On Oct. 12, the House Judiciary Committee’s Subcommittee on Crime, Terrorism, and Homeland Security held its first hearing on federal sentencing since March 2006. The title—”Uncertain Justice: The Status of Federal Sentencing and the U.S. Sentencing Commission Six Years After U.S. v. Booker“—apparently reflects the subcommittee majority’s views on post-Booker1 sentencing. Representative James Sensenbrenner, chairman of the subcommittee, stated that he is “deeply concerned” with federal sentencing due to the “increasing frequency of downward departures.”2 Those who commit child pornography or fraud offenses “are in luck,” he added, explaining that these crimes have seen particularly high rates of below-guideline sentences.3

Proposals to address these concerns have run the gamut. One witness recommended that the Sentencing Reform Act be repealed and the Sentencing Commission abolished.4 Another proposed that sentencing guidelines once again be presumptively applicable, circumventing Booker by relying on juries to make findings on aggravating factors.5 And Judge Patti Saris, chair of the Sentencing Commission and Massachusetts district judge, testified about the problems sentencing courts have encountered in the post-Booker sentencing regime and recommended a more robust system of appellate review.

The commission’s proposals, although understandable, did not address the real culprit—the guidelines themselves. In several areas, the guidelines, primarily as a result of congressional directives, no longer fairly reflect the varied purposes for which they were created.


Judge Saris’ testimony focused on the appellate standard of review. Describing Booker and its progeny as weak,6 she recommended “revitaliz[ing] appellate review” to ensure that the guidelines are given “respectful weight.”7

Her proposals, if implemented, would effectively overrule no fewer than four post-Booker Supreme Court decisions. First, Judge Saris recommended that Congress adopt a presumption of reasonableness for within-range sentences. This would create tension with the Supreme Court’s decision in Rita v. United States,8 which held that courts of appeals may, but are not required to, apply such a presumption. Second, Judge Saris asked that Congress require sentencing courts to provide more justification for sentences based on the degree of variance from the applicable guideline range.

While not explicitly overruling Gall v. United States,9 the commission’s proposal comes close to the proportionality test that the Supreme Court rejected in its opinion. Third, Judge Saris recommended that Congress require a more rigorous standard of review for non-guideline sentences based on policy disagreements. This recommendation attempts to blunt the force of Kimbrough v. United States10 and Spears v. United States,11 which allow district courts to reject the guidelines for exactly this reason, particularly when the guideline range is not a product of “empirical data and national experience.”

The Real Problem

The commission is charged with ensuring that the guidelines account for the multiple purposes of punishment, address aggravating and mitigating factors, and provide certainty and fairness by avoiding unwarranted sentencing disparities. Although Booker, Rita, Gall and Kimbrough have made it easier for district courts to issue outside-the-guideline sentences, the real problem is the guidelines themselves.

In the past two decades, the guidelines, as one observer has noted, have shifted from a defensible effort to promote uniformity and fairness in sentencing to one that is “a one-way upward ratchet increasingly divorced from considerations of sound policy” and “commonsense.”12 The commission has consistently increased both the base offense level of federal offenses and specific offense characteristics, often tying the adjustments to quantitative calculations—for example, drug quantity in trafficking and possession cases and the amount of loss in fraud cases—that are not necessarily well-correlated with a defendant’s culpability.

Though the commission promulgates the guidelines, Congress is largely responsible for the one-way upward ratchet. Congress has directed the commission to amend the guidelines over 100 times. Of these, only two—the 1994 drug “safety valve” and provisions in the Fair Sentencing Act of 2010—provided for decreases in sentencing ranges that were not paired with corresponding increases. The commission has implemented many of these directives with little or no explanation of how they promote the goals of §3553(a).13

The drug guideline, for instance, is a direct result of the mandatory minimum sentences that Congress enacted in 1986. As the commission explained in its recently issued report on mandatory minimum penalties, the guideline “extrapolates upward and downward from the mandatory minimum thresholds…for all drug quantities.”14 This approach favors proportional sentencing at the expense of low-level offenders, whose sentencing ranges are unnecessarily high because of the effort to link guideline ranges with statutory minimums. By adopting Congress’ “weight-driven scheme,” the commission did not use “an empirical approach based on data about past sentencing practices,” as it has for other guidelines.15

Like the drug guideline, the fraud guideline is tied primarily to quantitative factors—the amount of loss and number of victims. These are relevant factors, for sure, but ones that are not necessarily the best proxies for a defendant’s culpability.16 Yet in response to Congress’ directive in the Sarbanes-Oxley Act of 2002, the commission expanded the existing enhancements for amount of loss and number of victims and added other specific offense characteristics that are frequently applied in high-loss corporate fraud cases.17 And this was only two years after the commission significantly expanded the guideline range for many of the same factors. “These higher penalty levels,” the commission explained, were based on comments from the Justice Department and others that the guideline range for fraud “under-punish[ed]” offenders “ relative to…offenses of similar seriousness.”18 Though it is unclear which offenses the commission was referring to, drug-related offenses clearly fit the bill.

