The Yale Law Journal


Navigating Between “Politics as Usual” and Sacks of Cash

16 Feb 2024

abstract. Like other recent corruption reversals, Percoco was less about statutory text than what the Court deems “normal” politics. As prosecutors take the Court’s suggestions of alternative theories and use a statute it has largely ignored, the Court will have to reconcile its fears of partisan targeting and its textualist commitments.


When a lobbyist makes a call to persuade a state official to do his bidding on behalf of a paying client, that’s called America. What about when a governor’s right-hand man, who has temporarily stepped down from his high executive post to work on the governor’s campaign and is about to resume his official duties, gets paid to make such a call? How hard should we try to distinguish between the two for the purposes of the federal mail- and wire-fraud statutes?

Such was the issue before the Justices last Term in Percoco v. United States,1 and they thought it quite easy, at least when it came to overturning the conviction of Joseph A. Percoco, a top aide to former New York Governor Andrew M. Cuomo. After resigning his formal state position in order to manage the Governor’s re-election campaign, Percoco had been paid by a private developer to obtain favorable treatment from a state agency. Shortly thereafter, Percoco resumed his official position in the Governor’s office. The jury had been instructed that it could convict Percoco on mail- and wire-fraud charges on a bribery theory because of his “special relationship” with the state government and “dominat[ion] and control[]” of state business.2 Justice Alito—usually quite sympathetic to the statutory arguments of federal prosecutors—rejected this logic on behalf of a unanimous Court. The case’s facts were unique and might, the Court suggested, have supported conviction on a different theory. Even so, the decision raises critical questions about the sweep of the Court’s recent reversals of convictions in public-corruption cases and the effect of those decisions on corruption enforcement.

To be sure, the Supreme Court’s skepticism towards what it deems expansive prosecution theories is not limited to the public-corruption area. Recent years have seen a steady drip of federal criminal law decisions in which the Court’s close reading of statutory texts and concerns about vagueness have led it to overturn convictions under the Computer Fraud and Abuse Act,3 the Sarbanes-Oxley Act,4 the Chemical Weapons Convention Implementation Act of 1998,5 and even the Controlled Substances Act, at least when doctors are involved.6 As Justice Sotomayor put it last Term, in the course of a decision overturning a conviction under the aggravated-identity-theft statute7: “Time and again, this Court has prudently avoided reading incongruous breadth into opaque language in criminal statutes.”8 Still, the Court has taken a particularly keen interest in public-corruption cases, which occur at the intersection of its general concerns about prosecutorial exploitation of overbroad statutes and its specific concerns about keeping the political process free from excessive regulation and the risk of partisan prosecutions.

Among the questions Percoco raises, which will be explored here, are, first, how should we read the Supreme Court’s increasingly regular interventions into federal corruption prosecutions? Part I of this Essay suggests that the Court is driven by a concern that statutory interpretations criminalizing “normal” political activity would not lead to even-handed prosecutions of all such political actors but run the risk of partisan targeting. Second, how does the Court’s interpretative methodology in these cases square with the Court’s vaunted commitment to textualism? Part II shows that the close attention to statutory text that has dominated so many of the Court’s recent federal criminal law opinions has played little or no role in its public-corruption cases, which seem to deploy a textualist style that only casually connects with a text, a sort of “textless textualism.” Third, how much of a constraint will Percoco and other recent cases impose on the government’s ability to pursue corrupt conduct worthy of criminal prosecution (assuming a rough consensus on what such cases entail)? Part III, in its first section, highlights the uncertainty created when the Court accompanies its reversals of convictions with suggestions of alternative theories that might, had they been properly pursued, have captured the same conduct. In its second section, Part III shows how the Court’s studied refusal, so far, to engage with the expansive text of the Federal Program Bribery statute, 18 U.S.C. § 666, heightens the uncertainty.

I hope federal interest in pursuing corrupt arrangements far more nuanced than the exchange of sacks of cash for official favor continues. Should that happen, the Court will have to squarely confront the tension between its fears of the partisan targeting allowed by expansive statutes and its ostensible commitment to statutory text. For the statutes in this area are broad indeed, and the Court’s view of what constitutes “normal” politics may require recalibration.

I. the margiotta project

At the heart of Percoco and the Supreme Court’s other recent corruption cases is a concern about the special risks of statutory overbreadth affecting political actors. Statutory interpretation must avoid criminalizing “politics as usual” not because the Court really worries all actors engaging in the relevant conduct will be prosecuted. Rather, the worry is that only a handful will; a handful targeted for partisan reasons, rather than the egregiousness of their conduct.

One may be tempted to look to McNally v. United States9 in 1987—in which the Court rejected the “honest services” theory of mail- and wire-fraud that had hitherto been used to prosecute bribes and undisclosed conflicts of interests—as the historical origin of the Supreme Court’s recent efforts to pare back the sweep of federal corruption statutes. However, the story of Percoco and the Court’s concern about partisan targeting more appropriately starts five years earlier, with Second Circuit Judge Ralph Winter’s 1982 dissent in United States v. Margiotta.10

In Margiotta, the Second Circuit upheld the mail-fraud conviction of a local party chairman for violating a duty of “honest and faithful services” to the public when he failed to disclose a secret scheme to steer town business to an insurance agency that, in turn, kicked back a portion of its compensation to his political allies.11 Under the honest-services theory, which was widely accepted across circuits,12 the public official who accepted a bribe had committed mail fraud by deceitfully depriving the public, to whom he owed a fiduciary duty, of its intangible right to his honest services. It did not matter whether this fraud had caused any financial loss to the public; the deprivation of that intangible right was sufficient.

