published by Berkeley Law Scholarship Repository, 2012

25 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT C ONVICTION [Vol. 17:2 ITHOUT 283 P W mere preliminary to a judicial inquisition into the facts of the crime the 105 seeks to punish.” State actually and that sentences in the Witte The Court has affirmed in Watts modern, federal system punish only for the offense of conviction, and the Court has in some respects reinvigorated the jury right in the Apprendi line of cases, but the Court has not yet confronted the precise question addressed here and the endemic in the federal system: how do you know when judicial fact-finding at sentencing produces legitimate punishment for the offense of conviction, as opposed to producing unauthorized by any jury verdict or guilty plea additional punishment — — for some additional offense? III. I MPLEMENTATION The simple principle that a judge can punish only for the offense of conviction turns out to be quite hard to implement in many cases. It is tempting, for example, to apply some simple arithmetic rule, perhaps that facts found at sentencing cannot do more than double the sentence that would otherwise have applied. But no one has come up with a logical basis for any particular arithmetic rule, and criminal statutes have long provided for very wide sentencing ranges for a reason: violators of any particular criminal statute can vary enormously in their culpability, their motives, their prior records, their demonstrated dangerousness going forward, and the degree of harm done to name a few major — considerations. It is, therefore, very difficult to come up with a reliable rule to cover those cases where it just seems like the enhancement is “too big.” Nevertheless, there are cases where it is easy to implement the principle that courts must punish only for the offense of conviction. If these simple applications are embraced, perhaps then the federal courts will begin to apply the principle explicitly and make progress towards using it in more difficult cases as well. The two cases with which we began this article, United States v. 106 107 United States v. Kelly in the Fourth Circuit and in the Tenth Barber Circuit, present the easy case. They asked whether a judicial finding of premeditation could justify a departure from the presumptive second- degree murder range of the Guidelines. Both courts agreed that premeditation fell outside the core or “heartland” of second -degree 105 Id. at 306-07. 106 United States v. Barber, 119 F.3d 276 (4th Cir. 1997). 107 United States v. Kelly, 1 F.3d 1137 (10th Cir. 1993). https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 24 DOI: https://doi.org/10.15779/Z382D08

26 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 284 B OURNAL OF 108 This conclusion seems inescapable, considering that the murder. federal murder statute specifically includes premeditation as an element of first-degree murder and specifically defines second-degree murder as 109 The two courts diverged, all murder that is not first-degree murder. however, on whether the Commission had adequately considered premeditation in setting the punishment range for second-degree murder. The Tenth Circuit concluded that the Commission had obviously considered premeditation, even if the Commission did not explicitly 110 At the time of the appeal, first-degree murder carried a base address it. offense level of forty-three, compared to just thirty-three for second- 111 degree murder. Because only the presence or absence of premeditation distinguished first- and second-degree murder (in the context of that case), the Tenth Circuit reasoned that the Commission must have used premeditation to justify the ten-level disparity in punishment between 112 Since the Commission had first- and second-degree murder. adequately considered premeditation in setting the sentencing range, the trial judge could not increase Kelly’s sentence beyond that range u pon a finding of premeditation at sentencing. In contrast, Fourth Circuit Judge Wilkins, who happened to be the former chair of the Sentencing Commission itself, disagreed strongly 113 Wilkins’s opinion lacked formal legal with the Tenth Circuit’s analysis. authority, since it was written for half of an evenly divided en banc court, 108 Barber , 119 F.3d at 287 (Wilkins, J.); Kelly , 1 F.3d at 1140. The Sentencing Guidelines define this “core” as the “heartland” of the offense. According to the Guidelines, The Commission intends the sentencing courts to treat each guideline as carving out a “heartland,” a set of typical cases embodying the conduct that each guideline describes. When a court finds an atypical case, one to which a particular guideline linguistically applies but where conduct significantly differs from the norm, the . court may consider whether a departure is warranted U.S. S ENTENCING ANUAL ch. 1, pt. A, introductory cmt. 4(b) (1987). G UIDELINES M , the “heartland” concept served Under the mandatory Guidelines, and even after Booker and continues to serve to anchor every sentencing proceeding in the actual offense of conviction by compelling judges to use as the starting point in their analysis how the facts of the case at issue compare to the facts of the typical conviction of that offense. See Rita v. United States, 551 U.S. 338, 351 (2007); United States v. Christman, 607 F.3d 1110, 1118-19 (6th Cir. 2010); United States v. King, 604 F.3d 125, 142-43 (3d Cir. 2010); United States v. Martinez-Barragan, 545 F.3d 894, 900-902 (10th Cir. 2008); United States v. Moreland, 437 F.3d 424, 433 (4th Cir. 2006). 109 See supra notes 19-20 and accompanying text. 110 , 1 F.3d at 1141-42. Kelly 111 S ENTENCING G UIDELINES M ANUAL §§ 2A1.1 – 2A1.2 (1993). U.S. 112 , 1 F.3d at 1141-42. Kelly 113 Barber , 119 F.3d at 287. 25 Published by Berkeley Law Scholarship Repository, 2012

