United States v. Burns

500 F.3d 756 (2007)

UNITED STATES of America, Appellant,
v.
Travis Ray BURNS, Appellee.

Nos. 04-2901, 04-2933.

United States Court of Appeals, Eighth Circuit.

Submitted: September 27, 2006. Filed: August 27, 2007.

Counsel who presented argument on behalf of the appellant was Robert L. Teig, AUSA, Cedar Rapids, IA. Also appearing on appellant's brief was Shawn S. Wehd, AUSA, Sioux City, IA.

Counsel who presented argument on behalf of the appellee was Richard S. Rhinehart of Sioux City, IA.

Before LOKEN, Chief Judge, BRIGHT, WOLLMAN, ARNOLD, MURPHY, BYE, RILEY, MELLOY, SMITH, COLLOTON, GRUENDER, and BENTON, Circuit Judges, En Banc.

WOLLMAN, Circuit Judge.

The United States appealed as excessive the downward departure granted by the district court for Burns's substantial assistance. Burns cross-appealed, arguing that the district court erred by selecting an incorrect guidelines range as a departure starting point. After a divided panel of this court affirmed, United States v. Burns, 438 F.3d 826 (8th Cir.2006), we granted the government's petition for a rehearing en banc and vacated the panel opinion. We now reverse the district court's departure and affirm its starting point.

I.

Burns was indicted on one count of conspiracy to manufacture and distribute fifty grams or more of methamphetamine in violation of 21 U.S.C. §§ 841(a)(1), 841(b)(1)(A), and 846. The government notified Burns of its intention to seek enhanced penalties under 21 U.S.C. § 851 that would subject him to a mandatory life sentence. In the absence of the § 851 enhancement, Burns faced a sentencing range of 188 to 235 months. Burns cooperated with the government and the government, in return, moved for a downward departure under United States Sentencing Guidelines Manual (U.S.S.G.) § 5K1.1 and 18 U.S.C. § 3553(e).

At sentencing, the government told the court that Burns had immediately admitted his involvement and had been debriefed on two separate occasions, providing detailed information on several groups involved in manufacturing methamphetamine. The court was also made aware that Burns had testified on two occasions before the grand jury and had provided information pertaining to several ongoing investigations. His cooperation assisted the government in establishing drug-quantity evidence against one defendant, Brad Messerly. In addition, Burns served as a key witness against a second defendant, Victor DeFoe, thereby enabling the government to obtain an indictment which led to a guilty plea. The government characterized Burns's cooperation as timely, truthful, and complete and recommended a 15 percent departure from a 360-month sentence.

The district court rejected the government's recommendation and departed 60 percent to 144 months. The court summarized its consideration of the § 5K1.1 factors as follows:

I'm going to use 360 months as a starting point. In this case I specifically find under the 5K1.1 factors, factor number 5, the timeliness of the defendant's assistance, the defendant was exceptionally timely in this case. My understanding is he started cooperating as soon as he was arrested. To me that's exceptional timeliness. While some defendants start that early, virtually no defendants start earlier than that. . . .

So I find that his timeliness was exceptional and apparently started cooperating before he was advised of the impact of the United States Sentencing Guidelines, before he knew anything about how the guidelines might affect his sentence, before he exercised his Sixth Amendment right to have counsel present. So I think in this case the fifth factor weighs very heavily in favor of the defendant. . . .

Number 4 does not apply, any injury suffered or danger of risk because I haven't heard anything about that.

Number 3, the nature and extent of the defendant's assistance, in this case based on the representations of the assistant U.S. attorney, I find that the defendant provided every single bit of information he knew, so you couldn't—the extent of the defendant's assistance could not be greater in the sense that he provided all of the information he knew.

Now, it's true that some defendants have greater information which leads to indictments of more people. But I don't think that's necessarily the test. I think the test is did the defendant provide substantial assistance on everything he knew, and in this case he did. So the defendant scores very highly on the third prong.

Defendant scores very highly on the second prong, truthfulness, completeness, and reliability of the information. There's no information that the defendant's substantial assistance was anything but a hundred percent complete, a hundred percent truthful, and a hundred percent reliable. So Mr. Burns scores very highly on the second prong.

[T]he first prong is the Court's evaluation of significance and usefulness of the defendant's usefulness taking into consideration the government's evaluation of the assistance rendered. Here the government has indicated that the defendant testified twice in front of the grand jury, that he established the drug quantity on one defendant and led to the indictment and guilty plea of another defendant. I find that that was both very significant and very useful.

Now for some reason which the government refuses to disclose, they only recommend 15 percent, but they won't tell me why they only recommend 15 percent. And the government refuses to indicate how any one of the five 5K1.1 factors affect the 15 percent recommendation.

So while I do take into consideration the government's evaluation of the significance and usefulness, it's hard to put any weight on the 15 percent recommendation because the government refuses to disclose how they arrive at that recommendation.

And looking back on the other sentencings that I've had, that recommendation is in my view substantially lower than other recommendations the government has made for similarly situated defendants.

Having said all that, I have the independent right under 5K1.1 to evaluate the substantial assistance based on the 5K1 factors as I see it.

Having taken into consideration the fact that the defendant scores very, very highly on the second factor, the third factor, and the fifth factor, I'm going to reduce the defendant's sentence substantially beyond what the government recommends in this case. That ought to come as no surprise to the government because I have a ten-year history of doing that because I just evaluate the five factors differently than—than how the government does, and the government refuses to disclose how they do it to me.

Sent. Tr. at 12-15.

On appeal, the government argues that the district court ignored its recommendation for a 15 percent departure and that the court's 60 percent (ten offense level) departure was excessive in light of Burns's assistance. Burns cross-appeals, arguing that the district court should have departed from the 188 to 235 month guidelines range instead of from the 360-month presumptive life sentence.

II.

When determining a sentence, the district court must first ascertain the governing guidelines range, then consider any permissible departures within the guidelines's structure, and finally, post-Booker, decide whether a non-guidelines sentence is more appropriate under the circumstances. See United States v. Haack, 403 F.3d 997, 1002-03 (8th Cir.), cert. denied, 546 U.S. 913, 126 S. Ct. 276, 163 L. Ed. 2d 246 (2005). This case comes before us on appeal of the district court's guidelines departure and not from a post-Booker variance. We apply an abuse of discretion standard when we review for reasonableness the extent of a district court's sentence reduction below the sentencing guidelines or beneath the mandatory minimum sentence. See, e.g., United States v. Dalton, 404 F.3d 1029, 1032 (8th Cir.2005); United States v. Saenz, 428 F.3d 1159, 1161-62 (8th Cir.2005); United States v. Mashek, 406 F.3d 1012, 1017 (8th Cir.2005).

A § 3553(e) motion permits the court to sentence below a statutory minimum and a motion under § 5K1.1 authorizes the sentencing court to depart below the applicable advisory guideline range in determining the advisory guideline sentence. See United States v. Williams, 474 F.3d 1130, 1131 (8th Cir.2007) (citing Melendez v. United States, 518 U.S. 120, 128-29, 116 S. Ct. 2057, 135 L. Ed. 2d 427 (1996)). Where, as here, a district court imposes a sentence below the statutory minimum in response to a § 3553(e) motion, the departure must be based "exclusively on assistance-related considerations." Id. at 1130-31. Although § 5K1.1 provides assistance-related factors that a district court may consider, it is not an exhaustive list. Dalton, 404 F.3d at 1033. The § 5K1.1(a) factors include:

(1) the court's evaluation of the significance and usefulness of the defendant's assistance, taking into consideration the government's evaluation of the assistance rendered;

(2) the truthfulness, completeness, and reliability of any information or testimony provided by the defendant;

(3) the nature and extent of the defendant's assistance;

(4) any injury suffered, or any danger or risk of injury to the defendant or his family resulting from his assistance;

(5) the timeliness of the defendant's assistance.

U.S.S.G. § 5K1.1(a).

A district court abuses its discretion by issuing a ruling that fails to consider a relevant factor that should have received significant weight, gives significant weight to an improper or irrelevant factor, or considers only appropriate factors but nevertheless commits a clear error of judgment by arriving at a sentence that lies outside the limited range of choice dictated by the facts of the case. Haack, 403 F.3d at 1004.

We noted in Saenz that "[d]epartures under § 5K1.1 and reductions under § 3553(e) should not be untethered from the structure of the advisory guidelines." Saenz, 428 F.3d at 1162; see also United States v. Cammisano, 917 F.2d 1057, 1064 n. 5 (8th Cir.1990) (stating that to the extent possible, the extent of departure should be linked to the structure of the guidelines). Because departures and reductions based on § 5K1.1 and § 3553(e) appear contextually within a scheme designed "to reduce unwarranted sentence disparities among similar defendants," Saenz, 428 F.3d at 1162, and because the Sentencing Commission's sentence adjustments for mitigating or aggravating factors fall close to the recommended guidelines, "[a]n extraordinary reduction must be supported by extraordinary circumstances." See id. (alteration in original) (quoting Dalton, 404 F.3d at 1032); see also United States v. Maloney, 466 F.3d 663, 668 (8th Cir.2006) ("`[A]bsent exceptional facts,' imposition of a sentence that is `dramatically lower than that recommended by the guidelines is an abuse of the district court's discretion.'" (alteration in original) (quoting United States v. Goody, 442 F.3d 1132, 1134 (8th Cir.2006), petition for cert. filed (U.S. Aug. 14, 2006) (No. 06-6079))). We note that under the Seventh Circuit's measure of reasonableness, "the standard is whether the district court's explanation is sufficiently proportional to the extent of the variance from the guidelines range; in other words, `the greater the deviation, the more compelling the district court's explanation must be.'" United States v. Repking, 467 F.3d 1091, 1095 (7th Cir.2006) (per curiam) (quoting United States v. Wallace, 458 F.3d 606, 613 (7th Cir.2006), petition for cert. filed (U.S. Nov. 13, 2006) (No. 06-7779)). Likewise, we have observed while evaluating the departure from a presumptively reasonable guidelines range that "the farther the district court varies from the presumptively reasonable guidelines range, the more compelling the justification . . . must be."^[1]United States v. McMannus, 436 F.3d 871, 874 (8th Cir.2006); see also Maloney, 466 F.3d at 668; United States v. Bryant, 446 F.3d 1317, 1319 (8th Cir.2006); United States v. Gonzalez-Alvarado, 477 F.3d 648, 650 (8th Cir.2007).^[2] In sum, our extraordinary reduction/extraordinary circumstances formulation requires circumstances of a strength proportional to the extent of the deviation from reductions envisioned by the guidelines's structure. As discussed below, under this proportionality standard, Burns's ten-level and 60 percent reduction exceeds the bounds of reasonableness.

