Undue Influence: Defining the Scope of Legal Marketing under EKRA

Picking is your poison.

In 2018, the Eliminating Kickbacks in Recovery Act (EKRA) signified a massive change in how diagnostic labs and addiction treatment providers can market their services. Under EKRA, it is a federal crime for such companies to pay for referrals of patients, even those with private insurance, and even when the payments are made to bona fide employees of the payer.

But what is a "referral?" If a lab pays a doctor $1,000 for each patient sample the doctor sends to it, that would qualify. If the lab pay a marketing firm to put advertisements in an industry journal, probably not. Where exactly the line lies is enormously important for companies in these industries, but EKRA's text does not say, and there have been almost no court decisions interpreting EKRA, including as to this issue.

The Ninth Circuit may soon change that. A single panel of that court is currently considering a pair of cases raising the definition of "referral." In S&G Labs v. Graves, Case No. 19-cv-310 (D.Hawai'i Oct. 18, 2021), a lab paid a marketer to persuade doctors and addiction treatment providers to send patient samples to the lab for urinalysis. Similarly, in U.S. v. Schena, Case No. 20-cr-425 (N.D. Cal. May 28, 2022), a lab hired marketers to promote its Covid-19 testing services to allegedly "naïve" doctors, using purportedly misleading representations about the lab's capabilities.

In both cases, the parties have addressed on appeal whether the arrangement entailed payment for a referral, or simply for legitimate "marketing" services. There is no allegation that any compensation went to doctors.

To define "referral," the court may adopt the "undue influence" test developed to answer the same question in the older Anti-Kickback Statute (AKS), which applies to all health care companies, but only as to patients insured by federal government programs. 

For example, in U.S. v. Polin, 194 F.3d 863 (7th Cir. 1999), the defendant was paid to "recommend" pacemaker monitoring services for patients, and his recommendations were essentially always adopted by the treating doctors. Similarly, in U.S. v. Shoemaker, 746 F.3d 614 (5th Cir.), a nurse staffing business paid a hospital chairman and CEO to arrange for the nurses to service the hospital's patients. In both cases, the courts held that the payments were for "referrals," as they went to individuals with effective "decision-making" power or "undue influence" over who ended up providing and billing for the services.

By contrast, in U.S. v. Miles, 360 F.3d 472 (5th Cir. 2004), a home healthcare provider paid a marketing firm to merely provide promotional materials to, and arrange "informational" sessions with, prescribing doctors. And more recently, in U.S. v. Sorensen, Case No. 24-1557 (7th Cir. April 14, 2025), a supplier of orthopedic braces paid marketers to publish ads, collect information from patients who responded, and send those patients' information to doctors, a minority of whom prescribed the braces. In both cases, the courts found that the marketers who received the payments did not have "undue influence" over the ultimate decision of who would supply the patients; that decision remained in effect with the treating physicians.

The Fifth Circuit's decision in U.S. v. Marchetti, 96 F.4th 818 (5th Cir. 2024) drew the line between two different schemes within the same case. The defendant marketer promoted one laboratory's testing services through advertising and marketing strategy. But in a second arrangement, the marketer actively determined which lab the patient samples would be set to. The court held that the first scheme did not involve referrals and was thus legal, but that the second involved unlawful referrals under the AKS. Again, the difference came down to the marketer's control over the patients' samples.

At the oral arguments in S&G Labs and Schena, both held the same day, the Ninth Circuit panel (Thomas, Bress, De Alba) questioned counsel specifically about the issue, at some points specifically discussing Miles, Marchetti, and other AKS cases. There appeared to be a consensus that the "undue influence" test from those cases applied. If the court addressed the "referral" issue in its decision, it should be an important milestone in EKRA litigation going forward, and will provide guidance both to defendants in EKRA investigations and prosecutions and to labs or SUD treatment providers seeking to ensure that their marketing operations are compliant.

A Bridge Too Far?: New Supreme Court Decision May Undermine DOJ's Assault on DEI

Not quite there.

The DOJ's new Civil Rights Fraud Initiative plans to use False Claims Act lawsuits against universities with DEI programs, claiming their federal funding certifications are fraudulent. But the Supreme Court's recent decision in Kousisis v. United States, and particularly Justice Thomas's concurrence, suggests this strategy faces a fatal flaw: civil rights compliance appears immaterial to research grants' "fundamental purpose."

How is race-neutrality like painting a bridge? What sounds like a riddle is actually the question underlying a new Supreme Court precedent and an ambitious new Department of Justice program.

