Executive Power

The Prophylactic Executive Privilege

By Jonathan Shaub
Friday, June 14, 2019, 5:18 PM

Two seminal events have occurred in recent days in the ongoing oversight war between the House of Representatives and the Trump administration—and in the ongoing expansion of the doctrine of executive privilege. Although each incident warrants further individual analysis, together they suggest the “constitutionalization” of what I will call a “prophylactic executive privilege,” a view that the executive branch has absolute constitutional authority to protect and further the president’s qualified constitutional authority to assert executive privilege. And that absolute prophylactic authority is undiminished by—indeed, is empowered by—Congress’s exercise of its constitutional authority and takes no account of Congress’s countervailing interests. Although the departures from past precedent may seem subtle, they in fact represent an extreme conception of the president’s constitutional authority, one that renders congressional authority immaterial.

Oversight is a delicate area in which to take a legal position, particularly an extreme one. Any asserted authority or legal position may be asserted against one’s own party after the next election, and hypocrisy is rampant. And because so few disputes make it to court and most have been eventually resolved through compromise, legal scholars have not focused as much on the constitutional dispute between the branches in the context of oversight as they have in other disputed areas, such as war powers.

But over the past 40 years, the executive branch has been steadily developing a comprehensive, and largely consistent, doctrine of presidential control over information—one that has never been tested by an appellate court but that has influenced everything from signing statements, to the authority of inspectors general, to the payment of private attorneys for executive branch officials and, most importantly, to congressional oversight.

In a recently released formal Office of Legal Counsel (OLC) opinion, the executive branch advances its constitutional position a step further, and a significant one: It not only staunchly affirms the president’s authority to control information and his authority to control individuals who have information but also constitutionalizes any executive branch practice or preference that furthers or protects that authority. And in its constitutional analysis, Congress’s countervailing constitutional authority is irrelevant.

In recent days, at the request of the Justice Department, the president also made his second “protective assertion” of executive privilege, this time over a category of documents and communications relating to the inclusion of the citizenship question on the census. Protective assertions of executive privilege are a similar species of constitutional prophylaxis and permit a similar form of one-sided analysis. Unlike a traditional assertion of executive privilege, which, under Supreme Court precedent is qualified and may be overcome by congressional interests in the specific information, a protective assertion is not subject to any balancing. Its use has been justified by the executive branch as a preliminary step to protect the president’s authority to make a traditional assertion of executive privilege later. But the facts surrounding this protective assertion suggest that it may now become an independent constitutional authority, detached from the requirement of a pending formal assertion of privilege and from its attendant balancing.

These prophylactic executive privilege doctrines are designed for one purpose: to protect the president’s authority to assert executive privilege. Where executive branch policies and practices are at issue, that makes some sense. The executive branch should adopt and pursue policies and practices that further its interests and protect its constitutional authority. But when the prophylactic doctrines are elevated to the status of a constitutional authority to countermand a co-equal branch’s exercise of its own authority under the Constitution, further analysis is warranted. To date, the executive branch has not provided it. And the omission of that inquiry, particularly in a purportedly comprehensive formal OLC opinion, is startling.


First, OLC released a formal opinion providing the legal basis for the administration’s decision in April to instruct two officials to disregard congressional subpoenas requiring their deposition. I wrote about that instruction on Lawfare in April, observing that the administration was establishing a “new constitutional weapon in the executive branch’s oversight arsenal” and demonstrating a willingness to play “constitutional hardball.” But I also noted that the decision, though unprecedented, was “grounded in the executive branch’s long-standing, but expansive, view of executive privilege” and represented “the next logical step in the constitutional one-upsmanship that has characterized the actions of both branches over the past 30 years.” The newly released opinion explains in depth the legal basis for this new authority. It relies on past precedents to justify the use of this novel constitutional authority. But it also stakes out remarkably expansive new territory about the executive branch’s constitutional authority to protect the president’s authority to assert executive privilege.

