POSNER

Brian P. Netols, Jason A. Yonan, Office of the United States Attorney, Chicago, IL, for Plaintiff–Appellee. Michael D. Ettinger, Ettinger, Besbekos & Schroeder, Palos Heights, IL, Marc W. Martin, Marc Martin, Ltd., Chicago, IL, for Defendants–Appellants.
Before POSNER, RIPPLE, and KANNE, Circuit Judges.

POSNER, Circuit Judge.

Robert Brunt, John Farano, Charles Murphy, and Tracey Scullark were charged with mail and wire fraud and Brunt and Scullark also with money laundering and Farano also with theft of federal government funds. 18 U.S.C. §§ 641, 1341, 1343, 1957(a). All were crimes relating to an elaborate real estate financing fraud scheme in Chicago during the housing bubble of the early 2000s. A jury convicted the defendants, and the judge sentenced Brunt to 151 months in prison, Farano to 108, Murphy to 72, and Scullark to 78. He also ordered them all to pay restitution.

We need to describe the criminal scheme. The U.S. Department of Housing and Urban Development sells properties that it owns at a 10 to 30 percent discount if the buyer of a property is a certified nonprofit organization that agrees to resell the property to a low- or moderate-income person or family who intends to live there. The defendants obtained the properties from HUD by using a HUD-certified nonprofit named Westwood Community Development, which they corrupted (although surprisingly neither Westwood nor its principals appear to have been prosecuted), to buy the properties for them. Westwood in effect “fronted” for the defendants, who paid kickbacks, sometimes labeled “donations,” to Westwood personnel who assisted in the scheme. Having obtained the properties from Westwood, the defendants resold them not to persons of low or moderate income but instead to persons who wanted to invest in real estate rather than use the real estate they bought as a place to reside. To recruit these investors the defendants misrepresented the value of the properties by promising to rehabilitate them so that they would be worth even more, and to find tenants for the properties. Some of the properties the defendants did not rehabilitate at all; the others they gave “cosmetic rehabs” of little value.

The investor-buyers had little or no money, and so needed large mortgages to finance their purchases of the properties. The defendants obtained the mortgages for them by submitting false information to banks regarding the conditions of the properties and the investors' assets, income, employment, and intentions to occupy the properties. A loan officer at a mortgage brokerage company was bribed to assist in the scheme, and appraisers were bribed to submit fraudulent appraisals of the properties.

The defendants played different roles in administering the scheme. Brunt and Scullark recruited the investors (that is, the buyers of the properties) and the appraisers, and Brunt arranged for the “cosmetic rehabs,” while Farano and Murphy, who were lawyers, financed the transactions—the purchase (from HUD via Westwood) and sale of properties and the bank financing. Farano did the paperwork for many of the transactions. Many of the investors were ruined when the housing bubble collapsed and the banks lost money as a result of defaults because the properties were now worth less than the unpaid balances of the mortgages on them.

All the defendants except Brunt asked to be tried separately from the other defendants. The judge refused, justifiably, because severance would have caused massive duplication of effort. With severance the entire scheme in all its complexity would have had to be proved anew in each case by the government's witnesses, who included investors, appraisers, and HUD officers. The superior efficiency of trying defendants jointly in a complex criminal case (provided it isn't so complex that the jury can't understand it if there are multiple defendants) is justification for making the only criterion for severance “prejudice” to one or more of the defendants. Fed.R.Crim.P. 14(a).

The only claim of prejudice is that the judge's refusal to sever allowed the admission of very damaging inadmissible evidence against particular defendants in the form of testimony by their codefendants, notably Brunt. But their real complaint, as should be apparent from the word “inadmissible” in the preceding sentence, has nothing to do with the refusal to sever; it concerns, rather, the judge's rulings on the defendants' motions under Rules 403 and 404(b) to exclude specific testimony. As the Supreme Court explained in Zafiro v. United States, 506 U.S. 534, 540, 113 S.Ct. 933, 122 L.Ed.2d 317 (1993), “a defendant normally would not be entitled to exclude the testimony of a former codefendant if the district court did sever their trials, and we see no reason why relevant and competent testimony would be prejudicial merely because the witness is also a codefendant.” If the judge should have granted some of the motions to exclude that he denied, and the defendants have challenged those denials and we decide that his rulings were erroneous and the errors not harmless, the defendants have grounds for reversal of their convictions. Admissibility and severance are separate concerns.

