The Yale Law Journal


What Are Federal Corruption Prosecutions for?

16 Feb 2024

abstract. What explains the Supreme Court’s repeated rejection of public-corruption prosecutions over the last two decades? This Essay turns the lens on prosecutors, examining how their tendency to rely on broad theories of liability has paradoxically narrowed federal criminal law’s reach over public corruption. It investigates the dynamics of public-corruption prosecutions that push prosecutors towards breadth and away from the alternatives (a narrower theory or no prosecution at all). It considers how, relative to those alternatives, reversals at the Supreme Court have harmed the broader anticorruption project. And it proposes an alternative approach to the exercise of charging discretion in public-corruption prosecutions, one rooted in a wholesale reassessment of what those prosecutions should be for. The ultimate goal is not to find a theory through which corrupt acts are prosecutable federal crimes; the ultimate goal is to reduce corruption. This guiding principle should steer federal prosecution to where it does the greatest good: bringing to light those corrupt acts that would otherwise remain invisible to the public and thus immune from political or other consequences.


Prosecutors allege facts and charge crimes. To prevail, they must marry the two. That is, prosecutors must advance a theory of why and how the alleged facts make the defendant guilty of the charged crime. The prosecution’s theory of liability guides the selection and presentation of evidence; argument to the jury; jury instructions; and, should a conviction ensue, the appellate arguments for sustaining it. More than most federal criminal cases, public-corruption prosecutions tend to rise and fall on how prosecutors theorize liability—and they have mostly fallen in recent years, at least when reaching the Supreme Court.1

Behind these prosecutorial failures, some argue, is a Supreme Court wary of overcriminalization and federal infringement into state and local governance.2 While this is undoubtedly true, it is remarkable that these reversals have persisted across substantial Court personnel changes and have so often been unanimous.3 Among Justices of varied political commitments, interpretive approaches, and judicial philosophies—particularly on issues of federalism, criminal law, statutory interpretation, and regulating money in politics—there has been unusually strong agreement. For the Justices, these cases have not been close calls. It is worthwhile, then, to turn our attention away from the Court and consider the role prosecutors have played in these outcomes.

This Essay posits that prosecutors, and specifically, their choices of liability theories, have played an important and under-noticed role in the development of federal public-corruption law. Faced with a choice between broad and narrower theories of liability, federal prosecutors have tended to choose the former notwithstanding the appellate risks—even in cases where a narrower theory may have sufficed. The Essay considers how the dynamics of federal public-corruption prosecutions incentivize prosecutors to sometimes overreach, and the costs of those prosecutorial choices for corruption mitigation more broadly. It then offers an alternative approach, in which prosecutors construct and charge public-corruption cases keenly attentive to the larger goal of corruption mitigation and the potential for federal prosecution to undermine it.

The latest prosecutions to fall at the hands of the Court—Percoco v. United States4 and Ciminelli v. United States5—offer an ideal jumping-off point to consider the lure and peril of broad prosecutorial theories. In each, the prosecution adopted a theory of liability that, if adopted by the Supreme Court, would extend the reach of federal fraud statutes in public-corruption cases. In each, the Court handily rebuffed those efforts, weakening the laws’ capacity to deter corruption. And in neither was such an expansive theory of liability necessary; there were other, more circumscribed theories available to marry the facts to the charged crime. Why did prosecutors not pursue them?

Grappling with this question illuminates the prosecutorial pressures, incentives, and orientations that can sometimes make federal public-corruption prosecutions self-defeating. It also suggests a potential corrective to these forces. Federal prosecutors should consider prosecution’s capacity not only to deter corruption, but also inadvertently to enable it. The Court’s repeated rejection of broad liability theories defangs the deterrent effects of broadly drafted statutes, making abuse of power both more tempting and less risky for public office holders and their enablers. Federal prosecutors, then, should choose and charge cases with that potential risk in mind. They should reserve federal criminal accountability for clear violations of public authority irremediable by the political process, the state criminal process, or forms of civil regulatory accountability. And when a case meets that standard, prosecutors ought to proceed on the narrowest possible theory required to prevail.

The Essay proceeds in three parts. Part I discusses the broad theories of liability advanced in Percoco and Ciminelli, and the narrower alternatives. Part II illuminates how certain features of public-corruption cases—namely, trial risks and the federal prosecutorial role—can steer prosecutors to broad theories of liability often ultimately rejected by the Supreme Court. Part III considers the costs of those judicial rejections for the broader goal of reducing corruption, and argues that, to mitigate them, federal prosecutors should bring narrowly tailored charges for power abuses of which voters would not otherwise be aware.

