United States v. Tomasino, Michael J.

• Court: Court of Appeals for the Seventh Circuit
• Decided: March 15, 2000
• Docket: 99-2796
• Status: Published

Opinion

In the United States Court of Appeals For the Seventh Circuit

No. 99-2796

United States of America,

Plaintiff-Appellant,

v.

Michael J. Tomasino,

Defendant-Appellee.

Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 98 CR 956--Robert W. Gettleman, Judge.

Argued January 5, 2000--Decided March 15, 2000

Before Posner, Chief Judge, and Easterbrook and Ripple, Circuit Judges.

Posner, Chief Judge. The government appeals from the district court’s refusal (reported at 57 F. Supp. 2d 565 (N.D. Ill. 1999)) to enhance the defendant’s mail fraud sentence under U.S.S.G. sec. 2F1.1(b)(7)(B), which provides for a 4-level increase in sentence if the fraud "affected a financial institution and the defendant derived more than $1,000,000 in gross receipts from the offense." The appeal confronts us with the question, left open in United States v. Lauer, 148 F.3d 766, 768-70 (7th Cir. 1998), of the guideline’s validity.

The guideline was promulgated in response to section 2507(a) of the Crime Control Act of 1990, Pub. L. 101-647, 104 Stat. 4789, 4862, which commands increased punishment of persons who derive more than $1 million in gross receipts from mail fraud affecting "a financial institution (as defined in Section 20 of title 18, United States Code)." The list of financial institutions in section 20 does not include, either explicitly or implicitly, pension funds, yet the only "financial institution" affected by Tomasino’s fraud was a pension fund. However, the Sentencing Commission’s authority to promulgate sentencing guidelines is not limited to ones that interpret or apply a statute. On the contrary, the Commission’s major task is that of picking sentencing ranges within statutory minimum and maximum limits--a legislative rather than an interpretive task. So even though Congress did not command the Commission to increase the punishment of people who commit a mail fraud that affects a pension fund and yields the malefactor more than $1 million, the Commission was free to legislate such an increase (within the statutory maximum) by means of a guideline. United States v. Rutherford, 54 F.3d 370, 374 n. 11 (7th Cir. 1995). The question is whether in including pension funds in the punishment-increasing guideline the Commission was exercising its legislative authority or merely misreading section 2407(a) by overlooking the cross- reference to section 20. In the latter event, the guideline is invalid as applied to pension funds. See United States v. Lightbourn, 115 F.3d 291, 292-93 (5th Cir. 1997), where the general point is clearly stated.

There is no doubt that the guideline embraces pension funds. Application Note 14 (now 16) to guideline section 2F1.1 says so, and the application notes are part of the guidelines themselves, and not mere commentary on them. Stinson v. United States, 508 U.S. 36, 38 (1993); United States v. Joseph, 50 F.3d 401, 402-03 (7th Cir. 1995). But a background note to subsection (b)(7)(B), the subsection under which the government sought to increase Tomasino’s sentence, says that the subsection "implements the instruction to the Commission in Section 2507" of the Crime Control Act. Standing alone, this implies that the Commission thought erroneously that section 2507 requires the Commission to treat a pension fund as a financial institution within the meaning of the section. If this is what it thought, it presumably never considered whether such a guideline would be a good idea if not commanded by Congress.

History may cast some light on the question. A year before subsection (b)(7)(B) was promulgated, the Commission had added a similar subsection to section 2F1.1-- (b)(7)(A)--in response to section 961(m) of the Financial Institutions Reform, Recovery, and Enforcement Act of 1989 (FIRREA), Pub. L. 101-73, 103 Stat. 183, 501, which had similarly required the Commission to up the punishment of defrauders of financial institutions. That statute, however, was limited to "federally insured financial institutions," a term that, appearing as it did in a statute concerned with the savings and loan crisis, was unlikely to be thought to include pension funds. By Application Note 14 the Commission had made the guideline that it promulgated in response to section 961(m) of FIRREA include pension funds, but the background note had been explicit that in doing so the Commission was "implement[ing], in a broader form, the instruction to the Commission in section 961(m)" (emphasis added). It was clear, therefore, that the Commission believed that it was exercising its legislative power rather than merely interpreting the statute. When one year later the Commission added (b)(7)(B) to the guideline in response to the Crime Control Act, it did not change Application Note 14, and so (B) like (A) includes pension funds. But the background note to (B) omits "in a broader form."

