673 F.3d 638 (2012)
UNITED STATES of America, Plaintiff-Appellee,
v.
Anthony RAUPP, Defendant-Appellant.
No. 11-2215.
United States Court of Appeals, Seventh Circuit.
Argued November 2, 2011.
Decided March 9, 2012.
Matthew J. Rinka (argued), Attorney, Office of the United States Attorney, Indianapolis, IN, for Plaintiff-Appellee.
William E. Marsh, Juval O. Scott (argued), Attorneys, Indiana Federal Community Defenders, Inc., Indianapolis, IN, for Defendant-Appellant.
Before EASTERBROOK, Chief Judge, and POSNER and WOOD, Circuit Judges.
EASTERBROOK, Chief Judge.
Anthony Raupp pleaded guilty to possessing a firearm despite his status as a felon. 18 U.S.C. § 922(g)(1). The district court concluded that Raupp is a "career offender" under the Sentencing Guidelines because he has at least two other convictions for crimes of violence. U.S.S.G. § 4B1.1. This appeal, from the sentence of 100 months' imprisonment, presents a single question: Whether a conspiracy to commit robbery is a "crime of violence" under the Guidelines.
Robbery in Indiana is a "crime of violence" under the Guidelines and a "violent felony" under the Armed Career Criminal Act, 18 U.S.C. § 924(e). See United States v. Lewis, 405 F.3d 511, 514 (7th Cir.2005). Raupp was convicted under Ind.Code § 35-41-5-2 of conspiring to violate Ind.Code § 35-42-5-1, Indiana's robbery statute. An application note to U.S.S.G. § 4B1.2, which defines the phrase "crime of violence", tells us that an inchoate offense such as conspiracy is a "crime of violence" when the underlying crime is one. This note reads: "`Crime of violence' and `controlled substance offense' include the offenses of aiding and abetting, conspiring, and attempting to commit such offenses." That disposes of this appeal, as far as the Sentencing Commission is concerned.
Raupp asks us to ignore the application note. He contends that it has been superseded by Begay v. United States, 553 U.S. 137, 128 S.Ct. 1581, 170 L.Ed.2d 490 (2008), and later decisions. Begay held that the Armed Career Criminal Act treats as a "violent felony" only an offense in which violence is an element, see § 924(e)(2)(B)(i), or is similar to the introductory list in subsection (B)(ii) in the sense that the activity not only creates risks of injury but also is purposeful, violent, and aggressive conduct. 553 U.S. at 144-45, 128 S.Ct. 1581. There's nothing violent about conspiracy, Raupp insists; a conspiracy is an agreement, an exchange of words rather than an aggressive deed. Although James v. United States, 550 U.S. 192, 127 S.Ct. 1586, 167 L.Ed.2d 532 (2007), holds that attempted burglary is a violent felony under the Act, Raupp contends that we should not infer from James that all inchoate crimes can be classified the same way. Conviction for attempt usually requires proof of a substantial step toward the completed crime, while conviction for conspiracy does not. Some conspiracy statutes don't require proof of an overt act, see United States v. Shabani, 513 U.S. 10, 115 S.Ct. 382, 130 L.Ed.2d 225 (1994), and those that do could be satisfied by peaceable steps, such as casing the joint in advance of a robbery.
Begay and its successors interpret a statute, not the Guidelines. The final step in Raupp's argument is the proposition that the statute and the career-offender Guideline must be understood identically. We have held that, when the Guidelines and the Armed Career Criminal Act use the same language, they receive the same interpretation. See, e.g., United States v. Woods, 576 F.3d 400, 403-04 (7th Cir. 2009); United States v. Templeton, 543 F.3d 378, 380 (7th Cir.2008). The definition of "crime of violence" in U.S.S.G. § 4B1.2(a) is materially the same as the definition in § 924(e)(2)(B). It follows, Raupp contends, that Indiana's crime of conspiracy to commit robbery is not a "crime of violence" under the Guidelines. If that's right, then Raupp's correct sentencing range is lower than the one the district court calculated, and he would be entitled to a remand for resentencing.
The United States contends that conspiracy should be treated like attempt (and for that matter aiding and abetting) under both the statute and the Guidelines. We need not decide whether that is so. James reserved the question whether a particular inchoate offense may be so far distant from the completed crime that it should not be treated as a "violent felony" under the statute. 550 U.S. at 205-06, 127 S.Ct. 1586. But Raupp was not convicted under § 924(e) of being an armed career criminal. He was convicted of unlawfully possessing one firearm and sentenced as a career offender under the Guidelines. That makes a difference.
