United States v. Herbert Vederman

PRECEDENTIAL UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT _____________ Nos. 16-4397, 16-4410, 16-4411, 16-4427, 17-1346 _____________ UNITED STATES OF AMERICA Appellant in 17-1346 v. CHAKA FATTAH, SR., Appellant in 16-4397 KAREN NICHOLAS, Appellant in 16-4410 ROBERT BRAND, Appellant in 16-4411 HERBERT VEDERMAN, Appellant in 16-4427 _____________ On Appeal from the United States District Court for the Eastern District of Pennsylvania District Court Nos. 2-15-cr-00346-001, 2-15-cr-00346-002, 2-15-cr-00346-003, 2-15-cr-00346-004 District Judge: The Honorable Harvey Bartle III Argued January 18, 2018 Before: SMITH, Chief Judge, GREENAWAY, JR., and KRAUSE, Circuit Judges (Filed: January 16, 2019) Andrea G. Foulkes Eric L. Gibson Paul L. Gray Robert A. Zauzmer Office of United States Attorney 615 Chestnut Street Suite 1250 Philadelphia, PA 19106 Jonathan Ian Kravis [ARGUED] United States Department of Justice Criminal Division, Public Integrity Section 1400 New York Avenue, N.W. Washington, DC 20005 Counsel for the United States Mark M. Lee Bruce P. Merenstein [ARGUED] Samuel W. Silver 2 Schnader Harrison Segal & Lewis 1600 Market Street Suite 3600 Philadelphia, PA 19103 Counsel for Appellant Fattah Ann C. Flannery [ARGUED] Suite 2700 1835 Market Street Philadelphia, PA 19103 Lisa A. Mathewson Suite 810 123 South Broad Street Philadelphia, PA 19109 Counsel for Appellant Nicholas Alan Silber Pashman Stein Walder Hayden 21 Main Street Suite 200 Hackensack, NJ 07601 Counsel for National Association of Criminal Defense Lawyers, Amicus Appellant Nicholas Mira E. Baylson Barry Gross [ARGUED] Meredith C. Slawe Drinker Biddle & Reath One Logan Square 3 Suite 2000 Philadelphia, PA 19103 Counsel for Appellant Brand Henry W. Asbill Buckley Sandler 1250 24th Street, N.W. Suite 700 Washington, DC 20037 Glen D. Nager [ARGUED] Jacob M. Roth Julia W. M. F. Sheketoff Jones Day 51 Louisiana Avenue, N.W. Washington, DC 20001 Counsel for Appellant Vederman ________________ OPINION ________________ SMITH, Chief Judge. 4 Table of Contents I. Introduction ................................................................... 8 II. Background .................................................................. 9 A. The Fattah for Mayor Scheme ................................. 9 1. The Lord Loan and Its Repayment ..................... 10 2. The College Tuition Component of the FFM Scheme ................................................................ 27 3. The NOAA Grant and the Phantom Conference 28 B. The Blue Guardians Scheme.................................. 31 C. The Fattah–Vederman Bribery Scheme................. 33 D. The Indictment and Trial ....................................... 39 III. Juror Misconduct and Dismissal of Juror 12 ........... 58 A. Investigation of Alleged Juror Misconduct ........... 58 B. Dismissal of Juror 12 ............................................. 64 IV. The District Court’s Instructions Under McDonnell 70 A. The McDonnell Framework................................... 71 B. The Kirk Meeting ................................................... 76 C. Fattah’s Efforts to Secure Vederman an Ambassadorship.......................................................... 79 D. The Zionts Hiring .................................................. 82 E. Vederman’s Sufficiency Challenge to Counts 16–18 and 22–23.................................................................... 89 F. Blue Guardians ....................................................... 91 5 V. Sufficiency of the Evidence for the RICO Conspiracy Conviction ...................................................................... 95 VI. Variance from the Indictment and Sufficiency of the Evidence for Count 2.................................................... 110 VII. The District Court’s Instruction to the Jury on the Meaning of Intent ......................................................... 120 VIII. Sending the Indictment to the Jury ...................... 126 IX. The District Court’s Evidentiary Rulings .............. 129 A. The District Court’s Application of Rule 404(b) 129 B. Evidentiary Rulings Regarding Nicholas’s Defense ..................................................................... 134 1. The EAA Board Minutes .................................. 134 2. Jones’ Memory Regarding Other Contracts ..... 136 3. Exclusion of NOAA Evidence.......................... 137 C. The Cooperating Witness’s Mental Health Records ..................................................................... 138 1. The District Court’s Denial of Access to the Mental Health Records ..................................... 140 2. The District Court’s Grant of the Motion in Limine ............................................................... 142 X. The Government’s Cross-Appeal ............................ 147 A. CUMA is a Mortgage Lending Business............. 148 B. Sufficiency of the Evidence ................................. 155 XI. Prejudicial Spillover ............................................... 158 A. Fattah’s Claim of Prejudicial Spillover ............... 159 6 B. Vederman’s Assertion of Prejudicial Spillover ... 160 XII. Conclusion ............................................................ 164 7 I. Introduction Chaka Fattah, Sr., a powerful and prominent fixture in Philadelphia politics, financially overextended himself in both his personal life and his professional career during an ultimately unsuccessful run for mayor. Fattah received a substantial illicit loan to his mayoral campaign and used his political influence and personal connections to engage friends, employees, and others in an elaborate series of schemes aimed at preserving his political status by hiding the source of the illicit loan and its repayment. In so doing, Fattah and his allies engaged in shady and, at times, illegal behavior, including the misuse of federal grant money and federal appropriations, the siphoning of money from nonprofit organizations to pay campaign debts, and the misappropriation of campaign funds to pay personal obligations. Based upon their actions, Fattah and four of his associates—Herbert Vederman, Robert Brand, Bonnie Bowser, and Karen Nicholas—were charged with numerous criminal acts in a twenty-nine count indictment. After a jury trial, each was convicted on multiple counts. All but Bowser appealed. As we explain below, the District Court’s judgment will be affirmed in part and reversed in part. 8 II. Background1 During the 1980s and ’90s, Fattah served in both houses of the Pennsylvania General Assembly, first as a member of the House of Representatives and later as a Senator. In 1995, Fattah was elected to the United States House of Representatives for Pennsylvania’s Second Congressional District. In 2006, Fattah launched an unsuccessful run for Mayor of Philadelphia, setting in motion the events that would lead to his criminal conviction and resignation from Congress ten years later. A. The Fattah for Mayor Scheme Fattah declared his candidacy for mayor in November of 2006. Thomas Lindenfeld, a political consultant on Fattah’s exploratory committee, believed that “[a]t the beginning of the campaign, [Fattah] was a considerable . . . candidate and somebody who had a very likely chance of success.” JA1618. But Fattah’s campaign soon began to experience difficulties, particularly with fundraising. Philadelphia had adopted its first-ever campaign contribution limits, which limited contributions to $2,500 from individuals and $10,000 from political action committees and certain types of business organizations. Fattah’s fundraising difficulties led him to 1 The facts are drawn from the trial record unless otherwise noted. 9 seek a substantial loan, far in excess of the new contribution limits. 1. The Lord Loan and Its Repayment While serving in Congress, Fattah became acquainted with Albert Lord, II. The two first met around 1998, when Lord was a member of the Board of Directors of Sallie Mae. As the May 15, 2007 primary date for the Philadelphia mayoral race approached, Fattah met Lord to ask for assistance, telling Lord that the Fattah for Mayor (FFM) campaign was running low on funds. Fattah asked Lord to meet with Thomas Lindenfeld, a political consultant in Washington, D.C., and part-owner of LSG Strategies, Inc. (Strategies), a company that was working with the FFM campaign and that specialized in direct voter contact initiatives. Lindenfeld had been part of the exploratory group that initially considered Fattah’s viability as a candidate for mayor. Lindenfeld had known Fattah since 1999, when Fattah endorsed Philadelphia Mayor John Street. Through Fattah, Lindenfeld had also gotten to know several of Fattah’s associates, including Herbert Vederman, Robert Brand, and Bonnie Bowser. Herbert Vederman, a businessman and former state official, was the finance director for the FFM campaign. Robert Brand owned Solutions for Progress (Solutions), a “Philadelphia-based public policy technology company, whose mission [was] to deliver technology that directly assists low and middle income families [in obtaining] 10 public benefits.” JA6551. Bowser was Fattah’s Chief of Staff and campaign treasurer, and served in his district office in Philadelphia. Lord’s assistant contacted Lindenfeld to arrange a meeting, and Lindenfeld informed Fattah that he would be meeting with Lord. Lindenfeld, along with his partner, Michael Matthews, met with Lord and discussed Fattah’s need for funds to mount an intensive media campaign. After that meeting, Lindenfeld reported to Fattah that Lord wanted to help, but that they had not discussed a specific dollar amount. Approximately a week later, Fattah instructed Lindenfeld to meet with Lord a second time. Lord “wanted to know if he could give a substantial amount of money, a million dollars” to Fattah’s campaign. JA1630. That prompted Lindenfeld to reply that the amount “would be beyond the campaign finance limits.” Id. Lord proposed a solution: he offered to instead give a million dollars to Strategies in the form of a loan. To that end, Lindenfeld had a promissory note drafted which specified that Lord was lending Strategies $1 million, and that Strategies promised to repay the $1 million at 9.25% interest, with repayment to commence January 31, 2008. Lindenfeld later acknowledged that the promissory note would make it appear as though Lord’s $1 million was not a contribution directly to the Congressman, although he knew that it was actually a loan to the FFM campaign. Indeed, Lindenfeld confirmed with Fattah that neither 11 Lindenfeld nor Strategies would be responsible for repayment. With that understanding, Lindenfeld executed both the note and a security agreement purporting to encumber Strategies’ accounts receivable and all its assets. On May 1, shortly before the primary election, Lord wired $1 million to Lindenfeld. Lindenfeld held the money in Strategies’ operating account until Fattah told him how it was to be spent. Some of the money was eventually used for print materials mailed directly to voters. And, at Fattah’s direction, Lindenfeld wired a substantial sum to Sydney Lei and Associates (SLA), a company owned by Gregory Naylor which specialized in “get out the vote” efforts. Naylor had known Fattah for more than 30 years.2 During the campaign, Naylor worked as the field director and was in charge of getting out the vote on election day. 2 Naylor first worked with Fattah when he was in the state legislature. When Fattah was elected to Congress, Naylor worked in his Philadelphia office. Naylor met Nicholas when she joined Fattah’s staff at some point in the 1990s. After concluding her employment with Fattah’s office, Nicholas worked with the Educational Advancement Alliance (EAA), an education nonprofit entity founded by Fattah. This entity helped to recruit underrepresented students for scholarship and college opportunities. Around 2009, Naylor left Fattah’s office to work exclusively with SLA. Naylor also knew Brand. 12 On the final day of the campaign, Naylor worked with Vederman, who allowed Naylor to use his credit card to rent vans that would transport Fattah voters to the polls. As the primary date neared, Fattah and Naylor knew the campaign was running out of money. The campaign was unable to finance “media buys,” and Naylor needed money for field operations to cover Philadelphia’s more than one thousand polling places. In early May, Lindenfeld called Naylor to say that Lindenfeld “would be sending some money [Naylor’s] way.” JA3057. Within days, SLA received a six-figure sum for Naylor to use in the campaign and on election day. Naylor used the money to pay some outstanding bills, including salaries for FFM employees, and allocated $200,000 to field operations for election day. Fattah lost the mayoral primary on May 15, 2007. Afterward, Lindenfeld spoke with Fattah, Naylor and Bowser about accounting for the FFM campaign money from Lord that had been spent. They decided that the amounts should not appear in the FFM campaign finance reports, and Fattah instructed Naylor to have his firm, SLA, create an invoice. Naylor did so, creating an invoice dated June 1, 2007 from SLA to FFM, seeking payment of $193,580.19. Naylor later acknowledged that the FFM campaign did not actually owe money to SLA, and that the false invoice was created to “hide the transaction that took place earlier” and “make it look like [SLA] was owed money.” JA3075–76. Although FFM did not owe SLA 13 anything for the election day expenses, the FFM campaign finance reports from 2009 through 2013 listed a $20,000 in-kind contribution from SLA for each year, thereby lowering FFM’s alleged outstanding debt to SLA. Of the total $1 million Lord loan, $400,000 had not been spent. Lindenfeld returned that sum to Lord on June 3, 2007. He included a cover letter which stated: “As it turns out the business opportunities we had contemplated do not seem to be as fruitful as previously expected.” JA1254. Lindenfeld later admitted that there were no such “business opportunities” and that the letter was simply an effort to conceal the loan. In late 2007, faced with financial pressures, Lord asked his son, Albert Lord, III, to collect the outstanding $600,000 balance on the loan to Strategies. Lord III contacted Lindenfeld about repayment and expressed a willingness to forgive the interest owed if the principal was paid. Lindenfeld immediately called Fattah and informed him that repayment could not be put off any longer. Fattah told Lindenfeld more than once that “[h]e would take care of it,” JA1652, but Fattah did not act. Needing someone who might have Fattah’s ear, Lindenfeld reached out to Naylor and Bowser. Naylor talked to Fattah on several occasions and told him that Lindenfeld was under considerable pressure to repay the loan. Fattah told Naylor more than once that he was “working on it.” JA3082–83. 14 During his political career, Fattah had focused on education, especially for the underprivileged. Indeed, Fattah founded two nonprofit organizations: College Opportunity Resources for Education (CORE), and the Educational Advancement Alliance (EAA). EAA held the annual Fattah Conference on Higher Education (the “annual conference”) to acquaint high school students with higher education options. JA3079. Sallie Mae regularly sponsored the conference. According to Raymond Jones, EAA’s chairman of the board from 2004 through 2007, EAA offered a variety of programs to provide “marginalized students with educational opportunities so they could continue and go to college.” JA1360. EAA was funded with federal grant money which could only be spent for the purposes described in the particular grant. Karen Nicholas served as EAA’s executive director, handling the organization’s day-to-day administrative responsibilities. Nicholas had previously been a staffer for Fattah when he was a member of Pennsylvania’s House of Representatives. CORE was an organization that awarded scholarships to graduating high school students in Philadelphia who had gained admission to a state university or the Community College of Philadelphia. CORE received funding from a variety of sources, including Sallie Mae. Because CORE also received federal funds, and because EAA had experience working with federal grants, EAA received and handled the federal 15 funds awarded to CORE. In short, EAA functioned as a fiduciary for CORE. When money became a problem for the FFM campaign, Fattah’s involvement with EAA and CORE soon became less about helping underprivileged students, and more about providing an avenue for disguising efforts to repay the illicit campaign funds from Lord. On January 7, 2008, Robert Brand contacted Fattah by telephone. Shortly thereafter, Lindenfeld received an unexpected call from Brand proposing an arrangement for Brand’s company, Solutions, to work with Strategies. Solutions had developed a software tool called “The Benefit Bank,” which was designed to “assist low and moderate income families to have enhanced access to benefits and taxes.” JA1993. During the telephone call, Brand referred to The Benefit Bank and suggested a contract under which Strategies would be paid $600,000 upfront. JA1666. Shortly thereafter, on January 9, 2008, Brand followed up on his call to Lindenfeld with an email about “develop[ing] a working relationship where you could help us to grow The Benefit Bank and our process of civic engagement. While I know this is not your core business I would like to try to convince you to take us on as a client.” JA6427. Lindenfeld responded that he was interested. To Lindenfeld, “this was the way that Congressman Fattah was going to repay the debt to Al Lord.” JA1654. When Lindenfeld called Fattah and told him of the contact from Brand, Fattah simply replied that Lindenfeld “should just proceed.” JA1666–67. 16 A few days later, Brand emailed Nicholas at EAA a proposal from Solutions concerning The Benefit Bank, which sought EAA’s support in developing an education edition of The Benefit Bank and a $900,000 upfront payment. As the January 31 date for repayment of the balance of the $1 million Lord loan approached, a flurry of activity took place. On January 24, both Raymond Jones, chair of the EAA Board, and Nicholas signed a check from EAA made out to Solutions in the amount of $500,000. Although no contract existed between EAA and Solutions, the memo line of the check indicated that it was for a contract, and Nicholas entered it into EAA’s ledger.3 That same day, Ivy Butts, an employee of Strategies, emailed Lindenfeld the instructions Brand would need to wire the $600,000 balance on the Lord loan. 3 Raymond Jones, who was EAA’s Chairman of the Board from 2004 through 2007, recalled at trial that the Board had a limit on the amount that Nicholas could spend without board approval. JA1358, 1369. Nicholas was authorized to sign contracts on behalf of EAA for no more than $100,000. JA1369–71. Jones did not recall the contract between EAA and Solutions, nor did the EAA board minutes for December 2007, February 2008, or May 2008 refer to the EAA–Solutions contract or to the substantial upfront payment of half a million dollars upon execution of the agreement. JA6358–63; 6567. 17 Within minutes, Lindenfeld forwarded that email to Brand at Solutions. Brand then made two telephone calls to Fattah. By late afternoon, Brand emailed Nicholas, informing her that he had “met with all the people I need to meet with and have a pretty clear schedule of what works best for us. I am also seeing what line of credit we have to stretch out the payments until you get your line of credit in place.” JA6558. Brand asked if they could talk and “finalize this effort.” JA6558. On January 25 and 26, there were a number of calls between Fattah, Brand, and Nicholas. On Sunday January 27, at 5:46 pm, Brand telephoned Fattah. At 10:59 pm, Brand emailed Nicholas a revised contract between EAA and Solutions for the engagement of services. Brand indicated he would send someone to pick up the check at about 1:00 pm the following day. The revised contract called for the same $900,000 payment from EAA to Solutions, yet specified that $500,000 was to be paid on signing, with $100,000 due three weeks later, and another $100,000 to be paid six weeks out. No due date for the $200,000 balance was specified. The terms of the contract called for EAA to assist Solutions with further developing The Benefit Bank. In addition, under the contract, EAA would receive certain 18 funds from the Commonwealth of Pennsylvania for a program relating to FAFSA applications. 4 The same evening, Brand sent Lindenfeld a contract entitled “Cooperative Development Agreement to Provide Services to Solutions for Progress, Inc. for Growth of The Benefit Bank.” JA6569. The agreement proposed a working partnership in which Strategies would work with Solutions to identify and secure a Benefit Bank affiliate in the District of Columbia and two other states, and to facilitate introductions to key officials in other states where The Benefit Bank might expand. The terms of the agreement provided that Solutions would pay $600,000 to Strategies by January 31, 2008, which would “enable [Strategies’] team to assess opportunities and develop detailed work plans for each area.” JA6572. Brand copied Solutions’ Chief Financial Officer, Michael Golden. Lindenfeld responded to Brand’s email within a minute, asking if Brand had received the wiring instructions. Brand immediately confirmed that he had. Concerned that Solutions did not have $600,000 to pay Strategies, Golden talked to Brand, who informed him that Solutions would be receiving a check for $500,000 from EAA. Early the next morning, Nicholas responded to Brand’s email from the night before. She advised Brand that he could pick up the check, “but as I stated I am not 4 FAFSA is an acronym for Free Application for Federal Student Aid. 19 in a position to sign a contract committing funds that I am not sure that I will have.” Gov’t Supp. App. (GSA) 1. That same day, a $540,000 transfer was made from the conference account, which EAA handled, into EAA’s checking account. The conference account was maintained to handle expenses for Fattah’s annual higher education conference. Prior to this transfer, EAA had only $23,170.95 in its account. EAA then tendered a $500,000 check to Solutions, which promptly deposited the check before the close of that day’s business. EAA never replenished the $540,000 withdrawal from the conference account. Brand received the executed contract between Solutions and Strategies on January 28. Even though the contract called for Strategies to perform services in exchange for the $600,000 payment, Lindenfeld neither expected to do any work for the $600,000, nor did he in fact do any work. In sum, by January 28, Solutions had received $500,000 from EAA, but it still had to come up with $100,000 to provide Strategies with the entire amount needed to repay the Lord loan. Golden obtained the needed funds the following day by drawing $150,000 on a line of credit held by Brand’s wife. Brand and Fattah spoke four more times on the telephone on January 29. Trial evidence later showed that, during the month of January 2008, neither the FFM campaign bank account nor Fattah’s 20 personal account had a sufficient balance to fund a $600,000 payment. On the morning of January 30, frustrated by the delay, Lindenfeld sent Brand an email with a subject line “You are killing me.” JA6430. Lindenfeld stated that he had “made a commitment based on yours to me. Please don’t drag this out. I have a lot on the line.” Id. Brand responded late in the afternoon, stating: “just met with Michael. He does the transfer at 8 AM tomorrow. It should be in your account ($600K) early tomorrow morning.” Id. Lindenfeld replied: “The earlier the better.” Id. The following morning, Golden wired $600,000 from Solutions’ Pennsylvania bank account into Strategies’ Washington D.C. bank account. JA2745, 2874. Strategies in turn, wired the same amount from its Washington D.C. bank account to Lord’s bank account in Virginia. JA2874, 6549. Around noon, Brand telephoned Lindenfeld. In the days following the exhaustive efforts to meet the January 31 loan repayment deadline, four more telephone calls took place between Brand and Fattah.5 Naylor learned at some point that the loan had been paid off. When Naylor asked Fattah about details of the repayment, Fattah simply replied “[t]hat it went through EAA to Solutions and it was done.” JA3088. 5 By contrast, between October to December 2007, Brand and Fattah spoke by telephone only “once or twice [a] month.” JA2734. 21 Meanwhile, at some point in January, EAA received notice that the Department of Justice Office of the Inspector General (DOJ) intended to audit its books.6 DOJ auditors told EAA to provide, at the “entrance conference,” documentation containing budgetary and accounting information. EAA failed to produce any accounting information. Although Lindenfeld was no longer making demands of Brand, Brand was still owed the remaining $100,000 that Solutions had paid to satisfy the Lord loan. On March 23, 2008, Brand sent Nicholas an email outlining his efforts to contact her over the previous two weeks about documentation on the CORE work, how to proceed with the paperwork for the Commonwealth of Pennsylvania, and “how we can get our proposed contract signed and the outstanding payments made.” JA2749. Nicholas responded that evening, writing: I can appreciate your urgency however I do have EAA work that I continue to do, including the [usual] facilitation of programs, our financial audit, the start-up of two new programs[,] and of course the DOJ audit. I am still trying to obtain a line of credit without a completed 2007 audit and things are getting a 6 One of the terms and conditions of a federal grant is that the recipient “be readily prepared for an audit.” JA2314. 