Sentencing Co-Offenders

ABSTRACT

Tort law and criminal law are the two main vehicles utilized by the state to deter wrongful behavior. Despite the many similarities between the two legal fields, they differ in their treatment of collaborations. While tort law divides liability among joint-tortfeasors, criminal law abides by a no-division rule that imposes on each co-offender the full brunt of the sanction. Thus, each of two offenders who jointly steal $1,000, will be subject to the full corresponding penalty (rather than the divided penalty for stealing $500).

This Article demonstrates that in property and financial crimes, the no-division regime of criminal law harms both offenders and victims. Specifically, it creates three troubling distortions that have been overlooked by theorists and judges. First, the no-division rule violates the frugality principle, which mandates that sanctions be kept to the minimum level necessary to prevent the offense. Second, it disadvantages vulnerable victims, while favoring the well-to-do. Third, it prompts all potential victims to engage in excessive procurement of private precautions, to the detriment of society at large. The Article proposes two possible solutions to these problems. One option is to substitute the no-division rule with a division regime similar to the one endorsed by tort law. Alternatively, if lawmakers opt to retain the no-division rule, they can, and should, allocate greater public resources to the protection of the vulnerable.

Introduction

Criminal law bears important similarities to tort law. Both domains are designed to spur actors to behave in a socially desirable manner, by forcing them to internalize the cost of harms they inflict on others.^[1] As Judge Posner and others have observed, the two disciplines “frequently overlap. . . . [T]he crime of theft is the tort of conversion; the crime of assault is the tort of battery—and the crime of fraud is the tort of fraud.”^[2] It is thus not surprising that there are many doctrinal commonalities between the two fields.^[3]

Yet, tort law and criminal law critically diverge in their approach to transgressions and offenses perpetrated by more than one actor. When a crime is committed by co-offenders, criminal law ignores one of the most fundamental teachings of tort law. Under tort law, if multiple tortfeasors jointly harm a victim, their liability will be split.^[4] To be sure, vis-à-vis the victim, their liability is often joint and several, meaning that the victim can recover the full amount from either of them. But their duty to compensate is shared.^[5] Any portion of the harm paid by one tortfeasor will be deducted from the other’s liability. Each of two (similar) tortfeasors jointly liable for stealing $1,000 will ultimately pay $500 in damages.

Criminal law, by contrast, treats co-offenders as if each inflicted the full harm on the victim. Insofar as punishment is concerned, criminal law treats each of the co-offenders as though she committed the crime alone.^[6] Accordingly, if two offenders jointly steal $1,000, each will be subject to the penalty prescribed for stealing $1,000 (not $500).^[7] Given that both criminal law and tort law are designed to deter wrongful behavior, this difference is puzzling. Why would liability for harm be indivisible under criminal law if tort law always divides liability between joint tortfeasors? While our example focuses on theft, the question arises in a host of offenses in which the motivation for illicit behavior is the attainment of tangible goods.

Criminal law’s adherence to the no-division rule is more than a theoretical puzzle. Practically, it raises three important concerns. First, the rule of no-division is at odds with the fundamental “principle of frugality.”^[8] Moral dictates generally mandate that sanctions should not exceed the level needed to deter offenders.^[9] As Jeremy Bentham famously explained: “[T]he punishment ought in no case to be more than what is required . . . since, if it be, all that is above that quantity is needless.”^[10] But the no-division rule means that co-offenders are subject to excessive sanctions. Suppose that criminal law makers—the legislature or the courts—determine that the optimal penalty for deterring stealing $100,000 is a two-year imprisonment term, whereas the sanction for stealing $50,000 is one year in prison. Under the frugality principle, to deter two offenders from jointly stealing $100,000, it suffices to impose a penalty of one year in prison on each of them. After all, each expects to pocket only $50,000. Yet given criminal law’s no-division rule, if they commit the crime together, each offender would be sentenced to a two-year imprisonment term.

Even more troubling, the no-division rule also disadvantages vulnerable victims. A second concern emerging from the no-division rule is its disparate impact on underprivileged members of society. Collaboration among offenders is often a necessity when targeting well-off victims.^[11] However, the no-division rule makes collaboration particularly unattractive. Offenders operating together must share their gains. Each offender only pockets some of the proceeds obtained from the crime. But under the no-division rule, an apprehended co-offender is subject to the full, undivided penalty. The no-division rule thus imposes a “tax” on offenders’ collaboration, thereby providing greater protection to the well-off.

To gain a quick insight into how the no-division rule disfavors the vulnerable, consider two burglars who face the choice between stealing from a bank or from two small mom-and-pop shops. The bank is a large establishment, so a sole offender trying to break into the safe may be unable to see a police officer nearing the building until it is too late. The burglary thus requires two offenders. In the mom-and-pop shops, by contrast, a single burglar can monitor the entrance as he is breaking into the cashbox. Therefore, each offender can burglarize a mom-and-pop shop by himself. Suppose further that the amount deposited in the bank’s safe is $2,000, whereas the cashbox in a mom-and-pop shop contains $1,000. From the offenders’ standpoint, breaking into the bank (together) or into a mom-and-pop shop (alone) provides the same expected individual gain: $1,000. But since under the no-division rule the penalty for stealing $2,000 (together) is greater than for stealing $1,000 (alone), breaking into the mom-and-pop shops represents a more attractive option—the same payoff, with a lower sanction. The offenders will thus target the shops, leaving the bank unharmed.

This brings us to the third concern. Not only does the no-division rule disfavor poor victims, but it also reduces overall social welfare. Particularly, we suggest that the no-division rule induces potential victims to invest in private precautions. Current estimates evaluate victims’ annual expenditures on private precautions in the range of “$160 billion to $300 billion,” a figure exceeding “the entire public law enforcement budget.”^[12] But one may wonder why victims invest so much in precautions. After all, a home-security system does not prevent a break-in; it only makes it somewhat more complicated. We contend that the actual attractiveness of these private precautions often stems from their effect on offenders’ need to collaborate. A home security system requires offenders to team up. But this in itself does not change much if the expected gain from the break-in is sufficiently large. The real bite comes from the no-division rule. As we have seen, the no-division rule renders criminal collaboration particularly unattractive. While co-offenders must share the proceeds from crime, each will bear the entire penalty alone if apprehended. Thus, investment in private precautions allows victims to significantly raise the cost of crime by forcing collaboration among offenders. Because private precautions often only displace crime from one victim to another, other potential victims must follow suit and similarly invest in such measures.^[13] Ultimately, the no-division rule induces large investments that result in little social benefit.^[14]

Against this backdrop, this Article offers a new approach to the sentencing of co-offenders. We claim that the rule of no-division is not always justified. As we explain, splitting penalties among co-offenders in economically motivated crimes might be desirable for both offenders and victims. Particularly, the sentence should be split in accordance with the offenders’ individual share of the overall gains. The application of a division regime will eliminate all three distortions. We also appraise this solution in view of retributive goals, showing that these goals should not challenge our proposal. But even if the legal system chooses to retain the current no-division rule, its implications must be accounted for. We thus offer a second solution. We delineate how the government can neutralize the perverse effects of the no-division rule by allocating public resources to the economically disadvantaged.

This Article proceeds as follows. Part I describes the sentencing process set by the Federal Sentencing Guidelines and the centrality of the victim’s loss and offender’s gains in meting out penalties. Part II points to the systematic adherence to the no-division rule in property and financial crimes, showing its blanket application in both loss-based and gain-based sentences. Part III explores the implications of the no-division rule. Particularly, it shows that the rule gives rise to three distortions that harm both offenders and victims. Part IV discusses two possible ways to rectify these distortions: replacing the no-division rule with a regime that incorporates tort law’s division regime, and diverting governmental efforts to the less-privileged.

I. The Role of Loss and Gain in Sentencing

This Part elaborates on the principles set by the Federal Sentencing Guidelines in the context of property and financial crimes. Sentencing for such offenses is primarily driven by the loss to the victim. In some cases, the offender’s gain—rather than the victim’s loss—is the key determinant of the sentence. Importantly, in both categories, when offenders collaborate, each is subject to the sanction prescribed by the Guidelines for the full loss, or the full gain. The no-division rule accordingly applies in both loss-based and gain-based sentencing.

Section A begins by describing the sentencing process and its basic elements. Section B demonstrates that in the context of property and financial crimes, the victim’s loss is the predominant factor determining the sentence for most offenses. Section C then surveys the instances in which the sentence hinges on offenders’ gains.

A. The Federal Sentencing Guidelines

The organizing principle of the Federal Sentencing Guidelines,^[15] promulgated in accordance with the Sentencing Reform Act of 1984,^[16] is to establish a sentencing range for offenses based on two parameters: the gravity of the offense and the criminal history of the defendant.^[17]

The Guidelines achieve this goal by referring courts to a table known as the Sentencing Table.^[18] Each criminal offense is assigned a base offense level, which is a simple numeric value.^[19] The graver the offense, the higher the number. The base offense level is then increased or decreased based on specific offense characteristics listed in the Guidelines, such as whether or not a dangerous weapon was brandished,^[20] the quantity of drugs sold or manufactured,^[21] whether the crime was intended to harm the government,^[22] and the like. The Guidelines also prescribe offense level adjustments based on factors such as the victim’s vulnerability and the defendant’s role in the offense.^[23]

After the offense level has been adjusted, the court is left with a number that reflects the seriousness of the offense. The vertical axis of the Sentencing Table contains forty-three possible offense levels—the more serious the crime, the higher the offense level.^[24]

The horizontal axis of the Sentencing Table corresponds to the defendant’s criminal history^[25] and contains six criminal history categories. The rubric at the intersection between the vertical and horizontal axes provides a range of months of incarceration.^[26] This is the defendant’s sentencing range.

