Why Robert Mueller Is Right that the Obstruction Statutes Apply to the President

In recent weeks, Special Counsel Robert Mueller has come under sharp attack with respect to three aspects of Volume II of his concluding Report to the Attorney General.

First, several observers, from disparate perspectives, have condemned Mueller’s decision not to articulate any final conclusions about whether President Trump violated federal obstruction-of-justice statutes.

Second, President Trump’s defenders, and even some critics who are no fans of Trump, go one step further:  They accuse Mueller of acting inappropriately, or even unlawfully, for doing what he did in Volume II instead of making a call about Trump’s criminal culpability—namely, laying out more than 140 pages of detailed evidence about Trump’s conduct, and providing preliminary assessments about whether that conduct might or might not satisfy the elements of an obstruction-of-justice offense.

Third, some have argued that Mueller made a more fundamental legal mistake about the application of the obstruction-of-justice statutes to the President in the first instance.  According to these critics—most prominently Jack Goldsmith in a recent post–Mueller was wrong to conclude that the relevant obstruction statutes, in particular 18 U.S.C. § 1512(c)(2), apply at all to a President’s exercise of Article II authorities, such as when he exercises the power of removal or when he directs the Attorney General to investigate or prosecute—or not to investigate or prosecute—particular individuals.

All three of these critiques are mistaken, in my view.  I’ll discuss the first two of them in a forthcoming post.  This post is devoted to the third criticism, involving Mueller’s statutory interpretation.

The analysis in question appears on pp. 169-171 of Volume II of the Mueller Report.  On those pages, Mueller addresses what he calls a “threshold statutory construction principle that is unique to the presidency”: a so-called “‘clear statement rule’” that “‘general statutes must be read as not applying to the President if they do not expressly apply where application would arguably limit the President’s constitutional role’” (quoting Application of 28 USC. § 458 to Presidential Appointments of Federal Judges, 19 Op. O.L.C. 350, 352 (1995)).

In the March 2018 memorandum he wrote to DOJ in his personal capacity, William Barr argued that this “clear statement rule” foreclosed Mueller from investigating Trump for § 1512(c)(2) violations.

Mueller’s contrary analysis in Volume II of his Report is, in effect, a rejection of Barr’s view.  Jack Goldsmith, however, now offers support to Barr’s critique (albeit not every particular of Barr’s analysis), concluding that Mueller’s statutory interpretation “exposed the president to much more potential criminal liability than a proper analysis would allow.”

As I’ll explain presently, I think Mueller has the better of the argument, for three reasons:  First, there’s no such categorical “clear statement rule” to begin with; second, § 1512(c)(2) simply doesn’t bear the construction that Barr and Goldsmith propose; and third, even if there were a “clear statement rule” of construction, § 1512(c)(2), as Mueller construes it, does not implicate that rule, even as applied to Trump’s exercise of his removal and directory authorities.

* * * *

Before turning to the merits, however, it’s very important to stress the relative insignificance of this debate, and thus to apologize for prolonging it, which unfortunately distracts from the much more important lessons that Volume II of the Mueller Report offers.

For one thing, several of the incidents Mueller discusses did not involve Trump acting in his official capacity.  The application of § 1512(c)(2) to those incidents thus wouldn’t trigger the alleged “clear statement rule” that’s the subject of Jack Goldsmith’s critique of Mueller.  Most obviously this includes the efforts by Trump’s lawyers, presumably at his direction, to induce Paul Manafort and Michael Cohen (including possibly by dangling the prospect of pardons) to refuse to cooperate with Mueller’s investigation—and, in Cohen’s case, to provide false information to Congress.  It would also be a stretch to argue that Trump was even nominally exercising his Article II authorities when he implored Corey Lewandowski, a private citizen, to persuade Jeff Sessions to “unrecuse” himself and subsequently limit the scope of Mueller’s investigation; and when Trump personally directed Sessions to “unrecuse” himself even though Trump’s lawyers had informed him that Sessions’s recusal was legally required.  Trump’s statement to FBI Director Comey that he “needed loyalty” in connection with the investigation, and his direction to Don McGahn to create false records about Trump’s earlier directives to McGahn involving Muller’s removal, might also fall within this category.