The Supreme Court's recent decision in Kousisis v. United States couldn't have come at a more interesting time. Just as DOJ rolls out its new Civil Rights Fraud Initiative—wielding the False Claims Act against universities and contractors who allegedly discriminate through DEI programs—the Court has provided a refresher course on fraud's materiality requirement. And buried in Justice Thomas's concurrence is a discussion that should make DOJ's civil fraud lawyers very nervous.

What Did the Supreme Court Decide in Kousisis?

Let's start with what happened in Kousisis. Stamatios Kousisis and his company Alpha Painting secured $85 million in contracts to restore Philadelphia landmarks (including the Girard Point Bridge) by promising to use disadvantaged business enterprises (DBEs) for their paint supplies. Instead, they used a DBE as a mere pass-through—a paper pusher that tacked on a few percent while the real suppliers did the work. When caught, they argued that PennDOT got what it paid for: quality paint jobs at the agreed price. No economic loss, no fraud, right?

Wrong, said the Court. In rejecting the defendants' proposed economic-loss requirement, Justice Barrett's majority opinion confirmed that you can commit fraud even when delivering full value. Promise Yankees tickets and deliver Mets tickets of equal worth? That's still fraud if the misrepresentation was material.

But here's where it gets interesting for DOJ's new initiative. While the majority reserved the question of materiality's proper standard—the defendants hadn't contested it—Justice Thomas dove right in. And what he found should give pause to anyone planning to build a fraud prosecution empire on civil rights compliance certifications.

Thomas examined whether DBE requirements could meet the "essence of the bargain" materiality test from Universal Health Services v. Escobar. His key insight? The DBE provisions were "irrelevant to the contracts' fundamental purpose—bridge repair." PennDOT was buying paint jobs, not social policy. The work was completed satisfactorily regardless of DBE compliance. As Thomas noted, the DBE conditions "had no bearing on petitioners' ability to complete their projects."

How Does This Affect DOJ's Civil Rights Fraud Initiative?

This observation cuts to the heart of DOJ's Civil Rights Fraud Initiative. According to Deputy Attorney General Todd Blanche's memorandum, the Initiative will use the FCA to target federal funding recipients who certify compliance with anti-discrimination laws while allegedly "engaging in racist preferences, mandates, policies, programs, and activities." Universities are explicitly in the crosshairs.

But apply Thomas's logic here. When the National Science Foundation awards a grant for particle physics research, what's the "essence of the bargain"? Discovering new particles or ensuring the university doesn't use race-conscious admissions? When NIH funds cancer research, is it buying medical breakthroughs or HR policies? When the Department of Transportation contracts for highway construction, does it fundamentally care about concrete and asphalt or the contractor's diversity initiatives?

Under Thomas's framework, the answer seems clear: the government is buying research, roads, and services—not social policies. Civil rights compliance, however important as a policy matter, appears peripheral to these contracts' fundamental purposes.

How could Kousisis apply in practice?

Consider how this plays out. A university receives millions in federal research grants while maintaining allegedly discriminatory DEI programs in admissions or hiring. Under the Civil Rights Fraud Initiative, DOJ would argue the university's compliance certifications were fraudulent. But was Title VI compliance material to a grant studying quantum mechanics? Under Thomas's analysis, probably not. The "essence of the bargain" was quantum research, not the university's admission policies.

The government might counter that federal funds shouldn't support any discriminatory institutions, period. Fair enough as a policy position. But Thomas's concurrence suggests that policy preferences don't automatically translate into materiality for fraud purposes. If they did, the government could make any political priority "material" simply by adding it to funding certifications. Today it's DEI policies, tomorrow it could be environmental practices, next week labor relations. Where would it end?

Justice Sotomayor, concurring only in the judgment, tried to distinguish the DBE requirements in Kousisis by emphasizing that PennDOT explicitly made them material terms and faced potential federal sanctions for noncompliance. But this won't save DOJ's theory. First, most federal grants include boilerplate compliance certifications without making them explicit payment conditions. Second, and more fundamentally, Thomas's point is that regulatory requirements don't become material just because the government says so—they must relate to the contract's core purpose.

The timing is exquisite. Just as DOJ prepares to unleash the FCA on universities and contractors over DEI, the Supreme Court has reminded us that materiality is "demanding" and can't be satisfied by regulatory requirements unrelated to a contract's fundamental purpose.

How will this work across various federal projects?

