United States of America v. Larry Householder (23-3565); Matthew Borges (23-3566)

RECOMMENDED FOR PUBLICATION
Pursuant to Sixth Circuit I.O.P. 32.1(b)
File Name: 25a0114p.06
UNITED STATES COURT OF APPEALS
FOR THE SIXTH CIRCUIT
UNITED STATES OF AMERICA,
Plaintiff-Appellee,
v.
LARRY HOUSEHOLDER (23-3565); MATTHEW BORGES
(23-3566),
Defendants-Appellants.
┐
│
│
│
│
│
│
│
│
┘
Nos. 23-3565/3566
Appeal from the United States District Court for the Southern District of Ohio at Cincinnati.
No. 1:20-cr-00077—Timothy S. Black, District Judge.
Argued: February 5, 2025
Decided and Filed: May 6, 2025
Before: THAPAR, NALBANDIAN, and DAVIS, Circuit Judges.
_________________
COUNSEL
ARGUED: Steven L. Bradley, MAREIN & BRADLEY, Cleveland, Ohio, for Appellant
Householder. Dennis C. Belli, BELLI LAW, Columbus, Ohio, for Appellant Borges. Alexis J.
Zouhary, UNITED STATES ATTORNEY’S OFFICE, Cincinnati, Ohio, for Appellee.
ON BRIEF: Steven L. Bradley, MAREIN & BRADLEY, Cleveland, Ohio, Nicholas R. Oleski,
MCCARTHY, LEBIT, CRYSTAL & LIFFMAN CO., LPA, Cleveland, Ohio, for Appellant
Householder. Dennis C. Belli, BELLI LAW, Columbus, Ohio, for Appellant Borges. Alexis J.
Zouhary, UNITED STATES ATTORNEY’S OFFICE, Cincinnati, Ohio, for Appellee.
The court delivered a PER CURIAM opinion. THAPAR, J. (pp. 44–64), delivered a
separate concurring opinion.
>

-- 1 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 2
_________________
OPINION
_________________
PER CURIAM. Larry Householder was Speaker of the Ohio House of Representatives.
A jury found him guilty of conspiring to solicit and receive almost $60 million in return for
passing a billion-dollar bailout of a failing nuclear energy company. A jury also found lobbyist
Matthew Borges guilty of playing a role in Householder’s conspiracy. Because we find no
reversible error, we affirm their convictions.
Facts and Background
This case begins with two parties, each with a problem in need of a solution.
On the one side, there’s Larry Householder. Householder was an old hand in Ohio
politics: he served in the Ohio House of Representatives for four terms in the late 1990s and
early 2000s, two of those terms in the speaker’s chair. After a decade-long absence,
Householder decided to return to public life. In November 2016, the citizens of Ohio’s 72nd
district elected Householder to represent them in the Ohio House of Representatives. This time,
Householder returned to the Ohio Statehouse seeking to reclaim the speaker position. To do so,
he sought to recruit as many candidates as possible to run in the next election and support his bid
for the speakership.
On the other side, there’s FirstEnergy Corp., an Ohio-based public utility holding
company. In 2016, FirstEnergy was in dire financial straits because one of its wholly owned
subsidiaries, FirstEnergy Solutions, ran two failing nuclear plants. FirstEnergy Solutions was
“bleeding cash.” R. 194, Pg. ID 4770. So, FirstEnergy sought a “legislative or regulatory
solution[].” Id. at Pg. ID 4777. It hoped for a “guaranteed payment” to support the failing
plants. Id. at Pg. ID 4780. Put simply, FirstEnergy sought a taxpayer-funded bailout. When a
federal effort for the bailout failed, FirstEnergy turned its attention to Ohio. For this, FirstEnergy
would need the support of the leadership of the Ohio House, which it lacked.

-- 2 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 3
Larry Householder wanted to become Speaker of the Ohio House. And FirstEnergy
wanted a legislative bailout.
A. Meetings
Householder’s bid for speaker was an ambitious undertaking. He needed to recruit
enough candidates to run in the 2018 election who, once elected, would vote him in as speaker.
So, he hired a political strategist, Jeffrey Longstreth, to “quarterback” the operation. R. 217, Pg.
ID 7585. But he also needed someone to finance it or, as Longstreth called it, serve as a “main
benefactor.” Id. at Pg. ID 7630–31. Who could that be?
In November 2016, Householder bumped into FirstEnergy CEO Chuck Jones at Game
Seven of the World Series in Cleveland. The two discussed FirstEnergy’s “urgen[t]” need for
financial help. Def. App’x, Gov’t Ex. 212. Householder and Jones met again two months later
in Washington, D.C. There, Householder and Longstreth joined FirstEnergy CEO Jones and
Vice President Michael Dowling for two dinners while in the nation’s capital.
At these dinners, both parties laid out their problems—and a potential solution.
Householder relayed his plan to recruit candidates and win the speaker position. And
FirstEnergy executives outlined their need for a legislative bailout. Dowling told Longstreth that
FirstEnergy was “going to be very supportive” of Householder’s bid for speaker, and that
Longstreth needed to set up a 501(c)(4) entity so that Householder could receive “undisclosed
and unlimited contributions.” R. 217, Pg. ID 7636. Householder didn’t ask many questions at
these dinners; it appeared to Longstreth that “he already knew pretty much everything that was
being said.” Id. at Pg. ID 7639. After the D.C. trip, FirstEnergy pledged a million dollars to
Householder’s speaker bid.
Before FirstEnergy could send the money, Householder needed somewhere for it to go.
Following FirstEnergy’s advice, Householder told Longstreth to set up the 501(c)(4), which they
called Generation Now. That entity meant FirstEnergy could give Householder “unlimited
money,” as Longstreth described, and the funds “wouldn’t be traced back” to the company. Id.
at Pg. ID 7644. And because Generation Now wasn’t a political campaign subject to disclosure
requirements, “nobody would ever know” who was giving the funds. R. 302, Pg. ID 12340.

-- 3 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 4
This was the perfect setup for Householder and Longstreth. As the latter put it, Generation Now
would be the “vehicle” to “fund everything that we were trying to do.” R. 217, Pg. ID 7644–45.
About a month after the D.C. dinners, Householder and Jones again discussed the
proposed bailout and the million dollars that Jones had promised. That same day, Longstreth
sent Dowling the wiring instructions for Generation Now—“the organization that Chuck [Jones]
and Larry discussed.” R. 307, Pg. ID 12708. In March, Generation Now received the first of
four $250,000 installments from FirstEnergy. Another installment arrived in May.
A few months later, in July, Longstreth texted Dowling to ask “if there is anything we
can be doing for you guys.” Id. at Pg. ID 12710. Dowling responded, “I know you guys are
there for us.” Id.
Soon after, Longstreth met with Jones and Dowling at a resort in West Virginia. The two
sides took turns discussing their needs. Jones brought up the failing nuclear power plants and
reiterated FirstEnergy’s need for a “state solution”—a “bailout.” R. 217, Pg. ID 7623.
Longstreth, in turn, gave a “very detailed summary of where we stood in the Speaker’s race” and
“where we were with our candidate recruitment.” Id. at Pg. ID 7623–24. The FirstEnergy
executives “wanted to make sure that their donations were being well spent.” Id. at Pg. ID 7625.
After Longstreth had updated them, Jones added: “we have to get Larry in there . . . because I
know he won’t let anything bad happen to us.” Id. at Pg. ID 7624. Longstreth relayed this
conversation to Householder.
A few days later, Longstreth followed up with Dowling “regarding the next donation
installment.” R. 307, Pg. ID 12712. Like clockwork, another $250,000 from FirstEnergy
showed up in the Generation Now account. FirstEnergy wired the final $250,000 of the initial
million-dollar pledge in December 2017.
B. The 2018 Campaign
In the lead-up to the 2018 House election, Householder used FirstEnergy’s money to
amass political power. FirstEnergy, in turn, was cultivating a loyal ally in the Ohio Statehouse.

-- 4 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 5
Householder spent FirstEnergy’s money on Generation Now infrastructure, such as office
space, staff, and consultants. And he paid for various personal expenses, including to settle an
Alabama lawsuit.
Householder also dedicated time and resources to recruiting loyal candidates to run for
the House and support his speaker bid. His priority was to ensure that they “would be loyal and
vote” for him. R. 217, Pg. ID 7653. While Longstreth identified which candidates to support,
Householder had the “final say” in terms of whether they could be counted on. Id. at Pg. ID
7654. Householder then spent the FirstEnergy money on the candidates he had recruited.1
Householder called his political machine “Team Householder.” Id.
While FirstEnergy was bankrolling Team Householder, FirstEnergy Solutions went
bankrupt. After failing to get a bailout from Householder’s predecessor, FirstEnergy Solutions
hired a lobbyist named Juan Cespedes to figure out a path forward. Cespedes’ “first order
of business” was to put the right leadership in place in the Ohio Statehouse. R. 211, Pg. ID
6763–64.
The race for speakership was “most pivotal” to FirstEnergy Solutions. Id. at Pg. ID 6756.
Cespedes knew that Householder had a “close political relationship” with FirstEnergy, and that
he was “very, very good on our issue.” Id. at Pg. ID 6756–57. So, Cespedes obtained a list of
the Team Householder candidates and contributed to their campaigns.
With the 2018 election in sight, Householder (via Generation Now) continued to receive
money from FirstEnergy. In the spring, for example, after a series of calls between Householder
and Dowling, FirstEnergy paid Generation Now $400,000. In the summer, Generation Now
received another half a million from FirstEnergy. Householder was grateful and sent Jones a text
thanking him. Jones was also grateful for Householder’s support and said, “We are rooting for
you and your team!” R. 302, Pg. ID 12348. Householder replied, “I’m rooting for you as
well . . . we are on [the] same team.” Id.
1Householder also funneled money from Generation Now into several other 501(c)(4) entities that, in turn,
spent on these candidates.

-- 5 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 6
Meanwhile, FirstEnergy Solutions’ lobbyists were meeting with Householder to give him
a better sense of the company’s needs. Cespedes, for example, met with Householder in August
2018 to “explain to him what our issues were” and discuss a solution. R. 211, Pg. ID 6767. Bob
Klaffky, another FirstEnergy Solutions lobbyist, accompanied Cespedes. Cespedes recounted
that Householder already knew all about “our issue,” so the men got “granular.” Id. After the
meeting, Householder made clear “what he was expecting from FirstEnergy Solutions going
forward,” asking Klaffky for a “multiple hundred thousand dollar contribution.” Id. at Pg. ID
6768.
Cespedes and Klaffky met with Householder again in October 2018. They discussed the
House races, which were of “extreme importance” to FirstEnergy. Id. at Pg. ID 6780. During
the conversation, Klaffky slid a $400,000 check across the table to Householder. Klaffky
emphasized: “[M]y clients care very much about our issue.” Id. Opening the check,
Householder exclaimed, “well, yes, they do.” Id. at Pg. ID 6781. The men then discussed what
Householder could do for FirstEnergy if he were elected speaker.
FirstEnergy’s decision to hand Householder the check while they discussed the bailout
legislation was intentional. As Cespedes put it, “we were trying to establish the fact that . . . our
support was specifically tied to the legislation.” Id. at Pg. ID 6786.
But Team Householder was blowing through cash faster than FirstEnergy could write
checks. So, Householder and Longstreth arranged yet another meeting with the FirstEnergy
executives. They discussed potential races that the Team Householder candidates could lose,
which would have been “very bad for the whole plan.” R. 217, Pg. ID 7665. Jones’s reaction?
“I’ll help you with whatever you need.” Id. Jones didn’t need to give a reason for his blanket
support. “Everybody there” knew that it stemmed from his desire to receive a “bailout for the[]
nuclear power plants.” Id. at Pg. ID 7665–66. FirstEnergy soon wired another half million to a
Householder-affiliated entity.
Soon after, Cespedes asked Householder’s manager, Longstreth, if Householder was
available to meet. Cespedes had another $100,000 check to give Householder, and Cespedes
wanted to “tie the contributions directly back to our issue.” R. 211, Pg. ID 6790. But

-- 6 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 7
Householder was traveling, so Cespedes left the check with Longstreth and a message for
Householder to call the FirstEnergy Solutions president. Cespedes “wanted to make sure that
[Householder] understood where, in fact, the money was coming from.” Id. at Pg. ID 6791.
Throughout this process, FirstEnergy coordinated with Householder through a lobbyist
named Neil Clark, who styled himself as Householder’s “proxy.” Id. at Pg. ID 6834. Clark
called FirstEnergy “the bank.” R. 302, Pg. ID 12438. It was an “unlimited” source of money,
and Team Householder could turn to it anytime its funds dwindled. Id. at Pg. ID 12453.
All told, the FirstEnergy faucet poured $1.4 million into Generation Now in 2018, as well
as half a million more into other Householder-affiliated entities.
C. Goals Achieved
The bargain paid off. All but one of the Team Householder candidates won their
primaries. Once elected, those representatives all voted for Householder, who became Speaker
in January 2019. Householder texted Jones that evening: “Thank you for everything.” R. 307,
Pg. ID 12798. FirstEnergy also reaped its reward. In his first speech to the new session,
Speaker Householder declared his intention to create an energy-generation subcommittee.
FirstEnergy was thrilled. It became a “matter of when, not if,” the bailout legislation would be
introduced. R. 211, Pg. ID 6801. As Cespedes reflected: “That 500k investment seems very
wise right now . . . this is a good day.” R. 303, Pg. ID 12644. Or as Klaffky put it: “High risk,
high reward.” Id.
Once elected, Householder “went to war” for FirstEnergy. R. 302, Pg. ID 12423. To
start, he created a subcommittee filled with many Team Householder representatives. And over
the next few months, FirstEnergy helped draft bailout legislation. Cespedes would pick up hard
copies of the latest version and hand-deliver them to FirstEnergy executives. The executives
then would “edit, rewrite,” and give the bill back to Cespedes to return to Householder’s office.
R. 211, Pg. ID 6807–08. Householder and FirstEnergy used this courier system because they
wanted no trace of the draft legislation changing hands. This happened a dozen times.

-- 7 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 8
In April 2019, Team Householder members introduced the bailout legislation, known as
House Bill 6, in the Ohio House.
As part of this effort, FirstEnergy retained a lobbyist named Matthew Borges as a
consultant. Borges had known Cespedes (one of the FirstEnergy lobbyists who previously met
with Householder), and Cespedes had kept Borges in the loop about his efforts to get a bailout
deal. So, when the legislative session kicked into gear, Cespedes and FirstEnergy employed
Borges’s help—asking Borges to identify legislators that needed to be persuaded and to suggest
language for the bill.
The bill soon encountered opposition. No matter—Householder’s operation kicked in to
shut it down. After seeing a negative ad, Householder told Jones: “I hope [FirstEnergy
Solutions] is ready for a fight because the first shot was fired at us tonight . . . . Nobody screws
with my members.” R. 302, Pg. ID 12388. Meanwhile, FirstEnergy had retained a media
consulting company to persuade Ohioans of the merits of House Bill 6. But Householder’s right-
hand man, Longstreth, told them to fire the consultant. He demanded FirstEnergy put the money
into Generation Now “if [they] expected to . . . have continued support” of the bailout legislation.
R. 211, Pg. ID 6811–12. Longstreth made clear that these were “the Speaker’s wishes.” Id. at
Pg. ID 6813. FirstEnergy Solutions “had no choice.” Id. at Pg. ID 6815. Team Householder
“were the ones that were going to initiate [and] pass our legislation.” Id.
Over the next two months, FirstEnergy turned the faucet back on, sending approximately
$15 million into Generation Now. Householder and his team used the funds to run
ads supporting House Bill 6. In return for the $15 million, FirstEnergy enjoyed “the full support
of the Speaker.” Id. at Pg. ID 6817. As Jones told Householder, “I would say you are a
bargain—not cheap.” R. 302, Pg. ID 12392.
Householder’s support, though costly, had its rewards. The state House and Senate
passed House Bill 6, and the governor signed it into law in July 2019. FirstEnergy received a
$1.3 billion bailout, which included a fixed revenue stream of $20 to $50 million a year through
2024 for three of FirstEnergy’s utility companies. As Jones told Dowling, “We made a bbiiiiiiig
bet and it paid off.” Id. at Pg. ID 12406.