Judge Winter’s dissent bewailed not merely the imposition of a fiduciary duty to the public on a private individual without official responsibilities but the breadth of this new “catch-all prohibition of political disingenuousness.”13 His concern was not lack of notice as much as selective prosecution. With there being “no end to the common political practices which may now be swept within the ambit of mail fraud,”14 the risk was less of “wholesale indictment of candidates, public officials and party leaders,” than of “the degree of raw political power the freeswinging club of mail fraud affords federal prosecutors.”15

It’s a fair bet that Winter’s dissent in Margiotta loomed large in the Court’s decision to grant certiorari in Percoco. The Second Circuit relied extensively on what it repeatedly referred to as the binding Circuit precedent in Margiotta in affirming Percoco’s conviction.16 And Percoco’s lawyers had, in turn, understandably relied extensively on Winter’s dissent in their petition and, later, merits brief.17

Justice Alito’s opinion for a unanimous Court gave Judge Winter a posthumous victory lap. That opinion found that the imposition of a fiduciary duty to the public on private individuals exercising de facto influence on governmental operations opened too many ordinary political players to criminal prosecution.18 Citing Winter, Alito explained:

From time immemorial, there have been éminence grises, individuals who lacked any formal government position but nevertheless exercised very strong influence over government decisions. Some of these individuals have been reviled; others have been respected as wise counselors. The Margiotta test could be said to apply to many who fell into both of these camps. It could also be used to charge particularly well-connected and effective lobbyists.19

At the core of Judge Winter’s dissent and Justice Alito’s opinion is a concern about the special relationship between statutory overbreadth and selective prosecution when it comes to the conduct of political figures. Federal criminal statutes are famous for their breadth, particularly given that federal enforcement resources can never, and are not intended to, be deployed against all those who violate the legislative prohibition.20 And the fact that a defendant can therefore point to many similarly situated people not prosecuted is rarely a legal basis for relief. Indeed, selectivity is the hallmark of the federal system, with virtually all defendants targeted in a strategic deployment of enforcement resources, for deterrent or other reasons.21

But selectivity and overbreadth are particularly worrisome in the political sphere. When a criminal statute can be read to cover politics as usual, one might worry about First Amendment chill (if one assumes that the conduct arguably covered has constitutionally recognized value) or about the unfairness when someone is charged for conduct widely deemed acceptable. Yet, for Judge Winter and Justice Alito, the most troublesome part of an interpretation that covers politics as usual is that the process of selecting out the relative handful of political figures to charge will be inevitably, or at least is at great risk of being, political, even partisan.

This connection between overbreadth and partisan targeting did not play an explicit role in McNally in 1987,22 when the Supreme Court killed the honest-services line of cases in one blow (at least until Congress quickly intervened by passing § 1346). In McNally the chair of the Kentucky Democratic Party, Howard “Sonny” Hunt, had been given “de facto control over selecting the insurance agencies from which the Commonwealth would purchase its policies.”23 Hunt would give instructions to the Insurance Commissioner, who would “execute” them.24 But that was not the crime; it was just the way things worked in Kentucky. The crime occurred when, as a condition for letting an agency continue to handle the Commonwealth’s workers’ compensation business, Hunt required it to share commissions with other agencies, including one nominally owned by Charles McNally but really controlled by Hunt and James Gray, a high-ranking official in Kentucky state government.25

Justice White’s opinion for the Court in McNally did not address whether—as the Sixth Circuit had concluded, relying on Margiotta26—Hunt owed a duty of honest service to the public.27 Rather, it flatly rejected the general proposition that any public official could be charged under the mail-fraud statute for a fraud that involved not property but the intangible right to honest services. Justice White focused on the reference to “property” in at least one part of the statute and relied on the rule of lenity and principles of federalism. He explained that, although “fraud” in § 371 can encompass a wide range of efforts to deceitfully obstruct United States government processes that do not necessarily deprive the government of “property,”28 the “fraud” in § 1341 must involve a deprivation of property. Otherwise, White noted, the federal government would end up “setting standards of disclosure and good government for local and state officials.”29

After McNally, the Court made somewhat more of an effort to bring out the Margiotta concern about the relationship between the criminalization of “politics as usual” and the risk of partisan targeting. Common practices loomed quite large in United States v. Sun-Diamond Growers of California.30 There, in the prosecution of a trade association for giving the Secretary of Commerce U.S. Open tickets, luggage, and other things, the Court addressed the scope of 18 U.S.C. § 201(c)(1)(A), which prohibits giving “anything of value” to a federal official “for or because of any official act performed or to be performed by such public official.”31 The Court held that the government had to prove a link between a thing of value conferred upon the official and a specific “official act” for or because of which it was given.32 Not requiring a nexus to a specific, identified “official act,” Justice Scalia noted, “would criminalize, for example, token gifts to the President based on his official position and not linked to any identifiable act,” like “the replica jerseys” the President gets from “championship sports teams.”33 Were such a link not required, Scalia explained, “nothing but the Government’s discretion [would] prevent[]” the giving or acceptance of such gifts “from being prosecuted.”34 This reference to prosecutorial discretion highlights the real problem for Scalia: not that federal officials at all levels would suddenly wake up criminals but that prosecutors would use unspecified and possibly troubling criteria to select a few to treat as such.