This doesn't mean DOJ's initiative is dead on arrival. Some funding relationships might involve civil rights compliance as a core purpose—grants specifically for diversity programs, for instance, or contracts where discrimination would directly impair performance. But the vast majority of federal funding goes toward concrete deliverables: research results, construction projects, healthcare services. Under Thomas's framework, civil rights compliance seems incidental to these transactions.

The real lesson from Kousisis may be this: if you're going to build a fraud prosecution strategy around regulatory compliance, you'd better make sure the regulations relate to what the government is actually buying. Otherwise, you might find yourself explaining to a skeptical judge why a university's admission policies matter to its ability to sequence genomes or why a contractor's diversity training affects its capacity to pave highways.

As Justice Thomas warned, a demanding approach to materiality is "all that prevents an 'extraordinarily expansive view of liability' from rendering the federal wire-fraud statute nearly limitless in scope." DOJ's Civil Rights Fraud Initiative seems designed to test those limits. Thanks to Kousisis, we now know where at least one justice would draw the line: at the contract's fundamental purpose. And when that purpose is research, construction, or services—not social policy—civil rights compliance may be a bridge too far.

Not All Quids Are Created Equal

Keep spreading the news.

DOJ’s proposed dismissal of the criminal bribery indictment of New York Mayor Eric Adams, in return for his agreement to help with the Trump administration’s immigration crackdown, has drawn criticism from various quarters. Many of those have been on the right, including not only the conservative AUSAs who resigned in protest, but also commentators from the Wall Street Journal, the National Review, and others who usually back Trump’s positions—particularly on immigration.

But others, in defense of the decision, have analogized the arrangement to a typical DOJ cooperation deal with a criminal defendant. The typical rhetorical question: doesn’t DOJ agree to drop charges against defendants all the time in return for their help with prosecutions or other law enforcement proceedings?

The answer: Yes, but there is a fundamental difference here. Which explains why Adams, his counsel Alex Spiro, and Acting Deputy Attorney General Emil Bove have so strenuously denied that there was any such quid-pro-quo deal.

There are several situations in which federal prosecutors may cut a deal with a criminal target or defendant in return for cooperation. The most common is under a "cooperation plea agreement," otherwise known as a "5K" deal. The defendant agrees to plead guilty, to provide information to the government about others involved in the crime or about separate crimes, and to testify about those crimes and other persons in trial or before a grand jury, if necessary.

In return, the government agrees that it will tell the sentencing judge about the defendant's cooperation and, if the cooperation has been truthful and valuable, recommend that the judge reduce the defendant's sentence accordingly. The formal method for this recommendation is via a motion for a sentencing reduction under Section 5K1.1 of the Federal Sentencing Guidelines; hence the term "5K." The judge does not have to agree to lower the sentence, but usually will, because otherwise it would be hard to persuade future defendants to accept such deals.

In some cases, in return for cooperation, the government may agree not to charge the cooperator at all. This may involve a Deferred Prosecution Agreement or a Non-Prosecution Agreement, or in some cases granting the cooperator immunity from prosecution, as long as his or her testimony is truthful.

And in other cases, law enforcement agencies may actually put suspects on its payroll, having them work as confidential informants to provide information on ongoing crimes, in return for money, or perhaps informal promises not to pursue charges against them.

But there is a major difference between those situations and the deal that DOJ apparently cut with Adams. Ordinary, legitimate cooperators offer to act in their individual capacities--providing information or testimony--in return for the benefit of avoiding or reducing criminal liability.

Adams is a public official, who was elected to act on behalf of the people of New York, and use his powers only for their benefit. The agreement he apparently made here was to take actions in his official capacity--using his power as Mayor to influence city immigration policy, such as by opening the Rikers jail to ICE--in return for personal benefits (i.e., having the charges dropped). And the deal was also structured so that if he did not use his official power in that way, the charges could be re-filed.

Agreeing to act on your own behalf in return for a benefit to you is generally legal. Agreeing to use your official powers in return for a personal benefit is not. Ordinarily, DOJ does not enter into such deals--it prosecutes them. Indeed, that is what Adams was originally charged with.

Think of it another way: If someone were to offer Mayor Adams $10,000 in return for his taking some official action to help the payer—say, by granting a building permit—few would claim that was allowed. The illegality would lie in Adams using his position of public trust for private benefit. Why would it be any different if the benefit to Adams is his freedom rather than cash?

The issue is not that DOJ is trading off one federal law enforcement interest for another. It is that Adams, with DOJ’s help, is trading his office for a free pass.