87 Colum. L. Rev. 920, *

 
Copyright © 1987 The Columbia Law Review.

Columbia Law Review

JUNE, 1987

87 Colum. L. Rev. 920

LENGTH: 30598 words



ARTICLE: RICO: THE CRIME OF BEING A CRIMINAL, PARTS III & IV. *

SEC-NOTE-1:

* These are the second two parts of a four part Article.

NAME: Gerard E. Lynch **

BIO:

** Professor of Law, Columbia University. B.A. 1972, J.D. 1975, Columbia University.

SUMMARY:
  ... RICO: A NEW KIND OF BEAST? ... Prosecutors may have been induced to add RICO counts to ordinary narcotics conspiracy charges in order to avail themselves of its automatic forfeiture provisions. ... For example, one defendant was a juror who allegedly took a bribe to fix a prosecution of one of the members of the "crew"; another's entire involvement (limited to three of the eighty racketeering acts charged in the indictment) consisted of hiring members of the crew to bribe and ultimately to murder a witness against his son in a state prosecution entirely unrelated to the affairs of the enterprise; others were involved only in one of the many criminal affairs of the enterprise, and not at all in its more violent activities. ... The court thus seems to accept that an agreement to raise revenue through miscellaneous criminal activity is indictable as a conspiracy, holding only that the evidence here did not show such an agreement. ... Third, the value of the RICO forfeiture sanction suggests that increased use of this sort of penalty should be seriously considered. ... As we have seen, the original purpose of the RICO forfeiture remedy was the antitrust-type goal of extruding the racketeer from the legitimate enterprise he had infiltrated. ...  

TEXT:
 [*920]  III. RICO: A NEW KIND OF BEAST?

In the first portion of this study, 1 we saw that the Supreme Court in its 1981 Turkette decision 2 endorsed what was already the consensus view of the courts of appeals that a group of individuals associated in fact to pursue entirely illegitimate purposes could constitute a RICO enterprise. 3 Prosecutions of such associations have quickly become the leading use of the statute. It can be reliably estimated that more than forty percent of the reported appellate cases involving RICO indictments concern prosecutions in which the alleged enterprise was such an illicit association. 4 When the cases are classified by the nature of the predicate criminal acts rather than of the alleged enterprise, roughly the same figure appears: about forty percent of the cases involve allegations of concerted criminal activity by more-or-less organized criminal groups, with the rest divided among the various types of perversion of legitimate institutions discussed in the preceding sections. 5 These figures demonstrate that the principal use of RICO has not been to deal with the distinctive evil of infiltration by the mob into legitimate enterprises, but rather to add an additional weapon to prosecutors' efforts to attack organized criminal groups themselves for their primary illegal activities.

The illicit association cases present a broad spectrum of illegal activities. In many of the cases, the illicit activities charged as predicate acts involve repeated instances of the same general type of criminal conduct. The large number of RICO narcotics prosecutions typify this kind of case, although similar examples involving gambling and prostitution activity can be cited. 6 Other cases involve arson, extortion, and  [*921]  theft or fencing schemes. 7

The use of RICO in many of these cases appears indistinguishable from the use of ordinary conspiracy law. Although most of the RICO cases seem to involve particularly large or lucrative conspiracies, 8 some of them are strikingly ordinary criminal conspiracies. 9 Why have prosecutors invoked RICO so frequently? The cases in this area are so diverse that no single answer is possible. In some instances, particularly in simple cases involving schemes limited to one particular sort of crime, the answers parallel those we have already seen in the political corruption, white collar, and labor racketeering areas. 10
 
A. Jurisdictional Use of RICO

In theory, RICO can serve as a device to obtain federal jurisdiction to prosecute common-law crimes against persons or property that would normally be within the province of local law enforcement. A purse snatcher or mugger may seem the quintessential example of the sort of street criminal who is the responsibility of the local police force and the local district attorney to apprehend and prosecute. If two or three muggers, however, form a loosely knit gang and can be shown to have cooperated in two or more such robberies they have become -- assuming that some effect on interstate commerce can be fantasized -- a RICO enterprise, and can be prosecuted federally. 11 Given the higher public concern about street crime than about the less immediately threatening sorts of activity that compose the principal objects of federal law enforcement efforts, an ambitious United States Attorney might well want to grab a piece of this kind of crime-fighting action by using RICO to stake out a federal presence in the war on street robbery.

The use of RICO purely as a jurisdictional device to supplement or preempt local law enforcement efforts against ordinary street crime has so far been mostly a theoretical fear. The Department of Justice's RICO Guidelines emphasize the importance of avoiding encroachments on local law enforcement, 12 and specifically disapprove RICO  [*922]  indictments "where the predicate acts consist solely and only of state offenses" except in certain instances. 13 This might be dismissed as mere lip-service to the values of federalism and, in fact, the exceptions are sufficiently amorphous to provide little barrier to prosecutions of entirely local criminal activity. Nevertheless, pure cases of disregard for these principles are rare. 14

Still, the jurisdictional use of RICO is not without significance. Even in cases that present only one straightforward conspiracy involving a single type of criminal activity, jurisdiction and venue barriers may prevent prosecutors from presenting all the evidence relating to the single conspiracy in the same trial. In United States v. Winter, 15 for example, federal investigators had developed evidence of a wide-ranging  [*923]  scheme to fix horse races at various race tracks. Although the case was prosecuted in Massachusetts, the scheme affected racing in New Hampshire, Rhode Island, Pennsylvania, and New Jersey. Such multistate activity may be difficult to prosecute efficiently under conventional doctrines of criminal law and procedure. Substantive criminal acts may be committed in a variety of jurisdictions, making it impossible to join all related transactions in the same venue. While a conspiracy encompassing the entire scheme may be prosecuted in any district where any overt act was committed, 16 unless the substantive offenses can be prosecuted in the same district, the federal conspiracy statute may not provide penalties commensurate with the scope of the criminal conduct involved. 17 RICO may function, in cases like Winter, as a jurisdictional device to prosecute trans-jurisdictional schemes -- or, in effect, as an aggravated conspiracy statute. 18

Congressional expansion of federal criminal jurisdiction makes it possible for federal law enforcement agencies to reach most of the forms of organized criminal conduct that have been attacked using RICO, under certain circumstances. Theft, 19 arson, 20 extortion, 21 prostitution, 22 gambling, 23 and of course narcotics trafficking, 24 are all covered by federal statutes. While the use of RICO to assert jurisdiction over criminal conduct that has not yet been made subject to federal intervention could be seen as abusive, as in the area of political corruption, 25 the frequent resort by Congress to artificial jurisdictional devices to secure federal jurisdiction often makes the dividing line between federal and state jurisdiction arbitrary. As to any individual  [*924]  criminal act, this patchwork jurisdiction might not matter. Unlike the case of local governmental corruption, 26 there is no systemic reason to think that the absence of federal jurisdiction to prosecute an act of gambling or arson will lead to failure to prosecute the crime. 27 Where, however, the criminal conduct is part of a pattern of criminal transactions conducted by an organized criminal conspiracy, the fragmented pattern of jurisdiction may prevent any court from having jurisdiction over enough of the case to permit unified prosecution of the entire operation in a manner that makes significant penalties available. RICO has been of genuine value in permitting such prosecutions. 28
 
B. RICO as a Penalty Enhancer

In some cases, the impetus for the use of RICO in criminal enterprise cases appears to be, as in the white collar and labor cases, 29 its extreme, mandatory and procedurally simple financial penalties. At least some of the large number of narcotics cases in which RICO counts are included can probably be accounted for on this basis. 30 In most narcotics cases, of course, there is no need for RICO to enhance the incarceration penalties available; federal narcotics offenses carry potentially drastic sentences. 31 Nor do narcotics offenses present the jurisdiction and venue barriers to unified prosecution discussed in the preceding section. 32

