Earl Thomas Anderson was convicted of armed bank robbery and sentenced to forty-six months under the sentencing guidelines. The district court included in the calculation of Anderson's total offense level an upward adjustment for being an organizer, leader, manager, or supervisor under U.S.S.G. § 3B1.1(c), even though it assumed that Anderson was the only person criminally responsible for committing the offense. While the guideline itself says nothing about requiring more than one criminally responsible person, the Introductory Commentary instructs that § 3B1.1 applies only if the offense involves more than one participant, and Application Note 1 defines "participant" as a "person who is criminally responsible for the commission of the offense." This case requires us to determine the proper weight to be given to the sentencing guidelines' commentary and the correct interpretation of § 3B1.1(c) in light of the commentary. We vacate and remand.

* On December 17, 1987, Anderson robbed a bank at gunpoint, and his codefendant, Gary Lee Hambley, drove the getaway car. Police arrested both of them. After the arrest, Hambley told officers that he had not known that Anderson was going to rob the bank and that, instead, Anderson had picked him up on the ruse of going to work. Hambley said that when they parked the car near the bank, Anderson told him to open the trunk and wait for him; then, Anderson came out of the bank, jumped into the trunk, and told Hambley to drive to another individual's house. According to Hambley, Anderson did not tell him about robbing the bank until after they had arrived at the house.

Anderson pled guilty to armed bank robbery in violation of 18 U.S.C. §§ 2113(a) and (d). Hambley was charged with armed bank robbery and with the lesser offense of accessory after the fact in violation of 18 U.S.C. § 3. He pled not guilty, and his trial ended with a hung jury. He then pled guilty to misprision of a felony in violation of 18 U.S.C. § 4.1 Counsel for both sides in the Hambley proceeding stipulated that Hambley did not know that Anderson was robbing the bank until after they had arrived at the house and that Hambley then failed to notify authorities and drove Anderson to another person's home.

Anderson's presentence report calculated his total offense level at twenty-two, which included a two-level upward adjustment for role in the offense. The report stated:

Since there is strong evidence the defendant was in the process of recruiting a getaway driver prior to the robbery, and two people are known to have been recruited, coupled with the fact the defendant's and codefendant's statements indicate the codefendant was not aware of the robbery until after it occurred, it would appear the defendant had an aggravating role in the offense. Even if one believes the codefendant was tricked into being the getaway driver, the mere act of trickery to cause the codefendant to become a participant in the crime would appear to strongly forward the proposition the defendant played a leadership role. Pursuant to Section 3B1.1(c), two levels are added.

U.S.S.G. § 3B1.1(c) provides for a 2-level upward adjustment when the defendant is "an organizer, leader, manager, or supervisor" in the criminal activity.

Anderson moved to correct the presentence report to eliminate the two-level adjustment for aggravating role in the offense. He argued that U.S.S.G. § 3B1.1(c) only applies when the offense is committed by more than one criminally responsible person, and since Hambley was not criminally responsible for the robbery, it did not apply in his case. The guideline, in subsection (c), says nothing about any required number of criminally responsible persons. The Introductory Commentary, however, says that "[w]hen an offense is committed by more than one participant, § 3B1.1 or § 3B1.2 (or neither) may apply," and Application Note 1 explains that "[a] 'participant' is a person who is criminally responsible for the commission of the offense."

The district court denied Anderson's motion, concluding that § 3B1.1(c) applies regardless of whether Anderson was the sole "participant." The district judge explained: "I don't find anything that satisfies me under the governing provision that I think I have to follow that the other person or persons must be a knowing participant." Anderson's offense level was established at twenty-two, and he was sentenced to forty-six months in prison.

Anderson has appealed this sentence twice. The first time, we summarily remanded for resentencing in light of Gubiensio-Ortiz v. Kanahele, 857 F.2d 1245 (9th Cir.1988), vacated, 488 U.S. 1036, 109 S. Ct. 859, 102 L. Ed. 2d 984 (1989), in which this court had held the sentencing guidelines to be unconstitutional. Before Anderson's resentencing, the Supreme Court upheld the constitutionality of the guidelines. Mistretta v. United States, 488 U.S. 361, 109 S. Ct. 647, 102 L. Ed. 2d 714 (1989). The district court therefore resentenced Anderson to the same sentence.