-- 8 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 9
D. The Referendum
But “there was really no time to celebrate.” R. 211, Pg. ID 6844. Why? Ohio has a
mechanism by which citizens may repeal legislation: a voter referendum. If Ohio citizens
collect around 265,000 signatures in support of repealing a law, they can place the issue on the
ballot for voter approval, so long as the referendum isn’t targeting a tax. Right after Householder
passed the bailout, opponents tried to do just that. They had 90 days to get the signatures if they
wanted to stop the bailout. FirstEnergy and Householder were worried. But just as before the
bailout, both parties found a solution: each other.
Following the launch of the signature campaign, Householder hopped on a call with the
chairman of FirstEnergy Solutions, John Kiani, and assured Kiani that “he would do everything
in his power” to defeat the referendum. Id. at Pg. ID 6846. FirstEnergy was in “good hands”
with Generation Now, Householder emphasized, and he was even “prepared to introduce new
legislation.” Id.
Householder’s promises placated FirstEnergy. Kiani reported to FirstEnergy Vice
President Dowling that he had a “good call” with Householder. R. 302, Pg. ID 12468. And after
Dowling himself spoke with Householder, he texted Kiani: “I think you’re in excellent hands”
given Householder’s “personal involvement and engagement.” Id. at Pg. ID 12466.
Soon after these conversations, tens of millions of FirstEnergy dollars began pouring
into Generation Now. FirstEnergy and Householder kept the “same arrangement” as before—
FirstEnergy providing funds, Generation Now managing the operation. R. 211, Pg. ID 6848–49.
Between August and October 2019, Generation Now received about $38 million from
FirstEnergy and its affiliates. If Householder needed money, he would just call Jones and ask for
it. Following one call, Cespedes texted Longstreth, “CJ [Chuck Jones] $ is in route.” R. 307,
Pg. ID 12960. The next day, a $10,000,000 wire transfer to Generation Now arrived from
FirstEnergy.
Householder and the Generation Now team developed a multi-pronged strategy to defeat
the referendum campaign.

-- 9 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 10
First, Householder and Borges tried to cajole other state officials to help stop the
referendum. As part of this strategy, Householder and Borges each pressured Ohio Attorney
General Dave Yost to interpret House Bill 6 as a “tax,” since taxes aren’t subject to referenda.
They also pressured Yost to reject the petition language so the referendum campaign would have
less time to collect signatures. Yost initially rejected the organizers’ first attempt at a petition.
But the campaign submitted a revised petition, which Yost approved, despite more pressure from
Householder and his allies. In the end, this delay posed a “significant impediment,” since the
campaign had only 54 days to collect the necessary signatures. R. 216, Pg. ID 7494–95.
Second, Householder asked his staff to begin drafting new legislation characterizing
House Bill 6 as a tax that could not be subject to a referendum. That way, if the current bill were
repealed, Householder could just pass another bailout—one that organizers couldn’t thwart.
Soon, Householder had the “‘tax’ bill ready to go.” R. 303, Pg. ID 12580.
Third, Householder had his political machine try to disrupt the signature campaign, which
was run by a group called Ohioans Against Corporate Bailouts (OACB). Borges was the leader
of such efforts. He tried to make it hard for the organizers to get signatures. For one, Borges
hired operatives to research and create media stories about the OACB signature collectors.
The idea was to find out negative information about individual collectors and then run ads to
“discredit their efforts.” R. 212, Pg. ID 6894. Another lobbyist agreed to attempt to “buy[]
out” other signature collection firms so that OACB wouldn’t have anyone to use. Id. at Pg. ID
6906–07. Other Householder affiliates took to the streets, where signature collectors were
“stalked,” “intimidated,” “harassed,” and even “assaulted.” R. 216, Pg. ID 7497. As the CEO of
one signature-collection firm put it, “[i]t was like a war zone out there.” Id.
As part of his efforts to defeat the signature campaign, Borges wanted to monitor how
many signatures OACB collected so he could assess whether the initiative was likely to succeed.
As FirstEnergy recognized, this was “the most important piece of information” at issue. R. 212,
Pg. ID 6901. But the information wasn’t publicly available.
So, Borges tried a different approach. He told Cespedes that he knew one of the
employees at AMT, the firm OACB hired to collect the signatures. His name was Tyler

-- 10 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 11
Fehrman. Borges reported that he had a “long-time relationship” with Fehrman and would be
able to “approach[]” him for information. Id. at Pg. ID 6888. Borges thought Fehrman had
inside information about the signatures AMT collected. So, Borges asked Fehrman to meet for
coffee. At that coffee meeting, Borges told Fehrman that he’d be able to “take[] care” of all
Fehrman’s debt, and “all he needed was information on what was going on related to the House
Bill 6 repeal.” R. 224, Pg. ID 8125–26. Fehrman replied that he needed “time to think about it.”
Id. at Pg. ID 8126. He declined Borges’s offer later that day. Borges replied that Fehrman
shouldn’t tell anyone about their conversation.
Fehrman, shaken by the conversation, contacted the FBI. The FBI asked Fehrman to
cooperate and document future conversations with Borges. Fehrman agreed.
Now acting as a cooperator, Fehrman reached back out to Borges. Over a series of
conversations, Borges explained how he conceived of the effort to defeat the referendum. He
explained that “everyone else in the universe” was on OACB’s side. R. 303, Pg. ID 12537. But
there was an “unholy alliance” between Householder and FirstEnergy that would let people “get
fat off of” the dispute. Id. at Pg. ID 12538. Throughout the conversations, Borges insisted that
the conduct was legal. “[D]on’t steal money from a campaign,” he said, “set up a PAC.” Id. at
Pg. ID 12539. Borges wanted to know how many signatures AMT had collected. He asked that
Fehrman call—not email—with the information. That way, there was “never any record” of
their interaction. Id. at Pg. ID 12540. Borges paid $15,000 for the signature count information
and promised that he’d pay $10,000 once he got the data he requested. Fehrman never sent the
data.
In the end, the referendum campaign didn’t gather enough signatures, so the ballot
initiative failed. That day, FirstEnergy wired Generation Now another $3,000,000. House Bill
6 remained on the books. For Jones, Householder was “an expensive friend.” R. 307, Pg. ID
12723. But it looked like money well spent.
E. Federal Court
That all changed when a grand jury indicted Householder, Longstreth, Clark, Cespedes,
Borges, and Generation Now for participating in a conspiracy in violation of the Racketeer

-- 11 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 12
Influenced and Corrupt Organizations Act, more commonly known as “RICO.” 18 U.S.C.
§§ 1962(c) & (d).
To prove a RICO conspiracy charge, the government must make four showings: (1) that
an enterprise existed; (2) that the enterprise was engaged in, or that its activities affected,
interstate commerce; (3) that the defendant was employed by or associated with the enterprise;
and (4) that the defendant conspired to “conduct or participate, directly or indirectly, in the
conduct of such enterprise’s affairs through a pattern of racketeering activity.” Id.; see also
United States v. Fowler, 535 F.3d 408, 418 (6th Cir. 2008).
Here, the indictment alleged that Householder, Borges, and their affiliates formed an
enterprise. Acting on behalf of that enterprise, the defendants conspired to engage in a pattern of
“racketeering activity” consisting of multiple acts of seven predicate offenses.2 While a
substantive RICO charge requires the government to prove these predicate offenses, the RICO
conspiracy charge doesn’t. United States v. Hills, 27 F.4th 1155, 1174 (6th Cir. 2022). Rather,
the government only must show that the defendant “intended to further an endeavor which, if
completed, would” have resulted in the commission of two predicate offenses. Fowler, 535 F.3d
at 421 (internal quotation marks omitted).
Faced with these charges, Longstreth, Cespedes, and Generation Now pled guilty. Clark
sadly passed away. And Householder and Borges went to trial. After a 26-day trial, a jury found
them guilty.
On appeal, Householder and Borges bring a number of claims challenging their
convictions. Because each defendant’s arguments fail, we uphold their convictions. We address
Householder’s claims first, and then turn to Borges’s.
2The offenses were the following: Public-official honest services fraud and private honest services fraud
(18 U.S.C. §§ 1343, 1346); Hobbs Act extortion (18 U.S.C. § 1951); a Travel Act violation (18 U.S.C. § 1952);
money laundering (18 U.S.C. § 1956); engaging in monetary transactions in property derived from specified
unlawful activity (18 U.S.C. § 1957); and bribery under Ohio Revised Code § 2921.02.

-- 12 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 13
Larry Householder
Householder brings six claims: He argues that the jury instructions were erroneous; that
there was insufficient evidence to convict him; that the trial court twice violated his right to
counsel; that the court erred in admitting various pieces of evidence; that the judge was biased;
and that his sentence was procedurally and substantively unreasonable.
I. Jury Instructions
Householder first challenges the jury instructions. He disputes three aspects of the
instructions: the instructions on bribery under federal law, the instructions on bribery under
Ohio law, and the instructions about a witness’s prior inconsistent statement.
A. Federal-Law Bribery Instructions
Householder challenges the jury instructions that relate to the federal-law bribery
predicates for his RICO charge: Hobbs Act extortion and honest services fraud. The district
court’s instructions complied with the applicable law.
1. The Law
At trial, the government used two federal statutes as RICO predicates: Hobbs Act
extortion (in violation of 28 U.S.C. § 1951) and honest services fraud (in violation of 28 U.S.C.
§§ 1343, 1346).
We start with the required showing for extortion. The Supreme Court has told us that
extortion under color of official right requires that a public official “receive[] a payment in return
for his agreement to perform specific official acts.” Evans v. United States, 504 U.S. 255,
268 (1992). In other words, the Court has required the government to show a quid pro quo
agreement: that the official received a payment (the quid) and in return agreed to take official
action (the quo).
The Court’s quid pro quo requirement also applies to honest services fraud.3 While the
Supreme Court has never explicitly said so, the Court has limited honest services fraud to bribery
3Both parties operate under this assumption.

-- 13 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 14
and kickbacks. Skilling v. United States, 561 U.S. 358, 408–09 (2010). And bribery has long
proscribed an official from receiving something of value “in return for” official action.
18 U.S.C. § 201(b)(2). So, to give adequate notice to defendants, it would make sense to apply
the quid pro quo bribery requirement to honest services fraud. See United States v. Siegelman,
640 F.3d 1159, 1173 n.21 (11th Cir. 2011). Further, the need for a quid pro quo is critical in the
campaign contribution context, where it “is the corrupt agreement that transforms the exchange
from a First Amendment protected” contribution into bribery. Id. (emphasis in original). Thus,
honest services fraud requires a quid pro quo agreement rather than a more lenient showing of
general influence.4
Now, to show a quid pro quo, the government must show a “meeting of the minds and
specific, agreed-upon terms.” United States v. Sittenfeld, 128 F.4th 752, 770 (6th Cir. 2025).
That agreement can be formal or informal, written or unwritten, and express or implied. United
States v. Terry, 707 F.3d 607, 613 (6th Cir. 2013); Evans, 504 U.S. at 274 (Kennedy, J.,
concurring in part and concurring in the judgment). Regardless of the form, the key terms of the
agreement are that the bribe payor gave the gift to obtain a promise, and the bribe recipient made
the promise to obtain the gift. Sittenfeld, 128 F.4th at 770. But that’s not all. For a gift to
become a bribe, “the parties must understand that ‘official conduct will be controlled by’ the
bribe.” Id. (quoting McCormick v. United States, 500 U.S. 257, 273 (1991)). Thus, “the public
official must bind himself with some additional promise that the gift has induced.” Id. That
way, the law doesn’t criminalize giving a gift for something an official has already promised to
do.
Further, since an agreement can be informal, unwritten, or implied, the government can
prove the existence of the quid pro quo with circumstantial evidence. Id. at 771. Otherwise,
anyone could frustrate the law by “knowing winks and nods.” Id. at 769 (quoting Evans, 504
4Other circuits agree. The Ninth Circuit, for one, requires the government to show a quid pro quo when
proving honest services fraud in the form of bribery. United States v. Inzunza, 638 F.3d 1006, 1013 (9th Cir. 2011).
And several other circuits have assumed without deciding that the quid pro quo requirement applies to honest
services fraud. See, e.g., United States v. Benjamin, 95 F.4th 60, 67–68 (2d Cir. 2024); see also id. at 68 n.2
(collecting cases from the Eleventh, Seventh, and D.C. Circuits).

-- 14 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 15
U.S. at 274 (Kennedy, J., concurring in part and concurring in the judgment)). Thus, a jury can
infer an agreement from what the participants “say, mean and do.” Terry, 707 F.3d at 613.
In sum, Hobbs Act extortion and honest services fraud require the government to show a
quid pro quo—that is, the official received money in exchange for a promise to take specific
official action. See Sittenfeld, 128 F.4th at 770. There must be an unambiguous meeting of the
minds between the official and the payor. But the ways the government can go about showing
this agreement are manifold.
2. The Instructions
Considering that legal framework, Householder challenges three aspects of the federal-
law bribery instructions: (i) the instruction on the agreement necessary to convict him; (ii) the
instruction on the timing of the payments vis-à-vis the official action; (iii) and the instruction on
bribery by implication. Householder’s arguments fail.
i. Agreement Instructions
First, Householder argues that the instructions erroneously defined the agreement
necessary for a quid pro quo. He claims the instructions allowed the jury to convict if
he received money “knowing that the expectation was legislation in return.” Appellant Br. at
28 (quotation omitted). According to Householder, an “expectation” isn’t the “agreement”
necessary to show a quid pro quo. Id.
But the instructions didn’t erroneously describe the agreement requirement. Here, the
jury was told that bribery could include either (1) “a public official’s solicitations of things of
value in exchange for performing or agreeing to perform specific official action” or (2) “a public
official’s receipt of things of value when the public official knows that the person who gave the
thing of value was doing so in return for the public official performing or agreeing to perform a
specific official action.” R. 237, Pg. ID 9421.
These instructions come straight from the Supreme Court’s holding in Evans: an official
commits extortion when he “has obtained a payment to which he was not entitled, knowing that
the payment was made in return for official acts.” 504 U.S. at 268. The instruction below

-- 15 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 16
matched this language. It defined extortion as an official’s receipt of something of value
“know[ing]” it was given “in return for the public official performing or agreeing to perform a
specific official action.” R. 237, Pg. ID 9421. An instruction can’t misrepresent the law when it
hews so closely to Supreme Court precedent.5
Finally, the instructions also mirrored our circuit’s pattern jury instructions. See Sixth
Circuit Pattern Jury Instr. 17.02(1)(C) (including as an element of Hobbs Act extortion “that the
defendant knew the property was being obtained [accepted] [taken] [received] in exchange for an
official act” (alteration in original)). That’s relevant, because whether jury instructions track the
pattern instructions is “one factor in determining whether any particular instruction is misleading
or erroneous.” United States v. Damra, 621 F.3d 474, 499 (6th Cir. 2010). And here, the pattern
instructions draw on the Supreme Court’s pronouncement in Evans. This grounding in caselaw
means that we must give “deference” to the pattern instructions—and, by extension, the
agreement instructions here. United States v. Frei, 995 F.3d 561, 565 (6th Cir. 2021).
We thus find no error in the agreement instructions.
ii. Timing Instructions
Householder’s next objection is that the court refused to give his requested instruction
that the “quid pro quo agreement must exist at the time . . . the bribe was paid. It cannot be
formed later.” R. 237, Pg. ID 9354. Householder maintains that the failure to give this
instruction meant that the jury could have convicted him for agreeing to take any unspecified
action at any time.
But the instructions did not leave open that possibility. They specified that a public
official “intended to exchange a thing of value from the payor for a specific official action . . . or
that the public official knew the payor intended to exchange the thing of value for a specific
official act.” Id. at Pg. ID 9422–23 (emphases added).
5Householder argues that this loose definition of bribery is inconsistent with recent Supreme Court
precedent in the campaign finance realm. But even if it is, we’re not at liberty to disregard the Court’s holding in
Evans. If a Supreme Court precedent applies, “yet appears to rest on reasons rejected in some other line of
decisions,” we must “follow the case which directly controls.” Rodriguez de Quijas v. Shearson/Am. Express, Inc.,
490 U.S. 477, 484 (1989). After all, only the Supreme Court may overrule its decisions. Id.

