Orin Kerr is an insightful legal analyst, so when we are in disagreement I take his criticism seriously. Respectfully, however, his Lawfare critique of a recent column in which I took issue with Special Counsel Robert Mueller’s pleading practices is not his best work.
To recap my argument, Mueller’s tactic of charging sensational offenses and pleading them down to comparatively trivial crimes flouts guidelines that are prescribed in the U.S. Attorney’s Manual and that are followed by responsible U.S. attorneys’ offices. Kerr objects, contending that a provision in the manual that I did not discuss indicates there is more leeway in the guidelines than I let on.
I think he is wrong on this narrow point because the guideline he cites, which applies to non-prosecution agreements for potentially culpable witnesses when time is of the essence, is not pertinent to the situation I was discussing: viz., the plea deal of Richard Gates, who faced two indictments alleging financial-fraud felonies involving over $100 million in the aggregate, but was permitted to plead guilty to minor charges. Before we come to that, though, some underbrush needs clearing.
First, Kerr implies that I see the manual guidelines as legally binding. I don’t, and never have — not in 20 years living under them as a prosecutor, nor in the succeeding 15 years as a commentator. The thrust of my argument is that Mueller is not upholding critical Justice Department “standards” and “policy” (the words used in the column) that are expressed in the guidelines. If the guidelines were binding, there would be little point in arguing policy, as I have done; I would call for court enforcement. But the guidelines reflect internal DOJ standards; there is no judicial remedy for deviations.
In a different recent column, I highlighted the manual’s opening passages, which unambiguously declare that the guidelines are just that, guidelines. (See Section 1-1.100.) How strictly they are honored is up to federal prosecutors, their supervisors, and Justice Department leadership. In my experience, the manual’s guidelines are taken quite seriously — they certainly were in the U.S. attorney’s office for the Southern District of New York (SDNY), where I worked. But Mueller is effectively unsupervised, so no one is going to force him to adhere to them. Kerr’s description of my claim that Mueller is “breaking the rules” suggests that I’ve accused the special counsel of violating the law. No, I’ve accused him of abusing his discretion.
That this is not actionable does not make it right.
Second, while the guidelines obviously aim to ensure equal and ethical enforcement of federal law across the country, just as significant is their goal of promoting the most effective investigations and prosecutions possible under whatever circumstances prevail. My criticism is directed to this latter objective, which I think Kerr fails to give its due.
A defendant should be required to plead guilty to “the most serious readily provable” offense charged, as dictated by the guideline I cited (U.S. Attorneys Manual, Section 9-27.430), which is directly applicable to plea agreements. Kerr grouses that I don’t address the purportedly complicating factor of cooperation in discussing this provision (though he acknowledges that I do raise it elsewhere — we’ll come to that). But there is no need to address cooperation as if it complicated matters, because it doesn’t: The “most serious readily provable” standard applies to plea agreements regardless of whether cooperation is in the mix.
Requiring such a guilty plea not only ensures that the defendant is held appropriately accountable and is not given favorable treatment in comparison to others similarly situated; a plea to “the most serious readily provable charge” also makes the defendant a more compelling cooperating witness. By contrast, failing to require a plea to the most serious offense degrades the defendant’s testimony, which is usually offered to prove against other defendants the same serious offense on which the cooperating defendant has been given a pass.
A concrete example makes the point. A defendant who has committed bank-fraud conspiracy and pleads guilty to bank-fraud conspiracy is an effective witness — his admission of guilt goes a long way toward proving the existence of the scheme and makes it more likely that conspirators he implicates will be convicted. This helps the prosecution. On the other hand, if the prosecutor lets the defendant plead guilty to a minor (non-bank-fraud) conspiracy to induce his testimony against other bank-fraud conspirators, it signals to the jury that the bank-fraud conspiracy is not as serious as the indictment suggests; it opens the door to defense claims that the cooperator was given a major break to buy his testimony and exaggerate the culpability of the other conspirators. This hurts the prosecution.
I am not criticizing Mueller because I’m a contrarian who eschews special-counsel appointments (though I plead guilty to that). Regardless of my take on Mueller’s appointment, if Richard Gates committed $100 million in financial fraud, I want to see him commensurately punished for it. If Paul Manafort committed these same egregious crimes, I want to see Mueller effectively prosecute him. This should involve the main cooperating witness, Gates, pleading guilty to the most serious readily provable charge against both himself and Manafort.
