Reset A A Font size: Print

United States Court of Appeals,First Circuit.

UNITED STATES of America, Appellee, v. Daniel APPOLON; Ernst Appolon; Latoya Haltiwanger; J. Daniel Lindley; and Eric L. Levine, Defendants, Appellants.

Nos. 10–2243, 10–2350, 10–2266, 11–1130, 10–2313.

Decided: September 19, 2012

Before TORRUELLA, BOUDIN, and LIPEZ, Circuit Judges. Jeanne M. Kempthorne for appellant Daniel Apollon. Mark E. Howard for appellant Ernst Appolon. Tina Schneider for appellant Latoya Haltiwanger. James C. Rehnquist for appellant J. Daniel Lindley. Dana A. Curhan for appellant Eric L. Levine. Vijay Shanker, Attorney, United States Department of Justice, Criminal Division, Appellate Section, with whom Carmen Milagros Ortiz, United States Attorney, Victor A. Wild, Ryan M. DiSantis, and Mary Beth Murrane, Assistant United States Attorneys, Lanny A. Breuer, Assistant Attorney General, and Greg D. Andres, Acting Deputy Assistant Attorney General, were on brief, for appellee.

Appellants Daniel Appolon (“Daniel”), Ernst Appolon (“Ernst”), Latoya Haltiwanger, J. Daniel Lindley, and Eric L. Levine were players in the Boston real estate market. Along with six coconspirators, appellants devised and executed a mortgage fraud scheme which netted them illegal profits of nearly $2 million between May 2005 and June 2006. The scheme itself was uncomplicated: appellants and their coconspirators arranged for straw buyers to purchase real property at the asking price, falsified mortgage loan applications for the straw buyers to obtain financing for an artificially-inflated purchase price, and pocketed the difference. The loans secured by each of the properties involved in appellants' scheme eventually went into default, and most of the properties were forced into foreclosure at huge losses for the lenders.

Appellants and their coconspirators were indicted by a federal grand jury on May 15, 2008. Appellants were each charged with one count of conspiring to commit wire fraud in violation of 18 U.S.C. § 371 and with multiple counts of committing wire fraud in violation of 18 U.S.C. § 1343. (Daniel was charged with five counts of wire fraud, Ernst with thirty-four, Haltiwanger with seven, and Lindley and Levine each with forty-one.) In addition, Lindley and Levine were charged with nineteen counts of money laundering in violation of 18 U.S.C. § 1957. All of the counts in the indictment included a charge of aiding and abetting in violation of 18 U.S.C. § 2.

On June 2, 2010, after a six-week jury trial in the United States District Court for the District of Massachusetts, appellants were found guilty of the charges against them, except that Lindley was found not guilty of eleven counts of wire fraud and four counts of money laundering.1 Each appellant was sentenced to a term of imprisonment, followed by a period of supervised release.2

Appellants now raise a series of challenges to the district court's management of this case, evidentiary rulings, jury instructions, sentencing calculations, and other orders. Included in our analysis of these challenges is a rejection of Haltiwanger's argument that a district court may not undermine a jury's nullification power by explicitly instructing the jury that it has a duty to return a guilty verdict if it is convinced beyond a reasonable doubt of a defendant's guilt. We also explain the proper method of calculating loss in a mortgage fraud case such as this.

After carefully considering each challenge, we affirm.


We provide here only a brief synopsis of the essential facts of this case, in the light most favorable to the jury's verdict, see United States v. Mubayyid, 654 F.3d 35, 41 (1st Cir.2011), reserving additional detail for the analysis that follows.

Levine, a real estate attorney who had been suspended from the practice of law, shared office space in Boston with Lindley, another attorney. During Levine's suspension, he transferred components of his real estate practice in early 2005 to Lindley, who thereafter handled all property closings for Levine. Levine and Lindley sometimes worked with Haltiwanger, a residential mortgage broker at Topdot Mortgage Company, and with Ernst, a realtor at New England Merchants and a principal at Oligarchy Funding. Daniel, Ernst's brother, recruited clients and did odd jobs for New England Merchants.

In mid–2005, appellants hatched their mortgage fraud scheme, which began with Ernst identifying properties for sale in the Boston area and negotiating purchase prices with the sellers. Assisted by Daniel and others, Ernst then recruited straw buyers for these properties, targeting individuals with good credit scores who were willing to trade their complicity in appellants' scheme for a cut of the profits. The standard pitch delivered to a straw buyer was that, in exchange for the use of his or her name, the straw buyer could expect to earn approximately $10,000 and would not be responsible for any down payment or mortgage payments. Next, aided by Lindsay MacPhee (Levine's administrative assistant), Levine, Lindley, Ernst, and Haltiwanger prepared and filed falsified mortgage loan applications, purchase-and-sale agreements, and HUD–1 settlement statements on behalf of the straw buyers, misrepresenting the straw buyers' eligibility for the loans and overstating the purchase prices of the properties.3 (Haltiwanger also served as a straw buyer for one property.) Once the falsified applications were approved by unsuspecting mortgage lenders, the loan proceeds were wired to Lindley's Interest on Lawyers Trust Account (“IOLTA”), from which the actual purchase prices were paid to the sellers and the excess was disbursed to the conspirators, usually after passing through bank accounts held by Levine. The loans secured by the properties were then permitted to default. Most of the properties were forced into foreclosure, and one was burned for insurance money.

In all, twenty-one properties were involved in appellants' scheme. For each of these properties, two separate HUD–1 forms were created, with one form reflecting the actual purchase price and the other reflecting the artificially-inflated price listed on the falsified mortgage loan application. In addition, separate closings, presided over by Lindley, were held for many of the properties in order to distance the straw buyers from the sellers and keep secret the existence of the scheme. By June 2006, the conspirators had earned nearly $2 million in illegal profits, commissions, and fees.


Our analysis begins with Ernst's pretrial motion for severance. It then turns to the sufficiency of the evidence, moves to assorted evidentiary issues raised by appellants during trial, proceeds to the district court's jury instructions and sentencing calculations, and concludes with Levine's post-trial motion for an evidentiary hearing or discovery.

A. Ernst's Motion for Severance

Levine was represented at trial by Isaac Borenstein of the law firm Denner Pellegrino, LLP (“Denner Pellegrino”).4 During jury selection, Ernst's counsel notified the district court that, prior to trial, another Denner Pellegrino attorney, Jeffrey Denner, had discussed with Ernst the possibility of representing him in this case. Although Denner Pellegrino never took on Ernst's representation, Ernst's counsel asserted that a severance of Ernst from his co-defendants was necessary to preserve Ernst's constitutional right to testify in his own defense, see Rock v. Arkansas, 483 U.S. 44, 51, 107 S.Ct. 2704, 97 L.Ed.2d 37 (1987), since Ernst's interests at trial differed from Levine's, and Borenstein had the ability to cross-examine Ernst with confidential information that he disclosed to Denner.

Borenstein explained to the district court that he brought Levine's representation with him when he joined Denner Pellegrino from another law firm. When Borenstein arrived at Denner Pellegrino, Denner informed him that he had previously had a non-substantive conversation with “someone whose name [was] Appolon” about taking over his representation in this case. Borenstein and Denner then consulted with Levine, who waived any potential conflict of interest. Borenstein represented to the court that he had no further discussions with Denner “or anybody at Denner Pellegrino” about Ernst and had acquired no confidential information about him.

The district court then held an in camera hearing on the conflict claim followed by fact-finding. In 2008, while awaiting trial, Ralph Appolon (Ernst's brother and coconspirator in this case) became frustrated with his court-appointed counsel and contacted Denner, who had represented him in 2005 in an unrelated matter. Denner sent Paul Andrews, an associate at Denner Pellegrino, to meet with Ralph and Ernst, who also was represented by court-appointed counsel at the time. Denner also may have met personally with Ralph. He did not meet with Ernst, however. Neither Ralph nor Ernst retained Denner Pellegrino's services.

The district court found that it was unclear what exactly transpired in these preliminary meetings. Ralph testified that he provided one of the Denner Pellegrino attorneys with detailed written notes that he and Ernst had prepared. However, the attorneys stated that they had no notes from the meetings and that there were no relevant documents in Denner Pellegrino's files. The court found that “there must have been some discussion of the indictment, at least in general terms, ․ between Ernst and Andrews,” but that any discussion was “exploratory as to whether there would be representation” and could not have been “in-depth.” The court concluded that “the extent to which substantive confidential information was exchanged was probably not very great” and that no confidential information about Ernst was subsequently shared with Borenstein.