We are mindful that the appropriate degree of sentencing reduction cannot be calculated with mathematical precision and that there is a range of reasonableness available to the district court in any given case. Saenz, 428 F.3d at 1164-65. It may be that we have placed too great an emphasis on numerical or percentage lines in conducting the reasonableness inquiry with respect to sentencing reductions. Cf. United States v. Jensen, 493 F.3d 997, 1000-01 (8th Cir.2007) (explaining why guidelines range levels may be a metric superior to percentages when establishing a life sentence departure point); Wallace, 458 F.3d at 613 (explaining the Seventh Circuit's reluctance to "distill the reasonableness inquiry into a numbers game"); Maloney, 466 F.3d at 668 (same). It may also be that the use of the term "extraordinary" suggests a false dichotomy. The term as applied to reductions should not be read to suggest a true dichotomy in which the location of an imaginary line demarcating "ordinary" from "extraordinary" may be divined by a statistical inquiry.^[3] Rather than representing a term of art with unique legal significance, the "extraordinary" label more accurately serves as a convenient characterization of departures that we have considered particularly large relative to the two to four offense level adjustments generally envisioned by the structure of the sentencing guidelines for mitigating or aggravating circumstances.^[4]Saenz, 428 F.3d at 1162 ("[T]he Sentencing Commission has concluded that most adjustments for aggravating or mitigating circumstances should be in the amount of two, three, or four offense levels.").

In conducting the proportionality analysis, consideration of the offense levels traversed by the departure may be the method most "in keeping with our assigned role to further the objectives of the Sentencing Reform Act, because the guideline system established by the Act was designed to adjust sentences incrementally by offense level, rather than by percentages." Maloney, 466 F.3d at 668 (finding unreasonable a variance amounting to a seven-offense-level reduction from a life sentence); see also Wallace, 458 F.3d at 613 ("The percentage reduction will always seem larger if the overall number is a small one: 24 months less than a possible sentence of 25 months would be a 96% reduction; 24 months less than a possible sentence of 240 months would be a 10% reduction."). Nevertheless, evaluation by percentage may still be useful at times. Maloney, 466 F.3d at 668. Regardless of the method used, because Burns is a criminal history category IV offender, the departure from 360 to 144 months effectively constitutes a ten-level and 60 percent reduction that must be considered a substantial deviation. See, e.g., Jensen, 493 F.3d at 1001-02 (citing recent cases and their associated extraordinary reductions in terms of guideline levels); United States v. Meyer, 452 F.3d 998, 1001 (8th Cir.2006) (citing numerous recent Eighth Circuit decisions in which departures or variances above 50 percent were considered extraordinary), petition for cert. filed (U.S. Nov. 27, 2006) (No. 06-8085); Dalton, 404 F.3d at 1033 (finding a 75 percent (or twelve level) departure extraordinary); see also Saenz, 428 F.3d at 1162 (holding that a 68 percent (or eleven level) departure was unreasonable and noting that a 50 percent (or eight level) departure would be extraordinary); cf. United States v. Nelson, 491 F.3d 344, 347 (7th Cir.2007) (endorsing district judge's comment that a 40% reduction for substantial assistance was "extremely generous" and a "rare beast," as a 20% reduction was "more common based on sentences that he and other judges have imposed.").

Extraordinary circumstances are infrequently found and involve assistance going well beyond that provided by other defendants. When we have found the presence of circumstances with a strength proportional to a departure of the magnitude granted here, the assistance was of a quality that could not reasonably be characterized as anything but extraordinary. Cf. United States v. Pizano, 403 F.3d 991, 995-96 (8th Cir.2005) (providing timely and truthful cooperation, serving as a "key witness" against two co-conspirators, giving testimony that could be instrumental in seizing assets from a money laundering scheme, submitting himself to debriefings and providing grand jury testimony regarding both a close family member and a "major figure" in the conspiracy, and putting himself and his family at risk of harm from "dangerous people" by testifying). Determinations of the value of the assistance for any given circumstance must be viewed in light of the entire continuum of defendant behavior associated with that circumstance. See United States v. Coyle, 429 F.3d 1192, 1193 (8th Cir.2005). Requiring more qualitatively impressive substantial assistance to justify progressively larger departures furthers the goal of reducing unjustified sentencing disparities and yet recognizes that situations exist at the tapering edge of the assistance bell-curve that justify departures that substantially exceed the Sentencing Guidelines's range.

III.

The district court informed its departure by considering only the § 5K1.1(a) factors. We look, then, first at whether the court abused its discretion as it considered and weighed the circumstances of Burns's assistance and whether the district court's 60 percent (ten level) departure was reasonable given the circumstances present. We then address Burns's contention that the district court departed from the incorrect guidelines range.

A.

The district court cites Burns's immediate and unceasing cooperation from the moment of his arrest, the truthfulness and completeness of the information he shared, and his testifying twice before a grand jury as constituting "very, very" strong showings on three of the five § 5K1.1(a) factors and as warranting an extraordinary 60 percent departure. With all due respect to the court's reasoning, we conclude that the court applied improper tests to determine the weight of the circumstances for each of the three factors relied upon.

First, the district court improperly analyzed the timeliness of Burns's assistance under the fifth § 5K1.1(a) factor. The court found the timeliness and completeness of Burns's cooperation exceptional because few defendants cooperated fully in a more timely fashion than did Burns. Although few defendants may participate earlier than did Burns, if a sufficient number participate in a similarly timely manner, Burns's participation could not be viewed as extraordinary.^[5] The court's premise, therefore, excluded Burns from the ranks of the most extraordinarily timely defendants, while it remained silent regarding how his individual timeliness compares to that of the remainder of cooperating defendants, resulting in an overvaluation of his timeliness. Furthermore, we have previously reversed extraordinary departures under conditions of comparable timeliness where such timeliness did not yield a discernibly beneficial result. See Saenz, 428 F.3d at 1162-63 (noting that the timeliness of the defendant's full cooperation did not, in and of itself, compel anyone else to plead guilty).

For similar reasons, we cannot agree with the district court that the truthfulness and completeness of Burns's assistance constitutes a circumstance warranting a 60 percent departure under § 5K1.1(a)(2). The safety-valve provision of § 5C1.2(a)(5) of the Sentencing Guidelines provides a sentencing reduction for a defendant who, by the time of sentencing, "has truthfully provided to the Government all information and evidence the defendant has concerning the offense or offenses. . . ." U.S.S.G. § 5C1.2(a)(5). This guidelines section, and the underlying statute, 18 U.S.C. § 3553(f), permit relief from mandatory minimum sentences so long as the defendant provides the government all information that he has, regardless of its usefulness. Given the frequency with which defendants provide honest and complete assistance in compliance with § 5C1.2(a)(5) or with the hope that the government will enable an increased departure by making a § 5K1.1 or § 3553(e) motion, Burns's honesty and the completeness of his assistance can hardly be viewed as so exceptional as to merit the reduction he received.^[6] We are skeptical whether complete truthfulness could ever warrant a departure of this magnitude in a substantial-assistance context — even where the defendant could not have offered more complete or more honest assistance. See Saenz, 428 F.3d at 1163 ("we respectfully disagree with the . . . view that a strong showing in [timeliness and completeness] makes reasonable an extraordinary departure or reduction of more than 50 percent, or the equivalent of more than about eight offense levels for this defendant, without regard to the nature and extent of the defendant's assistance, the significance and usefulness of the assistance, or any danger or risk of injury suffered by the cooperating defendant."). Accordingly, although Burns's honesty and completeness might help to justify a more favorable departure than that recommended by the government, the district court abused its discretion by according that factor the undue weight that it did.

The district court also supported its departure on the basis of an improper analysis of the "nature and extent" factor — § 5K1.1(a)(3). It applied a test that compared Burns's assistance to the assistance it speculatively believed the defendant capable of providing. Such a test invites sentencing disparities in contravention to the Sentencing Guidelines's stated "goal [of] `reduc[ing] unjustified disparities[, thereby] reach[ing] toward the evenhandedness and neutrality that are the distinguishing marks of any principled system of justice.'" United States v. Petersen, 276 F.3d 432, 439 (8th Cir.2002) (quoting Koon v. United States, 518 U.S. 81, 113, 116 S. Ct. 2035, 135 L. Ed. 2d 392 (1996)). Under the district court's test, two defendants providing objectively identical assistance might well receive different departures based entirely on the respective judge's speculative conjectures. Just as "the proper analysis under § 5K1.1(a)(1) focuses on the actual significance and usefulness of the assistance, regardless of the defendant's desire, effort, or `fault[,]'" Saenz, 428 F.3d at 1163-64, the proper analysis under § 5K1.1(a)(3) focuses on the nature and extent of the actual, discrete, and specific activities constituting the assistance, regardless of the opportunities available or the defendant's willingness and ability to provide more. Cf. United States v. Mills, 329 F.3d 24, 34 (1st Cir.2003) (upholding the district court's recognition of only the defendant's activities in focusing on the nature and extent of the defendant's assistance). Among the § 5K1.1(a) factors, the "nature and extent" and "significance and usefulness" considerations are particularly noteworthy for their inclusion of a broad spectrum of conduct. See U.S.S.G. § 5K1.1, cmt. background (isolating the two factors based on their breadth). Accordingly, even though Burns appeared before two grand juries, any departure granted to him premised on this factor must allow room for more substantial and extensive assistance. See Saenz, 428 F.3d at 1163 (implying that any combination of listed activities set forth as examples would represent more substantial and extensive assistance than the cooperation present). While the nature and extent of Burns's assistance was not inconsiderable, it did not qualitatively justify the reduction awarded.

