Clary William Wester was formerly president, chairman of the board, and chief executive officer of First Service Bank for Savings ("First Service"), a federally insured bank in Leominster, Massachusetts. In the late 1980s, Wester arranged various transactions at First Service, including a series of loans by First Service to Webster's partners in a separate real estate venture, made with the understanding that the partners would use the loaned funds to buy out Wester's interest in the partnership. At trial, Wester was convicted by a jury of several different crimes. He now challenges the jury instructions and two adjustments to his sentence.

Although Wester does not directly dispute the sufficiency of the evidence, one of his claims as to jury instructions can be taken to raise the issue of sufficiency indirectly. For that reason, we begin by describing what the evidence would have permitted the jury to find. A reviewing court's perspective on the evidence depends on the claim of error being considered, and for a sufficiency claim, we take the evidence most favorable to the verdict. E.g., United States v. Dodd, 43 F.3d 759, 760-61 (1st Cir.1995).

In June 1986, Wester formed a partnership with three other men to construct a condominium project in Manchester, New Hampshire. The three others were Robert Fredo, senior vice president at First Service, Robert George, a developer, and Charles Morgan, a broker. Wester and Fredo supplied start-up money, and Wester helped arrange a $12.4 million loan organized by New England Financial Resources, Inc. ("NEFR"), a commercial real estate lender not affiliated with First Service. The loan was secured by land and future improvements and a personal guaranty of the debt from each of the partners.

Since Morgan had been a frequent borrower at First Service, NEFR was concerned that Wester's and Fredo's participation in the partnership might create conflicts of interest. As a condition of the loan NEFR required a certificate from First Service acknowledging that Wester and Fredo had disclosed their interest in the project to the board of directors of First Service. The certificate issued by First Service stated, inter alia, that the bank "was not involved in the financing of this project and would not, without specific prior approval, grant any additional loans to Messrs. Morgan or George."

In the fall of 1986, Morgan and George proposed another condominium project, this one in Massachusetts. Wester suggested that First Service participate in the project as a joint venturer, but said that he and Fredo would need to divest their interests in the earlier partnership. The four men agreed that Wester and Fredo would sell their interests to George and Morgan for $425,000 each, and be reimbursed for additional start-up money they had provided, all to be paid from future profits from the New Hampshire project. Before the details of the buyout plan had been resolved, First Service (through a subsidiary) joined the new project with Morgan and George, and the bank provided a $5 million loan to the venture.

By June 1987, the New Hampshire project had yet to begin earning profits. Wester grew impatient and told George and Morgan that he wanted his buyout payments. When George said this was not feasible because of cash flow problems, Wester offered to provide First Service loans to George and Morgan to fund the buyout payments. These loans, and the resulting buyout payments, became the basis for most of the later charges against Wester.

On June 12, 1987, George signed two promissory notes for unsecured loans by First Service totalling $200,000. That same day, George paid Wester and Fredo $100,000 each. Morgan received a $300,000 loan from First Service on June 12; several days later he paid Wester and Fredo $25,000 each, and George and Morgan (through the partnership) gave Wester and Fredo $250,000 for the start-up money previously contributed.

Neither Wester nor Fredo disclosed the true purpose of these loans to First Service's loan review committee, executive committee, or board of directors.1 Nor did the supporting documentation reveal that the loaned funds were being used to fund the buyout. In one instance, the loan set-up sheets stated that the purpose of the loan was to "finance acquisition of real property"; in other instances no purpose for the loan was provided. The jury could have found that the failure to disclose the purpose of the loans to the loan review committee was material, deliberate, and dishonest.

This process was repeated several times over in the following months, with Wester and Fredo arranging loans or letters of credit to Morgan, George or entities they controlled--and in one instance George's father--with portions of the proceeds returned to Wester and Fredo to satisfy the buyout. The last such loan was made on March 9, 1988. On March 10, 1988, the buyout agreement was executed, and Wester's and Fredo's interests in the partnership were terminated "retroactive" to January 1, 1987.

There was one more wrinkle of considerable importance. The buyout agreement included a provision for releasing Wester and Fredo from their personal guaranties on the earlier $12.4 million loan from NEFR. NEFR, however, was concerned about the financial health of the New Hampshire condominium project. It made clear that it would only consent to the release if the partnership obtained a $2.3 million bank loan or line of credit to provide additional security for the $12.4 million loan.

Ultimately, Wester and Fredo arranged a $2.3 million loan by First Service for the partnership, without any disclosure to other bank officials of the connection to the proposed release and without approval by First Service's executive committee or board of directors. Under the bank's rules, approval by the former was evidently required because of the size of the loan. This loan and the promised release were each specified as offenses in the subsequent indictment.