The court expressed some concern that Andrews played a “minor” role in Levine's defense after Borenstein joined Denner Pellegrino, including drafting proposed jury instructions and meeting with Levine on one occasion. However, the court reiterated that there was no evidence that any confidential information about Ernst was ever disclosed improperly. Accordingly, the court denied Ernst's motion for severance. As a prophylactic measure, however, it barred Andrews from participating further in Levine's representation. Ernst appeals the denial of his severance motion.

“[T]he general rule is that those indicted together are tried together to prevent inconsistent verdicts and to conserve judicial and prosecutorial resources.” United States v. SotoBeníquez, 356 F.3d 1, 29 (1st Cir.2004). However, a district court is empowered to sever co-defendants' trials if “there is a serious risk that a joint trial would compromise a specific trial right of one of the defendants, or prevent the jury from making a reliable judgment about guilt or innocence.” Zafiro v. United States, 506 U.S. 534, 529, 113 S.Ct. 933, ––––, 122 L.Ed.2d 317, –––– (1993); see also Fed.R.Crim.P. 14(a). “If the district court decides not to sever the trial, the defendant bears the burden of making a strong showing that prejudice resulted from the denial of severance, and prejudice in this context ‘means more than just a better chance of acquittal at a separate trial.’ “ United States v. DeCologero, 530 F.3d 36, 52 (1st Cir.2008) (quoting United States v. Boylan, 898 F.2d 230, 246 (1st Cir.1990)). We review a district court's denial of a motion for severance for manifest abuse of discretion, see United States v. Celestin, 612 F.3d 14, 19 (1st Cir.2010); DeCologero, 530 F.3d at 52, and any predicate findings of fact for clear error, see United States v. Allen, 491 F.3d 178, 189 (4th Cir.2007).

We discern no clear error in the district court's factual findings. To the extent that Ernst disclosed protected information to Andrews while discussing the possibility of Denner Pellegrino taking on his representation, there is ample support for the court's finding that no such information was relayed to Borenstein, who obviously could not have cross-examined Ernst with confidential information he did not possess. As a result, we see no reasonable basis for Ernst to have feared cross-examination more than any other criminal defendant. Therefore, we find no manifest abuse of discretion in the denial of his severance motion.

B. The Sufficiency of the Evidence (Lindley, Ernest, and Levine)

At the close of the government's case, the defendants moved for judgments of acquittal based on the insufficiency of the evidence against them. See Fed.R.Crim.P. 29(a). The court reserved its ruling, the defendants renewed their motions at the close of all the evidence, and the court denied the motions at that time. See Fed.R.Crim.P. 29(b). Only Lindley, Ernst, and Levine now challenge the court's ruling. We review their challenges de novo, appraising the evidence in the light most favorable to the jury's verdict, see United States v. Rodríguez–Vélez, 597 F.3d 32, 38 (1st Cir.2010), and giving equal weight to direct and circumstantial evidence, see United States v. Ortiz, 447 F.3d 28, 32 (1st Cir.2006). “The verdict must stand unless the evidence is so scant that a rational factfinder could not conclude that the government proved all the essential elements of the charged crime beyond a reasonable doubt.” Rodríguez–Vélez, 597 F.3d at 39.

1. Lindley

Lindley argues that the evidence adduced at trial was insufficient to prove that he either had actual knowledge of or was willfully blind to the mortgage fraud scheme. Such knowledge or willful blindness is an essential element of the thirty wire fraud counts, see 18 U.S.C. § 1343; United States v. Vázquez–Botet, 532 F.3d 37, 63 (1st Cir.2008), fifteen money laundering counts, see 18 U.S.C. § 1957; United States v. Carucci, 364 F.3d 339, 343 (1st Cir.2004), and one conspiracy count, see 18 U.S.C. § 371; United States v. Yefsky, 994 F.2d 885, 890 (1st Cir.1993), of which he was convicted.

a. Actual Knowledge

The following incidents, which are illustrative, were recounted in detail at trial by one of the government's star witnesses, Andre Junior Lamerique, and corroborated by other witnesses and documentary evidence. Lamerique, a coconspirator of appellants, testified at trial in exchange for the government's agreement to recommend a reduced sentence. However, it was well within the jury's province to choose to believe the testimony of a cooperating witness. See United States v. Rivera–Donate, 682 F.3d 120, 135 (1st Cir.2012); United States v. Calderón, 77 F.3d 6, 10 (1st Cir.1996).

For one of the properties involved in the scheme, the initial straw buyer, Shonette Tomlinson, failed to qualify for financing. As a fallback, appellants used a straw buyer they had worked with previously, Elio Garay. Lindley helped to prepare Garay's mortgage loan application package by signing the HUD–1 settlement statement and occupancy affidavit, which certified that Garay would occupy the premises upon close of escrow. However, Lindley also drafted and signed a side agreement between Tomlinson and Garay providing that Tomlinson would occupy the property, flatly contradicting Garay's affidavit. In addition, Lindley executed a construction holdback agreement between the mortgage lender and Garay providing that $41,000 of the loan proceeds would be held in escrow by Lindley for the purpose of bringing the property into compliance with the Massachusetts Building Code. There is nothing inherently suspicious about construction holdbacks, see Reyes v. Remington Hybrid Seed Co., 495 F.3d 403, 409 (7th Cir.2007), but Lindley failed to list this holdback on the HUD–1 form and, instead of using the funds for repairs pursuant to the agreement, he transferred them from his IOLTA to one of Levine's bank accounts.

For another property, $10,000 was wired from Lindley's IOLTA to the straw buyer, Kimyli Recca, although the HUD–1 form Lindley signed did not list a payment to Recca in that amount. Lindley also signed a $30,000 check drawn on his IOLTA for a fictitious construction holdback on this property. The check was made out to a shell company controlled by Levine.

For a third property, Lindley conducted a closing at which one of the sellers, Taslim Chowdhury, somehow discovered that the straw buyer's HUD–1 form reflected a purchase price that was $101,000 higher than the price on his own form. Chowdhury began arguing with Lindley about the discrepancy and refused to proceed with the sale until he was offered $20,000 by Levine and Ernst to offset any additional tax liability flowing from the higher purchase price. Lindley then signed a check to Chowdhury for $20,000 but did not record that amount on the HUD–1 form.

Lindley attempts to blunt the force of this evidence by chalking up his behavior to inattentiveness and professional incompetence. He points out that Lindsay MacPhee, Levine's administrative assistant, testified that she prepared the checks and much of the paperwork he signed, insisting that he “simply signed [what] his staff handed to him.” He also cites evidence that his signature was forged on a number of HUD–1 forms and other loan documents, and Lamerique's testimony that Levine instructed some of appellants' coconspirators not to discuss the mortgage fraud scheme with him. He notes that after a straw buyer, Elizabeth Son, disclosed the existence of the scheme to law enforcement officers in his presence, he gave agents from the Federal Bureau of Investigation free access to his records.5 On this basis, Lindley argues that the jury reasonably could have drawn an inference that he was “sloppy, but not [a] criminal.”

We acknowledge the reasonableness of that inference. However, it is not the only reasonable inference that could be drawn. When the evidence is construed in the light most favorable to the jury's verdict, see Rodríguez–Vélez, 597 F.3d at 38, this is not a case in equipoise, see Ortiz, 447 F.3d at 34. The evidence of Lindley's knowing participation in appellants' scheme is unusually strong and more than sufficient to prove his actual knowledge beyond a reasonable doubt. See Rodríguez–Vélez, 597 F.3d at 40 (“[W]here, as here, the evidence can be viewed in different ways, we must honor the jury's evaluative choice among plausible, albeit competing, inferences.”); United States v. Hughes, 211 F.3d 676, 681 (1st Cir.2000) (“[T]he jury is generally at liberty to select freely among a variety of reasonable alternative constructions of the evidence.” (internal quotation marks omitted)).

b. Willful Blindness

For the sake of completeness, and to emphasize the strength of the evidence against Lindley, we also address his argument that the evidence was insufficient to prove that he willfully blinded himself to the existence of appellants' scheme. As we have explained, “[w]illful blindness serves as an alternate theory on which the government may prove knowledge.” United States v. Pérez–Meléndez, 599 F.3d 31, 41 (1st Cir.2010). “The purpose of the willful blindness theory is to impose criminal liability on people who, recognizing the likelihood of wrongdoing, nonetheless consciously refuse to take basic investigatory steps.” United States v. Rothrock, 806 F.2d 318, 323 (1st Cir.1986). In order to establish a defendant's willful blindness to criminal activity, the government must show that (1) the defendant was aware of a high probability of wrongdoing and (2) consciously and deliberately avoided learning of the wrongdoing. See Pérez–Meléndez, 599 F.3d at 41; United States v. Azubike, 564 F.3d 59, 66 (1st Cir.2009). Direct evidence is not required; “what is needed are sufficient warning signs that call out for investigation or evidence of deliberate avoidance of knowledge.” Azubike, 564 F.3d at 66.