On Anderson's second appeal, the panel affirmed the sentence. United States v. Anderson, 895 F.2d 641 (9th Cir.1990). The majority concluded that the two-level adjustment under § 3B1.1(c) was appropriate because subsection (c) does not mention any requirement of more than one participant. Concluding that the commentary to the guidelines is to be treated like legislative history, which normally need not be considered in reviewing a statute that is clear on its face, the majority declined to consider the commentary. The dissent would have treated the commentary instead as an integral part of the guidelines and therefore would have interpreted subsection (c) to apply only when the offense is committed by more than one person who is criminally responsible for the commission of the offense.

We granted rehearing en banc to reconsider the interpretation of § 3B1.1 and to decide the broader question of the appropriate weight to give to the commentary when interpreting the guidelines.

Anderson's appeal is properly before us because it raises the issue of an incorrect application of the guidelines. 18 U.S.C. § 3742(a)(2). Because the facts are not in dispute and the interpretation of U.S.S.G. § 3B1.1 is a purely legal question, we review the district court's decision de novo. See United States v. McConney, 728 F.2d 1195, 1200-01 (9th Cir.) (en banc), cert. denied, 469 U.S. 824, 105 S. Ct. 101, 83 L. Ed. 2d 46 (1984).

As the panel opinion illustrates, the interpretation of § 3B1.1 and probably many other guidelines can turn on the weight given to the commentary. Section 3B1.1 itself does not say that any particular number of participants is required to trigger application of the guideline, so it is only by referring to the commentary that courts could conclude that such a requirement exists. Our first task, therefore, is to determine how much weight courts must give to the commentary when interpreting the guidelines.

* Congress created the United States Sentencing Commission and directed it to promulgate guidelines and general policy statements for use by the federal courts in determining criminal sentences. 28 U.S.C. §§ 991(a) and 994(a). In April 1987, the Commission submitted its original guidelines package to Congress for approval. Although the Commission was required only to submit "sentencing guidelines" to Congress, Pub.L. No. 98-473 § 235(a)(1)(B), 98 Stat. 1837, 2031-32, the Commission submitted commentary along with the guidelines, see United States Sentencing Commission, Sentencing Guidelines and Policy Statements (Apr. 13, 1987). Congress had six months to review the guidelines package, and the guidelines became effective November 1, 1987.2

Since 1987, the Commission has amended both the guidelines and the commentary. Although only required to submit "amendments to the guidelines" to Congress for approval, 28 U.S.C. § 994(p), the Commission has submitted its amendments to the commentary to Congress as well, United States Sentencing Commission, 1989 Annual Report 19. Each year, the Commission publishes its Guidelines Manual, and in that manual prints its commentary immediately before and after each guideline.

Apart from an amendment to 18 U.S.C. § 3553,3 Congress has not specifically addressed the commentary in its statutory provisions, and we have found nothing in the legislative history explaining its significance. We are therefore left to determine its status without Congress's guidance. On one end of the spectrum, the commentary could be viewed as equal to the guidelines themselves. In that case, courts would consider a guideline and its attendant commentary together and treat them as if they were different subsections of a single statute. In the event of a conflict, the language of the commentary could prevail over the language of the guideline as easily as the other way around.

On the other end of the spectrum, the commentary could be viewed as legislative history. In that case, it would never prevail over a guideline, and courts normally could ignore it if the language of the guideline is unambiguous. See, e.g., Rubin v. United States, 449 U.S. 424, 430, 101 S. Ct. 698, 701, 66 L. Ed. 2d 633 (1981) ("When we find the terms of a statute unambiguous, judicial inquiry is complete, except 'in "rare and exceptional circumstances." ' " (quoting TVA v. Hill, 437 U.S. 153, 187 n. 33, 98 S. Ct. 2279, 2298 n. 33, 57 L. Ed. 2d 117 (1978) (quoting Crooks v. Harrelson, 282 U.S. 55, 60, 51 S. Ct. 49, 50, 75 L. Ed. 156 (1930)))); Fernandez v. Brock, 840 F.2d 622, 632 (9th Cir.1988) ("Unless exceptional circumstances dictate otherwise, if we find the statutory language unambiguous, then we will not resort to the legislative history.").