This “it can’t be” strain of analysis protecting “normal” practices, not because they are intrinsically worthy of protection, or even because of notice concerns, but because their criminalization would create a dangerous risk of selective prosecution, would be deployed in more rugged territory in McDonnell v. United States.35 There, the Supreme Court acknowledged the “tawdry” facts of the case, which had the Governor of Virginia and his wife receiving a steady stream of expensive gifts from a businessman seeking state support for his nutritional product.36 However, it invalidated a conviction that might have rested on those gifts having been given in exchange for the governor’s commitment to “merely” make phone calls or set up meetings with state officials.37 As Chief Justice Roberts explained for a unanimous Court, “conscientious public officials arrange meetings for constituents, contact other officials on their behalf, and include them in events all the time.”38 And even though the Chief Justice acknowledged that the facts in the case might not “typify normal political interaction,” he refused to “construe a criminal statute on the assumption that the Government ‘will use it responsibly.’”39

That strain would play a less explicit role in a more recent unanimous decision, Kelly v. United States.40 Justice Kagan’s explanation for why the shenanigans of New Jersey Governor Chris Christie’s Administration didn’t amount to wire or federal-program fraud focused more on the lack of a cognizable “property” deprivation—the defendants’ deceitful effort to reduce access lanes to the George Washington Bridge “did not aim to obtain money or property”—than on the nature of the alleged deceit.41 But it’s a fair bet that the Court’s rejection of the government’s property-fraud theories was at least partially driven by its desire to avoid addressing a deceit theory that turned on the defendant officials’ failure to disclose the “true” partisan nature of their conduct. Acceptance of that theory, as one defendant reminded the Court—tapping into its concern about the relationship between overbreadth and political targeting—“would subject public officials to unending second-guessing and hand their political enemies the jailhouse key.”42 After all, the Court had only recently sidestepped an inquiry into the disjunction between President Trump’s own statements about his Muslim ban and the policy justifications offered by his administration.43 Even as it acknowledged Trump’s “extrinsic statements,” the Court had avoided doing its own second-guessing and looked to the “neutral” language of the presidential directive and the broad scope of presidential authority.44 And, in a country where states had been admitted to the Union for badly concealed partisan gain,45 having juries determine whether an official’s ostensible public justification for his actions concealed a self-interested partisan one would indeed open the door to selective prosecution for “politics as usual.”

The connection between overbreadth and fear of partisan targeting by federal prosecutions was more muted in Justice Alito’s Percoco opinion than it had been in Judge Winter’s Margiotta dissent. But, as we have seen, this fear has loomed large in the Court’s recent corruption decisions. And, particularly in the wake of the Trump Administration’s use of the criminal process to target its enemies and protect its friends,46 it likely looms large for many others.47

Is this fear truly what drove the Court in Percoco and the other cases to limit federal prosecutions to only the most blatant quid-pro-quo abuses of official power? It’s not hard to spin out an alternative theory of the Court’s acquiescence, even embrace, of the outsized role of money in politics, and of antipathy to regulation in the area. Perhaps the Court overstates its commitment to the eradication of “real” corruption. With one hand, the Court justifies limiting the sweep of prophylactic campaign finance legislation by demanding evidence of obviously prosecutable quid-pro-quo corruption.48 With the other, the Court, in cases like McDonnell and Percoco, seems to preclude the prosecution of what, to an ordinary citizen, would seem to be straightforward quid-pro-quo official bribery. Moreover, the current Court has so far eschewed efforts to hold itself to what many ordinary citizens would find to be straightforward ethical standards of conduct.49 Perhaps it just lacks an institutional commitment to the anticorruption project.

But at a time when trust in prosecutorial independence and neutrality has been sorely tested, there’s value in taking the Supreme Court at its word, and I’m inclined to do so. One can, as I do, believe current claims about the “weaponization of the justice system”50 to be an extraordinarily destructive partisan ploy and still at least provisionally assume that the Court’s articulated concerns about partisan prosecutions actually drive its constrained approach to corruption statutes. It’s also worth noting that support for the Court’s corruption decisions comes from Justices Kagan and Sotomayor, who have dissented from its campaign finance decisions.51

Of course, figuring out what politics are “usual” and what are not is no small thing, to put it mildly. Receiving $100,000 in cash in a briefcase, as Louisiana Congressman William J. Jefferson did in 2005 (keeping most of it in his freezer, where FBI agents found it) is pretty clearly over the line. But that cash was ultimately supposed to go to the Nigerian vice president.52 The bribes intended for Jefferson came in a form more typical for larger scale corruption in the United States: money and stock to his wife’s consulting firm, and a separate consulting arrangement for his brother.53 New York Assembly Speaker Sheldon Silver got referral fees for potential clients referred to his law firm by individuals to whom he steered state funding through official actions.54 The ways in which a public official can improperly benefit herself or her family simply defy easy categorization.55 And line drawing becomes even more challenging when the influence being sold is, as in Percoco, that of someone who does not hold a formal government office.

II. textless textualism

What informs the Supreme Court’s line drawing besides its asserted knowledge of how the world works? One might assume that the Court’s efforts to limit the risk of partisan targeting by protecting politics as usual would be driven by, or at least heavily informed by statutory text. After all, legislators surely have a better sense of political norms than do Justices. Yet, as we shall see, the Court has advanced its challenging but perfectly sensible policy project with opinions that, while sounding in statutory interpretation, owe little to the textualism that dominates so much of its other statutory work.56 To be sure, the Court’s resort to “strong-form substantive canons” like the major questions doctrine, or, in the criminal realm, the presumption of mens rea, is, as Justice Barrett recently put it, “in significant tension with textualism.”57 But one need only read the Court’s Talmudic exegeses in Yates v. United States,58 United States v. Van Buren,59 and Dubin v. United States60 (to name just a few recent cases) to get a sense of the intense textual focus that has ostensibly driven recent federal criminal law decisions—a focus notable for its absence when the Court encounters public-corruption statutes.