I. percoco and ciminelli: when good facts meet bad theories

Percoco and Ciminelli are the most recent in a series of cases in which the Court has considered the reach of the federal mail- and wire-fraud and honest-services-fraud statutes over political corruption. In prior cases, most notably against a former Virginia Governor who granted a business executive access to key decision makers in exchange for personal gifts,6 and political aides to a New Jersey Governor who closed state roadways as “payback” against a local politician who declined to endorse the Governor’s reelection campaign,7 the Court discarded federal prosecutors’ theories of liability as reaching beyond the statutes’ limited terms.

In McDonnell v. United States, the Court made clear that a scheme to defraud another of the intangible right to honest services must not only involve quid-pro-quo bribery or kickbacks,8 but that, in public-corruption cases, the quo must involve exercise of (or agreement to exercise) formal public authority in exchange for the quid.9 In Kelly v. United States, the Court reiterated that the wire-fraud statute is limited to schemes that aim to deprive another of property, a narrowly defined term that does not extend to state regulatory authority.10 Extending these statutes too broadly to cover state and local corruption, the Court explained in both cases, would upset the balance between state and federal police power and risk criminalizing ordinary (however distasteful) political wheel-greasing.11

It was against this backdrop that federal prosecutors constructed their cases against Joseph Percoco and Louis Ciminelli. Both cases arose out of former New York Governor Andrew Cuomo’s administration and were prosecuted by the U.S. Attorney’s Office for the Southern District of New York, an office with an active and aggressive public-corruption unit that, under the leadership of Preet Bahara, had felled powerful New York politicians including State Assembly Speaker Sheldon Silver and State Senate Majority Leader Dean Skelos.12 If McDonnell and Kelly collectively served as a warning to prosecutors to tread circumspectly when charging corrupt public officials under the federal wire-fraud and honest-services-fraud statutes, prosecutors in Percoco and Ciminelli did not fully heed it. To the contrary, in both cases, prosecutors pursued broad liability theories which sought to expand those statutes to novel applications.

The strategy might have made sense had they lacked alternative options. But the particular facts of these cases gave prosecutors some breathing room. These were both cases with what prosecutors refer to as “good facts”—that is, readily proven facts that would support a conviction under settled statutory definitions. Broad theories of liability were likely helpful to securing a conviction,13 but they probably were not necessary in light of the risks they carried.

Begin with Percoco. Joseph Percoco was Governor Cuomo’s Executive Deputy Secretary from 2011 to 2016, taking an eight month leave in 2014 to manage Governor Cuomo’s reelection campaign. The government charged Percoco with honest-services fraud (among other crimes) arising from two separate schemes: one that occurred entirely during his employ with the Governor, and one that began while he was managing the campaign but continued after he had signed and submitted the forms required for reinstatement as Executive Deputy Secretary following Cuomo’s reelection.14 The jury convicted Percoco of honest-services fraud arising from both schemes, but before the Supreme Court, Percoco challenged only his conviction on the latter.15 That scheme involved payments Percoco received from state contractors in July and October 2014 (while he was on leave) in exchange for his asking state officials in December 2014 to take actions that would benefit those contractors.16

A breach of fiduciary duty is an embedded element of honest-services fraud, because the right to honest services arises from a fiduciary relationship.17 The government’s theory of liability at trial and on appeal was that Percoco owed a fiduciary duty to the public throughout his leave from the government by virtue of his degree of influence as a former and potentially future deputy secretary; that he violated it by accepting payments in exchange for lobbying for private interests; and that this violation constituted honest-services fraud. A conception of fiduciary duty that extends to former public officials based solely on their continued influence, however, could sweep within its ambit the many lobbyists in state capitols and Washington who spend their careers traveling through the proverbial revolving door.

In its place, the prosecution had two alternatives. First, because Percoco’s conduct in December 2014 occurred after he had already formally filed all paperwork necessary for his reinstatement and just days before reinstatement became official, the prosecution could have limited its fiduciary claim to that time period, on the theory that Percoco was by then functionally a public official exercising clear public authority over government decision-making.18 Second, the prosecution could have argued that throughout his leave running Cuomo’s campaign, Percoco was functioning as Cuomo’s agent. Fiduciary duty would clearly lie under an agency theory,19 which had ample factual support in Percoco’s regular use of his Deputy Secretary office space, phone line, and administrative assistant; his continued participation in state operations and policy decisions, including instructing the Governor’s staff on noncampaign matters; and his authority to speak “for the Governor” on legislative matters.20 Either or both of these theories could have been argued (they were not contradictory) and, given the egregiousness of the facts, would have likely resulted in a conviction.