Anderson urges that the commentary be viewed as an integral part of the guidelines. He contends that courts must consider and place great importance on the commentary because it reflects the intent of the Commission. He further argues that ambiguity in the guidelines must be resolved in favor of the defendant under the "rule of lenity." The government counters that the commentary does not have equal status with the text of the guidelines and cannot be used to contradict otherwise clear language in the guidelines themselves. It argues that commentary is only a secondary interpretative aid akin to legislative history, but at the same time, it agrees with Anderson that the commentary is an integral part of the guidelines in that it cannot be ignored.

We believe that the weight of the commentary must fall somewhere in between that of ordinary legislative history and that of the guidelines themselves. In determining where on this spectrum the commentary should fall, we must decide whether courts must always consider the commentary, whether the commentary reflects the intent of the Commission, and whether it could prevail over a guideline in the event of a conflict.

We begin with the Commission's own view of how much weight should be given to the commentary. The guidelines provide:

The Commentary that accompanies the guideline sections may serve a number of purposes. First, it may interpret the guideline or explain how it is to be applied. Failure to follow such commentary could constitute an incorrect application of the guidelines, subjecting the sentence to possible reversal on appeal. Second, the commentary may suggest circumstances which, in the view of the Commission, may warrant departure from the guidelines. Such commentary is to be treated as the legal equivalent of a policy statement. Finally, the commentary may provide background information, including factors considered in promulgating the guideline or reasons underlying promulgation of the guideline. As with a policy statement, such commentary may provide guidance in assessing the reasonableness of any departure from the guidelines.

U.S.S.G. § 1B1.7 (citation omitted).

In this guideline, the Commission divided its commentary into three types: commentary that interprets a particular guideline, commentary that suggests circumstances warranting departure, and commentary that provides background information relevant to determining whether the degree of departure is reasonable. We are concerned here with the first type of commentary, that which interprets or explains a guideline. Regarding this type of commentary, the Commission states that failure to follow it "could constitute an incorrect application of the guidelines."4

Section 1B1.7 indicates that the Commission views the commentary as an interpretive aid, secondary to the guideline itself. The guideline does not say that failure to follow the commentary in itself could subject the sentence to reversal. Instead it says that such a failure could constitute an incorrect application of the guidelines, which would subject the sentence to possible reversal. Failure to follow the commentary, then, is only error to the extent that it leads to an incorrect interpretation of a guideline. This construction is consistent with 18 U.S.C. § 3742(a)(2); which provides for appeals from sentences that are imposed as a result of an incorrect application of the guidelines but not from sentences that are imposed as a result of an incorrect application of the commentary.

It is also consistent with the commentary to U.S.S.G. § 1B1.7, which explains:In stating that failure to follow certain commentary "could constitute an incorrect application of the guidelines," the Commission simply means that in seeking to understand the meaning of the guidelines courts likely will look to the commentary for guidance as an indication of the intent of those who wrote them. In such instances, the courts will treat the commentary much like legislative history or other legal material that helps determine the intent of a drafter.

Of course, we cannot say how much weight to give to this commentary until we have decided how much weight to give to commentary. It makes no difference in this instance, however, because whatever weight this commentary has, it further supports the conclusion that the Commission intends for the commentary to be a secondary interpretive aid. By saying that courts will "look to the commentary for guidance" and will treat it "much like legislative history," the Commission again has indicated that the commentary is an aid to correct interpretation of the guidelines, not a guideline itself or on a par with the guidelines themselves.

One would also expect that if the Commission had intended the commentary to carry the same weight as the guidelines, it would simply have put the information in the guidelines themselves. That it instead separated the commentary from the guidelines suggests that the Commission intended for courts to differentiate between the commentary and the guidelines.