27 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 285 P ITHOUT but it fairly represented the implicit position of most courts outside the Tenth Circuit. According to Wilkins, the Commission’s failure to discuss premeditation in the context of second-degree murder meant simply and plainly that the Commission had not considered it when setting the base 114 offense levels. Since premeditation fell outside the core of second- degree murder, the trial judge could use this factor to increase the 115 The ndant Hodge’s prison term beyond the presumptive sentence. defe Commission’s implicit “consideration” of premeditation, as understood by the Tenth Circuit, was irrelevant because the fact remained that the Commission had not explicitly discussed premeditation in the context of 116 second-degree murder. It is worth noting that even in the Fourth Circuit, half the judges disagreed with Wilkins and agreed with Kelly , thereby draining all legal authority from Wilkins’s opinion but leaving intact the trial court’s Wilkins-esque decision. The three-judge panel that initially heard the Kelly and held that the trial judge could not Barber appeal relied heavily on use premeditation to justify an upward departure: Given the statutory definition of second-degree murder as murder without premeditation, the Commission appears to have come to the sensible, perhaps inescapable, conclusion that premeditation should not be used as an aggravator of a crime that by definition lacks premeditation. To do otherwise would be to us e sentencing mechanisms to punish defendants for crimes 117 of which they have not been convicted. The Fourth Circuit vacated this opinion upon agreeing to hear the appeal 118 But the en en banc, thus reinstating the judgment of the district court. banc court ultimately divided evenly, half of its judges giving credence to the Tenth Circuit’s position. When the Tenth Circuit had an opportunity to reconsider its Kelly Barber , it reaffirmed its position that holding in light of the decision in 114 Id. 115 Id. at 287-88. 116 at 288 (“[I]t is obvious that the Commission ‘considered’ premeditation in Id. establishing the homicide guidelines as a whole. It is equally clear that this fact is irrelevant in determining whether a departure from the second-degree murder guideline is appropriate.”). 117 rev’d en banc , 119 F.3d United States v. Barber, 93 F.3d 1200, 1206-07 (4th Cir. 1996), 276 (4th Cir. 1997). 118 Barber , 119 F.3d at 279 n.1 (majority opinion). https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 26 DOI: https://doi.org/10.15779/Z382D08

28 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY B C RIMINAL L AW [Vol. 17:2 OURNAL OF J 286 premeditation could not justify departure from the presumptive 119 In fact, it stated the Guidelines range for second-degree murder. operative principle even more boldly: “[T]he sentencing court may not depart upward from the Guideline range for second-degree murder on 120 The degree murder.” grounds that recharacterize the offense as a first- defendant in that case, Michael Lee Hanson, pled guilty to second-degree murder. The evidence at sentencing, however, suggested that Hanson had both premeditated the murder and committed it in furtherance of a 121 robbery, each a statutory element of first-degree murder. The sentencing judge, though frustrated, denied the government’s motion for Kelly departure on either ground, citing the Tenth Circuit’s opinion in : [W]hen somebody offers somebody a plea for second degree, and that plea is accepted, then my ability to do what I think is justice in this case is limited by the guidelines. . . . [E]ven though I find myself in enormous agreement with the analysis of the heinousness and to simply treachery of these acts, I’m no longer free ignore [the second-degree murder guideline] that existed at the 122 time the plea was entered and to depart as I see fit. The Tenth Circuit panel that heard the appeal affirmed the judge’s denial on both grounds, not only reaffirming that judges could not use premeditation, but also holding that judges could not increase sentences upon a judicial finding of felony murder because that too redefined 123 As to premeditation, second-degree murder as first-degree murder. “To allow upward departure on the grounds that a second -degree murder was premeditated would permit the sentencing court to treat the offense of conviction (here, a murder that was not premeditated) as merely 124 establishing a floor offense level.” The panel emphasized the importance of maintaining the integrity of the offense of conviction and -finding worried about increased sentencing disparity should judges’ fact 125 turn identical convictions into different substantive crimes. 119 United States v. Hanson, 264 F.3d 988, 990 (2001). 120 Id. 121 Id. at 991. 122 Id. 5094) (Sentencing Hr’g, at 992 (quoting Brief for Appellant, App. at 183 (No. 00- May 3, 2000)) (emphasis in original). 123 Id. at 997. 124 Id. at 996. 125 Id. 27 Published by Berkeley Law Scholarship Repository, 2012