In McNally, Justice White, after deploying federalism and lenity canons to explain why “mail fraud” was different from “fraud,” asserted that “[i]f Congress desires to go further, it must speak more clearly than it has.”61 Indeed, clear statement rules have become the Court’s standard way to demand heightened congressional attention to legislation encroaching on spaces the Court deems worthy of protection.62 But Congress can act clearly and forcefully for all sorts of reasons, as it did when it quickly overturned the Court’s decision. Back in the fall of 1988, on the eve of a general election, would you want to be the legislator who voted against the “Anti-Drug Abuse Act of 1988,” which contained the so-called “McNally fix”?63 Fears of being labelled both “soft on crime” and “pro-corruption” might well have powered the “McNally fix” to swift passage.64 The provision, 18 U.S.C. § 1346, blithely defined “scheme or artifice to defraud” to include “a scheme or artifice to deprive another of the intangible right of honest services,” thus appearing to resurrect the considerable, and variegated, body of common law that McNally had sought to kill.

While not fully reinstating their pre-McNally jurisprudence, lower courts, with ample prodding by prosecutors, resumed their previous common-law lawmaking until, after considerable delay, the Supreme Court again intervened in 2010 in Skilling v. United States.65 As is often the case in the corruption area, the real question is less why the Court intervened on a matter of significant circuit variation and more why it took so long to do so.66 Now the Court wielded the cudgel of possible unconstitutional vagueness to justify surgery on the lower-court case law, paring the “honest services” duty down to a duty not to take bribes or kickbacks. The policy justifications for this move were eminently sensible, avoiding the sprawling issues of defining conflicts of interest worthy of criminal prosecution that had dogged the lower courts. But the move had little to do with textualism and far more to do with a conservative version of “choose your own adventure.” Justice Scalia (joined by Justices Kennedy and Thomas) called the Court out on its pretended deference to congressional will: “Among all the pre-McNally smörgåsbord-offerings of varieties of honest-services fraud, not one is limited to bribery and kickbacks. That is a dish the Court has cooked up all on its own.”67

Yet Scalia’s interpretations of corruption statutes had their own cooked-up quality. A decade before, in Sun-Diamond, when Scalia donned his textualist robe to show why the statute required the government to show a nexus between a gift to an official and a specific “official act,” his reasoning was somewhat less than persuasive. The statutory word “any,” he explained is best understood as referring to a “specific” official act, not the general class of acts, “just as the question ‘Do you like any composer?’ normally means ‘Do you like some particular composer?’”68 This reach for the interrogative form willfully ignored the fact that “I like any composer” is most naturally read to reflect an easy-going attitude towards music appreciation.

Future public-corruption cases would similarly be marked by a nonchalant approach to statutory text. McDonnell involved close textual analysis, but of a statute—18 U.S.C § 201—that was not charged in the case (and could not have been, as federal officials were not involved) and whose application the Court accepted based on party stipulation.69 The language of the offenses of conviction—wire fraud and the Hobbs Act—played virtually no part in the analysis.

More recently, Kelly, the “Bridgegate” case, did not turn on textual analysis. Rather, as it has done elsewhere, the Supreme Court took the statutory reference to “property,” a term that it has long understood to be a legislative incorporation of common law, and applied its own restrictive analysis. Here, as in other of the Court’s fraud decisions, the authoritative text came from the Court’s own precedents, not Congress.70 Indeed, the decision’s main brush with textualism was in its brief engagement with the charges under 18 U.S.C.§ 666. Reversing defendants’ convictions under that provision without a separate analysis, the Court ignored the Third Circuit’s focus on how they had “intentionally misapplied” the Port Authority’s property, even if they had not “obtained” it.71 The statute explicitly covers both. The Court dealt with this textual impediment to its analysis, however, by simply not quoting the “misapplication” part of the statute.72

Percoco carries on this proud tradition of looking beyond statutory text, with Justice Alito doing the same sort of doctrinal surgery the Court did in Skilling. But the discomfort and methodological slipperiness of a textualist Court working without a text provoked a sharp concurrence by Justice Gorsuch (joined by Justice Thomas), which echoed Justice Scalia’s Skilling concurrence. Noting that the Court had been effectively “writing [Section 1346] bit by bit in decisions spanning decades with the help of prosecutors and lower courts who present us with one option after another,” Gorsuch admonished his colleagues that “under our system of separated powers, the Legislative Branch must do the hard work of writing federal criminal laws.” He added, “Congress cannot give the Judiciary uncut marble with instructions to chip away all that does not resemble David.”73

One need not be a card-carrying textualist to appreciate the methodology’s value, not just for giving notice to the public but for guiding prosecutors and courts. The Supreme Court has forcefully intervened in public-corruption cases less to ensure adherence to text than to protect politics as usual from the danger of partisan targeting. In doing so, the Court has conveyed its deep interest in the area without reliably conveying what form that interest will take in its future cases or offering the guidance lower courts and prosecutors require. For such guidance, those critical actors in the application of federal criminal law are better advised to look, not to statutory text, but to what sort of governmental conduct the Court is likely (if it ever grants certiorari) to deem normal.

III. whither enforcement?

The most obvious question following the Supreme Court’s periodic corruption reversals is whether they will undermine the prosecution of public corruption. While Skilling and McDonnell required the reversal of many convictions,74 Percoco involved the less frequent scenario of a nonpublic official being held responsible to the citizenry,75 so the number of convictions overturned will likely be limited. On the other hand, many reversals after McDonnell in particular were generally based on the (understandable) failure of the jury instructions in those prior cases to predict how the Court would interpret the relevant statutes. That failure will presumably be cured in the future.

