The JOR Memo, the Myers Case, and the Theory of Executive Privilege

To assess the executive privilege argument of the JOR memo, we should begin by drawing three distinctions which are either blurred or ignored in that memo (and the subsequent Wolkinson memo). These are (1) constitutional immunity versus (common law or constitutional) privilege; (2) the protections available to the chief executive versus those available to lower level officials; and (3) subpoenas to appear and provide testimony versus subpoenas to produce documents.

These distinctions are suggested by a passage from a prominent legal treatise that Roberts quotes:

Subpoena to Executive. It is well settled that public officials are not bound to disclose state secrets or to submit public papers to judicial scrutiny. Partly on this ground, and partly because of the immunity of the executive from judicial control on account of the tripartite separation of powers, it seems now to be undisputed that courts cannot compel the attendance of the chief executive as a witness.

JOR memo at 4-5 (quoting 12 Corpus Juris 896 (W. Mack, ed. 1917)) (emphasis added).

This passage distinguishes between the right of public officials to protect state secrets, which is a common law privilege, and the constitutionally based theory of “immunity of the executive from judicial control on account of the tripartite separation of powers.” There were several governmental privileges widely understood to be available under common law, but these protections were limited in scope and generally qualified rather than absolute even when they applied. See Jonathan David Shaub, Common Law Executive Privilege(s) (forthcoming 2025).

The passage also addresses judicial subpoenas to the “chief executive” (i.e., a governor or president) and opines (with some exaggeration) that it is “undisputed” such subpoenas cannot be enforced. The rationale given is based partly on common law privileges and partly on state and federal separation of powers principles that prohibit compulsion of the chief executive. But the fact that the chief executive is (or may be) beyond the compulsory power of the court does not imply the same for lower-level officers.

Finally, the passage baldly asserts that public officials are not required to submit “public papers to judicial scrutiny.” Again this seems to be a significant exaggeration, but there is language in some of the cases and secondary sources cited by the treatise which suggests a broad discretion on the part of chief executives and (sometimes) other high-level public officials to withhold certain types of public papers in judicial proceedings.

The authorities speaking to these questions are a mix of English and state cases, along with some 19th century evidentiary treatises. No federal authority had addressed the issue as of the time of the JOR memo. Moreover, none of the authorities spoke to congressional proceedings at all. According to Roberts, however, Myers somehow supports the proposition that while executive officers “are subject to subpoena to testify before the United States House of Representatives or a committee thereof in connection with hearings for legislative purposes, the question as to whether the public interest requires them to refuse to answer a question or refuse to produce records in their custody or control is for their determination as a public officer.” JOR memo at 4.

As noted in my last post, why Roberts thinks Myers supports this result is somewhat obscure. There seem to be two possible arguments, which are discussed below.

Separation of Powers

 One possible argument is suggested by a brief that would be submitted by President Nixon to the Supreme Court some three decades after the JOR memo. Nixon, like Roberts, did not draw any direct analogy between the power of removal involved in Myers and the power to withhold information in response to a subpoena (which is not surprising since none exists). Nixon instead relied on Chief Justice Taft’s strict approach to the separation of powers to support the following proposition:

The doctrine of the separation of powers embodies the concept that each branch is independent of the others, except where some form of interaction flows from the regular operation of the government or where the Constitution or statutes explicitly provides to the contrary. The doctrine necessarily includes the right of the holder of the privilege to decide when it is to be exercised. It means, in this case, that compulsory process cannot issue against the President.

Br. for the Respondent, Cross-Petitioner Richard Nixon, President of the United States 81, United States v. Nixon, 418 U.S. 683 (1974), reproduced in Special Report of the J. Committee on Congressional Operations 177, 276 (1974); see also id. (quoting Myers that “the reasonable construction of the Constitution must be that the branches should be kept separate in all cases in which they were not expressly blended”).

The problem with this argument is that there is no necessary reason why even a strict approach to the separation of powers should imply that the right of one branch to obtain information (as recognized by the Court in McGrain) should give way to the asserted right of the other to withhold it. Indeed, this would seem to be a prime example of a case “where some form of interaction flows from the regular operation of the government,” to use Nixon’s words. In any event, the Supreme Court had little difficulty brushing off Nixon’s argument, noting that allowing the president “an absolute, unqualified privilege” in the face of a criminal subpoena would “plainly conflict with the function of the courts under Art. III.” Nixon, 418 U.S. at 707. It did not find the language of Myers relevant enough even to cite the case.