Second, in light of the House oversight committee’s planned contempt vote, the president asserted executive privilege over documents the committee had subpoenaed related to the inclusion of the citizenship question on the 2020 census. The assertion is actually a dual one: The president formally asserted executive privilege only over certain “priority documents” that were identified specifically in the committee’s subpoena. The priority documents include, among other documents, (1) a memo and note drafted by James Uthmeier, a senior counsel in the Department of Commerce, and hand delivered to John Gore, the acting head of the Department of Justice’s Civil Rights Division, and (2) the drafts of the Justice Department’s 2017 letter to the Department of Commerce formally requesting the inclusion of the citizenship question on the census as a mechanism for enforcing the Voting Rights Act. Both of those documents potentially relate to the heart of the contentious litigation over the inclusion of the citizenship question that is now pending before the U.S. Supreme Court, in which the plaintiffs contend—and the lower court concluded—that the rationale for the inclusion of the citizenship question provided in the Justice Department letter was pretextual.

The president’s actual assertion of executive privilege over the priority documents was accompanied by a protective assertion of executive privilege over all the other documents covered by the subpoena. This protective assertion was the second by President Trump and only the third protective assertion ever made. (The first was by President Clinton.) There appeared to be legitimate reasons to support Trump’s first use of this formerly unusual device: The House judiciary committee had given the Justice Department essentially two weeks to comply with a sweeping subpoena covering millions of pages of documents, many which included potentially sensitive information, including classified and grand jury information. The facts surrounding the new protective assertion of privilege are much different, however, and the decision to utilize this approach appears to rest on much more questionable grounds. As discussed below, this second assertion also reflects the potential that protective assertions of executive privilege may become ends in themselves, rather than mere means to protect actual assertions of privilege.

The OLC Opinion

The basic conclusion of the OLC opinion should not be all that surprising to anyone familiar with executive branch doctrine. But its constitutional analysis is strikingly absolute—and extreme. The opinion makes two principal constitutional claims: the committee’s refusal to allow an attorney representing the government to attend a deposition of a government official (1) “unconstitutionally interferes with the President’s right to control the disclosure of privileged information” (Part II.A of the opinion) and (2) “interferes with the President’s authority to supervise the Executive Branch’s interactions with Congress” (Part II.B). As it argues, “[e]ven if the President has not yet asserted a particular privilege, excluding agency counsel would diminish the President’s ability to decide whether a privilege should be asserted.” As a result, the subpoenas requiring depositions without agency counsel “exceeded the Committee’s lawful authority”; “lack[] legal effect”; and thus cannot be enforced through criminal, civil or inherent contempt procedures.

The opinion’s reasoning rests on three basic premises. First, there is a substantive authority or right that belongs to the president—the right to control disclosure of privileged information and the authority to supervise and control interbranch interactions, respectively. Second, the refusal to allow agency counsel to attend the deposition “interferes” with or “diminish[es]” that right. And third, a premise necessary to the second, the category of “privileged information” subject to the president’s absolute control includes not only information over which the president has asserted executive privilege because he has decided its disclosure would harm the public interest but also all information that the president could potentially decide to withhold. In other words, all information that fits within one of the various “components” of executive privilege recognized by the executive branch is subject to exclusive presidential control. On the basis of these three premises, the opinion concludes that the president has the authority to direct a lower branch official not to comply with a congressional subpoena. The opinion then concludes by rejecting the committee’s counterarguments, most of which focus on Congress’s constitutional authority, the longevity of the rule, and the committee’s interests in conducting depositions without agency counsel.

Each premise of the constitutional analysis warrants in-depth analysis that would be beyond the scope of this piece. Undoubtedly, many, if not most, legal scholars and others uninitiated in the “dark arts” of executive privilege would vigorously contest their validity. But to varying extents, the premises do reflect past doctrine. It is worth noting that the second premise has the least historical support. At the end of the Obama administration, OLC issued an opinion on the payment of private counsel for employees subpoenaed by a congressional committee to sit for a deposition. A footnote in that opinion noted that prohibiting agency counsel could raise constitutional “concerns” and “potentially undermine [both] the Executive Branch’s ability to protect its confidentiality interests” and “the President’s constitutional authority to consider and assert executive privilege where appropriate.” As the new opinion notes, the Justice Department has repeatedly objected to the practice of prohibiting agency counsel in past letters to congressional committees. But the executive branch has never before gone so far as to say publicly that the practice directly interfered with or impaired the president’s constitutional authority—let alone that the president has the constitutional authority, as a result of that impairment, to countermand the subpoena’s legality. Instead, the executive branch has made other arrangements, such as paying for an official’s private counsel.