After a portion of the $2.3 million was disbursed to the partnership, and before NEFR formally executed Wester's and Fredo's releases from the guaranties, the FDIC began investigating the goings-on at First Service. Wester and Fredo were subsequently fired. First Service honored its commitment to the partnership and released the balance of the $2.3 million loan proceeds. NEFR never executed the releases, but neither did it call upon Wester or Fredo to pay based on their guaranties.

On August 11, 1990, Wester and Fredo were named in a 22-count federal indictment charging them primarily with conspiracy, 18 U.S.C. § 371, misapplication of bank funds, 18 U.S.C. § 656, and bank bribery, i.e., the soliciting or receiving of bribes or rewards for the making of the loans, 18 U.S.C. § 215. The loans for the buyout payments and for the release were charged as misapplications under section 656; the payments and promised releases were charged as bribes or rewards under section 215.

Fredo pled guilty before trial to one count each of conspiracy, misapplication, and bank bribery, and testified for the government at Wester's trial. After a 13-day jury trial in July 1994, Wester was convicted of one count of conspiracy, five counts of misapplication, and six counts of bank bribery. He was acquitted of one count each of misapplication and bank bribery, and two tax evasion counts. In December 1994, Wester was sentenced to 46 months in prison. This appeal followed. For the reasons that follow, we affirm the convictions but remand for resentencing.

1. Wester's main challenge to the trial proceedings concerns the district court's jury instructions on the misapplication counts. In relevant part, 18 U.S.C. § 656 provides criminal penalties for "an officer, director, agent, or employee of ... national bank or insured bank ... [who] willfully misapplies any of the moneys, funds or credits of such bank." Formally, the dispute on appeal centers around the phrase "willfully misapplies"; in reality, Wester's argument also presents the question of whether the evidence was adequate.

The problem that has confronted and perplexed the courts is that there is no statutory definition or common law heritage that gives content to the phrase "willfully misapplies." United States v. Gens, 493 F.2d 216, 221 (1st Cir.1974). And to focus simply on the deprivation of property is hardly much help since it is a purpose of banks to lend money. In response, the case law has developed two notions that help to clarify and delimit the statute--one relating primarily to conduct and the other to intent.

First, "misapplication" has been taken by most courts to mean "wrongful" use of the bank's moneys. See 1 Sand, et al., Modern Federal Jury Instructions § 24.01 (1995). And, second, the courts have uniformly read back into the statute an earlier requirement, removed by a careless revisor, that the defendant have intended "to injure or defraud" the bank. E.g., United States v. Angelos, 763 F.2d 859, 861 (7th Cir.1985). Of course, the same facts can easily be the basis for deeming the conduct to be wrongful and the intent fraudulent; but both misapplication and scienter are required.

In this case, the district court's affirmative charge describing the offense of misapplication was for the most part conventional. What Wester objects to on appeal is the court's refusal to give certain additional language specifically requested by Wester. The language--which Wester believes to have been required by our decision in Gens--appears at two different points in Wester's requested instruction no. 37:

I instruct you that a loan to a financially capable person who fully understands that it is his responsibility to repay the loan does not constitute misapplication, even if the bank officer involved with the loan receives proceeds of the loan, or some other benefit. Thus, in this case, with respect to the loans charged, if the debtors were financially capable of repaying the loans and that [sic] they understood that it was their responsibility to repay the loans, Mr. Wester must be acquitted on those counts irrespective of whether or not he received proceeds, or any other benefit, from those loans....

. . . . .

Therefore, in this case, for each loan alleged in the Indictment as a misapplication of bank funds, if the named debtor was financially capable of repaying the loan and recognized his responsibility to repay the loan, there is no misapplication as a matter of law, even if proceeds of those loans or some other benefits were received by Mr. Wester.

Needless to say, such language would have been very useful to Wester. The government, it appears, did not try to show that any of the designated borrowers in the misapplication counts (E.g., George and Morgan) were fictitious or financially irresponsible or had never assumed liability for the loan. Wester suggests that for bona fide loans to financially responsible borrowers, there is no serious risk of harm to the bank and therefore, even apart from the authority of Gens, no reason to apply the statute.