22 little uncomfortable now as I try to keep us afloat. JA6576. Nicholas told Brand that the DOJ auditors were making demands and would soon be on site. She noted that “[t]hey are still very uncomfortable with your contract amongst other things and depending on their findings some of the funding received may have to be returned.” Id. Nicholas said that she had submitted the paperwork to the state, and she told Brand that “in the future . . . as a result of the DOJ audit I will not be in a position to do another contract such as this.” Id. Shortly after Nicholas’s reply to Brand, Nicholas forwarded the Brand–Nicholas email chain to Fattah. The body of the email stated, in its entirety: “I really don’t appreciate the tone of Bob’s email. I can appreciate that he has some things going on however I am doing my best to assist him. Some other things are a priority. He needs to back off.” GSA2. Later that night, Bowser sent Fattah an email with a subject line that read “Karen N” and a telephone number. JA2752. As the audit continued, the auditors found other deficiencies. During April of 2008, DOJ issued a notice of irregularity to EAA, which resulted in the audit being referred to DOJ’s Investigations Division for a more comprehensive review. On April 24, 2008, Brand emailed Nicholas asking for a time to update her on The Benefit Bank. In early May, 23 Brand sent another email to Nicholas attaching a revised EAA–Solutions contract proposal, which decreased the initial upfront cost from $900,000 to $700,000. Although Solutions and EAA had still not signed a contract, EAA paid Solutions another $100,000 in May. That money was obtained via a loan to EAA from CORE. Thomas Butler, who had worked for Fattah both when Fattah was in Congress and when he was in the General Assembly, was CORE’s executive director. Butler had been contacted in mid-May by Jackie Barnett, a member of CORE’s Board who had also worked with Congressman Fattah. Barnett informed Butler that Nicholas had requested a loan from CORE to EAA, and that Fattah, as Chairman of CORE’s Board, had approved it. Butler and Barnett withdrew funds from two CORE bank accounts and obtained a cashier’s check, dated May 19, in the amount of $225,000 and made payable to EAA. The withdrawals were from accounts used for Sallie Mae funds and other scholarship money. After EAA received the $225,000 check, EAA tendered a $100,000 check to Solutions. The check bore the notation “Commonwealth of Pennsylvania.” EAA repaid CORE the following month. Because EAA lacked sufficient funds of its own to cover this payment, EAA drew on grant money that it had received from NASA. Brand and Lindenfeld continued to communicate concerning The Benefit Bank. In July of 2008, a meeting was held at Solutions with Brand, Lindenfeld, Golden, and 24 other Solutions employees to discuss “an enormous amount of work” that Brand wanted Strategies to do. JA1670. Lindenfeld said in response “we’d be glad to do that, but . . . we would have to be paid.” Id. At that point, someone in the meeting stated that Strategies “had already been paid” $600,000. Id. Lindenfeld replied: “well, that was for Congressman Fattah, . . . that’s not for us. So if you want us to do work, we have to get paid for it separately.” Id. Brand became upset with Lindenfeld over his comment about being paid because his colleagues at Solutions were not aware of the reason for the $600,000 payment. Meanwhile, EAA was attempting to meet the demands of the DOJ auditors, who were focused on the relationship between EAA and CORE. DOJ served a subpoena upon Solutions to produce “[a]ny and all documents including, but not limited to, contract documents, invoices, correspondence, timesheets, deliverables and proof of payment related to any services provided to or payments received” from CORE or EAA. JA2350. Special Agent Dieffenbach, from the DOJ, interviewed Nicholas on July 14, 2008. During that interview, Nicholas discussed the relationship between EAA and CORE, how invoices were paid, and how consultants were handled. Nicholas also answered questions about EAA’s relationship with Solutions, including the payment of invoices. She did not inform 25 Agent Dieffenbach of the $500,000 payment in January or the subsequent $100,000 payment in May. Nor did the interview address the EAA–Solutions contract that purportedly required those payments, because the contract had yet to be produced. Solutions failed to comply with the subpoena, prompting an email from Agent Dieffenbach on August 26 asking for an update concerning Solutions’ reply to the DOJ subpoena. Solutions then produced an undated version of the EAA–Solutions contract that required the $600,000 upfront payment. Neither Brand nor Nicholas provided the auditors with the January and May checks from EAA to Solutions. Efforts to conceal the repayment of the Lord loan and to promote the political and financial interests of Fattah continued. The FFM campaign reports indicated in- kind contributions of debt forgiveness by SLA even though there had been no actual debt. In September of 2009, with EAA’s ledgers still under scrutiny, Nicholas altered the description of the entry for the $100,000 check to Solutions from “professional fees consulting” to “CORE Philly.” JA2546. Other FFM campaign debt was reduced further after Vederman negotiated with creditors. EAA never fully recovered from its payment of the $600,000 balance on the Lord loan and the audits that took place in 2008. It began laying off employees in 2011, and by June of 2012, only four employees remained. JA3659. EAA ceased operations at some point in 2012. JA1530. 26 2. The College Tuition Component of the FFM Scheme Although the FFM campaign was close to insolvent, it nevertheless made tuition payments for Fattah’s son, Chaka Fattah Jr., also known as Chip. Chip attended Drexel University, but had yet to complete his coursework because he had failed to pay an outstanding tuition balance. As the FFM campaign got underway in 2007, Fattah wanted Chip to re-enroll in classes at Drexel and get a degree. Fattah asked Naylor to help financially, and he did so by writing checks from SLA to Drexel toward Chip’s outstanding tuition. By October of 2007, Chip was permitted to re-enroll in classes. Although Naylor never directly addressed the issue with Fattah, he agreed to assist with Chip’s outstanding tuition with the expectation that SLA would be repaid. The first check to Drexel in the amount of $5,000 was sent in August of 2007, with $400 payments in the months that followed until August of 2008. At some point, Chip informed Naylor that the payee was no longer Drexel, but Sallie Mae. Naylor then began sending monthly checks from SLA to Sallie Mae. Those payments, in the amount of $525.52, began in March of 2009 and continued until April of 2011, after which Fattah told Naylor he no longer needed to make them. SLA’s payments to Drexel and Sallie Mae totaled $23,063.52. Naylor’s expectation of repayment was eventually realized. Beginning in January of 2008 and continuing 27 until November 2010, Bowser sporadically sent SLA reimbursement checks from the FFM campaign with a notation that payment was for “election day operation expenses.” JA3136. The FFM funds had been transferred from the Fattah for Congress campaign. These reimbursement checks totaled $25,400. In an effort to conceal the source of the payments to Drexel and Sallie Mae, and to make it appear that the younger Fattah had performed services for SLA, Naylor created false tax forms for Chip. Chip, however, had never performed services for SLA. 3. The NOAA Grant and the Phantom Conference In mid-December 2011, when EAA was experiencing serious financial difficulties, Nicholas submitted an email request to the educational partnership program of the National Oceanic & Atmospheric Administration (NOAA) for a grant “designed to provide training opportunities and funding to students at minority serving institutions” interested in science, technology, engineering, and math fields related to NOAA’s mission. JA3354–55. The request sought $409,000 to fund EAA’s annual conference scheduled for February 17–19, 2012. Jacqueline Rousseau, a supervisory program manager at NOAA, participated in a conference call with Nicholas shortly thereafter and advised Nicholas that the agency could not afford the $409,000 request but would consider a smaller grant. Rousseau advised Nicholas that EAA 28 would need to submit an application if it wished to be considered for a grant. Before submitting a grant application, Nicholas emailed Rousseau about sponsoring the conference. On January 11, 2012, Rousseau informed Nicholas that the “NOAA Office of Education, Scholarship Programs has agreed to participate and provide sponsorship funds of $50K to support the referenced conference.” JA6453. Rousseau also informed Nicholas that Chantell Haskins, who also worked with the student scholarship program, would be the point of contact for NOAA. In February 2012, EAA held its annual conference at the Sheraton Hotel in downtown Philadelphia. The conference had been held at the same location each year since 2008. Nicholas contacted Haskins at some point in early 2012, inquiring about the $50,000 grant. On May 8, 2012, Haskins sent Nicholas an e-mail which included information about submitting proposals to fund a conference for students. EAA then submitted a grant application, which Haskins reviewed. She advised Nicholas on June 28, 2012 that the grant could not be used to provide meals, and that the date of the conference would have to be pushed back, with the new date included in a modified application. When Nicholas asked if expenses from a previous conference could be paid from the new grant, Haskins informed her that this was not allowed. 29 In early July 2012, Nicholas sent a modified grant proposal to Haskins. It eliminated the budget item for food and changed the date of the 2012 conference to October 19–21, 2012 at the same Sheraton Hotel in Philadelphia where EAA’s annual conference had taken place earlier in the year. NOAA approved a $50,000 grant for the October 2012 conference—a conference that would never be held. Unaware that no October 2012 conference had taken place, NOAA allowed Nicholas access to the $50,000 grant in March of 2013. She then transferred the entire amount from NOAA to EAA’s bank account a few days later. Naylor had performed services for EAA for which he was still owed $116,590. JA3119. In discussions with Naylor, Nicholas had informed him that the likelihood of EAA’s being able to pay him was “[n]ot very good.” JA3120. Yet several days after EAA had received the $50,000 from NOAA, Nicholas sent Naylor a check for $20,000. JA3120, 4283. On April 3, 2013, Nicholas submitted a final report to NOAA concerning EAA’s use of the grant. Notably, page 4 of the report stated the conference had been held in February 2012, while page 17 stated that the conference had been held from October 19 to 21, 2012. NOAA issued a notice asking for clarification and for a list of students who had been supported at the conference. Nicholas failed to file either a clarifying report regarding the date of the conference or a timely report regarding the disbursement of the grant. Finally, in November of 2013, Nicholas 30 submitted the final Federal Financial Report in which she certified, falsely, that the $50,000 had been used for a project during the period from August 1, 2012 to December 30, 2012. B. The Blue Guardians Scheme In addition to functioning as the conduit for Lord’s $1 million loan to Fattah’s campaign, Lindenfeld’s company, Strategies, also performed services for the campaign. The work resulted in indebtedness from FFM to Strategies of approximately $95,000. Fattah made several small payments, but failed to pay the full amount due. Although Lindenfeld spoke to Fattah, Naylor and Bowser about the debt, no payments were forthcoming. During a meeting in Fattah’s D.C. office, Fattah told Lindenfeld “that [repayment] really wasn’t going to be possible because the campaign had been over for a long time” and the funds were not available. JA1693. Fattah then asked Lindenfeld if he could write off the debt on his FFM campaign finance reports. Id. Lindenfeld told Fattah that as long as he was paid, it was not his business how Fattah disclosed it on the campaign finance reports. JA1694. In lieu of repayment, Fattah suggested that Strategies could claim to be interested in setting up an entity to address environmental issues and ocean pollution along the coastline and in the Caribbean. Fattah explained that creating such an entity would make it possible to obtain an appropriation from the government. Hearing 31 this, Lindenfeld knew he was not going to be paid by the FFM campaign, and was amenable to receiving money from an appropriation instead. At a later meeting, Lindenfeld told Fattah that the name of the entity would be “Blue Guardians.” Lindenfeld consulted with an attorney about creating Blue Guardians as an entity to receive the federal grant. He emailed Fattah, asking questions about how to complete an application to the House Appropriations Committee. Fattah provided suggestions, and an application was eventually completed. It indicated that Blue Guardians would be “in operation for a minimum of ten years,” and, in accordance with Fattah’s guidance, requested $15 million in federal funds. JA1711– 13. Lindenfeld submitted the application to Fattah’s office in April of 2009. Afterward, a Fattah staffer contacted Lindenfeld to suggest that he change his Washington, D.C., address to Philadelphia because that was the location of Fattah’s district. Fattah later suggested to Lindenfeld that Brand might allow the use of his Philadelphia office address, a plan to which Brand agreed. In February 2010, Lindenfeld submitted a second application to the Appropriations Committee. In March, Fattah submitted a project request using his congressional letterhead and seeking $3,000,000 for the “Blue Guardians, Coastal Environmental Education Outreach Program.” JA6432. Within a month, Blue Guardians had both articles of incorporation and a bank account. Around 32 that time, a news reporter contacted Lindenfeld to discuss the new Blue Guardians entity. The inquiry made Lindenfeld uncomfortable, and he ultimately decided to abandon the Blue Guardians project. He continued to seek payment from Fattah, to no avail. Nonetheless, having obtained Lindenfeld’s acquiescence to writing off the campaign’s debt to Strategies, Fattah started falsifying FFM’s campaign reports. Beginning in 2009 and extending through 2013, the FFM campaign reports executed by Fattah and Bowser stated that Strategies made in-kind contributions of $20,000, until the debt appeared to have been paid in full. C. The Fattah–Vederman Bribery Scheme Vederman and Fattah were personal friends. Vederman was a successful businessman who had also served in prominent roles in the administrations of Ed Rendell when he was Mayor of Philadelphia and Governor of Pennsylvania. In November of 2008, Vederman was a senior consultant in the government and public affairs practice group of a Philadelphia law firm. His assistance to the FFM campaign included paying for rented vans used in the get-out-the-vote effort. After Fattah’s electoral defeat, the campaign still owed more than $84,000 to a different law firm for services performed for the campaign. Vederman approached that firm in the summer of 2008 asking if it would forgive FFM’s debt. Negotiations resulted in a 33 commitment from FFM to pay the firm $30,000 by the end of 2008 in exchange for forgiveness of $20,000, all of which would appear on the FFM campaign finance report. Vederman’s efforts also led to payment by Fattah of an additional $10,000 in 2009 to the law firm, in exchange for additional forgiveness of $20,000 of debt. It was not long after Vederman’s successful efforts to lower Fattah’s campaign debt, that Fattah wrote a letter to U.S. Senator Robert P. Casey recommending Vederman for an ambassadorship. At some point in 2010, Vederman again intervened on behalf of the FFM campaign. FFM remained in debt to an advertising and public relations firm owned by Robert Dilella. By late 2011, Vederman and Dilella had worked out a settlement to resolve the outstanding debt. Pursuant to that settlement, Dilella received partial payment from the FFM campaign: $25,000 in satisfaction of a $55,000 debt. Dilella testified at trial that he would not have agreed to retire a portion of the debt had he known the FFM campaign was paying college tuition for Fattah’s son. Vederman helped Fattah financially in other ways. Before the 2006 FFM campaign, Fattah and his wife, Renee Chenault-Fattah, sponsored a young woman named Simone Muller to live with them as an au pair exchange visitor. Muller was from South Africa, and her J-1 visa allowed her to serve as a nanny and to study in the United States. Muller later applied for and received a second visa, an F-1 student visa that indicated she had been accepted as 34 an international student at the Community College of Philadelphia. The application indicated that Muller would again be residing with the Fattahs. Notwithstanding this living arrangement, Fattah identified Vederman as the person who would be paying for Muller’s trip to the United States. By the beginning of 2010, Muller wished to transfer to Philadelphia University. This required her to submit verification that funds were available to pay for her study. Although the Fattahs were Muller’s sponsors, Fattah explained to the University’s Dean of Enrollment Services that he was submitting a letter of secondary support from Vederman. JA3754, 3763–65, 6504. Without Vederman’s January 2010 letter of support, the University would not have admitted Muller. In addition to this pledge of support, Vederman paid $3,000 of Muller’s tuition. Shortly thereafter, Fattah resumed his efforts to secure an ambassadorship for Vederman. In February of 2010, Fattah staffer Maisha Leek contacted Katherine Kochman, a scheduler for White House Chief of Staff Rahm Emanuel. Leek requested a telephone conference with Emanuel, Rendell, and Fattah to discuss Vederman’s “serving his country in an international capacity.” JA2893. In a follow-up email on March 26, Leek sent documents to Kristin Sheehy, a secretary to White House Deputy Chief of Staff James Messina. The documents included Fattah’s 2008 letter to Senator Casey and Vederman’s biography. After 35 participating in a telephone conference about Vederman with Fattah and Rendell, Messina sent Vederman’s biography to the White House personnel office for consideration. As the April 2010 tax deadline approached, Fattah still owed the City of Philadelphia earned income tax in the amount of $2,381. Just days before the filing deadline, Vederman gave a check to Chip Fattah for $3,500. The younger Fattah quickly deposited $2,310 into his father’s bank account. Fattah paid his tax bill on April 15. Without Chip’s deposit into his father’s bank account, the older Fattah would not have had sufficient funds to pay his tax bill. On October 30, 2010, Vederman gave Chip another check, this one for $2,800. That same day, Fattah hand- delivered a letter to President Obama recommending Vederman for an ambassadorship. A few weeks later, Fattah’s staffer, Leek, sent the letter that Fattah had given to President Obama to Messina’s office. That letter pointed out that both Rendell and Fattah had sent letters on behalf of Vederman, and that he was an “unquestionably exceptional candidate for an ambassadorship.” JA6291–92. Fattah’s efforts to secure Vederman an ambassadorship were unsuccessful. Fattah then shifted gears and sought to secure Vederman a position on a federal trade committee. Fattah approached Ron Kirk, who served as U.S. Trade Representative, and asked him to 36 speak with a constituent. In May of 2011, Leek followed up on that discussion by emailing Kirk and asking him to meet with Vederman. Kirk met with Vederman on June 5, 2011 and explained to him the role of the trade advisory committees. Although the two men “had a very nice conversation,” JA 3566, it soon became “pretty apparent to [Kirk and his staff] that [serving on a trade advisory committee was] not what Mr. Vederman was interested in.” JA3567. As Kirk put it, “it was obvious that [Vederman] was looking for something perhaps more robust in his mind or . . . higher profile than one of our advisory committees.” Id. Given Vederman’s lukewarm interest, no appointment to an advisory committee was forthcoming. In late December 2011, the Fattahs applied for a mortgage so they could purchase a second home in the Poconos. Shortly after applying for the mortgage, Fattah emailed Vederman, offering to sell him his wife’s 1989 Porsche for $18,000. Vederman accepted the offer. The next day, Vederman wired $18,000 to Fattah’s Wright Patman Federal Credit Union account. The Credit Union Mortgage Association (CUMA) acted as the loan processing organization for the home mortgage. Because CUMA is required to verify the source of any large deposits, CUMA’s mortgage loan processor, Victoria Souza, contacted Fattah on January 17, 2012, to confirm the source of the $18,000. Fattah informed Souza that the $18,000 represented the proceeds of the Porsche 37 sale. Souza requested documentation, including a signed bill of sale and title. That same day, Bowser emailed Vederman a blank bill of sale for the Porsche. After Vederman signed the bill of sale, Fattah forwarded it to Souza. The bill of sale was dated January 16, 2012, which was the day before Souza had requested the documentation. It bore the signatures of Renee Chenault-Fattah and Herbert Vederman, with Bonnie Bowser as a witness. Fattah also provided Souza with a copy of the Porsche’s title. It was dated the same day it was sent to Souza, and bore signatures of Chenault-Fattah as the seller and Vederman as buyer, along with a notary’s stamp. Neither Vederman nor Chenault-Fattah actually appeared before the notary. Vederman never took possession of the Porsche. Renee Chenault-Fattah continued to have the Porsche serviced and insured long after the purported sale had taken place. Moreover, the Porsche remained registered in Chenault-Fattah’s name, and was never registered to Herbert Vederman. When FBI agents searched the Fattahs’ home in 2014, the Porsche was discovered in the Fattahs’ garage. On January 24, 2012, the Fattahs wired $25,000 to the attorney handling the escrow account for the purchase of the vacation home. Without the $18,000 transfer from 38 Vederman, the Fattahs would not have had sufficient funds in their bank accounts to close on the home. Around the same time that the Fattahs were purchasing the house in the Poconos, Fattah’s Philadelphia office hired Vederman’s longtime girlfriend, Alexandra Zionts. Zionts had long worked for a federal magistrate judge in Florida. Near the end of 2011, the magistrate judge retired, leaving Zionts ten months shy of obtaining the necessary service required to receive retirement benefits. If Zionts could find another job in the federal government, her benefits and pension would not be adversely affected. Vederman assisted Zionts in her job search, which included calling Fattah. Fattah hired her, a move that put his congressional office overbudget. Zionts worked in Fattah’s office for only about two months, leaving to work for a congressman from Florida. Tia Watson, who performed constituent services for Fattah and worked on the same floor as Zionts in Fattah’s district office, testified she had no idea what work Zionts performed. Although Zionts contacted Temple University about archiving Fattah’s papers from his career in both the state and federal government, an employee from Temple University observed that Zionts’ work contributed nothing of value to the papers project. D. The Indictment and Trial Fattah’s schemes eventually unraveled. On July 29, 2015, a federal grand jury in the Eastern District of 39 Pennsylvania returned a twenty-nine count indictment alleging that Fattah and his associates had engaged in a variety of criminal acts. Fattah, Vederman, Nicholas, Brand, and Bowser were charged with unlawfully conspiring to violate the Racketeer Influenced and Corrupt Organizations Act (RICO), 18 U.S.C. § 1962(d). In addition to the RICO charge, the indictment alleged that Fattah and certain co-defendants had unlawfully conspired to commit wire fraud, 18 U.S.C. §§ 1343, 1349; honest services fraud, 18 U.S.C. §§ 1343, 1346, 1349; mail fraud, 18 U.S.C. §§ 1341, 1349; money laundering, 18 U.S.C. § 1956; and to defraud the United States, 18 U.S.C. § 371. Several defendants were also charged with making false statements to banks, 18 U.S.C. § 1014; falsifying records, 18 U.S.C. § 1519; laundering money, 18 U.S.C. § 1957; and engaging in mail, wire, and bank fraud, 18 U.S.C. §§ 1341, 1343, and 1344. The RICO charge alleged that the defendants and other co-conspirators constituted an enterprise aimed at supporting and promoting Fattah’s political and financial interests. The efforts to conceal the $1 million Lord loan and its repayment are at the heart of the RICO conspiracy and the Fattah for Mayor scheme. The indictment further alleged that the RICO enterprise involved: (1) the scheme to satisfy an outstanding campaign debt by creating the fake “Blue Guardians” nonprofit; and (2) the bribery scheme to obtain payments and things of value from Vederman in exchange for Fattah’s efforts to secure Vederman an appointment as a United States Ambassador. 40 A jury trial, before the Honorable Harvey Bartle III of the Eastern District of Pennsylvania, began on May 16, 2016, and lasted about a month. 7 Judge Bartle charged the jury on Wednesday, June 15, 2016, and deliberations began late that afternoon. The following day, after deliberating for only four hours, the jury sent a note to the judge. Written by the foreperson, the note read: Juror Number 12 refuses to vote by the letter of the law. He will not, after proof, still change his vote. His answer will not change. He has the 11 of us a total wreck knowing that we are not getting anywhere in the hour of deliberation yesterday and the three hours today. We have zero verdicts at this time all due to Juror Number 12. He will not listen or reason with anybody. He is killing every other juror’s experience. We showed him all the proof. He doesn’t care. Juror Number 12 has an agenda or ax to grind w/govt. JA5916. Shortly after receiving the foreperson’s note, the Court received a second communication—a note signed 7 The District Court dismissed one charge prior to trial: an individual money laundering count against Nicholas. 41 by nine jurors, including the foreperson. The second note read: We feel that [Juror 12] is argumentative, incapable of making decision. He constantly scream [sic] at all of us. Id. Judge Bartle met with counsel in his chambers and advised them of his intention to voir dire both the foreperson and Juror 12 in an effort to determine whether the juror was deliberating as required by his oath. The Judge also indicated that he would “stay away from the merits of the case,” and that whether he would voir dire more jurors “remain[ed] to be seen.” JA5917. Counsel for the defendants objected to the Court’s proposed inquiry. As a group, they indicated that while the note could be read as suggesting “a flat refusal to deliberate,” they were of the opinion that it sounded “more in the manner of a disagreement over the evidence.” JA5918. Nicholas’s counsel specifically argued that questioning the jurors so quickly after the start of deliberations would send a message that differences of opinion among a block of jurors could be resolved by complaining to the Court. Defense counsel acknowledged that the case law gave Judge Bartle wide discretion on how to proceed, but suggested that a “less intrusive” course of action was preferred. JA5918–19. They collectively urged 42 the Court to do nothing more than remind the jurors of their duty to deliberate. The Government agreed with Judge Bartle’s proposed voir dire. In the prosecution’s view, the Court had already given proper instructions to the jury on their duty to deliberate. The Government further argued that if Juror 12 had exhibited bias, as suggested in the notes, he would have lied during the voir dire process and his refusal to deliberate would be “further evidence of that and his unsuitability as a juror.” JA5921. With all counsel present, and over defense counsel’s objections, Judge Bartle ultimately questioned five jurors in chambers. He questioned Juror 2 (the foreperson), Juror 12 (the subject of the complaints), Juror 3, Juror 6, and Juror 1. Judge Bartle began each voir dire by informing the juror that he would ask a series of questions, but would not inquire into the merits of the case or how any juror was voting. Each juror was placed under oath, and Judge Bartle asked, among other questions, whether screaming was occurring; whether the jurors were discussing the evidence; whether Juror 12 was placing his hands on other jurors; and whether Juror 12 was unwilling to follow his instructions. The foreperson acknowledged that he had written the initial note during lunch earlier that day. He stated that Juror 12 was not willing to follow the law, but instead 43 “want[ed] to add his own piece of the law . . . which has nothing to do with it.” JA5927–28. The foreperson further testified that Juror 12 “was standing up screaming” and that “[i]t was everybody pretty much against [Juror 12].” JA5929. He testified that Juror 12 “has his own agenda,” and that Juror 12 put his hand on another juror. JA5930. The foreperson also stated that the jury had discussed only a single count since the day before, and that they were still discussing it. When the District Court responded that the jurors should understand that they could take as much time as they needed, the foreperson responded: “I understand that. . . . [W]e all understand it. But we feel that he’s just— he’s got another agenda.” JA5934. Judge Bartle advised counsel that he considered this “a very serious situation” and that he would proceed to voir dire Juror 12. JA5937. Fattah’s counsel renewed his objection to questioning Juror 12, which the Court overruled. Brand’s counsel argued that because the Court had decided to voir dire Juror 12, it should also voir dire an additional juror. The Court agreed to do so. When the Court questioned Juror 12, he admitted to having “yelled back” at others, but only when they raised their voices to him. JA5939. Juror 12 contended that he, in fact, was “the only one” deliberating. Id. When an initial vote was taken the previous afternoon, his vote “was different than everybody else’s.” Id. Juror 12 explained to the other jurors why his vote was different, bringing up specific evidence. In response, the other jurors said “that 44 doesn’t mean anything” and “pointed to the indictment.” JA5940. Juror 12 told the other jurors that the indictment is not evidence. Id. In response, the others “threatened to have [him] thrown off.” Id. Juror 12 testified that a similar sequence of events had taken place that morning. After a