Woods and Templeton hold that identical language implies identical interpretation, but the Guidelines contain some language that is not in the statute. The application note about the treatment of inchoate offenses is unique to the Guidelines. Section 924(e)(2)(B) of the statute corresponds to § 4B1.2(a); everything else in the Guidelines and the accompanying notes must be taken into account, not ignored. If the Sentencing Commission wants to have a list of qualifying offenses that differs from the one in the statute (as Begay reads § 924(e)), there's no reason why the judges should say nay. The Commission could have put the language of the note in § 4B1.2(a) as a new paragraph, and then Raupp's argument would be sunk. Likewise the Commission could have added offense levels for anyone whose record includes a conviction of conspiracy to commit robbery, whether or not that conviction is classified as a "crime of violence." Why should it matter that the Commission achieved the same end by using a note to elaborate on the meaning of "crime of violence"?
Decisions such as Auer v. Robbins, 519 U.S. 452, 461-63, 117 S.Ct. 905, 137 L.Ed.2d 79 (1997), and Homemakers North Shore, Inc. v. Bowen, 832 F.2d 408 (7th Cir.1987), tell us that, when an agency interprets one of its own regulations, the agency's understanding prevails unless it contradicts the text of the regulation. The Sentencing Commission's application notes carry the same force.
the commentary [should] be treated as an agency's interpretation of its own legislative rule. The Sentencing Commission promulgates the guidelines by virtue of an express congressional delegation of authority for rulemaking, and through the informal rulemaking procedures in 5 U.S.C. § 553, see 28 U.S.C. § 994(x). Thus, the guidelines are the equivalent of legislative rules adopted by federal agencies. The functional purpose of commentary (of the kind at issue here) is to assist in the interpretation and application of those rules, which are within the Commission's particular area of concern and expertise and which the Commission itself has the first responsibility to formulate and announce. In these respects this type of commentary is akin to an agency's interpretation of its own legislative rules. As we have often stated, provided an agency's interpretation of its own regulations does not violate the Constitution or a federal statute, it must be given "controlling weight unless it is plainly erroneous or inconsistent with the regulation."
Stinson v. United States, 508 U.S. 36, 44-45, 113 S.Ct. 1913, 123 L.Ed.2d 598 (1993) (most internal citations omitted without indication). Thus the Supreme Court treats application notes as authoritative glosses on the Guidelines, unless the notes conflict with the text. See, e.g., Melendez v. United States, 518 U.S. 120, 116 S.Ct. 2057, 135 L.Ed.2d 427 (1996). See also, e.g., United States v. Vizcarra, 668 F.3d 516, 519-20 (7th Cir.2012); United States v. Hill, 645 F.3d 900, 907-08 (7th Cir.2011).
The first application note to § 4B1.2 tells us that the Sentencing Commission deems inchoate and completed offenses to be the same for the purpose of identifying career offenders. Is this note inconsistent with anything in the text of a Guideline? Begay and similar decisions do not concern any Guideline, so they do not address that subject; they have nothing to say about the effect of language that differentiates the Guidelines from the Armed Career Criminal Act. Accord, United States v. Martinez, 602 F.3d 1166, 1173-75 (10th Cir.2010) (considering, and rejecting, the argument that Begay nullifies the first application note to § 4B1.2; also rejecting an argument that the application note conflicts with the text of the Guideline).
Forget Begay and ask directly whether the note conflicts with the Guideline. Here is § 4B1.2(a):
The term "crime of violence" means any offense under federal or state law, punishable by imprisonment for a term exceeding one year, that
(1) has as an element the use, attempted use, or threatened use of physical force against the person of another, or
(2) is burglary of a dwelling, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.
And here is the note:
"Crime of violence" and "controlled substance offense" include the offenses of aiding and abetting, conspiring, and attempting to commit such offenses.
There cannot be a conflict because the text of § 4B1.2(a) does not tell us, one way or another, whether inchoate offenses are included or excluded. The note says they are included.