-- 16 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 17
As above, these instructions parrot Evans. Evans held that a bribery offense is completed
“when the public official receives a payment in return for his agreement to perform specific
official acts.” 504 U.S. at 268. What was sufficient in Evans is sufficient here. Thus, the
instructions demanded the necessary link between FirstEnergy’s payments and Householder’s
conduct.
Relatedly, Householder objects to the instruction that the jury could convict if “the public
official understood the agreement was to take a specific official action on the payor’s behalf
when the opportunity presented itself.” Appellant Br. at 30 (quoting R. 237, Pg. ID 9423). But
we recently said this exact same “as opportunities arise” bribery instruction was proper. Hills,
27 F.4th at 1179. Householder cannot overcome this precedent.
The two cases he relies on also don’t help. One upheld the exact language he challenges.
United States v. Silver, 948 F.3d 538, 558–59 (2d Cir. 2020). And the second, United States
v. Skelos, featured a far broader instruction than the one here: it required only that the defendant
“be expected to perform official acts in exchange for . . . property.” 988 F.3d 645, 656 (2d Cir.
2021) (quotation omitted). Critically, this “left open the possibility that the jury could convict
even if Skelos was expected to take official action on any question or matter in return.” Id.
(emphasis in original). The Skelos instruction had no link to a specific official action.
All told, under any theory, Householder can’t show error.
iii. Bribery by Implication
Householder also argues that the instructions allowed for a conviction based on “bribery
by implication.” Appellant Br. at 31–32. Householder is right that they did.
But blackletter law holds that inferential and circumstantial evidence may support a
bribery conviction. That is, a bribery showing can rest on “an agreement . . . which can be
formal or informal, written or oral.” Terry, 707 F.3d at 613. To be sure, the agreement must be
“unambiguous” from the perspective of the payor and recipient; both parties must know the
terms of the proposed arrangement. See Sittenfeld, 128 F.4th at 772 & n.8. “But the existence of
that agreement is governed by the reasonable doubt standard and can be proved with

-- 17 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 18
circumstantial evidence.” Id. at 771 (emphasis in original). In other words, the government can
point to the usual types of evidence to show that both parties knew of an unambiguous
agreement. Id.; see also Benjamin, 95 F.4th at 68 (joining “[e]very other circuit to have
considered th[e] question” to hold that the “explicit quid pro quo requirement may be met by
implication from the official’s and the payor’s words and actions and need not entail an express
statement”). Householder’s challenge thus fails as a matter of precedent.
B. State Law Bribery & Harmless Error
Next, Householder challenges the jury instructions for the Ohio bribery charge. See Ohio
Rev. Code Ann. § 2921.02(B). Householder argues that the jury instructions defining Ohio
bribery flout the First Amendment because they didn’t require a quid pro quo. This showing,
Householder argues, is necessary for the government to turn otherwise legal campaign
contributions into an illegal bribe. While this challenge has merit, any error was harmless.
Here, the court instructed the jury that an Ohio state bribery conviction required three
findings: (1) that a “public servant . . . knowingly solicited or accepted for himself any valuable
thing or valuable benefit;” (2) that he “intended the valuable thing or benefit to corrupt or to
improperly influence him;” and (3) “that the corruption or influence was with respect to the
discharge of his duties as a public servant.” R. 237, Pg. ID 9434–35.
The Ohio bribery law, as defined in the jury instructions, thus criminalized less than a
quid pro quo. A contribution could be treated as illegal solely because the official “intended the
valuable thing or benefit to corrupt or to improperly influence him.” Id.
And that’s a problem. Why? If the FirstEnergy payments were campaign contributions,
then the only permissible ground for restricting them is to prevent “‘quid pro quo’ corruption or
its appearance.” FEC v. Cruz, 596 U.S. 289, 305 (2022). So, if the Ohio state law bribery
instructions allowed for a conviction for improper influence alone, those instructions were
erroneous.6
6The government points out that Householder used some of the money for personal expenses. Those
contributions don’t enjoy First Amendment protection. But the funds spent on Generation Now and Householder’s
other political entities are still covered by the First Amendment.

-- 18 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 19
But erroneous jury instructions are harmless if we can conclude, beyond a reasonable
doubt, that the jury would have convicted Householder of a RICO conspiracy anyway. See
Neder v. United States, 527 U.S. 1, 18 (1999). Applying that standard, a jury would have had to
convict Householder of a quid pro quo.
Here, the jury would have convicted regardless of the instructions on Ohio bribery. The
evidence was more than enough to show a quid pro quo—we know this because it was sufficient
to show just that under both the extortion and honest services fraud theories. And no party
argued that Ohio bribery would require anything less.
On the contrary, both parties characterized all three public-corruption predicates as
requiring a quid pro quo agreement. At closing, for example, the government stressed that the
public-corruption predicates turned on whether Householder “solicited or received money
knowing it was given in return for specific official action.” R. 238, Pg. ID 9485. Defense
counsel told the jury the same thing: “a bribery case requires that quid pro quo, an explicit
agreement, this for that.” Id. at Pg. ID 9587. And defense counsel understood the government’s
theory to require a quid pro quo. Indeed, on our count, the parties mentioned the phrase “quid
pro quo” ten times during closing. We can’t find a single instance where any party suggested the
Ohio predicate had a lesser standard. Thus, we can conclude beyond a reasonable doubt that,
had the jury been told Ohio bribery required a quid pro quo, it would’ve found Householder
guilty.
C. Prior Inconsistent Statement
Householder’s final challenge to the jury instructions involves an instruction on an
alleged inconsistency between a witness’s testimony and previous statement to a reporter.
This challenge involves a specific incident. Juan Cespedes, one of FirstEnergy’s
lobbyists, testified about a meeting where Householder received a $400,000 check from
FirstEnergy. Cespedes said that a second lobbyist, Robert Klaffky, had slid a check across the
table to Householder and stated, “my clients care very much about our issue.” R. 211, Pg. ID
6780. And the defense hoped to rebut that testimony by calling Klaffky. But after Cespedes
testified, Klaffky received a call from a reporter, who asked about Cespedes’ testimony. Klaffky

-- 19 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 20
responded that he did not “recall saying any of those things, but I’m not saying that I didn’t [do
or say that].” R. 228, Pg. ID 8454 (citation omitted).
After that call, Klaffky took the stand. On direct examination by Householder’s counsel,
Klaffky claimed that Householder and FirstEnergy did not reach an agreement during the
October 2018 meeting. So, during cross-examination, the government brought up Klaffky’s
previous statement to the reporter that he couldn’t recall what was discussed at the meeting.
Later on, the judge gave the jury the following instruction: “You have heard the
testimony of Robert Klaffky. You have also heard that before this trial he made a statement that
may be different from his testimony here in court. The earlier statement was brought to your
attention only to help you decide how believable his testimony was.” R. 237, Pg. ID 9444–45.
Householder challenges the court’s decision to apply this instruction to Klaffky’s
allegedly inconsistent statements. We review the district court’s decision for abuse of discretion.
United States v. Henderson, 2 F.4th 593, 597 (6th Cir. 2021).
The court didn’t abuse its discretion. Klaffky’s denial of any “pay to play” activity
conflicted with his earlier failure to recall what occurred at the October meeting. R. 228, Pg. ID
8426, 8454. Because a reasonable jury could find these two statements inconsistent, the court
didn’t abuse its discretion in giving an instruction on the use of prior inconsistent statements.
See United States v. Foster, 376 F.3d 577, 591 (6th Cir. 2004).
In response, Householder says the government had to admit Klaffky’s prior statement
into evidence. He cites United States v. Toney, where the court found use of prior inconsistent
statements improper because “there was no evidence before the jury that [the witness] had made
the inconsistent statements.” 161 F.3d 404, 410 (6th Cir. 1998). Here, by contrast, Klaffky
admitted on the record that he made the statement to the reporter. Thus, the government didn’t
need to separately admit the statement; it was “admitted” into evidence through Klaffky’s
testimony.
We therefore deny Householder’s challenge.

-- 20 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 21
* * *
In sum, all of Householder’s challenges to the jury instructions fail.
II. Sufficiency of the Evidence
Next, Householder brings a sufficiency-of-the-evidence challenge to his conviction. This
argument fails.
The government alleged that Householder’s enterprise committed dozens of racketeering
acts. At closing, the government divided these alleged acts into three broad categories: public-
official bribery, private-citizen bribery, and money laundering.
Householder challenges the sufficiency of the evidence underlying all three. To succeed,
he bears a “very heavy burden.” United States v. Robinson, 99 F.4th 344, 354 (6th Cir. 2024)
(internal quotation marks omitted). His conviction stands if “any rational trier of fact could have
found the elements of the crime beyond a reasonable doubt.” Id. at 353 (internal quotation marks
omitted). In assessing Householder’s challenge, we must “view[] the evidence in the light most
favorable to the government” and “draw all available inferences and resolve all issues of
credibility in favor of the jury’s verdict.” Id. (internal quotation marks omitted). We address
each category of racketeering activity in turn.
A. Public-Official Bribery
The evidence showed that Householder agreed to commit—and did commit—extortion
and honest services fraud. That is, he solicited and received millions of dollars from FirstEnergy
in exchange for passing the bailout legislation and saving that bailout from a voter referendum.
Start at the beginning. A rational jury could infer that the corrupt bargain began at the
D.C. dinners in January 2017. Longstreth told the jury that at these dinners, Householder
outlined to the FirstEnergy executives “the entire plan of how he was going to get elected
speaker.” R. 217, Pg. ID 7638. The FirstEnergy folks, in turn, laid out their “need to do
something at the state level” about the financial failings of FirstEnergy Solutions. Id. The jury
then learned that Longstreth set up Generation Now so Householder could receive “undisclosed
and unlimited contributions,” and FirstEnergy pledged a million dollars to it. Id. at Pg. ID 7636.

-- 21 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 22
Throughout 2017 and 2018, as Householder recruited candidates to run and vote for him
as speaker, he repeatedly turned to FirstEnergy to fund those races. The jury heard that
FirstEnergy was the “bank,” and the account was “unlimited.” R. 302, Pg. ID 12438, 12453.
During the lead-up to the 2018 election, FirstEnergy gave around $3 million to Generation Now
and its subsidiary entities.
To be sure, a “flow of benefits from one person to a public official . . . does not by itself
establish bribery.” Terry, 707 F.3d at 615. Rather, the benefits “must be part and parcel of an
agreement by the beneficiary to perform public acts for the patron.” Id.
But evidence abounds that Householder and FirstEnergy knew that this funding was in
exchange for bailout legislation if Householder were elected. As Longstreth recounted at trial,
Householder knew all about FirstEnergy’s financial troubles at the Washington dinners.
Householder also wanted updates as the election drew nearer. For example, Dowling texted
Jones, “Larry wants to hear about us – status of company, what’s important to us this year and
next year.” R. 302, Pg. ID 12345. And when Jones texted Householder, “We are rooting for you
and your team,” Householder replied, “we are on [the] same team.” Id. at Pg. ID 12348.
Householder’s commitment to FirstEnergy’s financial health also left an impression on
Jones and Dowling. As Dowling put it, “I know you guys are there for us.” R. 307, Pg. ID
12710. Or, as Jones confided to Longstreth, “I know [Householder] won’t let anything bad
happen to us.” R. 217, Pg. ID 7624. And, when Householder and Longstreth warned Jones a
month before the election that they were having “some trouble,” Jones offered, “I’ll help you
with whatever you need.” Id. at Pg. ID 7664–65. As Longstreth told the jury, Jones was all-in
because FirstEnergy was “in need of the state solution, the bailout for their nuclear power
plants.” Id. at Pg. ID 7665–66. FirstEnergy then wired another $500,000 to one of
Householder’s entities.
Meanwhile, FirstEnergy Solutions lobbyists were meeting with Householder. As one of
those lobbyists, Cespedes, testified, he gave checks to Householder in person to establish that
“our support was specifically tied to the [bailout] legislation.” R. 211, Pg. ID 6786. Having face

-- 22 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 23
time with Householder was critical to FirstEnergy Solutions. It allowed the lobbyists to “tie the
contributions directly back to our issue.” Id. at Pg. ID 6790.
The meetings also were productive. It was “obvious” that Householder knew all about
FirstEnergy’s financial woes. Id. at Pg. ID 6767. And Householder was “very affirmative” in
his support of FirstEnergy. Id. at Pg. ID 6783. He hashed out the details of what he could do in
the legislature once he became speaker.
The jury then heard how Householder “went to war” for FirstEnergy once elected
speaker. R. 302, Pg. ID 12423. He created a subcommittee and filled it with many Team
Householder members. He met with FirstEnergy Solutions lobbyists and “establish[ed] a
timeline” for the legislation. R. 211, Pg. ID 6799–6801. And he shuttled the draft legislation
back and forth with FirstEnergy so they could “edit” and “rewrite” the proposed bill. Id. at Pg.
ID 6806–08.
But the FirstEnergy payments weren’t over. When it became apparent that the proposed
legislation, House Bill 6, would encounter opposition, Longstreth made clear to Cespedes that
FirstEnergy needed to keep paying Generation Now if it expected to have the “continued
support” of Speaker Householder. Id. at Pg. ID 6811–12. And Longstreth left no doubt: “this
was the Speaker’s wishes.” Id. at Pg. ID 6813. Cespedes testified that FirstEnergy “had no
choice” but to comply. Id. at Pg. ID 6815. In return, FirstEnergy “would get the full support of
the Speaker and make sure this legislation was passed.” Id. at Pg. ID 6817.
Nor did FirstEnergy turn off the faucet once the legislation passed. In response to the
referendum campaign to repeal House Bill 6, FirstEnergy contributed around $35 million to
Householder via Generation Now. The jury heard that Householder promised the FirstEnergy
Solutions chairman to “do everything in his power to help defeat the referendum.” Id. at Pg. ID
6846.
What did Householder do to keep his promise? For one, he pressured Attorney General
Yost to reject the petition language. And Householder asked his staff to begin drafting new
legislation should the initiative succeed; the new legislation was “ready to go” before the
signature-collection period was over. R. 303, Pg. ID 12580.