 [*925]  On the other hand, as with other forms of criminal activity, the financial penalties attached to narcotics violations have not necessarily been commensurate with the extraordinary profit potential of such violations. The fines provided by federal narcotics laws, while steeper than those applicable to mail fraud, conspiracy, and other white collar offenses, until recently have not been remotely comparable to the profits of major drug dealers. 33 Even if the fines had been significantly higher, however, they would not be adequate to take the profit out of drug dealing. The collection of fines presents a far graver problem in the case of narcotics dealers, whose assets are generally underground, than in the case of white collar offenders, who more commonly have visible sources of payment. While narcotics offenses always carried the potential for civil forfeiture actions, 34 before the Comprehensive Crime Control Act of 1984, RICO's procedurally efficient criminal forfeiture remedy was only available under the narcotics statutes against the supervisors and managers of the ring subject to prosecution under the Continuing Criminal Enterprise statute. 35 Prosecutors may have been induced to add RICO counts to ordinary narcotics conspiracy charges in order to avail themselves of its automatic forfeiture provisions. 36

This hypothesis is supported by several inferences that can be drawn from the sample data. RICO indictments seem to be used in narcotics cases with some regularity where the narcotics ring involved was particularly extensive or profitable. In addition, the narcotics cases include a relatively large number of references to application of the forfeiture remedy, 37 or to the involvement of a legitimate business organization  [*926]  as a front for narcotics operations. 38 Such references suggest that RICO is used where an element of infiltration of legitimate business suggests that invocation of RICO is particularly apt, or where the availability of forfeiture offers an alluring additional sanction.

One might question, however, whether the use of RICO to achieve marginally greater penalties in cases in which extensive financial penalties and extremely harsh jail terms are already available to deter and punish justifies the existence of the statute. As in the case of white collar and labor offenses, if weak financial sanctions in narcotics laws hinder their effective enforcement, a direct approach to that problem that avoids the dangers and complexities of RICO is obviously preferable. 39 Moreover, Congress in 1984 responded to that problem by enacting a dramatic increase in the fines applicable to narcotics offenses, 40 providing an alternative fine equal to twice the gross profits or proceeds from narcotics where even the enhanced fines are insufficient, 41 and adapting the criminal forfeiture procedures pioneered by RICO to all felony narcotics cases. 42 If additional penalties are needed in the fight against narcotics dealers, RICO is not necessary to provide them.

In some cases, RICO prosecutions are predicated not only on narcotics violations, but on other offenses committed in furtherance of narcotics activity. One common additional offense in RICO prosecutions of criminal activity is corruption of law enforcement. In some cases, the extent of corruption related to narcotics, gambling or prostitution is so extreme that the law enforcement organization itself becomes the focus of prosecutorial interest. 43 In others, the corruption  [*927]  of law enforcement is merely an adjunct to a prosecution directed principally at the criminal group's primary activities. 44

In such cases, in addition to its value as a sentence enhancer, addition of a RICO charge also serves to bolster the prosecution's ability to join charges for trial. Even without RICO, crimes committed in order to obstruct an investigation might be successfully added to an indictment for the principal offenses as "parts of a common scheme or plan," 45 or treated as aspects of a continuing conspiracy. 46 The use of an overarching RICO offense, however, serves to clinch the prosecution's ability to present the corruption charges together with the underlying narcotics or other violations.

This motivation for use RICO is especially strong where the offenses sought to be joined are very serious and prejudicial, and thus particularly likely to be severed in the exercise of the trial judge's discretion. 47 Several cases in which the principal focus of the criminal enterprise was the distribution of narcotics exemplify this phenomenon. United States v. Thomas, 48 for example, involved "a huge narcotics ring run by a governing body called the 'Council.'" 49 In addition to narcotics offenses, the predicate acts charged against the defendants, all allegedly important members or associates of the Council, included the murder of workers in the narcotics enterprise thought to be potential informers or otherwise threatening to the power of its leaders. 50 Without RICO, the government's ability to charge these murders would  [*928]  have been dubious. First, except in unusual circumstances, murder is a state offense, not subject to federal jurisdiction. Second, even if jurisdiction could be obtained in some circumstances, 51 a strong argument could be made that the joinder of acts of violence would unfairly prejudice the jury's ability to judge the evidence of narcotics crimes. The effort to avoid these arguments against use of the murder evidence undoubtedly encourages use of RICO in such cases. 52 The use of RICO to incorporate collateral crimes of violence into indictments charging conspiracies to carry on narcotics or other "victimless" activities, however, like its use to overcome jurisdictional boundaries to joinder of related offenses, is only a minor instance of a phenomenon that reaches its fulfillment in massive indictments of organized crime members, or members of other sorts of criminal organizations, for a wide variety of separate types of criminal activities, united only by the common purpose of the organization to promote the long-range economic or political goals of its members. It is now time to consider that phenomenon.
 
C. The Multi-Faceted Criminal Enterprise

RICO has thus been used by federal authorities to make possible a wide variety of prosecutions that could not have been brought effectively without it. But while these cases show the use of RICO to obtain jurisdiction, to increase available sanctions, and to effect otherwise dubious joinder of offenses, the resulting trials would not look terribly strange to a criminal lawyer who had spent the last twenty years in a Rip Van Winkle-style slumber. For all of the advantages RICO has conferred upon prosecutors, the trials in the cases discussed so far have been easily recognizable as bribery trials, narcotics conspiracy cases, labor racketeering cases, and so on. In each, two or more reasonably closely related criminal acts, usually of the same general nature, have been charged against a reasonably manageable number of defendants. Although the jury is asked, at the end, not only to decide the defendants'  [*929]  guilt on the various specific acts, but also to make an overall determination whether those acts formed a "pattern" in connection with an "enterprise," the rules of procedure and evidence, and the ultimate tests of guilt, have hardly been revolutionary.

But what would our attorney Van Winkle make of United States v. Castellano? 53 The indictment in Castellano charged twenty-four defendants in seventy-eight counts. Count 1 of the indictment alone charged all of the defendants and a number of unindicted "co-racketeers" with violating section 1962(c), in that they conducted the affairs of a "crew" or subdivision of an organized crime family through a pattern of eighty separate acts of racketeering, 54 including "twenty-six murders, bribery, extortion, narcotics violations, thefts from interstate shipments, mail and wire fraud, obstruction of justice, transportation of stolen property, and transportation of women for purposes of prostitution," extending over a period of more than eleven years. 55 Several of the acts of racketeering themselves consisted not of individual actions, but of complex conspiracies and schemes, or were technically duplicitive in that they charged more than one crime arising out of the same criminal episode. 56 Numerous acts of racketeering charged could not have been prosecuted as criminal offenses in their own right, either because the statute of limitations had already run on those offenses, 57 or because they had been the subject of previous prosecutions. 58 Moreover, a large number of those acts could never have been charged independently in federal court for lack of jurisdiction, and in any event almost none of the racketeering acts could have been prosecuted in the government's chosen venue, because "virtually every significant racketeering act alleged in the indictment" occurred outside that district. 59

While the range of activities charged against the enterprise was vast, the involvement of many of the defendants in those activities could only be described as tangential. For example, one defendant was a juror who allegedly took a bribe to fix a prosecution of one of the members of the "crew"; another's entire involvement (limited to three of the eighty racketeering acts charged in the indictment) consisted of hiring members of the crew to bribe and ultimately to murder a witness against his son in a state prosecution entirely unrelated to the affairs of the enterprise; others were involved only in one of the many criminal  [*930]  affairs of the enterprise, and not at all in its more violent activities. 60 All of these defendants were to be tried together, despite the fact that little of the evidence in what could only be an extraordinarily lengthy trial would have any direct bearing on their own actions. Moreover, the government's proof would not be limited to the actions of the defendants on trial. The actions and fates of several alleged co-racketeers "not [named as] defendant[s] herein because [they were] murdered" would also be proved, as would the fact that the "defendants are part of organized crime, and particularly the Mafia." 61

Our Rip Van Winkle of a defense lawyer would probably regard the notion that such a case could be tried in this form as the product of a demented prosecutor with delusions of grandeur and no understanding of the rules of procedure and evidence. Yet the government's indictment, and its plan to try the case in a single proceeding, were sustained in virtually every particular by a scholarly and moderate district judge in a careful opinion that persuasively demonstrates that RICO permits all of these consequences. 62