James puts to rest any doubt about whether answering the question "are inchoate offenses included?" conflicts with the text. The Supreme Court held in James that attempted burglary (an inchoate offense) is a "violent felony" for the purpose of § 924(e)(2)(B)(ii). As we related above, James does not tell us whether conspiracy to commit a violent felony is treated as a violent felony. But James does show that there is no conflict between the text of § 924(e)(2)(B)(ii), or § 4B1.2(a), and a rule treating an inchoate offense the same as the substantive offense. If treating substantive and inchoate offenses alike created a conflict, then James would have come out the other way. Instead it grouped attempt with the substantive crimeand it found support in the Sentencing Commission's note to § 4B1.2, quoting with approval, 550 U.S. at 206, 127 S.Ct. 1586, the very language equating attempts and conspiracies that Raupp wants us to ignore. See also Kawashima v. Holder, ___ U.S. ___, 132 S.Ct. 1166, 1172-73, 182 L.Ed.2d 1 (2012) (treating the offense of aiding and abetting a violation of tax laws as having the same attributes as the underlying substantive offense).
Whether conspiracy should be treated the same way as attempt cannot be resolved by the rule that the Commission must not use notes to contradict the Guidelines' text. Instead the question "should conspiracy and attempt be treated the same?" concerns wise policy. Perhaps, as Raupp contends, attempts are more dangerous than conspiracies, because of the substantial-step element of attempt crimes. Perhaps one could say, to the contrary, that the conspiracies are more dangerous than attempts, because conspiracies always involve multiple actors, and criminal gangs are more likely to succeed than lone criminals are. See United States v. Manzella, 791 F.2d 1263, 1265 (7th Cir.1986). Deciding how to handle conspiracy is a question about wise policy, not about textual conflict.
Raupp may be assuming that the career-criminal Guidelines implement § 924(e). If that were so, then our interpretation of the Guidelines would be required to mirror § 924(e) as interpreted in Begay (and James). But the career-offender Guidelines don't depend on § 924(e). Instead they implement 28 U.S.C. § 994(h), which requires the Commission to establish guidelines that specify a range near the statutory maximum for career criminals. Section 994(h) defines a career criminal as a person 18 or over who commits a "crime of violence" or a specified drug offense, and who also has two or more prior felony convictions for a "crime of violence" or a drug felony. Congress left "crime of violence" in § 994(h) undefined, though it did furnish a list of qualifying drug offenses. Section 994(h) does not cross-reference § 924(e).
Thus the Commission is free to go its own way; it can classify as "crimes of violence" offenses that are not "violent felonies" under § 924(e). It can't do this by application notes that contradict the text of the Guideline, but what the first note to § 4B1.2 does is address a questionthe treatment of inchoate offensesleft open by the text of § 4B1.2, as it is also left open by the text of § 924(e) and the holding of James.
Section 924(e) uses the definition of "violent felony" to set 15-year minimum sentences. The Sentencing Commission does not prescribe such a stern and inflexible outcome by defining "crime of violence." The career-offender Guideline raises the offense level without setting a mandatory minimum. It is just another Guideline, so the judge is free to impose a sentence outside the Commission's preferred range after evaluating each defendant's arguments. See United States v. Corner, 598 F.3d 411 (7th Cir.2010) (en banc). Raupp was free to contend that a conviction for conspiracy to commit robbery does not imply the same level of dangerousness as a conviction for robbery, and to seek a lower sentence on that account. But he does not assert that the district judge misunderstood the extent of her discretion or exercised it unreasonably. His sole contention is that district judges must ignore the first application note to § 4B1.2, and that contention does not carry the day.
AFFIRMED
WOOD, Circuit Judge, dissenting.
The only point that Anthony Raupp has raised on this appeal is whether he was correctly classified as a career offender under the U.S. Sentencing Guidelines, §§ 4B1.1 and 4B1.2. This question turns on the definition of "crime of violence" that should govern here, and whether that definition covers Raupp's prior state conviction for conspiracy to commit robbery. My colleagues conclude that the Sentencing Guidelines in this instance have adopted a significantly broader definition than the one used in the Armed Career Criminal Act ("ACCA"), 18 U.S.C. § 924(e)(2)(B). With respect, I do not agree with them. Their holding is inconsistent with a long line of cases holding that the text of § 4B1.2 and the nearly identical text in the ACCA have the same meaning. Although there would be no problem if the commentary to the Guidelines on which my colleagues rely were merely explaining concepts within the boundaries established by the Guidelines themselves, there is a problem when the commentary strays outside those boundaries altogether. I am persuaded by Raupp's argument, and I would therefore vacate his sentence and remand for resentencing.