Presidential Authority over Subordinates

 A second argument, which seems to be the one that Roberts was trying to make, is that Myers recognized a broad presidential right of control over executive officers, one with which the other branches may not interfere, and that this right of control implies that the president may direct his subordinates to refuse to comply with compulsory process. See JOR memo at 4 (Myers “definitely settled” that “the executive branch of the Government is independent and that officers of that branch are appointed by the President . . . [and] that once appointed they exercise the functions of their office subject [only] to the control and direction of the President, and subject to removal by the President [alone].”); id. (“If the President of the United States, exercising his independent discretion as the Chief Executive, decides against [an executive] officer testifying or producing such records [demanded by a congressional subpoena], and orders him not to do so, then it is my opinion that the Congress of the United States has no power to compel such officer to do so.”).

Wolkinson makes essentially the same argument, focusing on language in Myers explaining that in many cases, such as dealings with foreign governments, executive officers are charged with exercising discretion that is constitutionally assigned to the president and that “[i]n this field his cabinet officers must do his will.” Myers, 272 U.S. at 134 (quoted in Wolkinson memo at 76a). In language quoted by Wolkinson, Chief Justice Taft states that “[i]n all such cases, the discretion to be exercised is that of the President in determining the national public interest and in directing the action to be taken by his executive subordinates to protect it.” Myers, 272 U.S. at 134 (quoted in Wolkinson memo at 76a). Wolkinson then twice paraphrases this language to conclude that in responding to compulsory process “the discretion is that of the President to determine the national public interest, in directing his executive subordinates to withhold confidential information from congressional committees.” Wolkinson memo at 76b, 100.

Roberts and Wolkinson have engaged in a basic failure of logic. Myers may be relevant to the question whether the president can properly direct subordinates as to how to exercise discretion in responding to compulsory process. But this says nothing about whether such discretion exists in the first place. If the officer lacks the discretion to refuse to comply with compulsory process, the president’s right to direct or control him is irrelevant. The president’s order does not give his subordinate the right to take an unlawful action. Thus, for example, the president cannot order the officer to commit perjury or destroy documents in response to compulsory process from the legislative or judicial branches.

Roberts and Wolkinson would presumably try to overcome this problem by citing historical and judicial precedent in support of the president’s immunity from compulsory process. Thus, Roberts asserts that it can be said “without fear of contradiction that the President of the United States, while subject to subpoena, will not be forced through the issue of judicial process to respond to same.” JOR memo at 8. But even if that were true, it does not follow that the president’s immunity would transfer to his subordinate. As we saw in my last post, Taft explained in his book that the “[t]he Supreme Court seems to make a broad distinction between issuing process against the President and against his subordinates under laws requiring the specific performance of a definite act.” Our Chief Magistrate 132. Nothing in Myers suggests otherwise.

Furthermore, when executive officers are responding to compulsory process they are acting in a capacity that is different than when they are carrying out statutory duties or other official functions. An individual who is subpoenaed as a witness in a proceeding, whether congressional or judicial, is subject to the “fundamental maxim that the public . . . has a right to every man’s evidence” and “the general duty to give what testimony one is capable of giving.” United States v. Bryan, 339 U.S. 323, 331 (1950) (quoting Wigmore, Evidence §2192 (3d ed. 1940)); see also Trump v. Mazars, 591 U.S. ___ (2020) (“When Congress seeks information ‘needed for intelligent legislative action,’ it ‘unquestionably’ remains ‘the duty of all citizens to cooperate.’”) (quoting Watkins v. United States, 354 U.S. 178, 187 (1957) (emphasis added by Mazars). Public officials, including executive officers, are subject to this duty as individuals and citizens and are not simply exercising the functions of their office. Thus, regardless of what authority the president has to direct executive officers in the exercise of their official duties, it does not extend to directing such officers not to perform a civic duty.

This would seem most apparent in cases where executive officers are called to testify in matters that have little or no connection to their offices, as for example where they are witnesses to ordinary crime or are involved in private litigation. It is not an infringement of the independence of the executive branch to demand that an executive official appear to testify as a witness in a murder trial or as a party to a divorce proceeding, even if it might be inconvenient for him to spare the time away from work. Nor does it seem likely that even Roberts or Wolkinson would claim that the witnesses in such cases have a general right to refuse to answer questions based on their assessment of the “public interest.”

The logical implication of the Roberts/Wolkinson position, however, is that the executive officer would have a right to refuse to answer questions, even in an otherwise unrelated proceeding, if they sought to probe into aspects of his official functions. For example, if in a divorce case the officer’s spouse wants to know how much he earns from his job, he could deem it not in the “public interest” to divulge this information. If the court is dissatisfied with this response, its only recourse (if we are taking the Roberts/Wolkinson theory seriously) would be to appeal to the president personally. Perhaps needless to say, neither Roberts nor Wolkinson present an example of such a theory being presented to, much less being accepted by, any court.