Moreover, if the major premise of the new OLC opinion is true—that the president has absolute constitutional authority to control the provision of information to Congress and to control officials disclosing information—then it’s hard to see why the refusal to allow agency counsel to attend would necessarily impair that authority. The individual would be subject to that direction whether agency counsel is there or not. As Neil Eggleston (later White House counsel to President Obama) did with Sara Taylor (a former adviser to President George W. Bush) after Bush’s claim of executive privilege, an official’s private counsel could assist in maintaining any privilege line established by the president or agency. Indeed, OLC’s constitutional analysis implies that private counsel would be required to do so because his or her client would be legally obligated to comply with the executive branch’s direction. If questions about privilege lines arose during the deposition, agency counsel could be in the next room and the official being deposed could take a break to consult with agency counsel as necessary. In other words, it is impossible to see the prohibition on agency counsel as a direct infringement on the asserted constitutional authorities, particularly given the executive branch’s first premise about the president’s absolute authority and control over the information. At most, it arguably burdens their practice, or, in the words of the opinion “diminishes” to some unspecified degree the executive branch’s ability to withhold information.

The basic syllogism of the opinion’s constitutional argument looks something like this, leaving out the minor premise about what information is privileged, which supports the second premise:

1. The President has constitutional authority to control information disclosure.

2. The refusal to allow agency counsel in a deposition diminishes or interferes with that authority.

∴ The President has the authority to direct a subordinate not to comply with a subpoena requiring a deposition without agency counsel.

Even accepting the premises as true, they are not alone sufficient to reach the conclusion at which OLC arrives. Instead, the conclusion of the syllogism requires the truth of another unexamined—and somewhat startling—premise: that the president’s constitutional authority includes the ability to countermand any congressional action that impairs, interferes with, or diminishes his other presidential authority no matter the basis for the congressional action and without consideration of Congress’s constitutional authority to take that action. In the opinion’s words, congressional action that interferes with or diminishes the president’s authority in any way “exceed[s] the Committee’s lawful authority” and “lack[s] legal effect.”

To understand why this is startling, it’s helpful to look to the precedents on which the opinion relies. The primary support for the new OLC opinion is a 2004 OLC opinion, written by Jack Goldsmith called “The Authority of Agency Officials to Prohibit Employees from Providing Information to Congress.” Over the six pages that constitute the constitutional analysis, the Goldsmith opinion is cited six times, and it is the primary citation provided to support both of the presidential authorities discussed. The Goldsmith opinion was written in response to a Congressional Research Service (CRS) memo concluding that the Department of Health and Human Services did not have the authority to prevent or prohibit officers or employees from providing information to Congress in light of the Lloyd-La Follete Act of 1912 and various other statutes. As the Supreme Court has noted, that act, passed in response to “gag” orders issued by Presidents Theodore Roosevelt and Taft that prohibited federal civil servants from providing information to Congress without authorization from their superiors, explicitly guaranteed that the right of civil servants “to furnish information to either House of Congress, or to any committee or member thereof, shall not be denied or interfered with.” The legislative history of the act explained that the legislation was intended “to protect employees against oppression and in the right of free speech and the right to consult their representatives.”

The Goldsmith opinion concludes that the Lloyd-La Follete Act and other statutes on which the CRS relied do not override the executive branch’s constitutional authority to control the dissemination of privileged information and direct the conduct of lower branch officials—nor could any other statute. The statutes could not “constitutionally be applied, as [CRS] would apply them, to the circumstance where a government official instructs a subordinate government employee” not to provide particular information. The new OLC opinion relies on the Goldsmith opinion and other related precedent—a good deal of which addresses the distinct legal basis for control of classified information—for the proposition that Congress cannot interfere with the president’s authority to control the disclosure of information to Congress.

But purposefully or not, the new OLC opinion ignores a crucial distinction between the prohibition on agency counsel and the circumstances addressed by the Goldsmith opinion and other past precedents. In those past circumstances, the “interference” with presidential authority was direct and absolute; if allowed, the interference would have nullified the president’s authority entirely. The Lloyd-La Follete Act, by its terms, did not allow the president to control the disclosure of information. The Goldsmith opinion concluded that applying the statute in that manner would be unconstitutional, reasoning that the president must have the ability to assert control and, when necessary, refuse to disclose certain information. That is the core of executive privilege. The “remedy” was that the statutes could not constitutionally be interpreted in that manner.

