How to flip a witness, as explained by a former federal prosecutor

Michael Cohen certainly seems like a guy who’s ready to turn on his old boss.

Last week, Cohen’s attorney, Lanny Davis, gave CNN a tape of Donald Trump and Cohen discussing a payment to former Playboy Playmate Karen McDougal in September 2016. The recording confirmed Trump was aware of the payoff at the time.

Just a few days later, sources told CNN that Cohen was willing to tell special counsel Robert Mueller that Trump knew in advance about the 2016 Trump Tower meeting, where Russians offered campaign “dirt” on then-candidate Hillary Clinton. Trump has repeatedly denied knowing about that meeting.

It was a stunning turn for Trump’s former attorney and fixer, who once said he would take a bullet for the president. The revelations came after months of speculation as to whether Cohen — who’s under federal investigation and appears to be in serious legal jeopardy — would “flip” on the president, striking a deal with prosecutors and spilling what he knows about Trump’s business dealings, the campaign’s ties to Russia, hush money payments to porn actress Stormy Daniels or McDougal, or whatever information prosecutors want.

But if this is a potential defendant after a deal, his strategy is a strange one, Daniel Braun, a former federal prosecutor, told Vox. Braun worked in the Justice Department for nearly two decades, first as an assistant US attorney in the Southern District of New York (SDNY), where he handled official corruption and economic crimes, and later at the Department of Justice in Washington, DC. He returned to private practice in 2016.

Braun walked me through the process of flipping and how cooperation plea agreements work, particularly in the Southern District in New York, which traditionally makes defendants plead guilty to all of their crimes. He offered some insight into what Cohen might be facing if he’s trying to flip — though Braun cautioned that when it comes to Cohen, and the Mueller case in general, expect a few twists.

“These incredibly high-profile cases, where the whole kingdom is watching, don’t always follow the same sorts of patterns. Different pressures are put on the system as a result of all the attention,” Braun said, “and so the system doesn’t always operate the same way it would in a more routine set of circumstances.”

This conversation has been edited for length and clarity.

When we talk about “flipping,” what does that actually mean? What’s actually happening?

What it normally means is that somebody who has had some involvement in criminal conduct decides to take responsibility for that conduct and to cooperate with investigators.

It doesn’t always mean that, but it typically does. Sometimes people who are not facing a meaningful threat of prosecution, but who may have other concerns, decide nonetheless that they will cooperate and assist government investigators.

But usually what people mean when they talk about flipping is someone who has real exposure, someone who has engaged in criminal conduct, saying, “I am going to admit that and cooperate with investigators in various ways.” Really, in any way prosecutors would request. And in exchange for that, the hope is they will obtain more lenient treatment at their own sentencing.

I would imagine this is the ultimate goal for a prosecutor: flipping a key witness or party to criminal conduct. There are probably some caveats, but is it fair to say that getting as many people to cooperate is generally a good thing?

It can be too much of a good thing. You don’t want to end up cooperating too many people whom you’re investigating. Typically, prosecutors want to do it to the extent it’s useful to advance the investigation and help their case. But you might approach a line at some point you don’t want to cross, if you sign up so many people that you end up cooperating nearly everyone in the case. That’s not always a concern, but it can be a concern.

It sounds like it’s a bit of a dance. Prosecutors are investigating a case, getting a sense who is involved, what they might know, whose testimony they might want. How does that work from a prosecutor or investigator’s standpoint?

I don’t know if I can give you one single answer that will always be an accurate description. It happens in different ways, depending on the circumstances. Different prosecutors and agents on the government side of the case will form different tactical judgments. Defense lawyers as well.

The conversation can be initiated from either side. Whichever side it comes from, they initially just want to explore whether or not there’s an interest. The prosecutor can reach out to the defense lawyers or — if the defendant or suspect is not represented — to the person themselves and suggest it would be in his or her interest to do this. The defense lawyer can reach out to the government on behalf of the client and say, “My client would like to come in.”

After that initial flirtation, and if the overture is well received, how do you proceed?

It’s in writing invariably, and has to be by DOJ policy. There really shouldn’t be any aspect of the agreement that’s not included in the cooperation plea agreement itself.