This omission could have been inadvertent, as the government argues. Another possibility, however, is that the Commission overlooked the cross-reference to 18 U.S.C. sec. 20, the provision which makes clear that section 2507 does not cover pension funds. The Commission was less likely to have made such a mistake with reference to FIRREA, the scope of which would have seemed clearer without looking up the definition of "federally insured financial institution."

The question is not whether the application note takes precedence over the background note, or vice versa. Stinson v. United States, supra, 508 U.S. at 43. There is no conflict. Unquestionably the Commission wanted Tomasino punished more heavily than he was, and this intention is expressed in both notes. The question is whether the Commission wanted this because it thought Congress had ordered it to want this, as it were, or because it had made its own independent legislative judgment--prompted by section 2507, but not commanded by it--that frauds affecting pension funds should be punished as severely as frauds affecting the financial institutions enumerated in 18 U.S.C. sec. 20. That is a question on which the application note casts no light. The only clue to the Commission’s thinking on whether it was engaged in legislation or interpretation is the background note. Read literally, the note says that the application note is interpretive ("implements the instruction to the Commission in Section 2507" of the Crime Control Act), not legislative. A textualist would stop there. And the history behind the note does nothing to dispel the impression that a literal reading creates.

If the Commission had not published a background note to the new subsection, the natural inference would be that it was doing just what it had done a year before--assimilating pension funds to financial institutions as a matter of legislative judgment. The Commission is not required to publish background notes, and in their absence courts will assume that the Commission knows when it is exercising a legislative judgment, especially since that is what it usually does in issuing guidelines. A presumption of regularity attends the Commission’s doings, as it does that of other official bodies. But if a background note said, "This is a dumb guideline we’re promulgating, and we’re doing it only because Congress has commanded us to," and this was wrong--Congress had not commanded, but merely authorized--the courts could not overlook the error. Discretion is abused when the decision maker erroneously thinks that he is compelled to make a particular decision and so fails to exercise judgment. The fact that the error might be concealed does not justify overlooking it when it is disclosed, blatantly in our example. And, although the Commission has like other administrative agencies a discretionary power that is legislative in character, we do not think the background note can be dismissed as mere "legislative history," or a demand for clarification as a demand for "subsequent legislative history," both being aids to interpretation that are derided in some influential quarters. Lacking the democratic legitimacy of legislatures, agencies do not have the same freedom to base decisions on arbitrary grounds, let alone on misunderstandings of their own powers. The issue here, moreover, is not whether the Sentencing Commission was legislating wisely, but whether it was legislating at all; if it thought it was interpreting, as the background note suggests, it was acting ultra vires, purporting to carry out a congressional command that had never been given.

Despite what we have said, it is of course possible that the Commission dropped the words "in a broader form" inadvertently. We are not so devoted to literalism in interpretation that we are unwilling to consider such a possibility. But before jacking up a defendant’s sentence on the basis of a legislative determination by the Sentencing Commission, a court should have at least minimal confidence, here lacking, that the Commission was not simply misinterpreting a statute. If we do not know whether the Commission was exercising its legislative judgment, we do not know whether Tomasino was lawfully sentenced.