brief period of deliberations, another vote was taken, and with the same result as the previous afternoon. A discussion ensued, and the other jurors again “point[ed] to the indictment.” Id. Juror 12 told them to “read the charge,” “[t]he indictment is not evidence.” Id. They read the charge, and Juror 12 again attempted to explain his view, but the other jurors paid little attention. Accordingly, Juror 12 told the others that if they did not want him there, he “[didn’t] want to be [there]”—he would be “[o]kay with it” if they wanted him taken off the jury. JA5941. Upon hearing this testimony, Judge Bartle again asked about the tone of deliberations. Juror 12 repeated that he raised his voice only in response to others who did so—he did “not want to yell at anybody.” JA5942. Judge Bartle then asked whether he had touched other jurors. Juror 12 replied that he had not hurt anyone. When asked if he had put his hand on anybody’s shoulder, Juror 12 answered: “I couldn’t remember to be honest with you.” JA5946. Following Juror 12’s voir dire, the Court summoned Juror 3 to chambers. Juror 3 testified that, after discussion of a particular count, there was one juror at odds with the 45 others. According to Juror 3, “the rest of the jurors pounced on the gentleman with the . . . dissenting opinion.” JA5948. Juror 3 testified that Juror 12 “got very defensive and just a little bit [] impatient” and that “the other jurors were very impatient with him.” Id. Juror 3 did not recall witnessing Juror 12 putting his hand on any other jurors. The Government requested that the Court voir dire another juror. Defense counsel objected, claiming that the questioning “threaten[ed] . . . the entire deliberative process.” JA5949–50. Judge Bartle reminded counsel that he had the authority to question each juror, and called for voir dire of Juror 6. Juror 6 testified that the jury had been discussing the case and reviewing the evidence, but that Juror 12 “wants to be seen” and was “being obstinate.” JA5951–52. According to Juror 6, Juror 12 “may not agree” with the conclusion of other jurors but “doesn’t give valid reasons as to why he may disagree with the charge.” JA5952. Juror 6 also revealed that Juror 12 was the first to raise his voice, and that he may have touched her and another juror. When asked to clarify what she meant by Juror 12 disagreeing with “the charge,” Juror 6 testified that Juror 12 was “reading maybe too deeply into it and putting his own emotions into it instead of just looking at what it says [and] what the facts are.” JA5952, 5955. According to Juror 6, Juror 12 “just continues to read past that into his own mind of what he feels it should be.” JA5955. Juror 6 testified 46 that Juror 12’s “justification for some of his responses [did not] seem to relate to what the matter [was] before [them].” JA5957. Judge Bartle chose to hear from yet another juror. Juror 1 was called and informed the Court and counsel that the jury “really [hadn’t] been able to even start the deliberation process” in light of the disruptive behavior of “one particular individual.” JA5958–59. The particular individual, according to Juror 1, was “very opinionated” and “[came] into the process with his view already established, refusing to even listen to any of the evidence . . . [being] very forceful . . . standing up, yelling, pointing his finger.” JA5959. When asked if this individual was willing to follow the Court’s instructions, Juror 1 testified that he “pours [sic] over the documents very well” but that he was adding other factors to answer the question on the verdict form, such as “what did this person feel.” JA5961. When Judge Bartle advised that intent was an appropriate consideration, Juror 1 agreed but said that Juror 12 was “trying to investigate . . . going way beyond the scope” of the evidence before them. JA5961–62. Juror 12, he said, “has an opinion and that opinion is established.” JA5962. He stated that Juror 12 was “not willing to listen to any sort of reason or any sort of what everyone else is saying” but instead, was “trying to force everyone else to get to his point of view.” Id. “[I]f he feels like he’s not getting there, he gets louder and louder and points and puts his hand on your shoulder . . . .” Id. 47 After questioning the five jurors (Jurors 1, 2, 3, 6, and 12), Judge Bartle heard argument from counsel. The attorney for the Government pointed out that the Court would have to make a credibility determination because Juror 12 stated that he did not recall touching anyone. In the Government’s view, Juror 12 was disrupting the process and should be removed. Defense counsel disagreed. They argued that Juror 12 was conscientious and was engaging with the evidence. They pointed out that despite the testimony that Juror 12 was reading too deeply into the instructions or introducing new factors for the jury to consider, Juror 6 had testified that the jurors “talked it through” and resolved the concern. JA5965. Defense counsel argued that the jury was discussing intent, an issue that was at the heart of the case. Defense counsel perceived no breakdown in deliberations and argued that dismissal would be premature. They suggested, instead, that the Court provide a supplemental instruction. Judge Bartle decided to adjourn for the afternoon. But before he left the courtroom, defense counsel brought two matters to his attention. First, in light of testimony during the voir dire, they asked that the jury be reinstructed that the verdict form and indictment were not evidence. Second, they apprised the Judge of the standard for juror dismissal set forth in United States v. Kemp, 500 F.3d 257 (3d Cir. 2007). Defense counsel stated that under Kemp, a request to discharge a juror must be denied if there is a possibility that the request stems from the juror’s view of the evidence. Judge Bartle expressed hesitation on 48 reinstructing the jury, but agreed that Kemp would control his determination as to whether dismissal was appropriate. With the following morning came a new revelation. With counsel in chambers, the Judge informed them that “additional significant evidence” had come to light since the previous day’s recess. JA5980. He placed his courtroom deputy under oath, and she proceeded to testify to an exchange that had occurred the previous day as she was escorting Juror 12 back to the jury room after he had been voir dired. According to the deputy, Juror 12 stopped her in the hallway, placed his hand on her shoulder, and looked her “straight in the eye.” JA5981. He then said: “I’m going to hang this jury.” Id. The deputy then related that before any further conversation could take place between Juror 12 and the deputy, Judge Bartle summoned Juror 12 back to his chambers. Later that day, however, Juror 12 and the courtroom deputy had another exchange. She testified that after all five jurors had been questioned, Juror 12 emerged from the jury room and told her “I really need to talk to you.” JA5982. She informed Judge Bartle and counsel that Juror 12 “said more about how they’re treating him and what he’s saying to them.” Id. He flatly stated that “it’s going to be 11 to 1 no matter what.” Id. There were no follow-up questions for the deputy. Instead, defense counsel suggested that what Juror 12 may have meant was that he was willing to hang the jury because of a lack of evidence. They requested that Juror 12 be asked about his comments to the deputy. 49 After once again summoning Juror 12 to his chambers, the Judge advised him that “[s]ome questions have arisen” about what he may have done after being voir dired the previous day. JA5985. Juror 12 acknowledged having conversations with the courtroom deputy. When asked “what happened” and “[w]hat occurred,” Juror 12 responded: “Basically, I said that there was a lot of name calling going on.” JA5985. He said comments had been made by other jurors about his service in the military. He specifically referred to other jurors’ suggesting that he had possibly “hit [his] head . . . hard a few times” while serving in a parachute regiment. JA5986. He testified he had conveyed these comments to the deputy and that he found them offensive. When asked if he said anything else to the deputy, Juror 12 responded: “I may have. I really can’t recall.” JA5987. And when Judge Bartle followed up by asking if he could recall anything else that he said to the deputy, Juror 12 simply replied: “No. To me, that was the most important thing.” Id. Juror 12 was then excused from chambers. Defense counsel next requested that the juror be asked directly whether he told the courtroom deputy that he was going to “hang this jury.” JA5988. Juror 12 was recalled to chambers, and the following back and forth took place: The Court: You may be seated. And, of course, [Juror 12], you know you’re under oath here from yesterday? 50 Juror 12: Yes, sir. The Court: . . . Did you say to [the courtroom deputy] that you’re going to hang this jury? Juror 12: I said I would. The Court: You did? Juror 12: I did. I said—I told her—I said, we don’t agree; I’m not just going to say guilty because everybody wants me to, and if that hangs this jury, so be it. .... Juror 12: I did say that, sir. The Court: You didn’t remember that before? Juror 12: I’m more concerned about people spitting on my military record. The Court: Did you say that you’d hang the jury no matter what? Juror 12: If they do—if we cannot come to— The Court: No. The question is what you said to her. Did you say to her you would hang the jury no matter what? Juror 12: I can’t really remember that. I did say that if we didn’t—a person—no matter what, I can’t recall that exactly. The Court: All right. Thank you very much. You can wait just out there in the anteroom. JA5989–90. Defense counsel continued to oppose Juror 12’s dismissal. They argued that the juror’s concern was about 51 the evidence, and that his comments to the courtroom deputy reflected a conviction that “he’s not going to agree just because others want him to agree.” JA5991. They also argued that nothing should be made of Juror 12’s failure to mention the comments when initially questioned by the Court, and that a supplemental instruction was all that was warranted given the early stage of the deliberations. The Government strongly disagreed. The Assistant United States Attorney argued that Juror 12 “should absolutely be removed” because “his demeanor ha[d] demonstrated a hostility . . . both to the other jurors and to the court.” JA5993. The Government also suggested that Juror 12’s comments that he would hang the jury meant that he was not participating in the deliberations and was ignoring the evidence and the law. Ruling from the bench, Judge Bartle announced: I find [the deputy clerk] to be credible. I find [Juror 12], not to be credible. I find that [Juror 12] did tell [the deputy clerk] that he was going to hang this jury no matter what. There have been only approximately four hours of deliberation. There’s no way in the world he could have reviewed and considered all of the evidence in the case and my instructions on the law. I instructed the jury to deliberate, meaning to discuss the evidence; obviously, to hold onto your honestly held beliefs, but at 52 least you have to be willing to discuss the evidence and participate in the discussion with other jurors. Juror number 12 has delayed, disrupted, impeded, and obstructed the deliberative process and had the intent to do so. I base that having observed him, based on his words and his demeanor before me. He wants only to have his own voice heard. He has preconceived notions about the case. He has violated his oath as a juror. And I do not believe that any further instructions or admonitions would do any good. I think he’s intent on, as he said, hanging this jury no matter what the law is, no matter what the evidence is. Therefore, he will be excused, and I will replace him with the next alternate . . . . JA5994–95. In response, defense counsel moved for a mistrial, which the judge promptly denied. He then informed the reconstituted jury that deliberations would need to start over, and reinstructed them on certain points of law, including that the verdict slip does not constitute evidence. Judge Bartle elaborated upon his decision to remove Juror 12 in two post-trial memorandum opinions. In the first, ruling on a media request for the sealed transcripts, he explained: 53 Here, there is no doubt that Juror 12 intentionally refused to deliberate when he declared so early in the process that he would hang the jury no matter what. This finding was predicated on the admission of Juror 12 as reported by the court’s deputy clerk. The facts became clear to the court after hearing the credible testimony of the deputy clerk and the less credible testimony of Juror 12. The demeanor of Juror 12 before the court confirmed the court’s findings. GSA23–24. The second opinion addressed motions for bail pending appeal from Nicholas and Brand. GSA25. There, Judge Bartle explained: The law is well-settled that the court has discretion to act as it did under these circumstances. See United States v. Kemp, 500 F.3d 257, 304 (3d Cir. 2007). The court, after taking testimony, specifically found that the juror, following only a few hours of deliberation, stated to the court’s courtroom deputy clerk that he would hang this jury no matter what. He could not possibly have reviewed all of the law and evidence of this five-week trial at the time he made his remark. The court examined the deputy clerk and the juror under oath in the presence of counsel for all parties. The undersigned 54 found the deputy clerk to be credible and the juror not to be credible. Based on the juror’s demeanor, it was clear he would not change his attitude and that his intent had been and would continue to be to refuse to deliberate in good faith concerning the law and the evidence. GSA32. After deliberating for approximately 15 hours, the jury returned with its verdicts on June 21, 2016, finding the defendants guilty on most counts. Fattah, Vederman, and Brand were convicted on all counts. The jury acquitted Bowser on sixteen counts, but found her guilty of the bribery conspiracy and the associated charges of bank fraud, making false statements to a financial institution, falsifying records, and money laundering (Counts 16, 19, 20, 21 and 22). The jury also acquitted Nicholas of wire fraud (Count 24). See Nicholas Supp. App. (NSA) 36. The following week, on June 27, the Supreme Court issued its opinion in McDonnell v. United States, 136 S. Ct. 2355 (2016). McDonnell provided new limitations on the definition of “official acts” as used in the honest services fraud and bribery statutes under which Fattah and Vederman had been convicted. Id. at 2369–72. Fattah and Vederman both moved to set aside their convictions. The District Court “acknowledge[d] that under McDonnell our instructions to the jury on the meaning of official act turned out to be incomplete and thus erroneous.” JA103. 55 But the Court held that “the incomplete and thus erroneous jury instruction on the meaning of official acts did not influence the verdict on the bribery counts” and upheld the verdict on Counts 16–18 and 22–23. JA107, 121. Fattah, Vederman, and Bowser had more success with their other post-verdict motions. The District Court, in a thoughtful opinion, granted relief under Federal Rule of Criminal Procedure 29, acquitting Vederman of the RICO conspiracy (Count 1) and Fattah, Vederman, and Bowser of bank fraud, making false statements to a financial institution, and falsifying records (Counts 19, 20, and 21). JA37–139. This appeal followed. 8 The defendants raise a variety of challenges to their convictions. All defendants but Bowser challenge the District Court’s decision to dismiss Juror 12. Fattah and Vederman argue that the District Court erred in upholding the jury’s verdict on the bribery and honest services fraud counts in light of the Supreme Court’s decision in McDonnell. Fattah, Brand and Nicholas challenge the sufficiency of the evidence underlying the RICO conviction. Several of the defendants contend the District Court erred in its instruction on intent and by sending the indictment out to the jury. There are 8 Fattah, Brand, Vederman, and Nicholas each filed a timely notice of appeal, but Bowser did not challenge her convictions. 56 also several evidentiary challenges.9 The Government cross-appeals from the District Court’s judgment acquitting Fattah and Vederman on Counts 19 and 20, arguing that the District Court erred in interpreting the definition of a “mortgage lending business” under 18 U.S.C. § 27. We address these arguments in turn. We hold that the District Court erred in upholding the jury verdict in light of McDonnell, and we will therefore reverse and remand for retrial on Counts 16, 17, 18, 22, and 23. We also hold that the District Court erred 9 Pursuant to Rule 28(i), “Fattah joins in the arguments of Herbert Vederman, Robert Brand, and Karen Nicholas to the extent their arguments on appeal apply to Mr. Fattah.” Fattah Br. 19 n.69. Federal Rule of Appellate Procedure 28(i) provides that a defendant, “[i]n a case involving more than one appellant . . . may adopt by reference a part of another’s brief.” Here, Fattah’s decision to join fails to specify which of the many issues of his codefendants he believes worthy of our consideration. Rather, it appears that he presumes we will scour the record and make that determination for him. This type of blanket request fails to satisfy Rule 28(a)(5)’s directive requiring that the “appellant’s brief must contain . . . a statement of the issues presented for review.” Fed. R. App. P. 28(a)(5). We conclude that expecting the appellate court to identify the issues to be adopted simply results in the abandonment and waiver of the unspecified issues. See Kost v. Kozakiewicz, 1 F.3d 176, 182 (3d Cir. 1993). 57 in acquitting Fattah and Vederman on Counts 19 and 20. Because the jury’s verdict was supported by the evidence, we will reinstate the convictions as to those counts. In all other respects, we will affirm the judgment of the District Court. III. Juror Misconduct and Dismissal of Juror 12 10 Defendant Fattah challenges the District Court’s decision to conduct an in camera inquiry into alleged juror misconduct and the ultimate dismissal of Juror 12. 11 We reject both challenges. The record reveals credible allegations of juror misconduct and a sufficient basis to support the finding that Juror 12 violated his oath. A. Investigation of Alleged Juror Misconduct We first consider whether the District Court erred in its handling of the two notes from jurors. A trial court’s response to allegations of juror misconduct is reviewed under an abuse of discretion standard. United States v. Boone, 458 F.3d 321, 326 (3d Cir. 2006) (citing United States v. Resko, 3 F.3d 684, 690 (3d Cir. 1993)). We conclude that the District Court did not abuse its discretion 10 The District Court had jurisdiction under 18 U.S.C. § 3231. We have jurisdiction under 28 U.S.C. § 1291 and 18 U.S.C. § 3742(a). 11 Vederman, Nicholas, and Brand adopt Fattah’s claim of reversible error concerning the dismissal of Juror 12. 58 in addressing the issues raised in the jurors’ notes to the Court. Trial courts are afforded discretion in responding to allegations of juror misconduct. This is so because “the trial court is in a superior position to observe the ‘mood at trial and the predilections of the jury.’ ” Resko, 3 F.3d at 690 (quoting United States v. Chiantese, 582 F.2d 974, 980 (5th Cir. 1978)). But this discretion is not unlimited. Once the jury retires to deliberate, the confidentiality of its deliberations must be closely guarded. An accused is constitutionally entitled to be tried before a jury of his peers. As ordinary citizens, jurors are “expected to speak, debate, argue, and make decisions the way ordinary people do in their daily lives.” Pena-Rodriguez v. Colorado, 137 S. Ct. 855, 874 (2017) (Alito, J., dissenting). To protect against intrusion into a defendant’s right to be judged only by fellow citizens, “the door to the jury room [is] locked.” Id. at 875. In Boone, this Court considered the threshold for intervention by a trial judge who is presented with allegations of juror misconduct during the course of deliberations. 458 F.3d at 327. We recognized that “[i]t is beyond question that the secrecy of deliberations is critical to the success of the jury system.” Id. at 329. But that secrecy abuts a competing interest—the jury’s proper execution of its duties. That is, “a juror who refuses to deliberate or who commits jury nullification violates the sworn jury oath and prevents the jury from fulfilling its 59 constitutional role.” Id. Recognizing these competing interests, we declined in Boone to adopt a sweeping limitation on a trial court’s ability to investigate allegations of misconduct during jury deliberations. See id. Consistent with the standard applied at other stages of criminal proceedings, Boone teaches that “where substantial evidence of jury misconduct—including credible allegations of jury nullification or of a refusal to deliberate—arises during deliberations, a district court may, within its sound discretion, investigate the allegations through juror questioning or other appropriate means.” Id. Fattah argues that the District Court had no basis to question any of the jurors. Fattah Br. 20. We disagree. In Boone, notes from the jury presented substantial credible evidence of misconduct. 458 F.3d at 330. Here, the initial note from the foreperson alleged that Juror 12 “refuse[d] to vote by the letter of the law,” would “not listen or reason with anybody,” and that he had “an agenda or ax to grind” with the Government. JA5916. The note contained allegations of both a refusal to deliberate and a suggestion of nullification. A refusal to deliberate is a violation of a juror’s oath. Boone, 458 F.3d at 329 (citing United States v. Baker, 262 F.3d 124, 130 (2d Cir. 2001) (“It is well- settled that jurors have a duty to deliberate.”)). Moreover, nullification—a juror’s refusal to follow the law—is a violation of the juror’s sworn oath to render a verdict according to the law and evidence. See United States v. Thomas, 116 F.3d 606, 614–18 (2d Cir. 1997) (discussing 60 both “benevolent” and “shameful” examples of juror nullification, but “categorically reject[ing] the idea that, in a society committed to the rule of law, jury nullification is desirable or that courts may permit it to occur when it is within their authority to prevent”). The second jury note, signed by nine jurors, supported the claim of misconduct by asserting that Juror 12 was “incapable of making decision[s]” and was “constantly scream[ing]” at the other jurors. JA5916–17. We conclude that the District Court did not abuse its discretion in deciding to initially question Juror 2, and subsequently, Jurors 12, 3, 6 and 1. Fattah also challenges the scope of the District Court’s questioning. He argues that the rights to an impartial jury and to a unanimous verdict “would be rendered toothless if trial courts had free rein to question jurors during deliberations.” Fattah Br. 36. Indeed, we acknowledged the legitimacy of such a concern in Boone. Despite adopting a modest “credible allegations” standard for investigating misconduct, we “ke[pt] in mind the importance of maintaining deliberative secrecy.” Boone, 458 F.3d at 329. Fattah asserts that the trial court’s questions to the five jurors were “intrusive and pointed” and “nothing like the questioning . . . approved in Boone.” Fattah Br. 38. But Fattah does not elaborate on how, in his view, the questions posed by Judge Bartle specifically intruded into deliberative secrecy. To be sure, Judge Bartle’s questioning of each juror was more extemporaneous than the juror questioning in 61 Boone. There, the district court asked a single juror four “concise and carefully-worded” questions. 458 F.3d at 330. Judge Bartle’s voir dire of each of the five jurors took on a more conversational tone. We take no issue with that approach. The substance of the judge’s questions was limited and mirrored that of questions we deemed appropriate in Kemp. There, the court conducted three rounds of questioning. In the first round, each juror was asked: (1) “Are you personally experiencing any problems with how the deliberations are proceeding without telling us anything about the votes as to guilt or innocence? If yes, describe the problem.” (2) “Are all the jurors discussing the evidence or lack of evidence?” (3) “Are all the jurors following the court’s instructions on the law?” Kemp, 500 F.3d at 273. In the second and third rounds, each juror was asked: (1) “Is there any juror or jurors who are refusing to deliberate?” (2) “Is there any juror who is refusing to discuss the evidence or lack of evidence?” (3) “Is there any juror who is refusing to follow the Court’s instructions?” Id. at 274. Here, Judge Bartle began his voir dire of each juror by stating that he did not wish for the juror to discuss 62 the merits of the case or to reveal the content of the deliberations that had taken place. He asked the jurors whether screaming was occurring, whether the jurors were discussing the evidence, whether Juror 12 was placing his hands on other jurors, and whether Juror 12 was unwilling to follow his instructions. Fattah points to no specific question posed or topic discussed that was inappropriate, and we see little to no substantive difference between the questions here and those asked by the trial judge in Kemp. As in Kemp, “the District Court took care to limit its questions to appropriate matters that did not touch on the merits of the jury’s deliberation, and expressly informed each juror on multiple occasions that he or she should not reveal the substance of the deliberations.” Id. at 302 (citing United States v. Edwards, 303 F.3d 606, 634 n.16 (5th Cir. 2002)). Fattah also argues that once the remarks of Juror 2 and Juror 12 revealed no further evidence of misconduct, the court had no basis to question other jurors. Fattah Reply 19. Yet, our cases make clear that a trial court may, in its discretion, examine each juror. Kemp, 500 F.3d at 302 (“We have recognized that there are times in which individual questioning is the optimal way in which to root out misconduct.”). Indeed, “the District Court must utilize procedures that will ‘provide a reasonable assurance for the discovery of prejudice.’ ” Id. (quoting Martin v. 63 Warden, Huntingdon State Corr. Inst., 653 F.2d 799, 807 (3d Cir. 1981)). 12 Judge Bartle, a very able and experienced district judge, was in the best position to determine what type of inquiry was warranted under the circumstances. We conclude that his questioning of the five jurors was not an abuse of discretion. See id. at 302. B. Dismissal of Juror 12 Fattah, joined by Vederman, Brand, and Nicholas, strongly contends that the District Court committed reversible error by dismissing Juror 12. “We review the dismissal of a juror for cause for abuse of discretion.” 12 Our cases do not suggest that a trial judge confronted with allegations that a jury’s deliberations are being obstructed by one of its members should always resort to interviewing jurors. Reinstructing the jury on its duty to deliberate will often be the better course at the first sign of trouble. Mere disagreement among jurors—even spirited disagreement—is no ground for intervention. Furthermore, intrusive or leading questions about the deliberative process may work against the twin goals of protecting that process and ensuring that jurors remain faithful to their oaths. We share the Eleventh Circuit’s preference of “err[ing] on the side of too little inquiry as opposed to too much.” United States v. Oscar, 877 F.3d 1270, 1287 (11th Cir. 2017) (quoting United States v. Augustin, 661 F.3d 1105, 1133 (11th Cir. 2011)). 