United States v. John Thomas Burnette

65 F.4th 591

Opinion

[PUBLISH]

In the

United States Court of Appeals For the Eleventh Circuit

____________________

No. 21-13990

____________________

UNITED STATES OF AMERICA, Plaintiff-Appellee,

versus JOHN THOMAS BURNETTE,

Defendant-Appellant.

____________________

Appeal from the United States District Court for the Northern District of Florida D.C. Docket No. 4:18-cr-00076-RH-EMT-3

____________________

2

2 Opinion of the Court 21-13990

Before JORDAN, ROSENBAUM, and NEWSOM, Circuit Judges. NEWSOM, Circuit Judge:

Real-estate developer John Burnette was convicted on multiple counts arising out of his alleged complicity in the bribery of Tallahassee City Commissioner Scott Maddox. On appeal, Burnette challenges his bribery-based convictions on several grounds, two of which require us to carefully examine the Supreme Court’s decision in McDonnell v. United States, 579 U.S. 550 (2016), which explained—and by all accounts narrowed to some degree—the category of “official acts” that can support a federal bribery charge. Burnette separately contests his conviction for making false statements to federal agents during the course of their investigation.

After careful consideration of Burnette’s McDonnell-related arguments, his challenges to two evidentiary rulings, and his attack on his false-statements conviction, we affirm.

I

John Burnette controlled a substantial real-estate syndicate in Tallahassee, Florida. In the course of his business, he became “friend[s]” with Tallahassee City Commissioner Scott Maddox. Doc. 461 at 35. In 2015, the FBI initiated an undercover operation to investigate public corruption in Tallahassee. Two agents created a fictitious company called Southern Pines and posed as a property developer, Michael Miller, and an investor, Michael Sweet. Doc. 456 at 141–43, 161. “Miller” and “Sweet” befriended Burnette and, over the course of several months, engaged in

3

21-13990 Opinion of the Court 3

discussions about development opportunities with him and Maddox —many of which the agents secretly recorded. Doc. 456 at 144–45; Doc. 440-2 passim.

Burnette, Miller, and Sweet together pinpointed two projects for further consideration. First, they would encourage Tallahassee officials to “annex” a parcel of land called Fallschase, which was situated just outside the city limits, in order to increase its value. Doc. 456 at 173–74 (Miller); Doc. 440-2 at 47–49 (Burnette). Second, they would aim to convince officials to approve a Request for Proposal authorizing the city to invite potential developers (like themselves) to bid for a city-owned property called Myers Park. Doc. 456 at 194–95 (Miller); Doc. 459 at 178–79 (Sweet); Doc. 457 at 14–15 (Miller); Doc. 440-2 at 7–9 (Burnette).

In recorded conversations in July and September 2016, Burnette instructed Miller and Sweet that they would need to pay Maddox for his votes on the Fallschase and Myers Park projects because he was “very transactional” and wanted his “piece of pie.” Doc. 440-2 at 6, 21. In the September conversation, Burnette told the agents that while they might be able to persuade the other commissioners “on the merits,” Maddox could “convince[]” his colleagues if the agents paid “$10,000 a month the next 3 years for [Maddox] to lobby” on their behalf. Id. at 46–48. Although at one point Burnette counseled Miller and Sweet to wait because he “hate[d] to see” them “spend money and not know exactly what [they were] doing,” he emphasized that it was “money well spent” if they were going to “do a deal here.” Id. at 121.

4

4 Opinion of the Court 21-13990

Maddox subsequently met with Miller and Sweet and agreed to “run interference” and help them with “whatever [they] needed”—so long as (1) Burnette remained “involved” and (2) they paid $10,000 a month to Governance Services, a company run by Maddox’s girlfriend, Paige Carter-Smith. Doc. 459 at 163–64; see also Doc. 440-2 at 99–102 (Maddox instructing Sweet to pay Governance “so I would not be conflicted out if you had shit coming up in front of me” and assuring Sweet that “J.T. [i.e., Burnette] will tell you who [Governance] is”); Doc. 453 at 246–50 (Carter-Smith testifying that she found it “very curious that [she] was getting paid and [she] was not being asked to do anything”). Burnette echoed Maddox’s request that Miller and Sweet “run [payments] through Governance.” Doc. 440-2 at 125–27; see also id. at 121–22 (same). In November 2016, consistent with Maddox’s instructions, the agents sent a $10,000 check to Governance, which Carter-Smith received . Doc. 453 at 228–30. Maddox told Burnette about the payment . Doc. 461 at 77–78.

Miller and Sweet arranged a trip to Las Vegas for themselves , Burnette, and Maddox in early December 2016. Doc. 460, 127–28. While in Vegas, the four discussed both Fallschase and Myers Park. In particular, Sweet and Burnette told Maddox (1) that they wanted to move forward with the annexation of Fallschase, Doc. 440-2 at 151–52, and (2) that they wanted Maddox to “throttle ”—i.e., slow-roll—the Myers Park project “until the time it’s appropriate for [them] to move on it,” id. at 158. Burnette told Miller and Sweet that Maddox would help to ensure that the city annexed

5

21-13990 Opinion of the Court 5

Fallschase and delayed the Myers Park RFP. Id. at 151, 155–58. One other Vegas-related incident bears brief mention here: Evidence in the record indicates that during the trip, Sweet bought Maddox either a private dance or oral sex (or perhaps both) at a strip club. As we’ll explain in due course, the district court’s decision to exclude some of that evidence forms the basis for one of Burnette’s challenges.

Following the Las Vegas trip, the agents sent two more $10,000 checks to Governance—one in mid-December and another in late January. Doc. 453 at 237, 239. To be sure, Burnette occasionally sent Miller and Sweet mixed messages about the payments . He twice insinuated, for instance, that he didn’t “want [Sweet] to think that [he] can effectively pay these people and get a[ ] vote” and that, if he did, Maddox would “just recuse himself, and [not] vote.” Doc. 440-2 at 172; see also Doc. 440-17 at 48–49 (similar). And Burnette emphasized his above-board wins with the commission, once telling Sweet: “5-0 vote, did not pay a $. Tallahassee is just about doing the right thing.” Doc. 460 at 30–31. At trial, Sweet testified—over objection—that he considered Burnette ’s comments to be “false exculpatory” statements. Doc. 460 at 5.

All the while, though, Burnette reiterated to Miller and Sweet that Maddox would move Fallschase through the city commission in exchange for their money and, in fact, warned the agents not to stop sending checks, for fear that Maddox—whom Burnette

6

6 Opinion of the Court 21-13990

called “god damn mafia” and “a revengeful mother fucker,” Doc. 440-2 at 193, 211—might engage in retribution. For instance—

Sweet: Are you sending the checks to Maddox?

Miller: Yeah. Sending what you told me to send.

Burnette: Let me tell you this, don’t stop that.

. . . It’ll get done. [Maddox will] get it run through [the city manager] . . . . It’ll be a 3-1 vote.

Id. at 192; see also, e.g., id. at 193 (Sweet: “Well at this point we’ve put him in a paycheck.” . . . Burnette: “You can’t take him out.”); id. at 196 (Burnette on January 9, 2017: “$10,000 a month, and it’s all going to be okay . . . [Maddox] isn’t going to vote, but he’s going to make sure that the votes are enough.”); id. at 198 (Sweet: “So . . . keep Maddox on the payroll.” Burnette: “Yeah.”); id. at 203 (Burnette on March 13, 2017: “I would continue to pay [Carter-Smith] the ten thousand dollars . . . .”); Doc. 461 at 87 (Maddox: “I told [Carter-Smith] that I thought they were going to move forward with Fallschase.”).

During a meeting in late March 2017, Maddox and Carter-

Smith both made statements to Miller and Sweet insisting that Maddox was only acting in the city’s best interests and that they considered the agents’ payments to be for legitimate purposes. Miller testified at trial that he thought Maddox was “tr[ying] to change” the “way he presented Governance . . . like it was an arm’s

7

21-13990 Opinion of the Court 7

length lobbying firm,” Doc. 457 at 59, and Sweet explained his view that Maddox “was really making false exculpatory statements claiming that he was just doing whatever was best for the city,” Doc. 460 at 29. At that point, Miller and Sweet decided to terminate the undercover operation and shifted to an overt investigation . Id. In May 2017, two different FBI agents interviewed Burnette about his dealings with Maddox, Miller, and Sweet. Doc. 440- 2 at 223–25.