With respect to those incidents in which Trump did actually or “facially” exercise Article II functions—such as removing Comey; directing McGahn to urge Acting Attorney General Rosenstein to remove Mueller; (possibly) directing Sessions to limit Mueller’s jurisdiction; or (possibly) asking Comey to “let[] Flynn go”[1]–Mueller apparently concluded (and I agree) that whether such conduct supports obstruction charges and, if so, whether to ask a grand jury to indict, are questions that should be left for the Attorney General to make after Trump leaves office (something I’ll discuss in my follow-up post).

That said, I think it’s almost inconceivable that any Attorney General would ever approve prosecuting a president on the basis of such conduct, even after his term in office, at least absent unequivocal evidence that the president would not have taken the actions but for a corrupt, i.e., self-serving, motive.  As I wrote when Barr’s 2018 memo was made public, an Attorney General would almost certainly conclude (and I agree) that “it’d be deeply unfortunate and unwise for DOJ to begin a practice of investigating whether a ‘facially legitimate’ exercise of the President’s Article II authorities was in fact motivated by illegitimate—i.e., self-interested—considerations, and whether such unconstitutional considerations were a ‘but for’ cause of the President’s actions.”  As I explained then, it’s very doubtful that any Attorney General “would conclude that it’s a question a criminal jury should determine, except perhaps in extreme circumstances.”  More broadly, there’s very good reason DOJ has never sought to prosecute a past president; and no future Attorney General would be eager to deviate from that norm, at least absent very compelling circumstances of unequivocal abuse of office. (See also page 179 of Volume II of Mueller’s Report, discussing the institutional safeguards that would protect a past president from baseless, abusive or politicized prosecutions after he leaves office.)

In any event, no current question, or decision, turns on whether such conduct did, or could, constitute a violation of § 1512(c)(2).

For these reasons, not much of practical significance turns on the debate about whether Mueller did or didn’t properly apply an alleged “clear statement rule” in construing § 1512(c)(2).

Of course, if Mueller had concluded at the outset that some of the incidents could not constitute obstruction violations as a matter of law, then this particular report to Barr, about Mueller’s prosecutorial decisions, would not have recounted them in such detail.  Even in that case, however, it would have been well within Mueller’s counterintelligence function to investigate such incidents, and to report his assessment of them to Barr, the Intelligence Community, and the congressional intelligence committees, because they are the central elements of the President’s concerted efforts to undermine the investigation of the Russian threat and to minimize or disregard the importance of that threat, which are in turn part and parcel of his grossly and inexplicably aberrant conduct toward Russia (e.g., Helsinki; refusing to allow witnesses to or transcriptions of his meetings with Putin; disclosing highly classified information to Russian diplomats in the Oval Office; etc.)—the cumulative nature of which naturally raises the question of whether Trump is compromised with respect to his duty to protect the nation from such foreign threats.  All of which is to say that Mueller would almost certainly have investigated and reported the incidents in question–albeit perhaps without the extensive legal analysis–even if he had concluded that § 1512(c)(2) did not apply to some (but not all) of the conduct described.

Which brings me to my final preliminary point:  As I’ve tried to stress several times on this site, assessing whether Trump did or did not violate the criminal obstruction statutes simply isn’t germane to—and unfortunately distracts attention from–the most important questions now facing the nation (Congress and the electorate, in particular) as a result of the Russia investigation, namely:  assessing the nature, scope and continuing implications of the Russian threat; assessing whether and how the President might be compromised in his ability to deal with that threat, and with Russia more broadly, going forward (with respect to which this new interview with HPSCI Chair Adam Schiff about the administration’s failure to brief the Intelligence Committees about the ongoing counterintelligence investigation is a must-read); and, most importantly for purposes of Volume II of the Report, assessing whether the President has violated his oath of office and failed to comply with his constitutional duty to take care that the laws are faithfully executed.

Volume II describes in extraordinary detail a presidential abdication of constitutional responsibility more extreme and more unrelenting than any other in our history.  Whether it also describes one or more violations of section § 1512(c)(2) is, if not quite irrelevant to assessing how we should now proceed, certainly trivial in comparison and secondary (at best) in importance.  (Oh, and by the way—and as Jack agrees—the Volume II narrative also makes crystal clear “that Trump committed impeachable offenses,” even if there’s no chance that the Republican majority in the Senate would ever convict him, indifferent as they are to the horrifying breakdown in constitutional norms and firewalls described in the Mueller Report.)

