Sorting Out White-Collar Crime

Article - Volume 97 - Issue 2

Our federal criminal code defines crimes, but declines to sort its fraud offenses according to degrees of harm or culpability. Although state prosecutors routinely charge crimes such as homicide or robbery in varying degrees, the federal code’s core fraud statutes are noticeably flat. There is no such thing as first- or second-degree fraud in the federal code.

Amidst a roiling debate as to whether the federal government overcriminalizes or underenforces white-collar crime, scholars have lost sight of the federal code’s lack of gradation. This Article seeks to remedy this neglect, particularly in regard to fraud crimes. Drawing examples from federal and state criminal codes, the Article analyzes the ways in which ungraded statutory regimes generate problematic and self-destructive expressive gaps. By lumping so much conduct under a single statutory umbrella, the federal code deprives the public of the ability to gauge the seriousness of a specific offense and of the will to discern those factors that separate the worst frauds from the merely bad ones.

If criminal law’s function is to distinguish wrongdoing and not solely to prohibit it, then our federal fraud statutes leave much to be desired. Reasonable people can debate the proper methodology for distinguishing bad from worse offenses, but it is quite another matter to abandon statutory sorting altogether. Accordingly, the Article closes by advocating the use of misdemeanor and low-level felony statutes to improve—and sort—the federal code’s fraud crimes.


What is white-collar crime’s greatest weakness? Some say it overcriminalizes, generating a mass of overlapping and redundant penal laws that punish too broadly, too precipitously, and too harshly.[1] Others claim just as passionately that it is underenforced: federal prosecutors[2] select for prosecution the easiest cases over the most serious, thereby shielding society’s most powerful from the punishment they so roundly deserve.[3]

This intractable and long-standing debate obscures a more vexatious problem: white-collar crime’s lack of statutory gradation. Unlike violent crimes and narcotics offenses, white-collar offenses include few meaningful statutory subdivisions. This is particularly true of the federal criminal code, the primary source of punishment for the triad of crimes known as fraud, bribery, and obstruction of justice.

The federal criminal code does not sort—at least not in any meaningful way—its fraud offenses. Despite its breadth and complexity, the code all but ignores the “degrees” that pervade state penal codes.[4] There is no such thing as first- or second-degree mail or wire fraud. Rather, all of the major fraud offenses, whether they threaten the evisceration of an entire industry or defraud an unfortunate few, fit under the same statutory umbrella. To be sure, different statutes apply depending on the offender’s use of mails, interstate wires, securities exchanges, or financial institutions to carry out his or her fraud. From a moral standpoint, however, federal fraud law remains “flat.”[5] To the extent the federal system sorts any of its fraud offenses, it does most of its work at the end, when the defendant is sentenced.[6] Even then, the sentencing phase focuses far more on sorting offenders and their individualized situations than it does on identifying those abstract factors that distinguish families of offenses.

Several consequences arise out of fraud’s lack of statutory gradation. The most notable is its expressive weakness. Crimes such as homicide, robbery, and rape feature well-understood labels, terms that convey qualitative distinctions among and between offenses.[7] Observers commonly speak of “murder in the first degree” and “manslaughter in the second degree.” One need not be an expert to know that the first-degree label reflects conduct more serious than a third- or fourth-degree crime, much less that manslaughter means something vastly different from murder.

By contrast, the federal fraud statutes define just one category of behavior: the “scheme or artifice to defraud”[8] undertaken through the use of some medium, such as the mails, the interstate wires, or the securities markets.[9] Because these distinctions are almost exclusively jurisdictional, they fail to educate the public, either in regard to the abstract factors that make one type of fraud worse than another or in regard to a specific offender’s culpability. As a result, the punishment the public expects to occur in a given case is often inconsistent with the punishment that actually does occur and inconsistent, yet again, with the punishment the public believes should occur.

Flatness is not merely an expressive problem; it is also a democratic one.[10] Graded statutes, such as homicide, do more than announce that killing someone is wrong. They reflect society’s views as to which types of killings are more serious and therefore more deserving of condemnation.[11] Ideally, these subdivisions reflect the transparent deliberation of a democratically elected body.[12] If criminal law is to embody this democratic ideal, then its subdivisions should be announced by legislatures, enshrined in written statutes, and communicated to the general public through the iterative process of prosecution, conviction, and judicial interpretation.[13] However frequently state codes fall short of this ideal, the federal code all but ignores it.

Defenders of the federal code might respond that the federal system’s sentencing regime obviates the need for the degrees we commonly associate with statutory grading. According to this line of argument, sentencing enhancements and offense-level reductions engineered under the United States Sentencing Guidelines (Sentencing Guidelines) more than adequately execute criminal law’s sorting function.[14] Indeed for some observers, the Sentencing Guidelines’ fine-grained provisions outperform a code’s blunter categories.[15] The juxtaposition of open-ended codes versus finely honed sentencing guidelines, however, perpetuates a false dichotomy. There is no need to choose one system over the other. To the contrary, a fair and equitable criminal justice system can employ statutory grading and sentencing guidelines, sorting its offenses as well as its offenders.[16]

Offense grading is a species of front-end sorting. It identifies in advance the offenses that merit less or more punishment. And it is a reflection, for good or bad, of a democratically elected legislature’s judgment. Back-end sorting—at least the kind that prevails under the Sentencing Guidelines—is more complicated, less salient, and less transparent;[17] its premises are grounded primarily in the judgments of an administrative agency.[18] Moreover, back-end sorting does not sort offenses in the abstract; rather, it sorts a bundle of factual circumstances relating to both the offense and the respective offenders who have engaged in that offense.[19] Thus, even under the best circumstances, sentencing complements grading but can never serve as a complete substitute for it.

How should one go about grading white-collar crime’s core fraud statutes? One might start with the courts, but extant interpretive doctrines barely encourage, much less mandate, the legislature’s grading of criminal statutes.[20] Perhaps a beefed-up conception of proportional punishment by the Supreme Court would force Congress to reexamine its statutory code, but at the end of the day, the solutions to this problem begin and end with the legislative branch.[21]

The legislature has always been the institution best positioned to define and grade crimes.[22] This view is captured by the “legality principle,” whose precepts demand that criminal law be enacted prospectively, in writing, and by a legislature.[23] But that only tells us that Congress ought to do something. It does not tell us what it is Congress should do. Congress might rewrite its federal fraud statutes from top to bottom, but history and common sense foretell the implausibility of a radical overhaul.[24] Accordingly, this Article concludes by suggesting Congress make greater use of both white-collar misdemeanors and low-level felonies (i.e., felonies capped at no more than two years’ imprisonment).[25]

The federal code already contains an impressive number of misdemeanor statutes, but it lacks an all-purpose provision applicable to lesser fraud cases. A set of misdemeanor and low-level felony statutes defining and distinguishing lesser frauds could set Congress on a fruitful path to introducing some needed gradation into one of its core white-collar offense categories. Moreover, this effort—if successful—might encourage legislators to divide and subdivide additional criminal offenses.

For good reasons, misdemeanor prosecutions have recently attracted substantial scholarly criticism.[26] Critics contend that they strip defendants of essential rights, dump offenders into overburdened state systems, and deny poverty-stricken defendants meaningful access to court. These concerns are valid, but not insurmountable for white-collar crimes. Federal courthouses command far greater resources than state courts and can provide better access and process to misdemeanants. Moreover, federal-misdemeanant offenders often enjoy the resources to more effectively exercise those rights.[27] Indeed, federal white-collar crime may be one of the few venues in which the misdemeanor statute can serve its intended purpose without generating so many of its negative externalities.

This Article advances criminal law’s discourse in several ways. First, it reveals the problems inherent in a system that fails to meaningfully distinguish its fraud offenses. Second, it uncovers and explicates statutory gradation’s educative and expressive functions. Finally, it proposes a venue in which the misdemeanor statute might be rehabilitated into a useful and principled tool of criminal justice.

The remainder of this Article unfolds as follows: Part I surveys several paradigmatic graded and ungraded criminal statutes, briefly recapping the historical factors that produced federal law’s flatness. Part II undertakes an institutional comparison, demonstrating the ungraded code’s failures along several dimensions. Parts III and IV confront the theoretical and practical challenges of introducing offense grading into an otherwise flat code. Part III explains why grading’s burden should lie first and foremost with the legislature and then proceeds to hypothesize a nuanced, multitiered rubric that incorporates mens rea and harm. Concluding that this idealized regime is politically implausible and potentially too complex, the Part ends with the recognition that a different fix is needed. Part IV therefore proceeds to hypothesize and sketch a misdemeanor fraud regime. Part V concludes.

I. Graded and Ungraded Crimes

Criminal law not only punishes, but also lumps and separates, sorting offenders according to culpability, harm, and a mix of other factors.[28] Through statutes, sentencing guidelines, and prosecutorial charging policies, the criminal justice system assigns criminal offenders a place on a penal continuum.[29] In its ideal form, it reserves its harshest punishments for only the most culpable and dangerous of offenders.[30]

The concept of graduated ranges of sanctions that correspond to increasing levels of harm and culpability is uncontroversial.[31] Disagreements arise, however, in the application of this sliding scale. It may be easy enough to conceptualize a punitive ladder for disparate offenses,[32] but differentiation grows more fraught for crimes that fall within the same family.[33] Since our nation’s founding, two types of regimes have attempted to address this problem. Graded criminal codes rely on legislatively written criminal statutes to make categorical distinctions. Ungraded codes collect a large amount of behavior under the same statutory umbrella.[34] The remainder of this Part highlights the differences between these two regimes, commencing first with a brief account of the divergent historical forces that pushed state legislatures to adopt statutory grading early on, and Congress to shift in the opposite direction during roughly the same time period.

A. A Brief History of Grading in Criminal Law

The first graded offense dates back to 1794, when Pennsylvania’s legislature divided its homicide statute into the two degrees of murder and manslaughter.[35] Other states soon followed course, in part to chip away at the then-prevailing common law rule that subjected all murder convictions to capital punishment.[36] Eventually, the intuition to subdivide crimes and label them by degree spread beyond homicide.[37] Today, gradation in state codes is the norm.[38]

Unlike most state codes, the federal criminal code, laid out in Title 18 and in additional titles such as Title 21, all but ignores the convention of labeling crimes by degree. A few statutes carve up federal law’s most serious offenses, such as murder and narcotics trafficking.[39] But the federal code’s white-collar offenses, particularly those that relate to fraud, feature little to no grading—and certainly none of the explicit labels we see in state codes.[40]

Why so little grading for federal crimes? One might start with the fact that federal criminal law barely existed in the years following the Revolution. As Kate Stith and Steve Koh have observed, “The first criminal statute was enacted in 1789 (even before the lower federal courts were established).”[41] Although other statutes would eventually follow[42]—most notably the mail fraud statute in 1872—federal criminal law nevertheless remained tightly moored to the protection of the federal government and its processes.[43] Accordingly, the preoccupation that drove early state law reform—namely, the fear that homicide and other felony offenses exposed too many offenders to capital punishment—was absent in the federal sphere.

Whereas the common law’s presence was still felt in state courts, federal courts declared early on that there was no such thing as federal criminal common law.[44] In some ways, federal criminal law benefitted from the courts’ explicit rejection of common law. Unlike its state-law counterparts, federal criminal law was a blank slate; Congress could shape its criminal laws—and criminal punishments—however it liked.[45] Early federal criminal statutes featured generic sentencing caps for imprisonment and within those caps, federal judges were granted the discretion to sentence how they saw fit.[46]

Where white-collar crimes were concerned, federal criminal law was relatively lenient. The mail-fraud statute’s statutory maximum sentence of imprisonment was just five years until 2002, when Congress increased it to twenty.[47] More generally, until the late 1980s, convictions for white-collar offenses did not necessarily result in prison sentences and even when they did, those sentences were not particularly onerous.[48] Thus, gradation was not a concern; the boundary that mattered most was the line between criminal and civil enforcement.[49]

None of this is to say that federal criminal law was perfect; it surely was not. Nevertheless, the forces that drove federal and state criminal justice reform differed substantially. Early state code reforms arose out of a concern that common law felonies were excessively punitive and insufficiently respectful of distinctions in moral culpability.[50] The way to fix this problem was to enact statutory provisions that distinguished the more and less serious variants of a given offense.[51] In doing so, Professor Lauren Ouziel observes, state criminal law morphed from a body of common law into a collection of statutes.[52]

Federal criminal law’s evolution followed a different path. Narrow at first, its scope broadened exponentially throughout much of the twentieth century,[53] owing in part to Congress’s enactment of substantive laws that overlapped state codes.[54] Many of these statutes represented cheap political reactions to singular events or scandals of the day.[55] Eventually, the common problems associated with federal criminal law in the modern era came to be its breadth, its overlap with state law, and its internal incoherence.[56] These critiques, however, either ignored or deemphasized the federal code’s lack of grading.[57]

Even in the modern era, federal and state criminal law continued to travel different paths. State codes (and state legislatures) benefitted from insights expressed in the American Law Institute’s 1962 Model Penal Code (MPC). The oft-praised MPC proved highly influential and led to the reformation of over thirty state codes.[58] Whereas the MPC streamlined[59] state codes by adopting uniform vocabulary and four basic mental states (purpose, knowledge, recklessness, and negligence),[60] it quite deliberately left intact the practice of legislatively articulating more and less serious variants of the same offense.[61] So, for example, although the MPC eliminated the multiple degrees many states had attached to the crime of murder, it still distributed homicide into three tiers depending on the offender’s culpability and the circumstances of the offense.[62] Indeed, the MPC’s opening section explicitly includes as one of its “general purposes” the “differentiat[ion] . . . between serious and minor offenses.”[63] Despite their effort to streamline overly formalistic statutes, the MPC’s architects still valued legislative grading, as did the architects of state codes.

The federal code, meanwhile, remained politically impervious to the MPC’s reforms. Ronald Gainer’s comprehensive history of this episode is highly instructive.[64] According to Gainer, for years, members of both houses sought reforms along the lines suggested by the Model Penal Code.[65] Had this effort succeeded, the federal code likely would have incorporated the MPC’s explicit embrace of statutory gradation, as well as its many other salutary characteristics. A more than decade-long effort to revise the federal code ultimately collapsed, however, and in its stead, Congress enacted a statute focused predominantly on sentencing.[66]

The 1984 Sentencing Reform Act sought to eliminate sentencing disparities and to correct the variance between a sentencing court’s formal judgment and the amount of time a prisoner actually served in prison.[67] The Act created and vested in the United States Sentencing Commission the responsibility for managing the federal government’s punishment apparatus.[68] Eliminating parole, the Act eclipsed the federal code’s broad, indeterminate[69] sentencing ranges with a highly regimented administrative sentencing regime, encapsulated by the United States Sentencing Guidelines.[70]

If the Guidelines’ intended purpose was to ensure greater sentencing uniformity,[71] its actual effect was to distance Congress from the nasty business of sorting related crimes.[72] From 1984 and afterwards, talk of federal code reform was effectively replaced with sentencing reform. If, prior to the Reform Act, there had been an interest in grading federal crimes, that interest was eclipsed by the Guidelines and its ubiquitous sentencing rubric.

B. State Crimes: Discrete and Subdivided

The preceding subpart traced the divergence between the evolution of state codes, whose architects experimented with grading as early as 1794, and the federal code, whose criminal statutes all but ignored the convention of organizing and labeling offenses in distinct degrees. The remaining two sections of this subpart investigate this contrast in greater detail.

1. Grading by Mens Rea: Homicide.—Homicide occurs when one or more persons cause the death of another person.[73] Murder—such as murder in the first degree—often occurs when an offender acts with the express purpose of bringing about another person’s death.[74] A number of jurisdictions further refine the first-degree murder offense to require the offender’s “willful, deliberate and premeditated” state of mind.[75] Courts have struggled to distinguish this triad from ordinary purpose or intent.[76] The Model Penal Code dispenses with the language altogether, preferring to rely solely on terms such as purpose or knowledge.[77] Nevertheless, the triad continues to be popular in several states, where it sometimes implies a planned or more wanton killing.[78]

Homicide law further subdivides intentional killings. In many jurisdictions, an unprovoked, intentional killing is treated more harshly than certain types of provoked killings, and a killing in the “heat of passion” or under “extreme emotional distress” reduces the offense from murder to voluntary manslaughter, a less serious offense.[79]

Further down the mens rea ladder,[80] homicide law subdivides reckless killings between those in which the offender deliberately ignored a substantial and unjustifiable risk,[81] and those evincing a “depraved mind” or “depraved heart.”[82] Depravity elevates the crime’s seriousness, as it implies the offender’s severe lack of care as to the outcome of his behavior.[83] Finally, a number of statutes punish negligent homicides, albeit these statutes often require a showing of gross negligence above the lack of care that would be sufficient for civil liability.[84]

To scholars and practitioners of criminal law, the preceding overview is well-known. That is not to say state codes have perfectly captured homicide and all of its distinctions. To the contrary, scholars strongly refute the premises underlying certain distinctions, and courts are often unable to settle on a stable definition of the term that separates one type of homicide from another.[85] Moreover, even when these doctrines successfully grade homicide, other doctrines, such as provocation or the felony-murder rule, can undermine the very distinctions courts and legislatures have so carefully crafted.[86]

Notwithstanding the foregoing, homicide’s grading system represents an unalloyed victory for front-end statutory sorting. It would be unthinkable today for a state legislature to collapse all of its homicide statutes into a single statute (e.g., “It shall be illegal to cause the death of another.”), and then leave all sorting of homicide-related crimes to sentencing judges or commissions.[87] Society understands a murder to mean something qualitatively different from a negligent homicide, and first-degree manslaughter to mean something different from third-degree manslaughter.[88] As I argue below in Part II, this means something—not just to scholars and jurists, but to the general public as well.

2. Grading by Harm: Robbery.—Robbery is the intentional taking of someone else’s property from another by force or threat of force.[89] Here, one finds state codes that subdivide the offense along lines of threatened or completed harm.[90] The Model Penal Code distinguishes a lesser crime such as theft from robbery in terms of physical injury or threat of such injury.[91] Although the Model Penal Code divides robbery into just two degrees,[92] states, such as New York, employ three or more degrees to distinguish more serious robberies from lesser ones.[93]

Consistent with its view of attempt, the Model Penal Code treats an attempted robbery as the equivalent of a completed one.[94] Many state codes have rejected this stance and continue to distinguish attempted robberies from completed ones, assigning a lower range of punishments to the incomplete offenses.[95] The distinction is desirable insofar as it promotes what economists refer to as “marginal deterrence.”[96] The separate and lesser punishment for an attempt incentivizes some criminals to abandon their efforts before they actually succeed.[97]

* * *

Across state criminal codes, gradation is common and proceeds according to different metrics; degreed offenses can reflect distinctions in the offender’s state of mind or differences in the degree of harm the offender imposes on his or her victim. Concededly, grading is not easy, and legislatures encounter difficulty when they attempt to incorporate more than one metric into their code.[98] Nevertheless, there remain many good reasons to grade a criminal code even if gradation produces imperfect outcomes. This becomes clearer when one considers the alternative, which is a code that either sporadically grades or does not grade at all.

C. Fraud Statutes: Overlap and Breadth

The federal criminal code (Title 18 of the United States Code) punishes many crimes, including crimes of violence. A separate set of statutes punish immigration offenses and narcotics trafficking.[99] Several of these offenses do in fact subdivide crimes according to degrees of harm (or, in the narcotics contexts, amounts of contraband), and these subdivisions do in fact correspond with accelerating statutory ranges of imprisonment and mandatory minimum sentences.[100] One would not be mistaken to conclude that federal grading—to the extent it exists—emerges primarily whenever Congress wishes to impose harsher punishments.[101]

Unlike narcotics and firearms offenses, the federal code’s white-collar crimes mostly ignore grading. There are a mix of statutes that criminalize domestic and federal bribery, statutes that prohibit many types of fraud, and over ten statutes that criminalize a variety of obstructive behavior.[102] One would be hard pressed, however, to divine a rigorous grading rubric from these statutes. Their maximum terms of imprisonment (five, ten, or twenty years) populate the Code in an almost ad hoc fashion, reflecting the latest congressional compromise of the moment rather than any attempt to arrange crimes in order of their wrongfulness or risk or degree of harm.

None of this would matter so much if the statutes themselves were rarely charged. But fraud is in fact one of the more popular of the federal white-collar-crime offenses.[103] Of those convicted and sentenced annually in federal court, at least 10% are fraud cases.[104] An array of federal statutes prohibit fraud in different contexts, but all statutes share certain characteristics relevant to the discussion of grading.

1. Specialization and Overlap.—The federal code’s fraud laws are both specialized[105] and yet prone to significant overlap.[106] A single offense—a doctor’s decision to submit false insurance reports by email or mail, for example—implicates multiple statutes. Depending on the circumstances, the doctor may be guilty of violating the mail or wire fraud statutes (or both), the federal health-care fraud statute, the criminal False Claims Act, one or more conspiracy statutes, and potentially other related statutes, such as the bank or securities fraud statutes.[107]

If the offending doctor asks his attorney which fraud violations are more serious than the others, his attorney will likely explain that the distinctions between these offenses are largely jurisdictional (using the mail or the interstate wires) or relate to the victim’s identity (defrauding the government, a health care institution, or a financial institution). Otherwise, they reflect relatively little in terms of society’s views as to which crimes are more morally wrongful or dangerous.

Moreover, none of the fraud statutes themselves reflect any gradation in punishment. Barring certain circumstances, the maximum sentence for a mail or wire fraud offense is twenty years’ imprisonment; the maximum for bank fraud, thirty years’ imprisonment; for securities fraud, depending on the statute, either twenty or twenty-five years’ imprisonment; and for health care fraud, ten years’ imprisonment.[108]

For most prisoners, these maximum caps are irrelevant.[109] The offender’s behavior will be analyzed, in the first instance, under the Federal Sentencing Guidelines,[110] and under the Guidelines, most of the charged offenses will not come even close to grazing the statute’s maximum term.[111] In 2016, twenty-four months’ imprisonment was the median sentence for those whose primary charge was a fraud offense and who also received a prison sentence.[112] For over a third of fraud and theft cases sentenced in 2016, the amount of actual or intended loss amounted to less than $95,000.[113] Thus, the fraud crimes for which most offenders are convicted have little to do with the imaginary offenses that animated Congress to set maximum penalties of a decade or more.