Nowhere in the guidelines has Congress micromanaged sentencing policy (with ever-harsher results) as much as it has with child pornography. The commission has amended the child pornography guidelines nine times, primarily at the behest of Congress, making them more severe each time. Congress’ intervention reached a crescendo with the PROTECT Act of 2003, where, for the first time since the commission was created, Congress directly amended the guidelines. Not only did Congress ignore the process for creating guideline amendments established by the Sentencing Reform Act, it passed the legislation “without any consideration of the views of the Judiciary.”19

Congressional involvement has only increased in recent years. Just last week, a series of amendments became effective, the first three of which are based on specific congressional directives. As an example, the Fair Sentencing Act of 2010, though widely praised for eliminating the five-year mandatory minimum for possessing crack cocaine, also directed the commission to promulgate new specific offense characteristics and take account of other aggravating and mitigating factors in drug trafficking cases. Although denominated as mere instructions to the commission, the legislation is so specific that Congress might as well have simply amended the guidelines. Notably, the commission did not even try to justify these amendments based on empirical data or its own expertise—for each amendment tied to a congressional directive, the commission tersely noted that the amendment responds to a congressional order.20

To illustrate the harshness of congressionally directed guideline adjustments, let’s consider an individual convicted of health care fraud resulting in a $9 million loss and 50 victims. Taking sophisticated means into account and ignoring mitigating factors, the guideline range for a first-time offender under the original (1987) guidelines would have been 30-37 months; the range under the 2011 guidelines is 188-235 months, a more than 600 percent increase.

Changes to the child pornography guidelines are even more drastic. As an example, we compared the recommended range in United States v. Dorvee, 616 F.3d 174 (2d Cir. 2010), with the 1987 guidelines. Ignoring mitigating factors, the 1987 guideline range was 18-24 months; the 2011 guideline range was 30 years to life, a 20-fold increase. However repulsive the offense conduct, this sort of increase is extraordinary.

No Respect

As a result of these steep increases, increasing numbers of sentencing courts have refused to follow the guidelines. As Judge Saris’ testimony indicates, there has been a marked increase in non-government-sponsored, below-range sentences since Booker. For instance, the rate of drug trafficking offenses sentenced below the guideline range (without government support) has almost quadrupled, jumping from 4.7 percent after Congress passed the PROTECT Act to 17.3 percent since the Court issued its decision in Gall. Similarly, the rate of non-government-sponsored, below-range sentences for fraud and child pornography offenses increased during the same period from 6.2 percent to 22.3 percent (for fraud) and from 12.7 percent to a whopping 42.4 percent (for child pornography).

Courts have been openly critical of sentencing ranges for these types of offenses. The fraud guidelines, which Eastern District Judge Frederic Block has called “a black stain on common sense,”21 have been increasingly ignored in high-loss white-collar crime cases. For instance, in sentencing a white-collar fraud defendant with a guideline range of life to only 42 months, Southern District Judge Jed Rakoff lamented “the utter travesty of justice that sometimes results from the guidelines’ fetish with abstract arithmetic.”22 More recently, Eastern District Judge John Gleeson rejected the fraud guidelines in issuing a sentence of five to 12½ years below the low end of the recommended range—stating that the guidelines can produce sentences “[un]moored to fairness.”23

Likewise for the drug guidelines. In a recent case, Judge Nancy Gertner in Massachusetts sentenced a street-level crack cocaine dealer to five years, more than 10 years lower than the bottom of the guideline range. She rejected the crack-powder disparity and varied the sentence based on the defendant’s minimal culpability, finding that the guideline adjustments for role “hardly offset the substantial impact of quantity in the other direction.”24

Since Kimbrough, the federal judiciary has engaged in a full-scale attack on the child pornography guidelines, culminating in the Second Circuit’s remarkable rejection, in Dorvee, of a within-range sentence as substantively unreasonable.25 Justin Dorvee pled guilty to one count of distributing child pornography. His sentencing range was 262 to 327 months, which was lowered to the 20-year statutory maximum pursuant to §5G1.1(a). He was sentenced to 20 years.

The Second Circuit traced the history of §2G2.2, explaining that this guideline is “fundamentally different from most” because it is not based on the commission’s expertise or “data about past sentencing practices.”26 The court overturned the sentence and concluded by encouraging district courts to “take seriously the broad discretion they possess” in the post-Booker world.27 (On remand, the district court re-sentenced Dorvee to 121 months.28)

Courts are increasingly heeding the Second Circuit’s advice. Many now turn to other factors to better comply with the statutory mandate, in §3553(a), that a sentence be “sufficient, but not greater than necessary.” These factors include predicted recidivism rates, collateral consequences, role in the offense, lack of personal gain, cooperation without a government motion, and others that are not adequately reflected in many of the current guidelines.