29 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT C ONVICTION [Vol. 17:2 ITHOUT 287 P W We find this analysis consistent with and compelled by the federal courts’ responsibility to defend the integrity of the Sixth Amendment jury right and the Fifth Amendment protection against double jeopardy, the SRA’s insistence on uniformity and proportionality, and the Guidelines’ consequent rejection of “real offense” sentencing. When Congress passed the federal murder statute, it enacted a graded offense structure, one that clearly distinguished between first- and second-degree murder and created two distinct offenses. In similar - fashion, the Sentencing Guidelines carve out a “heartland” for second 126 but this heartland does not exist in isolation, as the degree murder, Fourth Circuit proposed. It is bounded by the Guidelines’ sentencing rules for related offenses. The Sentencing Commission must, of course, write the Guidelines with due respect for Congress’s statutory grading of offenses, and judges must then read the Guidelines in light of that statutorily mandated grading structure. As clearly as premeditation falls outside the typical second-degree murder, it just as clearly falls within the typical first-degree murder. Finding premeditation in a second-degree murder, then, does more than contextualize the nature of the offense: it redefines the conduct as a typical first-degree murder, punishing the defendant for a crime other than the offense of conviction. Such a determination violates the Supreme Court’s longstanding directive that 127 courts punish only for the offense of conviction. Thus, at a minimum, federal courts should hold that a sentence for one offense cannot be enhanced by factual findings that redefine the offense, that render the enhancement a punishment for some offense different from the offense of conviction. Acknowledging that, in many cases, it will be hard to draw the line that best reflects this principle, the courts should at least begin by declaring a legal rule that a sentence for second-degree murder in the federal system cannot be enhanced by a finding of premeditation or a finding that the murder happe ned “in the 128 A perpetration of” one of the murder statute’s enumerated felonies. 126 See , supra note 108. 127 See supra Part III.B. , 128 18 U.S.C. 1111(a) (2006): Murder is the unlawful killing of a human being with malice aforethought. Every murder perpetrated by poison, lying in wait, or any other kind of willful, deliberate, malicious, and premeditated killing; or committed in the perpetration of, or attempt to perpetrate, any arson, escape, murder, kidnapping, treason, espionage, sabotage, aggravated sexual abuse or sexual abuse, child abuse, burglary, or robbery; or perpetrated as part of a pattern or practice of assault or torture against a child or children; or perpetrated from a https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 28 DOI: https://doi.org/10.15779/Z382D08

30 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY C RIMINAL L AW [Vol. 17:2 OURNAL OF 288 B J sentencing judge should not be able to recharacterize second-degree murder as first-degree murder to justify an enhanced sentence beyond 129 Guidelines range. To allow otherwise the presumptive, albeit advisory, reintroduces the element of lawlessness that was to be cured by the modern requirement that judges exercise their discretion only on the basis of explicit factual findings that are rationally connected to the offense of conviction. IV. B ECOND -D EGREE EYOND URDER : A PPLYING THE S M RINCIPLE TO O THER O FFENSES P Although this article focuses most concretely on premeditation in second-degree murder cases, the limiting principle espoused could and consistent with the Supreme Court’s should receive broader application sentencing jurisprudence. As the circumstances and facts of an individual case become more complex, the principle necessarily loses its bright-line quality. Nevertheless, the key tenet still stands. The conviction must provide a meaningful link between offense and punishment, and judges should not have authority to redefine the offense. First, we want to consider a case that is different from the premeditation cases but nearly as easy doctrinally, even as its terrifying facts tempt one to depart from principle. In this 2007 case, United States v. , the Tenth Circuit again cemented its lonely leadership among the Allen federal circuits by insisting on the principle that we have outlined in this article, both as a matter of proper interpretation of the Guidelines and as 130 In this case, the authorities had a Sixth Amendment matter. information that Leroy Eric Allen was seriously contemplating the 131 horrific crimes of raping, murdering, and perhaps torturing young girls. They had gathered enough evidence to suggest that Allen might be premeditated design unlawfully and maliciously to effect the death of any human being other than him who is killed, is murder in the first degree. Any other murder is murder in the second degree. 129 Although the Guidelines are now advisory, the judge is still required to justify any variance from the recommended Guidelines range. Rita v. United States, 551 U.S. 338, 351 (2007). The appellate courts, therefore, retain the power to invalidate sentences that rest on factual findings and reasoning that redefines the offense of conviction as some other offense. 130 See United States v. Allen, 488 F.3d 1244 (10th Cir. 2007). 131 Id. at 1246-47. 29 Published by Berkeley Law Scholarship Repository, 2012