The facts of Ciminelli likewise offered a less capacious theory of liability that the government eschewed at trial and belatedly adopted before the Supreme Court (although the theory may have had some evidential hurdles). Louis Ciminelli, the owner of a construction company in Buffalo, New York, had engaged in a bid-rigging scheme with a state agent who controlled the processes used to manage contractor selection. Together, Ciminelli, the state agent, and others ensured that the state’s request for proposals contained a set of requirements that aligned uniquely with Ciminelli’s firm, thereby ensuring—unbeknownst to the state agents awarding the contract—that the supposedly competitive bidding process was in fact rigged to award the contract to Ciminelli. Ciminelli and the other schemers were charged with wire fraud, which prohibits use of the interstate wires in a scheme to defraud others of money or property. But Ciminelli did the work contracted for, at the bid price—so what money or property was deprived?

At trial, the prosecution argued that the fraudulent scheme deprived the state of the right to control its assets. Ciminelli was convicted under this right-to-control theory of property, which had been endorsed on multiple earlier occasions by the Second Circuit.21 But such a capacious definition of property would seem to run afoul of the Supreme Court’s recent efforts, in a series of cases, to rein in similarly broad conceptions of “property” for purposes of the mail- and wire-fraud statutes.22 Indeed, in one such case, the Court seemed to have addressed and rejected precisely the right-to-control theory on which the government in Ciminelli relied, stating that “intangible rights of allocation, exclusion, and control amount to no more and no less than [a state’s] sovereign power to regulate,” and did not give rise to a property interest.23

In place of that theory, the prosecution had available a more straightforward argument: the $750-million contract had been obtained by fraud, because Ciminelli’s bid falsely stated that his company had not “retained, employed or designated” any person or organization “to attempt to influence the procurement process.”24 In fact, in its briefs and at oral argument before the Supreme Court, the government sought to defend Ciminelli’s conviction on this theory, acknowledging that the right-to-control theory defined “property” too broadly.25 But the Court rightly rejected affirmance on a theory different from what was advanced at trial, and on which the jury had not been instructed.26

The prosecution’s strategic trial choices resulted in the Supreme Court tossing out both fraud convictions. In each case, the Court rejected the theory of liability offered by trial prosecutors as overly broad. “Without further constraint,” the Court held in Percoco, the jury instructions did not “define the intangible right of honest services with sufficient definiteness that ordinary people can understand what conduct is prohibited, or in a manner that does not encourage arbitrary and discriminatory enforcement.”27 “The right-to-control theory” of property, the Court held in Ciminelli, exceeded the “limited [] scope” of the federal mail- and wire-fraud statutes, “vastly expand[ing] federal jurisdiction without statutory authorization.”28

On the one hand, these prosecutorial failures might be contextualized as relatively inconsequential: not many public officials take temporary leaves to work on their superiors’ campaigns, and few circuits had expressly adopted the right-to-control theory in mail- and wire-fraud cases.29 But legal context is one thing and expressive value another. Those looking to monetize power or to purchase favored treatment will surely take from these rulings a broader message: federal criminal laws still offer them ample space to maneuver, so long as they couch their actions carefully. This is unfortunate, because in fact the federal fraud laws, more circumspectly applied, probably would have been found to bar these defendants’ particular actions. So why did prosecutors not proceed with greater restraint?

II. breadth’s lure

Percoco and Ciminelli are merely the latest in a series of opinions handily discarding prosecutors’ broad theories of liability in public-corruption cases. Insistence on broad theories in the face of repeated rejections might seem puzzling. But it is less so when one considers the dynamics and motivations at play in federal public-corruption prosecutions. Key among these factors is how prosecutors perceive their chances of success, as well as their role, in public-corruption cases. As the following Sections elaborate, broad theories enhance the likelihood of trial convictions and thereby enable a more proactive federal prosecutorial role in corruption enforcement.

A. Improving the Chances of Success

Once a prosecutor elects to file charges in any case, she works to obtain a conviction. But public-corruption cases present particular challenges that, from a prosecutor’s perspective, can make her chances appear slimmer. One key challenge is the nature of the proof, particularly in prosecutions of honest-services fraud. The corrupt officials and favor seekers engaging in quid-pro-quo transactions rarely do so openly or obviously; payments are well disguised, usually via third parties who perform some service (however nominal) in return. And often the defendants—officials or government contractors well versed in the intricacies and loopholes of the relevant legal prohibitions—construct their schemes in ways that will permit technical defenses. It was almost certainly not fortuitous that Percoco took the requested action in the days before he formally reentered public office, or that Ciminelli performed the contracted-for services at market rate. These decisions enabled each defendant to argue that technically, they did not run afoul of the law.30