Nor is Castellano an aberration. The fourteen cases from our sample classified as involving "diversified syndicates," and most of the cases categorized as involving "violence and extortion" or "political violence," 63 involved criminal conduct almost as diverse, and problems of trial procedure almost as complex, as the Castellano prosecution. The cases present various patterns. In some, like Castellano, the RICO enterprise is explicitly identified as a traditional organized crime family -- in effect, an arm of the Mafia -- engaged in the provision of unlawful goods and services (prostitution, gambling, narcotics, loan-sharking), appropriation of the property of others (theft, fraud, extortion) and crimes of violence attendant on the operation and enforcement of such schemes. 64 Others involve groups of criminals that are smaller or that  [*931]  lack formal affiliation with traditional organized crime groups, but that engage in similar patterns of ongoing, organized criminality. 65 At least one case involves the efforts of a criminal group to control a legitimate industry through a pattern of criminal activity, in a classic pattern of violent infiltration of legitimate business. 66

In addition to these groups of organized criminals, RICO has been used against groups whose methods may sometimes resemble those of traditional organized crime syndicates, but whose goals are quite different -- political terrorists. In these cases, too, RICO has enabled prosecutors to link together a wide range of different offenses committed by numerous different individuals, linked together by common aims, overlapping patterns of complicity in different crimes and general awareness that others committed to the same goals were engaged in similar illegal acts, in ways that would be impossible under traditional rules of joinder, jurisdiction and venue. For example, in United States v. Bagaric, 67 prosecutors presented evidence of "at least fifty acts of extortion carried out in this country, two murders of extortion victims, several unsuccessful efforts to murder, approximately a dozen bombings and attempted bombings . . . and the transportation of weapons and explosives from coast to coast," committed by a terrorist group of Croatian nationalists. 68 Although venue was laid in New York, the activities of the ten defendants in the thirteen-week trial centered on Chicago, Los Angeles, Cleveland and Toronto, and included evidence of criminal acts in New York, Pittsburgh, San Francisco, Milwaukee, Akron and Bridgeport -- as well as Canada, Paraguay and West Germany -- over a  [*932]  seven-year period. 69 Although the testimony proved conclusively that the convicted defendants were all committed members of the group, and had each been involved in two or more of the specific violent acts alleged, the cast of characters involved in each act of racketeering proved by the government differed. 70 Several similar prosecutions, involving other extremist political sects, can be cited. 71

While the number of cases fitting this pattern is small in comparison to the number of cases involving government corruption, business fraud, labor racketeering or more specialized criminal organizations, 72 it is in these cases that RICO presents its most innovative face, and its most significant challenge to orthodox notions of criminal law, procedure and evidence.
 
D. The Transaction-Based Model of Crime

In order to understand RICO's value in prosecuting diversified illicit enterprises, and the potential abuses of such prosecutions, we must first understand the limits imposed on criminal prosecutions by our conventional understanding of what a crime is, and the potential of RICO to explode those limitations.

Fundamental to our traditional law of crimes, criminal procedure and evidence is a conception of crime that is transaction-bound. Synthesizers  [*933]  of the common-law tradition tell us that the core of any definition of crime is a particular act or omission. 73 That act or omission is conceived as taking place in an instant of time so precise that it can be associated with a particular mental state of intention, awareness of risk, or neglect of due care. 74 The verbs that form the heart of the definitions of particular offenses ("takes and carries away," "engages in sexual intercourse," "damages by starting a fire," "sells a controlled substance") typically refer to single rather than repeated actions, completed in a brief span of time. Where the verbs in penal statutes instead refer to causing a particular result ("causes the death of another human being," "causes serious physical injury") -- a process that can extend over a period of time -- the focus of inquiry into a defendant's culpability must nevertheless be a specific, momentary act or omission. 75 Even the crime of conspiracy, which in practice may permit an examination of an extended course of conduct by one or more individuals, does so in the guise of using that course of conduct as evidence from which to infer that a particular act of "agreement" occurred, presumably at a specific, if not precisely ascertainable, moment in time. 76

Of course, while the criminal act itself must generally meet this criterion, other elements in the definition of a crime sometimes allow expansion of the relevant time-frame. Attendant circumstance elements can have this effect. For example, in the crime of rape, the act of penetration provides an identifiable instant in which the crime is complete. But the required circumstance that the act of intercourse be the product of forcible compulsion, which really represents the crux of the offense, may significantly blur the time boundary of the inquiry, and make identification of the precise act that makes the conduct criminal (as opposed to the act that completes the offense) more difficult. In  [*934]  borderline cases of intimidation and implied threat, it may be difficult to determine which, if any, of a series of actions by an accused rapist constituted the culpable application of force. Similarly, defenses such as duress or self-defense make relevant courses of conduct, by the defendant or another, that break the boundaries of the brief transaction that constitutes the charged offense. Most dramatically, the insanity defense can make the defendant's entire life history the subject of a trial.

Most importantly, the mental element of most crimes (and, in the case of conspiracy, the mental act of agreement itself) will often make relevant a course of conduct extending beyond the specific criminal transaction, because of the need to prove such mental states inferentially. For example, intent or premeditation may be shown by prior activities that show planning for the crime. Proof of motive, which is also relevant to proving intent, may require an extensive inquiry into the background of the specific act charged. But almost all of these possible expansions of the scope of the inquiry are anchored to the particular act or transaction in question in the case. The expanded inquiry is always directed at ascertaining the circumstances or mental state of the accused at a particular instant identified by the act charged. Other acts or events are relevant only to the extent they support an inference about that question.

The focus on particular events in defining crimes is not merely a linguistic convention. The requirement that criminal punishment be based on a specific act has deep roots. The very nature of criminal punishment, as distinct from other uses of the compulsive power of the state (such as mandatory treatment for physical or mental illness), requires that a person not be punished for bad character, tendency to commit crime, or even a specifically formulated intention to commit some particular prohibited act. Before the state can deprive a citizen of liberty in a punitive way, the individual must manifest that character or tendency by the commission of some concrete prohibited act. 77

In significant part, the purpose of this limitation is the protection of an individual from punishment for thoughts or traits not yet exemplified by actual harmful conduct. 78 But the moral basis of the focus on particular acts extends beyond this problem. Even for those accused of committing what is unquestionably a concrete, particular offense, we are careful to guard against the possibility that a defendant may be convicted and punished for bad character rather than for the particular act charged. The insistence on incident-based liability thus has important consequences for our rules of procedure and evidence.

Since the crime with which a defendant is charged took place at a  [*935]  particular moment in time, the relevance of a defendant's actions prior to that moment is always problematic; some tendency to support an inference about what the defendant did or what he thought at the moment of the crime is always necessary for such a prior act or event to matter. 79 We are particularly concerned about the relevance of prior actions which themselves constitute crimes. Because we fear that a jury will "irrationally" conclude that a person who has committed prior crimes will be guilty of the offense for which he stands accused on a particular occasion, or will dismiss the very question of his present guilt in favor of a condemnation of his general bad character, evidence of prior crimes -- as well as evidence of general bad character or criminal associations -- is usually excluded from evidence. Such information is admissible only where its particular relevance to the specific act charged greatly outweighs the "prejudice" it occasions by distracting the jury from the only question properly before it -- the defendant's actions in the particular incident being examined. 80 For similar reasons, charges that a defendant is guilty of more than one offense, or that two or more defendants are guilty of joint crimes, may only be tried together where the charges are so closely related that it would be manifestly inefficient to have separate trials. 81 Where joint trials of different alleged offenses or offenders are permitted, we are -- in theory at least -- careful to guard against the danger that evidence relevant to one crime will unfairly "spill over" into what ought to be a clinically pure evaluation of the evidence concerning another. 82 A criminal trial thus tends to focus on a particular incident or transaction.

