RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 25a0031p.06

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

┐ UNITED STATES OF AMERICA, │ Plaintiff-Appellee, │ > No. 23-3840 │ v. │ │ ALEXANDER SITTENFELD aka P.G. Sittenfeld, │ Defendant-Appellant. │ ┘

Appeal from the United States District Court for the Southern District of Ohio at Cincinnati. No. 1:20-cr-00142-1—Douglas Russell Cole, District Judge.

Argued: May 9, 2024

Decided and Filed: February 11, 2025

Before: BUSH, NALBANDIAN, and MURPHY, Circuit Judges. _________________

COUNSEL ARGUED: Yaakov M. Roth, JONES DAY, Washington, D.C., for Appellant. Matthew Singer, UNITED STATES ATTORNEY’S OFFICE, Cincinnati, Ohio, for Appellee. ON BRIEF: Yaakov M. Roth, Harry S. Graver, Ryan M. Proctor, JONES DAY, Washington, D.C., James M. Burnham, KING STREET LEGAL, PLLC, Washington, D.C., for Appellant. Matthew Singer, Alexis J. Zouhary, Emily N. Glatfelter, Megan Gaffney Painter, UNITED STATES ATTORNEY’S OFFICE, Cincinnati, Ohio, for Appellee. L. Bradfield Hughes, PORTER, WRIGHT, MORRIS & ARTHUR, LLP, Columbus, Ohio, Joseph O. Masterman, COOPER LAW PARTNERS, Washington, D.C., Joshua H. Runyan, Patrick F. Linehan, Nicholas P. Silverman, Jennie A. Askew, M. Vito Arethusa, STEPTOE LLP, Washington, D.C., Charles C. Speth, WILMER CUTLER PICKERING HALE AND DORR LLP, Washington, D.C., Steven Garrett Tegrar, DEBEVOISE & PLIMPTON LLP, New York, New York, David A. O’Neil, DEBEVOISE & PLIMPTON LLP, Washington, D.C., John S. Moran, MCGUIRE WOODS, Washington, D.C., for Amici Curiae.

NALBANDIAN, J., delivered the opinion of the court in which MURPHY, J., concurred. MURPHY, J. (pp. 39–52), delivered a separate concurring opinion. BUSH, J. (pp. 53–71), delivered a separate dissenting opinion. No. 23-3840 United States v. Sittenfeld Page 2

_________________

OPINION _________________

NALBANDIAN, Circuit Judge. Every day in this country, politicians solicit donations to finance their campaigns. And every day, those same politicians make statements about what they believe in, what they’ve done, and what they promise to do once elected. Sometimes, even often, these solicitations and promises occur in the same place, at the same time. But though this speech and conduct are generally protected by the First Amendment, bribery remains illegal.1 When the bribery involves money flowing to a politician for his personal use, the crime is straightforward. But when a politician is accused of accepting campaign funds in exchange for the promise of official action, the line becomes blurrier. Still, the Supreme Court tells us there is a line. And Congress and the Court have entrusted juries with discerning between legitimate campaign donations and illegitimate bribes. We must respect that line even in hard cases.

This is one such case. A jury convicted former Cincinnati council member Alexander “P.G.” Sittenfeld of attempted Hobbs Act extortion and federal-program bribery. But this case comes with twists. All the major players, except for Sittenfeld, were working for or with the government—that is, these were paid actors working to incriminate Sittenfeld. And despite nearly every relevant conversation being recorded, the investigation didn’t yield overwhelming evidence. Still, a jury found that Sittenfeld solicited or accepted campaign donations in exchange for his promise to support a property development project. On appeal, Sittenfeld challenges the sufficiency of the evidence against him and argues that his indictment was constructively amended. But neither challenge succeeds, so we AFFIRM.

1 The Supreme Court has recently offered helpful context about how illegal bribes arise. Federal and state law distinguish between two kinds of payments to public officials— bribes and gratuities. As a general matter, bribes are payments made or agreed to before an official act in order to influence the official with respect to that future official act. American law generally treats bribes as inherently corrupt and unlawful. Snyder v. United States, 144 S. Ct. 1947, 1951 (2024). No. 23-3840 United States v. Sittenfeld Page 3

I.

A.

In 2018, Sittenfeld was a Cincinnati city council member who planned to run for mayor. Chinedum Ndukwe was a local developer who sometimes discussed his development projects with Sittenfeld. In September 2018, Sittenfeld called Ndukwe and let him know that “the majority of the developers in Cincinnati [we]re going to be giving [him] ten grand.” R. 266, Trial Tr. Day 6, pp. 59–60, PageID 6242–43. Sittenfeld then asked Ndukwe, “[C]an I count on you for ten?” Id. Although Ndukwe had known Sittenfeld since 2010, and had consistently donated to Sittenfeld’s past campaigns, Ndukwe claimed that this fundraising request was “jarring.” Id. at 60–61, PageID 6243–44. And what Sittenfeld didn’t know was that Ndukwe was working for the FBI at the time of that conversation.

Earlier in January 2018, the FBI had flagged Ndukwe as a possible source for some of their ongoing public-corruption cases. Ndukwe had previously given “money orders and cashier’s checks to local politicians in other individuals’ names,” so FBI agent Nathan Holbrook pursued Ndukwe as an asset. R. 264, Trial Tr. Day 4, p. 84, PageID 5888. By mid-March 2018, Ndukwe had agreed to provide information to the FBI in exchange for a commitment by the government not to pursue a criminal investigation against him. From there, Ndukwe began informing on Sittenfeld to Holbrook, starting with his September conversation with Sittenfeld. At that point, Holbrook told Ndukwe to record all his calls with Sittenfeld.

To catch Sittenfeld taking a bribe, the FBI introduced undercover agents, “Rob” and “Brian,” who pretended to be investors in Ndukwe’s real estate development project at 435 Elm Street in downtown Cincinnati. As it stood, the 435 Elm property was dilapidated, costing Cincinnati taxpayers hundreds of thousands of dollars each year to maintain. Ndukwe wanted to create a mixed-use development project on the site and had bought a leasehold mortgage note from the bank and air rights for it. But Sittenfeld already had 435 Elm on his radar because Ndukwe had discussed the project with him earlier in 2018. And separately in 2017, someone else had asked Sittenfeld about 435 Elm and he had directed them to the City of Cincinnati. No. 23-3840 United States v. Sittenfeld Page 4

Early in the investigation, Ndukwe had three important conversations with Sittenfeld, each recorded. The first call was on October 26, 2018, when Ndukwe told Sittenfeld that Rob and Brian would be a source of capital for 435 Elm. They discussed setting up a meeting with Sittenfeld so Rob and Brian could donate to the campaign before campaign-finance rules changed on November 6, 2018. Ndukwe agreed to try to schedule something before that date.

A second, more pivotal conversation took place on October 30, 2018. Ndukwe broke some bad news—Rob wouldn’t be in town until November 7, 2018. Instead, Ndukwe said he could try to “get some of [his] friends up in Columbus” to support Sittenfeld. R. 312, Appeal Exs., p. 8, PageID 7551. Ndukwe explained that he was trying not to donate in his own name, and Sittenfeld encouraged him to round up donations from others:

Sittenfeld: Just so, just so you know like, look I have, you know I, I love what you do as someone revitalizing our city creating jobs. I am fond of you as a friend. I also have like you know obligations to do the things I need to do to be a successful candidate so. Ndukwe: Absolutely. Sittenfeld: So, but what that means is I don’t really get like, if if you say look I don’t want to support you in the name of Chinedum Ndukwe, but some guy I’ve never met from Columbus is going to use a coup, you know, you know [your] network are going to a, round up a bunch of LLC checks. Like that’s great. I actually don’t care. But I mean the one thing I will say is like, you know I mean, you don’t want me to like be like ‘hey Chin like love you but can’t’ you know like, you know, I mean like, you know like. I, I, I want people to support me, that’s like . . . Ndukwe: Absolutely. Sittenfeld: . . . if a candidate doesn’t want people to support them, they’re a shitty dumb candidate . . . Ndukwe: Yeah, right, yeah, right. Sittenfeld: . . . and you know I’ve been (UI [unintelligible]) a lot of people have come through in a really big way that’s been awesome so far and I would love, I would love for you to be one of those people too. No. 23-3840 United States v. Sittenfeld Page 5

Ndukwe: I hear ya. I hear ya. So we’ll, we’ll figure, we’ll, we’ll make sure, trust me, we’ll, we’ll, we’ll make something happen sooner than later too. Um (UI). Sittenfeld: Well can you (UI), Can you even do it before the LLC thing? Ndukwe: I don’t know, let me, let me, let me, let me touch base with Jay again and see how, how we can make it work. Uhm I mean honestly just with, with everybody I’ve been trying to, you know, the other thing is these guys have a ton of LLCs. But I, I know that he’s not going to be in town . . . so when is the drop dead date? Is it the sixth? Sittenfeld: They can’t, those guys can’t use more than one LLC for each of them after, after the sixth, yeah. That’s why, so just to let you know North American has done twelve in LLCs, Uptown has done ten, Medpace has done ten, Model has done eight. Uhm I, I need to go down like the whole list. Ndukwe: Yeah, yeah, yeah. Sittenfeld: You know Eli has done five. So because people are like, you know, we got to get this done before this (UI) option goes away. Ndukwe: Yeah, yeah, yeah, for sure, for sure. Well let me, let me see— Sittenfeld: Even, even if you were able to like, you know, Columbus people, these guys round up five LLCs before next Tuesday it’d be big. Ndukwe: Yeah, okay, okay. All right, let me, let go to work on that. Let me go to work on that.

Id. at 8–9, PageID 7551–52 (ellipses in original) (emphasis added). Ndukwe and Sittenfeld then set up a time to meet with Rob. Just before the end of the call, Sittenfeld reiterated, “[Y]ou’re gonna deliver the goods before next Tuesday,” and Ndukwe responded, “All right, let me go to work on that.” Id. at 10, PageID 7553.

At trial, witnesses offered different interpretations of the “love you but can’t” comment. Ndukwe testified that he understood it to mean “that whether you donate or don’t donate,” it would “have an impact on” Sittenfeld’s “advocacy” for him. R. 266, Trial Tr. Day 6, p. 64, PageID 6247. In other words, Ndukwe thought it “was very clear that if I donated, he was going to support and be supportive in my efforts, and if I didn’t, he wasn’t going to be supportive.” Id. Sittenfeld, however, claimed that he was explaining, “I’m only going to be around here if I’m No. 23-3840 United States v. Sittenfeld Page 6

successful in this race.” R. 269, Trial Tr. Day 10, p. 110, PageID 6746. Due to term limits, Sittenfeld was in his last city council term, so to continue “advanc[ing] projects that were good for the city” he had to become mayor. Id.

After reviewing that call, the FBI instructed Ndukwe to talk with Sittenfeld again and “make a very clear and obvious offer of money in exchange for votes by Mr. Sittenfeld on 435 Elm.” R. 263, Trial Tr. Day 3, p. 55, PageID 5700. Ndukwe followed that advice in a third call on November 2, 2018. Ndukwe told Sittenfeld that the November 6 deadline would impact the donations he could solicit, but he could provide “close to twenty thousand” in “the next couple of weeks.” R. 312, Appeal Exs., p. 13, PageID 7556. Ndukwe then asked Sittenfeld to agree to a quid pro quo:

Ndukwe: . . . so and then for, for, and then for this meeting with Rob next week, I’m pretty sure he can get you ten this week. You know the biggest thing is, you know, if we do the ten, I mean, they’re gonna want to know that when it comes time to vote on 435 Elm, like whenever that, I don’t know if it’s next year, two years, three years, that it’s gonna be a yes vote, you know, without, without a doubt. I’ve shared that with them, that hey [I’ve] known PG for years, all this stuff, but they’re like all right we’ll get his attention. Sittenfeld: I mean, obv-, as you know, obviously nothing can be illegal like . . . illegally nothing can be a quid, quid quo pro [sic]. And I know that’s not what you’re saying either. But what I . . . Ndukwe: Yeah. Sittenfeld: . . . can say is that I’m always super pro-development and revitalization of especially our urban core. Ndukwe: Okay, no, I hear ya. I hear ya. And so they’re, he’ll probably come out— Sittenfeld: And we can, we, we, we can discuss that more in person. Ndukwe: Okay, okay. My guy, perfect, perfect. Sittenfeld: But I’m not, I’m not sure, I’m not they’re, I, in seven years I have voted in favor of every single development deal that’s ever been put in front of me so.

Id. at 13–14, PageID 7556–57 (ellipses in original) (emphasis added). No. 23-3840 United States v. Sittenfeld Page 7

At trial, Sittenfeld described this conversation as surprising, because he “had never once known [Ndukwe] to do something” that “sounded untoward or unethical.” R. 269, Trial Tr. Day 10, p. 111, PageID 6747. After Sittenfeld expressed his confidence that Ndukwe did not mean to propose a quid pro quo and Ndukwe agreed, Sittenfeld “thought that put the issue to bed.” Id. at 112, PageID 6748. Sittenfeld also claimed he thought Ndukwe “was probably misrepresenting” Rob. Id. Regardless, Ndukwe testified that he still thought that Sittenfeld’s “true intentions” involved bribery. See R. 266, Trial Tr. Day 6, p. 106, PageID 6289.

On November 7, 2018, Sittenfeld met with Rob and Ndukwe at Nada, a Cincinnati restaurant. There, Rob spoke about the investment group he represented. Ndukwe talked about his plans for 435 Elm and his idea of getting it from the city for a dollar or “a minimal amount.” R. 312, Appeal Exs., pp. 21–23, PageID 7564–66. Sittenfeld said he could “certainly shepherd the votes” and “promise the votes.” Id. at 24, 30, PageID 7567, 7573. Sittenfeld described the project as “a strategic development,” given that it was across the street from the convention center and “easy to support.” Id. at 27, 30, PageID 7570, 7573.

Later, still at Nada, Sittenfeld transitioned to fundraising. He noted that Rob would be “making good bets and good investments” by supporting him. Id. at 35, PageID 7578. Sittenfeld showed Rob three slides on his laptop, illustrating his broad support in Cincinnati. And Sittenfeld claimed that successful developers and business leaders had “already placed their bet with [him],” so “if anyone d[id] that going forward, they’re in good company.” Id. at 37, PageID 7580.

After hearing Sittenfeld’s pitch, Rob transitioned back to 435 Elm and explained that “we wanna try to set 435 up being veto-proof you know.” Id. at 38, PageID 7581. Sittenfeld responded that he could “move more votes than any single other person,” even the mayor, but he did not think the mayor would oppose the deal. Id. at 38–39, PageID 7581–82.

Sittenfeld and Rob then headed to Rob’s nearby apartment. Rob explained that his investors liked to donate anonymously and so tried to set up a bribe:

UC Rob: So [Chin] doesn’t want his name on anything. Sittenfeld: Right. No. 23-3840 United States v. Sittenfeld Page 8

UC Rob: Um you know we usually try to figure out a creative way to do something. Sittenfeld: Right, understood. UC Rob: Um, but at the same time, like we want to help, we want to make sure, we want to really get this thing, [Mayor] Cranley, I would feel comfortable tellin’ my guys like hey we’re we’re in . . . Sittenfeld: Right. UC Rob: . . . this deal with Chin [Ndukwe] if I know we’re Cranley-proof. Sittenfeld: Yeah. UC Rob: Um and so with that like Chin told me like hey you know I want to try to get uh P.G. 20,000 . . . Sittenfeld: Right. UC Rob: . . . and I’m like, hey man I I’m I’m if if if we can get this deal done, like fuckin’ let’s do it. Sittenfeld: Yeah. UC Rob: You know and so. Sittenfeld: Do you guys . . . UC Rob: What is the best way, what’s the best way for us to get that to you, to get that deal? You know what I mean, like . . . Sittenfeld: Yeah yeah yeah. I’m just, do you guys know that he’s gonna try and veto it?

Id. at 44, PageID 7587 (ellipses in original). Rob said Cranley was fine with the project so long as Ndukwe was not the face of it. Sittenfeld replied, “Honestly . . . I can sit here and say I can deliver the votes.” Id. at 45, PageID 7588. He said, “I can get it done.” Id.

Rob then jumped back to money. He said he had $10,000 cash on him and it would be easiest to just give it to Sittenfeld. So the two talked about the best way for Sittenfeld to accept the donation. At one point, Sittenfeld asked, “[W]hose name can stuff be in?” Id. at p. 46, PageID 7589. Rob said, “[W]e can come up with some names.” Id. Later, Sittenfeld talked about the “technicality” of attributing a donation to an individual and having that person “agree No. 23-3840 United States v. Sittenfeld Page 9

that that’s where it came from.” Id. at 47, PageID 7590.2 Sittenfeld did not want to “send up signals,” so he got advice from his political consultant who was helping him with fundraising at the time. Id. at 48–49, PageID 7591–92.

In the end, Sittenfeld decided not to take the cash. Instead, he and Rob arranged to transfer the funds through money orders. Sittenfeld explained that he did not want to “cause a headache” and preferred “to proceed” in this way “out of an abundance of caution.” Id. at 50, PageID 7593. Rob offered to mail the money orders to Ndukwe instead of directly to Sittenfeld. At the end of their meeting Rob said, “If you can pull those votes together that’s awesome,” and Sittenfeld responded, “Yeah . . . we definitely can.” Id. at 53, PageID 7596.

But Sittenfeld changed his mind. On November 21, 2018, Sittenfeld told Rob he could not take the donation through money orders. As an alternative, Sittenfeld proposed that Rob give to his political action committee (PAC) since “no one [wa]s going to be poking around . . . to find your names on it.” Id. at 55, PageID 7598.3

Sittenfeld, Rob, and Brian met a week later. Sittenfeld asked about 435 Elm and expressed his readiness “to shepherd the votes as soon as it gets to us at Council.” Id. at 59, PageID 7602. The FBI agents gave Sittenfeld two checks worth $5,000, payable to the PAC. Sittenfeld assured Rob that Sittenfeld’s “name [wa]s not connected to [the PAC] in any way” and that “[n]o one w[ould] ever know [about] this.” Id. at 60, PageID 7603. But on December 3, 2018, Sittenfeld spoke with Rob on the phone and alerted him to the fact that the PAC couldn’t accept the checks he’d sent because they were from corporations. The PAC could only accept checks from LLCs.

Money finally changed hands when Sittenfeld met with Rob and Brian on December 17, 2018. Rob gave Sittenfeld four checks in total worth $20,000. Each was from a different LLC. Those checks were later deposited into Sittenfeld’s PAC. At the same meeting, Sittenfeld again talked about 435 Elm. He expressed confidence that it would “happen” and emphasized his

2 Despite those cavalier statements, Sittenfeld testified at trial that he would expect fundraisers and donors to tell the truth about the actual source of a donation. 3 Sittenfeld mainly used the PAC to donate to candidates and organizations that he believed in. No. 23-3840 United States v. Sittenfeld Page 10

influence: “[D]on’t let these be my famous last words, but, um, I can always get a vote to my left or a vote to my right.” Id. at 70, PageID 7613. That same day Sittenfeld left a voicemail message for Ndukwe; he was “looking forward to doing what [he could] to help get 435 across the finish line, so never hesitate to reach out.” R. 263, Trial Tr. Day 3, p. 135, PageID 5780.

Sittenfeld, Rob, and Brian stayed in touch during 2019 and talked about 435 Elm several more times.4 Sittenfeld continued to accept more money from Rob. To justify his engagement with Rob and Brian, Sittenfeld said at trial that he engaged with people “from all walks of life.” R. 269, Trial Tr. Day 10, p. 121, PageID 6757. He saw Brian and Rob as “backing something that was good for the city,” and he trusted Ndukwe. Id.