As others have noted, even in those relatively few cases in which circumstances warrant a sentence exceeding the statutory maximum, federal prosecutors can almost always charge multiple offenses for the same chain of events.[114] Thus, when necessary, sentencing judges can construct long sentences consecutively over several counts.

2. Undifferentiated Mental States.—As subpart I(A) observed, state codes differentiate crimes such as homicide according to the offender’s mens rea. Purposeful behavior defines a crime more serious than knowing or reckless behavior, and so forth. By contrast, federal fraud law largely eschews these subdivisions. Instead, each of the major statutes criminalizing fraud employs a near-identical variant of the “willful [and] specific intent to defraud” test.[115]

Although the federal code requires a specific intent to defraud, courts define the term with surprisingly broad language. The least controversial articulation is the one that describes fraud as a purposeful scheme, designed to separate a victim from his money or tangible property, through deceptive conduct.[116] Unlike homicide, all purposeful fraud schemes receive the same statutory treatment, regardless of whether the offender acted upon impulse or carefully concocted a scheme over a period of days or weeks.[117] Thus, the heat-of-passion fraud is treated identically to the premeditated one. And whereas provocation may serve as a partial defense to homicide, it has no equivalent in federal fraud law. The planned fraud is charged no differently from the impulsive fraud; the provoked fraud receives the same charge as the wanton one. To the extent any of these distinctions matter, they are credited solely at sentencing.

But that is not all. Fraud criminalizes more than purposeful efforts to deceive; in a number of courts, it extends to reckless behavior. Indeed, the law’s treatment of recklessness is perhaps the most confusing aspect of this crime. Trial judges begin with the proposition that fraud requires purposefully deceptive conduct,[118] but those same courts agree that statements made in reckless disregard of their falsity can be used to infer a purposeful scheme to defraud.[119] From there, a number of courts stretch reckless disregard into an independent basis for liability, dropping the purposeful scheme language entirely.[120] As Samuel Buell has argued, these interpretive moves elide the distinction between a statement delivered with reckless indifference to the truth and irresponsible conduct that happens to result in someone else’s loss.[121] The former represents qualitatively different behavior from the latter, particularly if one’s aim is to punish purposely deceptive conduct.[122]

3. Reliance on Sentencing.—Because the federal criminal code declines to differentiate fraud up front—either by amount, mens rea, or degree of risk—whatever sorting there is of fraud offenses takes place at sentencing. The garden-variety con artist, the mid-level associate who played a bit part in a securities fraud, and the ringleader of a Ponzi scheme may all be charged with three or four counts of wire and mail fraud, but their recommended sentences will vary widely under the United States Sentencing Guidelines’ primary sentencing provision, § 2B1.1.[123]

Section 2B1.1 is the Guidelines’ provision that governs economic crimes such as fraud.[124] It sorts offenses by examining the amount of actual or intended loss to victims, which is laid out in the section’s loss table.[125] Additional provisions enhance or mitigate the offender’s offense level, although loss by far plays the most important role in determining the offender’s recommended sentence range.[126] Apart from § 2B1.1, other parts of the Guidelines further sort offenders. Conspirators who played relatively minor roles in the scheme are eligible for “role in the offense” reductions, while those who managed or supervised the offense are subject to enhanced punishment.[127]

The pervasive criticism of § 2B1.1 is not that the section fails to sort, but that it does so mindlessly.[128] It overemphasizes quantifiable factors such as loss amount, and all but ignores the qualitative state-of-mind distinctions that play such a big role in homicide grading.[129] Moreover, it pays insufficient attention to whether a scheme has been completed, was in its infant stages, or was entirely implausible.[130]

In 2015, in response to the observation that federal judges were increasingly diverging from § 2B1.1’s recommended sentencing ranges,[131] the United States Sentencing Commission convened a two-year listening tour that included comments from jurists, practitioners, and the Department of Justice.[132] The end result was a modestly revised guideline in which the Commission softened several of § 2B1.1’s harsher provisions, but otherwise left the Guideline’s emphasis on loss amount intact. Speaking for the Commission in her capacity as Chair, Judge Patti Saris stated in 2015, “[W]e [find] that in most cases, the fraud guideline works well to distinguish the more- and less-culpable offenders.”[133] Not everyone agreed.[134]

D. The Spoilers: Conspiracy and Accomplice Liability

Subparts B and C contrasted graded and ungraded criminal statutes. State codes rely on a mix of factors to statutorily distinguish offenses; federal fraud law relies primarily on the Sentencing Guidelines to do its sorting. The state code partially sorts its offenses at the front end. The federal system instead relies on its sentencing system to do all of the sorting at the back end.

Two doctrines confuse this neat dichotomy: conspiracy and modern accomplice liability. Both pervade state and federal criminal liability, and both are notably undifferentiated in terms of culpability or participation.[135] The typical conspiracy statute punishes a coconspirator for agreeing to commit a crime with another.[136] Accomplice liability punishes those who aid or abet an offense with the intent of facilitating or promoting it.[137]

Conspiracy and accomplice liability are flat statutes. Like the substantive fraud statutes themselves, neither state nor federal codes parse the coconspirator’s state of mind beyond his purposeful association with the venture. Nor does the law focus on his degree of assistance or the centrality of his participation to the success of the venture. Thus, the law does not care whether someone joined enthusiastically or reluctantly, whether he provided grudgingly small or remarkably strong assistance, or whether his conduct played an important or irrelevant role in the criminal enterprise.[138] To the contrary, the reluctant, ineffective conspirator suffers as much exposure as the gleeful, essential one.[139]

Modern accomplice liability is no different. Although complicity law once distinguished actors by their participation, today’s statutes treat accomplices and principals identically.[140] The prosecutor still must establish that the defendant acted with the purpose of promoting the underlying crime,[141] but the jury can infer such purpose from the accomplice’s knowing participation.[142] And, as Justice Kagan bluntly advised in Rosemond v. United States,[143] the reluctant accomplice is just as guilty as the enthusiastic one.[144]

Conspiracy and accomplice liability admittedly complicate the comparison of federal and state codes. If state legislatures care so much about front-end, statutory sorting, why do they permit these doctrines to collapse so many offenders into a single overly broad category? It’s a worthy question, but it is one best saved for another day.[145] Although conspiracy and complicity are counterproductive insofar as grading is concerned, they do not entirely undermine the statutory sorting scheme. Grades still play a role in determining the predicate crime an actor has either aided or conspired to commit. Moreover, these doctrines play no role whatsoever in determining the liability of the defendant who, acting by himself, commits a singular, discrete offense.[146] Accordingly, group-crime doctrines are something we should keep in mind when we consider remedies. They are not cause, however, for abandoning a grading project altogether.

II. The Advantages of Grading

Some regimes rely on their statutes to distinguish a family of offenses, while others rely primarily on their sentencing institutions to sort their offenders.[147] To what extent does this distinction matter? This Part undertakes an institutional analysis, examining grading’s expressive and systemic benefits along several dimensions.

A. Deterrence and Retribution

Scholars have long debated justifications for punishment. The two theories which receive the greatest attention are deterrence and retribution.[148]

Punishment deters when it persuades a putative offender not to engage in a given crime. Rational actors, in turn, take into account the costs of engaging in a particular crime and desist from engaging in it when its costs exceed its benefits.[149] Thus, the variables that matter most for deterrence theory are the sanction and its probability of occurrence.[150] To efficiently deter wrongdoing, the state should set the sanction, modified by the probability of punishment, just slightly higher than the offender’s expected gain, while taking into account other variables such as the social costs of policing and imposing punishment.[151]

To improve deterrence, the state can operate a mix of levers. It can increase sanctions dramatically and save money on enforcement.[152] Or, it can instead ramp up enforcement and keep sanctions relatively low—the common strategy for traffic and strict-liability offenses.[153] It can experiment with some scheme of gradually increasing sanctions matched with equally specialized amounts of enforcement. Or, if it wants to get creative, the state can alter opportunities to commit crimes through structural changes (e.g., making a road bumpy in order to reduce speeding around a bend).[154] Regardless of the levers it chooses, the state’s goal is always the same: the reduction of harmful behavior.

By contrast, retributive theory explicitly ties criminal punishment to desert.[155] A criminal conviction is appropriately retributive when it is deserved and when it is imposed proportionately.[156] Beyond this bromide, retributivists disagree on how often or how severely defendants should be punished. Under a limiting theory, moral desert sets the upper bounds for punishment, but yields to other considerations, including resource concerns and interests in crime reduction.[157] Other variations of the moral-desert theory obligate the state to punish the offender, even if doing so would be costly or make no difference in the future frequency or severity of wrongdoing.[158]

Does legislative grading reliably improve deterrence or retribution? One might conclude that it has no effect either way. Offense grading is, after all, just one aspect of a complex criminal-justice regime. For deterrence, the offender’s perceived probability of detection is probably most determinative.[159] As for retribution, actual punishment outcomes, including the collateral effects that accompany a given term of imprisonment, seem far more important than whether a criminal code features two or three variants of the same offense.

Scratch beneath the surface, however, and the graded system’s benefits begin to emerge. First, the graded system’s labels (“first degree X”) tend to be more salient and better understood than the ungraded system’s sentencing guidelines.[160] These labels, in turn, encourage some subset of defendants to engage in less harmful conduct, either because they fear the harsher stigma (e.g., being called a rapist) or because the label more effectively infiltrates the public lexicon, thereby altering the public’s view of the offense (e.g., driving while texting).[161] Accordingly, if a state employs consistent statutory labels to signal the seriousness of a crime and those labels consequently cause some nontrivial group of individuals to alter their behavior, we can attribute the improvement in deterrence to the code’s use of offense grading.[162]

The retributive argument for grading is less straightforward. The actual length and terms of confinement can be as just or unjust under a graded or ungraded scheme. Moreover, graded systems can feature just as much, if not more, irrationality, particularly if legislatures tinker with codes by adding new and inconsistent statutes and penalties over time.[163] In fact, several of the worst examples of disproportionate punishment hail from state criminal-justice systems whose codes conscientiously divide and subdivide crimes into degrees.[164]

On the other hand, if a given crime’s label carries an additional stigma,[165] and we consider that stigma to be an inherent component of punishment, then the graded system, by definition, sets forth a more precise punishment that is more likely to be reflective of the perpetrator’s actual conduct. Criminal philosophers refer to this as the “fair labeling” argument.[166] As Professor Stuart Green argues, “[L]abeling is important in determining what constitutes a proportional punishment, for evaluating prior convictions, and in sending the appropriate signal to the public.”[167] In sum, labels matter, and in criminal law, offense grades happen to be the mechanisms we most commonly use to convey those labels.

The graded system rewards the negligent killer with a label less daunting than murder. The ungraded system, by contrast, demands that the two-bit con artist bear the full stigmatic weight of a fraud conviction, alongside the Ponzi schemer and the CEO who presides over a massive accounting fraud. Compared to other consequences, this may seem like a minor matter, but to the putative offender, it can be quite significant.

Notwithstanding the labeling problem, it is doubtful that anyone would rush to revise an ungraded code solely on grounds of improving deterrence or retribution. Still, in regard to both goals, the graded system enjoys a slight edge over its ungraded counterpart. This edge grows wider when we consider additional factors discussed below.

B. Proof Rights and Ceiling Rights

Any well-run criminal-justice system will be judged not just on its punishment outcomes but also on its procedures. The graded system vests the defendant with two very important rights, which I refer to here as “ceiling rights” and “proof rights.” These rights and their effective dilution have already been the subject of much scholarly discussion,[168] but I repeat them here to demonstrate their connection to statutory grading.

The graded regime creates a ceiling right by guaranteeing that the defendant’s punishment will fall at or below some legislatively prescribed ceiling. If the state charges an individual with a crime known as negligent homicide, it cannot seek a sentence above negligent homicide’s prescribed maximum, which might be five years’ imprisonment.

At the same time, the graded system creates a corresponding proof right—the prosecution cannot bump the defendant into a higher sentencing category unless it proves, beyond a reasonable doubt, the existence of those additional elements that define the more serious offense.[169]

Under a graded system, proof rights and ceiling rights ideally enjoy a reciprocal relationship. That is, a statute featuring a high or nonexistent ceiling (e.g., life without parole) ought to require proof of either a high mental state (e.g., purpose) and/‌or particularly serious conduct. By contrast, the statute that features a low statutory ceiling (say, five years’ imprisonment) might well feature relatively weak proof obligations for the prosecutor (say, evidence the offender acted negligently or recklessly).[170]

Consider the following hypothetical: Joe accidentally kills Susan through his negligent behavior and is prosecuted under a negligent-homicide statute that carries a maximum of five years’ imprisonment. Joe’s ceiling right is strong, but his proof rights are correspondingly weak. The prosecutor does not have to demonstrate a difficult-to-prove mental state such as purpose or intent. Rather, it must show nothing more than Joe’s gross failure to adhere to commonly accepted norms of care. If, on the other hand, Joe intentionally kills Susan after premeditation, he may be eligible for a much more serious offense, such as first- or second-degree murder. His ceiling rights become less valuable (because the more serious crimes carry much higher maximum sentences), but his proof rights grow stronger.

Together, proof and ceiling rights restrain prosecutorial power.[171] That is, they make it more difficult for the prosecutor to threaten a minor offender with the type of sentence we ordinarily would reserve for more serious variants of the same offense. Granted, these rights offer little protection against outright prosecutorial misconduct (e.g., withholding exculpatory evidence) and may be overshadowed by other factors, such as the defendant’s lack of resources or the prosecutor’s plea-bargaining tactics.[172] Nevertheless, they restrain those prosecutors who view themselves as law-abiding because they require conscientious prosecutors to do more work: acquire more evidence, interview more witnesses, research more law, and seek more internal office approvals in advance of charging criminal defendants with more serious crimes.[173] Accordingly, over time, the ceiling and proof rights embedded in a graded code positively impact some defendants, at least some of the time.[174]

Ungraded systems abandon the foregoing safeguards. The ungraded statute is one whose statutory terms permit a punishment of anything from zero to as much as twenty or thirty years’ imprisonment. Because the same statute is intended to address serious as well as less serious misconduct, it eliminates the offender’s ceiling rights while also watering down his proof rights. Aggravating factors that in the graded system would be considered statutory elements are downgraded to mere sentencing factors, which can be proven by a mere preponderance of the evidence, provided the legislature steers clear of a mandatory-sentencing regime.[175]

Although the Supreme Court could mitigate this problem by declaring sentencing factors the equivalents of statutory elements, it has no need to do so now that it has declared the Sentencing Guidelines merely advisory.[176] Accordingly, unless a sentencing enhancement triggers a mandatory minimum sentence[177] or elevates the statute’s maximum allowable sentence,[178] it need only be proven by a preponderance of the evidence.[179] Thus, in the ungraded system, there are no ceiling rights or proof rights.

C. Salience and Legitimacy

As subpart A observed, many of the labels that populate state codes have been woven into popular culture. Laypeople are aware of the differences between manslaughter and murder, between robbery and theft, and even between rape and sexual misconduct. The graded system’s labels are salient.[180]

Salience intertwines with legitimacy. For the criminal justice system to properly function, it must be perceived by its citizens as morally and politically legitimate.[181] Legitimacy yields voluntary compliance with the law, as well as cooperation between citizens and law enforcement institutions.[182] Moreover, it serves as the underlying premise for the state-sanctioned punishment.[183] If we grant the state a monopoly in exerting force on its citizens, then we need substantive and procedural laws that render such force legitimate.[184]

At first glance, there is no reason to assume graded systems enjoy any greater sense of legitimacy than ungraded ones. Under both types of regimes, the legislature announces what is and is not a crime; in both instances, there is the possibility that the legislature will misfire by punishing the wrong behaviors or by punishing them too harshly.[185] Moreover, to the extent legitimacy derives from procedural justice,[186] grading seems to matter not at all. Mass incarceration can occur regardless of the number of subdivisions contained in a single statute,[187] and police and prosecutors can behave deplorably regardless of how many degrees a legislature builds into its exquisitely crafted code.

Nevertheless, to the extent one believes strongly in the separation of powers, the graded system outshines the alternative.[188] Graded systems reflect on-the-record legislative judgments as to which crimes are bad and worse. The authors of those judgments become accountable to the citizens who elect them.[189] These legislative judgments, in turn, receive further amplification when judges interpret them in pre- and post-trial motions, and when juries apply them in criminal trials. To the extent one views the jury as a rough proxy for the public, the criminal trial under the graded system represents a democratic ideal, with one representative body (the legislature) voicing its sorting principles in the abstract and another representative body (the jury) implementing its own sorting analysis on a case-by-case basis.[190]

By contrast, ungraded systems delegate judgments to unelected and unaccountable experts, such as line prosecutors, probation officers, and sentencing judges.[191] Superficially, the ungraded system’s code appears more streamlined since it dispenses with degrees. It redoubles its complexity at the back end, however, relying on its Guidelines technicians to sort, in relative obscurity, the criminal justice system’s offenses and offenders.[192]

Thus, the graded system is, at least in the abstract, preferable to the ungraded one. This preference, however, ought not to be blown out of proportion. A graded system can easily fall short of its ideals—its actors can violate procedural rules, twist the meaning of criminal statutes, or attach crude sentencing triggers that all but destroy the legislature’s formal gradations.[193] By the same token, enlightened lawmakers could just as plausibly alleviate the ungraded system’s legitimacy shortfalls by adopting a series of accountability measures, many of which have been proposed by a number of scholars in the past decade.[194] In criminal law, as in other domains, the actors who run the system are nearly as important as the system’s structure.

Nevertheless, if democratic legitimacy and participation are one’s guiding lights, then the graded code is clearly the better option. Graded statutes teach us something about the conceptual differences between closely related crimes, and they do it in a more effective way than the more technical, complicated, and diffuse sentencing-guideline regime that is the hallmark of the ungraded system. Grading at its best is better than its alternative.

D. Statutory Interpretation

Because criminal law relies so heavily on statutes, one might wonder how graded and nongraded codes fare in terms of statutory interpretation. At first glance, one might conclude that this is the one area where grading falters in comparison to flatter codes. Because it is so reliant on statutory terms, grading arguably provokes more pre- and post-trial litigation. Trial and appellate courts must distinguish the premeditated killing from the merely intentional one, decide if a robbery caused someone serious bodily injury, and so forth. The more grades the legislature introduces into its code, the greater the need for judges to interpret the terms that separate them. The opportunity for error—and for costly legal wrangling—increases with each new grade.

At the same time, as several recent Supreme Court cases demonstrate, interpretive problems still surface when a jurisdiction explicitly declines to subdivide its crimes. In the graded system, courts are tasked with interpreting the statutory language that purports to distinguish one degree from another. In the federal system, interpretive debates are fueled by unstated prototypes.[195] Defendants couch their claims in lenity or vagueness terms, but their unstated argument is that a case diverges too far from a given crime’s prototype to justify punishment.[196]

Consider several of the most notable white-collar cases decided by the Supreme Court over the past decade. Jeffrey Skilling, one of the architects of Enron’s infamous rise and demise, successfully challenged the government’s prosecution of his theft of “honest services” because his behavior appeared too dissimilar from a concealed payment of a bribe or kickback.[197] Meanwhile, John Yates, a commercial fisherman who unsuccessfully concealed his catch of too-small fish, persuaded five justices that a fish was not a “tangible object” as defined in an obstruction of justice statute.[198] And finally, the governor of Virginia was able to escape liability for accepting a Rolex and other valuables from a constituent because the invitations to parties that he provided in exchange diverged too much from the Court’s conception of what an “official act” looks like when the Court imagines a typical bribery case.[199]

In each of the foregoing cases, the extent to which the charged offense diverged from its presumptive prototype played an essential role in the Court’s statutory interpretation. The authors of the Skilling, Yates, and McDonnell opinions narrowed key statutory terms—sometimes beyond the point of recognition—to keep a criminal statute from diverging too far from its intended prototype. In each of the three cases, the Court transparently sought to protect the defendant from being branded with a certain label, and, in the Yates case, to alleviate the (infinitesimally small) risk of excessive punishment that might accrue to individuals who were situated similarly to Yates.[200]

Reasonable people can debate whether the Court should engage in such analysis, or whether the maintenance of prototypes should be left solely to prosecutors, juries, or legislators.[201] The more interesting question is whether the Court would undertake such work if the federal code were to subdivide its fraud, bribery, and obstruction statutes into easily recognizable degrees. Had John Yates, the commercial fisherman, been charged with substantially the same statute (i.e., the illegal concealment of a tangible object in the course of an investigation), but with a label such as “Fourth-Degree Interference in a Regulatory Investigation,” and exposed to the risk of nothing more than a very short maximum sentence (say, two years’ imprisonment), would the Yates plurality have been so put off by the government’s definition of tangible object?[202] Indeed, would the Court have even granted certiorari?[203]

To be sure, graded statutes do not guarantee good statutory interpretation. Ambiguity is a fact of life, and judges can interpret statutes incorrectly for a host of reasons. Nevertheless, by signaling from the very beginning that an offense is a minor one, the graded system avoids the spurious claim—made nearly as often by the government as by the defendant’s attorneys—that a given defendant “faces” a twenty-year sentence or worse.[204]

III. Implausible (and Ineffective) Remedies

Assume we reach consensus that, all things being equal, a graded criminal justice system is preferable to an ungraded one. How might we go about reforming an ungraded code? Should we focus on the entire code, undertaking a top-to-bottom revision, or should we settle on more incremental reforms? And should we rely solely on legislative reform? Might the judiciary have some salutary role to play in this reformatory effort?

This Part takes up several of these questions.