But before you get to that stage, there’s a relatively long process of debriefing that happens. Before you get to a cooperation agreement, where the obligations of both the government and the defendant are set forth in writing, you need to have, in most cases, a series of conversations in order for the prosecutor to evaluate the credibility and completeness of the information that this defendant or potential defendant is willing to provide. And that conversation will typically happen through something called a proffer agreement.

The conversation can start in the street, at someone’s home. This can be FBI agents knocking on someone’s door and saying at the outset, “We’re here to talk to you about the following things, and we’re not asking you to say anything, and actually please don’t. We have some things to share with you.”

That is typically an effort to convince someone they are in a great deal of trouble, that they need to listen carefully, and that when they do decide to talk, if they make that decision, they need to do so with an understanding of the gravity of the situation and the importance of being fully truthful.

The agents are going to have information with them, or they may ask the person to come with them to their office. They might play a tape, or show some photos, or share other evidence they’ve obtained during the course of their investigation to persuade the person it’s in their interest to cooperate — and, moreover, that they had better tell the truth because if they don’t, the agents and prosecutors will know.

Once the lawyers are involved, you would expect DOJ to use a proffer agreement. And that governs the dialogue that then takes place between prosecutors and agents on the one hand, and the defendant or prospective defendant on the other. It provides a set of ground rules for those interviews which then take place prior to a decision to enter into a cooperation plea agreement.

I imagine the proffer agreement is kind of what takes you, step by step, to get to the cooperation plea agreement. Is that a fair characterization?

It is. It provides a limited degree of protection to the person answering questions while both sides explore what’s going to happen in the future.

So, in dummy terms: If the cooperation agreement is the goal, the proffer exists to get both the defendant and the prosecutors to that point. Prosecutors want to make sure the defendant’s story checks out, he’s being truthful. But the defendant also gets some protection [in providing] the government information with their case?

Almost entirely correct. The one thing I wouldn’t agree with: What the agreement doesn’t say, ever, is that there’s only a certain amount of information that the government will be seeking. It doesn’t attempt to draw lines between what you are willing to say and what you are not willing to say. The interview itself — and this is covered it the agreement — is a voluntary one at this stage in the process. There’s no compulsion whatsoever, and the agreement makes that clear: You can walk out of here at any time.

But it doesn’t say we’re going to question you about A, B, and C, but not D, E, and F. And it says at the very top, in bold letters in the Southern District of New York: “This is not a cooperation agreement.” Meaning: Don’t get the wrong idea; this is not an agreement under the terms of which you agreed to plead guilty in return for the government taking certain steps on its side.

The proffer agreement starts off by saying this is voluntary. You’re going to sit down and answer questions and provide information so that we can consider that information in making prosecutorial decisions. If you’re not fully truthful, you lose the protection afforded to you under this agreement. But so long as you are fully truthful, we as the government agree that we won’t use your statements directly against you in any case that ultimately we one day might bring.

However, it says what we, as prosecutors, can do is use the information you provide to investigate. If through that investigation we obtain additional evidence that implicates you, that is fully admissible.

Moreover, if you take the stand in your own defense and you say something inconsistent with your proffer, we can use the proffer to cross-examine and contradict you. If your lawyers call other witnesses or make arguments at a future trial that are inconsistent with your proffered statement, we can again use the proffer for impeachment purchases, meaning to contradict that evidence or those arguments.

That’s the heart of it. There are other provisions too. But that’s the basic idea.

So prosecutors are really in the lead, pulling the defendant or witness, and making sure it’s worth the government’s while to enter into a cooperation plea agreement.

Yes, and to make sure what they’re getting is fully truthful, that it will be valuable, and that the person is willing to answer any question posed truthfully as the case moves forward. Because if the person is drawing lines between what he or she is willing to say and not willing to say, that’s the sort of thing that can cause testimony to blow up at trial.

If you’re going to sign a cooperation agreement as a defendant, there’s a real likelihood that you could end up on a witness stand in a courtroom one day. And if you do that, the prosecutor can’t take the risk that your credibility is going to be undermined.

Although I didn’t say that quite right: That’s always a risk. But prosecutors can’t take the chance that you’re going to be up there and suddenly say, “Oh, I’m not going to talk about that.”