The transaction-based model is so fundamental to our ways of thinking about criminal law that we tend to take it for granted. Professor Kelman has pointed out the importance of questioning the functions of such "unconscious interpretive constructs" that "shape the way we view disruptive incidents." 83 He views "narrow time-framing" and a tendency to treat incidents as "disjoined" or separate transactions (central aspects of the transaction-based model) as tools to "buttress the traditionally asserted intentionalism of the criminal justice system," and suppress the political choices inherent in imposing punishment. 84

Both the procedural and substantive manifestations of the model  [*936]  of crime based on specific incidents or acts are indeed associated with a particular conception of the individual as a moral actor. The careful elaborations in our penal codes of the precise nature of the prohibited acts, and the equally careful calibrations of the degree of blameworthiness to be attributed to different prohibited acts, seem to presuppose actors with a free will to avoid the prohibited conduct, who can fairly be apportioned different degrees of guilt or punishment based on the nature of the conduct in which they have chosen to engage. Indeed, our rules of procedure seem to carry this notion of moral freedom even further. The individual is implicitly conceived not only as free in principle to act in accordance with or in violation of defined norms, but also as free at any given moment to make choices at odds with any consistent character that may be deduced from his prior acts. To infer that a defendant committed the particular offense for which he is being tried from the fact that he has previously committed other crimes of a generally similar nature -- or, worse still, other crimes of an entirely different nature -- is not only unfair, but inconsistent with a fundamental supposition that criminal behavior is punishable because it represents a free choice at a particular moment in time to commit an immoral act. 85

Indeed, the power of this model of the individual is so strong that some proponents of the "just deserts" model of punishment have argued that the focus on the individual incident rather than on the character of the offender should be extended even into the sentencing process. On this view, a defendant's past conduct or overall character would have no relevance at all in determining an appropriate sentence, giving especially concrete content to the idea of punishing the crime and not the criminal. 86 At this point, however, our tradition until recently has balked, and the sentencing decision has been seen, within limits set by a vague principle of proportionality and by concrete maximum sentences devised by legislatures in correlation to the seriousness of particular offenses, as including appropriate attention to treatment and incapacitation goals based in part on the general character of the offender. 87 The prevalence of legislative proposals for less discretionary, more conduct-based sentencing systems may suggest that the retributive view of crime may be weakening even the citadel of sentencing discretion.

 [*937]  But Professor Kelman is wrong in his claim that these moral notions are the sole basis of the transaction model, and that they have served to obscure the function of criminal law as a means of social control. The historical roots of the transaction model reach back far beyond the relatively recent philosophical arguments with which Kelman associates it. Indeed, the notion that individuals (or at least nobles) should not be punished except for defined conduct -- according to the law of the land -- has roots in the explicitly political demands of the barons at Runnymede not to be subjected to punishment by arbitrary fiat from above. It is precisely because criminal punishment constitutes an exercise of power by the strong over the weak that the weak have demanded limitations on its exercise, including that punishment only be imposed on a showing of particular conduct, defined in advance.

If such a system is to function, a trial has to be about something relatively concrete. 88 The historical and political roots of the transaction model show, on one hand, that that model cannot merely be dismissed as a mask for fundamentally arbitrary exercises of power; they suggest, on the other, that to those not committed to a purely retributivist position, the attributes of that model are contingent -- part of the important but adjustable balance that the criminal law must always maintain between the exercise of social control and the maintenance of individual liberty.
 
E. The Enterprise Offense

RICO prosecutions of criminal enterprises present a serious challenge to the substantive and procedural implications of this transaction-based model of crime.

This challenge is partially apparent on the face of the statute. Ordinary criminal statutes, as we have seen, define the conduct they prohibit in terms of rather concrete actions that can be committed in an identifiable moment of time. Indeed, two of the three substantive prohibitions imposed by the RICO statute in essence follow this very model. Those sections make it a crime to "use or invest" money from particular sources in a particular way, 89 and to "acquire . . . any interest" in an enterprise by means of certain conduct. 90 While the necessity  [*938]  of proving a "pattern of racketeering activity" may well permit proof of a variety of (possibly only distantly related) criminal acts, the act that constitutes the offense is a single, specific action -- acquisition of a business interest. The particular moment at which an individual commits the prohibited act can, in theory and usually in practice, be identified. Past acts of racketeering are relevant to the offense charged only if they bear directly on the particular acquisition of an interest charged in the indictment. 91

Section 1962(c), in contrast, makes it a crime to "conduct or participate, directly or indirectly, in the conduct" of the affairs of any "enterprise[ ] . . . through a pattern of racketeering activity." 92 The very words of the statute reveal an intent to prohibit not any particular, time-bound action, but a course of conduct extending over a potentially lengthy period of time. Although the predicate acts of racketeering are conventional crimes, defined in terms of specific conduct, the actual RICO violation is not identifiable by the physical contours of a particular action or effect. Rather, the defining characteristic of the "pattern of racketeering" is the relationship of certain conduct to other conduct and to the "enterprise," which itself is an abstract construct of certain interpersonal relationships. Whether or not this definition is vague in the technical legal sense of the word, 93 the level of abstraction in the definition permits the offense to cover a wide variety of conduct for which ordinary language does not supply a single common term. 94

RICO is such an oddity among penal statutes that its exponents frequently claim that it is not really a criminal statute at all, arguing that "RICO is a remedial, as opposed to substantive, statute" because "[t]he  [*939]  provisions of section 1962 do not create 'new crimes' but serve as the prerequisites for the invocation of increased sanctions for conduct which is proscribed elsewhere in both federal and state criminal codes." 95 But this claim is misleading. In formal terms, RICO is plainly a criminal statute; each of its provisions, including section 1962(c), defines a certain cluster of behaviors as a new crime. Like section 1962(c), a statute establishing a higher penalty for certain murders defined as "first-degree" does not prohibit conduct previously lawful, but rather establishes "prerequisites for the invocation of increased sanctions for conduct which is [already] proscribed." 96 But such statutes are certainly substantive criminal laws, in any commonly understood meaning of the term.

The distinction between remedial and substantive statutes is not merely formal or rhetorical; it has serious procedural and substantive consequences. If RICO were truly only a remedial statute which added an additional sentencing consideration to affect the punishment meted out for what were elsewhere defined as criminal acts, there might well be no need to require that the sentence-enhancing element be proved to a jury beyond a reasonable doubt; the recidivist or organized enterprise element might be deferred to a post-trial sentencing hearing, with the necessary findings to be made before a judge subject to a lower standard of proof. 97

Moreover, it is the fact that RICO does define a crime that entails some of its most dramatic procedural and evidentiary consequences.  [*940]  Since section 1962(c) defines participating in the affairs of an enterprise through a pattern of racketeering as a crime separate and apart from the predicate acts, it does not merely enhance the statutory penalty for the predicate acts, but rather permits the imposition of consecutive sentences for the RICO offense and the predicates. 98 Because the RICO offense is a separate crime, the statute of limitations runs only from its completion; thus, every additional racketeering offense committed in furtherance of the enterprise's affairs within ten years of a previous one extends the statute of limitations for another five years for prosecution of the entire pattern. 99 A RICO indictment thus may hold a defendant accountable for acts that took place twenty or more years before the date of the indictment -- not for the penalty attached to the predicate crime, but for the separately defined RICO offense.

Even within the ordinary limits of the double jeopardy principle and the statute of limitations, a prosecutor can use section 1962(c) to place before a single jury in a single trial offenses that could not otherwise be included in the same indictment or admitted into evidence at the same trial. Suppose, for example, the authorities develop evidence that the same defendant from whom they have recently made an undercover purchase of narcotics is a member of an organized crime family who committed a contract killing three years earlier. Under our ordinary, transaction-bound rules of procedure and evidence, the defendant would have to be tried separately for each offense. Since the earlier crime is plainly not part of the same course of events as the later, joinder of the two crimes would not be possible; if the homicide had taken place in another state, jurisdictional or venue problems would also prevent joinder. 100

In a trial on the narcotics charge alone, moreover, the evidence of a prior homicide committed by the defendant would likely be excluded as irrelevant and highly prejudicial. Evidence that the defendant in a narcotics trial was part of the "Mafia" would surely be excluded as merely prejudicial evidence of the defendant's character and associations. And the prosecutor presumably would not even think about trying to elicit evidence of crimes that some other member of the same crime family had committed, in which this particular defendant was not personally involved. Evidence of the defendant's involvement in organized crime or of the murder he may have committed might finally surface after the defendant's conviction, as part of an argument for a  [*941]  severe sentence. 101

If the case could be indicted and tried under RICO, however, all of the evidence regarding this defendant's activities could easily be presented in the same trial. Since the government would have to allege and prove a pattern of racketeering activity, the murder and the narcotics offense could be alleged as elements of the same crime, the violation of section 1962(c). The rules precluding admission of evidence of other crimes, consequently, would simply have no application -- evidence of the homicide would not be evidence of a prior crime, but evidence of the very offense charged in the indictment.