A federal grand jury indicted Burnette, Maddox, and Carter-

Smith for their roles in the alleged bribery scheme. Maddox and Carter-Smith pleaded guilty; Burnette didn’t. Following a 17-day trial, a jury convicted Burnette on five counts: one of Hobbs Act extortion, 18 U.S.C. § 1951(a); two of honest-services mail fraud, 18 U.S.C. §§ 1341, 1346; one of using a facility of interstate commerce to facilitate unlawful activity, 18 U.S.C. §§ 1952(a)(3); and one of making a material false statement to the FBI, 18 U.S.C. § 1001(a)(2). The district court imposed a below-guidelines sentence of 36 months’ imprisonment followed by one year of supervised release.

This is Burnette’s appeal. He challenges (1) both the district court’s jury instructions and the sufficiency of the evidence related to his extortion and honest-services fraud convictions, (2) two of the district court’s evidentiary rulings, and (3) the sufficiency of the evidence underlying his false-statements conviction. We will consider Burnette’s arguments in turn.

8

8 Opinion of the Court 21-13990

II

Two of Burnette’s challenges—one pertaining to the district court’s jury instructions and another contesting the sufficiency of the evidence—turn on a definitional provision in the federal bribery statute, see 18 U.S.C. § 201, and, in particular, on the Supreme Court’s interpretation of that statute in McDonnell v. United States, 579 U.S. 550 (2016). Although Burnette wasn’t charged with bribery per se, all here agree—as they did at trial—that Burnette’s alleged attempt to bribe Maddox underlies the Hobbs Act extortion and honest-services-fraud charges, and that, for those purposes, § 201 provides the relevant definition of “bribery.” See Skilling v. United States, 561 U.S. 358, 404 (2010) (defining honest-services fraud to include bribery); Evans v. United States, 504 U.S. 255, 260, 269 (1992) (defining Hobbs Act extortion to include bribery). Before addressing Burnette’s particular arguments, we begin with a primer on § 201 and the Supreme Court’s decision in McDonnell.

In relevant part, the federal bribery statute’s operative provision makes it unlawful for anyone to “corruptly give[], offer[] or promise[] anything of value to any public official . . . with intent . . . to influence any official act.” 18 U.S.C. § 201(b)(1)(A). Importantly here, the statute defines the term “official act” as follows:

[T]he term “official act” means any decision or action on any question, matter, cause, suit, proceeding or controversy, which may at any time be pending, or which may by law be brought before any public

9

21-13990 Opinion of the Court 9

official, in such official’s official capacity, or in such official’s place of trust or profit.

Id. § 201(a)(3).

In McDonnell, the Supreme Court clarified—and narrowed —the meaning of the term “official act.” The defendant in that case was the former governor of Virginia, who (like Burnette) had been convicted of bribery-related Hobbs Act extortion and honest-services fraud—there, for accepting some $175,000 in gifts, loans, and other benefits from a local businessman in exchange for his promise to arrange meetings, organize events, and contact other government officials about his donor’s nutritional-supplement product. See 579 U.S. at 556–61.

Unpacking the statute’s language, the Supreme Court emphasized that “[t]he text of § 201(a)(3) sets forth two requirements for an ‘official act’”:

First, the Government must identify a “question, matter, cause, suit, proceeding or controversy” that “may at any time be pending” or “may by law be brought” before a public official. Second, the Government must prove that the public official made a decision or took an action “on” that question, matter, cause, suit, proceeding, or controversy, or agreed to do so.

Id. at 567 (quoting 18 U.S.C. § 201(a)(3)). The first of § 201(a)(3)’s two “requirements” pertains to what we’ll call the “matter” and the second to what we’ll call the “act.”

10

10 Opinion of the Court 21-13990

The Court went on to detail the characteristics of the covered matters and acts. With respect to matters, it made two important observations. First, it rejected the government’s contention that “nearly any activity by a public official qualifies.” Id. at 567–68. Rather, employing the familiar noscitur a sociis canon of construction, the Court interpreted the general terms “question [and] matter” by reference to the more specific terms “cause, suit, proceeding [and] controversy.” Id. at 568–69. Because “a typical meeting, call, or event arranged by a public official is not of the same stripe as”—i.e., is not as serious as—“a lawsuit before a court, a determination before an agency, or a hearing before a committee ,” the Court held those sorts of occurrences didn’t constitute covered “question[s or] matter[s].” Id. at 569.

Second, and separately, the Court held that a covered matter can’t be framed at too high a “level of generality”—like, in that case, “[e]conomic development.” Id. Rather, it must be more specific —i.e., “focused and concrete.” Id. at 570. In particular, the Court emphasized § 201(a)(3)’s condition that a matter be either “pending” or the sort of thing that “may by law be brought” before a “public official.” Id. (quoting 18 U.S.C. § 201(a)(3)). That language , the Court said, “suggest[s] something that is relatively circumscribed —the kind of thing that can be put on an agenda, tracked for progress, and then checked off as complete.” Id.

The McDonnell Court separately explained—albeit more briefly—what we have called § 201(a)(3)’s act requirement. Recall that § 201(a)(3) refers to “any decision or action on” a covered

11

21-13990 Opinion of the Court 11

matter. Although, in the abstract, the terms “decision” and “action ” could be read broadly to capture ordinary steps like “setting up a meeting, hosting an event, or calling another official,” the Court held that under its precedents, “something more is required .” Id. at 571–72. As examples of qualifying acts, the Court used illustrations relevant to the case before it—“[f]or example, a decision or action to initiate a research study[,] or a decision or action on a qualifying step, such as narrowing down the list of potential research topics.” Id. at 572. Some acts even more clearly qualify ; for instance, although not squarely at issue in McDonnell, no one disputes (or could) that casting or abstaining from a vote on a covered matter, or agreeing to do either, would constitute the sort of act that triggers § 201’s prohibition. See, e.g., United States v. Roberson, 998 F.3d 1237, 1251–52 (11th Cir. 2021).

* * *

In sum, then, McDonnell clarified that in order to implicate the bribery statute’s prohibition, a public official must either engage or agree to engage in (1) a sufficiently serious act—casting a vote being the quintessential example—(2) concerning a sufficiently serious and “concrete” matter.

A

Having set the table, we turn to Burnette’s McDonnell-

based arguments. We will first address his challenges to the district court’s jury instructions, which he contends violated McDonnell in two respects, and then turn to his assertion that the government

12

12 Opinion of the Court 21-13990

failed to present sufficient evidence to sustain his convictions on the bribery-related counts.

1

Burnette raises two challenges to the district court’s jury instruction defining the term “official act”—one general, the other more specific. First, and more generally, Burnette argues that the instruction failed to give the jury any meaningful guidance about what constitutes a § 201(a)(3)-qualifying matter. Second, and more specifically, he contends that the instruction elided the requirement —which he grounds in McDonnell and its progeny—that the government prove that Maddox agreed, in exchange for payment, to provide assistance with “concrete,” identified matters. McDonnell , 579 U.S. at 570. After setting out the official-act instruction’s language, we will consider Burnette’s arguments in turn.

In pertinent part, the district court’s instruction provided as follows:

An “official act” is a decision or action that involves the formal exercise of governmental power. This case involves only three kinds of possible official acts. Unless you find there was an official act of one of these three kinds, you cannot find there was an official act.

The first is a vote on any matter that was pending or might later come before a local governmental entity, including the City Commission, the Community Redevelopment Agency, or the Planning

13

21-13990 Opinion of the Court 13

Commission. To count as a matter that might later come before the governmental entity, a matter need not be identified at the time of the payment; it is sufficient if the payment is made in exchange for favorable treatment on any not-yet-known matter that might later come up for a vote.

The second kind of official act is abstaining from a vote—that is, not voting—on the same kind of matter .

The third kind of official act is pressuring or advising another official, including another commissioner or a staff member, on the same kind of matter. But Mr. Maddox’s advice to a commissioner or staff member was an official act only if Mr. Maddox knew or intended that the commissioner or staff member would take formal action based on that advice on a matter that could come before the local governmental entity . Talking to another commissioner or staff member or advocating a course of action, without more, is not an official act.

* * *

Before diving in, we note two hurdles that Burnette must overcome on appeal—the doctrines of “invited error” and “plain error.” As to the former, we have held that “[w]hen a party agrees with a court’s proposed instructions, the doctrine of invited error applies.” United States v. Frank, 599 F.3d 1221, 1240 (11th Cir. 2010). And “[i]t is a cardinal rule of appellate review that a party

14

14 Opinion of the Court 21-13990

may not challenge as error a ruling” that he invited. United States v. Love, 449 F.3d 1154, 1157 (11th Cir. 2006) (quoting United States v. Ross, 131 F.3d 970, 988 (11th Cir 1997)). All here agree that Burnette ’s trial counsel invited error regarding some aspect of the court’s official-act instruction; the only dispute, as we will explain, is how far the invited-error bar extends. See Oral Arg. at 1:40–2:40. (Burnette’s appellate counsel conceding as much).