Consider another hypothetical in the congressional context. As Roberts stresses, the witness at issue in the McGrain case was a private citizen, but he neglects to mention why the witness was being subpoenaed by the Senate. Mal Daugherty was an Ohio banker and the brother of Attorney General Harry Daugherty, and the Senate wanted to ask him about large amounts of cash which had been carried by Harry’s assistant to Ohio and deposited in accounts controlled by the Daugherty brothers. Suppose the witness had been Harry himself? Could he refuse to testify about the cash deposits on grounds that it would be contrary to the public interest? What if the questions started to get closer to the exercise of his “official functions” (like if it turned out the cash was received from selling protection from federal law enforcement to bootleggers)? Where is the line to be drawn and who gets to draw it? And if Harry’s testimony were treated differently because he happens to be an executive officer, what happens after he leaves office? Surely a former executive officer cannot claim the right to decide what is in the public interest.

It seems far more likely, and more sensible, that such issues be addressed like they would for any witness in a congressional or judicial proceeding. If the witness believes he has a privilege not to answer a question, he may assert it, the committee or court will rule on it, and then the witness has to decide whether to obey an adverse ruling or seek to challenge it further (which will often risk being held in contempt).

Executive Branch Records

The situation is arguably different, however, when what is at issue is the production of executive branch records, rather than oral testimony.  What records should be deemed to be in the “custody or control” of a particular witness is less clear-cut than what information is in the witness’s head. The mere fact that a witness has possession of or access to records is not enough to establish the authority to produce them in response to a subpoena. (For example, one could not subpoena the janitors to produce documents from the government or corporate offices they clean). Moreover, courts have generally recognized that decisions on producing government records should be made by senior officials and/or in accordance with rules established by the department or agency in question. In short, the production of executive branch records raises questions of official authority that are not present when a witness is simply responding to questions.

It is this distinction that lies at the heart of the executive branch’s traditional approach to responding to congressional (and, to a lesser extent, judicial) demands for papers. In response to the House select committee’s demand that the secretary of war produce records regarding the St. Clair expedition, President Washington’s cabinet asserted that “neither the committee nor House had a right to call on the Head of a Department, who and whose papers were under the President alone; but that the committee should instruct their chairman to move the House to address the President.” The ensuing century and a half of legislative-executive disputes over papers (not testimony) reflected the fact that Congress lacked any effective means to challenge the president’s claim of ultimate custody over executive branch papers.

The claim of superior custody over documents is not the same as a claimed right to control every aspect of a subordinate’s response to a congressional or judicial demand for information. This distinction is illustrated by President Cleveland’s 1886 message responding to the Senate’s demand to the attorney general for documents relating to Cleveland’s decision to remove certain U.S. attorneys from office. Cleveland directed the attorney general not to comply on the ground that the papers related to a matter (removal of executive officers) committed by the Constitution to his sole discretion and for which he was not accountable to the Senate. He explained:

Against the transmission of such papers and documents I have interposed my advice and direction. This has not been done, as is suggested in the committee’s report, upon the assumption on my part that the Attorney-General or any other head of a Department “is the servant of the President, and is to give or withhold copies of documents in his office according to the will of the Executive and not otherwise,” but because I regard the papers and documents withheld and addressed to me or intended for my use and action purely unofficial and private, not infrequently confidential, and having reference to the performance of a duty exclusively mine. I consider them in no proper sense as upon the files of the Department, but as deposited there for my convenience, remaining still completely under my control. I suppose if I desired to take them into my custody I might do so with entire propriety, and if I saw fit to destroy them no one could complain.

 8 Messages and Papers of the Presidents 375 (emphasis added). Thus, Cleveland explicitly disclaims an unlimited right to control the attorney general’s responses to congressional inquiries even as to the production of documents, much less with regard to testimony; instead, he maintains that the papers at issue are in his constructive possession even if they happen to be located at the Department of Justice.

Ironically, Roberts himself acknowledges the importance of presidential custody near the close of his memo:

[L]et me point out that if the governmental records are in the possession of the President of the United States, while it may be that the Congress has the power to issue a subpoena, as was said by Mr. Justice Marshall in the Burr case, there is no power in the Congress to force the President to act. It would seem that the only power Congress would have in such a situation would be either to refuse to appropriate moneys for the executive branch of the Government, thus bring (sic) the President to reason, or to commence impeachment proceedings.

JOR memo at 10. What Roberts overlooks or ignores is that all the supposed precedents he cites (as well as virtually all those discussed by Wolkinson in his longer and more detailed memo) can be explained by the combination of presidential immunity from compelled process and the president’s actual or constructive possession of executive branch records. This leaves little or no precedent, even in executive branch practice, to support the much broader Roberts/Wolkinson theory.

All of which would suggest that the JOR memo is as superficial and one-sided in its analysis as the subsequent Wolkinson memo. But in a surprise ending, the JOR memo does not actually apply its broad theory to conclude that the Federal Communications Commission and its officers are immune from congressional subpoenas. In my final (for the moment, anyway) post regarding the JOR memo, I will endeavor to explain that plot twist.

Leave a Reply

Your email address will not be published. Required fields are marked *