Jurisdictional and venue problems disappear, as well. It is irrelevant that the federal government lacks jurisdiction to prosecute ordinary homicides; the crime charged here is racketeering that affects interstate commerce, not murder. The single crime of racketeering, like any other crime, can be prosecuted in any district where a portion of the crime was committed, 102 so any venue problem with combining crimes committed in different districts disappears.

The government would also have to allege and prove that the crimes were committed in furtherance of the affairs of an enterprise, so the prosecution would be permitted to show the existence, purposes and structure of the organized crime family, and the defendant's membership in it. 103 Even if no other defendant were on trial, this may necessitate reference to criminal activities committed by other members of the organization, as examples of its continuing nature, hierarchical structure, or purposes as an entity; if the defendant were indicted along with several other alleged members of the same organized crime family, as is commonly done in RICO prosecutions, their crimes would of course have to be proved too. Joining those defendants in the same indictment would automatically be proper, of course; since the defendants were all jointly charged with the same crime -- the RICO violation -- we are faced not with the joinder of several separate offenses by different actors, but with a single offense all the defendants are alleged to have committed together. 104

All of these procedural consequences stem from the fact that violation  [*942]  of section 1962(c) is defined as a single crime, and our procedural system, for reasons alluded to above, attaches considerable importance to the concept of a crime as a unified event, distinct from other crimes. If RICO's effects are principally remedial, in the sense that the statute's importance lies not in the prohibition of certain conduct, but in the procedural and sentencing consequences of committing conduct already defined as criminal, it accomplishes most of those effects precisely by the fact that it is indeed, in formal terms, a substantive criminal statute.

In substance, as well as in form, section 1962(c) defines a substantive crime. The RICO offense is not reducible to the predicate acts of racketeering. If the jury determines beyond a reasonable doubt that the defendant committed those acts, it still must find an additional element before it can convict: that the predicate acts were committed in the conduct of the affairs of an enterprise. 105

The significance of this additional element varies considerably in different types of RICO cases. Where, as in the bulk of the labor, business and governmental crimes, 106 the enterprise is a more or less formal entity, the structure of the section 1962(c) offense is quite conventional: the prohibited conduct is the commission of the predicate acts, with the relation of the crimes to an enterprise serving as an aggravating factual circumstance. Like most such aggravating factors (possession of a weapon, causing injury, entry at night), the existence of the enterprise and the relation of the criminal conduct to it are relatively easily ascertainable, noncontroversial facts. 107

Where the enterprise is an illegitimate association-in-fact, however, the existence of the enterprise is not merely an easily established formal element of proof. Rather, the existence of the enterprise is both potentially controversial and genuinely significant in legally differentiating  [*943]  RICO offenses from mere aggregations of predicate crimes. Indeed, it can be argued that the enterprise element constitutes the essence of the crime. Operation of a criminal organization -- unlike operation of a business corporation -- is not morally neutral. Nor is it merely an incidental fact about the context in which a criminal act was committed. Rather, it constitutes a distinct species of social harm.

The arguments made by appellants in Turkette 108 and Elliott 109 illustrate the importance of the enterprise element. In each case, the defendants' legal and factual claims were that the government had, at best, shown that various individuals had committed various distinct crimes. In effect, they were asserting the factual accuracy and legal necessity of applying a transactional view of crime to their various antisocial acts. The courts rejected this argument, however, holding that the whole offense was indeed greater than, or at least distinct from, the sum of its parts -- that, at least on the facts of those cases, it was legitimate to hold the defendants guilty not only of a series of separate criminal transactions, but of entering into a relationship, exemplified by a course of conduct over a period of years, that itself was criminal. It is the operation of the criminal enterprise through criminal acts, not merely the commission of the acts themselves, that constitutes the crime of RICO.

But whether or not a group of individuals, who, in various combinations, committed a series of predicate offenses, constituted an enterprise, and whether each of the defendants was part of that enterprise, are not always questions of historical fact. Where a criminal group has a sufficiently tangible organization, it may be possible to confirm the existence of the enterprise, and to identify someone as a "member." 110 But one need not be a "member" of an organization to participate in the conduct of its affairs, 111 and, of course, not all illicit enterprises are so conveniently structured. As is often true with the "agreement" that is the actus reus of conspiracy, the jury is not necessarily being asked to decide whether a particular event occurred. Rather, it is being asked to impose a conceptual construct on the events that it finds took place. 112  [*944]  What the jury is being asked to decide is whether the defendant's acts should be treated as evidence of a commitment to a criminal association. 113

Such a commitment is not, in any conventional sense, an "act." The jury's task is to assess in a global way the nature of the defendant's involvement in a network of criminal activities and associations, to determine whether the total picture of the defendant's criminal career permits the judgment that he has become part of an underworld "enterprise." If character can be defined as the residue of a series of moral decisions, the jury in a very real sense is being asked to make a judgment on the defendant's character.

In making such a judgment, the jury is entitled to rely not only on evidence of the defendant's own crimes, but also on evidence of the crimes of those with whom he is alleged to have thrown in his lot. Such evidence is excluded from the transaction-model trial, precisely because it may distract the jury from its responsibility of deciding what the evidence shows about a particular act. In an illicit-enterprise RICO trial, it is admitted, precisely because the jury is asked to make a judgment not only about what discrete acts the defendant committed at particular moments in time, and what his intention was with respect to each act at those moments, but also about how those acts fit into his entire moral life: Were they parts of a pattern? Were they committed as part of his association with a subculture of crime? 114

RICO illicit association cases thus pose both a substantive and a procedural challenge to the transaction-based model of criminal law. Substantively, the standard legal texts tell us that a distinct act or omission is the core event constituting a crime, 115 and academic analyses of  [*945]  penal codes, including ones that in some ways radically attack the Anglo-American consensus, 116 are principally concerned with articulating the precise circumstances in which specified acts should be subject to condemnation as crimes. The distinctive nature of criminal punishment, we are told, is that it represents a societal response to and judgment upon particular moral actions, rather than to a person's character, status, or intentions. The RICO illicit association cases, in contrast, demand a more global judgment about a defendant's character and loyalties. To be found guilty, it is not enough that the defendant has committed specific criminal acts; those acts must be part of an ongoing commitment to the values of a criminal organization.

Our procedural and evidentiary rules support the substantive values of the transaction-based model of crime by rigorously focusing the trial process on information that bears directly on demonstrating what happened, in the physical world and in the defendant's consciousness, during the particular transaction under examination. RICO trials, however, permit a much wider exploration of the context of the particular predicate acts, both in the defendant's history, and within the institutions and communities of which he is a part.
 
F. RICO and Conspiracy

The challenge RICO presents to conventional criminal law thinking is not without precursors. Indeed, it can be argued that the practical and theoretical problems presented by RICO prosecutions have long been festering under the law of conspiracy. 117 There is considerable truth to this observation, although for reasons developed below, RICO constitutes a more extreme departure from the traditional model.