Charting the future is difficult for at least two reasons that relate to the cases decided and one relating to cases not decided. First, the Supreme Court has taken pains, even when reversing convictions, to suggest that the prosecution might have pursued the same conduct with the same statutes but different liability theories. But the decisions have still overturned convictions, and usually include dicta highlighting the need for constrained readings of the relevant statutes. The second reason for uncertainty is thus the challenge of predicting the degree to which prosecutors and lower courts will look to these dicta and the trend line of reversals, instead of the Court’s relatively narrow statutory holdings. The third, and perhaps, most important wild card—at least with respect to state and local corruption cases—is the Court’s studied failure so far to seriously engage with the scope of the Federal Program Bribery statute, 18 U.S.C. § 666, which potentially gives prosecutors a vehicle for pursuing some of the very conduct the Court has sought to keep other statutes from reaching.

A. Closely Reading the Cases

It is hard to definitively call the Supreme Court’s recent public-corruption reversals harbingers of a constrained enforcement environment when the Court sometimes explains the limited effects of its own holdings. Consider McDonnell. There, the unanimous Court made much of its holding that a mere telephone call from a government official cannot constitute the “official act” that must be the subject of the quid pro quo required for mail fraud and for Hobbs Act “under color” convictions.76 But then, perhaps aware that his analysis would immunize some egregious conduct, Chief Justice Roberts assured us that “setting up a meeting, hosting an event, or making a phone call” won’t always be an “innocent act.”77 If “in exchange for a thing of value” one official called another official and tried to “pressure or advise” the second official “on a pending matter,” the call could indeed be prosecuted under these statutes.78

Put differently, Virginia Governor Robert F. McDonnell’s calls to subordinates could not count as his “official acts” within the meaning of § 201, and therefore could not provide the basis for a quid pro quo that violated the Hobbs Act and mail-fraud statutes. To count them as such would, the Chief Justice explained, raise “significant federalism concerns” and lead public officials to “wonder whether they could respond to even the most commonplace requests for assistance.”79 But were the government to charge the same call as a corrupt effort to pressure another official to do an official act as the quo of the bribery scheme, then Hobbs and mail-fraud charges would be fine. This is so even though one could make the same argument of “chill” here that carried the day in McDonnell.

So what counts as “pressure” or “advice”? This will take some time to work out. When a high official makes the call to a subordinate, perhaps the analysis will favor the government, since a polite request might amount to an order in this context. But even that might depend on the official’s influence. After all, Virginia’s governor has but one term, and the facts in McDonnell highlight the defendant’s limited sway.80 The government’s case in Percoco rested on a very different picture of state governance, at least under Governor Andrew Cuomo in New York: his “longtime political associate” could immediately get an official decision reversed with a phone call even while he was (temporarily) out of government.81

What about “inter-branch” lobbying, as when a senator—having received things of value in exchange—puts pressure on an executive official, as has been charged in the most recent prosecution of Senator Robert Menendez (and was charged in his first prosecution)?82 Or when a mayor calls a district attorney on behalf of a donor? If quid pro quos involving “mere” telephone calls can easily be reconfigured, McDonnell is just a speed bump for prosecutors in a broad range of cases, and the Court’s solicitude for federalism and constituent service becomes empty rhetoric. Perhaps McDonnell will end up being more a matter of jury instruction than a restriction on the prosecution of corrupt deals involving phone calls or meetings.83 Indeed, the new Menendez indictment will likely require the trial court to work out precisely what “pressure” the Chief Justice had in mind in McDonnell.

Alternatively, the Supreme Court’s broad language will push lower courts to be hostile to prosecutions targeting some blatant influence peddling—not the mere communication of a constituent’s plea, but the implicit or even explicit demand for action by an official rented for the occasion who takes advantage of another official’s sense of felt obligation. In short, figuring out whether the Court’s recent decisions will lead prosecutors not to pursue certain conduct, or, if they do bring charges, lead lower courts to throw them out, is challenging especially because, on several occasions, the Court seems to minimize the enforcement gap it has created.

Even in Percoco, the Court took pains to avoid addressing several alternative theories of “honest services” liability, including that Percoco’s leaving office had been a “sham,” or that a duty to the public arises out of someone’s selection for a government post in the future.84 Without more—say, a clear agency relationship with the government85—party bosses and others with powerful political connections—like Margiotta, Hunt, and Percoco—owe no duty of “honest services.” But the limits of this immunity may be narrow depending on the surrounding facts.

On the other hand, the trend line of the Supreme Court’s decisions has been clear, and in the direction of limiting the sweep of federal corruption statutes. Lower courts that in the past have been persuaded by prosecutors touting sordid conduct to expand the law to cover it,86 may be less prone to do so now. Prosecutors will read the same tea leaves and may either limit the sweep of their liability theories or not charge at all. Moreover, the Office of the Solicitor General may play a key role, amplifying the Court’s signals into enforcement limitations. In Percoco, the Solicitor General’s Office pointedly refused to defend the reasoning in Margiotta even though the Second Circuit, following the U.S. Attorney’s Office, had relied on that long-standing circuit precedent.87 (The Office also refused to defend the Second Circuit’s reasoning in Percoco’s companion case.)88 Concern that the Solicitor General’s Office will pull the rug out from under them in the Supreme Court can only deter prosecutors and lower courts from charging forward as before, which is presumably that Office’s intention.

Sure, the odds that a defendant will not plead guilty, and that, should she go to trial, the Supreme Court would actually review her conviction, are low. But corruption prosecutions are not the daily grist of a U.S. Attorney’s Office, and when they are brought, the stakes are usually high. These are often high-profile cases whose reversals cast a considerable shadow on the public-integrity project, and perhaps on the confidence of lower courts that all levels of the Justice Department will stand by the arguments that prosecutors make to them, not just in these cases but in others.