Second, invited error aside, it is undisputed here that Burnette ’s trial counsel didn’t affirmatively object to the court’s instruction on either of the grounds that he argues on appeal. Accordingly , even with respect to any aspects of the instruction that aren’t covered by the invited-error bar, we may review only for plain error. To establish plain error, Burnette must show that “(1) an error occurred; (2) the error was plain; (3) it affected his substantial rights; and (4) it seriously affected the fairness of the judicial proceedings.” United States v. Ramirez-Flores, 743 F.3d 816, 822 (11th Cir. 2014).

a

Burnette first (and more generally) contends that the district court’s official-act instruction failed to narrow or adequately explain § 201(a)(3)’s “matter” requirement. The instruction, he complains , referred to “matter[s]” only in the starkest and most indefinite way: (1) It adverted to “any matter” that “was pending or might later come before a local government entity”; (2) it said that “a matter” needn’t be identified at the time of payment but, rather, that it was enough that a payment be made in exchange for

15

21-13990 Opinion of the Court 15

favorable treatment on “any not-yet-known matter” that might later come up for a vote; (3) and with respect to any theory of bribery that involved one public official seeking to influence another, the instruction said only that the former must know or intend that the latter would heed his advice on “a matter” that could later arise. What the instruction didn’t do, Burnette insists, is capture McDonnell ’s requirements that a § 201(a)(3)-qualifying matter be both sufficiently serious and sufficiently specific. With respect to the former , he complains, the instruction didn’t clarify that a covered matter must be “of the same stripe as a lawsuit before a court, a determination before an agency, or a hearing before a committee.” McDonnell, 579 U.S. at 569. And with respect to the latter, it didn’t say that a covered matter must be “relatively circumscribed—the kind of thing that can be put on an agenda, tracked for progress, and then checked off as complete,” advise the jury that a covered matter must be “focused and concrete,” or otherwise warn against defining the matter at too high a “level of generality.” Id. at 570. For all intents and purposes, Burnette asserts, the district court’s official-act instruction left it to the jury to define the term “matter” however it saw fit.

We hold that, at least in part, Burnette invited the error that he now seeks to challenge. Here’s why: At trial, the government proposed an official-act charge that tracked the Eleventh Circuit pattern jury instruction. The pattern instruction defines the term “official act” by express reference to the three illustrative examples mentioned in McDonnell itself—i.e., something similar to “a

16

16 Opinion of the Court 21-13990

lawsuit before a court, a determination before an agency, or a hearing before a committee.” See 579 U.S. at 569. The district court refused the government’s request. The court explained its decision this way:

Judges writing opinions are explaining results in a case. They are not always trying to express things in language that works in jury instructions. I get it that it’s generally safe to take the language from the Supreme Court and put it right in the jury instruction. So I understand why the government’s experience is, we do better on appeal if we just took the language right out of [McDonnell]. So part of the language you have is: It must be similar in nature to a lawsuit before a court, a determination before an agency, or a hearing before a committee. That’s clear as mud. I mean, it’s clear to a lawyer that reads all of these cases and figures out what kind of thing we're talking about.

I think what the Supreme Court was trying to do was to tell the Justice Department in bringing prosecutions and judges in ruling on—lower-court judges in ruling on cases what kind of thing we’re talking about, so that the government would know whether to charge something and courts would know how to handle the case. I don’t think anybody on the Supreme Court thought they were writing a standard jury instruction. So what I did in the case was take that language—like I said, I wrote this instruction

17

21-13990 Opinion of the Court 17

with [McDonnell] open on the computer, but I tried to apply it to the facts of this case.

So, “similar to a hearing before a committee,” I don’t know if any jurors have ever been to a hearing before a committee or knows what kind of thing comes before a committee. So, I hear you. And one of the judges of this district used to say he's never been reversed for giving a standard instruction. If my main goal in the case was not to be reversed, I would just give the standard instructions in the case.

Doc. 463 at 43–44 (emphasis added).

The government’s lawyer replied that he understood. Importantly , the court then made pointed inquiries of both sides:

Court: But my instruction doesn’t leave anything out, right?

[Government]: No, no. I think the court’s instruction is accurate, without a doubt.

Court: You would like it the way it is on the defense side, I take it. You didn’t object to the way I had it.

[Defense]: No, no. I think, Judge, that you captured it. Looking at [McDonnell ], I understand what the court did.

18

18 Opinion of the Court 21-13990

Id. at 44 (emphasis added).

To repeat, we have held that “[w]hen a party agrees with a court’s proposed instructions, the doctrine of invited error applies.” Frank, 599 F.3d at 1240. But we have also traditionally construed invited errors narrowly, so as to preserve the opportunity for appellate review in close cases. See, e.g., Cherry v. Dometic Corp., 986 F.3d 1296, 1301 (11th Cir. 2021) (holding that the invited-error bar “is triggered only by unambiguous statements or representations ”). And to that end, we have drawn some pretty fine lines distinguishing between invited and merely-unobjected-to errors in jury instructions. Compare, e.g., United States v. Silvestri, 409 F.3d 1311, 1337 (11th Cir. 2005) (holding that defense counsel’s statement that a district court’s jury instructions “covered the bases” triggered the invited-error rule), with, e.g., United States v. Dortch, 696 F.3d 1104, 1112 (11th Cir. 2012) (finding no invited error and reviewing for plain error where counsel said that he “d[id]n’t think” he needed to review the instructions again with the court), overruled in part on other grounds by Alleyne v. United States, 570 U.S. 99 (2013).

Under our precedent, Burnette’s trial counsel’s concluding statement—“[Y]ou captured it . . . . Looking at [McDonnell], I understand what the court did.”—clearly constitutes an “agree[ment]” to something, and thus clearly invited error with respect to something. What, though, is that something? The thrust of the colloquy between the judge and the lawyers was about whether the official-act instruction should include McDonnell’s

19

21-13990 Opinion of the Court 19

analogy to lawsuits, determinations, and hearings—which, as we’ve explained, bears on the required seriousness of a § 201(a)(3)- qualifying matter. Because Burnette’s lawyer expressly “agree[d]” that the official-act instruction didn’t need to include the analogy, we hold that any error that the court might have committed in refusing to include it was invited, and that we are accordingly powerless to review it. In keeping with our precedent, though, we will construe the invited error as extending only that far. We will consider the balance of Burnette’s contention that the instruction failed to adequately define covered “matter[s]”—namely, his argument that it failed to convey that a matter must be sufficiently specific and “concrete”—in connection with his second challenge.

We proceed, then, to that issue.

b

Burnette’s second challenge focuses on the following language in the district court’s instruction: “To count as a matter that might later come before the governmental entity, a matter need not be identified at the time of the payment; it is sufficient if the payment is made in exchange for favorable treatment on any not- yet-known matter that might later come up for a vote.” Doc. 433 at 7–8 (emphasis added). By so charging the jury, Burnette contends , the court impermissibly relieved the government of its burden to prove that the transacting parties themselves identified and agreed on a specific matter (or matters) on which Maddox would provide his assistance in exchange for payment.

20

20 Opinion of the Court 21-13990

As an initial matter, we don’t think that Burnette’s trial counsel affirmatively invited that alleged error. Whereas the lawsuit -determination-hearing analogy that Burnette’s lawyer gave away pertains to the seriousness of a § 201(a)(3)-qualifying matter, Burnette’s second challenge runs on the matter’s requisite specificity —i.e., whether it was sufficiently “focused and concrete,” 579 U.S. at 570—and the timing of its identification. In short, it tees up a different McDonnell-related issue. Be that as it may, it is undisputed that Burnette failed to object to the court’s official-act instruction on the ground that he now presses on appeal. Accordingly , we review only for plain error—which, again, requires proof that “(1) an error occurred; (2) the error was plain; (3) it affected [Burnette’s] substantial rights; and (4) it seriously affected the fairness of the judicial proceedings.” Ramirez-Flores, 743 F.3d at 822.