We do not wish to give the commentary more weight, or less, than the Commission intended. Our ultimate goal, whether we are interpreting a particular guideline or the overall approach to applying the guidelines, must be to interpret the guidelines in a way that most likely reflects the intent of the Commission. If the Commission did not expect that the commentary would have the same weight as the guidelines, it might not have given the same amount of time and care to drafting the commentary as it did to drafting the guidelines.

We therefore conclude that the commentary cannot be treated as equivalent to the guidelines. Because it is an interpretive aid for courts applying the guidelines, it must be treated as something less than the guidelines themselves.

At the same time, we also conclude that the commentary must be treated as something more than ordinary legislative history, which normally can be ignored if the statute is clear. Section 1B1.7 and its commentary assume that guidelines often will need further explanation. If the guidelines were clear and courts were expected simply to follow their plain meaning, the Commission would have had no reason to state that failure to follow the commentary could constitute an incorrect application of the guidelines.

Moreover, the commentary is different from legislative history in that it is passed by the whole Commission. Unlike legislative history, which could reflect just the views of Congressional staff or of a minority of Congress, the commentary is put out by the Commission itself and is approved by the Commission as a whole. Furthermore, Congress had the commentary before it when it approved the guidelines. Thus, the commentary is more likely to reflect the intent of the Commission and of Congress than is ordinary legislative history. See United States v. DeCicco, 899 F.2d 1531, 1536-37 (7th Cir.1990); United States v. Carroll, 893 F.2d 1502, 1508 (6th Cir.1990); United States v. Pinto, 875 F.2d 143, 144 (7th Cir.1989).

The commentary is analogous in this sense to the advisory committee notes that accompany the federal rules of practice and procedure (e.g., Federal Rules of Civil Procedure). The advisory committees draft rule changes, which are ultimately submitted to the Supreme Court and then to Congress, and each committee drafts the committee notes, which explain the purpose and intent of the proposed rule changes. The Supreme Court has held that the advisory committee notes are "of weight" in interpreting the rules. Torres v. Oakland Scavenger Co., 487 U.S. 312, 316, 108 S. Ct. 2405, 2408, 101 L. Ed. 2d 285 (1988); Mississippi Publishing Corp. v. Murphree, 326 U.S. 438, 444, 66 S. Ct. 242, 245, 90 L. Ed. 185 (1946).

Similarly, revision committees often draft revisions to Congressional statutes, and the committees publish, and submit to Congress, reviser's notes explaining the purpose of the revisions. Courts give considerable weight to reviser's notes primarily because they are before Congress when it considers the revisions. See Western P.R.R. Corp. v. Western P.R.R., 345 U.S. 247, 254-55, 73 S. Ct. 656, 659-60, 97 L. Ed. 986 (1953) (reviser's notes entitled to "great weight"); Ex Parte Collett, 337 U.S. 55, 65-71, 69 S. Ct. 944, 949-953, 93 L. Ed. 1207 (1949) (emphasizing that reviser's notes were before Congress throughout its consideration of revision to venue provisions of Judicial Code); Acron Invs., Inc. v. Federal Sav. & Loan Ins. Corp., 363 F.2d 236, 240 (9th Cir.) (reviser's notes are "authoritative" in interpreting statutes), cert. denied, 385 U.S. 970, 87 S. Ct. 506, 17 L. Ed. 2d 434 (1966). As with the advisory committee notes and reviser's notes, we believe the commentary is entitled to considerable weight, more so than ordinary legislative history, in interpreting the guidelines.

The commentary is also more readily available than ordinary legislative history. Whereas legislative history can be buried in volumes of reports and records, the commentary is printed alongside the guidelines. This distinction is important for two reasons. First, considering the commentary requires little additional time by attorneys or courts than looking at the guidelines themselves, so there is no practical reason for courts to ignore it. Second, members of the public are more certainly on notice of the contents of the commentary than they are of the contents of legislative history, so we do not have the same due process concern that we might have when relying on legislative history.