Until 2005, courts were obligated to hand down a sentence within a properly calculated sentencing range.^[27] In its 2005 United States. v. Booker decision, the Supreme Court found this requirement unconstitutional, and made the Guidelines advisory.^[28] Consequently, courts are currently allowed to depart from the Guidelines. Nonetheless, the Guidelines remain extremely significant in sentencing decisions.^[29] To begin with, courts must still commence the sentencing process by calculating the applicable Guidelines’ range.^[30] This mandate was imposed “to secure nationwide consistency.”^[31] The effect of this requirement is to anchor both the trial court’s discretion and the appellate review process.^[32] Second, if a sentence is within the Guidelines’ range, an appellate court may apply a presumption of reasonableness.^[33] Conversely, if the sentence is outside the Guidelines’ range, an appellate court may consider the degree of variance when reviewing the reasonableness of the sentence.^[34] Not surprisingly, the empirical evidence is clear: although courts are formally allowed to depart from the Guidelines, they adhere to them in the vast majority of cases,^[35] and the Guidelines remain “the lodestone of sentencing.”^[36]

B. The Role of the Victim’s Loss in Sentencing

Within the framework of the Guidelines, the most important determinant of penalties in property and financial crimes is the victim’s loss.^[37] The base offense level for such crimes is normally relatively moderate.^[38] For example, the base offense level for larceny, embezzlement and other forms of theft is six or seven,^[39] which calls for a sentence of zero to six months (assuming the defendant has no criminal record).^[40] As the loss to the victim increases, so does the offense level. If the loss to the victim exceeds $6,500, the offense level is increased by two points.^[41] If the loss exceeds $15,000, the offense level is increased by four points,^[42] and so on. There are no less than sixteen possible offense-level categories corresponding to different amounts of loss. In the highest loss category the base offense level is increased by no less than thirty points, which brings the sentencing range up to 188–262 months of incarceration.^[43]

Similarly, the base offense level for insider trading is eight.^[44] The corresponding sentencing range is zero to six months.^[45] Depending on the loss suffered, the Guidelines set sixteen different offense levels.^[46] The largest loss amount brings the offense level to thirty-eight, calling for a sentence of between 235–293 months of incarceration.^[47]

This pattern is the general rule for property and financial crimes: burglary,^[48] robbery,^[49] antitrust violations,^[50] and other related offenses all have a relatively low base offense level and a large number of possible increases corresponding to the victim’s loss. The ultimate sentencing range is determined predominantly by the amount of loss to the victim.

In fact, the effect of the victim’s loss on sentencing is even more pronounced than the preceding analysis suggests. The reason is that the larger the victim’s loss, the greater its impact on the sentence. At the lower end of the Sentencing Table, an increase of one point in the offense level only raises the sentence-range by several months. At the higher end, the same increase in offense level raises the range by several years.^[51] As Professor Bowman noted: “because of the logarithmic character of the 43-level Sentencing Table, each increase in offense level has an ever-greater absolute effect on sentence length, the higher one goes up the Table.”^[52]

The significance of the victim’s loss is evident not only standing alone, but also in comparison to the effect of other factors that may impact sentencing. Most of the additional factors enumerated in the Guidelines, if relevant, increase the offense level by a mere two to four points. Others may increase the offense level by more, but no other factor carries a potential offense-level increase of anything close to thirty points.^[53] The amount of loss to the victim overshadows all other factors in determining the sentence.

Not only is the victim’s loss the factor that bears the largest effect on sentencing, it is also invoked far more often than all other factors. A recent survey found that for offenders who were convicted of committing “basic economic crimes” (i.e., property and financial offenses), more than 85% were subjected to an enhancement of the offense level based on the victim’s loss.^[54] This was by far the most common offense characteristic invoked—it was nearly four times more common than the second most-common specific offense characteristic.^[55] For 50% of the convicted defendants, this was the only offense characteristic applied.^[56] For this reason as well, the victim’s loss is the key determinant of the Guidelines sentencing range.^[57]

While the focus of this Article is on the federal regime, much of the analysis applies to state sentencing as well. States differ in their sentencing mechanisms. Some have adopted guidelines that resemble the federal ones. Others have different mechanisms. Nevertheless, in all states the severity of punishment for at least some property and financial offenses hinges on the loss to the victim.^[58]

C. The Role of the Offender’s Gain in Sentencing

In some cases, the offender’s gain, rather than the victim’s loss, serves as the key sentencing determinant. This differing focus occurs in three different settings: First, when the victim’s loss cannot be determined, the offender’s gain substitutes for it as the sentencing touchstone. Second, for some property and financial offenses, the Guidelines set the offender’s gain as the yardstick for the gravity of the offense. Third, the offender’s gain becomes central when no tangible harm was caused to the victim, but the offender profited from an offense. We elaborate on these in order.

The first category involves settings in which loss was inflicted but cannot be assessed. Under the Guidelines, “[t]he court shall use the gain that resulted from the offense as an alternative measure of loss only if there is a loss but it reasonably cannot be determined.”^[59]

The second category in which the Guidelines set the offender’s gain as the determinant of the gravity of the offense includes a number of property and financial offenses. Gain-based sentencing is the rule for insider trading; bribery; commercial bribery; and offering, giving, soliciting or receiving a prohibited gratuity.^[60] Particularly, when a defendant is convicted of insider trading, the Guidelines instruct courts to determine the specific offense level based on the gain resulting from the offense (irrespective of the victim’s loss).^[61] In bribery cases, the offense level may be increased by up to thirty points based on “the benefit received,” when this benefit exceeds the loss to the government.^[62] Similarly, in commercial bribery cases, the base offense level is increased based on the greater of the value of the bribe or the improper benefit that was to be conferred.^[63] Likewise, when a prohibited gratuity is given, the offense level is determined based on the benefit conferred on the recipient (namely, her gain) rather than on the harm to the victim.^[64]

The third category encompasses offenses that involve a fraudulent scheme.^[65] Specifically, cases in which the offender rendered services or goods to the victim while falsely posing as a licensed professional, or while falsely representing the goods as approved by a governmental agency. In such cases, the sentence is based on the amount paid to the offender.^[66] Note that at least from the victim’s standpoint, the offender’s behavior often causes no tangible harm.^[67] The victim may well have received the service, the goods, or the property for which she bargained.^[68] The amount paid, however, is an approximation of the offender’s illicit gain resulting from the fraud.

II. The No-Division Rule

The preceding Part explained the dominance of the victim’s loss and the offender’s gain in determining sentences for property and financial crimes. Given the importance of the loss or gain, a crucial question in the sentencing context is whether losses and gains should be divided between co-offenders. If four co-offenders steal $100, to be split among them equally, should each be sentenced as if she stole $100, or as if she stole $25? Under the no-division rule, each co-offender will be sentenced as if she alone stole the full $100.

The Guidelines endorse the no-division rule explicitly. Per the Guidelines, the defendant’s sentence is to be determined based on the overall harm (or, when applicable, the overall gain) that resulted from the defendant’s conduct or from that of her co-offenders.^[69]

In accordance with the Guidelines’ instructions, courts have consistently held that losses and gains are indivisible in the sentencing context. The issue was explicitly addressed in United States v. Allen.^[70] In Allen, the defendant and two codefendants engaged in a fraud, selling insurance policies, without actually purchasing the policies from insurers. Instead, they pocketed the premiums paid by the victims.^[71] Allen, the defendant, argued that he was to receive only one-third of the proceeds, as the earnings were to be divided with his co-offenders. The Eighth Circuit rejected the argument, reiterating the Guidelines’ no-division rule:

It appears to us that Mr. Allen’s second “objection” was actually an inquiry with respect to a question of law—i.e., whether the amount of loss to the victims, for purposes of determining his offense level, depended on how much of the proceeds he personally received. Under the sentencing guidelines, the answer to that question . . . was that the amount of money paid by the victims . . . plus the amount of money owed to the victims for outstanding claims—and not the amount of money received by Mr. Allen individually—were the critical figures.^[72]

The no-division principle has been applied even in cases involving a large number of defendants. For example, in United States v. Riley, seven individuals conspired to pass fictitious financial instruments.^[73] The court sentenced the defendant, Riley, based on the total amount of loss, rather than on a fraction of this amount.^[74] And in United States v. Gaye, each of the twenty-two defendants were sentenced for the full amount of loss resulting from their fraud.^[75]

The no-division rule has been similarly followed in gain-based sentencing. For example, in United States v. DeVegter, the Eleventh Circuit ruled that each of the two defendants were to be sentenced based on the full amount of the benefit accruing to them as a result of the bribe.^[76] Similarly, in United States v. Anchor Foods, Inc., the court sentenced each of the four defendants based on the total amount of gains from the fraudulent scheme.^[77]

There is one exception, however, to the no-division rule—antitrust violations.^[78] The Guidelines deal with a single kind of antitrust violation—cartelistic agreements; that is, agreements whereby competitors agree to restrain or eliminate competition between themselves.^[79] When antitrust offenders are sentenced, each defendant’s offense level is set based on the “the volume of commerce done by him or his principal in goods or services that were affected by the violation.”^[80] Thus, each cartel member’s offense level is based not on the overall harm caused by the cartel, but rather on her individual market share. For example, if five cartel members jointly inflict a loss of $10 million on consumers, each cartel member will be sentenced as if she caused a loss of $2 million. If cartel members have different market shares, the sentence will be calibrated accordingly.

With the exception of antitrust offenses, the no-division rule underlies the sentencing for all property and financial offences. Whether focusing on losses or on gains, each offender is sentenced as if she committed the crime alone. Furthermore, the total loss emanating from the offense or the offenders’ overall gains significantly overshadow all other factors in determining the sentence.

III. The Distortive Effects of the No-Division Rule

In Part II, we established that the no-division rule is a cornerstone of the Guidelines. The rule makes sense as applied to offenses such as assault, battery, defamation, and manslaughter, in which the satisfaction of one co-offender does not detract from that of the others. If two offenders kill a witness who is supposed to testify in their respective trials, each nets the full gain from the crime. Or to take another example, if three co-offenders defame a person, each offender’s benefit (whatever that may be) is usually not affected by that of the others. In such cases, there is absolutely no reason to divide the sanction between the offenders. Splitting the sanction would lead to underdeterrence.