In the decision that the Supreme Court affirmed in Zafiro, we had said that “persons charged in connection with the same crime should be tried separately only if there is a serious risk that a joint trial would prevent the jury from making a reliable judgment about the guilt or innocence of one or more of the defendants,” as in “a complex case with many defendants some of whom might be only peripherally involved in the alleged wrongdoing. The danger is that the bit players may not be able to differentiate themselves in the jurors' minds from the stars.” 945 F.2d 881, 885 (7th Cir.1991) (emphases in original). This is not such a case. None of the defendants was a “bit player” in the conspiracy. In fact they were the principals; the bit players were the investors who misrepresented their ability to repay the mortgage loans that they needed in order to be able to buy properties from the defendants, the bribed loan officer, and the crooked appraisers.

Nor was the trial so long or complex that the jury's verdict cannot be thought reliable. One can imagine a trial expected to be so long that no employed person could take the time off from his job to serve on the jury, with the result that the jury might be unrepresentative. “Professionals often cannot afford (or their employers will not abide) jury service on protracted cases. Consequently, courts frequently excuse them upon a showing of undue hardship or extreme inconvenience. The hardship results from the projected loss of pay over a lengthier trial. Collectively, these two practices combine to seriously suppress the percentage of persons with higher education who serve on juries in complex cases.” Franklin Strier, “The Educated Jury: A Proposal for Complex Litigation,” 47 DePaul L.Rev. 49, 72–73 (1997); see also “Development in the Law—III. Jury Selection and Composition,” 110 Harv. L.Rev. 1443, 1454 (1997); Gordon Van Kessel, “Adversary Excesses in the American Criminal Trial,” 67 Notre Dame L.Rev. 403, 478–79 (1992). The trial in this case was long—nine weeks—but there is no indication that the jury was incompetent, by virtue of professional exemptions or hardship exclusions, to render a sound decision. The fact that the jury spent seven days deliberating and acquitted some of the defendants of some of the counts (and a fifth defendant of the only count against him) suggests that the jury was attentive and conscientious.

A second situation that we said in our Zafiro opinion may require severance is “where exculpatory evidence essential to a defendant's case will be unavailable—or highly prejudicial evidence unavoidable—if he is tried with another defendant.... There are cases in which a person would refuse to testify for a codefendant in a joint trial for fear of incriminating himself, yet if tried separately and convicted might thereafter be willing to testify and might give testimony exculpating the other defendant.” 945 F.2d at 886. That's really two situations, one in which the defendants would oppose severance and the other in which the government would oppose it. Neither one is our case.

Our case illustrates a third situation discussed in Zafiro, that of “mutual antagonism, finger-pointing, and other manifestations or characterizations of the effort of one defendant to shift the blame from himself to a codefendant”; but those things, we said, “neither control nor illuminate the question of severance.” Id. Consider Farano's complaints about Brunt's testimony that Farano had used a racial slur against him and Scullark, both of whom are black; Farano is white. But his lawyer didn't object to Brunt's testimony about the racial slur. He did object to Brunt's testimony that Farano had used his law license for theft; but the judge sustained the objection. Other objections by Farano to Brunt's testimony were overruled—but instead of alleging judicial error in ruling on objections, Farano insists mistakenly that his case should have been severed from Brunt's.