A. The Shortfalls of Interpretive Remedies

With certain narrow exceptions, the Supreme Court has long manifested a hands-off attitude towards substantive criminal law.[205] It occasionally exercises its muscle when substantive statutes undermine First Amendment rights,[206] are excessively vague,[207] or, as in the case of Yates, when prosecutors appear to stretch a federal criminal statute too far beyond its intended purpose. Scholars such as Shon Hopwood have urged the judicial branch to amplify these doctrines and impose on Congress a “clear statement” rule for its federal criminal statutes.[208] Stephen Smith similarly argues either for a beefed-up application of the rule of lenity or an interpretive rule that narrows the application of a criminal statute whenever a broader interpretation would result in the imposition of a disproportionate punishment.[209]

Whatever the independent merits of these doctrinal reforms, none of them solves the grading issues described in Parts I and II. To take Professor Hopwood’s proposal first, clear-statement rules stop short of mandating a legislature’s subdivision of criminal offenses. They might encourage some gradation here and there, but they just as easily might encourage Congress to state (and restate) in clarion terms just how broadly it would like a single statute to apply.

From the same perspective, Professor Smith’s proposals fare no better. A more robust lenity rule contracts the scope of an ambiguous statute, but it does not mandate offense grading. And a rule that narrows the application of a criminal law in light of its potentially disproportionate sentence is again premised on the assumption that the statute’s terminology is ambiguous enough to invite interpretive wrangling.[210] As the Yates case itself demonstrates, the exposure of an individual to a long maximum penalty may well affect the justices’ view of a given statute.[211] The primary target of decisions like these, however, is not the legislature so much as it is the executive branch. Cases like Yates do not spur the revision of federal obstruction statutes; rather, they cause federal prosecutors to think twice before pursuing minor outlier cases.

If vagueness and statutory interpretation are unlikely to spur grading, what will? The most obvious doctrinal candidate is the Eighth Amendment’s Cruel and Unusual Punishment Clause, which purportedly requires the government to avoid disproportionate punishments. Even here, however, the Court has signaled its unwillingness to intervene.[212] “[O]nly an extreme disparity between crime and sentence offends the Eighth Amendment.”[213] Given the Court’s jurisprudence on what counts as an extreme disparity, Congress need not worry about its fraud statutes, whose maximum penalties may appear extreme, but whose actual penalties are often modest, particularly in comparison to street-crime cases in state courts.[214] Most importantly, given the federal code’s history, it is inconceivable that the Supreme Court would strike down a federal white-collar crime statute solely on account of its lack of gradation.[215]

B. The Difficulty of a Full Overhaul

The potential judicial remedies described above are not only unlikely but also misguided because they focus excessive attention on the wrong institution. Primary responsibility for fixing the criminal code ought to lie with the legislature, the branch responsible for drafting penal statutes in the first place, and not the judiciary.

Were Congress to solve fraud’s flatness, it could do so either by revamping its fraud statutes completely or by undertaking a less radical alternative, which I describe more fully in Part IV.[216] Because a top-to-bottom overhaul is, in many critics’ minds, the ideal solution, I consider it first in this Part. What would the federal fraud statutes look like were Congress to take seriously its obligation to separate the worse from merely bad conduct?

Grading can be informed by one or more theories of punishment. Drafters of a legislative code might draw upon economic deterrence principles, seeking a grading system that optimally deters harm.[217] Alternatively, they might prefer the more philosophical retributive approach, which is neither “hard science [n]or a mathematical calculation; it requires discretion, moral reasoning, and the exercise of judgment.”[218] Finally, separately or in conjunction with the other approaches, a legislature might rely more directly on the public’s intuitive views of certain crimes.[219] In this vein, legislatures might find helpful studies of the sort published by Stuart Green and Matthew Kugler, who have surveyed individuals on their views of fraud, bribery and obstruction, and insider trading.[220]

Assuming Congress adopted one of these approaches, it likely would subdivide its mail, wire, and associated fraud statutes into a series of graded offenses. These subdivisions, one should note, should not reflect the miserably confusing additions and subtractions one currently encounters in the federal code’s highly specialized computer and identification fraud statutes.[221] These enhancements are simultaneously too numerous, too idiosyncratic, and reflect far too many ad hoc amendments to serve as a model for legislators.[222] No, in an idealized world, we would have some sort of fraud rubric that reflected nuanced considerations beyond some arbitrary loss amount.[223] Unfortunately, even a pared down one could grow quite complicated.

First, were one to draw guidance from the homicide and robbery statutes discussed in Part I, one might well pay attention to the offender’s state of mind and the degree or type of harm. These two concepts, however, create what Professor Kenneth Simons refers to as an “incommensurability” problem.[224] In terms of culpability, state of mind and risk of harm represent different, and in many ways, incomparable values.[225]

Putting aside the foregoing, one can see how a graded fraud statute would fission very quickly. A code that focused on the offender’s moral culpability might first distinguish the meticulously planned fraud from the impulsive one, and the purposeful lie from the statement made with reckless indifference to the truth. The same code might distinguish frauds that have exacted harm and inchoate schemes.[226] Finally, a code might further subdivide potentially successful inchoate frauds from wholly implausible ones.

As the foregoing discussion suggests, the grading rubric could quickly become rather complicated. Based on mens rea alone, one could imagine four separate degrees:

If one wanted to incorporate additional degrees to reflect a distinction between inchoate and completed frauds, the rubric would expand slightly, although some might argue that since an attempted fraud is one that, by definition, the actor purposely intends to bring about, one cannot create provisions for “reckless attempted fraud” because recklessness and attempted behavior are inherently incompatible.[227] Nevertheless, one could set up a distinction between attempted purposeful fraud and completed purposeful fraud. Thus, the rubric might look something like Figure 2:

To those familiar with federal criminal law, the level of differentiation in Figures 1 and 2 represents a substantial departure from current practice. To scholars familiar with state codes, however, the differentiation described above is rather modest, as state legislatures routinely group offenses in five, seven, or even nine degrees.[228]

In any event, the rubric is not the problem—or at least not the only problem. As Ronald Gainer points out, even in the best of times, federal code reform requires sustained attention and difficult tradeoffs.[229] And as Professor Erin Murphy has more recently recounted in her discussion of sexual assault, code reform becomes more difficult when rulemakers attempt to layer it onto an already existing set of statutes.[230] Add to that the difficulty of attaching arbitrary punishment ranges to the newly subdivided offenses,[231] and one likely ends up with an intractable political stalemate.

In a period in which bipartisan cooperation has grown increasingly rare, federal code reform takes on a quixotic cast. As Gainer bluntly observed in 2011, “Our current political environment does not seem to offer the circumstances required to engender reasoned and dispassionate congressional cooperation.”[232] If that was the case in 2011, it is difficult to muster much hope for code reform’s success in 2019.

C. The Multiple Counts Problem

Aside from code reform’s political implausibility, there is an additional problem that attaches to any grade scheme whose sanctions are smoothly graduated (i.e., not too much of a bump between the maximum sentence for one category to the next). That problem, previewed earlier in Part I, is the practice of count-stacking, a practice that is particularly problematic in the fraud context.[233] Count-stacking occurs when a prosecutor purposely charges a defendant in multiple counts in order to maximize the defendant’s maximum sentence and secure greater leverage in plea bargaining.[234]

Because the federal code punishes the use of the “mails” and “wires” to execute schemes, fraud often can be charged as more than one crime. As a result, prosecutors can avoid graded statutes and their statutory caps by charging a crime in two or more counts. In other words, even if reckless fraud is designated a third-degree felony, the prosecutor can still achieve a maximum sentence cap comparable to a first-degree felony if she charges the reckless fraud in multiple counts.

To confront this problem, legislatures could select from a menu of options: they could place direct restraints on a prosecutor’s charging of multiple related counts, redefine the unit of crime to make count-stacking more difficult, or severely limit the availability of consecutive sentences. As Andrew Manuel Crespo observes, procedural constraints on count-stacking do exist and have in fact been employed in state jurisdictions.[235] Until those changes migrate to the federal context, however, a full-blown graduated statutory grading system must give way to alternative statutory tools, one of which is white-collar misdemeanor discussed in Part IV.[236]

* * *

Congress would encounter a litany of political and practical challenges were it to take up the complex job of grading its fraud statutes. Critics would castigate the rubrics laid out in Figures 1 and 2 as confusing, unnecessary, and prone to litigation. The lesson, however, is not that reformers should give up. Rather, it is that reformers should instead seek alternative measures that capture the gist of statutory differentiation without imposing its costs. In the next Part, I propose such a mechanism.

IV. Fraud and the White-Collar Misdemeanor

Part III sketched an idealized fraud scheme that incorporated subdivisions based on mens rea and degrees of harm. It suggested differential treatment for frauds that were merely inchoate, that were highly implausible, that were brought about solely by reckless conduct, or that were undertaken on an impulse. These so-called “lesser” frauds reflect either a less condemnable state of mind (the wanton and well-planned crime is surely more abhorrent than the spur-of-the-moment one) or a substantially reduced likelihood of harm.

To incorporate all of these concepts, Congress would have to rewrite its fraud statutes, adopting some variant of the rubrics laid out in Part III. Such an ambitious reform, unfortunately, is likely beyond our reach. That’s not to say, however, that we should give up on grading altogether.[237] As I explain below, the adoption of one or more fraud misdemeanor statutes could plausibly capture the public intuition that some frauds are indeed less deserving of condemnation than others.

A. Reconsidering the Federal Misdemeanor

A federal misdemeanor is a crime whose maximum punishment does not exceed one year of imprisonment.[238] In lay terms, it is a minor offense. Its collateral consequences are (or are often presumed to be) less harsh than that of a full-blown felony.[239]

Over the past decade, misdemeanors have attracted uniformly negative attention.[240] According to critics, the standard misdemeanor conviction can significantly interfere with an individual’s liberty and increasingly threatens the security and property of those least able to challenge its application.[241] Throughout numerous state courts, misdemeanor prosecution has become synonymous with a lack of procedural protections, notable weaknesses in judicial oversight, and significant collateral consequences for poorer defendants.[242]

The federal code contains numerous misdemeanor provisions.[243] Some are silly and serve no purpose other than proving the overcriminalization critique.[244] Others are minor offenses that happen to occur on federal property or at the expense of a federal official.[245] Surprisingly, the code contains not a single all-purpose fraud misdemeanor, notwithstanding a few statutes relating to de minimis amounts of computer or identification fraud.[246] True, a prosecutor and defense attorney could agree to dispose of a lower level fraud case as something else, but this in itself would perpetrate a fraud on the court and the general public, and the United States Attorneys’ manual rightly forbids its prosecutors from either offering or accepting such fictional guilty pleas.[247]

Because there is no such thing as misdemeanor federal fraud, prosecutors and defense attorneys encounter a familiar dilemma for lesser violations: either drop the case altogether and hope for a sufficient civil punishment or proceed as a felony case despite the fact that the crime, although serious, is not as deserving of condemnation as the ordinary fraud. From a sentencing perspective, this is not the end of the world. Numerous low-level fraud offenders have received relatively modest sentences, either in conjunction with § 2B1.1 of the Sentencing Guidelines, or in a departure from the Guidelines’ recommended sentencing range. Nevertheless, the problem remains that the typical fraud statute covers everything from an impulsive filing of a fraudulent invoice, all the way up to and including a ten-year, multi-million-dollar Ponzi scheme. As discussed in Part II, this mixing of harms and conduct creates expressive problems on both a retail and wholesale level. The public learns nothing about the particular offender from his charge (or charges), and it learns remarkably little about the offense itself because Congress has declined to differentiate it in terms that matter.

Might a misdemeanor statute address these problems? Quite possibly, but some—particularly those who believe white-collar crime has been underenforced—might argue that gradation should begin at the top. That is, we might say that the offenders most in need of a new grade are those who wantonly perpetrate large or harmful schemes. If that is true, the upshot of this Article might be to urge Congress to draft an “aggravator” statute for meticulously planned frauds, frauds that take advantage of particularly vulnerable victims, or vast systemic frauds that threaten the integrity of our economy.

Which is more necessary: an aggravator statute or a misdemeanor statute? Putting aside some unhelpful rhetoric, we might answer this question by analyzing annual fraud convictions, as collected by the Sentencing Commission. As discussed in Part II, a surprising number of fraud offenses reflect, at best, relatively modest schemes. The median sentence among those who receive a sentence of imprisonment is roughly twenty-four months’ imprisonment, depending on the year.[248] Among all those convicted in fiscal year 2017 (many of whom received no prison sentence at all), it was fifteen months’ imprisonment.[249] In 2017, 15% of the year’s theft and fraud convictions involved intended or actual loss amounts of $6,500 or less.[250] Our culture emphasizes the multi-million-dollar Ponzi schemes and the heartless executives who preside over billion-dollar frauds; the bread and butter of our federal fraud docket, however, is the garden-variety fraud.

Accordingly, insofar as grading is concerned, the most pressing problems lie at the lower end of the fraud docket. If this year is like any other, a significant number of offenders will be charged under one of several fraud statutes. Those statutes will nominally expose these offenders to high statutory maximums, but prosecutors will plead out their cases to relatively modest sentences. To anyone lacking intimate familiarity with this system, these prosecutions will either appear heartlessly aggressive (at the beginning of the case) or instead inspire the cynical conclusion that the government was simply bluffing.

These information effects extend beyond minor frauds. A regime that charges all frauds under the same statute stymies democratic oversight; it is more difficult to figure out how many of those prosecutions represent serious frauds compared with moderate ones. Moreover, it contributes to the astoundingly small number of trials that take place within the federal system.[251] The risk-averse offender charged with a misdemeanor is more likely to take his chances at trial than the same offender charged with a felony offense carrying a twenty-year maximum.

The introduction of either a misdemeanor or very low-level felony fraud statute (e.g., a statute carrying a two-year maximum sentence of imprisonment) could usher in a new version of truth in sentencing, whereby the defendant’s statutory exposure (e.g., up to two years in jail) would finally bear some semblance to his actual sentence (e.g., fourteen months’ imprisonment). It could reinstate the very ceiling and proof rights defendants are forced to forgo in an ungraded system.[252] And finally, it might actually induce a few defendants to take their chances at trial. Even those defendants charged with a standard felony fraud count might be more willing to roll the dice on the theory that their attorney could persuade the jury to acquit them of the felony and instead find them guilty of a lesser included misdemeanor.[253]

B. Content

If we are sure that some frauds are less condemnable than others and that gradation at the lower end of the scale is both necessary and valuable, how do we go about identifying this subset? Which frauds should we deem worthy of misdemeanor or low-level-felony status?

To answer these questions, it is helpful to define the legislature’s objective. A legislature might seek to expand criminal fraud law’s reach, ratify the distinctions prosecutors already informally employ in their charging decisions, or differentiate anew conduct that has been swept into a single category. Consistent with the thrust of this Article, the remainder of this subpart presumes the second and third objectives.

Where might we attempt differentiation or ratification? Reformers have long observed the distinction between the bit players who are swept up in conspiracies and those participants whose behavior serves as an integral part of the scheme.[254] Others have questioned the equivalent treatment of wholly implausible fraud schemes alongside their more plausible (and therefore more dangerous) counterparts.[255] Rather than relying on the occasional well-argued judicial sentencing opinion, we might benefit more from an explicit debate over these scenarios, such as whether they reflect diluted culpability and reduced danger and whether they are distinct enough to merit their own statutory designation.

The point here is not to single out any particular type of fraud for misdemeanor treatment so much as it is to set up a framework for debate. Moreover, the category ought to hinge on something more meaningful than crude factors, such as the dollar amount of loss or the number of convictions an offender already has. Although a loss-based misdemeanor would be the most straightforward statute to carve out, it would also be the least valuable. Calling any fraud of “X amount or less” a misdemeanor ignores the culpability and degree-of-risk factors that play such an important role elsewhere in helping us sift the worst offenses from the merely bad ones. That is not to say loss should play no role in delineating misdemeanor treatment. It would be thoroughly unhelpful, however, if the misdemeanor category were to become synonymous with some arbitrary loss amount.

C. Form

Agreement on the general content of a misdemeanor statute would leave additional questions unanswered. Among them would be whether the misdemeanor statute would serve as a true limitation on the prosecution (a “mandatory misdemeanor”); whether it would function as a partial defense, in which case it would require the defendant’s proof by a preponderance of the evidence (a “misdemeanor defense”); or whether Congress would write the misdemeanor in such a way as to allow the prosecutor to choose freely between a misdemeanor or felony charge (a “misdemeanor option”). I address each of these possibilities in turn:

1. Mandatory Misdemeanor.—The most ambitious reform would be the creation of one or more statutes whose proscriptions limited the prosecutor from charging certain less serious conduct as felony violations. To achieve this effect, Congress would have to write one statute (the misdemeanor) and revise its relevant felony statutes to exclude the behavior now defined as a misdemeanor. It would thus be a mandatory misdemeanor because the prosecutor would have no choice but to charge the conduct as the lesser crime.

The mandatory misdemeanor powerfully subdivides an offense and limits the felony’s scope. Moreover, it punishes prosecutorial aggression. If the prosecutor overcharges a defendant with a felony offense, the defendant can request that the court advise the jury of the misdemeanor charge and its ability to convict the defendant of the lesser included offense.[256]

2. The Misdemeanor Defense.—Instead of a mandatory carveout, Congress might instead create a series of affirmative defenses that reduce felony charges to misdemeanors, provided the defendant proves their existence by a preponderance of the evidence. Such a scheme differentiates fraud, but it imposes on the defendant the burden of production and proof.[257]

Imagine, for example, a defendant has been charged in a billion-dollar fraud scheme. The fraud, however, is so silly that the likelihood of convincing anyone other than an undercover officer to hand over his money is slim to none. Under a mandatory scheme described in the preceding section, a prosecutor would have to prove beyond a reasonable doubt the scheme’s plausibility. Under an affirmative-defense regime, the prosecutor would have to prove only the existence of the scheme to defraud, leaving the defendant to prove (presumably by a preponderance) the scheme’s implausibility.

From the legislator’s perspective, the affirmative defense is more desirable. It subdivides fraud cases but is less fraught with risk. It requires only the drafting of language necessary to define the affirmative defense but leaves felony fraud statutes intact. On the other hand, from the defendant’s perspective, the affirmative misdemeanor defense is far less valuable. It places a burden of production (and likely proof) on the defendant, thereby favoring defendants with the resources and risk appetite to claim it. Most importantly, if we are concerned about the law’s expressive content, the misdemeanor defense is decidedly less powerful than a mandatory statute that explicitly deems certain conduct less deserving of condemnation.

3. A Prosecutor’s Misdemeanor Option.—Finally, Congress might enact a set of white-collar-misdemeanor statutes that define fraud with language identical to the felony statutes. Current felony fraud statutes would remain as is, and Congress would simply add language permitting but not requiring prosecutors to charge certain behavior as a misdemeanor.

This discretionary approach is controversial but not unprecedented. In the state of New York, coercion in the first and second degrees are identical and permit the prosecutor to choose between a misdemeanor and felony charge.[258] In defense of this approach, the New York Court of Appeals has reasoned that the second-degree offense “is apparently a ‘safety-valve’ feature included in the event an unusual factual situation should develop where the method of coercion [meets the definition of the felony], but . . . lacks the heinous quality the Legislature associated with such threats.”[259] Subsequent New York decisions have affirmed this concept, although other states have rejected similar statutes on multiple grounds.[260]

Whether this type of scheme would pass muster in federal courts is somewhat beside the point. The misdemeanor option is a far cry from the grading reforms described in Part III of this Article. It neither subdivides criminal offenses nor vests defendants with significant protections. For that reason, the remainder of this discussion presumes that Congress would adopt either a separate misdemeanor or an affirmative defense.

D. Potential Objections

Framed either as a separate offense carved out of a felony, or even as an affirmative defense, a well-drafted statute could generate the various expressive and adjudicative benefits described in the preceding section. Nevertheless, one could easily conjure a series of objections to such an approach. I address several of them here:

1. The Do-Nothing-at-All Scenario.—Some might conclude that misdemeanor fraud statutes would do little to change federal law’s landscape. Fearful of appearing too soft on crime, Congress would define misdemeanor statutes so narrowly as to make them meaningless, tie them to some arbitrary loss amount such as $10,000, or make them purely discretionary. Prosecutors would continue to charge all frauds as felonies, relying on misdemeanors solely as some sort of fallback position lest some previously charged case blow up on the eve of trial.

These critics are of course right. The misdemeanor remedy would fail to remedy fraud’s flatness if Congress in fact drafted a lesser fraud statute so cynically. That being said, mass incarceration has, over the past decade or so, become the interest of both the political left and right.[261] Accordingly, the belief that Congress might fashion a meaningful misdemeanor statute is neither fanciful nor naïve.

2. The Increased-Punishment Scenario.—Some criminal defense attorneys might worry that prosecutors would embrace the misdemeanor as a means of expanding fraud law’s reach, either to prosecute defendants engaging in marginally wrongful acts, or to pursue individuals who were not factually guilty but lacked the appetite to challenge a prosecution. This is the phenomenon Paul Crane has identified in regard to misdemeanor street crimes in state courts, and it is one that ought to give us pause, particularly if it affects the poorest and least powerful of those charged with committing fraud offenses.[262]

One reason to be less worried about the pathologies that arise in regard to state court misdemeanors is that the federal system truly differs from state systems. It convicts far fewer defendants per year and allocates greater resources per defendant.[263] Nevertheless, one way to definitively prevent this outcome—and prevent the strategic charging practice Crane identifies—is to ensure that federal felons and misdemeanants receive near-identical procedural rights in federal court.[264]

3. The Undermine-Deterrence Scenario.—Finally, law-and-order advocates might worry that a misdemeanor statute would tack too far in the opposite direction. Intended as an incremental remedy, the misdemeanor might unwittingly undermine deterrence. Cases previously classified as felonies might go uninvestigated and unprosecuted, as FBI agents and prosecutors refuse to expend resources on misdemeanor prosecutions.[265] This is not simply a matter of laziness or snobbery. By downgrading the fraud offense’s label, Congress would in fact be communicating to the law-enforcement community the public’s view that certain offenses are less deserving of concern. In and of itself, the reaction is rational and hardly blameworthy. When a new label causes too dramatic a fall in enforcement activity, however, it creates a problematic shortfall in deterrence.

The underdeterrence scenario is a problem, but only if prosecutors and FBI agents can predict in advance which investigations will produce felonies and which will produce only misdemeanors. Moreover, if supervisors were to detect a shortfall, they could encourage investigators and prosecutors to revive misdemeanor cases by creating a series of incentives (e.g., the creation of a specific misdemeanor-fraud budget).

