dissenting:
I dissent. For the reasons stated in the original opinion, United States v. Restrepo, 883 F.2d 781, 783-86 (9th Cir.1989), I believe that the Guidelines are ambiguous as to whether criminal conduct of which defendant Restrepo is not convicted can be aggregated to determine his total offense level.
The majority ignores the profound changes the Guidelines have brought about in federal criminal sentencing. Under the old sentencing scheme, trial judges had wide discretion in fixing appropriate sentences. Trial judges were permitted to take into account a variety of relevant unchallenged information, contained in the sentencing report, concerning defendants’ background and past behavior. Challenged information, when proved by a preponderance of the evidence, was also taken into account in imposing sentence. In most cases, no hearing was held on challenged information. Rather, pursuant to the Federal Rules of Criminal Procedure, the trial judge just considered the unchallenged information, and would often disregard challenged information. See Fed.R.Crim.P. 32(c)(3)(D).1 The judge, in the exercise of his or her discretion and depending on the circumstances of the particular case — including differences in the relative accuracy of information — would decide how much weight to give information contained in the sentencing report concerning the defendant’s past conduct. The judge might decide that the past conduct should enhance punishment or that it should have little if any effect on his or her sentencing decision.
In contrast, under the Guidelines the sentencing judge has very limited discretion. For example, adding a factor to the sentencing calculus — drug amounts from other uncharged counts — automatically and substantially increases the sentence that the judge is bound to give under the Guidelines. And, as one learned trial judge has pointed out, the Guidelines require sentencing judges to apply the factors “on the basis of any available information, whether or not that information was introduced at trial or proved beyond a reasonable doubt.” United States v. Davis, 715 F.Supp. 1473, 1476 (C.D.Cal.1989) (Letts, D.J.).
The majority’s contention that the approach it adopts is consistent with pre-Guideline sentencing practice is, therefore, wrong. While trial judges have long taken into account related criminal conduct in determining appropriate sentences, only now must they apply factors with fixed weights applicable to all cases without adjustment to reflect different degrees of certainty.2
The relevant inquiry is not whether under the old system judges considered defendants’ background and past conduct in fixing sentences, but whether the Commission, in restructuring the federal sentencing process, took “a ‘real offense’ or a *658‘charge offense’ approach in its provisions regarding multiple counts.” Restrepo, 883 F.2d at 785 n. 9. The Commission apparently took both approaches — one in the Relevant Conduct section and the other in the Multiple Counts section. This created the ambiguity on which the original Restrepo decision was based. The effect of aggregating drug amounts alleged in counts of which the defendant is convicted with drug amounts alleged in counts of which the defendant is not convicted is clear: the Multiple Counts section is applied not only to counts of which the defendant is convicted but also, as a practical matter, to counts of which the defendant is not convicted. And as a consequence punishment for the convicted counts is automatically increased. This conflicts with the plain language of the Multiple Counts section, which provides rules for “determining a single offense level that encompasses counts of which the defendant is convicted.” Id. at 785 (quoting, with added emphasis, Guidelines at 3.9).
The majority’s interpretation probably will make the prosecutors’ job easier. The government, in complex cases, could, if it wished, charge and prosecute defendants on easily provable offenses without having to prove beyond a reasonable doubt more serious offenses or charges where there appears to be a lack of sufficient proof. There is some incentive to do this because, under the majority’s holding, the government need only convict the defendant of one count to punish him or her for all related criminal behavior proved by a preponderance of the evidence. Moreover, the government could, for example, reach plea agreements with criminal defendants on minor charges in exchange for dropping all other counts, and then use the behavior underlying the dropped counts to increase dramatically the sentence, unless specifically foreclosed by the plea agreement. See United States v. Kamer, 781 F.2d 1380, 1386-87 (9th Cir.), cert. denied, 479 U.S. 819, 107 S.Ct. 80, 93 L.Ed.2d 35 (1986) (sentence imposed must comport with sentence bargained for).3
Restrepo’s case is instructive. He was convicted of two counts of possession of cocaine with intent to distribute, but was sentenced based on a drug quantity that included cocaine possessed by a co-defendant, Judith DeMaldonado. DeMaldonado initially was also charged with possession of cocaine, but the government dropped all charges against her in exchange for her testimony against Restrepo. Her testimony alone at the sentencing hearing was the basis of the district court’s conclusion that the cocaine she possessed was part of a common plan or scheme involving Restre-po; her testimony alone increased the amount of cocaine attributed to Restrepo from 37.5 grams to 103.33 grams. That substantially increased Restrepo’s guideline range — from 27-33 months to 41-51 months.
Given the severe consequences of applying the guidelines as interpreted by the majority, I believe that the government should, at the very least, be required to prove related criminal conduct as part of a common scheme by clear and convincing evidence, rather than by a preponderance of the evidence, as the majority holds. Upon factual findings, made without the benefit of a jury’s determinations, now rest significant deprivations of personal liberty.
. Rule 32(c)(3)(D) states in relevant part:
If the comments of the defendant and the defendant’s counsel or testimony or other information introduced by them allege any factual inaccuracy in the presentence investigation report ..., the court shall, as to each matter controverted, make (i) a finding as to the allegation, or (ii) a determination that no such finding is necessary because the matter controverted will not be taken into account in sentencing.
. The majority argues that sentencing judges may exercise discretion by departing downward from the Guidelines. It is true that the sentencing judge can, in rare circumstances, engage in the burdensome process of departing from the Guidelines. "A sentencing court can impose a sentence outside the [prescribed] range ... if ‘the court finds that there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines...." United States v. Michel, 876 F.2d 784, 786 (9th Cir.1989) (quoting 18 U.S.C. § 3553(b)). The Guidelines and our decisions, however, make clear that departures are a very narrow exception to the rule of determinate sentencing within the Guidelines. See Guidelines at 1.6-1.7; United States v. Hernandez-Vasquez, 884 F.2d 1314, 1315 (9th Cir.1989).
. Some courts have expressed concern about this. See, e.g., U.S. v. Scroggins, 880 F.2d 1204, 1213 (11th Cir.1989). One effect of these practices, of course, is to reduce criminal defendants’ incentive to enter into plea agreements. See Federal Courts Study Committee, Tentative Recommendations for Public Comment, 61 (1989) (”[T]he Guidelines have cast serious doubt on the continued ability of the federal courts to resolve a high percentage of criminal cases without trial.").