-- 23 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 24
The jurors also heard no shortage of evidence that the conspirators knew that they were
doing wrong. Clark, for one, described the relationship of Householder and FirstEnergy as “pay
to play.” R. 302, Pg. ID 12424. Borges likened FirstEnergy’s payments to “Monopoly money”
and described the relationship between Householder and FirstEnergy as an “unholy alliance.” R.
303, Pg. ID 12538, 12554. If “people [were] going to get fat off of this,” Borges mused,
“why . . . not us.” Id. at Pg. ID 12538. As Cespedes told Borges after one set of payments:
“Who would ever assume a bankrupt company is willing to spend 15 million. What a joke?
LOL.” R. 211, Pg. ID 6826. Or as Klaffky called FirstEnergy’s ploy: “High risk, high reward.”
R. 303, Pg. ID 12644. And when House Bill 6 passed, Jones bragged to Dowling that their
“bbiiiiiiig bet” had “paid off.” R. 302, Pg. ID 12406. Dowling’s response? “Huge bet.” Id.
What’s more, Householder tried to conceal his tracks along the way. It began with the
web of secret 501(c)(4) entities. He tried to cajole another representative into deleting text
messages about House Bill 6. He deleted his call logs with Yost during the referendum saga.
And he gave “unequivocally false” testimony, according to the district court. R. 285, Pg. ID
11182.
* * *
In sum, Householder committed multiple RICO predicates when he solicited and received
payments from FirstEnergy in exchange for specific official action. See 18 U.S.C. §§ 1343,
1346; 18 U.S.C. § 1951; Ohio Rev. Code Ann. § 2921.02. It was reasonable for a jury to
conclude that he received money from FirstEnergy to finance a political machine that would
elect him to the speakership; passed a billion-dollar bailout for FirstEnergy in return for those
contributions; and solicited more contributions in exchange for agreeing to take official action to
preserve that bailout.
Householder’s Counterarguments
Householder offers a slate of arguments to the contrary. They fail.
His first argument sounds in law: he maintains that in the context of campaign
contributions, the government can only show a quid pro quo through “unambiguous evidence of

-- 24 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 25
a corrupt agreement—evidence that cannot be explained as ordinary electoral politics.”
Appellant Br. at 46.
But that’s not the governing standard. Rather, “the government’s evidence need not rule
out all reasonable, alternative hypotheses to guilt.” Sittenfeld, 125 F.4th at 771. We trust juries
to “parse words and actions to discern the intent behind them,” even in the context of campaign
contributions. Id. To be sure, the terms of the agreement must be “unambiguous” to the parties
involved. Id. But “the existence of that agreement is governed by the reasonable doubt
standard.” Id. (emphasis in original). The lack of a heightened requirement makes sense:
Congress did not write into bribery statutes an unambiguous-evidence exception for public
officials who receive campaign contributions.
Householder cites one case to the contrary. It’s United States v. Benjamin, No.
21-CR-706, 2022 WL 17417038 (S.D.N.Y. Dec. 5, 2022), rev’d and remanded, 95 F.4th 60 (2d
Cir. 2024). Householder cites the district court’s opinion, which said an agreement predicated on
campaign contributions must be “clear and unambiguous.” Id. at *12. But the Second Circuit
reversed on appeal: joining “[e]very other circuit to have considered the question,” it held that
the “explicit quid pro quo requirement may be met by implication from the official’s and the
payor’s words and actions and need not entail an express statement.” Benjamin, 95 F.4th at 68.
Thus, this case isn’t helpful to Householder.
Householder also claims that the evidence had a “plausible explanation sounding in
politics.” Appellant Br. at 47. Householder points us to his “longtime support” of public utility
companies. Id. at 50. But Householder pointed the jury to this exact alternative at trial. Indeed,
his attorney began closing statements by emphasizing Householder’s “long-held political views
regarding the importance of energy generation in Ohio.” R. 238, Pg. ID 9541. We can’t disturb
the jury’s decision rejecting this explanation.
Householder further disputes Longstreth’s testimony about the Washington, D.C. dinners
because Jones’s travel records show that Jones wasn’t in Washington for the first dinner and was
elsewhere at the time of the second dinner. But this argument fails. For starters, the “credibility
of a trial witness is not relevant to our determination of the sufficiency of the evidence to support

-- 25 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 26
a conviction.” Hills, 27 F.4th at 1176. Householder’s attorneys were free to—and did—cross-
examine Longstreth about these dinners and the alleged inconsistency with Jones’s travel
records. Further, it’s irrelevant if Jones himself wasn’t at these dinners. Householder doesn’t
dispute that other FirstEnergy executives, including Vice President Dowling, attended.
In addition, Householder claims that the government failed to establish that either
Householder or the FirstEnergy executives “promised to do anything” at these dinners.
Appellant Br. at 49. But the jury heard subsequent evidence suggesting otherwise, from the
immediate creation of Generation Now to the repeated conversations between Householder and
FirstEnergy executives about contributions and potential legislation.
Householder also takes issue with the government’s emphasis on the size of his
contributions, which came out to just under $60 million. He argues that the size alone isn’t
evidence of a quid pro quo. But a jury could have reasonably inferred from this exorbitant
amount of money that FirstEnergy paid Householder in exchange for bailout legislation. Indeed,
a jury could find it unlikely that (1) FirstEnergy would squander tens of millions of dollars amid
financial difficulties with no promise of anything in return, and (2) that Householder would go
“to war” to pass a controversial, billion-dollar bailout in return for chump change. R. 302, Pg.
ID 12423.
The next objection is that the official action was not performed “at the time” that he
received the contributions. Appellant Br. at 52–53. Thus, Household argues, he didn’t agree to
take action in return for contributions. But “fulfillment of the quid pro quo is not an element of
the offense.” Evans, 504 U.S. at 268. Rather, “the offense is completed at the time when the
public official receives a payment in return for his agreement to perform specific official acts.”
Id. (emphasis added); see also McDonnell v. United States, 579 U.S. 550, 572 (2016) (reiterating
that a “public official is not required to actually make a decision or take an action”). Thus, the
government could show that Householder began receiving money in March 2017 even though he
didn’t act until January 2018. All that mattered was that Householder agreed to take official
action when he began receiving the funds.

-- 26 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 27
Finally, Householder cites United States v. Menendez for the proposition that the
government “utterly failed to connect the quid to the alleged quo.” Appellant Br. at 52 (citing
291 F. Supp. 3d 606 (D.N.J. 2018)). But Menendez had a critical difference: there was no
evidence the defendants “knew the terms of the quid pro quo” described in the indictment.
Menendez, 291 F. Supp. 3d at 630. Not so here. The government presented abundant evidence
that Householder knew FirstEnergy’s payments were in exchange for official action in passing
and preserving the bailout.
All told, none of Householder’s arguments justifies overturning the jury’s verdict.
B. Money Laundering
As Householder acknowledges, the sufficiency of the evidence supporting the money
laundering predicates depends on whether the payments from FirstEnergy to Generation Now
constitute bribery proceeds. Since they do, the money laundering convictions also stand.
C. Private-Citizen Bribery
Finally, Householder argues that there was insufficient evidence of the predicate act for
private-citizen bribery—Borges’s $15,000 bribe to Tyler Fehrman for inside information on the
signature-collection campaign. But the government never argued that the predicates relating to
private-citizen bribery applied to Householder. So, we need not address these arguments.
* * *
In sum, Householder’s sufficiency-of-the-evidence challenge fails. There was ample
evidence for the jury to conclude that Householder solicited and received millions of dollars
from FirstEnergy in return for passing and then preserving bailout legislation. We won’t disturb
that decision.
III. Right to Counsel
Householder next alleges two Sixth Amendment violations at trial. Neither succeeds.

-- 27 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 28
A. Dismissal of a Juror
Householder argues that the court’s dismissal of a juror without consulting his counsel
deprived him of his Sixth Amendment rights. He is incorrect.
A few days into trial, a juror tested positive for COVID-19. The court ordered a five-day
recess and requested that all jurors obtain a negative COVID test. But the afternoon before the
recess was set to end, the court emailed Householder’s counsel about a related issue: another
juror still hadn’t received his COVID test, did not know when he would receive it, and was not
inclined to take another test. This juror also refused to wear a mask in defiance of a new court
order and would need to be cut anyway if the deliberations lasted into early March. Thus, the
court told counsel, it had decided to remove this juror now, and to return the next day with
15 jurors.
Householder disputes whether the court gave notice to counsel before dismissing the
juror. But that’s irrelevant because the court didn’t need to confer with counsel before
dismissing the juror.
Courts may “replace any jurors who are unable to perform or who are disqualified from
performing their duties.” Fed. R. Crim. P. 24(c)(1). In so doing, the “consent of the parties is
not needed if the district court has ‘reasonable cause’ to replace the juror.” United States v. De
Oleo, 697 F.3d 338, 341 (6th Cir. 2012) (internal quotation marks omitted). When, as here, the
court has “thoroughly explained its concerns,” we defer to its decision. United States v. Helton,
35 F.4th 511, 523 (6th Cir. 2022).
As the court made clear in its email to counsel, it had three reasons to dismiss the juror:
his refusal to obtain a negative COVID test, his refusal to obey court orders by not wearing a
mask, and the possibility that he might not be available during the trial. The court didn’t err by
identifying these as grounds for dismissing the juror.
B. Speaking with Counsel
Householder brings another Sixth Amendment challenge, but he brought it too late.
More than three months after filing his opening brief, Householder moved to file a supplemental

-- 28 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 29
brief. He provided no legitimate procedural basis for his late filing and did not explain why he
couldn’t have made this argument earlier. He’s therefore forfeited it. See Island Creek Coal Co.
v. Wilkerson, 910 F.3d 254, 256 (6th Cir. 2018) (reiterating that an argument is “forfeited when it
was not raised in the opening brief” (quotation omitted)).
But even if we were to address Householder’s contention, it would fail. Householder
claims the district court prevented him from speaking to counsel during an overnight recess and
therefore violated his Sixth Amendment rights. But the court actually said: “Mr. Householder,
you’re not to discuss your testimony during the break. You can speak to your attorneys however,
understood?” R. 228, Pg. ID 8611. The most natural reading of this instruction is that
Householder couldn’t discuss his testimony during the overnight recess, except with his counsel.
Indeed, after this instruction, defense counsel did not ask for clarification. They evidently
understood that the district court meant what it said: Householder couldn’t talk about his
testimony but could speak to his attorney. Therefore, even if we were to reach Householder’s
claim, we would reject it.
IV. Evidentiary Rulings
Householder next disputes two evidentiary rulings at trial. He claims that the evidence
was irrelevant and unfairly prejudiced him.
A. Recordings
Householder argues that the district court erred by admitting two excerpts from recorded
conversations that he had with his co-conspirator, Neil Clark. He is wrong.
Householder took the stand in his own defense. When asked why he sought to return to
the Ohio House, Householder claimed that he “was discouraged by the divisiveness there was in
politics and frankly in the Ohio House of Representatives.” R. 228, Pg. ID 8494. The
government then played a short clip of a conversation between Householder and Clark in which
Householder proclaimed, “We like war, you know that, Neil?” Householder then asked Clark if
they should “make some kind of a movement on [two opposing politicians] just to . . . say that if

-- 29 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 30
you’re going to f*** with me, I’m going to f*** with your kids.” Def. App’x, Gov’t Ex. 913.
Defense counsel didn’t object to the airing of this recording.
Householder also testified that he tracked which donors supported him or his opponent
but denied that there were “consequences” for those that supported his opponent. R. 229, Pg. ID
8739–40. The prosecutor then played another brief clip in which Householder mentioned his
opponent’s donors and recommended “f***[ing] them over later.” Def. App’x, Gov’t Ex. 906.
This time, defense counsel did object.
Householder challenges the admission of these recordings as irrelevant and unfairly
prejudicial. Evidence is relevant if it “has any tendency to make a fact more or less probable
than it would be without the evidence” and “the fact is of consequence in determining the
action.” Fed. R. Evid. 401. This standard is “extremely liberal.” Dortch v. Fowler, 588 F.3d
396, 400 (6th Cir. 2009) (quotation omitted). The recordings easily meet that standard.
As the government argues, the recordings were relevant to undermine Householder’s
assertion that he “sought the Speakership to build bridges in politics.” Appellee Br. at 90.
Householder cites one case where introduced recordings were found to be irrelevant, but that
case is distinguishable. There, the government introduced recordings of a defendant using
“deeply offensive racist and misogynistic language.” United States v. Hazelwood, 979 F.3d 398,
402 (6th Cir. 2020). But the underlying charges involved mail fraud, and the recordings had
nothing to do with that. See id. Here, however, the recordings undermined Householder’s
claims about entering politics to lower the political temperature. The recordings thus were “of
consequence” in determining Householder’s political motives. See Fed. R. Evid. 401.
Householder’s argument about unfair prejudice also fails. A court “may exclude relevant
evidence if its probative value is substantially outweighed by a danger of . . . unfair prejudice.”
Fed. R. Evid. 403. There wasn’t any unfair prejudice here, however. Householder’s unpalatable
language aligned with lots of evidence that the jury received of Householder and his co-
conspirators’ foul language. These recordings wouldn’t have unduly prejudiced him. The
recordings also were inconsequential in the context of the whole trial; overwhelming evidence
demonstrated that Householder solicited and received money from FirstEnergy in exchange for

-- 30 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 31
official action. The probative value of these recordings, then, was not “substantially
outweighed” by the danger of unfair prejudice. Id.
B. Co-Conspirator Testimony
Householder’s other evidentiary challenge relates to the introduction of his co-
conspirators’ guilty pleas. This challenge also fails.
At trial, the government put Longstreth and Cespedes, who had both pled guilty, on the
stand. Householder and his co-defendant, Borges, agreed not to impeach Longstreth and
Cespedes with their plea agreements. But the government still elicited evidence from Longstreth
and Cespedes that they had pled guilty. Householder argues that this was prejudicial error.
Usually, guilty pleas “may be introduced into evidence if the co-conspirator or
co-defendant testifies at trial, so that the factfinder will have appropriate facts on hand to assess
the witness’s credibility.” United States v. Sanders, 95 F.3d 449, 454 (6th Cir. 1996). Here,
Householder argues that because he stipulated not to impeach the witnesses with their guilty
pleas, there was no need to bolster their credibility.
Householder is wrong. One of the jury’s primary tasks is to assess witness credibility.
See United States v. Kelly, 204 F.3d 652, 656 (6th Cir. 2000). The existence of a guilty plea goes
to that question. While our circuit hasn’t spoken on whether a defendant who stipulates that he
won’t attack a co-conspirator’s motives should still face testimony about the co-conspirator’s
guilt, at least three other circuits have upheld testimony about guilty pleas under similar
circumstances. See United States v. Montani, 204 F.3d 761, 766 (7th Cir. 2000); United States
v. Universal Rehab. Servs., 205 F.3d 657, 666 (3d Cir. 2000) (en banc); United States v. Kroh,
915 F.2d 326, 331 (8th Cir. 1990) (en banc). For good reason. Jurors would “naturally and
unavoidably wonder why” a witness was testifying about crimes in which he participated.
Montani, 204 F.3d at 766. So, by introducing the witness’s guilty plea, the government “does

-- 31 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 32
not ‘bolster’ the witness’s credibility”; rather, the government is answering a question that
“unavoidably would be raised in the jurors’ minds.” Id.7
Householder responds that he wasn’t challenging Longstreth or Cespedes’ credibility.
But jurors are instructed to evaluate the credibility of all witnesses who testify. “Indeed, they are
so instructed even in the absence of an affirmative challenge to witness credibility.” Universal
Rehab. Servs., 205 F.3d at 666. Case in point: the court here instructed the jury “to decide how
credible or believable each witness was.” R. 237, Pg. ID 9400.
Householder also argues that criminal convictions can be used only to attack a “witness’s
character for truthfulness.” Appellant Br. at 63 (citing Fed. R. Evid. 609(a)). But this point is
inapposite—the government wasn’t impeaching its own witness. Rather, the government was
providing a full picture of the witness it put on the stand. And a party is entitled “to extract the
complete testimony of his witness . . . rather than be forced to leave gaping holes.” United States
v. LeFevour, 798 F.2d 977, 983 (7th Cir. 1986) (Posner, J.). A rule about impeachment says
nothing about the government’s actions here.
In any event, the introduction of the guilty pleas wasn’t prejudicial. Longstreth and
Cespedes testified at length about the conspiracy and their role in it. The mere fact that they pled
guilty was inconsequential considering their exhaustive testimony. And the court gave the
proper limiting instruction that the jury could not consider the pleas as substantive evidence of
Householder’s guilt. See Sixth Cir. Inst. 7.08(3).
We thus find no error.
V. Judicial Bias
Householder further argues that the judge should have recused himself, and that the
judge’s failure to do so violated the Due Process Clause. Why? On the third day of trial,
Householder’s counsel mentioned that Householder had raised money for a 501(c)(4) “that was
7Faced with this consistent direction from our sister circuits, Householder cites almost no law holding that
a promise not to impeach should matter. The best he has are two unreported district court cases. See Appellant Br.
at 62 (citing United States v. Clark, No. 1:19-CR-148, 2020 WL 830057, at *14–15 (N.D. Ohio Feb. 20, 2020);
United States v. Nosal, No. CR-08-0237, 2013 WL 11327121, at *8 (N.D. Cal. Mar. 29, 2013)). That’s insufficient
to overcome consistent circuit-court caselaw.