So please, let’s not get too distracted about Mueller’s interpretation of § 1512(c)(2) to cover certain presidential actions.

* * * *

OK, but what about the merits of the § 1512(c)(2) question?

Section § 1512(c)(2) establishes criminal culpability for “[w]hoever corruptly . . . obstructs, influences, or impedes any official proceeding, or attempts to do so.”  How is it that that statute might be construed not to cover presidential actions undertaken in order to “attempt” to “corruptly obstruct” an official proceeding such as a grand jury investigation?

Jack Goldsmith, like Bill Barr before him, relies upon two statements from a 1995 OLC opinion signed by Walter Dellinger, the first of which Mueller himself quotes in the Report:

— “[G]eneral statutes must be read as not applying to the President if they do not expressly apply where application would arguably limit the President’s constitutional role.”

— “[S]tatutes that do not expressly apply to the President must be construed as not applying to the President if such application would involve a possible conflict with the President’s constitutional prerogatives.”

Barr and Goldsmith argue that this alleged canon of statutory construction categorically precludes application of § 1512(c)(2) to at least some types of presidential conduct, such where the President purports to exercise his (implicit) constitutional powers to remove officers and to direct criminal investigations and prosecutions.  (As I note in footnote 1, above, there’s some question whether the President has authority under current federal statutory law to direct the conduct of individual criminal investigations, but for present purposes I’ll assume that he does.)

When Barr’s argument about this so-called “plain-statement rule” emerged, I wrote this:

In my view the executive branch (including OLC) invokes this alleged “canon” far too indiscriminately.  But that’s a (complicated) topic for another day.  For now, what’s important is that Barr begs the critical question:  To be sure, an application of section 1512(c) to criminalize a proper exercise of, say, the President’s removal authority would be unconstitutional—indeed, such conduct wouldn’t satisfy the statutory standard itself because it wouldn’t be “corrupt.”  But if the President did not have the constitutional “prerogative” to use his Article II power in a particular way—say, for instance, firing an FBI Director in an effort to prevent DOJ from uncovering genuinely criminal conduct that’s worthy of investigation and possible prosecution—there’s no reason to construe the criminal statute not to apply to that purported (but illegitimate) exercise of presidential power, let alone to immunize the President categorically from the statute’s prohibition in all cases.  Once again, I’m confident any DOJ would be deeply reluctant to put to a jury the question of whether a legitimate, articulated presidential motive was a pretext for wrongful self-dealing.  But it doesn’t follow from that altogether salutary institutional reluctance that the statute itself should be construed not to cover any and all presidential exercises of an Article II function.

That basically captures the primary reason why I believe Barr and Goldsmith are wrong.  I discuss it in greater detail at the end of this post.  Yet there are also two other reasons to question the critique of Mueller’s statutory analysis.

First, I was actually a bit too generous in writing that the executive branch “invokes this alleged ‘canon’ far too indiscriminately.”  The truth is that there is no such established canon, or “clear statement rule,” notwithstanding its appearance in a handful of OLC opinions in the Clinton Administration.  (Disclosure: I worked at OLC then.)

To be sure, courts and the executive branch will sometimes construe statutes, at least where their language will bear such a reading, to preclude application to certain executive actions where such applications would raise serious constitutional questions.  That’s an ordinary example of the application of the “constitutional avoidance canon,” which the Dellinger opinion invokes as the principal justification for the more extreme “clear statement” requirement.  In other cases (such as the one Deputy Attorney General Silberman examined in a 1974 opinion Jack cites), the legislative history or context might fairly be read to support the view that Congress did not intend for a statute to apply to some or all presidential actions.

But a general rule that a statute must be read as not applying to the President if it doesn’t expressly do so “where application would arguably limit the President’s constitutional role”?  That particular, categorical canon of construction doesn’t exist.