64 Kemp, 500 F.3d at 303. That deferential standard compels us to affirm. Rule 23(b) of the Federal Rules of Criminal Procedure permits a trial court to excuse a deliberating juror for good cause. See id. (citing Fed. R. Crim. P. 23(b)). Good cause exists where a juror refuses to apply the law, refuses to follow the court’s instructions, refuses to deliberate with his or her fellow jurors, or demonstrates bias. See Kemp, 500 F.3d at 305–06; United States v. Oscar, 877 F.3d 1270, 1287 (11th Cir. 2017); Thomas, 116 F.3d at 617. Good cause does not exist when there is reasonable but sustained disagreement about how a juror views the evidence. The courts of appeals are emphatic that trial courts “may not dismiss a juror during deliberations if the request for discharge stems from doubts the juror harbors about the sufficiency of the government’s evidence.” Kemp, 500 F.3d at 303 (quoting United States v. Brown, 823 F.2d 591, 596 (D.C. Cir. 1987)); see also Oscar, 877 F.3d at 1287 (same); United States v. Symington, 195 F.3d 1080, 1087 (9th Cir. 1999); Thomas, 116 F.3d at 622. To reinforce a defendant’s right to a unanimous jury, we have adopted a high standard for juror dismissal. Kemp, 500 F.3d at 304 & n.26. “[D]istrict courts may discharge a juror for bias, failure to deliberate, failure to follow the district court’s instructions, or jury nullification when there is no reasonable possibility that the allegations of misconduct stem from the juror’s view of the evidence.” 65 Id. at 304 (emphasis added). This “no reasonable possibility” standard is “by no means lax.” Id. Rather, “[i]t corresponds with the burden for establishing guilt in a criminal trial.” Id. We first applied this standard in Kemp, but have not had occasion to do so since. There, the evidence supporting the district court’s removal decision was “overwhelming.” 500 F.3d at 304. Ten jurors separately and consistently reported that a juror was improperly biased, and did so only after three rounds of questioning and careful and correct instructions from the district court as to the distinction between appropriate skepticism and impermissible bias. Id. at 304–05; see id. at 275–76 (district court’s instruction). The testimony also showed that the juror in question refused to deliberate or to discuss the evidence with her fellow jurors. Id. at 305. Whether the evidence of misconduct in this case is as strong as that in Kemp is beside the point. After only four hours of deliberations, Juror 12 stated unequivocally to the courtroom deputy that he was “going to hang” the jury, and that it would be “11 to 1 no matter what.” JA5981–82 (emphasis added). These statements, coupled with the District Court’s finding that Juror 12 lacked credibility, provided a sufficient basis for Juror 12’s dismissal. As grounds for excusing Juror 12, the District Court found that he refused to deliberate in good faith, “delayed, disrupted, impeded, and obstructed the deliberative 66 process and had the intent to do so,” JA5995, and that he was “intent on . . . hanging this jury no matter what the law is, no matter what the evidence is.” Id. The District Court determined from this that Juror 12 had violated his oath as a juror and that no further instructions or admonitions could rehabilitate the juror. Id. The District Court based these findings on personal observation, including Juror 12’s words and demeanor, and making the specific finding that Juror 12 was not credible. That finding is amply supported by the record. In United States v. Abbell, 271 F.3d 1286, 1303 (9th Cir. 2001), the Ninth Circuit recognized that “the demeanor of the pertinent juror is important to juror misconduct determinations” because the “juror’s motivations and intentions are at issue.” That court emphasized, as we do, that a district judge is best situated to assess the demeanor of a juror. Id. Here, Juror 12 stated he could not recall putting his hand on another juror’s shoulder, while his fellow jurors’ testimony was consistent on this point. Juror 12 also failed, at first, to recall his troubling statements to the courtroom deputy despite having made those statements only the previous afternoon. When questioned a second time and asked directly about the statements, he admitted to saying that he would hang the jury but claimed he could not “really remember” saying “no matter what” the day before. JA5989–90. Juror 12’s spotty recollection of the previous day’s events further supports the District Court’s finding that he was not credible. 67 Fattah argues that the credibility determination was not, by itself, a sufficient reason to dismiss the juror because the record demonstrates more than a reasonable possibility that the complaints about his conduct stemmed from Juror 12’s own view of the Government’s case. Fattah Reply Br. 11; Fattah Br. 25, 28. Fattah claims that the District Court abused its discretion by dismissing Juror 12 “on the basis of, in effect, six words the juror purportedly said to the court’s deputy after he was verbally attacked by other jurors.” Fattah Br. 24. According to Fattah, the questioning of the other jurors “confirmed that there were no legitimate grounds for removing juror 12.” Id. at 26. We conclude otherwise. “A district court’s finding on the question whether a juror has impermissibly refused to participate in the deliberation process is a finding of fact to which appropriate deference is due.” Baker, 262 F.3d at 130. While district courts must apply a high standard for juror dismissal, their underlying findings are afforded considerable deference on appeal. Kemp, 500 F.3d at 304 (citing Abbell, 271 F.3d at 1302–03). We will reverse only if the decision to dismiss a juror was “without factual support, or for a legally irrelevant reason.” Abbell, 271 F.3d at 1302 (citation omitted). Here, the District Court had a legitimate reason for removing Juror 12. Refusal to deliberate constitutes good cause for dismissal. Although the judge did not expressly articulate the Kemp standard when he announced that he 68 would dismiss Juror 12, he did acknowledge the “no reasonable possibility” standard in his discussion with counsel. The unmistakable import of the District Court’s statement from the bench is that there was no reasonable possibility that Juror 12’s intransigence was based on his view of the evidence. See Oscar, 877 F.3d at 1288 n.16. Fattah contends that there is no record support for the finding that Juror 12 said “he was going to hang this jury no matter what.” Fattah Br. 29. To be sure, the courtroom deputy’s testimony is not that Juror 12 used the words “hang this jury” and “no matter what” in the same sentence. She testified that Juror 12 first stated “I am going to hang this jury,” then later stated “it is going to be 11 to 1 no matter what.” JA5981–82. This is a distinction without a difference. Likewise, Fattah challenges the District Court’s finding that Juror 12 was determined to hang the jury “no matter what the law is” and “no matter what the evidence is.” Fattah Br. 29. Although there is no evidence that Juror 12 uttered the phrases “no matter what the law is” or “no matter what the evidence is,” the District Court was describing the import of Juror 12’s statements. This was not error. Fattah expresses the concern that “[i]f jurors are asked the right questions or interrogated long enough, it would not be difficult for a trial court to elicit testimony from [a] majority [of] jurors that can be held up as evidence of a dissenting juror’s bias or refusal to deliberate.” Fattah Br. 22. He also worries that a group of 69 jurors might have an incentive to rid themselves of a juror who holds a different view. Id. These are valid concerns— but no basis existed for such concerns in this case. Juror 12’s own words provided most of the support for his eventual dismissal. Furthermore, his statements were made early in the deliberations, in a complex case, before any juror could reasonably be expected to have reached final verdicts on the twenty-nine counts before the jury. The able District Judge did not err in finding that Juror 12 refused to deliberate and therefore violated his oath. IV. The District Court’s Instructions Under McDonnell On appeal, Fattah and Vederman renew their challenge to the jury instructions given on Counts 3, 16, 17, 18, 22, and 23, concerning the meaning of the term “official act” as used in the bribery statute (pursuant to which both were convicted) and the honest services fraud statute (pursuant to which Fattah alone was convicted). In light of the Supreme Court’s opinion in McDonnell v. United States, 136 S. Ct. 2355 (2016), released the week after the jury verdict, the District Court conceded that its instructions were incomplete and erroneous, at least as to Counts 16–18. Nevertheless, the District Court held that the erroneous jury instructions had not influenced the verdict on the bribery counts, and declined to set aside Fattah and Vederman’s convictions. 70 As to Counts 16–18 and 22–23, we disagree, and will reverse the District Court’s judgment. The District Court’s judgment with respect to Count 3, which did not involve Vederman, will be affirmed. JA78–79. A. The McDonnell Framework In McDonnell, the Supreme Court interpreted the term “official act” as defined in 18 U.S.C. § 201(a)(3). 136 S. Ct. at 2368. The statute defines an “official act” as “any decision or action on any question, matter, cause, suit, proceeding or controversy, which may at any time be pending, or which may by law be brought before any public official, in such official’s official capacity, or in such official’s place of trust or profit.” 18 U.S.C. § 201(a)(3). The McDonnell Court distilled this definition into two requirements: First, the Government must identify a “question, matter, cause, suit, proceeding or controversy” that “may at any time be pending” or “may by law be brought” before a public official. Second, the Government must prove that the public official made a decision or took an action “on” that question, matter, cause, suit, proceeding, or controversy, or agreed to do so. 136 S. Ct. at 2368. Applying this two-step test to Governor Robert McDonnell’s convictions, the Supreme Court concluded that “the jury was not correctly instructed on 71 the meaning of ‘official act,’ ” and as a result, “may have convicted Governor McDonnell for conduct that is not unlawful.” Id. at 2375. Given that uncertainty, the Court “[could not] conclude that the errors in the jury instructions were ‘harmless beyond a reasonable doubt.’ ” Id. (quoting Neder v. United States, 527 U.S. 1, 16 (1999)). The Supreme Court, therefore, vacated Governor McDonnell’s convictions. Id. McDonnell lays out a clear path for the Government to follow in proving that an accused has performed an “official act.” First, the Government must “identify a ‘question, matter, cause, suit, proceeding or controversy’ that ‘may at any time be pending’ or ‘may by law be brought’ before a public official.” 136 S. Ct. at 2368 (quoting 18 U.S.C. § 201(a)(3)). This first step is divided into two sub-components. In Step 1(A), the Government must “identify a ‘question, matter, cause, suit, proceeding or controversy.’ ” Id. Step 1(B) then clarifies that the identified “question, matter, cause, suit, proceeding or controversy” be one that “ ‘may at any time be pending’ or ‘may by law be brought’ before a public official.” Id. Under Step 1(A), a “question, matter, cause, suit, proceeding or controversy” must be “a formal exercise of governmental power that is similar in nature to a lawsuit before a court, a determination before an agency, or a hearing before a committee.” Id. at 2372. Importantly, “a typical meeting, telephone call, or event arranged by a 72 public official” does not qualify as such a formal exercise of governmental power. Id. at 2368. Step 1(B) then requires us to ask whether the qualifying “question, matter, cause, suit, proceeding or controversy” was one that “ ‘may at any time be pending’ or ‘may by law be brought’ before a public official.” Id. As the McDonnell Court clarified, “ ‘[p]ending’ and ‘may by law be brought’ suggest something that is relatively circumscribed—the kind of thing that can be put on an agenda, tracked for progress, and then checked off as complete.” Id. at 2369; accord United States v. Repak, 852 F.3d 230, 252 (3d Cir. 2017) (quoting McDonnell, 136 S. Ct. at 2369). By contrast, matters described at a high level of generality—for example, “[e]conomic development,” “justice,” and “national security”—are not sufficiently “focused and concrete.” McDonnell, 136 S. Ct. at 2369. In McDonnell, the Court concluded that at least three questions or matters identified by the Fourth Circuit were sufficiently focused: (1) whether researchers at any of Virginia’s state universities would initiate a study of [a drug]; (2) whether the state-created Tobacco Indemnification and Community Revitalization Commission would allocate grant money for the study of [a chemical compound]; and (3) whether the health insurance plan for state employees in 73 Virginia would include [a specific drug] as a covered drug. Id. at 2370 (internal quotations omitted) (quoting United States v. McDonnell, 792 F.3d 478, 515–16 (4th Cir. 2015)). We provided guidance in the form of a fourth example in Repak, when we held that a redevelopment authority’s awarding of contracts was “a concrete determination made by the [redevelopment authority’s] Board of Directors.” 852 F.3d at 253. Step 2 requires the Government to prove that the public official made a “decision” or took “an action” on the identified “question, matter, cause, suit, proceeding or controversy.” McDonnell, 136 S. Ct. at 2368. The McDonnell Court explained: Setting up a meeting, hosting an event, or calling an official (or agreeing to do so) merely to talk about a research study or to gather additional information . . . does not qualify as a decision or action on the pending question of whether to initiate the study. Simply expressing support for the research study at a meeting, event, or call—or sending a subordinate to such a meeting, event, or call—similarly does not qualify as a decision or action on the study, as long as the public official does not intend to exert pressure on another official or provide advice, knowing 74 or intending such advice to form the basis for an “official act.” Id. at 2371. The Court further clarified: If an official sets up a meeting, hosts an event, or makes a phone call on a question or matter that is or could be pending before another official, that could serve as evidence of an agreement to take an official act. A jury could conclude, for example, that the official was attempting to pressure or advise another official on a pending matter. And if the official agreed to exert that pressure or give that advice in exchange for a thing of value, that would be illegal. Id. Here, Fattah was charged with engaging in three categories of official acts, which we analyze in accordance with the McDonnell framework. In Counts 16–18 and 22– 23, Fattah is alleged to have set up a meeting between Vederman and the U.S. Trade Representative, attempted to secure Vederman an ambassadorship, and hired Vederman’s girlfriend, all in return for a course of conduct wherein Vederman provided things of value to Fattah. In this case, as in McDonnell, the jury instructions were erroneous. We conclude that the first category of the charged acts—setting up a meeting between Vederman 75 and the U.S. Trade Representative—is not unlawful, and that the second category—attempting to secure Vederman an ambassadorship—requires reconsideration by a properly instructed jury. The third charged act—hiring Vederman’s girlfriend—is clearly an official act. But because we cannot isolate the jury’s consideration of the hiring from the first two categories of charged acts, we must reverse and remand the judgment of the District Court. B. The Kirk Meeting We turn first to Fattah’s scheduling of a meeting between Vederman and the U.S. Trade Representative, Ron Kirk. Under McDonnell, “setting up a meeting . . . does not, standing alone, qualify as an ‘official act.’ ” 136 S. Ct. at 2368. Fattah’s setting up the meeting between Vederman and Kirk was therefore not an official act, a concession implicit in the Government’s opening brief. See Gov’t Br. 32 (failing to mention the Kirk meeting as one of the “two categories” of allegedly “official acts”). But the jury was not properly instructed on this point. Without the benefit of the principles laid down in McDonnell, the jury was free to conclude that arranging the Kirk meeting was an official act—and it may have done so. The District Court’s erroneous jury instructions, therefore, cannot survive harmless error review. In a footnote in its brief to this Court, the Government argues that evidence about the Kirk meeting was offered only “because it established the strength of 76 Vederman’s desire to be an ambassador” and not because the Government was attempting to establish the meeting as an independent official act. Id. at79–80 n.6. But the record undercuts the Government’s post hoc justification. The indictment, provided to the jury in redacted form for use in its deliberations, lists Fattah’s setting up the Kirk meeting as an official act under the heading “FATTAH’s Official Acts for VEDERMAN.” JA494. Under this heading are three distinct subheadings: (1) “The Pursuit of an Ambassadorship,” (2) “The Pursuit of Another Executive Branch Position,” and (3) “Hiring the Lobbyist’s Girlfriend to the Congressional Staff.” JA494–95. The second subheading, “The Pursuit of Another Executive Branch Position,” describes the arrangement of the Kirk meeting. Quite clearly, then, this three-part structure demonstrates that setting up the Kirk meeting was one of three distinct categories of official acts alleged by the Government. Although there is some support for the Government’s argument that evidence of the Kirk meeting was presented at trial only to establish the extent of Vederman’s interest in becoming an ambassador, JA827, 852–53 (mentioning the Kirk meeting in close proximity to references to Fattah’s attempts to secure Vederman an ambassadorship), it is undermined by language in the redacted indictment itself, and by the way in which the Government presented its case at trial as a “pattern” of connected acts. 77 The redacted indictment, for example, refers to the Kirk meeting as “The Pursuit of Another Executive Branch Position.” JA495 (emphasis added). The use of the word “Another” strongly suggests that evidence about the Kirk meeting was not merely evidence of Fattah’s attempt to secure Vederman an ambassadorship, but was also evidence of a separate and distinct attempt to secure Vederman a position on a federal trade-related commission. The redacted indictment also notes that “[i]n or around May 2011, with little progress made on securing an ambassadorship for VEDERMAN, FATTAH turned towards obtaining for VEDERMAN an appointment in the Executive Branch to a federal trade commission.” Id. (emphasis added). The words “turned towards,” taken literally, clearly convey that arranging the Kirk meeting was presented as distinct from Fattah’s efforts to secure Vederman an ambassadorship. The District Court denied Fattah and Vederman a new trial on Counts 17 and 18, referring to evidence of the Kirk meeting as “de minimis” and noting that “Kirk’s testimony during this lengthy trial lasted a mere sixteen minutes.” JA97 n.14. In the District Court’s view, evidence of the Kirk meeting “played no role in the outcome” of the case. Id. Considering the record in light of McDonnell, we are not so sure. Although it is possible that evidence of the Kirk meeting played a minor role at trial when compared to the other acts on which the Government presented evidence, 78 the redacted indictment suggests that the Kirk meeting was a significant part of the Government’s case. The indictment dedicates five paragraphs to describing the Kirk meeting, but just three paragraphs to describing the hiring of Vederman’s girlfriend—a hiring that, as we explain below, is clearly an official act. JA495–96. While neither the number of minutes used at trial nor the number of paragraphs contained in an indictment is a dispositive unit of measurement for determining the significance of evidence, we conclude that the District Court’s erroneous jury instructions pertaining to the Kirk meeting were not harmless. We conclude, in accordance with McDonnell, that Fattah’s arranging a meeting between Vederman and the U.S. Trade Representative was not itself an official act. Because the jury may have convicted Fattah for conduct that is not unlawful, we cannot conclude that the error in the jury instruction was harmless beyond a reasonable doubt, and we must vacate and remand the convictions of Fattah and Vederman as to Counts 16, 17, 18, 22 and 23. C. Fattah’s Efforts to Secure Vederman an Ambassadorship The nature of Fattah’s efforts to secure Vederman an ambassadorship is less clear, and presents a closer question than the Kirk meeting. We ultimately conclude that the question warrants remand so that it may be answered by a properly instructed jury. On remand, the jury must decide whether Fattah’s conduct constituted a 79 “decision” or “action” under Step 2 of the McDonnell analysis. At the outset, it is clear to us that, under Steps 1(A) and 1(B), a formal appointment of Vederman as an ambassador would qualify as a “matter” that “may at any time be pending” before a public official. The formal appointment of a particular person (Vederman), to a specific position (an ambassadorship), constitutes a matter that is sufficiently focused and concrete. The formal appointment of an ambassador is a matter that is “pending” before the President—the constitutional actor charged with nominating ambassadors—as well Senators, who are charged with confirming the President’s ambassadorial nominations. U.S. Const. art. II § 2 (“[H]e shall nominate, and by and with the Advice and Consent of the Senate, shall appoint Ambassadors . . . .”). It is beyond cavil that the formal appointment of an ambassador satisfies both sub-components of McDonnell’s Step 1. Turning to Step 2, we consider whether Fattah’s efforts to secure Vederman an ambassadorship qualify as making a “decision” or taking “an action” on the identified “matter” of appointment. McDonnell, 136 S. Ct. at 2368. Although those efforts—three emails, two letters, and one phone call—do not themselves qualify as a “question, matter, cause, suit, proceeding, or controversy” under McDonnell’s Step 1, they may nonetheless qualify as the making of a “decision” or taking “an action” on the identified matter of appointment. Id. 80 McDonnell’s Step 2 requires us to determine whether Fattah’s efforts qualify as permissible attempts to “express[] support,” or impermissible attempts “to pressure or advise another official on a pending matter.” Id. at 2371. At trial, the jury was not instructed that they had to place Fattah’s efforts on one side or the other of this divide. The jury might even have thought they were permitted to find Fattah’s efforts—three emails, two letters, and one phone call—to themselves be official acts, rather than a “decision” or “action” on the properly identified matter of appointment. Such a determination would have been contrary to the dictates of McDonnell. Faced with such uncertainty, we cannot assume the jury verdict was proper. Although the jury might not have concluded that Fattah’s efforts were themselves official acts, and although the jury might not have concluded that those efforts crossed the line into impermissible attempts “to pressure or advise,” we are unable to conclude that the jury necessarily did so. Nor can we, on the cold record before us, determine whether Fattah’s efforts to secure Vederman an ambassadorship crossed the line. Determining, for example, just how forceful a strongly worded letter of recommendation must be before it becomes impermissible “pressure or advice” is a fact- intensive inquiry that falls within the domain of a properly instructed jury. Should the Government elect to retry these counts after remand, the finder of fact will need to decide whether Fattah’s efforts constituted permissible attempts to “express[] support,” or impermissible attempts “to 81 pressure or advise another official on a pending matter.” Id. D. The Zionts Hiring The third group of acts charged in the Fattah– Vederman scheme involves Fattah’s decision to hire Vederman’s girlfriend, Alexandra Zionts, as a congressional staffer. We conclude that the hiring was an official act. A brief analysis of McDonnell’s two steps suffices to show why this is so. Here, under McDonnell’s Step 1(A), the relevant “matter” is the decision to hire Zionts. Step 1(B) of the analysis is satisfied because the hiring decision was “pending” before Fattah himself. And that hiring was “focused and concrete,” “within the specific duties of an official’s position—the function conferred by the authority of his office.” Id. at 2369. Finally, McDonnell’s Step 2 requires that the “Government . . . prove that the public official made a decision or took an action ‘on’ [the identified] question, matter, cause, suit, proceeding, or controversy, or agreed to do so.” Id. at 2368. Fattah’s decision to hire Zionts clearly satisfies that requirement. We therefore conclude that the hiring of Zionts was an official act under McDonnell. Vederman concedes that the Zionts hiring was an official act. Oral Argument Transcript at 5–6. Fattah, for his part, maintains that “hiring someone for a routine, part- time, short-term position falls well outside [the] definition 82 [of ‘official act’] and is nothing like a lawsuit, agency determination, or committee hearing, even if each shares the happenstance that federal funds will be used.” Fattah Reply Br. 25. Fattah’s argument lacks traction. Official acts need not be momentous decisions—or even notable ones. Judges, for example, make “routine” evidentiary rulings every day, and yet it is beyond question that those rulings are official acts. In the realm of official acts, it is of no moment that Zionts provided only “part-time, short term” labor. When a public official hires an employee to work in his government office, he has engaged in an official act. * * * If we could conclude that the Zionts hiring was the only category of actions that the jury relied on when it found that Fattah performed an official act under Counts 16–18 and 22–23, remand would not be necessary. But, as we have explained, we cannot rule out that the jury erroneously convicted Fattah and Vederman based on other actions that were not official acts under McDonnell.13 13 More specifically, the incomplete, and therefore erroneous, instructions could have led the jury to commit at least one of three mistakes. First, the jury could have improperly convicted Vederman and Fattah based on the Kirk meeting alone, or misunderstood the Kirk meeting to 83 The Government argues that because the Zionts hiring was an official act, the effect of the erroneous jury instructions could be no more than harmless. The jury’s verdict, the Government contends, permits us to deduce that the jury necessarily concluded the Zionts hiring was an official act, and that this conclusion alone supported Fattah’s and Vederman’s convictions as to Counts 16–18 and 22–23—regardless of whether the jury erroneously found any unofficial acts to be official acts. We disagree. Fattah and Vederman objected to the definition of “official act” at trial. We thus apply the harmless error standard of review. McDonnell, 136 S. Ct. at 2375. The Government argues that because the jury convicted Fattah and Vederman of illegally laundering the proceeds of a “scheme to commit bribery” under Count 23, the jury found that the scheme must have encompassed only the Zionts hiring. JA531.That would mean that the jury did not conclude that the “scheme to commit bribery” included be a necessary component of an impermissible “pattern” of official acts. Second, the jury might have concluded that Fattah’s efforts to secure Vederman an ambassadorship were themselves official acts. Third, the jury might have concluded that Fattah’s efforts to secure Vederman an ambassadorship were merely attempts to “express[] support,” rather than to “exert pressure . . . or provide advice,” but nonetheless erroneously concluded that those expressions of support were official acts. McDonnell, 136 S. Ct. at 2371. 84 any acts that McDonnell now makes clear were unofficial. Yet the redacted indictment, jury instructions, and the fact that the Government presented its case under a “pattern” theory at trial compel us to reject the Government’s argument. The very first sentence under Count 23 of the redacted indictment incorporates all three categories of “Overt Acts” contained within paragraphs “58 through 95 of Count One.” 14 All three of these categories fall under a general heading within the redacted indictment titled “The Bribery and Fraud Scheme [redacted].” JA494. The jury had before it instructions for Count 23 which referred to “the alleged bribery scheme involving an $18,000 payment,” JA448 (emphasis added), and the redacted indictment which referred to “a scheme to commit bribery,” JA531 (emphasis added). The parallel language could well lead a rational jury to conclude that the relevant “scheme” included all three categories of acts listed under the general heading: “The Bribery and Fraud Scheme [redacted].” JA494 (emphasis added). 