The analysis changes dramatically, though, in the case of property and financial offenses. In these crimes, the offenders’ benefits have the structure of a zero-sum game. Each offender’s benefit comes at the expense of the other co-offenders. For example, if four offenders steal $100 together, the net gain of each will be $25.^[81] Yet, as shown in Part II, the no-division rule applies with equal force in this case, as well. Each offender will be sentenced as though she stole $100 from the victim and will be subject to the corresponding sanction.

This contrast is not a purely theoretical anomaly. We demonstrate that criminal law’s no-division rule, as applied to property and financial crimes, has several undesirable efficiency and distributive effects that have been overlooked by legislatures, courts, and criminal-law theorists.

First, the no-division rule is at odds with the frugality principle, according to which the sanction imposed on offenders should be the minimal penalty necessary to prevent the crime. Excessive penalties are particularly undesirable as they typically involve imprisonment. Disproportionate incarceration terms unduly deprive individuals of their freedom and consume valuable societal resources. Second, the no-division rule discriminates against vulnerable victims. It raises the cost of collaboration among offenders since they must split the proceeds from the offense, but each is subject to the full penalty under the law. Offenses against impecunious victims, which can be carried out alone, allow the perpetrator to keep the entire gain. Thus, all things being equal, the no-division rule renders property crimes against economically disadvantaged victims more attractive. Third, from a societal standpoint, the no-division rule spurs excessive investment in private precautions by victims. Private precautions often force offenders to collaborate. The no-division rule effectively allows potential victims to levy a tax on offenses against them. This, in turn, spikes similar investments by other potential victims, who fear that offenders will now target them. The ensuing arms race is a social waste. Hence, the no-division rule is a principal driver of excessive investment in private precautions.

A. Co-Offenders and the Frugality Principle

The frugality principle is a foundational tenet of criminal law. It originated with Jeremy Bentham, who famously annunciated that criminal sanctions should, as much as possible, be calibrated to the minimum necessary to prevent crime.^[82] For example, if a penalty of one year in prison suffices to prevent offenders from committing a crime that is expected to generate a gain of $1000, criminal law should not impose a stricter penalty. Excessive penalties are cruel to offenders and produce no benefit to society.^[83] Prima facie, therefore, there is no reason to deny a person her liberty for a long period of time when a shorter incarceration term can achieve the same result. Yet, under the no-division rule, this is exactly what the legal system does. If in the previous example the offense were committed by two co-offenders, each of whom had an expectation to gain $500 from the activity, a six-month imprisonment term should have sufficed to discourage them from breaking the law. Nonetheless, the no-division rule mandates an excessive sentence of one year in prison for each of the offenders.

To be sure, the loss to the victim is an important factor in sentencing decisions. A theft of $1000 cannot be subject to the same penalty as a theft of $1 million. A loss of $1 million is obviously more harmful to the victim than a loss of a $1,000. Hence, marginal deterrence requires a stricter penalty in the former case. Unsurprisingly, criminal law often focuses on the harm to the victim in meting out penalties.

In the case of co-offenders, however, focusing exclusively on the harm to the victim distorts the perspective. While marginal deterrence still requires imposing a stricter penalty as the harm to the victim increases, it does not follow that sanctions should remain uniform irrespective of the number of offenders. Co-offenders who steal $1 million should be penalized more harshly than co-offenders who steal $1,000. Yet, under the frugality principle, they should not each bear the full penalty—each should bear a penalty corresponding to her personal gain.

Nonetheless, it is possible to argue that the no-division rule is justified since co-offenders face a lower probability of detection.^[84] Acting in tandem, so the argument goes, allows offenders to benefit from division of labor. It permits certain members of the group to focus on committing the main illegal act (for example, breaking into the safe), while others stand guard and inspect their surroundings. One way to offset the benefit of lower detection is to step up the penalty on joint offenses. The no-division rule, so one could argue, counterbalances the benefits of division of labor.

While this explanation has surface appeal, it does not withstand scrutiny. To begin with, although co-offenders may face a lower probability of detection, they are subject to a higher probability of conviction when apprehended. This increased likelihood of conviction is due to the famous Prisoner’s Dilemma.^[85] When a group of offenders is interrogated by the police, there is a higher chance that individual members will rat on companions in order to reduce their own penalty.^[86] Hence, it is far from clear that committing offenses together with others reduces the overall risk of penalty.

In addition, there is no a priori reason to assume that the steeper penalties imposed on co-offenders perfectly offset the reduction in the probability of apprehension. For there to be a perfect offset, there must always be a symmetry between the increase in the penalty and the decrease in the probability of detection. Yet, the “tax” on collaboration remains uniform across the board, irrespective of the number of co-offenders and its effect on detection.

Furthermore, the lower-probability-of-apprehension argument fails to account for the interplay between collaboration and offenders’ individual gains across offenses. Recall that in offenses such as battery and assault, each perpetrator nets the full benefit of the crime. Accordingly, the full penalty for each offender in such crimes is mandated regardless of the benefit of collaboration. If collaboration decreases the chances of apprehension, the penalty for such offenses should be further enhanced and must exceed the full penalty for committing the crime alone. Yet, no such enhancement exists. Therefore, the lower-probability-of-apprehension argument fails to provide a general justification for the no-division rule.

Finally, the no-division rule comes with a significant price tag to society. Imprisonment terms are extremely costly from a societal standpoint. Reports estimate the average annual cost of incarceration per person at $37,449.^[87] In 2019, the number of inmates in the U.S. was 2.115 million.^[88] Co-offending, or group crime, is common in the U.S. Although it is difficult to estimate the precise percentage of co-offenders out of the entire inmate population, data show that “more than one-quarter of all federal criminal prosecutions and a large number of state cases involve prosecutions for conspiracy.”^[89] A back-of-the-envelope calculation reveals the potential contribution of the no-division rule to imprisonment costs. Even if one assumes that the number of co-offenders in each such case is only two, the cost of the no-division rule to society approaches $32 billion per year.^[90] The share of this amount attributable to property and financial offenses could have been allocated to other, socially more desirable, activities. This calculation only captures the quantifiable costs of the no-division rule. It does not account for the unnecessary suffering and emotional harms inflicted by the rule on inmates, their families, and communities.

B. Discrimination Against Vulnerable Victims

In Section A, we studied the effect of the no-division rule on offenders. In this section, we turn the spotlight to the distributive effects of the rule on victims.

Property crimes in the U.S. are disproportionately committed against poor victims.^[91] The no-division rule creates strong incentives to commit property crimes against impecunious victims. While other factors, such as spatial proximity, accessibility, and familiarity with one’s surroundings affect the choice of target,^[92] we contend that the no-division rule renders poor victims vulnerable to property crimes.

To see why, let’s return to the example from the Introduction. Assume an offender who contemplates breaking into a small mom-and-pop shop or a bank. Breaking into a mom-and-pop shop can be carried out without a partner, with an expected gain of $1000. Breaking into a bank requires help from a partner and the expected gain is $2000. In the relevant jurisdiction, the penalty for stealing $1000 is one year in prison and the penalty for stealing $2000 is two years. From the standpoint of the offender, the expected gain from each option is the same—$1000 (since in the second case, the gain of $2000 must be split with a partner). Yet, because of the no-division rule, the penalty is twice as harsh in the second case (a two-year imprisonment term for each co-offender). The mom-and-pop shop thus represents a much more appealing opportunity for potential offenders. Consequently, it is far more likely to be burglarized. The no-division rule thus disadvantages poor victims.

It is important to note that the regressive effect of the no-division rule only intensifies as the number of co-offenders increases. In the above example, if breaking into the bank necessitated three or four perpetrators, rather than just two, under the no-division rule the bank would hardly ever be targeted, while the mom-and-pop shop would fall prey to burglaries time and time again.

Empirical data bear out this prediction. As the renowned economist Steven Levitt has shown, crime distribution exhibits an unusual pattern.^[93] While homicides are largely distributed evenly among members of different segments of society, property offenses sadly fall disproportionately on the poor.^[94] This result poses a challenge to the standard explanations that emphasize the role of geographic proximity, accessibility, and familiarity with one’s surroundings. If these factors were indeed dispositive, arguably there should have been greater uniformity in the distribution of property crimes and murders (namely, one would expect homicides to be committed disproportionately against the poor). Yet, as noted, the data reveal a different picture.

Levitt’s findings are supported by recent data. According to the National Crime Victimization Survey (a leading source for statistics concerning crime distribution), low-income households are overrepresented as victims of financial and property crimes.^[95] For the ten most recent years in which data are available (2010–2019), households with an annual income of less than $15,000 constitute approximately 12%–13% of all households, but account for roughly 28% of domestic-burglary and trespass victims. By contrast, homes in the upper band of annual income (above $75,000), represent approximately 35% of the population, but account for a mere 17% of victims of these crimes. This pattern exists across all income levels and in all three property crimes included in the survey (burglary-trespass, theft, and motor vehicle theft).^[96]

The no-division rule can account for the observed pattern. As we explained, in the case of property crimes the gains are split between the co-offenders, while in cases of murder they are not; each perpetrator receives the full ill-gotten gain from committing this heinous crime.^[97] While conventional analysis in criminology has focused on situational and circumstantial factors as the main determinants of crime distribution, we suggest that legal doctrine also affects the selection of victims. The no-division rule makes marginalized members of society a far more attractive target of property crimes.

C. Private Investment in Precautions

In the two previous sections, we explained how the no-division rule affects offenders and actual victims. In this section, we turn to examine the effect of the no-division rule on potential victims. To explore this effect, attention must be focused on the pre-crime stage, at which potential victims invest in precautions in order to avoid being targeted.

Potential victims can reduce their exposure to crime by investing in a variety of preventive measures. For example, a homeowner can erect a fence around her property, install bars on the windows, purchase an alarm system, or place a security camera. These, and other measures, lower the risk that her home will be burglarized. While private precautions benefit the individual homeowner, they do not necessarily reduce crime altogether. The reason for this is displacement.