Or at the very least, the Supreme Court has never adopted it (and it didn’t really appear in such form in DOJ opinions before 1995, either). OLC has principally relied upon two Supreme Court opinions—but the “express statement” requirement doesn’t appear in either of them.  Which is hardly surprising, not only because such a rule would have dramatic effects upon many statutes, often where Congress obviously did mean to cover the President, but also because neither of those two cases involved any question of whether a statute could or should be applied to regulate the President’s exercise of an Article II authority.

Public Citizen v. DOJ was about whether the disclosure requirements of the Federal Advisory Committee Act applied to the American Bar Association’s reports to the Justice Department about prospective judicial nominees.  (There wasn’t any issue in the case involving whether FACA applied to the President himself.)  Relying in part upon the general constitutional avoidance canon—but primarily on the fact that the Court’s “review of the regulatory scheme prior to FACA’s enactment and the likely origin of the phrase ‘or utilized’ in FACA’s definition of ‘advisory committee’ reveal[ed] that Congress probably did not intend to subject the ABA Committee to FACA’s requirements when the ABA Committee offers confidential advice regarding Presidential appointments”—the Court held that the word “utilized” did not describe the Justice Department’s use of the ABA Committee and therefore FACA did not apply to that ABA function.

Franklin v. Massachusetts asked whether the Administrative Procedure Act’s provision for judicial review to consider whether an agency’s exercise of discretion is “arbitrary and capricious” applies to the President when he exercises a statutory (not constitutional) authority.  The Court held that the President is not an “agency” for such APA purposes:  Invoking “respect for the separation of powers and the unique constitutional position of the President,” the Court concluded that “[w]e would require an express statement by Congress before assuming it intended the President’s performance of his statutory duties to be reviewed for abuse of discretion.”

In neither case did the Court announce any categorical requirement that Congress must “expressly” provide for coverage of the President whenever application of a statute would (in OLC’s words) “involve a possible conflict with the President’s constitutional prerogatives.”  (Indeed, in Public Citizen, the Court extensively examined “arguments based on FACA’s text and legislative history,” and concluded that they “tend to show that Congress did not desire FACA to apply to the Justice Department’s confidential solicitation of the ABA Committee’s views on prospective judicial nominees.”  That entire analysis would have been unnecessary had the Court applied the “clear statement rule” that OLC purports to find in the Public Citizen decision.)

It’s also worth noting that in landmark cases implicating the President’s ability to perform his constitutional role, where one would most expect such a rule of construction to be applied, or at least debated, there’s nary a mention of such a clear statement rule.  In Clinton v. Jones, for example, no advocate or Justice even suggested that the Federal Rules of Civil Procedure–which do not expressly mention the President–should or must be construed to exclude the President from civil process in cases where he’s sued for conduct preceding his term in office.  And, most importantly, in the Nixon tapes case Richard Nixon’s lawyers proffered plenty of aggressive arguments in support of the President’s prerogatives, but they never thought to contend that Federal Rule of Criminal Procedure 17(c)–which requires witnesses “to produce any books, papers, documents, data, or other objects [a judicial] subpoena designates”–must be construed to exempt the President, even in a case where the subpoenaed tapes were recordings of conversations between the President and his closest advisors.  Nor did any Justice on the Court raise any such statutory construction argument.

Because the Supreme Court has never recognized the “clear statement rule” described in the 1995 OLC opinion; because the Court hasn’t mentioned it in important cases where it presumably would have been applicable; and because such a rule would lead to deeply counterintuitive results in some cases where Congress plainly did mean to regulate the President even without “expressly” saying so, it’s hardly surprising that OLC and Justice Department prosecutors have often not even mentioned the so-called “clear statement rule” in discussing the application of “generally applicable” statutes to conduct undertaken, directed or authorized by the President.

Consider, for example, the notorious 2002 OLC opinion on the torture statute.  That opinion contains plenty of erroneous and tendentious analysis, both statutory and constitutional—including its mistaken view that the Constitution forbids Congress from regulating how the President, acting as Commander-in-Chief, treats prisoners in an armed conflict.  Yet even in that very aggressive, audacious opinion, OLC did not rely upon the so-called “clear statement” rule to exclude presidentially authorized interrogations, even though the torture statute does not mention the President expressly, and even though application of the statute would have prevented the President from engaging in conduct that (at least according to the 2002 OLC) would have otherwise been a permissible exercise of the President’s Article II powers.  Indeed, the 2002 opinion did not even mention such a “clear statement rule,” because it would have been absurd to do so:  Of course Congress intended to foreclose the President from ordering torture—it didn’t need to say so “expressly.”