B.

In November 2020, a grand jury indicted Sittenfeld on six counts. Counts 1 and 2 charged honest-services wire fraud, in violation of 18 U.S.C. §§ 1343 and 1346. Counts 3 and 5 alleged bribery concerning a program receiving federal funds, in violation of 18 U.S.C. § 666. And counts 4 and 6 charged attempted Hobbs Act extortion, in violation of 18 U.S.C. § 1951. Each charge was tied to a date range, with counts 3 and 4 limited to activities between “about September 21, 2018, and December 17, 2018.”5

Count 3 of the indictment charged Sittenfeld with obtaining money from Rob: “[T]o wit, the defendant . . . corruptly solicited and demanded, and accepted and agreed to accept, payments to PAC for his benefit from UCE-1 . . . .” R. 3, Indictment, pp. 16–17, PageID 41–42. Count 4 had similar language: “[T]o wit, the defendant . . . solicited, obtained, agreed to accept, accepted, and received payments to PAC from UCE-1 . . . .” Id. at p. 17, PageID 42. The parties agree that UCE-1 was Rob. Counts 3 and 4 incorporated by reference Sittenfeld’s calls with Ndukwe on October 26, October 30, and November 2, 2018.

4 Sittenfeld’s 2019 meetings were relevant to other charges against him, but we do not describe them in detail here because his convictions were based on the 2018 quid pro quo. 5 Counts 1 and 2 concerned conduct between about September 21, 2018, and February 4, 2020, and specific phone calls on November 21, 2018, and September 25, 2019. And counts 5 and 6 concerned conduct between about July 8, 2019, and February 5, 2020. The jury instructions repeated these date ranges. No. 23-3840 United States v. Sittenfeld Page 11

The case went to trial in June 2022. Over the next two weeks, the government played audio and video recordings alongside testimony from Rob, Brian, Ndukwe, Sittenfeld, and other witnesses. For each count, the jury was told that the government had to prove an explicit (but not express) quid pro quo.

For the federal-program bribery counts, the instruction required the jury to find “that Mr. Sittenfeld solicited, demanded, accepted, or agreed to accept a thing of value from another person.” R. 251, Trial Tr. Day 11, pp. 34, 51–52, PageID 4974, 4991–92. And the Hobbs Act instruction required the jury to find “that Mr. Sittenfeld obtained, accepted, agreed to accept, or received property that he was not lawfully entitled to from another person with that person’s consent.” Id. at 38, PageID 4978.

On July 8, 2022, the jury returned a mixed verdict. It convicted Sittenfeld of counts 3 and 4 but acquitted him of the remaining charges. Sittenfeld moved for a judgment of acquittal under Rule 29 of the Federal Rules of Criminal Procedure and for a new trial under Rule 33. On the motion for judgment of acquittal, he claimed the government had not proven the existence of a “quid pro quo,” or an “agreed-upon official act,” and that the two statutes under which he was convicted were unconstitutional. R. 283, Op. & Order, pp. 2–3, PageID 7137–38. The court denied each claim.

The court also denied Sittenfeld’s motion for a new trial. He argued that the verdict was against the manifest weight of the evidence, that the government constructively amended his indictment, and that the district judge had issued erroneous jury instructions. The court again rejected each claim. This was the first time Sittenfeld had argued that the government exceeded the scope of the indictment. It was also the first time Sittenfeld argued that the jury instructions were erroneous because they suggested counts 3 and 4 could rest on a solicitation of “another person,” not just Rob, as the indictment required.

Ultimately, the district court sentenced Sittenfeld to sixteen months’ imprisonment.6 He timely appealed. Now he makes two arguments. First, he claims the government did not present

6 We granted Sittenfeld release pending appeal on May 15, 2024, after he had begun serving his sentence. No. 23-3840 United States v. Sittenfeld Page 12

sufficient evidence for the jury to find an explicit quid pro quo. Second, he argues that the government constructively amended the indictment. We address each in turn.

II.

We evaluate sufficiency claims de novo, asking “whether, after reviewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” United States v. Emmons, 8 F.4th 454, 477–78 (6th Cir. 2021) (internal quotation marks omitted). The evidence “need not remove every reasonable hypothesis except that of guilt.” United States v. Sadler, 24 F.4th 515, 539 (6th Cir. 2022) (internal quotation marks omitted). And this court may not “reevaluate the credibility of witnesses.” Id. (internal quotation marks omitted).

A.

Two Supreme Court cases guide our understanding of what the government must prove to show Hobbs Act extortion and federal-program bribery in the campaign-contribution context: McCormick v. United States, 500 U.S. 257 (1991), and Evans v. United States, 504 U.S. 255 (1992). Combined with controlling precedent from our circuit, it’s clear that the government must show that an elected official received campaign donations “in return for an explicit promise or undertaking.” McCormick, 500 U.S. at 273 (Hobbs Act); United States v. Inman, 39 F.4th 357, 365 (6th Cir. 2022) (federal-program bribery). That is, we ask whether the government presented sufficient evidence to show an explicit quid pro quo. Resolving that query turns on the form of the government’s evidence.

First, McCormick explained that proof of a quid pro quo was a prerequisite to a conviction for extortion under the Hobbs Act. In the early 1980s, a member of the West Virginia House of Delegates was convicted for violating the Hobbs Act. He had accepted cash payments, which he claimed were campaign contributions, in exchange for sponsoring legislation. 500 U.S. at 259–60, 268. The court of appeals affirmed his conviction by distinguishing between legal and illegal campaign contributions. Importantly, the appellate court concluded that a contribution could violate the law even without an explicit quid pro quo, so long as they “were No. 23-3840 United States v. Sittenfeld Page 13

never intended to be legitimate campaign contributions.” Id. at 271 (internal quotation marks omitted).

The Supreme Court reversed. It started by acknowledging the practical realities of campaigning: a legislator must finance his race while also engaging in the “everyday business of a legislator,” like supporting legislation and serving his constituents. Id. at 272. “Money is constantly being solicited on behalf of candidates, who run on platforms and who claim support on the basis of their views and what they intend to do or have done.” Id. Given this close relationship between campaign promises and financing campaigns, the Court made clear that legislators do not commit extortion merely by “act[ing] for the benefit of [their] constituents” shortly before or after soliciting or receiving campaign contributions. Id.

But the Court was unwilling to say that it was “impossible for an elected official to commit extortion in the course of financing” his election. Id. at 273. Instead, he can commit extortion “under color of official right, but only if” he receives payments in exchange for an “explicit promise or undertaking” to “perform or not to perform an official act.” Id. (emphasis added). In such a case, the official effectively asserts that “his official conduct will be controlled by the terms of the promise or undertaking.” Id. (emphasis added). Thus “proof of a quid pro quo [is] essential” to a conviction. Id. (internal quotation marks omitted).

The Supreme Court clarified McCormick the next year in Evans. 504 U.S. 255. At issue were alleged bribes paid to a member of the Board of Commissioners of DeKalb County, Georgia, in exchange for favorable zoning policies. An FBI agent, posing as a real estate developer, sought the commissioner’s help in rezoning a 25-acre tract of land for high-density residential use. Id. at 257. At a meeting, the FBI agent handed the commissioner $7,000 in cash and a check for $1,000 payable to his campaign—though there was no evidence that the commissioner had initiated the transaction. The commissioner reported the check on his campaign-finance disclosure forms but not the cash. For these actions, a jury convicted him of violating the Hobbs Act. Id.

Evans directly answered whether “passive acceptance” of a benefit—rather than an affirmative step toward fulfillment of the quid pro quo—sufficed for a Hobbs Act extortion No. 23-3840 United States v. Sittenfeld Page 14

conviction. Id. at 266–67. The Court answered yes for someone acting “under color of official right,” at least where there was evidence that the official knew that the payment was in exchange for some official action. Id. at 261, 265.

Importantly, we have understood Evans to clarify that McCormick’s quid pro quo is “satisfied by something short of a formalized and thoroughly articulated contractual arrangement,” so “merely knowing the payment was made in return for official acts is enough.” United States v. Blandford, 33 F.3d 685, 696 (6th Cir. 1994). This clarification “gave content to what the McCormick quid pro quo entails.” Id.

To synthesize Evans and McCormick, Blandford relied on Justice Kennedy’s Evans concurrence. He provided a crucial fifth vote and reiterated that the quid pro quo was an “essential element of the offense.” Evans, 504 U.S. at 275 (Kennedy, J., concurring in part and concurring in the judgment). But the form of the quid pro quo is not dispositive. The official need only “intend[] the payor to believe that absent payment the official is likely . . . to give the prospective payor less favorable treatment if the quid pro quo is not satisfied.” Id. at 274. Thus “the official and the payor need not state the quid pro quo in express terms, for otherwise the law’s effect could be frustrated by knowing winks and nods.” Id. Instead, the “inducement from the official is criminal if it is express or it is implied from his words and actions, so long as he intends it to be so and the payor so interprets it.” Id.

We interpreted Justice Kennedy’s concurrence as a “gloss on the McCormick Court’s use of the word ‘explicit’ to qualify [the] quid pro quo requirement.” Blandford, 33 F.3d at 696. Explicit “speaks not to the form of the agreement,” but to the “degree to which the payor and payee were aware of its terms.” Id. Thus, we read Evans to mean that “by ‘explicit’ McCormick did not mean ‘express,’” id., and therefore we do not require unambiguous evidence, so long as the jury can infer the content of the quid pro quo. See United States v. Terry, 707 F.3d 607, 612– 13 (6th Cir. 2013) (“[T]he question is one of inferences taken from what the participants say, mean and do, all matters that juries are fully equipped to assess.”). It’s for this reason that “‘[m]otives and consequences, not formalities,’ are the keys for determining whether a public official entered an agreement to accept a bribe.” Id. at 613 (quoting Evans, 504 U.S. at 274 (Kennedy, J., concurring in part and concurring in the judgment)). It’s also for this reason that No. 23-3840 United States v. Sittenfeld Page 15

circumstantial evidence can prove a quid pro quo. Illegal inducement can be “implied from [an official’s] words and actions.” Blandford, 33 F.3d at 696 (quoting Evans, 504 U.S. at 274 (Kennedy, J., concurring in part and concurring in the judgment)); see also United States v. Benjamin, 95 F.4th 60, 68–69 (2d Cir. 2024) (collecting cases holding that a quid pro quo can be implied). So it is enough “if the public official understood that he or she was expected to exercise some influence on the payor’s behalf as opportunities arose.” Terry, 707 F.3d at 612 (internal quotation marks omitted). In other words, there must be a meeting of the minds and specific, agreed-upon terms that could be reduced to writing.

But the government need not prove that the parties ever spoke about or wrote down those terms. “What is needed is an agreement, full stop, which can be formal or informal, written or oral.” Id. at 613. To be sure, like an ordinary contract, there must be an exchange—payment “in return for an explicit promise.” McCormick, 500 U.S. at 273. The bribe payor must give a gift to obtain the promise, and the bribe recipient must make the promise to obtain that gift. But a corrupt quid pro quo requires more. For a donation to become a bribe, the parties must understand that “official conduct will be controlled by” the bribe. Id.; Terry, 707 F.3d at 613 (“On the other hand, if a donor . . . makes a contribution so that an elected official will ‘do what I asked him to do,’ and the official . . . accepts the payment with the same understanding, the donor and the official have formed a corrupt bargain.” (citation omitted)).

As with any contract, the public official must bind himself with some additional promise that the gift has induced. And as with any contract, that means the public official must receive some consideration for his promise. See Oliver Wendell Holmes, The Common Law 293–94 (1881); Terry, 707 F.3d at 612 (quoting McCormick, 500 U.S. at 273). That consideration may simply be “because of this gift I will now be sure to keep my campaign-trail promise.” But if the donor delivers a gift not expecting the public official to “alter his position in any way,” then consideration is lacking. See Holmes, supra, at 294. A donor who walks away with vague hopes has not paid a bribe, even if those hopes later come to fruition. Terry, 707 F.3d at 613.

Thus, state of mind is at the heart of the inquiry and we rely on objective manifestations of intent to show a defendant’s subjective mindset. Restatement (2d) of Contracts § 71 cmt. b; see also Evans, 504 U.S. at 274 (Kennedy, J., concurring in part and concurring in the judgment). No. 23-3840 United States v. Sittenfeld Page 16

This means first, undercover agents can enter into bribery agreements even if they disguise their true motives, and second, a quid pro quo may exist even if the official does not uphold his end of the bargain. United States v. Carmichael, 232 F.3d 510, 520 (6th Cir. 2000); see also Terry, 707 F.3d at 613. It is bribery to knowingly accept a bribe even if the official does not intend to be influenced by the bribe in any official act despite the claimed promise to the contrary. See United States v. Abbey, 560 F.3d 513, 517–19 (6th Cir. 2009), abrogated on other grounds by Snyder, 144 S. Ct. 1947. So Sittenfeld ended up in a kind of Truman Show, accepting bogus bribes from pretend investors to aid a speculative development. But a jury could still convict him if he played along. See Evans, 504 U.S. at 268 (“[T]he offense is complete[] at the time when the public official receives a payment in return for his agreement to perform specific official acts. . . .”).

B.

Sittenfeld attempts to recast McCormick and Evans to require explicit evidence of an agreement and unambiguous evidence of the quid pro quo. According to Sittenfeld, the evidence is insufficient as a matter of law if it is susceptible to a “legitimate explanation”; holding otherwise would conflict with “the entire point” behind McCormick. Appellant Br. at 18. He doesn’t believe a jury can handle the responsibility of deciphering the “common sense” understanding of a remark’s “implication.” Id. at 18, 28.

But Sittenfeld’s view of McCormick is too limited. Campaign contributions will almost always have an inherent “legitimate alternative explanation.” Terry, 707 F.3d at 613. As the Supreme Court has already recognized,

to hold that legislators commit . . . extortion when they act for the benefit of constituents or support legislation furthering the interests of some of their constituents . . . is an unrealistic assessment of what Congress could have meant by making it a crime to obtain property from another . . . “under color of official right.”

McCormick, 500 U.S. at 272 (quoting 18 U.S.C. § 1951(b)(2)). This is also why “matters of intent are for the jury to consider.” Inman, 39 F.4th at 365 (quoting McCormick, 500 U.S. at 270). And we have held that the factfinder can parse words and actions to discern the intent behind them, even with respect to campaign contributions. Terry, 707 F.3d at 613. The Supreme No. 23-3840 United States v. Sittenfeld Page 17

Court has held the same. See, e.g., McCormick, 500 U.S. at 270 (“It goes without saying that matters of intent are for the jury to consider.” (quoting Cheek v. United States, 498 U.S. 192, 203 (1991))); see also Evans, 504 U.S. at 274 (Kennedy, J., concurring in part and concurring in the judgment) (“[T]he trier of fact is quite capable of deciding the intent with which words were spoken or actions taken as well as the reasonable construction given to them by the official and the payor.”). We may not deviate now. Contrary to Sittenfeld’s arguments, after McCormick, the general rule remains unchanged: the government’s evidence need not rule out all reasonable, alternative hypotheses to guilt. Sadler, 24 F.4th at 539.

And explicit evidence is also not a requirement. “Explicit” relates to the quid pro quo, not the evidence. Granted, “explicit” means “[n]ot obscure or ambiguous, having no disguised meaning or reservation. Clear in understanding.” Blandford, 33 F.3d at 696 n.13 (quoting Black’s Law Dictionary 579 (6th ed. 1990) (alteration in original)). This means that the government must prove a meeting of the minds between the parties and that the agreement must be unambiguous from their perspective. But the existence of that agreement is governed by the reasonable doubt standard and can be proved with circumstantial evidence.7

Sittenfeld’s attempts to recast McCormick and Evans also have no foundation in the law. One of Sittenfeld’s own citations confirms the point. He relies on United States v. Benjamin—a district court opinion—as a “prime example” where the government lacked sufficient evidence because the connection between the quid and the quo “must be shown by something more than mere implication.” No. 21-CR-706, 2022 WL 17417038, at *12 (S.D.N.Y. Dec. 5, 2022). But the Second Circuit reversed. Benjamin, 95 F.4th at 64. The panel held that a quid pro quo can be implied, relying on our decision in Terry and other circuits holding the same. Id. at 68–69 (citing Terry, 707 F.3d at 613). So we apply McCormick and Evans as they stand:

7 Sittenfeld and the amici invoke the First Amendment and Due Process Clause. But Sittenfeld frames constitutional concerns as justifying the McCormick standard, not as making any statute unconstitutional. And the McCormick standard, as interpreted by the Sixth Circuit and others, permits the jury to infer a quid pro quo from circumstantial and less-than-conclusive evidence. The district judge acknowledged this important boundary between bribery and protected First Amendment activity. He expressly charged the jury to distinguish payments which would qualify as bribery (those made as “part of an explicit promise or understanding by the public official”) and those protected by the First Amendment (“contributions to public officials, or to PACs with which a public official is associated”). R. 251, Trial Tr. Day 11, pp. 43–44, PageID 4983–84. No. 23-3840 United States v. Sittenfeld Page 18

unambiguous evidence is not required, circumstantial evidence can prove an agreement, and though an explicit agreement must be present, it need not be express.8

C.

We hold that the evidence construed in the government’s favor was enough to convict Sittenfeld of federal-program bribery and attempted Hobbs Act extortion. Two conversations, in context, support the jury’s verdict on both counts.

First, Sittenfeld arguably solicited a bribe on October 30, 2018, when he said, “[Y]ou don’t want me to like be like ‘hey Chin like love you but can’t.’” R. 312, Appeal Exs., pp. 8–9, PageID 7551–52. Sittenfeld may have had an innocent explanation for that conversation—but the jury could reject it if it found Ndukwe’s explanation more persuasive. “[L]ove you but can’t” could have reasonably implied, as Ndukwe testified, that Sittenfeld would advocate for Ndukwe’s development projects only if he rounded up campaign donations. Seen in that light, Sittenfeld offered a corrupt bargain, and Ndukwe accepted. Ndukwe’s response to the solicitation was “trust me . . . we’ll make something happen sooner than later too.” Id. at 9, PageID 7552. And at the end of the call Sittenfeld said, “And then you’re gonna deliver the goods before next Tuesday,” and Ndukwe agreed: “All right, let me go to work on that.” Id. at 10, PageID 7553. 8 As we see it, one way to understand whether the explicit quid pro quo requirement has been satisfied is to ask what the proposed arrangement is, describing the terms that both parties would have understood as binding, in plain English. And we can. Based on the circumstantial evidence, there are two possible agreements in this case. The first option, arising out of the October 30, 2018 conversation, is straightforward: “If you give me campaign donations, I will support your projects” or “If you don’t give me campaign donations, I won’t support your projects.” The second, based on Sittenfeld’s interactions with Rob, is just as clear: “If you give me $20,000 in campaign donations, I will deliver a veto-proof majority for your project.” There’s nothing ambiguous about these proposed agreements. That a jury can infer the existence of these agreements even if the evidence falls short of a smoking gun or an express statement is not unusual in our criminal justice system. Juries retain “broad discretion” to draw inferences from the evidence at trial. United States v. Cox, 871 F.3d 479, 490 (6th Cir. 2017) (quoting Coleman v. Johnson, 566 U.S. 650, 655 (2012) (per curiam)). The dissent’s proposed standard—although perhaps wise, or even one we might adopt if we were writing on a blank slate—is simply the heightened standard that we have already rejected. We must apply the law as it exists. For the same reason, whether we ought to require more of the government given the First Amendment interests and the realities of our political system is a question for the Supreme Court. 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. But for now “[a]s an intermediary appellate court, we must follow Supreme Court decisions until directed otherwise.” Witham v. United States, 97 F.4th 1027, 1035 (6th Cir. 2024). No. 23-3840 United States v. Sittenfeld Page 19

A reasonable juror perhaps could have seen this conversation as sufficient to convict on its own—at least with respect to count 3, which required only proof of solicitation. Even if they didn’t reach a bargain, Sittenfeld plainly asked for money and implied that negative consequences could result if Ndukwe didn’t come through for him. Viewing this call in the light most favorable to the government, Sittenfeld was threatening to use his office against Ndukwe.