Defendant Murphy objects to the introduction of evidence that he claims would not have been introduced had Brunt not been on trial with him. The evidence related to Brunt's rehab of a building on West 72nd Street in Chicago, in connection with which he obtained a loan of $155,000 from a title company and used most of it to buy a Rolls Royce. The evidence of his laundering the fraud-induced receipt of the $155,000 by using it to buy the Rolls was used by the government to bolster the money-laundering charge against Brunt; and Murphy was one of Brunt's partners in the West 72nd Street project. But he was not accused of participating in Brunt's fraud or money laundering with respect to the project. The judge ruled sensibly that the fact that “Murphy knew how to establish a proper rehabilitation loan [for a construction project] in which the funds were paid out incrementally only after submission of proof that the work had actually been done is also evidence that the other transactions in which he and Brunt collaborated to obtain large amounts of money from unsuspecting banks before any rehabilitation work was done were deliberately fraudulent” (emphasis in original)—contrary to Murphy's claim to be unsophisticated about real estate transactions. The procurement of the rehabilitation loan showed that Murphy knew what a lawful rehabilitation loan was, implying in turn a sophisticated knowledge of real estate transactions. This evidence would have been admissible against Murphy even if Brunt hadn't been on trial with him.

Brunt had been represented at one time by a lawyer named McDermott, and testified that Murphy had attended a meeting with him and McDermott during which it became clear that McDermott, who shared an office with Murphy, was Murphy's puppet. This testimony, Murphy argues, forced him to call McDermott as a witness. So McDermott testified—and obligingly denied having been Murphy's puppet. So far, so good, for Murphy. But on cross-examination McDermott acknowledged Murphy's presence in several of his meetingswith Brunt, though he insisted that his (that is, McDermott's) advice to Brunt had been based on his independent professional judgment. There is no ground for thinking that joinder with Brunt, by “forcing” Murphy to call McDermott as a witness, hurt Murphy's defense; for McDermott's testimony as a whole was favorable to Murphy.

There are other objections to the district judge's rulings at the trial, but they fall with the objections that we've already discussed. We add that the evidence of the defendants' guilt was overwhelming, and if there were any erroneous rulings (which we don't think there were) they were therefore harmless. We turn to the sentences.

Farano and Murphy each received a four-level enhancement of their guidelines sentencing range on the ground that each was an organizer or leader of criminal activity in which there were at least five participants or that was “otherwise extensive.” U.S.S.G. § 3B1.1(a). Both conditions were satisfied.

Brunt and Scullark, however, did not receive the enhancement. The omission is puzzling, but may reflect the ambiguity of Application Note 4 to the guideline. It states that “in distinguishing a leadership and organizational role from one of mere management or supervision ... the court should consider” factors that “include [but are not necessarily exhausted by] the exercise of decision making authority, the nature of participation in the commission of the offense, the recruitment of accomplices, the claimed right to a larger share of the fruits of the crime, the degree of participation in planning or organizing the offense, the nature and scope of the illegal activity, and the degree of control and authority exercised over others.”

Years ago we observed that this multifactor, open-ended, unweighted so-called test “contributes to the murk surrounding review of § 3B1.1 adjustments.” United States v. Mustread, 42 F.3d 1097, 1104 n. 3 (7th Cir.1994). We elaborated our concerns in United States v. Rosales, 716 F.3d 996, 997 (7th Cir.2013), complaining that “as with most multifactor tests, the application note's seven-factor test is none too clear. No weighting of the factors is indicated (so really the ‘multifactor test’ should be called a ‘list of factors'). And a majority of the factors are vague or redundant. That is true of ‘the nature of [the defendant's] participation in the commission of the offense,’ ‘the degree of participation in planning or organizing the offense,’ ‘the nature and scope of the illegal activity,’ and even ‘the degree of control and authority exercised over others.’ For what is the difference between ‘control’ and ‘authority’? And for that matter is there a difference between a ‘leader’ and an ‘organizer’? The phrase ‘a leadership and organizational role’ appears to fuse them.” We noted that “the apparent equation of ‘organizer’ to ‘leader’ is another oddity. The ‘organizer’ of an entire enterprise would usually be thought the person who had started it; he might not be running it; he might not be a leader.” Id. at 998.