The government identified a number of warning signs or “red flags,” United States v. Frigerio–Migiano, 254 F.3d 30, 35 (1st Cir.2001), that, uninvestigated, suggest Lindley's willful blindness. Three stand out in particular. First, MacPhee included in Lindley's files two sets of loan documents for nearly every property involved in appellants' scheme. Second, Lindley conducted several closings involving repeat buyers. For example, Son purchased two properties for $520,000 each within a six-day span, signing an occupancy affidavit each time, and Garay (or someone posing as him) purchased three properties within seven weeks. Third, Lindley conducted one closing involving a buyer, Andrew Caputo, who was listed on the HUD–1 form as living in a property which Levine had recently sold to a different straw buyer and for which Lindley had staged the closing.

Lindley attempts to explain away these red flags in various ways. He claims, for instance, that he simply did not notice Caputo's address on the HUD–1 form or other “individual details in closing packets often spanning hundreds of pages,” and that he never inspected his files and thus was unaware that they contained dual sets of loan documents for any properties.

Again, we acknowledge the plausibility of Lindley's explanations. The jury could have concluded that Caputo's address escaped his attention or that he did not carefully examine his own closing files. But the jury also was entitled to disbelieve those explanations, see Rodríguez–Vélez, 597 F.3d at 40, and to find, for instance, that Lindley was aware that his files held two sets of documents for certain properties and that, if Lindley did not compare the two to find out why, it was only because he was consciously avoiding the knowledge that they recorded different purchase prices or otherwise contained discrepancies.

Relying on our decision in United States v. Martin, 815 F.2d 818 (1987), Lindley also argues that the government failed to call any witnesses who could explain to the jury why these “business irregularities” should have alerted him to a high probability of wrongdoing, even if he had detected them. In Martin, the defendants were convicted of crimes arising out of their operation of a car theft ring. See 815 F.2d at 820–21. The government relied on three factors in arguing that one of the defendants, a used car salesman who had helped to resell the stolen vehicles, had willfully blinded himself to the fact that the vehicles' title certificates were forged: (1) the title certificates were duplicates; (2) the cars, which were received in Rhode Island, were resold in New York; and (3) the salesman had been directed by his supervisor not to list his own name on the title certificates as assignee. See id. at 826. Reversing the salesman's conviction, we noted that “[t]he trouble with these factors is that [a jury] not experienced in marketing used cars cannot tell what to make of them.” Id. We elaborated:

[T]he government was asking the jury to draw inferences beyond their unaided competence. It may well be that to an experienced eye each of the above factors would have been dead giveaways that the cars were likely stolen. But a jury is unable to draw such inferences from its ordinary experience. Evidence was required.


This case is distinguishable from Martin for two reasons. First, it was not necessary for a witness to explain to the jury the significance of some of the red flags in this case. A jury without experience in real estate closings could nevertheless infer that Lindley should have been alerted to a high probability of wrongdoing from the fact that multiple buyers purchased properties, signed occupancy affidavits declaring their intention to reside in those properties, and then turned around and bought new properties within the week. Second, the government did call witnesses to testify to the suspicious nature of certain less obvious red flags. For example, MacPhee testified that Lindley's files only contained dual sets of loan documents for closings involving New England Merchants, the real estate company where Ernst was employed. By contrast, Shirley David, a mortgage originator who was not connected to appellants' scheme, testified that only one set of documents was used in all of the twenty or so closings on which she and Lindley worked together. The jury thus could have concluded that the existence of two sets of documents for the properties involved in appellants' scheme was sufficiently unusual that it should have aroused Lindley's suspicion. We therefore hold that there was sufficient evidence to support Lindley's conviction on a willful blindness theory.6

2. Ernst

The scope of Ernst's sufficiency of the evidence challenge is narrow. Rather than disclaim any role in appellants' scheme, Ernst contests only the evidentiary basis for seven of his thirty-four wire fraud convictions, arguing that the government did not prove that he knowingly and willingly participated in the scheme with respect to the four properties implicated in these seven counts. See 18 U.S.C. § 1343.

The four properties in question are located in the Dorchester, Mattapan, and Hyde Park neighborhoods of Boston, and in nearby Brockton. Lamerique tied Ernst to each property. He testified that it was Ernst who identified the Dorchester property as suitable for appellants' scheme and delivered a pitch to the straw buyer, also instructing Samuel Jean–Louis, another of appellants' coconspirators, how to structure the loan application and purchase-and-sale agreement. Lamerique also testified that Ernst recruited and pitched the straw buyer for the Mattapan property, collecting fees of $11,125, and that he identified and negotiated a purchase price for the Brockton property, doctoring the purchase-and-sale agreement to conceal $20,000 in illegal profits. In addition, Lamerique testified that Ernst identified the Hyde Park property and prepared the purchase-and-sale agreement, earning a commission of $11,375.

In the context of this case, and against the backdrop of Ernst's general awareness of the mechanics of appellants' scheme, which is undisputed, the foregoing evidence was sufficient for the jury to convict him of the seven counts in question.

3. Levine

Levine challenges the sufficiency of the evidence for his convictions on eight of the forty-one wire fraud counts and three of the nineteen money laundering counts. Nevertheless, Levine concedes that “the evidence, viewed in the light most favorable to the government, certainly established that [he] involved himself in some [unlawful] activities” and “could certainly support a finding that [he] conspired with various players at New England Merchants to defraud various lenders.” He maintains, though, that he did not knowingly participate in wrongdoing with respect to the four properties implicated in the eleven challenged counts, see 18 U.S.C. §§ 1343, 1957, which are located in South Boston, Brockton, Hyde Park, and Dorchester. In support of this claim, he points to evidence that he became upset when he discovered that the straw buyer for the Hyde Park property had used a stolen identity.

The evidence shows that proceeds from the South Boston, Brockton, and Hyde Park properties passed through bank accounts held by Levine. The fact that he did not want to engage in identity theft with respect to one or more of these properties has no bearing on his knowing participation in other aspects of the mortgage fraud scheme. The evidence also establishes that the Dorchester property was burned by Lamerique and another of appellants' coconspirators for insurance money, that Levine advised Lamerique how to form a shell company to fraudulently conceal the insurance money from the property's mortgage lender, and that Levine personally handled $40,000 of the insurance money. There was, in short, sufficient evidence for the jury to infer the requisite knowledge to convict Levine of the eleven counts in question.

C. The Proceeds of the Scheme (Daniel)

Daniel, who by all accounts played a bit part in appellants' scheme, argues that the district court ran afoul of Federal Rule of Evidence 404(b) in allowing Lamerique to testify that Daniel had used some of the proceeds of the scheme to purchase, among other things, firearms and marijuana. Because Daniel failed to lodge a contemporaneous objection during Lamerique's testimony, our review is for plain error. See United States v. Meserve, 271 F.3d 314, 321 (1st Cir.2001). Accordingly, Daniel must show that (1) an error occurred (2) which was clear or obvious and which not only (3) affected his substantial rights, but also (4) seriously impaired the fairness, integrity, or public reputation of judicial proceedings. See id. at 321–22.

Federal Rule of Evidence 404(b) prohibits the admission of evidence of a defendant's other crimes or bad acts to establish his or her character or propensity to commit a crime. See Fed.R.Evid. 404(b); see also United States v. Fulmer, 108 F.3d 1486, 1501 (1st Cir.1997) (“Rule 404(b) is intended to ‘forbid judging a person on the basis of innuendo arising from conduct [that] is irrelevant to the charges for which he or she is presently standing trial, i.e., against finding present guilt based on a bad character profile.’ “ (quoting United States v. Cortijo–Díaz, 875 F.2d 13, 15 (1st Cir.1989))). We employ a two-part test to evaluate whether evidence is admissible under Rule 404(b). See United States v. Pelletier, 666 F.3d 1, 5 (1st Cir.2011). First, we determine whether the evidence has special relevance—“that is, whether it is relevant to any purpose other than to prove that a defendant has a propensity to commit a crime.” Id. If so, we look to Federal Rule of Evidence 403 to determine whether the probative value of the evidence is substantially outweighed by its danger of unfair prejudice. See id.; United States v. Rodríguez–Berríos, 573 F.3d 55, 64 (1st Cir.2009).