We may soon have one measure of how the Second Circuit will be affected when, in a significant corruption case, it reviews a trial court’s reconciliation of two Supreme Court cases involving the extortion “under color of official right” provision of the Hobbs Act, which the Court has long read to cover bribery.89 Shortly after Margiotta, the Court, in McCormick v. United States,90 was presented with a Hobbs Act prosecution involving a campaign contribution. It imposed what sounded like a strict requirement that prosecutors prove an express quid pro quo—evidence that payments were made “in return for an explicit promise [] to perform or not to perform an official act.”91 With typical concern for protecting “politics as usual,” the Court explained that “[t]o 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.”92

The next year, in Evans v. United States,93 in an opinion written by Justice Stevens, who had dissented in McCormick, the Court offered a far more flexible standard, without restricting it to the noncampaign context. Now prosecutors had to show only “that a public official has obtained a payment to which he was not entitled, knowing that the payment was made in return for official acts.”94 Concurring, Justice Kennedy took pains to disclaim the need for evidence of some “express” quid pro quo, “for otherwise the law’s effect could be frustrated by knowing winks and nods.”95 The sequence, and lack of clarity of these opinions, left lower courts unclear as to whether campaign contributions would be judged by a special, higher standard for purposes of a quid-pro-quo analysis in Hobbs Act cases (and perhaps mail- and wire-fraud cases as well, given the unified approach the Court took to the two statutes in McDonnell). Mindful of the subtleties of corrupt transactions in the real world, many lower courts read Evans to dispense with any “explicit” promise requirement across the board.96

Late last year, a well-respected judge in the Southern District of New York went in a very different direction when he threw out the bribery indictment of former New York Lieutenant Governor Brian A. Benjamin, concluding that the government had “not met the heightened legal standard for bribery and fraud charges in the particular context of a public official’s fundraising for a political campaign.”97 The judge’s decision rested on his reading of Second Circuit cases, which, while not involving campaign contributions, provided some support in dicta.98 The government has taken an appeal to the Second Circuit, which will perhaps give us a sense of how far the shadow of the Supreme Court’s recent corruption cases extends beyond their holdings.99 Certainly, much was made of that shadow in the amici brief filed on behalf of a number of present and former elected officials.100 One can expect similar readings to populate defense briefs in the future.101

We have thus seen the tension in the Supreme Court’s opinions between the lawyerly reminders of how the same conduct might be alternatively charged and the combination of broad dicta and the unidirectional trend of the decisions. We may therefore face the prospect of lower courts reading Roberts Court decisions as a broad mandate to privilege “politics as usual.” And the Court has not tried to limit that risk by balancing broad language protecting politics as usual with concern about the protean nature of corrupt deals and the evidentiary challenges of proving them.

B. Section 666 Waits in the Wings

Yet the greatest source of uncertainty when assessing the practical fallout of the Supreme Court’s recent public-corruption jurisprudence stems not from the cases it has decided but the ones it has not. More specifically, the Court has thus far studiously refused to engage with the scope of the federal program bribery statute, 18 U.S.C. § 666. This provision, enacted in 1984 but coming into its heyday only recently, broadly reaches any “agent” of any “organization, or [] State, local, or Indian tribal government” receiving more than $10,000 in federal benefits or support (which covers pretty much any governmental unit in the country). And it targets any such agent who “embezzles, steals, obtains by fraud, or otherwise converts to the use of any person other than the rightful owner or intentionally misapplies” more than $5,000 of the agency’s property or who “corruptly” demands or accepts anything of value in connection with transactions of $5,000 or more.102 As we will see, the breadth and textual specificity of this statute presents a challenge to the policy-driven approach the Court has so far taken with other corruption statutes.

Even as, since at least McNally and particularly since Sun-Diamond, the Court has limited corruption statutes in the name of lenity, federalism, and protecting “normal” political practices, lower courts have plunged ahead, with some variation, interpreting § 666 to dispense with many of the proof requirements that the Court imposed in the name of those canons. Most lower courts have, for example, found McDonnell’s requirement of an “official act” inapplicable to § 666 because its language is very different from that of § 201. As the Second Circuit explained, while § 201 restricts its coverage to “official acts,” defined as “acts on pending ‘question[s], matter[s], cause[s], suit[s], proceeding[s], or controvers[ies],” § 666 is “more expansive,” prohibiting individuals from seeking something of value “intending to be influenced or rewarded in connection with any business, transaction, or series of transactions of [an] organization, government, or agency.”103

These decisions have been powered less by disrespect for the Supreme Court than by attention to the text of § 666 and its distinct legislative authorization. The provision’s broad but specific language offers far less room for lenity. The First Circuit recently grabbed language from the Supreme Court’s jurisdictional decisions involving § 666 and noted that the “Supreme Court has repeatedly observed that § 666 uses ‘expansive, unqualified language’ in service of Congress’s unique interest in protecting federal funds from misuse.”104

Section 666’s basis in the Spending Clause and specific targeting of conduct affecting state and local governments reduces, perhaps even eliminates, the bite of federalism concerns (at least as a formal matter). And its deployment of expansive terms like “misapplication” opens the way to the very sort of conflict-of-interest prosecutions that Skilling barred the mail- and wire-fraud statutes from reaching.105 Given the chance to restrict the sweep of “misapplication” in Kelly, the Supreme Court avoided even using the word, even though it had been a key part of the Third Circuit’s analysis below.106 Lower courts have since reaffirmed their broad readings of the term, at least in cases involving the diversion of funds (not of the labor of public employees, as in Kelly).107

The Supreme Court’s sustained refusal to engage with the scope of § 666, save in the handful of rulings focusing on the jurisdictional amounts,108 took dramatic form in a 2016 criminal-procedure case in which it blithely dropped a footnote acknowledging, but not resolving, an important circuit conflict on whether § 666 requires a quid pro quo or can, instead, extend to gratuities.109 That conflict continues, as the Fifth Circuit recently noted when slightly evening out the “lopsided” circuit split on the issue.110

Why such a studied refusal? One reason may well be that in § 666 the Supreme Court finds a textual specificity missing in the mail- and wire-fraud statutes and the Hobbs Act—a specificity that limits (but doesn’t eliminate) its ability to cite congressional silence or invoke constitutional avoidance. Moreover, the Spending Clause basis for the statute limits the Court’s ability to rely on federalism canons, as it has done since McNally. Finally, in § 666, the Court also finds a provision that seems to be explicitly targeted at some of the “normal” practices it has strained to immunize from prosecution, including, in the absence of an “official act” requirement (and maybe not even a quid pro quo one), the “mere” phone calls in McDonnell.