Similarly, as Ben Wittes notes in his response to Jack, “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.”  Leon Jaworski didn’t do so, even though the presidential conduct he investigated included some exercises of Nixon’s directory authority.  Nor did Lawrence Walsh, when he investigated whether President Bush possibly violated the obstruction laws by exercising “his constitutional prerogative to pardon a former close associate to prevent further Iran/contra revelations.”

To be sure, Mueller himself assumed—perhaps because OLC insisted upon it—that there is such a canon of statutory construction.  And for the reasons discussed below, I think he was right to conclude that it wouldn’t preclude application of the obstruction statutes to the conduct described in Volume II of the Report.  My point here, however, is that this is all something of a false distraction:  There’s no reason to think Congress did not mean to make presidential acts subject to those prohibitions where they are done (as the statutes require) for a “corrupt” purpose; and such an application would not be unconstitutional.  That ought to be the beginning and end of the analysis.

Second, it’s noteworthy that in both Public Citizen and Franklin—purportedly the sources of the “clear statement rule”—the Court construed an actual term of the law (“utilized”; “agency”) to avoid an application to, or affecting, the President.

In the case of § 1512(c)(2), by contrast, the operative term is “whoever.”  If I understand them correctly, Barr and Goldsmith concede that this word can encompass the President in at least some cases (such as when he suborns perjury or threaten witnesses).  Yet there’s no language in the statute—certainly not the word “whoever”—that might be parsed to exclude only those instances in which the President purports to be exercising an Article II authority.[2]

This is what Mueller is getting at on page 170 of Volume II when he writes that such a gloss—distinguishing among types of presidential actions, such that some would be potentially covered by § 1512(c)(2) and others (e.g., removals) would be categorically exempted–“would be difficult to implement as a matter of statutory interpretation” because the relevant terms in the corruption statutes could not “easily bear that specialized meaning.”  Jack contends that this argument “makes no sense,” but I don’t see why that’s so.  It’s one thing to say that the President is not an “officer” or “agency” covered by a statute that uses such terms; it’d be quite another to say that the word “whoever” can be construed to cover certain presidential actions but not others.  Just as the ordinary constitutional-avoidance canon “is not a license for the judiciary to rewrite language enacted by the legislature,” Salinas v. United States, 522 U.S. 52, 59–60 (1997); see also Seminole Tribe of Fla. v. Florida, 517 U.S. 44, 57 n.9 (1996) (“We cannot press statutory construction to the point of disingenuous evasion even to avoid a constitutional question.”), the same presumably should and would be true of any such “clear statement” canon.[3]

(In this regard it’s noteworthy that in Public Citizen three of the Justices–Justice Kennedy, joined by Chief Justice Rehnquist and Justice O’Connor–concluded that the text of FACA could not fairly be construed to exclude the ABA Committee.  Instead of then applying any sort of “clear statement rule” to reach the same statutory result, however, they found it necessary to resolve the case on constitutional grounds.)

Third and finally, even if Walter Dellinger’s 1995 opinion were correct to recognize a canon that “statutes that do not expressly apply to the President must be construed as not applying to the President if such application would involve a possible conflict with the President’s constitutional prerogatives,” Mueller was right to conclude that § 1512(c)(2) comfortably covers the Trump conduct described in Volume II of the Report notwithstanding that rule of construction.

In the portion of his report discussing the “clear statement rule,” Mueller stresses that § 1512(c)(2) only prohibits corrupt efforts to obstruct or influence proceedings, which requires an “intent to secure an improper advantage “inconsistent with official duty and the rights of others.”  Thus, as Mueller explains in the next subsection of his report, the obstruction statutes properly construed “restrict presidential action only by prohibiting the President from acting to obstruct official proceedings for the improper purpose of protecting his own interests” (p.174; first emphasis added), such as “with the intention of shielding himself from criminal punishment, avoiding financial liability, or preventing personal embarrassment” (id.).

