Justice Thomas and Judge Rao: A Tale of Two Mazars Dissents

Justice Thomas’s dissent in Trump v. Mazars USA, LLP, 591 U.S. __ (2020), has been compared to Judge Rao’s dissent in the D.C. Circuit below, with the implication that this somewhat vindicates Rao’s widely panned opinion. However, the two dissents are in fact quite different, and it seems pretty clear that Justice Thomas was not persuaded by his former clerk’s opinion.

To be sure, there are similarities between the two dissents. Both ignore the presidency-centered arguments offered by President Trump’s personal legal team and the Department of Justice in favor of broader theories not raised by any party or amicus. Both evince skepticism if not outright hostility toward legislative investigations generally and clearly prefer the stance taken by the Court in Kilbourn v. Thompson, 103 U.S. 168 (1881) to that of McGrain v. Daugherty, 273 U.S. 135 (1927). Both indicate that these congressional subpoenas seeking the president’s personal financial information would be valid, if at all, only through the exercise of the impeachment power. Both rely to a great extent on historical practice, particularly a kind of negative historical practice (i.e., drawing conclusions from things that allegedly did not happen).

Despite these similarities, the two dissents employ different reasoning, rely on different “precedents,” and reach very different conclusions. The textual and structural lynchpin of Judge Rao’s analysis is the impeachment power, which she claims “provides the exclusive method for Congress to investigate accusations of illegal conduct by impeachable officials, particularly with the aid of compulsory process.” Trump v. Mazars USA, LLP, 940 F.3d 710, 751 (D.C. Cir. 2019) (Rao, J., dissenting). While Rao acknowledges Congress’s general legislative power to conduct investigations, this power in her view does not extend to matters within the “impeachment zone” (my term, not hers). How one determines what falls within the impeachment zone is somewhat unclear, but Rao finds that the subpoenas for Trump’s financial information are ones that can only be pursued through the impeachment power.

Much of Rao’s opinion is devoted to her contention that “consistent historical practice” supports her conclusion. Id. at 753. To wit, she endeavors to show that Congress has never investigated matters within the impeachment zone except through the use of the impeachment power. Id. at 758-67.

Thomas takes a different approach. Although he agrees with Rao that the personal financial records at issue may be obtained by Congress, if at all, only through the exercise of the impeachment power, he reaches this result because he believes Congress lacks the power to subpoena private, nonofficial documents in any legislative investigation. Unlike Rao, he does not claim that the impeachment power somehow displaces otherwise available legislative power to investigate.

A simple illustration of the difference between the two dissents is to imagine Donald Trump had never been elected president. (It’s easy if you try). Under Rao’s theory, Congress would be able to subpoena his financial records for legislative purposes, such as to investigate the operation of money-laundering laws (which was the asserted purpose of the subpoena issued by the Committee on Financial Services). Under Thomas’s theory, on the other hand, these records could never be subpoenaed for a legislative purpose.

Put another way, Thomas would proscribe a particular legislative tool (subpoenas for private documents) for all legislative investigations, while Rao would proscribe the use of any compulsory process for certain investigative subjects (legislative investigations of matters falling within the impeachment zone). Thus, Rao would allow subpoenas for private documents in legislative investigations outside the impeachment zone; Thomas would not. Thomas would allow subpoenas for testimony or official documents in legislative investigations within the impeachment zone; Rao would not. Though they produce the same result in this particular case, the two theories are entirely different.

Furthermore, Thomas evidently rejects Rao’s interpretation of historical practice. While Rao claims that investigations of wrongdoing by impeachable officials have occurred exclusively through the exercise of the impeachment power, Thomas makes this observation:

     For nearly two centuries, until the 1970s, Congress never attempted to subpoena documents to investigate wrongdoing by the President outside the context of impeachment. Congress investigated Presidents without opening impeachment proceedings. But it never issued a subpoena for private, nonofficial documents as part of those non-impeachment inquiries.

Trump v. Mazars USA, LLP, 591 U.S. __, slip op. at 20-21 (2020) (Thomas, J., dissenting) (citation omitted) (emphasis added).

As Thomas acknowledges, Congress has investigated wrongdoing by presidents (not to mention other impeachable officials) in “non-impeachment inquiries” both before and after the 1970s. Whatever the factual accuracy or legal significance of the claim that these non-impeachment inquiries did not subpoena private, nonofficial documents “until the 1970s,” Thomas clearly does not see the historical pattern asserted by Rao as the key to her dissent.

None of this is to say that Justice Thomas’s dissent is correct (it is not) or even more plausible than Judge Rao’s (we will leave that to another day). It is fair to say, though, that Thomas was unpersuaded by Rao’s analysis and finds his own to be more plausible.

Will the Mazars Court Overrule McGrain? (Part Two)

As suggested in my last post, the May 12, 2020 oral argument in Trump v. Mazars USA, LLP did not go well for the House, to put it mildly. Most of the tough questions for the House Counsel clustered around a single idea: what is the limiting principle that prevents Congress from prying into whatever it wants, whenever it wants? Before getting to that, however, let’s consider an even more fundamental issue raised by Justice Thomas.

Justice Thomas began his questioning of House Counsel Doug Letter by essentially asking what the constitutional basis is for recognizing the power to issue legislative subpoenas at all. Tr. 54-55. Letter responded by pointing to the long line of Supreme Court cases (which began with McGrain) holding that the power to conduct investigations and issue compulsory process is an inherent and integral part of the legislative power conferred by the Constitution.

Justice Thomas did not appear entirely satisfied with this answer, and he followed up by asking “can you give me the earliest example you have of Congress issuing a legislative subpoena?” Tr. 56. Letter pointed to the House’s 1792 investigation of General St. Clair’s failed expedition. This investigation was viewed by the McGrain Court as significant historical evidence of the existence of a constitutional power to issue legislative subpoenas. As the Court explained:

This power was both asserted and exerted by the House of Representatives in 1792, when it appointed a select committee to inquire into the St. Clair expedition and authorized the committee to send for necessary persons, papers and records. Mr. Madison, who had taken an important part in framing the Constitution only five years before, and four of his associates in that work, were members of the House of Representatives at the time, and all voted for the inquiry.