14 JA531. Paragraphs 58 through 95 of Count 1 refer to the three categories of allegedly official acts discussed above: (1) “The Pursuit of an Ambassadorship,” (2) “The Pursuit of Another Executive Branch Position,” and (3) the “Hiring of the Lobbyist’s Girlfriend to the Congressional Staff.” JA494–95. 85 Like the redacted indictment and jury instructions, the Government’s trial arguments referred to patterns and a course of conduct, and stressed that the jury need not connect specific payments to particular official acts. In its closing argument to the jury, the Government stated that the alleged “scheme took place over a period of several years. Over and over again you’re going to see the same pattern.” JA5383 (emphasis added). Then, in its rebuttal argument, the Government went out of its way to explicitly distinguish its “pattern” theory from an alternative theory that would have directly connected individual payments to individual acts. As the prosecutor argued to the jury: Ms. Recker appears to argue that each thing of value must coincide with some specific official act, but that is not the law and that is not what Judge Bartle is going to instruct you. Instead what he will tell you is that the government is not required to prove that Vederman intended to influence Fattah to perform a set number of official acts in return for things of value so long as the evidence shows a course of conduct of giving things of value, things of value to Fattah in exchange for a pattern of official acts favorable to Vederman. In other words a stream of benefits. These for those, not this for that. 86 JA5715–16 (emphases added). In closing to the jury, the Government made several other references to this “pattern” theory, 15 and the District Court referred to this “pattern” theory in its instructions to the jury. As Judge Bartle instructed: [I]t is not necessary for the government to prove that a defendant intended to induce a public official to perform a number of official acts in return for things of value. So as long as the evidence shows a course of conduct of giving things of value to a public official in exchange for a pattern of official acts favorable to the giver. JA5833–34 (emphasis added). On appeal, the Government changes course, asking us to assume that the jury ignored these repeated references to a “pattern of official acts” and 15 See, e.g., JA5389 (“And the exchange of an official act for a thing of value is called a bribe. There’s the pattern. Fattah needs money, Vederman gets an official act.”); JA 5393 (“That’s why you see the pattern over and over again. Fattah needs money, Vederman gets an official act.”); JA5400 (“The same pattern we saw over and over again. Fattah needs money, Vederman gets an official act.”); JA5409 (“[Y]ou know that these were bribes because of the pattern you saw over and over and over again. Fattah needs money, Vederman gets an official act, that makes these things a bribe.”). 87 instead considered the Zionts hiring and Vederman’s $18,000 payment to Fattah as an isolated quid pro quo. This is an invitation to speculate, and we decline to do so.16 The jury began its deliberations accompanied by a copy of the redacted indictment which alleged a pattern of official acts, consisting of any combination of three categories of acts: pursuing an ambassadorship, arranging the Kirk meeting, and hiring Zionts. In light of the erroneous instructions, and because only one category clearly 16 Providing some support to the Government’s ultimately unconvincing argument that the jury considered the Zionts hiring and $18,000 payment in isolation, we note that the redacted indictment does mention those two events side- by-side in paragraph 78 of the indictment’s Part V. JA497 (“On January 13, 2012, VEDERMAN wired $18,000 to FATTAH, and six days later, on January 19, 2012, BOWSER emailed VEDERMAN’s girlfriend, A.Z., welcoming her as a new employee to FATTAH’s Congressional Staff.”). But although paragraph 78 mentions the $18,000 wire transfer and the Zionts hiring in the same breath, paragraph 78 does not instruct the jury to connect these two events apart from the rest of the evidence presented at trial. In light of the other instructions and arguments indicating that the jury should not consider the Zionts hiring in isolation, but instead should consider the hiring as one part of a three-part scheme, paragraph 78 is not sufficient to avoid a reversal and remand on the convictions of Fattah and Vederman as to Counts 16–18 and 22–23. 88 qualifies as an “official act,” the jury’s deliberations were fraught with the potential for McDonnell error. We will vacate the convictions of Fattah and Vederman as to Counts 16, 17, 18, 22, and 23, and remand to the District Court. E. Vederman’s Sufficiency Challenge to Counts 16–18 and 22–23 Vederman argues that there is insufficient evidence to support a conviction, even if a jury were properly instructed under McDonnell. Specifically, Vederman argues that there is insufficient evidence to convict him and Fattah, after remand, on Counts 16–18 and 22–23 because “[a]t least seven of the eight alleged ‘official acts’ were, as a matter of law, not official at all.” Vederman Br. 35. As to the single act that Vederman implicitly concedes to be an official act—the Zionts hiring—Vederman argues that “[t]he only thing that even arguably associates” the Zionts hiring with Vederman was its timing in relation to Vederman’s sham purchase of the Fattahs’ Porsche. Id. According to Vederman, “the undisputed chronology precludes any inference that Vederman conferred this benefit on his friend as an illegal bribe.” Id. (emphasis omitted). Vederman is wrong. Sufficient evidence was produced at trial to have allowed a properly-instructed jury to convict Fattah and Vederman of Counts 16–18 and 22– 23. To begin with, even if the Zionts hiring had been the sole official act to survive this Court’s interpretation of 89 McDonnell, there would still be sufficient evidence to convict Fattah and Vederman. Zionts did not receive written notice of her official hiring until six days after the sham Porsche purchase. Moreover, the jury would not be restricted to considering the chronology of the sham purchase alone. It would be free to consider Vederman’s entire course of conduct. Under the general heading “VEDERMAN’S Payments and Things of Value to FATTAH,” the redacted indictment not only refers to the $18,000 wire transaction from Vederman to Fattah as part of the sham Porsche purchase, but also to Vederman’s $3,000 payment for the college tuition of Simone Muller, Fattah’s live-in au pair, as well as thousands of dollars in payments made by Vederman for Chip Fattah’s college tuition. JA496–97. And the Zionts hiring is not the only act to survive our application of McDonnell. As we explained, a jury could find that Fattah’s efforts to secure Vederman an ambassadorship—three emails, two letters, and a phone call—were an impermissible attempt to “pressure or advise” President Obama, Senator Casey, or both men.17 17 Although Fattah’s efforts to secure Vederman an ambassadorship present a jury question that is not for us to answer on appeal, we note that not one of these efforts alone could qualify as an official act itself. See McDonnell, 136 S. Ct. at 2372 (“Setting up a meeting, talking to another official, or organizing an event (or agreeing to do so)—without more—does not fit that definition of ‘official 90 This means that a properly instructed jury on remand, presented with evidence of Fattah’s efforts to secure an ambassadorship for Vederman and evidence of the Zionts hiring, could find more than a single official act. F. Blue Guardians In addition to the charges arising from his dealings with Vederman, Fattah was charged in Count 3 with act.’ ”). The relevant question for a jury to consider on remand, then, is whether these actions constituted “a ‘decision or action’ on a different question or matter”—to wit, the formal appointment of an ambassador. Id. at 2369 (emphasis omitted). Even though the emails, letters, and phone call are not, individually, official acts, it will be for a jury to decide if Fattah’s efforts to secure an ambassadorship for Vederman crossed the line from permissible “support” to impermissible “pressure or advice.” While we express doubt that some of Fattah’s efforts concerning the ambassadorship are, when considered in isolation, enough to cross that line, a properly instructed jury considering all of the facts in context might nonetheless conclude that other efforts—such as a hand-delivered letter to the President of the United States—indeed crossed that line. Further, a jury might find that in the aggregate, three emails, two letters, and a phone call crossed the line and therefore constituted a “decision or action” on the identified matter of appointment. 91 participating in a scheme with Lindenfeld to funnel money to a fraudulent nonprofit organization. In connection with this scheme, Fattah was convicted of conspiring to commit honest services fraud. Fattah owed Lindenfeld nearly $100,000 for work performed on Fattah’s 2007 mayoral campaign. In lieu of repayment, Fattah suggested that Lindenfeld create an entity, later named Blue Guardians, to which Fattah would direct $15,000,000 in public funds by using his position as a member of the House Committee on Appropriations. Nothing in McDonnell requires us to upset Fattah’s conviction on Count 3. Step 1(A) of our McDonnell analysis requires the Government to “identify a ‘question, matter, cause, suit, proceeding or controversy.’ ” 136 S. Ct. at 2368. Here, the “matter” is the appropriation of millions of dollars in public funds. See Repak, 852 F.3d at 253–54 (holding the awarding of redevelopment funds to be an official act). In particular, it was Fattah’s promise to perform this official act that was unlawful. As McDonnell makes clear: [A] public official is not required to actually make a decision or take an action on a “question, matter, cause, suit, proceeding or controversy”; it is enough that the official agree to do so. The agreement need not be explicit, and the public official need not specify the means that he will use to perform his end of the bargain. 92 136 S. Ct. at 2370–71 (internal citations omitted). That Fattah took steps to actually carry out his promise (e.g., by drafting and sending a formal appropriations request on official congressional letterhead) is evidence of his illegal promise. See id. at 2371. Step 1(B) requires the Government to establish that the “ ‘question, matter, cause, suit, proceeding or controversy’ . . . ‘may at any time be pending’ or ‘may by law be brought’ before a public official.” McDonnell, 136 S. Ct. at 2368. Appropriating public funds was not only a matter that was pending before Fattah as a member of the Appropriations Committee, it was also a matter that was pending before the Chairman and Ranking Member of an Appropriations Subcommittee to whom Fattah ultimately sent a formal written request. See id. at 2369 (“[T]he matter may be pending either before the public official who is performing the official act, or before another public official.”). Appropriating millions of dollars in response to the Blue Guardians request is “focused and concrete,” and “the kind of thing that can be put on an agenda, tracked for progress, and then checked off as complete.” Id. Given Fattah’s membership on the Appropriations Committee, this was “something within the specific duties of an official’s position—the function conferred by the authority of his office.” Id. Even if we were to assume, against all reason, that an appropriation is not “something within the specific duties” of either Fattah or the Chairman or Ranking Member of an Appropriations Subcommittee, 93 Fattah’s formal request for an appropriation was something that Fattah had the authority to do. Like the Executive Director in Repak, who lacked authority himself to award redevelopment funds but could request such funds from the Board, Fattah used his position as a Congressman to formally request appropriations for the Blue Guardians. 852 F.3d at 254 (“Repak had the power to, and indeed did, make recommendations to the [redevelopment authority. Step 2 of McDonnell requires the Government to “prove that the public official made a decision or took an action ‘on’ that question, matter, cause, suit, proceeding, or controversy, or agreed to do so.” 136 S. Ct. at 2368 (emphasis added). Here, Fattah agreed to request an appropriation for a bogus purpose. Unlike Fattah’s letters, emails, and phone call seeking an ambassadorship for Vederman, there is no potential for the jury to have made a mistake when it found Fattah’s Blue Guardians promise unlawful. Fattah argues that the Government presented “[n]o evidence . . . that would have allowed [the jury] to conclude that [he] made a decision or took an action, or could have done so, on the question whether Blue Guardians would receive a $15 million federal grant.” Fattah Br. 46. This argument misses the point. It was Fattah’s agreement to engage in the official act of formally requesting the appropriation that was illegal. See McDonnell, 136 S. Ct. at 2371. 94 Lindenfeld’s trial testimony provided sufficient evidence of Fattah’s illegal agreement. JA1694–96, 1954. Fattah’s letter provided additional evidence from which the jury could have concluded that Fattah illegally agreed to perform an official act. 18 In short, the agreement itself was illegal, and the Government provided sufficient evidence for the jury to conclude that the illegal agreement took place. The Government’s evidence in support of the Blue Guardians scheme meets the requirements of McDonnell, and the Count 3 verdict will stand. V. Sufficiency of the Evidence for the RICO Conspiracy Conviction The jury found Fattah, Vederman, Brand, and Nicholas guilty of the RICO conspiracy charged in Count 1 of the indictment, but acquitted Bowser. Vederman filed a post-verdict motion, and the District Court overturned his RICO conspiracy conviction. 18 Despite Fattah’s protestation to the contrary, there is evidence that Fattah took steps to carry out his official act. JA6432–33 (Letter from Congressman Fattah to House Appropriations Subcommittee members “request[ing] funding and support for the following projects and programs of critical importance,” including $3,000,000 for “Blue Guardians”). 95 On appeal, Fattah, Brand, and Nicholas challenge the sufficiency of the evidence supporting their RICO conspiracy convictions. We “review[] the sufficiency of the evidence in the light most favorable to the government and must credit all available inferences in favor of the government.” United States v. Riddick, 156 F.3d 505, 509 (3d Cir. 1998). If a rational juror could have found the elements of the crime beyond a reasonable doubt, we must sustain the verdict. United States v. Cartwright, 359 F.3d 281, 286 (3d Cir. 2004), abrogated on other grounds by United States v. Caraballo-Rodriguez, 726 F.3d 418 (3d Cir. 2013) (en banc). The indictment charged a RICO conspiracy in violation of 18 U.S.C. § 1962(d), which makes it “unlawful for any person to conspire to violate” § 1962(c). Section 1962(c) provides: It shall be unlawful for any person . . . associated with any enterprise engaged in, or the activities of which affect, interstate . . . commerce, to conduct or participate, directly or indirectly, in the conduct of such enterprise’s affairs through a pattern of racketeering activity . . . . 18 U.S.C. § 1962(c). In Salinas v. United States, 522 U.S. 52 (1997), the defendant was convicted of a § 1962(d) RICO conspiracy, but a jury acquitted him of the substantive RICO offense 96 under § 1962(c). Id. at 55. The Supreme Court rejected Salinas’s contention that his conviction had to be set aside because he had neither committed nor agreed to commit the two predicate acts required for the § 1962(c) offense. Id. at 66. The Court declared that liability for a RICO conspiracy under § 1962(d), “unlike the general conspiracy provision applicable to federal crimes,” does not require proof of an overt act. Id. at 63. A conspiracy may be found, the Court explained, “even if a conspirator does not agree to commit or facilitate each and every part of the substantive offense. The partners in the criminal plan must agree to pursue the same criminal objective and may divide up the work, yet each is responsible for the acts of each other.” Id. at 63–64 (citations omitted). This means that, if a plan “calls for some conspirators to perpetrate the crime and others to provide support, the supporters are as guilty as the perpetrators.” Id. at 64. Thus, opting into and participating in a conspiracy may result in criminal liability for the acts of one’s co-conspirators. Smith v. Berg, 247 F.3d 532, 537 (3d Cir. 2001). Accordingly, liability for a RICO conspiracy may be found where the conspirator intended to “further an endeavor which, if completed, would satisfy all of the elements of a substantive criminal offense, but it suffices that he adopt the goal of furthering or facilitating the criminal endeavor.” Salinas, 522 U.S. at 65. Because the substantive criminal offense here was conducting a § 1962(c) enterprise, the government had to prove: 97 (1) that two or more persons agreed to conduct or to participate, directly or indirectly, in the conduct of an enterprise’s affairs through a pattern of racketeering activity; (2) that the defendant was a party to or member of that agreement; and (3) that the defendant joined the agreement or conspiracy knowing of its objective to conduct or participate, directly or indirectly, in the conduct of an enterprise’s affairs through a pattern of racketeering activity. United States v. John-Baptiste, 747 F.3d 186, 207 (3d Cir. 2014). In United States v. Turkette, 452 U.S. 576 (1981), the Supreme Court instructed that an enterprise is a “group of persons associated together for a common purpose of engaging in a course of conduct.” Id. at 583. The government can prove an enterprise “by evidence of an ongoing organization, formal or informal, and by evidence that the various associates function as a continuing unit.” Id. In Boyle v. United States, 556 U.S. 938 (2009), the Supreme Court established that an “association-in-fact enterprise must have at least three structural features: a purpose, relationships among those associated with the enterprise, and longevity sufficient to permit these associates to pursue the enterprise’s purpose.” Id. at 946. The structure necessary for a § 1962(c) enterprise is not complex. Boyle explained that an enterprise 98 need not have a hierarchical structure or a “chain of command”; decisions may be made on an ad hoc basis and by any number of methods—by majority vote, consensus, a show of strength, etc. Members of the group need not have fixed roles; different members may perform different roles at different times. The group need not have a name, regular meetings, dues, [or] established rules and regulations . . . .While the group must function as a continuing unit and remain in existence long enough to pursue a course of conduct, nothing in RICO exempts an enterprise whose associates engage in spurts of activity punctuated by periods of quiescence. Id. at 948. Another element of a substantive § 1962(c) RICO enterprise is that the enterprise must conduct its affairs through a pattern of racketeering activity. Section 1961 defines racketeering activity to include various criminal offenses, including wire fraud, 18 U.S.C. § 1344, and obstruction of justice, 18 U.S.C. § 1511. See 18 U.S.C. § 1961(1). A pattern of such activity “requires at least two acts of racketeering activity.” Id. § 1961(5). The racketeering predicates may establish a pattern if they “related and . . . amounted to, or threatened the likelihood 99 of, continued criminal activity.” H.J. Inc. v. Nw. Bell Tel. Co., 492 U.S. 229, 237 (1989). Here, the District Court denied the post-trial sufficiency arguments raised by Fattah, Brand, and Nicholas. It reasoned: For a RICO conspiracy to exist, the conspirators must agree to participate in an enterprise with a unity of purpose as well as relationships among those involved. The evidence demonstrates that an agreement among Fattah, Brand, Nicholas, Lindenfeld, and Naylor existed for the overall purpose of maintaining and enhancing Fattah as a political figure and of preventing his standing from being weakened by the failure to be able to pay or write down his campaign debts. These five persons agreed to work together as a continuing unit, albeit with different roles. The Government established that Fattah, Brand, and Nicholas conspired along with Naylor and Lindenfeld to conceal and repay the 2007 illegal $1,000,000 loan to the Fattah for Mayor campaign. JA128–29. The District Court further determined that [w]hile each member may not have been involved in every aspect of the enterprise, its activities were sufficiently structured and 100 coordinated to achieve the purpose of maintaining and enhancing Fattah’s political standing and of preventing him from being weakened politically because of his campaign debts. A RICO conspiracy also requires an agreement to participate in an enterprise with longevity sufficient to pursue its purpose. This was established. In May 2007 the illegal loan was obtained and continued through its repayment in January 2008 and into at least 2014 when the last campaign report reducing a fake campaign debt to Naylor’s consulting firm was filed by Fattah. JA131. The defendants argue that the evidence is insufficient to show either an enterprise for purposes of § 1962(c) or an agreement as required for a § 1962(d) conspiracy. We disagree, and conclude that the District Court’s analysis is on the mark. We begin by considering whether there was an agreement. The evidence showed that Fattah knew each member involved in the scheme to conceal the unlawful campaign loan. When Lindenfeld learned of the $1 million loan, he informed Fattah that it exceeded campaign finance limits. In short, the transaction was unlawful, and the two knew it. The transaction nonetheless went forward, disguised as a loan, with Lindenfeld executing 101 the promissory note as Strategies’ officer and obligating Strategies to repay Lord $1 million. The concealment efforts continued as Lindenfeld funneled a substantial portion of the loan proceeds to Naylor for get-out-the-vote efforts. After the losing campaign, Lindenfeld spoke with Fattah and Naylor about accounting for the funds that had been spent. They decided not to include the amounts in the FFM campaign reports. Fattah instructed Naylor to prepare a fictitious invoice, and Naylor complied. The FFM campaign reports filed from 2008 to 2014 disclosed nothing about the unlawful $1 million loan. Instead, they falsely showed that Naylor’s consulting firm made yearly in-kind contributions of $20,000 in debt forgiveness, when in reality there was no debt to forgive. As Lindenfeld fretted over repaying the $600,000 balance of the Lord loan, Naylor assured him that Fattah had promised to take care of the repayment. And the evidence supports an inference that Fattah recruited both Nicholas and Brand in doing so. As EAA’s director, Nicholas could fund the repayment. Brand, through his company, Solutions, acted as the middleman: he received the payment from EAA pursuant to a fictitious contract, and then forwarded the balance due to Strategies pursuant to yet another fictitious contract. Nicholas and Brand continued in the spring and summer of 2008 to hide the fictitious agreement and the $600,000 payment to Lindenfeld to satisfy the Lord loan. 102 In short, this evidence shows that Fattah, Lindenfeld, Naylor, Brand, and Nicholas all agreed to participate in Fattah’s plan to conceal the unlawful campaign loan to maintain his political stature. Nicholas and Brand claim that they had no knowledge of the false campaign reporting aspect of the plan. But as Salinas instructs, conspirators need not “agree to commit or facilitate each and every part of the” conspiracy. 522 U.S. at 63. Rather, they “must agree to pursue the same criminal objective and may divide up the work, yet each [be] responsible for the acts of each other.” Id. at 63–64. Thus, a conspirator may agree to “facilitate only some of the acts leading to the substantive offense” yet still be criminally liable. Id. at 65. The evidence showed that a substantial amount of money was needed to repay Lord, and that the source of the repayment was EAA, a non-profit organization whose funds could be spent only for purposes consistent with the terms of the grants it received. It also showed that Nicholas was presented with a sham contract to legitimize the EAA–Solutions transaction. We conclude that the evidence is sufficient to support an inference that Nicholas knew at the start that the plan was unlawful. Yet she still agreed to provide the requisite funds and to play a role in concealing the illegal campaign loan so that Fattah could maintain his political stature. As to Brand, even if he did not know that false campaign reports were being filed, the evidence is 103 sufficient to show he played a key role in the enterprise. From the outset, Brand worked to disguise the repayment of the Lord loan as the consideration in a sham contract between EAA and Solutions. He then arranged for the transfer of funds to Strategies in satisfaction of a contractual term in another purported business agreement between Solutions and Strategies. The evidence reveals that Brand was the point man in the effort to meet the January 31, 2008 deadline to repay the Lord loan, and it amply shows that Brand also agreed to participate in the plan to hide the illegal campaign loan and its repayment to benefit Fattah politically. Fattah, Brand, and Nicholas attack their RICO conspiracy convictions on another front. They argue that those verdicts should be set aside because the evidence fails to show that the various schemes alleged in the indictment as part of the RICO conspiracy are connected. The RICO count, they assert, charges a hub-and-spoke conspiracy that is unconnected by a rim. In their view, Fattah is the hub, and the spokes consist of a series of independent schemes: the Vederman bribery scheme, the payment of the outstanding tuition debt of Fattah’s son Chip, the Blue Guardians plan, and the repayment of the illegal Lord loan to maintain Fattah’s political stature. They argue that, without a unifying rim, their actions cannot constitute an enterprise. Again, we disagree. In In re Insurance Brokerage Antitrust Litigation, 618 F.3d 300 (3d Cir. 2010), we concluded, in analyzing 104 one of plaintiffs’ RICO claims, that the alleged hub-and- spoke enterprise—comprised of broker hubs and insurer spokes—could not withstand a motion to dismiss because it did not have a unifying rim. Id. at 374. We explained that the allegations did “not plausibly imply concerted action—as opposed to merely parallel conduct—by the insurers, and therefore cannot provide a ‘rim’ enclosing the ‘spokes’ of these alleged ‘hub-and-spoke’ enterprises.” Id. Thus, the allegations did not “adequately plead an association-in-fact enterprise” because the hub- and-spoke conspiracy failed to “function as a unit.” Id. That is not the case here. The evidence showed that Fattah, Brand, and Nicholas agreed to conceal the illegal Lord loan. Each acted for the common purpose of furthering Fattah’s political interests. In short, they engaged in concerted activity and functioned as a unit. The jury convicted Fattah, Brand, and Nicholas of the RICO conspiracy based on the racketeering activity of wire fraud and obstruction of justice to conceal the unlawful transaction. Because the evidence shows that Fattah, Lindenfeld, Naylor, Brand, and Nicholas agreed to protect Fattah’s political status by acting to maintain the secrecy of the unlawful Lord loan, the alleged lack of a unifying “rim” is not fatal to this RICO enterprise. What matters in analyzing the structure of this enterprise is that it functioned as a unit. Boyle, 556 U.S. at 945; In re Ins. Brokerage Antitrust Litig., 618 F.3d at 374. That “basic requirement” was met. Id. 105 We turn next to the contention that the evidence fails to establish other components of an enterprise. We conclude that much of the evidence supporting the existence of an agreement also shows that there was an association-in-fact enterprise. Boyle made clear that an association-in-fact enterprise must have “a purpose, relationships among those associated with the enterprise, and longevity sufficient to permit these associates to pursue the enterprise’s purpose.” 