31 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 289 ITHOUT P prepared to carry through on his declared intentions but not enough to actually prosecute him for an attempt. In the end, they decided to take advantage of his concurrent involvement with drugs to set him up on a 132 charge of methamphetamine distribution. This proved the successful avenue, and the Government ultimately convicted Allen on the drug 133 The prosecution predictably sought to prove at sentencing that charge. Allen represented a serious danger to young girls in the future. The trial judge, suitably horrified by the evidence to this effect, embraced the Government’s position and enhanced Allen’s otherwise likely sentence of about twelve years to a full thirty years on the basis of his 134 dangerousness. The Tenth Circuit, however, adhered stoutly to principles of law in denying the U.S. Attorney’s office its desired shortcut. It insisted on what it called the “relatedness principle”— that is, that sentencing facts must be adequately related to the offense of conviction — which it located firmly in the Guidelines and the SRA, citing even Chairman/Judge Wilkins in support: The relatedness principle is fundamental because of our commitment to sentencing based on the seriousness of the actual offense proven or admitted. See 18 of the U.S.C. § 3553(a)(1) (‘‘the nature and circumstances ’’) (emphasis added);; offense § 3553(a)(2)(A) (‘‘the need id. of for the sentence imposed . . . to reflect the seriousness the offense . . .’’) (emphasis added);; W illiam W. Wilkins, Jr. & John R. Steer, Relevant Conduct: The Cornerstone of the – 99 Federal Sentencing Guidelines , 41 S.C. L. Rev. 495, 497 135 (1990). Moreover, Judge McConnell’s compelling opinion, recognizing that, after Booker , he could not rely solely on the Guidelines to control sentencing in the trial courts, located the relatedness principle in the Sixth Amendment as well: This is not unrelated to the Sixth Amendment principles Booker underlying . . . . When a sentencing court considers conduct related to the offense of conviction, the 132 Id. at 1248. 133 Id. at 1248-49. 134 at 1249-52. Id. 135 Id. at 1255. https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 30 DOI: https://doi.org/10.15779/Z382D08

32 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 290 OURNAL OF B objective is to determine the seriousness of the very crime found by the jury or admitted by the defendant. If the considered conduct has nothing to do with the offense of conviction, the court is effectively sentencing a defendant for a crime that was never proved to the jury, or admitted by the defendant. To allow this would empower the government to obtain punishment for any number of unrelated crimes, based on bench trial rather than jury trial. The relatedness principle thus keeps the system from straying too far beyond the Sixth 136 Amendment line. Just as in the second-degree murder cases, therefore, but now reaching out to other contexts, the Tenth Circuit held Allen’s sentence unreasonable because of the irrelevance to the drug conviction of the sentencing facts regarding Allen’s propensity to sexual violence. Aside from these Tenth Circuit cases and Barber in the Fourth Circuit, few cases have openly confronted such stark examples of 137 even punishment for facts other than the offense of conviction, though it seems clear that such cases exist. More typical is a troubling 138 but ambiguous case like United States v. Paster from the Third Circuit. Mitchell Frederick Paster was convicted of second-degree murder. The s entencing judge moved Paster’s sentence from a presumptive Guidelines sentence of about ten or eleven years to more than thirty 139 —in this case, Paster’s infliction years on the basis of “extreme conduct 140 In doing so, the judge of at least sixteen stab wounds on his wife. sentenced Paster in a range that might have applied if he had been convicted of first-degree murder, rather than pleading to second-degree 136 . We think the only reasonable reading of Allen is that a departure from the Id “relatedness principle” is a violation of the Sixth Amendment, even though the opinion See id. at 1262 (“ In a case involving a is less than explicit in summarizing the holding. variance of this magnitude, we hold that, whatever latitude a sentencing court may have Booker discretion, it may not discard to adjust a defendant’s sentence in an exercise of the advisory Guideline range and impose sentence, instead, on the basis of evidence of conduct, whether actually committed or the defendant’s uncharged, unrelated mis . contemplated for the future.”) 137 See, e.g. , United States v. Newsom, 508 F.3d 731, 734 (5th Cir. 2007); United States v. Ortiz, 431 F.3d 1035, 1040-41 (7th Cir. 2005); United States v. Leonard, 289 F.3d 984, 987-89 (7th Cir. 2002); United States v. Cross, 121 F.3d 234 (6th Cir. 1997); United States v. Kim, 896 F.2d 678, 682-84 (2d Cir. 1990). 138 United States v. Paster, 173 F.3d 206 (3d Cir. 1999). 139 Id. at 216. 140 Id. at 209-10. 31 Published by Berkeley Law Scholarship Repository, 2012