*           *          *

We are of opinion that the power of inquiry– with process to enforce it– is an essential and appropriate auxiliary to the legislative function. It was so regarded and employed in American legislatures before the Constitution was framed and ratified. Both houses of Congress took this view of it early in their history– the House of Representatives with the approving votes of Mr. Madison and other members whose service in the convention which framed the Constitution gives special significance to their action– and both houses have employed the power accordingly up to the present time.

McGrain v. Daugherty, 273 U.S. 135, 161, 174 (1927).

Still not satisfied, Thomas pressed further: “What’s the first example of Congress issuing a legislative subpoena to a private party for documents?” Tr. 56. Letter could not answer him directly, but referred him to the discussion of congressional investigatory history in Watkins v. United States, 354 U.S. 178 (1957).

The referenced passage in Watkins, I think, is the following:

Most of the instances of use of compulsory process by the first Congresses concerned matters affecting the qualification or integrity of their members or came about in inquiries dealing with suspected corruption or mismanagement of government officials. [Note: here the Court cites to Landis’s article]. Unlike the English practice, from the very outset, the use of contempt power by the legislature was deemed subject to judicial review.

     There was very little use of the power of compulsory process in early years to enable the Congress to obtain facts pertinent to the enactment of new statutes or the administration of existing laws. The first occasion for such an investigation arose in 1827, when the House of Representatives was considering a revision of the tariff laws. In the Senate, there was no use of a factfinding investigation in aid of legislation until 1859.

Watkins, 354 U.S. at 192-93.

This passage does not specifically answer Justice Thomas’s question, but it suggests why it may not have been exactly the right question. While courts pass on the validity of specific subpoenas, the scope of Congress’s investigatory authority is determined by reference to the investigation that is being conducted, not by the nature of an individual subpoena (e.g., whether it is directed to a private party or seeks documents).

Thus, for example, the investigation of the St. Clair expedition would be one of the inquiries involving “suspected corruption or mismanagement of government officials” referred to in Watkins, but that does not  mean the investigation lacked the power to compel the production of documents or other information from private parties. Indeed, one of the issues in the St. Clair investigation was the quality of military supplies provided by private contractors, and the committee received affidavits and other evidence from these contractors. See I Arthur M. Schlesinger, Jr. & Roger Bruns, eds., Congress Investigates: A Documented History: 1792-1974 95 (1983). Whether or not the committee actually issued compulsory process to a private party, there seems little doubt it had the authority to do so.

When was the first occasion on which a congressional committee actually issued a legislative subpoena to a private party for documents? The earliest I can verify is that in 1827 a House committee investigating John Calhoun’s prior administration of the War Department subpoenaed documents from an unsuccessful bidder on a government contract. See 3 Reg. of Debates in Cong. 1124 (Feb. 13, 1827). However, the House’s 1810 investigation of General James Wilkinson also obtained testimony and documents from a number of private individuals, at least some of which was obtained by compulsory process. I Congress Investigates at 119 & 170.

The passage quoted from Watkins does not distinguish between subpoenas directed to private parties and government officials, but it does suggest a distinction between (1) investigations of suspected government corruption or mismanagement (what would often be referred to as congressional oversight) and (2) inquiries to obtain facts relevant to enacting or amending legislation. Although both are “legislative” in nature, the Court implies that the latter requires more vigorous scrutiny to ensure that the information sought is pertinent to the investigation, particularly when the information sought would implicate the constitutional rights of private citizens.

This interpretation is consistent with the holding of Watkins, where a labor organizer summoned to testify before the House Un-American Activities Committee testified freely about his own activities and associations, but refused to answer questions about individuals whom he believed may have once been but no longer were members of the Communist Party. The Court reversed his conviction for contempt of Congress, holding that the committee violated his rights under the contempt statute and the due process clause by failing to clearly explain to him the pertinency of the questions to its investigation. It did not dispute that political opinions and associations protected by the Bill of Rights could nonetheless be a proper subject of congressional investigation, but “[p]rotected freedoms should not be placed in danger in the absence of a clear determination by the House or the Senate that a particular inquiry is justified by a specific legislative need.” Watkins, 354 U.S. at 198, 205.

The Court emphasized that it was not dealing with congressional oversight, noting that “[t]he public is, of course, entitled to be informed concerning the workings of its government.” Id. at 198. It explained:

     We are not concerned with the power of Congress to inquire into and publicize corruption, maladministration or inefficiency in agencies of the Government. That was the only kind of activity described by Woodrow Wilson in Congressional Government when he wrote: “The informing function of Congress should be preferred even to its legislative function.” From the earliest times in its history, Congress has assiduously performed an “informing function” of this nature. See Landis, Constitutional Limitations on the Congressional Power of Investigation, 40 Harv. L. Rev. 153, 168-194.

Watkins, 354 U.S. at 200 n.33 (citation omitted). The Court thus distinguishes the inquiry in Watkins from the type of congressional oversight involved in McGrain.

This distinction may help point the way to an answer to the question asked by many of the justices at the May 12 argument in Mazars, i.e., what stops Congress from investigating virtually anything on the basis that it has some connection to a subject on which legislation could potentially be had. See, e.g., Tr. 52-54 (Chief Justice Roberts); 57 (Justice Ginsburg); 64 (Justice Alito); 74 (Justice Kavanaugh). Letter had some difficulty answering this question, perhaps because judicial doctrine since McGrain has in fact been extremely deferential to Congress on this score. As the district judge in Mazars pointed out, the governing legal standards are so deferential that they “do not substantially constrain Congress.”