556 U.S. at 946. The purpose, as we have repeatedly observed, was to maintain and preserve Fattah’s political stature by concealing the illegal loan and its repayment. Though informal, there were relationships among those associated with the enterprise. Fattah was at the center of this association and he directed its activity. He knew each of the association’s members, and the members knew each other (except, perhaps, for Nicholas, who may not have known Lindenfeld).19 19 Nicholas’s lack of familiarity with Lindenfeld does not undermine her membership in this association-in-fact enterprise. We have previously explained that “[i]t is well- established that one conspirator need not know the identities of all his co-conspirators, nor be aware of all the details of the conspiracy in order to be found to have agreed to participate in it.” United States v. Riccobene, 709 F.2d 214, 225 (3d Cir. 1983), abrogated on other grounds by Griffin v. United States, 502 U.S. 46 (1991). 106 The Government also adduced sufficient proof of the longevity component required for an enterprise. The scheme began in mid-2007, when Lord made the campaign loan, directing the proceeds of the loan to Strategies. From the outset, Fattah, Lindenfeld, and Naylor all knew they needed to conceal this illegal transaction. They began by fabricating an explanation for the source of the funds they spent on election day. SLA created a fake invoice for the campaign, showing a fictitious debt that Naylor could later forgive by fictitious in-kind contributions existing only on Fattah’s campaign finance reports. The effort to disguise the Lord loan was not limited to filing false campaign reports. Nicholas and Brand, who joined the conspiracy a few months later than the other members, understood that they too had to make the fraudulent $600,000 payment by EAA to Solutions appear legitimate. Nicholas and Brand tried to disguise the sham contract as an ordinary transaction (even though it called for a six-figure upfront payment simply to support Solutions’ various projects), and they succeeded in keeping it out of the DOJ auditors’ view until August 2008. The ruse continued as Solutions funneled the $600,000 payment to Strategies under the guise of another sham contract (which also required an upfront six-figure payment). The scheme then continued as Fattah submitted false FFM campaign reports from 2008 through 2014. 107 Finally, we consider whether the enterprise conducted its affairs through a pattern of racketeering activity, as required for a § 1962(c) enterprise. Wire fraud and obstruction of justice may constitute “racketeering activity” under § 1961(1). As the Supreme Court instructed in H.J. Inc., the “multiple predicates within a single scheme” must be related and “amount[] to, or threaten[] the likelihood of, continued criminal activity.” 492 U.S. at 237. Here, the amount of the illegal loan to be concealed was substantial. The enterprise needed to write off the fictitious debt to Naylor’s consulting firm, and it was urgent that both the EAA–Solutions contract and the Solutions–Strategies contract be legitimized. We conclude the evidence was sufficient to establish that this enterprise conducted its affairs through a pattern of racketeering activity and that the predicate acts of wire fraud and obstruction of justice were related. The racketeering activity furthered the goals of maintaining the secrecy of this $1 million illicit campaign loan and of preserving Fattah’s political stature. Nicholas contends that the evidence fails to establish a pattern of racketeering activity because the actions to which she agreed did not “extend[] over a substantial period of time” as H.J. Inc. requires. 492 U.S. at 242. That case indeed instructs that the continuity requirement of a pattern is a “temporal concept,” and that “[p]redicate acts extending over a few weeks or months” do not satisfy the continuity concept. Id. But the Supreme Court explained that continuity may also be established by 108 showing that there is a “threat of continued racketeering activity.” Id. Here, the course of fraudulent conduct undertaken to secure and to conceal the $1 million Lord loan consisted of the creation of sham debts, fictitious contracts, and false accounting entries over the course of about a year. But because Fattah needed to appear able to retire his campaign debt, the enterprise needed to continue filing false campaign reports for several years, allowing the annual $20,000 in-kind debt forgiveness contributions to appear to satisfy Naylor’s fake $193,000 invoice. That evidence was sufficient to establish the requisite threat of continued criminal activity. See H.J. Inc., 492 U.S. at 242– 43. We conclude that the Government met its burden in proving that Fattah, Brand, and Nicholas 20 engaged in a RICO conspiracy in violation of § 1962(d). 20 Nicholas also asserts, in passing, that that her conviction under § 1962(d) should be set aside because that statutory provision is unconstitutionally vague as applied to her. According to Nicholas, a person of ordinary intelligence would not know that her actions constituted an agreement to participate in a RICO enterprise. See United States v. Pungitore, 910 F.2d 1084, 1104–05 (3d Cir. 1990). To the contrary, a person of ordinary intelligence, who had been employed by a prominent politician and then became the CEO of a nonprofit organization which that politician had founded (and, to some extent, continued to direct), would 109 VI. Variance from the Indictment and Sufficiency of the Evidence for Count 2 Brand and Nicholas challenge their convictions for conspiracy to commit wire fraud by arguing that the Government’s evidence at trial impermissibly varied from the indictment. Nicholas also challenges the sufficiency of the evidence to support her conviction for conspiracy to commit wire fraud. We address these contentions together. 21 Count 2 of the indictment alleged a single conspiracy. JA277–79. Brand and Nicholas assert that the Government’s evidence at trial did not support the existence of a single conspiracy but instead showed two independent conspiracies, only one of which involved the two of them. According to Brand and Nicholas, the only conspiracy with which they were involved ended more than five years before the Government charged them. That would mean that all their conduct falls outside the five- year limitations period for wire fraud conspiracy under 18 U.S.C. § 3282. realize that agreeing to participate with others in hiding an unlawful campaign loan of $1 million could constitute an unlawful RICO conspiracy. 21 In her briefing, Nicholas discusses variance in far less detail than Brand, so we refer primarily to Brand’s arguments. See Nicholas Br. 54–56. Her variance arguments fail for the same reasons that Brand’s fail. 110 “A conviction must be vacated when (1) there is a variance between the indictment and the proof presented at trial and (2) the variance prejudices a substantial right of the defendant.” Kemp, 500 F.3d at 287 (quoting United States v. Kelly, 892 F.2d 255, 258 (3d Cir. 1989)). We see no variance, and will affirm the District Court. A variance exists “if the indictment charges a single conspiracy while the evidence presented at trial proves only the existence of multiple conspiracies.” Id. “We must determine ‘whether there was sufficient evidence from which the jury could have concluded that the government proved the single conspiracy alleged in the indictment.’ ” Id. (quoting Kelly, 892 F.2d at 258). Viewing the record in the light most favorable to the Government, we consider three factors: (1) “whether there was a common goal among the conspirators”; (2) “whether the agreement contemplated bringing to pass a continuous result that will not continue without the continuous cooperation of the conspirators”; and (3) “the extent to which the participants overlap in the various dealings.” Id. (quoting Kelly, 892 F.2d at 259). Brand argues that the Government failed to establish a common goal among the conspirators. To determine whether the conspirators shared a common goal, “we look to the underlying purpose of the alleged criminal activity” in a fairly broad sense. United States v. Rigas, 605 F.3d 194, 214 (3d Cir. 2010) (en banc). In Rigas, we described the common goal of the defendants as 111 “enriching [themselves] through the plunder of [their corporate employer],” id., and we have similarly articulated the common goal in fairly general terms elsewhere. See United States v. Greenidge, 495 F.3d 85, 93 (3d Cir. 2007) (“There was certainly evidence of a common goal among these co-conspirators: to make money by depositing stolen and altered corporate checks into business accounts.”); Kelly, 892 F.2d at 259 (“[T]he common goal of all the participants was simply to make money selling ‘speed.’ ”). Importantly, a common goal may exist even when “conspirators individually or in groups perform different tasks in pursuing the common goal,” and a single conspiracy may “attract[] different members at different times” or “involve[] different sub- groups committing acts in furtherance of the overall plan.” United States v. Boyd, 595 F.2d 120, 123 (3d Cir. 1978). Here, the indictment described the purpose of the unified conspiracy in Count 2 at length: It was a purpose of the conspiracy to obtain an illegal campaign loan and to fraudulently repay that loan with hundreds of thousands of dollars of misappropriated charitable funds from Sallie Mae and federal grant funds from NASA which were intended for educational purposes. . . . . It was further a purpose of the conspiracy to present FATTAH to the public as a perennially viable candidate for public 112 office who honored his obligations to his creditors and was able to retire his publicly reported campaign debts. . . . . It was further a purpose of the conspiracy to promote FATTAH’s political and financial goals through deception by concealing and protecting the conspirators’ activities from detection and prosecution by law enforcement officials and the federal judiciary, as well as from exposure by the news media, through means that included obstruction of justice and the falsification of documents, including Campaign Finance Reports, false invoices, contracts, and other documents and records. JA277–78, ¶¶ 3–5. Brand characterizes the evidence at trial as establishing two distinct conspiracies. The first he labels the “diversion of funds scheme,” covering the misappropriation of funds by Nicholas, Brand, Lindenfeld, and Fattah to repay the Lord loan. Brand Br. 34. Brand calls the second conspiracy the “CFR scheme,” in which Fattah and Naylor filed the false campaign finance reports showing Naylor gradually forgiving a non- existent debt. Id. Brand argues that the only goal of the CFR scheme was to cover up how the funds from the illegal campaign 113 loan were spent, a goal he distinguishes from that of the diversion of funds scheme, which he characterizes as a plan to cover up the repayment of the loan with stolen funds. He also argues that the evidence does not establish he was involved in, or even aware of, the false campaign finance reports filed by Fattah. In Brand’s view, that necessarily means the evidence showed two separate conspiracies. In considering these arguments, we begin by noting that one conspiracy can involve multiple subsidiary schemes. Rigas, 605 F.3d at 214. It is true that the false campaign finance reports, in the narrowest sense, had the specific purpose of covering up how the illegal loan funds were used during the election. But the false campaign finance reports were also filed in furtherance of a broader goal shared by the conspirators involved in repayment of the Lord loan. They sought to promote Fattah’s political and financial goals by preserving his image as a viable candidate and making him appear able to repay or otherwise service his campaign debts without resorting to illegal means in doing so. The two subsidiary schemes worked in concert in furtherance of this overarching goal, and both were directed at covering up how the loan was truly repaid. The “diversion of funds scheme” hid the illegal (but real) loan repayment through the use of fake contracts; the “CFR scheme” showed the seemingly legal (but fake) loan forgiveness installments through the creation of fake invoices and campaign finance reports. The existence of two concealment schemes acting in 114 concert does not undermine the unity of the conspiracy of which they were both a part. We have no difficulty concluding that the false campaign finance reports and the concealed use of stolen funds to repay the Lord loan operated together in furtherance of a common goal. As for Brand’s argument that he was unaware of the false campaign finance reports and therefore could not be a part of any conspiracy involving them, it is well-settled that “each member of the charged conspiracy is liable for the substantive crimes his coconspirators commit in furtherance of the conspiracy even if he neither participates in his co-conspirators’ crimes nor has any knowledge of them.” United States v. Bailey, 840 F.3d 99, 112 (3d Cir. 2016) (citing Pinkerton v. United States, 328 U.S. 640 (1946)). The exceptions to that rule allow a defendant to escape liability for a co-conspirator’s crime if: (1) “the substantive offense committed by one of the conspirators was not in fact done in furtherance of the conspiracy,” (2) “the substantive offense committed by one of the conspirators ‘did not fall within the scope of the unlawful project,’ ” or (3) “the substantive offense committed by one of the conspirators ‘could not be reasonably foreseen as a necessary or natural consequence of the unlawful agreement.’ ” Id. (quoting Pinkerton, 328 U.S. at 647–48). There was, as we have concluded, a unity of purpose between the co-conspirators to further Fattah’s political and financial goals by secretly obtaining and repaying an illegal campaign loan with stolen funds. The filing of false campaign reports does not fit within any of 115 the recognized exceptions to co-conspirator liability, as it was in furtherance of the conspiracy’s shared goal, within the scope of the agreement to conceal the loan, and foreseeable to Brand and Nicholas. Neither Brand nor Nicholas briefed the other two factors we consider when determining whether the evidence impermissibly varied from the evidence, “whether the agreement contemplated bringing to pass a continuous result that will not continue without the continuous cooperation of the conspirators,” and “the extent to which the participants overlap in the various dealings.” Kemp, 500 F.3d at 287 (quoting Kelly, 892 F.2d at 258). The unified goal of promoting Fattah’s political career and maintaining secrecy surrounding the illegal loan and the misappropriated funds used to repay it required the continuous cooperation of the conspirators. Indeed, the efforts of several of them overlapped in every aspect of the scheme. And Lindenfeld and Fattah were, at a minimum, involved in some way in nearly every aspect of the origination of the loan, the false campaign finance reports, and the use of misappropriated funds to repay the loan. For his part, Naylor was involved in the use of the funds, the false campaign finance reports, and to a lesser extent, the repayment of the loan. Brand (as part of his variance argument) and Nicholas (as part of her sufficiency argument) argue that the Government did not prove they agreed to conceal their actions, and thus the false campaign reports would not be 116 sufficient to extend the duration of the conspiracy so that it fell within the statute of limitations. Acts of concealment, such as the false campaign reports, are not automatically “in furtherance” of a conspiracy. We must determine whether there was “an express original agreement among the conspirators to continue to act in concert in order to cover up, for their own self-protection, traces of the crime after its commission,” as opposed to “a conspiracy to conceal . . . being implied from elements which will be present in virtually every conspiracy case, that is, secrecy plus overt acts of concealment.” Grunewald v. United States, 353 U.S. 391, 404 (1957). If the indictment “specifically alleges a continuing conspiracy” to conceal the crime after the completion of the wire fraud, and such a conspiracy can be proven, the statute of limitations does not begin to run until the last overt act of concealment. United States v. Moses, 148 F.3d 277, 282 (3d Cir. 1998). Here, the evidence shows that the conspirators expressly agreed to conceal the loan and its repayment. As an initial matter, Brand’s only role in the conspiracy was to cover up the use of stolen funds by (1) serving as an intermediary between Nicholas and Lindenfeld; and (2) agreeing to create false documentation (the contracts) with both EAA and Strategies for the sole purpose of disguising the payments and covering up the wire fraud conspiracy. Nicholas could simply have paid Lindenfeld herself (or paid Lord) if she and Brand had not agreed to conceal the crime from the start. Additionally, and as 117 Brand acknowledges, the false campaign finance reports began before the loan was repaid, proving that concealment of the crime was contemplated and begun as a direct purpose of the conspiracy before Brand and Nicholas became involved in the repayment. Nicholas too agreed to conceal the repayment, as she implicitly acknowledged in her emails with Brand and Fattah. GSA2. Finally, when Lindenfeld briefly strayed from the conspiracy’s commitment to secrecy by mentioning the repayment in front of others who did not know of the scheme, Brand became “angry,” “took [Lindenfeld] out in the hallway,” and chastised him, saying that “[Lindenfeld] couldn’t say that sort of []thing” in front of other people. JA1670–71. We conclude that the evidence is consistent with the allegations in the indictment, which charge a single conspiracy consisting of an original agreement to conceal the illegal loan and its subsequent illegal repayment to further Fattah’s political career. Nicholas makes several arguments in passing. She suggests that the District Court upheld the conviction after trial “on a theory not submitted to the jury.” Nicholas Br. 51. This argument is, essentially, that the indictment and the District Court’s post-trial ruling described the conspiracy one way, but that the jury charge described the conspiracy differently. Nicholas argues that the jury was presented with the theory that the sole purpose of the false campaign reports under Count 2 was to “conceal[] the alleged scheme to defraud,” JA5849, rather than to support 118 Fattah’s political career, as the District Court described the purpose after trial, see JA74. Nicholas ignores that part of the jury charge which instructed that Count 2 required a finding “[t]hat two or more persons agreed to commit wire fraud as charged in the indictment.” JA5845 (emphasis added). The jury had access to the indictment, and as Nicholas points out, Nicholas Br. 45–46, the indictment outlines the offense in the same way the District Court later described it in its post-trial ruling. The District Court consistently described the count, and we see no reversible error. Nicholas also argues that the conspiracy charged in Count 2 has an objective—“to ‘present Fattah’ as ‘perennially viable’ ”—and that such an objective is not illegal. Nicholas Br. 53. But, of course, the jury was not instructed that it was illegal to be a Fattah supporter, or even to work on his campaign. The jury was charged specifically on the crime of wire fraud. We conclude that there was no impermissible variance between the indictment and the Government’s evidence at trial, and that there was sufficient evidence to support the convictions. We will affirm the convictions of Brand and Nicholas for conspiracy to commit wire fraud under Count 2. 119 VII. The District Court’s Instruction to the Jury on the Meaning of Intent Nicholas contends that the District Court improperly instructed the jury by using the disjunctive rather than the conjunctive at one point in its definition of intent. When providing its final charge to the jury, the District Court explained: Certain of the offenses charged in the indictment require that the government prove that the charged defendant acted intentionally or with intent. This means that the government must prove either that (1), it was the defendant’s conscious desire or purpose to act in a certain way or to cause a certain result; or (2), the defendant knew that he or she was acting in that way or it would be practically certain to cause that result. JA5787 (emphasis added). According to Nicholas, an accurate definition of intent required that the final “or” be an “and.” Nicholas argues that this was an error so grievous as to “effectively eliminate[] the intent element from each offense of conviction.” 22 Nicholas Br. 26. 22 The Comment to Third Circuit Model Criminal Jury Instruction 5.03 makes clear that the definition of intent encapsulates both “specific intent” (acting “purposely” or with “conscious object”) and “general intent” (acting 120 Our review of whether a jury instruction stated the proper legal standard is plenary. United States v. Petersen, 622 F.3d 196, 207 n.7 (3d Cir. 2010). At trial, Nicholas failed to object to this portion of the jury charge. Accordingly, our review must be for plain error. See United States v. Flores-Mejia, 759 F.3d 253, 258 (3d Cir. 2014) (en banc). To prevail on plain error review, Nicholas must establish that there was an error, that it was plain (i.e., clear under current law), and that it affected her substantial rights (i.e., whether there is a reasonable likelihood that the jury applied the challenged instruction in an impermissible manner). United States v. Olano, 507 U.S. “knowingly” or “with awareness”). Although Nicholas describes the alleged error as “essentially eliminating” the element of intent, we think Nicholas’s argument is better understood as a claim that the instruction given could have permitted a jury to conclude that she acted with only general intent (that she was aware of what she was doing), when her crimes require specific intent (that she had an illegal purpose). As her brief states, “[p]lainly she ‘knowingly’ wrote checks from EAA to [Solutions] and made record entries about them; the question was whether she intended to defraud EAA and NASA, or to obstruct justice, by doing so.” Nicholas Br. 24. We cannot agree with her characterization that the instruction resulted in the “effective omission” of the intent element from the jury instructions. 121 725, 733–34 (1993); United States v. Dobson, 419 F.3d 231, 239–40 (3d Cir. 2005). If these requirements are met, we may then exercise our discretion to address the error, but only if we conclude that the error seriously affected the fairness, integrity, or public reputation of the judicial proceeding. United States v. Andrews, 681 F.3d 509, 517 (3d Cir. 2012) (quoting Johnson v. United States, 520 U.S. 461, 467 (1997)). A failure to instruct the jury on a necessary element of an offense ordinarily constitutes plain error, unless the instructions as a whole otherwise make clear to the jury all the necessary elements of the offense. United States v. Stimler, 864 F.3d 253, 270 (3d Cir. 2017). Nicholas acknowledges, as she must, that the instruction given was a verbatim recitation of Instruction 5.03 of the Third Circuit’s Model Criminal Jury Instructions. She nonetheless contends that our Model Instruction is erroneous. Even if we were to accept Nicholas’s contention that the instruction is incorrect, a proposition we consider as highly doubtful, see Petersen, 622 F.3d at 208 (“We have a hard time concluding that the use of our own model jury instruction can constitute error . . . .”), we conclude that, considering the instructions as a whole, the District Court clearly and specifically instructed the jury on the intent element as it applied to each of Nicholas’s charged crimes. The disputed intent instruction was given at the beginning of the final charge, explaining the general 122 meaning of the intent applicable to “[c]ertain of the offenses charged.”23 JA5787. The District Court went on to instruct the jury in specific detail on the elements of each of the crimes of which Nicholas was accused, 23 The introductory definition did not end with the language Nicholas cites. The District Court elaborated that acting in good faith is a complete defense to the charges: The offenses charged in the indictment require proof that the charged defendants acted with criminal intent. If you find that a defendant acted in good faith that would be a complete defense to such a charge, because good faith on the part of the defendant would be inconsistent with his or her acting knowingly, willfully, corruptly, or with intent to defraud or intent to impede, obstruct, or wrongfully influence. JA5788–89 (emphasis added). This instruction undermines Nicholas’s claim that the jury could have reasonably concluded that she “ ‘knowingly’ wrote checks” but did not “intend[] to defraud . . . or to obstruct justice[] by doing so,” Nicholas Br. 24, as this instruction leaves little room for doubt that good faith is at odds with “criminal intent.” 123 explaining also the intent element of each. 24 See JA5791 (describing the third element of the RICO conspiracy charge as: “the particular defendant and at least one other alleged conspirator shared a unity of purpose and the intent to achieve the objective of conducting or participating in the conduct of an enterprise’s affairs through a pattern of racketeering activity”); JA5823 (regarding wire fraud, instructing that the government must prove “[t]hat the defendant under consideration acted with the intent to defraud”); JA5838–39 (regarding obstruction of justice, instructing that the defendant must have acted “with the intent to impair the record, document, or object’s integrity or availability for use in an official proceeding,” and must have acted corruptly “with the purpose of wrongfully impeding the due administration of justice”); JA5860 (explaining that falsification of records requires that “the defendant under consideration acted with the intent to impede, obstruct or influence the investigation or proper administration of a matter”). These instructions are consistent with both our Model Jury Instructions and our case law concerning the elements of these crimes. See Third Circuit Model Criminal Jury Instruction 6.18.1962D (RICO), 6.18.1343 (wire fraud), 6.18.1512A2 (obstruction of justice); United States v. Sussman, 709 F.3d 155, 168 (3d Cir. 2013) (obstruction of justice); United States v. 24 Nicholas did not object to the knowledge and intent instructions when the District Court discussed each of the individual charges, and does not identify a disagreement with any specific instruction on any particular charge. 