Another challenge is the defense’s orientation. Generally speaking, defendants in public-corruption cases are well-resourced public figures whose career interests often are better served by fighting the charges rather than pleading guilty, and who can procure the most aggressive legal talent to take their cases to trial.31 As they put the finishing touches on their indictments, federal prosecutors can hardly be faulted for envisioning the likes of Abbe Lowell or Ted Wells at the defense table, or the local equivalent (and there usually is a local equivalent—a defense attorney widely respected by prosecutors for his or her formidable trial skills).32 And as these things go, a prosecutor gearing up for indictment will have almost certainly already heard from that attorney about the many evidential and legal weaknesses that will make conviction unlikely (and that, back in their day at the Justice Department, would have never resulted in charges!).33 Percoco and Ciminelli are cases in point: each was represented by multiple former federal prosecutors, among them Barry Bohrer (for Percoco) and Paul Schectman (for Ciminelli), each former chiefs in the Southern District of New York and highly respected members of the white-collar-defense bar.34

Finally, not only do prosecutors need to contend with evidential weak points, experienced legal talent homing in on them, and defendants willing to go the distance—but they must also do so under intense media scrutiny. Public-corruption cases typically generate keen media attention. No prosecutor wants an acquittal, but an acquittal under the media’s glare is all the more discomfiting.35

In the face of these trial pressures, broad theories of liability are attractive. By expanding the scope of charged conduct, they allow prosecutors to prove all of the distasteful actions undertaken by the defendant, and in so doing give the jury a fuller—and more offensive—picture. In contrast to narrower theories, which might require in-the-alternative arguments or a cobbling together of different theories to cover different conduct allegations, broad theories offer prosecutors a single, overarching narrative of illegality. Such an overarching narrative is not only more compelling to a jury, but also staves off defense accusations that the prosecution is just throwing spaghetti at the wall and seeing what sticks. Finally, broad theories eliminate, or at least lessen the persuasiveness of, technical defenses. They make it harder for even the most talented lawyers to defend on the facts, leaving the defense instead to attack the theory itself as legally insufficient. And while those arguments have been successful at the Supreme Court, they have had less sway at the circuit-court level.36 Because so few prosecutions are ever ultimately taken up by the Supreme Court, trial prosecutors are understandably focused on winning lower-court battles.

All of these incentives were present in the Percoco and Ciminelli cases. Had prosecutors proceeded on either or both of the narrower theories available in Percoco—agency or limiting to post-reappointment conduct—they would have forgone the opportunity to present the jury with a single, overarching narrative of a greed-generated abuse of public trust. Had they proceeded on the fraudulent-inducement theory in Ciminelli, they would have opened the door to a technical legal defense, namely, that the state did not rely on Ciminelli’s misrepresentations in designating his firm as a preferred developer.37 In both cases, narrower theories would have presented the jury with a less compelling argument of malfeasance, seemingly missing the forest for the trees. Given the favorable Second Circuit case law, the candid assessment of the Assistant Solicitor General during oral argument in Ciminelli was on the nose: under the circumstances, it is hard to expect trial prosecutors to resist the temptations of breadth.38

Of course, broad liability theories are not a prosecutor’s only response to outcome uncertainty; alternatively, she can decline to bring charges in the first place. The likelihood of conviction is among the factors prosecutors weigh in exercising charging discretion, and, perhaps unsurprisingly, the declination rate in corruption cases is high.39 Yet from the Supreme Court’s perspective, it should be even higher. Why might federal prosecutors sometimes opt for breadth over declination in public-corruption cases? Part of the reason lies in the federal prosecutorial role in these cases and its effect on enforcement discretion. The next Section considers that interaction.

B. Role and Discretion in Public-Corruption Enforcement

For most areas of criminal enforcement, federal prosecutors play a relatively minor, supporting role. Drug trafficking, firearms, episodic street crime, financial frauds, identity theft—while these crimes collectively constitute a majority of the federal criminal docket, that docket is a small portion of such prosecutions overall.40 In any of these areas, federal prosecutorial declination often leads to local or state prosecution (provided declination was for reasons other than evidential insufficiency).

Not so with public corruption. In cases involving federal officials, state crimes are often not an easy fit, and federal officials’ misdeeds do not typically implicate core local or state interests. In such cases, federal prosecution is the only viable route to criminal accountability. State and local corruption squarely implicates state law crimes such as fraud and bribery. But state and local chief prosecutors might have conflicts of interest, or at least their appearance, in investigating state or local officials.41 State and local prosecutors are also almost all elected officials themselves, beholden to political party organizations with vested interests in investigations involving public officials of the same or opposing party.42 What’s more, building public-corruption cases is complex and time-consuming, and the conduct involved, even if proven, does not typically harm citizens in direct, tangible ways.43 Local prosecutors, saddled with enormous caseloads, may not readily pursue resource-intensive investigations of potential crimes that do not immediately impact their constituents.44