Wester's position is far from absurd, cf. United States v. Docherty, 468 F.2d 989 (2d Cir.1972), but in the end it reads the statute too narrowly. There is no indication that insider loans are inflexibly forbidden by federal law, but they obviously create a special set of dangers. At least one danger--quis custodiet ipsos custodes--is that the insider who approves or fosters the loan may do so too readily if he himself benefits by it. Controls on such loans, including authorizations and disclosures, are therefore pertinent to the safety of the bank.

Further, financial responsibility on the part of the borrower is not an absolute but a matter of degree. To say that the nominal borrower is at the outset financially capable of repayment hardly proves that the bank would have made the loan if it had been fully apprised of the risks and circumstances. Here, two members of First Service's executive committee testified that they would not have approved the loans if Wester had disclosed that the proceeds were going to fund the buyout of Wester and Fredo's interests.

In this instance, the jury could reasonably have found that Wester caused the loans to be made for his own benefit without obtaining approvals from the executive committee or board of directors required under the bank's own rules (as to the largest $2.3 million loan) and (as to it and all others) because he deliberately suppressed or withheld information that the purpose of the loans was one that the bank would not have approved.2 This wrongful conduct permitted the jury in turn to find that Wester had engaged in the "misapplication" of bank funds. All that remained was to find scienter.

The scienter requirement--an intent to injure or defraud--is stated in the alternative. In the Supreme Court's classic summary, "the words 'to defraud' commonly refer 'to wronging one in his property rights by dishonest means or schemes,' and 'usually signify the deprivation of something of value by trick, deceit, chicane or overreaching.' " McNally v. United States, 483 U.S. 350, 358, 107 S. Ct. 2875, 2881, 97 L. Ed. 2d 292 (1987) (citation omitted). Whether or not Wester intended to injure the bank, a jury could properly find that he intended to "defraud" the bank by causing it through consciously dishonest means to part with its property for his own benefit.

This brings us to Gens. In that case, the defendant Gens, a director of the bank, had persuaded others (e.g., one of his friends) to borrow from the bank and to transfer the funds to him; and bank officers working with Gens had approved the loans knowing that he would obtain use of the funds. As this court read the trial court's charge, it told the jury that misapplication had occurred "if it was found that [the officers] granted loans to the [nominal borrowers] knowing that the proceeds would be turned over to Gens." 493 F.2d at 221. The jury convicted Gens and he appealed.

On appeal in Gens, this court rejected the government's broad notion that "willful misapplication occurs whenever bank officials grant loans to parties with the knowledge that the proceeds will go to a third party." 493 F.2d at 223 (emphasis added). Our opinion pointed out that most of the pertinent cases under the misapplication statute involved loans to borrowers who were fictitious, unwitting, irresponsible or had not assumed liability. The contrary was so in Gens, except arguably as to one borrower; and the count as to that borrower was remanded for a new trial under new instructions.

Gens held that the government's "whenever" theory was overbroad but Gens did not bar a misapplication charge in every case where the straw happened to be a financially responsible borrower. We so noted in United States v. Brennan, 994 F.2d 918 (1st Cir.1993). There we said that a misapplication charge could be made out where a bank officer made loans to named debtors knowing that the proceeds would go to a third party and where the surrounding circumstances involved dishonesty (e.g., false entries in the bank's records). Id. at 923-24.

Wester's requested instruction 37 was thus not warranted by Gens because it would have converted a circumstance in Gens--financial responsibility of the borrower--into an automatic defense requiring acquittal regardless of other evidence of dishonesty. Misapplication and intent to defraud turn largely on the facts; the facts here were enough to convict; and the requested instruction was overbroad and was properly denied.

2. Wester's other complaint about the jury charge concerns the district court's instruction based on Pinkerton v. United States, 328 U.S. 640, 66 S. Ct. 1180, 90 L. Ed. 1489 (1946), that a conspirator may be accountable for actions of co-conspirators taken in furtherance of the conspiracy. Wester was charged with a conspiracy that had as its objects misapplication and bank bribery. Wester claims that the district court's Pinkerton instruction was mistaken in two respects.

First, Wester argues that the Pinkerton instruction allowed the jury to find him vicariously liable for the substantive crimes of a co-conspirator even if those crimes were not the object of the conspiracy or in furtherance of it. For example, he says that the jury could have found that Wester was guilty of the substantive crime of bank bribery because one of his co-conspirators committed that offense; yet the jury could have found that the conspiracy's object was limited to misapplication.

One wonders if Wester carefully read the transcript of the jury charge before making this argument. After correctly describing the other elements of the Pinkerton doctrine, the district court stated that the jury must find "that the substantive crime (attributed to the defendant vicariously) was committed pursuant to the common plan and understanding you found to exist among the conspirators," and that "the defendant could have reasonably foreseen that the substantive crime might be committed by his co-conspirator." In short, the instruction itself answers Wester's hypothetical.