33 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 291 P ITHOUT murder. Although the sentencing judge had not flouted any clear rule, Kelly sentencing judge had, the way the Tenth Circuit determined that the the Third Circuit nevertheless saw a similar dynamic at play: The vice of the nine-level upward departure imposed on Paster is that he has incurred for second degree murder a sentence that would be appropriate for first degree murder . . . . Thus, if the government had required Paster to plead guilty to first degree murder in order to escape the death penalty, . . . he would have faced a sentence in the range of 324-405 months, the median of which is the actual sentence that Paster received. This lack of disparity between Paster's actual sentence and one he could have received had he pleaded guilty to, or been convicted of, a more serious crime distorts proportionality, a critical U.S.S.G. Ch. objective of the Sentencing Guidelines. See 141 1, Pt. A, 3. In our view, the reversal in is a reasonable application of Paster the principle we are defending: that judges must sentence only for the offense of conviction and have no authority to redefine that offense for 142 sentencing purposes. But it is nevertheless a problematic application. a particularly heinous second-degree That is, it is not at all clear that murder should never be punished at a level that might overlap with some first-degree murders. Legislatures grade offenses for important reasons, but they do their grading imperfectly. Available punishments for first- 141 Id. at 220-21 (citations omitted). 142 Paster Note that, much like the court, which used the language of “proportionality,” found that the principle of avoiding Federal Sentencing Law & Practice the editor of “unwarranted disparity” justified the Tenth Circuit’s position in : Kelly The Tenth Circuit has the stronger position. If a court can depart upward based on a factor that differentiates the chapter two offense guidelines (such as premeditation in a second-degree murder case), -offense/real-offense compromise then the Commission’s conviction is negated. Under the rationale of the separate opinion in the Fourth Circuit case, the guidelines would have a floor based on the conviction offense, but each sentencing court could depart upward based upon the ‘real offense’ whenever the co urt wished. This hardly seems consistent with the goal — established by Congress and — of reducing unwarranted subscribed to by the Commission sentencing disparity. L H UTCHISON ET AL ., F EDERAL S W. HOMAS AW AND P RACTICE § 2A1.2 ENTENCING T cmt. 2(f) (2012 ed.). https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 32 DOI: https://doi.org/10.15779/Z382D08

34 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY J C RIMINAL L AW [Vol. 17:2 OURNAL OF 292 B and second-degree murders, like available punishments for so many offenses, overlap with each other quite a bit. And there is no reason to think that legislatures have fully defined the criteria of criminal grading when they have established the elements of all offenses, not as long as the legislatures and courts remain committed to wide sentencing ranges and substantial fact-finding at sentencing. The fact is that strict, formal in some tension, at times — with relatively grading of offenses coexists — loose sentencing practices amid overlapping statutory ranges. It is -degree murder was punishable at a plausible to think that Paster’s second level that overlapped with the available range for first-degree murder in this case because it was committed in a particularly “extreme” way. So the holding in is, as we say, problematic. It is neither logically Paster compelled nor logically prevented by any available principle. The Paster holding is nevertheless a plausible, pragmatic effort to employ appellate judicial power to incrementally enhance the rationality of grading and punishment in the modern, legalized era of sentencing. Although Congress has fastened on premeditation and an accompanying felony as the only elements that make for first-degree murder at the conviction stage, a court can make efforts to reinforce the grading structure in other respects at sentencing. That is, it might, as the Third Circuit effectively did in Paster , establish a strong presumption that no sentencing aggravator should enhance a second-degree sentence into the same range that presumptively applies to first-degree murder. Such a rule would contain real elements of arbitrariness, but it would represent a readily intelligible effort to give appellate review and the sentencing system as a whole a more predictable and legal character. Another difficult category of cases arises from what might be internal contradictions within the Guidelines. Although we think the Guidelines declare a basic commitment to punishing only for the offense of conviction, particular guidelines arguably violate this principle. A 143 A jury convicted 2009 case from the Fourth Circuit offers an example. Gary Williams on two counts of cocaine distribution and one count of 144 distributing fifty grams or more of cocaine base. At sentencing, however, the judge found Williams responsible for the first-degree 145 The judge then followed murder of an intended prosecution witness. the command of the guideline applicable to those convicted of drug 146 offenses, which required that the judge cross-reference the first-degree 143 United States v. Williams, 343 F. App’x 912 (4th Cir. 2009). 144 at 913. Id. 145 Id. 146 S ENTENCING G UIDELINES M ANUAL § 2D1.1(d)(1) (2011). U.S. 33 Published by Berkeley Law Scholarship Repository, 2012