Apart from invoking doomsday scenarios, some might object that the notion of a misdemeanor is inconsistent with any federal prosecution. That is, after all, the intuition that fuels the well-known maxim, “Don’t make a federal case out of it.”[266] Whatever the argument’s normative strength, it ignores federal criminal law’s reality. The Sentencing Guidelines are rife with provisions intended for so-called low-level federal offenders.[267] According to the U.S. Sentencing Commission, a noticeably large portion of those charged with federal fraud are responsible for relatively low loss amounts.[268] For this group, it is far from clear that we should eliminate federal liability. Regardless of the relative importance or wrongfulness associated with any given offense, the federal government may be the best and only venue for prosecuting certain crimes, particularly those that are intimately intertwined with federal programs, pertain to markets and industries closely monitored by federal enforcement authorities, or are otherwise of the type that would be ignored or underenforced by local authorities.

The bottom line is this: We are not about to eliminate federal fraud prosecutions in modest or small-scale cases. Accordingly, we might as well start thinking about how best to portray them accurately. Misdemeanor statutes may represent a second-best option, but it is an option we would be foolish to ignore.

V. Conclusion

Toward the end of her spirited dissent in Yates, Justice Kagan alighted upon the “real” reason five of her fellow justices had gone out of their way to declare the word “fish” outside the ambit of a statute that ostensibly punished the destruction of “any” “tangible object.” The obstruction statute wasn’t vague or ambiguous so much as it was overly broad, subjecting its most minor offenders to the same high statutory maximum reserved for its worst offenses. Kagan didn’t disagree with her colleagues’ assessment but couldn’t see how it could justify the Court’s rewrite of an otherwise clear legislative term. As Justice Kagan pointed out, the offending statute, while “too broad and undifferentiated,” was “unfortunately not an outlier, but an emblem of a deeper pathology in the federal criminal code.”[269]

When referencing the federal code’s “deeper pathology,” Justice Kagan could easily have been speaking of the federal code’s fraud statutes. They certainly operate no differently from § 1519. With the federal code’s fraud statutes firmly in mind, this Article has fleshed out the problem Justice Kagan underscored in her Yates dissent. Scholars have hotly debated whether white-collar crime has been overcriminalized or underenforced (or both). These critiques, dependent as they are on depictions of prosecutorial overreach in some cases and prosecutorial capture in others, de-emphasize the very code that makes white-collar crime’s enforcement possible in the first place.

Graded codes are admittedly imperfect. At their best, however, they reflect a democratically elected body’s view of which crimes are worse than others and which ones are slightly less deserving of condemnation. They bring us closer to the democratic ideal encapsulated by criminal law’s legality principle. Finally, they highlight intuitive distinctions in related crimes. “Murder in the first degree” does more than signal a predicted punishment or aggressive prosecutorial stance; it announces society’s considered viewpoint as to which type of killings merit the state’s harshest punishment, and it tells us something about the person who has been charged and convicted of such an offense. In the same manner, the best graded statutes do far more than sort crimes; they also educate and affirm communal values. Despite their many drawbacks, our state criminal codes partially, and even admirably, come close to meeting these basic functions. We should expect no less of the federal statutes that purport to define and punish white-collar crime.

