Federal Law Enforcement

In Defense of Mueller’s Obstruction Theory: A Reply to Jack Goldsmith

By Benjamin Wittes
Sunday, May 12, 2019, 4:01 PM

I am far too good a lawyer—despite not being one at all—to take on Jack Goldsmith on a matter combining statutory interpretation, the presidency, and the historic positions of the Justice Department’s Office of Legal Counsel (OLC) without a healthy dose of humility. So let me start by saying that I’m not saying Goldsmith’s critique of Mueller’s statutory analysis on obstruction of justice is wrong. Mueller actually needs to be persuasive to people like Goldsmith on a matter as fundamental as the application of the obstruction statutes to presidential action. If he isn’t, that is a good sign that his obstruction analysis is suboptimal.

With that concession up front, I do believe there is more to be said for Mueller’s position on obstruction than Goldsmith allows—perhaps because there’s more to be said for Mueller’s position than Mueller himself offered. Here are a few additional considerations that neither Mueller nor Goldsmith discussed.

An initial background observation is that Goldsmith’s view of the matter leads to some genuine absurdities. This is not to say that his view of the law is wrong, just that if it’s right, the law is an ass.

Consider Goldsmith’s analysis of the presidential plain statement rule as applied to the perjury statute, rather than to the obstruction statutes. They are, with respect to the presidential plain statement rule, similarly situated. Both perjury and obstruction, after all, are criminal laws of general applicability whose terms do not explicitly apply to the president. Both are also laws that are plausibly but not obviously, as then-Deputy Attorney General Laurence Silberman wrote in a legal opinion in 1974, “like the bribery statute ... where from the nature of the offense charged, no one, however exalted his position, should safely feel that he is above the law.” So I think it’s reasonable to say that if the obstruction statutes cannot be applied to presidential action when doing so may burden his Article II responsibilities, then the perjury statute cannot either, absent a contrary specification by Congress. (Goldsmith’s argument is similar to ones advanced by Josh Blackman here and elsewhere.)

In other words, if Goldsmith is right, it is lawful for the president to, say, walk into a grand jury and lie knowingly and intentionally and repeatedly as long as there’s some plausible argument, even an attenuated one, that the lies are intended to protect, say, a secret diplomatic initiative. I suspect this argument would come as a surprise to Bill Clinton, who was accused of perjury and apparently did not know that the only defense he really needed was a thin-reed claim that his self-protective lies under oath were actually designed to insulate his ability to conduct foreign policy. Clinton’s deposition in the Paula Jones civil lawsuit took place on January 17, 1998; in that very time frame, Clinton was meeting with the presidents of all three Baltic nations and the Israeli prime minister. As Goldsmith describes the law, I think a perfectly plausible defense of Clinton’s conduct would have been that—absent a clear statement by Congress—the perjury statute would not even have applied to him if he were lying under oath in order to make sure he had a strong hand in dealing with these foreign leaders. It strikes me as notable that Clinton’s lawyers, who made just about every defense lawyers could think of, never made this one.

Goldsmith may well be correct that this absurd outcome in which Clinton’s lies might have been lawful is a consequence of the best understanding of OLC’s position on the presidential plain statement rule—a matter to which I shall return. He is certainly correct that Mueller needed to pay careful attention to the matter because, as Mueller himself acknowledged, he is bound by OLC’s views. But OLC’s general discussion of this rule is not the only relevant body of executive branch work on the subject. Another body of work exists, and it uniformly has not adopted the position that Goldsmith faults Mueller for declining to take. To be precise, to the best of my knowledge, nobody actually charged with investigating a president for criminal conduct has taken the view that the presidential plain statement rule prevents application of the obstruction statutes to presidential action.