Raupp pleaded guilty to being a felon in possession of a firearm, made criminal by 18 U.S.C. § 922(g)(1). As part of the required calculation of his advisory guidelines sentence, the district court considered whether he should be classified as a career offender under § 4B1.1, which in relevant part says:
(a) A defendant is a career offender if (1) the defendant was at least eighteen years old at the time the defendant committed the instant offense of conviction; (2) the instant offense of conviction is a felony that is either a crime of violence or a controlled substance offense; and (3) the defendant has at least two prior felony convictions of either a crime of violence or a controlled substance offense.
If those criteria are met, then the defendant's offense level is governed by § 4B1.1(b) rather than the Guidelines that would otherwise apply. The only point of contention in Raupp's case is element (3): does he have two prior felony convictions that qualify as a crime of violence (or a controlled substance offense)? He does not, unless the conspiracy conviction counts. And so the question becomes whether that offense meets the definition set forth in the Guidelines. To answer that, one must turn to § 4B1.2, which defines the terms used in § 4B1.1.
Section 4B1.2(a) defines the term "crime of violence" as follows:
... any offense under federal or state law, punishable by imprisonment for a term exceeding one year, that
(1) has as an element the use, attempted use, or threatened use of physical force against the person of another, or
(2) is burglary of a dwelling, arson, or extortion, involves the use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.
The language of the ACCA does not track that definition perfectly, but it comes close. After stating that a person who violates § 922(g) and who has three previous convictions "for a violent felony or a serious drug offense" gets an enhanced sentence, the ACCA defines the term "violent felony" as follows:
... any crime punishable by imprisonment for a term exceeding one year [or certain juvenile dispositions not at issue here], that
(i) has as an element the use, attempted use, or threatened use of physical force against the person of another, or
(ii) is burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another. ...
18 U.S.C. § 924(e)(2)(B). The only differences between these definitions are (1) the use of the term "crime of violence" versus the term "violent felony"; (2) the addition of the phrase "of a dwelling" after the word "burglary"; and (3) the addition of the word "the" before the word "use" in subpart two. To say that these are inconsequential differences (with the possible exception of the mention of a dwelling, see United States v. Wenner, 351 F.3d 969, 973 n. 2 (9th Cir.2003), but that is not at issue here) is to flatter them. In all respects relevant to this case, the two provisions are essentially identical.
So far, so good. This court has regularly respected the substantive identity of these two provisions. See, e.g., United States v. Scanlan, 667 F.3d 896, 898 (7th Cir.2012); Narvaez v. United States, 641 F.3d 877, 879 (7th Cir.2011); United States v. Woods, 576 F.3d 400, 403-04 (7th Cir. 2009) (applying the Supreme Court's decision in Begay v. United States, 553 U.S. 137, 128 S.Ct. 1581, 170 L.Ed.2d 490 (2008), an ACCA case, to § 4B1.2; the dissenters disagreed only with respect to how Begay should be applied, not whether it should be); United States v. Templeton, 543 F.3d 378, 380 (7th Cir.2008) ("It would be inappropriate to treat identical texts differently just because of a different caption.").
Raupp argues here that the inchoate crime of which he was convictedconspiring to commit robbery in violation of Ind. Code § 35-42-5-2is not (1) a crime that has physical force as an element; (2) a crime listed in the Guideline; nor (3) does it involve conduct "that presents a serious potential risk of physical injury to another" (the so-called "residual clause"). Raupp's first two points are indisputable. Unlike the inchoate crime of attempted robbery, which requires a substantial step toward the completed crime, Indiana's crime of conspiracy may involve only nefarious conversations. Henderson v. State, 769 N.E.2d 172, 177 (Ind.2002); see also Coleman v. State, 952 N.E.2d 377, 382 (Ind.App.Ct.2011) ("[T]he agreement itself constitutes the criminal act."). Using the modified categorical approach that governs these cases, Raupp's crime was indivisible and so it does not matter how he actually carried out his conspiracy. See Woods, 576 F.3d at 405-06. Conspiracy cannot be used as a predicate for his career criminal calculation, then, unless it falls within the "residual clause."