Second, Wester argues that it was inappropriate to give a Pinkerton instruction at all, because "where there is evidence of various substantive offenses ... it raises the risk that the jury will resort to the inverse of Pinkerton and infer the existence of the conspiracy from the series of substantive criminal offenses." He says this risk was especially high here because the government concentrated its efforts on proving only the substantive charges. We agree neither with the premise nor the conclusion.

In a case like this one, some interplay between the jury's assessment of guilt on the substantive counts and the conspiracy charge is both natural and appropriate. Indeed, the fact that substantive crimes were carried out by the defendants, following discussions between them, may well make the fact of agreement more likely. Rossetti v. Curran, 80 F.3d 1, 5 (1st Cir.1996). This is so whether or not a Pinkerton charge is given; the charge is at most an added complication for the jury but one well within its ken.

Here, the government offered ample evidence of discussions between the four partners that provided a firm basis for the conspiracy charge. There were pages of testimony concerning the meetings among Wester, Fredo, George and Morgan, that led to the various loans and the payments back to Wester. This testimony provided grounds for the jury to find that Wester participated in the charged conspiracy. And, because Wester argued that he was unaware of many acts undertaken by his co-conspirators (i.e, the false entries on loan documents by Fredo), a Pinkerton instruction was especially apt.

3. Wester's challenges to his sentence have more merit. One argument is that the district court improperly calculated the victim loss figures for one of the bank bribery counts. The other concerns an adjustment for role in the offense. We address the claims in that order, describing at the outset the calculation of the sentence. Citations are to the 1987 edition of the guidelines which was applied in this case.

At sentencing, the misapplication and bank bribery counts were grouped as closely related counts under U.S.S.G. § 3D1.2(d); and the bribery guideline was used to determine the base offense level because its level is the higher of the two. Id. § 3D1.3(b). The base offense level for bank bribery is eight, id. § 2B4.1, to be increased based on the greater of the value of the bribe or the improper benefit conferred in return, according to the table at section 2F1.1 (fraud). In this case, the figure employed was the value of the bribe.

At the sentencing hearing, the district court found that Wester received, or intended to receive, bribes totalling $12,650,000. From the presentence report, it appears that this total reflected Wester's release from personal liability on the $12.4 million NEFR loan (in exchange for arranging the $2.3 million loan to Morgan and George), and the $250,000 in buyout payments he received from Morgan and George. This $12,650,000 figure subjected Wester to the maximum 11-level increase. U.S.S.G. § 2F1.1.

The resulting offense level of 19 (8 plus 11) was further adjusted upward by 4 levels to 23, reflecting Wester's role as an organizer or leader (a separate issue addressed below). There was no reduction for acceptance of responsibility. The resulting range, for a first time offender, is 46 to 57 months' imprisonment. The district court sentenced Wester, at the bottom of the range, to 46 months.

On appeal, Wester first maintains that the district court should not have included the $12.4 million figure as any part of the value of the bribes. He contends that the release from his personal guaranty on the $12.4 million loan should not count because NEFR did not consider the $2.3 million loan a quid pro quo for the release, an argument that he supports by pointing out that NEFR never formally executed the release. He also asserts that First Service would likely have made the $2.3 million loan to Morgan and George regardless of whether NEFR offered to release Wester and Fredo from personal liability.

18 U.S.C. § 215 makes it criminal corruptly to solicit, accept or agree to accept anything of value intending to be influenced or rewarded in connection with a bank transaction. The jury was entitled to find that Wester did foster the $2.3 million loan to NEFR on the understanding that he would be relieved of his personal guaranty. Whether the bank would have made the loan anyway, and whether Wester actually received the promised benefit, are of no moment under the statute; and the guidelines apply to a promised payment quite as much as to payment actually received.3

Wester is on more solid ground when he argues that, assuming that the promised release of his personal guaranty could be counted for sentencing purposes, the district court incorrectly valued the release at the $12.4 million figure, which represented the full amount of the loan. It is far from clear that this issue was properly preserved, a point to which we will return; but the issue was discussed in oral argument in this court, and the government has furnished us with the Eighth Circuit's helpful decision in United States v. Fitzhugh, 78 F.3d 1326, 1331 (8th Cir.1996).

