Regulating the Market for Snitches

Buffalo Law Review

Volumn 47, Issue 1, 1999, Pages 563-631

  Weinstein, Ian ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0039744813     PISSN: 00239356     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (297)

1. "Cooperation" is a term of art for the process by which a federal criminal defendant gains the possibility of sentence mitigation by providing assistance in the prosecution or investigation of others. "Snitching," "substantial assistance" and "ratting" are all synonymous with cooperation, if not quite interchangeable. The Oxford English Dictionary notes that "snitch" is slang and of obscure origin. Its first definition is "A fillip † on the nose)," the second is "The Nose" and the third is "An informer; one who turns King's or Queen's evidence." The first citation for the use of "snitch" as informer is to a 1785 dictionary entry, followed by an 1800 citation from Byron and many others. OXFORD ENGLISH DICTIONARY 858 (2d ed. 1989); see also Daniel C. Richman, Cooperating Clients, 56 OHIO ST. L.J. 69, 72 n.13 (1995) (associating "snitch" with the nose and so with informing).
2. "Cooperation" is a term of art for the process by which a federal criminal defendant gains the possibility of sentence mitigation by providing assistance in the prosecution or investigation of others. "Snitching," "substantial assistance" and "ratting" are all synonymous with cooperation, if not quite interchangeable. The Oxford English Dictionary notes that "snitch" is slang and of obscure origin. Its first definition is "A fillip † on the nose)," the second is "The Nose" and the third is "An informer; one who turns King's or Queen's evidence." The first citation for the use of "snitch" as informer is to a 1785 dictionary entry, followed by an 1800 citation from Byron and many others. OXFORD ENGLISH DICTIONARY 858 (2d ed. 1989); see also Daniel C. Richman, Cooperating Clients, 56 OHIO ST. L.J. 69, 72 n.13 (1995) (associating "snitch" with the nose and so with informing).
3. I am not aware of any data on cooperation rates before the Sentencing Commission began collecting and analyzing data pursuant to its statutory directive, 28 U.S.C. § 995(12)-(16) (1994 & Supp. 1997), but the relatively sparing use of cooperation in the first years of the Guidelines is consistent with the sense of experienced practitioners that cooperation was not as widespread before the Guidelines as it is now. In 1989, the first year data was available, only 3.5% of the cases involved substantial assistance departures. In 1990 the number climbed to 7.5%, and continued to move up to 11.9% in 1991, 15.1% in 1992 and 16.9% in 1993. In 1994 the figure reached 19.5% and has remained at about that level, 19.7% in 1995 and 19.2% in 1996. See UNITED STATES SENTENCING COMM'N, 1996 SOURCEBOOK OF FEDERAL SENTENCING STATISTICS 39 fig.G (1996) [hereinafter 1996 FED. SENT. SRCBK.]. This information is also available at the United States Sentencing Commission Homepage (Jan. 30, 1999) 〈http://www.ussc.gov〉 (on file with author and the Buffalo Law Review). A federal judge and former federal prosecutor has pointed to these same statistics to support the view that cooperation has increased since the introduction of the Sentencing Guidelines. See John Gleeson, Supervising Criminal Investigations: The Proper Scope of the Supervisory Power of Federal Judges, 5 J.L. & POL'Y 423, 424 n.9 (1997).
4. I am not aware of any data on cooperation rates before the Sentencing Commission began collecting and analyzing data pursuant to its statutory directive, 28 U.S.C. § 995(12)-(16) (1994 & Supp. 1997), but the relatively sparing use of cooperation in the first years of the Guidelines is consistent with the sense of experienced practitioners that cooperation was not as widespread before the Guidelines as it is now. In 1989, the first year data was available, only 3.5% of the cases involved substantial assistance departures. In 1990 the number climbed to 7.5%, and continued to move up to 11.9% in 1991, 15.1% in 1992 and 16.9% in 1993. In 1994 the figure reached 19.5% and has remained at about that level, 19.7% in 1995 and 19.2% in 1996. See UNITED STATES SENTENCING COMM'N, 1996 SOURCEBOOK OF FEDERAL SENTENCING STATISTICS 39 fig.G (1996) [hereinafter 1996 FED. SENT. SRCBK.]. This information is also available at the United States Sentencing Commission Homepage (Jan. 30, 1999) 〈http://www.ussc.gov〉 (on file with author and the Buffalo Law Review). A federal judge and former federal prosecutor has pointed to these same statistics to support the view that cooperation has increased since the introduction of the Sentencing Guidelines. See John Gleeson, Supervising Criminal Investigations: The Proper Scope of the Supervisory Power of Federal Judges, 5 J.L. & POL'Y 423, 424 n.9 (1997).
5. I am not aware of any data on cooperation rates before the Sentencing Commission began collecting and analyzing data pursuant to its statutory directive, 28 U.S.C. § 995(12)-(16) (1994 & Supp. 1997), but the relatively sparing use of cooperation in the first years of the Guidelines is consistent with the sense of experienced practitioners that cooperation was not as widespread before the Guidelines as it is now. In 1989, the first year data was available, only 3.5% of the cases involved substantial assistance departures. In 1990 the number climbed to 7.5%, and continued to move up to 11.9% in 1991, 15.1% in 1992 and 16.9% in 1993. In 1994 the figure reached 19.5% and has remained at about that level, 19.7% in 1995 and 19.2% in 1996. See UNITED STATES SENTENCING COMM'N, 1996 SOURCEBOOK OF FEDERAL SENTENCING STATISTICS 39 fig.G (1996) [hereinafter 1996 FED. SENT. SRCBK.]. This information is also available at the United States Sentencing Commission Homepage (Jan. 30, 1999) 〈http://www.ussc.gov〉 (on file with author and the Buffalo Law Review). A federal judge and former federal prosecutor has pointed to these same statistics to support the view that cooperation has increased since the introduction of the Sentencing Guidelines. See John Gleeson, Supervising Criminal Investigations: The Proper Scope of the Supervisory Power of Federal Judges, 5 J.L. & POL'Y 423, 424 n.9 (1997).
6. 18 U.S.C. § 3553(e) (1994). See also U.S. SENTENCING GUIDELINES MANUAL § 5K1.1 (1997) [hereinafter U.S.S.G.] (stating the policy regarding criminals who provide substantial assistance to authorities).
7. The Sentencing Commission has reported sentencing statistics on a fiscal year basis, October 1 to September 30, since 1990. See 1996 FED. SENT. SRCBK., supra note 2, at 39 fig.G n.1. Between October 1, 1995, and September 30, 1996, 19.2%, or 7845 of the total 40,879 criminal sentences imposed under the Sentencing Guidelines involved sentence mitigation (downward departures from the otherwise applicable range established by the Sentencing Guidelines) to reward substantial assistance. See id. at tbl.26.
8. For a discussion of sentencing disparities, see infra pp. 601-17.
9. For the impact of excessive cooperation on the adversary system, see infra pp. 617-22.
10. For discussion of our strong moral ambivalence about snitching, see infra pp. 622-25.
11. For discussion of why some cooperation may not have significant law enforcement value, see infra pp. 613-14.
12. The problem of the externalization of the costs of prosecution are discussed by Robert L. Misner, along with his proposal for a market incentive based approach to regulating discretion. See Robert L. Misner, Criminal Law: Recasting Prosecutorial Discretion, 86 J. CRIM. L. & CRIMINOLOGY 717, 766-71 (1996) (urging tying prosecutorial discretion to the availability of prison resources to internalize costs and provide an incentive to maximize strategic law enforcement planning). Two classic examples of economic analysis of discretion in the criminal justice system are Frank H. Easterbrook, Criminal Procedure as a Market System, 12 J. LEGAL STUD. 289, 290 (1983) (discussing the incentives favoring plea bargaining and arguing that prosecutorial discretion is an efficient and fair regulator) and Stephen J. Schulhofer, Criminal Justice Discretion as a Regulatory System, 17 J. LEGAL STUD. 43 (1988) (explaining that prosecutorial discretion in charging, bargaining and sentencing does not lead to the most efficient results because of problems of public management; even if it were most efficient, fairness concerns would still justify some controls on discretion).
13. The problem of the externalization of the costs of prosecution are discussed by Robert L. Misner, along with his proposal for a market incentive based approach to regulating discretion. See Robert L. Misner, Criminal Law: Recasting Prosecutorial Discretion, 86 J. CRIM. L. & CRIMINOLOGY 717, 766-71 (1996) (urging tying prosecutorial discretion to the availability of prison resources to internalize costs and provide an incentive to maximize strategic law enforcement planning). Two classic examples of economic analysis of discretion in the criminal justice system are Frank H. Easterbrook, Criminal Procedure as a Market System, 12 J. LEGAL STUD. 289, 290 (1983) (discussing the incentives favoring plea bargaining and arguing that prosecutorial discretion is an efficient and fair regulator) and Stephen J. Schulhofer, Criminal Justice Discretion as a Regulatory System, 17 J. LEGAL STUD. 43 (1988) (explaining that prosecutorial discretion in charging, bargaining and sentencing does not lead to the most efficient results because of problems of public management; even if it were most efficient, fairness concerns would still justify some controls on discretion).
14. The problem of the externalization of the costs of prosecution are discussed by Robert L. Misner, along with his proposal for a market incentive based approach to regulating discretion. See Robert L. Misner, Criminal Law: Recasting Prosecutorial Discretion, 86 J. CRIM. L. & CRIMINOLOGY 717, 766-71 (1996) (urging tying prosecutorial discretion to the availability of prison resources to internalize costs and provide an incentive to maximize strategic law enforcement planning). Two classic examples of economic analysis of discretion in the criminal justice system are Frank H. Easterbrook, Criminal Procedure as a Market System, 12 J. LEGAL STUD. 289, 290 (1983) (discussing the incentives favoring plea bargaining and arguing that prosecutorial discretion is an efficient and fair regulator) and Stephen J. Schulhofer, Criminal Justice Discretion as a Regulatory System, 17 J. LEGAL STUD. 43 (1988) (explaining that prosecutorial discretion in charging, bargaining and sentencing does not lead to the most efficient results because of problems of public management; even if it were most efficient, fairness concerns would still justify some controls on discretion).
15. One general argument is that prosecutors find the immediate benefits of plea bargaining so attractive that they overshadow relatively weak concerns about its systemic costs. See Stephen J. Schulhofer, Plea Bargaining as Disaster, 101 YALE L.J. 1979 (1992) (providing an example of one of the most insightful and persistent criticisms of plea bargaining, arguing for the elimination of plea bargaining). For discussions of pre-Guidelines plea bargaining, see generally Albert W. Alschuler, The Changing Plea Bargaining Debate, 69 CAL. L. REV. 652 (1981) (criticizing plea bargaining as irrational, unfair and reaching results unrelated to the proper objectives of criminal prosecution) and PAMELA J. UTZ, SETTLING THE FACTS: DISCRETION AND NEGOTIATION IN CRIMINAL COURT (1978) (defending plea bargaining).
16. One general argument is that prosecutors find the immediate benefits of plea bargaining so attractive that they overshadow relatively weak concerns about its systemic costs. See Stephen J. Schulhofer, Plea Bargaining as Disaster, 101 YALE L.J. 1979 (1992) (providing an example of one of the most insightful and persistent criticisms of plea bargaining, arguing for the elimination of plea bargaining). For discussions of pre-Guidelines plea bargaining, see generally Albert W. Alschuler, The Changing Plea Bargaining Debate, 69 CAL. L. REV. 652 (1981) (criticizing plea bargaining as irrational, unfair and reaching results unrelated to the proper objectives of criminal prosecution) and PAMELA J. UTZ, SETTLING THE FACTS: DISCRETION AND NEGOTIATION IN CRIMINAL COURT (1978) (defending plea bargaining).
17. One general argument is that prosecutors find the immediate benefits of plea bargaining so attractive that they overshadow relatively weak concerns about its systemic costs. See Stephen J. Schulhofer, Plea Bargaining as Disaster, 101 YALE L.J. 1979 (1992) (providing an example of one of the most insightful and persistent criticisms of plea bargaining, arguing for the elimination of plea bargaining). For discussions of pre-Guidelines plea bargaining, see generally Albert W. Alschuler, The Changing Plea Bargaining Debate, 69 CAL. L. REV. 652 (1981) (criticizing plea bargaining as irrational, unfair and reaching results unrelated to the proper objectives of criminal prosecution) and PAMELA J. UTZ, SETTLING THE FACTS: DISCRETION AND NEGOTIATION IN CRIMINAL COURT (1978) (defending plea bargaining).
18. A total of 60,255 defendants were "disposed of" in the federal courts in 1996. Dismissals accounted for 7,083 of those defendants. Of the remaining 53,172 defendants, 4976 (9.4% of the defendants whose cases were not dismissed) went to trial. Of those 4976 defendants who went to trial, 902 (18% of all defendants who went to trial) were acquitted and 4074 (82% of all defendants who went to trial) were convicted. The remaining 48,196 (90.6% of the defendants whose cases were not dismissed) entered pleas of guilty or nolo contendre. BUREAU OF JUSTICE STATISTICS SOURCEBOOK OF CRIMINAL JUSTICE STATISTICS, 448 tbl.5.27 (1996) [hereinafter BUREAU OF JUSTICE SRCBK.]. The rate of trials in the state courts appears similar, with an average plea rate of 89% reported for felony convictions in state courts in 1994. Id. at 471 tbl.5.52.
19. See, e.g., Santobello v. New York, 404 U.S. 257, 261-62 (1971) (taking the inducement to plead as a given in holding that plea agreements are enforceable). That inducement, in the form of sentence mitigation, has been formalized in the Sentencing Guidelines system. See U.S.S.G. § 3E1.1. The Guidelines permit the judge to reduce a defendant's "offense level" by two or three points, depending on the severity of the offense level and the judge's evaluation of the degree of the acceptance of responsibility. The maximum of the range at a given offense level exceeds the minimum by six months or 25%, whichever is greater. The maximum of the range of each offense level (above level 12) is the mid-point of the next higher offense level range and the minimum of the range two levels higher. Thus, a two point reduction makes what was formerly the minimum sentence now the maximum sentence, while a three point reduction puts the maximum about 14% lower than the former minimum. The overall inducement can be as low as no mitigation (the judge can impose the same sentence as the maximum after plea as could have been imposed as the minimum after trial, assuming a two point reduction) to an almost 40% reduction in sentence, comparing the high end of the range at the unreduced offense level with the low end of the range three offense levels lower. Defendants may even receive modest relief from statutorily mandated mandatory minimum sentences for drug offenses if they agree to plead guilty, disclose their criminal activities, are first time offenders and meet certain other requirements. This safety valve provision is codified at 18 U.S.C. § 3553(f) (Supp. 1997). See Philip Oliss, Mandatory Minimum Sentencing: Discretion, the Safety Valve, and the Sentencing Guidelines, 63 U. CIN. L. REV. 1851, 1883 (1995) for a discussion of the safety valve provision.
20. The Supreme Court has noted on many occasions that charging decisions are within the: "special province" of the Executive. Heckler v. Chaney, 470 U.S. 821, 832 (1985). The Attorney General and United States Attorneys retain " 'broad discretion' " to enforce the Nation's criminal laws. Wayte v. United States, 470 U.S. 598, 607 (1985) (quoting United States v. Goodwin, 457 U.S. 368, 380, n.11 (1982)). They have this latitude because they are designated by statute as the President's delegates to help him discharge his constitutional responsibility to "take Care that the Laws be faithfully executed." U.S. Const., Art. II, § 3; see 28 U.S.C. §§ 516, 547. As a result, "[t]he presumption of regularity supports" their prosecutorial decisions and, "in the absence of clear evidence to the contrary, courts presume that they have properly discharged their official duties." United States v. Chemical Foundation, Inc., 272 U.S. 1, 14-15 (1926). In the ordinary case, "so long as the prosecutor has probable cause to believe that the accused committed an offense defined by statute, the decision whether or not to prosecute, and what charge to file or bring before a grand jury, generally rests entirely in his discretion." Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978). United States v. Armstrong, 517 U.S. 456, 464 (1996) (holding that discovery or a selective prosecution claim is only required after a threshold showing that others of a different race who were otherwise similarly situated were not prosecuted).
21. The sentencing reforms of the 1980s have increased the relative importance of prosecutorial discretion. Although those changes have reduced discretion generally, they have much more significantly reduced judicial sentencing discretion, leaving prosecutors with a greater degree of control over sentencing than they enjoyed under the old law. See Cynthia K.Y. Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, 50 RUTGERS L. REV. 199 (1997); Misner, supra note 9, at 719. See generally Cynthia Kwei Young Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, 42 UCLA L. REV. 105 (1994) (arguing that a prosecutor's discretion over sentencing issues is not related to discretion regarding bargaining or charging). Others have argued that power has shifted to Congress. See James B. Burns et al., We Make the Better Target (But the Guidelines Shifted Power from the Judiciary to Congress, Not from the Judiciary to the Prosecution), 91 NW. U. L. REV. 1317 (1997).