A variety of private precautions have the effect of shifting criminal activity from one potential victim to another, or from one place to another, without lowering overall crime rates.^[98] In the previous example, burglars may simply break into the nearest home whose owner has not invested in similar preventive measures. As should be clear to the reader, crime displacement has the effect of creating an arms race among potential victims. The understanding that being less protected than one’s neighbors makes one more susceptible to crime, creates an incentive not to fall behind. In this case, one does not only want to “keep up with the Joneses,” but one’s clear desire is to stay ahead of them.

That precautionary measures can result in the displacement of crime has been documented extensively. Numerous empirical studies have shown that for a variety of crimes, when commission of the offense against specific victims or in specific locations becomes more difficult, offenders tend to shift their illicit activity to other victims or places.^[99] Thus, rather than reducing crime, precautions may only displace it. While benefiting the specific potential victims who have taken the precautions, these investments in crime “prevention” are often socially wasteful.

An important, yet overlooked, effect of the growing expenditure on private precautions concerns its influence on collaboration among offenders. As private properties become increasingly fortified, individual offenders acting alone might not succeed in targeting them.^[100] Moving to the next available target may likewise prove impossible. When all homeowners invest in private precautions, there will be no attractive or easy targets. The only viable option left for offenders is to collaborate with others. This brings us back to the no-division rule.

As we will show, the no-division rule makes investment in private precautions attractive. To see why, assume a legal regime that divides the penalty for property crimes among co-offenders. Under such a division regime, investment in private precautions would have little effect on offenders’ incentives. Offenders operating under said regime would share both the profits and the sanction. The two effects would cancel each other out.

The following hypothetical is illustrative. Consider an offender who can break into a house by himself and gain $100 if not apprehended. The probability of apprehension is 10% and the sanction is $800. Under these assumptions, the offender’s expected gain is $90 (90% × $100) and the expected sanction is $80 (10% × $800). Since the expected gain exceeds the expected sanction, the offense would be committed.

Suppose now that the homeowner can maintain a security system for an annual cost of $50. The system cannot be overcome by a single offender; only two individuals working in tandem can commit the burglary. As the two co-offenders must share the proceeds, each co-offender’s expected gain would be $45 (90% × $100 × ½). Under a division rule, the expected sanction to be imposed on each co-offender would be $40 (10% × $800 × ½). Note that the cost-benefit ratio did not change as a result of the security system. In both scenarios, the offense has a positive expected value and would therefore be committed. In the absence of a security system, the expected gain was $90 and the expected cost was $80. With the system in place, the expected gain for each offender stood at $45, while the expected cost equaled $40. The homeowner, thus, derives little benefit from her investment in private precautions. Particularly, as the offenders’ cost-gain ratio remains intact, the homeowner’s investment in precautions did not reduce the risk of burglary.

This result changes, however, under existing law’s no-division rule. In the previous example, the offenders would still split the gains (such that each pockets $45), yet each would now bear the full sanction if apprehended. Accordingly, each of them faces an expected sanction of $80 (10% × $800). This expected sanction far outweighs the expected gain, rendering the offense unprofitable. The following table summarizes the difference in offenders’ cost-benefit ratio under both regimes:

Table 1: Offenders’ Cost-Benefit Ratio

(Benefit: 100; Sanction: 800; Probability of Apprehension: 10%)

	No Security System (offender acts alone)	Security System (offenders must collaborate)
Division	80 : 90	40 : 45
No-Division	80 : 90	80 : 45

As is evident, the homeowner will invest in a security system only under the no-division regime. The system, under this rule, renders the expected penalty greater than the expected benefit. Ostensibly, maintaining the security system is desirable: a potential crime was averted. It would thus seem that the no-division rule maximizes social welfare. But this conclusion is premature. As we explained earlier, investment in precautions does not prevent crime, but may simply displace it.

The possibility of displacement acts as a one-way ratchet, prompting potential victims to increase their investment in private precautions. If one homeowner invests in a security system, her neighbors would follow suit and invest in such a system too. The no-division rule greatly intensifies the incentive to invest in socially undesirable precautions. The reason is simple: by increasing the potential sanction for collaboration, the no-division rule makes private precautions more beneficial from an individual perspective. From a social perspective, the expenditure is actually harmful.

To demonstrate this point, consider again the previous example with a slight tweak. As before, the payoff from a successful burglary into a home is $100, and the probability of detection is 10%. The sanction remains $800. Again, as before, in the absence of a security system, an offender can burglarize a property singlehandedly; if such a system is in place, two offenders must collaborate.

Suppose now that there are three homeowners and two offenders. One home has fewer windows than the other two. Accordingly, the annual cost of maintaining the security system for the homeowner is $50, while the annual cost for the other two is $70 each.

For ease of exposition, suppose that a burglar targets one house per year.^[101] In this case, victims’ precautions are in fact socially undesirable. The annual aggregate costs of maintaining the security systems in all three homes is $190 ($50 + $70 + $70). This figure exceeds the annual loss resulting from burglaries; each of the two burglars will break into one house—a total expected loss of $180 ($90 + $90).

Under a division rule, the socially desirable outcome will prevail. As we have shown (Table 1), homeowners do not gain from maintaining the security systems, since those have no effect on offenders’ incentive to commit the crime. Offenders will simply team up and burglarize together (sharing both the proceeds and the sanction).

But turn now to the no-division rule. Given that there are three houses and two offenders, if none of the homeowners maintain the security system, each faces a ²/₃ probability of being victimized. The expected harm to each of the homeowners is, therefore, $60 (²/₃ × $90). Table 2 contrasts the homeowners’ expected harm from burglary with their respective costs of maintaining a security system:

Table 2: Expected loss from burglary, and cost of precautions

(Benefit: 100; Sanction: 800; Probability of Apprehension: 10%)

	Expected Loss from Burglary	Cost of Security System
Homeowner 1	60	50
Homeowners 2, 3	60	70

Under the no-division rule, all three homeowners will in fact invest in the undesirable precautions. As we have seen (Table 1), under the no-division rule, a security system renders the offense unprofitable for the offenders. Initially, only the first homeowner would maintain the system; it is in her interest to incur a $50 cost to avert an expected loss of $60. The other two homeowners will not invest $70 to avoid an expected loss of $60. Once the first homeowner protects her house, however, the calculus of the two remaining homeowners changes dramatically. Now, each of them faces a 100% probability of being victimized—recall that there are two offenders and two remaining homes. If neither homeowner takes precautions, each looks at an expected loss of $90. Given this fact, each would choose to incur a $70 cost to prevent a burglary into her house. Indeed, the investment pays off, as no burglary would occur. Each homeowner acted rationally from her own narrow perspective. But from a social perspective, the homeowners’ arms race was wasteful.

This undesirable result stems directly from the no-division rule. As we have shown, under a division regime, the homeowners would derive little benefit from the security systems.^[102] Surely, we do not fault the homeowners for acting rationally to protect their properties. Yet, in the aggregate, the no-division rule makes the homeowners (and society) worse off.

This example demonstrates a general trait of criminal law’s no-division rule. It augments the displacement problem by greatly increasing the cost of criminal collaboration. This in turn incentivizes victims to invest in private precautions that force offenders to collaborate, even when the result is merely the migration of criminal activity.

* * *

It bears emphasis that the problems attributable to the no-division rule are of modern vintage. Specifically, they are the result of two recent developments. The first is the great increase in the number of penalty levels arising from the Sentencing Guidelines. The common law traditionally recognized only two forms of theft, distinguishing between petty theft and grand theft.^[103] It had no rule prescribing a categorical enhancement of penalties based on the amount stolen.^[104] The Guidelines deviated from past tradition, introducing an elaborate scheme of penalty enhancements based on losses and gains. Furthermore, in the course of the evolution of the Guidelines, the number of penalty levels associated with amounts of losses and gains has significantly increased.^[105] The introduction of these multiple degrees of enhancements renders the no-division rule problematic. The more levels there are, the greater the potential disparity between a co-offender’s individual gain and the enhancement to which she is subject.

The second development is the restriction on judges’ discretion. While since 2005 judges may de jure deviate from the Guidelines following the Supreme Court’s ruling in United States v. Booker,^[106] for the reasons discussed in Part I, de facto, the Guidelines have become the standard for sentencing. Courts abide by the Guidelines’ formulations, meting out penalties based on the total amount of loss or gain. Accordingly, even when there is a large number of co-offenders, each is subject to the full penalty.

IV. Policy Implications

We have shown that the broad application of the no-division rule violates the frugality principle. We have also shown that it has undesirable distributive implications, as well as welfare reducing effects. In this Part, we explore possible policy responses to eliminate, or at least ameliorate, these troubling outcomes. The two central options we discuss are: (a) abolition of the no-division rule in the context of financial and property crimes; and (b) diverting public funds to the economically disadvantaged.

A. Splitting Losses and Gains Among Co-Offenders

One possible solution to the distortions created by the no-division rule is to abolish it with respect to property and financial crimes. Implementation of this measure would restore the frugality principle in sentencing, prevent unnecessary deprivations of liberty, and economize on social resources. Under the regime we propose, the amount dictating the enhancement will be divided among co-offenders in accordance with each offender’s individual share of the gains. If A and B steal $100 from a victim and split the proceeds equally, each of them will receive a penalty commensurate with a $50 theft.

At the outset, it should be emphasized that we do not support a wholesale abolition of the no-division rule. The rule is justified as applied to all offenses in which one violator’s gain does not detract from the gain of other co-offenders. In such cases, negating the gain of each individual co-offender requires imposing upon her the full penalty prescribed by law—dividing the penalty would result in under-deterrence.