On the merits, the parties vigorously dispute McDonnell’s meaning, as well as its implications for this case. For his part, Burnette emphasizes the Supreme Court’s own statements about § 201(a)(3)-qualifying matters and their identification. With respect to the nature of a covered matter itself, Burnette points to the Court’s requirement that it be specific—i.e., “focused and concrete .” 579 U.S. at 570. And with respect to identification, he points out that the Court observed both (1) that “the Government must identify” the covered matter and (2) that “the Government must prove”—as relevant here—that “the public official” “agreed” to “t[ake] an action ‘on’ that” matter. Id. at 567. The upshot, he contends, is clear: In order to secure a bribery-related conviction,

21

21-13990 Opinion of the Court 21

the government must establish that the transacting parties agreed among themselves that, in exchange for payment, a public official would provide assistance on a concrete, identified matter. The district court’s official-act instruction in this case, he says, failed to communicate that burden, because it told jurors (1) that the § 201(a)(3)-covered matter “need not be identified at the time of payment” and (2) that it was “sufficient” that they find that a payment was made to Maddox “in exchange for favorable treatment on any not-yet-known matter that might later come up for a vote.” The instruction, Burnette complains, left the issue of the matter’s identification completely up in the air, impermissibly communicating to the jurors that they could convict him even if they concluded that Maddox never agreed to assist him, Miller, and Sweet with any specific, identified matter.

The government offers several arguments in response.

First, it asserts that “a legislator’s promise to vote on future not- yet-known legislative bills” in a way that favors his benefactor “will always constitute a promise to undertake an official act” because “[w]hatever the eventual subject of those bills, the legislator’s vote” will necessarily concern a sufficiently concrete matter. Br. of Appellee at 25–26. Thus, the argument goes, because the district court’s instruction here properly adverted to McDonnell-covered acts—e.g., “vote[s]”—it must also implicitly have narrowed the range of covered matters. Second, the government insists that nothing in either McDonnell or our follow-on decision in Roberson necessarily requires it to prove that the parties themselves

22

22 Opinion of the Court 21-13990

identified a particular matter at the time of their agreement or precludes it from “identify[ing]” the pertinent matter after the fact, so to speak, “at trial.” Br. of Appellee at 28 (quoting United States v. Van Buren, 940 F.3d 1192, 1204 (11th Cir. 2019)). Finally, and perhaps most significantly, the government emphasizes that it “relied on an ‘as-the-opportunities-arise’ theory of bribery in this case”— which, it explains, “occurs when a person bribes an individual or entity in exchange for a continuing course of conduct.” Br. of Appellee at 18 (quoting Roberson, 998 F.3d at 1245 n.10). And the government notes, correctly, that we observed in Roberson that McDonnell “did not reject the retainer theory of bribery,” 998 F.3d at 1251—a fraternal (if not quite identical) twin of the as-the-opportunities -arise theory.

As the parties’ competing contentions reflect, there is a lot to be said about the meaning and import of the Supreme Court’s decision in McDonnell. And there is, we confess, some temptation to attempt an overarching explanation of how McDonnell applies in cases like this one and, in particular, what it might portend for the traditional retainer and as-the-opportunities-arise theories of bribery. But the passive virtues are virtues for a reason. We conclude that we needn’t definitively decide at Step 1 of the plain-error standard whether the district court’s instruction impermissibly relieved the government of its burden to prove that Maddox agreed to assist Burnette with particular, identified matters, or even, at Step 2, whether any such error was “plain”—because we hold that

23

21-13990 Opinion of the Court 23

any error that occurred didn’t affect Burnette’s substantial rights at Step 3.

“To show that an instructional error affected his substantial rights” at Step 3 of the plain-error analysis, “a defendant must show that the error ‘was probably responsible for an incorrect verdict.’” United States v. Iriele, 977 F.3d 1155, 1179 (11th Cir. 2020) (quoting United States v. Whyte, 928 F.3d 1317, 1332 (11th Cir. 2019)). Put slightly differently, the defendant must establish “a reasonable probability of a different result but for the error.” Id.

Burnette hasn’t met that standard. For reasons that we will detail more thoroughly in the next section, addressing Burnette’s sufficiency-of-the-evidence challenge, the proof at trial convincingly demonstrated that Burnette facilitated the $10,000 payments to Maddox in November 2016, January 2017, and February 2017. See Doc. 453 at 228–30, 239, 242. Importantly, the evidence also demonstrated that Maddox understood at the time that those payments were made in exchange for his assistance with two specific development projects—Fallschase and Myers Park. See, e.g., Doc. 461 at 77–78 (Maddox testifying that he told Burnette about the November check to Governance); Doc. 440-2 at 151–52, 155–58 (Maddox and Burnette discussing Fallschase with the agents in Las Vegas after the November check was paid); id. at 192 (Burnette confirming that, in exchange for the checks, “It’ll be a 3-1 vote” to annex Fallschase); id. at 196 (Burnette stating that it’s “$10,000 a month, and it’s all going to be okay” and that Maddox “isn’t going to vote, but he’s going to make sure that the votes are enough”).

24

24 Opinion of the Court 21-13990

No one disputes that the Fallschase and Myers Park projects constitute concrete, identifiable “matters” within the meaning of § 201(a)(3) as interpreted in McDonnell. Accordingly, even if the district court had instructed the jury as Burnette says it should have—as relevant here, by clarifying that the government had to prove that the parties identified and agreed on the particular matter (s) that Maddox would seek to influence—it is overwhelmingly likely that the jury would have convicted Burnette anyway. And where, as here, “the defendant’s guilt would have been clear under the correct instruction, he loses under the substantial rights third prong of plain error review.” Iriele, 977 F.3d at 1179.

* * *

For all these reasons, we find ourselves constrained to reject Burnette’s challenges to the district court’s official-act instruction, though for reasons having little to do with their merits (or demerits ). First, to the extent that Burnette now contends that the court erred in declining to include McDonnell’s lawsuit-determinationhearing analogy as a means of explaining that a § 201(a)(3) matter must be sufficiently serious, we hold that his trial counsel invited the error. Second, and separately, we hold that we needn’t decide whether the court should have more clearly instructed the jury that the government had to prove that the parties agreed, in advance, that Maddox would provide assistance on concrete, identified matters because, in any event, its failure to do so didn’t affect Burnette ’s substantial rights.

25

21-13990 Opinion of the Court 25

B

Burnette separately—but relatedly—contends that he was entitled to a judgment of acquittal on the bribery-related counts because the government failed to prove that Maddox agreed to assist Burnette with a particular “matter,” as McDonnell defined (and as we have explained) that term. We disagree.

In considering sufficiency-of-the-evidence claims like Burnette ’s, we employ an indulgent standard. “[W]e view the evidence in the light most favorable to the prosecution and draw ‘all reasonable inferences and credibility choices’ in its favor.” United States v. Fleury, 20 F.4th 1353, 1367 (11th Cir. 2021) (citations omitted ). The evidence need not “be inconsistent with ‘every reasonable hypothesis except guilt.’” Id. (citations omitted). Rather, it is enough that “any rational trier of fact could have found the essential element of the crime beyond a reasonable doubt.” Id. (quoting Jackson v. Virginia, 443 U.S. 307, 319 (1979)). Accordingly, a guilty verdict need only “be reasonable, not inevitable, based on the evidence presented at trial.” United States v. Browne, 505 F.3d 1229, 1253 (11th Cir. 2007).

Under that standard, the evidence was sufficient to permit a reasonable jury to conclude that Maddox agreed to perform an official act in connection with a McDonnell-qualifying matter. Burnette makes two arguments to resist that conclusion. Neither convinces us.

26

26 Opinion of the Court 21-13990

First, he cites snippets of trial testimony that he claims show that the undercover agents purported only “to buy generic access ”—an “arrangement with Maddox divorced from any particular matter.” Br. of Appellant at 24–25. But the government presented ample evidence from which a reasonable jury could have found, to the contrary, that Burnette met with both Maddox and the agents to arrange for a series of $10,000 payments to Maddox’s girlfriend’s company in order to secure Maddox’s vote (or abstention ) in connection with the Fallschase and Myers Park projects. See, e.g., supra at 23 (collecting record citations supporting the three payments); Doc. 440-2 at 87–89 (Sweet asking how to “fill Maddox’s coffers” and Burnette answering that “Maddox has a consulting [firm]”); id. at 99–100 (Maddox discussing Myers Park with Sweet on October 4, 2016 while mentioning “somebody [Sweet] can hire” that was “not [Maddox]”); id. at 110 (Maddox: “[T]here’s more than that just one deal. There’s [sic] two or three deals here that make total sense.”); id. at 151–59 (Maddox, Burnette, and Sweet discussing Fallschase and Myers Park in December 2016, after the first payment); id. at 51 (Burnette and Sweet discussing sealing the real-estate deal because “the check’s already been written in the back door to Scott Maddox”); id. at 121–22 (Burnette: “[I]t’s definitely for Maddox, there’s nobody else in Governance other than Paige [Carter-Smith], which is Maddox effectively, indirectly .”); id. at 125–27 (Sweet: “[D]oes he not want paid?” Burnette : “No, no, he does. He wants to get paid. . . . Mike and I talked about that and run that through Governance.”); Doc. 461 at 87

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 27 of 51

21-13990 Opinion of the Court 27

(Maddox: “I told [Carter-Smith] that I thought they were going to move forward with Fallschase.”).