Any risk of underpunishing Tomasino can be avoided by the district court’s deferring, on remand, resentencing Tomasino until the Sentencing Commission has had a reasonable opportunity to clarify its understanding in issuing the pension-fund guideline under which he was sentenced. (The needed clarification, of course, is not of the application note itself, which is pellucid, but of the background note.) Although the membership of the Commission has turned over since the guideline was promulgated, we assume that like any administrative agency it has an institutional memory that enables it to clarify an action that a court cannot evaluate without clarification. A clarifying guideline can lawfully be applied retroactively, U.S.S.G. sec. 1.B1.11(b)(2); United States v. Goudy, 78 F.3d 309, 314-15 (7th Cir. 1996); United States v. Patasnik, 89 F.3d 63, 70 (2d Cir. 1996), as distinct from one that increases the defendant’s punishment. United States v. Bullis, 77 F.3d 1553, 1561 (7th Cir. 1996). But we do not think the formality of issuing a new guideline, whether as a guideline formally so denoted, an application note, or even a background note, is required here. The application note under which Tomasino was sentenced gave him clear notice of the Commission’s intention that he be sentenced within the range used by the judge in sentencing him, and so there is no issue of retroactivity. The background note does not prescribe punishment, and the question of its meaning is unrelated to retroactivity. The question whether Tomasino was sentenced under a valid guideline is the question that requires clarification.

We vacate Tomasino’s sentence and remand the case to the district court to resentence him after giving the Commission a reasonable opportunity to clarify its intentions in promulgating U.S.S.G. sec. 2F1.1(b)(7)(B). Vacated and Remanded, with Directions.

Easterbrook, Circuit Judge, dissenting. Our case arises from the interaction of decisions by two policy-making bodies. First, there is a decision of the Sentencing Commission that fraud affecting a bank should be treated identically to fraud affecting a pension fund. Second, there are decisions of Congress in 1989 and 1990 that change the penalties for persons who defraud banks. When implementing the first statute, the Commission adopted parity between bank fraud and pension-fund fraud. When responding to the second statute, the Commission amended sec.2F1.1(b) and parts of sec.2B1.1 and sec.2B4.1, each already subject to a declaration that frauds affecting pension funds are treated the same as frauds affecting banks. It is easy for us to respect both the Commission’s decision and Congress’s. Yet my colleagues dishonor one of these decisions--the decision by the Sentencing Commission to treat pension-fund fraud the same as bank fraud--on the ground that the Commissioners might have misunderstood their options in 1991. This elevates speculation about the Commissioners’ mentation over the Commission’s deeds. Normally a court will say that clear legislative history cannot contradict a clear statutory text; here my colleagues say that an unclear administrative history prevails over a pellucid text.

If this case depends on whether the Sentencing Commission’s Background Note to the 1991 amendment fully reveals the Commission’s understanding of its options, then Tomasino has been sentenced improperly. The Background Note does not show whether the Commission understood its discretion to limit the higher sentences for fraud affecting "financial institutions" to the entities listed in 18 U.S.C. sec.20. But the sentence ought not depend on the extent to which the Commission discussed the options it rejected, because we know which one it embraced. Application Note 16 (formerly 14) to U.S.S.G. sec.2F1.1(b)(7)(B) provides:

"Financial institution," as used in this guideline, is defined to include any institution described in 18 U.S.C. sec.sec. 20, 656, 657, 1005-1007, and 1014; any state or foreign bank, trust company, credit union, insurance company, investment company, mutual fund, savings (building and loan) association, union or employee pension fund; any health, medical or hospital insurance association; brokers and dealers registered, or required to be registered, with the Securities and Exchange Commission; futures commodity merchants and commodity pool operators registered, or required to be registered, with the Commodity Futures Trading Commission; and any similar entity, whether or not insured by the federal government. "Union or employee pension fund" and "any health, medical, or hospital insurance association," as used above, primarily include large pension funds that serve many individuals (e.g., pension funds of large national and international organizations, unions, and corporations doing substantial interstate business), and associations that undertake to provide pension, disability, or other benefits (e.g., medical or hospitalization insurance) to large numbers of persons.