That is what would have been required, at a minimum, in the U.S. attorney’s office where I worked. More likely, an SDNY prosecutor would have insisted that Gates plead guilty to all of the offenses that he and Manafort were indicted for committing jointly. This prudent practice was the standard during my years in the office, and I am pleased to say that it continued in the years after I left. (See, e.g., Byron York’s recent Washington Examiner column in which he quotes former SDNY U.S. attorney Preet Bharara: “When we had evidence against somebody and wanted them to flip, we made them plead guilty to every bad act that they had ever done, especially if we were later going to be alleging other people had engaged in that activity as well.”)
Because of the way federal sentencing law works (which we’ll come to), good prosecutors tend to demand more, not less, when a plea involves cooperation. The point is not just to get cooperation from defendants in Gates’s position; it is to make them the most effective witnesses they can be in any resulting cases. My beef with Mueller is that his deviation from this practice undermines the ability of other Justice Department prosecutors to honor the guidelines and demand guilty-plea terms that are vital to effective prosecutions — either that, or Mueller does not have strong enough evidence to justify the grave charges he has filed.
With that as background, I’m puzzled that Professor Kerr is puzzled about a passage he excerpts from my column. In it, I explained why it is unnecessary for a federal prosecutor to agree to a sweetheart plea in order to induce cooperation:
If, as all appearances suggest, Mueller’s goal is to get Gates to cooperate, such a plea [to the most serious, readily provable charge — in this instance, bank fraud], besides honoring Justice Department guidelines, would have provided plenty of incentive. Under federal law, the prosecutor does not need to sell out the case for a song to induce cooperation. The prosecutor can demand a guilty plea that reflects the gravity of the defendant’s actual offenses. Then, if the defendant cooperates fully and truthfully, the law permits the prosecutor to ask the judge to impose a sentence beneath the severe term that would otherwise be called for — a sentence of little or no jail time.
Because Kerr is under the misimpression that I claim Mueller is legally bound to follow the guidelines, he is perplexed by this passage, which implicitly concedes Mueller’s discretion to deviate from them. Again, I am not saying Mueller is lawfully bound to follow the guidelines; I am saying he is duty-bound to do so. But that aside, this excerpt refutes Kerr’s suggestion that Mueller may have had no choice but to sell out his case if he wanted Gates’s cooperation.
Under the sentencing guidelines, a prosecutor never has to give away the store to obtain cooperation from a serious offender — something a prosecutor should never want to do, because it makes the cooperator a less effective witness. The prosecutor can insist that the defendant plead guilty to the most serious charges and then, after the defendant cooperates, the prosecutor can file a motion that empowers the court to sentence the defendant to little or even no jail time — regardless of what the sentencing guidelines and statutes of conviction would otherwise have mandated. (See U.S. Sentencing Guidelines, Section 5K1.1, here at pp. 482–83.) That is to say, under Justice Department practice, a cooperating defendant should not get sentencing leniency until he has earned it by providing cooperation (including testimony) and demonstrating truthfulness, usefulness, and contrition. Under Mueller’s approach, in stark contrast, the defendant is rewarded with a low-ball cap on the sentence before he has earned it.
As noted above, Kerr infers that federal prosecutors have more leeway than I have suggested by pointing to a manual guideline I did not discuss. Specifically, Kerr cites the manual’s section 9-27.600, which relates to non-prosecution agreements in exchange for cooperation. This guideline provides that, in exigent circumstances, a suspect may be given complete immunity from prosecution — i.e., not made to plead guilty to any charges — in exchange for information and testimony.
Mueller could have demanded an appropriate plea and pressured Gates to earn sentencing leniency by cooperating.
Preliminarily, I’m confused by Kerr’s pivot to inference-drawing. The main thrust of his argument is that Justice Department guidelines are discretionary (a position in which I concur). If that is so, a prosecutor would not be bound to follow the guideline calling for a plea to “the most serious readily provable charge”; why, then, the need to draw an inference, from a different non-binding guideline, to construct this authority to deviate?