The case for abolishing the no-division rule is most straightforward in gain-based sentencing. As previously explained,^[107] for some financial and property offenses, the offender’s gain (rather than the victim’s loss) is the key factor. In these cases, deterrence is clearly the Guidelines’ predominant consideration. Insider trading, bribery, commercial bribery, and prohibited gratuities are all offenses in which the penalty is set in accordance with the amount offenders gained (or sought to gain).^[108] Similarly, when the offense entails a fraudulent scheme, offenders’ sentence hinges on their proceeds.^[109]

Importantly, in these cases, offenders’ gains are not a proxy for victim’s loss that cannot be assessed. They are the a priori preferred determinant of the sentence. Even if the victim’s loss is clearly established, offenders’ gains remain the touchstone of sentencing. In all of these offenses, the victim’s loss may be far smaller or greater than the offender’s gain. Such is the case, for example, when a skillful bidder in a public tender bribes an official to overlook her criminal record, which bars her from participating. If the bidder performs her duties under the tender, the public entity suffers no direct harm. In fact, the public entity may ultimately pay less for the work than it otherwise would have.^[110] Similarly, in cases of insider trading, the harm to the stockholders and investors at large may far exceed the illegal gain to the corporate fiduciary.^[111] But the touchstone for sentencing nonetheless remains the offenders’ gains. When offenders’ gains are the preferred sentencing determinant, the argument for replacing the no-division rule with a division regime is evident.

Even when the victim’s loss is the dominant sentencing factor, the case for switching to a division regime can be supported by the Guidelines themselves. The idea of splitting a sentence driven primarily by the victim’s loss is not foreign to the Guidelines. Antitrust offenses present an example of precisely such a sentencing regime. In antitrust cases, the Guidelines explicitly set the fine to correspond to the approximated loss inflicted by cartels, not to cartel members’ estimated gains.^[112] Nonetheless, the Guidelines calibrate the offense level based on each offender’s volume of commerce, not on the overall harm inflicted on the victims.^[113]

For some, the proposed shift to a division regime in loss-based sentencing may appear objectionable on just-desert grounds. Particularly, retributivists could claim that each individual co-offender should bear the full brunt of the law. Just-desert approaches emphasize blameworthiness and moral reprehensibility.^[114] Accordingly, they place particular weight on offenders’ knowledge of the victim’s harm and intent to inflict it. Retributivists focus neither on deterrence nor on the frugality principle, both of which are not at the heart of their analytical framework.^[115]

We contend, however, that an unadulterated retributivist perspective raises difficulties as applied to co-offenders. The reason is that when the same harm is committed by multiple offenders, a retributivist approach acts as a multiplier of the appropriate sentence, leading to harsh results for both offenders and society. Suppose that two co-offenders committed a property crime that carries a penalty of five years in prison. Maintaining that each co-offender should bear the full sanction, irrespectively of her gain, retributivists would sentence each joint-offender to five years in prison. Assume now that ten offenders committed the same crime. A retributivist philosophy may support sentencing each of them to a five-year imprisonment term, meting out fifty years in prison to the group. The same logic would apply if twenty offenders were to participate. The lesson is clear: retributivism in itself offers no limiting principle. In the context of property and financial crimes, this leads to the imposition of harsh penalties. The proposed division regime offers such a limiting principle.

Finally, it is worth noting that courts have in fact adopted a division regime in a closely related retributivist context. Punitive damages are, per the Supreme Court, primarily designed to serve retributive functions.^[116] One of the fundamental rules set by the Court is the “single digit multiplier,” under which punitive damages generally cannot exceed nine times (and usually, four times) the compensatory damages.^[117] Importantly, this cap is subject to a division regime. If a number of tortfeasors are subject to punitive damages, the single-digit cap is not multiplied by the number of injurers; the overall retributory payment is split among them.^[118] The idea of division thus does not seem to undermine the attainment of retribution goals.

B. Diverting Public Resources to the Vulnerable

Even if lawmakers choose to retain the no-division rule, they must address the consequences of this decision. As we have shown, the no-division rule disadvantages offenders and victims. Perhaps most troubling, the no-division rule has a disparate impact on poor victims. While the issue of equality is complex, and some inequalities may not mandate governmental intervention, the inequality resulting from the no-division rule is different.

One of the most important services the government must provide is protection from criminal activity. The protection of life and property is enshrined in the Constitution.^[119] Importantly, this duty of the government comes with unique powers. The government alone holds the authority to criminalize activities and to impose sanctions on transgressors. Criminal law is a governmental creation. Thus, just like any other governmental service, provision of protection against crime must be carried out in an egalitarian manner.^[120]

In the case of property and financial crimes, the disadvantage of poor victims arises in large part from the penalty regime that applies to co-offenders, an artifact of the government’s design of criminal doctrine. It is only just, therefore, to impose on the government the burden of remediating the inequality it created itself.

Two possible approaches can be taken to the problems caused by the no-division rule. First, to offset the heightened exposure of poor victims to criminal activity, it is possible to enhance the penalties imposed on co-offenders who target poor victims. The increased sanction would make crime against the poor less attractive, leveling the playfield between the well-off and the poor. While the no-division rule favors affluent victims who can invest greater resources in private protection measures, the steeper sanctions for targeting poor victims can serve as an equalizer. This result can be achieved by expanding the scope of the term “vulnerability” in the Sentencing Guidelines. Although, to date, courts have rarely considered socioeconomic factors as a “vulnerability” per se,^[121] nothing prevents them from revisiting this policy.

Notwithstanding the potential of steeper penalties to alleviate the plight of poor victims, it comes with a high price and there are compelling reasons not to adopt it. The incarceration rate in the U.S. is staggering. Depriving individuals of their liberty is an extreme measure of last resort. Worse yet, studies show that imprisonment of individuals usually does not lead them to abandon crime, but perversely, increases the chances of future incarceration.^[122] This finding is not surprising. Spending time in jail dramatically decreases the opportunities available to an inmate upon release. As the non-criminal options become fewer and smaller, the lure of crime increases and sometimes it becomes the only viable option for former prisoners. This pattern, in turn, helps explain the high recidivism rates in the U.S.^[123] Thus, not only the unjust cruelty of excessive punishment, but also its social cost justifies abandoning a rule that multiplies offenders’ jail time.

A second approach that may be used to cure the distortive effects of the no-division rule is to allocate greater public resources to impecunious victims. Protection from crime is not limited to the formal penalty. The government can bolster the protection provided to vulnerable victims by adopting myriad tactics, from increased enforcement efforts to provision of better anti-crime infrastructure, such as security cameras and street lighting, funding of community services, rehabilitation, and education. These measures can prevent crime from occurring in the first place and therefore help both potential offenders and victims. As importantly, these measures significantly diminish the need for private investments in precautions and public investment in jails.

By funneling greater resources to crime prevention in poor areas, the government would not only neutralize the heightened exposure to crime it created by adopting the no-division rule, but would also generate long term positive effects for the residents. While the no-division rule advantages the rich by taxing collaboration, government investment in anti-crime measures of the sort we identified will reduce criminal activity in vulnerable communities, benefitting those who are made worse off by the no-division rule.

Conclusion

Collaboration among offenders is a common reality. According to current doctrine, co-offenders are sentenced under the no-division rule that requires courts to impose the full penalty on each collaborator. To date, criminal law theorists have largely overlooked the implications of the no-division rule, treating it as an inconsequential doctrinal feature. Against this background, this Article demonstrates the important implications of the no-division rule for offenders, victims, and society at large.

The analysis in the Article suggests that a no-division rule introduces three distortions into the criminal justice system. First, it violates the frugality principle by subjecting offenders to excessive and unnecessary sanctions. Second, it exposes vulnerable victims to a heightened risk of crime. Third, it fuels a wasteful arms race of investment in private precautions among potential victims. These undesirable effects call into question the desirability of the no-division rule. While the rule is justified as applied to offenses in which collaborators do not share ill-gotten gains, it triggers serious concerns when applied to offenses in which perpetrators must divide their gains.

To address these distortions, this Article calls for the replacement of the no-division rule with a division regime akin to that adopted by tort law. Dividing criminal penalties among co-offenders in appropriate cases can promote deterrence without violating retributive goals. Alternatively, if social planners elect to retain the no-division rule, they should allocate greater resources to protect economically disadvantaged victims. Providing greater protection to the vulnerable is not merely a desirable goal in itself, but also an affirmative duty of the state.

Appendix A – List of State Statutes

Ala. Code §§ 13A-8-3 to 13A-8-23 (LexisNexis through Acts 2021);

Alaska Stat. §§ 11.46.100–.150 (LexisNexis through 2021 legislation);

Ariz. Rev. Stat. § 13-1802 (LexisNexis through all 2021 legislation);

Ark. Code Ann. § 5-36-103 (LexisNexis through Act 1112 of the 2021 Reg. Sess.);

Cal. Penal Code §§ 486–88 (Deering, LexisNexis through Chapter 1-100, 102, 103, 105–110, 112, 114, 115, 117–123, 125–142, 145, 150, 152, 164, 174, and 180 of the 2021 Reg. Sess);

Colo. Rev. Stat. § 18-4-401 (LexisNexis through all 2021 Reg. Sess. bills with safety clauses);

Conn. Gen. Stat. §§ 53A-122 to 53A-125 (LexisNexis through all Acts from the 2021 Reg. and June Special Sess.);

Del. Code Ann. tit. 11, § 841 (LexisNexis through 83 Del. Laws c. 208);

D.C. Code § 22-3212 (LexisNexis through all permanent laws effective as of Oct. 1, 2021);

Fla. Stat. Ann. § 812.014 (LexisNexis through the 2021 Reg. and 1st Extraordinary Sess.);

Ga. Code Ann. § 16-8-12 (LexisNexis through 2021 Reg. Sess.);

Haw. Rev. Stat. Ann. §§ 708-830 to –833 (LexisNexis through 2021 Legislative Sess. and 1st Special Sess.);

Idaho Code Ann. § 18-2407 (West through 2021 1st Reg. Sess.);

720 Ill. Comp. Stat. Ann. 5/16-1 (LexisNexis through P.A. 102-297 of the 2021 Sess.);

Ind. Code Ann. § 35-43-4-2 (LexisNexis through Public Laws 1-220 (end) of the 2021 1st Reg. Sess.);

Iowa Code § 714.2 (LexisNexis through legislation from the 2021 Reg. Sess.);

Kan. Stat. Ann. § 21-5801 (LexisNexis through 2021 Reg. Sess.);