22. The sentencing reforms of the 1980s have increased the relative importance of prosecutorial discretion. Although those changes have reduced discretion generally, they have much more significantly reduced judicial sentencing discretion, leaving prosecutors with a greater degree of control over sentencing than they enjoyed under the old law. See Cynthia K.Y. Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, 50 RUTGERS L. REV. 199 (1997); Misner, supra note 9, at 719. See generally Cynthia Kwei Young Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, 42 UCLA L. REV. 105 (1994) (arguing that a prosecutor's discretion over sentencing issues is not related to discretion regarding bargaining or charging). Others have argued that power has shifted to Congress. See James B. Burns et al., We Make the Better Target (But the Guidelines Shifted Power from the Judiciary to Congress, Not from the Judiciary to the Prosecution), 91 NW. U. L. REV. 1317 (1997).
23. The sentencing reforms of the 1980s have increased the relative importance of prosecutorial discretion. Although those changes have reduced discretion generally, they have much more significantly reduced judicial sentencing discretion, leaving prosecutors with a greater degree of control over sentencing than they enjoyed under the old law. See Cynthia K.Y. Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, 50 RUTGERS L. REV. 199 (1997); Misner, supra note 9, at 719. See generally Cynthia Kwei Young Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, 42 UCLA L. REV. 105 (1994) (arguing that a prosecutor's discretion over sentencing issues is not related to discretion regarding bargaining or charging). Others have argued that power has shifted to Congress. See James B. Burns et al., We Make the Better Target (But the Guidelines Shifted Power from the Judiciary to Congress, Not from the Judiciary to the Prosecution), 91 NW. U. L. REV. 1317 (1997).
24. For problems with direct regulation of prosecutorial discretion, see infra pp. 626-29.
25. For a discussion of lowering defendants' incentives, see infra pp. 629-31.
26. For an evaluation of measures to limit the demand for cooperators, see infra pp. 630-32.
27. Gleeson, supra note 2, at 424.
28. See Graham Hughes, Agreements for Cooperation in Criminal Cases, 45 VAND. L. REV. 1, 8-14 (1992) (tracing the history of inducements to snitching); The Whiskey Cases, 99 U.S. 594 (1878) (discussing historical and late nineteenth century practices and holding that a defendant could not plead compulsion to testify against an accomplice as a formal bar to prosecution but could delay trial and apply for executive clemency to which he had equitable title).
29. For the history of the development of the professional prosecutor in America, see Misner, supra note 9, at 728-31.
30. This Article does not discuss the use of immunity agreements, requests for which peaked in 1986 at 2550 and have fallen somewhat steadily to 1493 in 1996. See BUREAU OF JUSTICE SRCBK., supra note 11, at 418 tbl.5.1. Although they are fairly viewed as the "extreme variant" of the cooperation agreements discussed here, Richman, supra note 1, at 151 n.14, they involve the dynamic of witness and prosecutor - a very different relationship from that of defendant and prosecutor, particularly when the prosecutor has power over sentence mitigation.
31. See generally Hughes, supra note 19, at 10-13 (discussing the development of modern practices and the importance of the sanctioning of plea bargaining in Santobello v. New York, 404 U.S. 257 (1971)).
32. Sentencing Reform Act of 1984, Pub. L. No. 98-473, 98 Stat. 1987 (codified as amended at 18 U.S.C. §§ 3551-3559, 3561-3566, 3571-3574, 3581-3586, and at 28 U.S.C. §§ 991-998 (1994 & Supp. 1996)).
33. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
34. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
35. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
36. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
37. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
38. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
39. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
40. The statutory directive to draft the Guidelines is codified at 28 U.S.C. § 994 (1994 & West Supp. 1998). Much has been written about the impact and wisdom of the Guidelines. See generally THOMAS W. HUTCHISON ET AL., FEDERAL SENTENCING LAW AND PRACTICE (1994); PHYLLIS SKLOOT BAMBERGER, PRACTICE UNDER THE FEDERAL GUIDELINES (1993); HARRY I. SUBIN ET AL., THE CRIMINAL PROCESS: PROSECUTION AND DEFENSE FUNCTIONS, § 10, at 131-61 (1992). Academic debate has raged about the fairness of the Guidelines system. Compare Albert W. Alschuler, The Failure of the Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization), Daniel J. Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize), and Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); with Frank O. Bowman, III, The Quality of Mercy Must Be Restrained, and Other Lessons in Learning to Love the Federal Sentencing Guidelines, 1996 WIS. L. REV. 679 (defending the Guidelines), and Julie R. O'Sullivan, In Defense of the U.S. Sentencing Guidelines' Modified Real-Offense System, 91 NW. U. L. REV. 1342 (1997) (defending the fundamental choice of modified, as opposed to pure charged offense sentencing, as necessary to impose sentences meaningfully related to the underlying wrongful conduct).
41. See Ian Weinstein, Judicial Discretion and the Federal Sentencing Guidelines, 79 B.U. L. Rev. (forthcoming Spring 1999) (manuscript on file with author and the Buffalo Law Review) (describing old law discretionary sentencing). Criticism of unfettered judicial discretion was central to the demise of that system. A federal judge was the most influential critic of unfettered judicial sentencing. See generally MARVIN E. FRANKEL, CRIMINAL SENTENCES (1973).
42. See Ian Weinstein, Judicial Discretion and the Federal Sentencing Guidelines, 79 B.U. L. Rev. (forthcoming Spring 1999) (manuscript on file with author and the Buffalo Law Review) (describing old law discretionary sentencing). Criticism of unfettered judicial discretion was central to the demise of that system. A federal judge was the most influential critic of unfettered judicial sentencing. See generally MARVIN E. FRANKEL, CRIMINAL SENTENCES (1973).
43. See Kate Stith & Steve Y. Koh, The Politics of Sentencing Reform: The Legislative History of the Federal Sentencing Guidelines, 28 WAKE FOREST L. REV. 223, 268 (1993) (describing the battle between those who successfully advocated mandatory Guidelines and those who would have let judges retain greater discretion).
44. Sentences are now longer but it took time for this major reform to take effect. See Bowman, supra note 24, at 732 n.186 (explaining that sentences are longer under the Guidelines); Stith & Koh, supra note 26 (describing how the Commission exercised the discretion conferred upon it to increase the severity of sentences); David Yellen, What Juvenile Court Abolitionists Can Learn from the Failures of Sentencing Reform, 1996 WIS. L. REV. 577, 585-90 (using the development of the federal Guidelines to illustrate how sentencing reform politics inevitably lead to harsher punishments). The Guidelines had to be drafted, take effect and withstand challenge before they governed federal sentencing. See Mistretta v. United States, 488 U.S. 361 (1989) (upholding the constitutionality of the Sentencing Guidelines).
45. Anti-Drug Abuse Act of 1986, Pub. L. No. 99-570, 100 Stat. 3207 (codified as amended 21 U.S.C. §§ 841-848 (1994 & West Supp. 1998)).
46. The most important mandatory minimum sentence provisions are codified at 21 U.S.C. §§ 841, 844, 860 (1994 & West Supp. 1998) and 18 U.S.C. § 924(c) (1994 & West Supp. 1998). There are mandatory minimum sentences for offenses other than narcotics sentences but as the Sentencing Commission noted in 1991, of the 60 mandatory minimum statutes then on the books, the four relating to particular instances of narcotics trafficking, importation and possession and weapons offenses committed in the course of another crime accounted for the vast bulk of the convictions and sentences. UNITED STATES SENTENCING COMM'N, SPECIAL REPORT TO THE CONGRESS, MANDATORY MINIMUM PENALTIES IN THE FEDERAL CRIMINAL JUSTICE SYSTEM 11 (1991) [hereinafter MANDATORY MINIMUM PENALTIES] (indicating that four statutes, setting mandatory minimum penalties for certain drug trafficking, drug importation, drug possession and firearms offenses accounted for 94% of all statutory mandatory minimum sentences; the overall argument of this report stresses the incompatibility of mandatory minimums with the Guidelines system). See generally Orrin G. Hatch, The Role of Congress in Sentencing: The United States Sentencing Commission, Mandatory Minimum Sentences, and the Search for a Certain and Effective Sentencing System, 28 WAKE FOREST L. REV. 185 (1993) (emphasizing the legislature's ultimate responsibility for setting sentences and suggesting mechanisms for greater oversight of the Commission in place of mandatory minimums); Gary T. Lowenthal, Mandatory Sentencing Laws: Undermining the Effectiveness of Determinate Sentencing Reform, 81 CAL. L. REV. 61 (1993) (discussing and evaluating the adoption of statutory mandatory minimum sentences in both the state and federal systems); Stephen J. Schulhofer, Rethinking Mandatory Minimums, 28 WAKE FOREST L. REV. 199 (1993) (arguing that although mandatory minimums meet some of their goals, those goals can be achieved by other means that avoid the irrefutable ill effects of the statutory minimums).
47. The most important mandatory minimum sentence provisions are codified at 21 U.S.C. §§ 841, 844, 860 (1994 & West Supp. 1998) and 18 U.S.C. § 924(c) (1994 & West Supp. 1998). There are mandatory minimum sentences for offenses other than narcotics sentences but as the Sentencing Commission noted in 1991, of the 60 mandatory minimum statutes then on the books, the four relating to particular instances of narcotics trafficking, importation and possession and weapons offenses committed in the course of another crime accounted for the vast bulk of the convictions and sentences. UNITED STATES SENTENCING COMM'N, SPECIAL REPORT TO THE CONGRESS, MANDATORY MINIMUM PENALTIES IN THE FEDERAL CRIMINAL JUSTICE SYSTEM 11 (1991) [hereinafter MANDATORY MINIMUM PENALTIES] (indicating that four statutes, setting mandatory minimum penalties for certain drug trafficking, drug importation, drug possession and firearms offenses accounted for 94% of all statutory mandatory minimum sentences; the overall argument of this report stresses the incompatibility of mandatory minimums with the Guidelines system). See generally Orrin G. Hatch, The Role of Congress in Sentencing: The United States Sentencing Commission, Mandatory Minimum Sentences, and the Search for a Certain and Effective Sentencing System, 28 WAKE FOREST L. REV. 185 (1993) (emphasizing the legislature's ultimate responsibility for setting sentences and suggesting mechanisms for greater oversight of the Commission in place of mandatory minimums); Gary T. Lowenthal, Mandatory Sentencing Laws: Undermining the Effectiveness of Determinate Sentencing Reform, 81 CAL. L. REV. 61 (1993) (discussing and evaluating the adoption of statutory mandatory minimum sentences in both the state and federal systems); Stephen J. Schulhofer, Rethinking Mandatory Minimums, 28 WAKE FOREST L. REV. 199 (1993) (arguing that although mandatory minimums meet some of their goals, those goals can be achieved by other means that avoid the irrefutable ill effects of the statutory minimums).
48. The most important mandatory minimum sentence provisions are codified at 21 U.S.C. §§ 841, 844, 860 (1994 & West Supp. 1998) and 18 U.S.C. § 924(c) (1994 & West Supp. 1998). There are mandatory minimum sentences for offenses other than narcotics sentences but as the Sentencing Commission noted in 1991, of the 60 mandatory minimum statutes then on the books, the four relating to particular instances of narcotics trafficking, importation and possession and weapons offenses committed in the course of another crime accounted for the vast bulk of the convictions and sentences. UNITED STATES SENTENCING COMM'N, SPECIAL REPORT TO THE CONGRESS, MANDATORY MINIMUM PENALTIES IN THE FEDERAL CRIMINAL JUSTICE SYSTEM 11 (1991) [hereinafter MANDATORY MINIMUM PENALTIES] (indicating that four statutes, setting mandatory minimum penalties for certain drug trafficking, drug importation, drug possession and firearms offenses accounted for 94% of all statutory mandatory minimum sentences; the overall argument of this report stresses the incompatibility of mandatory minimums with the Guidelines system). See generally Orrin G. Hatch, The Role of Congress in Sentencing: The United States Sentencing Commission, Mandatory Minimum Sentences, and the Search for a Certain and Effective Sentencing System, 28 WAKE FOREST L. REV. 185 (1993) (emphasizing the legislature's ultimate responsibility for setting sentences and suggesting mechanisms for greater oversight of the Commission in place of mandatory minimums); Gary T. Lowenthal, Mandatory Sentencing Laws: Undermining the Effectiveness of Determinate Sentencing Reform, 81 CAL. L. REV. 61 (1993) (discussing and evaluating the adoption of statutory mandatory minimum sentences in both the state and federal systems); Stephen J. Schulhofer, Rethinking Mandatory Minimums, 28 WAKE FOREST L. REV. 199 (1993) (arguing that although mandatory minimums meet some of their goals, those goals can be achieved by other means that avoid the irrefutable ill effects of the statutory minimums).
49. The most important mandatory minimum sentence provisions are codified at 21 U.S.C. §§ 841, 844, 860 (1994 & West Supp. 1998) and 18 U.S.C. § 924(c) (1994 & West Supp. 1998). There are mandatory minimum sentences for offenses other than narcotics sentences but as the Sentencing Commission noted in 1991, of the 60 mandatory minimum statutes then on the books, the four relating to particular instances of narcotics trafficking, importation and possession and weapons offenses committed in the course of another crime accounted for the vast bulk of the convictions and sentences. UNITED STATES SENTENCING COMM'N, SPECIAL REPORT TO THE CONGRESS, MANDATORY MINIMUM PENALTIES IN THE FEDERAL CRIMINAL JUSTICE SYSTEM 11 (1991) [hereinafter MANDATORY MINIMUM PENALTIES] (indicating that four statutes, setting mandatory minimum penalties for certain drug trafficking, drug importation, drug possession and firearms offenses accounted for 94% of all statutory mandatory minimum sentences; the overall argument of this report stresses the incompatibility of mandatory minimums with the Guidelines system). See generally Orrin G. Hatch, The Role of Congress in Sentencing: The United States Sentencing Commission, Mandatory Minimum Sentences, and the Search for a Certain and Effective Sentencing System, 28 WAKE FOREST L. REV. 185 (1993) (emphasizing the legislature's ultimate responsibility for setting sentences and suggesting mechanisms for greater oversight of the Commission in place of mandatory minimums); Gary T. Lowenthal, Mandatory Sentencing Laws: Undermining the Effectiveness of Determinate Sentencing Reform, 81 CAL. L. REV. 61 (1993) (discussing and evaluating the adoption of statutory mandatory minimum sentences in both the state and federal systems); Stephen J. Schulhofer, Rethinking Mandatory Minimums, 28 WAKE FOREST L. REV. 199 (1993) (arguing that although mandatory minimums meet some of their goals, those goals can be achieved by other means that avoid the irrefutable ill effects of the statutory minimums).
50. At the time the Anti-Drug Abuse Act of 1986 was passed, this statutory language gave judges the discretion to impose a sentence of no incarceration in most cases: Upon entering a judgment of conviction of any offense not punishable by death or life imprisonment, any court having jurisdiction to try offenses against the United States when satisfied that the ends of justice and the best interest of the public as well as the defendant will be served thereby, may suspend the imposition or execution of sentence and place the defendant on probation for such period and upon such terms and conditions as the court deems best. 18 U.S.C. § 3651 (1982), repealed by Sentencing Reform Act of 1984, Pub. L. No. 98-473, Title II, § 212(a) (2), 98 Stat. 1987 (applying to offenses committed prior to Nov. 1, 1987).
51. The Narcotics Control Act of 1956, Pub. L. No. 84-728, 70 Stat. 651 (1956), included mandatory minimum sentences of five to thirty years for a variety of narcotics offenses, which were virtually all repealed by the Comprehensive Drug Abuse Prevention and Control Act of 1970, Pub. L. No. 91-513, 84 Stat. 1236 (1970). For a discussion of the history of mandatory minimums from 1790 through 1990, see MANDATORY MINIMUM PENALTIES, supra note 29, at 6-12.
52. The five and ten year mandatory sentences of 21 U.S.C. § 841 are keyed to the weight of a "mixture or substance containing a detectable amount" of a controlled substance. Chapman v. United States, 500 U.S. 453, 457 (1991) (holding that the statute requires that the weight of the carrier medium be included in determining the applicability of the statutory mandatory minimum for trafficking in LSD). The twenty year mandatory sentence applies to individuals convicted of offenses involving the required quantity of drugs who have also been previously convicted of a narcotics offense.
