Blog
Cooperation in Specific Case Types
Cooperation is not one thing. The word appears in proffer agreements and plea negotiations, in corporate enforcement policies and regulatory settlements, and in each context it carries a different weight, a different set of risks, and a different calculus of what the government requires in exchange for what it is willing to offer. The type of case determines the structure of the cooperation. A business owner facing an antitrust investigation inhabits a different procedural world than one facing a healthcare fraud inquiry, even if both are told by counsel that cooperation is advisable. The architecture of cooperation shifts with it.
The New Department-Wide Enforcement Policy
In March 2026, the Department of Justice released its first unified Corporate Enforcement and Voluntary Self-Disclosure Policy. The policy establishes a single framework for how prosecutors evaluate voluntary self-disclosure, cooperation, and remediation across all corporate criminal matters. Companies that disclose misconduct early, cooperate in full, and remediate can receive a declination, meaning the Department agrees not to prosecute. The policy applies to all corporate criminal matters across the Department, with one significant exception: antitrust.
The Antitrust Division retains its own Leniency Policy, a framework that has operated since the early 1990s and that functions on principles distinct enough to resist consolidation. For an executive or business owner attempting to understand what cooperation means in their case, the first question is jurisdictional: which division, which agency, which enforcement regime governs the conduct at issue. The answer determines the rules of the cooperation, the scope of what must be disclosed, and the nature of what can be gained.
That exclusion is not incidental.
The Race That Only One Company Wins
Antitrust leniency operates on a principle that no other federal enforcement regime shares: there is one grant of immunity per conspiracy, and only one. The Antitrust Division awards leniency to the first corporation that self-reports its participation in a cartel, and the competition for that position has produced what practitioners refer to as a race to the courthouse. The race can be decided by hours. A company that arrives second receives cooperation credit, but not immunity. Subsequent cooperators receive smaller fines and expose fewer executives to indictment, but the distance between first and second is substantial.
The mechanism that initiates this process is the marker. Counsel contacts the Division, identifies the client and the nature of the conduct, and requests a hold on the leniency slot while the company conducts an internal investigation. We have seen three cases in the past eighteen months where the marker window was extended, and two where it was not.
In antitrust, the company that cooperates can shield its executives; in most other federal investigations, the company cooperates in order to expose them. Type A leniency, available when the Division has not received information about the activity from any other source, guarantees immunity for all directors, officers, and employees who admit involvement and cooperate. Type B leniency, available after an investigation has commenced, offers protection to the corporate entity but leaves the coverage of individuals to the Division’s discretion. The distinction is temporal, and the temporal window is not announced. The marker is good for thirty or forty-five days, sometimes extended, sometimes not.
Cooperation in FCPA and Bribery Investigations
In a bribery case, the arithmetic reverses. The company cooperates against its own people. The Department’s stated first priority is to prosecute individuals whose decisions produced the misconduct, and the company earns its cooperation credit by facilitating that prosecution. This is the structural logic of the Foreign Corrupt Practices Act enforcement framework, and it creates a tension that antitrust leniency does not: the individuals whose conduct gave rise to the investigation are the very people the company must identify, investigate, and present to the government.
Under the revised Corporate Enforcement Policy (which, before its department-wide expansion, originated in the FCPA context as a pilot program in 2016), a company that disclosed bribery to the DOJ before the conduct came to the Department’s attention, cooperated in full, and remediated could expect a declination. The first resolution under the new department-wide policy, announced just weeks ago, involved a French medical device manufacturer (which, it should be noted, had restructured its compliance program and terminated the business relationships at issue before the Department had completed its own investigation, a sequence that suggests the company’s internal timeline moved faster than the government’s, and that this speed was credited). The company that disclosed the bribery and cooperated received a declination; the individuals who carried it out were indicted.
Healthcare Fraud, Securities Enforcement, and the Regulatory Agencies
Healthcare fraud cooperation resembles the FCPA model in its broad strokes but differs in the details that determine outcomes. The Department’s Health Care Fraud Unit applies the same Corporate Enforcement Policy, but the practical dynamics of healthcare cases introduce variables that bribery cases do not. The data is voluminous. The regulatory overlay (Medicare, Medicaid, CMS reporting obligations) means that cooperation with DOJ often runs in parallel with obligations to other agencies. And the misconduct itself frequently involves systems and processes rather than discrete acts, which complicates the narrative a cooperating company must construct.