But we need not decide whether the call alone sufficed because this evidence doesn’t stand alone. The government had even stronger evidence to support the convictions for both counts 3 and 4. It argues that Sittenfeld agreed to accept donations on November 7, 2018, when he knew that Rob was offering him a bribe. Rob said that Ndukwe wanted “to get [Sittenfeld $]20,000” and Rob had agreed—“if we can get this deal done, like fuckin’ let’s do it.” Id. at p. 44, PageID 7587. Rob then asked explicitly, “[W]hat’s the best way for us to get that to you, to get that deal? You know what I mean.” Id. Sittenfeld didn’t answer the question—saying only “[y]eah yeah yeah.” Id. Instead, Sittenfeld pivoted to ask whether Mayor Cranley would veto the 435 Elm project. He claims that he “missed the corrupt pitch.” Appellant Br. at 31.9 But the jury could have concluded otherwise. Video damningly shows Sittenfeld nodding along as he listened to Rob, and he responded to Rob’s question “[y]eah yeah yeah.” R. 312, Appeal Exs., p. 44, PageID 7587. Nor did Sittenfeld dodge Rob’s question about the “best way” to “get that deal.” A little later in the conversation, Sittenfeld assured Rob, “I can sit here and say I can deliver the votes.” Id. at 45, PageID 7588. And he reiterated: “I can get it done.” Id. Rob responded, “Okay” and explained that he had “brought 10,000 cash with [him].” Id. And then the two talked at length about how Sittenfeld could accept the donation.

When we look at the surrounding circumstantial evidence, three more points lead us to conclude that a jury could have reasonably believed that Sittenfeld understood Rob’s intentions for a quid pro quo and still agreed to accept the money. First, construed in the government’s favor, Sittenfeld’s earlier conversations with Ndukwe laid important groundwork and offered significant context for Sittenfeld’s later interactions with Rob. Second, Sittenfeld obsessed over

9 Sittenfeld argues that Rob’s trial testimony confirms this explanation. Rob simply acknowledged that he did not use the precise words “quid pro quo” and that Sittenfeld asked about the mayor vetoing the development deal. See R. 265, Trial Tr. Day 5, p. 230, PageID 6157. Rob’s testimony does little to clarify Sittenfeld’s state of mind. No. 23-3840 United States v. Sittenfeld Page 20

the form of Rob’s donation yet demonstrated none of the same concern for its substance. And finally, Rob and Sittenfeld worked hard to keep the donations secret.

Start first with the groundwork Ndukwe laid with Sittenfeld in their phone calls. On October 30, Sittenfeld arguably agreed to accept a bribe. Then on November 2, Ndukwe presented Rob as someone who wanted to trade money for a “yes vote” on 435 Elm. Sittenfeld understood that Ndukwe’s words, at least at face value, asked for a quid pro quo. So Sittenfeld had reason to believe that at the November 7 meeting, Rob might bribe him—a bribe Sittenfeld later worked hard to, and did, accept.

This context also helps explain why Sittenfeld could promise support for 435 Elm before making his fundraising pitch. Based on the October 30 call, Sittenfeld knew that Ndukwe would pay him a bribe. And based on the November 2 call, he knew that Rob would provide the funds. A reasonable jury could conclude that “reciprocity was understood” when he promised to shepherd votes for 435 Elm. See United States v. Inzunza, 638 F.3d 1006, 1015–16 (9th Cir. 2011).

And during the November 7 meeting, Sittenfeld and Rob showed apparent disregard for the true source of the donations. Sittenfeld asked, “[W]hose name can stuff be in?” and Rob replied, “[W]e can come up with some names.” R. 312, Appeal Exs., p. 46, PageID 7589. Later, Sittenfeld talked about the “technicality” of attributing a donation to an individual and having that person “agree that that’s where it came from.” Id. at 47, PageID 7590. At the same time, Sittenfeld wanted to avoid “signals” of impropriety. See id. at 48, PageID 7591. And though he insisted that donations take the correct form, he did not care whether Rob used straw donors.

And finally, Rob and Sittenfeld wanted to keep the donation secret. Rob made it clear that in general Ndukwe did not “want his name on anything.” Id. at 44, PageID 7587. Rob wanted to avoid mailing money orders directly to Sittenfeld. And Sittenfeld later assured Rob and Brian that “[n]o one w[ould] ever know” about their PAC donation. Id. at 60, PageID 7603. Legal efforts to hide the source of a campaign donation represent fairly weak evidence alone, but they offer some probative value as to the participant’s intent. Criminals “rarely seek to No. 23-3840 United States v. Sittenfeld Page 21

perpetrate felonies before larger-than-necessary audiences.” United States v. Correia, 55 F.4th 12, 32 (1st Cir. 2022) (quoting United States v. Patch, 9 F.4th 43, 47 (1st Cir. 2021)).

The three cases Sittenfeld cites also do not persuade us. We can distinguish two on their facts. In United States v. Menendez, the best evidence for a quid pro quo was mere closeness in time between a contribution and an official action. 291 F. Supp. 3d 606, 629–32 (D. N.J. 2018). And in Inzunza, there was no evidence about the contents of a meeting between a public official and political donor, so it was unclear what the public official knew about the donor’s expectations of the official. 638 F.3d at 1025–26. Here, by contrast, a reasonable jury could interpret the video evidence and recordings as showing that Sittenfeld asked Ndukwe for a bribe and he heard both Ndukwe and Rob use language that strongly suggested bribery. And as we’ve explained, the district court’s decision in Benjamin is not an accurate description of prevailing law. 95 F.4th at 74–75.

Based on all the evidence, a reasonable juror could conclude that Sittenfeld understood exactly what Rob was offering and agreed to accept a bribe on November 7. Even if some juries might disagree, that does not change the fact that Sittenfeld’s conviction was not unreasonable. So Sittenfeld’s sufficiency-of-the-evidence claim fails.

III.

Sittenfeld also alleges his indictment was constructively amended. He argues that given the trial evidence, jury instructions, and closing arguments, the jury could have convicted him on both counts 3 and 4 based solely on the October 30, 2018, call with Ndukwe. But because both counts specify the bribe came from Rob, not “another person” (like Ndukwe) as the jury instructions allowed, Sittenfeld argues reliance on this call was impermissible. The government counters that the October 30 call was featured in the indictment within the overall charged bribery scheme and therefore poses no constructive-amendment problem. Since Sittenfeld failed to timely object, the government also maintains that we should review for plain error. Although we agree that the jury instructions, standing alone, were problematic, Sittenfeld must show more to establish a constructive amendment, which he hasn’t done. No. 23-3840 United States v. Sittenfeld Page 22

A.

The Fifth Amendment ensures that “[n]o person shall be held to answer for a . . . crime, unless on a presentment or indictment of a Grand Jury.” U.S. Const. amend V. So a defendant has a right to be “heard on the specific charges of which he is accused” in the indictment. Dunn v. United States, 442 U.S. 100, 106 (1979). Inherent in this constitutional protection is a limitation that only the grand jury that first issued an indictment can later amend it. Stirone v. United States, 361 U.S. 212, 215–16 (1960) (“[A]fter an indictment has been returned its charges may not be broadened through amendment except by the grand jury itself.”). Any change to the indictment outside of this process implicates other, related constitutional concerns—the defendant’s Fifth Amendment protection from double jeopardy and Sixth Amendment right to notice of the charges against him. 1 Wharton’s Crim. Proc., Constructive Amend. § 5:17, Westlaw (14th ed., database updated June 2023).

Whether an indictment has been amended is a flexible inquiry dependent upon how an alteration arises: actual amendment, constructive amendment, or variance. Actual amendments arise when the prosecutor “actually changes the text of the indictment.” United States v. Budd, 496 F.3d 517, 521 (6th Cir. 2007). By contrast, if the indictment remains literally unchanged but “the evidence at trial proves facts materially different from those alleged in the indictment,” a variance occurs. Id. (quoting United States v. Prince, 214 F.3d 740, 756 (6th Cir. 2000)). A variance, however, is reversible only if the defendant can show his substantial rights were prejudiced by it. Id.

More complicated are constructive amendments. Here the indictment remains unchanged, but the terms of the indictment are “in effect altered” because events at trial raise a “substantial likelihood that the defendant may have been convicted of an offense other than the one charged in the indictment.” Id. (quoting United States v. Smith, 320 F.3d 647, 656 (6th Cir. 2003)). The typical case arises because of the combined effect of the “presentation of evidence and jury instructions.” Id.; see also Stirone, 361 U.S. 212; United States v. Cusmano, 659 F.2d 714, 717–19 (6th Cir. 1981). No. 23-3840 United States v. Sittenfeld Page 23

But we have also recognized the possibility of a constructive amendment “where jury instructions differ from an indictment, even in the absence of varied evidence,” such that the effect was to charge the jury on a “separate offense that was not listed in [the] indictment.” United States v. Kuehne, 547 F.3d 667, 685 (6th Cir. 2008) (citing United States v. Combs, 369 F.3d 925 (6th Cir. 2004)). When the jury instructions alone differ from the indictment to charge a different means for committing the same crime, “a mere variance occurs and a defendant must demonstrate prejudice.” Id.10

In sum, the defendant carries a heavy burden to prove a constructive amendment. First, we look at whether the jury instructions broaden the indictment, and if so, we must ask whether the evidence also varied from the indictment. If it did, we have a constructive amendment. But if it didn’t, we must ask separately whether those jury instructions were so broad as to have allowed the jury to convict on a different crime—rather than a different means to commit the same crime. That answer, in turn, determines whether the defendant must prove prejudice. With

10 Our cases regularly describe the overarching question for whether a constructive amendment happened as whether the jury could have convicted the defendant of a separate “offense.” United States v. Smith, 320 F.3d 647, 656 (6th Cir. 2003); United States v. Budd, 496 F.3d 517, 521 (6th Cir. 2007); United States v. Prince, 214 F.3d 740, 757 (6th Cir. 2000); United States v. Kuehne, 547 F.3d 667, 683 (6th Cir. 2008). But it isn’t entirely clear what separate “offense” refers to. It could mean a separate “crime” or it could mean the same “crime” committed with different means. This imprecision, however, manifests in only some cases. When we have broad jury instructions but no varied evidence, the question is whether the defendant was injured because the jury was charged with another crime. As we noted, if the problem with the jury instructions is that they allowed guilt based on different means, the defendant must show prejudice. But when there is both varied evidence and overly broad jury instructions, that same line doesn’t hold. Both Stirone and Cusmano involved varied evidence and overly broad jury instructions. Both cases contemplated whether the defendant’s indictment had been constructively amended. And both cases answer that question “yes.” But in neither case was the defendant found guilty of a new “crime” based on the alterations at trial. Instead, both men were found guilty based on a change to the “means” alleged versus the “means” proved at trial. Take Stirone. The indictment charged interference with commerce through the importation of sand. But the proof at trial and the instructions charged the jury with the interference through the importation of steel. Both seem to involve the “means” of satisfying the same essential element: interference with interstate commerce. Still, the Court explained that the “crucial question” was whether the defendant had been “convicted of an offense not charged in the indictment.” Stirone, 361 U.S. at 213 (emphasis added). And found that it had. See also United States v. Cusmano, 659 F.2d 714, 719 (6th Cir. 1981) (alleging the essential element of “extortion” through “threats of economic loss” and proving the alternative means at trial through “wrongful use of force and fear”). Thus, the precise difference between “means” and “offense” is not as crucial where there is both varied evidence and overly broad jury instructions. Although this generally helps defendants arguing for constructive amendments, none of this helps Sittenfeld because, as we explain, even under the broadest readings of Stirone and Cusmano, he doesn’t prevail. No. 23-3840 United States v. Sittenfeld Page 24

the law clarified, we pivot briefly to determine the appropriate standard of review before addressing the merits of Sittenfeld’s case.

B.

We typically review a challenge based on a constructive amendment de novo. United States v. Hynes, 467 F.3d 951, 961 (6th Cir. 2006). But when the challenging party fails to timely object before the district court, we review for plain error. Budd, 496 F.3d at 528. The government contends that plain error applies here.

A party preserves a claim of error by objecting “when the court ruling or order is made or sought.” Fed. R. Crim. P. 51(b); see also United States v. Margarita Garcia, 906 F.3d 1255, 1268–69 (11th Cir. 2018). And the objection must include an indication of the “true basis for his objection.” United States v. Bostic, 371 F.3d 865, 871 (6th Cir. 2004) (quoting United States v. LeBlanc, 612 F.2d 1012, 1014 (6th Cir. 1980)). “A specific objection provides the district court with an opportunity to address the error in the first instance and allows this court to engage in more meaningful review.” Id.

Although there may be different ways for a defendant to raise a timely, constructive- amendment objection, Sittenfeld’s post-trial motion was not enough. Thus, plain-error review applies. As we’ve discussed, a constructive amendment typically occurs based on a confluence of events. But whether it’s the combined effect of varied evidence and broad jury instructions or the jury instructions themselves that allow the jury to convict on an unindicted crime, the constructive amendment is apparent once the parties agree on the appropriate instructions. So, at a minimum, the defendant must raise a constructive-amendment claim at the time the jury instructions are agreed upon and on grounds that they would permit the jury to convict based on a constructively amended indictment.11

11 It’s possible that earlier objections may be appropriate. If the constructive amendment is based on varied evidence, objecting at the time that such evidence is admitted on the grounds that the evidence is irrelevant and might result in a constructive amendment is another time to raise the issue. See, e.g., Cusmano, 659 F.2d at 715–16 (objecting to evidence on relevance grounds, which the trial court construed and addressed as presenting a constructive amendment issue); see also 3 Charles Alan Wright & Arthur R. Miller, Fed. Prac. & Proc. Crim. § 516, Westlaw (5th ed., database updated June 2014) (collecting cases). No. 23-3840 United States v. Sittenfeld Page 25

Thus, the jury-instruction conference is a point of no return for the parties. By then, the parties and the court know exactly what evidence has been admitted. So the defendant can spell out whether the proposed instructions themselves or their combined effect with the admitted evidence will permit the jury to convict for something not in the indictment. See, e.g., Cusmano, 659 F.2d at 717, 717 n.5. For this reason, courts have repeatedly applied plain error when a defendant fails to object to the jury instructions and argues on appeal that his indictment was constructively amended as a result. See e.g., United States v. Leon, 841 F.3d 1187, 1190, 1192 (11th Cir. 2016) (“At no time during trial did [the defendant] challenge the government’s theory of prosecution or object to the jury instructions given by the district court . . . . Because [she] did not raise her constructive amendment argument in the district court, our review is for plain error.”); United States v. Brandao, 539 F.3d 44, 57 (1st Cir. 2008) (“The district court distributed its draft jury instructions to counsel more than a week before the jury was charged and held two conferences on the instructions in the interim, yet [the defendant] did not object . . . . As an unpreserved objection, [the] constructive amendment claim is subject to plain error review.”); United States v. Remsza, 77 F.3d 1039, 1043 (7th Cir. 1996) (“[B]ecause [the defendant] objected to neither the evidence nor the jury instruction in question, he waived the objection on appeal. Therefore, we review for plain error.” (footnote omitted)).

Here Sittenfeld could have raised his constructive-amendment claim at the time that the instructions were being finalized but didn’t—a point he doesn’t dispute. Instead, Sittenfeld relies on his objection to the government’s proposed instruction for count 3 in a pretrial filing (and that the same objection was understood to apply to count 4). But that objection argued vagueness, not a possible constructive amendment. These are two distinct grounds. If Sittenfeld intended to

A defendant could also bring the problem to the trial court’s attention in a motion for acquittal before the case is submitted to the jury. 3 Wright & Miller § 516 (collecting cases). But the “grounds for th[e] objection,” are especially important here. Fed. R. Crim. P. 51(b) (emphasis added). That the insufficiency rests on an impermissible amendment to the indictment is a substantively different claim from a general sufficiency-of-the- evidence claim. It’s for this reason that, although Sittenfeld made a Rule 29 motion for judgment of acquittal, renewed the motion once the defense rested, and renewed the motion again after the jury returned a guilty verdict, he still did not preserve the constructive-amendment claim. At no point did he suggest that the indictment had been constructively amended. Instead, he repeated the more general theory that the government failed to meet its burden of proof on all six counts. No. 23-3840 United States v. Sittenfeld Page 26

preserve both, he had an obligation to specify as much before the trial court. Bostic, 371 F.3d at 871; R. 98, Prop. Jury Instrs., p. 20, PageID 1043.

The defense also did not raise any objection at the charging conference, which was intended to “allow the parties to make any objections” to the instructions “for the record.” R. 272, Trial Tr. Day 9, pp. 2, 7–8, PageID 6931, 6936–37; United States v. Semrau, 693 F.3d 510, 527–28 (6th Cir. 2012) (reviewing instructions for plain error when defendant failed to object after “the court made clear that objections would be ‘sought’ and its ruling would be ‘made,’ or after the jury was charged”); United States v. Blood, 435 F.3d 612, 625–26 (6th Cir. 2006) (reviewing objection to jury instruction for plain error when defendant failed to renew it despite district court’s warning).

That Sittenfeld raised the constructive-amendment question in a post-trial motion does not change our plain-error conclusion. A post-trial motion is too little, too late. Margarita Garcia, 906 F.3d at 1268–69 (applying plain error on appeal when defendant raised constructive- amendment claim for the first time when she moved for a new trial); Brandao, 539 F.3d at 57 (applying plain error on appeal when defendant “first raised the constructive amendment issue in a post-trial motion, which the district court denied”).

Applying plain-error review in these situations encourages contemporaneous objection, which ensures that the trial court has an opportunity to “remediate or cure the errors.” Margarita Garcia, 906 F.3d at 1269; Puckett v. United States, 556 U.S. 129, 134 (2009) (“[Plain error] limit[s] appellate-court authority [and] serves to induce the timely raising of claims and objections, which gives the district court the opportunity to consider and resolve them. . . . In the case of an actual or invited procedural error, the district court can often correct or avoid the mistake so that it cannot possibly affect the ultimate outcome.”); United States v. Jackson, 877 F.3d 231, 236 (6th Cir. 2017) (“Where a defendant has failed to preserve [an] objection by first giving the district court the opportunity to address and remedy it, we review only for plain error.” (internal quotation marks omitted)). After trial, the district judge was no longer able to rectify any problem, so the post-trial motion did not preserve the objection. No. 23-3840 United States v. Sittenfeld Page 27

Sittenfeld’s other arguments for de novo review also do not persuade us. It’s true we have said that “there can be no forfeiture where the district court nevertheless addressed the merits of the issue.” Cartwright v. United States, 12 F.4th 572, 580 n.6 (6th Cir. 2021) (quoting United States v. Clariot, 655 F.3d 550, 556 (6th Cir. 2011)). But that rule applies when the district judge raises an issue in a ruling that was not previously raised by the parties—not when the court addresses a forfeited argument after a criminal trial.12

Finally, on count 4 only, the government claims Sittenfeld invited the jury-instruction error that is the crux of his constructive-amendment claim. It’s true that we may forgo appellate review if the defendant invites an error based on his proposed instructions. United States v. Howard, 947 F.3d 936, 944–45 (6th Cir. 2020). And according to the government, because Sittenfeld contributed to the alleged constructive-amendment problem by suggesting “another person” instead of Rob’s name in the count 4 instructions, appellate review should be entirely foreclosed. We decline this invitation. Even when a defendant invites error, “the interests of justice will typically favor reviewing an argument where the government and the defendant are equally at fault and the defendant claims a violation of his constitutional rights.” United States v. Montgomery, 998 F.3d 693, 699 (6th Cir. 2021). Because the government also proposed using “another person,” and the claimed error implicates Sittenfeld’s Fifth and Sixth Amendment rights, the interests of justice favor reviewing the claimed error. Yet, as with count 3, our review is deferential because of Sittenfeld’s failure to object. Howard, 947 F.3d at 945.

