Cooperation in Federal Drug Conspiracy Cases

Cooperation in Federal Drug Conspiracy Cases

The Proffer Session

Cooperation in a federal drug conspiracy case begins before the defendant has decided to cooperate. It begins in a room at the United States Attorney’s office, with a proffer agreement on the table and a pair of federal agents who already know most of what the defendant is about to say. The document itself, sometimes called a “queen for a day” letter, is two or three pages long. Its promises are narrow. Its exceptions are not.

A proffer agreement offers limited use immunity: the government agrees not to introduce the defendant’s own statements in its case in chief at trial. That phrase, “case in chief,” carries less protection than it appears to. The government retains the right to use anything said during the proffer to develop new leads, identify new witnesses, and construct new charges from information the defendant may not have realized was discoverable. If the defendant later testifies at trial in a manner inconsistent with the proffer, the entire session becomes admissible for impeachment. The protection, in other words, runs in one direction until it does not.

Most proffer sessions last several hours. The agents ask questions they already possess answers to, measuring the defendant’s truthfulness against what wiretaps, surveillance, and prior cooperators have already established. The defendant who minimizes involvement or withholds names is detected, usually before the session concludes. The consequence of dishonesty is not merely the collapse of the cooperation agreement; it is a possible prosecution under 18 U.S.C. § 1001 for false statements to a federal agent.

The decision to proffer is not a decision to talk. It is a decision to disclose everything, to every person in that room, with the understanding that the disclosure cannot be retrieved. A defendant who proffers and then proceeds to trial has surrendered the architecture of any defense that contradicts what was said. Defense counsel in that posture may find themselves unable to cross-examine certain witnesses, unable to present certain arguments, unable to place the defendant on the stand, all because the proffer agreement they signed three months earlier has transformed the trial into a structure with walls the defense did not construct.

Substantial Assistance and the 5K1.1 Motion

The sentence reduction most defendants envision when they hear the word “cooperation” is governed by Section 5K1.1 of the United States Sentencing Guidelines. Under this provision, the government may file a motion requesting that the sentencing court impose a sentence below the advisory guideline range, based on the defendant’s substantial assistance in the investigation or prosecution of another person. The motion belongs to the government. Only the government may file it. The defense cannot compel it, and the court cannot grant the departure without it.

This asymmetry is the central fact of federal cooperation. A defendant may provide information, testify before a grand jury, wear a recording device, and appear at trial as a government witness, and the prosecutor retains complete discretion over whether to characterize that assistance as “substantial.” The Supreme Court affirmed this discretion in Wade v. United States, holding that absent an unconstitutional motive, the government is under no obligation to file a 5K1.1 motion regardless of the defendant’s efforts. The exceptions (racial discrimination, breach of a plea agreement, conduct that shocks the conscience) are invoked more often than they succeed.

When the government does file a 5K1.1 motion, the court considers five factors in determining the extent of the departure: the significance and usefulness of the assistance, the truthfulness and completeness of the information provided, the nature and extent of the cooperation, any danger or risk of injury to the defendant or the defendant’s family, and the timeliness of the assistance. That last factor is the one most defendants discover too late. A defendant who waits six months to cooperate while co-defendants have already provided the same information possesses a commodity that has, if we are being precise, already been sold.

And the 5K1.1 motion, standing alone, permits the court to sentence below the guideline range but does not permit the court to sentence below a statutory mandatory minimum. For that, the government must also file a motion under 18 U.S.C. § 3553(e), a separate statutory authority that empowers the court to disregard the mandatory floor. In practice, prosecutors who file a 5K1.1 motion will usually file the companion 3553(e) motion. But not always. I have seen cases in which the government filed the guideline departure and declined the statutory one, leaving the defendant at the mandatory minimum despite months of cooperation that had already altered the defendant’s relationships, safety, and standing in every community that mattered to him. The written plea agreement controls what the government has obligated itself to do, and the language of that agreement deserves the kind of scrutiny one would apply to a contract with an adversary, because that is what it is.

The range of sentence reductions under a 5K1.1 departure varies. In drug conspiracy cases involving mandatory minimums of ten or twenty years, a cooperating defendant might receive a sentence somewhere in the range of four to eight years, though the figures are approximate and the sentencing court exercises broad discretion.

One dimension of cooperation that receives insufficient attention is its duration. Cooperation agreements require the defendant to remain available for debriefings, grand jury testimony, and trial testimony in cases that may not be filed for years. A cooperating defendant sentenced this year may be called to testify in 2029 or later. The obligation persists. The risk persists with it.

The Safety Valve Under Section 3553(f)

The safety valve is not cooperation in the sense that most defendants understand the term. It does not require the defendant to assist in the prosecution of another person. It does require the defendant to disclose, truthfully and completely, all information concerning the offense and any related conduct, but the disclosure serves the court’s assessment of the defendant’s own situation. The distinction is material.