We conclude, then, that the departure exceeded that which was justified by the circumstances.^[7] The district court did not merely reduce the sentence by, say, 20 or 30 percent, but by 60 percent. This equated to a reduction of not, say, three to five offense levels, but ten levels. Given the sizeable qualitative difference between the totality of the assistance provided by Burns and that provided by Pizano, to uphold a reduction so far removed from the presumptively reasonable guidelines range would frustrate our goal of preventing departures from becoming untethered from the structure of the guidelines and would not reflect any meaningful sense of proportionality. As was the case in Coyle, there is also here "a good deal of room between the government's modest recommendation and the district court's generous departure to recognize this defendant's assistance without at the same time skewing the degree of reduction that must be granted to future defendants whose performance on the continuum of substantial assistance deserves more credit than [Burns's]." Coyle, 429 F.3d at 1194.

B.

Finally, we reject Burns's contention on cross-appeal that the district court departed from the wrong guidelines range. Even though Burns acknowledges that he was subject to a mandatory life sentence, he contends that the government's filing of a substantial assistance motion "waived" the mandatory life sentence, leaving the district court with a sentencing range of 188 to 235 months from which to depart. We do not agree. As the panel opinion held, the government's substantial assistance motion did no more than permit the district court to sentence below the mandatory life sentence and did not negate the § 851 enhancement that made Burns eligible for a life sentence. Burns, 438 F.3d at 831. Accordingly, the district court properly departed from a presumptive life sentence of 360 months.^[8]

The sentence is vacated, and the case is remanded to the district court for resentencing consistent with the views set forth in this opinion.

BRIGHT, Circuit Judge, with whom BYE, Circuit Judge, joins, dissenting in part and concurring in part.

I concur in Judge Bye's well-reasoned and thoughtfully written dissent. Judge Bye has capably discredited the majority's statistically inaccurate and potentially harmful "extraordinary" standard for substantial assistance departures. Moreover, he has carefully described the level of assistance Burns provided to justify the district court's departure. I write separately to express my views because I rely on a more pragmatic approach than statistical comparisons. For these additional reasons, I dissent from the majority's decision to vacate the sentence. I concur in the majority's decisions as to the calculation of the Sentencing Guidelines starting point.^[9]

I.

The majority of my colleagues today grant more deference and discretion to prosecuting attorneys than to the district judges of this country. That is what is "extraordinary" in this case, not the departure afforded Burns by the district court. The majority opinion is itself "extraordinary" because it perpetuates a stringent and feckless standard for departures not based in the Federal Sentencing Guidelines. This court's efforts to micro-manage the district court's sentencing function rest on the slippery surface of a myth — the myth that circumscription of departures in this manner will reduce unwarranted sentencing disparity.

II.

"Extraordinary" is a meaningless word unless one describes what is "ordinary." Although the court rhetorically attempts to avoid describing "ordinary" by reference to statistics or as a bright line, the majority continues to give the impression that it views with nearly unsurmountable skepticism a departure in excess of fifty percent. The court cites only one case when it has affirmed a downward departure for substantial assistance in excess of fifty percent: United States v. Pizano, 403 F.3d 991 (8th Cir.2005). By contrast, in its opinion today alone, the court cites three cases in which it has reversed a substantial assistance departure above fifty percent. See United States v. Coyle, 429 F.3d 1192, 1193-94 (8th Cir.2005) (reversing seventy-three percent substantial assistance departure); United States v. Saenz, 428 F.3d 1159, 1162, 1165 (8th Cir.2005) (vacating sixty-five percent departure); United States v. Dalton, 404 F.3d 1029, 1031, 1034 (8th Cir.2005) (reversing a seventy-five percent substantial assistance departure); cf., United States v. Haack, 403 F.3d 997, 1005-06 (8th Cir.), cert. denied, 546 U.S. 913, 126 S. Ct. 276, 163 L. Ed. 2d 246 (2005) (vacating forty-three percent departure from 180 months to seventy-eight months). The court's "proportionality principle" in practice creates an unjustified normative baseline for departures at a two to four level adjustment and the impression of a ceiling at fifty percent. Any district judge aware of these cases would discern that our court only rarely affirms a sentence exceeding these measures. See, e.g., United States v. Jensen, 493 F.3d 997, 1001-02 (8th Cir.2007) (affirming six level downward departure); United States v. Christenson, 424 F.3d 852 (8th Cir.2005) (en banc) (affirming seventy-five percent downward departure).

Moreover, the baseline established by the majority relies on out of context statistics and percentages for what the Guidelines dictate should be a discretionary and individualized inquiry.^[10] Statistics should be used for illumination, not for support.^[11] The background commentary to the Guidelines' provision § 5K1.1 relating to substantial assistance departures sets forth that the "significance of assistance can involve a broad spectrum of conduct that must be evaluated by the court on an individual basis." U.S.S.G. § 5K1.1, cmt. background (emphasis added). The majority's comparative endeavor is therefore misguided. Furthermore, the background commentary of § 5K1.1 also states: "Latitude is, therefore, afforded the sentencing judge to reduce a sentence based upon variable relevant factors. . . ." Id. By this language, the Guidelines endorse a sentencing court's discretion to consider these factors.

The subtext of the majority's opinion is a disdain for downward departures.^[12] The majority's disfavor for downward departures is demonstrated by their recitation of the word "extraordinary." By their own words, the majority considers the reduction of a sentence beyond the government's recommendation without any data to be an exceptional event.

This disfavor for downward departures cannot be found in the Guidelines. There is no mention in the Guidelines of a substantial assistance departure being exceptional regardless of its magnitude. In fact, the Commission provides sentencing judges various avenues for both upward and downward departures when warranted. See, e.g., U.S.S.G. § 5K2.0(a)(3)-(4) (permitting departures for circumstances not already adequately taken into consideration and characteristics not ordinarily relevant). It is this court, not the Sentencing Guidelines, that have rarified such departures.

The majority holds that the exceptional nature of a departure over fifty percent is based on the Guidelines' two, three, or four offense level adjustments for mitigating or aggravating circumstances. That the Sentencing Commission specifically delineated in § 3B1.1 (aggravating role) and § 3B1.2 (mitigating role) offense level increases or decreases should discourage this court from circumscribing the degree of substantial assistance departures when the text of the Guidelines does not provide similar limitations. As also noted by Judge Bye, there is no warrant for imposing the levels proscribed by § 3B1.1-2 on Part K. The specific mention of the number of offense levels in one section demonstrates that the Commission knew how to proscribe adjustments, but declined to do so in favor of the discretion and "latitude" owed the sentencing judge for substantial assistance departures.

In this appeal, the majority rejects the district court's statement of reasons offered for its sentencing decision but also denigrates the percentage methodology offered by the prosecutor. It then adds to the recipes by citing United States v. Jensen, 493 F.3d 997 (8th Cir.2007), a case which advocates the use of Guideline ranges over statistics as a proper measure for sentencing departures. Need I remark, that sort of measure finds no support in the text of the Guidelines and was not argued by the parties in any of the briefs in this case. Its introduction into the majority's analysis only adds confusion.

III.

The Sentencing Guidelines operate as a sophisticated algorithmic-like puzzle for lawyers and judges. Within such a system, the only thing preventing sentencing judges from becoming automatons is their reason and judgment. The majority has effectively decimated the authority of a district judge to exercise that good judgment through its mathematical oversight.

The Guidelines do not demand of us such lifeless application of its advice. The Guidelines, as written, describe a human and interactive process between prosecutor, defense attorney, and sentencing judge for determining the mitigating effects substantial assistance should have upon a defendant's sentence. First, the Guidelines set forth that the government may initiate a substantial assistance motion. This is the first part of the conversation: the government speaks. The Guidelines then list a set of non-exclusive and non-exhaustive factors that the court might consider when making its evaluation. See U.S.S.G. § 5K1.1. In one of the factors, the Guidelines recommend the court consider the government's evaluation of the assistance rendered. U.S.S.G. § 5K1.1(a)(1). Similarly, in one application note, the Guidelines explain that the court should give "substantial weight" to the government's evaluation of the extent of the assistance given. Id. at cmt. n. 3. After the government speaks, the Guidelines give "latitude" to the district judge to exercise a reasoned judgment about the mitigating effects the assistance should have on a defendant's sentence. The Guidelines envision a dynamic process, by which the district court weighs information and carefully considers its impact on a person about to serve a devastating sentence.

But what happens, as here, when the government's recommendation appears arbitrary because it cannot or will not provide a basis or explanation for how it arrives at the recommendation it makes to the district court? Is the integrity of the Guidelines process tainted when the government stifles the conversation envisioned in § 5K1.1?^[13] What are the implications of the prosecutor and this appellate court circumventing the cooperative and discretionary process described in the Guidelines? What will become of advisory guidelines when an appellate court treats a list of non-exhaustive and nonexclusive factors as a checklist and a defendant like a point on a line graph? What are the dehumanizing effects of this court mechanizing the function of a district court — a court which meets the defendant and the prosecutor and can, as charged, neutrally evaluate the mitigating effects assistance should have on a sentence? The process of sentencing cannot and should not be undertaken by a computer!^[14]

IV.