Ky. Rev. Stat. Ann. § 514.030 (LexisNexis through the end of the 2021 Reg. Sess.);

La. Stat. Ann. § 14:67 (West through the 2021 Reg. Sess. and Veto Sess.);

Me. Rev. Stat. Ann. tit. 17-A, § 353 (West through the 2021 1st Reg. Sess. and 2021 1st Special Sess..);

Md. Code Ann., Crim. Law § 7-104 (West through all legislation from the 2021 Reg. Sess.);

Mass. Ann. Laws ch.266, §30 (LexisNexis through Chapters 1–38, and 40–50 of the 2021 Legis. Sess.);

Mich. Comp. Laws Serv. § 750.356 (LexisNexis through Act 83 of the 2021 Reg. Legis. Sess.);

Minn. Stat. § 609.52 (LexisNexis through 2021 Reg. Sess.);

Miss. Code Ann. §§ 97-17-41 to -43 (Lexis through 2021 Reg. Sess.);

Mo. Rev. Stat. § 570.030 (LexisNexis through 101st General Assembly, 1st Reg. Sess., except for 2021 HB 271);

Mont. Code Ann. § 45-6-301 (LexisNexis through 2021);

Neb. Rev. Stat. Ann. § 28-518 (LexisNexis through all Acts of the 1st Reg. Sess. and LB 14 of the 1st Special Sess. of the 107th Leg.);

Nev. Rev. Stat. Ann. § 205.0835 (LexisNexis through 2021 Reg. Sess.);

N.H. Rev. Stat. Ann. § 637:11 (LexisNexis through 2021 Reg. Sess.);

N.J. Stat. § 2C:20-20 (LexisNexis through 2021 219th 2nd Ann. Sess., L. 2021, ch. 188, and J.R. 3);

N.M. Stat. Ann. § 30-16-1 (LexisNexis through 2021 Reg. Sess.);

N.Y. Penal Law §§ 155.20–155.42 (Consol., LexisNexis through 2021 released Chs.1-429);

N.C. Gen. Stat. § 14-72 (LexisNexis through Sess. Laws 2021-162);

N.D. Cent. Code § 12.1-23-05 (LexisNexis through all acts approved by the governor through the end of the 2021 67th Legis. Assembly);

Ohio Rev. Code Ann. § 2913.02 (LexisNexis through file 47 of the 134th (2021) Gen. Assembly);

Okla. Stat. Ann. tit. 21, § 1704 (LexisNexis through 2021 1st Reg. Sess.);

Or. Ann. Stat. §§ 164.045–.055 (LexisNexis through laws effective as of Sept. 1st, 2021);

18 Pa. Stat. And Cons. Stat. Ann. § 3903 (West through 2021 Reg. Sess. Act 70);

11 R.I. Gen. Laws § 11-41-5 (LexisNexis through all legis. of the 2021 Sess.);

S.C. Code Ann. § 16-13-30 (West through 2021 Act. No. 116);

S.D. Codified Laws §§ 22-30A-17 (LexisNexis through 2021 Gen. Sess.);

Tenn. Code Ann. § 39-14-105 (LexisNexis through 1st Extraordinary and Reg. Sess.);

Tex. Penal Code Ann. § 31.03 (LexisNexis through 2021 Reg. Sess.);

Utah Code Ann. § 76-6-412 (LexisNexis through 2021 1st Spec. Sess.);

Vt. Stat. Ann. tit. 13, §§ 2501–2502 (LexisNexis through Act 16 and Municipal Act M-1 of the 2021 Sess.);

Va. Code Ann. §§ 18.2-95 (West through 2021 Reg. Sess. and Spec. Sess. I);

Wash. Rev. Code Ann. §§ 9A.56.030–.050 (LexisNexis through 2021 Reg. Sess.);

W. Va. Code Ann. § 61-3-13 (West through 2021 1st Spec. Sess.);

Wis. Stat. Ann. § 943.20 (West through 2021 Act 59-79);

Wyo. Stat. Ann. § 6-3-402 (LexisNexis through 2021 Gen. Sess.).

1. . See, e.g., Noah M. Kazis, Tort Concepts in Traffic Crimes, 125 Yale L.J. 1131–32 (2016) (“Legal scholars have long understood tort and criminal law as parallel mechanisms for sanctioning private behavior.”); Gail Heriot, An Essay on the Civil-Criminal Distinction with Special Reference to Punitive Damages, 7 J. Contemp. Legal Issues 43, 52 (1996) (“Both [civil and criminal liability] intentionally inflict a cost on the defendant in order to alter his behavior and in that sense are punitive. Both help to educate the public as to which acts are considered inappropriate. Both help deter the public from committing those acts for fear of suffering the same fate as the defendant.”); People v. Singh, 2008 WL 5064918 at *5 (Cal. Ct. App. Dec. 2, 2008) (“criminal cases and tort cases share the concern of deterring harmful acts.”). ↑

2. . United States v. Bach, 172 F.3d 520, 523 (7th Cir. 1999); Thomas C. Grey, Accidental Torts, 54 Vand. L. Rev. 1225, 1244 (2001) (“Most crimes (the ones that have victims) are also torts.”). ↑

3. . Gill v. Arab Bank, PLC, 893 F. Supp. 2d 474, 496 (E.D.N.Y. 2012) (quoting Antiterrorism Act of 1990: Hearing on S. 2465 Before the S. Comm. on Cts. & Admin. Practice, 101st Cong. 136–37 (1992) (statement of Joseph A. Morris, former general counsel of the U.S. Information Agency and the U.S. Office of Personnel Management)) (“[T]ort law system generally tracks, and usefully tracks, criminal law doctrines.”). For concrete doctrinal similarities, see, for example, Paroline v. United States, 572 U.S. 434, 444 (2014) (“The concept of proximate causation is applicable in both criminal and tort law, and the analysis is parallel in many instances.”); State v. Pollander, 706 A.2d 1359, 1363 (Vt. 1997) (citations omitted) (“This Court has recognized the necessity defense in criminal cases . . . and in tort litigation.”); Breland v. Schilling, 550 So. 2d 609, 611 (La. 1989) (“[I]n both the criminal and tort fields, ‘intent’ refers to the consequences of the act, rather than to the act itself.”). ↑

4. . See Dan B. Dobbs, Paul T. Hayden & Ellen M. Bublick, Hornbook on Torts (2d ed. 2015) § 35.1–.3 (describing the principle of apportionment among joint tortfeasors, and the modern trend of shifting from several to joint-and-several liability). ↑

5. . Id. (explaining the doctrinal elements of apportionment, and that under either several, or joint-and-several liability, joint tortfeasors share their responsibility to the victim). ↑

6. . See infra Part III. ↑

7. . The offenders may also be charged for the crime of conspiracy. But with respect to sentencing, the charge of conspiracy has little bearing. See Neal Kumar Katyal, Conspiracy Theory, 112 Yale L.J. 1307, 1309 (2003) (noting that under the Federal Sentencing Guidelines, “it is not generally possible to punish someone for conspiring to commit a crime and for committing it.”). Furthermore, in an increasing number of states, as well as in the Model Penal Code, the conspiracy offense merges with the completed crime. Id. at 1369. Thus, as this Article shows, it is largely the rule that each offender is sentenced for the full amount that penalizes collaboration. ↑

8. . Jeremy Bentham, An Introduction to the Principles of Morals and Legislation 142 (Batoche Books 2000) (1781) (introducing the idea of the “frugality principle.”). ↑

9. . The “frugality principle” is sometimes referred to as “the principle of parsimony.” See Richard S. Frase, Punishment Purposes, 58 Stan. L. Rev. 67, 68 (2005) (discussing the principle and its moral underpinnings); Michael Tonry, Intermediate Sanctions in Sentencing Guidelines, 23 Crime & Just. 199, 206–07 (1998) (noting the wide consensus in support of the principle). ↑

10. . Bentham, supra note 8, at 147. ↑

11. . This point has long been a key motivation for penalizing conspiracy. See Callanan v. United States, 364 U.S. 587, 593–94 (1961) (explaining co-offenders’ benefit in carrying out crimes collaboratively, particularly when crimes are complex). ↑

12. . Robert A. Mikos, “Eggshell” Victims, Private Precautions, and the Societal Benefits of Shifting Crime, 105 Mich. L. Rev. 307, 308 (2006). ↑

13. . See infra note 98 and accompanying text (presenting empirical studies on crime displacement). ↑

14. . See infra Part III.C. ↑

15. . U.S. Sent’g Guidelines Manual (U.S. Sent’g Comm’n 2018) [hereinafter USSG]. ↑

16. . 28 U.S.C. § 994(a). ↑

17. . USSG, supra note 15, at 2; Frank O. Bowman, III, “Loss” Revisited: A Defense of the Centerpiece of the Federal Economic Crime Sentencing Guideline, 82 Mo. L. Rev. 1, 2 (2017) [hereinafter Bowman, “Loss” Revisited]. ↑

18. . USSG, supra note 15, at 407. ↑

19. . Id. § 1B1.1(a). ↑

20. . E.g., id. §§ 2A2.2(b)(2), 2A5.2(b)(1). ↑

21. . Id. § 2D1.1(a)(5). ↑

22. . Id. § 2C1.3(b)(1). ↑

23. . Id. § 3A & 3B. On the distinction between specific offense characteristics and adjustments, see U.S. Sent’g Comm’n, An Overview of the Federal Sentencing Guidelines 2–3, https://www.ussc.gov/sites/default/files/pdf/about/overview/Overview_Federal_Sentencing_Guidelines.pdf [https://perma.cc/BUN7-7LQH]. ↑

24. . USSG, supra note 15, at 12. ↑

25. . Id. § 1B1.1(a)(6). ↑

26. . Id. § 1B1.1(a)(7). ↑

27. . 18 U.S.C. § 3553(b)(1). ↑

28. . 543 U.S. 220, 220 (2005). See also Jillian Hewitt, Fifty Shades of Gray: Sentencing Trends in Major White-Collar Cases, 125 Yale. L.J. 1018, 1026–29 (2016) (discussing the pre-Booker sentencing regime and Booker’s implications for the status of the Guidelines). ↑