A resolution from last year involving a healthcare technology company illustrates the point. The company did not receive credit for voluntary self-disclosure under the CEP because the disclosure was made to CMS, not to DOJ, before the conduct reached the Department’s attention. The distinction is procedural, but it determined the range of available outcomes. The company received a twenty percent discount off the bottom of the applicable Sentencing Guidelines fine range, a reduction that reflected both the value of its cooperation and the limitations of it. The failure to preserve and produce key documents in the early stages of the investigation prevented the company from receiving the maximum reduction. The result was an eighteen-month non-prosecution agreement, which in the context of healthcare fraud enforcement is a considered outcome.
Every cooperation decision is, at its foundation, a wager on information. What one can offer, and when, determines the return.
Securities enforcement operates through a different institutional structure. The SEC maintains its own Cooperation Program, distinct from DOJ, with its own menu of formal agreements: cooperation agreements, deferred prosecution agreements, and non-prosecution agreements. The Commission’s analytical framework for evaluating cooperation traces to the Seaboard Report of 2001, which articulated factors the Commission considers in determining leniency for investigated companies. I am less certain about how the new department-wide DOJ policy will interact with the SEC’s own framework than the preceding analysis might suggest; the two regimes have historically operated in parallel, and parallel operation produces gaps. What we have observed, in cases involving both the SEC’s formal program and its informal credit, is that the formal agreement tends to offer more predictability but less flexibility. The formal program and the informal credit exist in parallel, and which one applies depends on facts that are often unclear at the moment one must decide.
What a Proffer Feels Like from the Other Side of the Table
Before any cooperation agreement is executed, before any commitment is made, there is the proffer. The session takes place at the United States Attorney’s office. The defendant sits with counsel. Across the table are the prosecutor and one or more federal agents. The agents ask questions. They already know some of the answers, and they use the known answers to calibrate the truthfulness of the unknown ones. The proffer agreement, sometimes called a queen-for-a-day letter, provides that statements made during the session cannot be used directly at trial.
The protections of that letter are narrower than the phrase suggests. The statute is not entirely clear on the protections the proffer letter affords, which is part of the problem. A defendant who mentions the cooperation to a fellow inmate, a defendant who misstates a date or a figure by enough margin to raise doubt, a defendant whose account fails to corroborate on some detail the government already possesses: any of these can cause the entire cooperation to collapse.
It is, if we are being precise, not an agreement at all but an audition. The government evaluates whether the defendant’s information is valuable enough to warrant a cooperation agreement, and the evaluation happens while the defendant is already providing the information. The room is a controlled environment, and everything said in it is recorded and evaluated regardless of what the letter provides. The derivative use exception means the government can use what it learns from the proffer to discover other evidence, even if it cannot introduce the proffer statements themselves. The information the defendant provides can be used to derive other evidence, to impeach the defendant if later testimony contradicts the proffer, and to inform the probation department for sentencing purposes.
We begin the cooperation analysis before the first proffer session, not after. The preparation involves constructing an inventory of what the client knows, assessing how that information maps onto what the government is likely pursuing, and identifying the points where the client’s account is vulnerable to the kind of minor inconsistency that, in the Southern District in particular, prosecutors treat as a credibility event rather than a factual correction. Some offices approach the proffer as a fact-finding exercise; others approach it as a test. The distinction matters, and it is not always apparent in advance.
And there is the question that the proffer itself cannot resolve. If the government decides the cooperation was not substantial, no Section 5K1.1 motion will be filed. The defendant cannot compel it. The Second Circuit held in United States v. Knight that where the cooperation agreement grants the government sole and unfettered discretion over the filing decision, judicial review is limited to unconstitutional motive. Whether the court intended the rule to function this way, or whether it simply failed to prevent this outcome, is a question that has not received a satisfying answer.
Timing and Sequence
The value of cooperation degrades with time. A defendant who cooperates within the first ninety days of an investigation provides information that shapes the case; a defendant who cooperates after the indictment provides information the government may already possess. In something like forty percent of the cases we have handled, though the sample is not scientific, the difference between early and late cooperation corresponded to a difference of several years in the ultimate sentence. The government weighs timeliness as one of the five factors under Section 5K1.1, and in practice it functions as a multiplier on every other factor.
In multi-defendant cases, the dynamic compounds. The first defendant to cooperate provides the narrative; the second provides confirmation. Confirmation is worth less. That much, at least, is consistent across every case type.
The particular enforcement regime, the particular agency, the particular moment in the investigation: these determine whether cooperation is a path to a declination or a path to a slightly shorter sentence. The gap between those two outcomes is the gap between retaining a business and losing one. What one knows, what one can prove, and what the government needs at the moment one offers it: these determine the outcome more than any policy framework.
The decision to cooperate cannot be reversed once it begins. The first conversation about that decision should be with counsel who has conducted the analysis specific to the case type, the agency, and the current enforcement posture. A consultation is where that conversation begins.