The defendant bears the burden of proving whether a constructive amendment or variance has occurred. Kuehne, 547 F.3d at 683. And to succeed on plain error, the defendant must show (1) an error, (2) that is plain, (3) that affects substantial rights, and (4) that “seriously affects the fairness, integrity, or public reputation of judicial proceedings.” Id. (internal quotation marks

12 See, e.g., Heyward v. Cooper, 88 F.4th 648, 655 (6th Cir. 2023) (motion to dismiss); Hudak v. Elmcroft of Sagamore Hills, 58 F.4th 845, 857–58 nn.4–5 (6th Cir. 2023) (motion to remand); Bledsoe v. Tenn. Valley Auth. Bd. of Dirs., 42 F.4th 568, 588 n.4 (6th Cir. 2022) (summary judgment); Cartwright v. United States, 12 F.4th 572, 580 n.6 (6th Cir. 2021) (habeas petition); Owens v. Parris, 932 F.3d 456, 458 (6th Cir. 2019) (habeas petition); Raines v. United States, 898 F.3d 680, 687 (6th Cir. 2018) (habeas petition); United States v. Clariot, 655 F.3d 550, 556 (6th Cir. 2011) (suppression motion). But see United States v. Hofstetter, 31 F.4th 396, 412 (6th Cir. 2022) (forfeited issue considered by district court in denying motion for acquittal), cert. granted, judgment vacated on other grounds, 143 S. Ct. 351 (2022); Bavelis v. Doukas, 835 F. App’x 798, 810 n.6 (6th Cir. 2020) (issue raised by district court in reviewing bankruptcy court decision). No. 23-3840 United States v. Sittenfeld Page 28

omitted). For the error to be “plain,” it must be “clear or obvious, rather than subject to reasonable dispute.” Puckett, 556 U.S. at 135. With this standard in mind, we turn to the merits of Sittenfeld’s claim.

C.

To avoid a constructive amendment, Sittenfeld must have been convicted for the specific crimes alleged in the indictment. The government did not rely on varied evidence in the way Cusmano or Stirone discuss, and though the jury instructions were broader than the indictment, this alone is not reversible error. Sittenfeld cannot plainly show that the overly broad instructions resulted in a conviction for another, unindicted crime rather than another means of committing the same crime. So Sittenfeld’s challenge fails.

1.

We start with the indictment, which begins by describing Sittenfeld’s “scheme” through quotes from transcripts of phone calls like the one on October 30, 2018. It also alludes to a series of solicitations beyond the ones listed in the to-wit clauses of each count. That said, the to-wit clauses make clear that Sittenfeld’s innocence or guilt was tied to specific solicitations. Sittenfeld was charged in count 3:

to wit, [Sittenfeld] while a member of Cincinnati City Council, corruptly solicited and demanded, and accepted and agreed to accept, payments to PAC for his benefit from [Rob], who was posing as a businessman supporting [435 Elm], while intending to be influenced and rewarded in connection with business, transactions, and series of transactions involving the City of Cincinnati and [435 Elm].

R. 3, Indictment, pp. 16–17, PageID 41–42 (emphasis added). And in count 4:

to wit, [Sittenfeld] solicited, obtained, agreed to accept, accepted, and received payments to PAC from [Rob], with [Rob’s] consent, knowing the payments were made in exchanged for the defendant’s specific official action in his role as a member of the Cincinnati City Council to further [435 Elm], under color of official right.

Id. at 17, PageID 42 (emphasis added). Emphasizing these clauses, Sittenfeld contends that he can only be validly convicted based on the payments Rob made to his PAC to secure his favor in No. 23-3840 United States v. Sittenfeld Page 29

the 435 Elm project. A conviction based on any phone call or solicitation with Ndukwe is legally insufficient. And he says this is what happened here.

In Cusmano, we interpreted a to-wit clause as alleging “only one means of extortion” and creating a constructive-amendment problem when a different means of extortion was raised at trial. 659 F.2d at 715, 719.13 There, the indictment alleged extortion through economic harm. Id. at 715, 717 nn.5–6. But the evidence at trial varied from that to-wit clause because it suggested extortion through physical harm. We reversed because there was a substantial likelihood that the conviction rested on an unindicted crime—extortion based on physical harm. Kuehne, 547 F.3d at 685 (citing Cusmano, 659 F.2d at 718). And other circuits have similarly reversed convictions when admitted evidence differed from the to-wit clause.14

Under Cusmano, the to-wit clauses in Sittenfeld’s indictment restricted and confined the charges against him. The government fights the limiting effect of the to-wit clause, theorizing that it can rely on any evidence that appeared earlier in the indictment if it was incorporated into the charged count. But an indictment does not charge everything it describes. An indictment may include significant background information and circumstantial evidence, yet a defendant may justifiably rely on a to-wit clause to particularize the crime. So we reject the government’s contention that we can, in effect, view the entire indictment as one big to-wit clause.15

13 We have also said that a “general expression in an indictment may be restricted and confined to a precise and definite fact by a description under a videlicet,” preventing an indictment from being duplicitous. Beauchamp v. United States, 154 F.2d 413, 415 (6th Cir. 1946). The term “videlicet” is abbreviated “viz.,” Black’s Law Dictionary (11th ed. 2019), and it is used to describe phrases like “to wit,” “that is to say,” and “namely.” See State v. Sudrala, 116 N.W.2d 243, 244 (S.D. 1962); see also People v. Hartfield, 208 N.E.3d 1223, 1229 (Ill. App. Ct. 2022) (“Where an alleged fact is preceded by ‘to-wit,’ it is said to be laid under a videlicet.” (quoting People v. Wilson, 182 N.E.2d 683, 684 (Ill. 1962))); Commonwealth v. Hart, 76 Mass. (10 Gray) 465, 468 (1858) (recognizing that “to wit” acted as a videlicet that served “to isolate, to distinguish, and to fix with certainty, that which was before general”). 14 See, e.g., United States v. Davis, 854 F.3d 601, 604–06 (9th Cir. 2017); United States v. Chambers, 408 F.3d 237, 240–41, 247 (5th Cir. 2005); United States v. Willoughby, 27 F.3d 263, 266–67 (7th Cir. 1994); United States v. Weissman, 899 F.2d 1111, 1115–16 (11th Cir. 1990). The Second Circuit illustrates a notable exception of this trend, viewing to-wit clauses instead as merely “illustrative.” See, e.g., United States v. Agrawal, 726 F.3d 235, 261 (2d Cir. 2013); United States v. Khan, 726 F. App’x 73, 75 (2d Cir. 2018) (applying Agrawal). We find the reasoning of the other circuits more persuasive and reject Agrawal as inconsistent with our caselaw. 15 Because a to-wit clause limits the scope of a defendant’s charge, we also reject the view that Sittenfeld’s conviction could stand based on a “conduit theory.” Any theory of liability that relies on the transitive property to tie Sittenfeld to Rob improperly broadens the scope of the defendant’s liability. Treating a bribe from Ndukwe as if No. 23-3840 United States v. Sittenfeld Page 30

This means that despite the indictment’s description of the conversation with Ndukwe, the to-wit clause only charged Sittenfeld with accepting a bribe from Rob—not Ndukwe—and he could reasonably expect that counts 3 and 4 were so limited. Thus, we agree with Sittenfeld that the to-wit clauses in his indictment require his conviction to rest on payments from Rob in exchange for favor on 435 Elm.

2.

Next, we consider the jury instructions. We ask whether they could be read to permit a conviction on grounds broader than what was alleged in the indictment. As we previewed above, the answer is yes.

The court instructed the jury that it could find Sittenfeld guilty on count 3 if it found “that Mr. Sittenfeld solicited, demanded, accepted, or agreed to accept a thing of value from another person.” R. 251, Trial Tr. Day 11, p. 34, PageID 4974 (emphasis added). And the instruction for count 4 required the jury to similarly find “that Mr. Sittenfeld obtained, accepted, agreed to accept, or received property that he was not lawfully entitled to from another person with that person’s consent.” Id. at 38, 51, PageID 4978, 4991 (emphasis added). By referring to “another person” and not specifying Rob—in a case that involved several personalities and several solicitations—the jury instructions failed to clarify that the jury could not convict based on any perceived agreement between Sittenfeld and Ndukwe.

3.

But as we noted above, whether the instructions permit the jury to convict on grounds beyond the indictment does not end the analysis. We still must compare the government’s evidence to the scope of the indictment to see if there was a variance. Though we rejected the government’s theory that the general scheme alleged in the indictment can broaden the scope of a to-wit clause, that question is distinct from whether the trial evidence varied from the indictment.

it had come from Rob fundamentally misreads these to-wit clauses. If the government chooses to charge a defendant using a to-wit clause, the government must be prepared to demonstrate the defendant’s guilt accordingly. No. 23-3840 United States v. Sittenfeld Page 31

On this latter point, we read indictments “as a whole,” United States v. McAuliffe, 490 F.3d 526, 531 (6th Cir. 2007), and can consider the “full scope of [the] indictment” to analyze whether the trial evidence varied from the indictment, United States v. Bradley, 917 F.3d 493, 503 (6th Cir. 2019). This includes the indictment’s description of the scheme which details Sittenfeld’s (and others’) “specific involvement” in the events related to each count. Id. The “manner and means,” the “overt acts,” and the “scheme” are the government’s “specification of the ways in which the defendant[] sought to accomplish” his crime. United States v. Mubayyid, 658 F.3d 35, 53–54 (1st Cir. 2011). The facts alleged in this portion inform the defendant of the “specific offence . . . with which he is charged.” Hamling v. United States, 418 U.S. 87, 117–18 (1974).

For example, in Stirone, the government did not rely on a to-wit clause to cabin the scope of the indictment, but it described specific acts that violated the Hobbs Act. The indictment alleged that the defendant “caused supplies and materials (sand) to move in interstate commerce,” and did “unlawfully obstruct, delay (and) affect interstate commerce between the several states . . . by extortion.” Stirone, 361 U.S. at 213–14. Then, at trial, the district judge allowed the government to admit evidence and to include in the jury instructions reference to the defendant’s interference with steel shipments. Id. at 218–19. The extortion was an essential element of the crime and proving it with an entirely different set of facts (steel vs. sand) effectively allowed the defendant to be convicted of an unindicted crime. Cusmano, 659 F.2d at 719 (alleging the essential element of “extortion” through “threats of economic loss” and proving the alternative means at trial through “wrongful use of force and fear”).

Sittenfeld theorizes that his case is like Stirone and Cusmano. On his view, these cases demonstrate that the government improperly admits varied evidence whenever its evidence is not expressly listed in the to-wit clause. But this represents too limited a view of indictments and Sittenfeld’s case overall. It’s true that once the government describes the “factual basis for an element of the crime, the prosecution may not rest its proof of that element at trial on other facts.” Budd, 496 F.3d at 522 (emphasis added) (internal quotation marks omitted). This still leaves the government with room to make its case before the judge or jury by relying on the No. 23-3840 United States v. Sittenfeld Page 32

specific (relevant) facts it alleges in the indictment’s background about how the defendant committed his crime.

In both Cusmano and Stirone, the government proved facts beyond those alleged in the indictment. That’s not what happened here. Under the indictment’s scheme and alleged manner and means, a series of events between Sittenfeld and Ndukwe resulted in the corrupt solicitation between Sittenfeld and Rob and ended with the transfer of $20,000. It was this latter crime for which Sittenfeld was indicted in the to-wit clause, but that did not foreclose the government from admitting evidence that explained how Sittenfeld and Rob got together—consistent with the indictment. A look to the government’s closing arguments shows how the government hewed to the scheme it alleged in counts 3 and 4.

Start with count 3. The indictment alleged that Sittenfeld solicited, demanded, or agreed to accept a thing of value from Rob. The October 30 phone call may well have provided sufficient evidence for a conviction on count 3. And we acknowledge that when the government discussed count 3 in its closing,16 it explicitly said that Sittenfeld “corruptly solicited Mr. Ndukwe on October 30th.” R. 251, Trial Tr. Day 11, p. 97, PageID 5037. Clearly the October 30 call was central to the case overall—the relationship between Sittenfeld and Ndukwe set the stage for how Sittenfeld and Rob were connected, as the “scheme” alleged in the indictment suggests.

But talking about the October 30 call wasn’t the government’s final word on count 3. And the indictment reflects that. Instead, the government explained count 3 to the jury by

16 We have repeatedly acknowledged in several contexts that counsel’s closing arguments are not evidence, but they are essential to the trial through which the parties can “argue reasonable inferences from the evidence” to the jury. United States v. Crosgrove, 637 F.3d 646, 663–64 (6th Cir. 2011). It’s for this reason that courts regularly look to closing arguments in constructive-amendment claims to see how the government conceptualized the evidence and the indictment, and how the evidence was presented to the jury. See, e.g., United States v. Jaimez, 45 F.4th 1118, 1127 (9th Cir. 2022) (“While the government’s closing argument is not evidence, it is useful to consider in evaluating both the permissible inferences that can be drawn from the evidence and how the government built its money laundering conspiracy case against [the defendant].” (citation omitted)); United States v. Miller, 891 F.3d 1220, 1236 (10th Cir. 2018) (finding a constructive amendment where the trial evidence, jury instructions, and prosecution’s closing argument revealed the jury could find guilt based on an unindicted false statement); United States v. Mariano, 729 F.3d 874, 882–83 (8th Cir. 2013) (concluding that the combined effect of the jury instructions and the government’s closing arguments did not create a substantial likelihood that the defendant was convicted of uncharged conduct). No. 23-3840 United States v. Sittenfeld Page 33

relying on three separate “explicit quid pro quos,” any one of which could sustain the conviction. First, the government cited the October 30 call, then, the November 7 exchange where Rob offered Sittenfeld $20,000 for his PAC, and finally, December 17, when Sittenfeld finally received the $20,000 in exchange for the promise to deliver votes. Id.

And the same is true of count 4. The indictment alleged that Sittenfeld had knowingly obtained or agreed to accept payments from Rob. Again, the October 30 call and its “love you but can’t” were key to the government’s narrative, a point the government acknowledges. And the district court pointed to the call (without distinguishing between counts 3 and 4) as “[p]erhaps the best example” of evidence that “could reasonably be interpreted to give rise to a finding of intent to enter an explicit deal.” R. 283, Op. & Order, pp. 3–4, PageID 7140–41.

But as the government explained the evidence to the jury, it never argued that the October 30 call alone constituted a completed agreement.17 R. 251, Trial Tr. Day 11, pp. 60–72, PageID 5000–12. Far from relying only on the October 30 call as Sittenfeld suggests, the government explained count 4 to the jury by relying on two other conversations—on November 2 between Ndukwe and Sittenfeld and on November 7 between Sittenfeld and Rob. Id. at 60–72, PageID 5000–12. The government described “the second express quid pro quo in 5 days”—November 2 “money for votes,” and November 7 getting “P.G. $20,000.” Id. at 70, PageID 5010. As we’ve detailed, supra Part II.C., the November 7 interaction with Rob provided the strongest evidence for the government on both counts 3 and 4. Rob agreed to “get [Sittenfeld $]20,000” and promised “if we can get this deal done, like fuckin’ let’s do it.” Rob asked Sittenfeld about the details of getting him the $20,000 and Sittenfeld—rather than offering a direct answer—nodded along to the discussion, saying “Yeah yeah yeah.” As we’ve explained, that whole exchange led to a promise by Sittenfeld: “I can sit here and say I can deliver the votes. . . . I can get it done.”18

17 The government asks, “how could Sittenfeld agree to anything based on Ndukwe’s reply” to Sittenfeld’s “love you but can’t” solicitation. Appellee Br. at 68–69. We do not find this argument persuasive. Sittenfeld was the offeror. The jury could have believed that he solicited a bribe, and that Ndukwe agreed to provide one, thereby accepting Sittenfeld’s offer. 18 Finally, Sittenfeld argues that the jury acquitted on count 1 precisely because it was relying on the October 30 call with Ndukwe. But it’s not clear or obvious that the jury made such a decision. That the jury resulted in a split verdict tells us nothing of the jury’s reasoning. Cf. Bravo-Fernandez v. United States, 580 U.S. 5, 13 (2016) (“When a jury returns irreconcilably inconsistent verdicts, . . . it is just as likely that ‘the jury, convinced No. 23-3840 United States v. Sittenfeld Page 34

The line between the indictment’s allegations—in the scheme, the manner and means, and the substantive counts—and the government’s evidence and arguments to the jury, is clear and direct. The government did not rely on varied evidence to prove the indicted crimes.

4.

Given that we have broad jury instructions but no varied evidence, neither Cusmano nor Stirone control. That said, the question remains whether a bribery agreement with Ndukwe amounted to a “separate offense” from a bribery agreement with Rob, or whether it represented another way to commit the same crime. See Kuehne, 547 F.3d at 685. If the former, the instructions may have caused a constructive amendment, but if the latter, they would only have caused a variance.19 Id.

Determining whether a statute sets out multiple offenses or alternative means for committing the same offense is inherently one of statutory interpretation. Combs, 369 F.3d at 931 (“The statutory text, legislative history, and requisite proof argue for the . . . perspective that 18 U.S.C. § 924(c) criminalizes two separate offenses . . . .”). We look to two bodies of law to cabin our inquiry. The first are cases that analyze whether an indictment is “duplicitous.” Duplicitousness is similarly a search for distinct offenses, because, like constructive amendments, they threaten a defendant’s Sixth Amendment right to notice of the charges against

of guilt, properly reached its conclusion on [one count], and then through mistake, compromise, or lenity, arrived at an inconsistent conclusion on the [related] offense.’” (quoting United States v. Powell, 469 U.S. 57, 65 (1984))). 19 If a variance occurred, we could ask whether Sittenfeld was prejudiced. But he doesn’t make this argument. Prejudice requires proof that the defendant’s “substantial right[s]” were affected. Kuehne, 547 F.3d at 683. With a variance, the defendant’s substantial rights refer to his “ability to defend himself at trial, to the general fairness of the trial, or to the indictment’s sufficiency to bar subsequent prosecutions.” Id. (quoting United States v. Hynes, 467 F.3d 951, 962 (6th Cir. 2006)). Sittenfeld couldn’t make that showing anyway. As we explained supra Part II.C and Part III.C.3, the convictions for both counts 3 and 4 went beyond the October 30 call. And the detail of each phone call and meeting—dates, locations, participants, discussions, and excerpts—are all alleged chronologically and succinctly in the indictment, telling a larger story of the Ndukwe-Rob-Sittenfeld relationship and scheme. Sittenfeld was on notice of every detail that the government planned to (and did) prove at trial. Viewed in this light, Sittenfeld’s Fifth and Sixth Amendment rights to fair notice and a fair trial were properly protected by the grand jury and subsequent indictment. No. 23-3840 United States v. Sittenfeld Page 35

him and his Fifth Amendment right against double jeopardy.20 See United States v. Davis, 306 F.3d 398, 415–17 (6th Cir. 2002) (citing 2 Charles Alan Wright & Arthur R. Miller, Fed. Prac. and Proc. § 142 (3d ed. 1999)); see also Combs, 369 F.3d at 935; United States v. Aguilar, 756 F.2d 1418, 1420 n.2 (9th Cir. 1985).