There have been three such investigations prior to Mueller’s: Watergate, Iran-Contra and the Starr investigation. The matter is complicated because two of the three investigations predated OLC’s adoption of its current articulation of the presidential plain statement rule, and they predated as well the Supreme Court’s decision in 1992 in Franklin v. Massachusetts, in which it applied the rule. As Goldsmith aptly summarizes the Supreme Court jurisprudence:

The leading Supreme Court decision on the presidential clear statement rule is Franklin v. Massachusetts (1992). The court reasoned that “[o]ut of respect for the separation of powers and the unique constitutional position of the president,” the term “agency” in the Administrative Procedure Act (APA) did not apply to the president—even though the express exceptions to the term did not include the president. The Supreme Court cited “respect for the separation of powers and the unique constitutional position of the President” as a basis for requiring “an express statement by Congress” to subject presidential decisions to the APA’s abuse of discretion standard. Another decision is Public Citizen v. Department of Justice (1989), where the Supreme Court applied the clear statement rule (among other tools) to determine that the president and Justice Department did not “utilize” the American Bar Association under the Federal Advisory Committee Act when it sought its advice on judicial nominations. The court said that reading “utilized” to apply to the executive’s relationship with the ABA would raise serious questions about whether the statute “infringed unduly on the President's Article II power to nominate federal judges.” The court added: “we are loath to conclude that Congress intended to press ahead into dangerous constitutional thickets in the absence of firm evidence that it courted those perils.”

What’s more, none of the three special prosecutors at issue—Leon Jaworski, Lawrence Walsh and Kenneth Starr—bothered to write a memo explaining his view of the interaction between Article II and the obstruction statutes, so we don’t know quite how any of them, as opposed to Mueller, justified their actions.

That said, all three investigated the presidents then in office for obstruction of justice in situations in which the matters in question touched their Article II management of the executive branch at least in some significant part. None, to my knowledge, took the view that the presidential plain statement rule precluded application of the obstruction statutes to presidential conduct.

Nixon’s case is the most dramatic. Special Prosecutor Leon Jaworski indicted the president’s top aides for a conspiracy to obstruct justice. He named Nixon as an unindicted co-conspirator in that case. The conspiracy included matters directly related to the management of the executive branch. These included, as Jaworski’s office later summarized in a memo describing the evidence against Nixon, “obtaining by dishonest means information from the Department of Justice for furthering the concealment conspiracy.” As the memo described, “the President began a series of almost daily meetings with [Criminal Division head] Henry Petersen,” in which he “repeatedly sought information from Petersen about the progress of the investigation and, especially, about the evidence being accumulated against Haldeman and Ehrlichman.” Note that the president’s alleged conduct here involved his supervision of an ongoing criminal investigation, a core Article II prerogative. The memo makes other allegations that would similarly be precluded by application of the presidential clear statement rule to the obstruction and perjury statutes—as would, notably, any allegation about the so-called “smoking gun” tape, in which Nixon ordered that the CIA be recruited to stop the FBI’s investigation.

Whatever OLC may have since said about the presidential plain statement rule, the Justice Department has never intimated that Nixon’s participation in this conspiracy—which involved in key aspects the exercise of Nixon’s Article II powers to manage the executive branch—was, after all, lawful. Note that this prosecution was ongoing when Silberman wrote his Justice Department opinion citing the bribery statute as the paradigmatic case of a law that, notwithstanding Congress’s failure to explicitly apply it to the president, applies anyway. I find it somewhat implausible to believe that Silberman was, even while watching the Watergate case progress, imagining the bribery statute as some singular exception to a broad rule, rather than as standing for a class of statutes of which laws like obstruction and perjury were also members. Jaworski, to my knowledge, never laid out a legal theory, and what became the presidential clear statement rule was, in Silberman’s articulation, still in an early form that predated Franklin. So the Jaworski example is not dispositive. It is, however, striking. The consequence of Goldsmith’s reading of the law is that a lot of Nixon’s obstructive activity, though not all of it, was lawful.

Iran-Contra Independent Counsel Lawrence Walsh also investigated presidential obstruction of investigations. Here is how he summarized his investigation on this point:

A principal focus of this section centers on the activities of the President, the Vice President, Regan and Meese during November 1986 when the public disclosure of the Iran arms sales created a political furor, generating demands for a congressional investigation and the appointment of an independent counsel....

Independent Counsel concluded that President Reagan, Vice President Bush, [Don] Regan, [Edwin] Meese and other senior Administration officials in November 1986 undertook to “rearrange the record,'' as Secretary of State George P. Shultz put it in a conversation with his senior advisers, in an effort to protect the President and themselves from accusations of possible violations of law.