However, the real constraint on Congress is that enforcing a subpoena is extremely cumbersome and therefore legal sanctions for contempt are virtually never imposed. This is in part because the Court in cases like Watkins has imposed technical and procedural requirements for criminal contempt to address the very issue raised in the Mazars argument. See Watkins,  354 U.S. at 204 (expressing concern that the committee “can radiate outward infinitely to any topic thought to be related in some way” to its mandate, that “[r]emoteness of subject can be aggravated by a probe for a depth of detail even farther removed from any basis of legislative action,” and further that “investigators [can] turn their attention to the past to collect minutiae on remote topics, on the hypothesis that the past may reflect upon the present”).

As a consequence, any witness who wishes to contest a congressional subpoena has far more leverage than the formal legal standards would imply. In addition, witnesses have the right to assert privileges, including the privilege against self-incrimination. Congress also has political incentives which further constrain its exercise of the subpoena power. Thus, the hypotheticals advanced by the justices are, for the most part, very unlikely to occur. See, e.g., Tr. 85-86 (Justice Alito) (suggesting the possibility that one house of Congress might subpoena personal records relating to a member of the other house).

Of course, some of these safeguards are inoperative in the Mazars case because it presents the fairly rare scenario of Congress seeking non-privileged records from third parties with no interest in contesting the subpoenas. Whether this  creates a significant potential of congressional abuse is debatable. After all, if Congress were to attempt to exercise this authority in an excessive or abusive manner, banks and other third party record keepers would have an incentive to contest subpoenas to protect the interests of their clients. (This is why the Court would have been wise to consider this blog’s suggestion that only the third parties themselves should have standing to contest the validity of the subpoenas).

Nonetheless, there are undoubtedly instances where Congress investigates particular factual questions which seem tenuously related to a legislative need. It is difficult to see, for example, why Congress would need to know whether a particular baseball player used steroids in order to legislate on the general subject. One could reasonably argue that if Congress is merely seeking information as a case study of a particular social, economic or national security problem, it ought to explain not only how the information is pertinent to potential legislation but why there is a legislative need to explore one specific example out of many. This should be more than adequate to protect against some of the other hypotheticals raised in the May 12 argument, such as the idea that Congress could subpoena an individual’s medical records on the ground it was considering healthcare legislation. See Tr. 65 (Justice Sotomayor).

On the other hand, there is no need for Congress to provide any additional justification for conducting oversight of government agencies and officials. As explained in McGrain and Watkins (and detailed in Professor Landis’s article, among other places), Congress has conducted searching probes into the conduct of government officials and operations since its earliest days. Such investigations are inherently justified by the need to inform itself and the public as to the working of the federal government and to uncover corruption, maladministration and inefficiency of every kind.

This distinction is reflected in Justice Kagan’s suggestion that there may be reasons for treating differently the three congressional subpoenas involved in the consolidated Mazars and Deustche Bank cases. See Tr. 88. Although all three seek similar types of information (financial records relating to President Trump’s private business interests), there are significant differences in the nature of the investigation to which each subpoena relates. The investigation by the Financial Services committee seeks the information simply to use it as a case study of a much more general problem (money laundering) in the financial sector. By contrast, the subpoena from the Intelligence committee is for the purpose of determining whether the president has financial ties to Russia or other foreign actors that might create a conflict of interest or give such actors leverage over his official decision making. The latter falls squarely within the province of congressional oversight while the former constitutes a pure case study investigation that may require additional justification.

The subpoena from the House Oversight committee falls somewhere in the middle. It is defended in part on the ground that it will assist the committee in determining whether to recommend changes to disclosure laws applying to federal officials generally. This is arguably closer to a case study approach, although it seems self-evident why the committee would focus on the highest-ranking federal official, particularly when it has gathered substantial evidence that he has been less than truthful in his private financial disclosures. In addition, the subpoena can be justified on the pure oversight grounds of determining whether the president has financial conflicts of interest or is in violation of the Foreign Emoluments Clause.

The line suggested by Justice Kagan would allow the Court to uphold at least one and likely two of the congressional subpoenas, while sending the other(s) back for further proceedings. It seems to me this would be a reasonable compromise that would satisfy the concerns expressed by the justices (with the possible exception of Justice Thomas) without fundamentally disturbing the legal standards established by McGrain and applied in subsequent cases.

Unlike Kagan (and several of her colleagues), however, I would be loathe to establish a special protection applicable only to the president. Historically the Court’s concerns about over broad congressional investigations focus on protecting the affairs of private citizens from arbitrary scrutiny. Even Judge Cochran, who would have applied these principles to an inquiry into the conduct of (then former) Attorney General Daugherty, claimed only that these principles applied as much to federal officials as to private citizens, not that the former were entitled to additional protection. (To date only Judge Rao, in her Mazars dissent in the D.C. Circuit, has advanced the remarkable proposition that impeachable officials enjoy an immunity from legislative investigation that is unavailable to private citizens). If the Court believes that changes are needed to the doctrine governing congressional case study investigations to avoid arbitrary intrusions into private affairs, such should apply to all citizens, not just the one who happens to sit in the Oval Office.

Whatever the Court ends up deciding in Mazars, let us hope they emulate the McGrain Court in one way but not another: the first by achieving unanimity or something close to it; and the second by not taking more than two years to issue a decision.

Will the Mazars Court Overrule McGrain? (Part One)

Nearly a century ago the Supreme Court decided the landmark case of McGrain v. Daugherty, 273 U.S. 135, 174 (1927), in which the Court declared that “the power of inquiry– with process to enforce it– is an essential and appropriate auxiliary to the legislative function.” In so holding, the Court dispelled doubts raised by Kilbourn v. Thompson, 103 U.S. 168 (1880), where, as we discussed here, the Court had expressed skepticism whether Congress could issue compulsory process outside the context of its judicial functions (such as impeachment and disciplining its members). McGrain settled this issue in Congress’s favor and, along with subsequent cases, established such a deferential judicial stance toward the validity of congressional investigations  that no congressional investigation since has been held to exceed Congress’s legislative powers. After listening to the oral argument in Trump v. Mazars USA, LLP, however, one has to wonder whether this will soon change.