  1. .See, e.g., Todd Haugh, Overcriminalization’s New Harm Paradigm, 68 Vand. L. Rev. 1191, 1223–29 (2015) (suggesting that overcriminalization results in “inconsistent enforcement and overly harsh adjudication of criminal laws,” thus eroding the legitimacy of criminal law and ultimately causing more criminal conduct); Julie O’Sullivan, The Federal Criminal “Code” Is a Disgrace: Obstruction Statutes as Case Study, 96 J. Crim. L. & Criminology 643, 656–59 (2006) (critiquing the frequent ambiguity of criminal mens rea requirements and the proliferation of strict liability offenses); Ellen S. Podgor, Overcriminalization: New Approaches to a Growing Problem, 102 J. Crim. L. & Criminology 529, 530 (2012) (“You end up adding more laws to the existing ones, without discarding any in the process.”); Paul H. Robinson et al., The Disutility of Injustice, 85 N.Y.U. L. Rev. 1940, 1960–61 (2010) (remarking on the federal criminal code’s “‘unprecedented expansion’ in recent years,” particularly in regard to regulatory offenses). Scholars have used the term as a shorthand for a number of concepts. See, e.g., Stuart P. Green, Why It’s a Crime to Tear the Tag Off a Mattress: Overcriminalization and the Moral Content of Regulatory Offenses, 46 Emory L.J. 1533, 1545–46 (1997) (citing two forms of overcriminalization: (1) when “sanctions are over-authorized by Congress and [state] legislatures” and (2) when authorized sanctions “are over-applied by prosecutors and courts”). On the political factors that drive criminal law’s expansion, see, e.g., Paul J. Larkin, Jr., Public Choice Theory and Overcriminalization, 36 Harv. J.L. & Pub. Pol’y 715, 735–37 (2013) (explaining legislators’ incentives to enact new statutes rather than reduce crime rates through other approaches); William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 509 (2001) (sketching the political dynamic that induces legislators to generate new substantive criminal statutes). For the most recent argument that federal law in particular has been overcriminalized, see Stephen F. Smith, Overfederalization, in 1 Reforming Criminal Justice: Introduction and Criminalization 39, 40 (Erik Luna ed., 2017) (observing that the federal government’s expansion of criminal statutes “raises serious problems of its own”). But see Susan R. Klein & Ingrid B. Grobey, Debunking Claims of Over-Federalization of Criminal Law, 62 Emory L.J. 1, 5 (2012) (rejecting claims that federal caseloads have changed in response to additional federal statutes).
  2. .On white-collar crime’s federal nature, see Daniel C. Richman & William J. Stuntz, Al Capone’s Revenge: An Essay on the Political Economy of Pretextual Prosecution, 105 Colum. L. Rev. 583, 601–02 (2005) (“[H]igh-end white-collar crime is (with a few rare exceptions) a federal preserve . . . .”). Edwin Sutherland first coined the phrase in the 1930s to denigrate high-status persons engaging in occupation-based crimes; the phrase has since transformed into a more general term denoting nonviolent crimes “characterized by deceit, concealment, or violation of trust.” Miriam H. Baer, Confronting the Two Faces of Corporate Fraud, 66 Fla. L. Rev. 87, 100 (2014) (quoting Fed. Bureau of Investigation, U.S. Dep’t of Justice, White Collar Crime: A Report to the Public 3 (1989)). For more on the term’s historical usage and its different meanings, see Gerald Cliff & Christian Desilets, White Collar Crime: What It Is and Where It’s Going, 28 Notre Dame J.L. Ethics & Pub. Pol’y 481, 482–87 (2014) (acknowledging lack of consensus within disciplines).
  3. .See generally Brandon L. Garrett, Too Big to Jail: How Prosecutors Compromise with Corporations (2014) (critiquing kid-gloves treatment of corporations and their executives); see also Todd Haugh, The Most Senior Wall Street Official: Evaluating the State of Financial Crisis Prosecutions, 9 Va. L. & Bus. Rev. 153, 167 (2015) (observing the dearth of prosecutions of corporate executives in the wake of the financial crisis). On the public’s perception of underenforcement, see Peter J. Henning, Making Sure “The Buck Stops Here”: Barring Executives for Corporate Violations, 2012 U. Chi. Legal F. 91, 100 (“In the area of corporate crime, the perception of rampant lawlessness in the upper strata of society certainly appears to have firmly taken hold.”); Joseph W. Yockey, Beyond Yates: From Engagement to Accountability in Corporate Crime, 12 N.Y.U. J.L. & Bus. 407, 409 (2016) (observing that “fifty-three percent of respondents believe that not enough was done to prosecute bankers” responsible for the subprime-mortgage crisis).
  4. .“The classic white collar crimes—bank fraud, mail fraud, and wire fraud—are not predicated on lower level crimes with a lesser degree of culpability or extenuating circumstances. The individual is either guilty or not guilty of the designated offense.” Ellen S. Podgor, The Challenge of White Collar Sentencing, 97 J. Crim. L. & Criminology 731, 757 (2007); see also Miriam H. Baer, Unsophisticated Sentencing, 61 Wayne L. Rev. 61, 79–80 (2015) (“Federal crimes do not parse mental states or offense conduct the way state statutes do.”); Frank O. Bowman III, Sentencing High-Loss Corporate Insider Frauds After Booker, 20 Fed. Sent’g Rep. 167, 171 (2008) (“A peculiar feature of American property crimes generally and federal fraud crimes in particular is that . . . statutory law does not rank the severity of offenses according to differences in the mental states of defendants.”).
  5. .By referring to federal fraud law as flat, this Article joins other scholars who have used this and similar metaphors to reflect a criminal code’s lack of gradation. See, e.g., Andrew von Hirsch et al., Criminal Deterrence and Sentence Severity: An Analysis of Recent Research 41–43 (1999) (analyzing sentencing “gradients” in England and the United States and their effect on deterrence); Jonathan Simon, How Should We Punish Murder?, 94 Marq. L. Rev. 1241, 1249 (2011) (arguing that flattening homicide law creates a risk of excessive punishment); Ronald F. Wright & Rodney L. Engen, Charge Movement and Theories of Prosecutors, 91 Marq. L. Rev. 9, 10 (2007) [hereinafter Wright & Engen, Charge Movement] (explaining connection between a criminal code’s “depth” and plea bargaining); Ronald F. Wright & Rodney L. Engen, The Effects of Depth and Distance in a Criminal Code on Charging, Sentencing, and Prosecutor Power, 84 N.C. L. Rev. 1935, 1939–40 (2006) [hereinafter Wright & Engen, Depth and Distance] (introducing notions of “depth” and “distance” in criminal codes).
  6. .A minor exception to this rule can be found in the part of the U.S. Code that criminalizes computer, credit card, and identification fraud and aggravated identity theft. See 18 U.S.C.§§ 1028–1030 (2012). These statutes set forth a grab bag of additional punishments relating either to the defendant’s criminal history (as in computer fraud) or to the underlying crime (e.g., whether the fraud occurred in connection with a narcotics-trafficking scheme). Id. They do not, however, divide these crimes into salient degrees; as a result, the Department of Justice has generated a charging manual to aid its own prosecutors. See Comput. Crime & Intellectual Prop. Section, Dep’t of Justice, Prosecuting Computer Crimes 2–3 (2010), https:/‌/‌‌sites/‌default/‌files/‌criminal-ccips/‌legacy/‌2015/‌01/‌14/‌ccmanual.pdf [https:/‌/‌‌547Z-EACA] [hereinafter Prosecuting Computer Crimes] (setting forth a table of penalties).
  7. .As of 2007, thirty-six states employed a formal grading system of degrees to delineate distinctions between and among felonies and misdemeanors. Model Penal Code: Sentencing § 6.01 cmt. b (Am. Law Inst., Proposed Final Draft 2017) (observing that all but fourteen states and the federal system employed the standard degree methodology). The American Law Institute’s 2017 Sentencing draft contends that ungraded schemes frequently devolve into “a patchwork of authorized punishments, with no clear rationale for the assignment of penalties to specific crimes when compared one to another.” Id. The federal computer and identity theft statutes, see supra note 6, aptly instantiate this critique.
  8. .18 U.S.C. § 1346 (2012).
  9. .For examples of federal fraud statutes, see 15 U.S.C. § 78j (2012) (securities fraud); 18 U.S.C. § 1341 (2012) (mail fraud); 18 U.S.C. § 1343 (2012) (wire fraud); 18 U.S.C. § 1344 (2012) (bank fraud); 18 U.S.C. § 1347 (2012) (health-care fraud); 18 U.S.C. § 1348 (2012) (securities fraud); 17 C.F.R. § 240.10b-5 (2018) (securities fraud).
  10. .See Michael Serota, Proportional Mens Rea and the Future of Criminal Code Reform, 52 Wake Forest L. Rev. 1201, 1210 (2017) (arguing that criminal law’s legality principle “compel[s] the direct legislative resolution of the most significant issues of penal policy by statute”).
  11. .See James Chalmers & Fiona Leverick, Fair Labelling in Criminal Law, 71 Mod. L. Rev. 217, 222 (2008) (“[A] description of conduct as manslaughter draws explanatory value from the fact that it has in some way been differentiated from murder.”). The concept has been explored by British and European criminologists at length. See, e.g., Hilmi Zawati, Fair Labelling and the Dilemma of Prosecuting Gender-Based Crimes at the International Criminal Tribunals 31–32 (2014) (noting that one justification for the principle of fair labeling is to ensure that defendants are not unfairly stigmatized). For an exploration of the concept as it relates to property-law crimes such as theft, see Stuart P. Green, Thirteen Ways to Steal a Bicycle: Theft Law in the Information Age 18, 52–54 (2012) (arguing that the Anglo-American shift toward consolidating theft crimes undermined the principle of fair labeling).
  12. .This was one of the animating philosophies of the Model Penal Code. “[T]he model code was meant to assist democratic legislators in fulfilling their responsibility as representatives of the people in making the value choices necessary to enact a code.” David Wolitz, Herbert Wechsler, Legal Process, and the Jurisprudential Roots of the Model Penal Code, 51 Tulsa L. Rev. 633, 660 n.193 (2016) (explaining Professor Wechsler’s desire for a prudent yet democratic criminal code).
  13. .See discussion infra Part II.
  14. .See Frank O. Bowman, III, Mr. Madison Meets a Time Machine: The Political Science of Federal Sentencing Reform, 58 Stan. L. Rev. 235, 245 (2005) (recounting the “admirable distribution of sentencing authority” envisioned by the Sentencing Reform Act of 1984).
  15. .Adam Kolber’s work foreshadows this defense, although Kolber is more concerned with the abruptness of increases in punishment rather than with the presence or absence of statutory subdivisions in offenses. See Adam J. Kolber, The Bumpiness of Criminal Law, 67 Ala. L. Rev. 855, 856 (2016) (asserting a preference for a smooth relationship between offense and punishment in which a gradual increase in the seriousness of an offense is accompanied by a proportionate increase in the severity of punishment).
  16. .A number of states whose criminal codes are graded also employ a guidelines system to guide trial judges in their sentencing decisions. See Michael L. Perlin & Alison J. Lynch, “In the Wasteland of Your Mind”: Criminology, Scientific Discoveries and the Criminal Process, 4 Va. J. Crim. L. 304, 310 & n.10 (2016) (citing sources indicating that over half the states employ some type of guideline process for sentencing); see also supra note 7 and accompanying text.
  17. .On the Guidelines’ complexity, see Kate Stith & José A. Cabranes, Judging Under the Federal Sentencing Guidelines, 91 Nw. U. L. Rev. 1247, 1266 (1997) (describing various technical issues the Guidelines have spawned).
  18. .See generally Ronald F. Wright, The United States Sentencing Commission as an Administrative Agency, 4 Fed. Sent. Rep. 134 (1991) (summarizing implications of “[t]he most remarkable aspect of the 1984 Sentencing Reform Act”—the decision to entrust responsibility for federal sentencing to an administrative agency); see also Mistretta v. United States, 488 U.S. 361, 396 (1989) (referring to the Sentencing Commission as an “independent agency”). Because of its unique structural characteristics, jurists continue to grapple with its status. See, e.g., United States v. Havis, No. 17-5772, 2018 WL 5117187, at *2 (6th Cir. Oct. 22, 2018) (describing the Commission as “a sort of hybrid body that does not fit squarely within any of the three branches of government”).
  19. .See, e.g., Pepper v. United States, 562 U.S. 476, 487–88 (2011) (articulating the principle that a sentence ought to “fit the offender and not merely the crime”) (quoting Williams v. New York, 337 U.S. 241, 247 (1949)). For a normative defense of the principle, see Richard A. Bierschbach & Stephanos Bibas, Constitutionally Tailoring Punishment, 112 Mich. L. Rev. 397, 429 (2013) (arguing that a fair criminal justice system must individualize an offender’s punishment on the basis of his “own blameworthiness and dangerousness, the harm he has done to a victim, his efforts to make amends and apologize, and so on”).
  20. .See discussion infra subpart III(A).
  21. .Serota, supra note 10, at 1211 (arguing that a legislatively enacted grading scheme “constitutes a primary means of ensuring that the overall distribution of punishment meted out by a criminal justice system reflects the community’s norms”). Serota links the legislature’s obligation to ensure “proportional” punishment with its obligation to grade offenses according to offenders’ differing states of mind. Id. at 1202, 1210–13. This Article rejects this claim insofar as it unduly constrains the legislature’s discretion to grade offenses. Although mens rea provides one basis upon which to subdivide offenses, the severity of harm or risk provide equally plausible bases for gradation. See Paul H. Robinson & Jane A. Grall, Element Analysis in Defining Criminal Liability: The Model Penal Code and Beyond, 35 Stan. L. Rev. 681, 683 n.6 (1983) (“Both the degree of harm and the level of culpability are relevant to the grade of an offense.”); discussion infra Part III.
  22. .See Stuart Green, Is there Too Much Criminal Law?, 6 Ohio St. J. Crim. L. 737, 741 (2009) (book review) (distinguishing legislatures, who “are obliged to consider, prospectively, the extent to which certain kinds of harm and wrong are associated with certain types of conduct” from the legal actors who consider specific offenders and their factual situations); see also Dan Markel, Are Shaming Punishments Beautifully Retributive? Retributivism and the Implications for the Alternative Sanctions Debate, 54 Vand. L. Rev. 2157, 2207–08 (2001) (describing the ex ante approach to retributive punishment, which considers the “gravity of the wrongdoing as a general matter”).
  23. .See John Calvin Jeffries, Jr., Legality, Vagueness, and the Construction of Penal Statutes, 71 Va. L. Rev. 189, 190 (1985) (remarking that the legality principle “stands for the desirability in principle of advance legislative specification of criminal misconduct”); Paul H. Robinson, Fair Notice and Fair Adjudication: Two Kinds of Legality, 154 U. Pa. L. Rev. 335, 336 (2005) (noting that in its modern form, the legality principle means “that criminal liability and punishment can be based only upon a prior legislative enactment of a prohibition that is expressed with adequate precision and clarity”). 
  24. .See discussion infra subpart I(A).
  25. .There is no conceptual reason to require these statutes to be misdemeanors. A low-level felony with a maximum of two years’ imprisonment would accomplish the same result.
  26. .See infra notes 238–44 and accompanying text.
  27. .On the differences between federal and state criminal justice systems, see Alexandra Natapoff, The Penal Pyramid, in The New Criminal Justice Thinking 71, 72 (Sharon Dolovich & Alexandra Natapoff eds., 2017) (conceptualizing the criminal justice system as a pyramid and arguing that its top “is the world of federal offenses [and] serious cases . . . [in which] rules dominate” and its bottom consists of state and local systems where “offenses are petty and caseloads number in the thousands”).
  28. .See Richard Bierschbach & Stephanos Bibas, What’s Wrong with Sentencing Equality?, 102 Va. L. Rev. 1447, 1477–79 (2016) (distinguishing “centripetal” and “centrifugal” elements of criminal punishment).
  29. .Gerard E. Lynch, Towards a Model Penal Code, Second (Federal?): The Challenge of the Special Part, 2 Buff. Crim. L. Rev. 297, 302 (1998) (observing that for most criminal defendants, the abstract “line between guilty and not guilty . . . matters far less than where the case will be placed on the continuum of possible punishments”).
  30. .See Bierschbach & Bibas, supra note 28, at 1450 (“The big fish deserve more punishment than the medium and small fry, even if they all violated the same statute.”).
  31. .The argument that offenses should be subdivided into degrees is distinct from the one that punishments should increase gradually in relation to culpability. Ronald Wright and Rodney Engen refer to the former as “depth” and the latter as “distance.” Wright & Engen, Depth and Distance, supra note 5, at 1954. The present Article argues for a “deeper” federal criminal code but takes no position on how gradually punishments should increase. For an argument that criminal justice systems should avoid precipitous jumps, see Kolber, supra note 15, at 856 (reasoning that “[w]hile the law must draw difficult lines, the lines need not have such dramatic effects”).
  32. .For evidence that societies instinctively distinguish crimes, see, e.g., Paul H. Robinson, The Ongoing Revolution in Punishment Theory: Doing Justice as Controlling Crime, 42 Ariz. St. L.J. 1089, 1107 (2010–2011) (observing a broad consensus “at least with regard to the core of wrongdoing”). This claim is not without controversy. See Michael T. Cahill, Retributive Justice in the Real World, 85 Wash. U. L. Rev. 815, 819 n.12 (2007) (citing debate over empirical desert claims).
  33. .See Paul H. Robinson et al., The Modern Irrationalities of American Criminal Codes: An Empirical Study of Offense Grading, 100 J. Crim. L. & Criminology 709, 718–28 (2010) (citing Pennsylvania’s criminal code as one whose grades conflict with its citizens’ views of relative desert). Whether two crimes are similar enough to fall within the same family for purposes of classification and grading is beyond the scope of this project.
  34. .Concededly, many states employ both mechanisms, or rely on a parole system to differentiate offenders at a later point in time. There exists, however, a difference between a regime that relies solely on its sentencing stage to sort offenders and one that includes statutory grading and additional sorting mechanisms.
  35. .McGautha v. California, 402 U.S. 183, 198 (1971) (tracing society’s “rebellion against the common law rule imposing a mandatory death sentence on all convicted murderers”); Paul J. Larkin, Jr., The Demise of Capital Clemency, 73 Wash. & Lee L. Rev. 1295, 1301 (2016); see also Model Penal Code § 210.2 cmt. 2 (Am. Law Inst. 1980) (noting the most significant reform prior to the Model Penal Code was Pennsylvania’s division of murder into degrees); Guyora Binder, The Origins of American Felony Murder Rules, 57 Stan. L. Rev. 59, 119 (2004) (detailing Pennsylvania’s legislative reform of state homicide laws); Simon, supra note 5, at 1263–64 (same). On the application of capital punishment to crimes other than homicide, see Lauren Ouziel, Beyond Law and Fact: Jury Evaluation of Law Enforcement, 92 Notre Dame L. Rev. 691, 713 n.84 (2016) (noting that “[d]eath was the sentence for many felonies”); Kathleen “Cookie” Ridolfi & Seth Gordon, Gubernatorial Clemency Powers: Justice or Mercy?, Crim. Just., Fall 2009, at 26, 29 (“At the time of the Declaration of Independence in 1776, more than 200 crimes carried mandatory death sentences in England.”).
  36. .Larkin, supra note 35, at 1301 (“Juries . . . disliked seeing a mandatory death sentence in cases where the offender did not deserve to die and would refuse to convict a defendant if doing so would send him to the gallows.”); see also Ouziel, supra note 35, at 714 (“Second degree murder was developed . . . largely as a concession to capital-averse juries.”).
  37. .Ouziel, supra note 35, at 713–14 (noting the effort, first in Pennsylvania and then in other states, “to enact criminal laws that offered more nuanced degrees of guilt and punishment”).
  38. .See Sam Kamin & Justin Marceau, Vicarious Aggravators, 65 Fla. L. Rev. 769, 776 (2013) (“[T]oday, nearly every state uses degrees of murder as the first slice at determining which murderers should live and which should die.”).
  39. .See, e.g., 18 U.S.C. §§ 922–931 (2012) (setting forth an array of offenses for possession or use of firearms in different circumstances); 21 U.S.C. §§ 841865 (2012) (describing the offenses and penalties for trafficking in different amounts and types of controlled substances).
  40. .An arguable exception to this is the U.S. Code’s distinction between domestic bribery and the giving and receiving of gratuities. See 18 U.S.C. § 201(b)–(c) (2012) (setting out a fifteen-year maximum penalty for the former but only a two-year maximum penalty for the latter); see also Stuart P. Green, Official and Commercial Bribery: Should They Be Distinguished?, in Modern Bribery Law: Comparative Perspectives 39, 40–42 (Jeremy Horder & Peter Allridge eds., 2013) (describing and distinguishing major bribery statutes in the federal code). Even here, the federal code does not distinguish the offenses by degrees.
  41. .Kate Stith & Steve Y. Koh, The Politics of Sentencing Reform: The Legislative History of the Federal Sentencing Guidelines, 28 Wake Forest L. Rev. 223, 226 n.7 (1993).
  42. .David Schwartz describes an array of crimes that were enacted in the 1790 Crimes Act, many of which related to the federal government’s enumerated powers. David S. Schwartz, A Question Perpetually Arising: Implied Powers, Capable Federalism, and the Limits of Enumerationism, 59 Ariz. L. Rev. 573, 643 (2017) (describing the Act’s criminalization of “counterfeiting, treason, piracy, felonies on the high seas,” and violent felonies occurring within the federal government’s exclusive jurisdiction).
  43. .Smith, supra note 1, at 40–41 (describing federal criminal law in the early period following the nation’s founding when prosecution was limited to “offenses involving criminal activity that either occurred outside of state jurisdiction or uniquely threatened the operations, property, or personnel of the federal government”). Although there is little legislative history regarding the 1872 act that produced the mail-fraud statute, the criminalization of mail fraud was consistent with “the expansion of federal authority that came about in the wake of the Civil War” and reflected a concern with the “growth in large-scale swindles and frauds” made possible by a more sophisticated economy. Mark Zingale, Fashioning a Victim Standard in Mail and Wire Fraud: Ordinarily Prudent Person or Monumentally Credulous Gull?, 99 Colum. L. Rev. 795, 802 (1999).
  44. .See Screws v. United States, 325 U.S. 91, 152 (1945) (Murphy, J., dissenting) (citing United States v. Gooding, 25 U.S. (12 Wheat.) 460 (1827); United States v. Hudson, 11 U.S. (7 Cranch) 32 (1812)) (“It was settled early in our history that prosecutions in the federal courts could not be founded on any undefined body of so-called common law.”); see also Hudson, 11 U.S. (7 Cranch) at 34 (holding that Congress “must first make an act a crime [and] affix a punishment to it”). For an account of the common law’s demise in regard to federal crimes, see Michael J. Zydney Mannheimer, On Proportionality and Federalism: A Response to Professor Stinneford, 97 Va. L. Rev. In Brief 51, 55–59 (2011) (describing early debates which culminated in the Supreme Court’s definitive rejection of a federal common law in United States v. Hudson). Despite the Court’s rejection of a body of federal criminal common law, judges have continued to engage in interstitial lawmaking. See Dan M. Kahan, Lenity and Federal Common Law Crimes, 1994 Sup. Ct. Rev. 345, 347 (“[F]ederal criminal law, no less than other statutory domains, is dominated by judge-made law crafted to fill the interstices of open-textured statutory provisions.”).
  45. .Daniel J. Richman et al., Defining Federal Crimes 25 (2014) (“Because the field [of federal criminal law] is almost entirely optional, Congress is free to use criminal law in innovative ways.”).
  46. .“From the beginning of the Republic, federal judges were entrusted with wide sentencing discretion . . . .” Stith & Koh, supra note 41, at 225 & n.6 (citing extensive pre-Guidelines sentencing discretion which federal judges enjoyed under federal criminal statutes); see also Mistretta v. United States, 488 U.S. 361, 364 (1989) (“Congress early abandoned fixed-sentence rigidity . . . and put in place a system of ranges within which the [judge] could choose the precise punishment.”).
  47. .Peter J. Henning, The Changing Atmospherics of Corporate Crime Sentencing in the Post-Sarbanes-Oxley Act Era, 3 J. Bus. & Tech. L. 243, 246 (2008) (describing criminal-law changes brought about by the Sarbanes-Oxley Act of 2002). The Act similarly increased the statutory maximum sentence for wire fraud and for conspiracy to commit wire or mail fraud. Id.
  48. .See Daniel Richman, Federal White Collar Sentencing in the United States: A Work in Progress, 76 Law & Contemp. Probs., no.1, 2013, at 53, 55 (describing white-collar sentencing practices in the 1970s). “Congress was especially concerned that prior to the Sentencing Guidelines, ‘[m]ajor white collar criminals often [were] sentenced to small fines and little or no imprisonment.’” United States v. Martin, 455 F.3d 1227, 1240 (11th Cir. 2006) (quoting S. Rep. No. 98-225, at 76 (1983), as reprinted in 1984 U.S.C.C.A.N. 3182, 3259). In recent years, some have questioned the conventional wisdom that white-collar offenders routinely avoided jail sentences in the 1970s. See Sam W. Buell, Is the White Collar Offender Privileged?, 63 Duke L.J. 823, 833 n.22 (2014) (citing a study from the 1970s indicating that more than 40% of offenders received some sentence of incarceration); see also Alan M. Dershowitz, Background Paper, in Fair and Certain Punishment 67, 105–06 (Report of the Twentieth Century Fund Task Force on Criminal Sentencing, 1976) (citing 1972 sentencing study conducted by the United States Attorney’s Office for the Southern District of New York in which only 36% of white-collar offenders received a prison sentence).
  49. .The distinction between criminal and civil liability continues to preoccupy academics. See, e.g., Samuel W. Buell, Capital Offenses: Business Crime and Punishment in America’s Corporate Age 40 (2016) (discussing the fine lines between “ordinary commerce” and “criminal wrongdoing”); Gerard E. Lynch, The Role of Criminal Law in Policing in Corporate Misconduct, 60 Law & Contemp. Probs., no. 3, 1997, at 23–27 (exploring what it means to argue whether a matter is genuinely a “criminal case”).
  50. .See Dershowitz, supra note 48, at 83–87 (describing early colonial-era codes that eventually gave way to post-Revolution reforms in criminal punishment). For early examples of gradation in non-homicide statutes, see Green, supra note 11, at 11–12 (noting that even relatively early state codes assigned different punishments to different variants of theft).
  51. .The reliance on state legislatures to subdivide state statutes was not foreordained. England’s common law system relied on its judiciary to “spare defendants from capital punishment and sentence them to some lesser form of punishment” by relying on the “benefit of clergy” doctrine. Kahan, supra note 44, at 358. The doctrine, however, slowly disappeared in the years leading up to and following the American Revolution. Id. at 358–59 (observing that the benefit of clergy doctrine “was virtually extinct in the states and was completely unknown to federal law”). For more on the doctrine and its use in previous centuries to temper the rule that all felonies were subject to capital punishment, see Williams v. United States, 569 A.2d 97, 101 n.9 (D.C. 1989) (describing the doctrine’s use from the fourteenth through seventeenth centuries); Shon Hopwood, Clarity in Criminal Law, 54 Am. Crim. L. Rev. 695, 714 (2017) (charting the doctrine’s rise and ultimate demise); see also Simon, supra note 5, at 1257–58 (discussing the benefit of clergy doctrine and use of royal pardons to avoid death sentences for early homicide prosecutions in England).
  52. .Ouziel, supra note 35, at 714 (charting the move “from pre-existing, natural law to be ‘found’ by courts, to positive, legislatively-defined law”). In some cases, the move towards a body of statutory law resulted in less gradation, as related crimes were streamlined into a single offense. See, e.g., Green, supra note 11, at 8–9 (analyzing series of reforms that consolidated disparate common law property offenses into a “single ‘unitary’ offense of theft”).
  53. .Stuntz, supra note 1, at 514–15 (“[The] expansion of federal criminal law generally focused on vice in the first third of the twentieth century, regulatory crimes and racketeering in the second third, and violence and drugs (plus yet more white-collar offenses) in the last third.” (citations omitted)).
  54. .For more on federal criminal law’s growth and consequent overlap with state law, see Daniel C. Richman, The Changing Boundaries Between Federal and Local Law Enforcement, in 2 Criminal Justice 2000: Boundary Changes in Criminal Justice Organizations 81, 83–84 (Charles M. Friel ed., 2000).
  55. .Stuntz, supra note 1, at 531–32 (describing “symbolic stands” legislators take by enacting redundant statutes in response to particularly notable or upsetting crimes).
  56. .See, e.g., Rachel E. Barkow, Clemency and Presidential Administration of Criminal Law, 90 N.Y.U. L. Rev. 802, 822 & nn.119–20 (2015) (citing commonly voiced criticisms of the federal criminal code’s bloat); Smith, supra note 1, at 41 (“[T]he loose collection of statutes known as ‘federal criminal law’ is sprawling and virtually limitless in its reach . . . .”); Stuntz, supra note 1, at 515 (observing that “anyone who studies contemporary state or federal criminal codes is likely to be struck by their scope, by the sheer amount of conduct they render punishable”); see also, e.g., O’Sullivan, supra note 1, at 643 (declaring the “code” a “haphazard grab-bag of statutes”).
  57. .Stephen Smith is a partial exception in that he cites disproportionate punishment as one of overcriminalization’s fruits. Stephen F. Smith, Overcoming Overcriminalization, 102 J. Crim. L. & Criminology 537, 540 (2012) (highlighting the ways in which overly broad codes can give an “unwarranted and perhaps unintended sweep to criminal laws and threaten disproportionately severe punishment”). Smith’s proposed remedy, however, lies primarily with judicial interpretation of extant statutes. See discussion infra at subpart III(A).
  58. .Lynch, supra note 29, at 297–98 (observing the MPC’s success in colonizing state penal codes). Despite its initial success, the MPC’s influence has waned over the years. See Paul H. Robinson & Michael T. Cahill, The Accelerating Degradation of American Criminal Codes, 56 Hastings L.J. 633, 635–44 (2005) (using the examples of Illinois and Kentucky to demonstrate how American penal codes have deteriorated).
  59. .Serota, supra note 10, at 1201 (observing the MPC’s success in clarifying mens rea classifications).
  60. .Model Penal Code § 2.02 (Am. Law Inst. 1985) (setting forth four mental states in defining “General Requirements of Culpability”); see also Robinson & Grall, supra note 21, at 691 (praising the promulgation of the four mental states as “the most significant and enduring achievement of the Code’s authors”).
  61. .According to several critics, the MPC eliminated too many distinctions. See Green, supra note 11, at 20–21, 40 (criticizing the MPC’s consolidation of theft law in the United States); Serota, supra note 10, at 1215–16 (lamenting the MPC’s failure to carry forward, in other substantive offenses, the nuanced approach it used to define homicide); Simon, supra note 5, at 1247–48 (criticizing the MPC’s treatment of homicide as insufficiently graded).
  62. .See Model Penal Code art. 210, intro. note (Am. Law Inst. 1980) (acknowledging that Article 210 “abandons the degree structure that has dominated” states’ codes since Pennsylvania’s first reform of its own law and instead “classifies all criminal homicides into the three basic categories of murder, manslaughter, and negligent homicide”). Notwithstanding this innovation, many states clung to their degree structure, even while adopting much of the MPC’s language and general approach. See, e.g., N.Y. Penal Law §§ 125.00–.60 (McKinney 2009) (creating degrees within the various categories laid out by the MPC).
  63. .Model Penal Code § 1.02(1)(e) (Am. Law Inst. 1985); see also Robinson & Grall, supra note 21, at 682 (arguing that a “precise, principled code” must, among other things, “provide[] the distinctions among degrees of harm and degrees of culpability that create the foundation of a fair sentencing system”).
  64. .Ronald L. Gainer, Federal Criminal Code Reform: Past and Future, 2 Buff. Crim. L. Rev. 45, 111–29 (1998) (tracing the twelve-year effort to revise the federal code leading to its ultimate failure in 1982).
  65. .Id.
  66. .Id. at 118–20 (describing a 12-year effort to revise the federal code).
  67. .See Paul J. Hofer et al., The Effect of the Federal Sentencing Guidelines on Inter-Judge Sentencing Disparity, 90 J. Crim. L. & Criminology 239, 239–40 (1999) (“The sentencing reforms of the past twenty-five years have had several goals, including ‘truth in sentencing,’ control of prison populations, and reduction of unwarranted disparity.”).
  68. .See Rachel E. Barkow, Separation of Powers and the Criminal Law, 58 Stan. L. Rev. 989, 1028 (2006) (noting that Congress purposely created “a federal criminal agency modeled in crucial respects after traditional administrative agencies”); Paul H. Robinson, One Perspective on Sentencing Reform in the United States, 8 Crim. L.F. 1, 13–14 (1997) (describing the Commission’s history, resources, and intended function); Ronald F. Wright, Sentencers, Bureaucrats, and the Administrative Law Perspective on the Federal Sentencing Commission, 79 Calif. L. Rev. 1, 9–10 (1991) (describing early preferences for the creation of a sentencing commission).
  69. .An indeterminate sentencing system is one that sets forth a broad penalty range for an offense but otherwise permits the judge to punish the offender within that prescribed range. See, e.g., Richard S. Frase, Punishment Purposes, 58 Stan. L. Rev. 67, 69 (2005) (“In an indeterminate sentencing system, the legislature usually only provides a very general, all-inclusive list of sentencing purposes and limitations, giving little or no guidance to system actors . . . as to how these principles should be defined and applied in specific cases.”).
  70. .Although the Federal Sentencing Guidelines represent the most famous of the guideline regimes, their administrative approach originates in earlier state practices. See Robinson, supra note 68, at 10 (“The sentencing reform movement began at the state, not the federal, level.”). For an empirical examination of federal pleas and sentencing immediately following the Guidelines’ adoption, see Stephen J. Schulhofer & Ilene H. Nagel, Negotiated Pleas Under the Federal Sentencing Guidelines: The First Fifteen Months, 27 Am. Crim. L. Rev. 231, 232 (1989) (recognizing that the Guidelines produced “[a] new era in federal criminal practice”).
  71. .“Congress’ basic goal in passing the Sentencing Act was to move the sentencing system in the direction of increased uniformity.” United States v. Booker, 543 U.S. 220, 253 (2005). Booker struck down, on Sixth Amendment grounds, the component of the Guidelines that made its sentencing ranges mandatory. Id. at 245. Post-Booker, judges still turn to the Guidelines as an initial starting point but then adjust sentences within their statutorily prescribed sentencing ranges. Beckles v. United States, 137 S. Ct. 886, 894 (2017) (citing Gall v. United States, 552 U.S. 38, 49–50 (2007) (noting that courts may sentence defendants within the statutory range in accordance with the general purposes of sentencing set forth in 18 U.S.C. § 3553)).
  72. .See Wright, supra note 68, at 10 (“[Congress] hoped that the Sentencing Commission could remove sentencing issues from politics to some degree . . . .”).
  73. .Sanford H. Kadish et al., Criminal Law and its Processes: Cases and Materials 447–50 (10th ed. 2016) (citing the Pennsylvania and New York statutory definitions of criminal homicide).
  74. .Not all variants of first-degree murder require an express purpose to kill. See infra note 86 (discussing felony murder). Moreover, not all states employ the MPC’s mens rea terminology. For example, in states such as California, murder is the killing of another with “malice aforethought.” Six states continue to employ this term, and of these six, all but one “divide the offense into first and second degrees.” Anders Walker, The New Common Law: Courts, Culture, and the Localization of the Model Penal Code, 62 Hastings L.J. 1633, 1639 & nn.25–26 (2011).
  75. .See Kadish et al., supra note 73 at 458 (citing number of jurisdictions that employ the premeditation language).
  76. .Kadish et al., supra note 73, at 458–59 (discussing the courts’ conflation of premeditated behavior with intentional conduct falling short of premeditation).
  77. .See Model Penal Code § 210.6 cmt. 4(b) (Am. Law Inst. 1980) (explaining that “[p]rior reflection” may demonstrate a “tortured conscience rather than exceptional depravity”); see also Larry Alexander & Kimberly Kessler Ferzan, Culpable Acts of Risk Creation, 5 Ohio St. J. Crim. L. 375, 400–01 (2008) (identifying premeditation’s over- and under-inclusiveness); Kimberly Kessler Ferzan, Plotting Premeditation’s Demise, Law & Contemp. Probs. no. 2, 2012, at 83,84–86 (2012) (mounting normative arguments against the doctrine’s continued use).
  78. .Kadish et al., supra note 73, at 460–61 (citing courts that require the state to prove “actual reflection” in order to distinguish first-degree murders from less culpable homicides). Efforts to set these murders apart from other killings have admittedly failed in some jurisdictions. See Simon, supra note 5, at 1264–65 (observing that although “premeditated” and “deliberate” were originally intended to reflect the killer’s “settled determination to kill,” this limiting principle eventually disappeared in some jurisdictions).
  79. .See, e.g., Model Penal Code § 210.3(1)(b) (Am. Law Inst. 1985) (defining manslaughter as a criminal homicide that would otherwise be murder but which is “committed under the influence of extreme mental or emotional disturbance for which there is reasonable explanation or excuse”). The provocation defense has attracted its share of criticism. See Aya Gruber, A Provocative Defense, 103 Calif. L. Rev. 273, 278–83 (2015) (summarizing the scholarly debate regarding provocation); Samuel H. Pillsbury, Misunderstanding Provocation, 43 U. Mich. J.L. Reform 143, 143–46 (2009) (highlighting common misunderstandings of provocation). For a defense of and argument in favor of expanding provocation’s partial defense to crimes other than homicide, see generally Paul H. Robinson, A General Mitigation for Disturbance-Driven Crimes? Psychic State, Personal Choice and Normative Inquiries (U. Pa. Law Sch., Pub. Law Research Paper No. 18-29) (Oct. 19, 2018), https:/‌/‌‌abstract=3244770 [https:/‌/‌‌R477-AEE5].
  80. .On the use of the ladder metaphor to distinguish homicides, see generally Victor Tadros, The Homicide Ladder, 69 Mod. L. Rev. 601 (2006) (commenting on England’s proposed statutory homicide reforms); see also Simon, supra note 5, at 1251 (arguing that homicide law ought to separate “terrible violence into morally meaningful substantive crimes” and should “link these crimes through a ladder principle”).
  81. .See, e.g., Michael T. Cahill, Attempt, Reckless Homicide, and the Design of Criminal Law, 78 U. Colo. L. Rev. 879, 880 & n.1 (2007) (explaining that “reckless homicide” is the conceptual equivalent of “involuntary manslaughter”).
  82. .“Depraved heart murder elevates a reckless killing, which typically would be punished as manslaughter, to murder based on an exceptional, extreme recklessness that demonstrates an indifference to the value of human life.” Dora W. Klein, Is Felony Murder the New Depraved Heart Murder? Considering the Appropriate Punishment for Drunken Drivers Who Kill, 67 S.C. L. Rev. 1, 2–3, 3 n.8 (2015) (surveying different state approaches).
  83. .Model Penal Code § 210.2 cmt. 4 (Am. Law Inst. 1980) (explaining the different treatments of ordinary recklessness and recklessness that “manifests extreme indifference to the value of human life”).
  84. .See Kadish et al., supra note 73, at 493–95.
  85. .See, e.g., Gruber, supra note 79 (summarizing scholarly criticisms of the heat-of-passion defense). On practical difficulties in interpretation, see supra notes 77, 79 and accompanying text.
  86. .The felony-murder rule’s threat to homicide’s grading system has incited strong criticism. See Paul H. Robinson & Tyler Scot Williams, Mapping American Criminal Law: Variations Across the 50 States 53–63 (2018) (discussing the status of felony-murder laws in each U.S. state, including some that have rejected the felony-murder rule).
  87. .Even the federal murder statute divides into two different degrees. See 18 U.S.C. § 1111 (2012) (distinguishing first-degree from second-degree murder). Years ago, Professor Robinson hypothesized a single-sentence statute, but his proposal was a parcel of a more radical reformation of criminal law, wherein legislators would promulgate two separate codes, one for laypersons and one for adjudicators. Paul H. Robinson, Structure and Function in Criminal Law 183–84 (1997).
  88. .See Tadros, supra note 80, at 601–02 (“[T]he law provides public guidance about how we should perceive [a] killer . . . .”). Concededly, criminals are often ignorant of the exact wording of criminal statutes. See Paul H. Robinson & John M. Darley, Does Criminal Law Deter? A Behavioral Science Investigation, 24 Oxford J. Legal Stud. 173, 174 (2004) (observing that offenders are often ignorant of legal rules). Nevertheless, over time, rules can and do infiltrate popular discourse. See Tracey L. Meares et al., Updating the Study of Punishment, 56 Stan. L. Rev. 1171, 1182 (2004) (“The educational impact of the criminal law is . . . one that works through a complex process of social interaction.”). For an overview of theoretical work studying criminal perceptions of criminal law and their impact on deterrence, see Von Hirsch et al., supra note 5.
  89. .18 U.S.C. § 1951 (2012) (providing that “[w]hoever in any way or degree obstructs, delays, or affects commerce . . . by robbery . . . shall be fined under this title or imprisoned not more than twenty years, or both”). For a comparison of federal and New York robbery statutes, see Susan R. Klein et al., Why Federal Prosecutors Charge: A Comparison of Federal and New York State Arson and Robbery Filings, 2006–2010, 51 Hous. L. Rev. 1381, 1396–97 (2014).
  90. .Section 160.15 of the New York Penal Code states in relevant part:A person is guilty of robbery in the first degree when he forcibly steals property and when, in the course of the commission of the crime or of immediate flight therefrom, he or another participant in the crime:1. Causes serious physical injury to any person who is not a participant in

    the crime; or

    2. Is armed with a deadly weapon; or

    3. Uses or threatens the immediate use of a dangerous instrument; or

    4. Displays what appears to be a pistol, revolver, rifle, shotgun, machine gun

    or other firearm . . . .