53. It was also an unwise change according to many. The Sentencing Commission and many commentators have argued that statutory mandatory minimums have no role in a comprehensive sentencing scheme and introduce anomalous results. See generally Lowenthal, supra note 29 (discussing conflicts between determinate sentencing schemes and mandatory minimums); MANDATORY MINIMUM PENALTIES, supra note 29 (noting conflicts between the Guidelines and mandatory minimums). Others have argued that mandatory minimums are not cost effective.
54. See generally JONATHON P. CAULKINS ET AL., RAND DRUG POLICY RESEARCH CENTER, MANDATORY DRUG SENTENCES: THROWING AWAY THE KEY OR THE TAXPAYERS' MONEY (1997) (providing an empirical analysis of costs and benefits of mandatory minimums). Some of the worst effects of the mandatory minimums have been addressed, albeit rather weakly, by the safety valve provision, 18 U.S.C. § 3553(f) (1994 & West Supp. 1997), permitting sentences below the mandatory minimums for some first offenders. See generally Oliss, supra note 12, at 1892 (calling the provision a small first step). Mandatory sentencing provisions have proliferated despite those criticisms.
55. See Ronald Weich, The Battle Against Mandatory Minimums: A Report from the Front Lines, 9 FED. SENTENCING REP. 94 (1996) (chief counsel to Senator Kennedy writing on the difficulties of opposing mandatory minimum sentence provisions during the 104th Congress, which passed immigration and child pornography bills including mandatory minimums but did not approve proposed end of term legislation including higher mandatory minimum sentences for methamphetamine and new mandatory minimums for Rohyphynol (the "date rape" drug)).
56. The statutes which fostered widespread cooperation, Pub. L. No. 99-570, §§ 1007, 1008, 100 Stat. 3207 (1986), were passed as a pair of amendments buried deep in the Anti-Drug Abuse Act of 1986. There was no debate in either chamber and no mention of the change appears in any of the reports or other legislative history of that act, save a brief comment on the floor by Senator Dole, 132 CONG. REC. 31,409 (1986); see also United States v. Severich, 676 F. Supp 1209, 1212 (S.D. Fla. 1988) (stating that the "legislative history appears silent as to any appropriate and articulated rationale for the substantial assistance feature.").
57. See Anti-Drug Abuse Act of 1986, Pub. L. No. 99-570, § 1007, codified at 18 U.S.C. § 3553(e) (1994).
58. 28 U.S.C. § 994(n) (1994 & West Supp 1998).
59. 18 U.S.C. § 924(c) (1994 & West Supp. 1998) offers a counter example in its requirement that a defendant convicted of using or carrying a firearm during or in relation to any crime of violence or drug trafficking be sentenced to a five year term of imprisonment, not more or less, to run consecutive to any other sentence imposed.
60. See id. § 3553(f); supra note 12 (discussing the "safety valve" provision).
61. The overall complexity of the Guidelines has been a subject of particularly interesting and insightful criticism. Kevin R. Reitz, Sentencing Guideline Systems and Sentence Appeals: A Comparison of Federal and State Experiences, 91 NW. U. L. REV. 1441 (1997) (arguing that the complexity and rigidity of the Guidelines have led the courts to take a high enforcement/low judicial creativity approach to sentencing); Kate Stith & Jose Cabranes, Judging Under the Federal Sentencing Guidelines, 91 NW. U. L. REV. 1247, 1281 (1997) (criticizing the Commission's efforts to "micro-manage" sentencing).
62. The overall complexity of the Guidelines has been a subject of particularly interesting and insightful criticism. Kevin R. Reitz, Sentencing Guideline Systems and Sentence Appeals: A Comparison of Federal and State Experiences, 91 NW. U. L. REV. 1441 (1997) (arguing that the complexity and rigidity of the Guidelines have led the courts to take a high enforcement/low judicial creativity approach to sentencing); Kate Stith & Jose Cabranes, Judging Under the Federal Sentencing Guidelines, 91 NW. U. L. REV. 1247, 1281 (1997) (criticizing the Commission's efforts to "micro-manage" sentencing).
63. The guideline range is determined by finding the defendant's place on the Sentencing Table, a matrix defined by a vertical axis determined by rules keyed to the severity of the defendant's conduct and a horizontal axis determined by prior criminal record. Each box in the matrix is assigned a sentencing range, expressed as a minimum and maximum number of months within which the defendant must be sentenced. See U.S.S.G., supra note 3, at § 5A Sentencing tbl.
64. Judges may depart downward from the range, mitigating the sentence, or they may depart upward, making the sentence harsher. See U.S.S.G., supra note 3, at § 1A4(b). Departures must be distinguished from adjustments; adjustments are factors related to the crime or the criminal which raise or lower the offense level within the Guidelines scheme, while departures are factors created by the Commission to take the case outside the Guidelines scheme all together. Id. at § 3. The Guidelines separate departure factors into three categories: those that judges are encouraged to use, those that they are discouraged from using and those that they are forbidden from using. See id. at § 1A4(b); see also Koon v. United States, 518 U.S. 81, 116 (1996) (discussing the roles of forbidden, encouraged, discouraged and unmentioned departure factors and holding that abuse of discretion is the proper standard of review of district decisions to depart). Forbidden factors such as race, sex, national origin, creed, religion and socio-economic status may never be the basis for a departure. See U.S.S.G., supra note 3, at § 5H.10. Others are encouraged, such as upward departures in cases where the defendant's criminal history category does not adequately reflect the seriousness of his or her past conduct. See id. at § 4A1.3. Discouraged factors, such as age, education, mental and emotional conditions, physical condition, employment record, family ties and responsibilities, military, civic, charitable or public service and lack of guidance as a youth are not "ordinarily relevant" to determining the sentence but may be considered if the situation is not ordinary. See id. at §§ 5H1.1-1.6, 1.11-12. This set of factors has spawned much litigation, with disparate results. Compare United States v. Galante, 111 F.3d 1029 (2d Cir. 1997) (affirming departure based on discouraged factor of family circumstances), with United States v. Wilson, 114 F.3d 429 (4th Cir. 1997) (reversing departure based on family circumstances). See generally Weinstein, supra note 25 (manuscript at 49-79) (comparing differing circuit court approaches to departures after Koon, 518 U.S. at 81). Departures based upon factors not mentioned in the Guidelines must be considered on a case-by-case basis and cannot be categorically excluded. See Koon, 518 U.S. at 108.
65. Under the Guidelines, defense lawyers are able to make much more accurate sentence predictions much earlier in the case than they could in the era of discretionary sentencing. This has a significant impact on defendants' decision making. Under the old system, defendants could make their decisions with the hope of a lenient sentence and a lawyer could not tell them such a sentence was legally impossible. No matter how bad the odds, in a discretionary regime there is always a chance. That is no longer the case. Defendants facing the Guidelines or a statutory mandatory minimum cannot even gamble on leniency. The role of greater predictability in shaping the decision making of defendants, who must make very important decisions under very difficult circumstances, is probably quite significant, although its effects may be hard to separate out from the impact of longer sentences.
66. Despite predictions that the Guidelines would result in more trials, that does not seem to be the case. Although trial rates increased in the first three years, they have since retreated. Two data sets show the same trend. The Commission collects data for the mode of conviction, a number that does not account for acquittals. By that measure, the rate of trial was 11.9% in 1989, 12.3% in 1990, 14.6% in 1991, 13.0% in 1992, 11.5% in 1993, 9.5% in 1994, 8.1% in 1995 and 8.3% in 1996. 1996 FED. SENT. SRCBK., supra note 2, at 15 fig. C. Using the numbers from Defendants Disposed of in U.S. District Courts, in BUREAU OF JUSTICE SRCBK., supra note 11, at 448 tbl.5.27, the rate of trial among all defendants whose cases were not dismissed was: 16% in 1989, 16% in 1990, 15% in 1991, 14% in 1992, 12% in 1993, 11% in 1994, 10% in 1995 and 9% in 1996.
67. Despite predictions that the Guidelines would result in more trials, that does not seem to be the case. Although trial rates increased in the first three years, they have since retreated. Two data sets show the same trend. The Commission collects data for the mode of conviction, a number that does not account for acquittals. By that measure, the rate of trial was 11.9% in 1989, 12.3% in 1990, 14.6% in 1991, 13.0% in 1992, 11.5% in 1993, 9.5% in 1994, 8.1% in 1995 and 8.3% in 1996. 1996 FED. SENT. SRCBK., supra note 2, at 15 fig. C. Using the numbers from Defendants Disposed of in U.S. District Courts, in BUREAU OF JUSTICE SRCBK., supra note 11, at 448 tbl.5.27, the rate of trial among all defendants whose cases were not dismissed was: 16% in 1989, 16% in 1990, 15% in 1991, 14% in 1992, 12% in 1993, 11% in 1994, 10% in 1995 and 9% in 1996.
68. See U.S.S.G., supra note 3, at § 5A Sentencing tbl. These ranges are for a hypothetical defendant facing narcotics trafficking charges involving one kilogram of heroin or five kilograms of cocaine or fifty grams of cocaine base who had two prior non-drug convictions. Assuming no adjustments, that defendant would be at offense level 32, criminal history category III, upon conviction after trial for a sentencing range of 151-88 months. Assuming that he or she received the three point downward adjustment for acceptance of responsibility, § 3E1.1(b), the sentence would be at level 29, criminal history category III, and ten year mandatory minimum would determine the bottom of the sentencing range, which would be 120-35 months.
69. The reduction from the ten year mandatory minimum reflects sentence credit for good behavior. Federal prisoners serving sentences longer than one year, but less than life (including those serving statutory mandatory minimum sentences) can receive 54 days of "good time" credit on any sentence longer than one year. See 18 U.S.C. § 3624(b) (1994 & West Supp. 1998).
70. This figure is calculated by subtracting 15% good time credit from 151 months yielding roughly 129 months or ten years and nine months.
71. Defendants also have a strong incentive to litigate the many detailed calculations underlying the determination of the sentencing range including sentence mitigation and proper sentence calculation, depending on the underlying facts related to each case. See discussion infra nn.180 & 181 (outlining Guidelines calculations issues that are often used to manipulate sentences). For a broad historical discussion of mitigation, see generally Roscoe Pound, Discretion, Dispensation and Mitigation: The Problem of the Individual Special Case, 35 N.Y.U. L. REV. 925 (1960) (tracing the history of the development of the equitable powers of dispensation and mitigation and arguing for greater individualization of legal results). The other significant source of mitigation is prosecutorial charging decisions.
72. See, e.g., Ilene H. Nagel & Stephen J. Schulhofer, A Tale of Three Cities: An Empirical Study of Charging and Bargaining Practices Under the Federal Sentencing Guidelines, 66 S. CAL. L. REV. 501 (1989) (describing prosecutorial charging practices in three districts);
73. Stephen J. Schulhofer & Ilene H. Nagel, Plea Negotiations Under the Federal Sentencing Guidelines: Guideline Circumvention and its Dynamics in the Post-Mistretta Period, 91 NW. U. L. REV. 1284 (1997) (discussing the mechanisms prosecutors use to manipulate the Guidelines);
74. Stephen J. Schulhofer & Ilene H. Nagel, Negotiated Pleas Under the Federal Sentencing Guidelines: The First Fifteen Months, 27 AM. CRIM. L. REV. 231 (1989) (comparing prosecutorial practices before and after the Mistretta decision).
75. Weinstein, supra note 25 (manuscript at 30-32) (explaining judici discretion in non-substantial assistance departures).
76. The statutory and Guidelines requirements for a government motions this instance as required by 18 U.S.C. § 3553(e) and U.S.S.G. § 5K1.1 combin to create a key element in the shift in sentencing discretion from judges prosecutors. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 107-08; Lee, From Gatekeep to Concierge: Reigning in the Federal Prosecutor's Expanding Power Ov Substantial Assistance Departures, supra note 14, at 234-38; Misner, supra no 9, at 757-58 n.295. There are very limited avenues for judges to recogni cooperation in the absence of a motion. See Lee, Prosecutorial Discretio Substantial Assistance, and the Federal Sentencing Guidelines, supra note 1 at 144 n.175 (discussing cases in which judges have departed without gover ment motion to recognize extraordinary acceptance of responsibility or to r ward a defendant for encouraging another to plead and so assisting the court).
77. The statutory and Guidelines requirements for a government motions this instance as required by 18 U.S.C. § 3553(e) and U.S.S.G. § 5K1.1 combin to create a key element in the shift in sentencing discretion from judges prosecutors. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 107-08; Lee, From Gatekeep to Concierge: Reigning in the Federal Prosecutor's Expanding Power Ov Substantial Assistance Departures, supra note 14, at 234-38; Misner, supra no 9, at 757-58 n.295. There are very limited avenues for judges to recogni cooperation in the absence of a motion. See Lee, Prosecutorial Discretio Substantial Assistance, and the Federal Sentencing Guidelines, supra note 1 at 144 n.175 (discussing cases in which judges have departed without gover ment motion to recognize extraordinary acceptance of responsibility or to r ward a defendant for encouraging another to plead and so assisting the court).
78. The statutory and Guidelines requirements for a government motions this instance as required by 18 U.S.C. § 3553(e) and U.S.S.G. § 5K1.1 combin to create a key element in the shift in sentencing discretion from judges prosecutors. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 107-08; Lee, From Gatekeep to Concierge: Reigning in the Federal Prosecutor's Expanding Power Ov Substantial Assistance Departures, supra note 14, at 234-38; Misner, supra no 9, at 757-58 n.295. There are very limited avenues for judges to recogni cooperation in the absence of a motion. See Lee, Prosecutorial Discretio Substantial Assistance, and the Federal Sentencing Guidelines, supra note 1 at 144 n.175 (discussing cases in which judges have departed without gover ment motion to recognize extraordinary acceptance of responsibility or to r ward a defendant for encouraging another to plead and so assisting the court).
79. The statutory and Guidelines requirements for a government motions this instance as required by 18 U.S.C. § 3553(e) and U.S.S.G. § 5K1.1 combin to create a key element in the shift in sentencing discretion from judges prosecutors. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 107-08; Lee, From Gatekeep to Concierge: Reigning in the Federal Prosecutor's Expanding Power Ov Substantial Assistance Departures, supra note 14, at 234-38; Misner, supra no 9, at 757-58 n.295. There are very limited avenues for judges to recogni cooperation in the absence of a motion. See Lee, Prosecutorial Discretio Substantial Assistance, and the Federal Sentencing Guidelines, supra note 1 at 144 n.175 (discussing cases in which judges have departed without gover ment motion to recognize extraordinary acceptance of responsibility or to r ward a defendant for encouraging another to plead and so assisting the court).
80. There were 40,879 sentences imposed under the Guidelines in 1996. S 1996 FED. SENT. SRCBK., supra note 2, at 41 tbl.26. A total of 28,445 (69.6%) those were within the Guidelines range. Id. Of those sentences, 43.8% we within the first quarter of the range, 9.6% were in the second quarter, 3.3 were in the third quarter and 9.1% were in the fourth quarter.
81. See 1996 FE SENT. SRCBK., supra note 2, at 44 tbl.27. A total of 7845 sentences (19.2% of t total) involved departures based on substantial assistance. Other downwa departures were involved in 4201 (10.3% of the total number) and 388 cas (0.9%) involved upward departures.
82. 1996 FED. SENT. SRCBK., supra note 2, at tbl.26.
83. The 10% of sentenced defendants who receive the benefit of oth downward departures typically present compelling facts which are evident fro the start of the decision process, or are sentenced by a lenient judge.
84. Once a judge finds a ground for departing downward from the guidelines, the magnitude of the departure is in his or her discretion. See, e.g., United States v. Robinson, 1997 U.S. App. LEXIS 3641, at *3-4 (6th Cir. Feb. 24, 1997) (explaining that the magnitude of departure is in judge's discretion). Judges may impose non-incarceratory sentences in most cases, despite the provisions of 18 U.S.C. § 3561(a) (1), which forbids probation to anyone convicted of a crime carrying a statutory maximum of 25 years or more. See United States v. Elliott, 971 F.2d 620 (10th Cir. 1992) (holding that a sentence of no jail is not a probationary sentence and is permitted despite 18 U.S.C. § 3561(a) (1)). But see United States v. Greene, 1996 U.S. App. LEXIS 19056, at *1 (4th Cir. June 18, 1996) (finding no difference between a sentence of probation and no imprisonment but noting that cooperators, but not other relevant departure recipients, may receive non-incarceratory sentences). For discussion of the problems engendered by the lack of consistency in practices governing the magnitude of departures see Patti B. Saris, Below the Radar Screens: Have the Sentencing Guidelines Eliminated Disparity? One Judge's Perspective, 30 SUFFOLK U. L. REV. 1027, 1045-46 (1997) (commenting on warranted versus unwarranted disparity and the failure to control the magnitude of departures from the perspective a United States District Court Judge for the District of Massachusetts).