Contrary to its stated attempts to prevent unwarranted disparity, the majority's decision actually promotes disparity. The majority explains that Burns's participation, although completely truthful and exceptionally timely, could not be viewed as extraordinary because some other defendants, none of whom participated in the instant drug conspiracy, also frequently or sometimes provided one hundred percent truthful information in a similarly timely manner.

This rationale nonsensically creates a moving target for departures not based in the Guidelines' text or envisioned by the Guidelines' dictate for an "individual" inquiry. See U.S.S.G. § 5K1.1, cmt. background. Moreover, it improperly dilutes the weight of the timeliness, completeness, truthfulness, and reliability factors. For example, imagine three defendants, convicted of the same offense, in the same sentencing range, but for separately committed crimes. Suppose that defendant number one assists the government in a timely manner and is one hundred percent truthful. He receives a thirty percent departure. The next day, defendant number two assists the government in the same timely manner and is also one hundred percent truthful. He will receive only a twenty percent departure because his timeliness and truthfulness are less exceptional by virtue of the prior participation of defendant number one. On the third day, defendant number three also participates in the same timely manner and is one hundred percent truthful. Again, under the majority's theory, defendant number three is likely only to receive a ten percent departure because his identical timeliness and truthfulness have become comparably less remarkable. The Sentencing Guidelines do not require such an absurd result. How can one be more than one hundred percent truthful? What kind of truthfulness would the majority contend is exceptional?

The majority also rejects Burns's departure because the district court factored Burns's capability of assistance when evaluating the nature and extent of his assistance under § 5K1.1(a)(3). This standard, as applied by the majority, operates to make substantial assistance departures too heavily favored for the biggest criminals. Those at the bottom of the crime totem pole, so to speak, are often ineligible for an "exceptional" departure because they have less information to sell. See United States v. Jones, 145 F.3d 959, 966-67 (8th Cir. 1998) (Bright, J., dissenting); see also Maria Limbert, Problems Associated with Prosecutorial Control Over Filing Substantial Assistance Motions and a Proposal for a Substantial Assistance Pre-Sentence Hearing, 27 J. Legis. 251, 258 (2001) ("Since drug organizations are now so large and diverse, one can be involved as an unloader, a seller, a mule, or a courier, and thus they are insulated and do not know who the principals are, so very often the smaller dealers have nothing that they can really offer the government.")

Moreover, "[e]ven among defendants with equal access to useful information, the availability of a substantial assistance departure may hinge primarily on the timing of their arrests and plea bargains. Those apprehended early in an investigation can provide assistance by informing on their associates, while the associates arrested later are likely to add nothing new to the authorities' knowledge." Gerald W. Heaney, The Reality of Guidelines Sentencing: No End To Disparity, 28 Am.Crim. L.Rev. 161, 199 (1991). As I have said before, our system of justice should not condone such topsy-turvy results.

V.

The majority repeatedly bases its "extraordinary" decision on reducing "unjustified sentencing disparities."^[15] First, as Judge Bye points out in his opinion, and as I have illustrated above, the court's opinion today may actually serve to increase unwarranted disparity, not lessen it. But more significantly, the court's justification sounds a hollow note: the notion that stiff adherence to the Sentencing Guidelines range reduces unwarranted sentencing disparities is a myth!

Long before a sentencing judge considers a government motion for a substantial assistance departure, a defendant's Guidelines sentence is rife with factors which will increase disparity. The Guidelines range assigned a criminal defendant invariably depends upon law enforcement officers, the prosecutor's charging decision, the defendant's decision to plead guilty, the probation officer's report, the quality of defense counsel,^[16] and numerous other factors. See Albert W. Alschuler, Disparity: The Normative and Empirical Failure of the Federal Guidelines, 58 Stan. L.Rev. 85, 111-13 (2005). By way of example, one study randomly selected forty-six probation officers and asked them to apply the relevant conduct guideline to four offenders who had participated in the same drug conspiracy. The officers' conclusions resulted in substantial variation. "For the least culpable of the offenders, the probation officers' calculations ranged from one to five years in prison." See id. at 111-12 (citing Pamela B. Lawrence & Paul J. Hofer, An Empirical Study of the Application of the Relevant Conduct Guideline 1B1.3, 4 Fed. Sent'g Rep. 330 (1992)).

Similarly, prosecutorial discretion significantly impacts the sentence a criminal offender will serve.^[17]See Heaney, supra, at 165 ("Although both the legislative and executive branches have routinely attributed sentencing disparity to the discretionary actions of individual judges, the lion's share of discretion in the sentencing process has been and continues to be exercised by the executive and legislative branches.") Approximately 74.9% of federal judges and 58.6% of probation officers think that the prosecutor has "the greatest influence on the final guideline sentence." See Limbert, supra, at 258 (quoting Linda Drazga Maxfield & John H. Kramer, Substantial Assistance: An Empirical Yardstick Gauging Equity in Current Federal Policy and Practice, 5 n. 11 (1998), at http://www.ussc.gov/publicat/5kreport.pdf (visited July 31, 2007)). The impact of prosecutorial discretion on a sentence is particularly acute in the context of substantial assistance departures, which can only be granted upon motion by the government and where government policies regarding cooperation vary widely by district. See Limbert, supra, at 259. This is particularly troublesome when prosecutorial discretion in the context of substantial assistance departures is virtually unreviewable. See id. at 260-61.

The enactment of the Sentencing Guidelines has not produced less disparity. In fact, it has only increased it. One author has divined the following illuminating statistics about the impact of the Sentencing Guidelines on sentencing disparity:

In the pre-Guidelines period, an offender could expect the jurisdiction in which his case arose to make a difference of about 5.24 months in his sentence. With the Guidelines in effect, the expected difference in sentence attributable to the jurisdiction in which a case arose increased to 12.94 months.

. . . [T]he court in which a case was heard accounted for 1.81% of the total variation in sentences in the pre-Guidelines period. It accounted for 5.64% of this variation in the post-Guidelines period. In temporal terms, the expected variation per case was 7.70 months before the Guidelines and 17.97 months after.

. . . Geographic disparity more than tripled after implementation of the Guidelines.

. . . .

. . . [a]lthough blacks constitute 48% of the offenders who appear to qualify for a mandatory firearms enhancement in drug cases, they constitute 64% of the offenders who receive it. . . . prosecutors seek "substantial assistance" departures for blacks and Latinos less often than for whites. . . .

. . . .

. . . The time served by men in federal prisons before the Guidelines exceeded that served by women by about nine months or 50%. In the years since the Guidelines were implemented, the gender gap has grown. The time served by men increased 96% after the Guidelines while that served by women increased 75%. Men now serve 51 months on average and women 28. The previous nine-month gender gap has grown to 23 months.

Alschuler, supra, at 101, 104-5 (footnotes omitted).

Numerous law professors, judges, and other commentators have observed that the Sentencing Guidelines simply do not create less disparity, but more. Thus, the majority's opinion rests on the slippery surface of a myth. The myth that by reducing the number and size of downward departures, disparity will improve. The means to eliminate or reduce disparity has thus far eluded the Sentencing Commission, sentencing judges, prosecutors, and apparently continues to escape the reason of this court of appeals today.^[18]

VI.

Thus far, my criticism has focused on the majority's opinion; it is not alone responsible for the outcome in this case. A prosecutor is an officer of the court, responsible for administering justice, but also for assisting the trial judge. The process envisioned by the Guidelines § 5K1.1, as described above in Section III., calls on the prosecutor to satisfy this obligation. When, as here, the government refuses or is unable to provide the court reasons for its recommendation, it hampers the district court's ability to evaluate the assistance rendered and disregards its duties to assist the court.

In the instant case, after describing the assistance Burns provided, the prosecutor recommended on behalf of the government a fifteen percent downward departure for substantial assistance. The following colloquy between the judge and prosecutor ensued:

THE COURT: How did the defendant's timeliness affect the 15 percent recommendation?

[ANSWER]: It was one of — those were factors that all went through in the mix. I guess the term would be in totality.

THE COURT: Are you in a position to tell me why, if the defendant was as timely as you indicate he was, you're only recommending 15 percent?

[ANSWER]: The overall substantial — or the overall cooperation he provided, the U.S. attorney decided that in totality these — under the 5K factors this was the percentage they would come up with.

THE COURT: Let's try and see if we could short-circuit this. Are you in a position to tell me how any of the specific 5K1.1(a)(1) through (5) factors affected the 15 percent recommendation?

[ANSWER]: I can't give you a percentage of what each one and how it was arrived at 15, no, but in looking at all of the facts with those factors, the U.S. attorney recommended 15 percent.

THE COURT: Well, for example, why — if the defendant was exceptionally timely in his cooperation, why isn't he getting a 25 percent or a 50 percent recommendation?

. . . .

[ANSWER]: Each is a case by case, but there are — I don't know how to put a percentage on it.

THE COURT: . . . Would you agree that except for those very rare situations where a defendant comes forward and cooperates before he's arrested that this defendant cooperated in about as timely a manner as one could? Is that a fair statement?

[ANSWER]: In this case with this defendant, as soon as he was arrested, he told everything; he was complete and truthful, yes. Those are the facts as I understand it.

THE COURT: And my question is why doesn't that justify a 20, 30, 40, 50, or 60 percent recommendation from your office?

[ANSWER]: Looking at that individual factor, I can't answer your question, but in overall, everything that the defendant did, the defendant in this case assisted with the indictment of one individual at this time and all these other facts that we presented.

THE COURT: Did the defendant provide you all the information he knew about in terms of substantial assistance?

[ANSWER]: From what I know from [the AUSA], yes.