    N.Y. Penal Law § 160.15 (McKinney 2010). By contrast, § 160.10 states in relevant part:

    A person is guilty of robbery in the second degree when he forcibly steals property and when:

    1. He is aided by another person actually present; or

    2. In the course of the commission of the crime or of immediate flight

    therefrom, he or another participant in the crime:

    (a) Causes physical injury to any person who is not a participant in the

    crime; or

    (b) Displays what appears to be a pistol, revolver, rifle, shotgun,

    machine gun or other firearm . . . .

    N.Y. Penal Law § 160.10 (McKinney 2010). Concededly, §§ 160.15(4) and 160.10(2)(b) define identical behavior. Thus, if an offender brandishes a loaded pistol, a prosecutor apparently may choose between a first- or second-degree charge.

  91. .See Model Penal Code § 222.1(1)(a)–(b) (Am. Law Inst. 1980) (“A person is guilty of robbery if, in the course of committing a theft he: (a) inflicts serious bodily injury on another; or (b) threatens another with or purposely puts him in fear of [such] injury.”). Readers may dispute whether theft and robbery should be grouped in the same family, or whether they should be treated as different and unrelated crimes. See, e.g., Stuart P. Green, Lying, Cheating, and Stealing: A Moral Theory of White-Collar Crime 221 (2006) (contending that “robbery is theft by force”).
  92. .Model Penal Code § 222.1(2) (Am. Law Inst. 1980) (defining robbery as a second-degree offense, “except that it is a felony of the first degree if in the course of committing the theft the actor attempts to kill anyone, or purposely inflicts or attempts to inflict serious bodily injury”).
  93. .See N.Y. Penal Law §§ 160.05–.15 (McKinney 2010) (defining first, second, and third degree robbery). Moreover, the Model Penal Code’s gradation of robbery (between first and second degree) hinges more on the robber’s state of mind than on either the severity or risk of harm. Compare id. § 160.15 (elevating robbery to first degree when the robber causes serious physical injury or displays what appears to be a deadly weapon or firearm), with Lynch, supra note 29, at 306 (observing that the Model Penal Code “bases its grading distinction not on the actual degree of injury inflicted, but (primarily) on the robber’s state of mind”).
  94. .Model Penal Code § 5.05(1) (Am. Law Inst. 1985) (treating attempt, solicitation, and conspiracy as “crimes of the same grade and degree” as their completed variants, except that capital and first-degree felony offenses are reduced to the second degree); id. § 5.05 cmt. 2 (rationalizing equal treatment on criminal actor’s “antisocial disposition”). Despite the MPC’s aspiration to treat attempts on the same plane as completed offenses, “most states . . . reject this approach and grade completed offenses higher than attempt.” Paul H. Robinson et al., Realism, Punishment, and Reform, 77 U. Chi. L. Rev. 1611, 1614 & n.16 (2010) (referencing an earlier survey indicating that thirty-seven states had rejected the Model Penal Code’s treatment of attempt).
  95. .See, e.g., Colo. Rev. Stat. Ann. § 18-2-101 (West 2013) (classifying attempts one level below actual crime); Me. Rev. Stat. Ann. tit. 17-A, § 152 (2006) (same). For a deeper treatment of attempt and its criminalization, see generally Gideon Yaffe, Criminal Attempts, 124 Yale L.J. 92 (2014).
  96. .See Tracey L. Meares et al., Updating the Study of Punishment, 56 Stan. L. Rev. 1171, 1173–74 (2004) (“The marginal deterrence argument . . . is one about creating incentives for individuals to refrain from committing the same crime on a greater scale.”); Richard A. Posner, An Economic Theory of the Criminal Law, 85 Colum. L. Rev. 1193, 1207 (1985) (arguing that there is no marginal deterrence “[i]f robbery is punished as severely as murder” because “the robber might as well kill his victim to eliminate a witness”). For more formal treatments, see Steven Shavell, A Note on Marginal Deterrence, 12 Int’l Rev. L. & Econ. 345, 345 (1992) (explaining how the threat of sanctions “influences which harmful acts undeterred individuals choose to commit”). For the earliest treatment, see generally George J. Stigler, The Optimum Enforcement of Laws, 78 J. Pol. Econ. 526 (1970) (introducing the concept into modern discourse).
  97. .See Posner, supra note 96, at 1218 (“If the punishment for attempted murder were the same as for murder, one who shot and missed . . . might as well try again . . . .”).
  98. .On the incommensurability of culpability and harm, see Tadros, supra note 80, at 602 (asking rhetorically, “Is it as bad to rape another as it is to risk the death of another?” and a series of similar questions); see also Kenneth W. Simons, When is Strict Criminal Liability Just?, 87 J. Crim. L. & Criminology 1075, 1093–94 (1997) (describing the incommensurability problem as pervasive in criminal law).
  99. .8 U.S.C. § 1326 (2012) (criminalizing illegal reentry); 21 U.S.C. §§ 841–844 (2012) (enumerating narcotics offenses and penalties).
  100. .See, e.g., 18 U.S.C. § 924(e) (2012) (setting mandatory minimum penalties for offenders convicted of a firearms offense who have previously been convicted of three violent felonies or a “serious drug offense”); 21 U.S.C. § 841 (b)(1)–(3) (2012) (setting minimum statutory penalties according to amounts trafficked in enumerated substances).
  101. .See generally U.S. Sentencing Comm’n, An Overview of Mandatory Minimum Penalties in the Federal Criminal Justice System (2017) (detailing mandatory minimum statutes and their effect on sentencing).
  102. .See 15 U.S.C. § 77x (2012) (setting forth criminal penalties for willful violations of securities statutes); 18 U.S.C. § 201(b)–(c) (2012) (criminalizing bribery of public officials and witnesses); 18 U.S.C. §§ 1341, 1343–1344, 1347–1348 (2012) (proscribing mail, wire, bank, health-care, and securities fraud); 18 U.S.C. §§ 1503–1520 (2012) (defining offenses relating to obstruction, including criminal and civil penalties).
  103. .See Jed S. Rakoff, The Federal Mail Fraud Statute (Part I), 18 Duq. L. Rev. 771, 771 (1980) (“To federal prosecutors of white collar crime, the mail fraud statute is our Stradivarius, our Colt 45, our Louisville Slugger, our Cuisinart—and our true love.”).
  104. .U.S. Sentencing Comm’n, Overview of Federal Criminal Cases: Fiscal Year 2015, at 3 (2016).
  105. .Specialization comes about when Congress enacts a narrower statute alongside a more general one. See Julie Rose O’Sullivan, The Federal Criminal “Code”: Return of Overfederalization, 37 Harv. J.L. & Pub. Pol’y 57, 60 (2014) (criticizing “Congress’ proclivity for responding to events by just passing new statutes—regardless of whether they are redundant or not”). “[E]nacting a multitude of specific statutes to address every new instance of fraud is absurd.” Ellen S. Podgor, Do We Need a “Beanie Baby” Fraud Statute?, 49 Am. U. L. Rev. 1031, 1032 (2000) (warning of excessive specialization in fraud statutes).
  106. .“Given the breadth and variety of the federal criminal code, it is likely that a defendant’s behavior will potentially violate a multitude of overlapping criminal statutes, especially where white-collar crime is involved.” Michael L. Seigel & Christopher Slobogin, Prosecuting Martha: Federal Prosecutorial Power and the Need for a Law of Counts, 109 Pa. St. L. Rev. 1107, 1120 (2005) (explaining how overlap enables a prosecutor to charge multiple counts for a single episode of misconduct). For an argument linking statutory overlap to prosecutorial power, see Russell M. Gold et al., Civilizing Criminal Settlements, 97 B.U. L. Rev. 1607, 1617–24 (2017) (lamenting the growth of prosecutorial leverage due in part to overlapping criminal codes).
  107. .The number of counts can add up quite quickly. See, e.g., Fumo v. United States, No.13-3313, 2014 WL 2547797, at *1 (E.D. Pa. June 5, 2014) (describing a Pennsylvania state senator’s conviction on “one hundred thirty-seven counts of conspiracy, fraud, obstruction of justice, and aiding and abetting the filing of false tax returns of a tax-exempt organization”).
  108. .See supra note 9 and accompanying text.
  109. .Frank O. Bowman, III, Pour Encourager Les Autres? The Curious History and Distressing Implications of the Criminal Provisions of the Sarbanes-Oxley Act and the Sentencing Guidelines Amendments that Followed, 1 Ohio St. J. Crim. L. 373, 383–84 (2004) (explaining that caps are usually irrelevant because “a single criminal scheme so often consists of a multitude of acts separately chargeable as federal crimes”). When an offender’s Guideline range exceeds the statutory cap for one offense, the sentencing court can apportion the sentence over several counts, requiring the offender to serve consecutive terms. See, e.g., United States v. Rutherford, 599 F.3d 817, 820 (8th Cir. 2010) (citing standard procedures under the Guidelines and the federal code when multiple counts are involved).
  110. .On the origins of the Guidelines, see supra subpart I(A). Regarding the proper procedures for calculating and taking account of the Guidelines’ advisory recommended sentencing ranges, see infra notes 123–26.
  111. .See, e.g., United States v. Ulbricht, 858 F.3d 71, 130–31 (2d Cir. 2017) (noting rarity of life-sentences for federal drug offenses even when recommended by the Guidelines).
  112. .U.S. Sentencing Comm’n, 2016 Sourcebook of Federal Sentencing Statistics app. B (2017), https:/‌/‌‌sites/‌default/‌files/‌pdf/‌research-and-publications/‌annual-reports-and-sourcebooks/‌2016/‌stats_Nat.pdf [https:/‌/‌‌UQN3-2RGH] (analyzing national data for Guidelines sentencing during fiscal year 2016). The average sentence (reflecting much higher sentences at the very top) was thirty-four months’ imprisonment. Id. Roughly one-sixth of those sentenced for fraud offenses in 2016 received no term of imprisonment at all. See id. (reporting that 5,078 of the 6,517 fraud offenses in 2016 resulted in prison terms). Taking into account those who received no prison sentence at all, the mean and median sentence terms are twenty-five months and thirteen months, respectively. Id. at tbl.13, https:/‌/‌‌sites/‌default/‌files/‌pdf/‌research-and-publications/‌annual-reports-and-sourcebooks/‌2016/‌Table13.pdf [https:/‌/‌‌PDC8-A8GN] (analyzing the sentence length in each primary offense category for fiscal year 2016).
  113. .U.S. Sentencing Comm’n, Use of Guidelines and Specific Offense Characteristics, Offender Based: Fiscal Year 2016, at 10, https:/‌/‌‌sites/‌default/‌files/‌pdf/‌research-and-publications/‌federal-sentencing-statistics/‌guideline-application-frequencies/‌2016/‌Use_of_SOC_Offender_Based.pdf [https:/‌/‌‌RWN9-UKEP] [hereinafter Use of Guidelines].
  114. .See, e.g., Bowman, supra note 109; Seigel & Slobogin, supra note 106; see also U.S. Sentencing Guidelines Manual § 5G1.2(c)–(d) (U.S. Sentencing Comm’n 2016).
  115. .See, e.g., United States v. Dearing, 504 F.3d 897, 903 (9th Cir. 2007) (observing that the “intent to defraud” element “is common to the federal fraud statutes” and analyzed in the same way). “Federal wire, mail and securities fraud statutes all have scienter requirements that purport to require some type of dishonest mental state . . . .” Hilary J. Allen, The Pathologies of Banking Business as Usual, 17 U. Pa. J. Bus. L. 861, 912 (2015) (observing similarities among the statutes’ state-of-mind provisions); see also Zingale, supra note 43, at 799 (describing the unity of elements for mail and wire frauds, other than the jurisdictional element).
  116. .See Samuel W. Buell, What Is Securities Fraud?, 61 Duke L.J. 511, 526, 529 (2011) (treating as “core fraud” purposeful efforts to deceive).
  117. .United States v. Herzig, 26 F.2d 487, 489 (S.D.N.Y. 1928) (“A scheme to defraud may be simple in its plan and execution . . . .”).
  118. .See Buell, supra note 116, at 557–58 (cataloguing cases which have held that “specific intent to defraud” includes “reckless” behavior). For a helpful discussion of how the reckless state of mind has evolved with regard to insider-trading prosecutions, see Donald C. Langevoort, “Fine Distinctions” in the Contemporary Law of Insider Trading, 2013 Colum. Bus. L. Rev. 429,436–37 (contrasting courts who treat recklessness as extreme versions of negligence from those who treat it as a form of “conscious avoidance”).
  119. .Buell, supra note 116, at 556–59, 558 nn.170–71; see also Dearing, 504 F.3d at 903 (“We have repeatedly held that the intent to defraud may be proven through reckless indifference to the truth or falsity of statements.”); United States v. Coyle, 63 F.3d 1239, 1243 (3d Cir. 1995) (citations omitted) (observing that the intent to defraud “may be found from a material misstatement of fact made with reckless disregard for the truth”).
  120. .See, e.g., United States v. DeSantis, 134 F.3d 760, 764 (6th Cir. 1998) (stating a prosecutor may prove state of mind element by demonstrating a purpose to defraud or “that the defendant was reckless” and made an “extreme departure” from ordinary standards of care in omitting certain information (internal quotation marks and citation omitted)); United States v. Henderson, 446 F.2d 960, 966 (8th Cir. 1971) (appearing to embrace as fraud “reckless misrepresentations made to induce innocent victims to part with their money”); see also United States v. Boyer, 694 F.2d 58, 59–60 (3d Cir. 1982) (refusing to find a jury charge improper where the charge stated that the jury could find criminal liability either if the defendant knowingly lied or if his false statement was “due to recklessness on his part”).
  121. .Buell, supra note 116, at 530–31 (explaining the difference between “goal-oriented” behavior and misrepresentations that create loss).
  122. .Id. at 526–27 (distinguishing “core fraud” cases from those that rest upon determinations of non-goal-oriented deception).
  123. .For an overview and empirical study of § 2B1.1’s history, recent reform efforts, and federal judges’ increasing tendency to depart from its recommended sentencing ranges, see generally Mark W. Bennett et al., Judging Federal White-Collar Fraud Sentencing: An Empirical Study Revealing the Need for Further Reform, 102 Iowa L. Rev. 939 (2017).
  124. .U.S. Sentencing Guidelines Manual § 2B1.1 (U.S. Sentencing Comm’n 2016).
  125. .Bennett et al., supra note 123, at 944–45. “[T]he core criticism of section 2B1.1 is that loss, however defined, adds too many offense levels to defendants’ guideline calculations and thus increases sentence length by unduly large amounts.” Frank O. Bowman, III, Recalibrating the Federal Economic Crime Guideline: An Admiring Rejoinder to Judge Bennett and Friends, 102 Iowa L. Rev. Bull. 205, 227 (2017).
  126. .For more on § 2B1.1 and its respective enhancements, see Bennett et al., supra note 123, at 985–86.
  127. .See U.S. Sentencing Guidelines Manual §§ 3B1.1–.2 (U.S. Sentencing Comm’n 2016) (defining the circumstances under which adjustments for supervisory or minor roles in offense are appropriate).
  128. .See, e.g., Douglas A. Berman, Fiddling with the Fraud Guidelines as Booker Burns, 27 Fed. Sent’g Rep. 267, 267 (2015) (articulating common criticisms of § 2B1.1). Many of § 2B1.1’s ills can be traced more generally to the Guidelines’ well-intentioned effort to eliminate sentencing disparities by relying almost exclusively on objective criteria. See Bierschbach & Bibas, supra note 28, at 1478 (“[T]he more guidelines sought to equalize outcomes and minimize disparities, the more detailed and restrictive they were.”); see also Richard A. Bierschbach & Stephanos Bibas, Notice-and-Comment Sentencing, 97 Minn. L. Rev. 1, 6 (2012) [hereinafter Bierschbach & Bibas, Notice and Comment] (observing that the Guidelines have been denigrated as “complex, unintelligible, and unjust”).
  129. .Judge Rakoff has been particularly instrumental in bringing these issues to the forefront. See United States v. Gupta, 904 F. Supp. 2d 349, 351 (S.D.N.Y. 2012) (Rakoff, J.) (warning that the Guidelines’ approach to securities fraud has unduly crowded out factors other than monetary loss); United States v. Adelson, 441 F. Supp. 2d 506, 512 (S.D.N.Y. 2006) (Rakoff, J.) (arguing that the Guidelines’ emphasis on quantitative factors can devolve into a “fetish with abstract arithmetic”).
  130. .See, e.g., Daniel S. Guarnera, A Fatally Flawed Proxy: The Role of “Intended Loss” in the U.S. Sentencing Guidelines for Fraud, 81 Mo. L. Rev. 715, 739 (2016) (“Actual and intended loss often point in the same direction, but they do not always do so.”).
  131. .See, e.g., Jillian Hewitt, Note, Fifty Shades of Gray: Sentencing Trends in Major White-Collar Cases, 125 Yale L.J. 1018, 1025 (2016) (revealing findings that a “significant majority of defendants in major white-collar cases today receive sentences shorter than the Guidelines range”).
  132. .Berman, supra note 128, at 267–68.
  133. .Chief Judge Patti Saris, Chair, U.S. Sentencing Comm’n, The 2015 Economic Crime Amendments, Keynote Address at the Regulatory Offenses and Criminal Law Conference (Apr. 14, 2015), in 27 Fed. Sent’g Rep. 322, 323 (2015).
  134. .Berman, supra note 128, at 268 (contending that the revised Guidelines proposals were, “in a word, anti-climactic”). Even under the revised Guidelines, intended loss still “includes intended pecuniary harm[s]” that are either implausible or impossible. Hewitt, supra note 131, at 1034 (citation omitted).
  135. .See Serota, supra note 10, at 1222 (criticizing the Model Penal Code’s “dichotomous, all or nothing approach” to accomplice liability). Although earlier doctrines distinguished principals and accessories, modern accomplice liability treats all accomplices equally. Id. at 1222 n.108.
  136. .See, e.g., United States v. Rigas, 605 F.3d 194, 206 n.9 (3d Cir. 2010) (“The specific elements of [a] conspiracy to violate federal law are: (1) an agreement to commit an offense proscribed by federal law; (2) the defendants intentionally joining in the agreement; (3) one of the conspirators committing an overt act; and (4) an overt act in furtherance of the conspiracy.”).
  137. .Rosemond v. United States, 134 S. Ct. 1240, 1245 (2014).
  138. .“Conspiracy [effectively] ‘collapses the distinction between accessories and perpetrators’ [in that it treats] conspirators as principals in any substantive offense committed in furtherance of the conspiracy, whether or not they directly participated in that offense.” United States v. Alvarez, 610 F.2d 1250, 1253 (5th Cir. 1980) (citing George P. Fletcher, Rethinking Criminal Law 674 (1978)).
  139. .Id. at 1253–54 (observing that “criminalization of conspiracy eradicates common-law and theoretical methods of distinguishing the degree of liability of various participants in criminal enterprises”).
  140. .Rosemond, 134 S. Ct. at 1245 (observing that a person who “‘aids, abets, counsels, commands, induces or procures’[]the commission of a federal offense ‘is punishable as a principal’” (quoting 18 U.S.C. § 2 (2012))).
  141. .See Model Penal Code § 2.06(3) (Am. Law Inst. 1985) (stating that an accomplice must act “with the purpose of promoting or facilitating the commission of the offense”). In some states, this requirement has been downgraded to mere knowledge. See, e.g., United States v. Valdivia-Flores, 876 F.3d 1201, 1207–08 (9th Cir. 2017) (comparing Washington State’s more inclusive knowledge standard with federal law’s intent requirement).
  142. .In some states, knowledge is a stand-alone requirement; in others, it provides an inference of intent. Compare State v. D. H., 643 P.2d 457, 460 (Wash. Ct. App. 1982) (“Knowing assistance of another in the commission of a crime is a predicate for accomplice liability.”), with United States v. Heras, 609 F.3d 101, 107 (2d Cir. 2010) (observing that although knowledge in and of itself is insufficient to establish liability, it may be used to demonstrate “that the defendant has adopted the known goal of the crime as his own”). For criticisms of complicity doctrines that have eroded the distinction between knowledge and purpose, see Kit Kinports, Rosemond, Mens Rea, and the Elements of Complicity, 52 San Diego L. Rev. 133, 153 (2015) (raising the concern that Rosemond may erode the previously understood rule that “knowledge and purpose are distinct mental states”).
  143. .134 S.Ct. 1240 (2014).
  144. .Rosemond, 134 S. Ct. at 1250 (“The law does not, nor should it, care whether [the accomplice] participates with a happy heart or a sense of foreboding.”).
  145. .Despite raising this issue in the past, scholars have not experienced any success in securing meaningful reform. In the mid-1980s, Professor Dressler argued for the differentiation of causal from noncausal accomplices. Joshua Dressler, Reassessing the Theoretical Underpinnings of Accomplice Liability: New Solutions to an Old Problem, 37 Hastings L.J. 91, 93, 120 (1985). Persons in the latter group would still be subject to criminal prosecution, but they would be convicted of a lesser offense. Id. at 120–21. Dressler later wrote, “This causation approach received some academic attention, but no legislative interest.” Joshua Dressler, Reforming Complicity Law: Trivial Assistance as a Lesser Offense?, 5 Ohio St. J. Crim. L. 427, 430 (2008).
  146. .Cf. John C. Jeffries, Jr. & Honorable John Gleeson, The Federalization of Organized Crime: Advantages of Federal Prosecution, 46 Hastings L.J. 1095, 1107 (1995) (commenting on the “prototypical state prosecution, which focuses on a discrete, isolated criminal act,” and contrasting it with the ongoing criminal cases that characterize organized crime prosecutions).
  147. .For purposes of this Article, I consider a regime that employs both statutory grading and sentencing guidelines to be a graded regime.
  148. .“Current punishment theorists identify four distinct rationales for punishment . . . : retribution, deterrence, incapacitation, and rehabilitation.” John F. Stinneford, The Original Meaning of “Cruel”, 105 Geo. L.J. 441, 459 (2017). Legislatures are free to select or combine these rationales. See Graham v. Florida, 560 U.S. 48, 71 (2010) (“Criminal punishment can have different goals, and choosing among them is within a legislature’s discretion.”).
  149. .See generally Gary S. Becker, Crime and Punishment: An Economic Approach, 76 J. Pol. Econ. 169 (1968) (using economic analysis to evaluate how to discourage criminal behavior).