Daniel apparently had no legitimate source of disposable income. Therefore, evidence that he used money derived from appellants' scheme to buy “marijuana, clothes, vehicles, and firearms” had special relevance because it established his motive for participating in the scheme—his need to finance a lavish lifestyle. See United States v. Cole, 631 F.3d 146, 155–56 (4th Cir.2011) (evidence of defendant's “lavish spending” was probative of motive for violating tax laws); United States v. Jackson–Randolph, 282 F.3d 369, 376–80 (6th Cir.2002) (evidence of defendant's shopping and gambling expenses showed motive for committing financial crimes); United States v. Kadouh, 768 F.2d 20, 21 (1st Cir.1985) (evidence of defendant's use of cocaine, an expensive drug, offered financial motive for trafficking in heroin). As such, the evidence had significant probative value.

The evidence also was not unfairly prejudicial. “For purposes of Rule 403, ‘unfair prejudice’ occurs where there is ‘an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one.’ “ United States v. Symonevich, No. 11–1236, 2012 WL 3083491, at *7 (1st Cir. July 31, 2012) (quoting Fed.R.Evid. 403 advisory committee's note). Here, the reference to Daniel's purchase of firearms and marijuana was brief and unaccompanied by hints of either actual gun violence or drug abuse. Therefore, we find no error in the admission of the testimony, plain or otherwise.

D. The Summary Evidence Objection (Daniel and Haltiwanger)

1. The Summary Charts

Daniel and Haltiwanger argue that the district court erred by admitting into evidence four charts summarizing the reams of financial data in this case. As with other evidentiary rulings, our review is for abuse of discretion. See United States v. McElroy, 587 F.3d 73, 80 (1st Cir.2009); United States v. Stierhoff, 549 F.3d 19, 27 (1st Cir.2008).

The government's final witness was Thomas Zappala, an auditor employed by the United States Attorney's Office. Through Zappala, the government introduced four charts that he created outlining the mechanics of appellants' scheme. The first chart depicted, for each of the twenty-one properties involved in the scheme, the buyer and seller, the difference between the actual sale price and the falsely-inflated price represented on the mortgage loan applications, and the proceeds laundered through Lindley's IOLTA. The second chart listed, for each property, the profits accruing to each conspirator. The third aligned the wire fraud counts in the indictment with the corresponding wire transfers and property sales. The fourth chart did the same for the money laundering counts.

The government argued that Zappala's charts were admissible under Federal Rules of Evidence 611 and 1006. Over objections from Daniel and Haltiwanger, the district court admitted the charts into evidence “under the rules and law relating to summaries” and made them available to the jury. The court also denied requests for Daniel and Haltiwanger for contemporaneous limiting instructions, although it did provide a limiting instruction before jury deliberations, admonishing the jury that summaries should be scrutinized closely.

As we have explained, various summary tools may be used “to clarify complex testimony and evidence” for a jury. McElroy, 587 F.3d at 81; see also United States v. Milkiewicz, 470 F.3d 390, 396–98 (1st Cir.2006); Fraser v. Major League Soccer, L.L.C., 284 F.3d 47, 67 (1st Cir.2002). Of particular relevance, Federal Rule of Evidence 1006 provides that a party may summarize the contents of voluminous writings, recordings, or photographs which cannot conveniently be examined in court, so long as the summary is accurate, the underlying documents are made available to the other parties, and both the summary and the source materials are admissible. See Fed.R.Evid. 1006; see also Milkiewicz, 470 F.3d at 396–98. A Rule 1006 summary may be offered into evidence and made available to the jury. See 31 Charles A. Wright & Victor J. Gold, Federal Practice and Procedure § 8044 (2000).

No precise test dictates when source materials are sufficiently indigestible to permit summarization under Rule 1006. Instead, district courts are advised to carefully weigh the volume and complexity of the materials. These two factors have an inversely proportionate relationship: as either the volume or complexity increases, relatively less is required of the other factor. See id . The ultimate question, of course, is whether summarization will remove logistic or cognitive barriers to the jury's discharge of its duties, see United States v. Bakker, 925 F.2d 728, 736 (4th Cir.1991); United States v. Johnson, 594 F.2d 1253, 1255 (9th Cir.1979), and we are poorly positioned to second guess a district court's on-the-spot answer to this question, see Fraser, 284 F.3d at 67 (“It is hard to imagine an issue on which a trial judge enjoys more discretion than as to whether summary exhibits will be helpful.”).

The summary evidence in this case obviated the need for the government to introduce, and the jury to sift through, mortgage and sale records for each of the twenty-one properties involved in appellants' scheme, and also facilitated tracing the scheme's proceeds through Lindley's IOLTA. As such, it comported with the purpose of Rule 1006. See Bakker, 925 F.2d at 736 (“The purpose of Rule 1006 is to provide a practicable means of summarizing voluminous information.”). However, Daniel and Haltiwanger argue that the summary evidence nevertheless should have been excluded for two reasons: (1) it impermissibly summarized testimony, in addition to documents, see McElroy, 587 F.3d at 80 (noting that Rule 1006 “only allows the introduction of summary evidence that summarizes documents, as opposed to evidence that summarizes testimony”); and (2) it was based in part on inadmissible source materials, see Colón–Fontánez v. Municipality of San Juan, 660 F.3d 17, 29–30 (1st Cir.2011) (noting that for summary evidence to be admissible, the materials on which it is based also must be admissible). These arguments are non-starters.

As to the first argument, there has been no showing that the summary charts in fact summarized testimony. To the contrary, the charts, in certain respects, summarized documentary evidence that corroborated witness testimony. For example, Haltiwanger claims that the charts should not have linked her receipt of $17,000 to her participation in appellants' scheme because the only evidence drawing that connection was Lamerique's testimony. However, the documentary evidence on which the charts were based reveals that Haltiwanger received $17,000 in two installments from a bank account associated with appellants' scheme within three months of purchasing one of the properties involved in the scheme.

Similarly, Daniel asserts that the summary charts should not have indicated that he received three checks totaling $16,000 because the only proof that he earned any money from appellants' scheme came from Lamerique's testimony. The checks themselves, though, corroborated that testimony and were appropriately provided to the jury. Two of the checks were endorsed by Daniel, and the third was made out to him by Levine. That a particular fact may have been the subject of testimony does not mean that it cannot be corroborated through admissible documents summarized under Rule 1006.

As to the second argument, we understand Daniel and Haltiwanger to be contending that the summary evidence relied upon source materials, such as the sale records for the properties involved in the scheme, that were inadmissible for the truth of the matter asserted and thus barred by the hearsay rule codified in Federal Rule of Evidence 801. See Milkiewicz, 470 F.3d at 398 n. 15 (“The records summarized must ․ be admissible.”). This contention misses the point. The sale records were admissible, not to prove that the purchase prices reflected the true value of the properties, but for the limited purpose of showing that these prices were lower than those listed on the mortgage loan applications. Accordingly, they were not within the scope of the hearsay rule. See DeCologero, 530 F.3d at 58 (1st Cir.2008) (“ ‘[I]f the significance of an offered statement lies solely in the fact that it was made, no issue is raised as to the truth of anything asserted, and the statement is not hearsay.’ “ (quoting Fed.R.Evid. 801(c) advisory committee's note)). There was no abuse of discretion in the admission of the summary charts under Rule 1006.7

2. Zappala's Testimony

Citing our decision in United States v. Flores–De–Jesús, 569 F.3d 8 (1st Cir.2009), Daniel and Haltiwanger also challenge Zappala's testimony as a summary witness. In Flores–De–Jesús, we expressed concern that the government's use of a summary witness can generate imbalances in a trial if the summary witness endorses the credibility of other witnesses. See 569 F.3d at 18–19; see also United States v. Moore, 651 F.3d 30, 56 (D.C.Cir.2011); United States v. Fullwood, 342 F.3d 409, 413–14 (5th Cir.2003). That problem did not arise here. Zappala did not bolster the trial testimony of other witnesses. His testimony was limited to introducing and explaining the summary charts he prepared. There was no abuse of discretion in permitting him to testify.