-- 32 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 33
critical” of the judge’s Ohio Supreme Court campaign in 2000. R. 197, Pg. ID 4949. Counsel
then asked “if the Court holds personal animosity” towards Householder, and the judge
responded that he didn’t. Id.
The Constitution mandates recusal only in an “extraordinary situation.” Caperton v. A.T.
Massey Coal Co., 556 U.S. 868, 887 (2009). The relevant question is “whether, as an objective
matter, the average judge in his position is likely to be neutral, or whether there is an
unconstitutional potential for bias.” Williams v. Pennsylvania, 579 U.S. 1, 8 (2016) (internal
quotation marks omitted). For example, the Court has required recusal where the judge received
significant campaign contributions from a litigant. Caperton, 556 U.S. at 872. So too where the
judge “had significant, personal involvement as a prosecutor in a critical decision regarding the
defendant’s case.” Williams, 579 U.S. at 8.
That’s not this case. Householder’s fundraising occurred two decades before his
prosecution. There’s no evidence that the judge himself knew that Householder donated to that
organization. Not only that, but there also isn’t any evidence that the organization to which
Householder donated opposed this judge’s campaign. Thus, the facts of this case are a far cry
from Caperton, where the litigant gave money to a judge’s campaign as his case was pending.
See 556 U.S. at 873. In short, Householder hasn’t shown an objective likelihood for bias.
VI. Sentencing
Finally, Householder challenges the procedural and substantive reasonableness of his
sentence. Householder received the statutory maximum under RICO: twenty years. We uphold
his sentence.
A. Procedural Reasonableness
A sentence must be procedurally reasonable. That is, the court must “properly calculate
the [G]uidelines range” and “adequately explain” its sentence. United States v. Rayyan, 885 F.3d
436, 440 (6th Cir. 2018). Householder brings three arguments here: (1) that the court’s
calculation of the value of the bribe was erroneous; (2) that the court’s application of an

-- 33 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 34
enhancement for obstruction of justice was erroneous; and (3) that the Guidelines overstated the
seriousness of his offense.
1. Value of the Bribe
The Sentencing Guidelines provide an enhancement for a bribery-related conviction
based on the value of the bribe. U.S.S.G. § 2C1.1(b)(2). Based on the roughly $59 million
Householder received from FirstEnergy, the district court added 22 levels to Householder’s
offense according to the loss tables in U.S.S.G. § 2B1.1(b). This enhancement raised
Householder’s Guidelines range by two decades, from 21 to 27 months to 235 to 293 months.
See U.S.S.G. Ch. 5, Pt. A.
As part of the procedural reasonableness requirement, the district court must “adequately
explain why it chose the sentence.” Rayyan, 885 F.3d at 440. This explanation allows for
“meaningful appellate review” and “promote[s] the perception of fair sentencing.” Gall
v. United States, 552 U.S. 38, 50 (2007). Further, the district court must find the facts giving rise
to a sentencing enhancement by a preponderance of the evidence. United States v. Poulsen,
655 F.3d 492, 505 (6th Cir. 2011).
The explanation here was cursory at best. But even if there were error in the district
court’s explanation, it was harmless because it did not affect Householder’s sentence.
The district court explained it “sat through the entirety of the jury trial,” “heard all of the
evidence,” and could “say with absolute certainty that the government” showed that all the
contributions from FirstEnergy to Householder were bribes. R. 285, Pg. ID 11180–81. But this
explanation was conclusory—the court made no specific findings on the record and didn’t refer
to the Presentence Report or the government’s sentencing memorandum detailing the
FirstEnergy contributions. See United States v. Ferguson, 518 F. App’x 458, 467 (6th Cir. 2013)
(observing that procedural reasonableness requires “more than a simple and conclusory judicial
assertion”). This cursory explanation hardly “promote[s] the perception of fair sentencing.”
Gall, 552 U.S. at 50. District courts in the future should point to specific factual findings to
support the application of a sentencing enhancement—especially one that adds two decades to a
defendant’s Guidelines range.

-- 34 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 35
The court also cited the jury’s verdict, but that too provided little explanation of how the
court reached the $59 million number. As the court announced, “all computations [were]
supported by the evidence as ruled by the . . . jury.” R. 285, Pg. ID 11181. But the jury made no
finding about a total bribe value. Rather, as the government argued to the jury, “you only need
two acts of racketeering to convict.” R. 238, Pg. ID 9533. And the jury returned a general
verdict without any calculated amount.
The government, for its part, points to the district court’s observation that it also could
have imposed the loss-value enhancement based on a bribe value of more than $1 billion. Why?
Because the loss-value Guideline allows courts to use the larger of the “value of the payment”
and “the benefit received or to be received in return for the payment.” U.S.S.G. § 2C1.1(b)(2).
Here, the benefit FirstEnergy received was over $1 billion. While the government is right that
the court could’ve gone down this path, the court didn’t. The possibility that the court could’ve
used a larger bribe value therefore says nothing about how the court calculated $59 million as the
benefit Householder received from FirstEnergy.
Even so, we cannot vacate Householder’s sentence because any error would be harmless.
The district court maintained that it would have imposed the same sentence “no matter . . . the
guideline computation.” R. 285, Pg. ID 11219. And the court emphasized that the loss
calculation was “academic” because even if it used Householder’s proposed loss number, the
Guidelines range would be “at or above” the statutory maximum. Id. at Pg. ID 11182. In such a
situation, any error in calculating the Guidelines range is harmless. United States v. Morrison,
852 F.3d 488, 491 (6th Cir. 2017).
2. Obstruction of Justice
Next, because the court found that Householder gave false testimony, it applied a
two-level enhancement for obstruction of justice. See U.S.S.G. § 3C1.1. Householder
challenges that enhancement. But even without the increase, Householder’s offense level was
well above the Guidelines’ maximum offense level of forty-three. His imprisonment range also
was well above the RICO statutory maximum. Therefore, any error was harmless.

-- 35 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 36
3. Seriousness of the Offense
Householder further claims that the court failed to address his argument in the Sentencing
Memorandum that the Guidelines “vastly overstate the seriousness” of his offense. R. 279,
Pg. ID 11038. Householder didn’t raise this argument at his sentencing hearing, so we review
for plain error.
There wasn’t any error, let alone a plain one. The court did address the seriousness of the
offense. Discussing the § 3553(a) factors, for example, it emphasized the magnitude and
severity of Householder’s offense, referring to it as an “assault on democracy and the betrayal of
everyone in Ohio.” R. 285, Pg. ID 11217.
B. Substantive Reasonableness
In addition, Householder challenges the substantive reasonableness of his sentence. This
is “a complaint that the court placed too much weight on some of the § 3553(a) factors and too
little on others.” Rayyan, 885 F.3d at 442. Because sentencing is a “matter of reasoned
discretion, not math,” our review is highly deferential. Id. Indeed, we presume that a within-
Guidelines sentence is reasonable. United States v. Graham, 622 F.3d 445, 464 (6th Cir. 2010).
This sentence was within the Guidelines.
Householder argues that the Court discounted the minimal effect a sentence would have
on deterrence since Householder was sixty-four at the time. He also claims the sentence creates
unwarranted sentencing disparities.
But that’s wrong. The court observed that “the risk of recidivism is likely on the low end
here.” R. 285, Pg. ID 11220. But the court also had “no sense” that Householder grasped the
“harm that [he] caused” and emphasized the need for “adequate general deterrence.” Id. at Pg.
ID 11220–21.
And the court addressed any disparities. It looked at “similarly situated defendants in
similar cases” and found that “sentence[s] at or near the statutory maximum” were “not out of
the norm.” Id. at Pg. ID 11219–20. It emphasized that any sentencing disparity “would not be

-- 36 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 37
unwarranted on the evidence.” Id. And the court stressed that this was the “biggest corruption
case in Ohio’s history.” Id. at Pg. ID 11215.
Ultimately, Householder believes the court should have imposed a different sentence.
But our review “focuses on whether the sentence is reasonable, not whether we would have
imposed the same sentence.” United States v. Sanders, No. 24-3249, 2024 WL 4579446, at *4
(6th Cir. Oct. 25, 2024) (citing United States v. Sexton, 512 F.3d 326, 332 (6th Cir. 2008)). We
therefore reject his claim.
* * *
At the end of the day, Householder’s conviction stands.
Matthew Borges
Matthew Borges also brings several challenges to his conviction. Most of Borges’s
arguments relate to whether the government proved that Borges committed various predicates, or
whether particular predicate offenses were correctly instructed or charged. He makes these
arguments about several predicates: the Travel Act, honest services fraud, and money
laundering. But his arguments fail.
I. Sufficiency of the Evidence
We need not address whether the government adequately proved that Borges committed
any predicate offenses under the Travel Act, honest services fraud statutes, or money laundering
laws. While a defendant can only be convicted of a substantive RICO offense if he performs two
predicate acts creating a pattern of racketeering activity, proof of a defendant’s involvement in
two predicates is not required for a RICO conspiracy offense. Salinas v. United States, 522 U.S.
52, 65 (1997). For a conspiracy charge, the government must only prove that the defendant
“knew about and agreed to facilitate” the underlying scheme. Id. at 65. We have consistently
said the government does not need to prove that the defendant committed any predicates himself
or even that any overt acts have been committed—a conspiracy charge rests on the unlawful
agreement to further an endeavor which, if completed, would satisfy all the elements of a

-- 37 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 38
substantive criminal offense. United States v. Saadey, 393 F.3d 669, 676 (6th Cir. 2005); see
also Hills, 27 F.4th at 1174.
There is ample evidence that Borges knew and agreed to facilitate the illegal activity
involved in the Householder enterprise. While the bailout bill was being drafted, Cespedes
regularly communicated with Borges about the legislation and both Householder and
FirstEnergy’s involvement in the deal. When the bill faced opposition, FirstEnergy hired Borges
to identify legislators that needed to be persuaded and to suggest language for the bill. And
during the referendum campaign, Borges tried to pressure Yost to interpret House Bill 6 as a
“tax,” since taxes aren’t subject to referendum. Borges also led Householder’s efforts to disrupt
the signature campaign. He hired operatives to research and create media stories about the
OACB signature collectors. All told, Borges had a deep knowledge of (and involvement in)
Householder’s bribery scheme.
And, to the degree that Borges claims he did not know the enterprise was engaging in
illegal activity, that argument fails. Borges attempted to make this argument throughout the trial.
But the jury ultimately rejected it. And this court can’t second guess that conclusion.
That brings us to the Fehrman payment. Borges argues this payment did not violate the
Travel Act or amount to private honest services fraud. But it did not have to. Even if the
Fehrman payment is not a valid predicate, it still evinces Borges’s intent to further the
Householder enterprise because intent to further the enterprise need not be shown through
predicates or even criminal activity. See United States v. Bailey, Nos.
19-2280/2281/2354/20-1235, 2022 WL 2444930, at *3 n.1 (6th Cir. July 5, 2022). All this
evidence makes clear that Borges knew the object of the enterprise was to get the bailout
deal done and knew the enterprise was using illegal means to make it happen. R. 211, Pg. ID
6787–88 (Cespedes explaining that he kept Borges apprised of all the Householder team’s
efforts); R. 303, Pg. ID 12538 (Borges telling Fehrman they could “get fat off of this” “unholy
alliance” between Householder and FirstEnergy). That’s sufficient to establish co-conspirator
liability. See Salinas, 522 U.S. at 65.

-- 38 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 39
This was also clear to the jury. In closing arguments, the government told the jury it
could convict Borges even if he did “not personally commit any racketeering acts at all” but
“simply . . . agree[d] that another conspirator would commit two acts of racketeering activity and
agree[d] to be part of the enterprise and further its efforts.” R. 238, Pg. ID 9482. The jury
instructions echoed this statement of law. In its description of a “pattern of racketeering
activity,” the instructions explained: “The government is not required to prove that the
defendant actually committed the two acts of racketeering activity, or any acts at all. But the
government must prove beyond a reasonable doubt that the defendant agreed that either he or
another member of the conspiracy would commit at least two racketeering activities.” R. 237,
Pg. ID 9418. Therefore, any error related to either the Travel Act predicate or the private honest
services fraud predicate would be harmless because it was clear Borges did not need to commit
any predicate acts for the jury to find him guilty of a RICO conspiracy. Saadey, 393 F.3d at
676–77 (finding a trial court’s failure to strike an invalid predicate harmless because RICO
conspiracy does not require proof that the defendant committed any predicates so “the fact that
the conduct charged . . . cannot constitute a predicate offense is immaterial”).
So, all of Borges’s arguments about the sufficiency of the evidence fail.
II. Evidentiary Challenges
Borges also challenges the district court’s evidentiary rulings. He makes three claims.
All fail.
A. Meetings with Counsel
First, Borges objects to the district court’s treatment of seven pieces of testimony. The
evidence related to several individuals—Borges, Cespedes, and more—meeting with lawyers
during the attempt to pass House Bill 6. Seven instances are at issue.8 The meetings with
counsel largely discussed what Borges calls “fund raising [sic] activities.” Appellant Br. at 54.
8R. 191, Pg. ID 4603; R. 206, Pg. ID 6261–67; R. 208, Pg. ID 6529–31; R. 208, Pg. ID 6548–49; R. 212,
Pg. ID 6935; R. 212, Pg. ID 6991; R. 219, Pg. ID 7836.

-- 39 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 40
The district court didn’t abuse its discretion when it sustained the prosecution’s
objections to this testimony. See Gen. Elec. Co. v. Joiner, 522 U.S. 136, 141 (1997). Why? The
court was concerned that Borges was attempting to raise a pseudo-advice of counsel defense.
An advice of counsel defense is an affirmative defense in which a defendant says he
didn’t have the intent to do the unlawful act because he followed the advice of counsel in good
faith. See Williamson v. United States, 207 U.S. 425, 453 (1908). To be eligible for an “advice
of counsel” defense, a defendant must have (1) fully disclosed all pertinent facts to his lawyers
and (2) relied on the advice of counsel in good faith. United States v. Lindo, 18 F.3d 353,
356 (6th Cir. 1994).
Here, the court was right to be concerned about a pseudo-advice of counsel defense.
Borges never made a complete disclosure of all pertinent facts to the lawyers, nor did he act in
good-faith reliance on the advice of counsel. Id. Thus, any attempt to introduce evidence about
the attorney meetings would only confuse the jury.
In response, Borges says that the attorney meetings were necessary to understand the
background of the charged offenses. Such evidence is commonly known as res gestae evidence.
But the “background circumstances exception” is “not an open-ended basis to admit any
and all other act evidence the proponent wishes to introduce.” United States v. Hardy, 228 F.3d
745, 748 (6th Cir. 2000). Indeed, even the “very definition of what constitutes background
evidence contains inherent limitations.” Id. We must look at whether the background facts are
“inextricably intertwined with the charged offense or those acts, the telling of which is necessary
to complete the story of the charged offense.” Id.
Here, the evidence that Borges put forward—a few meetings with lawyers—isn’t
inextricably intertwined with his actions in the conspiracy. The meetings with lawyers—which
Borges claims occurred in the ordinary course of fundraising—are not relevant to whether
Borges tried to join a conspiracy of bribery and money laundering. As far as Borges argues that
some evidence was of such debatable use that it should have been admitted, or else was
necessary to get the full picture of the event in question, that’s simply a disagreement with how
the judge weighed the evidence. Because we review for abuse of discretion, that argument fails.