We don't understand the singling out of Farano and Murphy as the sole leaders or organizers of the fraudulent scheme. The gang had officers and enlisted men. The officers were our four defendants. The enlisted men were the buyers who misrepresented their financial situation, the bribed appraisers, and the bribed loan officer. The four defendants were all leaders, though they led in different domains. The lawyers, Farano and Murphy, played the decisive role in the financial negotiations with Westwood and the banks. Brunt and Scullark were responsible for recruiting investors (the buyers of the properties) and appraisers. Brunt also dealt with Westwood and managed the cosmetic rehabs, which doubtless meant that he was in charge of workers who didn't know they were assisting in a fraud. But in determining how extensive a criminal enterprise is for purposes of the leader/organizer enhancement, “unknowing ... outsiders” assisting the enterprise count. U.S.S.G. § 3B1.1, Application Note 3.

Scullark might have been thought a manager or supervisor, a type of leader who receives only a three-level enhancement. U.S.S.G. § 3B1.1(b); see also United States v. Figueroa, 682 F.3d 694, 695–97 (7th Cir.2012). The government sought a four-level enhancement for Brunt, but no enhancement for Scullark, though her presentence report had recommended the three-level enhancement for her. The judge imposed no section 3B1.1 enhancement on either one. This is especially puzzling with regard to Brunt, who according to his presentence report had “organized and led the nominee buyers' activities, the submission of loan applications, and gave specific directives to Defendants Scullark, D'Aifallah, and Jackson. In addition, Defendant Brunt was primarily responsible for identifying properties to acquire.... Although each defendant participated in the scheme of their own volition, it was Defendant Brunt that brought the parties together. Without Defendant Brunt, neither Defendant Farano or Defendant Murphy had knowledge of the specific properties or had relationships with an appraiser and loan officer who would readily participate in the scheme; not to mention Defendant Brunt's relationship with the individuals who served as nominee buyers, who appear to have been easily duped by the defendants.” But the government has not appealed the denial of the enhancement.

The length of the defendants' prison sentences was influenced by the amount of loss that the district judge found had been caused by the fraudulent scheme. See U.S.S.G. § 2B1.1. There were losses of two types: losses to the banks that made mortgage loans to the investors because of misrepresentations of the investors' financial situation, and losses to HUD equal to the 10 to 30 percent discount at which it sold properties to Westwood that were never used to provide low- or moderate-income housing.

All four defendants object that the losses to the banks often occurred after the properties that secured the bank loans had been refinanced. (In the case of five such properties, the losses attributed to the defendants were $768,450.) Suppose one of the investors had bought a property for $80,000 nominally from Westwood but actually from the defendants, obtained a $100,000 mortgage on the basis of a phony appraisal, and later refinanced the property by obtaining a $110,000 mortgage from another bank, replacing the original mortgage. And suppose that later still, when the bubble burst, the mortgagor defaulted and the mortgagee foreclosed, but at the foreclosure sale the property brought only $60,000. The government in such a case would assess the loss as $40,000 ($100,000 original loan amount, minus $60,000 foreclosure sale proceeds), but would ignore the further $10,000 loss to the refinancing bank. The $40,000 loss would also have been incurred by the refinancing banks rather than by the banks with which the defendants had dealt; and the defendants argue that refinancing breaks the causal link between the original fraud and the loss.

Not true. United States v. Morris, 80 F.3d 1151, 1171–72 (7th Cir.1996). Actual loss is defined as “reasonably foreseeable pecuniary harm that resulted from the offense,”and “reasonably foreseeable pecuniary harm” as “pecuniary harm that the defendant knew or, under the circumstances, reasonably should have known, was a potential result of the offense.” U.S.S.G. § 2B1.1, Application Notes 3(A)(i), (iv). The defendants knew that the properties were overpriced and that the buyers (the “investors”) were underfinanced. They could therefore foresee the likelihood of default and consequent loss.