29. . Frank O. Bowman, III, Dead Law Walking: The Surprising Tenacity of the Federal Sentencing Guidelines, 51 Hous. L. Rev. 1226, 1268 (2014) (“The endurance of the Guidelines, but more particularly the degree to which they continue to drive actual sentences, has surprised nearly everyone.”) [hereinafter Bowman, Dead Law Walking]. ↑

30. . Rita v. United States, 551 U.S. 338, 347–48 (2007). ↑

31. . Gall v. United States, 552 U.S. 38, 49 (2007). ↑

32. . Peugh v. United States, 569 U.S. 530, 549 (2013). ↑

33. . Rita, 551 U.S. at 347. ↑

34. . Gall, 552 U.S. at 47. ↑

35. . U.S. Sent’g Comm’n, The Influence of the Guidelines on Federal Sentencing: Federal Sentencing Outcomes, 2005–2017, at 7 (2020), https://www.ussc.gov/sites/default/files/pdf/research-and-publications/research-publications/2020/20201214_Guidelines-Influence-Report.pdf [https://perma.cc/Z2Z5-MP75]; Peugh, 569 U.S. at 543. ↑

36. . Peugh, 569 U.S. at 544; Bowman, Dead Law Walking, supra note 29, at 1229–30, 1235. ↑

37. . 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, 716 (2016) (“The most important driver of sentences for economic crimes under the Guidelines is the amount of pecuniary harm that the defendant either actually caused or intended to cause.”); Patti B. Saris, The 2015 Economic Crime Guideline Amendments, 12 N.Y.U. J.L. & Bus. 1, 7 (2015) (“The loss table is probably the most important example of the way that particular offense level characteristics can affect the advisory guideline range.”); see also id. at 16–17 (presenting data regarding specific offense characteristics); Bowman, “Loss” Revisited, supra note 17, at 2, 5–6. ↑

38. . See Saris, supra note 37, at 7. ↑

39. . USSG, supra note 15, § 2B1.1(a)(1), (2). ↑

40. . Or otherwise falls into Category I on the horizontal axis (Criminal History Category) of the Sentencing Table. Id. at 407. ↑

41. . Id. § 2B1.1(b)(1). ↑

42. . Id. ↑

43. . Id. at 407. ↑

44. . Id. § 2B1.4(a). ↑

45. . Id. at 407. ↑

46. . Id. § 2B1.4(b)(1). ↑

47. . Assuming, again, that the defendant falls into the Category I on the horizontal axis (Criminal History Category) of the Sentencing Table. Id. at 407. ↑

48. . Id. § 2B2.1(a) & § 2B2.1(b)(2). ↑

49. . Id. § 2B3.1(a) & § 2B3.1(b)(7). ↑

50. . Id. § 2R1.1(a) & § 2R1.1(b)(2). ↑

51. . Id. at 407. ↑

52. . Frank O. Bowman, III, Damp Squib: The Disappointing Denouement of the Sentencing Commission’s Economic Crime Project (and What They Should Do Now), 27 Fed. Sent’g Rep. 270, 271 (2015) [hereinafter Bowman, Damp Squib] (emphasis omitted). ↑

53. . USSG, supra note 15, § 2B1.1(b)(1)–(19). ↑

54. . Saris, supra note 37, at 16–17. ↑

55. . Id. ↑

56. . Id. at 16. ↑

57. . See generally supra note 37. ↑

58. . For example, in all states the penalty for theft (or larceny, as it is referred to in some jurisdictions) hinges on the amount stolen. For a full list of the relevant statutes in the different states see Appendix A. ↑

59. . USSG, supra note 15, § 2B1.1, cmt. n.3(B). On the question of when offenders’ gains are to be used as substitutes for the victim’s loss, see, for example, United States. v. Dickler, 64 F.3d 818, 825–26 (3d Cir. 1995), United States v. Schneider, 930 F.2d 555 (7th Cir, 1991). ↑

60. . USSG, supra note 15, § 2B1.4(b)(1) (insider trading); § 2C1.1(b)(2) (bribery); § 2B4.1(b)(1) (commercial bribery); § 2C1.2(b)(2) (prohibited gratuities). On the distinction between bribery and commercial bribery, compare United States v. DeVegter, 439 F.3d 1299 (11th Cir. 2006) (commercial bribery), with United States v. Huff, 609 F.3d 1240 (11th Cir. 2013) (bribery). ↑

61. . USSG, supra note 15, § 2B1.4(b)(1). ↑

62. . Id. § 2C1.1(b)(2). ↑

63. . Id. § 2B4.1(b)(1). ↑

64. . Id. § 2C1.2(b)(2). A paradigmatic case is a public official receiving a prohibited gratuity. In this case, the penalty would be based on the value of the gratuity regardless of the harm to the government. ↑

65. . Id. § 2B1.1(b)(1). ↑

66. . Id. § 2B1.1(b)(1), cmt. n.3(F)(v) (equating the “loss” with the offender’s benefit). ↑

67. . See, e.g., United States v. Anchor Foods, Inc., 2020 WL 8167426, at *6 (E.D.N.Y. May 10, 2020) (“In fact, there is no evidence that a single consumer indicated a preference for octopus over squid, other than the actual purchases.”). ↑

68. . See Thaler v. United States, 706 F. Supp. 2d 361, 370 (S.D.N.Y. 2009) (“[A] fraudulently induced purchase of certain assets does not cause loss equaling the entire purchase price if the assets actually have some value greater than zero.”) (describing the holding of United States v. Leonard, 529 F.3d 83 (2d Cir. 2008)). ↑

69. . USSG, supra note 15, 1B1.3(a). When the defendant is accountable for an amount that greatly exceeds his or her gain, the Guidelines allow for a “minimal participant” or “minor participant” adjustment (Id. § 3B1.2). But this does not undermine the general no-division principle. To begin with, all co-offenders who did not play a minor role in the offense will be sentenced based on the full gain or loss. Second, the offense-level decrease for a minor participant is only 2–4 points. This adjustment is trivial compared to most enhancements associated with the amount of loss or gain. ↑

70. . 75 F.3d 439 (8th Cir. 1996). ↑

71. . Id. at 440. ↑

72. . Id. at 441. ↑

73. . 335 F.3d 919, 923 (9th Cir. 2003). ↑

74. . Id. at 928. ↑

75. . 902 F.3d 780, 789–90 (8th Cir. 2018). ↑

76. . United States v. DeVegter, 439 F.3d 1299 (11th Cir. 2006). ↑

77. . United States v. Anchor Foods, Inc., 2020 WL 8167426, at *6 (E.D.N.Y. May 10, 2020). ↑

78. . There are also two cases in which the defendant’s individual gain is considered. See USSG, supra note 15, § 2B1.1(b)(17)(A) & cmt. n.13(A); § 2B4.1(b)(2)(A) & cmt. n.4(A). But these are not an exception to the no-division rule. In both of these cases individual gain is an additional factor, to be considered alongside loss to the victim (which remains indivisible). And the individual gain (which comes into play only if it exceeds $1 million) increases the offense level by 2 points, while the loss to the victim may increase the offense level by up to 30 points. See also Hewitt, supra note 28, at 1033–34 (noting that these characteristics are insignificant compared to loss). ↑

79. . USSG, supra note 15, § 2R1.1 cmt. background (“There is no consensus, however, about the harmfulness of other types of antitrust offenses, which furthermore are rarely prosecuted and may involve unsettled issues of law. Consequently, only one guideline, which deals with horizontal agreements in restraint of trade, has been promulgated.”). ↑

80. . Id. § 2R1.1(b)(2). ↑

81. . Of course, there is no reason to assume equal division of the gains. It is entirely possible that one of the four co-offenders will receive $40 and the other three $20 each. This, however, does not affect the basic point that a dollar pocketed by one of the offenders cannot be appropriated by others. ↑

82. . Bentham, supra note 8. ↑

83. . See, e.g., Samuel B. Lutz, The Eighth Amendment Reconsidered: A Framework for Analyzing the Excessiveness Prohibition, 80 N.Y.U. L. Rev. 1862, 1862 (2005) (“[I]t is widely accepted that the Eighth Amendment operates as a broad prohibition against excessive criminal sanctions . . . .”). ↑

84. . “The conventional understanding of conspiracy law regards group crime as significantly more dangerous than individual crime because criminal organizations can take advantage of greater resources, economies of scale, specialized labor, and perhaps greater commitment to the criminal enterprise on the part of their members.” Daryl J. Levinson, Collective Sanctions, 56 Stan. L. Rev. 345, 399 (2003). ↑

85. . See Katyal, supra note 7, at 1312 (“The more conspirators, the more witnesses there are to flip and the more ominous the prisoners’ dilemma for a conspirator.”). Furthermore, in some cases, criminal collaboration may even increase the prospects of apprehension, since “conspiracy is more vulnerable to being detected because of the scale of its activities.” Richard A. Posner, An Economic Theory of the Criminal Law, 85 Colum. L. Rev. 1193, 1219 (1985). ↑

86. . The prospects of co-offenders incriminating one another are attenuated when they are members of a criminal organization. But even in this case, there is no inherent justification for the no-division rule. Offenders already face an increased expected sanction under the Racketeer Influenced and Corrupt Organizations Act, 18 U.S.C. §§ 1961–68. ↑

87. . Annual Determination of Average Cost of Incarceration Fee (COIF), 84 Fed. Reg. 63891, 63891–92 (Nov. 19, 2019). For a state-by-state comparison see Vera, Prison Spending in 2015 (2015), https://www.vera.org/publications/price-of-prisons-2015-state-spending-trends/price-of-prisons-2015-state-spending-trends/price-of-prisons-2015-state-spending-trends-prison-spending [https://perma.cc/5KU8-MKD9]. ↑

88. . Jacob Kang-Brown, Chase Montagnet & Jasmine Heiss, Vera Institute of Justice, People in Jail and Prison in 2020 (2021). The number of inmates dropped in 2020, as individuals were released from jails and prisons in an attempt to prevent the COVID-19 virus from spreading. See id. at 3. ↑