THE COURT: Why doesn't that entitle the defendant to a greater recommendation than 15 percent?

[ANSWER]: I can't answer your question. I don't know how to answer that question, Judge.

This excerpt clearly shows that the government stifled the Guidelines process by its refusal or inability to assist the court in making its determination. The government's answers to the court's questions during sentencing were worthless! And, in its appeal before this court, the government asserts an astounding and misleading response in defense of its refusal to participate in the Guidelines process. The government states: "Subsequent to each inquiry of the count [sic], the government asserted the deliberative process privilege." Appellant's Br. at 4. Where? I have searched the sentencing transcript in vain for any mention of any such privilege.

The government, in its brief, also criticizes the district court for inquiring as to how it arrived at its recommendation of fifteen percent. See Appellant's Br. at 27-28. It calls the district court's questions "an attempt to infringe on the executive branch's deliberative process and the separation of powers doctrine."^[19] In support of its incredible assertion that it need not provide further explanation to the court, it cites United States v. Moeller, 383 F.3d 710 (8th Cir.2004). Moeller concerned the government's decision not to file a § 3553(e) motion, which permits the court to sentence a defendant below a statutory minimum. Here, however, the government recommended a § 5K1.1 departure. Although there is some indication that the government in Moeller may have been evasive when it explained its decision not to file a § 3553(e) motion, see id. at 712-13, Moeller does not stand for the proposition that "We looked at the factors" or "I don't know" are sufficient responses to the court once the government files a § 3553(e) motion, or in this case recommends a § 5K1.1 departure.

The government further argues that "the government's refusal to divulge further information was not a reason for the district court to disregard the government's recommendation." This criticism of the district court can only be described as hubris. The district court is not a rubber stamp for what may be an arbitrary whim of the United States Attorney. Failure to provide any assistance to the court in making its evaluation leaves the district court with little choice but to undergo its own evaluation of the assistance provided and reach its own conclusion. It is tremendously disingenuous of the government to chastise the district court's analysis when the government refuses to engage in its own. Perhaps if the government were less evasive to the trial court, the government would be more satisfied with that court's decisions.

In its appellate brief, the government at long last attempts to give a reason for its recommendation: "the modest nature of defendant's assistance." Appellant's Br. at 17. The government never used this descriptive term, "modest," during the sentencing hearing. To the contrary, during sentencing, the prosecutor described Burns's assistance as "consistent," "key," "truthful," "complete," and "timely." Never "modest." The government's late assertion that defendant's assistance was modest lacks support in the record.

I can divine several possible (non-exclusive) explanations for the prosecution's conduct in this case. One, the prosecutor truly didn't know how his superiors arrived at the recommendation. His superiors directed him to recommend fifteen percent and he declined to inquire further. Second, the prosecutor truly didn't know the basis for the recommendation because it was arbitrary.^[20] Another possibility is that the prosecutor truly believed that the government is not required to advise the court.^[21] The United States Attorney and the assistants view the government's "recommendation" to the court as a command, in part because of knowledge that this court will most likely reverse any departures that exceed the advice of the prosecutor.

The final explanation, and probably the likely one, is in my view the most egregious. The government knew its reasons for the recommendation, but withheld those reasons from the sentencing judge as a matter of strategy in order to sandbag the district judge's reasoned opinion and allow the prosecutor to call a "late foul" on appeal. The government believes that it can circumvent the sentencing court by refusing to assist the judge and then come to this court with criticism of the district judge's analysis.

I cannot and will not join the majority in support of such prosecutorial conduct. It is an affront to the judicial system and a conscientious district judge, not to mention treating the defendant fairly.

VII.

While Section VI. of this opinion critically analyzes the prosecution for its particular conduct in this case, I write this separate section to describe why, as a matter of law, the procedure utilized here must not continue.

The successful function of our judicial system relies largely on procedure. Both modern and ancient conceptions of justice are founded on the concept of due process, which is essentially procedure.

Sentencing is no exception. Our current federal sentencing system requires collaboration between the probation office, the prosecutor's office, law enforcement, defense counsel, and the court. These individuals operate together to effectuate a process: a process by which a reasonable and effective sentence may be determined. Although a judge makes the ultimate conclusion, a sentence should only be upheld where the judge follows the correct process. For example, if the trial judge fails to consider certain factors or considers factors that have been deemed impermissible, the process is tainted and the sentence will likely be reversed. In this way, our judicial system honors process and method — sometimes even more than the result.

The government in at least the Northern District of Iowa appears to be engaging in the routine practice of refusing to give reasons or a basis for its recommended departure.^[22] The by-product of the majority's affirmance today may serve to perpetuate the government's practice and interfere with the proper process by which we sentence criminal defendants.

The Guidelines state that the court will give "substantial weight" to the "government's evaluation of the extent of the defendant's assistance, particularly where the extent and value of the assistance are difficult to ascertain." U.S.S.G. § 5K1.1, cmt. n. 3. This note reaffirms that § 5K1.1 asks the government to assist (not direct) the court with an evaluation (not a conclusion). Yet the government refuses to provide such assistance to the court. Moreover federal prosecutors, both as officers of the court and administrators of justice, have a duty of candor and assistance to the court. Yet the government refuses to engage in either candor or assistance.

In Saenz, this court recognized that although the district court must give substantial weight to the government's recommended evaluation of the extent of assistance, it does not similarly owe such deference to the "valuation" afforded that assistance. See 428 F.3d at 1164. This court stated: "We are less persuaded that the court must give substantial weight to the government's valuation of the assistance, particularly where the government does not adequately explain its reasoning." Id. (emphasis in original). Similarly, in Haack, this court noted:

It is obvious that the sentencing judge was frustrated by the government's identical recommendations of ten percent departures in each of these three dissimilar cases. We expressed similar concerns to the Assistant United States Attorney who argued the cases. We had difficulty discerning how three such dissimilar cases could all result in the identical recommendation for departure. A recommendation by the government that does not adequately explain its reasoning is entitled to less weight, in the court's view, than a more fully explained recommendation.

403 F.3d at 1005 n. 2; see also Pizano, 403 F.3d at 996 ("serious consideration needs be given the government's recommendation, but that it is certainly not controlling"). Thus, this court's precedent and the text of the Guidelines belie the government's position that it need not explain, when asked, how it arrived at its recommendation.

Moreover, the government's refusal is problematic in large part because it is singularly unhelpful to the district court, this court, and the sentencing process. The government's silence creates a sentencing process through which defendants, their counsel, sentencing courts, and this court must traverse without any signposts or guides. Indeed, this court has commented on this very concern as it relates to our review:

While we recognize that a single United States Attorney brings to bear the broad perspective of one who has evaluated numerous cooperating defendants within a judicial district (indeed, at least during comparable tenure, a broader perspective than a single district judge in a multi-judge district), we have no assurance that United States Attorneys in different districts apply consistent methodologies for valuing substantial assistance and arriving at sentencing recommendations.

Saenz, 428 F.3d at 1164 (emphasis added). Similarly, the district court has no assurance that United States Attorneys have applied a consistent methodology — or a methodology at all. What good is the government's recommendation to the district court if the government cannot back it up? Moreover, without any guideposts or explanations, defense counsel has no idea how to reasonably advise the client of a potential assistance departure.

Finally, the government's silence impedes accountability and judicial review. How can the district court ever properly evaluate the government's recommendation when it is backed up by nothing? How can this court review the district court's consideration of the government's evaluation when from the record we cannot even discern on what basis the government made its recommendation? Because this court often reverses downward departures on appeal, there is little or no accountability for the government in this respect. The government can continue to provide unexplained recommendations, which the district court will then have difficulty assessing, the government will appeal, and this court will reverse. This cat and mouse game is a waste of our judicial resources; this upside-down process is certainly a strange one. It can perhaps be captured by the words of the musical Anything Goes:

The world has gone mad today And good's bad today, And black's white today, And day's night today, . . . . Anything goes.

Cole Porter, Anything Goes, in Anything Goes (1934).

Persons of intelligence, integrity, and good faith can disagree on the reasonableness of a sentence, the value or extent of assistance, or the degree of a departure. My colleagues and my disagreement as to the reasonableness of a sentence, and this court's disagreement with the district court's departure, should not effectively incentivize prosecutors to disengage from the sentencing process in the district court. Secrecy has no place in the courtroom. I believe it would be proper for the government on remand to make a new recommendation to the district court, providing the judge with a candid explanation of the methodology employed in reaching the recommendation. The district court should then consider that evaluation when re-sentencing Burns.

VIII.

Who should have the power to sentence in cases of this kind? The prosecutor, who has an unreasoned opinion for the appropriate sentence? This court, who has never met the defendant and has been given only a limited record by the government but often supports the position of the prosecutor?

We should recognize the discretion of district judges who give thoughtful consideration to a defendant's sentence, subject to review for abuse of discretion.

In the case before us, I would affirm the reasoned decision of the district court to grant Burns a substantial assistance departure and reduce his sentence to 144 months (twelve years). That is still a very long sentence.

I dissent.

BYE, Circuit Judge, with whom BRIGHT, Circuit Judge, joins, concurring in part and dissenting in part.

The majority holds the district court's sixty percent downward departure was excessive and unreasonable under the circumstances presented by this case. Because the district court appropriately exercised its sentencing discretion, I respectfully dissent from the majority's decision to vacate the sentence and remand for resentencing. I join the majority's opinion in other respects. I also join in Judge Bright's dissent.