The second set of cases grapple with “individuating crimes” in an indictment based on the defendant’s victims.21 Some contexts indicate that different victims constitute separate crimes. We count murders by “counting bodies.” United States v. Newell, 658 F.3d 1, 24 (1st Cir. 2011). And we’ve suggested a constructive amendment could result if a conviction relied on a different overdose victim, see United States v. Davis, 970 F.3d 650, 659 (6th Cir. 2020), or a different defrauded financial institution, see United States v. Nixon, 694 F.3d 623, 638 (6th Cir. 2012).22

But a different victim does not always indicate a different crime. For example, the victim’s identity was “irrelevant” to a conviction for transporting money obtained by fraud in interstate commerce. United States v. Von Stoll, 726 F.2d 584, 585, 587 (9th Cir. 1984) (finding no constructive amendment). And according to the Second Circuit, the government does not have to prove that a “particular person” was in fact defrauded by counterfeit bonds to convict a defendant of passing counterfeit bonds “with intent to defraud.” United States v. Mucciante, 21 F.3d 1228, 1235 (2d Cir. 1994) (quoting 18 U.S.C. § 479) (finding no constructive amendment).

20 Both issues also implicate a defendant’s Sixth Amendment right to a unanimous jury verdict. See United States v. Campbell, 279 F.3d 392, 398 (6th Cir. 2002); Ford, 872 F.2d at 1236–37; see also United States v. Lasley, 917 F.3d 661, 664–65 (8th Cir. 2019). 21 We recognize that Rob and Ndukwe were not victims in the traditional sense. But an extorted person may be considered a victim in at least some cases. See United States v. Capo, 817 F.2d 947, 954 (2d Cir. 1987) (en banc); Evans v. United States, 504 U.S. 255, 283 (1992) (Thomas, J., dissenting). 22 Other examples abound. We have said in dicta that “[s]eparate allegedly false statements” under 18 U.S.C. § 1001 “are entirely separate offenses.” United States v. Dedman, 527 F.3d 577, 600 n.10 (6th Cir. 2008). And we have asked whether assaults under 18 U.S.C. § 111 take place in “distinct successive criminal episodes, rather than two phases of a single assault,” to determine whether distinct crimes were committed. United States v. Shumpert Hood, 210 F.3d 660, 663 (6th Cir. 2000) (quoting United States v. Segien, 114 F.3d 1014, 1022 (10th Cir. 1997)). No. 23-3840 United States v. Sittenfeld Page 36

These cases highlight that whether a different victim constitutes a different crime is statute-specific and therefore crime-specific. As for Sittenfeld, we look to the two statutes (§ 666 and the Hobbs Act) to identify if different victims—or more specifically the corrupt counterparty—makes for different offenses or a different means to commit the same offense.

Start with § 666.23 We permit the government to aggregate transactions to reach the $5,000 jurisdictional minimum when they are “part of a single scheme.” United States v. Sanderson, 966 F.2d 184, 189 (6th Cir. 1992). And other circuits follow this logic for both § 666(a)(1)(A) and (B). United States v. Hines, 541 F.3d 833, 837 (8th Cir. 2008) (permitting aggregation to reach jurisdictional minimum); United States v. Cruzado-Laureano, 404 F.3d 470, 484 (1st Cir. 2005); United States v. Yashar, 166 F.3d 873, 876 (7th Cir. 1999); United States v. Doty, 832 F. App’x 174, 179, 180 n.4 (4th Cir. 2020).

But we have not found (and the parties do not point us to) any case invoking a similar “aggregation” principle when it comes to corrupt counterparties. The closest discussion of the relationship between these theories comes from the First Circuit. Though that court has permitted the government to aggregate transactions to reach the $5,000 minimum, Cruzado- Laureano, 404 F.3d at 484, it has not extended that reasoning to duplicitousness under § 666(a)(1)(A) when individual transactions exceeded $5,000, Newell, 658 F.3d at 26. Reasoning that the principles motivating a rule for aggregation did not dictate the outcome on duplicitousness, it concluded that an indictment that charged “distinct violations of § 666(a)(1)(A)” within the same count was duplicitous. Newell, 658 F.3d at 28. Retrieving funds from two agencies created different transactions to be charged separately. See id. at 26, 28.

If it’s one bribery scheme between the same two people, the answer is straightforward. But what if, as here, multiple parties are involved? Sittenfeld’s corrupt agreement with Ndukwe can be understood as part of the same general scheme that led to Sittenfeld accepting a bribe from Rob. Rob framed his donation as something that Ndukwe wanted. The parties also

23 Someone violates 18 U.S.C. § 666(a)(1)(B) when, acting as an agent of a local, state, or tribal government he “corruptly solicits or demands for the benefit of any person, or accepts or agrees to accept, anything of value from any person, intending to be influenced or rewarded in connection with any business, transaction, or series of transactions of such organization, government, or agency involving anything of value of $5,000 or more.” No. 23-3840 United States v. Sittenfeld Page 37

probably expected Rob’s payments to satisfy Ndukwe’s October 30 promise to funnel donations to Sittenfeld. And both corrupt agreements had 435 Elm in view.

But in the end, it’s not clear that having two distinct corrupt counterparties gives rise to distinct crimes under our constructive-amendment caselaw, especially given our lack of relevant § 666 precedent. And we need not answer this difficult question directly because even if it were an error, it would not have been plain that the identity of the corrupt counterparty could constitute a conviction for a second, unindicted crime.

We come to the same conclusion on the Hobbs Act.24 We again look to the First Circuit. That court has held that “the identity of the target is not an element of a robbery or conspiracy to commit robbery under the Hobbs Act.” United States v. Katana, 93 F.4th 521, 533 (1st Cir. 2024). That court also noted that the defendant failed to identify precedent “suggesting that robbery of an individual is a different offense than robbery of that individual’s home business.” Id. at 535.

This case is similar. Sittenfeld does not point to precedent suggesting that extortion of a business owner (like Ndukwe) must be charged as a separate offense from the extortion of an investor in the same business (like Rob). Perhaps distinct instances of extortion should be charged separately, even if related. But our precedent does not require that, and the answer is unclear. So even if the identity of the target is an element of attempted Hobbs Act extortion, we see no plain error.

Thus it was not plain that the possibility of a conviction for the bribery agreement with Ndukwe constituted a separate offense (rather than a different means of committing the same crime) from the one he was indicted for. Because that distinction was not plain, we cannot find that a constructive amendment occurred based on the breadth in the jury instructions.

24 A violation of the Hobbs Act is established by demonstrating a person “in any way or degree obstructs, delays, or affects commerce or the movement of any article or commodity in commerce, by robbery or extortion [as defined in § 1951(b)(2)] . . . to any person or property in furtherance of a plan or purpose to do anything in violation of this section.” 18 U.S.C. § 1951(a). No. 23-3840 United States v. Sittenfeld Page 38

* * *

In the end, when we review the evidence, we can’t conclude that Sittenfeld has shown— under plain-error review—that there is a substantial likelihood he was convicted of an offense not charged in the indictment. The absence of varied evidence forecloses any relief for Sittenfeld on his constructive-amendment claim. The jury instructions, though impermissibly broad, did not plainly result in a conviction for a different crime not alleged in the indictment.

IV.

We AFFIRM Sittenfeld’s conviction. No. 23-3840 United States v. Sittenfeld Page 39

__________________

CONCURRENCE __________________

MURPHY, Circuit Judge, concurring. Suppose that a local official decides to run for the U.S. Senate on a platform of repealing a well-known law—say, the Affordable Care Act from 2010 or the Tax Cuts and Jobs Act from 2017. Suppose further that the candidate’s campaign literature solicits money from supporters by promoting the promise to repeal the law: “Donate to my campaign today because I will vote to repeal the Affordable Care Act if elected. Together we can overturn this law.” If a supporter chooses to donate based on the candidate’s pledge to take that official act, have the parties exchanged a bribe that violates the Hobbs Act (18 U.S.C. § 1951) and the federal bribery law governing state and local officials (18 U.S.C. § 666)? Treating that campaign donation as an illegal “quid pro quo” (the payment of money for the pledged repeal) would raise serious concerns under the First Amendment. In our republic, political candidates regularly make promises about what they will do if elected, and citizens who agree with those promises regularly support the candidates with their votes, their time, and their money. See McCutcheon v. FEC, 572 U.S. 185, 191–92 (2014) (plurality opinion). The Free Speech Clause exists to protect this political speech and association. See id. at 203; Buckley v. Valeo, 424 U.S. 1, 14–23 (1976) (per curiam); Robert H. Bork, Neutral Principles and Some First Amendment Problems, 47 Ind. L. J. 1, 29 (1971). Thankfully, then, the government disavowed any intent to prosecute generic campaign donations tied to campaign pledges. Oral Argument 22:19–23:10.

But the government did not do the best of jobs distinguishing that hypothetical prosecution from this real one. While serving as a member of Cincinnati’s city council, Alexander “P.G.” Sittenfeld says he never saw a development project he didn’t like. So when running for mayor, he used this pro-development platform to solicit campaign donations from developers, including those who (purportedly) sought to redevelop a specific property: 435 Elm Street. As he explained when asking for funds from this project’s proponents (who turned out to be undercover agents): “in seven years I have voted in favor of every single development deal that’s ever been put in front of me[.]” Tr., R.312, PageID 7557. Federal prosecutors No. 23-3840 United States v. Sittenfeld Page 40

nevertheless claimed that Sittenfeld’s actions crossed the line from campaign fundraising to bribery soliciting. They asserted that he entered an illegal quid pro quo by requesting donations with the promise to support the 435 Elm project. Yet why is Sittenfeld’s alleged pledge to vote on this project any different from the hypothetical pledge to repeal the Affordable Care Act? And why doesn’t Sittenfeld’s conduct implicate similar First Amendment concerns? I do not have good answers to these questions because the Supreme Court has adopted a “vague” “line” to separate protected political speech from illegal bribery. McCutcheon, 572 U.S. at 209 (plurality opinion).

As I shall explain, however, I am not sure that either law in this case—when properly interpreted—would require courts to get close to this constitutional line. I nevertheless concur in Judge Nalbandian’s excellent majority opinion because existing precedent seems to interpret these two statutes in a way that maximizes (rather than minimizes) the constitutional concerns.

I

A

Start with the Hobbs Act. This 1946 law makes it illegal for a person to “affect[] commerce” by “extortion” or attempted extortion. 18 U.S.C. § 1951(a). It defines “extortion” as “the obtaining of property from another, with his consent, induced by wrongful use of actual or threatened force, violence, or fear, or under color of official right.” Id. § 1951(b)(2). A perceptive reader (and even a not-so-perceptive one) should notice that this text lacks “even a colorable allusion to campaign contributions or quid pro quos.” McCormick v. United States, 500 U.S. 257, 277 (1991) (Scalia, J., concurring). So what does the Hobbs Act mean when it bars “the obtaining of property from another . . . under color of official right”? The prepositional phrase “under color of official right” modifies the gerund phrase “obtaining of property.” This language suggests that public officials must “tak[e] . . . money by color of office” to violate the Hobbs Act. Evans v. United States, 504 U.S. 255, 284 n.4 (1992) (Thomas, J., dissenting). Police officers who have a duty to protect everyone, for example, might commit this type of extortion if they falsely claim that the law requires private parties to pay a fee for police protection. See id. at 279–82. No. 23-3840 United States v. Sittenfeld Page 41

In contrast, I tend to agree with those who have read the Hobbs Act to exclude garden- variety bribery (paying money to influence an official act). See id. at 279–84 (Thomas, J., dissenting); United States v. Cerilli, 603 F.2d 415, 434–35 (3d Cir. 1979) (Aldisert, J., dissenting); see also Ocasio v. United States, 578 U.S. 282, 300 (2016) (Breyer, J., concurring); John T. Noonan, Jr., Bribes 585–86 (1984). Public actors who take bribes do not typically claim that their office gives them the right to the money. See Evans, 504 U.S. at 283 (Thomas, J., dissenting). Rather, bribed officials accept the money unofficially in exchange for taking some other “act” that is “by color of office.” Id. at 284 n.4. But “under color of official right” in the Hobbs Act modifies “the obtaining of property”; it does not modify “the taking of an act” (a phrase nowhere to be found in the law). So the obtaining of money itself must be the claimed official act. That is why the common law distinguished extortion (in which the payor was a victim) from bribery (in which the payor was an accomplice). See id. at 284; Cerilli, 603 F.2d at 435 (Aldisert, J., dissenting).

Then how did an extortion law transmogrify into a bribery law? Like other dubious doctrines, this result arose “seemingly by accident.” FDA v. All. for Hippocratic Med., 602 U.S. 367, 403 (2024) (Thomas, J., concurring). For decades, courts refused to extend the Hobbs Act to cover voluntary (if illegal) agreements to exchange a payment for an official act. See McCormick, 500 U.S. at 279–80 (Scalia, J., concurring). It was not until the 1970s that a few (lightly reasoned) decisions included “classic bribery” within the Hobbs Act’s scope. United States v. Braasch, 505 F.2d 139, 151 (7th Cir. 1974); see United States v. Kenny, 462 F.2d 1205, 1229 (3d Cir. 1972). This expansion soon spread to the other circuit courts. But these other courts merely cited the earlier decisions without providing any “reasoned elaboration” of their own. Cerilli, 603 F.2d at 427 & n.5 (Aldisert, J., dissenting); see McCormick, 500 U.S. at 277– 78 (Scalia, J., concurring).

To make matters worse, when the Supreme Court got involved in McCormick, it granted certiorari on a subsidiary question that required it to assume that the Hobbs Act covered bribery (the “logically antecedent” issue). Warner Chappell Music, Inc. v. Nealy, 601 U.S. 366, 374 (2024) (Gorsuch, J., dissenting); see McCormick, 500 U.S. at 276–77 (Scalia, J., concurring). By then, some courts had enlarged the Hobbs Act even further to cover campaign contributions even No. 23-3840 United States v. Sittenfeld Page 42

when a payor and public official did not agree to a quid-pro-quo exchange of a donation for an official act. See McCormick, 500 U.S. at 271. While assuming that the Hobbs Act could reach bribery, the Court rejected the view that an official violates the Act merely by accepting a campaign contribution knowing that the contributor expects to benefit from the payment. Id. at 268 n.6, 271–74. The Court instead held that campaign contributions can violate the Act “only if the payments are made in return for an explicit promise or undertaking by the official to perform or not to perform an official act.” Id. at 273. The Court elsewhere suggested that the government must prove an “explicit quid pro quo” or just a “quid pro quo” to show a violation. Id. at 271, 274.

It was not until Evans that the Court held (rather than assumed) that the Hobbs Act covered bribery. Yet the parties there did not even brief the issue. See Ocasio, 578 U.S. at 300 (Breyer, J., concurring). Rather, the Court granted certiorari over whether the verb “induced” in the statutory definition of extortion modified “under color of official right” and thus required public actors to ask payors for the property. 18 U.S.C. § 1951(b)(2); Evans, 504 U.S. at 256. Evans held that the Act did not require this inducement largely because common-law extortion did not require a “demand” for money. 504 U.S. at 260. In a later section, the Court rejected the dissent’s narrower view that the Act required proof that public officials took money under the pretense that they had a right to it. See id. at 269–70. The Court instead held (in all of three paragraphs) that officials violate the Act whenever they receive a fee that they are not “entitled to” for their official duties. Id. at 270. In support, the Court noted that the defendant had not advocated for the dissent’s interpretation and that no circuit court had accepted it. Id. at 270–71. All told, then, the Hobbs Act confirms that “unexamined assumptions” in judicial opinions “have a way of becoming, by force of usage, unsound law.” McCormick, 500 U.S. at 280 (Scalia, J., concurring).

B

Sittenfeld accepts that the Hobbs Act covers bribery—as he must, given Evans. But he asks us to clarify what McCormick meant when it used the adjective “explicit” to describe the campaign donations that might violate the Act. See id. at 271, 273 (majority opinion). Sittenfeld believes that McCormick sought to give breathing space to political speech when prosecutors No. 23-3840 United States v. Sittenfeld Page 43

argue that the “the purported quid” in a quid pro quo “is a campaign donation” (rather than a “stack of cash,” a “Rolex,” or another item for personal use). Appellant’s Br. 21. He interprets the word “explicit” to require “objective evidence” that a public official agreed to exchange an official act for a campaign donation. Id. at 23. When ambiguous evidence would allow a jury to find either a legal donation or an illegal bribe, this argument goes, the “benefit of any doubt” should protect the First Amendment activity at the heart of most donations. FEC v. Wis. Right to Life, Inc., 551 U.S. 449, 469 (2007) (lead opinion). I agree with the majority opinion that we have rejected Sittenfeld’s reading that McCormick requires explicit evidence. See United States v. Blandford, 33 F.3d 685, 695–99 & nn.13, 16 (6th Cir. 1994). In fact, we have read McCormick to not even require an “express agreement” “in a campaign contribution case.” Id. at 698 n.16.

But I see room for debate over the proper test. To start, McCormick is an opaque opinion that articulated the governing standard in different ways. The Court first used the phrase “explicit quid pro quo” to describe the test that the Fourth Circuit had rejected. McCormick, 500 U.S. at 271. Because the Court disagreed with the Fourth Circuit’s view, perhaps its opinion could be read to adopt a requirement that the quid pro quo (that is, the agreement) be “explicit.” Later, though, the Court suggested that the Hobbs Act might cover a campaign donation when made for a public actor’s “explicit promise” to take an official act. Id. at 273 (emphasis added). This second formulation might suggest that only part of the agreement (the “quo”) must be “explicit.” Still later, the Court did not use the word “explicit” at all and simply said that the lower court had erred by not requiring a “quid pro quo” for campaign contributions. Id. at 274. This third formulation might suggest that the statute lacks an “explicitness” element. In short, McCormick leaves unclear what (if anything) must be “explicit” for campaign contributions to violate the Hobbs Act.

And what did McCormick mean by the word “explicit” anyway? One might have read that adjective to mean that the alleged quid pro quo (or the alleged promise) could not be “implied” and instead had to be “plain in language,” “clear,” or “express.” Webster’s New International Dictionary of the English Language 897 (2d ed. 1934). Our caselaw, though, has not read the term that way. We have reasoned that, while an agreement must be “explicit,” it No. 23-3840 United States v. Sittenfeld Page 44

need not be “express.” See Blandford, 33 F.3d at 696. Given this law, the district court in Sittenfeld’s case correctly instructed the jury that “while the government need not prove that the quid pro quo was express, the government must prove that the quid pro quo was explicit.” Instr., R.202, PageID 3214, 3231. Yet these two words are synonyms. See Webster’s, supra, at 897. So what divides an explicit agreement from an express one? I doubt many jurors would understand this subtle distinction. And I doubt even more that courts should be sending people to prison based on it.