85. The other path to serving no prison time, immunity from prosecution, is very rarely taken. One imprecise measure is the 1493 grants of immunity by federal prosecutors in 1996, BUREAU OF JUSTICE SRCBK., supra note 11, at 448 tbl.5.27, a number which includes grants to witnesses whose prosecution was not contemplated but does not include agreements to forego prosecution that do not include immunity. Potential defendants are only occasionally immunized from all prosecutions to induce them to assist the government. The prosecution of several Texaco executives on charges stemming from their alleged efforts to destroy evidence subpoenaed in a civil case illustrates the difficult judgments prosecutors must make. A Texaco official tape recorded incriminating conversations and then offered the tapes to the government in return for immunity from prosecution. The government refused immunity and indicted him instead. He and his co-defendant were acquitted and the government failed to convict anyone. See United States v. Lundwall, 97 Cr. 211 (BDP) (S.D.N.Y. 1997); Ex- Texaco Treasurer Indicted in Race Discrimination Case, N.Y. TIMES, June 28, 1997, at A36; Kurt Eichenwald, Texaco Witness is Said to be Talking to U.S., N.Y. TIMES, Jan. 8, 1997, at D1.
86. The other path to serving no prison time, immunity from prosecution, is very rarely taken. One imprecise measure is the 1493 grants of immunity by federal prosecutors in 1996, BUREAU OF JUSTICE SRCBK., supra note 11, at 448 tbl.5.27, a number which includes grants to witnesses whose prosecution was not contemplated but does not include agreements to forego prosecution that do not include immunity. Potential defendants are only occasionally immunized from all prosecutions to induce them to assist the government. The prosecution of several Texaco executives on charges stemming from their alleged efforts to destroy evidence subpoenaed in a civil case illustrates the difficult judgments prosecutors must make. A Texaco official tape recorded incriminating conversations and then offered the tapes to the government in return for immunity from prosecution. The government refused immunity and indicted him instead. He and his co-defendant were acquitted and the government failed to convict anyone. See United States v. Lundwall, 97 Cr. 211 (BDP) (S.D.N.Y. 1997); Ex-Texaco Treasurer Indicted in Race Discrimination Case, N.Y. TIMES, June 28, 1997, at A36; Kurt Eichenwald, Texaco Witness is Said to be Talking to U.S., N.Y. TIMES, Jan. 8, 1997, at D1.
87. The other path to serving no prison time, immunity from prosecution, is very rarely taken. One imprecise measure is the 1493 grants of immunity by federal prosecutors in 1996, BUREAU OF JUSTICE SRCBK., supra note 11, at 448 tbl.5.27, a number which includes grants to witnesses whose prosecution was not contemplated but does not include agreements to forego prosecution that do not include immunity. Potential defendants are only occasionally immunized from all prosecutions to induce them to assist the government. The prosecution of several Texaco executives on charges stemming from their alleged efforts to destroy evidence subpoenaed in a civil case illustrates the difficult judgments prosecutors must make. A Texaco official tape recorded incriminating conversations and then offered the tapes to the government in return for immunity from prosecution. The government refused immunity and indicted him instead. He and his co-defendant were acquitted and the government failed to convict anyone. See United States v. Lundwall, 97 Cr. 211 (BDP) (S.D.N.Y. 1997); Ex- Texaco Treasurer Indicted in Race Discrimination Case, N.Y. TIMES, June 28, 1997, at A36; Kurt Eichenwald, Texaco Witness is Said to be Talking to U.S., N.Y. TIMES, Jan. 8, 1997, at D1.
88. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
89. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
90. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
91. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
92. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
93. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
94. The prisoner's dilemma game theory model has become the most familiar in the law. See WILLIAM POUNDSTONE, PRISONER'S DILEMMA (1992). For further explanation of this model, see generally DOUGLAS G. BAIRD ET AL., GAME THEORY AND THE LAW (1994), reviewed in Stephen W. Salant & Theodore S. Sims, Game Theory and the Law: Ready for Primetime?, 94 MICH. L. REV. 1839 (1996), which has become a standard tool for teachers and students of negotiation theory. See generally ROBERT M. BASTRESS & JOSEPH D. HARBAUGH, INTERVIEWING, COUNSELING AND NEGOTIATION, 351-57 (1990); DONALD G. GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The model posits two co-defendants facing identical charges. If neither cooperates with the government, both will get a moderate sentence. If one cooperates and the other does not, the cooperator gets a very lenient sentence while the other defendant receives a harsh sentence. If both cooperate against the other, both receive harsh sentences. See DONALD GIFFORD, LEGAL NEGOTIATION THEORY AND APPLICATIONS 26-28 (1989). The prisoner's dilemma models a negotiation between parties with equal power and equal access to information about everything but the other negotiator's intentions. It teaches students that in many negotiations, defection (usually deceiving or breaking commitments) can gain great benefits, but carries the risk that the other side will behave similarly and both will suffer the worst possible result. Despite its name, the classic game theory/prisoner's model sheds less light upon the actual choices of typical prisoners. In most cases the decision to cooperate turns on the negotiation between the prisoner and the government and is independent of the decision of other defendants. After all, the government has very broad discretion to set the terms of the cooperation. There are some cases involving two equally culpable defendants, both of whom have no other useful information for the government besides the offer of testimony against the co-defendant. If the government chooses to accept cooperation from only one of them and the government could not otherwise convict the other defendant of the most serious crime the two committed, the structure of the prisoner's dilemma is presented, but even here reality diverges from the model. The government does not provide perfect information about the strength of its case or the sentences that will result from each course of action, and can change the pay out as the game proceeds. See generally Richman, supra note 1, at 89-91 (explaining how the reality of cooperation diverges from the model). The prisoner's dilemma much better models the risks and benefits of cooperation and competition in two party negotiation than it models the actual problem faced by most prisoners.
95. See 1996 FED. SENT. SRCBK., supra note 2, at 14 tbl.9. The link between substantial assistance departures and narcotics offenses reflects both the harshness of the Narcotics Guidelines. See, e.g., U.S.S.G., supra note 3, at § 2D1.1, and the mandatory minimums. Government motions pursuant to U.S.S.G. § 5K.1.1 for a sentence below the Guidelines range and government motions pursuant to 18 U.S.C. § 3553(e) for a sentence below the statutory mandatory minimum are distinct. The government is free to make one or the other or both, as it deems appropriate. See Melendez v. United States, 518 U.S. 120 (1996) (holding that the Guidelines did not create a unitary system - a government motion to depart from the Guidelines does not confer authority upon the court to ignore the mandatory minimum and a separate government motion pursuant to 18 U.S.C. § 3553(e) is required).
96. The Commission's Distribution of Drug Guidelines Offenses chart reports a total of 42,436 cases. Of these, 17,267 were sentenced under the various drug Guidelines while 16,196 (38.2% of all cases) were sentenced under the narcotics trafficking portion of the Guidelines pursuant to U.S.S.G. § 2D1.1. An additional 1071 (2.5% of all cases) were sentenced under other drug offense Guidelines, with about half that group (550 cases) involving misdemeanor prosecutions for simple possession (including some crack possession cases that can carry a five year mandatory sentence, 21 U.S.C. § 844(a)) and the remainder involving the following: § 2D1.2 Protected Locations 329 cases; § 2D1.5 Continuing Criminal Enterprise 54 cases; § 2D1.8 Renting or Managing Drug Establishments 41 cases). Non-narcotics prosecutions accounted for 25,169 cases (59.3% of all cases). See 1996 FED. SENT. SRCBK., supra note 2, at 45 fig.H.
97. Of the 40,818 cases for which the Commission had sufficient data to analyze the departure rate by primary offense category, 15,882 were drug trafficking cases and 5041 of them (31.7%) involved substantial assistance departures in 1996. See 1996 FED. SENT. SRCBK., supra note 2, at 14 tbl.9.
98. The Sentencing Commission analyzed 264 defendants in 64 randomly selected narcotics conspiracies prosecuted in 1992. More than 65% of those defendants offered assistance. LINDA DRAZGA MAXFIELD & JOHN H. KRAMER, UNITED STATES SENTENCING COMM'N, SUBSTANTIAL ASSISTANCE: AN EMPIRICAL YARDSTICK GAUGING EQUITY IN CURRENT FEDERAL POLICY AND PRACTICE 26 exh.5 (1998) [hereinafter SENT. COMM. SUBST. ASS. STUDY].
99. Couriers may be arrested carrying significant quantities of narcotics. Low and middle level dealers may have made several sales to undercover agents or confidential informants wearing recording devices. Higher level dealers and more sophisticated operations often face hundreds of hours of tapes and months or years of surveillance. While capable defense attorneys can gain acquittals in each type of case, the fact remains that 82% of the defendants who went to trial in 1996 were convicted. See Defendants Disposed of in U.S. District Courts, in BUREAU OF JUSTICE SRCBK., supra note 11, at 448 tbl.5.27.
100. Of the 17,165 cases sentenced under the narcotics Guidelines in 1996, 6365 (37.1%) involved a mandatory minimum sentence of ten years or more. In addition, 4985 (28.9%) involved a five year mandatory minimum and 5842 (34%) involved a mandatory minimum of 12 months or less or no mandatory minimum. Of the 4602 crack cases, 2643 (57.4%) involved a ten year mandatory minimum or greater, while 501 (28.4%) of the 1766 heroin cases, 1919 (42.9%) of the 4471 powder cocaine cases and 411 (9.4%) of the 4249 marijuana cases involved ten year or greater mandatory minimums. 1996 FED. SENT. SRCBK., supra note 2, at 53 tbl.38.
101. Some defendants receive mitigation through charging decisions or other downward departures. See infra pp. 57-65.
102. The mean decrease in the length of the sentences for the 4791 narcotics trafficking defendants who received 5K1.1 departures in fiscal 1994 was 63.9 months. The mean sentence imposed on the cooperators was 51.6 months. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 32 exh.11. Comparison of cooperation and non-cooperation sentences is impossible with this data because the overall mean sentence reported for narcotics trafficking cases includes the substantial assistance departures. The length of the departure correlates with the length of the predeparture sentence. Id. at 17-18, 21. For defendants whose guideline sentence would have been less than five years, the mean decrease was 13.7 months. For those with sentences between 60 and 120 months, the mean decrease was 31 months and sentences over 120 months received a mean decrease of 97.9 months. Id. at 32 exh.11.
103. The mean decrease in the length of the sentences for the 4791 narcotics trafficking defendants who received 5K1.1 departures in fiscal 1994 was 63.9 months. The mean sentence imposed on the cooperators was 51.6 months. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 32 exh.11. Comparison of cooperation and non-cooperation sentences is impossible with this data because the overall mean sentence reported for narcotics trafficking cases includes the substantial assistance departures. The length of the departure correlates with the length of the predeparture sentence. Id. at 17-18, 21. For defendants whose guideline sentence would have been less than five years, the mean decrease was 13.7 months. For those with sentences between 60 and 120 months, the mean decrease was 31 months and sentences over 120 months received a mean decrease of 97.9 months. Id. at 32 exh.11.
104. Of the 5789 fraud defendants, 913 (15.8%) received substantial assistance departures. See 1996 FED. SENT. SRCBK., supra note 2, at 14 tbl.9. Some other federal crimes such as immigration and larceny offenses, show very low rates of cooperation. Of the 4715 immigration cases, only 134 (2.8%) involved substantial assistance. Of the 2311 larceny cases, only 141 (6.1%) received substantial assistance departures. Id. Some crimes simply do not lend themselves to snitching. The typical federal immigration offense involves a lone defendant or a group of equally culpable people whose conviction is straightforward.
105. The absence of mandatory minimum sentences for fraud defendants opens the possibility of the full range of Guidelines arguments and ordinary Guidelines departures. See supra note 41.
106. 1996 FED. SENT. SRCBK., supra note 2, at 22 tbl.13.
107. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 18.
108. The mean decrease for fraud cases was 5.2 months. Id. at 32 exh.11. Of course, some fraud cases do involve long prison sentences. See, e.g., United States v. Hoffenberg, 1997 U.S. Dist. LEXIS 2394, at *1 (S.D.N.Y. Mar. 5, 1997) (sentencing opinion imposing a twenty year term in a multi-million dollar securities fraud case involving the failure of Towers Financial Corporation, subject to a sentencing hearing). Other fraud defendants facing relatively little jail time are very sensitive to the possibility of avoiding incarceration completely and may cooperate even if facing relatively modest sentences. Twenty percent of fraud defendants received probation only sentences, compared to 2.8% of the narcotics defendants. An additional 16.4% of the fraud defendants received a combined probation and confinement sentence. 1996 FED. SENT. SRCBK., supra note 2, at 21 tbl.12.
109. See generally KENNETH MANN, DEFENDING WHITE COLLAR CRIME: A PORTRAIT OF ATTORNEYS AT WORK (1985) (describing the results of an in-depth study of some members of the white collar defense bar in New York).
110. The general moral discomfort with cooperation is discussed infra nn. 75-80. It is commonly assumed that other prisoners see snitches in a very bad light. An illustration of this attitude can be found in the words of an inmate discussing the worst aspects of prison in a fascinating inside look at day-to-day life in prison from both the inmates' and penological perspectives. Robert Blecker, Haven or Hell? Inside Lorton Central Prison: Experiences of Punishment Justified, 42 STAN. L. REV. 1149, 1156 (1990) (presenting an inmate's discussion about the worst parts of prison including among them the problem posed by officers who try to make snitches out of prisoners they see as weak); see also Richman, supra note 1, at nn.35-38.
111. See Richman, supra note 1, at n.35 (discussing the oath of secrecy in classic organized crime cases and the prosecution of John Gotti, which turned largely on the testimony of his former confederate, Salvatore Gravano).
112. I have not considered the innocent defendant, who may see cooperation as offering too good a deal to pass up. This is a special case of the innocence problem in plea bargaining. See Robert E. Scott & William J. Stuntz, Plea Bargaining as Contract, 101 YALE L.J. 1909, 1946-66 (1992) (discussing the problem of innocent defendants taking advantageous pleas and suggesting reforms to improve plea bargaining); Schulhofer, supra note 10, at 1980-86 (1992) (arguing in response to Scott and Stuntz that the innocence problem is more theoretical than real but other structural flaws mandate the elimination of plea bargaining).
113. See Steven S. Nemerson, Coercive Sentencing, 64 MINN. L. REV. 669, 734 (1980) (explaining that fear of retaliation is the biggest disincentive to cooperation). The Substantial Assistance Guideline recognizes threats of harm as a specific factor judges should consider in evaluating cooperation. See U.S.S.G., supra note 3, § 5K1.1.(4).
114. For some examples of retaliation, see MICHAEL H. GRAHAM, WITNESS INTIMIDATION: THE LAW'S RESPONSE 3-8 (1985).
115. There are also defendants who fear economic reprisal.
116. A single buyer of a good which has multiple sellers is a monopsonist. See Jeffrey Standen, Plea Bargaining in the Shadow of the Guidelines, 81 CALIF. L. REV. 1471, 1472-74 (1993) (analyzing the government as the sole purchaser of defendants' convictions and arguing that the only effective counterweight to the power of that position is to redisperse sentencing power by returning sentencing discretion to judges).
117. 504 U.S. 181 (1992). Judicial constraints are only one kind of limit on prosecutorial discretion. See Standen, supra note 75, at n.7 (explaining that there are many factors that affect prosecutorial discretion including the impact of supervisors, the desire to win and other role related pressures).
118. Wade, 504 U.S. at 181, is consistent with earlier cases such as Wayte v. United States, 470 U.S. 598 (1985), in requiring that defendants make a substantial threshold showing before a court will order discovery or an evidentiary hearing on a claim that the prosecutor's refusal to move for a substantial assistance departure, in the absence of any agreement between the parties, violated constitutional prohibitions on prosecutors acting out on invidious motives or in ways not rationally related to any government end. See also Chapman v. United States, 500 U.S. 453 (1991) (holding that including the weight of the carrier medium in assessing penalties is not irrational).
119. 504 U.S. at 187 (citations omitted).