Finally, sound reasons of public policy favor consideration of the commentary. Because the commentary is printed alongside the guidelines, if a court ignores the commentary and interprets the guideline in a way inconsistent with the commentary, the inconsistency is readily apparent to the defendant and anyone else who is aware of the decision. In that case, even if the court explains its reason for ignoring the commentary, the sentence may appear arbitrary or unfair.

Because the commentary cannot be treated as equivalent to the guidelines themselves but also cannot be treated merely as legislative history, which can be ignored, it must be treated as something in between. Several principles should guide courts in steering the middle course.

First, courts should always consider the commentary, regardless of how clear the guideline may appear on its face. As a practical matter, it makes sense to consider the commentary because it is a part of the Guidelines Manual. In addition, the Commission expects that guidelines often will not explain themselves and that courts therefore will consider the commentary when interpreting them. See U.S.S.G. § 1B1.7 and commentary.

Second, courts should construe a guideline and its commentary so as to be consistent, if that is possible. If the guideline read in light of the commentary is still not entirely clear, courts should construe the guideline and commentary together so that they make sense in terms of the underlying purposes and overall structure of the guidelines. In addition, where the guideline and commentary appear headed in different directions, they should be construed so as to be consistent with the scheme of other guidelines (and their commentary) within the Part as a whole. Giving effect to the commentary in this way reflects the fact that the commentary is a relatively reliable source of information about the drafters' thinking, more so than legislative history, and will usually accurately reflect the intent of the Commission.

This rule of construction comports with the approach taken by other circuits. See, e.g., United States v. Bierley, 922 F.2d 1061, 1066 (3d Cir.1990) (agreeing with Judge Kozinski's dissenting opinion in Anderson, 895 F.2d at 647, and citing United States v. Ofchinick, 877 F.2d 251, 257 (3d Cir.1989) for proposition that commentary is more accurate reflection of intent than ordinary legislative history); United States v. Smith, 900 F.2d 1442, 1446-47 (10th Cir.1990) (commentary is "essential in correctly and uniformly applying the guidelines"; it reflects intent of the Commission; analysis begins with language of the guideline, which is to be followed if it is clear and unambiguous, in absence of express contrary intent; and when guideline is less than clear in light of other guidelines, commentary sheds light on meaning); United States v. DeCicco, 899 F.2d 1531, 1535-37 (7th Cir.1990) (introductory commentary, background commentary and application notes read together to ascertain intent of Commission); United States v. Rutter, 897 F.2d 1558, 1561 (10th Cir.), (commentary is "essential in correctly interpreting and uniformly applying the guidelines on a national basis"), cert. denied, --- U.S. ----, 111 S. Ct. 88, 112 L. Ed. 2d 60 (1990); United States v. White, 888 F.2d 490, 497 (7th Cir.1989) (application notes are contemporaneous explanations of the guidelines by the authors, entitled to substantial weight); United States v. Smeathers, 884 F.2d 363, 364 (8th Cir.1989) (commentary accompanying each section of guidelines reflects intent of Commission; Commission's instruction in commentary in determining how guideline is to be applied cannot be disregarded); United States v. Ortiz, 878 F.2d 125, 128 (3d Cir.1989) (interpretation of guideline consistent with commentary); United States v. Pinto, 875 F.2d 143, 144 (7th Cir.1989) (application notes are more reliable than legislative history).

Finally, because the commentary is not equal to the guidelines, in the unusual case in which the commentary and a guideline cannot be construed consistently, the text of the guideline must prevail. In so holding we are in accord with most of the other circuits that have considered the issue. See United States v. Bogas, 920 F.2d 363, 368 (6th Cir.1990) ("nothing in the manual suggests that the express command of a guideline section may be countermanded by the commentary"); United States v. Brunson, 882 F.2d 151, 157 (5th Cir.1989) (if interpretation of commentary would conflict with guideline itself, guideline prevails); United States v. Pinto, 875 F.2d 143, 144 (7th Cir.1989) (application notes are not formally binding and text would be followed over notes in event of conflict); see also Smith, 900 F.2d at 1446 ("As with statutory interpretations, our analysis must begin with the language of the guidelines in question. In the absence of express contrary intent we must follow the clear and unambiguous language of the guidelines."); United States v. Williams, 922 F.2d 737, 740 n. 4 (11th Cir.1991) (commentary does not have force of law but serves as aid in interpreting guidelines provision, as legal equivalent of policy statement, or as equivalent of legislative history).