Nor can we fall back on the text of the Hobbs Act to clarify McCormick’s ambiguities. McCormick did not even try to mesh its requirement of an (explicit?) quid pro quo with the text: “the obtaining of property from another . . . under color of official right.” 18 U.S.C. § 1951(b)(2). Rather, the Court “simply made up” the quid-pro-quo element to ensure that the Act did not have “substantial overbreadth.” Evans, 504 U.S. at 286–87 (Thomas, J., dissenting). The Court worried that the Fourth Circuit’s expansive reading (that the Hobbs Act barred a campaign donation if the contributor hoped to “benefit”) would prohibit “conduct that has long been thought to be well within the law[.]” McCormick, 500 U.S. at 272. After all, legislators routinely seek to benefit constituents, and constituents routinely give donations hoping for these benefits. See id. The Court thus adopted the quid-pro-quo element for policy reasons to cut off liability for payments that it did not believe should be illegal. Id. at 272–73. As a result, nothing in the statutory language can help us decide whether the quid-pro-quo element must be “explicit,” whether a public official’s promise must be “explicit,” or whether this “explicitness” element should exist at all.

To be sure, the Court in Evans later tried to ground McCormick’s quid-pro-quo element in the common-law definition of extortion. 504 U.S. at 260. Evans reasoned that extortion at common law was “the rough equivalent of what we would now describe as ‘taking a bribe.’” Id. But this sentence conflicted with Evans’s broad view of the common-law crime. According to Evans, a public official committed extortion at common law whenever the official took “money that was not due to him for the performance of his official duties.” Id. This definition would cover far more than quid-pro-quo bribery. Evans’s test, for example, would likely cover my original hypothetical: the political candidate would have taken the campaign donation (money No. 23-3840 United States v. Sittenfeld Page 45

that was not due) for the promised vote to repeal the Affordable Care Act (the performance of official duties). Not only that, Evans’s definition would likely read the Hobbs Act to cover what the Court has since called mere “gratuities”: payments made for official acts after the fact without a quid pro quo. Snyder v. United States, 603 U.S. 1, 5–6 (2024). If read literally, therefore, Evans would bar campaign donations paid to public officials whenever the officials took an official act with which a contributor agreed. So its expansive reading of the common- law crime also offers no help to resolve what McCormick meant with its narrower (atextual) view of the Hobbs Act.

All this said, I doubt that the Hobbs Act permits Sittenfeld’s proposal to require explicit evidence of a quid pro quo. He would reserve this test for claims that public officials wrongly received campaign contributions (not personal gifts). Appellant’s Br. 21–23. Like McCormick, however, this distinction between campaign donations and other “quids” rests on pure policy. Perhaps Sittenfeld means to invoke the canon of constitutional avoidance, given the First Amendment concerns with chilling ordinary campaign donations and speech requesting them. But that canon acts as a tiebreaker between two “plausible interpretations of a statutory text[.]” Clark v. Martinez, 543 U.S. 371, 381 (2005) (emphasis added). Sittenfeld identifies no plausible reading in which the Hobbs Act could mean one thing for campaign donations and another for personal items. And basic rules of interpretation bar us from turning the Act into “a chameleon” that takes “its meaning” based on “the presence or absence of constitutional concerns in each individual case.” Id. Either an “explicitness” element exists in all cases or it exists in none.

Regardless, Sittenfeld’s proposed test may well not fix the First Amendment problems he identifies. Consider my hypothetical again. If a candidate pledges to repeal the Affordable Care Act in campaign literature requesting campaign contributions, and if contributors disclose that they have supported the candidate because of this promise, why wouldn’t there be an explicit quid pro quo? Indeed, the district court’s definition of “quid pro quo” here would seem to render that hypothetical campaign contribution illegal. The court defined that phrase to cover either “a public official’s solicitations of things of value in exchange for performing or agreeing to perform specific official action,” or “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 No. 23-3840 United States v. Sittenfeld Page 46

official performing or agreeing to perform a specific official action.” Instr., R.202, PageID 3213. Why hasn’t the hypothetical candidate asked for funds “in exchange for performing” an official duty? Id.

Yet I have a hard time seeing how the First Amendment would allow the government to prosecute the political speech and association of the candidate. See McCutcheon, 572 U.S. at 203–04 (plurality opinion). To be sure, the Court has long recognized that the First Amendment allows the government to prohibit “the financial quid pro quo: dollars for political favors.” McCutcheon, 572 U.S. at 192 (plurality opinion) (quoting FEC v. Nat’l Conservative Political Action Comm., 470 U.S. 480, 497 (1985)). But dating back to Buckley, the Court has never defined what that Latin phrase requires with any precision. See id. at 192, 207–09; Citizens United v. FEC, 558 U.S. 310, 359 (2010); Buckley, 424 U.S. at 26–28. The Court likely had in mind traditional bribery. That crime requires the government to prove that an official was “influenced” by the quid (the payment) to take the quo (the official act). 18 U.S.C. § 201(b)(2)(A) (emphasis added). Or, as Blackstone put it, an official commits bribery if the official takes “any undue reward to influence his behaviour in his office.” 5 St. George Tucker, Blackstone’s Commentaries 139 (1803) (emphasis added). This element is missing from my hypothetical because the candidate has promised to repeal the Affordable Care Act whether or not any contributor donates funds to the campaign. The donation did not influence that promised act, even if the contributor gave money “in return for” the promise. Instr., R.202, PageID 3213. But this narrowing element is also missing from the Hobbs Act’s text or from Evans’s broad view of common-law extortion.

So where does this leave us? I agree with Sittenfeld that the current reading of the Hobbs Act raises First Amendment concerns. I disagree, though, that we are the right court to address the concerns. For one thing, the Supreme Court created this dilemma by adopting an ambiguous test seemingly tied to policy rather than text. Only that Court can resolve what it meant by its “quid pro quo” element, what it meant by “explicit,” and how these elements comport with the Hobbs Act’s text. McCormick, 500 U.S. at 271–74. Fittingly for present purposes, one of the earliest decisions about whether the federal judiciary has the power to create common-law crimes involved bribery. See Tucker, supra, at 140 n.26. Because the two judges split on that No. 23-3840 United States v. Sittenfeld Page 47

question, the defendant stood convicted of trying to bribe Tench Coxe in the Treasury Department. See United States v. Worrall, 2 U.S. 384, 394–95 (C.C.D. Pa. 1798). Only later did the Supreme Court hold that federal courts lack “implied powers” to “exercise . . . criminal jurisdiction in common law cases[.]” United States v. Hudson, 11 U.S. 32, 34 (1812). But Sittenfeld’s explicit-evidence test would take us one step more down the road of molding a forbidden common-law crime of bribery under the guise of “interpreting” the Hobbs Act. See Noonan, supra, at 585–86.

For another thing, I see only one reading that can eliminate the First Amendment concerns while adhering to a “plausible” view of the Hobbs Act’s text: the view that the Act does not reach garden-variety bribery. Clark, 543 U.S. at 381; see Evans, 504 U.S. at 279–84 (Thomas, J., dissenting). Here again, however, only the Supreme Court can take that course. Nine years ago, Justice Breyer identified other “problems” that Evans created by conflating extortion and bribery. Ocasio, 578 U.S. at 300 (Breyer, J., concurring). Yet he could not address the problems because the parties accepted Evans as valid. See id. at 300–01. Sittenfeld likewise accepted Evans as valid in our court because “vertical stare decisis” represents an “absolute” mandate. Ramos v. Louisiana, 590 U.S. 83, 124 n.5 (2020) (Kavanaugh, J., concurring in part). But nothing prohibits him from asking the Supreme Court to reassess the Act’s scope in light of three decades’ worth of precedent finding campaign donations entitled to strong First Amendment protection.

II

Turn to the other federal law at issue in this appeal: 18 U.S.C. § 666. This law separately punishes a state or local official who “corruptly solicits or demands for the benefit of any person, or accepts or agrees to accept, anything of value from any person, intending to be influenced or rewarded in connection with any business, transaction, or series of transactions” of the state or local government that has a value of at least $5,000. 18 U.S.C. § 666(a)(1)(B). Both parties have assumed on appeal that McCormick’s quid-pro-quo test for the Hobbs Act (including whatever explicitness element it contains) applies in an identical way to § 666. Indeed, their briefs do not distinguish between these two laws at all. This assumption makes little sense to me. The two laws contain different language, and courts must respect their textual differences No. 23-3840 United States v. Sittenfeld Page 48

under ordinary interpretative rules. See, e.g., Loughrin v. United States, 573 U.S. 351, 358 (2014).

We thus must ask again whether § 666’s language supports Sittenfeld’s McCormick- based claim that the law applies to campaign contributions only if explicit evidence shows that an official entered a quid-pro-quo agreement for the contributions. Here too, I agree with the majority opinion that our caselaw has already rejected this test for § 666. See United States v. Porter, 886 F.3d 562, 565–66 (6th Cir. 2018); United States v. Abbey, 560 F.3d 513, 520–21 (6th Cir. 2009), abrogated on other grounds by Snyder, 603 U.S. at 10–18. In all events, the test suffers from the same textual problems for § 666 as it does for the Hobbs Act. I see no “plausible” reading of § 666 that allows courts to adopt an explicit-evidence requirement for campaign contributions but not for other items “of value” that fall within the law. Clark, 543 U.S. at 381; 18 U.S.C. § 666(a)(1)(B).

That said, I agree with Sittenfeld that § 666 may well raise First Amendment concerns like those raised by the Hobbs Act when prosecutors bring charges against state or local officials over campaign donations. Modifying my hypothetical slightly, I would find this law problematic if it prohibited a city councilmember’s campaign literature from “solicit[ing]” campaign contributions (something “of value”) with the pledge that the candidate plans to repeal an unpopular local ordinance if reelected (“business” of the local “government”). 18 U.S.C. § 666(a)(1)(B). Indeed, most political candidates likely find each citizen’s vote “valuable” because candidates need those votes more than campaign cash to win the election. So one could read § 666 to bar candidates from soliciting citizens to give them their vote (“anything of value”) with promises of official acts. Id. Yet the First Amendment allows political candidates to discuss the actions they will take if elected and to associate with others who share their goals. See Citizens United, 558 U.S. at 359.

Aside from Sittenfeld’s proposed test, though, courts could avoid such problematic readings of § 666 in textually “plausible” ways. Clark, 543 U.S. at 381. Of most note, the Supreme Court recently clarified that the law reaches only bribes (“payments made or agreed to before an official act in order to influence the official with respect to that future official act”)— not gratuities (“payments made to an official after an official act as a token of appreciation”). No. 23-3840 United States v. Sittenfeld Page 49

Snyder, 603 U.S. at 5, 19–20. And the hypothetical city councilmember’s campaign literature would not have been soliciting a bribe because the councilmember would have lacked the “corrupt state of mind” of “intending to be influenced” by the donation to repeal the ordinance. Id. at 11 (emphasis added).

To be sure, Snyder may not eliminate all First Amendment concerns. Later in the opinion, the Court read the word “rewarded” in § 666 to eliminate the “defense that [an] official would have taken the same act anyway and therefore was not ‘influenced’ by the payment.” 603 U.S. at 19. Might this reading eliminate the city councilmember’s defense that he committed to repeal the ordinance with or without the donations? I am not sure. But § 666 can be read to contain other safeguards. For example, the law includes the following safe harbor: “This section does not apply to bona fide salary, wages, fees, or other compensation paid, or expenses paid or reimbursed, in the usual course of business.” 18 U.S.C. § 666(c). A properly documented campaign contribution permitted under state law might qualify as “bona fide . . . compensation paid . . . in the usual course of business.” Id. That reading would take lawful campaign donations off the table for § 666 prosecutions and offer a “plausible” path to distinguish those contributions from, say, under-the-table cash payments or other gifts for an official’s personal benefit. Clark, 543 U.S. at 381.

And even if one disagrees with this view, officials still violate § 666 only if they “corruptly” solicit or accept campaign contributions. 18 U.S.C. § 666(a)(1)(B). What does this adverb mean? Relying on Arthur Andersen LLP v. United States, 544 U.S. 696 (2005), the government in Snyder read it to be “‘associated with wrongful, immoral, depraved, or evil’ conduct.” Brief for the United States at 38–39, Snyder, 603 U.S. 1 (No. 23-108) (quoting Arthur Andersen, 544 U.S. at 705). And since the adverb serves as the statutory mens rea, the government added that a defendant must have acted with “consciousness of wrongdoing” when soliciting or accepting the payment. Id. at 39 (quoting Arthur Andersen, 544 U.S. at 706). At the same time, the Court in Snyder relied on this adverb to support its holding that § 666’s text reaches only quid-pro-quo agreements, not one-sided gratuities. 603 U.S. at 10–12, 18. Does Snyder’s holding leave room for “corruptly” to perform any additional work? I think so. The phrase “intending to be influenced or rewarded” independently conveys the need for the official No. 23-3840 United States v. Sittenfeld Page 50

to subjectively intend the quid pro quo. 18 U.S.C. § 666(a)(1)(B). And, as Judge Silberman suggested in a similar context, the adverb corruptly must add “something” on top of the other statutory elements. United States v. North, 910 F.2d 843, 940–41 (D.C. Cir. 1990) (Silberman, J., concurring in part and dissenting in part). While an ordinary bribe involving a stack of cash might be “inherently, that is, as a matter of law—corrupt,” a campaign contribution for a campaign promise may well not be. Id. at 941. So perhaps federal prosecutors should have to show that an official knew that a particular solicitation of a campaign donation was wrongful (say, because state law prohibited it). See id.

To sum up, Sittenfeld rightly recognizes that § 666 can raise constitutional concerns as applied to campaign donations. But he wrongly relies on an atextual gloss to reduce the concerns.

That said, I disagree with the government that “[t]he jury instructions” for the § 666 charges in this case “required the government to meet McCormick’s” test. Appellee’s Br. 35. Although the instructions comported with our existing caselaw, I doubt that caselaw adequately protected the First Amendment interests at stake. At the time of Sittenfeld’s trial, we did not even read § 666 to require a quid pro quo. See Abbey, 560 F.3d at 520. Snyder overruled this view. 603 U.S. at 10, 19–20. And to the district court’s credit, it did require the government to prove “that there was a quid pro quo agreement between the parties[.]” Instr., R.202, PageID 3224. So Sittenfeld has not asked us to reverse his § 666 conviction based on Snyder alone.

Yet I find our caselaw (and the jury instructions) problematic in other ways. We have not read § 666’s text to require proof that a contributor gave money for a specific act. See Abbey, 560 F.3d at 520–21. Rather, we have said that officials can violate the law if they accept money “with the corrupt intent to use [their] official influence in [the contributor’s] favor” at some unknown point. Id. at 521. Here, then, while the district court imposed a quid-pro-quo element, the court instructed the jury that it could find a nebulous “quo.” That is, it explained that the jury need not find that an “agreement existed as to any specific official act” and that “the agreement could be something as nebulous as, ‘If you contribute to me, I will use my official powers to take care of you once I am in office.’” Instr., R.202, PageID 3224. But the First Amendment does No. 23-3840 United States v. Sittenfeld Page 51

not allow the government to restrict campaign donations in order “to limit the general influence a contributor may have over an elected official.” FEC v. Cruz, 596 U.S. 289, 305 (2022).

Next, we have given the adverb “corruptly” an unclear definition. We have said that this adverb covers any “act done with an intent to give some advantage inconsistent with official duty and the rights of others.” Abbey, 560 F.3d at 520 n.7 (quoting United States v. Aguilar, 515 U.S. 593, 616 (1995) (Scalia, J., concurring in part and dissenting in part)). Consistent with this law, the district court instructed the jury that it must find that “Sittenfeld intended to provide the contributor, in exchange for the contributions, some advantage inconsistent with official duty and the rights of others.” Instr., R.202, PageID 3224. Although this opaque definition flows out of a dictionary definition of “corruptly,” see Ballentine’s Law Dictionary 276 (3d ed. 1969), I confess I have little idea what it requires (and I doubt the jury would have had much of an idea either).

For what it is worth, these instructions appear out of step with the government’s current view of § 666. In Snyder, it told the Court that § 666’s language contains a “parallel requirement” to the federal bribery statute’s “official act” language. 18 U.S.C. § 201(b)(2)(A), (c)(1)(B); Brief for the United States, supra, at 37. The bribery statute requires prosecutors to connect a contributor’s payment to a “specific ‘official act’” of the official. United States v. Sun- Diamond Growers of Cal., 526 U.S. 398, 414 (1999). Likewise, § 666 requires the government to connect a payment to a specific “business, transaction, or series of transactions” of the official’s employer. 18 U.S.C. § 666(a)(1)(B); Brief for the United States, supra, at 37–38. This reading conflicts with the notion that the law reaches a “nebulous” agreement to exchange a payment for a generic “use” of “official powers[.]” Instr., R.202, PageID 3224.

As noted, moreover, the government in Snyder read the adverb “corruptly” to require “‘wrongful, immoral, depraved, or evil’ conduct.” Brief for the United States, supra, at 38–39 (quoting Arthur Andersen, 544 U.S. at 705). And the dictionary that led to our precedent’s vague definition starts with a word—“[w]rongfully”—that this definition overlooked. Ballentine’s Law Dictionary, supra, at 276. Yet this wrongfulness element also did not make its way into the jury instructions in this case. So the district court did not tell the jury that it could convict Sittenfeld only if it found that he recognized the wrongfulness of his campaign solicitations. No. 23-3840 United States v. Sittenfeld Page 52

Sittenfeld did not raise these instructional issues in our court, so they give us no basis to reverse his single § 666 conviction. He instead rested his appeal of both counts on his explicit- evidence theory. That all-or-nothing strategy may have made sense because these § 666-specific issues would not have affected his separate Hobbs Act conviction. But the Supreme Court’s evaluation of this theory’s validity must begin with a proper reading of the statute at issue. The government in Snyder suggested that “if an appropriate case arose, the Court could make clear the boundaries of prosecutions for gratuities disguised as campaign contributions.” Brief for the United States, supra, at 34. Although the Court rejected the government’s view that the statute covered gratuities, this case shows that the need for these clear boundaries continues to exist.

* * *

All told, I share Judge Bush’s concerns with Sittenfeld’s prosecution in this case. Yet I part ways with the dissent’s conclusion that courts should depart from their usually deferential sufficiency-of-the-evidence test to police the boundary between a protected campaign donation and an illegal bribe. I see no need to rely on “pragmatism” over “principle” because a principled reading of the laws could eliminate any First Amendment concerns. But only the Supreme Court can provide that reading. Under the law as it exists now, I concur in the majority opinion. No. 23-3840 United States v. Sittenfeld Page 53

_________________

DISSENT _________________

JOHN K. BUSH, Circuit Judge, dissenting. Alexander Sittenfeld’s aggressive campaign fundraising caught the attention of the FBI. Over the course of more than year, a sting operation tried to catch him saying or doing something that might serve as evidence of a bribe or extortion masked as a campaign contribution. Only a few words—the “love you but can’t” remark—in his early conversation with an undercover informant created ambiguity as to his intent. But Sittenfeld’s subsequent statements and dealings with both the informant and undercover FBI agents clarified that his aim was to obtain a legitimate contribution, not an improper payment. Viewed in its totality, the proof at trial was insufficient to sustain Sittenfeld’s guilty verdict under McCormick v. United States, 500 U.S. 257 (1991).