120. The cases are discussed below, see infra nn. 93-106.
121. The attempt to negotiate a cooperation deal usually begins with a proffer, or debriefing session attended by the defendant, his or her lawyer, the prosecutor and the investigating agents, in which the defendant describes his or her own conduct and outlines the kind of aid he or she can offer the government. They are generally conducted pursuant to agreements which give the defendant limited use immunity. Often the government agrees not to use the defendant's statements in its case in chief, should there be a trial, but reserves the right to use the information to get other evidence and for cross examination, as well as in prosecutions for perjury or false statements. The Supreme Court captured the essence of the contemporary proffer session more than 100 years ago in describing the dynamics of the early meetings at which the value of a proposed cooperator's information is evaluated: Of all others, the prosecutor is best qualified to determine that question, as he alone is supposed to know what other evidence can be adduced to prove the criminal charge. Applications of the kind are not always to be granted, and in order to acquire the information necessary to determine the question, the public prosecutor will grant the accomplice an interview, with the understanding that any communications he may make to the prosecutor will be strictly confidential. Interviews for the purpose mentioned are for mutual explanation, and do not absolutely commit either party; but if the accomplice is subsequently called and examined, he is equally entitled to a recommendation for executive clemency. Promise of pardon is never given in such an interview, nor any inducement held out beyond what the before-mentioned usage and practice of the courts allow. Prosecutors in such a case should explain to the accomplice that he is not obliged to incriminate himself, and inform him just what he may reasonably expect in case he acts in good faith, and testifies fully and fairly as to his own acts in the case, and those of his associates. The Whiskey Cases, 99 U.S. 594, 603-04 (1878).
122. A formal cooperation agreement is a final, integrated agreement that supercedes any proffer or other preliminary agreements and sets out the charges to which the defendant will plead, his or her other obligations and specifies the government's reciprocal obligations. Formal agreements are discussed infra pp. 587-93.
123. At least as measured by their ultimate sentence, if they suffer no collateral consequences from the attempt to cooperate. They have lost something in giving information and getting nothing in return, but there is no other buyer for the information so they cannot, and could not, have sold it elsewhere.
124. Other discretionary mechanisms prosecutors use to mitigate sentences are discussed below, see infra pp. 608-11.
125. U.S.S.G., supra note 3, § 3C1.1.
126. The degree of protection afforded statements made during cooperation is determined by whatever agreements the parties make. Wade offers an example of a proffer agreement providing limited use immunity but permitting the government to use the defendant's statements to gather other evidence which would then be admissible against him or her and to use the defendant's own statements on cross examination, should a trial occur. Wade, 504 U.S. at 186-87. The Guidelines provide limited protection for information revealed during cooperation. U.S.S.G., supra note 3, § 1B1.8 That Guideline only protects information if the defendant enters into an agreement which prohibits use of the information and then only offers limited protection.
127. Most defendants simply do not have valuable enough information to bargain effectively with the prosecutor. Even those who do may calculate that they are better off winning the prosecutor's favor and gaining the greatest benefit rather than worrying about legal rights that will come into play if the deal goes awry and they are forced to litigate against the party they had hoped would become their supporter. One countervailing factor is that many districts use a standard agreement which incorporates favorable terms that the few relatively strong defendants managed to negotiate.
128. Although some have argued that the power imbalance inherent in plea bargaining renders all plea agreements contractually deficient, see Scott & Stuntz, supra note 71, at 1909 n.4 & 1912 n.11, a critique that would certainly apply to substantial assistance motions, Scott & Stuntz convincingly argue that the fundamental norms underlying contract law support the enforcement of contracts between parties of vastly unequal bargaining power, including defendants and prosecutors. Id. Less convincing is their argument that problems arising from strategic interactions of the parties is a relatively easily remedied structural problem. See Schulhofer, Plea Bargaining as Disaster, supra note 10.
129. 513 U.S. 196 (1995).
130. See Fed .R. Evid. 410. Mr. Mezzanatto was arrested on federal narcotics charges and made an initial proffer to the government. He signed an agreement granting him limited immunity and permitting the government to use any statements he made that day to impeach him if he took the case to trial and testified. During the session he minimized his role in the case and the meeting ended after he was confronted with surveillance evidence that contradicted his statements. At his trial Mr. Mezzanatto was confronted with damaging admissions he had made at his meeting with the government that contradicted his testimony on direct examination. 513 U.S. at 199.
131. 513 U.S. at 202 (quoting 21 C. WRIGHT & K. GRAHAM, FEDERAL PRACTICE AND PROCEDURE § 5039, at 207-08 (1977)).
132. Defendants bear the risk that any breach on their part will result in termination of the agreement. Some cases provide clear justification for the prosecutor's decision to walk away from the deal including the the following cases: those involving escape from custody, United States v. Holmes, No. 96-6080, 1998 WL 124053 (6th Cir. Mar. 30, 1998) (holding that defendant's escape from prison violated agreement), the commission of other crimes, United States v. Fernandez, 127 F.3d 277 (2d Cir. 1997) (declaring that cooperating witness violated his agreement in four tape recorded conversations with former codefendant offering to fix case for $10,000 and rejecting argument that government received benefit of cooperation through eventual plea by codefendant so cooperator should receive his benefit), the withholding of information, United States v. Garcia-Velilia, 122 F.3d 1 (1st Cir. 1997) (explaining that defendant's refusal to disclose names of suppliers of cocaine used by defendant while on pre-trial release was sufficient ground for breach and court did not have to consider whether agreement implicitly required that defendant not commit other crimes), the telling of lies, United States v. Armstrong, 842 F. Supp. 92 (S.D.N.Y. 1994) (holding that defendant's information proved unreliable therefore the agreement was terminated), or the refusal to testify, United States v. Hon, 17 F.3d 21 (2d Cir. 1994) (holding that defendant's refusal to testify violated agreement). Others involving apparently less serious conduct illustrate the allocation of virtually all of this risk to the defendant. See United States v. Brechner, 99 F.3d 96 (2d Cir. 1996) (illustrating how the court accepted government's decision to terminate agreement when defendant denied a key fact and then admitted it after immediate consultation with attorney).
133. Typically, the terms of the agreement obligate the defendant to do the following: plead guilty to the most serious count against him or her, be completely truthful, provide whatever assistance the government seeks, commit no other crimes and waive legal protections offered by the rules of evidence, statutes of limitations and any other additional doctrines added to this list by the government. The government typically agrees that the particular U.S. Attorney's Office entering into the agreement will not further prosecute the defendant for the conduct he or she has revealed to the government and will inform the sentencing judge of the defendant's assistance. See Richman, supra note 1, at 99 (offering examples of agreements which make the defendant's obligations clear and the expected reward vague).
134. See, e.g., United States v. Harris, 1997 U.S. App. LEXIS 16555, at *1, *3 (4th Cir. July 7, 1997) (holding that cooperation agreement gave government discretion to determine whether to make the motion); Shehee v. United States, 1997 U.S. App. LEXIS 2886, at *1-2 (6th Cir. Feb. 14, 1997) (explaining the agreement was that the government "may" move for a downward departure).
135. However, the government's discretion can be relinquished and, in fact, "once the government uses its § 5K1.1 discretion as a bargaining chip in the plea negotiation process, that discretion is circumscribed by the terms of the agreement." United States v. Conner, 930 F.2d 1073, 1075 (4th Cir.). See also Nasman v. United States, 1997 U.S. App. LEXIS 6389, at *1, *5-6 (6th Cir. Apr. 1, 1997) (remanding with instructions that the government must fulfill the agreement and make a downward departure motion at the resentencing); United States v. Courtois, 131 F.3d 937 (10th Cir. 1997) (government may bargain away discretion); United States v. Hernandez, 17 F.3d 78 (5th Cir. 1994) (quoting United States v. Watson, 988 F.2d 544, at 552-53 (5th Cir. 1993), for proposition that government can bargain away its discretion). The government may also limit its discretion by accepting an explicit obligation to exercise "good faith" in deciding whether or not to make the motion. See United States v. Gonsalves, 121 F.3d 1416 (11th Cir. 1997) (holding that the agreement included an express governmental obligation of good faith which was satisfied in this case).
136. It is difficult to even speculate about how often this happens. One case offers evidence that at least one judge thinks it a problem in the Southern District of Texas. In a case reversed because the defendant was mistakenly told at arraignment that the court had greater review power over the government's decision than it came to believe it could exercise at sentence, the judge is quoted in a footnote saying, "And this happens with some frequency, where there is lots of talk of Section 5K.1.1 at time of plea, and then that magically disappears at time of sentencing." The judge goes on to suggest that the United States Attorney's Office be clearer in setting out its expectations for cooperators. United States v. Amaya, 111 F.3d 386, 388-89 n.4 (5th Cir. 1997).
137. See, e.g., United States v. Garcia-Bonilla, 11 F.3d 45, 47 (5th Cir. 1993) (government's refusal to make a departure motion was held unreviewable absent an unconstitutional motive where agreement "expressly provides that the government retains absolute discretion to move for a downward departure"). Other courts have held that absolute governmental discretion is the default rule unless the agreement explicitly limits its decision making. See, e.g., United States v. Deollos, 1998 U.S. App. LEXIS 160 (10th Cir. Jan. 7, 1998) (holding that government retains unfettered discretion unless plea agreement includes an express promise); United States v. Rockwell Intern. Corp. 124 F.3d 1194 (10th Cir. 1997) (hodling the same); United States v. Courtois, 131 F.3d 937 (10th Cir. 1997) (holding the same).
138. See, e.g., United States v. Copeland, 122 F.3d 1063 (4th Cir. 1997) (stating that the refusal to make motion was found not rationally related to any government purpose where district court had extensive opportunities to see cooperator testify and had basis to find that her refusal to implicate one individual was justified and credible); United States v. Pipes, 125 F.3d 638 (8th Cir. 1997) (explaining that review is limited by Wade but includes rationality test; case remanded for hearing, with dissent arguing that defendant did not make substantial showing required for a hearing); Sullivan v. United States, 11 F.3d 573, 575 (6th Cir. 1993) (discussing that an agreement obligated the government to make the motion if the defendant provided substantial assistance and also explaining the government's reasons for determining the assistance not substantially "rational and acceptable").
139. Those courts have limited Wade to cases in which there was no plea agreement and hold the government to a general duty of good faith under contract principles. The Second Circuit has ruled that although the prosecutor is in the best position to evaluate the defendant's cooperation, "the prosecutor's discretion is not unfettered. '[W]here the explicit terms of the cooperation agreement leave the acceptance of the defendant's performance to the sole discretion of the prosecutor, that discretion is limited by the requirement that it be exercised in good faith.' " United States v. Resto, 74 F.3d 22, 26 (2d Cir. 1996) (quoting United States v. Khan, 920 F.2d 1100, 1105 (2d Cir. 1990) (additional citations omitted)). See also United States v. Knights, 968 F.2d 1483 (2d Cir. 1992) (remanding case for further findings where trial court's finding that government's refusal to move for substantial assistance downward departure was made in good faith was not supported by record which showed defendant performed by testifying at a trial and government's reasons for denying motion included four factors that were known at the time the agreement was signed; the government's reasons also included allegation that trial testimony was inconsistent with others but there was no allegation on the record that the testimony was untruthful); United States v. Isaac, 141 F.3d 477 (3d Cir. 1998) (applying contract principles to permit court review of whether government acted in good faith in evaluating defendant's cooperation pursuant to an agreement). The Eleventh Circuit has criticized this approach as too invasive of prosecutorial discretion. See United States v. Forney, 9 F.3d 1492, 1501-03 (11th Cir. 1993).
140. See Pipes, 125 F.3d at 638 (holding that a hearing is required to determine rationality of decision based on hearsay allegations of cooperator's lack of reliability); United States v. Rounsavall, 128 F.3d 665 (8th Cir. 1997) (remanding case for hearing to determine if the government's refusal to make the motion stemmed irrationally from the defendant's brother's refusal to cooperate); United States v. Leonard, 50 F.3d 1152, 1154 (2d Cir. 1995) (stating that hearing required to develop record to permit district judge to evaluate government's claim that cooperator changed his story); Knights, 968 F.2d at 1483 (remanding case for hearing); United States v. Ganz, 806 F. Supp. 1567 (S.D. Fl. 1992) (ordering government to make substantial assistance motion after court reviewed and evaluated defendant's cooperation and government's actions in refusing motion).
141. See, e.g., United States v. Canada, 960 F.2d 263, 269 (1st Cir. 1992) (explaining that there was a breach of agreement to make a specific sentence recommendation when prosecutor followed recommendation by urging the court to "impose a lengthy period of incarceration" and to "send a strong message").
142. See United States v. Jimenez, 992 F.2d 131, 135 (7th Cir. 1993) (finding no breach where prosecutor made the promised sentence recommendation but questioned the validity of the defendant's assertion of threat of physical harm, made for the first time at the sentencing hearing); United States v. Benchimol, 471 U.S. 453 (1985) (deciding that plea agreement does not bind the government to express any particular level of enthusiasm for the recommended sentence and there is no breach of the agreement if the government leaves the sentencing court with the impression that it is less than enthusiastic).
143. The magnitude of the departure might be seen as an aspect of the price paid for cooperation that is controlled by the judge, but his or her ability to effect the market (influence defendants' decisions about cooperation) through that pricing power is limited. Only the prosecutor can open the door to a substantial assistance departure and the defendant often does not know the identity of the judge when he or she has to decide whether or not to cooperate. In addition, prosecutors exercise a fair degree of influence over the magnitude of the departure. Most federal judges are understandably inclined to listen carefully to the prosecutor's characterization of the value of the cooperation and use it to calibrate the magnitude of the departure.
144. See 18 U.S.C. § 3742 (1997) (permitting appeals of sentences imposed in violation of the law, resulting from incorrect application of the Guidelines and outside the Guideline range, but not sentences within a lawful and correctly determined Guideline range). See, e.g., United States v. Ramos-Rosa, No. 96-1912, 1997 WL 312177, at *1 (1st Cir. June 9, 1997) (determining that decision about whether and how much to depart from government's sentence recommendation is unreviewable so long as court recognized and exercised its discretion); United States v. Light, No. 96-5482, 1997 WL 720386, at *4 (6th Cir. Nov. 10, 1997) (reasoning that judge's decision not to depart from government's sentence recommendation, after government motion, is not cognizable on appeal).
145. See United States v. Winters, 117 F.3d 346 (7th Cir. 1997) (utilizing judge's own evaluation of defendant in sentencing process); United States v. Lucas, 17 F.3d 596 (2d Cir. 1994) (refusing a downward departure based on judge's review of defendant); United States v. Foster, 988 F.2d 206 (D.C. Cir. 1993); United States v. Dellinger, 986 F.2d 1042 (6th Cir. 1993); United States v. Mittlestadt, 969 F.2d 335 (7th Cir. 1992); United States v. Dobynes, 905 F.2d 1192 (8th Cir. 1990); United States v. Cloughley, 901 F.2d 91 (8th Cir. 1990).
146. See United States v. Casiano, 113 F.3d 420 (3d Cir. 1997) (determining that sentencing court is not limited to the factors set out in § 5K1.1 and properly considered seriousness of underlying offense in limiting substantial assistance departure to three levels).
147. See United States v. Carnes, 945 F.2d 1013 (8th Cir. 1991) (limiting extent of departure by consideration of plea bargain benefit); United States v. Sutherland, 890 F.2d 1042 (8th Cir. 1989); United States v. Taylor, 868 F.2d 125 (5th Cir. 1989).
148. See Richman, supra note 1 (analyzing extensively and insightfully lawyers' incentives and disincentives to cooperate).
149. See Schulhofer, Criminal Justice Discretion as a Regulatory System, supra note 9, at 50-59 (discussing agency costs for both prosecution and defense); Schulhofer, Plea Bargaining as Disaster, supra note 10.
150. See Schulhofer, Criminal Justice Discretion as a Regulatory System, supra note 9, at 50-59 (discussing agency costs for both prosecution and defense); Schulhofer, Plea Bargaining as Disaster, supra note 10.
151. Although the hypothetical narcotics defendant discussed above, supra note 44, who faces a long sentence and against whom the government has strong evidence may seem to have a relatively straightforward decision that is likely to lead to cooperation, much may depend on how the lawyer presents the situation. For example, the lawyer may highlight the dangers and risks or the benefits of cooperation. More subtly the chances of cooperation may be maximized by presenting the mandatory minimum first and emphasizing the dangers of trial. Conversely, the chance of cooperating may be minimized by a counseling strategy that emphasizes the choice between trial and plea with a passing mention of cooperation near the close, once a decision between those options has begun to form. The lawyer will have relatively less influence on a strong-willed client who perceives himself or herself as well informed but can have great influence, and may exercise it consciously or unconsciously, on the many defendants who are terrified by their situation and mystified by the complexities of our legal system.
152. Richman, supra note 1, at 117.
153. One extreme example is a case involving a lawyer paid by someone other than the defendant. This lawyer actively discouraged his client from cooperating by telling her she would receive no benefit from doing so. Furthermore, he advised her to falsely exculpate herself at the one proffer session conducted at the client's insistence. See United States v. Gonzalez-Bello, 10 F. Supp.2d 232, 235 (E.D.N.Y. 1998).