How frequently do things fall apart in the proffer stage?

All the time.

Really?

Oh, yes. Plenty of people come in and they want to cooperate, but they actually don’t really want to cooperate. They’re hoping to get something in exchange for giving up this, but not the other thing. They’re trying to make a calculation as to how far they have to go.

Sometimes they’re not willing to tell the truth at all. They’re actually just coming in to present a defense, or version of events, that they want the prosecutor to consider in standing down. That’s sometimes called an “innocence proffer,” and it would typically be governed by a different sort of agreement that doesn’t provide the same level of protection.

How long can this process, from proffer to cooperation take? Months? Years?

It varies so widely. It’s really hard, if not impossible, to generalize. Years seems really unusual to me, but months is certainly not unusual. Sometimes cases are very straightforward: Agents who have a warrant go into a location and find that it’s an active drug mill. People are there packaging and bundling heroin to distribute, and everyone’s there with the drugs and the money and the counters and the cutting agents and all of that. Everyone’s arrested.

After that, the prosecutor’s likely to get some phone calls from some of the defendant’s lawyers saying, “My guy wants to come in.” And he wants to be the first in because the fifth guy in isn’t going to be able to help much, and their exposure is so clear.

Now, even in a case like that, the debriefing process can take a while because you have to make sure they’re not only being truthful about what they were doing in the room, they have to be truthful about everything. When did you get into this business? How? Through whom? Who did you work with? Who else was involved?

If the conduct is limited enough, you can cover it in a couple of sessions. Rarely one. Sometimes it takes many, and it can go on for weeks, if not months. It’s unusual for someone to be proffering for years. But again, if you’re a prosecutor or an agent, you have to cover everything.

And in the Southern District of New York, unlike in some other US attorneys’ offices or Justice Department components around the country, the practice is for a defendant to plead guilty to all conduct for which he or she has criminal exposure in the cooperation agreement.

So if they’re under investigation for one crime and they decide to admit that and explain what happened, before this process is over, the prosecutor will say: “Now tell me about any sort of other crimes that you’ve committed.” Unless it’s beyond the statute of limitations or unprosecutable, they will have to plead guilty to those crimes in addition to the crime that was under investigation.

I was wondering about that. There was some discussion, for example, after former National Security Adviser Michael Flynn was charged with lying to the FBI as to whether that’s the only thing he did or if he got leniency for cooperation. Why does the SDNY espouse that practice?

That is the Southern District’s practice, and it’s not the practice of most US attorneys’ offices. It depends on how judges in the districts sentence. Because the Southern District is rather open and flexible when it comes to sentencing, someone can plead guilty to everything they’ve ever done that’s against the law, and the judges will impose a sentence that doesn’t put them in an even worse situation than the one they would have been in before they cooperated.

A lot of federal districts use a formula for cooperation sentences. If you’re truthful and your information is important, you get X percent. If you take the stand and testify, it’s X plus 10. If you cooperate proactively and make tapes, particularly in a way that is risky for you, then you’re going to get an even higher percentage of a discount.

The Southern District doesn’t do that. For districts that do, and there are many of them, if someone increases their sentencing exposure by admitting everything and pleading guilty to it, they don’t make up for that through the discounting. You need to be in front of a judge that basically says, “Okay, I understand what the prosecutors are doing, and I have to make sure that the penalty imposed is a sufficiently strong incentive to actually reward cooperation.”

Do you think that SDNY’s practice — of basically throwing the whole book at a defendant — helps or hurts in getting a witness to cooperate? Or maybe the better question is, why do it that way?

I think the office’s position is that it helps in having more credible witnesses because if the witness takes the stand, [he or she] will have admitted and pleaded guilty to everything, and essentially can present as an open book.

And the prosecutor will say they admitted everything, and their deal is right here in the agreement, and what it says is that by cooperating in this way the government will write a letter to the judge that sentences them — it’s often, but not always the judge who’s presiding over the trial at which they testify. The prosecutors say, “This is the good stuff, this is the bad stuff,” and the judge gets to take all of that into account in deciding what sentence to impose.