E. The Willful Blindness Instruction (Lindley and Haltiwanger)

Lindley and Haltiwanger argue that the district court should not have instructed the jury on willful blindness.8 “We have not definitively resolved what standard of review we apply to the district court's decision to give a willful blindness instruction,” United States v. Anthony, 545 F.3d 60, 64 (1st Cir.2008), and have oscillated between “de novo and deferential standards of review,” Azubike, 564 F.3d at 66 n. 5. However, “[w]e need not determine the issue in this case, because applying either standard, the evidence supported the district court's decision to charge the jury on willful blindness.” United States v. Mitrano, 658 F.3d 117, 123 (1st Cir.2011); see also Anthony, 545 F.3d at 64.

A willful blindness instruction is appropriate if three requirements are met: “ ‘(1) a defendant claims a lack of knowledge, (2) the facts suggest a conscious course of deliberate ignorance, and (3) the instruction, taken as a whole, cannot be misunderstood as mandating an inference of knowledge.’ “ Mitrano, 658 F.3d at 123 (quoting Azubike, 564 F.3d at 66). Of the second requirement, which is the only one contested in this case, we have said that, “[i]n determining whether the facts suggest the type of deliberate avoidance warranting a willful blindness instruction, we must consider whether the record evidence reveals flags of suspicion that, uninvestigated, suggest willful blindness.” Id. (internal quotation marks omitted); see also Azubike, 564 F.3d at 66 (“Direct evidence of willful blindness is not required; what is needed are sufficient warning signs that call out for investigation or evidence of deliberate avoidance of knowledge.”).

We have already established that the evidence adduced at trial was sufficient to prove beyond a reasonable doubt that Lindley willfully blinded himself to the existence of appellants' scheme. This holding necessarily means that a willful blindness instruction was appropriate as to Lindley. See Azubike, 564 F.3d at 69.

The instruction also was appropriate as to Haltiwanger, an experienced residential mortgage broker. The evidence shows that Ernst and two of appellants' coconspirators pitched the mortgage fraud scheme to Haltiwanger, promising that if she served as a straw buyer she “should be able to make ․ around $10,000, and that the mortgage would be taken care of.” While waiting for the conspirators to identify a property for which she could be the straw buyer, Haltiwanger processed falsified mortgage loan applications on behalf of other straw buyers through her employer, Topdot Mortgage Company, and collected commissions. On two of these loan applications, Haltiwanger misrepresented the straw buyer's employer as Oligarchy Funding, where Ernst was a principal. She did the same on her own loan application. The government argues, and we agree, that, “[g]iven this ample evidence that ․ Haltiwanger proceeded with multiple transactions in the face of circumstances that, as [a] real estate professional [ ], [she] could only have deemed fraudulent, the district court did not err in permitting the jury to consider whether [her] actions evinced willful blindness.”

Haltiwanger protests that the government forfeited its right to a willful blindness instruction by stating in its closing argument that she “knew what was going on,” implying that she had actual knowledge of appellants' scheme. “These theories can coexist,” however. Griffin, 524 F.3d at 79; see also Azubike, 564 F.3d at 67–69; United States v. Cassiere, 4 F.3d 1006, 1024 (1st Cir.1993) (“Although the government's main contention at trial was that all three defendants were knowing participants in the scheme, the government presented evidence from which the jury could have concluded that if they did not know what was going on, it was only because they chose to turn a blind eye.”). The evidence against Haltiwanger could support either a finding of actual knowledge or a finding of willful blindness and “did not require the jury to make a binary choice between actual knowledge and innocence.” Azubike, 564 F.3d at 68. As a result, the district court did not err by giving a willful blindness instruction.

F. Haltiwanger's Jury Nullification Challenge

Haltiwanger argues that the district court should not have instructed the jury that it had a duty to return a guilty verdict if it concluded that the government had proven its case beyond a reasonable doubt.9 She complains that this instruction wrongly suggested that the jury lacked the power to nullify. See United States v. Thomas, 116 F.3d 606, 615 (2d Cir.1997) (defining nullification as “a practice whereby a juror votes in purposeful disregard of the evidence, defying the court's instructions on the law”). Because no objection to this instruction was made at trial, our review is for plain error. See United States v. Troy, 618 F.3d 27, 33 (1st Cir.2010). Here, there was no error at all.

“We have consistently held that a district court may not instruct the jury as to its power to nullify.” United States v. Manning, 79 F.3d 212, 219 (1st Cir.1996); see also United States v. Bunchan, 626 F.3d 29, 34 (1st Cir.2010) (“Neither the court nor counsel should encourage jurors to exercise their power to nullify.”); United States v. Sepulveda, 15 F.3d 1161, 1190 (1st Cir.1993) (“Though jury nullification has a long and sometimes storied past, the case law makes plain that a judge may not instruct the jury anent its history, vitality, or use.” (internal citation omitted)); United States v. DesMarais, 938 F.2d 347, 350 (1st Cir.1991) (“[I]t would [be] improper to urge the jury to nullify applicable law.”). This is because “jurors may have the power to ignore the law, but their duty is to apply the law as interpreted by the court, and they should be so instructed.” United States v. Boardman, 419 F.2d 110, 116 (1st Cir.1969); see also Bunchan, 626 F.3d at 34 (“A juror's duty is to apply the law as provided by the court.”); Sepulveda, 15 F.3d at 1190 (“The applicable rule is that, although jurors possess the raw power to set an accused free for any reason or for no reason, their duty is to apply the law as given to them by the court.”).

In light of these precedents, it is hardly a stretch to hold explicitly, as we now do, that a district court may instruct a jury that it has a duty to return a guilty verdict if convinced beyond a reasonable doubt of a defendant's guilt on a particular charge. See United States v. Carr, 424 F.3d 213, 219–20 (2d Cir.2005) (affirming instruction implying that nullification was not an option, because “[n]othing in our case law begins to suggest that the court cannot ․ tell the jury affirmatively that it has a duty to follow the law, even though it may in fact have the power not to”); United States v. Pierre, 974 F.2d 1355, 1357 (D.C.Cir.1992) (per curiam) (“[I]t was proper for the district court to instruct the jury that it had a duty to find appellant guilty if the government proved beyond a reasonable doubt every element of the offense with which he was charged.”).

G. Lindley's Closing Argument Challenge

During its closing argument, the government suggested that Lindley joined the mortgage fraud scheme in order to stay in Levine's good graces and thereby preserve for himself the lucrative real estate practice he inherited from Levine, which the government valued at “a quarter million dollars.” Lindley objected that the government's valuation was unsupported by the evidence,10 and later moved for a new trial on the same basis. The district court overruled the objection and denied the motion for a new trial, explaining that the reference to Lindley's financial motive was borne out by the record. We review de novo whether the challenged portion of the government's closing argument was improper and, if so, whether it was harmful, but we review the denial of Lindley's motion for a new trial only for manifest abuse of discretion. See United States v. Manor, 633 F.3d 11, 16–17 (1st Cir.2011); United States v. Nelson–Rodriguez, 319 F.3d 12, 38 (1st Cir.2003).

As a general principle, the government is permitted in its closing argument to attribute to a defendant a motive that is consistent with the evidence, see United States v. Torres–Rosario, 658 F.3d 110, 113–14 (1st Cir.2011); United States v. Meadows, 571 F.3d 131, 145 (1st Cir.2009), particularly where the defendant first placed in issue his or her motive or lack thereof, see United States v. Derman, 211 F.3d 175, 180 (1st Cir .2000). In this case, we see nothing improper about the government's reference to Lindley's financial motive, which was a fair rejoinder to Lindley's earlier exhortation to the jury to “follow the money” and his declaration that he had no pecuniary stake in appellants' scheme. Levine's administrative assistant testified that Levine had earned approximately $20,000 per month, or $240,000 per year, from the real estate practice he later transferred to Lindley. That these figures necessarily were estimates, and perhaps were extrapolated from unusually busy periods in Levine's professional career, does not negate the inference that Lindley was drawn to the scheme by the allure of “a quarter million dollars,” or thereabouts. There was no mistake in overruling Lindley's objection, and no manifest abuse of discretion in the denial of his motion for a new trial. See Manor, 633 F.3d at 19.

H. Loss Calculation (Ernst and Levine)

Ernst and Levine challenge the eighteen-level enhancements that the district court added to their sentences pursuant to U.S.S.G. § 2B1.1(b)(1)(J) for engendering losses between $2,500,000 and $7,000,000. See United States v. Innarelli, 524 F.3d 286, 290 (1st Cir.2008). We review the district court's loss calculation methodology de novo. See United States v. Walker, 234 F.3d 780, 783 (1st Cir.2000) (“The appropriate method for calculating loss amounts ․ is a prototypical question of legal interpretation, and we review de novo.”). The mathematical application of this methodology is reviewed only for clear error. See Vázquez–Botet, 532 F.3d at 65; United States v. Cacho–Bonilla, 404 F.3d 84, 92 (1st Cir.2005).

