What Is An Agreement? A Central, and Hard to Define, Concept in US White-Collar Criminal Law

A central question in many white-collar criminal cases – ranging across conspiracy, bribery, kickback, and price-fixing cases – is whether or not there is an agreement. In Sherman Act antitrust cases the issue is fundamental: the statute specifies that the crime is the agreement. So, too, with the conspiracy statute. And in honest services fraud prosecutions, now limited by Skilling v United States, 561 U.S. 358, 412 (2010) to “encompass only bribery and kickback schemes”, once again the question of whether or not there is an agreement, a quid pro quo, is elemental.

For decades, courts in such cases have struggled to cleave fact patterns that evince an agreement from those that do not. At root, the difficulty centres around the so-called implicit agreement. Take United States v Penn, et al, Case No 20-cr-0152 (D.Colo. 2020), a multi-defendant Sherman Act price-fixing case in the USD95 billion per year broiler chicken industry tried three times in the District of Colorado  in 2021 and 2022 (the first two trials ended in deadlocked juries for all defendants; the third trial resulted in acquittals for all defendants). The evidence showed that the defendants’ prices often fluctuated in similar ways, that they had phone calls with each other, and that, at times, the defendants even shared pricing information. The defendants denied that there was any agreement, pointing to market factors to explain price movements, and arguing that communication among competitors was a normal business practice.

The Court in Penn instructed the jury on implicit agreements, explaining that “the evidence need not show that the members of the conspiracy entered into any express, formal, or written agreement… [t]he agreement itself may have been entirely unspoken” (Case No. 20-cr-0152, Dkt. 1421 at 25). But the Court then expended close to ten pages across four instructions in an attempt to convey the full universe of different facts and circumstances that might establish an implicit agreement. Many of the instructions were equivocal or even contradictory. For example, the jury was instructed that pricing information had been admitted “to assist you in deciding whether a defendant entered into an agreement” but just a few sentences later was admonished that “[t]he mere fact that some or all of the defendants may have engaged in similar or parallel pricing does not by itself establish the existence of a conspiracy among them.” This was clearly an attempt to instruct the jury based on established legal principles but, respectfully, how can a jury be reasonably expected to make sense of such inherently confusing guidance?

To understand the reach of this conundrum, think further about bribery, now, post-Skilling, the heart of honest services cases. As the Second Circuit has explained, “it is the requirement of an intent to perform an act in exchange for a benefit—i.e., the quid pro quo agreement—that distinguishes [bribery] from both legal and illegal gratuities” (United States v Ganim, 510 F.3d 134, 146–7 (2d Cir. 2007)). As a result, understanding the limits of such cases first requires an understanding of the sorts of conduct that can rightly be described as agreements – the same problem as in antitrust cases.

One need not look far to see further examples of courts wrestling with the unsettled meaning of an agreement. The Supreme Court tried to bring some clarity to the issue in the seminal civil antitrust case Bell Atl Corp v Twombly, holding that “showing parallel conduct or interdependence, without more, mirrors the ambiguity of the behavior: consistent with conspiracy, but just as much in line with a wide swath of rational and competitive business strategy unilaterally prompted by common perceptions of the market” (550 U.S. 544, 554 (2007)). But the upshot was somewhat circular: what is the “more” in that context? “[E]nough fact to raise a reasonable expectation that discovery will reveal evidence of illegal agreement” (at 556).

Similarly, in Evans v United States, 504 U.S. 255, 257 (1992), the Court held that an agreement could be shown by evidence of an “implicit promise”. But several years later, in McDonnell v United States, the Court tightrope walked around the definition of an “official act” in 18 U.S.C. § 201 to avoid holding that “nearly anything a public official accepts—from a campaign contribution to lunch—counts as a quid; and nearly anything a public official does—from arranging a meeting to inviting a guest to an event—counts as a quo” (579 U.S. 550, 575 (2016)). The Court in McDonnell defined “[a]n ‘official act,’ for purposes of the federal bribery statute… [as] a decision or action on a question, matter, cause, suit, proceeding or controversy, and the question, matter, cause, suit, proceeding or controversy… [that] must involve a formal exercise of governmental power that is similar in nature to a lawsuit before a court, a determination before an agency, or a hearing before a committee; it must also be something specific and focused that is pending or may by law be brought before a public official” (at 574). Even with this definitional guidance, the questions of what exactly is a formal exercise of governmental power, and how specific the tie between the quid and the quo must be are inherently difficult to grasp. Indeed, neither case addressed the meaning of an agreement head on – at most, they provided guideposts.

As one would expect, these porous definitions have created headaches for district courts. In United States v Silver, the Second Circuit found (harmless) error in an honest services fraud case where a district court “only required the jury to find that [defendant] understood, at the time that he accepted any quid, that he was expected to exercise official influence or take official action for the benefit of the payor… [For] an illegal quid pro quo under the ‘as the opportunities arise’ theory of bribery requires more than… an open-ended promise to perform official actions ‘for the benefit of the payor’” (948 F.3d 538, 559 (2d Cir. 2020)).

The Supreme Court emphatically reinforced this point in Percoco v United States, finding reversible error where the court instructed the jury that it could convict a defendant of conspiracy to commit honest-services fraud based on his accepting payment to take nebulous official action to benefit the briber “as opportunities arose” (598 U.S. 319, 332 (2023)). Before being reversed on other grounds, the Second Circuit also concluded that the instruction was “‘too open-ended’ because it failed to convey that the defendants could not be convicted of honest-services fraud unless they promised to undertake official action on a specific question or matter as the opportunities arose.” (United States v Percoco, 13 F.4th 180, 190 (2d Cir. 2021) (citing Silver, 948 F.3d at 569).)

There is one suggestion: what if courts explicitly instruct juries on a sliding-scale of factors they should consider in deciding whether the government has proved the existence of an implicit agreement? These factors could include, at minimum:

The key here is that these factors will allow courts to move away from trying to summarise the wide range of factual scenarios that have been found sufficient to show the existence of an implicit agreement. Instead, courts can equip jurors with a list of relevant considerations and step aside. This proposal is no panacea – courts will still struggle with the timeless proposition that it is difficult to define concepts that are difficult to define – but perhaps this approach might simplify the jury instruction process and help juries in these cases grapple with the ephemeral concept of what constitutes an agreement.