The majority concludes otherwise. It finds that this single, ambiguous line—by a politician who refused contributions that did not comply with the law, resisted frequent attempts by the FBI’s pretend donors to cajole him with personal gifts, and repeatedly reminded these donors of how to contribute to his campaign legally—was sufficient to criminalize an otherwise lawful and constitutionally protected campaign contribution. Granted, the FBI agents and their informant, Chinedum Ndukwe, testified to their belief that there was a corrupt bargain. But all of the secret audio and video recordings of Sittenfeld after the “love you but can’t” remark captured normal political activity. They reveal that Sittenfeld himself viewed the transaction as lawful and that he did what one would expect for a legal contribution.

The evidence, viewed in whole, fell short of the minimum required for a rational juror to find beyond a reasonable doubt that Sittenfeld had the corrupt intent for bribery or extortion. The proof cannot sustain the explicit quid pro quo that McCormick requires to convict. Indeed, one can only reasonably infer Sittenfeld viewed the payment as a campaign contribution and nothing else. I therefore respectfully dissent.1

1 Because the evidence is insufficient to uphold the verdict, it is not necessary to address Sittenfeld’s argument that the prosecution constructively amended the indictment at trial. No. 23-3840 United States v. Sittenfeld Page 54

I. The Court’s Responsibility to Police the Line of Sufficiency

The government is correct that Sittenfeld “bears a very heavy burden” to show insufficiency of the evidence. United States v. Emmons, 8 F.4th 454, 478 (6th Cir. 2021) (citation omitted). But at least three aspects of this case distinguish it from the vast majority of sufficiency challenges to bribery or extortion convictions. First, Sittenfeld’s conviction involves only a campaign contribution and nothing that would directly benefit him financially or otherwise unambiguously show criminal intent. Second, Sittenfeld’s conduct that the government calls proof of an illegal quid pro quo is equally consistent with lawful activity that one would expect for a legitimate campaign contribution. Third, most of the government’s proof comes from audio and video recordings instead of testimony.

A. Unusualness of this Case

First, consider how unusual this case is among bribery or extortion cases that involve campaign contributions. Perhaps every such case at the federal appellate level that sustained a guilty verdict since McCormick has had at least one of the following characteristics: (1) the defendant received personal gifts (cash, vacations, dinners, etc.) in addition to campaign contributions, (2) the government proved that the campaign contributions were part of an independently criminal scheme, or (3) there was direct, unambiguous evidence of a quid pro quo. Sittenfeld’s case has none of them.

In each of these circumstances, a sufficiency challenge faces an uphill battle because there is little risk of the jury mistakenly finding that the defendant had corrupt intent. For example, in the first scenario, money going directly into a politician’s pocket does not parallel a legitimate campaign contribution, to say the least. See, e.g., United States v. Ring, 706 F.3d 460, 464 (D.C. Cir. 2013) (“Ring . . . relied heavily on campaign contributions to maintain relationships with elected officials and promote [his] clients’ political interests. But it was Ring’s other lobbying tactics that got him in trouble. These tactics chiefly included treating congressional and executive branch officials to dinners, drinks, travel, concerts, and sporting events.” (emphasis added)). As to the second scenario, one should not worry about jurors erroneously imputing corrupt intent to, say, a judge who decides a motion without reading the No. 23-3840 United States v. Sittenfeld Page 55

briefs, a legislator who encourages campaign contributions via illegal straw donors, or a local councilwoman who pressures a prosecutor to bring charges against an individual. See United States v. Terry, 707 F.3d 607 (6th Cir. 2013) (judge); United States v. Benjamin, 95 F.4th 60 (2d Cir. 2024) (legislator); United States v. Lee, 919 F.3d 340 (6th Cir. 2019) (local councilwoman). And as to the third, it is self-evident that direct and unambiguous evidence of a quid pro quo can prove the existence of the same. None of these fact patterns are present here.

Consequently, Evans v. United States, 504 U.S. 255 (1992) and United States v. Blandford, 33 F.3d 685 (6th Cir. 1994) prove less helpful than they may first appear. In Evans, an FBI agent gave the defendant not only $1,000 in campaign contributions but also $7,000 for personal use, thus placing the case squarely within the first scenario of a direct payment to the politician, as explained above. See Evans, 504 U.S. at 257. Blandford is less relevant still. It involved no campaign contributions at all, only a personal bribe—an even more obvious indicia of corrupt intent. It is also difficult to find authoritative the Blandford court’s interpretation of Evans when that court denied that Evans applied to its case-at-hand. See Blandford, 33 F.3d at 696. Indeed, in United States v. Abbey, we suggested Blandford’s reading of Evans only applies outside the campaign contribution context, and we have not cited Blandford in a campaign contribution case since. See 560 F.3d 513, 517–18 (6th Cir. 2009).

By contrast, consider the “thin legal ice” of a prosecution based on a campaign contribution alone, such as the one here. Empress Casino Joliet Corp. v. Balmoral Racing Club, Inc., 831 F.3d 815, 825 (7th Cir. 2016). Citizens make campaign contributions to influence policymaking, and these contributions constitute political speech under the First Amendment. See Citizens United v. FEC, 558 U.S. 310, 359 (2010) (“a substantial . . . if not the only reason . . . to make a contribution to, one candidate . . . is that the candidate will respond by producing those political outcomes the supporter favors” (citation omitted)); McCutcheon v. FEC, 572 U.S. 185, 191 (2014) (plurality op.). For this reason, it is particularly worrisome when, as here, a criminal conviction derives solely from a purported campaign contribution that, if legitimate, would be constitutionally protected.

A politician may take a legislative position, fundraise off it, and then fundraise more after successfully passing laws that advance it. An official does not break the law when he adopts a No. 23-3840 United States v. Sittenfeld Page 56

position that aligns with a donor’s interest. See Brown v. Hartlage, 456 U.S. 45, 55–56 (1982). This happens all the time, and nobody doubts its legality. No case suggests a politician can express policy positions only when he wholeheartedly thinks they are good ideas. And no case suggests it is illegal for a politician to take a position because he believes his constituents will support it or because adopting it increases the likelihood that donors will contribute to his campaign. To be sure, if a donor approaches a politician and makes a direct trade of dollars for official action, then it is illegal under McCormick. But bribery and extortion are extremely difficult to prove where, as here, a politician receives no direct payment into his pocket but only indirectly benefits from a contribution made to his campaign. See Empress Casino, 831 F.3d at 824–25.

We therefore need to consider Sittenfeld’s prosecution in light of its unusual facts—a conviction based on an allegedly corrupt campaign contribution and not anything else. This case presents the most troubling context for application of the bribery and extortion laws: when conduct arguably protected by the Constitution is at issue. In these circumstances, a case will likely come close to the line of sufficiency, so we must police that line vigorously.

B. Evidence of Sittenfeld’s Alleged Corrupt Intent Is Consistent with Lawful Motive

The second reason to zealously guard the line of sufficiency here is that the government’s proof of Sittenfeld’s alleged corrupt intent is entirely consistent with his having a lawful motive.

As the majority correctly says, “state of mind is at the heart of the inquiry.” Majority Op. at 15. For bribery, the government must prove that Sittenfeld had a “corrupt state of mind and the intent to be influenced in [an] official act.” Snyder v. United States, 603 U.S. 1, 12 (2024). Though “it goes without saying that matters of intent are for the jury to consider,” McCormick, 500 U.S. at 270, the law does not give jurors full autonomy to decide corrupt intent. That is because it can be very difficult to distinguish a legitimate contribution from a corrupt bargain. Thus, as an additional guardrail, the government must show an explicit quid pro quo to reach a jury. Snyder, 603 U.S. at 23 (Jackson, J., dissenting) (citing United States v. Sun-Diamond Growers, 526 U.S. 398, 404 (1999)) (for federal programs bribery); McCormick, 500 U.S. at 273 (for Hobbs Act extortion). “A quid pro quo means a specific intent to give or receive something No. 23-3840 United States v. Sittenfeld Page 57

of value in exchange for an official act. So, for a payment to constitute a bribe, there must be an upfront agreement to exchange the payment for taking an official action.” Snyder, 603 U.S. at 23–24 (Jackson, J., dissenting) (cleaned up) (quoting Sun-Diamond Growers, 526 U.S. at 404– 05).

A helpful way to analyze a quid pro quo is through the lens of contract formation. After all, a quid pro quo is a contract: as the majority correctly notes, there must be “a meeting of the minds,” Majority Op. at 15, and the official “must receive some consideration for his promise,” id. An agreement becomes illegal only if there is mutual consent for a specific contribution to “control[]” the politician’s vote or other official act. McCormick, 500 U.S. at 273. But to anyone other than the politician, it can be extremely difficult to answer when such mutuality exists, particularly when, as here and in Evans, the government has only circumstantial evidence to prove its case.

Not only is it difficult to discern whether there is an illegal bargain in a case like Sittenfeld’s, but there is also a high cost to getting it wrong. “The right to participate in democracy through political contributions is protected by the First Amendment[.]” McCutcheon, 572 U.S. at 191. And this constitutional guarantee is of the highest caliber. See FEC v. Cruz, 596 U.S. 289, 302 (2022) (“The First Amendment has its fullest and most urgent application precisely to the conduct of campaigns for political office.” (citation omitted)). Prosecution of campaign contributions therefore runs a substantial risk of chilling political speech. See NAACP v. Button, 371 U.S. 415, 433 (1963) (“The threat of sanctions may deter . . . almost as potently as the actual application of sanctions.”); Virginia v. Hicks, 539 U.S. 113, 119 (2003) (“[T]he threat of enforcement of an overbroad law may deter or ‘chill’ constitutionally protected speech— especially when the overbroad statute imposes criminal sanctions.”); Americans for Prosperity Found. v. Bonta, 594 U.S. 595, 618–19 (2021) (“[T]he protections of the First Amendment are triggered not only by actual restrictions on an individual’s ability to join with others to further shared goals. The risk of a chilling effect on association is enough because First Amendment freedoms need breathing space to survive.” (citation omitted)).

Consequently, we need to meaningfully enforce the line of sufficiency in cases such as this one, where a jury could mistake lawful—indeed, constitutionally protected—conduct for No. 23-3840 United States v. Sittenfeld Page 58

evidence of a crime. See Counterman v. Colorado, 600 U.S. 66, 75 (2023) (“A speaker . . . may worry that the legal system will err, and count speech that is permissible as instead not.”) In this regard, an analogy from antitrust law, though imperfect, can guide us. Like bribery and extortion cases involving campaign contributions, caselaw concerning allegedly illegal restraints of trade can present facts capable of being viewed both ways as to whether an unlawful bargain was struck. As a result, the Supreme Court has articulated particular doctrines for deciding which suits may proceed to discovery and which of those have sufficient evidence to reach a jury. When parties in those cases are accused of unlawful collusion, “it is of considerable importance that [legal] independent action . . . be distinguished from [illegal] . . . agreements.” Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 763 (1984). This task bears striking similarity to our job to discern the difference between politicians and donors engaged in lawful parallel acts (official acts and campaign contributions) that advance overlapping interests, on the one hand, and such actors when they strike corrupt bargains, on the other.

In restraint-of-trade cases, the former circumstance is known as conscious parallelism, a lawful “common reaction of firms in a concentrated market that recognize their shared economic interests and their interdependence.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 553 (2007) (cleaned up). Parallel behavior can be evidence of an illegal bargain, but it may also simply reflect lawful, rational decisionmaking by the parties involved. This presents a deep problem in the field of antitrust:

[J]udged from a distance, the conduct of the parties in the various situations can be indistinguishable. For example, the fact that a manufacturer and its distributors are in constant communication about prices and marketing strategy does not alone show that the distributors are not making independent pricing decisions. A manufacturer and its distributors have legitimate reasons to exchange information about the prices and the reception of their products in the market.

Monsanto, 465 U.S. at 762. Put succinctly, the problem is that actions suggesting illegal collusion are often “consistent with conspiracy, but just as much in line with a wide swath of rational and competitive business strategy.” Twombly, 550 U.S. at 554.

Switch “political strategy” for “business strategy” and one can see our problem: it is perfectly normal—and legal—for politicians and their donors to have shared interests and act on No. 23-3840 United States v. Sittenfeld Page 59

them in parallel, yet it is illegal to make an unambiguous agreement exchanging official acts for campaign contributions. And just as “mistaken inferences in [restraint-of-trade] cases . . . are especially costly, because they chill the very conduct the antitrust laws are designed to protect,” Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986), so too are mistaken inferences in prosecutions of campaign contributions because they chill conduct the First Amendment protects.

In the face of such precarious challenges, the Supreme Court does not permit judges in restraint-of-trade cases to sit back and let juries decide right from wrong when there is a mere possibility of inferring illegality from circumstantial evidence. Cf. Twombly, 550 U.S. at 573 (Stevens, J., dissenting) (“[T]here is a risk that jurors may mistakenly conclude that evidence of parallel conduct has proved that the parties acted pursuant to an agreement when they in fact merely made similar independent decisions.”). Instead, “antitrust law limits the range of permissible inferences from ambiguous evidence.” Matsushita, 475 U.S. at 588. For a collusion allegation to succeed, there must be “evidence that tends to exclude the possibility that the” alleged conspirators “acted independently,” which, as discussed above, cannot be the mere fact that the two parties talked about the subject at issue. Monsanto, 465 U.S. at 764. Therefore, “business behavior is admissible circumstantial evidence from which the fact finder may infer agreement. But this Court has never held that” conscious parallelism “establishes agreement.” Theatre Enters., Inc. v. Paramount Film Distrib. Corp., 346 U.S. 537, 540–41 (1954) (emphasis added) (citations omitted). In other words, “conduct as consistent with permissible competition as with illegal conspiracy does not, standing alone, support an inference of antitrust conspiracy.” Matsushita, 475 U.S. at 588. And just like McCormick’s quid pro quo requirement, these safeguards come from judicial construction of a broad text.

Consider In re Text Messaging Antitrust Litigation, 782 F.3d 867 (7th Cir. 2015), which is strikingly similar to our case. The plaintiffs there alleged that cellphone service providers illegally colluded to raise the price of text messages. Like the government in our case, “the plaintiffs presented circumstantial evidence consistent with an inference of collusion, but . . . equally consistent with independent parallel behavior.” Id. at 879. And like the government in our case, the plaintiffs had an additional arguable “smoking gun”: emails from a T-Mobile No. 23-3840 United States v. Sittenfeld Page 60

employee who worked on text message pricing saying “at the end of the day we know there is no higher cost associated with messaging. The move [the latest price increase by T–Mobile] was colusive [sic] and opportunistic.” Id. at 871–72 (brackets in original). These ambiguous emails were of a type akin to the “love you but can’t” line that the majority finds so damning to Sittenfeld. But it was even worse than that. The T-Mobile employee also tried to delete the emails, id. at 873, which evinces a guilty mind far more than Sittenfeld trying to honor the FBI agents’ request to keep their contributions anonymous.

The Seventh Circuit refused to let that case reach a jury. Using other evidence in the record to contextualize the emails, the court found that they “need not imply express collusion.” Id. As for the alleged coverup, the court found no reason to characterize it as such because “there is . . . an equally plausible reason for the deletion of the emails in question,” namely that the employee was afraid of “abusing one’s corporate superiors.” Id.

In re Text Messaging is one of many cases in which courts of appeals have held that evidence of an isolated, ambiguous “smoking gun” communication is not enough to create a triable issue of collusion when the totality of proof shows only conscious parallelism. See also Anderson News, LLC v. Am. Media, Inc., 899 F.3d 87, 108 (2d Cir. 2018) (email said “we should start simultaneously using our collective resources and influence to direct business”); Kleen Prods. LLC v. Georgia-Pac. LLC, 910 F.3d 927, 932 (7th Cir. 2018) (emails encouraged purported colluders to avoid “anti-trust issues” and said “everybody needs to do the same thing”); In re Citric Acid Litig., 191 F.3d 1090, 1098 (9th Cir. 1999) (handwritten note from a meeting of industry leaders said the “[u]ndertaking is a confidential agreement to maintain price” and meeting minutes suggested illegal collusion was plainly proposed and then immediately rejected). It is instructive that so many courts of appeals have found insufficient evidence in collusion cases with facts similar to those here.

It is not far-fetched to import principles from antitrust law to other legal contexts such as the one here. Indeed, civil procedure syllabi routinely include Twombly, Matsushita, and Monsanto because they set standards for sufficiency to reach discovery (Twombly), trial (Matsushita), and verdict (Monsanto)—they are not outliers specific to antitrust. Their lesson in how to police the line of sufficiency is widely applicable. It is all the more urgent that our court No. 23-3840 United States v. Sittenfeld Page 61

in particular heed this lesson given the similarity of our case to those seminal civil cases and the greater consequences at stake in a criminal prosecution. Indeed, it would be strange to allow the government to imprison Sittenfeld on evidence that would not reach the jury in a civil lawsuit. And it would be stranger still to treat core political speech less carefully than common corporate activity.

In analogizing to antitrust, I do not seek to impose a foreign standard of review in this case. Instead, restraint-of-trade cases can prove useful in policing the sufficiency line because that field of law has more frequently faced Sittenfeld-like circumstances than the bribery/extortion field has. Under the standard of proof applicable here (beyond a reasonable doubt for any rational trier of fact), the government’s case surely fails if it only produces evidence “consistent with permissible [political activity] as with illegal conspiracy,” Matsushita, 475 U.S. at 588, or does not “exclude the possibility that the” alleged conspirators “act[ed] independently,” Monsanto, 465 U.S. at 764.

As the Supreme Court observed less than three years ago, “[t]o be sure, the line between quid pro quo corruption and general influence may seem vague at times, but the distinction must be respected to safeguard basic First Amendment rights. And in drawing that line, the First Amendment requires us to err on the side of protecting political speech rather than suppressing it.” Cruz, 596 U.S. at 308 (citations omitted). Respectfully, while the majority hesitates to police that line, I believe that is precisely what the Constitution requires us to do.

C. Audio and Video Recordings Capture Sittenfeld’s Communications

Finally, there is a third reason here to guard the sufficiency line that is not present in all appeals of a bribery or extortion conviction: namely, the FBI introduced evidence of its interactions with Sittenfeld through audio or video recordings rather than testimony alone. When witnesses testify about events that are not recorded, a jury can entirely credit or ignore the testimony. In that situation we, as the appellate court, must defer to the jury: we review a verdict by crediting testimony that incriminates and ignoring what exculpates. See United States v. Osborne, 886 F.3d 604, 608 (6th Cir. 2018) (“All reasonable inferences and resolutions of credibility are made in the jury’s favor.” (citation omitted)). No. 23-3840 United States v. Sittenfeld Page 62

But when the evidence is recorded, the calculus changes. Like with the email smoking guns of In re Text Messaging and the other cases noted above, the court does not need a witness to know what was said. There is nobody’s credibility to evaluate. The Supreme Court also made this point in Scott v. Harris, 550 U.S. 372 (2007). The question was whether a police officer used excessive force in a car chase when he forced the plaintiff’s car to crash. On a summary judgment appeal for denial of qualified immunity, Justice Scalia, writing for eight justices, admonished the appeals court for uncritically accepting the plaintiff’s interpretation of events (that he was driving responsibly and did not pose a threat to the public) when the video evidence blatantly contradicted it. Id. at 378–80.

In this case, therefore, when we construe evidence in a light most favorable to the verdict, we should “view[] the facts in the light depicted by the videotape,” id. at 381, instead of through the subjective interpretation of the FBI-agents and their informant. And we should derive only “reasonable inferences” from the recordings in the verdict’s favor, Osborne, 886 F.3d at 608 (emphasis added) (citation omitted).