I think for Southern District prosecutors, the experience has been that that’s the best way to present a witness like this, and that it makes for the most effective testimony. But I will tell you, that having left the Southern District and worked at DOJ headquarters and seen all the variations of this practice, everyone tends to be happiest with their way of doing things. There are plenty of prosecutors who think the Southern District’s way of doing this doesn’t make sense and wouldn’t work; at least, it wouldn’t work for them because their judges behave differently.

It’s interesting to hear the logic behind SDNY’s practice.

If it is logical. A lot of it is practice that’s evolved over the years, but it’s also experience and people developing a level of comfort with their practice, seeing that by and large it works and that it’s effective from their perspective.

If you’re a defendant, you might well hesitate before signing on to an agreement that requires you to plead guilty not just to what the government was investigating when you were arrested, but everything else you’ve ever done.

But defendants make that decision with advice, hopefully, from a capable, experienced lawyer who can say, “This is the practice, these are our options, and in my experience, this is how the judges respond to a case of this nature.” Lawyers can’t make any promises, but they can certainly give the defendant a deeper sense of comfort that even though there’s sentencing exposure — their maximum sentence is going through the roof and then some — that the judges understand this practice, and that in imposing a sentence, the benefit the defendants receive will more than make up for the increased exposure.

And that’s not always the case, right? It’s sometimes limited just to the events that are under investigation. Often it is in the white-collar context, where you’re not talking about someone who would’ve been engaging in lots of other criminality. It just depends on the individual.

What about immunity from crimes?

You can get immunity, and we have a federal statute that authorizes immunity. It’s something that a prosecutor has to apply to offer. It’s an application, it goes up to the approval chain within the office, and then it goes to Washington.

The case that you have to make is why it makes sense to immunize this witness, for purposes of this prosecution. You’re not talking about flipping and cooperating with someone in the way you and I have been talking about. You’re talking about obtaining an immunity order that compels them to testify. Then they have to show up, either before the grand jury or the trial court or both, subject to the order to provide truthful testimony.

It is extremely unlikely that that that person will face prosecution themselves. It’s not impossible to prosecute someone who’s been immunized. It’s extremely difficult, but because of how difficult it is, what the government is basically deciding to do is to take a pass on that opportunity — if there even is one.

How as a prosecutor do you make sure that you’re basically picking the right witness to flip? Is it a matter of putting out feelers for a lot of different people, figuring out who has the most information? What is the strategy of trying to make sure that that person stands up in court to make the case most effectively?

It’s a great question, and it’s a difficult judgment call for prosecutors to make. You are focused on whether or not the person can testify credibly and whether or not they will present as a credible witness.

The most important factor in making that decision tends to be the strength of the corroborating evidence that you can offer alongside that witness’s testimony. It’s hard to imagine a situation in which a prosecutor would present the uncorroborated testimony of a cooperating witness and expect that to be accepted as proof beyond a reasonable doubt by a unanimous jury.

The court will instruct the jury that the cooperator’s testimony needs to be scrutinized with particular care. And the prosecutor will typically embrace that and say: “That’s absolutely right. How do you know who’s telling the truth? You’re going to listen to the witness. You’re going to evaluate what you see and what you hear, what makes sense to you. How do they strike you?

“And, of utmost importance, you’re going to look at all of the other evidence that was introduced over the course of the trial. You’re going to decide what’s true. We’re not asking you to like this person or simply to take their word for it. The question that you’re being asked to decide is not solely whether or not you believe this person. It’s whether all of the evidence taken together amounts to proof beyond a reasonable doubt.”

The prosecutor is thinking about the arguments they’re going to make to the jury and the arguments that the defense lawyer will make. Prosecutors understand the cooperator might be vulnerable, based not only on the behavior they’ve engaged in, but perhaps the way they testify.

There are different arguments that prosecutors make. For example, in the conspiracy-type setting, one argument you’ll often hear a prosecutor making: “Look, we didn’t pick Mr. So and So. The defendant did. If you’re going to understand the way a criminal enterprise worked, then you’re going to have to hear from some of the people who participated in it.

“You can only get that perspective from them. And those people are not going to be angels. And if they were, they wouldn’t be able to provide you with an honest and valuable account of what this scheme was all about and how it was carried out. There’s only one place you can get that. From somebody on the inside.”