154. See Richman, supra note 1, at 121-22 (reasoning that a defendant considering cooperation as an option might want a lawyer with some experience in such matters while those defendants who believe their interests are best served by insuring that no one among them cooperates will seek out lawyers with strong reputations for never representing cooperators).
155. See William J. Stuntz, The Uneasy Relationship Between Criminal Procedure and Criminal Justice, 107 YALE L.J. 1, 150 n.124 (1997) (presenting collected studies showing that defendants represented by appointed counsel plead at a higher rate than those represented by retained counsel).
156. There is a tension between the goal of minimizing punishment through cooperation and the culture of institutional defense lawyers which Stuntz characterizes as "valuing fighting the system and gaining victories, whole or partial in any legal way possible." Id. at 155 n.132 (claiming public defenders see themselves as fighting the system). This clash explains why one of the nation's most highly regarded and aggressive public defender offices, Public Defender Services of Washington, D.C., decided not to represent any cooperators. See generally Kim Taylor-Thompson, Individual Actor v. Institutional Player: Alternating Visions of the Public Defender, 84 GEORGETOWN L.J. 2419 (1996).
157. Bill Stuntz argues that institutional defenders are best understood as triaging their cases, maintaining an aggressive adversarial stance in the relatively few cases that appear winnable and taking a more conciliatory, case-processing approach in others. Stuntz, supra note 113, at 159 n.145.
158. See Standen, supra note 75, at 1472 (stating that a prosecutor is a monopsonist or single buyer). For a brief history of the offices of the United States Attorneys, their relationship to Washington and the selection of federal prosecutors, see Griffin Bell & Daniel Meador, Appointing United States Attorneys, 9 J.L. & POL. 247 (1993) (calling for merit selection of U.S. Attorneys controlled by the Attorney General).
159. Michael Fortier's cooperation in the Oklahoma City bombing case and Salvatore Gravano's testimony against convicted crime boss John Gotti are two examples. See Jo Thomas, Swift, Hard Attack in Bombing Trial, N.Y. TIMES, May 12, 1997, at A3; Joseph P. Fried, Ex-Mob Underboss Given Lenient Term for Helping as Witness, N.Y. TIMES, Sept. 27, 1994, at A.
160. Michael Fortier's cooperation in the Oklahoma City bombing case and Salvatore Gravano's testimony against convicted crime boss John Gotti are two examples. See Jo Thomas, Swift, Hard Attack in Bombing Trial, N.Y. TIMES, May 12, 1997, at A3; Joseph P. Fried, Ex-Mob Underboss Given Lenient Term for Helping as Witness, N.Y. TIMES, Sept. 27, 1994, at A.
161. See generally Daniel Richman, Cooperating Defendants: The Costs and Benefits of Purchasing Information from Scoundrels, 8 FED. SENT. REP. 292, 293 (1996).
162. See Daniel Richman, Old Chief v. United States, 83 VA. L. REV. 939, 965-68 (1997) (discussing prosecutors' motivation and arguing that prosecutors care less about winning than about not losing).
163. See id. at 142 n.99 (citing, inter alia, Sidney I. Lezak & Maureen Leonard, The Prosecutor's Discretion: Out of the Closet, Not Out of Control, in CARL F. PINKELE & WILLIAM C. LOUTHAN, DISCRETION, JUSTICE AND DEMOCRACY: A PUBLIC POLICY PERSPECTIVE 44, 46 (1985) ("Winning is important not only for the ego satisfaction and enhancement of reputations that victory brings, but also because a record of winning makes it easier to dispose of cases by favorable plea bargains in the future.")).
164. See id. at 142 n.99 (citing, inter alia, Sidney I. Lezak & Maureen Leonard, The Prosecutor's Discretion: Out of the Closet, Not Out of Control, in CARL F. PINKELE & WILLIAM C. LOUTHAN, DISCRETION, JUSTICE AND DEMOCRACY: A PUBLIC POLICY PERSPECTIVE 44, 46 (1985) ("Winning is important not only for the ego satisfaction and enhancement of reputations that victory brings, but also because a record of winning makes it easier to dispose of cases by favorable plea bargains in the future.")).
165. Id.
166. Indeed many more cases are dismissed than are tried in the federal criminal justice system. See supra note 11.
167. Federal prosecutors must prosecute some classes of cases which can be very difficult to win, such as criminal civil rights cases against police officers and some prosecutions of political figures. Dismissal or very lenient pleas in those cases, in order to minimize the risk of acquittal, would send a very dangerous message to other potential defendants.
168. Perhaps this only seems like an insight to a defense lawyer who represents indigent clients, but it was a revelation to me when a prosecutor first told me that he was supposed to win and the defense was supposed to lose. In addition, while I needed feel no shame in losing a case that had survived his discretionary charging decision, he had the burden of only going forward with cases he believed he could and should win. Although I wish the world were that simple, the system works well enough to give that view more than a grain of truth.
169. See generally Richman, supra note 119, at 292.
170. There are Department of Justice policies which should inform these decisions. See David Robinson, Jr., The Decline and Potential Collapse of Federal Guideline Sentencing, 74 WASH. U. L.Q. 881 (1996) (arguing that the change in Department of Justice charging Guidelines from the more restrictive Thornburgh Memorandum to the more lenient Reno Bluesheet introduced another source of unwarranted disparity). Those directives appear to have little impact beyond their occasional service as post hoc justifications. For a general, if somewhat dated discussion of the problems with developing and enforcing policies to control prosecutorial discretion, see Norman Abrams, Internal Policy: Guiding the Exercise of Prosecutorial Discretion, 19 UCLA L. REV. 1 (1971).
171. There are Department of Justice policies which should inform these decisions. See David Robinson, Jr., The Decline and Potential Collapse of Federal Guideline Sentencing, 74 WASH. U. L.Q. 881 (1996) (arguing that the change in Department of Justice charging Guidelines from the more restrictive Thornburgh Memorandum to the more lenient Reno Bluesheet introduced another source of unwarranted disparity). Those directives appear to have little impact beyond their occasional service as post hoc justifications. For a general, if somewhat dated discussion of the problems with developing and enforcing policies to control prosecutorial discretion, see Norman Abrams, Internal Policy: Guiding the Exercise of Prosecutorial Discretion, 19 UCLA L. REV. 1 (1971).
172. See Weinstein, supra note 25 (manuscript at 19-25). The reduction in judicial discretion eliminated an important outlet for prosecutorial discretion. To the significant degree that prosecutors influenced judicial sentencing decisions under the old law, they could use that influence to control case outcomes. That option is no longer available.
173. See generally Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14 (discussing the shift of power from judges to prosecutors occasioned by the sentencing reforms of the 1980s); Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, supra note 14 (tracing the expansion of prosecutorial power under the federal Sentencing Guidelines); Standen, supra note 75, at 1472.
174. See generally Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14 (discussing the shift of power from judges to prosecutors occasioned by the sentencing reforms of the 1980s); Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, supra note 14 (tracing the expansion of prosecutorial power under the federal Sentencing Guidelines); Standen, supra note 75, at 1472.
175. Substantial assistance does not always ultimately mitigate the defendant's sentence. It may, for example, be a way for a prosecutor to structure a plea to a very serious charge and give a relatively modest sentence break instead of charge bargaining to a non-mandatory minimum or otherwise more desirable charge.
176. See infra pp. 608-11.
177. See Schulhofer & Nagel, Plea Negotiations Under the Federal Sentencing Guidelines, supra note 47, at 1292 (calling charge bargaining the most important vehicle for Guideline evasion).
178. See Lisa M. Farabee, Disparate Departures Under the Federal Sentencing Guidelines: A Tale of Two Districts, 30 CONN. L. REV. 569, 629 (1998) (comparing, in-depth, the districts of Massachusetts and Connecticut detailing the interplay of actors and factors in shaping local practices and noting that Connecticut prosecutors used substantial assistance departures less than others because their judges departed more for given reasons than other judges).
179. Perhaps in an ideal world we would want a prosecutor's only motivation to be a commitment to the law and the public interest, although that posits a world in which there is only one view on where that commitment would take one. Whatever that world would look like, we live in a world in which that motivation is filtered through the prosecutor's self-interest and general concern about reputation. See Schulhofer, Criminal Justice Discretion as a Regulatory System, supra note 9, at 62-63.
180. See Albert J. Krieger, In Drug Defense, Stress Constitution Jurors Must Recognize a Defense Counsel's Obligation to Uphold All Civil Rights, NAT'L L.J., Sept. 22, 1997, at C6 (convincing jurors that cooperators have a motive to lie); SUBIN ET AL., supra note 24, § 18.9 (presenting an example of a cross examination of a cooperating witness by two different attorneys, one of whom must pursue the theory that self-interest is motivating the snitch to lie).
181. Richman has written about how "making cooperation" is a "leap into uncertainty." See Richman, supra note 1, at 94. For the defendant serves the prosecution's interest in maintaining control over him or her between the entry of the plea and the time of sentence.
182. See, e.g., United States v. Wallach, 935 F.2d 445 (2d Cir. 1991) (reversing convictions growing out of the Wedtech scandal because of cooperator's false testimony about his gambling habit); United States v. Griffin, 856 F. Supp. 1293 (N.D. Ill. 1994) (reversing convictions against the notorious El Rukn gang because of outrageous government actions in offering benefits to cooperators and not disclosing them to the defense).
183. See Stephen Trott, Words of Warning for Prosecutors Using Criminals as Witnesses, 47 HASTINGS L.J. 1381, 1383-86 (1996) (discussing the dangers of perjury by cooperators and offering two examples of cooperator perjury that resulted in wrongful convictions).
184. Much of the commentary has favored greater individualization. See Albert W. Alschuler, The Failure of Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization); Kevin Cole, The Federal Sentencing Guidelines: Ten Years Later - The Empty Idea of Sentencing Disparity, 91 NW. U. L. REV. 1336, 1336 (1997) (arguing that the Commission's failure to make fundamental theoretical choices makes the goal of reducing disparity incoherent); Daniel Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize); Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); Stith & Cabranes, supra note 39 (explaining that the Guidelines are excessively rigid). The Guidelines also have their supporters. See Bowman, supra note 24 (arguing that the Guidelines have struck the right balance and only judges whose power has been curtailed dislike them); O'Sullivan, supra note 24 (noting that the Guidelines system properly reflects differences in the underlying criminal conduct). The strongest supporters of uniformity have been the courts of appeal. See Reitz, supra note 39 (describing how the courts have taken a strongly enforcement oriented approach to the Guidelines); Weinstein, supra note 25 (manuscript at 54-63) (stating that most circuit courts have favored uniformity over individualization in their analyses of non-substantial assistance downward departures).
185. Much of the commentary has favored greater individualization. See Albert W. Alschuler, The Failure of Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization); Kevin Cole, The Federal Sentencing Guidelines: Ten Years Later - The Empty Idea of Sentencing Disparity, 91 NW. U. L. REV. 1336, 1336 (1997) (arguing that the Commission's failure to make fundamental theoretical choices makes the goal of reducing disparity incoherent); Daniel Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize); Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); Stith & Cabranes, supra note 39 (explaining that the Guidelines are excessively rigid). The Guidelines also have their supporters. See Bowman, supra note 24 (arguing that the Guidelines have struck the right balance and only judges whose power has been curtailed dislike them); O'Sullivan, supra note 24 (noting that the Guidelines system properly reflects differences in the underlying criminal conduct). The strongest supporters of uniformity have been the courts of appeal. See Reitz, supra note 39 (describing how the courts have taken a strongly enforcement oriented approach to the Guidelines); Weinstein, supra note 25 (manuscript at 54-63) (stating that most circuit courts have favored uniformity over individualization in their analyses of non-substantial assistance downward departures).
186. Much of the commentary has favored greater individualization. See Albert W. Alschuler, The Failure of Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization); Kevin Cole, The Federal Sentencing Guidelines: Ten Years Later - The Empty Idea of Sentencing Disparity, 91 NW. U. L. REV. 1336, 1336 (1997) (arguing that the Commission's failure to make fundamental theoretical choices makes the goal of reducing disparity incoherent); Daniel Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize); Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); Stith & Cabranes, supra note 39 (explaining that the Guidelines are excessively rigid). The Guidelines also have their supporters. See Bowman, supra note 24 (arguing that the Guidelines have struck the right balance and only judges whose power has been curtailed dislike them); O'Sullivan, supra note 24 (noting that the Guidelines system properly reflects differences in the underlying criminal conduct). The strongest supporters of uniformity have been the courts of appeal. See Reitz, supra note 39 (describing how the courts have taken a strongly enforcement oriented approach to the Guidelines); Weinstein, supra note 25 (manuscript at 54-63) (stating that most circuit courts have favored uniformity over individualization in their analyses of non-substantial assistance downward departures).
187. Much of the commentary has favored greater individualization. See Albert W. Alschuler, The Failure of Sentencing Guidelines: A Plea for Less Aggregation, 58 U. CHI. L. REV. 901 (1991) (analyzing the conflicting provisions and arguing for more individualization); Kevin Cole, The Federal Sentencing Guidelines: Ten Years Later - The Empty Idea of Sentencing Disparity, 91 NW. U. L. REV. 1336, 1336 (1997) (arguing that the Commission's failure to make fundamental theoretical choices makes the goal of reducing disparity incoherent); Daniel Freed, Federal Sentencing in the Wake of the Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 YALE L.J. 1681 (1992) (analyzing the conflicting provisions and arguing for greater judicial power to individualize); Stephen J. Schulhofer, Assessing the Federal Sentencing Process: The Problem is Uniformity, Not Disparity, 29 AM. CRIM. L. REV. 833 (1992) (calling for more individualization within the Guidelines system); Stith & Cabranes, supra note 39 (explaining that the Guidelines are excessively rigid). The Guidelines also have their supporters. See Bowman, supra note 24 (arguing that the Guidelines have struck the right balance and only judges whose power has been curtailed dislike them); O'Sullivan, supra note 24 (noting that the Guidelines system properly reflects differences in the underlying criminal conduct). The strongest supporters of uniformity have been the courts of appeal. See Reitz, supra note 39 (describing how the courts have taken a strongly enforcement oriented approach to the Guidelines); Weinstein, supra note 25 (manuscript at 54-63) (stating that most circuit courts have favored uniformity over individualization in their analyses of non-substantial assistance downward departures).
188. Specifically, the Commission was directed to consider "the need to avoid unwarranted sentencing disparities among defendants with similar records who have been found guilty of similar conduct." 18 U.S.C. § 3553 (a) (6) (1996 & Supp. 1998).
189. See Bowman, supra note 24, at 714-24 (stating that there is no convincing case for increased disparity under the Guidelines); Gerald W. Heany, The Reality of Guidelines Sentencing: No End to Disparity, 28 AM. CRIM. L. REV. 161 (1991) (arguing that disparity is just as great under the Guidelines); Schulhofer & Nagel, Plea Negotiations Under the Federal Sentencing Guidelines, supra note 47, at 1284 (stating that plea practices under the Guidelines lead to some disparity but less than under the old law); Stith & Koh, supra note 26, at 287 (stating that sentencing disparity is as great under Guidelines as was before).
190. See Bowman, supra note 24, at 714-24 (stating that there is no convincing case for increased disparity under the Guidelines); Gerald W. Heany, The Reality of Guidelines Sentencing: No End to Disparity, 28 AM. CRIM. L. REV. 161 (1991) (arguing that disparity is just as great under the Guidelines); Schulhofer & Nagel, Plea Negotiations Under the Federal Sentencing Guidelines, supra note 47, at 1284 (stating that plea practices under the Guidelines lead to some disparity but less than under the old law); Stith & Koh, supra note 26, at 287 (stating that sentencing disparity is as great under Guidelines as was before).
191. See Bowman, supra note 24, at 714-24 (stating that there is no convincing case for increased disparity under the Guidelines); Gerald W. Heany, The Reality of Guidelines Sentencing: No End to Disparity, 28 AM. CRIM. L. REV. 161 (1991) (arguing that disparity is just as great under the Guidelines); Schulhofer & Nagel, Plea Negotiations Under the Federal Sentencing Guidelines, supra note 47, at 1284 (stating that plea practices under the Guidelines lead to some disparity but less than under the old law); Stith & Koh, supra note 26, at 287 (stating that sentencing disparity is as great under Guidelines as was before).
192. There is no statistically significant correlation between a district's substantial assistance rate and its overall average or median sentences or its average or median narcotics trafficking sentences. These results are discussed infra pp. 611-17, and in the Statistical Appendix, infra p. 633.
193. There are 4036 total sentences in those districts and 297 total substantial assistance departures cases in those districts. See Statistical Appendix infra tbl.1.
194. Actually, the Northern Mariana Islands have a cooperation rate of 85% for 14 cases. However, this is completely atypical so I have excluded that district from my calculation. The districts I used are ranked 84 to 93. There are 3803 cases in those districts and 1485 total substantial assistance departures. Statistical Appendix, infra tbl.1.