I

We review the extent of a substantial assistance downward departure for an abuse of discretion. United States v. Coyle, 429 F.3d 1192, 1193 (8th Cir.2005). "[T]here is no bright line percentage or mathematical formula to determine when the extent of a substantial assistance motion becomes unreasonable." United States v. Pepper, 486 F.3d 408, 411 (8th Cir.2007) (Pepper II). In an effort to divine what is unreasonable, we have broadly stated "some proportionality must exist between the defendant's assistance and the extent of the departure; for example, an extraordinary departure must be supported by extraordinary circumstances." Id. (citing United States v. Saenz, 428 F.3d 1159, 1162-65 (8th Cir.2005) (Saenz I)). This unremarkable premise, however, offers scant assistance to district courts charged with the responsibility of imposing reasonable sentences. For proportionality to have meaning, we must engage in a substantive discussion of "extraordinary departures" and "extraordinary circumstances." Along the substantial-assistance-bell-curve described by the majority, we need to determine what reductions in sentences — ranging from 0 to 100 percent — are ordinary and which are extraordinary. After establishing what ordinary and extraordinary reductions are, we can begin to measure the quality and quantity of assistance necessary to obtain one or the other.

A

Despite cases eschewing a bright line rule, we have held, without explanation, that departures or variances in the range of fifty percent are extraordinary. See United States v. Meyer, 452 F.3d 998, 1001 (8th Cir.2006) (holding a fifty percent variance extraordinary); United States v. Bryant, 446 F.3d 1317, 1319-20 (8th Cir. 2006) (holding a fifty-seven percent variance extraordinary); United States v. Larrabee, 436 F.3d 890, 892-93 (8th Cir.2006) (holding a fifty-four percent variance extraordinary); United States v. Kendall, 446 F.3d 782, 785 (8th Cir.2006) (holding a fifty percent departure extraordinary); Saenz I, 428 F.3d at 1162 (noting a fifty percent departure would be an extraordinary sentence reduction); United States v. Dalton, 404 F.3d 1029, 1033-34 (8th Cir. 2005) (citing United States v. Enriquez, 205 F.3d 345, 348 (8th Cir.2000) (describing a fifty percent downward departure as extraordinary in light of the government's recommended twenty percent reduction)). This "fifty percent rule" debuted in Dalton, and quickly became, without benefit of critical analysis, the guiding principle used to evaluate the reasonableness of departures and variances. Despite no discernable resistance to its adoption, the rule resides as a squatter in our court's jurisprudence, owing its apparent legitimacy to having been oft repeated.

As ably recounted by Chief Judge Bennett^[23] in United States v. Saenz, 429 F. Supp. 2d 1081 (N.D.Iowa 2006) (Saenz II), the genesis for the rule lies in dicta found in Enriquez.

[T]he comment in Enriquez upon which the Circuit Court relied in Dalton and Saenz was not made in the context of a determination of whether a district court's downward departure for substantial assistance was "unreasonable" or "extraordinary." Instead, in Enriquez, the defendant argued "that he should be allowed to withdraw his plea because the government failed to keep an alleged promise to recommend an above-average sentence reduction for his wife." Enriquez, 205 F.3d at 348. The Circuit Court concluded that the district court had not erred in finding that there was no such promise. Id. Only then did the Circuit Court comment, apparently as an afterthought, that, "as a practical matter, Mrs. Enriquez actually did receive an extraordinary sentence reduction of 50 per cent., as opposed to the 20 per cent. expected." Id. I cannot believe that an off-hand statement, which is plainly dicta in the decision in which it appears, could possibly have been intended to establish the benchmark for what constitutes an "unreasonable" or "extraordinary" downward departure in the very different context of a reduction for substantial assistance. Certainly, there was no attempt in Enriquez (or indeed, in the subsequent decisions in Dalton and Saenz) to provide a reasoned basis for establishing a 50 percent reduction as the benchmark for "unreasonableness" of a reduction, whether the reduction is for substantial assistance or for any other reason. Moreover, to the extent that the court in Enriquez provided any basis for labeling a 50 percent reduction "extraordinary," the basis was that only a 20 percent reduction was "expected," and even then, there was no explanation of the basis on which only a 20 percent reduction was "expected." Id. Thus, the decision in Enriquez does not reasonably stand for the proposition that a 50 percent reduction in sentence for any reason, let alone for substantial assistance, is "extraordinary."

Id. at 1090-91.

Despite its dubious origins, the rule's simplicity and ease of application make it enticing. Our responsibility, however, is to find what is reasonable — not expedient — and, like Judge Bennett, I find no reasoned basis in Enriquez for concluding a fifty percent departure is ipso facto extraordinary. Indeed, Judge Bennett's well-reasoned discussion in Saenz II, convinces me a fifty percent departure for substantial assistance is not extraordinary. 429 F.Supp.2d at 1091-95. Rather, departures in the range of fifty percent represent the nationwide median afforded defendants providing substantial assistance, and are, therefore, quite ordinary. The U.S. Sentencing Commission (Commission), Special Post-Booker Coding Project, (data extracted February 22, 2006; table prepared February 23, 2006) (Special Post-Booker Coding Project), compared the percentage of decrease in sentences for substantial assistance — calculated from the minimum U.S. Sentencing Guidelines (Guidelines) range sentence — in 9,061 cases nationwide. Overall, the median decrease was 49.9 percent. A review of 5,754 drug trafficking cases revealed a nationwide median decrease in those cases of 45.8 percent.^[24] This data informs our present discussion by demonstrating, in concrete and reliable terms, that our reliance upon Enriquez and the "fifty percent rule" it spawned was misguided. We now know, "far from being `extraordinary,' a 50 percent reduction for substantial assistance very nearly approximates `the median' and, as such, is actually and necessarily `ordinary.'" Saenz II, 429 F.Supp.2d at 1092.

Today's decision, while citing cases which held departures in the range of fifty percent ipso facto extraordinary, offers another explanation for having vacating those sentences. The majority rejects "aggregate statistics" as a guide for determining what departures are extraordinary, arguing instead "the `extraordinary' label more accurately serves as a convenient characterization of departures that we have considered particularly large relative to the two to four offense level adjustments generally envisioned by the structure of the sentencing guidelines for mitigating or aggravating circumstances." This alternate method for gauging what is extraordinary was first alluded to in Saenz I, 428 F.3d 1159, where the court concluded

Departures under [Guidelines] § 5K1.1 and reductions under [18 U.S.C.] § 3553(e) should not be untethered from the structure of the advisory guidelines. They take place, rather, within the framework of an advisory guideline scheme designed to reduce unwarranted sentence disparities among similar defendants, and we are mindful that the Sentencing Commission has concluded that most adjustments for aggravating or mitigating circumstances should be in the amount of two, three, or four offense levels.

Id. at 1162 (internal citations omitted).

I seriously question whether our earlier decisions intended a definition of extraordinary tethered to adjustments for mitigating and aggravating circumstances. Nevertheless, assuming the court had such considerations in mind when it decided those cases, this alternate benchmark for ordinary or extraordinary should also be rejected.

As the Special Post-Booker Coding Project proves, the median sentence reduction nationwide is nearly fifty percent. In other words, along the bell curve representing possible departures, those in the range of fifty percent are ordinary, with extraordinary departures falling somewhere to the far left and far right of ordinary. By defining extraordinarily high departures as those exceeding two, three or four offense levels, the court arbitrarily moves ordinary to the left of the bell curve. Adjustments of two, three and four levels approximate sentencing reductions of twenty-four, thirty-four, and forty-two percent, respectively. Applying the majority's newly-minted test, extraordinary departures now fall on either side of twenty-four percent and forty-two percent, with departures exceeding forty-two percent displacing fifty percent departures as ipso facto extraordinary. By this reasoning, departures deemed extraordinarily low nationwide are ordinary in our court, and departures lower than the national median are extraordinarily high. This skewed view of what is ordinary is directly contradicted by the data from the Special Post-Booker Coding Project, but is consistent with our court's demonstrated tendency to closely scrutinize reductions in sentences while readily affirming increases. See United States v. McDonald, 461 F.3d 948, 960 (8th Cir.2006) (Bye, dissenting); United States v. Meyer, 452 F.3d 998, 1000 n. 3 (8th Cir.2006).

Additionally, the majority's reliance on adjustments for mitigating and aggravating circumstances as a benchmark for what is ordinary in § 5K1.1 departures is contradicted by the structure of the Guidelines. When considering adjustments for mitigating and aggravating conduct, a district court is called upon to determine whether an adjustment is factually supported. The degree to which conduct supporting the adjustment exists is not at issue. Instead, once the threshold fact finding is made, a defendant's base offense level is decreased or increased by a preset number of offense levels as dictated by the Commission.

Conversely, under § 5K1.1, the sentencing court is required to rate the value of a defendant's substantial assistance along a continuum of extraordinarily low, ordinary, and extraordinarily high, using a non-exclusive list of factors, which include the significance, usefulness, truthfulness, completeness, reliability, nature, extent, timeliness, etc., of a defendant's assistance. The district court then awards an appropriate reduction in sentence ranging from 0 to 100 percent. The evaluative process anticipated by § 5K1.1 does not lend itself to the same limitations which constrain district courts applying sections dealing with mitigating and aggravating circumstances. Were this not true, the Commission would have imposed a similar framework upon substantial assistance departures. It could easily have defined ordinary assistance as deserving reductions of between two and four offense levels, with extraordinary assistance garnering preset minimum and maximum reductions. Instead, the Commission, recognizing the broad range in the quality and quantity of possible substantial assistance, left it to district courts to evaluate the assistance and determine the appropriate reduction. By imposing the same construct on substantial assistance departures as applies to mitigating and aggravating adjustments, the majority restricts the discretion of district courts in a manner neither authorized nor intended by the Commission.