Second, and seemingly in acknowledgment of the fact that the record contains evidence regarding Fallschase and Myers Park, Burnette contends that “[t]o the extent the agents discussed particular projects with Burnette or Maddox, those projects did not require Maddox’s help.” Br. of Appellant at 25. Two problems. For one, in the same way that bribery liability doesn’t depend on whether a public official ultimately performed an official act, cf. Evans , 504 U.S. at 268, it doesn’t turn on the bribe’s actual or expected effectiveness. All that matters is that, in exchange for something of value, the official agreed to perform an act concerning a sufficiently serious and concrete matter. See McDonnell, 579 U.S. at 572; cf. United States v. Kimbrew, 944 F.3d 810, 815–16 (9th Cir. 2019) (“In short, execution is immaterial. It logically follows, then, that § 201 liability is not limited by the odds of success of the quo at issue.”). 1 For another, there was ample evidence in the record from which a reasonable jury could have concluded that neither the annexation of Fallschase nor the “throttl[ing]” of the Myers Park RFP was quite the fait accompli that Burnette now suggests—or, at the very least,

1 This case is the flip side of Kimbrew. There, the court rejected a defendant’s contention that he couldn’t be held liable under McDonnell where there was no evidence that he could have actually achieved the task that he allegedly was bribed to do. See 944 F.3d at 814–16. By parity of reasoning, Burnette can’t avoid liability by arguing that there was no evidence that Maddox was needed for the task that Burnette facilitated bribing him to undertake.

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 28 of 51

28 Opinion of the Court 21-13990

that the payments to Maddox provided some assurance that he would undertake official acts in connection with those matters. See, e.g., Doc. 456 at 195 (Miller: “[Burnette is] explaining to us that Scott Maddox is one of those votes and essentially controls two additional votes.”); Doc. 440-2 at 192 (Burnette confirming that, in exchange for the checks, “It’ll be a 3-1 vote”); id. at 196 (“[Maddox is] going to make sure that the votes are enough.”); id. at 151–59 (Maddox, Burnette, and Sweet discussing Fallschase and Myers Park in December 2016, after the first payment); id. at 51 (Burnette and Sweet talking about sealing the real-estate deal because “the check’s already been written in the back door to Scott Maddox”).

Viewing the evidence in the light most favorable to the government —and the verdict—we conclude that a rational jury could have found that Burnette agreed to facilitate the bribery of Maddox to act on a matter (or matters) that satisfied McDonnell’s requirements .

* * *

Burnette has presented serious McDonnell-based challenges to his extortion and honest-services-fraud convictions. In the end, though, we conclude (1) that any uninvited error the district court made in instructing the jury regarding the nature of § 201(a)(3)- qualifying “official act[s],” even if plain, didn’t affect Burnette’s substantial rights, and (2) that the evidence at trial was sufficient to support Burnette’s conviction on a proper understanding of the term “official act.”

29

21-13990 Opinion of the Court 29

III

Burnette next challenges his convictions on two evidentiary grounds. First, he contends that the district court erroneously excluded salacious evidence about Sweet’s conduct during the undercover investigation. Second, Burnette asserts that the court wrongly admitted Sweet’s testimony accusing him of making “false exculpatory statements” in an effort to exonerate himself once he suspected that Sweet might be operating undercover. We review both rulings for abuse of discretion. See United States v. Henderson , 409 F.3d 1293, 1297 (11th Cir. 2005). Under the deferential abuse-of-discretion standard, “we will affirm even if ‘we would have decided the other way if it had been our choice.’” Yellow Pages Photos, Inc. v. Ziplocal, LP, 846 F.3d 1159, 1163 (11th Cir. 2017) (citation omitted).

A

Shortly before trial began, the government disclosed to Burnette ’s defense team that Maddox had informed prosecutors that Sweet bought him a private dance and oral sex at a strip club during the Las Vegas trip. Confronted with Maddox’s allegations, Sweet initially denied them. Prosecutors then sent Sweet an audio recording from the evening in question in an effort to refresh his recollection . On the recording, Sweet can be heard saying, “I just paid $750 for [Maddox] to get fucked,” and then someone—either Sweet or another agent, the district court couldn’t tell—said, “Well, not get fucked.” Having listened to the recording, Sweet admitted to

30

30 Opinion of the Court 21-13990

buying Maddox a dance but denied paying for him to receive oral sex.

In response to the government’s motion in limine, the district court ruled that Burnette’s lawyers could ask Sweet whether he bought Maddox a dance, but it prohibited them from questioning him about the oral sex and it excluded the audio recording. The court based its decision on Federal Rules of Evidence 608(b) and Rule 403. We will address those Rules and their application here in turn. Although we are skeptical of the district court’s Rule 608(b) ruling, we conclude that Burnette can’t surmount the double hurdle of Rules 608(b) and 403 and that, in any event, the court didn’t abuse its discretion in excluding the evidence.

In pertinent part, Rule 608(b) states that “extrinsic evidence is not admissible to prove specific instances of a witness’s conduct in order to attack . . . [a] witness’s character for truthfulness.” Fed. R. Evid. 608(b). Burnette insists that he sought to introduce the oral-sex evidence not to impugn Sweet’s general “character for truthfulness” but, rather, to “undercut Sweet’s credibility as a supposedly unbiased witness in this case.” Br. of Appellant at 50. In particular, Burnette contends (1) that “Sweet’s actions suggested that he was willing to go to extraordinary lengths—even to break the law—to get Maddox and Burnette,” (2) that Sweet’s alleged lawbreaking “gave him a motive to shade his testimony to harm Maddox and Burnette in order to justify his unlawful actions or undermine the credibility of the allegations against him,” and (3) that “his apparent lie to the prosecutors further revealed his bias and

31

21-13990 Opinion of the Court 31

would have destroyed his credibility as the government’s star witness .” Id. at 51.

For its part, the government asserts that “[d]espite what he claims on appeal,” Burnette actually sought to attack Sweet’s “character for truthfulness” within the meaning of Rule 608(b). Br. of Appellee at 41. For support, the government points to Burnette’s lawyers’ statements to the district court that they would use the oral-sex evidence to “impeach” Sweet and show that he had “lied” to prosecutors about his conduct during the investigation. Id. at 39, 42. But those descriptions do not ipso facto show a violation of Rule 608(b): “[I]mpeach[ment]” evidence—even impeachment evidence aimed specifically at demonstrating a witness’s “lie[s]”— could bear just as easily on the witness’s credibility and bias as on his general character for truthfulness.

The distinction between a witness’s credibility or bias, on the one hand, and his character for truthfulness, on the other, is real, as both the Supreme Court and this Court have recognized. See, e.g., United States v. Abel, 469 U.S. 45, 56 (1984) (distinguishing between extrinsic evidence presented to demonstrate a witness ’s “bias,” which Rule 608(b) permits, and evidence presented to show his lack of “veracity,” which the Rule prohibits); United States v. Drury, 396 F.3d 1303, 1315 (11th Cir. 2005) (“An ‘attack’ that consists only of ‘[g]overnment counsel pointing out inconsistencies in testimony and arguing that the accused’s testimony is not credible does not constitute an attack on the accused’s reputation for truthfulness within the meaning of Rule 608.’”) (quoting United

32

32 Opinion of the Court 21-13990

States v. Danehy, 680 F.2d 1311, 1314 (11th Cir. 1982)). And indeed , Rule 608(b) was amended in 2003 specifically to underscore that distinction; its language was altered to substitute the phrase “character for truthfulness” for the word “credibility.” See Fed. R. Evid. 608(b) comm. note. Accordingly, under the amended Rule, “the absolute prohibition on extrinsic evidence applies only when the sole reason for proffering th[e] evidence is to attack or support the witness’ character for truthfulness.” Id. (emphasis added); see also United States v. Carthen, 906 F.3d 1315, 1325 (11th Cir. 2018) (W. Pryor, C.J., concurring) (emphasizing that, as amended, “[t]he Rule does not speak to anything other than the use of extrinsic evidence to support or attack a witness’s character for truthfulness”).