89. . Katyal, supra note 7, at 1310. ↑

90. . If one-fourth of the cases involve co-offenders, then even if there are only two co-offenders per such case, co-offenders would comprise 40% of the inmate population (of 100 cases, the 25 conspiracy cases would result in 50 inmates, while the remaining 75 cases would result in 75 inmates; accordingly, co-offenders would account for 50/125, or 40%, of the inmate population). Total annual incarceration costs are in the tune of $80 billion ($37,449 × 2.115 million). 40% of the annual cost is $32 billion. ↑

91. . See infra text accompanying notes 92–96. ↑

92. . For a recent review of the literature, see Badi Hasisi, Simon Perry, Yonatan Ilan & Michael Wolfowicz, Concentrated and Close to Home: The Spatial Clustering and Distance Decay of Lone Terrorist Vehicular Attacks, 36 J. Quantitative Criminology 607, 610–11 (2020). ↑

93. . Steven D. Levitt, The Changing Relationship Between Income and Crime Victimization, 5 FRBNY Econ. Pol’y Rev. 87 (Sept. 1999). ↑

94. . Id. at 87 (finding that “[b]y 1994, the poor households were 60 percent more likely to be burglarized than the rich households,” while finding no similar pattern for violent crimes). More recent reports on the distribution of burglary victims corroborate this finding. See, e.g., Jennifer Hardison Walters, Andrew Moore, Marcus Berzofsky & Lynn Langton, Bureau of Just. Stat., Household Burglary, 1994–2011, at 9 (2013) (finding that “[i]n 1994, 2001, and 2011, households with an income of $14,999 or less were victimized at a higher rate than households with higher incomes.”). ↑

95. . We produced these figures based on two sources of information. The National Crime Victimization Survey provides data on the distribution of crimes across households. NCVS Victimization Analysis Tool, Bureau of Just. Stats., https://www.bjs.gov/index.cfm?ty=nvat [https://perma.cc/69BF-DSKK]. The percentage of these households in the population is retrieved from the Population Census for the same years. Current Population Survey, U.S. Census Bureau, https://www.census.gov/data/tables/time-series/demo/income-poverty/cps-hinc/hinc-06.2010.html [https://perma.cc/2JLR-H32U]. ↑

96. . Id. ↑

97. . Interestingly, Levitt suggested that the high percentage of low-income victims in the context of property crimes may relate to financial inability to invest in private precautions. Levitt, supra note 93, at 91. As we claim, private precautions are particularly effective because of the no-division rule. ↑

98. . For an explanation of this phenomenon, see Omri Ben-Shahar & Alon Harel, The Economics of the Law of Criminal Attempts: A Victim-Centered Perspective, 145 U. Pa. L. Rev. 299, 309–10 (1996). ↑

99. . For a literature review, see John E. Eck, The Threat of Crime Displacement, 25 Crim. Just. Abstracts 527 (1993); René B.P. Hesseling, Displacement: A Review of the Empirical Literature, in 3 Crime Prevention Studs. 197 (Ronald V. Clarke ed., 1994); Kate J. Bowers & Shane D. Johnson, Measuring the Geographical Displacement and Diffusion of Benefit Effects of Crime Prevention Activity, 19 J. Quantitative Crimonology 275 (2003); see also Jerry H. Ratcliffe & Clairissa Breen, Crime Diffusion and Displacement: Measuring the Side Effects of Police Operations, 63 Pro. Geographer 230, 231–34 (2011) (reviewing studies on displacement). ↑

100. . Precautions may also increase offenders’ chances of being caught, or otherwise make the commission of the crime less attractive. See, e.g., Matthew T. Clements, Precautionary Incentives for Privately Informed Victims, 23 Int’l Rev. L. & Econ. 237, 239 (2003). As we point out, however, even in cases in which offenders can overcome these effects by teaming up, the no-division rule still makes investments in precautions beneficial for victims. ↑

101. . Of course, we may assume that burglars may target homes more or less frequently. This will not change the analysis. It will only imply a higher or lower annual investment in the security system. ↑

102. . Under a division regime, a security system carries a benefit to the homeowner; however, this benefit is a trivial one as compared to the benefit under the no-division rule. To illustrate, if the first homeowner has a security system, the offenders now face two options: targeting her home together, or splitting up and separately targeting the two remaining homes. Thus, the system reduces her probability of being targeted from 2/3 to 1/2 (a reduction of 1/6). But this benefit (1/6 × $90) is smaller than the cost of maintaining the security system ($50). Thus, under a division rule, none of the homeowners will invest in private precautions. ↑

103. . See, e.g., Frank O. Bowman, III, Coping with Loss: A Re-Examination of Sentencing Federal Economic Crimes under the Guidelines, 51 Vand. L. Rev. 461, 476–77 (1998) [hereinafter Bowman, Coping with Loss] (describing the development of the common-law regime and the two categories of theft). ↑

104. . The common law’s two-level categorization is still in effect in some states. See e.g., Mass. Ann. Laws ch. 266, § 30 (Lexis through 2021 Legis. Sess.); D.C. Code § 22-3212 (Lexis through all permanent laws effective as of Oct. 28, 2021). However, the modern trend follows the path of the Guidelines, introducing an increasing number of categories. For example, in New York the law differentiates among five categories of theft. See N.Y. Penal Law §§ 155.20–.42 (Lexis through 2021 released Chs. 1-451, 453–484). Colorado’s law goes even further, establishing nine different categories. See Colo. Rev. Stat. § 18-4-401 (Lexis through 2021 Reg. Sess.). ↑

105. . See Derick R. Vollrath, Losing the Loss Calculation: Toward a More Just Sentencing Regime in White-Collar Criminal Cases, 59 Duke L.J. 1001, 1008–12 (2010) (describing the evolution of the Guidelines and the expansion of the number of loss-categories). ↑

106. . United States v. Booker, 543 U.S. 220 (2005). ↑

107. . See supra Part I.C. ↑

108. . See supra text accompanying notes 60–64. Recall that for bribery this is the case when the amount obtained exceeds the harm to the victim. ↑

109. . See supra text accompanying notes 65–68. ↑

110. . There may be other losses associated with awarding the contract to the bidder: the public entity obviously has a preference (or duty) not to deal with such contractors; the entity may need to bear the expenses of a new tender it will need to initiate; and so on. But these losses are not correlated to the bidder’s profits. Retribution may thus justify a relatively low penalty, or a very large one. ↑

111. . On the protection of investors at large as one of the main goals of securities laws, see Abraham J.B. Cable, Mad Money: Rethinking Private Placements, 71 Wash. & Lee L. Rev. 2253, 2267–75 (2014). ↑

112. . USSG, supra note 15, § 2R1.1, cmt. n.3. ↑

113. . Id. § 2R1.1, cmt. n.7. See supra text accompanying note 80. ↑

114. . See, e.g., Gerard V. Bradley, Retribution: The Central Aim of Punishment, 27 Harv. J.L. & Pub. Pol’y 19, 20 (2003); Kevin M. Carlsmith, John M. Darley & Paul H. Robinson, Why Do We Punish? Deterrence and Just Deserts as Motives for Punishment, 83 J. Pers. Soc. Psych. 284 (2002). But see Ken Levy, Why Retributivism Needs Consequentialism: The Rightful Place of Revenge in the Criminal Justice System, 66 Rutgers L. Rev. 629, 636–48 (2013) (arguing that retributivism should incorporate consequential considerations). ↑

115. . See, e.g., Michael S. Moore, Placing Blame: A General Theory of the Criminal Law 153 (2010) (“Punishment is justified, for a retributivist, solely by the fact that those receiving it deserve it. . . . Punishment may deter future crime, incapacitate dangerous persons, educate citizens. . . . Yet for a retributivist these are a happy surplus that punishment produces and form no part of what makes punishment just. . . .”). ↑

116. . As the Court emphasized, rather than deterrence, it “is the degree of reprehensibility of the defendant’s conduct” that is “[p]erhaps the most important” constitutional guidepost in determining the appropriateness of a punitive damages award. See BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 575 (1996); see also Dobbs et al., supra note 4, at 870 (“Courts applying the Supreme Court’s guideposts have focused considerable attention on reprehensibility, at times calling it the most important indicium of the reasonableness of a punitive damages award.”). ↑

117. . State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 424–25 (2003). ↑

118. . See, e.g., Planned Parenthood of Columbia v. Am. Coal. of Life Activists, 422 F.3d 949, 960–63 (9th Cir. 2005) (discussing the application of the single-digit cap in cases of multiple injurers and endorsing a regime under which injurers’ overall punitive liability is similarly capped). ↑

119. . U.S. Const. amend. XIV § 1 (“No state shall . . . deprive any person of life, liberty, or property, without due process of law.”). ↑

120. . For an exploration of this duty, see Alon Harel, Efficiency and Fairness in Criminal Law: The Case for a Criminal Law Principle of Comparative Fault, 82 Calif. L. Rev. 1181, 1204–07 (1994); see also Alon Harel & Gideon Parchomovsky, On Hate and Equality, 109 Yale L.J. 507, 524–29 (1999) (discussing the government’s obligation to employ criminal doctrine in a manner ensuring that potential victims face equal risks). ↑

121. . For a discussion of the case law on the interpretation of vulnerability in the Guidelines, see Harel and Parchomovsky, supra note 120, at 530–31. ↑

122. . See Jeremy Travis, Bruce Western & F. Stevens Redburn, The Growth of Incarceration in the United States: Exploring Causes and Consequences 33–69, 130–56 (2014) (presenting findings on the effects of incarceration on criminal activity). ↑

123. . See, e.g., Francis T. Cullen, Cheryl Lero Jonson & Daniel S. Nagin, Prisons Do Not Reduce Recidivism: The High Cost of Ignoring Science, 91 Prison J. 48S, 53S–57S (2011) (finding high recidivism rates among former inmates and establishing the effect of imprisonment on recidivism). ↑