The upshot is that public-corruption cases, whether they involve federal, state, or local officials, implicate core federal prosecutorial interests: holding federal officials accountable and ensuring effective prosecution of criminal conduct.45 In many such cases, a decision to decline federal prosecution will mean the public official is not ultimately held criminally accountable. And when it comes to public corruption, a lack of accountability is particularly pernicious because of the pressures unchecked corruption imposes on noncorrupt actors to join in making, or acceding to, corrupt demands.46 Unchecked corruption can quickly metastasize, putting at risk effective and legitimate governance.47

For a federal prosecutor choosing whether to prosecute or decline, these realities weigh heavily. Ultimately, they may counsel in favor of prosecuting, even in harder cases. And in those harder cases, the ones likely to go to trial, a broad theory of liability can help smooth the prosecutor’s way.

But the lure can sometimes prove disastrous—not only to the individual convictions reversed when the Court ultimately rejects such broad theories, but also to the broader corruption-mitigation project. The next Part elaborates on this systemic risk, and suggests an approach that might reduce it.

III. an alternative approach

This Part considers an alternative approach to the exercise of charging discretion in public-corruption prosecutions. Section A considers the harms of the status quo and argues that prosecutors are the institutional actor best suited to address them. Section B considers how an alternative approach to charging and theorizing liability might have played out in four public-corruption cases recently reversed by the Court. Section C proposes organizational reforms to the Department of Justice’s charging process in corruption cases.

A. Identifying the Problem

Some might agree with my description of prosecutorial dynamics in corruption cases but query whether it identifies a problem in need of fixing. Why, they might ask, should prosecutors trim their sails in public-corruption cases? Given the stakes for governance, don’t we want federal prosecutors to push the envelope on liability in these cases and leave it to the courts to reign in excess? Is the status quo not preferable to a more restrained prosecutorial approach, in which some seriously improper but close-to-the-legal-line conduct will be given a pass? And given the relative infrequence of certiorari grants, why should prosecutors shy away from aggressively pursuing serious misconduct? Others might locate the problem less with prosecutors than legislators: prosecutors are simply doing what one would expect prosecutors to do when gifted broadly worded statutes, and so the solution is more circumspect lawmaking.

Though reasonable minds will differ, I do see a problem with the status quo, and believe it best solved by prosecutors rather than legislators or courts. Broad laws and liability theories have obvious and well-documented drawbacks.48 But they also have important benefits. As Samuel Buell has argued, an upside of criminal overbreadth—a key reason Congress drafts broad penal statutes and prosecutors seek to apply them to novel contexts—is that it “fortif[ies] the perception that the social environment is one in which refusal to participate in largely voluntary compliance is costly.”49 Broad laws, in other words, help deter those who would circumvent more narrowly tailored restrictions. And while breadth risks overdeterrence—chilling socially beneficial conduct for fear of penal sanction—when it comes to corruption, overdeterrence is less of a concern. On balance, governance and public perceptions of it are better served when public officials err far wide of the illegality line than when they barely skirt it.50

We should not be too quick, then, to point the finger at legislators as the problem or the solution. But neither should we resign ourselves to the current dynamic in public-corruption cases, in which capacious theories of liability instigate hard judicial limits on the scope of federal prosecutorial authority.51 Those limits may not be announced with regularity,52 nor always with specificity.53 But they are harmful to the broader anticorruption project. Every ruling overturning a conviction narrows federal penal law’s reach over corrupt abuses of official power. Supreme Court decisions rebuffing broad liability theories both push back the line between illicit and licit corruption and draw it more clearly, making it easier for public officials to push right up to its limits. In turn, they make future power abuses more tempting to public officials who might have otherwise steered clear of the line. In this way, prosecutorial envelope-pushing can ultimately backfire.

Counterintuitively, prosecutors might best reap the benefits of broad laws by exercising greater prosecutorial restraint. Prosecutors have two alternatives to broad liability theories: they may proceed on a narrower theory (with the attendant risks of acquittal), or they may decline to prosecute entirely. Granted, acquittals and declinations don’t exactly advance the corruption-mitigation project. But they also don’t hinder it to quite the same degree as would a court’s outright rejection of a liability theory. Neither acquittals nor declinations set legal precedent. They are inscrutable, heavily fact dependent, and therefore difficult to extrapolate to future cases. Because there is effectively no basis to challenge them, they almost never generate appellate court decisions.54 In these respects, they do not announce to a broad audience, in the way judicial reversals do, the precise loopholes through which corruption can evade criminal sanction.