The “remedy” in the new OLC opinion, by contrast, is an order by the president directing an official to refuse to comply with a congressional subpoena. And the opinion reasons that the president has that affirmative authority because of a committee rule that might impair the president’s asserted authority and imposes a burden on that authority. The rule prohibiting deposition counsel does not, in any way, nullify the ability to assert privilege.

To put the two situations on equal footing: The Goldsmith opinion supports the president’s authority to direct an official not to comply with a duly enacted statute requiring disclosure of information in particular circumstances. The new OLC opinion supports the president’s authority to direct an official not to comply with a duly issued subpoena requiring the individual to appear, not because an appearance would require the disclosure of particular information and nullify the president’s prerogative, but because the executive branch will not be able to have a representative in the room. In other words, the former combats a direct contravention of the asserted authority to control information disclosure. The latter combats a procedure that, at most, poses a danger to the president’s authority because an official may refuse to follow his superiors’ directions, may have a loose tongue, or may not have a counsel who is able to object when questions intrude into privileged areas. In reality, the denial of agency counsel mostly creates discomfort—because of the possibility that an official will ignore orders once she is alone with the committee—and inconvenience.

It is of course possible that a burden on or inconvenience to the exercise of a constitutional authority may amount to an unconstitutional action. But that requires analysis of both the extent of the burden and any countervailing constitutional authorities that permit the action. The new OLC opinion omits that analysis entirely. It never acknowledges the difference between the direct nullification of a constitutional authority and a burden on that authority. Nor does the opinion consider, as part of the constitutional analysis, how to balance the respective constitutional authorities of Congress. If Congress is exercising its valid constitutional authority to enact laws or rules for its procedures, and that exercise makes it more difficult or more inconvenient for the executive branch to perform some constitutional function, does that mean the executive branch can, unilaterally, nullify Congress’s actions? Or is more analysis required, balancing the need for Congress to exercise its constitutional authority against the extent of the burden on the executive branch’s authority in some manner?

The opinion takes the former position. Its conclusion relies on the unacknowledged assumption that any congressional action or rule that potentially diminishes or threatens the president’s implied authority to control the dissemination of information is unconstitutional and exceeds Congress’s authority. Full stop. The opinion rests its reasoning on that assumption without providing a theoretical defense for it and without any meaningful examination of Congress’s constitutional authority.

Particularly telling is the opinion’s use of the Supreme Court’s decision in United States ex rel. Touhy v. Ragen. In that case, the Supreme Court held that an inferior executive branch official was justified in withholding documents because he had been ordered to do so by the attorney general. The new OLC opinion first cites Touhy as indirect support for the statement that the president has constitutional authority to “control communications with, and information provided to, Congress on behalf of the Executive Branch.” The opinion also boldly claims that the court “recognized in Touhy” that the “head of an agency may properly bar subordinate officials from disclosing privileged agency information.” Those statements are true in a literal sense. But even the most cursory review of the Touhy opinion reveals that the Supreme Court found the attorney general’s direction “proper” solely on the basis of a statute and expressly declined to reach the constitutional question about the attorney general’s authority to withhold the information. The court held it “was appropriate for the Attorney General, pursuant to the authority given him by 5 U.S.C. 22, to prescribe regulations not inconsistent with the law” for the disclosure of information (emphasis added). The court made it clear that it was concerned “only with the validity of [the attorney general’s] order” and that “[t]he constitutionality of the attorney general’s exercise of a determinative power as to whether or on what conditions or subject to what disadvantages to the Government may he refuse to produce government papers under his charge must await a factual situation that requires a ruling.”

Touhy was thus a statutory case and did not, in any way, address the constitutional balance of authority between the branches. Touhy concluded that Congress had authorized the attorney general to issue such an order, a conclusion that Congress later amended the statute at issue to rebut. And even that statutory holding was qualified by the requirement that the order not be “inconsistent with the law.”