  150. .See Miriam H. Baer, Linkage and the Deterrence of Corporate Fraud, 94 Va. L. Rev. 1295, 1306–07 (2008) (“[T]he enforcement probability term is itself a composite of several probabilities, including . . . (i) detection; (ii) subsequent apprehension; (iii) conviction in a court of law; and (iv) sanctions.”).
  151. .Becker, supra note 149, at 180–81 (constructing a social welfare function that “gives due weight to the damages from offenses, the costs of apprehending and convicting offenders, and the social cost of punishment”). The success of any deterrence strategy, it should be noted, depends on the public’s perception of sanctions and probabilities of punishment. A high penalty for a given crime will not deter anyone if it is unknown or widely perceived as a meaningless bluff. See von Hirsch et al., supra note 5, at 6 (“To the extent that changes in actual penal policies do not alter potential offenders’ beliefs about the likelihood or severity of punishment, they cannot generate any marginal deterrence.”).
  152. .Becker, supra note 149, at 184 (theorizing why some countries pair high sanctions with low rates of “capture and conviction”). On why this may be a suboptimal strategy, see, e.g., A. Mitchell Polinsky & Steven Shavell, On the Disutility and Discounting of Imprisonment and the Theory of Deterrence, 28 J. Legal Stud. 1, 4–7 (1999) (demonstrating how the decreasing disutility of imprisonment can cause risk-preferring offenders to discount increases in sanctions); see also Yair Listokin, Crime and (with a Lag) Punishment: The Implications of Discounting for Equitable Sentencing, 44 Am. Crim. L. Rev. 115, 117 (2007) (using discount theory to demonstrate the reduced effectiveness of prison sentences imposed long after the commission of the crime).
  153. .As Professor Gilbert observes, a regulator can also enact a series of “insincere” rules (e.g., a rule requiring a slower traffic speed than the regulator actually prefers). Insincere rules heighten sanctions for rule violators because they cause the offending behavior to appear substantially worse than the residual behavior permitted under the newly enacted statute. See Michael Gilbert, Insincere Rules, 101 Va. L. Rev. 2185, 2194–99 (2015) (analyzing “insincere” rules’ punitive effect).
  154. .See, e.g., Edward K. Cheng, Structural Laws and the Puzzle of Regulating Behavior, 100 Nw. U. L. Rev. 655, 662 (2006) (describing structural laws as regulating “indirectly and ex ante by subtly shaping the physical, social, or other arrangements that enable the behavior to occur in the first place”); see also Leandra Lederman, Statutory Speed Bumps: The Roles Third Parties Play in Tax Compliance, 60 Stan. L. Rev. 695, 697–99 (2007) (discussing structural mechanisms used in the federal income tax system to facilitate compliance with the law).
  155. .Stephen R. Galoob, Retributivism and Criminal Procedure, 20 New Crim. L. Rev. 465, 466 (2017) (“Retributivism is the view that criminal punishment is justified in large part (and perhaps entirely) by an offender’s deserving to be punished.”).
  156. .See Tison v. Arizona, 481 U.S. 137, 149 (1987) (“The heart of the retribution rationale is that a criminal sentence must be directly related to the personal culpability of the criminal offender.”); see also Cahill, supra note 32, at 826–30 (surveying several versions of retributive theory); Stinneford, supra note 148, at 450 (“Retribution asks whether the punishment matches the offender’s culpability.”).
  157. .See, e.g., Richard S. Frase, Limiting Retributivism, in The Future of Imprisonment 83, 83 (Michael Tonry ed., 2004) (describing Norval Morris’s famous limiting concept).
  158. .“Retributivist literature is rife with references to the principle of desert-based punishment as a moral duty and to the corresponding claim that the retributive principle does not merely authorize punishment but affirmatively calls for its imposition on those who deserve it.” Cahill, supra note 32, at 826 (summarizing the absolutist view of retributivist thought).
  159. .Daniel S. Nagin, Deterrence in the Twenty-First Century, 42 Crime & Just. 199, 201 (2013) (reporting the social-research finding that probability of detection is the variable that most affects likelihood of offending).
  160. .An issue is salient when it is easily recalled by one’s mind. See Cass R. Sunstein, What’s Available? Social Influences and Behavioral Economics, 97 Nw. U. L. Rev. 1295, 1297 (2003) (explaining the cognitive underpinnings of the availability heuristic).
  161. .To the extent criminal law shapes an individual’s preferences, salience is a desirable characteristic. See, e.g., Kenneth G. Dau-Schmidt, An Economic Analysis of the Criminal Law as a Preference-Shaping Policy, 1990 Duke L.J. 1, 37 (modeling criminal law as a tool for shaping preferences).
  162. .One can make the same argument for grading attempts as lesser offenses. See Steven Shavell, Criminal Law and the Optimal Use of Nonmonetary Sanctions as a Deterrent, 85 Colum. L. Rev. 1232, 1251–53 (1985) (articulating the economic argument for assigning an attempted crime a lesser sanction).
  163. .See Robinson et al., supra note 33, at 734 (describing the degradation of codes that occurs when legislatures enact statutes for purely political reasons).
  164. .See, e.g., Ewing v. California, 538 U.S. 11, 17–18, 20 (2003) (rejecting the claim that the defendant’s twenty-five-year sentence of imprisonment for stealing three golf clubs was grossly disproportionate); Rummel v. Estelle, 445 U.S. 263, 265–66 (1980) (affirming a sentence of life imprisonment that was triggered by the offender’s theft of $120.75 by false pretenses). Both of these were state cases, and both dealt with legislative codes (California and Texas) one likely would categorize as graded.
  165. .“For life, he will bear the stigma of having a federal felony conviction.” Yates v. United States, 135 S. Ct. 1074, 1080–81 (2015) (commenting on the stigma the defendant will suffer despite his thirty-day sentence for conviction); see also supra note 11 and accompanying text.
  166. .Stuart P. Green & Matthew B. Kugler, When Is it Wrong to Trade Stocks on the Basis of Non-Public Information? Public Views of the Morality of Insider Trading, 39 Fordham Urb. L.J. 445, 451 (2011) (embracing fair labeling as “the idea that ‘widely felt distinctions between kinds of offences and degrees of wrongdoing are respected and signaled by the law, and that offences are subdivided and labeled so as to represent fairly the nature and magnitude of the law-breaking.’” (quoting Andrew Ashworth, Principles of Criminal Law 88 (5th ed. 2006))).
  167. .Stuart P. Green, Consent and the Grammar of Theft Law, 28 Cardozo L. Rev. 2505, 2517 (2007). Green goes on to argue that “the acts underlying larceny, fraud, extortion, and embezzlement should, at least for purposes of fair labeling, be regarded as morally distinct.” Id. at 2518.
  168. .See, e.g., Eli K. Best, Elements, Sentencing Factors, and the Right to a Jury Trial: An Analysis of Legislative Power and its Limits, 29 Yale L. & Pol’y Rev. 633, 643–44 (2010) (observing the ways in which sentencing factors can greatly impact a defendant’s sentence under the Sentencing Guidelines despite their weaker burden of proof).
  169. .The graded system codifies more formal statutory elements, which a factfinder must find beyond a reasonable doubt. See In re Winship, 397 U.S. 358, 364 (1970) (“[T]he Due Process Clause protects the accused against conviction except upon proof beyond a reasonable doubt of every fact necessary to constitute the crime with which he is charged.”). An ungraded system—particularly one that relies upon an indeterminate sentencing regime—incorporates fewer statutory elements and instead reserves to the judge the discretion to sentence according to factors not proven beyond a reasonable doubt. See McMillan v. Pennsylvania, 477 U.S. 79, 85–86 (1986) (distinguishing statutory elements from sentencing factors). Although the Supreme Court decided subsequent to its holding in McMillan that a mandatory Guidelines regime impermissibly eroded the defendant’s right to have a jury decide his guilt beyond a reasonable doubt, the Court nevertheless went on to preserve advisory Guidelines regimes and the preponderance standard that governs factfinding at sentencing. See United States v. Booker, 543 U.S. 220, 233 (2005) (affirming the judge’s “broad discretion” to sentence the defendant “within a statutory range”); United States v. Mustafa, 695 F.3d 860, 862 (8th Cir. 2012) (stating that under the advisory Guidelines scheme, “due process never requires applying more than a preponderance-of-the-evidence standard” (citation omitted)).
  170. .Professor Doron Teichman conceptualizes this framework as one that correlates the severity of the sanction with the certainty that the defendant engaged in wrongdoing. See Doron Teichman, Convicting with Reasonable Doubt: An Evidentiary Theory of Criminal Law, 93 Notre Dame L. Rev. 757, 758 (2017) (arguing that graded statutes distribute sanctions according to the “degree of certainty” of wrongdoing). Although provocative, Teichman’s thesis does not describe all instances of offense grading. For example, we might know for certain that an individual recklessly killed another individual by driving while texting on his cell phone. If we charge that individual with manslaughter instead of murder, we are doing so because we believe accidental killings deserve less condemnation than intentional ones, and not because we are uncertain of the killer’s mental state.
  171. .Cf. Wright & Engen, Charge Movement, supra note 5 (concluding that “criminal codes matter, even in a world of bargained justice”).
  172. .See generally Gold et al., supra note 106 (describing ways in which prosecutors can induce defendants to accept suboptimal plea agreements). But see Andrew Manuel Crespo, The Hidden Law of Plea Bargaining, 118 Colum. L. Rev. 1303, 1305–06, 1316 tbl.1, 1375–76 tbls.2 & 3 (2018) (analyzing the subconstitutional rules of procedure which rein in prosecutorial overreach).
  173. .For evidence that prosecutors respond to legally imposed obligations, see Jenia I. Turner & Allison D. Redlich, Two Models of Pre-Plea Discovery in Criminal Cases: An Empirical Comparison, 73 Wash. & Lee L. Rev. 285, 294 (2016) (finding that prosecutors did in fact increase discovery to defense counsel in response to a change in the state’s discovery laws).
  174. .I do not mean to overstate my claim here. Ceiling rights and proof rights aid the defendant in some instances, but the availability of multiple charges and viable “landing spots” aids the prosecutor in plea negotiations. See Wright & Engen, Charge Movement, supra note 5, at 16–17, 36 (explaining how “deep” codes “empower” prosecutors by enabling them to facilitate plea deals in part because they can pick from several landing spots when considering a menu of felony options). Thus, it is difficult to say as an absolute matter that defendants always fare better under graded statutes. They may feel better off, however, in a system that maximizes and respects their ceiling and proof rights.
  175. .Best, supra note 168, at 634.
  176. .See supra note 69 and accompanying text.
  177. .Alleyne v. United States, 133 S. Ct. 2151, 2160 (2013) (holding that a fact that triggers the mandatory minimum is an element that must be proved to the jury beyond a reasonable doubt).
  178. .Apprendi v. New Jersey, 530 U.S. 466, 490 (2006) (declaring that facts that increase the statutory maximum penalty are elements of the crime that must be proved to the jury beyond a reasonable doubt).
  179. .See supra note 169.
  180. .One of the reasons these terms are salient is that there are relatively few of them. Five or six categories are easier to remember than an encyclopedic list of aggravating and mitigating factors. For the seminal work on categorical numerosity and working memory, see generally George A. Miller, The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information, 63 Psych. Rev. 81 (1956).
  181. .Stuart P. Green & Matthew B. Kugler, Public Perceptions of White Collar Crime Culpability: Bribery, Perjury, and Fraud, 75 Law & Contemp. Probs. 33, 33 (2012) (“[T]he ability of criminal law to stigmatize, to achieve legitimacy, and to gain compliance ultimately depends on the extent to which it enjoys moral credibility and recognition in the broader lay community.”).
  182. .See generally Tom R. Tyler, Why People Obey the Law (1990) (demonstrating that individuals are more likely to voluntarily comply with the law when they view the legal system as procedurally just). For more recent applications of the procedural-justice concept in the policing context, see Tom R. Tyler & Jeffrey Fagan, Legitimacy and Cooperation: Why Do People Help the Police Fight Crime in Their Communities?, 6 Ohio St. J. Crim. L. 231, 263 (2008) (examining policing’s effect on community members’ feelings of legitimacy and willingness to cooperate with the police).
  183. .Legitimacy also serves as the strongest argument against the adoption of insincere laws. See, e.g., Gilbert, supra note 153, at 2185–86 (theorizing a system in which rule-makers purposely adopt insincere rules on the books with the intention of producing on the ground behavior that falls short of the written rule but that otherwise accords with the rule-makers’ true preferences). However defensible the insincere strategy might be in the regulatory context, it is difficult to defend in the criminal domain. See id. at 2218 (conceding that the strategy may yield negative outcomes and “runs into deontological objections” when those who fail to comply with the rule receive criminal sanctions that appear overly harsh or undeserved).
  184. .“[T]he simple fact that the state is pursuing its ends through the use of superior force raises concerns in our political system . . . .” Alice Ristroph, Just Violence, 56 Ariz. L. Rev. 1017, 1025 (2014) (citing Max Weber, Politics as a Vocation, in From Max Weber: Essays in Sociology 77, 78 (H. H. Gerth & C. Wright Mills eds., 2001) (defining a state by its monopoly on the legitimate use of force within its territory)); see also Guyora Binder & Robert Weisberg, What Is Criminal Law About?, 114 Mich. L. Rev. 1173, 1175 (2016) (“Criminal law regulates violence by asserting a public monopoly on vengeance.”); David A. Sklansky, The Private Police, 46 UCLA L. Rev. 1165, 1188 (1999) (repeating the Weberian concept of government as having a monopoly over “the coercive use of force, in order to ensure public peace, personal security, and the use and enjoyment of property”).
  185. .See William J. Stuntz, Self-Defeating Crimes, 86 Va. L. Rev. 1871, 1872 (2000) (describing as “self-defeating” those instances in which criminalization contravenes societal norms).
  186. .Tyler, supra note 182, at 106.
  187. .Recent work by John Pfaff attributes mass incarceration to prosecutorial decision-making in state and local courts, as opposed to federal mandatory minimum sentencing statutes. See, e.g., John F. Pfaff, Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform 131–32 (2017).
  188. .Ordinarily, the separation-of-powers argument is accompanied by a call for less criminal lawmaking by federal judges and a more robust use of the rule-of-lenity, void-for-vagueness, and other interpretive doctrines. See, e.g., Hopwood, supra note 51, at 742–43 (arguing that federal courts should employ lenity and void-for-vagueness doctrines to strike down ambiguous criminal laws); Smith, supra note 57, at 541 (urging courts to adopt narrowing interpretations of criminal statutes to combat the overbreadth created by vast codes).
  189. .Serota, supra note 10, at 1206–07 (arguing that a penal system’s policies ought to reflect the people’s will); see also Robinson, supra note 79, at 11 (explaining that “part of the value in having a codified provision [for a partial provocation defense] is that it requires a principled analysis of” what the law should be).
  190. .See Robinson, supra note 79, at 10 (arguing in favor of a codified provocation defense because statutory codification places the normative question before the jury, who is most qualified to make such a normative judgment); Rachel E. Barkow, Sentencing Guidelines at the Crossroads of Politics and Expertise, 160 U. Pa. L. Rev. 1599, 1603 (2012) (“[A]t the heart of any criminal justice system are questions of morality and justice that are not amenable to charts and data but rather are suited for juries comprised of members of the community.”); see also Rachel E. Barkow, Recharging the Jury: The Criminal Jury’s Constitutional Role in an Era of Mandatory Sentencing, 152 U. Pa. L. Rev. 33, 61–65, 78–82 (2003) (heralding jurors’ potential role in checking prosecutorial power).
  191. .Bierschbach & Bibas, Notice and Comment, supra note 128, at 2–3 (observing that experts play a much larger role in setting criminal sentences than judges or laypeople).
  192. .See, e.g., Stith & Cabranes, supra note 17 (describing various technical issues that the Guidelines have spawned); cf. Erik Luna, Gridland: An Allegorical Critique of Federal Sentencing, 96 J. Crim. L. & Criminology 25, 27 (2005) (criticizing a system that plots defendants on “a sentencing formula” that doles out punishment according to a “two-dimensional grid”).
  193. .See, e.g., Simon, supra note 5, at 1282 (arguing that although homicide gradations exist in California law, the collapse of parole and capital punishment have effectively transformed its murder statute into a flattened, two-level regime).
  194. .See, e.g., Ronald F. Wright, Reinventing American Prosecution Systems, 46 Crime & Just. 395, 395 (2017) (“The prosecutorial function can be reimagined with more effective legal, institutional, and internal cultural constraints that would produce responsive prosecutorial services.”); id. at 399–401 (setting forth eleven reforms to make prosecutors’ charging decisions more transparent and democratically accountable); see also Stephanos Bibas, Prosecutorial Regulation Versus Prosecutorial Accountability, 157 U. Pa. L. Rev. 959, 988–90 (2009) (pitching feedback reforms to make prosecutors more accountable to the public); Bierschbach & Bibas, Notice and Comment, supra note 128, at 41–42, 48–49 (proposing accountability mechanisms for sentencing).
  195. .This problem—that a certain type of misconduct, albeit wrongful, strays too far from the prototypical misconduct the court has in mind—is distinct from the more familiar situation in which the public questions whether certain behavior should be subject to criminal liability at all. See, e.g., Green & Kugler, supra note 166 (“[W]hite collar crimes often raise the issue of whether an act is wrong at all.”).
  196. .Professor Lawrence Solan’s work mines the implications of linguistics for statutory interpretation. According to Professor Solan, individuals associate statutory terms (including those appearing in criminal statutes) with prototypes. Lawrence M. Solan, Law, Language, and Lenity, 40 Wm. & Mary L. Rev. 57, 66–70 (1998) (citing studies indicating that individuals’ “knowledge of concepts is better characterized in terms of prototypes for categories”). Thus, when a statute uses the term fraud, a given prototype comes to mind. “[P]rototype analysis explains why it is that we might make category errors. If we focus on the prototype, we may wrongly use an overinclusive or underinclusive category, only to discover later the poor fit between disputed events and statutory categories.” Id. at 68.
  197. .Skilling v. United States, 561 U.S. 358, 408–09, 413 (2010). Skilling construed the honest services fraud statute to apply only to bribery and kickback schemes. Id. at 408–09.
  198. .“John Yates, a commercial fisherman, caught undersized red grouper in federal waters in the Gulf of Mexico. To prevent federal authorities from confirming that he had harvested undersized fish, Yates ordered a crew member to toss the suspect catch into the sea.” Yates v. United States, 135 S. Ct. 1074, 1078 (2015). Yates was prosecuted under 18 U.S.C. § 1519, which includes a twenty-year statutory maximum term of imprisonment, and 18 U.S.C. § 2232(a), which carries a five-year maximum term. Id. at 1078–79. He challenged his conviction under § 1519, which punishes the destruction or concealment of a “tangible object,” but not his conviction under § 2232(a), which prohibits the destruction of “property” that a government investigator seeks to take into custody. Id.
  199. .McDonnell v. United States, 136 S. Ct. 2355, 2363–71 (2016). Another way to view McDonnell and Yates is to conclude, as Aziz Huq and Genevieve Lakier do in a recent article, that these cases reflect the judiciary’s inexorable move to an apparent fault-based theory of federal crime, whereby courts interpret statutes in such a way as to avoid punishing individuals who might have been unaware of the moral wrongfulness of their misconduct. See Aziz Z. Huq & Genevieve Lakier, Apparent Fault, 131 Harv. L. Rev. 1525, 1563–64 (2018). The present Article adds an additional dimension to Huq and Lakier’s fault theory. However queasy courts feel when federal prosecutors purport to punish innocent behavior under broad criminal statutes, those feelings intensify when the statute in question is an ungraded statute. Accordingly, the jurisprudence of apparent fault owes some of its existence to federal criminal law’s lack of gradation.
  200. .Yates, 135 S. Ct. at 1087–88 (contrasting the obstruction statute’s twenty-year maximum with the Model Penal Code’s misdemeanor provision).
  201. .Compare Smith, supra note 57, at 582 n.169 (arguing that “taking proportionality considerations into account in interpreting federal crimes is no more perilous than in [legislative] contexts”), with Kahan, supra note 44, at 397 (arguing that “federal criminal law works best when Congress is permitted to cede a certain portion of its criminal lawmaking to courts”).
  202. .Justice Scalia’s own question during the Yates oral argument suggests not. See Ellen S. Podgor, “What Kind of a Mad Prosecutor” Brought Us This White Collar Case, 41 Vt. L. Rev. 523, 523 & n.1 (2017) (citing Transcript of Oral Argument at 28, Yates v. United States, 135 S. Ct. 1074 (2015) (13-7451), wherein Justice Scalia asked: “What kind of a mad prosecutor would try to send this guy up for 20 years or risk sending him up for 20 years?”).