35 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT W C ONVICTION [Vol. 17:2 293 ITHOUT P murder guideline and sentence the defendant according to that latter guideline, not the one normally applicable to the distribution 147 The Fourth Circuit affirmed Williams’s life sentence on convictions. appeal, rejecting the defendant’s claim that, given the size of the 148 sentencing enhancement, the trial judge should have found the 149 sentencing facts by a heightened standard of proof. The principle espoused in this article would have provided the defendant with a strong argument that the judge imposed an unreasonable and thus illegal sentence because he redefined the drug offense as some other offense, first-degree murder, for which there was no conviction. Indeed, one could argue that a sentence is illegitimate any time the Guidelines call for a judge to cross-reference a guideline for a judge did, separate, unconvicted offense in the way the Williams calculating the sentence ultimately without reference to the offense of conviction at all but entirely on the basis of an offense found for the first time at sentencing. This argument would not imply that the judge should ignore a finding that the defendant caused the witness’s dea th. Clearly, a drug crime involving a death deserves more serious punishment than a less harmful drug crime; it nonetheless deserves to be punished as a drug Williams judge did not simply enhance crime, not a murder. The punishment, but rather sentenced the defendant as if he had been 150 indicted, tried, and convicted of first-degree murder. Even more importantly, the judge’s written analysis treated the drug offense as irrelevant to the sentence once the murder was brought to light. Nothing indicated that the punishment reflected appropriate contextualization of the drug offense; everything in the analysis suggested that the defendant was punished for murder, despite the 151 The Tenth Circuit has absence of a conviction for that offense. succinctly described the vice in such a sentencing: 147 United States v. Williams, No. WDQ-07-0402, 2008 WL 2091152, at * 4-5 (D. Md. May 16, 2008). 148 The case report does not specify what the sentence would have been without the finding of murder, but the judge’s sentencing memorandum does note that, “The presentence report calculates for Williams a base offense level of 30 and a criminal history category of VI. The corresponding advisory Guidelines range is 168-210 Id. at *4 n.4. months.” 149 Williams , 343 F. App’x at 913. 150 Note that in rare cases a federal murder defendant might possibly be sentenced to See generally 18 U.S.C. §§ 3591-99 (2006 & Supp. II 2008). death. 151 The judge’s sentencing memorandum relies on the drug convictions only to justify the use of the relaxed procedures of a sentencing hearing in finding Williams responsible for a first- degree murder. That is, he brushed off Williams’s o bjection to being, in effect, tried for murder without any trial rights by noting that Williams had https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 34 DOI: https://doi.org/10.15779/Z382D08

36 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D ERKELEY C RIMINAL L AW [Vol. 17:2 OURNAL OF 294 B J If the government, federal or state, believes [the defendant] committed a crime in his dealings [which did not form the basis of the current conviction] . . . , it is free to bring a prosecution for that conduct. In such a proceeding, [the defendant] would be entitled to put the Booker government to its proof. Despite the wide latitude granted district courts, we do not believe it sanctions an 152 end-run around this fundamental process. Use of the first-degree murder guideline should have alerted the trial court that it was on dangerous ground, that it risked crossing the line from enhancing punishment to redefining the offense of conviction. Instead, the court uncritically applied the murder guideline, allowing the o ffense of conviction to provide a mere “floor” from which the court 153 rapidly built the sentence upward. To sentence as the Williams court did would seem to redefine the offense of conviction and violate the Sixth Amendment jury right. One last example will introduce a much harder class of cases, one that we do not attempt to handle in this article. In run-of-the-mill drug cases, sentence length is determined largely by the type and quantity of the drug involved. These cases often involve several counts of drug possession, distribution, and/or conspiracy. It is not uncommon for juries to acquit on one or more counts or for a plea bargain to knock out one or more charges, only to have the sentencing judge find that the offender was indeed responsible for all of the drugs underlying all of the charges. In such a case, the Guidelines will often call on the judge to sentence the offender exactly as if she or he had been convicted on all charges, rather than one or two, and the judge will often comply. Such a proceeding looks like the court’s seizing on one conviction as a mere technical device that allows it to punish primarily for other, unconvicted offenses. But, again, there may be arguments that justify what first appear to be illogical sentences in terms of the principles of the Guidelines, the SRA, and the Sixth Amendment. And, even if no argument can adequately justify identical sentences for meaningfully different convictions, it will be extremely difficult to determine exactly already been convicted at trial for the drug offense and therefore no longer had the rights of an unconvicted person. With that business out of the way, the judge then focused entirely on the murder and the guideline for first-degree murder in coming to a sentence. Williams , 2008 WL 2091152, at *4-5. 152 United States v. Allen, 488 F.3d 1244, 1261 (10th Cir. 2007). 153 See H UTCHISON ET AL ., supra note 142. 35 Published by Berkeley Law Scholarship Repository, 2012