In any event, the two guidelines address very different situations. I discussed Gates in the context of Section 9-27.430 (“most serious readily provable offense” guideline) because it is directly applicable to his situation — the section is entitled “Selecting Plea Agreement Charges” and Gates is an indicted defendant who negotiated a plea agreement to those charges. I did not discuss the guideline Kerr cites, Section 9-27.600, because it has nothing to do with Gates’s situation. That section is entitled “Entering into Non-prosecution Agreements in Return for Cooperation — Generally.” Gates did not enter a non-prosecution agreement (non-pros); he entered a plea agreement.
Kerr’s own description of the non-pros guideline illustrates why it is inapposite:
The idea is that in some cases, the only way to get a defendant to cooperate quickly may be to make a deal: Cooperation in exchange for no prosecution at all. The Manual says that is an option if “the person’s timely cooperation appears to be necessary to the public interest and other means of obtaining the desired cooperation are unavailable or would not be effective.”
The situation to which Kerr refers usually arises when the prosecutor cannot yet prove a suspect has committed crimes but time is of the essence — perhaps, to take a common example, the most culpable suspects in the investigation are about to flee. In that posture, the prosecutor might grant immunity to a suspect in order to pry the information needed to make the case against the higher-ups. The downside is that the prosecutor will eventually learn that the suspect has committed serious crimes and now can’t be charged; that is why, as Kerr concedes, the guidelines indicate that non-pros agreements should be avoided if there are alternatives that would hold the accomplice accountable for his crimes, or would otherwise trade away less immunity in exchange for cooperation. Where there are no viable alternatives, however, a non-pros may yield a worthy case against the main offenders, when such a prosecution would otherwise have been impossible.
This was not the case with Gates. To the contrary, by Mueller’s own account, Gates had committed known, serious, readily provable offenses. Applying the manual passage that Kerr himself quotes, it cannot be said that unless Mueller declined to prosecute (or at least dropped the serious charges), “other means of obtaining the desired cooperation [were] unavailable.” Mueller could have demanded an appropriate plea and pressured Gates to earn sentencing leniency by cooperating (which is standard operating procedure); or he could have convicted him at trial and pressured him to cooperate in exchange for a reduced sentence (the usual alternative to the standard procedure).
In addition, while Gates’s cooperation in the prosecution of Manafort (and in the Russia investigation generally) is doubtless welcome, it is presumably unnecessary. The fact that Mueller indicted Gates and Manafort necessarily means the special counsel has a good-faith belief that his evidence is strong enough to persuade a rational jury to convict both of them. And in this instance, the narrative style of the indictment conveys the prosecutor’s assessment that the proof of guilt is quite strong. So, not only is time not of the essence here. Gates is not of the essence. Mueller was fully prepared to prosecute Manafort without any help from Gates. Putting Manafort aside, if Gates has other information relevant to Russia’s interference in the 2016 election (as the New York Times intriguingly intimated this week), neither Gates, Russia, nor President Trump is going anywhere.
Contrary to the non-pros situation, here Mueller was holding all the cards. And Mueller had given every indication that Gates is guilty of major crimes that could land him in a penitentiary for decades. Thus, if Gates was to merit leniency, the onus should have been on him, not Mueller. It should have been Gates’s burden to plead guilty to what he did and earn leniency through cooperation. Mueller did not need to dump the most serious charges to entice the plea. He did not need to structure a plea deal that enables Gates to avoid the penalties that Congress has prescribed for the conduct — e.g., money-laundering and bank fraud — that Mueller tells us Gates has committed.
Professor Kerr is a very smart guy with solid Justice Department experience (among his other impressive credentials). It is hard for me to fathom that he sees nothing irregular in a defendant’s being charged one day with nine-figure financial-fraud felonies and then permitted, just 24 hours later, to plead out to minor charges. If the charges in Special Counsel Mueller’s indictment are what they are represented to be, the public interest was undermined by a plea that does not hold Gates accountable and that could hurt the case against Manafort. Kerr is right that I have qualms about Mueller’s appointment; but I have deeper objections to alleged bad guys getting sweetheart deals that make it harder for prosecutors across the nation to enforce prudent Justice Department practices.