But the majority has concluded that it does not need to address the question whether conspiracy "presents a serious potential risk of physical injury to another." Instead, it plays a trump card against Raupp that comes from the commentary to § 4B1.2. Application Note 1, paragraph 1, says that "`[c]rime of violence' and `controlled substance offense' include the offenses of aiding and abetting, conspiring, and attempting to commit such offenses." If the Sentencing Commission is entitled to broaden the Guideline so that it applies to non-violent crimes such as the version of conspiracy that Indiana has adopted, then my colleagues are correct that this language checks Raupp's argument. In order to reach that result, they assume that the treatment of inchoate offenses is left open by § 4B1.2, and that all the Commission has done in the Application Note is to fill in a blank. In my view, however, the inclusion of all conspiracy offenses is inconsistent with the language of the Guideline, and thus the expansion implicit in the Application Note is incorrect under established principles of administrative law.
I begin with James v. United States, 550 U.S. 192, 127 S.Ct. 1586, 167 L.Ed.2d 532 (2007), since it is the only Supreme Court decision to discuss inchoate offenses. There the Court held that Florida's attempted burglary offense was properly viewed as a "violent felony" under the ACCA. The attempt offense at issue required that the burglar take "an overt act directed toward entering or remaining in a structure," id. at 206, 127 S.Ct. 1586, and this, the Court found, "creat[ed] a risk of violent confrontation comparable to that posed by finding him inside the structure itself," id. at 203-04, 127 S.Ct. 1586. The holding in James left untouched different statutes that do not require a similar substantial step toward the completion of the crime, such as those at issue in United States v. Strahl, 958 F.2d 980, 986 (10th Cir. 1992), United States v. Martinez, 954 F.2d 1050, 1054 (5th Cir. 1994), and United States v. Weekley, 24 F.3d 1125 (9th Cir. 1994). See also United States v. Martinez, 602 F.3d 1166, 1171-72 (10th Cir.2010) (concluding that Arizona's attempted burglary statute is not a violent felony and distinguishing James).
This is consistent with the approach to the ACCA taken in Sykes v. United States, ___ U.S. ___, 131 S.Ct. 2267, 180 L.Ed.2d 60 (2011), where the Court stressed that risk of violence is the dispositive factor for the residual clause. Id. at 2275. In so holding, it reinforced what it already had said in James: "We conclude that nothing in the plain language of clause (ii), when read together with the rest of the statute, prohibits attempt offenses from qualifying as ACCA predicates when they involve conduct that presents a serious potential risk of physical injury to another." 550 U.S. at 198, 127 S.Ct. 1586 (emphasis added). The Court's caution is most sensibly read as an acknowledgment that some inchoate offenses will create the necessary "serious potential risk of physical injury to another" and others will not. Conspiracy offenses are often a further step away from any physical dimension, as the majority recognizes. Ante at 640 (conviction for conspiracy may not require proof of an overt act, or the overt act might be satisfied by peaceable steps). We must take care not to conflate the analysis of attempt offenses, which in Indiana require a "substantial step" toward completion of the robbery, with the analysis of conspiracy offenses, which do not in that state. Calvert v. State, 930 N.E.2d 633, 640 (Ind. App.Ct.2010). Indeed, Indiana courts have stressed that a conviction for conspiracy "does not require proof that the defendant or a cohort actually committed or even attempted to commit the underlying crime." Coleman, 952 N.E.2d at 382. And the Indiana legislature has declined to treat the crime of conspiracy as a crime of violence under state law. See id. at 383 ("The legislature reasonably may have concluded that because conspiracies do not necessarily result in actual harm to a victim, and often require less proof of detrimental conduct in order to convict a defendant than as to the completed crime, it would not include conspiracies to commit a `crime of violence' within the definition of `crime of violence.'").
The majority postulates that the Sentencing Commission might have chosen to include inchoate offenses in the text of § 4B1.2 itself. I agree with them that there was nothing forcing the Commission to use the same language as that found in the ACCA, and I also agree that Raupp's argument would have no merit if the Commission had written a Guideline that expressly included all inchoate offenses. But, to state the obvious, that is not what the Commission did. Instead, it elected to promulgate a Guideline that mimics the ACCA and then to "interpret" that Guideline expansively in the Application Note. That is where it went astray.