In addition, the investigation looked at George H.W. Bush’s withholding, while vice president and later while president, of his daily diary that shed light on Iran-Contra. The late disclosure of Bush’s diary, wrote Walsh, “prompted a special investigation into why the diary had not been produced previously, and the substance of the diary.” The Iran-Contra final report does not specify what precise legal theory Walsh took toward obstruction by the president, I suspect partly because Walsh never got close to alleging criminality on Reagan’s or Bush’s part. That said, a great deal of the conduct at issue in this probe involved presidential management of the executive branch.

So it’s unclear how Walsh’s view of the obstruction law interacted with Article II prerogatives—except in one critical area. Walsh’s investigation looked specifically at Bush’s 11th hour, pre-trial pardon of former Defense Secretary Casper Weinberger as a potential obstruction of justice; this inquiry specifically involved the use of a core Article II power.

The Iran-Contra prosecutor’s final report is clear that the evidence did not support charges against Reagan either for any involvement on his part in his subordinates’ lying to Congress or for any of his own statements. And the investigation didn’t find evidence either that Bush’s pardon of Weinberger was an obstruction. But there is not the remotest suggestion in the report that the office regarded either the obstruction or false statements statutes as not applying to Reagan or Bush in the exercise of their official functions as president. Indeed, Walsh wrote in explaining why he let the pardon matter drop, “The question before Independent Counsel was ... whether President Bush exercised his constitutional prerogative to pardon a former close associate to prevent further Iran/contra revelations. In the absence of evidence that the pardon was secured by corruption, Independent Counsel decided against taking the matter before the Grand Jury.”

Again, this all predates Franklin and OLC’s key 1995 opinion.

Independent Counsel Kenneth Starr’s case is a little more complicated, for a few reasons that do not all cut in the same direction. Starr did not formally evaluate whether Clinton committed crimes in the famous “Starr Report,” after all. That document was a referral to Congress concerned not with criminality but with whether the evidence the investigation developed might support an impeachment. What’s more, Clinton’s conduct in the Monica Lewinsky matter was, generally speaking, somewhat remote from his duties as president. And it thus arguably wasn’t shielded by the presidential plain statement rule from application of the obstruction statutes.

Except, that is, when Clinton’s conduct wasn’t remote from presidential management of the executive branch. In particular, Starr concluded that Clinton obstructed justice—he specifically framed it in the language of the criminal law—in the president’s efforts to influence his secretary, Betty Currie, and in his lies to his senior staff. “Rather than lie to the grand jury himself, the President lied about his relationship with Ms. Lewinsky to senior aides, and those aides then conveyed the President’s false story to the grand jury,” the Starr Report recounts.

This example is complicated because it’s possible to see this conduct as just garden-variety witness tampering, entirely outside the president’s Article II duties. But it’s also possible, under Goldsmith’s theory, to see it as the application of the obstruction statutes in a fashion that genuinely burdens the president’s ability to communicate unimpeded with his most immediate staff. After all, if the perjury and obstruction statutes don’t specifically apply by their terms to the president’s exercise of Article II prerogatives, why shouldn’t he lie to his staff? For that matter, if he doesn’t want, say, his foreign policy or domestic agenda encumbered by a messy investigation, why shouldn’t he coax his staff to lie to make things go away? And why shouldn’t he tell them lies so that they can mislead the grand jury without lying themselves? After Starr stepped down, his successor, Robert Ray, contemplated bringing a criminal case against Clinton. Though the exact parameters of the charges he would have brought are unclear, I think it’s fair to say that Starr’s and Ray’s office never considered the possibility that the obstruction statutes may not apply to presidential conduct if application would burden Article II prerogatives.

Starr’s case is messy because it vividly illustrates how hard it can be to disentangle presidential conduct within Article II from conduct outside of it. On the other hand, the Starr investigation’s relevant work also postdates Franklin and the OLC articulation of the presidential plain statement rule. It shows that even as OLC’s doctrine was developing, the line of special prosecutors assuming and acting on the assumption that the obstruction laws apply to presidential action remained unbroken. 

