Blog
Cooperating in Public Corruption Investigations
Cooperation in a public corruption investigation is not a negotiation. It is a surrender conducted under terms the government will dictate, that your attorney can influence but cannot control, and that will reshape your relationship to every person you have ever worked with in public life.
The instinct, when federal agents make contact, is to assume that talking will resolve the matter. Talking is the beginning of a process that, once entered, constrains every subsequent decision for months or years. The proffer session, the cooperation agreement, the testimony, the sentencing: each stage removes an option. What begins as a choice becomes, by the second meeting, something closer to an obligation, though the government will never describe it that way.
This is the part that most attorneys, if we are being precise, do not communicate with sufficient clarity at the outset. The question is not whether to cooperate. The question is whether you possess information the government considers valuable enough to justify the cost of keeping you as a witness, and whether the exposure you will accept in exchange is something your life and career can absorb.
What the Government Requires
Federal prosecutors in public corruption cases do not want your regret. They do not want your explanation of why the payments seemed reasonable at the time. They want testimony that can be corroborated by documents, recordings, or the testimony of others, delivered by someone who will not collapse under cross-examination.
The distinction between useful cooperation and insufficient cooperation is one the government makes on its own. Section 5K1.1 of the United States Sentencing Guidelines permits the court to impose a sentence below the guideline range when the government certifies that a defendant has provided substantial assistance. The word substantial does the work in that sentence. What qualifies is, in practice, whatever the assigned prosecutor determines it to be. A cooperating defendant may furnish names, dates, account numbers, the location of documents no one else knew existed, and still receive no 5K1.1 motion because the resulting prosecutions did not materialize or because the information proved redundant.
The arrangement places the determination of adequacy entirely with the party that did not perform. The government will expect complete truthfulness, which means not merely answering the questions posed but volunteering information the prosecutors have not yet thought to request. Omission, in a cooperation agreement, functions as dishonesty. A cooperating witness who conceals a fact that later surfaces has not merely failed to cooperate; that witness has committed a separate federal offense under 18 U.S.C. § 1001.
The Proffer and What It Does Not Protect
The proffer session is the threshold. It takes place, in most districts, at the United States Attorney’s Office, in a room that is smaller and less formal than one expects. Your attorney sits beside you. Across the table sit the assigned AUSA and the investigating agents, who in corruption matters are typically FBI.
A proffer agreement (the document itself runs to two or three pages, sometimes four) provides limited use immunity. The government agrees not to introduce your statements in its direct case at trial against you. That is the protection. It is narrower than it appears. The government retains the right to use your proffer statements for impeachment if you later testify in a manner that contradicts the proffer, and it retains the right to pursue any leads your statements generate. The derivative use exception is the clause that matters most and the one clients understand least: the government cannot quote your words, but it can act on everything your words reveal. If you disclose the existence of a bank account in Zurich, the account itself becomes evidence. Your statement about the account does not.
In the Southern District, the standard proffer letter includes language broad enough to permit the government to introduce proffer statements whenever the defendant offers any evidence or makes any factual assertion that contradicts the proffer. The Second Circuit addressed this scope in United States v. Barrow, holding that the triggering language encompasses factual assertions made through cross-examination and opening argument. The practical consequence is that a proffer, once given, forecloses most avenues of defense at trial. A defendant who proffers and then declines to plead guilty enters trial with an opponent who possesses a full account of the defendant’s own version of events and may introduce it the moment the defense departs from that account.
This is why the timing of the proffer matters as much as its content.
Seven cases in the past two years where we represented individuals considering proffers in corruption matters ended with the client declining to proffer after reviewing the government’s evidence through informal channels first. The reason was the same in each: the government’s case was weaker than the client believed. The agents’ confidence during the initial approach (which is, if one is being candid, a form of salesmanship that federal agents execute with considerable skill, not because they intend to deceive but because the institutional posture of an investigation communicates certainty whether or not certainty exists) had created an impression of evidentiary strength that the actual discovery did not support.
The proffer is not a conversation. It is a record, and everything said in that room will be measured against everything said afterward, in every forum, for the duration of the case.