-- 40 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 41
B. Testimony from Signature Collectors
Next, Borges argues that the district court erred when it admitted testimony from two
AMT employees. At first glance, it’s difficult to discern exactly what Borges argues about these
employees. At trial, he objected to their statements as hearsay. Here, he first says that the
witnesses’ statements weren’t based on personal knowledge, and then says they were unduly
prejudicial. But Borges’s arguments about these witnesses also fall short.
Because Borges didn’t make this objection about personal knowledge or prejudice
at trial, we review for plain error. Fed. R. Crim. P. 52(b); United States v. Young, 470 U.S. 1,
15–16 (1985).
“A witness may testify to a matter only if evidence is introduced sufficient to support a
finding that the witness has personal knowledge of the matter.” Fed. R. Evid. 602. Testimony is
excluded for lack of personal knowledge when no reasonable juror could believe that the witness
had the ability and opportunity to perceive the event he testified about. United States v. Hickey,
917 F.2d 901, 904 (6th Cir. 1990).
Borges objects to testimony from two AMT employees. The first is testimony from
AMT CEO Michael Roberson, who testified that his employees were harassed, yelled at, stalked,
and bombarded with text messages and phone calls from opposition workers. He also said that
the employees were “poached, offered money and a plane ticket to leave the state or offered
money and an opportunity to go work for the opposition.” R. 216, Pg. ID 7500. Next, Borges
challenges testimony from AMT consultant Douglas Gray. The witness offered similar
testimony. Gray explained that “trackers” would follow around signature collectors. Id. at Pg.
ID 7524. They would often have a counter-petition and try to persuade a person to sign their
alternate petition, with the “ultimate goal” of having that person “throw up [his] hands and say I
don’t know, I’m not signing either one.” Id. at Pg. ID 7523. Gray also testified the police were
called out of concern for his employees’ safety. Id. at Pg. ID 7525.
The district court didn’t err when it admitted this testimony. Both individuals knew
about these events because they supervised, managed, and were intimately involved with the
signature-collection efforts. Thus, both Roberson and Gray had extensive personal knowledge of

-- 41 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 42
the staffing challenges related to the petition initiative and were more than qualified to testify
about the harassment that caused many of their employees to quit.
Next, consider Borges’s arguments that Roberson and Gray’s testimony was unfairly
prejudicial. Since he did make this objection at trial, we review the district court’s determination
for an abuse of discretion. Joiner, 522 U.S. at 141.
Borges’s argument hinges on Federal Rule of Evidence 403. Recall that a district court
can exclude evidence under Rule 403 only if its relevance is “substantially outweighed by the
danger of unfair prejudice.” United States v. Hans, 684 F.2d 343, 346 (6th Cir. 1982). The trial
judge’s discretion in balancing the probative value of evidence against its potential for unfair
prejudice is “very broad.” United States v. Dabish, 708 F.2d 240, 243 (6th Cir. 1983). “[T]he
test is strongly weighted toward admission.” Hazelwood, 979 F.3d at 412 (quotation omitted).
Thus, a litigant who makes a Rule 403 challenge faces a steep hill. Borges can’t make the climb.
While there’s little doubt that the testimony portrayed Borges and his co-conspirators in a
bad light, it was probative of the tenor of the anti-referendum campaign and wasn’t unfairly
prejudicial. Roberson and Gray both testified about how the anti-referendum campaign worked.
As Roberson put it, the employees were stalked, intimidated, harassed, and some were assaulted.
Neither Roberson nor Gray’s recollection was unnecessary—rather, they recounted the basic
facts of what occurred. Were some of those actions ugly? Yes. But they were also integral to
understanding how the Householder enterprise and the anti-referendum campaign worked. Thus,
the district court didn’t err by allowing such testimony.
Borges, for his part, argues that the evidence was unduly prejudicial because it conveyed
“gangster-style conduct that the average juror would associate with racketeering.” Appellant Br.
at 56. But such conclusory labels don’t help his case. Instead, what matters is whether the
testimony was so prejudicial that it substantially outweighed any probative value. Hans, 684
F.2d at 346. Here, it wasn’t.
Thus, Borges’s arguments about the testimony of the AMT employees fails.

-- 42 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 43
C. Guilty Pleas
Borges also argues that the district court erred when it admitted evidence of Longstreth
and Cespedes’s guilty pleas. This is the same challenge as the one that Householder brought.
And just like in Householder’s case, Borges’s argument has no merit. See supra, Part IV.B. We
thus reject it.
* * *
We affirm.

-- 43 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 44
_________________
CONCURRENCE
_________________
THAPAR, Circuit Judge, concurring. As the per curiam opinion explains, conspiracy law
forecloses Matthew Borges’s arguments. See Salinas v. United States, 522 U.S. 52, 65 (1997).
That same framework also ties Borges’s conviction to Larry Householder’s bribery scheme.
Householder’s guilt is clear under current law. But if the Supreme Court revisits its bribery
cases and undermines the foundation of Householder’s conviction, Borges’s conviction is also
ripe for reconsideration.
I.
Understanding Borges’s arguments requires a brief discussion of Householder’s
conviction. A jury found Householder guilty of RICO conspiracy. One of the charged
predicates was extortion under the Hobbs Act. The Supreme Court has said that this statute
prohibits a public official from receiving money in exchange for official action—in other words,
a quid-pro-quo bribery agreement. United States v. Sittenfeld, 128 F.4th 752, 787–89 (6th Cir.
2025) (Murphy, J., concurring).
Where does the Supreme Court’s reading of the statute come from? Not the text. The
Hobbes Act criminalizes “extortion,” which is “the obtaining of property from another, with his
consent, induced by wrongful use of actual or threatened force, violence, or fear, under color of
official right.” 18 U.S.C. § 1951(b)(2). In other words, it criminalizes the conduct of an official
who “claim[s] that their office gives them the right to the money.” Sittenfeld, 128 F.4th at
788 (Murphy, J., concurring). That’s textbook extortion, where an official makes a victim of the
payor by forcing him to fork over the money. And that act is different than bribery, which
requires two willing partners in crime. Id. at 787–88.
Yet the Supreme Court has turned the Hobbs Act into a bribery law. Id. at 788 (citing
Evans v. United States, 504 U.S. 255 (1992)). In so doing, the Evans Court “simply made up”
the quid-pro-quo requirement. Evans, 504 U.S. at 286 (Thomas, J., dissenting). It held that an

-- 44 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 45
official commits extortion when he “receives a payment in return for his agreement to perform
specific official acts.” Id. at 268 (majority opinion).
This distinction matters. Imagine a candidate promises to oppose a dam’s construction
and hosts a fundraiser to publicize his opposition. A rancher attends and gives the candidate a
$10,000 check. Handing over the check, the rancher says to the candidate, “I’m giving you this
money because you promised to oppose that dam.” That’s not extortion. Instead, it’s a classic
campaign contribution, as the government acknowledged at oral argument. But under one
reading of Supreme Court caselaw, this routine political activity is illegal. The candidate has
“receive[d] a payment in return for his agreement to perform specific official acts.” Id.
As Judge Murphy explained, expanding the Hobbs Act to potentially cover such activity
poses serious constitutional issues. See Sittenfeld, 128 F.4th at 790–91 (Murphy, J., concurring).
After all, the First Amendment “has its fullest and most urgent application precisely to the
conduct of campaigns for political office.” McCutcheon v. FEC, 572 U.S. 185, 192 (2014)
(opinion of Roberts, C.J.) (quoting Monitor Patriot Co. v. Roy, 401 U.S. 265, 272 (1971)). But
one reading of Evans suggests that the Hobbs Act criminalizes at least some campaign
contributions. That’s inconsistent with the Constitution’s ironclad protection of political speech.
Indeed, these concerns apply with full force in Householder’s case. The district court’s
jury instructions—although supported by one reading of Evans—swept in everyday political
activity. The instructions described an illegal quid pro quo as “a public official’s receipt of
things of value when the public official knows that the person who gave the thing of value was
doing so in return for the public official performing or agreeing to perform a specific official
action.” R. 260, Pg. ID 10605–06. That definition captures everyday politics. For example, in
the candidate and rancher hypothetical above, the candidate received the rancher’s check
“know[ing]” that the rancher “was doing so in return for” the candidate’s promised opposition to
the dam. Id. at Pg. ID 10606. Thus, under the district court’s instructions, both the candidate and
the rancher could face up to 20 years in prison for Hobbs Act extortion.
Worse, the instructions eliminated an obvious defense: that the candidate would’ve
opposed the dam regardless of the rancher’s contribution. The instructions made clear that “it is

-- 45 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 46
not a defense to bribery . . . even if the public official would have performed the official action
anyway.” Id. So, in our rancher hypothetical, the instructions told the jury to convict a
candidate even if he would have supported the dam anyway. That raises serious First
Amendment issues.1
All’s to say, I join the chorus of judges encouraging the Supreme Court to revisit Evans.
See Silver v. United States, 141 S. Ct. 656, 656–57 (2021) (Gorsuch, J., joined by Thomas, J.,
dissenting from denial of certiorari) (calling on the Court to reconsider Evans); Ocasio v. United
States, 578 U.S. 282, 300 (2016) (Breyer, J., concurring) (agreeing that Evans “may well have
been wrongly decided”); Sittenfeld, 128 F.4th at 772 n.8 (Nalbandian, J.) (“At this point,
McCormick and Evans are nearly 35 years old and it may be time for the Court to revisit or
refine the doctrine.”); id. at 792 (Murphy, J., concurring) (encouraging the defendant to ask the
Court to “reassess the [Hobbs] Act’s scope in light of three decades’ worth of precedent finding
campaign donations entitled to strong First Amendment protection”); id. at 806 (Bush, J.,
dissenting) (emphasizing that “[i]t would be helpful for the Supreme Court to provide guidance
here”). Should the Court act, Householder’s conviction may well fall.
II.
If it does, that will have a trickle-down effect on Borges’s conviction. For Borges’s
conviction for RICO conspiracy to stand, the government had to prove that he intended to further
an endeavor that involved two (or more) predicate criminal acts. 18 U.S.C. § 1961; H.J. Inc.
v. Nw. Bell Tel. Co., 492 U.S. 229, 237 (1989). Here, there were several predicates: the Travel
Act, private honest services fraud, money laundering, and a trio of public-official bribery
statutes—Hobbs Act extortion, public honest services fraud, and Ohio bribery.
1While Evans remains good law, district courts must ensure juries aren’t convicting defendants based on
protected conduct. It’s critical that district courts distinguish between legitimate contributions and illegal bribes.
One way to do so is to focus on causation. If an official wouldn’t have taken the action but for the contribution, then
the payment is a bribe. There’s also another option: a bribe occurs when the official receives the money knowing
his conduct is “controlled by the terms of the promise.” United States v. McCormick, 500 U.S. 257, 273 (1991).
Courts can also instruct the jury to consider whether the official would’ve performed the action anyway, which
means the payment wasn’t a bribe. See United States v. Sittenfeld, Case No. 1:20-cr-142, Jury Instrs. at 26, 39.
Unfortunately, the instruction here had none of these safeguards.

-- 46 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 47
And if the public-official bribery convictions can’t stand, one or more predicate offenses
would be invalid. Sometimes, that requires overturning the whole conviction. The key question
is whether the invalid predicate is “inextricably intertwined” with the valid predicate offenses.
Baugh v. United States, 64 F.4th 779, 782 (6th Cir. 2023). Do the two predicates stem from the
exact same act? If yes, the two questions are inexplicably intertwined. If not, they aren’t.
In other words, the key inquiry is whether the “record furnish[es] the jury with any basis
to conclude” that a defendant committed the invalid predicates but not the valid ones. Id. at 782
n.2 (emphasis in original). If it did, the conviction can’t stand. But if the underlying conduct is
the same, then the conviction is proper.
Here, if Householder’s public-official bribery predicates are invalid, then Borges’s
conviction rests on shaky ground to say the least. Why? Because there would not be enough
valid predicates for Borges’s conviction to stand.
To start, Borges makes persuasive arguments that doom his Travel Act predicate. Next,
while the honest services fraud predicate rests on firmer theoretical ground, it too raises serious
concerns. If Householder’s public-official bribery predicates were taken out of the picture as
well, then Borges’s conviction would not stand.
A.
Borges’s role in Householder’s saga started out small. Borges knew one of the
FirstEnergy lobbyists, Juan Cespedes, and Cespedes kept Borges in the loop about his efforts to
procure a bailout deal for FirstEnergy by getting Householder elected speaker. So, once
Householder won the speakership, Cespedes and FirstEnergy turned to Borges for help getting
the bailout legislation passed. They began by asking Borges to suggest language for the bill and
to identify legislators who were on the fence about the bailout.
As the bailout legislation passed the Ohio House of Representatives and proceeded to the
state senate, Borges’s involvement increased. He thought the bailout might trigger a citizen-led
effort to repeal the bill. So, he took preventative steps to make the bill referendum-proof.

-- 47 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 48
But when the referendum did materialize, Borges quarterbacked a multi-pronged effort to
defeat it. The proponents of the referendum needed a certain number of signatures within 90
days of the bailout’s passage to get on the Ohio ballot. To prevent this result, Borges tried to get
information about the signature-collection campaign and disrupt its activities. One way he did so
was by asking Tyler Fehrman, who worked at the signature-collection firm the referendum
campaign hired, how many signatures his group had collected. Borges paid $15,000 to persuade
Fehrman to find out how many signatures the firm obtained, and promised that he’d pay $10,000
once he received the data that he requested. Fehrman never sent the data.
B.
The bulk of Borges’s arguments concern this payment. The government argued that this
payment could serve as a predicate offense in two ways: as a violation of the Travel Act and a
form of honest services fraud. Borges disagrees. First, Borges points out that he violated the
Travel Act only if he also violated a different Ohio election law—one that the state has never
charged anyone with violating. And he argues that even if Ohio wanted to use that election law
against him, it would have needed to go through a lengthy administrative process—a process the
state never even began. Thus, he believes that the federal government wrongly usurped the
state’s remedial scheme. Second, he argues that the government’s honest services fraud theory
has several flaws. These include that it violated recent caselaw requiring a recognized property
interest, that Borges’s jury instructions were deficient, and that Borges didn’t induce a breach of
a fiduciary duty. Borges’s arguments about both predicates have merit.
1.
The Travel Act prohibits using interstate commerce to commit “any unlawful activity,”
defined to include various types of state laws. 18 U.S.C. § 1952(a)–(b). The Ohio law invoked
by the government prohibits an individual from giving something of value to someone working
for (or against) a referendum campaign to influence how that person does his job. Thus, the
Travel Act charge required the government to prove that Borges used interstate commerce to

-- 48 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 49
bribe an employee or agent of the referendum campaign—with the “bribe” being Borges’s
payment to Fehrman.2
The federal government’s reliance on this state-law predicate raises a host of challenging
questions. For one, what state crimes suffice as Travel Act predicates? Is a violation of an Ohio
election law “bribery” under the Travel Act? And if the answers to these questions aren’t clear,
then how is the defendant on sufficient notice to be held criminally accountable for his acts?
In three cases, the Supreme Court recognized that the Travel Act implicates these
concerns but failed to give the lower courts clear guidance. In United States v. Nardello, the
defendant was indicted for violating the Travel Act when he extorted private citizens. 393 U.S.
286 (1969). The state extortion law didn’t make that a crime—it only criminalized extorting
public officials, as at common law. Id. at 288–89. But the Supreme Court said the Travel Act
still applied, even if the state extortion statute didn’t. See id. at 287. Why? Because the state’s
“blackmail crime” could cover the conduct in question. Id. The Court reasoned that disputes
about the state’s labels for criminal conduct shouldn’t matter. Id. at 293–94. That’s because the
Travel Act’s purpose was to “aid local law enforcement” in convicting “top men . . . whose
influence extends over State and National borders.” Id. at 293, 290–91.
But the Court soon appeared to walk back its broad reading of the Travel Act in Rewis
v. United States. 401 U.S. 808 (1971). Rewis concerned whether a Georgia resident who
crossed state lines to gamble, in violation of a Florida statute, was guilty under the Travel Act.
The Court said no, because an expansive reading of the Act would “alter sensitive federal-state
relationships, could overextend limited federal police resources, and might well produce
situations in which the geographic origin of customers, a matter of happenstance, would
transform relatively minor state offenses into federal felonies.” Id. at 812. Pointing to the
2Technically, the Travel Act prohibits using “any facility in interstate or foreign commerce” with the intent
to “promote, manage, establish, carry on, or facilitate . . . any unlawful activity.” 18 U.S.C. § 1952(a). “[U]nlawful
activity,” in turn, includes “extortion, bribery, or arson in violation of the laws of the State in which committed.” Id.
§ 1952(b)(i)(2). The state law predicate that the government charged Borges with violating was Ohio Rev. Code.
Ann. § 3517.22(A)(2). That’s a statute entitled “unfair activities in issue campaign[s].” Id. It prohibits promising,
offering, or giving a “valuable thing or valuable benefit to any person who is employed by or is an agent of a
committee in advocacy of or in opposition to the adoption of any ballot proposition or issue, for the purpose of
influencing the employee or agent with respect to the improper discharge of the employee’s or agent’s campaign
duties or to obtain information about the committee’s campaign organization.” Id.