The McGrain case arose from a Senate resolution calling for a broad investigation into the activities of Attorney General Harry Daugherty (our old friend) and his associates at the Department of Justice, including, but by no means limited to, Daugherty’s failure to pursue legal actions against individuals linked to the Teapot Dome scandal. Suspicions regarding Daugherty’s negligence or favoritism with regard to Teapot Dome, however, were the least of the attorney general’s troubles. Senate hearings in March 1924 featured blockbuster testimony from witnesses who claimed Daugherty and his associates had received large amounts of illicit cash which were deposited in a small Ohio bank run by Daugherty’s brother, Mally (“Mal”) Daugherty. The hearings led to Attorney General Daugherty’s forced resignation on March 28, 1924 and to a subsequent testimonial subpoena requiring Mal to appear before the Senate committee investigating his brother. When Mal refused to appear, the Senate ordered him taken into custody, and he immediately petitioned for a writ of habeas corpus in the federal district court for the Southern District of Ohio. (Fun fact: the judge who initially received the habeas petition was Smith Hickenlooper grandfather of the former Colorado governor and presidential candidate).

At this point matters stood at something of a crossroads. With Daugherty’s resignation, the major figures in the scandals of the Harding administration were out of office, and the new Coolidge administration (President Harding having passed away in 1923) was eager to disassociate itself from them. On the other hand, many Republicans argued that the congressional investigations into these scandals were political and excessive, and members of the bar warned that such investigations threatened civil liberties. Chief Justice Taft and Senator George Pepper, a well regarded Republican lawyer, were among the luminaries expressing skepticism about the investigations. See J. Leonard Bates, The Teapot Dome Scandal and the Election of 1924, 60 Am. Hist. Rev. 303, 317 (Jan. 1955).

While Mal Daugherty’s case was pending in the district court, a Harvard law professor named Felix Frankfurter wrote an article in the New Republic entitled “Hands off the Investigations,” which was reprinted in the Congressional Record on the day it was published. See 65 Cong. Rec. 9080-82 (May 21, 1924) (introduced by Senator Ashurst). Professor Frankfurter “came out squarely for the unlimited power of congressional investigations.” Louis B. Boudin, Congressional and Agency Investigations: Their Uses and Abuses, 35 Va. L. Rev. 143, 146 (Feb. 1949).

Frankfurter proclaimed “[i]t is safe to say that never in the history of this country have congressional investigations had to contend with such powerful odds, never have they so quickly revealed wrongdoing, incompetence, and low public standards on such a wide scale, and never have such investigations resulted so effectively in compelling correction through the dismissal of derelict officials.” 65 Cong. Rec. 9081. He sniggered at the suggestion that the Daugherty hearings were unfair because the witnesses who  testified were disreputable (sound familiar?), noting “[i]t is the essence of the whole Daugherty affair that the Attorney General of the United States was involved in questionable association with disreputable characters.” He also rejected the notion that congressional investigations should be subject to rules of evidence or other technical limitations applicable in court, asserting that “[t]he procedure of congressional investigation should remain as it is.” 65 Cong. Rec. 9082.

Just ten days later (May 31, 1924), Mal Daugherty’s habeas petition was granted by US District Judge Cochran (to whom the case for some reason had been reassigned). The court found that the Senate investigation of the (now former) attorney general was beyond the Senate’s constitutional power. See Ex Parte Daugherty, 299 Fed. 620 (S.D. Ohio 1924). Following the reasoning of Kilbourn, Judge Cochran expressed “very serious doubt” whether the Senate had the power to issue compulsory process in any legislative investigation, but he found it unnecessary to rest his decision on that ground. Instead, he reasoned that the Senate was not conducting a proper legislative investigation, but rather it was making an improper attempt to put Harry Daugherty on trial. See id. at __ (“What the Senate is engaged in is not investigating the Attorney General’s office; it is investigating the former Attorney General.”). This was a judicial function that could only be performed by a court or by the House of Representatives pursuant to its impeachment power. The court explained:

[T]he Senate has no power to impeach any Federal officer at the bar of public opinion, no matter what possible good may come of it. It is not within its province to harass, annoy, put in fear, render unfit, or possibly drive from office any such officer, high or low, by instituting such impeachment proceedings against him. The power to impeach under the Federal Constitution resides solely in the House of Representatives, and it has power to impeach solely at the bar of the Senate.

Id. at __.

Judge Cochran’s analysis in many respects mirrors that of Judge Rao in her Mazars dissent in the D.C. Circuit. Indeed, Judge Rao makes a point of identifying her position with that of Judge Cochran. See Trump v. Mazars USA LLP, No. 19-5142, slip op. at 49-50 n. 16 (D.C. Cir. Oct. 11, 2019). She claims that the Supreme Court did not disagree with the district judge on legal principle, but “simply disagreed with the district court’s characterization of the proceedings, which were not about the wrongdoing of the Attorney General but the administration of the Department of Justice as a whole.” Id. This betrays a lack of familiarity with the McGrain case since Mal Daugherty had no connection to the Department of Justice other than his knowledge of his brother’s wrongdoing.

In any event, Judge Cochran’s decision was music to the ears of Harry Daugherty’s defenders and critics of the congressional investigations. One can easily imagine that the Coolidge administration was tempted to endorse the decision (which would have undermined future congressional oversight) or at least to decline to get involved on the Senate’s side. Instead, however, Harlan F. Stone, Daugherty’s successor as attorney general, undertook to represent the Senate on appeal to the Supreme Court, thereby putting both political branches squarely on the side of congressional investigatory authority. Conveniently, though, briefing and oral argument did not take place until after the presidential election of 1924. (Stone’s opening brief was filed six days after the election).