Because Application Notes are treated as part of the Guidelines, Stinson v. United States, 508 U.S. 36 (1993), the rule governing Tomasino’s sentence is explicit: he committed a fraud from which he derived more than $1 million in gross receipts; one of his victims was a pension fund; thus under sec.2F1.1(b)(7)(B) his offense level must be increased. Tomasino was sentenced to 31 months in prison, but under sec.2F1.1(b)(7)(B) his range is 51-63 months. "Unquestionably the Commission wanted Tomasino punished more heavily than he was" (majority op. 4). Because the Application Note is clear, our job is over-- there is no ambiguous text requiring interpretation--unless the Application Note is invalid. If the Commission were an ordinary agency, the Application Note, and thus sec.2F1.1(b)(7)(B), might be deemed invalid under the Administrative Procedure Act, because the Commission did not adequately explain why it decided to jack up the punishment of persons who defraud pension funds, when Congress required an increase only for certain crimes that have banks as victims. The Commission did not do this in 1991, when it adopted sec.2F1.1(b)(7)(B) as part of Amendment 364, or for that matter in 1990, when as part of Amendment 317 it promulgated what is now sec.2F1.1(b)(7)(A) and the Application Note defining pension funds as "financial institutions". All the Commission said in 1990 is that it was implementing "in a broader form" a legislative command to increase penalties for jeopardizing banks’ financial soundness. No court would let the Federal Energy Regulatory Commission adopt a rule with such elliptic explanation. But the portion of the APA that authorizes judicial review of the adequacy of agencies’ explanations does not apply to the Sentencing Commission. It must follow some of the APA’s procedures, see 28 U.S.C. sec.994(x), incorporating 5 U.S.C. sec.553, but its substantive decisions are not subject to review under the APA’s standards. The 1991 amendment therefore cannot be deemed invalid as inadequately explained.

My colleagues do not hold the Commission to a minimum standard of explanation. They would enforce the Application Note as written if the Commission had kept silent. But because it spoke, and because the 1990 Background Note ("Subsection (b)(7)(A) implements, in a broader form, the instruction to the Commission in section 961(m) of Public Law 101-73.") differs from the 1991 Background Note ("Subsection (b)(7)(B) implements the instruction to the Commission in section 2507 of Public Law 101-647."), my colleagues wonder whether omission of "in a broader form" implies that the Commission misunderstood its obligation under sec.2507 of Pub. L. 101-647. Puzzlement leads to a refusal to enforce the guideline and its note unless within a "reasonable" time (whatever that may mean) the Commission clarifies its "understanding" of the issue. Slip op. 6.

I agree in principle with the proposition that a Background Note could show that the Guideline is invalid, for the same reason that the FERC’s explanation of a regulation in the Federal Register could show that its rule is irrational. But for reasons discussed later I do not think that the omission of "in a broader form" demonstrates a misunderstanding of the legal rules governing the amendment of the rules in 1991. What we have is not a clear legal error but uncertainty. How should a court deal with an ambiguous background note? My colleagues’ approach, a form of "remand" in spirit to the Commission (but without any actual remand), is unprecedented and invites evasion of the Commission’s enabling legislation. The unprecedented part is clear enough; our decision is the first to hold that a sentence must be kept in limbo pending new action by the Commission. The Commission, which had no members between October 1998 and the end of November 1999, is just beginning to address a substantial backlog of old business and cannot attend to our demands on short notice. Why should my colleagues (or the Commissioners) think that this issue is entitled to priority over other matters? Yet even if it had nothing else to do the Commission could not revise guidelines with dispatch. Proposals must be published for comment under sec.994(x); after receiving comments the Commission must promulgate a text and send it to Congress; the change does not take effect for at least 180 days, during which Congress may legislate. 28 U.S.C. sec.994(p). The Commission generally promulgates amendments in April and makes them effective the next November 1. No proposal added to the Commission’s agenda today could go into force until November 1, 2000, no matter how abbreviated the schedule for publication, public comment, and promulgation. November 1, 2001, would be a more realistic target. By inviting the Commission to issue a new background note (or to clarify matters in some other, unspecified, way, perhaps by a press release) rather than to reissue the guideline or an application note, my colleagues invite evasion of these time limits. I am not aware of any prior effort by the Commission to change the legal effect of a guideline--and that is what the majority contemplates--without using the processes of sec.994. If the Commission rather than a court came up with this back door method of changing the rules, I doubt that we would permit the gimmick to work.