U.S.S.G. § 2B1.1 increases a defendant's base offense level for fraud according to the amount of pecuniary loss caused by the defendant. As a general rule, this amount is “the greater of actual loss or intended loss.” U.S.S.G. § 2B1.1 cmt. n. 3(A); see also Innarelli, 524 F.3d at 290. As the term implies, actual loss is the reasonably foreseeable loss that actually resulted from an offense. See U.S.S.G. § 2B1.1 cmt. n. 3(A)(i). The extent of actual loss may depend on fortuities that minimize or exacerbate the effects of the defendant's fraudulent conduct. Intended loss is the loss that the defendant could have reasonably expected to occur at the time he or she perpetrated the fraud. See Innarelli, 524 F.3d at 290; see also United States v. McCoy, 508 F.3d 74, 79 (1st Cir.2007). In that respect, intended loss is frequently a better measure of culpability than actual loss. See McCoy, 508 F.3d at 79.

In cases where a defendant has pledged collateral to secure a fraudulent loan, actual loss usually can be calculated by “subtracting the value of the collateral—or, if the lender has foreclosed on and sold the collateral, the amount of the sales price—from the amount of the outstanding balance on the loan.” United States v. James, 592 F.3d 1109, 1114 (10th Cir.2010); see also United States v. Parish, 565 F.3d 528, 535 (8th Cir.2009); McCoy, 508 F.3d at 79. “[T]he damage wrought by fraud is sometimes difficult to calculate,” however. United States v. Agboola, 417 F.3d 860, 870 (8th Cir.2005). If actual loss cannot be determined, a district court may safely use intended loss in its computations, and vice versa. Of course, if both can be determined, the Guidelines require the use of the larger amount. If neither actual loss nor intended loss can be gauged, a district court may use, as a last resort, “the gain that resulted from the offense as an alternative measure of loss.” U.S.S.G. § 2B1.1 cmt. n. 3 (B).

The application of these principles in mortgage fraud cases must account for the fact that the original mortgage lender frequently is not the lender who forecloses on a property and receives the proceeds from the foreclosure sale. See James, 592 F.3d at 1115. Even if the original lender sells the mortgage to a successor lender, though, and there are subsequent transactions of the same kind, actual loss is always the difference between the original loan amount and the final foreclosure price (less any principal repayments). The commentary to U.S.S.G. § 2B1.1 “does not direct us to focus on harm to any particular victim; rather, it mandates that we focus on the total loss resulting from the commission of fraud to the extent the total loss is reasonably foreseeable.” Id. at 1117 (Lucero, J., concurring) (citing U.S.S.G. § 2B1.1 cmt. n. 3(A)(i)); see also United States v. Snow, 468 F. App'x 830, 840 (10th Cir.2012). Thus, provided that the total actual loss is reasonably foreseeable, its apportionment as between the original lender and a successor lender (or other downstream purchaser) does not matter. The same is true of intended loss. The focus is not on any particular lender to the exclusion of others, but rather on the total degree of loss that the defendant could have reasonably expected to occur. See United States v. Bonanno, 146 F.3d 502, 509–10 (7th Cir.1998) (“[T]he relevant inquiry is ․ ‘How many dollars did the culprits' scheme put at risk?’ ”).

With these principles in mind, we turn our attention to the case at bar. Most of the mortgages at issue here were sold by the original lenders to successor lenders (and, in some instances, resold by the successor lenders to other downstream purchasers) prior to the commencement of foreclosure proceedings. The district court assumed that it was precluded from making an actual loss determination due to its inability to ascertain which entities had suffered losses, and in what amounts, for any given property. Instead, the court relied on intended loss, arriving at a loss amount in excess of $2,500,000.11

The court was wrong to assume that it was incapable of making an actual loss determination merely because it could not tell which entities had lost what amounts of money. The relevant metric is total actual loss, not loss to any particular victim. See Snow, 468 F. App'x at 840; James, 592 F.3d at 1117 (Lucero, J., concurring).12 Thus, for each property, the court should have calculated actual loss by subtracting from the outstanding balance on the mortgage loan either the sum recouped via foreclosure or, if there was no foreclosure, the property's fair market value at the time of sentencing. See James, 592 F.3d at 1114; Parish, 565 F.3d at 535. Nevertheless, this error did not have any prejudicial effect and, hence, does not require resentencing. See United States v. Roman–Portalatin, No. 11–1542, 2012 WL 1418504, at *2 (1st Cir. Apr.25, 2012) (“Prejudice is ordinarily a necessary condition for any order for resentencing․”); United States v. Larios, 593 F.3d 82, 89 (1st Cir.2010). If actual loss was lower than intended loss, it was correct for the court to rely on intended loss. See U.S.S.G. § 2B1.1 cmt. n. 3(A). If intended loss was lower, the court's error benefitted Levine and Ernst.

In fact, however, the intended loss formula employed by the court was substantially similar to the actual loss formula described above: the court subtracted from the original mortgage loan amount for each property either the foreclosure sales price or, if there was no foreclosure, an estimate of the property's assessed value at the time of sentencing.13

Ernst argues that this formula overstates his culpability because the substantial disparity between the original loan amounts and the properties' final values was the result of a real estate market collapse that he could not reasonably have expected:

The collapse of the sub-prime market—indeed, even the conventional credit market—had a devastating effect on real estate values in precisely the time frame these properties were being resold or auctioned. The extent of the decline was not one that was reasonably foreseeable and it is unfair and unreasonable to hold Ernst accountable for those declines in the length of his sentence.

This issue poses a philosophical concern larger than the immediate effect on Ernst's sentencing. If extreme market fluctuations are allowed to dictate a defendant's sentence in a case of this nature, then the corollary will also be true. In a case of an identical nature, with an identical modus operandi and intended result, where the market remains nearly static, then the offenders will be punished significantly less because the values of the properties did not decline. It makes no penological sense to impose disparate punishment on similar situated defendants who engage in identical behavior with the same intended gain.

Thus, without a fair way to assess loss, Ernst contends that the court should have used the gain that appellants derived from their scheme as an alternative measure. See U.S.S.G. § 2B1.1 cmt. n. 3 (B). The formula that he proposes for calculating gain is based on the difference between the purchase prices negotiated with the properties' sellers and the artificially-inflated prices reported to the original mortgage lenders. Levine advocates the same formula. This approach would yield a sum of $1,770,000, which in turn would entitle Ernst and Levine to a two-level reduction in their total offense levels. See id. § 2B1.1(b)(1)(I).

There is no need to resort to gain here. The district court's intended loss formula was a reasonable proxy for culpability in the circumstances of this case. See McCoy, 508 F.3d at 79 (“Intended loss was therefore the value of the loans less the expected value of the properties.”). Levine and Ernst were veterans of the real estate industry. They knew that the mortgage loans on the properties involved in their scheme would enter default and that most, if not all, of the properties would be forced into foreclosure. They could reasonably have anticipated that the properties would be grossly devalued as a result. Even if the deterioration of the Boston real estate market during the recent recession also played some macroeconomic role in that outcome, Levine and Ernst could reasonably have expected that they were contributing to the emergence of those poor market conditions. See Parish, 565 F.3d at 535 (in measuring actual loss, explaining that defendants could have reasonably foreseen the depressing impact their mortgage fraud scheme would have on local markets and property values); United States v. Shattuck, 961 F.2d 1012, 1016–17 (1st Cir.1992).

However, the intended loss formula used by the court will not work where the real estate market outperforms a defendant's expectation that a property will be devalued. In that scenario, subtracting the property's final value from the original mortgage loan amount will not accurately reflect the defendant's culpability and, hence, will be an unreliable gauge of intended loss. A different intended loss formula will be necessary to punish the defendant for the full amount of loss he or she could reasonably have expected to occur rather than the more modest loss that actually occurs. See Innarelli, 524 F.3d at 291 (“Where, as here, the defendant reasonably should have expected that loss would result, he can and generally should be punished more severely to account for his greater level of moral culpability, even where the victim has managed to make money in spite of the fraud.”).