We part company only with the Third Circuit, which applied an application note instead of a guideline when confronted with an inconsistency in a particular case. See United States v. Ofchinick, 877 F.2d 251, 257 (3d Cir.1989). We may also be giving less weight to the commentary than the Fourth Circuit. See United States v. Jones, 907 F.2d 456, 472-73 (4th Cir.1990) (Wilkinson, J., dissenting) (criticizing majority opinion for elevating application note to status of guideline), cert. denied, --- U.S. ----, 111 S. Ct. 683, 112 L. Ed. 2d 675 (1991). We believe the view that guidelines must prevail over commentary more accurately reflects the fact that the Commission may not have drafted the commentary as carefully as the guidelines since its own statement indicates its belief that the commentary would be treated as an interpretive aid and not be given as much weight as the guidelines. U.S.S.G. § 1B1.7 and comment.

In summary, in order to reflect the general reliability of the commentary while still recognizing its limitations, we hold that courts shall:

(1) consider the guideline and commentary together, and

(2) construe them so as to be consistent, if possible, with each other and with the Part as a whole, but(3) if it is not possible to construe them consistently, apply the text of the guideline.5

Having established the standard for use of the commentary, we now apply it to the interpretation of U.S.S.G. § 3B1.1(c), the particular guideline involved in this case. We join the Sixth, Seventh, and Eleventh Circuits and hold that § 3B1.1 only applies when the offense is committed by more than one criminally responsible person.

* If we were to look only at the wording of the guideline itself, we would conclude that an adjustment could be proper regardless of how many criminally responsible persons are involved. U.S.S.G. § 3B1.1 provides:

Based on the defendant's role in the offense, increase the offense level as follows:

(a) If the defendant was an organizer or leader of a criminal activity that involved five or more participants or was otherwise extensive, increase by 4 levels.

(b) If the defendant was a manager or supervisor (but not an organizer or leader) and the criminal activity involved five or more participants or was otherwise extensive, increase by 3 levels.

(c) If the defendant was an organizer, leader, manager, or supervisor in any criminal activity other than described in (a) or (b), increase by 2 levels.

The guideline on its face does not require any particular number of participants for an adjustment to be given under subsection (c). While the word "participants" appears in subsections (a) and (b), it is omitted in subsection (c). Under ordinary rules of statutory construction, we would presume that the omission was intentional and that subsection (c) therefore applies regardless of how many participants are involved. See, e.g., Russello v. United States, 464 U.S. 16, 23, 104 S. Ct. 296, 300, 78 L. Ed. 2d 17 (1983).

The next step, however, must be to consider the guideline and the commentary together, construing them so as to be consistent with each other and with the Part as a whole. The Introductory Commentary to Part B of Chapter Three states: "When an offense is committed by more than one participant, § 3B1.1 or § 3B1.2 (or neither) may apply. Section 3B1.3 may apply to offenses committed by any number of participants." Application Note 1 explains that "[a] 'participant' is a person who is criminally responsible for the commission of the offense, but need not have been convicted."

Anderson argues that this commentary makes clear that the Commission intended that § 3B1.1 only apply to offenses that are committed by more than one criminally responsible person because the Introductory Commentary creates a threshold requirement of more than one participant and Application Note 1 specifically defines participant to mean a person who is criminally responsible for the commission of the offense. The government counters that the Introductory Commentary does not create a threshold requirement of more than one criminally responsible person. The government argues that simply because § 3B1.1 may apply when an offense involves more than one participant does not mean that it cannot apply when an offense does not involve more than one participant.