Meanwhile, Frankfurter’s camp was preparing legal scholarship to support the Senate. In December 1924, as the McGrain case was being argued, the Harvard Law Review published a student note critical of Judge Cochran’s decision. See Note, The Power of Congress to Subpoena Witnesses for Non-Judicial Investigations, 38 Harv. L. Rev. 234 (Dec. 1924). Among other things, the note took issue with Cochran’s conclusion that the impeachment power implicitly limited the Senate’s power to conduct legislative investigations of executive wrongdoing. See id. at 238 (“Impeachment is a ponderous method of rectifying gross misconduct and consequently has been seldom employed.  By limiting the exercise of this extraordinary remedy, the Constitution could not have intended to restrict more common powers of investigation shown by experience to be necessary to the practical exercise of a federal power.”).

Although the note is unsigned, there is little doubt it reflects Frankfurter’s influence. The articles editor was Thomas G. Corcoran, a Frankfurter protege who would go on to clerk for Justice Oliver Wendell Holmes at Frankfurter’s recommendation during the 1926-27 term. (Another fun fact: Corcoran in later life became a lobbyist who notoriously once “lobbied” the Supreme Court on behalf of a client. See Bob Woodward & Scott Armstrong, The Brethren 79-86 (1979)).

A more significant piece of scholarship came from Professor Landis, Frankfurter’s Harvard colleague and frequent co-author. See James M. Landis, Constitutional Limitations on the Congressional Power of Investigations, 40 Harv. L. Rev. 153 (Dec. 1926). Landis argued that the meaning of the legislative power conveyed by the Constitution could only be understood in light of historical experience; he then marshaled British and colonial history to demonstrate that “[a] legislative committee of inquiry vested with power to summon witnesses and compel the production of records and papers is an institution rivaling most legislative institutions in the antiquity of its origin.” Id. at 159. When combined with the unbroken practice of legislative investigations since the adoption of the Constitution, he concluded that “[t]he Daugherty inquiry of 1924 is thus a direct descendant of a more ancient lineage, ancient enough, when constitutional history begins for the United States in 1789, to demand recognition as a convention entitled to constitutional standing.” Id. at 193-94.

Many years later, during the conference in Watkins v. United States, 354 U.S. 178 (1957), then Justice Frankfurter remarked that “Landis’s article on investigations turned the trick in the Daugherty case in this Court and led it to uphold the powers of Congress.” The Supreme Court in Conference (1940-1985) 299 (Del Dickinson, ed. 2001). Whether this is exactly true or not (see below), Landis’s article seems to have had a powerful effect on legal thinking about the subject of congressional investigations by “completely demolish[ing]” the historical and logical foundations of Kilbourn‘s cramped reading of the legislative power of inquiry. Boudin, 35 Va. L. Rev. at 147; see also id. at 165-66.

Several factors thus converged to support the Senate’s position before the Supreme Court in McGrain. Politically, there was little motivation for anyone to defend the conduct of the Harding administration, particularly after President Coolidge won reelection in 1924. The fact that both the executive and legislative branches agreed on a common legal position likely weighed heavily in the Senate’s favor. The intellectual firepower of Harvard law school surely did not hurt either.

Nonetheless, it appears that the outcome in McGrain was, like Waterloo, a damn close run thing. Although it was argued in December 1924, it was not decided until January 1927. (Another strike against Professor Jonathan Turley’s theory that the courts will resolve such issues quickly). This in itself suggests more internal dissension than betrayed by the ultimate unanimous decision (Harlan Stone, who was appointed to the Court during the intervening period, did not for obvious reasons participate). Cf. McGrain, 273 U.S. at 154 (“We have given the case earnest and prolonged consideration because the principal questions involved are of unusual importance and delicacy.”).

According to this March 1927 letter to Frankfurter from John Gorham Palfrey, a longtime aide to Justice Holmes, in an earlier vote on the case Justices Holmes and Brandeis were “standing out against the whole bunch,” apparently meaning that the other justices would have affirmed the district court. Although Palfrey indicated that Holmes had read “Jim’s article” and that Brandeis had distributed it to other justices including Justice Van Devanter, who was assigned the opinion, he did not believe that was the real reason for the majority switch. Instead, “Van Devanter, who has been away behind on his opinions, go around to writing the opinion for the majority a couple months ago– and found he couldn’t do it to reach the majority result.”

Whatever the true reason, Van Devanter ultimately produced a strong and unanimous opinion in support of a broad congressional investigatory authority, one that has driven a largely deferential judicial attitude toward congressional investigations ever since.

Until now. We will turn to that in our next post.

Roger Cramton’s Memorandum Surfaces

Remember the Roger Cramton memorandum we discussed a few months ago? (Of course you do, scarcely a waking moment goes by when you don’t think “I wonder what ever happened with that Roger Cramton memorandum?”). This was one of the memoranda cited by the Office of Legal Counsel in footnote 1 of its opinion declaring that former White House counsel Don McGahn was absolutely immune from having to appear in response to a congressional subpoena.

As we have discussed, OLC’s argument for absolute immunity is based in large part on “precedent” consisting of its own prior statements on the subject. But, as two federal judges have now pointed out, OLC cannot create precedent simply on its own say-so. Last month Judge Ketanji Brown Jackson wrote, in her scathing rejection of McGahn’s claim of immunity, that OLC’s initial theory of absolute immunity set forth in the 1971 Rehnquist memorandum “was seemingly formed out of nothing” and “it appears that an endorsement of the principles that OLC espouses would amount to adopting the absolute testimonial immunity for senior-level presidential aides by ipse dixit.” Comm. on the Judiciary, U.S. House of Representatives v. McGahn, No. 19-cv-2379, slip op. at 99, 102 (D.D.C. Nov. 25, 2019); see also Comm. on the Judiciary, U.S. House of Representatives v. Miers, 558 F. Supp.2d 53, 86 (D.D.C. 2008) (rejecting OLC’s opinions on absolute immunity as “conclusory and recursive”). Furthermore, as both Judge Jackson and Judge Bates noted, the original justification for immunity set forth in the Rehnquist memorandum would not apply to former White House officials at all. See McGahn, slip op. at 100; Miers, 558 F. Supp.2d at 88 n. 36.