What principally concerns me, however, is interpretive method. Congress has specified a means to clear up ambiguities and prevent unjust applications of the Guidelines. The Commission may direct the retroactive application of amendments. 18 U.S.C. sec.3582(c)(2). The right thing for a court to do when it is unsure whether the Commissioners understood all of their options is to apply the text as written and to draw the subject to the Commission’s attention. Compare Chapman v. United States, 500 U.S. 453 (1991) (applying the LSD guideline even though it is harsh and can work erratically), with Neal v. United States, 516 U.S. 284 (1996) (noting that the Commission later amended the LSD guideline). If the Commission believes that a guideline has been applied in an inappropriate way, it may change the rules and apply that amendment to all affected persons. Even relatively short sentences such as Tomasino’s are long enough in relation to the amending cycle that the defendant may receive the benefit. Thus the right thing for us to do is remand with instructions to resentence Tomasino within the range prescribed by sec.2F1.1(b)(7)(B) and send a copy of the opinion to the Commission. If the Commission concludes that pension funds should be excluded from the application of sec.2F1.1(b)(7)(B), it can put that change into effect by November 1, 2001, and Tomasino will be eligible for release after serving 31 months (the right sentence without the extra levels). My colleagues’ approach, by contrast, spurns the method Congress provided while inviting violation of sec.994.

Unless the Commission uses sec.3582(c)(2), Tomasino should serve a sentence within the guideline range. By ruling otherwise my colleagues curtail the effect not only of sec.2F1.1(b)(7) but also of sec.2B1.1 and sec.2B4.1, which were altered in the same way and at the same time by Amendments 317 and 364. A Background Note is a form of administrative history. Just as legislative committees sometimes explain why they enacted the law, and what they expect it to accomplish, the Commission sometimes explains why it amended the guidelines. An explanation from either Congress or the Commission may suggest that the matter has not been thought through fully. Does this authorize either refusal to enforce a rule or "construction" of a text to confine it to the situation that the judges deem well thought out? No. We have it on the highest authority that when a legal text is unambiguous a court must disregard its background. Robinson v. Shell Oil Co., 519 U.S. 337, 340 (1997); United States v. Ron Pair Enterprises, Inc., 489 U.S. 235, 241 (1989); Burlington Northern R.R. v. Oklahoma Tax Commission, 481 U.S. 454, 461 (1987); United States v. Locke, 471 U.S. 84, 95-96 (1985); Rubin v. United States, 449 U.S. 424, 430 (1981); Packard Motor Car Co. v. NLRB, 330 U.S. 485, 492 (1947); United States v. Shreveport Grain & Elevator Co., 287 U.S. 77, 83-84 (1932); Wisconsin R.R. Commission v. Chicago, Burlington & Quincy R.R., 257 U.S. 563, 588-89 (1922); Caminetti v. United States, 242 U.S. 470, 490 (1917). See also In re Sinclair, 870 F.2d 1340 (7th Cir. 1989). The text of the Application Note could not be clearer; there is no excuse for resort to the legislative history. A textualist does not "stop" with the language of this Background Note (majority op. at 4) because a textualist does not reach the language of this Background Note! (Not, that is, as an interpretive tool.)