While the government's construction is logically accurate, it would make the Introductory Commentary meaningless. This commentary distinguishes §§ 3B1.1 and 3B1.2 from § 3B1.3, which may apply to offenses committed by any number of participants, including offenses committed by only one participant. If the Commission had intended §§ 3B1.1 and 3B1.2 also to apply to offenses committed by only one participant, it would likely not have separated out § 3B1.3. The Commission could simply have said that all three sections may apply to offenses involving any number of participants, or said nothing about the number of participants at all. We presume that the Commission did not intend to make a meaningless statement in the Introductory Commentary, and therefore it must mean that §§ 3B1.1 and 3B1.2 only apply to offenses committed by more than one participant. See, e.g., United States v. Mehrmanesh, 689 F.2d 822, 829 (9th Cir.1982) ("We may not construe a statute so as to make any part of it mere surplusage.").

Anderson's construction is also consistent with the Part read as a whole. According to its introductory commentary, Part B provides for adjustments based on the role the defendant played in committing the offense. Section 3B1.1 allows a two to four level increase in offense level for an aggravating role, while § 3B1.2 permits a two to four level decrease for a mitigating role. The Background Commentary to § 3B1.1 indicates that the guideline applies to defendants who are part of a "criminal organization," and Application Note 1 to § 3B1.2 similarly refers to defendants involved in "concerted activity" and "the conduct of a group." Section 3B1.3 pertains to abuse of trust, and § 3B1.4 makes clear that no adjustment for role in the offense in any other case is permissible. So, while § 3B1.1(c) is something of a catchall, meant to include any small criminal enterprise other than one involving five or more participants or that is "otherwise extensive," see Background Commentary to § 3B1.1, to construe it so as to catch a single criminal actor appears at odds with what Part B aims to accomplish--adjusting for relative culpability in committing the offense. It would also run afoul of § 3B1.4 and its commentary, which observe that "[m]any offenses are committed by a single individual or by individuals of roughly equal culpability so that none of them will receive an adjustment under this Part."

If § 3B1.1 and its commentary are to be construed consistently, so that § 3B1.1 applies only to offenses committed by more than one participant, it follows that it applies only to offenses committed by more than one criminally responsible person. Application Note 1 specifically defines "participant" to mean a person who is criminally responsible for the commission of the offense. Because we construe the guideline and the commentary together, we presume that the Commission intended that the term have the same meaning in the guideline and its accompanying commentary. Cf. S & M Inv. Co. v. Tahoe Regional Planning Agency, 911 F.2d 324, 328 (9th Cir.1990) ("When the same word or phrase is used in different parts of a statute, we presume that the word or phrase has the same meaning throughout."), cert. denied, --- U.S. ----, 111 S. Ct. 963, 112 L. Ed. 2d 1050 (1991).6 Construing the guideline so as to be consistent with the commentary as well as the Part as a whole, § 3B1.1 (including subsection (c)) appears to apply only when the offense involves more than one person who is criminally responsible for the commission of the offense.

Based on this construction of the guideline, we have to conclude that the district court incorrectly applied § 3B1.1(c) so as to adjust Anderson's offense level upward by two points on the assumption that the person with respect to whom he was a leader, organizer, supervisor or manager need not have been criminally responsible for the commission of the offense.7

We are also persuaded for prudential reasons to join other circuits in holding that what the Commission has said evinces its determination that adjustments for role in the offense are only appropriate when more than one person is criminally responsible for committing the offense.

On facts similar to this case, the Seventh Circuit held that a defendant could not receive an adjustment under § 3B1.1(c) because the defendant was the only criminally responsible person involved in the crime. United States v. DeCicco, 899 F.2d 1531, 1535-37 (7th Cir.1990). The court relied on the Introductory Commentary and Application Note 1 and concluded that § 3B1.1 did not apply because the only people the defendant organized were "the numerous unknowing victims of his fraudulent schemes as well as unwitting family members who provided him with transportation home from one of his fraudulent transactions." Id. at 1535. The court found additional support in the Background Commentary to § 3B1.1, which states that the guideline "provides a range of adjustments to increase the offense level based upon the size of a criminal organization (i.e., the number of participants in the offense) and the degree to which the defendant was responsible for committing the offense," and that subsection (c) is intended to cover "relatively small criminal enterprises." Id.