You’ve got me convinced.

Well, there are good arguments to make on the defense’s side, too. Look at the defendants’ motives. Look at what they were willing to do when less was on the line, ladies and gentlemen. When their freedom was not directly at stake, look at the choices they made. If they couldn’t help the government, they wouldn’t be here.

“And whether it’s substantial assistance, is not up to the judge. Ultimately, it’s up to the government lawyers, who have already made up their minds about what has happened here, and who have accused my client of engaging in this behavior. The [cooperating] witness has to make these people happy. And there’s only one way to do that.”

So the Michael Cohen case. Everyone’s talking as if he’s going to flip on Donald Trump, and cooperate with the Mueller investigation. Is that even a fair assessment for outside observers to make?

I’m an outside observer for sure. That’s important. I don’t know anything that all the people don’t know. I probably haven’t been following it as closely in the papers and other media as other people have. But if someone has their home and offices searched, it means that the government has demonstrated to a judge that there is probable cause to believe that there’s evidence of certain crimes that is likely to be located in those locations.

Most search warrants make a pretty compelling case. That’s why judges authorize them. I would tend to think based on that, that the government has significant evidence implicating Cohen. What’s in his home and office that prosecutors believe — and have persuaded a judge — is likely to be evidence of criminal behavior?

That seems significant to me, as far as him being in the crosshairs. But the likelihood that he would then flip and cooperate? I don’t know how to make a prediction with respect to that.

Could Michael Cohen just be guilty of his own crimes, that have nothing to do with Trump? Or is it likely that he’s in the crosshairs because he has something valuable on someone more important?

There’s no way to evaluate that, he could be guilty of his own stuff. He could be guilty of something that involves other people and not the president. He might be guilty of engaging in criminal behavior with the president and be unwilling to talk about that or to help authorities with that investigation.

There are some occasions, certainly, in which even confronting very serious evidence, people make the decision not to cooperate. I think it’s very difficult to jump to conclusions in this area.

It’s my understanding that SDNY’s Office of Public Corruption is leading this case. Is there something to gleaned from that?

It’s hard to glean too much from it other than that, in at least in the initial assessment phase, someone in the office thought that those issues were there and were part of the investigation.

It typically means that there are allegations of official corruption that are at issue in an investigation. Cases move across the office. There were cases that ended up in the public corruption unit because someone brought them to the public corruption unit from a different part of the office.

The lines between the different units within the US attorney’s office are not so bright and strongly drawn. The Eliot Spitzer case was handled out of the public corruption unit, even as we all realized it wasn’t really a traditional public corruption case.

Last week, man: First there was the release of the Michael Cohen tape, and then the reports that Cohen was prepared to say that Trump knew in advance about the Trump Tower meeting. If Cohen is looking for a deal, I’m imagining prosecutors wouldn’t be too happy about that.

Nope. Not at all happy. Because often prosecutors want to act on information they’re getting, they want to investigate, they often feel that the best opportunity they have to find the truth occurs from investigative steps they can take before the whole world was paying attention. That’s often from a prosecutors’ perspective the best way of finding the truth if they have the opportunity, again, it’s before it’s all the way out there.

This is a really unusual situation, where you have information — and I don’t know where it’s coming from, of course. I don’t know if it’s coming from Cohen; I don’t know if it’s coming from Trump. Who the heck knows? But if anybody is even thinking about cooperating, putting information out there like that is really strange.

But here’s the thing. This is just a weird universe that you’re in right now. You’re not covering a typical criminal case. When I was in law school, I took a great class called the “History of the Common Law,” and it’s all about the foundations of our legal system.

One of the first cases we read, if not the very first one, was something called Throckmorton’s case. It was a case involving high treason in England. What the professor explained is that, as a historical source, you have to be careful about the conclusions that you draw from looking at that material. Because, as he put it, “When the whole kingdom is watching, things tend to happen a bit differently.”

So you’re dealing with a version of that. These incredibly high-profile cases, where the whole kingdom is watching, don’t always follow the same sorts of patterns. Different pressures are put on the system as a result of all the attention, and so the system doesn’t always operate the same way it would in a more routine set of circumstances.