II. The Government’s Insufficient Evidence of an Explicit Quid Pro Quo

With these distinguishing aspects of the case in mind, consider the relevant facts. For the immense effort the FBI put into this sting, it found scant proof to support its case that Sittenfeld agreed to a corrupt bargain. The sting lasted over a year and involved two undercover FBI agents and one of Sittenfeld’s friends, Chinedum Ndukwe, a former football player for the Cincinnati Bengals who had long contributed to Sittenfeld’s political campaigns. The FBI recorded every discussion during the sting.

Prosecutors could only point to a single ambiguous remark in a single early phone call as standalone evidence that Sittenfeld was doing anything out of the ordinary. This, as noted, was the “love you but can’t” phrase prominently discussed in the parties’ briefs and the majority opinion. According to the prosecution’s interpretation, Sittenfeld used the October 30th call to threaten Ndukwe and, by proxy, Ndukwe’s purported business partner, the undercover FBI agent named Rob: if Rob or his associates failed to contribute to the campaign, then Sittenfeld would refuse to help them as mayor. Appellee’s Br. at 38–39. Remember, the government indicted No. 23-3840 United States v. Sittenfeld Page 63

Sittenfeld for extorting and receiving a bribe from Rob, not Ndukwe. So to be certain it did not constructively amend the indictment, the government must argue that any threat to Ndukwe was actually a threat to Rob. See Majority Op. at 36–37.

The majority believes the “love you but can’t” line is sufficient to convict for bribery. It also believes Ndukwe approached Sittenfeld with a quid pro quo offer from Rob on a November 2nd phone call and that Sittenfeld agreed to it when he met Rob on November 7th at Nada restaurant. I disagree with both points.

Beginning with the supposed coercion of Ndukwe on the October 30th call, it is important to note how little we can glean from the cryptic words “love you but can’t” themselves. Sure, Ndukwe testified that he took the line as a threat. Crediting that testimony only proves Ndukwe interpreted the remark that way. Even if one could ignore Ndukwe’s natural bias as a government informant to interpret Sittenfeld’s statement in the worst possible way, the recordings contradict his account. The other evidence at trial goes against Ndukwe’s interpretation, so much so that to adopt his interpretation would be to make an unreasonable inference. And a review for sufficiency of the evidence requires us to review the entire trial record, not just this single statement in isolation. See United States v. Vichitvongsa, 819 F.3d 260, 270 (6th Cir. 2016). That comprehensive review, I respectfully submit, the majority opinion fails adequately to do.

The complete context includes the recorded November 7th meeting with Rob at the aptly named restaurant, Nada. That conversation negates whatever value the government claims “love you but can’t” says about Sittenfeld’s intent. The recording demonstrates that Sittenfeld clearly did not think he had coerced Ndukwe (and Rob, by proxy) into a corrupt bargain by saying “love you but can’t.” Instead, he practically begged Rob for a campaign donation. Sittenfeld marshalled evidence that he was an influential legislator to convince Rob that he could get his 435 Elm plan enacted. He even brought a slideshow to make the point. Why would Sittenfeld make such effort to convince Rob to donate if had he bullied Ndukwe into promising a donation from Rob a week earlier? No. 23-3840 United States v. Sittenfeld Page 64

Next consider the theory that, outside the context of “love you but can’t,” Sittenfeld agreed to a quid pro quo at Nada. The recording suggests nothing in the way of corrupt intent on Sittenfeld’s part. He only tried to convince Rob to donate to him because he was an influential, pro-business politician—it is far from illegal for Sittenfeld to suggest his influence and positions could benefit Rob’s interests. See Citizens United, 558 U.S. at 359; Cruz, 596 U.S. at 308 (“influence and access embody a central feature of democracy” (citation omitted)). When Rob tried to frame his potential contribution as an explicit bribe, he flubbed. Even he admitted that Sittenfeld took his invocation of a “deal” to refer to the 435 Elm development deal he wanted with the city as opposed to a deal for a bribe. No rational juror could find that Sittenfeld evinced corrupt intent there when even Rob believed Sittenfeld did not take the conversation to be about bribery.

In fact, everything Sittenfeld said at the November 7th Nada meeting aligned with the terms of the legitimate campaign contribution that he had described to Ndukwe on the November 2nd phone call, captured by the FBI on audiotape. On that call, Ndukwe made the first and only clear request for a bribe, but Sittenfeld rebuffed him. Rather than accept the bribe, Sittenfeld avowed his pro-business bona fides like he later did with Rob. There is nothing illegal about that. Sittenfeld suggested that though a contribution could not control his action, his preconceived policy positions made him, to quote something he later told Rob, a “good bet[] and good investment[].” The benefit to Ndukwe and company would not be control over Sittenfeld’s action or a guarantee per se, but instead having a reliably pro-business mayor in office who was not predisposed against them like Mayor Cranley supposedly was. In the context of the whole record then, neither the isolated and ambiguous “love you but can’t” statement nor any subsequent conversation can prove extortion or bribery beyond a reasonable doubt, even under Evans’s permissive circumstantial evidence rule.

The majority cites other actions and words of Sittenfeld to suggest the existence of a quid pro quo, but they all suffer from the same flaw: each is “consistent with” what we would expect a law-abiding, pro-business politician to do. Twombly, 550 U.S. at 554. They are emblematic, at most, of the conscious parallel conduct that the Supreme Court has said is insufficient to take an unlawful collusion case to a jury. See id. at 553–54. No. 23-3840 United States v. Sittenfeld Page 65

First, the majority suggests Sittenfeld came to the Nada meeting ready to support 435 Elm because he knew from the November 2nd call that Rob was amenable to a bribe. See Majority Op. at 19–20. But any politician who plans a dinner with a potential donor will highlight the politician’s support for the interests that drive the donor. And the clarity with which Sittenfeld laid out what a contribution would and would not accomplish diminishes the relevance of this argument to a point that it is an unreasonable inference. Next the majority finds it incriminating that Sittenfeld can be seen nodding along to Rob’s points. But common experience teaches that one nods in a conversation to indicate one’s attention and understanding to what is said, not necessarily agreement with it.

Sittenfeld also, according to the majority, “showed apparent disregard for the true source of the donations” while caring meticulously about the donation procedure at the Nada meeting. Majority Op. at 20. By this the majority implies that the potential for Rob to use straw donors did not bother Sittenfeld. The recording shows otherwise. Until the Nada meeting, Sittenfeld and Ndukwe only discussed donating through LLCs, which was unquestionably legal. Sittenfeld heard at Nada for the first time that Rob wanted to donate through perhaps eighteen personal checks or money orders because he was afraid LLC donations could be traced to him. As the majority says, Sittenfeld responded by reminding Rob of the rules for attributing personal donations to donors. But because Rob told Sittenfeld earlier in the conversation that his investment group had more than twenty investors, it was obvious that Rob could round up eighteen checks without resorting to straw donors. A later piece of the conversation confirms Sittenfeld did not understand Rob to have proposed straw donors. When Rob mentioned giving Sittenfeld “money orders from you know John Doe,” Sittenfeld stopped him to make sure he meant “a real real person . . . whose [sic] part of your guys’ network.” Appellate R. 23, p. 84. Rob affirmed that he did. Viewed in whole then, the conversation does not support a reasonable inference that Sittenfeld was fine with straw donors.

The majority also leaves out that earlier in the conversation Sittenfeld tried to convince Rob to stick with the original plan of LLC donations, but directed toward Sittenfeld’s PAC instead of his campaign. That, according to Sittenfeld, would achieve Rob’s goal of staying anonymous. And in the end, Rob did contribute legally to Sittenfeld’s PAC via LLCs. No. 23-3840 United States v. Sittenfeld Page 66

The majority sees the potential to fault Sittenfeld there, too. It finds “at least some probative value” in Sittenfeld’s eagerness to help Rob and company stay anonymous. Majority Op. at 20. But that cannot possibly incriminate Sittenfeld. The right to donate anonymously to political organizations (like a political action committee, as here) is one of the strongest protections of the First Amendment’s right to freely associate. That freedom applies whether the beneficiaries are black activists in 1950s Alabama, conservative activists in 2010s California, or make-believe real estate developers in present-day Cincinnati. See NAACP v. Alabama, 357 U.S. 449 (1958); Americans for Prosperity Found. 594 U.S. at 595. It cannot possibly be that politicians must or even should ask their followers why they want to keep their contributions as hidden as the law allows.

And that is the full extent of what Sittenfeld offered to do for his donors. Each time Sittenfeld saw potential impropriety in how Rob wanted to contribute, he reminded Rob how to do so legally and refused a donation when it did not comply with the law.

To prove a quid pro quo, there must be evidence of an explicit agreement on both sides of the transaction, not just from the side conducting the sting operation. See Blandford, 33 F.3d at 696. Viewed in the light most favorable to the government, the evidence shows only a single ambiguous statement made by Sittenfeld to Ndukwe that could be read to suggest improper intent. But the totality of the evidence vitiates that statement’s probative value. To be sure, Ndukwe and the FBI agents testified to their belief that there was a corrupt bargain. But there is no proof that Sittenfeld himself viewed the campaign contribution that way. In fact, the video and audio recordings show just the opposite—that at every turn he did what one would expect to ensure its legality.

Even though the government tried to get Sittenfeld to take the bribery bait, he never bit. No illegal agreement ever materialized. Sittenfeld accepted the campaign contribution for what he thought it was—a campaign contribution—and not the product of any explicit quid pro quo for official acts. No. 23-3840 United States v. Sittenfeld Page 67

III. Reconsidering McCormick

Because the prosecuted conduct here is solely a purported campaign contribution, this case falls in the danger zone that surrounds the sufficiency line of bribery and extortion cases. Given that concern, and given post-McCormick caselaw that more strongly protects campaign contributions under the First Amendment, it would be helpful for the Supreme Court to provide guidance here. The lower courts need to know the extent of McCormick’s protections in cases where the only allegation of illegality relates to corrupt, but otherwise lawful, campaign contributions.

To understand the danger here, consider the majority’s assertion that McCormick would condemn a politician who says, “because of this gift I will now be sure to keep my campaign- trail promise.”2 Majority Op. at 15. If a court can reach this conclusion based on the logic of McCormick, then perhaps the Supreme Court should clarify the reasoning of that decision.

This example is much like common political horse-trading. Legislators routinely promise to vote for a colleague’s bill in consideration of a colleague’s vote on their bill. Interest groups routinely promise the votes of their members if a politician pledges to take a certain policy position. And candidates for executive office routinely endorse their opponents, give flattering speeches at campaign events, and wind up with a cabinet post. Surely these are things of value. Under the majority’s reading of McCormick, each of these examples contravenes the plain meaning of the bribery and extortion statutes. Further, campaign contributions are similar to each of these because “[they] are valuable only as a means to get votes . . . . A legislator who receives a contribution has increased her expected number of votes by a certain amount . . . . The legislator does not get anything more out of the contribution than that.” David A. Strauss, Corruption, Equality, and Campaign Finance Reform, 94 Colum. L. Rev. 1369, 1373 (1994). Indeed, Judge Murphy notes that Evans’s interpretation of the Hobbs Act appears to go at least this far. Concurring Op. at 44–45.

2 One obvious critique of this claim is that necessarily the politician already made the promise, so reiterating it cannot qualify as contract-like consideration to complete the quid pro quo. No. 23-3840 United States v. Sittenfeld Page 68

So what separates campaign contributions from the other items in the list? McCormick’s pragmatic explanation is that “a proposal to trade one public act for another, a form of logrolling,” United States v. Blagojevich, 794 F.3d 729, 734 (7th Cir. 2015), is so common that Congress could not have meant to outlaw it. See McCormick, 500 U.S. at 272; see also Blagojevich, 794 F.3d at 735 (“It would be more than a little surprising to Members of Congress if the judiciary found in the Hobbs Act . . . a rule making everyday politics criminal.”).3 Indeed, the Court “did not purport to discern that requirement in the common law or statutory text, but imposed it to prevent the Hobbs Act from effecting a radical (and absurd) change in American political life. ‘To hold otherwise would open to prosecution not only conduct that has long been thought to be well within the law but also conduct that in a very real sense is unavoidable so long as election campaigns are financed by private contributions or expenditures, as they have been from the beginning of the Nation.’” Evans, 504 U.S. 286–87 (Thomas, J. dissenting) (quoting McCormick, 500 U.S. at 272–73).

So, a reasonable inference from McCormick is that pragmatism rules this area of law much more so than principle. Cf. Sun-Diamond Growers, 526 U.S. at 412 (Regarding laws about bribery, “precisely targeted prohibitions are commonplace, and [] more general prohibitions have been qualified by numerous exceptions. Given that reality, a statute in this field that can linguistically be interpreted to be either a meat axe or a scalpel should reasonably be taken to be the latter.”) This buttresses my belief that courts in cases like this one must firmly guard against interpretations of McCormick that allow a prosecutor to reach a jury merely by showing the combination of campaign contribution and an official act benefitting that contributor.

Judge Murphy believes the better move is to refrain from employing pragmatism when a principled approach would prove superior. Concurring Op. at 52. I agree that we should prefer a textualist interpretation to the policy-centric reasoning of McCormick. I also agree with Judge Murphy that, in lieu of such a change, further Supreme Court guidance would help lower courts, particularly for cases like this one where there is no unambiguous evidence of a quid pro quo and

3 I agree with Judge Murphy and the three dissenting justices in Evans that the Hobbs Act, properly construed, should apply only to officials who tell people that their office entitles them to a payment. See Evans, 504 U.S. at 282–83 (Thomas, J., dissenting). No. 23-3840 United States v. Sittenfeld Page 69

no independent indicia of corrupt intent, like personal gifts or a connection to an independently criminal scheme.

However, I respectfully disagree with Judge Murphy on one point. He recognizes as I do that the Court’s vague doctrine in this field leaves the exercise of some First Amendment rights in a precarious position. Concurring Op. at 46–47. But I do not believe we must do nothing until the Court does something. To the contrary, courts routinely apply precedents that do not cleanly address the situation at hand. If we can choose to apply controlling precedent in one way that protects First Amendment rights or in another that imperils them, I believe our duty, both to the Constitution and the Court’s precedent, compels us to choose the former.

We need not contradict McCormick or Evans to do so here. McCormick did not mark itself as the fullest extent of protection for campaign contributions against Hobbs Act prosecutions. In fact, it is more precise than many acknowledge. McCormick ultimately decided that “a quid pro quo is [] necessary for conviction under the Hobbs Act when an official receives a campaign contribution.” 500 U.S. at 274. The Court also held that contributions violate the Hobbs Act “only if the payments are made in return for an explicit promise or undertaking by the official to perform or not to perform an official act.” Id. at 273 (emphasis added). In other words, a quid pro quo, is necessary but not sufficient to convict. Therefore, the Court in McCormick did not explicitly settle the full extent of First Amendment protection for campaign contributions in the context of a bribery or extortion case. Judge Murphy’s analysis suggests the McCormick Court saw this as a feature, not a bug; it only wanted to decide the narrow issue before the Court. Concurring Op. at 41–42. It did not foreclose an interpretation of the Hobbs Act that includes additional safeguards to prevent convictions for common politicking. Neither did Evans, which in its most generous reading only explained McCormick’s quid pro quo requirement without purporting to modify it. And neither did Justice Kennedy’s Evans concurrence that the majority cites.

The path of the caselaw after McCormick strongly suggests we should construe First Amendment principles broadly because they “treat the act of giving money as one of the central means by which citizens participate in politics.” Deborah Hellman, A Theory of Bribery, 38 Cardozo L. Rev. 1947, 1988 (2017). In 2010, for example, the Court recognized that “a No. 23-3840 United States v. Sittenfeld Page 70

substantial and legitimate reason, if not the only reason, to cast a vote for, or to make a contribution to, one candidate over another is that the candidate will respond by producing those political outcomes the supporter favors.” Citizens United, 558 U.S. at 359 (citation omitted). In 2014, Chief Justice Roberts, for a plurality of the Court, wrote “[r]epresentatives are not to follow constituent orders, but can be expected to be cognizant of and responsive to those concerns. Such responsiveness is key to the very concept of self-governance through elected officials.” McCutcheon, 572 U.S. at 227. The Chief Justice also wrote that “[t]here is no right more basic in our democracy than the right to participate in electing our political leaders” and among the methods of participating are voting and contributing to political campaigns. Id. at 191. The Court adopted much of the McCutcheon plurality’s reasoning in Cruz, 596 U.S. at 289.

These cases that post-date McCormick and Evans reflect the strength with which the Constitution guarantees the expression of speech through contributions to a campaign. These precedents “protect[] contributions because contributing is a form of political participation.” Hellman, at 1987. Because that is the case, politicians and donors deserve more protection from prosecutors who seek to reach a jury merely by showing a donor’s campaign contribution on one side and that donor’s “self-interest reflected in a candidate’s commitment” on the other. See Brown, 456 U.S. at 56.

IV. Conclusion

This case is unusual, but it is by no means an aberration. If Sittenfeld’s verdict is allowed to stand, it may incentivize more prosecutions of campaign contributions in the future. The costs of not policing the sufficiency line may prove grievous. Writing for a unanimous court, the Chief Justice addressed the threat of overzealous corruption prosecutions:

The basic compact underlying representative government assumes that public officials will hear from their constituents and act appropriately on their concerns—whether it is the union official worried about a plant closing or the homeowners who wonder why it took five days to restore power to their neighborhood after a storm. The Government's position could cast a pall of potential prosecution over these relationships if the union had given a campaign contribution in the past or the homeowners invited the official to join them on their annual outing to the ball game. Officials might wonder whether they could No. 23-3840 United States v. Sittenfeld Page 71

respond to even the most commonplace requests for assistance, and citizens with legitimate concerns might shrink from participating in democratic discourse.

McDonnell v. United States, 579 U.S. 550, 575 (2016) (emphasis omitted).

The majority decides that the best way to handle the task facing us—one with no precedent directly on point, a high chance of jury confusion as to what the First Amendment protects, and a high cost if the jury gets it wrong—is to do nothing. This should not come as a surprise: no caselaw has grappled with circumstances as close to the line as these, and sufficiency challenges are almost always losers to begin with. Faced with the very real possibility that Sittenfeld assented to only a “good investment” in his candidacy rather than an agreement that “control[ed]” his actions, the majority responds that “[c]ampaign contributions will almost always have an inherent legitimate alternative explanation” and that “[t]his is also why matters of intent are for the jury to consider.” Majority Op. at 16 (citation omitted).

But this is exactly the wrong solution to the problem that the majority correctly identifies. That a jury could interpret a raft of legitimate actions against an official suggests a high likelihood that an innocent politician could be convicted. This should count as less reason, not more, to give such an issue to a jury because of the constitutional concerns at stake. Judges, not juries, decide when conduct falls within the protection of the Constitution. See Sparf v. United States, 156 U.S. 51, 87 (1895) (“[T]he law in criminal cases is to be determined by the court. In this way we have our liberties and rights determined . . . not by a tribunal ignorant of the law, but by a tribunal trained to and disciplined by the law . . . .”). Consequently, judges bear the responsibility to overturn a verdict when, as in Sittenfeld’s prosecution, the risk of conviction for constitutionally protected conduct is too great.

Because I believe the government did not present evidence sufficient to convict Sittenfeld of federal programs bribery or Hobbs Act extortion, I respectfully dissent.