37 Berkeley Journal of Criminal Law, Vol. 17, Iss. 2 [2012], Art. 3 L IETER F ALL 2012 EONARD /D UNISHMENT ITHOUT C ONVICTION [Vol. 17:2 P 295 W how much of a discount one offender should get when convicted of only some charges, relative to a co-conspirator who is convicted of all charges. We thus leave for another day the challenge of figuring out exactly where the line is in such cases between appropriate contextualization at sentencing and illegitimate punishment for unconvicted offenses. C ONCLUSION This article recognizes that in the current federal sentencing system, there is no ready way to eliminate the tension between the conviction stage of our criminal process and the sentencing stage. At the conviction stage, offenses have relatively simple definitions and a defendant, presumed innocent, can rely on a number of constitutional rights that make conviction difficult. After conviction, however, our system insists that substantial additional fact-finding is essential to properly contextualize the offense of conviction and produce a just and effective sentence within typically vast statutory sentencing ranges. This fact-finding may bring along some constitutional protections for the offender. But our system firmly rejects any claim to the full procedural rights of trial since the defendant has already been convicted and the remaining task is merely to judge the context of the offense for sentencing purposes. This disjunction between the trial and sentencing stages has permitted routine imposition of sentences that punish defendants for offenses of which the offender was never convicted. What can be done about that short of insisting on full trial rights for every sentencing fact? We are sympathetic to arguments for increased procedural protections at sentencing, although scholarship that has argued for the entire elimination of acquitted or unconvicted conduct has seemed to us We offer a different approach that one-dimensional and unrealistic. recognizes the legitimacy of the use of unconvicted conduct at sentencing when properly disciplined. That practice can be disciplined if the courts develop rules that exclude from sentencing discrete kinds of fact-finding and reasoning that clearly circumvent the imperative that a defendant must be punished only for the offense of conviction. This article makes clear why one such rule would exclude a finding of premeditation or of an underlying felony in a federal sentencing for second-degree murder. Judges should not have authority to make such findings and thus punish a second-degree murderer as if she or he was a first-degree murderer. To do so denigrates the role of https://scholarship.law.berkeley.edu/bjcl/vol17/iss2/3 36 DOI: https://doi.org/10.15779/Z382D08

38 Leonard and Dieter: Punishment Without Conviction: Controlling the Use of Unconvicted L IETER F ALL 2012 EONARD /D AW ERKELEY OURNAL OF C RIMINAL L B [Vol. 17:2 296 J conviction in the criminal process and runs counter to the purpose and structure of the Sentencing Guidelines, the Sentencing Reform Act, and the Supreme Court’s recent constitutional cases on double jeopardy and the jury right. This article also begins to explain what other related rules for disciplining punishment might look like, even as it embraces significant judicial fact-finding and discretion at sentencing. In the end, though, we rely on the idea that firm adoption of a few readily available rules will vindicate and entrench a principle — no sentencing for — that the Court has quietly articulated but never unconvicted offenses fully developed. Only when that principle is made operative and reliable in the clearest cases will courts and scholars be in a position to work out the full contours of the rules in harder cases. 37 Published by Berkeley Law Scholarship Repository, 2012

Related documents

Climate Risk in Southeast and Texas

Climate Risk in Southeast and Texas

The Bottom Line on Climate Change JULY 2015 COME HEAT AND HIGH WATER: CLIMATE RISK IN THE SOUTHEASTERN U.S. AND TEXAS

More info »
why competition in the politics industry is failing america

why competition in the politics industry is failing america

SEPTEMBER 2017 WHY COMPETITION IN THE POLITICS INDUSTRY IS FAILING AMERICA A strategy for reinvigorating our democracy Katherine M. Gehl and Michael E. Porter

More info »
OurBicentennialCrisis

OurBicentennialCrisis

OUR BICENTENNIAL CRISIS

More info »
20180430 Through Growth to Achievement Text

20180430 Through Growth to Achievement Text

Through Growth to Achievement Report of the Review to Achieve Educational Excellence in Australian Schools March 2018 Review Panel Mr David Gonski AC Mr Terrey Arcus AM Dr Ken Boston AO Ms Valerie Gou...