We do not share Ernst's “philosophical concern” with this problem because the Sentencing Guidelines anticipate it. If a reliable intended loss formula cannot be devised, total actual loss must be used. If the resulting actual loss amount “substantially understates the seriousness of the offense, ․ an upward departure may be warranted.” U.S.S.G. § 2B1.1 cmt. n. 19(A). With the availability of an upward departure, the fortuities of the market that might make actual loss a poor proxy for culpability can be addressed.

In summary, there was no error in the district court's loss calculation methodology and none in its mathematical application of this methodology, which produced an intended loss amount within the range contemplated by U.S.S.G. § 2B1. 1(b)(1)(J).

I. Levine's Role–in–the–Offense Challenge

Levine argues that the district court should not have incorporated into his sentence a four-level enhancement pursuant to U.S.S.G. § 3B1.1(a) for his role as an organizer or leader of appellants' scheme. We review the imposition of this particular sentencing enhancement, and any predicate factual findings, for clear error. See United States v. Alfonzo–Reyes, 592 F.3d 280, 295 (1st Cir.2010) (“A court's decision to impose a sentencing enhancement for a leadership role based on the facts is reviewed for clear error.”); United States v. Martínez–Medina, 279 F.3d 105, 123 (1st Cir.2002) (“We review role-in-the-offense determinations, steeped in the facts of the case, for clear error.”); United States v. Wright, 873 F.2d 437, 443–44 (1st Cir.1989) (explaining rationale for applying clear error review).

“In order to invoke § 3B1.1(a), a district court must make a finding as to scope—that the criminal activity involved five or more participants or was otherwise extensive—and a finding as to status—that the defendant acted as an organizer [or] leader of the criminal activity.” United States v. Arbour, 559 F.3d 50, 53 (1st Cir.2009); see also United States v. TejadaBeltran, 50 F.3d 105, 111 (1st Cir.1995). Levine challenges only the district court's status finding, protesting that he was merely an ancillary participant in appellants' scheme and that others were equally or more culpable. There is no merit to this challenge.

A defendant acts as a leader if he or she exercises some degree of dominance or power in a criminal hierarchy and has the authority to ensure that others will follow orders. See United States v. Aguasvivas–Castillo, 668 F.3d 7, 15 (1st Cir.2012); Arbour, 559 F.3d at 55. A defendant qualifies as an organizer if he or she “coordinates others so as to facilitate the commission of criminal activity.” Tejada–Beltran, 50 F.3d at 111. Factors that are relevant to determining the supervisory nature of a defendant's role include: (1) the exercise of decision-making authority; (2) the nature of the participation in the commission of the offense; (3) the recruitment of accomplices; (4) the claimed right to a larger share of the fruits of the crime; (5) the degree of participation in planning or organizing the offense; (6) the nature and scope of the illegal activity; and (7) the degree of control and authority exercised over others. See U.S.S.G. § 3B1.1 cmt. n. 4; Aguasvivas–Castillo; 668 F.3d at 15. Of course, “[t]here need not be proof of each and every factor before a defendant can be termed an organizer or leader.” Tejada–Beltran, 50 F.3d at 111.

The evidence clearly establishes that Levine masterminded appellants' scheme. Levine inducted Lindley into the scheme and largely guided his actions. See United States v. CarreroHernández, 643 F.3d 344, 352 (1st Cir.2011); cf. Arbour, 559 F.3d at 56 (“[A] defendant needs only to have led or organized one criminal participant, besides himself of course, to qualify as a leader or organizer under § 3B1.1(a).”). Levine also directed the flow of the scheme's proceeds from Lindley's IOLTA to the conspirators, usually via his own bank accounts, see AguasvivasCastillo, 668 F.3d at 15 (holding that defendant qualified as an organizer or leader in part because of his financial control over fraudulent scheme), and he dictated who was authorized to discuss the scheme with whom. As a result, even if others, too, had supervisory roles, the district court did not clearly err in determining that Levine was an organizer or leader and sentencing him accordingly. See U.S.S.G. § 3B1.1 cmt. n. 4; (“There can, of course, be more than one person who qualifies as a leader or organizer of a criminal association or conspiracy.”); see also United States v. Casas, 356 F.3d 104, 109 (1st Cir.2004) (“The mere fact that [the defendant] was subordinate to [a coconspirator] does not establish, without more, that [the defendant] was not an organizer or leader of the conspiracy.”).

J. Levine's Motion for an Evidentiary Hearing or Discovery

Levine argues that the district court erred in denying his post-trial motion for an evidentiary hearing or discovery. Our review is for abuse of discretion. See United States v. Cartagena, 593 F.3d 104, 112 (1st Cir.2010); United States v. Theodore, 354 F.3d 1, 7 (1st Cir.2003).

During the trial, Levine's counsel, Isaac Borenstein, began to suspect that a paralegal on Levine's defense team, Melanie Abbruzzese, was romantically involved with a postal inspector on the government's prosecution team, Joseph McGonagle. Although Borenstein warned Abbruzzese that it would be inappropriate for her to have a romantic relationship with McGonagle, he did not initially take any further action or inform Levine of his suspicions.

Two weeks later, Abbruzzese disclosed to Borenstein that she and McGonagle had spoken briefly about this case. The next day, Borenstein alerted both Levine and the district court to what Abbruzzese had told him. The court promptly held a closed hearing at which Abbruzzese and McGonagle denied under oath that they were romantically linked and swore that their relationship was strictly professional. They also swore that they had not exchanged any confidential information. On the basis of this testimony, the court permitted the trial to proceed.

After Levine's conviction, but before his sentencing, Borenstein informed the government that Abbruzzese and McGonagle had lied at the closed hearing concerning the nature of their relationship.14 Borenstein's law firm, Denner Pellegrino, then withdrew from Levine's representation. Levine's replacement counsel moved for an evidentiary hearing or discovery to determine whether to pursue a new trial, or any other remedy, on the ground that Levine's defense had been compromised and his constitutional right to conflict-free counsel infringed. See Wood v. Georgia, 450 U.S. 261, 271, 101 S.Ct. 1097, 67 L.Ed.2d 220 (1981) (“Where a constitutional right to counsel exists ․ there is a correlative right to representation that is free from conflicts of interest.”). The motion asserted that an evidentiary hearing or discovery was necessary for Levine “to obtain evidence from sources currently unavailable to him—including Abbruzzese, McGonagle, other members of the prosecution team, and documents that have been requested from, but withheld by, the [government].” Attached to the motion were excerpts from the depositions of two Denner Pellegrino employees describing the romantic relationship between Abbruzzese and McGonagle.

At an initial status conference on Levine's motion, the court surmised that Abbruzzese and McGonagle would invoke their constitutional privilege against self-incrimination, see U.S. Const. amend. V, and refuse to testify at an evidentiary hearing. The court also reminded the government of its continuing obligation under Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), to disclose any material evidence favorable to Levine, and ordered the government to submit affidavits stating whether McGonagle had shared confidential information derived from his relationship with Abbruzzese with any other member of the prosecution team. The government responded with seven affidavits from prosecution team members making clear that no such information had been received.

At a second status conference, the court emphasized that it was not requiring Levine “to actually demonstrate prejudice” and was, instead, merely looking for a threshold showing that confidential information from Abbruzzese was conveyed by McGonagle to the rest of the prosecution team. The court then explained that the government's affidavits, which it accepted as true, established that the prosecution team had not knowingly received any such information, and that the only other possibility was that “in some way unknown to the prosecution team they were fed information that worked to Mr. Levine's detriment.” The court concluded, however, that the evidence against Levine was “so overwhelming” that it was “virtually impossible” that any information the prosecution team received unknowingly could have influenced the outcome of this case. As a result, the court found that further inquiry was not warranted and denied Levine's motion.15

We recognize the seriousness of the misconduct at issue here. However, Levine has not shown how either an evidentiary hearing or discovery would have been likely to produce any evidence previously unavailable to him and, thus, help him establish a claim for relief. The seven affidavits submitted by the government assert that McGonagle did not share confidential information from Abbruzzese with the rest of the prosecution team. The only people who were in a position to contradict that assertion were McGonagle and, to a lesser extent, Abbruzzese. However, we have no reason to second guess the district court's assumption that McGonagle and Abbruzzese would not have testified at an evidentiary hearing in light of their potential criminal liability. In these circumstances, the incremental value of further inquiry is dubious at best, and we cannot say that the court abused its discretion in denying Levine's motion. See United States v. Vigneau, 337 F.3d 62, 70 (1st Cir.2003) (affirming denial of motion for evidentiary hearing where defendant's “motion and brief spoke only in general terms about the new evidence available to him”); United States v. Rodriguez, 162 F.3d 135, 148 (1st Cir.1998) (“Conclusory allegations and pure speculation, without more, do not merit an evidentiary hearing.”); Shattuck, 961 F.2d at 1015 (“At no time did appellant identify any evidence which would be presented at a hearing, so as to enable the district court to evaluate the usefulness of an evidentiary hearing.”).