Enter the aforementioned Cramton memorandum of December 21, 1972 to “the Honorable John W. Dean, III,” Counsel to the President. Although OLC cited this memorandum in its opinion on McGahn, it did not make it public at the time, nor did it bother to mention that this memorandum differed in an important respect from the argument that it was making. We know this now because OLC has just posted it on its website. Hat tip: @kpolantz and @EricColumbus.

To wit, the Cramton memorandum concludes that former White House officials should not be entitled to the same absolute immunity as current officials. It states:

We have one caveat with respect to our conclusion. While we believe that an assertion of Executive privilege with respect to specific testimony on the subject of advice given by the former staff member to the President is entirely proper, we have some reservations about the propriety of invoking the privilege to direct the former staff member not to appear at all. This aspect of the Executive privilege has in the past been claimed only for the President and his most intimate, immediate advisers. One of the justifications that has been advanced for an immediate adviser declining to appear is that he is presumptively available to the President 24 hours a day; the necessity to appear before congressional committees therefore could impair that availability. This consideration would obviously not justify a refusal to appear by a former staff member. However, this justification is in our view neither the only nor the best one. An immediate assistant to the President may be said to serve as his alter ego in implementing Presidential policies. On this theory, the same considerations that were persuasive to former President Truman would apply to justify a refusal to appear by such a former staff member, if the scope of his testimony is to be limited to his activities while serving in that capacity.

In conclusion, we believe that an invocation of the privilege with respect to particular testimony by a former staff member on the subject of advice given the President is quite clearly proper; on the other hand, we believe an invocation of the privilege as a basis for refusal to appear at all is a closer question. An intention to invoke the privilege with respect to particular testimony could certainly be announced. This as a practical matter may solve the problem. If, however, the interrogation is expected to extend to non-privileged matters, a decision that the former staff member should not appear at all would not, in our opinion, be justified.

Memorandum of 12-21-1972 at 6-7 (emphasis added).

To be sure, this language does not foreclose a refusal to appear by a former White House official if the testimony is expected to involve only privileged matters (though it suggests this is a “closer question”). If, on the other hand, non-privileged matters are involved, it indicates that such a refusal would not be justified. This position is inconsistent with OLC’s current stance, which is that former officials are absolutely immune from any questioning about their official activities, regardless whether they are privileged. As OLC “explained” in its McGahn opinion, “the concept of immunity is distinct from, and broader than, the question whether executive privilege would protect a witness’s response to any particular question.” 5-20-19 Opinion at 17. Moreover, it asserted that “consistent with our prior precedents, we find no material distinction between the compelled congressional testimony of current and former senior advisers to the President.” Id. at 16. This again is inconsistent with the Cramton memorandum.

Furthermore, the Cramton memorandum implicitly rejects OLC rationales for extending immunity to former officials. If allowing such officials to testify about non-privileged matters will not impair the president’s ability to obtain confidential advice, there is no reason why they should not appear and invoke the privilege on a question by question basis (like every other executive official outside the White House). Moreover, Cramton obviously did not believe that allowing former officials to appear would adversely impact the president’s “autonomy.”

It seems to me that if you are going to rest an argument on ipse dixit, you ought at least to be honest about the ipse.  Maybe the D.C. Circuit will have some questions about this too.

Judge Leon’s Ruling in the Kupperman Case Could be Important Even if it Does not Reach the Merits

The lawsuit brought by former deputy national security advisor Charles Kupperman continues, for the moment, despite the House’s withdrawal of its subpoena. Most likely, Judge Leon will end up dismissing the case as nonjusticiable on one ground or another. However, it could matter a good deal which ground(s) the court relies upon.

If the case is dismissed as moot due to the withdrawal of the subpoena, it would be of little consequence. On the other hand, if the court were to base its dismissal on the president’s lack of authority to direct Kupperman not to appear in response to the subpoena, its ruling is potentially of much greater significance. As Jonathan Shaub has noted in connection with the House’s lawsuit against former White House counsel Don McGahn, a judicial ruling that the president lacks authority to direct former officials how to respond to congressional subpoenas might be more important than a ruling on the merits of the absolute immunity issue. While the latter would affect only the relatively small group of senior White House advisors who allegedly are protected by absolute immunity, the former “could be far-reaching, encompassing all disputes involving former officials whether they are grounded in immunity or executive privilege.”

Kupperman’s complaint alleges that he “has a duty to abide by a lawful constitutional assertion of immunity by the President and a lawful instruction by the President that he decline to testify before Congress concerning his official duties as a close advisor to the President.” Complaint ¶ 41. Note that this arguably constitutes two distinct assertions. At one level, it is an assertion that if the claimed immunity exists, it belongs to the president, not to the subordinate official, and therefore Kupperman cannot or should not waive it contrary to the president’s instruction. This makes sense to me. Since the immunity (if it exists) is designed to protect the presidency, it should be the president’s decision whether to assert or waive it.

Of course, as Eric Columbus has pointed out, former officials not infrequently choose to disclose confidential information regarding their government service in medial interviews or tell-all books. Indeed, former national security advisor John Bolton, who is currently declining to testify before Congress based on the president’s assertion of “absolute immunity,” has a book deal in which he will presumably discuss many of the matters allegedly covered by that immunity. (As one Twitter wag put it, absolute immunity is a monarchical doctrine so naturally it has a “royalty exception.” Ok, that wag was me.). While there is a tension between this fact and the non-waiver principle, in my view it simply illustrates that the executive branch has no means of punishing former officials who violate a duty not to disclose non-classified information (about which more below).