More info »
improving skills through americas workforce development system

improving skills through americas workforce development system

Improving Skills Through America’s Workforce Development System Kevin Bauman and Cody Christensen SEPTEMBER 2018 AMERICAN ENTERPRISE INSTITUTE

More info »
Main Street Matters Ideas for Improving Small Business Financing

Main Street Matters Ideas for Improving Small Business Financing

Main Street Matters Ideas for Improving Small Business Financing AUGUST 2018

More info »
2015 COP sexual abuse layout web 1

2015 COP sexual abuse layout web 1

THE SEXUAL ABUSE TO PRISON PIPELINE: THE GIRLS’ STORY Human Rights Project for Girls Georgetown Law Center on Poverty and Inequality Ms. Foundation for Women VERTY and INEQU CENTER ON PO ALITY

More info »
All the People, All the Places - Rural and Small-Town Civic Engagement (2018).pdf

All the People, All the Places - Rural and Small-Town Civic Engagement (2018).pdf

All the People, All the Places A Landscape of Opportunity for Rural and Small-Town Civic Engagement 2018 by Ben Goldfarb

More info »
Attorney General’s Advisory Committee on American Indian/Alaska Native Children Exposed to Violence:

Attorney General’s Advisory Committee on American Indian/Alaska Native Children Exposed to Violence:

Attorney General’s Advisory Committee on American Indian/Alaska Native Children Exposed to Violence: Ending Violence so Children Can Thrive NOVEMBER 2014

More info »
TC A Bid For Better Transit Publication 20170925 Digital

TC A Bid For Better Transit Publication 20170925 Digital

A Bid for Better Transit Improving service with contracted operations

More info »
Ending Violence So Children can Thrive

Ending Violence So Children can Thrive

Attorney General’s Advisory Committee on American Indian/Alaska Native Children Exposed to Violence: AMERICAN INDIAN AND ALASKA NATIVE CHILDREN EXPOSED TO VIOLENCE Ending Violence so Children Can Thri...

More info »
protectourspeciesprimerandactiontoolkit

protectourspeciesprimerandactiontoolkit

Earth Day 2019 Primer and Action Toolkit ® RTH K OR EA TW Y NE DA Protect Our Species Primer and Action Toolkit ® 1 K RTH Y NE TW OR DA EA

More info »
Review of the available remote sensing tools, products, methodologies and data to improve crop production forecasts

Review of the available remote sensing tools, products, methodologies and data to improve crop production forecasts

Review of the available Remote sensing tools, pRoducts, methodologies and data to impRove cRop pRoduction foRecasts

More info »
EPRS STU(2016)558777 EN

EPRS STU(2016)558777 EN

- The Cost of Non Europe in the Sharing Economy Economic, Social and Legal Challenges and Opportunities STUDY EPRS | European Parliamentary Research Service Author: Pierre Goudin Unit European Added V...

More info »
Where the Money Goes: Understanding Litigant Expenditures for Producing Electronic Discovery

Where the Money Goes: Understanding Litigant Expenditures for Producing Electronic Discovery

Institute for Civil Justice A RAND LAW, BUSINESS, AND REGULATION INSTITUTE CHILDREN AND FAMILIES The RAND Corporation is a nonprofit institution that helps improve policy and EDUCATION AND THE ARTS de...

More info »
Assessment of Minority Voting Rights Access

Assessment of Minority Voting Rights Access

AN ASSESSMENT OF MINORITY VOTING RIGHTS ACCESS IN THE UNITED STATES An Assessment of MINORITY VOTING RIGHTS ACCESS in the United States ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ 018 STATUTORY REPORT 2 U.S. COMMISSION ON CI...

More info »
grocerygap.original

grocerygap.original

One Penn Center, Suite 900 Headquarters: 1617 John F. Kennedy Blvd. 1438 Webster Street Philadelphia, PA 19103 Suite 303 t 215 575-0444 Oakland, CA 94612 f 215 575-0466 t 510 663-2333 f 510 663-9684 w...

More info »
New Booklet Single Final.indd

New Booklet Single Final.indd

A Guide to Quality in Online Learning

More info »
rr11

rr11

Read About It: Scientific Evidence for Effective Teaching of Reading Kerry Hempenstall Edited by Jennifer Buckingham Research Report | March 2016

More info »