Similarly, the Eleventh Circuit held that a defendant could not be given an adjustment under § 3B1.1(c) when he had not organized any "participants," as that term is defined in Application Note 1. United States v. Markovic, 911 F.2d 613, 616-17 (11th Cir.1990). In that case, the only other person involved in the crime was someone who did not understand English well enough to form the necessary criminal intent to support a criminal conviction.8

In United States v. Carroll, 893 F.2d 1502, 1507-09 (6th Cir.1990), the Sixth Circuit held that § 3B1.1 does not apply if the defendant only organizes or leads undercover government agents. The court explained, "Because Carroll only 'organized' law enforcement authorities--persons who could not have been 'criminally responsible'--it is undisputed that Carroll's crime involved only one offender and, therefore, no criminal organization or enterprise." Id. at 1507; see also United States v. Bierley, 922 F.2d 1061, 1065-66 (3d Cir.1990) (section 3B1.2 held inapplicable when the defendant dealt only with an undercover agent on the ground that more than one culpable individual is required for mitigating adjustment); United States v. Daughtrey, 874 F.2d 213, 216 (4th Cir.1989) (opinion by Wilkins, J.) (explaining that "[s]ince many offenses are defined in general terms and encompass a variety of conduct, a judge should be able to make a distinction at sentencing between individuals who violated the same statute but nevertheless participated in the crime in ways so different that significant degrees of culpability exist," and noting that pursuant to congressional directive, 28 U.S.C. § 994(d)(9), the Commission promulgated §§ 3B1.1 and 3B1.2 to provide offense level adjustments for aggravating and mitigating roles in offenses committed by more than one participant).

Several other circuits, though not faced with the issue of how many "criminally responsible" persons are required, have concluded that § 3B1.1 does not apply when the offense is committed by a single individual without the aid of other people as in, for example, a possession offense. See United States v. Streeter, 907 F.2d 781, 792 (8th Cir.1990); United States v. Pettit, 903 F.2d 1336, 1341 (10th Cir.), cert. denied, --- U.S. ----, 111 S. Ct. 197, 112 L. Ed. 2d 159 (1990); United States v. Fuller, 897 F.2d 1217, 1220 (1st Cir.1990); United States v. Williams, 891 F.2d 921, 925 (D.C.Cir.1989).

The Fifth Circuit may have gone the other way in United States v. Mejia-Orosco, 867 F.2d 216, 220-22 (5th Cir.), cert. denied, 492 U.S. 924, 109 S. Ct. 3257, 106 L. Ed. 2d 602 (1989), although its holding is not entirely clear. In Mejia-Orosco, the court upheld a finding that the defendant was an organizer, leader, supervisor, or manager in the criminal activity of alien smuggling on the footing that regardless of whether relatives of the illegal aliens were criminally responsible or were duped, the enterprise was extensive within the meaning of § 3B1.1. In so holding, the court recognized that the Introductory Commentary "clearly indicates that there must be more than one participant involved in the criminal activity for this section to apply," id. at 220, but it did not address Application Note 1's definition of "participant."

We join those circuits that have squarely addressed the issue and hold that § 3B1.1 only applies when the offense is committed by more than one person who is criminally responsible for the commission of the offense. Construing this guideline compatibly with the interpretation given by most other courts will further Congress's purpose of avoiding unwarranted sentencing disparities. See 28 U.S.C. § 991(b)(1)(B). Even were we inclined to write differently if the slate were clean, the interpretation of this single guideline is not so important as to warrant splitting the circuits in light of the need for uniformity in sentencing decisions. Finally, if the Commission believes we have all gone astray, it can rearticulate its intent and amend the guidelines.

In the meantime, because we conclude that the district court erred in its application of the guidelines by including a two-level adjustment for aggravating role in the offense without finding that Anderson was an organizer, leader, supervisor, or manager in a criminal activity involving another criminally responsible individual, we remand for resentencing. We emphasize, however, that we express no view on whether Hambley was in fact criminally responsible for the commission of the offense.

VACATED AND REMANDED.