The additional loss—the $10,000 in our example for which the defendants were not punished—had resulted from the decision to refinance, a decision that may well have been influenced by factors, such as declining interest rates, that the defendants couldn't be assumed to have foreseen; and so for that incremental loss they would not be held responsible. But there could be no doubt that they were responsible for the $40,000 loss, which would have been a foreseeable loss to the first lender had the borrower not refinanced.

The judge assessed the loss to HUD as the amount of the discounts that HUD gave Westwood on the sales of properties intended, HUD thought—as it was the motivation for the discount—for low- and moderate-income housing, yet never used for that purpose because the defendants sold the discounted properties to investors who had no interest in providing low- or moderate-income housing. The defendants argue that HUD lost nothing because it would have sold the properties at a discount to someone else. The argument is unsound. HUD was led to believe that it was buying something of value to it for the amount of the discounts, namely low- and moderate-income housing; and the defendants stole the money, thus depriving HUD of a thing of value that had been promised. It's as if you order a product, pay for it in advance, but never receive it, because the product is stolen en route to you. The money you pay to buy a product you never receive is a loss.

Of course the defendants could not have foreseen the precise loss to the lenders (as distinct from the loss to HUD), because, while they must have known that the buyers of the properties were very likely to default, having overpaid for the properties with fraudulently obtained mortgage loans, they could not know what the properties would bring at a foreclosure sale, given uncertainty about future real estate prices. Far more financially sophisticated people than the defendants, employed at companies like Bear Stearns, Lehman Brothers, and American International Group, also failed to predict the housing crash. But remember that the loss guideline requires only a “pecuniary harm that the defendant knew or, under the circumstances, reasonably should have known, was a potential result of the offense.” U.S.S.G. § 2B1.1, Application Note 3(a)(iv). The key word is “potential,” which means “could happen.” The losses to the lenders were a “potential result” of the defendants' fraud, as the defendants well knew.

The last issue involves the judge's calculation of the amount of restitution that the defendants were ordered to pay. Remember that in calculating loss for purposes of determining the defendants' prison sentences, the judge had ignored the loss to banks that refinanced properties sold by the defendants, as distinct from the loss that the banks that had originally financed the sales would have incurred had those banks' mortgages not been refinanced. Yet the lesser loss, the loss the judge used in calculating the defendants' prison sentences, had actually been borne by the refinancing banks, as we know; for the original lenders had been repaid when their loans were refinanced. And so the judge also included as victims entitled to restitution the refinancing banks, but deemed them entitled to restitution only for the initial loss, the loss the first tier of banks would have borne had their loans not been repaid. In our example of a property on which the first bank had lent $100,000 and the refinancing bank $110,000 and the property went into default and brought only $60,000 at the foreclosure sale, the loss independent of the refinancing was $40,000, and that would be the amount of restitution ordered.

In arguing for restitution to the refinancing banks, the government confounds loss amount, which requires only “pecuniary harm that the defendant knew or, under the circumstances, reasonably should have known, was a potential result of the offense,” U.S.S.G. § 2B1.1, Application Note 3(A)(iv), with restitution, authorized only for a “victim,” defined as “a person directly and proximately harmed as a result of the” defendant's crime. 18 U.S.C. § 3663A(a)(2); see United States v. Shepard, 269 F.3d 884, 886–87 (7th Cir.2001); cf. United States v. Donaby, 349 F.3d 1046, 1053 (7th Cir.2003); United States v. Marino, 654 F.3d 310, 319 (2d Cir.2011). The refinancing banks may have been caught up in the housing bubble and willing therefore to make “liars' loans,” that is, loans made without requiring proof of ability to repay. The $110,000 loan in our example may have been such a loan. But banks making such loans would be coauthors of their loss, since had they required proof of financial soundness from the borrowers they would not have refinanced the properties in question.

