More on Internal Investigations, Yates Memo

Last week we looked at a recent federal appeals court ruling, Gilman v. Marsh McLennan, that affirms a considerable amount of power for compliance and legal officers conducting internal investigations. The decision has enough potential implications to warrant a follow-up post, so let’s keep going.

I won’t rehash my earlier post here. Suffice to say the court ruled that insurance company Marsh McLennan was within its rights to fire two employees, William Gilman and Edward McNenney, who had refused to cooperate with an internal investigation. The employees were also the targets of a criminal probe in the state of New York, as was Marsh McLennan itself.

They argued that when Marsh fired them for non-cooperation, the company did so as part of a deal to save itself from criminal prosecution—which meant, in essence, that Marsh was an investigating agent of the New York attorney general. Therefore, the company’s threat to fire them, and its refusal to provide them legal counsel during the investigation, violated the Fifth and Sixth Amendment rights they would normally have when facing government prosecution.

OK, the appellate court shot down that logic. In this instance, the court said, Marsh was not an instrumentality of the government because it had already begun investigating Gilman and McNenney and had warned them about getting fired. The company cut a deal with prosecutors only after those threats were made, so it was not in cahoots with then-Attorney General Elliot Spitzer.

What bugs me about the ruling is that the court offered a counter-example from 10 years ago, United States v. Stein, where it did rule that the government overstepped its bounds. In that instance, prosecutors investigating some KPMG partners demanded that the firm to stop covering their legal costs (which had been part of their employment contract) if the firm wanted to avoid prosecution itself.

So we have two cases that reached different conclusions. We all know what that means. When the next case comes along—which inevitably it will—the court will apply a test to see whether the dispute is a Gilman case or a Stein case. Compliance officers should review their programs to make sure you land on the right side of things. That leads to a few points…

Require cooperation with internal investigations as part of your Code of Conduct. In a roundabout way, this point that won the case for Marsh. I don’t know whether the company required cooperation in its Code of Conduct at the time, but employment law for Delaware companies does specify that refusal to obey “a direct, unequivocal, reasonable order of the employer” is cause for termination—so you might as well bake that direct, unequivocal, reasonable order right into the Code of Conduct all employees certify.

Open investigations promptly and follow an investigations policy. The first part of this point is what Marsh did well; the second part is what KPMG did poorly. The theme through both parts is this: what are you doing of your own volition, and what are you doing at the government’s behest? Because Marsh had already opened an investigation into Gilman and McNenney’s misconduct, that raised the possibility that they might get fired before New York state authorities struck a cooperation deal with the company. And that was the circumstance that led the appeals court to say no, Marsh wasn’t acting as an instrumentality of the government to violate the employees’ constitutional rights.

Contrast that to KPMG in the Stein case. In that instance, KPMG had a longstanding policy of covering legal defense fees for employees under investigation. Prosecutors demanded that the firm cease those payments if it wanted to avoid prosecution itself, so the firm did. In other words, the government forced KMPG to do something it would not normally do (change its investigations policy) to save the firm, with the explicit result that the accused would have less ability to defend themselves.

That’s the key to what I’ll call “the Stein test,” which boils down to three factors. First, is the company changing its normal course of action? Second, is that change “a direct consequence of the government’s overwhelming influence,” as the court phrased it? Third, would that change of course not have happened but for prosecutors’ conduct?

An investigations policy puts you on the right side of all three questions. Even if prosecutors still pressure you to change the policy in some what that might push you into Stein territory—at least you can seek redress in court, if it comes to that. Without a policy, good luck articulating a defensible position.

Beware requests for legal counsel. When the Yates Memo first came out last fall, my initial objection was that heavy-handed internal investigations might cause trouble on Fifth Amendment grounds: employees forced to incriminate themselves. I still believe that will happen, but for better or worse, the courts have shown a path forward to allow it.

Instead, the real headache might be a person’s Sixth Amendment right to counsel. For example, say you’re investigating a conspiracy of two employees: one in a union, the other not. The unionized employee probably has right to counsel as part of the union contract, while the non-unionized employee doesn’t. So now you’re targeting two employees for the same misconduct, but giving right to counsel to only one. Even if the letter of the law allows it, that double standard should leave any compliance officer uncomfortable. Why?

Beware how this power dynamic will affect your culture. Because employees, like all humans, value a sense of fairness and self-control, that’s why—and the Yates Memo, plus the Gilman decision, will inevitably leave employees feeling less control. Might that drive them to cooperate and help you root out misconduct? Yes. Might it also tempt them not to report misconduct, play dumb, quit rather than cooperate, or accuse the innocent to save their own hide?

Oh, nah. I mean, that never happens in real life, right?