If the 1991 Background Note is legislative history, then what the Commissioners might say in 2000 or 2001 about this Background Note (or about its "understanding" of its options) is subsequent legislative history, which is even less useful. Pierce v. Underwood, 487 U.S. 552, 566-68 (1988). At least contemporaneous legislative history may illuminate the social or political context that produced the text being construed. Yet no member of the Sentencing Commission now serving has been there for more than four months. Only one has any way to retrieve what members serving in 1991 "intended." (Commissioner Steer was the Commission’s General Counsel in 1991 and may be able to consult his memory or files on the subject.) We would be better off asking the members from a decade ago to file affidavits narrating what they (or their staffs) thought about the matter. But that approach would be improper for too many reasons to count; the ban on scrutinizing public officials’ mental processes, see United States v. Morgan, 313 U.S. 409, 421-22 (1941), is only the most prominent. Thus if "intent" matters, it is too late to do reconstruction work. If the Guideline and Application Note were not valid when Tomasino committed his crime, they cannot be resuscitated retroactively, whether by statements of "understanding" or by new rulemaking. Bowen v. Georgetown University Hospital, 488 U.S. 204, 208-09 (1988). But if they were valid when issued, no new declarations are called for. A declaration by the sitting Commissioners that they think in 2000 that their predecessors thought a particular thing in 1991 is irrelevant. West Virginia University Hospitals, Inc. v. Casey, 499 U.S. 83, 100-01 & n.7 (1991). When a legislative body amends a rule, its declarations tell us only what the new text means. Lawmakers do not "construe" or "interpret" the work of their predecessors in office. Rivers v. Roadway Express, Inc., 511 U.S. 298, 304-13 (1994); Mojica v. Gannett Co., 7 F.3d 552, 562-64 (7th Cir. 1993) (concurring opinion).

Actually, nothing in the 1991 Background Note implies that the Commissioners misunderstood their marching orders. The Commission said that the 1991 amendment "implements the instruction to the Commission in section 2507 of Public Law 101- 647." That it does. Amendment 364 also increases the penalty for people who receive more than $1 million in gross receipts of fraud, and have a pension plan among their victims. The Commission did not add "and we really mean it!", but even fanciers of legislative history don’t require that of a lawmaker. See Harrison v. PPG Industries, Inc., 446 U.S. 578, 592 (1980); Swain v. Pressley, 430 U.S. 372, 378-79 (1977). Unlike my colleagues, I do not find it odd or jarring that in 1990 the Commission said that it was implementing a statutory directive "in a broader form" yet omitted that phrase in 1991. The reason for the difference is that Amendment 317, the 1990 change, added Application Note 14 (now 16) to sec.2F1.1 (and added identical notes to sec.2B1.1 and sec.2B4.1). The Commission had to decide in 1990 whether to treat pension funds like banks. It considered the subject and answered "yes." Having taken that decision in 1990, the Commission did not have to remark on it again in 1991.

If we are to give any weight to Commissioners’ unrevealed thoughts, plans, intent, or general mental state in 1991--an inquiry that I think inappropriate for reasons already given--we should ask these homunculi the right question. It is whether the Commissioners believed in 1991 (or would have concluded, had they given the subject any thought) that the sentences for persons who defraud pension plans should be lower than the sentences for persons who defraud banks, holding constant the size of the fraud. That is, would the need to comply with sec.2507 of Pub. L. 101- 647 have led the Commission to reverse its decision, made a year earlier, to treat the two alike for sentencing purposes? Nary a clue suggests that any Commissioner thought, or could under truth serum have been induced to say, that the sentences for those who defraud pension plans should be lower. Every mention of financial institutions everywhere in the Sentencing Guidelines treats the two alike. Similarly, Chapter 8 of the Guidelines groups pension funds with other organizations (including banks) for the purpose of organizational sentencing. U.S.S.G. sec.8A1.1 Application Note 1. Some guidelines apply exclusively to offenses against banks, and others exclusively to offenses against pension funds (e.g., U.S.S.G. sec.2E5.1), but wherever the victim of the crime is a generic "financial institution" the Guidelines treat banks and pension funds identically. (Only sec.2B3.1(b)(1), which uses the term "financial institution" without definition, is a possible counterexample; but still there is no hint that the Commission provided a lower punishment for persons who defraud pension funds.) The Commission’s failure to add the words "in a broader form" to a Background Note in 1991 does not imply that this equation should be jettisoned. To sentence Tomasino, all the court need know is whether persons who defraud pension funds properly receive lower sentences than persons who defraud banks. The Sentencing Guidelines answer in the negative.