When deciding on charges and theories of liability to support them, then, prosecutors ought to think long and hard about the cost of Supreme Court narrowing. How might they think about exercising discretion in cases that carry this narrowing risk? The first step is to consider, from a broader regulatory perspective, what federal prosecutions of corruption ought to be for. The ultimate goal is not to find a theory through which individual corrupt acts are prosecutable federal crimes; the ultimate goal is to deter corruption.55 Sometimes broad theories of prosecution will serve that goal; but often enough, they will backfire.

The key task for federal prosecutors, then, is assessing whether prosecution of a given case is the best, or perhaps the sole, means to constrain the type of corruption the case involves. For most types of corruption, federal prosecution serves its greatest purpose when other avenues for accountability have proven unviable—when without it, the conduct at issue would remain both invisible to voters and irremediable through political process, state criminal process or other forms of regulatory accountability. These, broadly speaking, are the federal corruption prosecutions that merit the risk of possible appellate reversals. Upon identifying a case as worthy, prosecutors should then seek to minimize that risk by proceeding on the narrowest possible theory of liability likely to result in a conviction.

The following Section considers how this more purposive approach would guide prosecution of recent cases reversed before the Court.

B. A Thought Experiment: Recent Reversals Reconsidered

The corrupt abuses of power at issue in Kelly v. United States56 were well known to the public before a federal criminal investigation began. Two aides to New Jersey’s then-Governor, Chis Christie, had snarled traffic for days in Fort Lee, New Jersey in retaliation for its mayor endorsing Christie’s political opponent. Within four months of the aides’ decision to close two major traffic lanes, the press reported on a leaked email between the aides—sent upon learning of the endorsement—that it was “[t]ime for some traffic problems in Fort Lee.”57 In response to the media maelstrom that ensued, Christie denied involvement, fired the responsible aides, and ordered an internal investigation of his administration’s actions.58 New Jersey’s legislature opened an investigation and held public hearings; investigations were also opened by the Port Authority of New York and New Jersey and the United States Senate Committee on Commerce, Science and Transportation.59 Long before any indictments were ever returned, Christie’s approval ratings plummeted and his future political endeavors fizzled (although Christie was ultimately never charged in connection with the scheme).60

In the face of all this, was a federal criminal case against the two fired aides the best use of federal prosecutorial power—particularly in light of the capacious theory of liability needed to reconcile the conduct to federal statutes and the Supreme Court’s seeming aversion to it?61 Prosecutors charged Christie’s aides with wire fraud and federal-program fraud based on the theory that the “property” deprived (a requisite element under both statutes) was the state’s roadways, or employee labor required to close them, or both. In reversing the convictions, a unanimous Court distinguished property interests from regulatory interests and found the defendants’ conduct infringed only on the latter.62 The Kelly decision was hardly a surprise, given that an earlier decision had narrowly construed “property” under the mail-fraud statute,63 distinguishing a state’s license-issuing power as regulatory in nature and thus outside the statute’s ambit.64 In hindsight, the Kelly prosecution ultimately did more to undermine public-corruption enforcement than to enable it. Officials tempted to abuse state regulatory powers (and even to co-opt state employee labor to that end) can now do so without fear that their actions may constitute a federal crime.

Or consider the case of former Virginia Governor Bob McDonnell, who was charged with bribery in connection with a gifts-for-access scheme.65 McDonnell and his wife received lavish gifts and cash from the chief executive of a Virginia-based nutritional supplement company.66 In apparent exchange, McDonnell set up meetings between the executive and state officials to discuss state-funded research studies of one of the company’s supplements.67 The scandal was first broken by the Washington Post in March 2013, while McDonnell was still governor.68 In the wake of the news, McDonnell’s approval ratings plummeted, local prosecutors opened a criminal investigation, McDonnell was forced to abandon his designs on the presidency and, though McDonnell himself was term-limited, voters turned against his endorsed gubernatorial candidate, Ken Cuccinelli.69

McDonnell was being held publicly accountable for his corruption. Nevertheless, federal prosecutors proceeded to charge McDonnell and his wife with honest-services fraud and extortion charges (among others), pushing local prosecutors to bow out in the process.70 Federal statutes, however, were a poor fit for McDonnell’s transgressions. Prosecutors charged McDonnell with honest-services fraud and Hobbs Act extortion but defined those alleged crimes with reference to the federal bribery statute (a reasonable approach in view of the facts of the case and prior Supreme Court rulings).71 The problem for the government was that the conduct charged—essentially, providing access to state officials—did not clearly run afoul of the bribery statute’s terms as the Supreme Court had interpreted them in prior cases.72

Given the appellate risks to bringing such a case, an open investigation by local prosecutors, and the political price McDonnell was already paying, was a federal prosecution on a gifts-for-access theory of bribery the best choice? The prosecution ultimately resulted in a unanimous Supreme Court decision reversing the convictions of McDonnell and his wife and holding that providing access—setting up meetings, hosting events or making phone calls on a constituent’s behalf—does not constitute an “official act[],” payment for which is proscribed under the federal bribery statute. Perhaps McDonnell v. United States merely blessed the sort of purchased favors many politicians had long considered politics-as-usual. But in blessing them, the Court removed a statutory ambiguity that surely gave some politicians pause. After all, exchanging political favors for Rolexes, rides in Ferraris, and $175,000 in cash was “far from” the “normal political interaction between public officials and their constituents” at the time.73 Yet McDonnell clarified that these sorts of exchanges are permissible under federal law.