Under 18 U.S.C. § 3553(f), a defendant convicted of certain drug offenses may be sentenced without regard to the statutory mandatory minimum if five criteria are satisfied. The defendant must have a limited criminal history. The defendant must not have employed violence or possessed a dangerous weapon. The defendant must not have served as an organizer, leader, manager, or supervisor. The offense must not have resulted in death or serious bodily injury. And the defendant must have provided the government with all information and evidence at the defendant’s disposal concerning the offense by the time of sentencing.

The safety valve also confers a two-level reduction in the offense level under Section 2D1.1(b)(18) of the Sentencing Guidelines, which compounds with the avoidance of the mandatory minimum. For a first-time offender with a low criminal history category facing a five-year mandatory minimum on a mid-level drug quantity, the resulting guideline range can fall well below what the statute would otherwise require.

The fifth criterion, truthful disclosure, often requires a proffer session with the government, which introduces some of the same risks described above concerning impeachment and derivative use. But the posture differs. The defendant is not offering to testify against others or to participate in ongoing investigations. The defendant is describing what happened. The court, not the prosecutor, determines whether the disclosure satisfies the statutory requirement.

A defendant who qualifies for the safety valve and also provides substantial assistance to the government may receive both benefits. The reductions can compound. Whether to pursue one, both, or neither depends on the defendant’s criminal history, the defendant’s role in the conspiracy, the quality and exclusivity of the information available, and the physical consequences that full cooperation would impose. I am less certain than most practitioners about whether the compounding strategy is advisable in every case where it is technically available, because the risks of full cooperation do not diminish simply because the safety valve has already been secured.

After Pulsifer

In March 2024, the Supreme Court decided Pulsifer v. United States and narrowed the safety valve’s reach. The question concerned how to read the criminal history requirement under § 3553(f)(1), which the First Step Act of 2018 had amended. Justice Kagan, writing for a six-to-three majority, held that each of the statute’s three criminal history conditions is an independent disqualifier. A defendant who triggers any one of them is excluded from safety valve relief, regardless of whether the other conditions are satisfied.

Mark Pulsifer had pleaded guilty to distributing methamphetamine and faced a fifteen-year mandatory minimum. He possessed two prior three-point convictions but no prior two-point violent offense. He argued that the statute’s use of “and” required all three conditions to be present before disqualification could attach. The Court disagreed, treating § 3553(f)(1) as a checklist. Justice Gorsuch, joined by Justices Sotomayor and Jackson, dissented on the ground that the rule of lenity should govern a genuinely ambiguous penal statute. Whether the dissent or the majority perceived the statute more faithfully is a question worth considering.

The practical consequence is that defendants with moderate criminal histories are now foreclosed from the safety valve entirely. For those defendants, a 5K1.1 cooperation framework may represent the only mechanism for obtaining a sentence below the mandatory minimum. Pulsifer did not alter the landscape for first-time offenders with clean records. It constricted the corridor for everyone else.

The Arithmetic of Cooperation

In a multi-defendant drug conspiracy case, cooperation becomes a problem of game theory that most defendants fail to recognize until the advantage has already shifted. The first defendant to cooperate possesses information the government has not yet obtained from another source. That information has value. The second defendant to cooperate possesses, in significant part, information the government has already received. The value has diminished. By the fifth cooperator, the government may regard the information as redundant, which means the 5K1.1 motion (which defendants of good faith believed they had earned through genuine and complete disclosure) may never be filed at all.

Prosecutors understand this dynamic and do nothing to discourage the urgency it creates. The defendant who is told that co-defendants have already met with the government is receiving accurate information, and that information is designed to accelerate a decision that deserves deliberation. The tension between the need to act quickly and the need to act with adequate counsel is, I suspect, the defining difficulty of federal drug conspiracy defense in its opening weeks.

Six months after the arrest, the cooperation window has not closed, but its value has contracted. The information a defendant holds has been corroborated or contradicted by others. The government has assembled testimony that renders the defendant’s account supplemental rather than essential. The proffer session that would have produced a substantial departure in the first month may produce a more modest one in the sixth, or none at all.

The cooperation decision is made once. Its consequences arrive for years.

Whether to cooperate implicates years, sometimes decades, of a life. It implicates the safety of families who remain in communities where the conspiracy operated. It implicates relationships that will not survive what happens inside that room at the United States Attorney’s office. The sentencing data confirm that cooperation produces shorter sentences, in the aggregate, than the alternative. What the data cannot capture is whether it produces shorter sentences for the particular defendant in the particular case, given what that defendant knows, what the co-defendants have already disclosed, and what the government assembled before anyone agreed to talk.

There is no formula for this determination. There are inputs: the strength of the government’s evidence, the defendant’s criminal history, the defendant’s role in the conspiracy, eligibility for the safety valve, the specificity of the plea agreement’s language, and the question of personal safety that no legal analysis can fully absorb. There are outputs: a sentence, a set of obligations that may extend for years, a set of risks that do not appear in the agreement.

A consultation is where the analysis begins. It costs nothing, and it assumes nothing beyond the recognition that the window in which every option remains available is not indefinite. The first weeks matter. We conduct this assessment at the outset of representation, because the decisions that close doors tend to be the ones made before a defendant understood how many doors were open.