The Commission did not say anything like "this is a dumb guideline that we’re promulgating only because Congress held a gun to our head." That proposition, if uttered, would indeed demonstrate that the amendments made in 1991 are invalid. My colleagues proceed as if the Commission had written something of the sort, but it did not. All the Background Note does is relate a matter of fact: Amendment 364 implements sec.2507 of Pub. L. 101-647. The majority errs in writing that the Background Note "interprets" sec.2507; what it says, rather, is that the amendments "implement" the law, which is true (and not ambiguous). We then must choose what to make of the omission of the words "in a broader form". Should we assume that the difference in language between the 1990 and 1991 Background Notes is attributable to the fact that the fundamental decision ("do we equate banks with pension funds?") had been made in 1990 rather than 1991? Or should we assume that the 1991 amendment reflects incompetence--that the Commissioners issued Amendment 364 without bothering to read either the cross-reference in sec.2507 of Pub. L. 101-647 or the Application Notes they had adopted the year before? My colleagues say that because the Background Note is ambiguous, we should adopt the reading that treats the Commissioners and their staff as bumblers. Yet there is a longstanding norm of administrative law that when the record is silent or ambiguous, we treat the agency as competent. A presumption of regularity, judges call it. Like other presumptions, it can be upset by contrary indicators, but here there are none. The Background Note stands alone. If the Sentencing Commission were an informal body composed of lay members like a draft board under the old Selective Service System, then maybe we should assume that silence implies legal error-- though judges did not assume this even about draft boards. United States v. Neckels, 451 F.2d 709, 712 (9th Cir. 1971); United States v. Harris, 436 F.2d 775, 777 (9th Cir. 1970). But the Commission is part of the Judicial Branch of government. 28 U.S.C. sec.991(a); Mistretta v. United States, 488 U.S. 361, 384-97 (1989). At least three of its seven members must be federal judges, and the Commission also has a large legal staff. The Commissioners who promulgated Amendment 364 included two circuit judges, two district judges, and one professor of law./* To suppose that none of them looked up the cross-reference in sec.2507, that the legal staff did not alert them to its significance, that as a result the Commissioners wrongly thought that they had been compelled to increase the sentences of persons who defraud pension funds, and that but for this legal error the Commission would have reversed the 1990 decision mandating identical sentences for frauds committed against pension funds and banks, is fantastic. Doubtless a presumption of regularity sometimes looks implausible. But here a presumption of regularity is fitting, and the contrary presumption imputes foolishness (or extreme carelessness) to too many intelligent people.

In the end, my colleagues depart from what the Commission adopted because they are not sure what it thought. We ought, instead, enforce its rules. Deeds prevail over thoughts (or the lack of thoughts). Section 2F1.1(b)(7)(B) and its Application Note cover Tomasino’s conduct directly. There is no ambiguity. We should apply the Sentencing Guidelines as written, and leave to the Commission under sec.3582(c)(2) the decision whether to reduce the sentences of persons in Tomasino’s position.

/* The Chairman of the Commission in 1991 was Circuit Judge William W. Wilkins. The other members were Julie E. Carnes (appointed to the district court in 1992), Helen G. Corrothers, Michael S. Gelacak, Circuit Judge George E. MacKinnon, District Judge A. David Mazzone, and Professor Ilene H. Nagel. The General Counsel was John R. Steer, who was appointed to the Commission as a member in November 1999. Carol Pavilack Getty and Paul L. Maloney served as members ex officio.