Now consider Percoco and Ciminelli again. The conduct leading to both prosecutions was uncovered by law enforcement, not the media. With respect to Ciminelli in particular, media organizations had attempted to glean information about the contractor selection process for the Western New York revitalization project for which Ciminelli’s firm was hired, but were rebuffed.74 Without the federal investigations and subsequent trials, the public would not have learned of the backroom deals and undisclosed payments from favor-seekers to Cuomo administration insiders.75 Bid rigging is notoriously hard to see and, when a state is mired in the practice (as the trials revealed), difficult for state prosecutors to confront. In short, the Percoco and Ciminelli cases involved the sort of corruption where federal prosecution can make a real impact; without it, accountability would be limited or nonexistent. Federal prosecution of these cases was worth the appellate risks.

But those risks could have been mitigated had prosecutors proceeded on narrower theories of liability. That is where the second step of the proposed approach comes in: once a case is determined worthy of federal prosecution, prosecutors should proceed on the most straightforward theory of liability the facts and law will allow—that is, one in which the jury instructions are not premised on (and so an eventual affirmance will not require the Supreme Court to adopt) capacious conceptions of statutory terms. In both Percoco and Ciminelli, such a route seems to have been available. Prosecutors could have pursued wire-fraud charges against Ciminelli on the more straightforward theory that his company’s $750 million contract had been obtained by fraud due to Ciminelli’s false statement in his bid application that his company had not “retained, employed or designated” any person or organization “to attempt to influence the procurement process.”76 And prosecutors could have limited their case against Percoco to actions he took while in public office and after having filed for reinstatement, or could have pursued fully out-of-office conduct on the theory that Percoco functioned as the Governor’s agent while running his campaign.

Prosecutors likely eschewed these narrower theories because they believed (correctly) that their broader theories would be upheld on immediate appeal, and because the broader theories undoubtedly gave them greater leverage at trial. With Ciminelli in particular, there were evidentiary challenges to proceeding on a fraudulent-inducement theory, which would have required proof that Ciminelli’s misrepresentation went to “the essence of the contract,” thus depriving the state of the contract’s fair value.77 Still, the added leverage these theories gave was not worth their high risk of reversal by the Supreme Court. In declining even to defend the government’s trial theory before the Court in Ciminelli, and in advancing the alternative soon-to-be-reinstated theory for affirmance in Percoco, the Solicitor General’s Office appears to have implicitly recognized this.

Prosecutors could have almost certainly convicted Percoco on either of the narrower theories, and there appeared to be sufficient evidence for a jury to find Ciminelli had fraudulently induced his contract.78 Had prosecutors gone those routes, they would have had a harder road at trial, to be sure. But failure at trial would have been less impactful, in the long run, than the reversals that ultimately ensured. And though on paper the unpursued theories remain viable against future influence-peddlers, the failure to bring them was a missed opportunity. Success under those theories at trial would have likely endured through every stage of review. And in enduring, it would have sent the deterrent message prosecutors had intended: that corrupt influence-peddling harms democracy, and federal prosecutors will hold those who engage in it to account.79 Instead, the takeaway from Ciminelli and Percoco is that officials can profit off their power and contractors can grease the wheels—so long as they put out just enough roadblocks to dissuade prosecutors from the straightforward path.

C. Operationalizing a More Purposive Approach to Federal Corruption Prosecutions

To be clear, my proposal is not aimed at curtailing federal investigations of public corruption; the feds should continue to diligently lift the hood on the workings of local, state, and federal government lest corruption escape notice and thus the criminal and noncriminal forms of accountability that might flow from it. My proposal, rather, is directed to the exercise of federal charging discretion, and to the construction of liability theories when federal charges are warranted. In that regard, it is important to assess the current charging process in federal public-corruption cases and how it might be improved upon.