Having said that, the line between evidence used to impeach a witness on the ground that he is biased or lacks credibility and evidence presented to show that he has a tendency to lie more generally is a fine (and hazy) one. The one can very easily bleed into— and reasonably be understood as focusing on—the other. And for that reason, we have been reluctant to hold that district courts have abused their discretion in deciding Rule 608(b) issues. See, e.g., United States v. Ochoa, 941 F.3d 1074, 1094 (11th Cir. 2019); Carthen, 906 F.3d at 1320–21; Drury, 396 F.3d at 1315; United States v. Novaton, 271 F.3d 968, 1004–07 (11th Cir. 2001); United States v. Gonzalez, 71 F.3d 819, 836 (11th Cir. 1996), abrogated on other grounds, Davis v. United States, 564 U.S. 229 (2011); United States v. Smalley, 754 F.2d 944, 951 (11th Cir. 1985). So too here. The district court reasonably (if perhaps incorrectly) concluded

33

21-13990 Opinion of the Court 33

that questioning and extrinsic evidence related to Sweet’s alleged conduct in purchasing oral sex for Maddox, as well as his subsequent denial, bore on his “character for truthfulness” within the meaning of Rule 608(b). Even if we might “have decided the other way if it had been our choice,” Yellow Pages Photos, 846 F.3d at 1163, we cannot say that the district court abused its discretion.

That is especially so in light of the court’s invocation of Rule 403 alongside Rule 608(b). The former provides that “[t]he court may exclude relevant evidence if its probative value is substantially outweighed by a danger of one or more of the following: unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.” Fed. R. Evid. 403. We have repeatedly held that evidentiary determinations under Rule 403’s balancing test are “largely committed to the discretion of the district court[s].” United States v. Lopez, 649 F.3d 1222, 1247 (11th Cir. 2011).

Burnette insists that “[e]vidence that Sweet broke the law to ingratiate himself with and/or ensnare Maddox and then lied after being caught was highly probative of his credibility and bias with respect to this investigation and its targets.” Reply Br. of Appellant at 22. There are, however, two substantial counters. First, the probative value of the oral-sex evidence—particularly to Burnette’s theory that Sweet sought to “ensnare” either him or Maddox—is diminished by the fact, confirmed by Burnette’s lawyers at trial, that they were not pursuing an entrapment defense. Second, introduction of the evidence most definitely risked (at the very least)

34

34 Opinion of the Court 21-13990

“confusing the issues,” “misleading the jury,” and “undu[ly] delay [ing]” the proceedings. Fed. R. Evid. 403; see also Noel Shows, Inc. v. United States, 721 F.2d 327, 329 (11th Cir. 1983). The district court reasonably concluded that if it allowed questioning and admitted evidence about Sweet’s alleged sex-act purchase, the proceedings would have devolved into a sideshow mini-trial—for instance , about what the recording actually revealed: Even if Sweet said, “I just paid $750 for [Maddox] to get fucked,” who—Sweet or someone else—uttered the follow-up, “Well, not get fucked”? And if it was Sweet, what import—that he had only purchased oral sex rather than intercourse, that he had caught himself and retreated having realized that he was on tape, or something else? Cf. Anderson v. WBMG-42, 253 F.3d 561, 567 (11th Cir. 2001) (excluding evidence that “would have in effect generated a mini-trial on collateral issues” based on Rule 403). Given the evidence’s limited probative value and the specter of a “trial within a trial,” we can’t say that the district court abused its considerable discretion in concluding that Rule 403 justified its exclusion.

B

Burnette separately challenges the district court’s refusal to exclude portions of Sweet’s testimony in which he opined that Burnette made “false exculpatory statements” when, having guessed that Miller and Sweet were undercover, he seemed to backtrack on the bribery scheme. For instance, during a recorded December 2016 telephone call, Burnette told Sweet that he didn’t want him “to think that [Sweet could] effectively pay these people and get a[]

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 35 of 51

21-13990 Opinion of the Court 35

vote.” On direct examination, Sweet characterized Burnette’s remark as a “false exculpatory statement”:

I immediately thought [that] this was what we call a false exculpatory statement. “Exculpatory” means to remove from guilt; “false exculpatory” means it’s a false statement being made about guilt, removing one’s self from guilt.

As soon as I heard this phone call and I knew Mr. Maddox’s concerns about us being potential FBI agents, maybe even having recorded him, I immediately thought to myself, okay, this is the false exculpatory statement that they are making to me.

Doc. 460 at 5.

Sweet followed up by saying that he thought Burnette’s statement was part of a “completely . . . false exculpatory telephone call.” Id. at 13. In the same way, Sweet testified that a text message in which Burnette wrote that he “did not pay a $” for votes was a “false exculpatory comment[].” Id. at 31.

Burnette contends that by allowing Sweet’s testimony, the district court impermissibly allowed him to opine on another’s truthfulness in violation of Federal Rules of Evidence 401, 602, and 608(a). 2 For support, he points, for instance, to United States v.

2 One housekeeping item: The government asserts that Burnette failed to preserve his challenge to Sweet’s “false exculpatory” testimony, both because his

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 36 of 51

36 Opinion of the Court 21-13990

Schmitz, in which we held that Rule 608(a) “does not permit a witness to testify that another witness was truthful or not on a specific occasion,” that Rule 602 prohibits lay witnesses from testifying to matters with respect to which they lack “personal knowledge,” and that under Rule 401’s general relevance standard, “one witness’s opinion that another person has or has not lied does not make it more or less likely that the person actually lied.” 634 F.3d 1247, 1268–69 (11th Cir. 2011); see also, e.g., United States v. Rivera, 780 F.3d 1084, 1097 (11th Cir. 2015) (observing that a “prosecutor . . . cannot” ask one person—there, the defendant himself— “whether a particular witness was lying”).

For its part, the government points to United States v. Henderson , in which we held that even though, as a general matter, one witness can’t testify about another witness’s truthfulness, a district court doesn’t err when it permits a law-enforcement officer to

lawyers didn’t object every time Burnette used that term and because, when they did, they objected on “speculation” grounds. We hold that Burnette’s lawyers did enough. First, having brought the false-exculpatory issue to the district court’s attention and obtained rulings on their objections, Burnette’s lawyers didn’t need to continue to beat the drum. See Fed. R. Evid. 103(b); United States v. Hoffer, 129 F.3d 1196, 1202–03 (11th Cir. 1997) (holding that objections were preserved where they “were sufficient to allow the district court to correct any errors”). Second, Burnette’s lawyers’ “speculation”-based objections were sufficiently related to—and thus adequate to preserve—the improper-opinion-related arguments that Burnette now makes on appeal. See, e.g., United States v. Schmitz, 634 F.3d 1247, 1268– 69 (11th Cir. 2011) (holding that one witness’s opinion that another has lied was improper because it lay “beyond the personal knowledge of the witness”).

37

21-13990 Opinion of the Court 37

explain the manner in which he conducted his investigation, even if his testimony incidentally bears on another witness’s credibility. 409 F.3d 1293, 1299 (11th Cir. 2005). That, the government says, is how to understand Sweet’s false-exculpatory testimony here: Throughout the trial, Burnette insinuated that the agents had coaxed him to engage in bribery and, indeed, that they had continued to pursue him even in the face of his statements disclaiming a desire (or need) to pay Maddox. The agents persisted in their investigation , Sweet testified, because they thought that Burnette was feeding them “false exculpatory statements.”

We needn’t choose between the parties’ competing interpretations of Sweet’s testimony or decide whether the district court erred in allowing it because we conclude that any error that might have occurred was harmless. “An evidentiary error ‘is harmless unless there is a reasonable likelihood that [it] affected the defendant ’s substantial rights.’” United States v. Frediani, 790 F.3d 1196, 1202 (11th Cir. 2015) (quoting United States v. Hands, 184 F.3d 1322, 1329 (11th Cir. 1999)). Accordingly, “w[e] need not reverse [Burnette’s] conviction if the [alleged] error had no substantial influence on the outcome and sufficient evidence uninfected by the error supports the verdict.” Id. (quoting Hands, 184 F.3d at 1329). In making that determination, we must “weigh[] the record as a whole, examining ‘the facts, the trial context of the [alleged] error, and the prejudice created thereby as juxtaposed against the strength of the evidence of defendant’s guilt.’” Hands, 184 F.3d at

38

38 Opinion of the Court 21-13990

1329 (quoting United States v. Reed, 700 F.2d 638, 646 (11th Cir. 1983)).