Kupperman also appears to be making a second and stronger assertion. He seems to be claiming that a former official has a duty to obey the president’s instruction, regardless of whether the former official agrees with the president’s legal position. As Shaub points out, though, it is not clear where the president gets the authority to direct a private citizen’s response to a congressional subpoena. OLC’s past pronouncements suggest it believes the president has this authority, but it fails to “offer any constitutional analysis to support that conclusion.” (Shaub, this might be a good place to note, is a former OLC lawyer).

If Judge Leon were to conclude the president lacks authority to direct Kupperman’s response to the subpoena, he could dismiss the case without reaching the merits. Kupperman claims to be facing “irreconcilable commands” from the executive and legislative branches, but if he is not bound to obey the president’s command, the alleged conflict disappears and can provide no basis for him to sue. He then would be in a posture no different than any other congressional witness who asserts a potentially valid privilege. He can choose to assert absolute immunity if he wishes and, when the committee (properly) rejects that assertion, he can decide whether to comply or risk the possibility of a contempt proceeding. There is no reason why he, any more than any other congressional witness in this situation, should be entitled to an advance court ruling to forestall contempt.

A somewhat narrower approach the court might take is to side step the question of legal duty entirely. Instead, the court might ask what injury Kupperman would suffer should he choose to ignore the president’s directive not to testify. Kupperman alleges that “an erroneous judgment to appear and testify in obedience to the House Defendants’ subpoena would unlawfully impair the President in the exercise of his core national security responsibilities,” Complaint ¶ 2, but it is hard to see how this constitutes an injury to Kupperman. As suggested earlier, there do not appear to be any practical repercussions to a former official who reveals confidential but non-classified information, whether before Congress or in a tell-all book. In the absence of any adverse consequence Kupperman will suffer as a result of disregarding the president’s order, it would seem he lacks standing to sue regardless of whether the president has the authority to issue the order.

Even if Kupperman has a legal duty to assert absolute immunity when instructed to do so by the president, it does not follow that he is obligated to go into contempt to protect the president’s privilege. For example, a lawyer who is subpoenaed by a congressional committee to provide privileged information of a current or former client is obligated to assert the privilege if her client so instructs, but she is not obligated to go into contempt in order to fulfill her professional obligations. See D.C. Bar Ethics Opinion 288 (Feb. 1999). There is no reason why a former government official should be required to do more when instructed by the president; after all, the president has ample other tools, including filing his own lawsuit, to protect whatever confidentiality interests are at issue.

In short, a non-merits dismissal of Kupperman v. House could still have a significant (and beneficial) effect on the House’s ability to get information in the current impeachment inquiry and/or in future information disputes between the political branches.

Kilbourn and Chapman and Rao. Oh my.

As it happens, the decision in Trump v. Mazars USA (D.C. Cir. Oct. 11, 2019) coincides nicely with our discussion of the congressional contempt power. Although Mazars involved the validity of a congressional subpoena, not the exercise of the contempt power per se, the cases we are about to encounter are quite relevant to an analysis of the D.C. Circuit’s opinion, particularly with regard to Judge Rao’s remarkable dissent.

Kilbourn v. Thompson

Following its decision in Anderson v. Dunn, 19 U.S. 204 (1821), the Supreme Court next had occasion to consider the contempt power in Kilbourn v. Thompson, 103 U.S. 168 (1880). Kilbourn arose out of a House special committee’s investigation into the bankruptcy of Jay Cooke & Co., a private firm with large investments in a somewhat shady “real estate pool”; the House resolution  establishing the committee recited that the government of the United States was a creditor of the bankrupt firm as the result of “improvident deposits by the Secretary of the Navy” of public moneys at the firm. 103 U.S. at 171. The resolution further recited that the bankruptcy trustee  “has recently made a settlement of the interest of the estate . . . to the disadvantage and loss, as it is alleged, of the numerous creditors of said estate, including the government of the United States, and . . . the courts are now powerless by reason of said settlement to afford adequate redress to said creditors.” Id.

Hallet Kilbourn, a real estate broker with knowledge of the private investments in question, was subpoenaed by the House to provide testimony and documents regarding the matter. He declined to do so, denying “the right of the House to investigate private business arbitrarily,” but stated that “if either the committee or the House would assert that the production of his private papers, or the revelation of his private business, would promote any public interest, or if any private individual would assert on oath that the papers asked for would lead to the detection of corruption, he would respond freely to all demands for information or papers.” 2 Hinds’ Precedents § 1609.

The House then ordered that the Speaker issue an arrest warrant for Kilbourn, pursuant to which the recalcitrant witness was brought before the bar of the House. When he continued to refuse to answer, the House held him in contempt and ordered the Sergeant-at-Arms to keep him in custody until such time as he was willing to provide the information demanded. 103 U.S. at 175; 2 Hinds’ Precedents § 1609.

While Kilbourn was in custody, he was indicted by a federal grand jury under the criminal contempt of Congress statute. This precipitated a conflict between the legislative branch and the executive/judicial branches when the U.S. marshal, with a warrant from the D.C. court, attempted to take custody of Kilbourn from the Sergeant-at-Arms. 2 Hinds’ Precedents § 1609. The Sergeant-at-Arms refused, and the House actually considered a Blackstonian resolution that would have asserted that the House, not the courts, had the ultimate right to determine the disposition of the prisoner. Id. The House rejected this resolution, however, and authorized the Sergeant-at-Arms to obey the court’s writ of habeas corpus. Id. The court eventually determined that Kilbourn should be released by the Sergeant-at-Arms and taken into custody by the U.S. marshal. Id.

Kilbourn subsequently sued the House for false imprisonment. Perhaps due to the prior tension with the executive branch, the House was represented by private counsel in the case. See Representation of Congress and Congressional Interests in Court: Hearings Before the Subcomm. on Separation of Powers of the Senate Comm. on the Judiciary, 94th Cong., 2d sess. 511-12 (1974-75). For whatever reason, the House’s position was much less warmly received when it reached the Supreme Court than it had been in Anderson.