As we discern no error by the district court, appellants' convictions and sentences are affirmed.

So ordered.


1.  Five of appellants' coconspirators pleaded guilty before trial. The sixth, Ralph Appolon, was tried separately because his attorney withdrew shortly before trial. He was convicted on February 22, 2011, and his appeal is currently pending before this court. See United States v. Appolon, No. 11–1627.

2.  Daniel was sentenced to forty-two months of imprisonment and three years of supervised release; Ernst to one hundred and twenty months of imprisonment and three years of supervised release; Haltiwanger to thirty months of imprisonment and two years of supervised release; Lindley to seventy-two months of imprisonment and three years of supervised release; and Levine to one hundred and forty-four months of imprisonment and two years of supervised release.

3.  The Real Estate Settlement Procedures Act of 1974, 12 U.S.C. §§ 2601–2617, requires that a HUD–1 settlement statement be used in every real estate settlement “involving a federally related mortgage loan in which there is a borrower and a seller.” 24 C.F.R. § 3500.8(a). Among other things, the HUD–1 form is meant to “conspicuously and clearly itemize all charges imposed upon the borrower and all charges imposed upon the seller in connection with the settlement.” 12 U.S.C. § 2603; see also BancOklahoma Mortg. Corp. v. Capital Tide Co., 194 F.3d 1089, 1096 (10th Cir.1999) (“HUD–1 forms provide a detailed account of the disbursements of money [that] borrowers are to receive.”).

4.  Levine was also represented by another Denner Pellegrino attorney, Bruce Levin. Ernst does not challenge Levin's participation in Levine's defense.

5.  In his briefing, Lindley claims that, “after being told by Elizabeth Son that she was not in fact the true owner of the house she had ‘bought,’ he took her to the Boston Police.” The record reveals a different scenario. The relevant testimony came from Son. She explained that, after someone was murdered in her house, Lindley accompanied her to a condemnation hearing. In the course of that proceeding, Lindley and Son met with two homicide detectives from the Boston Police Department. It was during that meeting that Son revealed to the police officers that she “just had the house underneath [her] name.” There was no testimony at trial that Lindley brought Son to the meeting for the purpose of making that disclosure or that they had discussed ahead of time the fact that Son was a straw buyer.

6.  The government also argues that other red flags existed. For example, for seventeen of the twenty-one properties involved in appellants' scheme, MacPhee gave Lindley checks to sign that split the loan proceeds due to the seller into two payments. However, a lay jury does not have the specialized knowledge to appreciate why splitting loan proceeds into two payments should have warned Lindley of a high probability of wrongdoing. None of the witnesses called by the government explained whether or not sellers usually receive loan proceeds in one lump sum. Therefore, it was beyond the jury's “unaided competence” to view the check-splitting as a red flag. Martin, 815 F.2d at 826.

7.  The court did not specify that it was admitting the summary charts under Rule 1006, simply stating that it was admitting the charts “under the rules and law relating to summaries.” This statement is obviously broad enough to encompass Rule 1006. Still, it would be a better practice if the court specified which evidentiary rule it was relying upon because these summaries are subject to different rules with different requirements and purposes. See Milkiewicz, 470 F.3d at 395–98.

8.  Drawing on pattern jury instructions, see Pattern Criminal Jury Instructions for the District Courts of the First Circuit § 2.15 (1998), the court instructed the jury as follows:When considering whether a defendant has acted knowingly, you may infer that a person has knowledge of a particular fact if you find that that person deliberately closed his eyes to a fact that otherwise would have been obvious to him. A conscious and deliberate attempt to avoid information or enlightenment is sometimes called willful blindness. That is, someone who is willfully blind to a fact that would under ordinary circumstances otherwise be obvious to him or her.In order to infer the fact of a person's knowledge of something by reason of willful blindness, you must find two things have been established: First, that that person, the defendant, was aware of a high probability that the fact was true; and second, that the defendant conscientiously and deliberately avoided learning the fact. That is to say, the defendant willfully made himself blind to the fact. If that's happened, you may attribute knowledge to that person by reason of this doctrine.Now, that does not mean that a person who was careless or negligent in learning what the fact are or makes a mistake in learning the facts would be guilty of having that knowledge. There must be a deliberate attempt, an effort made to remain ignorant of the fact.As we have noted previously, although pattern instructions may be helpful, they are only a guide and “have not been officially adopted by th[is] court.” United States v. Charlton, 502 F.3d 1, 3 n. 2 (1st Cir.2007); see also United States v. Jadlowe, 628 F.3d 1, 17 (1st Cir.2010).

9.  In pertinent part, the court instructed the jury:The burden of proof rests with the government. A defendant assumes no burden to prove that he is innocent. The question is never which side has convinced me but, rather, has the government convinced me beyond a reasonable doubt that the defendant is guilty? If the answer to that question is yes, then the government is entitled to your verdict of conviction. If that answer is no, then the defendant is entitled to be and must be acquitted.․The government must establish each element of an offense by proof that convinces you and leaves you with no reasonable doubt and thus satisfies that you can consistent with your oath as jurors base your verdict on it. Again, if you are so convinced, it is your duty to return a verdict of guilty. If, on the other hand, you have a reasonable doubt as to whether the defendant is guilty of any crime charged, the defendant must be given the benefit of that doubt and you must find him not guilty.

10.  Lindley's objection was not made until the next day. Nevertheless, because closing arguments had not yet concluded, the district court could have taken any necessary corrective action. As a result, and as the government concedes, the objection was timely. As we observed in United States v. Mandelbaum, “[a]stricture governing the timing of objections should not be employed woodenly, but should be applied where its application would serve the ends for which it was designed.” 803 F.2d 42, 44 n. 1 (1st Cir.1986) (internal quotation marks omitted); see also United States v. Azubike, 504 F.3d 30, 39 n. 9 (1st Cir.2007).

11.  The court did not pinpoint an exact amount of loss.

12.  Ernst and Levine cite James for the proposition that only actual losses incurred by the original mortgage lenders, and not the successor lenders, should be counted. They contend that the district court's inability to determine the extent of the original lenders' losses thus precluded an actual loss determination. This argument is wrong. James involved a scheme identical to the one in this case. See 592 F.3d at 1111. However, the sole reason that only the original lenders' actual losses were tallied in James was that the district court had made an uncontested factual finding that the successor lenders were not reasonably foreseeable victims, see id. at 1112, as the concurrence explained, see id. at 1118 (Lucero, J., concurring), and subsequent cases have reiterated, see Snow, 468 F. App'x at 840; United States v. Washington, 634 F.3d 1180, 1184–85 (10th Cir.2011). There was no such finding in this case and, hence, no cause for assessing only losses to the original lenders and not the successor lenders. See Snow, 468 F. App'x at 840.

13.  The only difference between this formula and the actual loss formula is the use of the original mortgage loan amount rather than the outstanding loan balance as the baseline figure. See Snow, 468 F. App'x at 839 n. 6. Because it is unclear from the record whether any principal repayments were made on the loans, we cannot say with certainty whether these figures differ in this case and, if so, to what extent. However, given the relatively short lifespan of the loans and the fact that appellants' scheme was based on allowing the loans to default, any difference between the original loan amounts and the outstanding balances is probably not significant.

14.  The government subsequently charged Abbruzzese and McGonagle with obstruction of justice, in violation of 18 U.S.C. § 1503, and perjury, in violation of 18 U.S.C. § 1623. Abbruzzese pleaded guilty on November 30, 2011, and was sentenced on March 14, 2012. McGonagle pleaded guilty on January 26, 2012, and was sentenced on May 3, 2012. They were each sentenced to a prison term of one year and a day, to be followed by a two-year period of supervised release.

15.  After filing his direct appeal in this case, Levine filed a petition for post-conviction relief under 28 U.S.C. § 2255. The district court denied Levine's petition without prejudice to its renewal upon resolution of this appeal. On April 13, 2012, we denied Levine's request for a certificate of appealability from that order. See Levine v. United States, No. 11–1940.

LIPEZ, Circuit Judge.

Copied to clipboard