124 Moyer, 674 F.3d 192, 208–09 (3d Cir. 2012) (falsification of records); United States v. Pelullo (Pelullo I), 964 F.2d 193, 216 (3d Cir. 1992) (wire fraud). The District Court also provided a separate definition of the knowledge element of each charge, illustrating the difference between knowledge and intent. See JA5793 (explaining that the evidence must show that a RICO defendant “knowingly agreed to facilitate or further a course of conduct, which if completed would include a pattern of racketeering activity”); JA5823 (wire fraud means that the defendant “knowingly devised a scheme to defraud a victim . . . by materially false or fraudulent pretenses”); JA5860 (falsification of records has as an element “[t]hat the defendant under consideration knowingly concealed, covered up, falsified or made false entries in a document or record”). These instructions made clear that knowledge and intent are separate considerations, undermining Nicholas’s contention that the jury was led to believe that “knowledge is sufficient to prove intent.” Nicholas Br. 24. The District Court provided each member of the jury with more than 100 pages of instructions before deliberations began. Viewing those instructions as a whole, we are satisfied that the jury was apprised of the correct meaning of intent as an element of the crimes with which Nicholas was charged, as well as the distinction between knowledge and intent. We perceive no error, 125 much less error that is plain, in the District Court’s instructions to the jury. 25 VIII. Sending the Indictment to the Jury At trial, Vederman, Nicholas, and Brand objected to the District Court’s decision to give the jury a redacted copy of the indictment to use during its deliberations. Only Nicholas and Brand raise this issue on appeal. In Nicholas’s view, sending the indictment to the jury unfairly prejudiced her because it contained unsupported allegations that she had obstructed federal agencies and referred to a nonexistent certification requirement for Sallie Mae funds. Brand argues that he was prejudiced by the indictment’s references to “schemes” and “fake” contracts, and because it mentioned Brand’s spouse and that she was a former member of Fattah’s congressional staff. Nicholas and Brand together assert that the indictment included legal theories on which the jury was not instructed. They contend that the indictment’s narrative of the Government’s case set out a roadmap that omitted any averments relating to the defense theory and allowed the Government to yet again present its case. To buttress that argument, Nicholas and Brand cite the testimony of Juror 12, who described the jury’s initial 25 Accordingly, we need not consider the merits of Nicholas’s argument that Model Criminal Jury Instruction 5.03 is erroneous. 126 deliberations and alleged that the jurors viewed the indictment as evidence. In United States v. Todaro, 448 F.2d 64, 66 (3d Cir. 1971), we held that the decision to allow “jurors to have a copy of the indictment with them during their deliberations . . . is a matter within the discretion of the District Judge, subject to a limiting instruction that the indictment does not constitute evidence, but is an accusation only.” Subsequently, in United States v. Pungitore, 910 F.2d 1084, 1142 n.83 (3d Cir. 1990), we acknowledged that the District Court has the power to redact the indictment if doing so would be appropriate to avoid prejudice to the defendant. See also United States v. Roy, 473 F.3d 1232, 1237 n.2 (D.C. Cir. 2007) (noting that court may redact an indictment before submitting it to the jury). While both Nicholas and Brand objected in general terms to the District Court’s decision to provide the indictment to the jury, they have not directed us to any specific request to redact the information they now claim is prejudicial. And the District Court provided a limiting instruction on four occasions during its charge, repeatedly emphasizing that the indictment was not evidence. JA5765, 5767, 5782, 5880. The Court instructed the jury on its duty to base its verdict “solely upon the evidence in the case.” JA5764. Just before the jury retired to deliberate, the Court reiterated that the purpose of the 127 indictment is to set forth the charges, and that it is “merely an accusation.” JA5909. “[J]uries are presumed to follow their instructions . . . .” Richardson v. Marsh, 481 U.S. 200, 211 (1987). In our view, Juror 12’s assertion that the indictment was being considered evidence does not, standing alone, establish that his fellow jurors actually did so. We reject the notion that the jury, after hearing weeks of testimony and having viewed substantial documentary evidence, went on to ignore the Court’s limiting instruction concerning the indictment. 26 Accordingly, we conclude 26 We acknowledge that our case law provides minimal guidance to district courts concerning the practice of sending an indictment to the jury for their use during deliberations. We are also aware that some courts have disapproved the practice of sending the indictment out with the jury. See United States v. Esso, 684 F.3d 347, 352 n.5 (2d Cir. 2012); Roy, 473 F.3d at 1237 n.2. We emphasize that this practice is committed to the sound discretion of the district judge. Todaro, 448 F.2d at 66. In our view, such an exercise of a judge’s discretion should be informed by considering the nature of the case, the number of defendants, the length of the indictment, the extent of the factual recitation supporting the criminal charges, and most importantly, whether the indictment (especially if lengthy and fact-laden) will be useful to the jury, in light of the judge’s own carefully tailored jury 128 that the District Court did not abuse its discretion in sending the indictment out to the jury. IX. The District Court’s Evidentiary Rulings Vederman, Nicholas, and Brand each challenge evidentiary rulings by the District Court. We conclude that none of these contentions warrants setting aside their convictions. A. The District Court’s Application of Rule 404(b) Vederman argues that the District Court misapplied Federal Rule of Evidence 404(b) when it excluded evidence of Vederman’s prior gift-giving. 27 This Court reviews a district court’s application of Rule 404(b) for abuse of discretion. United States v. Daraio, 445 F.3d 253, instruction, as supplemented by a verdict slip. See Esso, 684 F.3d at 352 n.5. 27 Although Rule 404(b) determinations are usually in response to attempts to introduce “bad” acts evidence, Vederman’s attempt to introduce “good” acts of gift- giving is properly analyzed under the same rule. Ansell v. Green Acres Contracting Co., 347 F.3d 515, 520 (3d Cir. 2003) (“The evidence admitted in this case differs from garden variety Rule 404(b) matter because it is evidence, not of a prior bad act in a criminal case, but of a subsequent good act in a civil case. Nonetheless, this evidence is encompassed by the plain text of Rule 404(b) which addresses ‘other . . . acts,’ not just prior bad acts.”). 129 259 (3d Cir. 2006); Ansell v. Green Acres Contracting Co., 347 F.3d 515, 519 (3d Cir. 2003). A trial court commits “[a]n abuse of discretion . . . when [the] district court’s decision rests upon a clearly erroneous finding of fact, an errant conclusion of law or an improper application of law to fact.” Pardini v. Allegheny Intermediate Unit, 524 F.3d 419, 422 (3d Cir. 2008) (quoting P.N. v. Clementon Bd. of Educ., 442 F.3d 848, 852 (3d Cir. 2006)). Federal Rule of Evidence 404(b) provides in part: (b) Crimes, Wrongs, or Other Acts. (1) Prohibited Uses. Evidence of a crime, wrong, or other act is not admissible to prove a person’s character in order to show that on a particular occasion the person acted in accordance with the character. (2) Permitted Uses; Notice in a Criminal Case. This evidence may be admissible for another purpose, such as proving motive, opportunity, intent, preparation, plan, knowledge, identity, absence of mistake, or lack of accident. Fed. R. Evid. 404(b)(1)–(2). 130 At trial, Vederman sought to present a witness from American University who would have testified that “Vederman agreed, on more than fifty instances, to financially assist students [at American] who needed help with tuition, book money, or travel funds to visit their families.” Vederman Br. 42 (emphasis omitted). According to Vederman, the testimony was relevant to refuting the Government’s argument that he agreed to guarantee the tuition expenses of Fattah’s au pair as a way of bribing the congressman. In excluding this evidence under Rule 404(b), the District Court stated at sidebar: I sustain the government’s objection to calling a representative of American University to testify on behalf of Herbert Vederman. In my view the testimony runs afoul of Rule 404(b)(1) of the Federal Rules of Evidence. I find it to be propensity evidence. He or she would be testifying about Mr. Vederman’s financial generosity with respect to students of American University. The issue here is payment of partial tuition of a student at the Philadelphia University. I see no connection between the generosity at American University and the situation with Philadelphia University. 131 JA4459–60. Vederman argues that because the proposed evidence related to Vederman’s intent, and not solely his propensity to perform good acts, we should conclude that the District Court abused its discretion. We see no error in the District Court’s ruling. Vederman challenges as arbitrary the District Court’s “assertion that support for American University students is too remote from support for Philadelphia University students” such that it constitutes inadmissible evidence. Vederman Reply Br. 23. This distinction was far from arbitrary. Vederman may well have financially supported American University students because of connections he had to that school or to the D.C. community at large—connections Vederman did not have to Philadelphia University. And the excluded testimony appears to have described support for students Vederman did not previously know. By supporting Fattah’s au pair, Vederman was helping an employee of a man whom he knew quite well. JA889 (“[Fattah and Vederman] spent a lot of time together traveling back and forth to Washington, in the case of a death in the family attending certain ceremonies that were important, and above all spending time with each other and their families together.”). Vederman’s decision to help Fattah’s au pair, who wished to attend Philadelphia University, seems more like a departure from, rather than a continuation of, his pattern of support for American University students. 132 As the party seeking admission of evidence under Rule 404(b), Vederman bore “the burden of demonstrating its applicability” and “identifying a proper purpose.” United States v. Caldwell, 760 F.3d 267, 276 (3d Cir. 2014). By failing to explain sufficiently why the factual distinctions discussed above were not material, Vederman failed to meet his burden. In particular, although Vederman argues that he offered evidence of his prior gift- giving to prove intent—“a proper non-propensity purpose”—he failed to show why the proposed testimony was “relevant to that identified purpose.” Id. at 277. 28 As we noted in Ansell v. Green Acres Contracting, an employment discrimination case on which Vederman heavily relies, “[t]here is. . . no bright line rule for determining when evidence is too remote to be relevant.” 347 F.3d at 525. As such, a district court’s determination under Rule 404(b) “will not be disturbed on appeal unless it amounts to an abuse of discretion.” Id. The District 28 Under Rule 404(b), “prior act evidence is inadmissible unless the evidence is (1) offered for a proper non- propensity purpose that is at issue in the case; (2) relevant to that identified purpose; (3) sufficiently probative under Rule 403 such that its probative value is not outweighed by any inherent danger of unfair prejudice; and (4) accompanied by a limiting instruction, if requested.” United States v. Caldwell, 760 F.3d 267, 277–78 (3d Cir. 2014). 133 Court did not abuse its discretion in excluding evidence of Vederman’s support for students at American University. B. Evidentiary Rulings Regarding Nicholas’s Defense Nicholas argues that the District Court rendered three erroneous evidentiary rulings that prejudiced her defense. We do not find any of her arguments convincing. 1. The EAA Board Minutes In support of its theory that Nicholas defrauded EAA, the Government introduced minutes from EAA’s Board. Minutes from 2005 revealed that the Board limited Nicholas’s signing authority to $100,000. Minutes from December 2007, February 2008, and May 2008 failed to reference either the EAA–Solutions contract or the checks, drawn from EAA’s account for $500,000 and $100,000, that were purportedly paid pursuant to the contract. Nicholas contends that the Board minutes were erroneously admitted because they constituted improper hearsay which failed to satisfy either the exception for business records under Federal Rule of Evidence 803(6) or the absence of records exception under Federal Rule of Evidence 803(7). “We review the District Court’s evidentiary ruling[s] for abuse of discretion, but also ‘exercise plenary review . . . to the extent [the rulings] are based on a legal interpretation of the Federal Rules of Evidence.’ ” Repak, 852 F.3d at 240 (citations omitted) (second alteration in original). 134 Applying those standards here,29 we conclude there was sufficient basis to admit exhibit EAA-48 under Federal Rule of Evidence 803(6) as a business record. Although allowing the prosecution to inquire over Nicholas’s earlier objection based on Federal Rule of Evidence 803(7) to the absence of any mention of the $500,000 check in the 2008 Board Minutes was error, it was not reversible error. Other evidence in the record overwhelmingly showed that Nicholas made the $500,000 and $100,000 payments to Solutions with the requisite fraudulent intent. That evidence included the 2005 policy limiting Nicholas’s authority to approve expenditures to $100,000, the accountant’s testimony about the disbursement policy requiring the completion of a check request form for the issuance of a check, and Jones’ testimony that he never saw a request for the $500,000 check. That evidence, coupled with Nicholas’s tendering of the upfront half-million dollar payment before the terms of the purported agreement had been finalized or executed, her failure to mention the $600,000 in payments to Solutions during her interview with FBI Agent 29 We note that the argument section of Nicholas’s brief on this issue specifically cited to the page in the record where exhibit EAA-48 was admitted without objection by any defense counsel. See Appellant’s Br. at 57 (citing JA4381- 82). Nonetheless, there was an earlier objection to the admission of exhibit EAA-48 that was preserved and that was referenced in a string citation elsewhere in Nicholas’s brief. 135 Dieffenbach, and her alteration of EAA’s general ledger to conceal the $100,000 payment, makes it “highly probable” that the jury would have reached the same result. See United States v. Friedman, 658 F.3d 342, 352 (3d Cir. 2011). Nicholas also asserts that the Board minutes were unfairly prejudicial. We disagree. Any possible prejudice was minimized by the fact that the Board minutes make no reference to either the EAA–Solutions contract or to any financial matters whatsoever. Indeed, given these lacunae in the Government’s proof, a reasonable factfinder might well have concluded that the Board’s intention to limit Nicholas’s signing authority had not been implemented and that Nicholas had not concealed the contract from the Board. 2. Jones’ Memory Regarding Other Contracts Nicholas next asserts that the District Court erred during her cross-examination of Board Chairman Jones by sustaining the prosecution’s objection to her inquiry into whether he remembered other contracts in excess of $100,000 being brought to the Board. See JA1386–87. The basis of the prosecution’s objection seemed to be that Nicholas’s line of inquiry was beyond the scope of the direct testimony. JA1387 (“I showed checks concerning what’s going on, not other programs.”); see Fed. R. Evid. 611(b). The District Court sustained the objection, declaring that “it has absolutely nothing to do with this case.” JA1387. Nicholas contends that if Jones did not 136 recall whether other large contracts had been presented to the Board, his inability to recall the EAA–Solutions contract would have been “unremarkable rather than evidence of fraud or concealment.” Nicholas Br. 61. We acknowledge that whether Jones remembered other large contracts requiring Board approval had some relevance under Rule 401. Yet any error by the District Court in prohibiting Nicholas’s counsel from pursuing this line of inquiry is harmless. Jones admitted that he did not know if the Board ever implemented the policy requiring its approval of contracts exceeding $100,000. He also conceded that the EAA Board focused less on the financial side of EAA than on its programs. JA1383–85. Nicholas could not have been prejudiced by the District Court’s ruling. 3. Exclusion of NOAA Evidence Nicholas defended against the criminal charges arising out of the non-existent October 2012 conference by asserting that she acted in “good faith in spending the NOAA funds on EAA expenses,” Nicholas Br. 64, that the difference in the dates in the paperwork was not material, and that NOAA had received the benefits of the sponsorship because its logo was displayed on the signage used at the February conference. Nicholas succeeded in presenting testimony and introducing photographs that showed NOAA’s logo on the February 2012 annual conference bags, padfolios, and name tags. The Court excluded a photograph of a NOAA intern at the February 137 2012 conference, other photographs of the February conference signage, and some checks that pertained to the February conference. Nicholas claims that her inability to introduce those exhibits frustrated her ability to present her good faith defense. We are not persuaded. The photographs were excluded as cumulative, the sort of ruling to which we afford trial judges very broad discretion. See Fed. R. Evid. 403; United States v. Dalfonso, 707 F.2d 757, 762 (3d Cir. 1983). It was not error to exclude the student intern’s photograph. The conference brochure included photographs from previous conferences, and the witness from NOAA was unable to testify to the year the student served as an intern. Finally, the checks tendered for the travel expenses incurred for the February conference were excluded as irrelevant to whether Nicholas had a good faith belief that NOAA sponsored the October conference. C. The Cooperating Witness’s Mental Health Records During discovery, Brand learned that a cooperating witness was diagnosed with bipolar II disorder and was taking medication to treat that condition. Brand subpoenaed mental health records kept by the witness’s current and former psychiatrists in hopes of using those records to attack the witness’s memory, truthfulness, and credibility. The witness and the Government both filed motions to quash the subpoena, arguing that the witness’s mental health records were protected by the psychotherapist–patient privilege recognized by the 138 Supreme Court in Jaffee v. Redmond, 518 U.S. 1 (1996). The Government also filed a motion in limine seeking to restrict the scope of cross-examination to prevent Brand from questioning the witness about his mental health. Alongside his motion to quash, the witness voluntarily produced for the Court his mental health records. The Court concluded that the psychotherapist– patient privilege would ordinarily apply to the mental health records, but that the privilege was not absolute, especially when invoked in response to a criminal defendant’s efforts to obtain through discovery evidence that is favorable to his case. Following the procedure set forth in Pennsylvania v. Ritchie, 480 U.S. 39 (1987), the District Court conducted an in camera review of the mental health records to determine if they contained material evidence—that is, evidence that would “give[] rise to a reasonable probability that it [would] affect the outcome of the case.” JA149. The District Court found “nothing in the mental health records of the [witness] . . . material for this criminal action,” noting that “[t]he records reveal nothing that calls into question [the witness’s] memory, perception, competence, or veracity.” JA150. Accordingly, the Court entered an order granting the motions to quash the subpoena. The District Court also granted the Government’s motion in limine and restricted the scope of cross- examination, ruling that “no reference may be made to [the witness’s] bipolar disorder or the medications he takes to 139 manage it.” JA142, 156. The Court reasoned that bipolar disorder varied in its effects from person to person, and concluded that Brand had not shown that the effects of the disorder had any bearing on the witness’s credibility. The District Court ruled that cross-examination would not serve any valid impeachment purpose. Brand claims that the District Court’s order ran afoul of the Due Process Clause of the Fifth Amendment and the Confrontation Clause of the Sixth Amendment. We review a district court’s rulings to quash a subpoena and to limit the scope of cross-examination for abuse of discretion. United States v. Tykarsky, 446 F.3d 458, 475 (3d Cir. 2006); NLRB v. Frazier, 966 F.2d 812, 815 (3d Cir. 1992). Here, the District Court did not abuse that discretion. 1. The District Court’s Denial of Access to the Mental Health Records In claiming that the District Court’s decision to review the mental health records in camera before ruling on their admissibility violated his rights under both the Fifth and Sixth Amendments, Brand specifically argues that his right to confront the witness was impeded because he was denied access to records he could have used to impeach the witness. This very argument was considered and rejected by a plurality of the Supreme Court in Ritchie, which noted that “the effect [of the argument] would be to transform the Confrontation Clause into a constitutionally compelled rule of pretrial discovery. . . . [T]he right to 140 confrontation is a trial right, designed to prevent improper restrictions on the types of questions that defense counsel may ask during cross-examination.” 480 U.S. at 52. We follow the Ritchie plurality, and conclude that the Confrontation Clause did not require the District Court to grant Brand access to the witness’s mental health records. Brand next challenges the District Court’s decision to quash the subpoena as a violation of the Fifth Amendment’s Due Process Clause. He concedes that Ritchie’s Due Process holding allowed the District Court to review the mental health records in camera without disclosing them to him. See id. at 59–60 (“A defendant’s right to discover exculpatory evidence does not include the unsupervised authority to search through [the Government’s] files. . . . We find that [the defendant’s] interest . . . in ensuring a fair trial can be protected fully by requiring that the [privileged] files be submitted only to the trial court for in camera review.”). Brand instead argues that the District Court abused its discretion by focusing on “irrelevant facts and spurious symptoms. . . . such as ‘hallucinations,’ ” and by “refus[ing] to consider evidence of cognitive impairment and memory issues.” Brand Br. 30. The record reveals, however, that the District Court reviewed the mental health records and determined that they “reveal[ed] nothing that calls into question [the witness’s] memory, perception, competence, or veracity.” JA150. This hardly amounts to a refusal to consider evidence of cognitive impairment or memory issues. 141 Brand also challenges the legal standard applied by the District Court, arguing that the court “focused solely on whether disclosure would ‘change the outcome’ of Brand’s trial,” Brand Br. 29 (quoting JA148), rather than considering “whether the ultimate verdict is one ‘worthy of confidence.’ ” Id. (quoting United States v. Robinson, 583 F.3d 1265, 1270 (10th Cir. 2009)). Brand misleadingly quotes from the District Court’s opinion. The District Court considered, in accordance with Ritchie, “whether there is a reasonable probability that disclosure would change the outcome” of Brand’s trial, JA148 (emphasis added), not whether disclosure would necessarily change the outcome. As articulated in Ritchie, a “ ‘reasonable probability’ is a probability sufficient to undermine confidence in the outcome.” 480 U.S. at 57 (quoting United States v. Bagley, 473 U.S. 667, 682 (1985) (Blackmun, J.)). The District Court applied the correct standard. 2. The District Court’s Grant of the Motion in Limine In granting the Government’s motion in limine, the District Court ruled that Brand could not “reference . . . [the witness’s] bipolar disorder or the medications he takes to manage it.” JA156. Yet that ruling placed no restriction on Brand’s ability to cross-examine the witness with respect to “his memory, competence, or truthfulness.” Id. Brand argues, nevertheless, that his Sixth Amendment right “to be confronted with the witnesses against him” was violated. U.S. Const. amend. VI. 142 The Confrontation Clause protects a defendant’s right to cross-examine a witness with respect to any testimonial statements made by that witness. United States v. Berrios, 676 F.3d 118, 125–26 (3d Cir. 2012) (citing Crawford v. Washington, 541 U.S. 36, 51 (2004), and Davis v. Washington, 547 U.S. 813, 823–24 (2006)). But the scope of cross-examination is not unlimited, and “[a] district court retains ‘wide latitude insofar as the Confrontation Clause is concerned to impose reasonable limits on such cross-examination based on concerns about . . . harassment, prejudice, confusion of the issues, the witness’ safety, or interrogation that is repetitive or only marginally relevant.’ ” John-Baptiste, 747 F.3d at 211 (quoting United States v. Mussare, 405 F.3d 161, 169 (3d Cir. 2005)). We review limitations on cross-examination for abuse of discretion, and reverse “only when the restriction ‘is so severe as to constitute a denial of the defendant’s right to confront witnesses against him and . . . is prejudicial to [his] substantial rights.’ ” Id. (alternation in original) (quoting United States v. Conley, 92 F.3d 157, 169 (3d Cir. 1996)). In United States v. Chandler, 326 F.3d 210, 219 (3d Cir. 2003), we analyzed whether a district court’s decision to limit cross-examination with respect to a witness’s motivation for testifying violated the Confrontation Clause. See also Mussare, 405 F.3d at 169; John-Baptiste, 747 F.3d at 211–12. Consistent with Delaware v. Van Arsdall, 475 U.S. 673 (1986), we first concluded that “the exposure of a witness’ motivation in testifying is a proper 143 and important function of the constitutionally protected right of cross-examination.” Chandler, 326 F.3d at 219– 20 (quoting Van Arsdall, 475 U.S. at 678–79). We also noted that the Confrontation Clause does not prevent a trial judge from imposing reasonable limits on cross- examination. Id. In reviewing a district judge’s imposition of such limitations, we apply a two-part analysis. As we have since described, “we inquire into: ‘(1) whether the limitation significantly limited the defendant’s right to inquire into a witness’s motivation for testifying; and (2) whether the constraints imposed fell within the reasonable limits that a district court has the authority to impose.’ ” John-Baptiste, 747 F.3d at 211–12 (quoting Mussare, 405 F.3d at 169). The same analytical framework is appropriate when determining whether a restriction on the cross- examination of a witness with respect to his memory and perception violates the Confrontation Clause. See Davis v. Alaska, 415 U.S. 308, 316 (1974); Greene v. McElroy, 360 U.S. 474, 496 (1959); United States v. Segal, 534 F.2d 578, 582 (3d Cir. 1976). Memory and perception, like motivation for testifying, are central issues affecting the credibility of any witness, and unreasonable limitations on the right to cross-examine on those subjects cannot be countenanced. We therefore ask, paraphrasing Chandler: (1) whether the District Court’s decision to put the witness’s diagnosis and medications off limits significantly impaired Brand’s right to inquire into the witness’s memory and perception; and (2) whether the 144 ruling fell within the reasonable limits that the District Court has the authority to impose. We conclude that the District Court did not err. As an initial matter, the District Court permitted Brand to cross-examine the witness about his memory and perception, and limited cross-examination only with respect to the witness’s bipolar disorder and the medications he was taking to treat that condition. Brand was free to question the witness about his memory and perception, and indeed did so. The restriction on asking the witness about his bipolar disorder was not a significant limitation of Brand’s right to inquire into the witness’s memory or perception. Moreover, as the District Court pointed out, Brand failed to show how inquiry into the witness’s bipolar disorder would be useful for impeachment purposes. See JA154. Given that failure, the District Court’s limits on cross-examination were reasonable. The Court concluded, after reviewing the evidence submitted by Brand and the witness’s mental health records, that any mention of the witness’s bipolar disorder would “only be designed to confuse the jury or to stigmatize him unfairly because of a ‘mental problem’ without any countervailing probative value.” JA155. The District Court did not abuse its discretion in limiting Brand’s cross-examination on a topic that would be far more prejudicial than probative. See Tykarsky, 446 F.3d at 476–77 (“[T]he District Court acted 145 well within its discretion to restrict irrelevant and confusing testimony.”). All of this is not to suggest that a witness’s mental health is always off limits. The appropriate course in any given case must be determined from the facts and circumstances surrounding that case and the witness’s particular condition. See United States v. George, 532 F.3d 933, 937 (D.C. Cir. 2008) (“The days are long past when any mental illness was presumed to undermine a witness’s competence to testify. . . . [M]ental illness [is] potentially relevant in a broad[] range of circumstances . . . . [But] some indication is needed that a particular witness’s medical history throws some doubt on the witness’s competence or credibility.”). Here, Brand failed to show, through mental health records or otherwise, any particularized reason to doubt the credibility of the witness for medical reasons. Brand states that the witness provided “the only evidence offered” on the intent element of his conspiracy conviction and that he should therefore be entitled to unrestricted cross-examination. Brand Br. 12 n.3. Yet no matter the importance of a witness to any party, a district court may always place reasonable limits on cross- examination to avoid “harassment, prejudice, confusion of the issues, the witness’ safety, or interrogation that is repetitive or only marginally relevant.” John-Baptiste, 747 F.3d at 211 (citation omitted). 