In theory, conspiracy, unlike section 1962(c), functions as an inchoate crime, criminalizing an agreement to perform prohibited acts without regard to the consummation of the criminal plan. If two or more people sit at a table and expressly agree to rob a bank the next day, they are guilty of conspiring to rob the bank. 118 Structurally, the crime fits the transaction model: although the "act" of agreeing is a somewhat bloodless one, at least in this simple hypothetical it is clear that the defendants have done something beyond engaging in antisocial  [*946]  thoughts or having deformed characters. They have taken a particular step toward accomplishment of social harm that can, at least in theory, be demonstrated to have occurred at a specific time and place. 119

In practice, however, three principal complicating factors undermine this theory, and bring conspiracy closer in its effects to RICO. First, even where such a simple express agreement to join in a criminal activity has been made, direct evidence of such an agreement will not often be available. 120 Absent an informer or electronic surveillance, the authorities will never know when and where the agreement was made, or what were its precise terms. Accordingly, the making of the agreement will ordinarily have to be inferred from the actions of the parties to it. After the bank robbery has occurred, one may determine from the apparently planned coordination of the robbers' actions that an agreement had been made. 121 Moreover, the agreement need not be express at all -- the agreement to commit the crime, though actual, may be made without words. 122 This further complicates the difficulty of inferring its terms.

Second, the scope of possible conspiratorial agreements is both wide and not clearly defined. It is well established that a single conspiracy can include among its objects the commission of several crimes -- either multiple violations of the same statute or violation of several statutes. 123 Thus, it is perfectly legitimate to charge defendants with conspiring not merely to rob one bank, but to rob several, or to rob a bank and buy drugs with the proceeds, or to commit mail fraud, evade taxes, and obstruct justice. Whether the evidence shows one conspiracy with multiple objects or several distinct conspiracies is essentially a question of fact: "the precise nature and extent of the conspiracy must be determined by reference to the agreement which embraces and defines its objects." 124

 [*947]  This would present little problem for traditional theory if we had a tape recording of the meeting at which a group of criminals agreed to their unlawful plan. But when the breadth of potential conspiratorial agreement is combined with the likely reliance on circumstantial evidence to prove the agreement, and the possibility that the agreement was only implicit in any event, the concept of agreement begins to lose its moorings. In the context of a typical narcotics conspiracy prosecution, it is plain that all of the participants at various levels of a distribution network have not "agreed," even implicitly, on any precise series of actions; at best, various individuals at different levels have agreed to engage in certain specific acts of possession or distribution, with the understanding that the acts and agreements of others at other levels are necessary for the success of their venture. The nature of the required agreement has subtly changed from the paradigm with which we began: the typical conspiracy to distribute narcotics does not involve an express agreement to engage in specific acts, but a series of mutually unconnected decisions to engage in a business known to involve a high degree of mutual interdependence. 125

The requirement of agreement is further diluted by the frequently repeated doctrine that a conspirator does not need to agree to, or even to know about, all of the objects of the conspiracy in order to be liable for joining it. 126 This doctrine is virtually an inevitable consequence of allowing the formulation of indictments charging conspiracies with multiple objects that extend over a period of time. It is extremely likely that the parties to such an agreement will change over time, and that adherents may be recruited to execute portions of the plan without being made aware of all of its contours. 127 It serves the convenience of prosecutors and courts to say that these new recruits are "members" of  [*948]  the conspiracy. But when each conspirator knows only some of the objects of the "agreement," it becomes difficult to see what reality remains to the notion of agreement. Moreover, because the overall agreement itself is usually merely an inference from the concrete offenses committed by the putative conspirators, what can be found in retrospect to be a single "conspiracy" embracing multiple objects may well consist merely of a series of crimes united only by rather casual links among the perpetrators.

Third, as with RICO, the procedural and evidentiary consequences directly or indirectly associated with a conspiracy charge make conspiracy charges attractive to prosecutors, and create possibilities of abuse. 128 Charging that various substantive offenses are objects of a unitary conspiracy usually permits joinder of those offenses in a single trial, permits selection of a favorable venue, and facilitates introduction of hearsay evidence. 129 These procedural advantages more readily account for the widespread use of conspiracy charges in federal criminal cases than any difference between the substantive elements of conspiracy and consummated offenses.

When these factors are considered, the widespread assumption that RICO vastly expands the ambit of possible conspiracy prosecutions 130 becomes somewhat puzzling. Let us assume that our hypothetical conspirators sitting at the table have more resources, and more ambition, than a handful of potential bank robbers. The proposal on the table, in consequence, is that the conspirators would pool their existing networks for narcotics distribution, gambling and prostitution; divide up the city into territories in which each would have an exclusive franchise; intimidate potential informers or complainants by murder and arson; and diversify their activities by infiltrating members of the ring into a securities firm in an effort to steal negotiable bonds and launder the proceeds of their illicit activities. Had the FBI been recording this meeting on a court-authorized "bug," is there any reason why those present at the meeting could not be charged with conspiring to commit a series of state or federal crimes?

 [*949]  There is no readily apparent reason why not. Certainly, the diversity of the criminal objects of the conspiracy should not be an obstacle. As noted above, so long as the conspirators have joined in a single common agreement, that agreement can have as its object the violation of several statutes. There is no conceptual reason why an agreement to commit fraud, evade taxes and obstruct justice should be indictable as conspiracy, while an agreement to spread the net more widely and commit a broader range of crimes should not.

United States v. Elliott, 131 the decision that popularized the notion of RICO as a super-conspiracy statute, contains the most extensive judicial effort to distinguish between conspiracy law and sections 1962(c) and (d). 132 The Elliott court announced that "RICO has displaced many of the legal precepts traditionally applied to concerted criminal activity. Its effect in this case is to free the government from the strictures of the multiple conspiracy doctrine and to allow the joint trial of many persons accused of diversified crimes." 133 But the court's discussion of just how this is so is perplexing.

The crux of the court's "doubt that a single conspiracy could be demonstrated" rests essentially on two observations. 134 First, the court points out that various subgroups of the alleged conspirators had no contact with each other. 135 But as we have seen, and as the Elliott court elsewhere concedes, "'a party to a conspiracy need not know the identity, or even the number, of his confederates'" -- let alone have direct contact with them. 136 This could hardly be a basis for rejecting a jury finding of a single conspiracy.

Second, and more significantly, the court states that "[t]he activities allegedly embraced by the illegal agreement in this case are simply too diverse to be tied together on the theory that participation in one activity necessarily implied awareness of others." 137 Essentially the court here is holding that the prosecution has failed to meet its burden  [*950]  of proving that a single agreement existed. If the government is relying on inference to show the overall agreement, the inference must be sufficient to persuade a reasonable jury beyond a reasonable doubt that the common plan existed, and it is not unreasonable to argue that mere proof of a collection of entirely unrelated crimes is insufficient for this purpose.

But then, how does RICO come "to the [r]escue"? 138 The problem, according to the court, is that in organized crime cases, the need to infer a common agreement "inhibited mass prosecutions because a single agreement or 'common objective' cannot be inferred from the commission of highly diverse crimes by apparently unrelated individuals." 139 The solution is that "RICO helps to eliminate this problem by creating a substantive offense which ties together these diverse parties and crimes." 140 Since the RICO objective is to "participate in the affairs of an enterprise through a pattern of racketeering activity," and not to commit the particular substantive crimes that serve as the predicate acts, the lack of apparent relation between the crimes is irrelevant "so long as we may reasonably infer that each crime was intended to further the enterprise's affairs." 141

This is simply double talk. According to the court's analysis, what was missing for conviction of a "traditional" conspiracy was not a legally sufficient objective, but evidence sufficient to support a finding that that objective existed in this particular case. Explaining the inadequacy of the evidence under conventional conspiracy law, the court had stated:

Even viewing the "common objective" of the conspiracy as the raising of revenue through criminal activity, we could not say, for example, that Foster, when he helped to conceal stolen meat, had to know that J. C. was selling drugs to persons unknown to Foster, or that Delph and Taylor, when they furnished counterfeit titles to a car theft ring, had to know that the man supplying the titles was also stealing goods out of interstate commerce. 142
 
The court thus seems to accept that an agreement to raise revenue through miscellaneous criminal activity is indictable as a conspiracy, holding only that the evidence here did not show such an  [*951]  agreement. 143