While the Anderson Court embraced the key arguments of the “pro-contempt” side of congressional debates (particularly the argument that the contempt power was an absolute necessity to protect the functioning of Congress), the Kilbourn Court adopts many of the principal arguments of congressional opponents of contempt. It begins with the observation that Congress’s powers are “dependent solely on the Constitution,” “either expressly or by fair implication.” 103 U.S. at 182. As no express power to punish contempts is granted, “advocates of this power have, therefore, resorted to an implication of its existence founded on two principal arguments . . . (1) its exercise by the House of Commons of England, from which country we, it is said, have derived our system of parliamentary law, and (2) the necessity of such a power to enable the two Houses of Congress to perform the duties and exercise the power the Constitution has conferred on them.” Id. at 182-83. Continue reading “Kilbourn and Chapman and Rao. Oh my.”

Must Congress Allow Agency Counsel to Attend Depositions?

Last week I wrote a piece for Just Security regarding the State Department’s refusal to produce certain officials for depositions requested by several House  committees (relating, of course, to the Ukraine affair). Among other things, I addressed the question of whether Congress is required to allow government counsel to attend depositions of the current or former officials of their department or agency. OLC (surprise!) says the answer is yes. In my view, the correct answer is no.

Anyway, I thought it would be helpful to collect the materials in one place. Here is the link to the piece. And here are the relevant House rules and regulations.

What to Look For in the Mazars Oral Argument

Tomorrow the Mazars subpoena case will be argued before the D.C. Circuit (Judges Tatel, Millett and Rao). In this appeal, President Trump (in his private capacity) has advanced a novel and audacious theory in support of his contention that the congressional subpoena at issue lacks a “legitimate legislative purpose.” Trump is asking the D.C. Circuit to reverse the decision of the district court, which denied Trump’s application for an injunction to prevent Mazars, his accounting firm, from producing financial records responsive to the subpoena.

As Judge Mehta pointed out in his decision below, the legal standards employed by the courts to review congressional subpoenas and investigations are generally so deferential that they “do not substantially constrain Congress.” Rather than fighting this proposition head on, Trump’s lawyers focus primarily on seeking a carve-out from Congress’s broad investigatory and oversight authority for those who hold constitutional offices (i.e., the president and Supreme Court justices).

The lynchpin of this argument is the distinction between Congress’s legislative and judicial powers. Although it has long been understood that Congress must have some ability to obtain information needed to consider and craft legislative measures, the scope of this power was once highly controversial. Following the Supreme Court’s decision in Kilbourn v. Thompson, 103 U.S. 168 (1880), which stressed that Congress, unlike the British Parliament, was not a court of general jurisdiction, it was widely believed that Congress’s authority to compel the production of information for legislative purposes was limited at best. This reading of Kilbourn, however, was undermined by later Supreme Court cases beginning with McGrain v. Daugherty, 273 U.S. 135 (1927). These subsequent cases, Judge Mehta noted, render Kilbourn “largely impotent as a guiding constitutional principle.”

Trump attempts to revive Kilbourn for a limited purpose. His theory goes something like this. Congress generally has broad legislative and oversight authority with respect to the departments and agencies of government, including offices created by statute. With respect to constitutional offices, however, Congress’s legislative authority is “severely constrained.” Trump Br. at 4, 38. Thus, Trump claims that Congress cannot rely on its legislative authority to seek his financial records because these records are not relevant to a subject on which legislation may be had.

To be sure, Trump acknowledges that Congress also has powers of a judicial nature. But, following Kilbourn, he stresses that these are not of general scope, but are limited to those areas expressly identified in the Constitution, namely judging elections, disciplining members, and, most importantly, impeachment. For those who hold constitutional offices, therefore, impeachment is the only method (Trump argues) by which Congress can investigate alleged wrongdoing.

Thus, “[w]hile Congress could presumably use subpoenas to advance these non-legislative powers, the Committee has not invoked them.” Trump Br. at 45. In fact, Trump takes pains to demonstrate that Speaker Pelosi has disavowed any intent to go down the path of impeachment. Trump Br. at 47-48.

The novelty of this argument is illustrated by the fact that one of the principal “authorities” cited in Trump’s brief is a three-page law student note that is more than a century old. See Note, Congressional Power to Punish for Contempt, 30 Harv. L. Rev. 384 (1917) (cited in Trump Br. at 17-18). The apparent utility of this note from Trump’s perspective is that it vividly distinguishes congressional contempt in the context of impeachment from that in the context of legislative oversight:

A committee of the House considering an impeachment is like a Grand Jury hearing evidence which may lead to the return of an indictment– it is a judicial body and it is one provided for by the Constitution. It must surely have the power to call witnesses, and the power is of little avail if these witnesses may contemptuously refuse to respond, or may be influenced or intimidated by outside contempts of the body before whom they are testifying. If the House is to sit in a judicial capacity, it must have the protection that a court has.

Note, 30 Harv. L. Rev. at 385. Yet even this note, which predates McGrain, concludes that “[t]he power to legislate . . . by necessary implication include[s] the power to examine witnesses and to compel them to respond by contempt proceedings.” Id. at 386.

Why would the Trump team rely on what seems like a very long shot argument? The short answer is I don’t know, but three possibilities come to mind. First, Trump’s lawyers may simply believe this is the best argument available to them. Second, it may be thought that stressing the lack of impeachment proceedings helps Trump’s cause beyond the confines of this particular case (e.g., with public opinion and/or with courts that will be ruling on more difficult subpoena enforcement issues in the months ahead). Third, Trump’s legal team may believe that advancing a bold legal theory is the best way to get the Supreme Court to grant certiorari, which strings out the proceedings even if it is unlikely to deliver an ultimate victory.

The two things to look for in tomorrow’s argument: (1) how much interest does the panel show in impeachment and why the House has declined to initiate impeachment proceedings to date, and (2) whether Judge Rao (the lone Republican appointee) seems at all open to Trump’s arguments. A unanimous panel opinion will make further review less likely.