-- 49 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 50
legislative history, the Court noted that Congress “did not intend that the Travel Act should apply
to criminal activity solely because that activity is at times patronized by persons from another
state.” Id. And the Court emphasized that “ambiguity concerning the ambit of criminal statutes
should be resolved in favor of lenity.” Id. In Rewis, then, the Court recognized that federalism
and notice principles counseled against stretching the Travel Act too thin.
Finally, in Perrin v. United States, the Court again revisited the breadth of the Travel Act.
444 U.S. 37 (1979). Perrin considered several defendants who were convicted of bribing an
employee to steal confidential data from his employer. Id. at 39–40. On appeal, the Perrin
defendant, like the Nardello defendant, argued that “bribery” only referred to public-official
bribery—as at common law—and not private employees. Id. at 41. In an effort to distinguish
his case from Nardello, the Perrin defendant argued that the “federalism concerns” in Rewis
meant courts shouldn’t read the Travel Act’s invocation of state bribery predicates too broadly.
See id. at 49.
But the Court disagreed and reaffirmed Nardello. The Perrin Court concluded that the
Travel Act reflected Congress’s “clear and deliberate intent . . . to alter the federal-state balance
in order to reinforce state law enforcement.” Id. at 50. So long as the defendant’s conduct had
an interstate nexus, the Court explained, courts could uphold convictions under the Travel Act
even if the “bribery” or “extortion” predicate swept broader than at common law. As the Court
put it, “[i]n defining an ‘unlawful activity,’ Congress has clearly stated its intention to include
violations of state as well as federal bribery law.” Id. (citation omitted). As in Nardello, the
Court said that Congress intended to sweep beyond the common law.
Adding all that up, in Nardello, the Court looked to the conduct at issue, not state labels.
393 U.S. at 293. In Rewis, the Court cautioned against a broad reading of the Travel Act and
declined to transform minor state-law violations into federal felonies. 401 U.S. at 812. And in
Perrin, the Court focused on the crime’s interstate nexus. 444 U.S. at 49–50.3
3In the forty-six years since Perrin, the Supreme Court hasn’t revisited or clarified its interpretation of the
Travel Act. It has, however, emphasized Nardello’s expansive reading of the Travel Act. See Scheidler v. Nat’l
Org. for Women, Inc., 537 U.S. 393, 409–10 (2003); Taylor v. United States, 495 U.S. 575, 594–95 (1990); see also

-- 50 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 51
Applying this framework to Borges’s claim is no small feat.
2.
Borges raises a compelling argument about his Travel Act charge. In effect, the federal
government’s use of a state election law turned a statute the state has never used into a magic
bullet to convict Borges.
The crux of Borges’s claim is that federalism and due-process principles mean the
government can’t use a rarely charged state election law to convict him of bribery. (Recall that
“bribery” is one of the available state law predicates under the Travel Act. See 18 U.S.C.
§ 1952(b).)
Start with the statute. The indictment alleged, and the jury was instructed, that Borges
violated the Travel Act because he violated section 3517.22(A)(2) of the Ohio Revised Code.
That statute is a state election law. It sits not within the state’s criminal provisions but in a
separate chapter governing elections.
And the statute has procedural protections that the federal government never pursued. If
Ohio wanted to prosecute a defendant under this law, it would have to conduct administrative
proceedings before it could ever charge a defendant in Borges’s position. See id. § 3517.22(C).
Thus, Borges couldn’t be convicted under this state law without going through an administrative
process—meaning his federal conviction under this law simply couldn’t have occurred at the
state level. In this way, the federal government has usurped Ohio’s ability to write, regulate, and
prosecute its laws. Ohio prescribed a specific method for enforcing its elections code. But the
federal government has used the Travel Act to steamroll Ohio’s process.
Perhaps the strongest indication that this Travel Act charge stretches too far is that, to my
knowledge, Ohio has never prosecuted an individual defendant under the Ohio statute here.
James v. United States, 550 U.S. 192, 221–22 (2007) (Scalia, J., dissenting), overruled by Johnson v. United States,
576 U.S. 591 (2015).

-- 51 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 52
Through the Travel Act, the federal government has managed to do with Ohio law what Ohio has
never done.
To be sure, Nardello and Perrin suggest that courts should discount labels if the conduct
matches a crime’s generic definition. But even setting aside the statutory differences listed
above, the Perrin indictment specified “bribery,” which it based on Louisiana’s commercial
bribery statute. 444 U.S. at 312 n.3. While Ohio has a bribery law that applies to public
officials, it does not have a commercial bribery statute that applies to private individuals. Cf.
Ohio Rev. Code Ann. § 2921.02. Thus, unlike Perrin, there is no underlying bribery law that
applies to private persons. And here, the indictment referenced a state election law, not a bribery
statute.
In addition, for those who find such arguments convincing, Congress’s purpose in
enacting the Travel Act confirms that the government’s charge upset the federal-state balance.
As Perrin said, the Travel Act was designed to help local law enforcement. But if, as here, the
state has never used the law to prosecute someone (and didn’t even begin administrative
proceedings against Borges), then it is unclear why the state would need help.
Further, in Nardello’s parlance, the government wanted to go after the “top men” who
“resided in one State but conducted their illegal activities in another.” 393 U.S. at 290 (citation
omitted). Here, this is a quintessential Ohio crime—paying a fellow Ohioan to influence the
results of an Ohio referendum on behalf of an Ohio company. And Borges never “traveled” at
all. Thus, it’s hard to say that federal “assistance” helps state law enforcement here. Nothing
about Nardello’s or Perrin’s reasoning justifies using the state law predicate here.
On the other hand, the district court upheld the Travel Act predicate because RICO
requires only that a predicate offense act be “chargeable” or “indictable” in a particular context.
United States v. Householder, No. 1:20-CR-77, 2023 WL 24090, at *6 (S.D. Ohio Jan. 3, 2023)
(quoting 18 U.S.C. § 1961(1)) (emphases omitted). Since Ohio could hypothetically charge the
state election law, the district court believed that was enough.
But this approach rests on a faulty reading of RICO. What matters is whether the Travel
Act was “indictable” in a particular context. See 18 U.S.C. § 1961(1)(B) (defining “racketeering

-- 52 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 53
activity” to include “any act which is indictable under . . . section 1952 [the Travel Act]”). So,
the relevant question is whether the Travel Act—not state law—was chargeable. And here, the
Travel Act cannot capture Borges’s conduct. Why? As I’ve explained, all evidence shows the
underlying state law wasn’t chargeable: the state has never charged an individual defendant with
it, and there were no administrative proceedings. See Rewis, 401 U.S. at 812. Thus, the
government couldn’t use the Travel Act in this context because the underlying conduct wasn’t
captured by its sweep.
What’s more, even if the Travel Act were chargeable, it’s debatable whether a given
defendant would be on notice that the Travel Act captured his conduct. Here, there are two
plausible readings of how the Travel Act applies to Borges’s conduct. Under the government’s
reading, Borges’s conduct fell within the Travel Act’s sweep. Under Borges’s plausible reading,
the Travel Act doesn’t sweep this broadly.
In such a situation, we often side with the defendant. The Tinkerbell strategy—clap if
you believe he’s guilty—doesn’t cut it when we think about criminal convictions. Instead, the
government must prove that its interpretation of a statute is the best reading of the law. See Gary
Lawson, Proving the Law, 86 Nw. U. L. Rev. 859, 888 (1992). Under this logic, the government
must show that a law covers the defendant’s conduct. See United States v. Erker, 129 F.4th 966,
970 (6th Cir. 2025). Courts also apply this principle through the substantive canon known as the
rule of lenity. Id. This rule dates to the sixteenth century. See Amy Coney Barrett, Substantive
Canons and Faithful Agency, 90 B.U. L. Rev. 109, 128–30 (2010). In the American tradition, it
demands that the accused be on notice of what the law criminalizes. Id. at 130. Otherwise,
citizens will be left “entirely at sea to guess” what the statute proscribes. Snyder v. United
States, 603 U.S. 1, 16 (2024). And punishing a defendant for guessing wrong sits uncomfortably
next to fundamental due process principles. Wooden v. United States, 595 U.S. 360, 389 (2022)
(Gorsuch, J., concurring).
Those concerns apply in full force here. The state statute at issue gives a defendant the
right to administrative proceedings before any criminal charge. Few readers of the Travel Act
are on notice that an obscure provision of Ohio’s election code could serve as a state law bribery
predicate. More likely, “any fair reader of this statute would be left with a reasonable doubt

-- 53 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 54
about whether it covers the defendant’s charged conduct.” Snyder, 603 U.S. at 20 (Gorsuch, J.,
concurring). As Justice Gorsuch said, “[w]hen that happens, judges are bound by the ancient
rule of lenity to decide the case . . . not for the prosecutor but for the presumptively free
individual.” Id.
*
All told, when it comes to the Travel Act, Borges makes compelling arguments about a
concerning federal statute. He’s right that the government charged a little-used state law
predicate in a way that violated fundamental notions of federalism and notice. But as the per
curiam explains, we need not decide these thorny issues today. The broad sweep of conspiracy
law is sufficient to affirm Borges’s sentence.
3.
Next, Borges wades into another thicket: honest services fraud.
i.
Title 18 of the United States Code criminalizes using the mails or wires in furtherance of
“any scheme or artifice to defraud” or “for obtaining money or property by means of false or
fraudulent pretenses, representations, or promises.” 18 U.S.C. §§ 1341, 1343. Most violations
of these provisions are what we commonly think of as mail or wire fraud—misrepresentations
and omissions that deprive a victim of his property or money. But Congress also made clear that
the term “scheme or artifice to defraud” includes a “scheme or artifice to deprive another of the
intangible right of honest services.” Id. § 1346. This statute, and its relationship with mail fraud
and wire fraud, lead to the crime called “honest services fraud.”
The elements of honest services fraud are deceptively simple. To secure a conviction, the
government must prove a few elements: (1) a scheme to defraud and intent to defraud another;
(2) through use of an interstate carrier; and (3) of money, property, or honest services. Cf.
United States v. Faulkenberry, 614 F.3d 573, 581–83 (6th Cir. 2010).
Often, this is simple. Start with property. The property element applies to both tangible
property and intangible interests, so long as those interests have “long been recognized as

-- 54 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 55
property.” Carpenter v. United States, 484 U.S. 19, 26 (1987). But if they haven’t, then there’s
no property interest and thus no wire or mail fraud. For example, the Court has explained that a
state lacked a property interest in video poker licenses because they were just variations on the
state’s sovereign right to issue such licenses and didn’t bestow a traditional property interest.
See Cleveland v. United States, 531 U.S. 12, 20 (2000). Thus, depriving a person of these
licenses wasn’t fraud. Id. One can’t commit wire or mail fraud unless the object of the fraud is a
specific property interest long recognized as such.
What are “honest services”? There’s no easy answer. Instead, to understand this strange
phrase, we must embark on a legal odyssey.
When the wire fraud statute was passed nearly seventy-five years ago, its plain text didn’t
prohibit honest services fraud. But courts began to interpret the fraud statutes to
cover deprivation of so-called “intangible rights.” See, e.g., United States v. Bryza, 522 F.2d
414, 421–22 (7th Cir. 1975); United States v. Odom, 736 F.2d 104, 116 n.13 (4th Cir. 1984).
Most of those cases involved public officials who “made governmental decisions with the
objective of benefitting themselves or promoting their own interests, instead of fulfilling their
legal commitment to provide the citizens of the State or local government with their loyal service
and honest government.” McNally v. United States, 483 U.S. 350, 362–63 (1987) (Stevens, J.,
dissenting). Even though the statute required a deprivation of money or property, courts
expanded it to include public officials who engaged in self-dealing, thus depriving constituents
of the right to the “honest services” of their representatives.
But in 1987, the Supreme Court in McNally “stopped the development of the
intangible-rights doctrine in its tracks.” Skilling v. United States, 561 U.S. 358, 401 (2010)
(citing McNally, 483 U.S. at 360). In McNally, the government charged several defendants with
mail fraud. 483 U.S. at 352. The defendants were Kentucky party officials who would choose
the insurance agencies from which Kentucky bought insurance. Id. Those insurers then gave
those officials kickbacks through the officials’ private companies. Id. The federal government
prosecuted these officials, arguing they committed mail fraud by depriving state citizens of “their
intangible rights to honest and impartial government” by misusing their office “for private gain.”
Id. at 355. But the Supreme Court disagreed. It explained that the fraud statute was limited “to

-- 55 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 56
the protection of property rights.” Id. at 360. Any broader reading would leave the statute’s
“outer boundaries ambiguous and involve[] the Federal Government in setting standards of
disclosure and good government.” Id. So, the Court confined the mail and wire fraud statutes to
their plain text.
Congress responded by passing 18 U.S.C. § 1346, the statute at issue. That law explained
that “the term ‘scheme or artifice to defraud’” in the wire fraud statute includes “depriv[ing]
another of the intangible right of honest services.” 18 U.S.C. § 1346. Thus, Congress intended to
criminalize some sort of conduct. But exactly what conduct remained unclear. After all,
Congress didn’t define what “honest services” meant.
Prosecutors began using the new statute to charge defendants. Most of these cases
involved “either bribery” or “failure to disclose a conflict of interest, resulting in personal gain.”
United States v. Sawyer, 85 F.3d 713, 724 (1st Cir. 1996).
But trouble arose. The circuits fractured in trying to figure out what kind of conduct
could be charged under this statute. For our part, the Sixth Circuit determined that an honest
services fraud conviction required that the defendant “foresaw or reasonably should have
foreseen that [a victim] might suffer an economic harm as a result of the breach of fiduciary
duty.” United States v. Frost, 125 F.3d 346, 366 (6th Cir. 1997). Other courts of appeals
asserted that an individual could be convicted only if his conduct also violated a state law. See
United States v. Brumley, 116 F.3d 728, 734 (5th Cir. 1997) (en banc). One court adopted the
view that a defendant just needed to misuse his position for private gain. United States v. Sorich,
523 F.3d 702, 707–08 (7th Cir. 2008). Still other courts asked whether a defendant’s
misrepresentation was material, or likely to cause an employee to change his behavior. United
States v. Cochran, 109 F.3d 660, 667 (10th Cir. 1997). What united these approaches? Nothing,
other than confusion.
After the honest services fraud statute had been on the books for about two decades,
Justice Scalia summed up the chaos: there was no “coherent limiting principle” to “separate[]
the criminal breaches, conflicts, and misstatements from the obnoxious but lawful ones.” Sorich
v. United States, 555 U.S. 1204, 1310 (2009) (Scalia, J., dissenting from denial of certiorari).