But if there was insufficient proof to infer a single agreement to "rais[e] revenue through criminal activity," how can the proof in turn be sufficient -- as the court says it must be -- to permit finding "agreement on an overall objective" to "further the enterprise's affairs"? 144 After all, the enterprise in this case consists of nothing other than an association to raise revenue by committing crimes. 145 Conversely, if it is permissible to infer from a defendant's participation in particular concrete crimes that he simultaneously agreed to participate in the affairs of an overarching enterprise, why not just rename the "enterprise" a "conspiratorial agreement with multiple criminal objects" and infer his agreement to that? 146

Within the court's confusion, however, lurks a truth. Although the borders of permissible conspiracy prosecutions remain unclear, and in theory encompass an agreement to form a gang and commit whatever crimes the leadership orders, the broad agreement by courts and commentators that RICO still somehow expands the range of conduct that can be prosecuted as a conspiracy strongly suggests that courts and prosecutors recognized limits, if not on the kinds of agreement that could in principle be indicted as conspiracies, at least on the kinds of criminal conduct that would be permitted in practice to support an inference of a unifying scheme. The Elliott court was no doubt both sincere and accurate in stating that it would not have permitted the defendants there to have been convicted of a simple conspiracy. And whatever courts might have accepted if tested, few precedents can be found in "traditional" conspiracy cases for agreements of the breadth and complexity of RICO illicit association cases involving diversified criminal syndicates. 147 RICO thus may be better seen as the occasion  [*952]  for a change in judicial and prosecutorial policy than as a provider of new theoretical concepts.

Three points need to be made about this analysis, however. First, the change in policy itself is significant. Whether RICO conspiracies involve the recognition of a new conspiratorial objective or greater latitude for prosecutors to charge and juries to find agreements that -- had they been express and proven by direct evidence -- would always have been recognized as illegal conspiracies, the prosecutions that result are distinctly broader than had previously been undertaken. Earlier conspiracies, however large and long-lasting they were alleged to be, typically were bounded by a single type of illegal activity (such as narcotics or gambling) or by an easily described intermediate-range objective involving a continuous flow of activity (such as conspiracies to commit securities fraud and evade taxes, or to hijack trucks, kidnap drivers, and bribe policemen). RICO prosecutions, as we have seen, have not been so limited.

Second, by incorporating state crimes as predicate activity, RICO does in fact expand the range of criminal objects that can be prosecuted in a single federal prosecution. Yoking the plenary subject matter jurisdiction of the states to the plenary geographic jurisdiction of the federal government permits unification of offenses that could not previously have been brought together in a conspiracy prosecution in any jurisdiction. This is particularly significant in organized crime cases, where it permits unified prosecution of multiple crimes of violence that would previously have been regarded as unrelated.

Third, by creating a substantive offense worded in terms of a course of conduct, RICO constitutes a theoretical break with the transactional model of crime. Section 1962(c) explicitly recognizes a crime that can be described as a course of conduct involving relationships with criminal groups, rather than as a single moral act. Although conspiracy prosecutions in practice permit presentation of a course of conduct  [*953]  in a single trial, in theory at least the crime remains defined in terms of a single act of agreement.

This conceptual change may have practical consequences. However attenuated the concept of agreement becomes in the actual administration of conspiracy law, the need to anchor the crime in a hypothetical instant of agreement may have helped courts to maintain some boundary around the kind of conduct that would be permitted to be tried as a unitary conspiracy. The fiction that all of the crimes charged were part of a specific agreement could only be maintained if it plausibly could be imagined that the core players at least could have sat together and agreed -- even if it was clear that they did not in the particular instance do so. "Participation in the affairs of an enterprise" more accurately captures the reality of what fringe participants in a conspiracy do than "agreeing" to the overall objectives of its core members, and therefore permits easier inferences of guilt.

While the Elliott court was concerned with the relation between ordinary conspiracies and RICO conspiracies under section 1962(d), the effect of sections 1962(c) and 1962(d) in these respects is essentially identical. Indeed, the difficulty of distinguishing between section 1962(c) and section 1962(d) in illicit association cases reflects the radical difference between those cases and other RICO violations.

A single individual can violate section 1962(a) or (b) or (c) in the context of a legitimate enterprise, without the aid of accomplices. A conspiracy to violate these sections therefore reflects a comprehensible concept and, subject to the debate about whether conspiracy ever identifies a harm distinct from the substantive offense that is its object, constitutes a distinct crime. In the illicit enterprise cases, however, the overlap between the completed and conspiracy offense is total. At least if a single criminal cannot be both an enterprise and a defendant, 148 the criminal enterprise whose affairs are conducted through a pattern of racketeering can only exist to the extent its members have agreed to form it. Each individual member, moreover, can only be guilty of conducting his affairs through a pattern of racketeering to the extent that he knows that his particular crimes are part of a larger criminal enterprise with which he has voluntarily associated himself -- precisely the mental state required to join the RICO conspiracy. 149

 [*954]  This overlap does not mean that courts were wrong to reject the arguments that section 1962(d) is an incoherent effort to penalize "conspiring to conspire," 150 or that multiple punishment for violating sections 1962(c) and (d) violates the double jeopardy clause. 151 An inchoate conspiracy to violate section 1962(c) is a comprehensible crime, if one rarely to be encountered. Had they been arrested immediately after their conclave, the conspirators deciding to found an organized crime syndicate hypothesized above would have violated section 1962(d), but not section 1962(c). 152 And so long as the Blockberger rule is applied at the level of abstract elements of the offense rather than in the context of the facts of a particular case, 153 it remains true that section 1962(c) is a separate offense from section 1962(d) because an agreement is not a formal element of the substantive offense. But it does mean that in the context of illicit enterprise prosecutions, the prohibition of conspiracies serves no purpose and should not be included in a modified statute.

Traditional conspiracy law, in short, is an important precursor of RICO, and indeed in theory might well encompass many of the results that have been reached with that statute. The limits of traditional conspiracy law are to be found not in the types of goals that a conspiratorial agreement could in theory have, but in judicial policing of the inferences that would be drawn from participation in various kinds of crimes. Within those limits, conspiracy law has fostered an erosion of the transaction model of criminal law and criminal procedure. Like RICO, conspiracy permits prosecutors to present complex events in a single criminal proceeding, and to avoid focusing on particular, identifiable acts or transactions. 154 But RICO has been the vehicle by which  [*955]  courts have considerably expanded those limits, and thus permitted still greater departures from the transaction model.
 
G. The RICO Model Evaluated

1. The Challenge to Substantive Law. -- Much of the criticism of RICO implicitly assume the immutable validity of the conventional model of criminal law and procedure; the critics apparently believe that to note the ways in which RICO departs from the conventional is sufficient to condemn it. 155 But the transaction-based model of criminal law is not beyond challenge. To evaluate the legitimacy of the RICO illicit enterprise cases, we must question why we have that model, whether we abide by it in practice, and what it costs us to preserve it.

One value served by the transaction model of crime is its preclusion of punishment in the absence of behavior manifesting a concrete threat of harm to legitimate social interests. The notion that a defendant is being punished merely for his character or status, or for the danger that he potentially represents, is often said to be offensive to our concepts of fairness. 156 A person may be a seething mass of antisocial ideas, repulsive character traits, dangerous tendencies, and even concrete evil plans, but he is neither a criminal nor subject to punishment until he commits specific proscribed acts. Even if society might gain by engaging in preemptive strikes against such villains, it is essential to the liberty and security of ordinary citizens that government not be permitted to deprive them of freedom unless they violate clearly defined norms. "Character" or "predicted danger" are flexible and unpredictable standards of decision, too easily used as tools of oppression.

These substantive concerns, however, are not directly violated by RICO. Although the distinguishing features of RICO are its somewhat amorphous associational and course of conduct elements, a fundamental prerequisite of a substantive RICO violation is the commission of particular criminal acts. These predicate racketeering acts are themselves conventional, transactionally defined crimes, requiring the commission of particular conduct for their violation. RICO does not permit  [*956]  a person to be convicted on the basis of his thoughts, tendencies, intentions, or character alone; defined and concrete antisocial acts are a part of the definition of the crime.