Application Notes in the commentary to the Sentencing Guidelines should "be treated as an agency's interpretation of its own legislative rule." Stinson v. United States, 508 U.S. 36, 44, 113 S.Ct. 1913, 123 L.Ed.2d 598 (1993). And, "provided an agency's interpretation of its own regulations does not violate the Constitution or a federal statute, it must be given `controlling weight unless it is plainly erroneous or inconsistent with the regulation.'" Id. at 45, 113 S.Ct. 1913 (quoting Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414, 65 S.Ct. 1215, 89 L.Ed. 1700 (1945)). This type of deference today is most commonly referred to as "Auer deference," alluding to the leading case of Auer v. Robbins, 519 U.S. 452, 117 S.Ct. 905, 137 L.Ed.2d 79 (1997). The majority acknowledges Auer and Stinson, but it fails adequately to consider whether "the guideline which the commentary interprets will bear the construction." Stinson, 508 U.S. at 46, 113 S.Ct. 1913.
In order to assess the scope of the Guideline properly, we must take into account the fact that its language is identical to, and in fact came directly from, a statute passed by Congress. The definition found in the ACCA was put in place in 1986. See 18 U.S.C. § 924(e)(2)(B). The current language in Section 4B1.2 of the Guidelines was adopted by the Sentencing Commission in its 1989 amendments, with the explanation that "[t]he definition of crime of violence used in this amendment is derived from 18 U.S.C. § 924(e)." It is therefore fitting for courts, as ours consistently has done until now, to look to ACCA case law to understand the corresponding language in the Guidelines. See, e.g., United States v. Lockley, 632 F.3d 1238, 1243 n. 5 (11th Cir.2011) ("Though ACCA's `violent felony' enhancement and the Guidelines' career offender enhancement differ slightly in their wording, we apply the same analysis to both."); United States v. Archer, 531 F.3d 1347 (11th Cir. 2008) (applying the analysis of ACCA cases to conclude that carrying a concealed weapon is not a crime of violence under the Guidelines, after the Supreme Court granted certiorari in Archer and then remanded in light of its ACCA decision in Begay). The majority dismisses the relevance of the ACCA decisions because they concern the statute; they do not concern any Guideline. Ante at 640. I reject this effort to create a new division unjustified by either the text of the two provisions or by the agency's own rationale for choosing to adopt the statutory language as its own.
When an agency borrows language that originated with Congress, we should not ignore the meaning of the congressional language when we evaluate the agency's interpretation of its rules. For example, in United Fire Ins. v. C.I.R., this court found that deference to the IRS's interpretation of its own regulation was not justified. 768 F.2d 164, 169 (7th Cir. 1985) ("We ordinarily accord great deference to an agency's interpretation of its own regulations. But even so, we need not follow the agency's interpretation when there are compelling indications that it is wrong."). In that case, we noted that "the grounds for deference are diminished by the fact that the language of the regulations in question did not originate with the Commissioner but was taken almost verbatim from the Senate Report on the Revenue Act of 1942." Id. Because the regulatory language originated in a congressional report, we took the position that "our inquiry is really into the intent of Congress," id., and that the evidence of congressional intent did not support the interpretation held by the Commissioner.
Often, perhaps even usually, when courts apply Auer deference, they conclude that an agency's interpretation of its own regulation is permissible. See, e.g., Chase Bank USA, N.A. v. McCoy, ___ U.S. ___, 131 S.Ct. 871, 881-82, 178 L.Ed.2d 716 (2011) (applying Auer deference to the Federal Reserve Board's interpretation of its regulation); United States v. Larionoff, 431 U.S. 864, 872-73, 97 S.Ct. 2150, 53 L.Ed.2d 48 (1977) (deferring to Navy's interpretation of Department of Defense regulations because they were "not plainly inconsistent with the wording of the regulations"). But "usually" is not "always," and so it is not hard to find cases when a court has found that an agency's interpretation goes beyond the bounds of the regulation, even when the agency's language did not originate with Congress. See, e.g., Pettibone Corp. v. United States, 34 F.3d 536, 541 (7th Cir.1994) (refusing to defer to IRS's interpretation of its own regulation and emphasizing that an agency's power to interpret its regulations "must not be confused with a power to rewrite"); Public Citizen, Inc. v. Mineta, 343 F.3d 1159, 1166 (9th Cir.2003) (refusing to defer to the National Highway Traffic Safety Administration's understanding of when one of its rules was issued); Municipal Resale Serv. Customers v. FERC, 43 F.3d 1046, 1053 (6th Cir.1995) (refusing to defer to the Federal Energy Regulatory Commission's interpretation of one of its regulations).