Directing an official to disregard an admittedly valid congressional subpoena is, without question, a direction that is “inconsistent with the law”—unless of course there is a constitutional defense. Touhy’s statutory conclusion that the direction to the inferior officer was lawful is the precise analysis that the OLC constitutional conclusion omits. Even if it’s true that the deposition procedures would reduce or diminish the president’s constitutional control to some contested degree, why, as a matter of constitutional law, does that potential diminishment provide the president the affirmative authority to nullify an exercise of Congress’s constitutional authority? Answering this question would seem to require at least some analysis of Congress’s constitutional authority relative to the executive branch in general and specific balancing of that authority against the extent of the burden. Unlike Touhy, the question here involves competing constitutional authorities, not a statutory exercise of authority blessed by Congress. The new OLC opinion’s unstated assumption is that the executive branch’s authority is so vastly superior that any infringement, however slight, means Congress’s authority must yield. This is striking.

In short, the opinion arms the executive branch with an expansive new constitutional authority—a prophylactic power grounded in executive privilege—with barely a mention, let alone analysis, of the constitutional authority of Congress that is being nullified. And the historical basis for this constitutional ruling is almost wholly the policies and practices formerly pursued and advocated (not always successfully) by the executive branch to protect its interests. The opinion rejects any need to analyze the extent of the burden on the president’s authority or to balance Congress’s countervailing interests or constitutional authority. As a method of constitutional interpretation, that approach appears incomplete, to say the least—and its misleading use of precedents like Touhy to fill the gap speaks volumes. As a matter of practice, this mode of analysis has the potential for wide ramifications on oversight. It creates a new constitutional barrier to congressional oversight, the sole purpose of which is to ward off any potential hindrances to or burdens on the president’s constitutional authority to assert executive privilege. Under this regime, disputes over procedures—such as Attorney General William  Barr’s insistence that the House judiciary committee not allow staff attorneys to question him at a public hearing—become constitutional disputes. And, under the analysis in the OLC opinion, the executive branch’s constitutional interests always prevail.

Assertions of Privilege Over the Census Documents

At this point in the ongoing oversight war, the basic contours of traditional assertions of executive privilege have been discussed exhaustively. And the president’s formal assertion of executive privilege over the “priority” census documents follows that traditional approach. The drafts and memo are almost certainly deliberative in nature and pre-decisional, and they would thus likely fall within the scope of the deliberative process component of executive privilege. Deliberative process has been the primary argument of most modern executive privilege assertions, including for the 2012 assertion by President Obama in the “Fast and Furious” case. Thus, even though Congress has long disputed that the executive branch may withhold information on the basis of deliberative process, an argument they lost in the litigation over the Fast and Furious assertion, the executive branch’s decision to do so here is nothing new. And although, in litigation, the qualified deliberative process privilege is somewhat easy to overcome (pursuant to an ad hoc balancing test), the executive branch has long taken the position that the showing of need required to overcome a claim of executive privilege, even one based on deliberative process, is much higher. Barr’s formal letter requesting the privilege assertion does not break much new ground in this area. The executive branch’s standard is that the congressional committee must “point[] to ... specific legislative decisions that cannot responsibly be made without access to [the privileged] materials,” a standard derived from the Watergate case in which the Senate select committee sought access to the White House tapes already obtained by the House judiciary committee pursuant to its impeachment inquiry.

It is worth noting that, in the Fast and Furious dispute, the Justice Department did offer to make available pre-decisional, deliberative drafts of a letter it sent to Congress. But the fact that the department has in the past turned over draft letters does not prevent it from withholding drafts in another dispute pursuant to an executive privilege claim. And it would likely claim that its actions in Fast and Furious reflected the extraordinary circumstance that the department had sent a letter to the committee that the department later acknowledged to be inaccurate.

The argument in favor of the protective assertion of privilege over the remaining subpoenaed materials has much less, if any, historical support. Facing the planned contempt vote, the Justice Department, the Commerce Department and the White House did not winnow down the subpoenaed documents and make available everything over which the president could not formally assert executive privilege as past practice dictates. Instead, the administration asserted privilege over the specific documents mentioned in the subpoena and included a protective assertion of executive privilege over the remaining categories of documents, allowing it to withhold everything. This combination of an assertion of executive privilege with a protective assertion of executive privilege is unprecedented. And in a war of constitutional hardball, that combination may provide the executive branch a lethal new weapon.