The government’s willingness to offer a proffer session is itself a signal worth interpreting. Prosecutors do not extend proffers to individuals whose cooperation they consider unnecessary. The invitation implies that the government’s case against the larger targets has gaps. Whether that positioning constitutes an advantage or an exposure depends on facts that, at the proffer stage, neither party possesses in full.
The physical setting of these sessions tends to produce a particular psychology. Conference rooms in federal buildings are constructed to project authority. The flags, the seal, the absence of anything resembling comfort. Clients who have sat through regulatory depositions and state proceedings describe the proffer room as different in kind. The agents do not raise their voices.
The Quiet Arithmetic of Section 5K1.1
The sentencing benefit of cooperation is real, when it arrives. A 5K1.1 motion, coupled with a motion under 18 U.S.C. § 3553(e), authorizes the sentencing court to depart below both the guideline range and any applicable mandatory minimum. The reduction can be measured in years.
Courts weigh five factors: the significance and usefulness of the assistance, the truthfulness and reliability of information provided, the nature and extent of cooperation, the risk of injury to the defendant or the defendant’s family, and the timeliness of the assistance. The list suggests a balancing test. In practice, the government’s evaluation receives what the guidelines call “substantial weight,” which means the court defers. The Supreme Court confirmed in Wade v. United States that the government’s refusal to file a 5K1.1 motion is reviewable only for unconstitutional motive. That standard, in the ordinary corruption case, amounts to no review at all.
We approach this phase differently. Before any proffer discussion begins, we conduct what amounts to a reverse assessment: an independent evaluation of the government’s case strength, the probable value of the client’s information to the prosecution, and the likely sentencing exposure with and without cooperation. If the difference between cooperating and proceeding to trial is modest (because the probable sentence is manageable, or because the client’s information is unlikely to produce indictments), cooperation may cost more in exposure and personal risk than it returns at sentencing. The number of clients who arrive at our office convinced they must cooperate and leave having decided otherwise is not trivial, though I am less certain of the precise proportion than I would like to be. What I can say is that the assumption of necessity is almost always premature.
Retaliation and Exposure
The personal cost of cooperating in a corruption investigation is not hypothetical. Public corruption cases implicate public figures, and public figures maintain constituencies, employees, and associates who regard the cooperator as a betrayal.
In government contracting cases, cooperators report difficulty obtaining future public work. What is less anticipated is the professional isolation that accompanies cooperation in cases involving elected officials. Colleagues who were not themselves targets create distance. Referral sources disappear. The cooperator’s name appears in court filings that are, in most districts, public record. In a community of any size, the effect on professional reputation is immediate.
The Witness Security Program, administered by the U.S. Marshals Service, exists for cooperators whose physical safety is endangered. It was constructed for organized crime and narcotics prosecutions, not for the contractor who testified against a county commissioner. In corruption matters, the risk is social and economic rather than physical, and no federal program addresses it.
Timing and the Decision Itself
The moment at which cooperation is considered determines its value. A defendant who approaches the government before indictment, with information the government does not yet possess, occupies a stronger position than one who offers to cooperate after the evidence has been assembled and the charges filed. The DOJ’s Corporate Enforcement Policy treats voluntary self-disclosure as a factor in determining whether a declination or reduced charges are appropriate. The same principle applies to individuals, though less formally and with fewer guarantees.
Six months after a contract is signed is too late to dispute its terms. Six months after federal agents have served subpoenas is too late to control the narrative of cooperation.
The pressure to delay is understandable. Cooperation requires admitting conduct that may be criminal. It requires trusting a system designed to prosecute. It requires placing your sentencing fate in the hands of a prosecutor whose institutional incentives do not align with your interests in every respect, though they temporarily overlap.
The practical question we pose to clients is this: what does the government already know, and what can you tell them that they cannot discover independently? If the answer to the second question is substantial, cooperation becomes a calculation with identifiable terms. If the answer is thin, cooperation is a confession with uncertain return.
Most of the individuals who contact our office about cooperation in public corruption matters have already waited longer than they should have. The letter has arrived. The agents have called. A colleague has been indicted, and the client’s name appeared in the complaint as an unindicted co-conspirator or, sometimes worse, as a witness the government intends to call. By that point the options have contracted. The corridor is narrower than it was six months prior, and the light at the end of it is the government’s.
A consultation is where this assessment begins, and it assumes nothing beyond the facts the client is prepared to share.