The majority believes that the Tenth Circuit's Martinez decision, supra, 602 F.3d 1166, which held that attempted robbery is a crime of violence under the Guidelines even though it is not under the ACCA, supports its result. But Martinez's holding is premised on an idea that this court has decisively rejected and that even the majority does not endorse: that courts should treat the ACCA and Guidelines language differently because the Sentencing Commission "chose to use a different termcrime of violence, rather than violent felony"in its caption. Id. at 1173 (emphasis in original). Martinez actually highlights the inconsistency of this approach: the Tenth Circuit found that under the ACCA, attempted robbery does not "present[] a serious potential risk of physical injury to another," James, 550 U.S. at 198, 127 S.Ct. 1586, but when that same definition is copied into the Guidelines, somehow attempted robbery does present a serious enough risk. Those two conclusions, drawn from the identical text, are inconsistent. Nothing in the different captions affects the proper way to assess risk. The distinction that Martinez tries to draw is unconvincing in the opinion, and is equally unconvincing as an agency interpretation.
Indeed, there are signs on the horizon that the Supreme Court may be about to revisit Auer and endorse a more skeptical review of agency interpretations of their own regulations. The Court has before it the case of Christopher v. SmithKline Beecham Corp., ___ U.S. ___, 132 S.Ct. 760, 181 L.Ed.2d 480, cert. granted, Nov. 28, 2011. The first question presented in Christopher is "[w]hether deference is owed to the Secretary [of Labor]'s interpretation of the Fair Labor Standards Act's outside sales exemption and related regulations." This case arises against the backdrop of criticism of the Auer rule from Justices Scalia and Thomas. In Talk America, Inc. v. Michigan Bell Tel. Co., ___ U.S. ___, 131 S.Ct. 2254, 180 L.Ed.2d 96 (2011), Justice Scalia wrote in a concurring opinion that "deferring to an agency's interpretation of its own rule encourages the agency to enact vague rules which give it the power, in future adjudications, to do what it pleases. ... We have not been asked to reconsider Auer in the present case. When we are, I will be receptive to doing so." Id. at 2266; see also Thomas Jefferson University v. Shalala, 512 U.S. 504, 524-25, 114 S.Ct. 2381, 129 L.Ed.2d 405 (1994) (Thomas, J., dissenting and criticizing the application of Auer/Bowles deference); John F. Manning, Constitutional Structure and Judicial Deference to Agency Interpretations of Agency Rules, 96 COLUM. L.REV. 612, 615 (1996). All of this said, I recognize that formal changes to Auer lie down the road, if they take place at all. My view rests squarely on the law as it currently stands. I take note of Christopher only to make the point that it cannot weaken, and may strengthen, the points I am making here.
This debate between an agency's adoption of formal regulations (or, as here, the Guidelines) and its interpretations of those regulations is not an exercise in empty formality. There is a significant difference between the procedures that the Sentencing Commission uses when it promulgates the Guidelines and those that it uses when it writes commentary or policy statements. See 28 U.S.C. § 994(p); USSC Rules of Practice and Procedure 2-3 (2007), available at http://www.ussc.gov/Meetings_and_Rulemaking/Practice_Procedure_Rules. pdf. Proposed Guidelines or changes to Guidelines must be submitted to Congress no later than May 1 of a calendar year, where they must sit for 180 days to give Congress an opportunity to modify or disapprove them. In contrast, "[a]mendments to policy statements and commentary may be promulgated and put into effect at any time." Id. at 3 (Rule 4.1). The Commission must comply with the notice and comment rules in section 553 of the Administrative Procedures Act when promulgating Guidelines, but it is under no such obligation when promulgating commentary and policy statements. Id. (Rule 4.3). This calls to mind the distinction that the Supreme Court has drawn between Chevron deference (owed to regulations issued under formal notice-and-comment procedures) and Mead/Skidmore consideration for things like interpretations contained in policy statements, agency manuals, and enforcement guidelines. See United States v. Mead, 533 U.S. 218, 234, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001); Skidmore v. Swift & Co., 323 U.S. 134, 65 S.Ct. 161, 89 L.Ed. 124 (1944).
When an agency like the Sentencing Commission uses a regulation as a springboard for an "interpretation" that goes beyond the boundaries of the original regulation, Auer and Stinson tell us that it has gone too far. That is exactly what the Sentencing Commission did here, when it decided that the phrase "presents a serious potential risk of physical injury to another" could be stretched to include Indiana's inchoate offense of conspiracy to commit robbery. In my opinion, it cannot, and so I would find that Raupp is entitled to be resentenced without being classified as a career offender. I therefore respectfully dissent.