  203. .Yates himself did not challenge his conviction under 18 U.S.C. § 2232(a), which sets forth a five-year maximum sentence. In its comparison of § 1519 to a similar evidence-tampering provision in the Model Penal Code, the plurality pointedly observed that the latter crime was a misdemeanor. Yates, 135 S. Ct. at 1087–88 (“Section 1519 conspicuously lacks the limits built into the [Model Penal Code] provision.”).
  204. .See, e.g., id. at 1088 (embracing the defendant’s lenity argument where defendants might otherwise be subject to a broad statute that “exposes individuals to 20-year prison sentences for tampering with any physical object that might have evidentiary value in any federal investigation”). As Justice Kagan suggests, the defendant in Yates never realistically faced a twenty-year sentence. Id. at 1100 (Kagan, J., dissenting) (“Let’s not forget that Yates’s sentence was not 20 years, but 30 days.”). Admittedly, some white-collar offenders do face draconian sentences, but these tend to be accounting-fraud cases within publicly held companies. See, e.g., Samuel W. Buell, Reforming Punishment of Financial Reporting Fraud, 28 Cardozo L. Rev. 1611, 1644–45 (2007) (explaining how an accounting fraud sentence can quickly reach stratospheric heights).
  205. .William J. Stuntz, The Political Constitution of Criminal Justice, 119 Harv. L. Rev. 780, 782 (2006) (“Since the 1960s, the Supreme Court has regulated policing and trial procedure aggressively, while leaving substantive criminal law . . . to the politicians.”).
  206. .See, e.g., United States v. Alvarez, 567 U.S. 709, 729–30 (2012) (striking down the Stolen Valor Act for its interference with free speech).
  207. .“A criminal statute is [vague] if it ‘fails to give a person of ordinary intelligence fair notice that his contemplated conduct is forbidden.’” United States v. Batchelder, 442 U.S. 114, 123 (1979) (citation omitted). Although less prominent in the vagueness literature, the Supreme Court has held that the doctrine pertains not only to the fact of criminal prohibition but also to its consequences. Id. (“[V]ague sentencing provisions may pose constitutional questions if they do not state with sufficient clarity the consequences of violating a given criminal statute.”).
  208. .Hopwood, supra note 41, at 700 (promoting a “clear-statement rule that would apply systematically to a lack of clarity in federal criminal law––to statutory provisions that are vague, ambiguous, or both”). For an earlier argument in favor of clear statement rules, see Joseph E. Kennedy, Making the Crime Fit the Punishment, 51 Emory L.J. 753, 763 (2002) (proposing an interpretive rule requiring “clear evidence of congressional deliberation about the punishment to be imposed on offenders whose moral culpability is not clear”).
  209. .Smith starts by proposing a default application of the rule of lenity to all arguably ambiguous statutes. Smith, supra note 57, at 579–81 (citing the benefits of a “rigidly enforced rule of lenity”). He then adopts the following as a fallback position: “In cases where an expansive interpretation would threaten to visit disproportionate punishment on convicted offenders . . . a narrow reading is the appropriate response unless the plain meaning of the statute commands a broader interpretation.” Id. at 581–82.
  210. .Smith’s interpretive rule explicitly calls for a narrow reading “unless the plain meaning of the statute commands a broader interpretation.” Smith, supra note 57, at 581–82 (emphasis added).
  211. .At oral argument, Justice Ginsburg inquired whether the United States Attorneys’ Manual guided prosecutors in their choice of overlapping statutes. Transcript of Oral Argument at 28, Yates v. United States, 135 S. Ct. 1074 (2015) (No. 13-7451). Upon hearing that prosecutors were expected to charge the most serious provable offense, Justice Scalia responded, “Well, if that’s going to be the Justice Department’s position, then we’re going to have to be much more careful about how extensive statutes are.” Id. at 28–29.
  212. .See, e.g., Ewing v. California, 538 U.S. 11, 23 (2003) (citing an earlier Kennedy opinion for the contention that the Eighth Amendment “does not require strict proportionality” between a crime and its punishment, but instead “forbids only extreme sentences that are ‘grossly disproportionate’ to the crime” (quoting Harmelin v. Michigan, 501 U.S. 957, 1001 (1991) (Kennedy, J., concurring))).
  213. .United States v. Odeneal, 517 F.3d 406, 414 (6th Cir. 2008).
  214. .See U.S. Sentencing Comm’n, supra note 112, at app. B (showing that one-sixth of offenders received no sentence of imprisonment and that a median sentence for fraud crimes was twenty-four months’ imprisonment).
  215. .Proportionality review is itself difficult to implement because what is proportional is itself a subjective normative judgment. Avlana K. Eisenberg, Mass Monitoring, 90 S. Cal. L. Rev. 123, 138 (2017) (“To a large extent, the proportionality of a punishment is in the eye of the beholder, and criminal justice norms differ across individuals and jurisdictions.”).
  216. .In his argument in support of “proportional mens rea,” Michael Serota similarly contrasts radical and more incremental treatments. See Serota, supra note 10, at 1220–21 (explaining the differences between “thick” and “thin” models). Because scholars already use these terms to describe the codes themselves, see supra note 5 and accompanying text, this Article avoids such terminology in its discussion of potential reforms.
  217. .For classic examples of the law-and-economics school’s approach to criminal punishment, see generally Posner, supra note 96 (developing an economic analysis of substantive criminal law); Shavell, supra note 162 (analyzing the use of nonmonetary sanctions as deterrents in the context of criminal law).
  218. .Alice Ristroph, How (Not) to Think Like a Punisher, 61 Fla. L. Rev. 727, 737 (2009). Ristroph was, ironically, describing the Model Penal Code’s proposed sentencing reforms. Id. For more on the moral contours of criminal punishment, see Green, supra note 1, at 1547 (dividing criminal law’s moral content into “three broad and often overlapping categories referred to as: (1) culpability, (2) social harmfulness, and (3) moral wrongfulness”).
  219. .See, e.g., Paul H. Robinson & Robert Kurzban, Concordance and Conflict in Intuitions of Justice, 91 Minn. L. Rev. 1829, 1830 (2007).
  220. .Green & Kugler, supra note 181, at 34, 37.
  221. .See supra note 6 (describing convoluted considerations set forth by statute for charging computer, credit card, and identification fraud under 18 U.S.C. §§ 1028–1030 (2012)).
  222. .For example, one winds up with a harsher punishment for trafficking in a birth certificate or driver’s license or in any document issued by a federal authority. 18 U.S.C. § 1028(b)(1)(A) (2012) (subjecting an offender to fifteen years’ maximum imprisonment instead of five). Meanwhile, the computer fraud’s penalty provisions are so numerous and confusing that the Department of Justice has organized them into its own helpful table to assist local prosecutors. Prosecuting Computer Crimes, supra note 6, at 3 (delineating nine different crimes and more than twenty different statutory maximums, depending on the relevant circumstances).
  223. .Some of the state codes that punish fraud tend to grade statutes according solely to the amount of loss involved. See, e.g., N.Y. Penal Law §§ 177.05–.25 (McKinney 2010) (defining health-care fraud in five degrees in accordance with the amount defrauded over a single year).
  224. .According to Professor Simons, we can compare different levels of mens rea (i.e., purposeful versus reckless behavior) and different degrees of harm (serious injury versus death), but we have a difficult time ranking crimes when both variables are in play. Simons, supra note 98 (describing incommensurability as pervasive in criminal law).
  225. .Id.
  226. .The federal code’s fraud statutes currently punish schemes to defraud; as such, they punish inchoate behavior. See Geraldine Szott Moohr, The Balance Among Corporate Criminal Liability, Private Civil Suits, and Regulatory Enforcement, 46 Am. Crim. L. Rev. 1459, 1467 (2009) (observing that mail fraud is an inchoate offense and effectively “a crime of attempt”).
  227. .Cahill, supra note 81, at 896–901 (describing the near-universal rejection of an attempted-reckless-murder offense because the concept of attempt is incompatible with an unintended result).
  228. .The State of New York’s code, for example, hosts four identity-theft statutes, three of which are felonies and one of which is a misdemeanor. N.Y. Penal Law §§ 190.77–.80-a (McKinney 2010). It also includes five health-care-fraud statutes (four felonies and one misdemeanor). Id. §§ 177.05–.25. In an article constructing a hypothetical grading scheme for arson, Professor Cahill effortlessly generates seven degrees. Michael T. Cahill, Grading Arson, 3 Crim. L. & Phil. 79, 83 (2009) (observing that “a scheme recognizing seven offense levels—say, four or five degrees of felony, and two or three degrees of misdemeanor—would be in the mainstream of current American offense-grading systems”).
  229. .Ronald L. Gainer, Remarks on the Introduction of Criminal Law Reform Initiatives, 7 J.L. Econ. & Pol’y 587, 589–90 (2011).
  230. .Erin Murphy, Writing on an Unclean Slate: Challenges in Substantive Reform of a Penal Code 2–4 (2017) (Theorizing Criminal Law Reform, Conference Paper), https:/‌/‌‌sites/law/files/attachments/Murphy%20%20Writing%20on%20an%20Unclean%20Slate%20Challenges%20in%20Substantive%20Reform%20of%20a%20Penal%20Code.pdf [https:/‌/‌‌2H3F-GN8J] (drawing upon her experience as Associate Reporter in the American Law Institute’s project to reform the Model Penal Code’s sexual offense provisions).
  231. .Id. at 5 (describing the “essentially arbitrary task of fixing criminal punishments”).
  232. .Gainer, supra note 229, at 590.
  233. .On count-stacking generally, see Gold et al., supra note 106, at 1619 (explaining how count-stacking increases prosecutorial leverage to induce guilty pleas).
  234. .See Crespo, supra note 172, at 1313 & n.31 (describing the practice, which he refers to as “piling on,” and surveying scholarship critiquing it).
  235. .Id. at 1315–17.
  236. .For count-stacking reforms generally, see Seigel & Slobogin, supra note 106, at 1128–30 (proposing a “law of counts” that would involve judicial review of prosecutorial charging decisions aimed at curtailing “redundant” or excessive charging).
  237. .See Darryl K. Brown, Criminal Law’s Unfortunate Triumph over Administrative Law, 7 J.L. Econ. & Pol’y 657, 670 (2011) (“[S]impler, smaller-scale strategies exist for substantially improving federal criminal law.”).
  238. .“Misdemeanors, under federal law, are defined as any offense carrying a term of imprisonment not to exceed one year, 18 U.S.C. § 3581(b), and a fine not to exceed $100,000, 18 U.S.C. § 3571(b).” United States v. Talkington, 32 F. Supp. 2d 1262, 1263 (D. Kan. 1998). Statutes that impose a one-year maximum term of imprisonment are Class A misdemeanors; those that cap imprisonment at no more than six months are Class B misdemeanors; and those that cap punishment at thirty days or less are Class C misdemeanors. 18 U.S.C. § 3559 (2012). For a Class A misdemeanor not resulting in death, the maximum fine is $100,000; for Class B and C misdemeanors, the maximum fine drops to $5,000. 18 U.S.C. § 3571 (b)(5)–(6) (2012).
  239. .On differences in collateral consequences for felonies and misdemeanors, see Kelly Holt, Comment, Congressional Guidance on the Scope of Magistrate Judges’ Duties, 84 U. Chi. L. Rev. 909, 943–44 (2017).
  240. .See, e.g., Alexandra Natapoff, Misdemeanors, 85 S. Cal. L. Rev. 1313, 1315–16 (2012) (citing numerous problems with the misdemeanor system, most notably, defendants’ access to justice); see also Paul T. Crane, Charging on the Margin, 57 Wm. & Mary L. Rev. 775, 778–80 (2016) (analyzing ways in which prosecutors can strategically benefit by purposely filing misdemeanor over felony charges); Irene Oritseweyinmi Joe, Rethinking Misdemeanor Neglect, 64 UCLA L. Rev. 738, 741 (2017) (illustrating harms caused by misdemeanor charges); Issa Kohler-Hausmann, Managerial Justice and Mass Misdemeanors, 66 Stan. L. Rev. 611, 613–14 (2014) (using a systemic study of New York City courts to demonstrate the phenomenon of prosecutors and police “flooding” courts with mass misdemeanors, thereby employing misdemeanors as a means of “managing people over time through engagement with the criminal justice system”).
  241. .Crane, supra note 240, at 778 (“[L]egislatures have increasingly attached severe collateral consequences to misdemeanor offenses—consequences that formerly were triggered only by felonies.”); see also Eisha Jain, Prosecuting Collateral Consequences, 104 Geo. L.J. 1197,
    1200–01 (2016) (
    examining prosecutorial discretion in light of criminal law’s far-reaching collateral consequences).
  242. .See Crane, supra note 240, at 781 (discussing the “acute docket pressures” in misdemeanor courts and “voluminous caseloads” of defense attorneys staffed on misdemeanor cases).
  243. .See generally U.S. Sentencing Comm’n, Table of Federal Misdemeanors (2012), https:/‌/‌‌sites/‌default/‌files/‌pdf/‌training/‌annual-national-training-seminar/‌2012/‌4_Table_Federal_Misdemeanors.pdf [https:/‌/‌‌U57X-BCAJ].
  244. .See, e.g., 18 U.S.C. § 46 (2012) (penalizing the transportation of water hyacinths); 18 U.S.C. § 710 (2012) (regarding cremation urns for military use); 18 U.S.C. § 916 (2012) (criminalizing false impersonation of 4-H club members). For more on the federal code’s embarrassingly trivial crimes, see Green, supra note 1, at 1540 (discussing the “much lampooned” crime of tearing tags off mattresses); Stuntz, supra note 1, at 517 (citing a statute criminalizing misuse of the “Woodsy Owl” image).
  245. .E.g., 18 U.S.C. § 111 (2012) (penalizing simple assault of federal officials); 18 U.S.C. § 113 (2012) (punishing assault within maritime and territorial jurisdiction).
  246. .See 18 U.S.C. § 1028(b)(6) (2012) (providing the misdemeanor offense for identification fraud); 18 U.S.C. § 1030(c)(2)(A) (2012) (providing the misdemeanor offense for fraud arising from intentional unauthorized access of a computer).
  247. .U.S. Dep’t of Justice, Justice Manual § 9-27.430 cmt. 1 (2018) (“The charge or charges to which a defendant pleads guilty should be consistent with the defendant’s criminal conduct, both in nature and in scope.”).
  248. .See supra note 112 and accompanying text. According to the Sentencing Commission staff’s own data, these low sentences reflect the relative modesty of the charged schemes. See U.S. Sentencing Comm’n, Economic Crime Public Data Briefing fig.4 (2015), https:/‌/‌‌sites/‌default/‌files/‌pdf/‌amendment-process/‌public-hearings-and-meetings/‌20150109/‌fraud_briefing.pdf [https:/‌/‌‌LJQ2-GRZ2] (indicating for fiscal year 2012 that 53.9% of § 2B1.1 offenders fell within loss categories involving less than $120,000).
  249. .U.S. Sentencing Comm’n, 2017 Sourcebook of Federal Sentencing Statistics tbl.13, https:/‌/‌‌sites/‌default/‌files/‌pdf/‌research-and-publications/‌annual-reports-and-sourcebooks/‌2017/‌Table13.pdf [https:/‌/‌‌RA6M-SPUD].
  250. .U.S. Sentencing Comm’n, Use of Guidelines and Specific Offense Characteristics: Offender Based Fiscal Year 2017, at 10, https:/‌/‌‌sites/‌default/‌files/‌pdf/‌research-and-publications/‌federal-sentencing-statistics/‌guideline-application-frequencies/‌2017/‌Use_of_SOC_Offender_Based.pdf [https:/‌/‌‌5P4U-44FV] (indicating 14.7% of convicted fraud offenders were responsible for actual or intended loss amounts of $6,500 or less).
  251. .On the disappearance of trials from the federal criminal system, see Darryl K. Brown, How to Make Criminal Trials Disappear Without Pretrial Discovery, 55 Am. Crim. L. Rev. 155, 155 & n.1 (2018) (parsing data indicating 97% of federal criminal cases resolve in guilty pleas). For a broader overview, see generally Marc Galanter, The Vanishing Trial: An Examination of Trials and Related Matters in Federal and State Courts, 1 J. Empirical L. Stud. 459 (2004).
  252. .See supra subpart II(B).
  253. .Regardless of how a prosecutor charges a crime, a defense attorney can request at trial a jury instruction advising of the existence of a lesser offense, which permits the jury to convict the defendant of a less serious crime. A lesser offense is one whose elements form a complete subset of the more serious charged offense. See Fed. R. Crim. P. 31(c) (“A defendant may be found guilty of . . . an offense necessarily included in the offense charged . . . .”); Schmuck v. United States, 489 U.S. 705, 716, 719–21 (1989) (“To be necessarily included in the greater offense the lesser must be such that it is impossible to commit the greater without first having committed the lesser.” (internal quotation marks and citation omitted)); Carter v. United States, 530 U.S. 255, 258–59 (2000) (citing Schmuck’s “subset of the elements” test in holding that 18 U.S.C. § 2113(b) is not a lesser included offense of § 2113(a)). Because a lesser offense instruction expands the jury’s menu of options, it affords the defendant a valuable right that potentially restrains the prosecutor from overcharging the defendant. See Crespo, supra note 172, at 1364–66 (explaining relevant incentives).
  254. .United States v. Adelson, 441 F. Supp. 2d 506, 512 (S.D.N.Y. 2006) (commenting on the distinction between a “belated entrant” into a fraud conspiracy and an “active leader”).
  255. .United States v. Corsey, 723 F.3d 366, 377 (2d Cir. 2013) (Underhill, J., concurring) (arguing that the Guidelines’ analysis premised on intended-loss amount is “valueless” because the fraud scheme at issue was “more farcical than dangerous”).
  256. .See supra note 253 and accompanying text; Schmuck, 489 U.S. at 717 n.9.
  257. .Congress can require this of the defendant without violating his due process rights so long as the affirmative defense does not negate a required element of the statute. Compare Patterson v. New York, 432 U.S. 197, 201, 210 (1977) (holding that requirement that defendant prove affirmative defense did not deprive him of his due process rights), with Mullaney v. Wilbur, 421 U.S. 684, 703–04 (1975) (concluding that where a Maine statute required the prosecutor to prove that the defendant acted without provocation, the court could not require the defendant to prove his heat-of-passion defense).
  258. .People v. Eboli, 313 N.E.2d 746, 747 (N.Y. 1974) (“[E]xcept for a minor variation in language, the crimes of coercion in the first and second degree are identical when the coercion is committed by instilling a fear that a person will be physically injured or that property will be damaged.”) Similar dynamics have been observed and unsuccessfully challenged in regard to other parts of the New York penal code. See People v. Vicaretti, 388 N.Y.S.2d 410, 413 (N.Y. App. Div. 1976) (“While it may be true that under certain circumstances the crimes of rape in the first degree and sexual misconduct may be identical, that fact alone does not . . . amount to a denial of equal protection.”).
  259. .Eboli, 313 N.E.2d at 749 (advising that statutory overlaps “and the opportunity for prosecutorial choice they represent, is no bar to prosecution”).
  260. .For a recent affirmation of the Eboli court’s approach in New York, see People v. Finkelstein, 68 N.E.3d 64, 66 (N.Y. 2016). For courts that have rejected the New York view, see, e.g., People v. Estrada, 601 P.2d 619, 621 (Colo. 1979) (en banc) (“We find a penalty scheme that provides widely divergent sentences for similar conduct and intent to be irrational . . . .”).
  261. .See United States v. Tarango, No. CR 07-2443, 2015 WL 10401775, at *24 (D.N.M. Oct. 29, 2015) (“Everywhere one turns, one can see a judge, sentencing commission, Attorney General, Senator, Representative, academic, or layperson talk about the high costs of imprisonment, overcrowded prisons, and the need to have fewer people in prisons and jails. Conservatives and libertarians have joined liberals [calling] for reductions in incarceration.”).
  262. .See Crane, supra note 240, at 780 (arguing that prosecutors choose misdemeanors because they trigger fewer procedural obligations than felonies).
  263. .As John Pfaff’s work shows, a relatively small number of the nation’s prisoners are federal prisoners, and the cost of prosecuting and punishing them represents a tiny percentage of the annual budget. See John F. Pfaff, Federal Sentencing in the States: Some Thoughts on Federal Grants and State Imprisonment, 66 Hastings L.J. 1567, 1573, 1576–77 (2015) (“[T]he federal system simply is not that large.”).
  264. .In some respects, federal misdemeanants already enjoy more rights than state and local misdemeanants. See Crane, supra note 240, at 805 (“[I]n the federal system most (though not all) discovery rules apply equally to [misdemeanants and felony offenders].”).
  265. .This apparently was a problem for environmental offenses, despite the fact that Congress upgraded certain misdemeanors to felonies in the 1990s. See Judson W. Starr, Turbulent Times at Justice and EPA: The Origins of Environmental Criminal Prosecutions and the Work that Remains, 59 Geo. Wash. L. Rev. 900, 909 (1991) (“[T]he FBI had little interest in supporting criminal enforcement under the other environmental statutes that still provided only for misdemeanor criminal penalties.”).
  266. .On the history of the idiom, see Judith Resnik, Revising Our “Common Intellectual Heritage”: Federal and State Courts in Our Federal System, 91 Notre Dame L. Rev. 1831, 1859 & n.140 (2016).
  267. .See, e.g., U.S. Sentencing Guidelines Manual § 2B1.1(b)(1)(A)–(D) (U.S. Sentencing Comm’n 2016) (providing guidelines for theft and fraud cases involving loss amounts less than $95,000); id. § 2B1.5(b)(1) (providing guidelines for theft or harm of cultural resources worth less than $6,500 but more than $2,500); id. § 2B2.1(a)(2), 2B2.1(b)(2)(A)–(C) (providing guidelines for burglaries of nonresidences and involving amounts of less than $95,000); id. § 2B2.3(b)(3) (providing guidelines for trespass of a protected computer resulting in a loss less than $6,500); id. § 2B3.3(b)(1) (providing guidelines for blackmail in which the greater of the amount obtained or demanded exceeded $2,500 but was less than $6,500).
  268. .See supra notes 112–13.
  269. .Yates v. United States, 135 S. Ct. 1074, 1101 (2015) (Kagan, J., dissenting).