In Fitzhugh, the court was concerned with valuing the improper benefit conferred on the borrower by a loan obtained by bank bribery. The trial court had taken this value to be simply the face amount of the loan; but as the Eighth Circuit explained, citing authority and examples, "[t]he value of a transaction is often quite different than the face amount of that transaction." 78 F.3d at 1331. Indeed, the current guideline commentary makes clear that (depending on the facts) the value of a loan might be no more than the value of a lower interest rate procured through the bribe. Id.

Obviously, if Wester had been bribed with a one-dollar lottery ticket for a million dollar prize, no one would claim that the ticket should be valued at the full potential winnings. So, too, if he had been given a million-dollar term life insurance policy. Here, the actual value of Wester's promised release from his personal guaranty for the $12.4 million loan depends on such factors as the likelihood of default and the worth of the collateral securing the loan. It is unlikely that the economic value of the release comes close to $12.4 million.

At sentencing, neither the parties nor the probation officer made any attempt to develop the information necessary to estimate reasonably the value of the release. It appears that in the district court Wester's primary concern was to exclude any consideration of the release (on grounds we have already rejected); and neither the probation officer nor the government seems to have noticed the underlying problem with using face value when the presentence report was prepared. Thus, the district court was not fairly alerted to the issue.

Nevertheless, we think that the miscalculation should be noticed as plain error. United States v. Olano, 507 U.S. 725, 113 S. Ct. 1770, 123 L. Ed. 2d 508 (1993). Prejudice exists since it is almost certain that the misevaluation affected the guideline range, quite possibly to a significant extent; for example, eliminating the $12.4 million figure entirely would lower the range to 30 to 37 months. And while an appeals court is not required to notice every such unpreserved error in sentencing, Olano, 507 U.S. at 736, 113 S.Ct. at 1778, we think that this is a proper case for us to notice a significant mistake. United States v. Whiting, 28 F.3d 1296, 1312 (1st Cir.), cert. denied, 513 U.S. 956, 115 S. Ct. 378, 130 L. Ed. 2d 328 (1994).

Wester's other main claim as to his sentence is that the district court erred in adjusting his offense level up four levels for his role in the offense, under U.S.S.G § 3B1.1(a). This provision provides for a four-level enhancement if the court finds that "the defendant was an organizer or leader of a criminal activity which included five or more participants or was otherwise extensive." On appeal, Wester's only developed challenge is to the latter requirement that the activity include five or more participants or be "otherwise extensive."

At the sentencing hearing, the district judge found that Wester was an organizer or leader, based on his capacity as a top official at First Service and because the scheme likely could not have taken place without Wester's leadership. From this the court concluded that the enhancement was warranted, without making any additional record finding as to whether the enterprise involved five or more participants or was "otherwise extensive."

The court did adopt the presentence report by checking the appropriate box, but the report is itself a source of uncertainty. The initial report appears to rely on the "five or more participants" prong, stating that Wester was the organizer of criminal activity involving himself, Fredo, George, Morgan, and then naming several other individuals such as Morgan's accountant, George's lawyer, officials at NEFR, and employees of First Service. There was no reliance on other variables such as duration, number of episodes, or amount.

Before sentencing, Wester objected to this finding on the grounds that the necessary five participants must each be criminally responsible, not merely involved, see United States v. Graciani, 61 F.3d 70, 75 (1st Cir.1995), and that none of the persons named in the report beyond the four main actors were criminally responsible for the relevant actions. In response, the probation officer prepared an amended report that took the position that Wester's activities were "otherwise extensive" because of the number of individuals directly involved and the necessary use of other unknowing employees of First Service in order to effect the scheme.

But the amended report did not clearly abandon the earlier position that there were also five or more participants, and the district judge did not make clear which of the report's two alternative grounds he was adopting. The problem is not that independent detailed findings by the district court are required; rather, it is that we cannot effectively review the decision to impose the four-level increase without knowing the ground on which it rests. United States v. Anh Van, 87 F.3d 1, 2-4 (1st Cir.1996).

None of this would matter if the undisputed facts required a finding that there were five criminally responsible participants or that the activity was otherwise extensive. But that is not the case here. On appeal, the government concedes that the five participant requirement cannot be met, and, in our view, the district court was not compelled to find that the activity was "otherwise extensive," a label that incorporates a number of variables primarily within the ken of the district court. Anh Van, 87 F.3d at 4.

On remand, the district court should address the "otherwise extensive" issue in the course of resentencing. The court is free to make new findings in support of its earlier determination or to reconsider the adjustment entirely, as it sees fit. Since resentencing will likely be required based on the re-valuation of the bribes, we affirm the convictions but vacate the existing sentence and remand for resentencing.

It is so ordered.