146 We conclude that the District Court did not abuse its discretion in restricting the scope of Brand’s cross- examination of the cooperating witness. X. The Government’s Cross-Appeal The jury convicted Fattah, Vederman, and Bowser of bank fraud, 18 U.S.C. § 134430 (Count 19) and making false statements to a financial institution, 18 U.S.C. § 1014 31 (Count 20). In response to post-trial motions, the District Court granted a judgment of acquittal on both counts under Fed. R. Crim. P. 29, concluding that the 30 “Whoever knowingly executes, or attempts to execute, a scheme or artifice . . . to defraud a financial institution . . . shall be fined not more than $1,000,000 or imprisoned not more than 30 years, or both.” The definition of “financial institution” for purposes of § 1344 is set forth at 18 U.S.C. § 20, and includes “a credit union with accounts insured by the National Credit Union Share Insurance Fund” and “a mortgage lending business (as defined in section 27 of this Title).” 18 U.S.C. §§ 20(2), (10). 31 “Whoever knowingly makes any false statement or report, or willfully overvalues any land, property or security, for the purpose of influencing in any way the action of . . . a Federal credit union . . . any institution the accounts of which are insured by . . . the National Credit Union Administration Board . . . or a mortgage lending business . . . shall be fined not more than $1,000,000 or imprisoned not more than 30 years, or both.” 147 evidence was insufficient to establish that the Credit Union Mortgage Association (CUMA), the entity to whom Fattah, Vederman, and Bowser made the false statements, is a “financial institution,” or, more specifically, a “mortgage lending business” as defined in 18 U.S.C. § 27. The Government claims that, viewing the evidence in the light most favorable to it, the District Court erred and that CUMA is, indeed, a “mortgage lending business.” We agree. Because the evidence is sufficient to support the jury’s verdict, we will remand so Fattah and Vederman may be resentenced on these charges.32 A. CUMA is a Mortgage Lending Business In reviewing the District Court’s post-verdict judgment of acquittal under Rule 29 of the Federal Rules of Civil Procedure, we consider whether the evidence, when viewed in a light most favorable to the government, supports the jury’s verdict. United States v. Dixon, 658 F.2d 181, 188 (3d Cir. 1981). Our standard of review is the same as that applied by the District Court, and we must uphold the jury’s verdict unless no reasonable juror could accept the evidence as sufficient to support the defendant’s guilt beyond a reasonable doubt. United States v. Coleman, 811 F.2d 804, 807 (3d Cir. 1987). 32 Because the Government did not file an appeal as to Bowser, the cross-appeal is limited to Fattah and Vederman. The judgment of acquittal as to Bowser is therefore unaffected by our ruling today. 148 Initially, the grand jury’s indictment alleged that CUMA is a financial institution because it is federally insured. JA302–03. At trial, however, the jury was instructed that CUMA could qualify as a financial institution either because it is federally insured or because it is a “mortgage lending business.” See JA111, 401–02. A “mortgage lending business” is “an organization which finances or refinances any debt secured by an interest in real estate, including private mortgage companies and any subsidiaries of such organizations, and whose activities affect interstate or foreign commerce.” 18 U.S.C. § 27. At trial, CUMA’s president and CEO, Eddie Scott Toler, testified that CUMA is not federally insured. JA4235. The Government therefore attempted to prove that CUMA is a “mortgage lending business” by presenting evidence that CUMA funds mortgages and then sells them in a secondary market. Toler also testified that CUMA is a “credit union service organization”—a for-profit company owned by 48 credit unions, which serves small credit unions that do not have the infrastructure or in-house expertise to handle mortgage loans themselves. JA4235. According to Toler, “[CUMA] exclusively provide[s] First Trust Residential Mortgage loaning [sic] services, all the way from the origination of the mortgage loan through processing, underwriting, closing and access to the secondary market where—and we’re selling the mortgage loan on the secondary market.” JA4236–37. In jurisdictions in which 149 CUMA is licensed, 33 CUMA holds the mortgage for a limited period, generally from two to thirty days, and then sells the mortgage either to a partner credit union or on the secondary market. JA4240. The District Court concluded that CUMA is not a “mortgage lending business” because “[t]he record is devoid of any evidence that CUMA finances or refinances any debt.” JA113. Concluding that CUMA “simply is a loan processor for various credit unions which do the financing or refinancing,” id., the District Court ruled that CUMA’s “activity does not constitute the financing or refinancing of any debt. CUMA is not the mortgagee. It is merely selling the debt instrument to a third party.” JA114. We cannot agree with the District Court’s view of the evidence. Toler testified that in “Maryland, D.C., and Virginia . . . all of the loans are closed in the name of CUMA.” JA4238–39. As Toler described it, CUMA borrows on a line of credit to fund the loan, and when the loan is sold, CUMA pays off its line of credit. JA4239–40. So contrary to the District Court’s assessment, the evidence, viewed in a light most favorable to the Government, shows that CUMA is indeed the mortgagee—at least during the time from closing until the loan is sold to a partner credit union or on the secondary market. The fact that CUMA funds the closing and then 33 CUMA is licensed in Maryland, Washington, D.C., and Virginia. JA4238. 150 holds the mortgage, even for a brief time, is sufficient to support a conclusion that CUMA is “an organization which finances or refinances any debt secured by an interest in real estate.” 18 U.S.C. § 27. Fattah and Vederman attempt to refute the argument that CUMA engages in financing mortgages by focusing on Toler’s testimony that CUMA “doesn’t actually have any money to fund these mortgage loans.” JA4239; see Fattah Reply Br. 38, Vederman Reply Br. 36. But Toler testified that CUMA employs a credit line to borrow the funds necessary to close on mortgages. See JA4239. That CUMA incurs debt to finance mortgages hardly undermines a conclusion that CUMA finances mortgages. Indeed, it is the very nature of modern banking that financial institutions do not hold cash reserves equal to the full amount of their liabilities. See, e.g., Timothy C. Harker, Bailment Ailment: An Analysis of the Legal Status of Ordinary Demand Deposits in the Shadow of the Financial Crisis of 2008, 19 Fordham J. Corp. & Fin. L. 543, 561 (2014) (“[F]ractional reserve banking . . . is the de facto standard for all modern banks.”). Vederman also argues that, even if CUMA acts as a mortgage lending business in some transactions, it was not acting as a mortgage lending business in this transaction. Vederman points to Toler’s testimony that, in a state in which CUMA is not licensed, the mortgage is closed in the name of a credit union. In such cases, the credit union, and not CUMA, owns the mortgage for the short period before 151 the loan is sold on the secondary market. JA4241. CUMA is not licensed in the Commonwealth of Pennsylvania. See id. Thus, according to Vederman, CUMA was acting in its capacity as a mortgage servicing company for Fattah’s vacation home purchase and did not—and could not— finance Fattah’s mortgage. That would mean that CUMA could not have been a victim of a crime against a financial institution in this instance: “When an entity is not functioning as a mortgage lender, the ‘pertinent federal interest’ behind the statutes is not implicated.” Vederman Reply Br. 38 (citation omitted). The Government responds that neither of the statutes of conviction requires that the fraud or false statement occur in connection with the same transaction that places the entity within the definition of “financial institution.” Gov’t Fourth Step Br. 4. We agree with the Government. Both § 1344 and § 1014 protect entities that fall within the definition of “financial institution” and are otherwise quite broad in their application. See Loughrin v. United States, 134 S. Ct. 2384, 2389 (2014) (interpreting § 1344 as not requiring specific intent to defraud a bank); United States v. Boren, 278 F.3d 911, 914 (9th Cir. 2002) (“[Section 1014’s] reach is not limited to false statements made with regard to loans, but extends to any application, commitment or other specified transaction.”). Neither statute is expressly limited in the manner that Vederman suggests. Williams v. United States, 458 U.S. 279, 284 152 (1982) (“To obtain a conviction under § 1014, the Government must establish two propositions: it must demonstrate (1) that the defendant made a ‘false statement or report,’ . . . and (2) that he did so ‘for the purpose of influencing in any way the action of [a described financial institution] upon any application, advance, . . . commitment, or loan.’ ”); United States v. Leahy, 445 F.3d 634, 646 (3d Cir. 2006) (“The purpose of the bank fraud statute is to protect the ‘financial integrity of [banking] institutions.’ ”) (citing S. Rep. No. 98-225, at 377 (1983), as reprinted in 1984 U.S.C.C.A.N. 3517), abrogated on other grounds by Loughrin, 134 S. Ct. at 2389. In support of his position, Vederman relies on United States v. Devoll, 39 F.3d 575 (5th Cir. 1994), in which the Fifth Circuit concluded that § 1014 (false statements to a financial institution) is not intended to capture fraud unrelated to an entity’s lending activities, and therefore held that it “applies only to actions involving lending transactions.” Id. at 580. The Fifth Circuit stated: [W]e are not persuaded that the statute imposes liability whenever a defendant’s false statement was intended to interfere with any activity of a financial institution; such a broad interpretation of section 1014 presumably would encompass fraud or false representations having nothing to do with financial transactions, such as fraud in an employment contract or, for example, in a 153 contract to provide goods or services for custodial care, premises repair, or renovation. Id. Yet a majority of circuits, including our own, have declined to follow Devoll’s suggestion that § 1014 is restricted to lending transactions. As the Ninth Circuit has held, “we join at least six of our sister circuits—the First, Third, Fourth, Sixth, Seventh, and Tenth—in holding that 18 U.S.C. § 1014 is not limited to lending transactions, and reject the minority rule to the contrary.” Boren, 278 F.3d at 915. And even if we were to adopt Devoll’s narrow construction of § 1014 to lending transactions, that would not resolve the more specific question of whether the defrauded entity must be defined as a “mortgage lending business” by virtue of the specific transaction in which the false statements arose. Recently, the Eighth Circuit addressed precisely this issue. In United States v. Springer, 866 F.3d 949 (8th Cir. 2017), that Court considered the defendant’s appeal from the district court’s denial of a Rule 29 motion on grounds that GMAC, the entity defrauded, was not a “financial institution.” The Court upheld the district court’s determination that the evidence was sufficient to establish that GMAC is in the mortgage lending business because there was testimony that “it had made hundreds or thousands of loans secured by mortgages in 2010 and 2011 in states all across the country,” which established that its activities affect interstate commerce. Id. at 953. It was not 154 determinative that GMAC did not own the specific loan at issue in the case: “we discern no requirement in the definition of ‘mortgage lending business’ that the business own the particular loan in question; it need only finance or refinance any debt secured by an interest in real estate, or, in other words, be in the interstate mortgage lending business in general.” Id. In our view, the Eighth Circuit’s analysis is correct. We therefore adopt that Court’s reasoning in Springer and conclude that it is of no moment that CUMA did not finance the mortgage at issue in Fattah’s case. CUMA is a “mortgage lending business,” and that alone suffices to support the convictions under §§ 1014 and 1344. B. Sufficiency of the Evidence Finally, Vederman argues that, even if CUMA is a financial institution, the judgment of acquittal should stand because the Government did not put forth any evidence that he made a false representation to CUMA.34 Specifically, Vederman argues that the title to the Porsche was actually changed to his name, making it a “true sale” as a matter of law, without regard to whether Fattah’s wife 34 Although Vederman presented this argument in his Rule 29 motion, the District Court did not need to reach it in the context of Counts 19 and 20 because the Court granted the motion on the ground that CUMA is not a financial institution. The District Court rejected the argument as to Counts 16, 17, and 18. See JA100–02. 155 continued to retain possession. See United States v. Castro, 704 F.3d 125, 139 (3d Cir. 2013) (holding in another context that “the government must be able to show that [the defendant] made a statement to government agents that was untrue, and the government cannot satisfy that burden by showing that the defendant intended to deceive, if in fact he told the literal truth”); see also 75 Pa. Cons. Stat. § 102 (defining “owner” as “[a] person, other than a lienholder, having the property right in or title to a vehicle”). The Government responds that, regardless of whether it is legally possible for one person to hold a title while a different person possesses the vehicle, the jury was permitted to consider all the circumstances in deciding whether the Porsche sale was a sham. We agree. First, as the District Court observed, it was unclear as to whether the title had been properly executed under Pennsylvania law. For instance, Fattah’s wife never appeared before a notary. 35 JA101. In addition, title 75, section 1111(a) of the Pennsylvania Consolidated Statutes requires that, “[i]n the event of the sale or transfer of the ownership of a vehicle within this Commonwealth, the owner shall . . . deliver the certificate to the transferee at 35 Vederman argues that it is of no significance that the parties did not appear before a notary as the statute requires, but he offers cases only from states other than Pennsylvania to support this proposition. 156 the time of the delivery of the vehicle.” And, the transferee must, within twenty days of the assignment of the vehicle, apply for a new title. See 75 Pa. Cons. Stat. § 1111(b). Neither of these requirements was fulfilled. Finally, Vederman never registered the Porsche in his name with the Department of Motor Vehicles. See id.; JA4254. Second, and more importantly, even if the title had been properly transferred to Vederman, the title provisions of the Pennsylvania Motor Vehicle Code “were [not] designed to establish conclusively the ownership of an automobile.” Weigelt v. Factors Credit Corp., 101 A.2d 404, 404 (Pa. Super. Ct. 1953). Indeed, “[t]he purpose of a certificate of title is not to conclusively establish ownership in a motor vehicle, but rather to establish the person entitled to possession.” Speck Cadillac-Olds, Inc. v. Goodman, 95 A.2d 191, 193 (Pa. 1953). Thus, a title provides evidence of ownership; it is not dispositive of the issue. Wasilko v. Home Mut. Cas. Co., 232 A.2d 60, 61 (Pa. Super. Ct. 1967). Vederman’s argument that the title in his name constitutes conclusive evidence of ownership rests upon an erroneous conclusion that the jury was prohibited from considering all the circumstances of the transfer. As the District Court observed, though, Pennsylvania’s Commonwealth Court has held that “[w]hether a transferor has transferred ownership of a motor vehicle to a transferee is a factual determination to be made by the court below.” Dep’t. of Transp. v. Walker, 584 A.2d 1080, 157 1082 (Pa. Commw. Ct. 1990). Thus, the signed certificate of title was appropriately treated as one piece of evidence for the jury to consider in assessing the validity of the vehicle transfer. Considered in the light most favorable to the Government, the totality of the evidence is sufficient to support the jury’s conclusion that the Porsche sale was a sham. XI. Prejudicial Spillover Finally, Fattah, Vederman, Nicholas, and Brand each contend that their convictions on various counts resulted from prejudicial spillover. We are not persuaded. We exercise plenary review over a district court’s denial of a claim of prejudicial spillover, United States v. Lee, 612 F.3d 170, 178–79 (3d Cir. 2010), and we apply a two-step test when reviewing such a claim. United States v. Wright, 665 F.3d 560, 575 (3d Cir. 2002). First, a court must consider “whether the jury heard evidence that would have been inadmissible at a trial limited to the remaining valid count[s].” Id. (quoting United States v. Cross, 308 F.3d 308, 317 (3d Cir. 2002)). The second step requires that we “ask whether that evidence (the ‘spillover evidence’) was prejudicial.” Id. We consider four factors: “whether (1) the charges are intertwined with each other; (2) the evidence for the remaining counts is sufficiently distinct to support the verdict on these counts; (3) the elimination of the invalid count [will] significantly change[] the strategy of the trial; and (4) the prosecution used language of the sort to arouse a jury.” Id. (quoting 158 United States v. Murphy, 323 F.3d 102, 118 (3d Cir. 2003)); see also United States v. Pelullo (Pellulo II), 14 F.3d 881, 898–99 (3d. Cir. 1994). These four factors are considered in a light “somewhat favorable to the defendant.” Wright, 665 F.3d at 575 (quoting Murphy, 323 F.3d at 122); see also Gov’t Br. 198 (same). A. Fattah’s Claim of Prejudicial Spillover Fattah argues that he suffered prejudicial spillover on the remaining counts of conviction in light of (1) evidence pertinent to the alleged Vederman bribery schemes that is now arguably inadmissible under McDonnell; and (2) “the government’s flawed RICO conspiracy theory.” Fattah Br. 50, 64. Fattah’s argument is undercut substantially because of our determination that McDonnell requires a new trial for Counts 16, 17, 22, and 23 and our decision to affirm the RICO conspiracy conviction. The only possible spillover left to consider is the evidence pertaining to Fattah’s arranging a meeting between Vederman and the U.S. Trade Representative, Ron Kirk, which in light of McDonnell is now arguably inadmissible.36 The evidence of the Kirk meeting admitted during this five-week trial was limited. Although this evidence 36 Nothing in this opinion is intended to foreclose the possibility that evidence of the Kirk meeting may be admissible on retrial for some purpose other than as proof of an official act. 159 was part of the Government’s proof as to both the RICO and the bribery related charges, there is more than sufficient—and distinct —evidence to support Fattah’s conviction on all the other counts. In our view, eliminating any evidence of the Kirk meeting would not have altered the strategy of the trial, nor should it significantly change the strategy for any new trial that may be held. Because Fattah has not pointed us to any argument by the prosecution relating to this meeting that could have inflamed the jury, we conclude that Fattah’s prejudicial spillover claim fails. Like the District Court, we presume that the jury followed the Court’s instructions to consider and weigh separately the evidence on each count as to each defendant and not to be swayed by evidence pertaining to other defendants. 37 B. Vederman’s Assertion of Prejudicial Spillover Because the District Court acquitted Vederman of the RICO charge, Vederman argues that he was “severely prejudiced by the presentation to the jury of a legally flawed racketeering conspiracy charge,” and as a consequence his bribery and money laundering 37 We likewise reject Brand’s prejudicial spillover arguments. See Brand Br. 6 (“Brand adopts the significant issue advanced by his co-appellant pursuant to Fed. R. App. P. 28(i) that improper jury instructions and the resulting spillover of related improperly admitted evidence and argument unfairly prejudiced Brand.”). 160 convictions should be overturned. Vederman Br. 46. In response to the Government’s appeal of the District Court’s Rule 29 acquittal on Counts 19–20 involving CUMA, Vederman asserts that these two counts also were affected by spillover evidence because the Government’s theory tied the bribery charges to the actions taken to defraud CUMA. In that we are vacating Vederman’s convictions of Counts 16–18 and 22–23 based on McDonnell and remanding for further proceedings, we need address only Vederman’s argument of prejudicial spillover as it relates to the charges involving CUMA in Counts 19–20, charges that we will reinstate. The District Court’s acquittal of Vederman on the RICO count establishes that step one of the Wright spillover test has been met. “[T]he jury heard evidence that would have been inadmissible at a trial limited” to the bribery and CUMA-related counts. Wright, 665 F.3d at 575 (quoting Cross, 308 F.3d at 317). Wright’s second step requires “ask[ing] whether that evidence (the ‘spillover evidence’) was prejudicial.” Id. Vederman submits that the RICO, bribery, and CUMA- related charges were intertwined “in that the acts relating to the alleged bribery scheme were also charged as ‘predicates’ under RICO.” Vederman Br. 49. We disagree. To be sure, the RICO, bribery, and CUMA Counts are related to one another. But in this instance, mere relatedness is not enough to demonstrate the foundation necessary for spillover. This is so because the bribery 161 charges were a predicate to the RICO charge. In other words, the jury had to determine if Vederman was guilty of bribery, and the jury then used that “predicate” to consider whether he was also guilty of the RICO conspiracy. Thus, the necessarily tiered structure of the questions presented to the jury refute Vederman’s argument that the counts were intertwined. That the bribery charges were predicates for the RICO conspiracy further demonstrates that the “evidence for the different counts was sufficiently distinct to support the verdict on other separate counts.” Pelullo II, 14 F.3d at 898. Regardless of the evidence pertaining solely to the RICO conviction, the evidence supporting both the bribery charges and the charges involving CUMA in Counts 19– 20 would have remained the same. The next factor we address is “whether the elimination of the count on which the defendant was invalidly convicted would have significantly changed the [defendant’s] strategy of the trial.” Id. As Vederman argues, “the RICO charge interfered with Vederman’s central defense to the bribery charge—that his gestures toward Fattah ‘were motivated purely by friendship.’ ” Vederman Reply Br. 28 (citing Gov’t Br. 200). In other words, the “RICO count made it dangerous to unduly emphasize [Vederman’s] close friendship” with Fattah. Id. From Vederman’s perspective, “a bribery-only trial would have reduced this danger and allowed a freer presentation of the defense.” Id. 162 It is quite likely that Vederman’s claim of friendship would have been less risky as a litigation strategy if he had not been facing a RICO charge. But Vederman nevertheless chose to take that risk and fully presented his friendship argument to the jury. Moreover, while Vederman’s reliance on friendship might have helped him defend against the bribery charges, that friendship would not have altered the evidence pertaining to Counts 19–20 involving CUMA. Whether done for friendship or some other reason, submitting fraudulent information to a financial institution is unlawful. Finally, we “examine the charges, the language that the government used, and the evidence introduced during the trial to see whether they are ‘of the sort to arouse a jury.’ ” Pelullo II, 14 F.3d at 899 (quoting United States v. Ivic, 700 F.2d 51, 65 (2d Cir. 1983)). Vederman points out that Fattah was presented as “a backslapping, corrupt party boss,” with “predictable spillover to his friend and associate, Vederman.” Vederman Br. 50 (quoting United States v. Murphy, 323 F.3d 102, 118 (3d Cir. 2003)). But this description was of Fattah, not Vederman. Vederman cites other examples of prejudicial, pejorative language in the Government’s closing arguments. At one point, the Government referred to “conspirators engaged in what can only be described as a white collar crime spree from Philadelphia all the way to Washington, D.C.” and promised “to untangle the webs of lies and deception that these conspirators spun.” Vederman Br. 51 (quoting JA5295, 5297). Whatever rhetorical flair these words 163 contained, they did not obscure the evidence which independently supported the convictions for bank fraud at Count 19 and for making false statements to CUMA at Count 20. Accordingly, because we presume that the jury followed the District Court’s instruction to consider and to weigh separately the evidence on each count and as to each defendant, and because the evidence supporting the CUMA-related charges in Counts 19–20 is sufficiently distinct from the RICO conspiracy, we conclude that Vederman’s spillover argument is unavailing.38 XII. Conclusion We will vacate the convictions of Chaka Fattah, Sr. and Herbert Vederman as to Counts 16, 17, 18, 22, and 23. 38 Nicholas adopted “pertinent portions” of the prejudicial spillover arguments advanced by Vederman and Fattah. Nicholas Br. 65. Her spillover claim has no more merit than theirs. Nicholas’s involvement in the RICO conspiracy was distinct from the bribery charges, which did not unfairly influence the other counts. As to Nicholas’s assertion that the NOAA charges did not belong in the indictment and should have been tried separately, we fail to see how this relates to a claim of prejudicial spillover. To the extent it challenges the District Court’s denial of Nicholas’s motion for a severance, Nicholas has failed to provide legal support for such a contention. See Fed. R. App. P. 28(a)(8)(A); United States v. Irizarry, 341 F.3d 273, 305 (3d Cir. 2003). 164 Fattah and Vederman may be retried on these counts before a properly instructed jury. We will also reverse the District Court’s judgment of acquittal on Counts 19 and 20. The convictions of Chaka Fattah, Sr. and Herbert Vederman will be reinstated, and the case will be remanded for sentencing on those counts. In all other respects, the judgments of the District Court will be affirmed. 165