Let us return to Percoco, and the “éminence grises” whose activities the Court strained to keep out of the range of § 1346 bribery. Percoco was acquitted on § 666 bribery charges relating to the scheme forming the basis for the § 1346 counts that the Court addressed.111 His lawyer speculated that the jury had been unable to find an “agency” relationship.112 Could someone, in the state or local context, be charged under § 666 with giving such an informal power broker money in exchange for a commitment to influence an officeholder in her thrall? A subsection targets anyone who

corruptly gives, offers, or agrees to give anything of value to any person, with intent to influence or reward an agent of an . . . [entity receiving federal funds] 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.113

Given that the provision targets the giving of money to a “person” to influence a government “agent” (thereby envisioning that the “person” may not be an “agent”) the provision avoids criminalizing the retention of a lobbyist only if “corruptly” does considerable work separating innocent conduct from culpable. But of course, “corruptly” has been an interpretative workhorse for the Court when drawing just this distinction.114 The word has done yeoman work in lower courts as well (as we have seen in the recent “obstruction of Congress” cases from January 6),115 and we can expect it to do that here. Such a limiting interpretation would also protect the power broker from being charged as an accessory.116

The Supreme Court, if it chose to truly engage with § 666 (in contrast to its drive-by analysis in Kelly), would doubtless be open to this move. Indeed, at the Percoco oral argument, several justices focused on the “mens rea” of this “outsider” purchaser of influence.117 But the effort would require the close statutory interpretation absent in its recent public-corruption cases. And Justice Gorsuch could not so easily dismiss § 666 the way he did § 1346, when he demanded that Congress “must do more than invoke an aspirational phrase and leave it to prosecutors and judges to make things up as they go along.”118 Absent aggressive constitutional-avoidance maneuvering, the Court’s ostensible embrace of textualism and legislative supremacy might come home to roost.

In the end, however, and even once one includes § 666, the public-corruption statutes, as foreseeably interpreted, will do little to capture the conduct of the power brokers featured in Percoco. On the facts of the case, a properly instructed jury might, the Court suggested, have properly convicted Percoco based on his intent to resume his official position, or perhaps even on the theory that his leaving was a “sham.”119 But not even the Solicitor General’s “functional” theory of government office would have reached that shadowy figure lurking just behind governmental structures—the one whose communications with actual governmental officials are treated not as “official directives” but merely as suggestions that they would fail to follow at their peril.120

IV. the challenge of influence peddling

Particularly in the current partisan environment, the Supreme Court’s strenuous efforts to protect “politics as usual”—less for fear that all participants will face prosecution than for fear that a few will be improperly singled out—are quite understandable. Yet its efforts to preclude statutory readings that create such a risk never seem balanced with a concern that cramped readings of corruption statutes will make it harder to bring the very prosecutions that the Court thinks should substitute for prophylactic regulation. As we shall see, the challenges of capturing corrosive influence peddling, and indeed even the most blatant sales of official favor, are considerable.

Consider a major donor of “hard” or “soft” money benefitting an elected official. Or maybe her spouse or close informal advisor—the one for whom the office door is always open and often used. The person who all the official’s subordinates know has the ear, even the hand, of the official, and whose wishes are backed, not by the force of law or bureaucratic hierarchy, but by the official’s vindictiveness toward those ignoring the “requests” of her closest associates and her favor toward those heeding them. That this individual’s influence is underregulated by criminal law and lacks direct democratic accountability is troubling but inevitable in a world that accepts the diverse human relationships that sustain us all. To what extent will we compound the problem of such unaccountable power by allowing this close associate to sell or rent it to others? What do we do about a modified version of McNally itself—one in which Hunt alone, without partnering with a real government official, sets up the shell agency that other insurance companies have to kick back to, but one in which the governor or some other high government official has still given Hunt the ability to order the insurance commissioner around?

The government tried to capture these scenarios in Percoco by following the relevant statutes and using terms like “fiduciary duty” or “agency.” But those terms get strained beyond recognition when applied to people like an official’s spouse or close advisor. Spiderman’s salutary axiom—recently noted by Justice Kagan121—about power and responsibility just doesn’t translate well into formal legal doctrine. But isn’t it odd to focus exclusively on Hunt’s obligations, instead of those of the governor or other high state official who either knowingly allowed his authority to be informally but effectively delegated for Hunt to sell, or willfully disregarded the likelihood that Hunt was doing that?122

It is tempting to propose an executive nondelegation doctrine, somehow restricting the powers that an official can effectively hand over to those out of government, at least when those outsiders resell it for cash. At the core of the legislative nondelegation doctrine (whatever its contours) is the idea that Congress has indefeasible powers, for whose exercise it must be held accountable. And a seminal case involved effective delegation, through the President, to private actors.123 Why not a similar doctrine, perhaps, in the federal context, based on the Appointments Clause, that bars officials from wholly outsourcing their authority, even as an informal matter?