In short, while Goldsmith may well be reading OLC’s opinions correctly with respect to the optimal application of the presidential plain statement rule in a formal sense, there is a significant body of executive branch practice, up to and including formal statements to the courts and to Congress, in the specific context of prior criminal investigations of presidents. This body of material, which both predates and postdates Franklin and OLC’s 1995 opinion, uniformly cuts the other direction. This is not, to be sure, dispositive. As I say, none of these prosecutors wrote a public memo justifying their apparent view of the obstruction statutes as applied to the presidency, as Mueller has done. So we don’t know their precise legal theories, only what they did. We actually don’t know if any of them even considered the question. But it’s safe to say that if Mueller is defying clear executive branch law in concluding that the obstruction statutes apply to the president even absent a plain statement—and that they apply even to his official acts as president—he seems to be following in an unbroken line of special prosecutors.

This brings us to the key question: Is Mueller’s reading really at odds with what OLC has said? I’m not so sure.

Let’s take another look at that 1995 OLC memo by Walter Dellinger, which postdates and references Franklin and, as Goldsmith describes, elaborates on the earlier Silberman opinion in distinguishing the bribery statute from those laws that cannot be applied to the president without a plain statement of intent to do so. And let’s take a look at it specifically in light of the concurrent development of executive branch practice with respect to obstruction—of which Dellinger was certainly aware. The lengthy footnote reads in relevant part: “Application of [the bribery statute] raises no separation of powers question, let alone a serious one. The Constitution confers no power in the President to receive bribes; in fact, it specifically forbids any increase in the President’s compensation for his service while he is in office, which is what a bribe would function to do. See U.S. Const, art. II, § 1, cl. 7. Moreover, the Constitution expressly authorizes Congress to impeach the President for, inter alia, bribery. Id. §4.”

I find it very hard to interpret OLC’s action here as having declared a significant aspect of the Nixon conspiracy indictment, along with the Walsh investigation, defective—all in a footnote on an unrelated matter (dealing with nepotism and federal judges) and all without referencing obstruction at all.

The question really comes down to whether Dellinger, and Silberman before him, were suggesting that the bribery statute is some unique animal—a law in the criminal code that, because it specifically maps onto both the language of the impeachment clauses and the language that fixes presidential compensation, is exempt from the presidential plain statement rule—or whether OLC here was describing, as I suggested above, a class of statutes of which bribery is the most vivid example. The language of the Silberman opinion suggests the latter, at least to me, though it’s impossible to know what he was intending. Silberman described the bribery statute as not requiring a plain statement of application to the president because “from the nature of the offense charged, no one, however exalted his position, should safely feel that he is above the law.” Whatever he meant with this throw-away line, Silberman’s comment is just as true of the obstruction statutes as it is of the bribery statutes.

But what of Dellinger’s language adding flesh to the bare bones Silberman sketched out a decade earlier? I actually think Dellinger’s additional language, albeit less perfectly than with bribery, applies reasonably comfortably to obstruction as well. Let’s consider both elements.

The impeachment clauses don’t merely make bribery impeachable. They also make “other High Crimes and Misdemeanors” impeachable. And obstruction of justice certainly qualifies, at least in many cases, as a potentially impeachable offense. (Jaworski and Starr certainly both regarded it as such.) So to the extent Dellinger’s bribery distinction hinges on the impeachment clauses, obstruction is different only in the specificity of the Constitution’s reference to it, not in the fact of the Constitution’s making it grounds for impeachment.

To the extent Dellinger is relying on the Constitution’s fixed compensation language, there is—to be sure—no similar language regarding obstruction. But there is the Take Care Clause and the presidential oath, both of which require faithful execution of the laws and the latter of which requires a promise “to the best of my Ability [to] preserve, protect and defend the Constitution of the United States.” So while the specific words Dellinger wrote about the bribery laws cannot be written about the obstruction statutes, something strikingly similar can. It would read something like this:

Application of obstruction statutes to the president raises no separation of powers question, let alone a serious one. The Constitution confers no power in the President to obstruct justice; in fact, it specifically requires him to swear to faithfully execute his office and to preserve, protect and defend the Constitution, and it requires him as well to take care that the laws are faithfully executed. Moreover, the Constitution expressly authorizes Congress to impeach the President for high crimes and misdemeanors, which past practice of both the executive branch and the Congress has taken to include obstruction of justice.