Most federal public-corruption prosecutions are investigated and charged via the ninety-four U.S. Attorney’s Offices around the country.80 Main Justice’s Public Integrity Section handles limited categories of cases on its own,81 and it consults with U.S. Attorney’s Offices as a matter of course on any case involving bribery of federal public officials, election-related crimes,82 purchase or sale of federal public office, and any matter involving a sitting member of Congress or congressional staff.83 Lawyers from the Public Integrity Section may get involved in cases that are particularly sensitive or involve multiple prosecuting jurisdictions, or at a U.S. Attorney’s Office’s request.84 But for investigations of most state and local corruption, and even some federal corruption, the Justice Department’s investigation and prosecution of public corruption is decentralized.

This means that many decisions about when, whether, and how to charge a given instance of public corruption are made by individual Assistant U.S. Attorneys. Depending on the targets and the offenses, some of those decisions may be made with input from Main Justice, but that is often a prerogative of the U.S. Attorney’s Office (which is to say, of the Office supervisor or supervisors signing off on the decision to charge or decline).

More importantly, the Department’s Public Integrity Section is a trial unit, not an appellate unit. There is no requirement that U.S. Attorney’s Offices or the Public Integrity Section consult with either the Solicitor General’s Office (which handles all Department cases before the Supreme Court) or the Criminal Division’s Appellate Section (which supervises the Department’s appeals of adverse lower-court rulings and works closely with the Solicitor General’s Office) before making key trial-level decisions in a public-corruption matter. When trial attorneys must decide whether to charge a particular case, what theory or theories of liability to advance, and how to instruct the jury, input from experienced appellate attorneys—and in particular, Supreme Court litigators—comes only if those trial attorneys choose to ask for it.

In light of the setbacks the Department has encountered in public-corruption cases before the Court, the absence of regularized and early-stage Solicitor General’s Office input is a problem. Assistant U.S. Attorneys and prosecutors in the Public Integrity Section, both line attorneys and supervisors, are first and foremost trial lawyers. Their focus is on winning the cases before them and protecting those wins on immediate appeal; they are less inclined to worry about how the Supreme Court might view their case in the unlikely event it were to grant certiorari. By contrast, the Solicitor General’s Office is closer to the Court’s thinking. It is better positioned not only to read the tea leaves on how the Court might view a given liability theory, but also to weigh the systemic risk to the Department of advancing it. It can also suggest useful legal strategies for minimizing the risk of Supreme Court review and reversal.

To reduce the risk of Supreme Court reversal, the Department should institute a more formalized, centralized approval process in public-corruption cases, one incorporating the Solicitor General’s Office at the pre-charging, charging, and trial stages. Prosecutors in the districts sometimes chafe at the intrusion of Main Justice attorneys into their cases, but unlike their peers in Main Justice’s criminal trial sections, attorneys in the Solicitor General’s Office offer a perspective and expertise that trial attorneys necessarily lack. Assessments of the strengths or weaknesses of particular legal theories through that lens will help focus investigations on the most likely avenues for both trial and appellate success,85 and will guide more bulletproof construction of liability theories. Input on proposed jury instructions will further protect those constructions. The government’s belated shift in liability theories before the Court in both Percoco and Ciminelli86 is both a symptom of the current lack of early-stage Solicitor General’s Office involvement in public-corruption cases and an illustration of its potential benefits. Those alternative theories came too late; had they been successfully advanced at the trial stage, they may have saved the judgments from reversal.


In recent years, the Supreme Court has steadily limited federal penal law’s reach over public corruption. Federal prosecutors may disagree with the Court’s predilection in public-corruption cases—and they are not alone87—but they neglect it at their peril.

True, pursuit of statutory breadth in public-corruption cases does help the anticorruption cause for a time. Breadth, after all, comes not only from legislative drafting but also judicial interpretation, and successfully pushing novel theories at the circuit court level has certainly strengthened federal prosecutors’ arsenal. But in pushing too far, prosecutors have ceded those gains at the Supreme Court, at times leaving them in a weaker position than before their lower-court victories. Mounting losses at the Court continue to push back the line between licit and illicit corruption and draw it more clearly. This undermines the deterrent value of broad laws, taking a toll on the broader anticorruption project.

Counterintuitively, federal prosecutors can best leverage broad laws by using those laws more circumspectly. Key to this approach is to reassess where and when federal prosecutorial power is best put to use. Prosecutors should reserve their fire for corruption that is otherwise impossible to fully see, and therefore likely to go unchecked—whether by the political process, the state criminal process, or state or federal regulatory processes. And in the cases where federal prosecution is worth the risk of potential failure, prosecutors should aim to minimize that risk by constructing the narrowest theory of liability likely to result in conviction.

James E. Beasley Professor of Law, Temple University Beasley School of Law. Thanks to Miriam Baer and Dan Richman for helpful comments on prior drafts, and to the editors of the Yale Law Journal Forum for their terrific editorial feedback.