This court has often noted the need to reduce unwarranted sentencing disparities among similarly situated defendants as the overarching concern when vacating substantial assistance departures. See, e.g., Saenz I, 428 F.3d at 1164; see also 18 U.S.C. § 3553(a)(6) (requiring the sentencing court to consider "the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct"). Paradoxically, our continued allegiance to the fifty percent rule or today's rule defining ordinary as the equivalent of a two to four level adjustment, exacerbates the very problem it claims to prevent. Nationwide defendants providing ordinary substantial assistance receive departures ranging in the area of fifty percent. In our court the same defendants receive departures that are in comparison extraordinarily low.

B

As shown above, the Special Post-Booker Coding Project offers a meaningful basis for determining where ordinary and extraordinary fall along the curve of possible departures. Our next task is to define the quality and quantity of substantial assistance necessary to earn a given departure. I concur with the majority's observation that statistics are not as useful to this inquiry. Instead, we need to compare the circumstances offered in support of the departure under review to the circumstances presented in prior decisions. After comparing the assistance provided by Burns to the assistance provided by other defendants, I am satisfied the district court's sixty percent departure was reasonable.

My review of the district court's departure is guided by our decisions in United States v. Haack, 403 F.3d 997, 1002-03 (8th Cir.), cert. denied, 546 U.S. 913, 126 S. Ct. 276, 163 L. Ed. 2d 246 (2005), United States v. Pizano, 403 F.3d 991, 995 (8th Cir.2005), as well as Dalton, Saenz I, and Pepper II. In Haack, we reversed a fifty-seven percent downward departure, finding it unreasonable in light of the defendant's limited assistance. 403 F.3d at 998. There the defendant made incriminating statements when first arrested which helped officers obtain a search warrant. Id. at 999. For several months thereafter, however, the defendant refused to cooperate. When he finally did cooperate, his information was of limited use to the government and he was not available to provide useful grand jury testimony. Id. at 1005. We also expressed reservations about comments the district court made at sentencing suggesting the departure may have been based in part on the court's dissatisfaction with the sentencing guidelines. Id. at 1006.

In Dalton, we concluded the district court's seventy-five percent downward departure was unreasonable because the defendant 1) provided only corroborative testimony before the grand jury, 2) did not implicate a large number of people, 3) did not serve as a primary government witness against anyone, and 4) absconded from custody while on pretrial release. 404 F.3d at 1033.

In Saenz I, we reversed a sixty-eight percent downward departure where the defendant provided substantial but only "modest" assistance. 428 F.3d at 1163. In Saenz I, the defendant's corroborating testimony at a co-defendant's sentencing hearing supported a two-level increase in the Guideline offense level. Id. Saenz also attempted, unsuccessfully, to assist the government with an investigation into drug trafficking in California. "She did not, however, play a lead role in building a case on another offender, participate in undercover work (such as by wearing a recording device or making controlled purchase), give testimony in a grand jury or at a trial, or experience significant risk of injury or death." Id.^[25]

Conversely, in Pizano, we affirmed a seventy-five percent downward departure where the defendant began his cooperation immediately, supplied grand jury testimony against a family member and a major figure in the conspiracy, and put himself at risk. 403 F.3d at 995-96. In Pepper II, we affirmed a forty percent downward departure where the defendant's assistance consisted of information about two individuals' involvement with illegal guns and drugs, which the district court characterized as "pedestrian" or "average." 486 F.3d at 411; see also United States v. Pepper, 412 F.3d 995, 996-97 (8th Cir. 2005) (Pepper I) (vacating and remanding Pepper's initial sentencing).

The reversals in Haack, Dalton, and Saenz I, are easily distinguishable and support the district court's sixty percent departure in this case. Burns began his cooperation immediately, before asserting his right to counsel and before being made aware of the impact the Guidelines would have on his sentence. Additionally, he continued to cooperate through sentencing, provided key grand jury testimony leading to the indictment and conviction of another defendant, and offered detailed information about several groups who were manufacturing methamphetamine. The significance of Burns's cooperation was not diminished by a subsequent refusal to cooperate, Haack, 403 F.3d at 1005, a serious lapse in judgment as demonstrated in Dalton, 404 F.3d at 1033, or any suggestion the district court took into account improper or irrelevant factors, Haack, 403 F.3d at 1006. The district court concluded Burns's assistance fell between the assistance provided in Pizano and Pepper II. Though not so substantial as to warrant the seventy-five percent departure upheld in Pizano, Burns provided markedly greater assistance than the pedestrian assistance provided by the defendant in Pepper II. After reviewing those cases, I find no basis to conclude the district court's departure—which approximates the middle ground between Pizano and Pepper II—was unreasonable.

C

The majority also takes exception to the weight the district court afforded three of the five enumerated § 5K1.1(a) factors.

As a preliminary matter, the majority opinion discusses the various § 5K1.1(a) factors considered by the district court in isolation, concluding none of them individually supports the court's sixty percent departure. The record in this case makes clear the district court's departure was not based on any single factor. Rather, taken together, the court concluded the quality and quantity of Burns's assistance warranted a sixty percent departure. Section 5K1.1 provides: "The appropriate reduction shall be determined by the court for reasons stated that may include, but are not limited to, consideration of the following [enumerated factors]." Nothing in the structure of § 5K1.1 suggests that for a reduction to be appropriate it must be justified by a single factor or by each factor standing alone. Nor does § 5K1.1 suggest all the factors must be present or present to the same degree. Instead, it is apparent district courts are to consider the entire universe of a defendant's assistance to determine the appropriate departure. The majority's analysis of the district court's consideration of the § 5K1.1 factors is, of course, further colored by its unfounded belief that departures exceeding approximately forty-two percent should be deemed extraordinary. When properly analyzed, the district court's sixty percent departure is reasonable if Burns provided sufficient assistance to warrant a departure falling on the high side of ordinary.

First, the majority finds the district court improperly analyzed the timeliness factor. At sentencing, the government conceded ninety-nine percent of defendants do not cooperate before they are arrested. Sent. Tr. 6. Nonetheless, the majority concludes: "Although few defendants may participate earlier than did Burns, if a sufficient number participate in a similarly timely manner, Burns's participation could not be viewed as extraordinary." The majority further concludes the district court failed to consider the "government's unaddressed and uncontested insistence that many defendants cooperate in a similarly timely fashion." Under the facts of this case, however, Burns's timeliness must be considered extraordinary.

At sentencing, the government was asked whether, aside from "those very rare situations where a defendant comes forward and cooperates before he's arrested [did] this defendant cooperate[] in about as timely a manner" as possible? Sent. Tr. 7. In response, the government's attorney stated: "In this case with this defendant, as soon as he was arrested, he told everything; he was complete and truthful, yes." Id. Thus, according to the government, the only way Burns's cooperation could have been more timely was if he had come forward prior to his arrest— which, again, according to the government, only one percent of defendants do. This concession by the government renders irrelevant its claim "that many defendants cooperate in a similarly timely fashion." Whether that number be many or few, under no application of § 5K1.1 can it be reasonable to conclude a defendant's cooperation is extraordinary only if it exceeds the cooperation of ninety-nine percent of all other defendants. Finally, I am unpersuaded by the majority's reliance on Saenz I, 428 F.3d at 1162-63, as supporting its rejection of the district court's timeliness evaluation. In Saenz I, though timely, the defendant's cooperation was discounted because it did not compel anyone else to plead guilty. Id. Here it is uncontested Burns's cooperation led to the indictment and guilty plea of another defendant.

Next, the majority takes exception to the district court's evaluation of the truthfulness and completeness of Burns's assistance. At sentencing, the district court considered the truthfulness, completeness, and reliability of Burns's information, and concluded it was "a hundred percent complete, a hundred percent truthful, and a hundred percent reliable." Sent. Tr. 14. The government, as noted above, concedes Burns "told everything; he was complete and truthful." Sent. Tr. 7. The majority, however, concludes Burns's absolute compliance with § 5K1.1(a)(2) is insufficient to be deemed extraordinary. When evaluating Burns's timeliness, the majority concluded only one percent of defendants who are the most timely are extraordinary. When evaluating truthfulness, completeness, and reliability, the majority concludes even 100 percent compliance is insufficient to be extraordinary. I can conceive of no reasonable application of § 5K1.1 which would require a defendant to do the impossible, i.e., provide more than 100 percent compliance, for his cooperation to be deemed extraordinary.

Finally, the majority faults the district court's evaluation of the nature and extent of Burns's cooperation, and its evaluation of the usefulness and significance of his cooperation. According to the majority, the district court "applied a test that compared Burns's assistance to the assistance it speculatively believed the defendant capable of providing." It further contends the district court's departure based on Burns's "not inconsiderable" assistance, leaves too little room for more extensive assistance based departures.

I am unable to divine from the record any basis for concluding the district court's evaluation of the nature and extent of Burns's cooperation was speculative. The district court, while recognizing other defendants provide greater information, found Burns provided "every single bit of information he knew." Sent. Tr. 13-14. Conversely, there is nothing in the record suggesting Burns did not do everything he could do or that was requested and asked of him.

The district court found also Burns's cooperation both very significant and very useful; Burns continued to cooperate through sentencing, provided key grand jury testimony leading to the indictment and guilty plea of another defendant, and offered detailed information about several groups who were manufacturing methamphetamine. I find nothing unreasonable about the district court's evaluation of the significance and usefulness of Burns's cooperation and the role it played in support of the court's sixty percent departure. The majority's concern about the court's departure leaves too little room for departures of greater magnitude is belied by the availability of departures in excess of sixty percent. Though rarely traversed in our circuit, the range of departures between 60 and 100 percent represents nearly one half of the total range of possible departures available to defendants providing substantial assistance. The belief about too little unclaimed ground remains after the district court's departure results from this court's steadfast refusal to extend its gaze beyond a limited horizon. The remaining territory is more than sufficient to accommodate defendants who provide greater substantial assistance.

II

For the foregoing reasons, I respectfully dissent from the court's decision vacating the district court's sentence and remanding for resentencing. I concur in footnote six and part III. B. of the majority's opinion.