195. As is thematic throughout this section, local variations in practice and other factors counsel caution in working with these numbers. The Commission's statistics probably understate the real use of substantial assistance departures. For example, the fifth lowest reported rate of cooperation is in the Eastern District of Virginia, famous for the "rocket docket." Apparently most cooperation in that district is rewarded with a sentence reduction pursuant to FED. R. CRIM. P. 35(b), which permits the government to request a post-sentence reduction for post-sentence cooperation. Although judges in that district will not postpone sentences to permit cooperation to mature, they will ignore the requirement that 35(b) reductions account only for post-sentence cooperation. See Daniel Richman, The Challenges of Investigating § 5K1.1 in Practice, 11 FED. SENTENCING REP. 75 (1998). These statistics do not include such reductions. If the Eastern District of Virginia is excluded as anomalous (although the practice could be common) and the 11th district (the Western District of Arkansas) were included, the rate is 7.5% (205 out of 3351 cases). Although there are likely other examples of local practices which counsel caution in the use of these numbers, the overall picture still reveals significant disparities.
196. As is thematic throughout this section, local variations in practice and
197. The actual percentage is 9.3% of all sentences (3817 out of 40,879 total sentences imposed). See Statistical Appendix, infra tbl.1.
198. Those that ranked 32 to 68 fell within this range. See Statistical Appendix infra, tbl.1.
199. I have chosen pairs of neighboring districts as natural experiments, with the underlying notion that contiguous districts might be more fairly compared than others chosen to highlight disparity or uniformity. These comparisons are suggestive, although these districts could be compared in other ways. For example, Connecticut and Massachusetts also share a border and have similar rates. In the Guidelines scheme, however, any disparity is troubling, while some consistency proves little.
200. In New Hampshire, 53.5% of narcotics trafficking defendants received substantial assistance departures (46 out of 86) compared to 33.3% (21 out of 63) in Maine. UNITED STATES SENTENCING COMM'N, 1996 FEDERAL SENTENCING STATISTICS BY STATE tbl.9 (1996) [hereinafter 1996 SENT. STATS. BY STATE] (also available at Sentencing Commission home page 〈http//www.ussc.gov/ judpack/jp1996.htm〉 (on file with author and the Buffalo Law Review). The mean sentence for drug trafficking in Maine was 61.2 months compared to a mean of 47.9 months in New Hampshire. Id. at tbl.7. All district-specific data on lengths of sentences and rates of departures for specific offenses are from these files, which contain federal sentencing data by individual state and district and offer more data than the district summary tables available in the 1996 FED. SENT. SRCBK., supra note 2. The national data is found in both sources.
201. In New Hampshire, 53.5% of narcotics trafficking defendants received substantial assistance departures (46 out of 86) compared to 33.3% (21 out of 63) in Maine. UNITED STATES SENTENCING COMM'N, 1996 FEDERAL SENTENCING STATISTICS BY STATE tbl.9 (1996) [hereinafter 1996 SENT. STATS. BY STATE] (also available at Sentencing Commission home page 〈http//www.ussc.gov/ judpack/jp1996.htm〉 (on file with author and the Buffalo Law Review). The mean sentence for drug trafficking in Maine was 61.2 months compared to a mean of 47.9 months in New Hampshire. Id. at tbl.7. All district-specific data on lengths of sentences and rates of departures for specific offenses are from these files, which contain federal sentencing data by individual state and district and offer more data than the district summary tables available in the 1996 FED. SENT. SRCBK., supra note 2. The national data is found in both sources.
202. In New Hampshire, 53.5% of narcotics trafficking defendants received substantial assistance departures (46 out of 86) compared to 33.3% (21 out of 63) in Maine. UNITED STATES SENTENCING COMM'N, 1996 FEDERAL SENTENCING STATISTICS BY STATE tbl.9 (1996) [hereinafter 1996 SENT. STATS. BY STATE] (also available at Sentencing Commission home page 〈http//www.ussc.gov/ judpack/jp1996.htm〉 (on file with author and the Buffalo Law Review). The mean sentence for drug trafficking in Maine was 61.2 months compared to a mean of 47.9 months in New Hampshire. Id. at tbl.7. All district-specific data on lengths of sentences and rates of departures for specific offenses are from these files, which contain federal sentencing data by individual state and district and offer more data than the district summary tables available in the 1996 FED. SENT. SRCBK., supra note 2. The national data is found in both sources.
203. Massachusetts' overall rate of substantial assistance departures was 25.6%, with a rate of 37.5% in narcotics trafficking cases (63 out of 168 cases), and an average imprisonment length for narcotics trafficking cases below the averages. The mean sentence length was 64.2 months, compared with a national mean of 86.7 months and the median was 44.5 months, compared with the national median of 60 months. Rhode Island's substantial assistance rate of 4.6% includes a rate of 3.9% in narcotics trafficking cases (2 out of 51 cases) and average imprisonment lengths closer to, but still below the national mean, at 76.5 months, with a median of 60, matching the national median. See 1996 SENT. STATS. BY STATE, supra note 148, tbls.7 & 9. A comparison of the distribution of case types and other basic data suggests no obvious difference, such as distribution of case types, between Rhode Island and its neighbor that would explain the different rates.
204. Connecticut's drug sentences were quite long, with a mean of 104.8 months and a median of 87 months. Vermont's mean sentence in drug trafficking cases (56.6 months) was below its median sentence, equal to the national median of 60 months. See id. at tbl.7.
205. See id. at tbl.9.
206. The percentage of substantial assistance departures for narcotics trafficking varied greatly throughout the three districts in Illinois. In the Central District of Illinois Central, 51.8% (59 out of 114) of narcotics defendants received departures. In the Northern District of Illinois, 36.8% (39 out of 106) of narcotics defendants received departures, and in the Southern District of Illinois, 6.8% (17 out of 250) of narcotics defendants received departures. See id. The average sentences also varied throughout the districts. In the Central District of Illinois, the mean sentence for narcotics trafficking was 98 months compared with 119.7 months in the Northern District and 109.5 months in the Southern District. Id.
207. Although the four California Districts were all low cooperation districts, in the Central District 6.9% of defendants received departures compared to 8.8% in the Eastern District, 12.2% in the Northern District and 9.5% in the Southern District. See 1996 FED. SENT. SRCBK., supra note 2, at 43 tbl.26.
208. Id.
209. The number of sentences mitigated are 388 out of 817 total. Although 67.3% of the narcotics trafficking defendants received substantial assistance departures, the median drug sentence length matched the national mean of 60%, and the median was only 6 months shorter, 80.6 months compared to 86.7 months. The average percentage of defendants receiving substantial assistance departures in the Middle District of Pennsylvania was 37.3%, compared to 20.2% in the Western District of Pennsylvania, 30.9% in New Jersey, 15% in Delaware, and 14.3% in the Virgin Islands. See 1996 SENT. STATS. BY STATE, supra note 148, at tbls.7& 9.
210. See 1996 FED. SENT. SRCBK., supra note 2, at 43 tbl.26.
211. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 124-28 (discussing mechanisms and policies in several different districts); Saris, supra note 52, at 1045, 1048 (noting differing policies on whether or not to recommend specific sentences and differing policies on when to make the motion).
212. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 6-7 (describing methodology).
213. The lowest rate of complete consistency, 44.4%, was for the 36 districts that reported that the U.S. Attorney, him or herself, reviewed the decision, while the highest rate, 63.2%, was for districts that had review or approval by an Assistant U.S. Attorney. The highest rate of inconsistency, 41.2% was in districts that had a review committee. SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 8 exh.3.
214. Although not every district would consider each activity worthy of the same reward. See Saris, supra note 52, at 1045-46.
215. Eighty-eight districts responded to the mail survey. All 88 noted that they would consider testimony under oath; 87 would consider participation in the investigation of another and information in the prosecution of others. Eighty-one would consider information about the criminal activity of others and 43 would consider information about the defendant's own activities. SENT. COMM. SUBST. ASS. STUDY, supra note 58, at exh.4.
216. Id. There is one potentially significant ambiguity in this section of the report. Although it clearly notes that most prosecutors thought those who testified should receive a motion, it goes to say some "used self-incriminating information in considering substantial assistance." Id. at 9 exh.4. Although the report treats both as answering the question of whether the one activity alone was enough for a motion, if the question was posed as the report suggests, some prosecutors may have been indicating that they would consider self-incriminating information as one factor so that a defendant who gave information about others and self-incriminatory information freely might do better than one who only gave information about others. Id.
217. Id. There is one potentially significant ambiguity in this section of the report. Although it clearly notes that most prosecutors thought those who testified should receive a motion, it goes to say some "used self-incriminating information in considering substantial assistance." Id. at 9 exh.4. Although the report treats both as answering the question of whether the one activity alone was enough for a motion, if the question was posed as the report suggests, some prosecutors may have been indicating that they would consider self- incriminating information as one factor so that a defendant who gave information about others and self-incriminatory information freely might do better than one who only gave information about others. Id.
218. The explicit statutory and Guideline language specifies that the assistance be in the "investigation or prosecution of another." 18 U.S.C. § 3553(e) (1994); U.S.S.G. § 5K1.1.
219. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 10 n.21 (describing this aspect of the study).
220. All § 5K1.1 recipients were assumed to have provided assistance but only those non-section 5K1.1 recipients whose files had actual documentation of assistance were counted as having attempted to cooperate. Those who provided assistance not reflected in the case file and did not receive the motion were not counted, making the 67.5% figure a conservative lower bound estimate. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at exh.5 n.21.
221. Id.
222. See id. The data does not indicate whether the two who testified or provided tangible evidence but did not receive the motion violated their agreement in some fashion and so did not receive the benefit for that reason.
223. See id.
224. See id.
225. Id. at 11. In addition, some defendants lose eligibility for § 5K1.1 motions by committing new offenses, lying or otherwise breaking their agreement with the government.
226. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at Exh.11.
227. See Schulhofer & Nagel, supra note 47, at 1292 (concluding that charging and bargaining undercut the Guidelines). 173. See, e.g., United States v. Gonzalez-Bello, 10 F. Supp.2d 232 (E.D.N.Y. 1998) (discussing sentence factor bargaining over Guideline level adjustments).
228. The Guidelines sentence is supposed to reflect all of a defendant's "relevant conduct." See U.S.S.G., supra note 3, at § 1B1.3 (listing, among other things, instances of similar misconduct which may go beyond those charged and conduct carried out by a co-defendant or co-conspirator if it was reasonably foreseeable). If a defendant is charged with narcotics trafficking, the Guideline sentence may be increased to reflect narcotics dealt by co-conspirators or transactions other than those charged in the indictment. See generally Edwards v. United States, 118 S. Ct. 1475 (1998) (concluding that judge determines whether conduct is part of same course of conduct); Paul M. Secunda, Cleaning Up the Chicken Coop of Sentencing Uniformity: Guiding the Discretion of Federal Prosecutors Through the Use of the Model Rules of Professional Conduct, 34 AM. CRIM. L. REV. 1267, 1275 (1997) (discussing how relevant conduct rules may be used by prosecutors to exercise discretion).
229. The Guidelines sentence is supposed to reflect all of a defendant's "relevant conduct." See U.S.S.G., supra note 3, at § 1B1.3 (listing, among other things, instances of similar misconduct which may go beyond those charged and conduct carried out by a co-defendant or co-conspirator if it was reasonably foreseeable). If a defendant is charged with narcotics trafficking, the Guideline sentence may be increased to reflect narcotics dealt by co-conspirators or transactions other than those charged in the indictment. See generally Edwards v. United States, 118 S. Ct. 1475 (1998) (concluding that judge determines whether conduct is part of same course of conduct); Paul M. Secunda, Cleaning Up the Chicken Coop of Sentencing Uniformity: Guiding the Discretion of Federal Prosecutors Through the Use of the Model Rules of Professional Conduct, 34 AM. CRIM. L. REV. 1267, 1275 (1997) (discussing how relevant conduct rules may be used by prosecutors to exercise discretion).
230. See Michael Gelacek et al., Departures Under the Federal Sentencing Guidelines: An Empirical and Jurisprudential Analysis, 81 MINN. L. REV. 299, 361 (1996) (finding that those districts that were relatively lenient before the Guidelines remain relatively lenient).
231. There is a statistically significant negative correlation (r = -0.27 (p<0.01)) between substantial assistance and other downward departures. See Statistical Appendix, infra tbl.2.
232. See Farabee, supra note 132, at 596 (supporting the interpretation of the nationwide pattern by a case study which compares Connecticut, a low substantial assistance but high other departure district, with Massachusetts, a low other departure but high substantial assistance district; Massachusetts typifies the first and Connecticut the second).
233. See, e.g., United States v. Wallace, 114 F.3d 652 (7th Cir. 1997) (detailing a fact pattern where government recommended three level downward departure for cooperation and judge only departed one level); United States v. Jimenez, 992 F.2d 131 (7th Cir. 1993) (specifying an agreement that the government would recommend a 25% departure). See generally Saris, supra note 52, at 1046-49 (determining magnitude of departure by variations in prosecutorial and judicial practices).
234. Although it is only one anecdote, the judicial attitude is captured in United States v. Garcia, 1997 U.S. App. LEXIS 30633 (6th Cir. 1997) (affirming as within his discretion the district judge's refusal to depart because so many cases involve § 5K1.1 motions; therefore the instant case was not unusual).
235. The only significant correlation was that both the mean and median for fraud sentences increased as substantial assistance increased. Otherwise there was no correlation. See Statistical Appendix, infra tbl.2. This and other district wide sentencing numbers only compare district wide results and do not reveal any information about the actual distribution of the individual sentences in the district.
236. There is a statistically significant correlation between the rate of other downward departures and the mean for all sentences (r= 0.34 (p<0.01)), the median for all sentences (r= 0.25 (p<0.05)), the mean for all narcotics trafficking sentences and the median for all narcotics trafficking sentences (r= 0.33 (p<0.01)). See Statistical Appendix, infra tbl.2.
237. Although this runs counter to the intuition that "harsh" or law enforcement oriented judges tend to reward cooperators highly.
238. This distribution would suggest that those against whom cooperators provide assistance receive longer sentences balancing out the mitigated sentences of the snitches.
239. The Commission has recommended that prosecutors be required to file a statement of reasons for their motions to permit better study of this area. SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 21.
240. Women, non-minority defendants, U.S. citizens, defendants with high school educations and younger defendants all showed a higher statistically significant probability of receiving the departure than their paired counterparts. SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 19 exh.9. Except for age, all those factors also correlated with longer departures of from 9% for women to 2% for high school graduates. Age correlated negatively with the magnitude of the departure. Id. at exh.12.
241. Women, non-minority defendants, U.S. citizens, defendants with high school educations and younger defendants all showed a higher statistically significant probability of receiving the departure than their paired counterparts. SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 19 exh.9. Except for age, all those factors also correlated with longer departures of from 9% for women to 2% for high school graduates. Age correlated negatively with the magnitude of the departure. Id. at exh.12.
242. See Oliss, supra note 12, at 1858 (stating that lower level defendants will end up with higher sentences than their more culpable and better informed supervisors); Schulhofer, supra note 29, at 212 (same); but see Bruce H. Kobayashi, Deterrence with Multiple Defendants: An Explanation for "Unfair" Plea Bargains, 23 RAND J. ECON. 507 (1992) (arguing that overall deterrence is maximized because increase in penalties for subordinates outweighs reduction in other sentences); Richman, supra note 1, at 80 n.49 (discussing inverted sentencing and citing two judicial expressions of concern).
243. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 11 exh.6.
244. This assumption is theoretically problematic. See infra p. 61.
245. See SENT. COMM. SUBST. ASS. STUDY, supra note 58, at exh.7.
246. Id. at 13.
247. Valuing cooperation presents very difficult questions. What are the relative merits of a cooperator whose testimony is sufficient to convict a mid-level player and cooperator who provides intelligence, but little or no admissible testimony against a high level criminal? Should we factor in the break to the cooperator and determine that mitigation of low-level defendant's sentence to convict a mid-level player results in greater overall value than mitigation of a mid-level defendant's sentence to convict the same criminal?
248. This "crime to time ratio" is of central importance in understanding how prosecutors allocate resources. Stuntz, supra note 120, at 24 ("the effects of the crime to time ratio . . . almost certainly dwarf the effects of variations in legal doctrine on both the number of charges filed and conviction rates").
249. This assumes that courts could handle more cases. Although cooperating defendants may assist prosecutors to speed some cases, in many districts crowded court dockets are more likely limiting factors on the number of cases that are prosecuted and the speed with which they are handled.
250. A case might be better if it involves more serious violators, more serious violations, averts more harm or has a greater deterrent effect than would the prosecution of the case to which it is compared.