This is admittedly a little more elaborate than what Mueller actually wrote, which—as Goldsmith quoted—was this:

Under OLC’s analysis, Congress can permissibly criminalize certain obstructive conduct by the President, such as suborning perjury, intimidating witnesses, or fabricating evidence, because those prohibitions raise no separation-of-powers questions. The Constitution does not authorize the President to engage in such conduct, and those actions would transgress the President’s duty to “take Care that the Laws be faithfully executed.” In view of those clearly permissible applications of the obstruction statutes to the President, Franklin’s holding that the President is entirely excluded from a statute absent a clear statement would not apply in this context.

If I do say so myself, my version is a little better than Mueller’s. But it’s not different in substance. It seems to me consistent with OLC’s past practice, and also consistent with nearly five decades of executive branch practice in investigating presidents under the obstruction laws. And it doesn’t lead to the absurdity that a president can walk into a grand jury and lie with impunity.

Mueller then goes on to make an argument that Goldsmith criticizes and that I agree is weak. As Goldsmith summarizes:

The exception should mean that the obstruction statutes apply, at most, to presidential actions that raise no separation of powers problems. In this light, one might think (as the report says) that it is appropriate to apply a “more limited application of a clear-statement rule to exclude from the obstruction statutes only certain acts by the President—for example, removing prosecutors or ending investigations for corrupt reasons,” both of which concern presidential action under Article II.  (A simpler way of saying this is that one might think the exception to the clear statement rule applies only to actions that fall within the exception.) Under the clear statement rule, any application of the obstruction statutes that would arguably limit or possibly conflict with the president’s constitutional prerogatives still must be excluded from the statutes, even if other non-Article II actions are not excluded.

In response to this, the Mueller report makes an argument that I have never seen in Supreme Court or OLC jurisprudence on the presidential clear statement rule. The report acknowledges that Article II gives the president removal authority and “broad discretion to direct criminal investigations.”  And yet despite the clear statement rule, the report declines to exclude actions that would burden these Article II authorities from the obstruction statutes’ scope.  The report reasons that excluding the burdened Article II actions “would be difficult to implement as a matter of statutory interpretation” since the relevant terms in the corruption statutes could not “easily bear that specialized meaning.”

But note that under my formulation, this additional argument is unnecessary. The obstruction statutes apply to the president for the same reason that the bribery statutes apply, and they apply in whole because obstruction is impeachable and because corrupt obstruction of lawful investigations, even if the obstructive acts are otherwise valid exercises of Article II powers, violates the obligation of faithful execution of the law. The common element of the obstruction offenses that the act must be “corrupt” puts a high burden on prosecutors to prove state of mind, and no sensible prosecutor would bring a case in which there were a plausible argument that the president acted for a legitimate Article II reason.

I think I can even demonstrate that this point is adequate to protect the value that the presidential clear statement rule is intended to protect, and that there is no more risk of criminalizing Article II activity by applying the obstruction statutes than there is by applying the bribery laws to the president. The reason everyone is comfortable applying the bribery law to the president is that all bribery is outside of the president’s Article II responsibility. My submission similarly is that all truly corrupt obstructions of justice are outside of the president’s Article II power, that there is literally not a single hypothetical that anyone can generate of a truly corrupt interference with a judicial proceeding or criminal probe that is, despite its corruption, a legitimate exercise of presidential power.

So here’s my challenge: Can you think of a case of unambiguously corrupt presidential interference, with the statutes’ criminal intent element wholly satisfied, that you think Article II nonetheless authorizes? If you can, perhaps my argument is circular and Goldsmith is right. If not, I think the reason is that the same theory that makes the bribery law applicable to the president notwithstanding the absence of a clear statement applies to the obstruction laws as well.

Put simply, I favor reading OLC’s position here with a deep awareness that it developed alongside another body of executive branch work, one that—unlike the presidential plain statement rule—specifically deals with the precise issue at hand. I cannot fault Mueller for trying to harmonize the two bodies of work.