-- 56 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 57
This lack of a clear rule invited “abuse by headline-grabbing prosecutors in pursuit of local
officials, state legislators, and corporate CEOs who engage[d] in any manner of unappealing or
ethically questionable conduct.” Id.
Soon after, the Supreme Court tried to answer Justice Scalia’s criticisms. In Skilling
v. United States, the Court both acknowledged that the vague phrase “honest services fraud”
raised due process concerns and tried to craft a judicial solution by giving it a “limiting
construction.” 561 U.S. at 408, 402.
The Court began by “survey[ing]” the “body of pre-McNally honest-services” caselaw.
Id. at 404 (citation omitted). Why? Because, the Court reasoned, when Congress passed § 1346
it referred to and incorporated that corpus. From this study, the Court reached two
conclusions: (1) that the “honest-services decisions preceding McNally were not models of
clarity or consistency” and, (2) that despite the inconsistency, § 1946 encompassed a “solid core”
of cases involving “offenders who, in violation of a fiduciary duty, participated in bribery or
kickback schemes.” Id. at 405, 407. But see id. at 416–20 (Scalia, J., concurring in part and
concurring in the judgment). With that in mind, the Supreme Court held that § 1346 was limited
to this “core” of bribery and kickback cases. Id. at 368, 409 (majority opinion).
What “fiduciary duty” is necessary to commit honest services fraud? The Court didn’t
say. But it did list a few examples of fiduciary relationships that lower courts had found to be
“beyond dispute”: “public official-public, employee-employer, and union official-union
members.” Id. at 407 n.41 (cleaned up). And what “bribery and kickback” schemes sufficed?
The Court didn’t say that, either. It instead remarked that those terms would “draw[] content”
from federal statutes and the “core” of pre-McNally case law. Id. at 404, 412.
As Justice Scalia put it, the Court’s interpretation of 18 U.S.C. § 1346 was “a dish the
Court . . . cooked up all on its own.” Id. at 422 (Scalia, J., concurring in part and concurring in
the judgment). Since Skilling, courts have struggled to digest the novelty that the Court
produced. Why? At bottom, “even with the bribery and kickback limitation, the statute does not
answer the question, ‘What is the criterion of guilt?’” Id. at 421. Courts have been grappling
with these questions ever since.

-- 57 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 58
ii.
Against this legal backdrop, Borges makes three relevant arguments.4 First, he claims
that his indictment failed to allege, and that the jury instructions failed to specify, that he offered
the bribe to Fehrman in exchange for a recognized property interest. Second, he says that the
district court’s instructions erroneously required Fehrman’s employer to have suffered a
foreseeable economic harm rather than that Borges intended to cause such a harm. Third, he
says that Fehrman didn’t owe a fiduciary duty to his employer, and so Borges didn’t induce a
necessary breach of fiduciary duty. Each contention raises tricky and unresolved issues in honest
services fraud jurisprudence.
Property Interest. Borges first argues that his indictment and jury instructions should
have required that a bribe be offered in exchange for a recognized property interest. He points to
a spate of Supreme Court cases suggesting that common-law fraud required a property interest.
Thus, his argument goes, honest services fraud, which is part of the traditional fraud statutes,
must require a property interest too. That’s a fair point. But the statute’s text and Supreme
Court interpretations make clear that no property interest is required to be guilty of honest
services fraud.
Supreme Court precedent tells us that “honest services fraud” does not involve a property
interest. In Skilling, the Court explained that, from the birth of the honest services fraud
doctrine, the betrayed party “suffered no deprivation of money or property.” 561 U.S. at 400.
Rather, the “actionable harm lay in the denial of that party’s right to the offender’s ‘honest
services.’” Id. (citation omitted).
Indeed, this much is clear from the statutory text: the mail- and wire-fraud statutes
criminalize using the mail or wires in furtherance of “any scheme or artifice to defraud.”
18 U.S.C. § 1341 (mail fraud), § 1343 (wire fraud). And the honest-services statute, § 1346,
clarifies that a “scheme or artifice to defraud” includes “a scheme or artifice to deprive another
of the intangible right of honest services.” Thus, the government need not show a deprivation of
a property interest to prove honest services fraud under § 1346.
4He also brings a sufficiency-of-the-evidence argument. See Per Curiam Op. at 37.

-- 58 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 59
In response, Borges argues that our caselaw mandates that honest services fraud involve a
property interest. He points to United States v. Frost, in which we said that “[d]espite the literal
terms of § 1346 . . . the intangible right to honest services in the private sector [is] ultimately
dependent upon the property right of the victim.” 125 F.3d at 369. But that case predates
Skilling, along with its explanation that honest services fraud involves non-property interests.
What’s more, Frost considered whether the defendant foresaw that he would cause economic
harm—a standard not involving property rights at all. See United States v. Douglas, 398 F.3d
407, 419 (6th Cir. 2005); see also United States v. Quintanilla, 114 F.4th 453, 470 (5th Cir.
2024) (rejecting argument that “honest services wire fraud and bribery require bribery
appurtenant to a property interest” (citation omitted)).
Borges’s last argument is that the federal wire and mail fraud statutes should only apply
to recognized property interests because the honest services fraud language is otherwise too
vague to be constitutional. In making this argument, he says we must reject the government’s
broad reading of the statute because its approach is out-of-step with the “structure and history” of
the federal fraud statutes. Ciminelli v. United States, 598 U.S. 306, 315 (2023).
The problem, though, is that Skilling’s reading of the honest services statute remains
good law. And, as a lower federal court, we can’t overrule Skilling. If a Supreme Court
precedent applies, “yet appears to rest on reasons rejected in some other line of decisions,” we
must “follow the case which directly controls.” Rodriguez de Quijas v. Shearson/Am. Express,
Inc., 490 U.S. 477, 484 (1989); see also Agostini v. Felton, 521 U.S. 203, 237 (1997) (same).
Thus, because Skilling still stands, Borges’s argument on this claim fails.
Foreseeable Economic Harm. Borges’s next argument concerns the jury instructions.
He argues that the district court erred when it instructed the jury that it could convict him of
honest services fraud if he could have reasonably foreseen that a breach of a fiduciary duty
would create an identifiable economic harm to the victim. Instead, he argues that the
government needed to show intent to create an economic harm to AMT. In making this
argument, Borges presses against yet another area of legal uncertainty: whether our reasonable
foreseeability test for honest services fraud, which this court created before Skilling, survives
Skilling.

-- 59 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 60
After Congress passed § 1346, lower courts adopted various tests for honest services
fraud. This circuit held that honest services fraud required proof that the defendant “foresaw or
reasonably should have foreseen that [the victim] might suffer an economic harm as a result of” a
payment. Frost, 125 F.3d at 368. Here, that would require that Borges foresaw or reasonably
should have foreseen that the victim—AMT—would suffer harm because Fehrman received the
bribe.
But Frost was pre-Skilling, and Skilling complicates things. In Skilling, the Court
emphasized that the honest services theory targets corruption when “the betrayed party [has]
suffered no deprivation of money or property.” Skilling, 561 U.S. at 400. Rather, the
“actionable harm [lies] in the denial of that party’s right to the offender’s ‘honest services.’”
Id. (citation omitted). To translate to the facts here, the injury lies in Borges’s attempt to get
Fehrman to deprive AMT of its right to Fehrman’s honest services.
So, does the “reasonable foreseeability” test survive Skilling? There are arguments in
favor: Skilling addressed whether § 1346 survived a vagueness challenge and said nothing to
discount a reasonable foreseeability requirement. Indeed, some courts continue to apply the
reasonable foreseeability test. See, e.g., United States v. Lusk, No. 2:15-CR-00124, 2017 WL
508589, at *11 (S.D. W. Va. Feb. 7, 2017).
On the other hand, dicta in Skilling could be read to eliminate the reasonable
foreseeability test. In Skilling, the Court wrote that honest services fraud applies to corruption
even when the victim enjoys “a money or property gain.” 561 U.S. at 400 (emphasis added). If
the victim could actually have a financial gain from the honest services fraud, then it makes no
sense to say a defendant must reasonably foresee harm to a business. For example, had Borges
promised Fehrman that he would give AMT business that would offset any loss from the
reputational harm to AMT, under Skilling, that would still constitute honest services fraud.
Thus, the reasonable foreseeability of economic harm isn’t a requirement of honest
services fraud. It follows, then, honest services fraud also doesn’t require an intent to cause
economic harm.

-- 60 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 61
Fiduciary Duty. Finally, Borges argues the government didn’t establish that Fehrman
owed a fiduciary duty to his employer, AMT. Thus, says Borges, Fehrman couldn’t have
breached any duty by taking a bribe from Borges. In so doing, Borges raise a question about
Skilling to which there is no good answer—namely, what “fiduciary duty” suffices to make
someone guilty of honest services fraud?
In Skilling, the Court explained “honest services fraud” encompassed a “solid core” of
cases “involv[ing] offenders who, in violation of a fiduciary duty,” participate in “bribery or
kickback schemes.” Id. at 407. The Court explained that pre-McNally duties were “usually”
beyond dispute. Id. at 407 n.41. What were some of those indisputable duties? The Court listed
a few examples in a single footnote: “public official-public, employee-employer, and union
official-union members.” Id. (cleaned up).
Here, because Fehrman was an employee of AMT, the government argues that Fehrman
had an employer-employee fiduciary duty.
In response, Borges asserts that (1) the fiduciary duty needs to spring from an
independent source of law, and (2) that the correct source of law is Ohio law, which says a
typical employer-employee relationship isn’t a fiduciary relationship.
To the first issue: what kind of fiduciary relationship suffices? Courts have struggled to
apply this part of Skilling’s call to look to “fiduciary” relationships. It’s not clear whether the
Court meant “fiduciary” in the technical sense or an informal sense.
On the one hand, the term “fiduciary” conjures up notions of trust law, corporate law, and
property law. And in those contexts, fiduciaries have defined meanings and structured
relationships.
On the other hand, courts routinely say that there can be informal fiduciaries. United
States v. Pappert, 112 F.3d 1073, 1080 (10th Cir. 1997) (“[T]here is not a bright line between
formal or informal fiduciary relationships, and run-of-the-mill commercial relationships.”). And
Skilling invoked the “established doctrine that a [fiduciary duty] arises from a specific
relationship between two parties.” 561 U.S. at 407 n.41 (alteration in original) (quoting

-- 61 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 62
Chiarella v. United States, 445 U.S. 222, 233 (1980)). Chiarella, in turn, compared a “fiduciary
duty” to “similar relation[s] of trust and confidence.” 445 U.S. 222, 228 (1980) (quotation
omitted). That citation suggests the Court meant fiduciary duty in a broad, non-technical sense,
to include duties of trust and confidence that arise out of a specific relationship.
There’s also a second issue. Assuming some fiduciary relationship is required, it would
need to be grounded in some source of law—and Skilling never said what that law was. Is it
federal, state, or common law? Courts are unsure. Some courts look only to state law. See, e.g.,
United States v. Brumley, 116 F.3d 728, 734 (5th Cir. 1997). Others look to common-law
principles or “inherent” duties they say stem from a given relationship, in effect creating a
federal common law of fiduciary duty. See, e.g., United States v. Nelson, 712 F.3d 498,
509 (11th Cir. 2013) (citation omitted). And some courts don’t analyze the source of the duty at
all. See, e.g., United States v. Nouri, 711 F.3d 129, 137 n.1 (2d Cir. 2013).
Which is right? It’s hard to say. A “duty” typically arises from a specific source of
law—usually state law. For example, in bankruptcy, courts often “must look to state law” when
considering whether a fiduciary duty exists. See In re Interstate Agency, Inc. 760 F.2d 121,
124 (6th Cir. 1985); Aguillino v. United States, 363 U.S. 509, 512–13 (1960). What’s more,
Skilling purported to narrow the pre-McNally caselaw to a “core” that foreclosed broad liability.
501 U.S. at 407. So, perhaps the “fiduciary duty” language is narrower than the pre-McNally
general relationship of trust and confidence.
On the other hand, Skilling itself is more functionalist than formalist. It lists a broad
range of “fiduciary duties,” including public official-public, employee-employer, and union
official-union member. Id. at 407 n.41. It also cites Chiarella as helping define a fiduciary duty.
And Chiarella itself equates such a duty with a broad relationship of trust and confidence.
Further, many pre-McNally cases didn’t define sources of duties.
Borges, for his part, approaches his argument from a formalist reading of Skilling. His
argument goes like this: first, we should look to state law for the requisite fiduciary duty. In
Ohio, the average employee is not a fiduciary. See, e.g., Lombardo v. Mahoney, No. 92608,

-- 62 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 63
2009 WL 3649997, at *4 (Ohio Ct. App. 2009). Thus, Borges says, he didn’t try to make
Fehrman breach a fiduciary duty. So he can’t be liable under Skilling.
But even if Borges is right—and he may well be—his claim fails. Why? Until the
Supreme Court says otherwise, Skilling is good law. And Skilling appears to establish that an
employee-employer relationship is sufficient regardless of what state law says. 561 U.S. at 407
n.41. Borges is right that this arguably creates a federal common law of fiduciary relationships,
and that such a federal common law can’t exist. Cf. United States v. Hudson & Goodwin,
11 U.S. (7 Cranch) 32 (1812) (explaining that there is no federal common law of crimes). But,
as a lower court, we are bound to follow it until the Supreme Court says otherwise.
*
All told, Borges raises compelling arguments about his honest services predicate. After
all, the Supreme Court has explained that “the intangible right of honest services must be defined
with the clarity typical of criminal statutes and should not be held to reach an ill-defined category
of circumstances.” Percoco v. United States, 598 U.S. 319, 328–29 (2023) (cleaned up). And
here, Borges has a good argument his conduct fell within a murky middle: perhaps
objectionable, but not clearly illegal. Until the Supreme Court revises its caselaw, however, we
must follow its precedent.
C.
Borges also challenges his money laundering and public-official bribery predicate
offenses. Start with money laundering under 18 U.S.C. § 1957. That law forbids using money
derived from unlawful activity. In particular, the statute requires that (1) the defendant
knowingly attempted to engage in a monetary transaction; (2) the defendant knew the transaction
involved property or funds that were the proceeds of some criminal activity; (3) the property had
a value of more than $10,000; (4) the property was in fact proceeds of a specific unlawful
activity, and (5) the transaction took place in the United States. Id.
Borges argues that his funds didn’t sit downstream of anything illegal, so they weren’t
“proceeds of a criminal activity.” In making these arguments, he covers much of the ground that

-- 63 of 64 --

Nos. 23-3565/3566 United States v. Householder, et al. Page 64
Householder does, contending that there weren’t any bribes at all. But for the reasons stated in
the per curiam opinion, there were.
* * *
At bottom, Borges asks whether it’s right for a single individual, who made a $15,000
payment, to be on-the-hook for the same $60 million RICO conspiracy charge that Householder
faced. That’s a policy question, and it’s one that defendants and scholars have raised when faced
with cases like this. Cf. Gerard E. Lynch, RICO: The Crime of Being a Criminal, Parts I & II,
87 Colum. L. Rev. 661, 664 (1987).
But we are a court of law. By passing RICO, Congress determined that the United States
should punish large criminal enterprises. In so doing, it enacted sweeping statutes that implicate
all actors in broad conspiracies. That was its prerogative. Our task, on the other hand, “is to
interpret the statute as best we can, not to second-guess the wisdom of the congressional policy
choice.” Mansell v. Mansell, 490 U.S. 581, 594 (1989).
Thus, when faced with a defendant in Borges’s situation, we can’t circumvent Congress’s
mandate and the Supreme Court’s interpretation of it. All we can do is follow the law. Here,
following the law means affirming Borges’s sentence. Thus, I concur in the per curiam opinion
doing just that.

-- 64 of 64 --