251. First there are cooperators who give information about their own cases. The revelation of cooperators might be assumed to persuade some co-defendants to plead guilty. Although the rate of pleas has risen slightly, and some low cooperation districts have relatively higher trial rates, that may be explained as the result of defendants not being induced to plead guilty with cooperation deals, rather than a lack of persuasive evidence to induce pleas in non-cooperating defendants. Cooperation does have that effect in some cases, but there is no obvious evidence that it has resulted in appreciably higher plea rates, particularly comparing some high and low cooperation districts. In many narcotics cases the evidence is already strong and defendants choose to go to trial because the sentence is so high, even on a plea, that they have little incentive to spare government resources. Other cooperators have evidence about other individuals not yet arrested in that same case or who have committed other crimes. Should those cases be pursued without comparison against the vast pool of other unprosecuted cases? Cooperation requires resources that are not being directed toward the prosecution of another case. The problem for federal prosecutors is not the detection of crime, it is the selection of crime. Finally cooperation might result in higher quality cases. Perhaps cooperators are in the best position to give information on the next level up in the organization, and so on. If this were generally true, the high district to district variations would be more surprising, unless it turned out that districts with low rates of departures are gaining most of the benefits without paying much for cooperation by not making the motion in most cases. For a defendant facing a very long sentence after either a regular plea or a trial, a small chance at significant mitigation may be adequate inducement, especially if it is the same small chance everyone else gets. The question is whether widespread cooperation alone is better than other techniques such as undercover work and confidential informants along with more judicious use of cooperators.
252. The Sentencing Commission and some of the courts of appeal are among the strongest proponents of uniformity through their strict adherence to the Guidelines sentencing structure. See Reitz, supra note 39, at 1465-66 (explaining that many of the appellate courts have taken a high enforcement/low creativity approach to the Guidelines); Stith & Cabranes, supra note 39, at 1272-74 (criticizing the Commission's refusal to share interpretive authority); Weinstein, supra note 25 (manuscript at 19-31) (distinguishing the uniformity and individualization approaches to the Guidelines).
253. Among the staunchest advocates for individualized sentencing are the Supreme Court as illustrated by its decision in Koon v. United States, 518 U.S. 81, 116 (1996) and the Second Circuit. See Weinstein, supra note 25 (manuscript at 48-54) and other commentators, supra note 138 (offering examples of the rhetoric of individualized sentencing).
254. This is exemplified by Koon, 518 U.S. at 116. See supra note 41 and accompanying text.
255. At least one court has expressed that view. "The desire to dictate the length of a defendant's sentence for reasons other than his or her substantial assistance is not a permissible basis for exercising the government's power under § 3553(e)." United States v. Stockdall, 45 F.3d 1257, 1261 (8th Cir. 1995). However, another judge, in the role of commentator, urged that cooperation departures be used to avoid "ultra-uniformity" early in the development of Guideline sentencing. Bruce M. Selya & John C. Massaro, The Illustrative Role of Substantial Assistance Departures in Combating Ultra-Uniformity, 35 B.C. L. REV. 799 (1994) (discussing areas of flexibility in the developing substantial assistance area). Another court has argued that sentence factor bargaining, by contrast, is properly viewed in the tradition of charging decisions as legitimate exercises of prosecutorial discretion. See United States v. Gonzalez-Bello, 10 F. Supp.2d 232, 237 (E.D.N.Y. 1998) (calling the Guidelines ambivalent on such bargaining and arguing that it should be expressly permitted).
256. Gleeson, supra note 2, at 452-55 (explaining how prosecutors come to feel compassion for cooperating witnesses and reward them more than their agreements might require).
257. Stuntz argues that the notion that institutional defenders are just case-processors has been largely rejected to account for evidence that the adversarial role conception persists and controls the few cases that appear to merit the investment in resources. Stuntz, supra note 113, at 41 n.145 (citing ROGER A. HANSON ET AL., NATIONAL CTR. FOR STATE COURTS, INDIGENT DEFENDERS GET THE JOB DONE AND DONE WELL 66-69 (1992); LISA J. MCINTYRE, THE PUBLIC DEFENDER: THE PRACTICE OF LAW IN THE SHADOWS OF REPUTE 46-49, 148-50 (1987)).
258. Stuntz argues that the notion that institutional defenders are just case- processors has been largely rejected to account for evidence that the adversarial role conception persists and controls the few cases that appear to merit the investment in resources. Stuntz, supra note 113, at 41 n.145 (citing ROGER A. HANSON ET AL., NATIONAL CTR. FOR STATE COURTS, INDIGENT DEFENDERS GET THE JOB DONE AND DONE WELL 66-69 (1992); LISA J. MCINTYRE, THE PUBLIC DEFENDER: THE PRACTICE OF LAW IN THE SHADOWS OF REPUTE 46-49, 148-50 (1987)).
259. Michael McConville & Chester L. Mirsky, Criminal Defense of the Poor in New York City, 15 N.Y.U. REV. L. & Soc. CHANGE 581, 582-83 (1986-87) (studying the failure of appointed criminal defense lawyers to engage in substantive motion practice and concluding they were primarily case-processors interested in efficient disposition of cases). Regardless of whether Mirsky and McConville are correct in generalizing their conclusions to all defenders and cases, it seems likely that most institutional defenders have different goals (and a different role conception informing their work) for the majority of cases they triage as losers.
260. I once represented an older man charged with committing immigration fraud by entering into several sham marriages with women who then used the marriages to regularize their immigration status. It was an early Guidelines case and although their enforcement was still fluid, my client faced the possibility of a prison sentence of 12 to 18 months. He decided to try to cooperate. Our first meeting began rather poorly, with the prosecutor seeming uninterested in his information against the women involved and rather badly disposed toward his acts of bigamy. After a time my client asked the prosecutor if she would mind if he removed his false teeth, as they were causing him some discomfort. She smiled and assured that he should make himself comfortable. As my client removed his teeth, the prosecutor and I exchanged smiles and the meeting took a friendlier turn. He entered into an agreement and received probation. Although I will never know, I have always thought of that as the turning point of the case - the moment my client became an old man whose false teeth hurt, instead of, or in addition to being a criminal to that prosecutor. Although it may not have been her conscious thought, perhaps it was then that she decided to become his advocate for sentence mitigation.
261. This meeting occurs while the defendant usually has minimal legal protection. See supra notes 80 & 85.
262. See United States v. Ming He, 94 F.3d 782, 793-95 (2d Cir. 1996) (discussing the role of defense counsel in opinion invoking supervisory power to prevent cooperating witnesses' statements made without counsel in a cooperation debriefing session from being used to negatively evaluate cooperator's assistance); Gleeson, supra note 2, at 442 (criticizing the use of supervisory power in United States v. Hammond, 858 F.2d 834 (2d Cir. 1988) and offering an alternative view of the defense lawyer's role in cooperation cases).
263. Gleeson, supra note 2, at 450-53 (discussing how defense attorneys may limit their attendance at meetings to avoid the possibility of becoming witnesses themselves but more fundamentally because they regard the prosecutor as the client's ally).
264. Prosecutors must explore the defendant's prior bad acts and other sensitive areas to properly evaluate the cooperation, meet disclosure obligations and prepare for cross examination. See Brady v. Maryland, 373 U.S. 83, 84-86 (1963) (holding that government's failure to disclose exculpatory evidence requires reversal); Giglio v. United States, 405 U.S. 150 (1972) (explaining that the government's failure to disclose promise of leniency to cooperator is a violation of the defendant's due process rights). These discussions are just one source of tension in a relationship between two people who would typically be very unlikely to choose each other's company if each had the choice.
265. Taylor-Thompson, supra note 114, at 2457 (discussing the decision of the highly regarded District of Columbia Public Defender Service not to represent any cooperators); Richman, supra note 1, at 116-17 (giving examples of lawyer's ideological resistance to cooperation).
266. Gleeson, supra note 2, at 454 (arguing that prosecutors have a propensity to favor cooperators even if they may not deserve it). 210. 18 U.S.C. § 4 (1998).
267. Richman, supra note 1, at 83 (discussing the glorification of McCarthy era cooperators).
268. Id.
269. MODEL CODE OF PROFESSIONAL RESPONSIBILITY DR 1-103 (1983) (requiring lawyers to report violations of the Code to the relevant authority); MODEL RULES OF PROFESSIONAL CONDUCT Rule 8.3 (1995) (same).
270. MODEL CODE OF PROFESSIONAL RESPONSIBILITY DR 1-103 (1983) (requiring lawyers to report violations of the Code to the relevant authority); MODEL RULES OF PROFESSIONAL CONDUCT Rule 8.3 (1995) (same).
271. See generally Cynthia L. Gendry, Ethics - An Attorney's Duty to Report the Professional Misconduct of Co-Workers, 18 S. ILL. U. L.J. 603, 606-15 (1994) (arguing that deterrents to reporting the misconduct of lawyers in one's own law firm are particularly strong); Gerard E. Lynch, The Lawyer as Informer, 1986 DUKE L.J. 491 (1986) (stating that the duty to report ignores the moral status of informing and is unenforceably vague and over-inclusive).
272. See generally Cynthia L. Gendry, Ethics - An Attorney's Duty to Report the Professional Misconduct of Co-Workers, 18 S. ILL. U. L.J. 603, 606-15 (1994) (arguing that deterrents to reporting the misconduct of lawyers in one's own law firm are particularly strong); Gerard E. Lynch, The Lawyer as Informer, 1986 DUKE L.J. 491 (1986) (stating that the duty to report ignores the moral status of informing and is unenforceably vague and over-inclusive).
273. The Unabomber case offers the contemporary hard case. Even in the face of the ongoing murders of innocent people, many questioned the morality of a brother turning in a brother.
274. GEORGE P. FLETCHER, LOYALTY 20-21 (1993).
275. Id. at 17-21. There has to be some other relationship against which the special relationship can be measured.
276. Id. at 18.
277. The people most likely to tell that story are judges, United States v. Ming He, 94 F.3d 782 (lauding the cooperator's contrition) and prosecutors, Gleeson, supra note 2, at 453 (stating that most prosecutors have a far more favorable view of accomplice witnesses stemming from a combination of valuing their contrition and the prosecutor's own desire to "make cases").
278. More theoretically, if Fletcher is right about the roots of loyalty, then seeking personal benefit is the most basic betrayal of the sense of common good that flows from according some relationships special deference. If one must disadvantage a member of the group at least one should forego weakening the group further by increasing the inequity through advantaging oneself - if the group is one whose cohesion we wish to encourage. Disloyalty that threatens the efficacy of the criminal group is the very thing for which substantial assistance departures provide an incentive. The question is whether, or perhaps how much of that social utility strikes the right balance against our moral ambivalence.
279. Fletcher's argument provides a strong basis for this concern by placing loyalty at the core of our civic virtues.
280. This positive spin on the morality of the government's use of cooperators is not universally shared. Three judges in the United States Court of Appeals for the Tenth Circuit reversed the judgment of the district court and ordered suppression of a cooperator's testimony because, the panel ruled, the government committed criminal bribery, in violation of 18 U.S.C. § 201(c) (2), by promising leniency in exchange for cooperation. See United States v. Singleton, 144 F.3d 1343 (10th Cir. 1998). Labeling the government's routine conduct in one-fifth of the federal criminal prosecutions criminal is a fundamental attack on the morality of the practice. The panel's decision was vacated when the Tenth Circuit granted en banc review on its own motion. See id. at 1361. The Circuit affirmed the district court, rejecting the panel's opinion and holding that the long history of rewarding cooperators with leniency, stretching back to common law, puts this conduct outside of the reach of the bribery statute. United States v. Singleton, No. 97-3178, 1999 WL6469, at *1 (10th Cir. Jan. 8, 1999) (en banc).
281. The modern example of a society overrun by government endorsed snitching was the former East Germany. The extraordinary revelation of the huge numbers of neighbors, colleagues and friends reporting on each other to a massive security bureaucracy should give us pause. Although there is no comparison between that experience and the current use of cooperators in federal law enforcement, we might well reflect on how pervasive we think snitching should be.
282. See Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, supra note 14, at 245 (identifying nationwide prosecutorial Guidelines as the most promising reform); Saris, supra note 52, at 1050 (suggesting that the Department of Justice should develop and publish principled justifications for prosecutorial discretion in substantial assistance cases); SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 22 (recommending that prosecutors be required to file a statement of reasons for their motions to permit development of more systematic approaches).
283. See Lee, From Gatekeeper to Concierge: Reigning in the Federal Prosecutor's Expanding Power Over Substantial Assistance Departures, supra note 14, at 245 (identifying nationwide prosecutorial Guidelines as the most promising reform); Saris, supra note 52, at 1050 (suggesting that the Department of Justice should develop and publish principled justifications for prosecutorial discretion in substantial assistance cases); SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 22 (recommending that prosecutors be required to file a statement of reasons for their motions to permit development of more systematic approaches).
284. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 177 (asserting that judges, not prosecutors, should have the ultimate authority to determine who does and does not receive a substantial assistance departure); SENT. COMM. SUBST. ASS. STUDY, supra note 58, at 21 (suggesting that a choice be made between regulating departure magnitude on an absolute or proportional basis).
285. See Stith & Cabranes, supra note 39, at 1281 (discussing the dangers of rigidity and micro management).
286. See Stuntz, supra note 113, at 22-26 (asserting that prosecutors determine how scarce enforcement resources are allocated).
287. See Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, supra note 14, at 160 (asserting that prosecutorial discretion is an integral part of a system in which the legislature typically over-criminalizes and over-punishes).
288. These considerations only apply to legislative efforts to control charging decisions. Attempts to directly regulate the relationship between cooperators and prosecutors face other problems, particularly the risk of compromising the prosecutor's current ability to control cooperators by deferring benefits and forcing them to accept most of the risks of their misbehavior.
289. One commentator has convincingly argued that the best way to control discretion is to disperse it rather than regulate it. Thus, the antidote for the excessive concentration of power that has arguably resulted from the reduction in judicial discretion at the same time as prosecutorial discretion has been increased is to disperse sentencing power by restoring judicial sentencing discretion. See Standen, supra note 75, at 1531-37 (calling for a return of judicial sentencing discretion to disperse and so counterbalance the power now concentrated in the prosecutor).
290. See supra note 29 (discussing the conflicts between the Guidelines and mandatory minimums).
291. Mitigation includes any non-substantial assistance downward departures and downward adjustments under the Guidelines that would take the sentence below the statutory punishment. Mandatory minimum sentences are also criticized for shifting power to prosecutors by making sentencing subject to charging decisions.
292. See Yellen, supra note 27, at 585-90 (explaining that political pressures create harsher sentencing regimes). Even so, some who support harshness in sentencing recognize the problems with mandatory minimums. See Hatch, supra note 29, at 192-97 (recognizing problems with mandatory minimums but concluding that the legislature's ultimate responsibility is to restore the voters' confidence in the criminal justice system).
293. See generally Weich, supra note 33 (opposing mandatory minimum sentences in Congress is politically unpopular but good policy).
294. It would be useful to determine if there is any relationship between rates of substantial assistance and any indicators of law enforcement efficiency or success, such as convictions per prosecutor or crime rates. Although clear relationships seem unlikely in such a complicated system, these relationships should be explored.
295. There are 31 districts that report cooperation rates at 15% and below. See Statistical Appendix, infra tbl.1.
296. Some will be concerned about limiting useful law enforcement oriented cooperation. I have argued that differential rates in neighboring districts and other evidence suggests that much of the cooperation is of marginal or no benefit to achieving crime control goals. The unpersuaded may at least agree that the differences in rates show that if there is a great deal of useful cooperation in all districts, prosecutors have a variety of mechanisms for encouraging its use. Cooperation involving information but not testimony raises fewer issues about keeping defendants honest through uncertainty during an extended period of time and some of that cooperation might be expected to be rewarded with charge and sentence bargaining. Indeed, that kind of evasion is to be expected, given the degree of evasion and manipulation that characterizes the Guidelines. A strict limit on substantial assistance departures would not equal a strict limit on snitching. Even so, limited evasion is not worrisome because the goal is to encourage fewer, more carefully chosen instances of cooperation overall, not to strictly control or monitor each choice.
297. Federal law permits the government to seek authority for wiretaps that "may provide" evidence of a very wide array of crimes, 18 U.S.C. § 2516 (1994 & West Supp. 1998), and law enforcement officers "may see what may be seen 'from a public vantage point where [they have] a right to be.' " Florida v. Riley, 488 U.S. 445, 449 (1989) (quoting California v. Ciraolo, 476 U.S. 207, 213 (1986) (holding that warrantless helicopter observation of marijuana plants not observable by ground observation from outside of home does not violate Fourth Amendment)).