Amending the guidelines to reflect the commission’s reasoned judgment and empirical analysis is the most logical way to reign in a sentencing system gone awry. Guidelines where quantities often drive the outcome can result in disproportionately harsh punishment. And as a result of years of congressional directives to the commission, the judiciary is rightfully pushing back from what it perceives as unjustifiably severe sentences. Even the Justice Department has acknowledged that Congress must address this problem because some of the guidelines have “lost the backing of a large part of the judiciary.”29

Judge Saris’ proposals try to balance the conflicting goals of sentencing discretion and meaningful appellate review. Though greater clarity may be needed at the appellate level, the commission, with Congress’ support, must fix the real problem. The commission needs to once again perform its traditional role of promulgating guidelines grounded in empirical data and experience. To the extent a guideline is based on a congressional directive, Congress must pass new legislation, or at least provide an explanation for how the current guidelines reflect the goals of §3553(a). This may well be politically unrealistic; but until Congress reconsiders its ever-increasing role in pushing federal sentences ever higher, courts will continue to impose below-guideline sentences in order to achieve greater fairness in sentencing.

Alan Vinegrad, former U.S. Attorney for the Eastern District of New York, is a partner at Covington & Burling. Jason Levine is an associate at the firm.


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

2. Hearing Before the Subcomm. on Crime, Terrorism, and Homeland Security Committee of the H. Comm. on the Judiciary, 157 Cong. Rec. D1086, webcast at ~14:30-16 min. (daily ed. Oct. 12, 2011), available at http://judiciary.house.gov/hearings/hear_10122011.html [hereinafter "Subcomm. Hearing"].

3. Id. at ~16:50 min.

4. Written Testimony of William G. Otis, Adjunct Professor of Law, Georgetown University Law Center, Subcomm. Hearing, at 6-7.

5. Written Testimony of Matthew S. Miner, partner, White & Case, Subcomm. Hearing, at 4.

6. Written Testimony of Judge Patti B. Saris, Chair, U.S. Sentencing Commission, Subcomm. Hearing, at 12.

7. See id. at 55; Subcomm. Hearing, webcast at ~1:40:30 min.

8. 551 U.S. 338 (2007).

9. 552 U.S. 38 (2007).

10. 552 U.S. 85, 109 (2007) (citation omitted).

11. 555 U.S. 261 (2009).

12. Frank O. Bowman III, “The Failure of the Federal Sentencing Guidelines: A Structural Analysis,” 105 Colum. L. Rev. 1315, 1319-20 (2005).

13. 18 U.S.C. §3553(a).

14. U.S. Sentencing Commission, “Mandatory Minimum Penalties in the Federal Criminal Justice System,” at 54 (October 2011).

15. Kimbrough, 552 U.S. at 96.

16. See United States v. Emmenegger, 329 F.Supp.2d 416, 428 (S.D.N.Y. 2004) (“amount of loss” for fraud offenses is “a kind of accident” and therefore “a relatively weak indicator of [] moral seriousness”).

17. U.S. Sentencing Guidelines Manual, App. C, Amend. 653 (effective Nov. 1, 2003).

18. U.S. Sentencing Guidelines Manual, App. C, Amend. 617 (effective Nov. 1, 2001).

19. See 2003 Year-End Report on the Federal Judiciary by Chief Justice William Rehnquist, 16 Fed. Sent’g Rep. 143, 143 (2003).

20. See U.S. Sentencing Guidelines Manual, App. C, Amend. 749-51 (effective Nov. 1, 2011).

21. United States v. Parris, 573 F.Supp.2d 744, 754 (E.D.N.Y. 2008).

22. United States v. Adelson, 441 F.Supp.2d 506, 512 (S.D.N.Y. 2006), aff’d 237 Fed App’x 713 (2d Cir. 2008).

23. United States v. Ovid, 2010 WL 3940724, at *1-2 (EDNY Oct. 1, 2010).

24. United States v. Whigham, 754 F.Supp.2d 239, 245 (D. Mass. 2010).

25. United States v. Dorvee, 616 F.3d 174, 176 (2d Cir. 2010).

26. Id. at 184.

27. Id. at 188.

28. Minute Order, United States v. Dorvee, 1:08-cr-514 (NDNY March 25, 2011).

29. Letter from Jonathan J. Wroblewski, Director, Office of Policy and Legislation, Department of Justice, to William K. Sessions III, Chair, U.S. Sentencing Commission, at 3 (June 28, 2010) (discussing guidelines for child pornography and fraud).