We think it exceedingly unlikely that the jury’s verdict in this case rose or fell on its assessment of Sweet’s false-exculpatory testimony. In one of the cases on which Burnette principally relies, we observed that “[t]his [C]ircuit has found even prejudicial nonconstitutional error harmless in criminal cases in which the government has presented highly convincing, admissible evidence of a defendant ’s guilt, such as . . . audiotapes or videotapes of the defendant engaging in or discussing the alleged criminal activity . . . .” Id. (emphasis added) (citing United States v. Wilson, 149 F.3d 1298, 1302 (11th Cir. 1998)). There, we found it significant that “[t]he government presented no similarly compelling pieces of evidence.” Id. Here, by contrast, as already explained in detail in connection with Burnette’s other challenges, there is ample evidence—including plenty of recorded audio—detailing Burnette’s involvement in the plans to bribe Maddox. See supra at 23, 26–28.

IV

Burnette raises one final challenge: He contends that the evidence was insufficient to support his conviction under 18 U.S.C. § 1001(a)(2) for making false statements to FBI agents during a May 2017 interview regarding his involvement in the bribery scheme. For the following reasons, we disagree.

Burnette’s false-statements conviction was based on answers that he gave in response to five questions the agents put to

39

21-13990 Opinion of the Court 39

him. Burnette asserts that all five of the agents’ questions were ambiguous—three of them “fundamentally” so and the remaining two at least “arguably” so. The distinction between fundamental and arguable ambiguity is important because it affects our review of the jury’s general verdict on the false-statements count, which doesn’t distinguish among Burnette’s five responses.

The parties here agree that, under our precedent, if a question “is so vague as to be ‘fundamentally ambiguous,’ the answers associated with” it are “insufficient as a matter of law to support [a] perjury conviction”—or, as here, a false-statements conviction. United States v. Manapat, 928 F.2d 1097, 1099 (11th Cir. 1991) (quoting United States v. Lighte, 782 F.2d 367, 375 (2d Cir. 1986)). Accordingly, and in light of the jury’s general verdict, if even one of the agents’ questions to Burnette was fundamentally ambiguous , the false-statements conviction cannot stand. Cf. United States v. Pendergraft, 297 F.3d 1198, 1210 (11th Cir. 2002) (vacating a general verdict for a multiple-object conspiracy on the ground that one of the conspiracy theories was contrary to law).

By contrast, if “a question is only arguably ambiguous and ‘an answer would be true on one construction of [the] question but false on another . . . the defendant’s understanding of the question is a matter for the jury to decide.’” United States v. Swindall, 971 F.2d 1531, 1553 (11th Cir. 1992) (quoting United States v. Bell, 623 F.2d 1132, 1136 (5th Cir. 1980) (quotation marks omitted)). “In such a situation, we review under the same sufficiency-of-the-evidence standard used for a jury’s determination of falsity.” Id.

40

40 Opinion of the Court 21-13990

Accordingly, absent any fundamental ambiguity, the jury’s general verdict must stand if the evidence is sufficient to support a conviction based on any of the statements that Burnette made, even in response to a question that might have been arguably ambiguous. Cf. Griffin v. United States, 502 U.S. 46, 56–57 (1991) (distinguishing between legal and factual insufficiencies in relation to general criminal verdicts).

We first consider whether any of the questions on which Burnette focuses was, as he contends, fundamentally ambiguous. A question satisfies the high fundamental-ambiguity bar only if it lacked “a meaning about which men of ordinary intellect could agree” and couldn’t “be used with mutual understanding by a questioner and answerer unless it were defined at the time [the answers] were sought.” Manapat, 928 F.2d at 1100 (quotations and citations omitted). Burnette highlights three instances in which he says the agents asked fundamentally ambiguous questions. First, and with our emphasis added—

Agent: Okay. Who did you recommend?

Burnette: I recommended, uh, Paige Carter-

Smith—and—that was it.

Agent: Okay. Who, is she with a firm or with . . . Is she on her own?

Burnette: She . . . used to be with—I don’t even know the name of the firm.

41

21-13990 Opinion of the Court 41

Doc. 440-2 at 223. Second—

Burnette: Um, but, like, I mean, I hate to say it.

Like, they didn’t know, like, who the right—I mean—like, they hired, like, Adam Corey, okay?

Agent: They did hire Adam Corey?

***

Burnette: So, you know, Adam is a lobbyist. I don’t—I mean, he’s not like the best lobbyist in the world. (Laughter)

Burnette: Um, so, I mean, they just—they didn’t really seem to know what they were doing .

Agent: Okay. Do you know did they retain anybody?

Burnette: I do not know who they retained.

Id. at 223–24. And finally—

Agent: Sorry. Corey introduced them to Maddox and you did as well?

Burnette: I was told that Maddox, they—that they were introduced to Maddox.

***

42

42 Opinion of the Court 21-13990

Agent: Do you know the content of their conversa —what did they want from Maddox or what did—

Burnette: I don’t know the answer to that.

Id. at 224.

None of the questions to which Burnette points was fundamentally —as opposed to just arguably—ambiguous. The last seems to us the most ambiguous of the three: “Do you know the content of their conversa—what did they want from Maddox or what did—?” At worst, though, this question asked one of two things: (1) what the agents discussed with Maddox, or (2) what Burnette understood the agents wanted from Maddox. To be sure, the inquiry was imperfectly framed and executed. But it wasn’t fundamentally ambiguous within the meaning of our precedent. Compare the following question, which we deemed only arguably ambiguous in Swindall:

Do you have a specific recollection about engaging in a conversation wherein it was discussed a money trail or the ramifications of the money trail from [an agent’s] sources to ultimately you and what potential liability would exist or you would be exposed to?

971 F.2d at 1553. If that twisted, asyntactical garble was only arguably ambiguous, we don’t see how the third of the three questions

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 43 of 51

21-13990 Opinion of the Court 43

that Burnette emphasizes—to say nothing of the first and second— could be fundamentally ambiguous. 3 Because we aren’t confronted with any fundamental ambiguity , the question in reviewing the jury’s general verdict is simply whether the evidence was sufficient to permit a reasonable factfinder to conclude that Burnette made a false statement in response to any of the agents’ five queries. In short, it was. Take, for instance , the following exchange, which not even Burnette contends included a fundamentally ambiguous question:

Burnette: [S]o all [Sweet] would do is he hired lobbyists. . . . So they work with lobbyists in Tallahassee—

Agent: Who did he—did he—oh, he did? Who did he hire?

Burnette: I—let me say this. I don’t know who they ever wrote a check to.

3For the sake of completeness, neither of the first two questions is fundamentally ambiguous, either. With respect to the first, Burnette asserts that because Carter-Smith owned and operated Governance, it was fundamentally ambiguous whether she was “on her own” or “with a firm.” Ordinary people, though, could clearly agree that the agent wanted to know whether Carter- Smith worked as a solo practitioner or with others. With respect to the second , Burnette contends that the word “retain[]” is fundamentally ambiguous. Again, ordinary people could certainly agree that it simply meant “hire.” And remember, Burnette told the undercover agents to “keep Maddox on the payroll .” Doc. 440-2 at 198.

44

44 Opinion of the Court 21-13990

Doc. 440-2 at 223. A reasonable jury could certainly conclude that Burnette’s answer to the agent’s question was false. Audio recordings entered into evidence show (1) that Burnette directed Miller and Sweet to write checks to Governance, which Burnette knew full well Maddox’s girlfriend Paige Carter-Smith owned, and (2) that he knew that the checks would ultimately find their way to Maddox. See supra at 23, 26–28.

Because the FBI agents interviewing Burnette didn’t ask him any fundamentally ambiguous questions, and because the jury could reasonably conclude that Burnette lied in response to at least one of the agents’ queries, we must reject his sufficiency-of-the-evidence challenge to his false-statements conviction.

V

In sum, we hold as follows: 1. We needn’t decide whether the district court erred in instructing the jury regarding the meaning and application of the term “official act,” as used in 18 U.S.C. § 201(a)(3) and interpreted in McDonnell, because (1) Burnette invited one of the errors that he now alleges and (2) he failed to object to the other and hasn’t shown that it affected his substantial rights. 2. The evidence presented at trial was sufficient to permit the jury to conclude that Burnette assisted in bribing Maddox in connection with a § 201(a)(3)-qualifying “official act.”

45

21-13990 Opinion of the Court 45

3. The district court did not abuse its discretion in excluding evidence pertaining to an FBI agent’s conduct during the undercover investigation. And any error that the court might have committed in admitting the agent’s testimony that Burnette had made “false exculpatory statements” was harmless. 4. The evidence was sufficient to permit the jury to conclude that Burnette made actionable false statements to FBI agents in the course of their official investigation, in violation of 18 U.S.C. § 1001(a)(2). Accordingly, Burnette’s convictions are AFFIRMED.