United States v. Yeung, 672 F.3d 594, 603 (9th Cir.2012), held that banks that had bought fraudulently obtained mortgages were victims of the fraud entitled to restitution because as part of the defendant's fraud “the borrower for each loan had misrepresented his financial ability to repay the loan” and the banks had relied on that fraudulent information, which had been concocted by Yeung. There is as yet no evidence in this case of such reliance by the refinancing banks, and in its absence those banks cannot be counted as “victims” for restitution purposes, though their loss is loss under section 2B1.1 and thus usable in determining the defendants' prison sentences.

A second question about restitution is the method of calculating it. The defendants argue that the method used was wrong, citing United States v. Robers, 698 F.3d 937, 939 (7th Cir.2012), cert. granted, ––– U.S. ––––, 134 S.Ct. 470, 187 L.Ed.2d 316 (2013). The victims in that case were a mortgage lender and a mortgage guarantor (for simplicity we'll call them both “mortgagees”). They correspond to the defrauded banks in our case. As a result of Robers's fraud, the properties were worth less than the balance of the mortgages and so the mortgagees could not recover the balance when the mortgagors defaulted. The restitution statute entitles a victim to “the value of the [victim's] property on the date of sentencing, less the value (as of the date the property is returned) of any part of the property that is returned.” 18 U.S.C. §§ 3663A(b)(1)(B)(i)(II), (ii). The question was whether “the date the property is returned” should be the date of the foreclosure or, later, when the property is sold (presumably by the mortgagee, on the assumption that the mortgagee had acquired title to the property at the foreclosure sale). We held that it was the latter. Four circuits agree with this approach, United States v. Statman, 604 F.3d 529, 538 (8th Cir.2010); United States v. James, 564 F.3d 1237, 1244–47, (10th Cir.2009); United States v. Innarelli, 524 F.3d 286, 294 (1st Cir.2008); United States v. Himler, 355 F.3d 735, 745 (3d Cir.2004), and maybe a fifth. See United States v. Boccagna, 450 F.3d 107, 115–20 (2d Cir.2006). Two disagree. United States v. Holley, 23 F.3d 902, 914–15 (5th Cir.1994); United States v. Smith, 944 F.2d 618, 625–26 (9th Cir.1991). The Supreme Court will decide; until it does, we'll stick with our Robers decision.

Since the refinancing banks probably were not victims and therefore were not entitled to restitution, the choice of the transaction date to use to measure the restitution that would have been due them had they been victims is irrelevant. But not all the banks awarded restitution were refinancers—some were owners of the original mortgages, and thus entitled to restitution. Robers would make the amount of that restitution the loss the banks suffered when they sold the properties that they acquired after they foreclosed their mortgages, just as the district court calculated.

That completes our analysis of the colorable issues presented by the defendants. We affirm the judgments against them except for the order of restitution to the refinancing lenders, which we vacate. The district judge will have to consider evidence on whether the refinancing banks that are seeking restitution had based their refinancing decision in whole or part on fraudulent representations by the defendants, and, depending on the outcome of that consideration, the judge may have to recalculate the restitution that the defendants owe. And of course should the Supreme Court reverse or modify our decision in the Robers case, the judge will be guided by the Court's opinion rather than by this opinion in calculating the amount of restitution.

We would be remiss if we ended this opinion without expressing our concern with the length of time that this case has taken to reach us—six and a half years since indictment. The initial delays were attributable largely to the complexity of the government's investigation of the defendants, which continued after the indictment was returned and resulted in a superseding indictment that added Murphy as a defendant. Trial was scheduled to begin on January 31, 2011—already more than three years since the initial indictment—but on the eve of trial Brunt switched lawyers, and as a result the trial was postponed nine months; it shouldn't have been postponed that long for that reason. The trial took nine weeks, and after that the defendants were allowed two months for briefing their motions for acquittal, which was followed seven months later by sentencing. The defendants filed their notices of appeal in July 2012, other than Murphy, who filed in October of that year. The defendants' lawyers than withdrew, and new counsel were appointed. The appeal was not argued until 19 months after the original notices of appeal. The delay troubles us, especially as there must now be further proceedings in the district court.

Affirmed in Part, Reversed in Part, and Remanded.