A comparison with the unconstitutional statute in Robinson v. California 157 is instructive. California law made it a crime to be a narcotics addict, without requiring any particular narcotics-related action. The Supreme Court's opinion could be read as suggesting two possible grounds for finding this statute offensive: the absence of a concrete proscribed act and the defendant's arguable inability to control his actions. We now know that the first of these grounds was controlling; nothing in the Robinson principle precludes punishment of the addict's possession or use of narcotics. 158 The critical defect of the statute was that it did not require proof that the defendant had actually done anything within the state's boundaries.

RICO does not contravene the Robinson principle. It is not a crime under section 1962(c) to have the character or status of a racketeer, but to be a racketeer who commits acts of racketeering. Nor would RICO raise the problems associated with the second branch of the Robinson case. To the extent that the elements of association with an illicit enterprise and pattern of criminal conduct define a "status" of "racketeer," that "status" is neither passive (unlike addiction, it cannot be acquired in utero or by medical treatment) nor uncontrollable (unlike addiction, continued involvement in the criminal enterprise is not the product of physiological compulsion).

Nevertheless, RICO does make aspects of a defendant's background and associations an aggravating factor in a defined crime. But our law has never held that such factors are irrelevant to the quantum of punishment that can be administered to one who has violated concrete norms. The practical operation of law enforcement, in fact, frequently aims to do precisely what a purely retributive model of criminal law tells us is illegitimate: to punish based on character and propensity rather than on the condemnation of specific actions. Considerations of "character" and "danger to the community," often said to be banished from the definition of criminal conduct, pervade the critical discretionary decisions of all three of the agencies responsible for applying those definitions: prosecutors, juries, and sentencing judges. 159

 [*957]  A defendant's prior criminal acts and associations with organized criminal groups, as well as all sorts of more amorphous features of his character, are permissible factors for a judge to take into account in imposing sentence, and are regularly used for that purpose. 160 The current discontent with broad judicial sentencing discretion does not, for the most part, dispute the wisdom of basing sentencing decisions on the character of the offender as well as of the offense; rather, the goal is to reduce the perceived unfair disparity of sentencing by systematizing the value to be assigned to various factors. 161 The Supreme Court's recent death penalty jurisprudence, indeed, requires capital sentencing to take account of a broad range of mitigating factors. 162

The acceptability of such considerations in the exercise of prosecutorial discretion and jury nullification is less well documented, since those discretionary features of the system are even less subject to appellate oversight than sentencing. But any experienced plea-bargainer knows that a defendant's prior record is a critical determinant of the treatment the defendant will be offered. 163 Nor is this practice highly controversial. Few would condemn a prosecutor for offering leniency to a young bank teller with no criminal record and a stable background who embezzled a sum of money, while insisting on prosecuting another teller who had committed a similar act, but who had prior convictions  [*958]  and who was believed to be a member of a youth gang. Such discretion is clearly not based on the relevance of the prior conduct to any element of the potential charge, 164 but on a substantive judgment about the seriousness to be attributed to the defendant's misconduct in light of a factual context beyond the temporally and spatially limited transaction with which the penal code is concerned.

Jurors' exercise of their power to enter a verdict of acquittal notwithstanding the facts and the law is an even murkier topic, 165 but it seems clear that a conviction is less likely if the jury is not persuaded that the defendant is a bad actor who deserves condemnation. Certainly this is the practical belief of most prosecutors and defense attorneys. Trial lawyers devote considerable effort to influencing the jury's perception of a defendant's character. Apparently these efforts pay off. Studies of jury behavior indicate that jurors' reactions to the characteristics of the defendant play a major role in explaining verdict disagreement between judges and juries. 166

The importance of these considerations to the substantive concerns of the criminal law is obvious. The protection of the public from criminal conduct is a central purpose of organized society. Given the substantial evidence that a large proportion of the criminal acts are committed by a relatively small number of repeat offenders, 167 it makes sense to single out such offenders for special attention from prosecutors, investigators and sentencers. Organized criminal groups in particular make possible the infliction of greater harm than can be committed by individuals. 168 Critics of conspiracy law have justly pointed out that this "group danger" argument does not provide a sufficient rationale for the separate punishment of every agreement covered  [*959]  by the conspiracy concept. 169 But many of the RICO illicit association cases exemplify precisely the sort of case in which even the critics implicitly acknowledge that the "group danger" argument makes sense. 170 Many of the crimes charged in those cases could not have been committed without the existence of an organized enterprise.

Moreover, criminal punishment is not only aimed at controlling or reducing the tangible harms that flow from particular forms of antisocial conduct. An important function of criminal justice is the reinforcement of the solidarity of the law-abiding community by the formal condemnation of offenders. 171 This function is most effectively served where it is manifest that the criminal being punished is not merely a person who has committed a single mistake, but someone who can plausibly be cast as a person of deeply flawed character who has chosen to live outside the value system of the dominant group.

This last perception, indeed, may be particularly relevant to the wrong committed by organized criminals. Professor Fletcher has argued strongly that recidivist criminals should not be punished more severely than first-offenders out of resentment towards their rebelliousness. Unlike parents of defiant teenagers, he contends, lawmakers in a liberal society "are not entitled to react to a 'persistent' criminal as though their personal authority were challenged." 172 But a criminal who chooses to give his allegiance to the code of the Mafia rather than to the code of civil society 173 is not merely challenging the "personal authority" of lawmakers. 174 He is inflicting a deep injury on the social fabric, both by declaring his own intent to prey upon those  [*960]  who agree to live within the law and by challenging the law-abiding citizen's belief that adherence to the norms of society is required, protected, and rewarded.

For the criminal justice system to reinforce adhesion to social norms, it must not only distinguish those to be condemned from ordinary citizens who merely yield to temptation, it must also actually and forcefully condemn those it finds worthy of condemnation. Little is more debilitating to the willingness of ordinary citizens to live within the law than a perception that persistent violators of the law are unpunished. In the area of organized crime, the disparity between social reality and legal reality can become particularly acute, as the legal system strives to isolate a defendant's particular acts from their personal and social context. Respect for law is hardly fostered when the legal system myopically focuses on isolated, perhaps minor, offenses of individuals whose entire lives make plain their complete commitment to a career of organized lawbreaking.

If facts extrinsic to a particular criminal transaction are significant to the proper operation of the criminal justice system, it is not clear why they should be confined to the shadows of its "discretionary" components. Delineation of society's official code of morality is one purpose of a penal code, and the character of the offender is clearly irrelevant to this purpose. But a penal code also shares the purposes of the law enforcement system of which it is a part. 175 So long as the values protected by the requirement that concrete conduct be a part of the definition of crime are preserved, there is no special reason why factors relating to the defendant's character and associations should not be included in the definition of crimes.

Rather, removing these factors from the shadows of discretionary decisionmaking can serve valuable ends. If there is broad social agreement that considering the context of violations is legitimate, 176 making that consideration explicit can only increase public understanding and approval of the system. Moreover, express inclusion of aggravating circumstances based on professional criminality furthers the goal of consistent and accurate decisionmaking, by subjecting allegations of organized crime involvement to the same procedural safeguards as other factors bearing on guilt. There is no procedural regularity at all to a prosecutor's determination that a particular defendant is associated with organized crime; a jury's judgment of character in a trial from which evidence going beyond the transaction at issue is in principle excluded can only be fragmentary and unreliable, and even a judge's fact-finding at sentencing is required to meet only minimal procedural standards. If membership in organized crime is relevant to our judgment of  [*961]  a defendant, engaging that issue in a public trial process, subject to ordinary procedural safeguards, is not in principle inappropriate.

2. The Challenge to Criminal Procedure. -- From the perspective of the traditional model of criminal prosecution -- especially the perspective of defense attorneys within that tradition -- it is difficult to see RICO trials of the illicit association type as anything but abominations. As described above, 177 the definition of the elements of a RICO violation of this type leads inexorably, by application of the ordinary rules of evidence and procedure, to the undermining of the principles underlying those rules, and the reversal of their ordinary consequences. Thus, for example, rules requiring that evidence be relevant to some element of the particular offense charged i