NOTES

[1] We observe that, following Rita v. United States, ___ U.S. ___, 127 S. Ct. 2456, 2465, 168 L. Ed. 2d 203 (2007), the presumption of reasonableness for a sentence within a properly calculated guidelines range exists only at the appellate level.

[2] Even though a similar analysis has thus far been used to weigh the reasonableness of substantial assistance departures and of post-Booker variances, we do not mean to suggest through our citations to both types of sentencing cases that the substantial assistance departure standard should always, going forward, mirror the variance standard. Because post-Booker variances must be reasonable with regard to § 3553(a), while substantial-assistance reductions are considered under the rubric of the sentencing guidelines or of § 3553(e) (which itself is not affected by Booker, see United States v. Williams, 474 F.3d 1130, 1132 (8th Cir.2007)), the general analytical approaches and associated standards may diverge. We do not now consider whether the analysis should diverge. Nevertheless, we emphasize here that the present case is entirely a substantial assistance departure case.

[3] Instead of relying on aggregate statistics, when there is any uncertainty concerning the strength of circumstances required to justify a given departure level, the strength of the particularized circumstances present in, or required by, prior decisions dealing with comparable departures should be given controlling weight.

[4] Accordingly, our observation that extraordinary departures require extraordinary circumstances should be read to represent a shorthand for a generally applicable proportionality principle despite the fact that the literal terminology involved may facially appear limited to the largest of departures.

[5] There is reason to believe this is the case. The district court's conclusion fails to consider the government's unaddressed and uncontested insistence that many defendants cooperate in a similarly timely fashion.

[6] The government also alleges that the district court abused its discretion by adhering to its bright-line rule of granting a 50 percent minimum departure for any defendant supplying entirely truthful, complete, and reliable assistance. In Saenz, we rejected this bright-line rule in part because to allow it would deter the government from moving for sentence reductions in otherwise borderline assistance cases. Saenz, 428 F.3d at 1163. Although the district court did not mention the 50 percent departure rule at Burns's sentencing hearing, we cannot say with certainty that the rule did not affect the district court's thinking. Accordingly, we restate here our conclusion in Saenz that any reliance on the bright-line rule would also amount to an abuse of discretion.

[7] In reaching this conclusion, we have considered and find to be without merit the government's allegation that the district court failed to give substantial weight to the government's evaluation. The commentary to § 5K1.1 requires that "substantial weight" be given to the government's evaluation of the extent of the assistance. U.S.S.G. § 5K1.1 cmt. n. 3. The government's recommendation is not controlling, however, and it is the district court's responsibility to determine an appropriate reduction. Haack, 403 F.3d at 1005. Although the district court stated that "it's hard to put any weight on the 15 percent [recommendation of the government]," we conclude that the court overcame its reluctance and accorded the government's evaluation substantial weight, for the court ultimately justified its departure by citing the strength of factors other than § 5K1.1(a)(1) despite its stated inclination to weigh the significance of Burns's assistance heavily in his favor. Furthermore, the district court's comments indicate that it fully considered that which it could make out of the government's evaluation. Cf. Haack, 403 F.3d at 1005 n. 2 (stating that opaque government recommendations deserve less weight). We do not consider final departures substantially at variance with unexplained government recommendations to be de facto proof of discretionary abuse.

[8] We note that prior panels have held that it was not erroneous for district courts to use starting points in excess of 360 months when departing from life sentences. See, e.g., United States v. Keller, 413 F.3d 706, 711 (8th Cir.) (470 months), cert. denied, 546 U.S. 1053, 126 S. Ct. 786, 163 L. Ed. 2d 609 (2005); United States v. Selby, 184 Fed.Appx. 589, 591 (8th Cir.2006) (unpublished per curiam) (405 months); see also United States v. Jensen, 493 F.3d 997, 1000-02 (8th Cir.2007) (expressing doubt about the adequacy of a 360-month starting point). Because the government has not appealed the 360-month starting point, we do not address the appropriateness of departing from a presumptive life sentence at, or in excess of, 360 months.

[9] Although I concur in the majority's affirmance of a 360-month starting point for Burns's life sentence, I note that I believe this measure appropriate only in cases where 360 months does not exceed the actual life expectancy of the defendant.

[10] The majority claims to eschew reliance on aggregate statistics. However, the majority notes that "prior decisions dealing with comparable departures should be given controlling weight" and analyzes the nature of Burns's departure by comparing it to prior percentages rejected by this court as extraordinary. The effect of this type of analysis is akin to placing Burns's departure on a line graph alongside other percentages and engaging in a statistical comparison rather than an individualized consideration of the defendant's assistance. Moreover, if such a comparison is valid, should it not be raised in the prosecutor's presentation to the district court rather than for the first time on appeal?

[11] "He uses statistics as a drunken man uses lamp-posts — for support rather than illumination." Attributed to Andrew Lang, quoted in Alan L. Mackay, Scientific Quotations: The Harvest of a Quiet Eye (1977).

[12] Notably, as mentioned by Judge Bye, the line of cases cited in the majority's opinion begins to resemble that previously described by Judge Heaney, where he notes this court's trend in reversing downward departures. See, e.g., United States v. Meyer, 452 F.3d 998, 1000 n. 3 (8th Cir.2006) ("Affirming upward variances at a rate of 92.3% while affirming downward variances at a rate of 15.8% could hardly be viewed as uniform treatment, and seems contrary to 18 U.S.C. § 3553(a)(6)'s concern with eliminating unwarranted disparity.")

[13] I write further in Section VI. of this dissent about the government's refusal or inability to assist the district court by providing a basis or reasons for its recommended departure.

[14] See Nancy Gertner, From Omnipotence to Impotence: American Judges and Sentencing, 4 Ohio St. J.Crim. L. 523 (2007) (comparing role of a federal sentencing judge to a common law civil code clerk or an "operator of a machine designed and built by legislators") (quoting John Henry Merryman, The Civil Law Tradition: Introduction to the Legal System of Western Europe and Latin America, 36 (2d ed.1985)).

[15] The majority also maintains that it is attempting to further the goal of preventing departures from becoming "untethered" from the Guidelines, citing Saenz, 428 F.3d 1159. The "untethering" and "reduction in disparity" goals are one in the same, as the reason stated in Saenz for preventing untethering is a reduction in disparity. See id. at 1162. In any event, departures also become "untethered" from the Guidelines when appellate judges invent an unwarranted "extraordinary" component to the explicit authority for individualized departures found in the Guidelines.

[16] "No state and, indeed, no other jurisdiction in the world makes an offender's sentence as dependent on the quality of his counsel as do the federal courts under the Guidelines. Finding what is relevant to a case in the 629-page Guidelines Manual, the 1100 pages of appendices explaining Guidelines provisions, and the endless judicial decisions interpreting the Guidelines takes a very good lawyer, and not every federal defendant has one. The disparate performances of the attorneys generate sentencing disparities that fly beneath the radar." Albert W. Alschuler, Disparity: The Normative and Empirical Failure of the Federal Guidelines, 58 Stan. L.Rev. 85, 111 (2005).

[17] "The prosecutor's role in the sentencing process has been enhanced by guidelines sentencing. Not only does he continue to determine who should be charged and what the charge should be, but the information that he controls largely determines the time to be served by an offender. No longer is a sentence subject to reduction by the Parole Commission, and no longer does the court retain its traditional ability to moderate the effect of the prosecutor's decisions by ultimately controlling the sentence imposed. A district court must consider the relevant conduct and the sentencing facts as presented to it and must impose a sentence within a given range if the appropriate facts are established by reliable evidence. The prosecutor's control over the ultimate sentence increases the prosecutor's bargaining power in plea negotiations." Heaney, supra, at 190.

[18] The Guidelines generate criticism by neutral and informed observers, including many federal sentencing judges. After more than twenty years of judicial turmoil, it is time for an improved federal sentencing procedure that will make criminal sentencing fair, reasonable, and effective. It is time to check the immense power of prosecutors and restore federal sentencing to judges. As I have said before, "Is anyone out there listening?" United States v. Alatorre, 207 F.3d 1078, 1080 (8th Cir.2000) (Bright, J., concurring).

[19] The prosecutor's mental process is of no concern, whether good or bad. It is the reasons, facts, and standards for sentencing that should be disclosed to the district court.

[20] The district judge commented during sentencing that this is not the first time the government has refused to disclose to him how it arrives at its recommendation. This observation suggests that the prosecutors in that district engage in guesswork, not reasoning, in making recommendations for a sentencing reduction.

[21] This explanation is more likely, considering the government's position to this court about its "deliberative process privilege."

[22] See, e.g., United States v. Saenz, 429 F. Supp. 2d 1081, 1089-90 (N.D.Iowa 2006) (referring to "the routine practice by the United States Attorney's Office in this District of making ridiculously stingy recommendations concerning the extent to which the court should depart downward for a defendant's substantial assistance, with no explication of the basis for such recommendations, accompanied by unfounded assertions that the court must then give such recommendations substantial deference").

[23] Chief Judge Bennett, whose sentencing decision we review in this case, has served as a United States District Court Judge since 1994 and is responsible for sentencing over 1,400 defendants. See United States v. Saenz, 429 F. Supp. 2d 1081, 1091 (N.D.Iowa 2006) (Saenz II).

[24] Owing to small revisions in the Special Post-Booker Coding Project, the numbers of cases referenced herein (9,061 & 5,754) are slightly higher than the numbers used by Judge Bennett (8,854 & 5,660) in Saenz II. The percentages of decrease in sentences, however, are identical.

[25] Notably, on remand for resentencing, Judge Bennett imposed an identical sentence. Saenz II, 429 F.Supp.2d at 1108. The government's appeal of Saenz's resentencing was later voluntarily dismissed.