So, for example, although the obstruction statutes could be applied to forbid removal of an official for a corrupt reason—in Mueller’s words, “a reason grounded in achieving personal rather than official ends”—such a “narrow and discrete limitation on removal . . . would leave ample room for all other considerations, including disagreement over policy or loss of confidence in the officer’s judgment or commitment.  A corrupt-purpose prohibition therefore would not undermine the President’s ability to perform his Article II functions.”  [UPDATE:  As I noted in my preliminary remarks above, and in previous posts, it’s inconceivable that any Attorney General would ever seek charges unless there were evidence beyond a reasonable doubt that a former President’s “corrupt” motives were a necessary, i.e., a “but for,” cause of his or her actions.  And I don’t read Mueller to be suggesting that prosecution would be proper otherwise.]

Indeed, construed as such, the law actually “furthers, rather than hinders,” the President’s ability to perform his constitutional functions, which include taking care to ensure “the impartial and evenhanded administration of the law” (id.)  Mueller is right that the Constitution “does not mandate that the President have unfettered authority to direct investigations or prosecutions, with no limits whatsoever, in order to carry out his Article II functions” (id.).  In fact, as David Pozen and I discussed here two years ago, the presidential oath and his constitutional duty to take care the laws are faithfully executed prohibit the President from using his Article II powers for self-interested ends—especially when the President tries to undermine enforcement of the law for such purposes, as he did repeatedly with the Russia investigation.  Thus, as Mueller emphasizes at page 8 of Volume II of his Report, the “corruptly” standard in § 1512(c)(2) actually “aligns with the President’s constitutional duty to faithfully execute the laws.”

Because § 1512(c)(2), as so (properly) construed by Mueller, would not even raise a serious constitutional question, there’d be no compelling reason for applying the OLC “clear statement rule” of statutory construction, even accepting OLC’s rationale for the rule.  See 19 Op. O.L.C. at 352 (“The principle that general statutes must be read as not applying to the President if they do not expressly apply where application would arguably limit the President’s constitutional role has two sources,” one of which is the “long-recognized ‘cardinal principle’ of statutory interpretation that statutes be construed to avoid raising serious constitutional questions.”).  Indeed, OLC specifically concluded (id. at 357 n.11) that “[t]he clear statement principle we have identified does not apply with respect to a statute that raises no separation of powers questions were it to be applied to the President.”

Even if one were to assume that the “clear statement rule” applies here, however, it’s easy to see why applying § 1512(c)(2) to the President, even in the context of removal or direction of law enforcement, would not violate that “rule”—would not require an “express” reference to the President in order to apply to Trump’s conduct described in the Report–because such an application would not, in Dellinger’s words, “arguably limit the President’s constitutional role.”  The statutory requirement of proving that Trump acted corruptly–to shield himself from criminal punishment, avoid financial liability, or prevent personal embarrassment, rather than to advance public endsmeans that the statute can’t be applied to “limit the POTUS’s constitutional role,” because such conduct isn’t within his constitutional role to begin with, even apart from the obstruction statute.  In fact, the Constitution prohibits it.


[1] I’ve included the parenthetical “possibly”s in that sentence because it’s actually not clear that the President has authority to direct DOJ officials about whom to prosecute or not prosecute, at least apart from cases in which the President believes that such a prosecution would be unlawful.  It’s a closer and much more complicated question than many people assume.  It’s unnecessary to resolve that question here, however:  My analysis proceeds on the assumption that the President has such a “directory” authority.

[2] As I discuss below, however, the word “corruptly” does exclude from § 1512(c)(2)’s coverage the proper exercise of the President’s Article II powers—which is another reason why it’s untroubling to apply the statute to the President and thus why the so-called “clear statement rule” would be no obstacle.

[3] Jack is right that in a couple of opinions OLC has construed a statute not to apply to the President even without purporting to construe a particular statutory term that might arguably be understood to exclude the President.  That’s another reason to be skeptical of those OLC opinions—but in any event, those OLC opinion didn’t purport to do what would be necessary here, namely, to construe statutes to cover some categories of presidential conduct but not others.



About the Author(s)

Marty Lederman

Professor at the Georgetown University Law Center. He was Deputy Assistant Attorney General at the Office of Legal Counsel from 2009-2010, and Attorney Advisor at the Office of Legal Counsel from 1994-2002. You can follow him on Twitter (@marty_lederman).