46

21-13990 Jordan, Rosenbaum & Newsom, JJ., Concurring 1

JORDAN, ROSENBAUM, and NEWSOM, Circuit Judges, concurring:

To say that the parties “vigorously dispute” the meaning and proper application of the Supreme Court’s decision in McDonnell v. United States, 579 U.S. 550 (2016), might be an understatement. See Maj. Op. at 20. From their competing briefs, we think it’s fair to say that Burnette views McDonnell as a sea-change, while the government views it as a ripple. As is often the case, the truth, we think, lies somewhere in between.

To be fair, Burnette’s reading of McDonnell finds ample support in the language in the Court’s written opinion. For instance, the Court was adamant (1) that a § 201(a)(3)-qualifying “matter” must be “focused and concrete,” (2) that “the Government must identify” the covered matter, and (3) that “the Government must prove,” as relevant here, “that the public official . . . agreed” to “t[ake] an action ‘on’ that . . . matter.” Id. at 567–70. Knitting those three requirements together, they do seem to impose a burden on the government to establish that the transacting parties agreed that, in exchange for a thing of value, a public official would provide assistance on a particular, ex-ante-identified matter. And it’s no stretch to conclude that, by that measure, the district court’s instruction here was erroneous: It not only told jurors that the § 201(a)(3)-covered matter “need not be identified at the time of payment,” but also (and potentially more problematically) told them that it was “sufficient” that they find that a payment was made to Maddox “in exchange for favorable treatment on any not- yet-known matter that might later come up for a vote.”

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 47 of 51

2 Newsom, Rosenbaum & Jordan, JJ., Concurring 21-13990

We must take care, though, not to overread the Supreme Court’s opinion. On the most muscular version of Burnette’s interpretation , McDonnell requires the government to prove in every bribery-related case that the transacting parties made a particular , discrete, identifiable project—like, say, Fallschase or Myers Park—the explicit focus of their agreement. The government is rightly worried that such an understanding of McDonnell would upend decades’ worth of settled bribery law. It is particularly concerned , it seems, about the continuing viability of what it calls the “as-the-opportunities-arise” and “retainer” theories of bribery— both of which, as it explains, refer to circumstances in which a payment is made to a public official in exchange for “a continuing course of conduct” rather than a specific, distinct project. Br. of Appellee at 18 (quoting United States v. Roberson, 998 F.3d 1237, 1245 n.10 (11th Cir. 2021)). 1 Before McDonnell, the government correctly says, courts applying the as-the-opportunities-arise and retainer theories had “coalesced around the principle that the government need not prove that the briber and the public official had, at the time of payment, identified a particular item for the public official to influence through these official acts.” Id. at 18–19; see

1 Courts (and the parties here) sometimes refer to the as-the-opportunities- arise and retainer theories of bribery interchangeably, although without further explanation. See, e.g., Roberson, 998 F.3d at 1245 n.10 (“The ‘retainer,’ ‘as opportunities arise,’ or ‘stream of benefits’ theory of bribery[] occurs when a person bribes an individual or entity in exchange for a continuing course of conduct.”). We don’t think any distinction that might exist between the two matters for present purposes.

48

21-13990 Jordan, Rosenbaum & Newsom, JJ., Concurring 3

United States v. Whitfield, 590 F.3d 325, 353 (5th Cir. 2009) (stating before McDonnell that the “overwhelming weight of authority” concluded that § 201 does not require “identif[ying] a particular case that would be influenced” at the time of the agreement). An important question, then, is whether McDonnell “thr[e]w all this out the window.” Br. of Appellee at 26.

Our view is that McDonnell is best understood as having tweaked, but not scrapped, the as-the-opportunities-arise and retainer theories. As for the “not scrapped” part, we’ve already said as much. In United States v. Roberson, we stated flatly that McDonnell “did not reject the retainer theory of bribery.” 998 F.3d at 1251. For support, we pointed to the Second Circuit’s post- McDonnell decision in United States v. Silver, 948 F.3d 538 (2d Cir. 2020), in which that court concluded that the as-the-opportunities arise theory of bribery had survived McDonnell, albeit in modified form. In particular, while it disagreed with the defendant’s contention that “McDonnell eliminated th[e] so-called ‘as the opportunities arise’ theory” outright, it “agree[d]” that any application of that theory had to account for the fact “that [McDonnell] requires identification of a particular question or matter to be influenced.” Id. at 552 (emphasis omitted). Accordingly, the court clarified, in the wake of McDonnell, “a public official must do more than promise to take some or any official action beneficial to the payor as the opportunity to do so arises; she must promise to take official action on a particular question or matter as the opportunity to influence that same question or matter arises.” Id. at 552–53.

49

4 Newsom, Rosenbaum & Jordan, JJ., Concurring 21-13990

So, if McDonnell didn’t scrap the as-the-opportunities-arise and retainer theories, exactly how did it tweak them? In what form do they survive McDonnell’s requirement that the government prove that a public official “agreed” to “t[ake] an action ‘on’” a “focused and concrete” matter? 579 U.S. at 567–70. As we see it, McDonnell likely invalidated one brand of these continuingcourse -of-conduct theories but left another intact—and the distinction lies in the very nature of bribery, whose essence is an agreement , the corrupt bargain itself. Cf. Evans v. United States, 504 U.S. 255, 268 (1992) (“[T]he offense is completed at the time when the public official receives a payment in return for his agreement to perform specific official acts; fulfillment of the quid pro quo is not an element of the offense.”).

First, then, the invalid brand of as-the-opportunities-arise and retainer liability: A bribery-related agreement, like any other agreement, requires a bargained-for exchange—a promise of an identifiable quid in return for an identifiable quo. If, in exchange for a thing of value, a public official pledges only, say, to vote for his benefactor on some unspecified, future project that might someday come before him, his promise is so vague as to be illusory —there has been no meeting of the minds, no promised exchange of the sort that is a prerequisite to an agreement. After McDonnell, it seems to us that so hazy a promise can’t form the basis of a bribery conviction.

Now for the valid brand: If, instead of promising his assistance only on some undefined future project, the official commits

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 50 of 51

21-13990 Jordan, Rosenbaum & Newsom, JJ., Concurring 5

to vote in favor of his benefactor’s pet projects every time one comes up for consideration—or, for that matter, to vote in accordance with his benefactor’s interests every time an issue comes up, no matter whose particular project is involved—he has, to our minds, made a corrupt bargain. He has done the dirty deed. In those circumstances, there has been a meeting of the minds, and the consideration (so to speak) at the heart of the agreement is tangible and identifiable—no less so, in fact, than had the official committed to vote for his benefactor on a specific, discrete, identifiable “Project X.” Concurring in Silver, Judge Lohier addressed this very scenario, and his words are worth repeating:

The payment is not in exchange for a vague promise to act in the payor’s general interests at discrete moments to be determined only at the official’s discretion . . . . Instead, it solicits a promise that the official will filter every official act through the lens of the payor’s interests. In other words, the promise is not vague, amorphous, or subject only to the official’s discretion of when and where to act. Although the matter that is the subject of the promise is broad in scope, its contours are clearly defined.

Silver, 948 F.3d at 578–79 (Lohier, J., concurring).

And to be clear, not only does this “every time” brand of the as-the-opportunities-arise theory square with the first principles of contract law that underlie bribery liability—it comports with common sense. What a topsy-turvy world it would be if the federal bribery statute criminalized an agreement in which a public official

USCA11 Case: 21-13990 Document: 65-1 Date Filed: 04/11/2023 Page: 51 of 51

6 Newsom, Rosenbaum & Jordan, JJ., Concurring 21-13990

pledged to back his benefactor’s interests on a single, discrete, one- off project but didn’t cover an agreement in which that same official pledged to do so without fail on a series of numerous projects. The every-time scenario isn’t just “worse” in an absolute sense than the one-off scenario—it’s a seriatim repetition of it.2

* * *

Our point is a simple one: McDonnell arose against a unique set of facts, and the Supreme Court was understandably worried about sanctioning the criminalization of vast swaths of day-to-day lobbying and constituent-relations activity. See 579 U.S. at 574–77. There is a risk, though, of over-rotation—i.e., that we might misread some of the language in the Court’s opinion as having done more than the Court intended. Those of us on what the Constitution calls “inferior courts,” U.S. Const. art. III, § 1, would do well to tread lightly and await further direction from our bosses before concluding that McDonnell revolutionized bribery law as we have long known it.

2 The district court seems to have been concerned about the “every time” phe-

nomenon, as well. Defending its instruction post-trial, it questioned whether the following could possibly not constitute bribery: “I will pay you $10,000 per month starting now if you will agree that you will vote in my favor when my future projects, whatever they may be, come before the commission.” Doc. 539 at 18.