Blog
The Cooperation Process
The Cost of Candor
Cooperation with the federal government is the single most consequential decision a defendant will make, and it is nearly always made too early. The word itself carries an institutional warmth that obscures what the process demands: a complete and irreversible surrender of information, delivered under conditions the government designs, in exchange for a benefit the government is not obligated to provide.
Most defendants who cooperate do so because their attorney has assessed the evidence as overwhelming. Federal conviction rates remain above ninety percent in most districts, a figure that leaves little room for optimism about trial. The calculus is not complicated. What is complicated is everything that follows.
The Proffer Agreement
Before the first meeting, before any information changes hands, the defendant signs a document the government calls a proffer agreement. It is sometimes referred to as a “Queen for a Day” letter, a name that communicates the wrong thing about the arrangement. The letter is not a grant of immunity. It is not a plea agreement. It is a one-sided contract, prepared entirely by the government, that establishes a narrow corridor of protection and a wide field of exposure.
The protection: statements made during the proffer session cannot be used in the government’s case-in-chief at trial. The exposure encompasses everything the protection does not. The government retains the right to pursue any investigative lead the defendant’s statements suggest. If those leads produce independent evidence, that evidence is admissible against the defendant. If the defendant later testifies at trial and says anything inconsistent with what was said in the proffer, the statements become available for impeachment. If the government concludes that the defendant was less than fully truthful, the agreement collapses, and the information disclosed can form the basis of a prosecution for false statements under 18 U.S.C. § 1001.
The proffer agreement contains no promise of a particular outcome.
The contract sat in a government filing cabinet the way a deed sits in escrow: technically in someone’s possession, operationally in limbo. The defendant had signed it. The government had signed it. Neither party yet understood what it would cost.
In the Southern District of New York, the expectation has long been that a cooperating defendant will disclose all criminal conduct throughout their lifetime as a precondition to the agreement. Not the conduct related to the charged offense. All of it. Every jurisdiction calibrates this differently, and the variance is not published in any manual a defendant is likely to consult. But in districts that follow the SDNY model (who, it should be noted, had formalized their own Corporate Enforcement and Voluntary Self-Disclosure Program just weeks ago, with cooperation requirements that extend three years past the final declination), the scope of that disclosure is something many defendants do not fully appreciate until they are seated across the table.
What Happens in the Room
The session itself takes place at the United States Attorney’s office or the local field office of the investigating agency. Present are the defendant, defense counsel, the assigned prosecutor, and one or more federal agents. In most districts, the session is not recorded, though agents take notes that become part of the government’s permanent file.
The government will ask questions the defendant expects. It will also ask questions the defendant does not expect, because the government’s interest extends past the boundaries of the immediate case. A proffer session is not a conversation about one investigation.
The silence after a difficult question in a proffer session is longer than it should be. The defendant looks at counsel. Counsel cannot answer for them. The agent and the prosecutor wait. In that silence, the defendant is deciding not just what to say but what to become: a person who has given the government everything, or a person who has given the government almost everything, which the government treats as equivalent to nothing at all.
I have sat through enough of these sessions to recognize the moment a client understands that cooperation is not a negotiation. It is a disclosure. The government is not asking what you would like to share. The government is asking what you know, and the agreement you signed requires the answer to be complete.
Whether the government always exercises this discretion with the restraint the process implies is a question worth sitting with.
Section 5K1.1 and Prosecutorial Discretion
The reward for cooperation, when it arrives, takes a specific procedural form. Under Section 5K1.1 of the United States Sentencing Guidelines, the government may file a motion stating that the defendant provided substantial assistance in the investigation or prosecution of another person. If the court grants the motion, it may impose a sentence below the applicable guideline range.
Two features of this mechanism deserve close attention. The first is that only the government can file the motion. The defendant cannot request it. Defense counsel cannot compel it. The court cannot initiate it on its own. If the government determines that the assistance was not substantial, no motion will be filed, and the sentence proceeds as calculated.
The second feature is more technical and, for the defendant facing a mandatory minimum, more consequential. A 5K1.1 motion, standing alone, permits the court to go below the guidelines but not below a statutory minimum. To descend below the mandatory floor, the government must file a separate motion under 18 U.S.C. § 3553(e). The government sometimes files one and not the other. The distinction matters. A defendant who cooperated in the expectation of a sentence below a ten-year mandatory minimum and receives, instead, a sentence of exactly ten years has learned the distinction in the most concrete way available.
The five factors the court weighs in determining the extent of a reduction are codified: significance of the assistance, its truthfulness, nature and extent, danger to the defendant or the defendant’s family, and timeliness. That final factor operates in a direction that penalizes deliberation. The defendant who cooperates first receives the most credit. The defendant who pauses to assess the government’s case, to consult with counsel more carefully, to weigh the personal consequences, cooperates later and receives less for it. The system rewards speed, and speed is not the condition under which most people make their best decisions.
At this firm, we approach the timing question with more caution than the system seems to encourage. The instinct to cooperate immediately, to establish good faith by arriving first, often produces disclosures that are premature and positions that cannot later be revised. We prefer, where circumstances permit, to obtain and review discovery before advising a client to proffer. There are exceptions to this, though in practice the exceptions tend to confirm the architecture. The delay is modest. The difference in the quality of the decision is not modest at all.
The Corporate Framework
In March of this year, the Department of Justice announced a unified Corporate Enforcement and Voluntary Self-Disclosure Policy applicable across all criminal components except the Antitrust Division. The policy replaces the collection of component-specific and district-level frameworks that had governed corporate cooperation for most of the past decade. Under the new structure, a company that voluntarily self-discloses misconduct, cooperates fully with the investigation, and remediates the conduct may receive a declination of prosecution, provided no aggravating circumstances exist.
The policy formalizes what practitioners already understood about the government’s preferences: early disclosure receives greater credit than thorough disclosure alone. A company that reports misconduct before the government discovers it occupies a different position than a company that cooperates only after a subpoena arrives. I am less certain about the corporate side than the individual side, if only because the policy is weeks old and untested in the ways that matter. The framework conditions full cooperation credit on the company identifying all individuals involved in the misconduct and disclosing relevant facts about each of them. The cooperation of the entity, in other words, is measured in part by its willingness to expose its people.
When the Answer Is No
Not every defendant should cooperate.
This is the sentence most articles on the subject omit, because it complicates the narrative and reduces the sense of agency a cooperation article is designed to provide. Cooperation is the wrong decision when the government’s evidence is weak enough that trial remains a genuine option. It is wrong when the defendant’s information concerns individuals who present a credible risk to physical safety. It is wrong when the defendant cannot be fully truthful, because the consequences of incomplete disclosure exceed the consequences of silence. And cooperation is, if we are being precise, the wrong decision when it is undertaken to satisfy a feeling rather than to achieve an outcome: the feeling of doing something, of participating in one’s own resolution, when the wiser course would have been to wait.
The decision requires counsel who understands the particular district, the particular prosecutors, and the particular manner in which cooperation credit is valued where the case will be resolved. A five-level departure in the Eastern District of Virginia does not produce the same result as a five-level departure in the Central District of California. The same assistance produces different outcomes depending on who receives it.
A first consultation is the beginning of that assessment. The call costs nothing and assumes nothing. What it establishes is whether the facts support cooperation, or whether they support something else entirely.
Most people who contact us about cooperation have already decided they wish to cooperate. That is understandable. The question we return to them is not whether they want to, but whether the process, as it actually operates in their jurisdiction, will deliver what they believe it will. The answer is sometimes yes. The gap between sometimes and always is where the work of representation occurs.