Yet the discontents and vagaries of current efforts to devise a legislative nondelegation doctrine124 suggest that any such effort would be doomed to incoherence and failure. So we may be left with a more troubling sort of tolerationof an official formally accountable to the public disregarding her responsibility to it. Had she sold her office, we would have § 666, or § 1346, or the Hobbs Act, or § 201 (if she had federal responsibilities),125 or any number of other statutes.126 But she is free to give it away, even financially profit from the informal delegation so long as, perhaps, she has a long horizon (a soft landing after leaving office) or there is some other impediment to drawing the tight nexus required for chargeable bribes or gratuities.

The foregoing exploration of the “power-broker” problem highlights a fourfold challenge to public-corruption prosecutions. First is the legislative challenge. Capturing betrayals of the public trust with criminal statutes that are both attentive to common notions of “corruption” and politically acceptable to legislators enmeshed in common political practice is challenging enough.

Second is the judicial, as any legislative response will inevitably leave room for judicial interpretation, especially when statutes draw on common-law terms like “fiduciary duty” and “agency” that courts take as their special province.127 When courts follow recent Supreme Court cases and shun fact-sensitive engagement with betrayals of the public trust in favor of restrictive interpretations protecting “normal” politics, the burden on enforcers only increases.

Third is the evidentiary. Even were the law more receptive, all corruption prosecutions would be rare commodities. Perhaps an informant or a flipped defendant ready to help will give the FBI the predication and interest to wire up an undercover agent or even get authorization for Title III surveillance. Perhaps an investigative reporter will break a story with a partial investigative roadmap (although the odds of this may decline, given the financial pressure on local investigative journalism). Perhaps pressure from an investigation, possibly unrelated, will prod a conspirator to beat his fellows in the race to the U.S. Attorney’s Office. Maybe an honest citizen will complain. But none of these can be counted on, even for conduct easily covered by the most constrained statutory interpretations.

The evidentiary issues lead to the final challenge: enforcer commitment. A recent study found that between 1986 and 2018, U.S. Attorneys’ offices in New Jersey and Eastern and Southern New York obtained fifteen percent of all federal corruption convictions, and the top twenty offices account for the “vast majority” of all such convictions.128 These are target-rich environments, but that fact alone does not tell the full story. I don’t diminish the efforts of so many smaller offices around the country, but I’m also confident that these numbers are far more a function of U.S. Attorney and FBI interest than of corruption base rates. Evidence must be found and restrictive law navigated, but commitment and readiness to push beyond the drugs, guns, child pornography, and immigration cases that fill the dockets of all too many offices is likely the dominant factor. Reports of conspicuous betrayals of the public trust may make prosecutorial blood boil, but the prospect that the Supreme Court will narrow the statutes one plans to use can still chill, making costly cases seem even less promising.

Enforcer commitment is even more of a challenge for the state and local authorities that would optimally coordinate with federal prosecutors. The state statutes, which often broadly reach “official misconduct,” are certainly there to be used.129 Efforts to encourage their enforcement are underway, but there is a long way to go.130 It’s often going to be a federal prosecution—perhaps with the help of state enforcers—or nothing. And a Supreme Court preoccupied with avoiding the criminalization of normal politics that does not equally worry about capturing the diverse and nuanced ways official favor gets sold will allow the corrupt to drift into the normal.


Tensions in the Supreme Court’s public-corruption jurisprudence are inevitable and probably inherent in the project. On one hand, principles of federalism have combined with concerns about vagueness and First Amendment chill to regularly produce (when the Court decides to take a case) restrictive statutory interpretations that heighten the evidentiary burden on the prosecution. These interpretations sometimes seem to demand that an official receive a sack of cash while reciting some B-movie script.

On the other hand, more subtle cases will keep coming. When rolling out its Strategy on Countering Corruption, the Biden Administration explained:

When government officials steal from public coffers or fix a contract to reward a political crony, these actors directly transfer funding from essential services to private interests. Corruption also indirectly contributes to reduced public trust in state institutions, which in turn can add to the appeal of illiberal actors who exploit popular grievances for political advantage.131

Much of the Strategy’s focus was on corruption abroad, but it was careful to give due attention to the problem at home, for corruption corrodes democracy and is inimical to good public policy here as well. The relative political insulation of federal enforcers has long given them the capacity to pursue abuses of power at the state and local level, and perhaps even the federal level. Yet because nailing down the precise dimensions of a corrupt arrangement can be difficult indeed, and because the arrangements themselves can be nuanced and elaborate, prosecutors will push for expansive charging theories.

Occasionally, the Court has demonstrated keen awareness of this tension. In Evans, as we have seen, Justice Kennedy was concerned that “knowing winks and nods” might avoid prosecution. Yet, eighteen years later, in Citizens United, Kennedy could blithely observe that the fact that donors “may have influence over or access to elected officials does not mean that these officials are corrupt . . . . The appearance of influence or access, furthermore, will not cause the electorate to lose faith in our democracy.”132

It would be wrong to suggest any titanic battle between a Justice Department pursuing a core enforcement priority and a Court concerned about overcriminalization and partisan prosecutions in the political arena. After all, plenty of officials get convicted and rely heavily on the Court’s precedents but find their convictions nonetheless affirmed and their certiorari petitions denied. Still, the Court’s recent interventions, however sporadic, have all been in one direction. The message of Percoco privileged politics as usual over the private monetization of government authority, and it remains to be seen whether legislative text will make a difference.

Paul J. Kellner Professor of Law, Columbia Law School. Huge thanks to Kate Stith, Michael Dreeben, and Lauren Ouziel for their enormously helpful comments on a previous draft.