Given the facts, there are substantial reasons to wonder if this protective assertion of executive privilege waters down the concept and reduces it to another highly manipulable tool the executive branch can use to delay and to insulate executive branch officials from any penalty. In the one historical precedent prior to this administration—President Clinton’s protective assertion over subpoenaed White House counsel documents in 1996—the president made a protective assertion on May 8, 1996, that was designed “to ensure [the president’s] ability to make a final decision, after consultation with the Attorney General, as to which documents are deserving of a conclusive claim of the privilege.” He made the final determination about two weeks later.

In the first protective assertion, Barr’s letter noted that the subpoena encompassed “millions of pages of classified and unclassified documents bearing upon more than two dozen criminal cases and investigations, many of which are ongoing.” The subpoena was issued on April 18 and imposed a deadline of May 1. The contention that the Justice Department and the White House needed more time to review and consider all of that information before handing it over thus seemed to have some validity.

This new protective assertion is much different. The subpoenas were issued on April 2 and sought a list of specific documents, plus about a full year of documents and communications from the Justice Department and the Commerce Department “regarding the request to add a citizenship question.” Although there are undoubtedly numerous emails and communications that fall into that category (Barr’s letter puts the number of “pages” in the tens of thousands), agencies process such requests all the time under the Freedom of Information Act. The congressional request is limited to emails related to the citizenship question, and Barr’s letter does not suggest any of the communications would implicate the kinds of classified information or law enforcement sensitivities that were implicated by the subpoena for the special counsel’s materials. The administration has had two months to go through the stack of emails and documents and redact or remove deliberative information or information protected by other privileges such as the attorney-client or attorney work product privileges. Although burdensome, this work is certainly no more burdensome than similar tasks the executive branch has undertaken—an example being the enormous amount of material processed for similar privileges during the nominations of Neil Gorsuch and Brett Kavanaugh to the Supreme Court. And the administration already knows, without a doubt, the most sensitive communications that it would not disclose.

More importantly, there has already been extensive discovery of similar material in the ongoing litigation. It strains credibility to say the litigating divisions have not already reviewed most, if not all, of the requested communications as part of that litigation. And Barr’s letter acknowledges that some of the materials have already been identified as having privileged information.

In other words, the need for the protective assertion here appears to be grounded solely on the executive branch’s unwillingness to expend the effort to review the documents before the contempt vote. But the availability of a protective assertion means there is little reason to bother expending that effort in the first place. The protective assertion of privilege provides another way for the executive branch to delay and stymie congressional demands. And it’s particularly potent because, by its nature, it relieves the executive branch of the burden of justifying the type of documents withheld or balancing congressional need for the information. All of that belongs to formal assertions. Moreover, the protective assertion practically results in the protection of all documents, even those the executive branch acknowledges are not privileged.

The combination of a protective assertion with an actual assertion makes it even easier to utilize. One historical limitation on the president’s authority to assert executive privilege has been political. The act of asserting executive privilege involves the use of some political capital, and such assertions are almost always derided by journalists (who tend to favor transparency) and scholars (who are skeptical of executive privilege in general). A protective assertion will likely be reported no differently than a traditional assertion, as the coverage of the president’s first protective assertion demonstrated, and impose the same political costs. But there are no additional costs to a protective assertion if the president is also formally asserting privilege over specific documents.

Notably, Barr’s first protective assertion letter added two words to the justification from the Clinton protective assertion, stating that the protective assertion was “designed to ensure the [president’s] ability to make a final assertion, if necessary, over some or all of the subpoenaed materials” (emphasis added). Thus, Barr’s letter contemplates that a formal assertion may not even be necessary but that negotiations will continue. And that appears to have in fact occurred with respect to the first assertion; the House and the Justice Department seem to have reached a preliminary agreement on those documents, and the committee agreed to hold its contempt vote in abeyance. The protective assertion thus appears now to be a stand-alone entity, unmoored from the necessity of a later actual assertion over the covered materials and distinct from a traditional assertion. This protective assertion does not require balancing congressional interest or any analysis of the withheld information; yet it is also sufficient to immunize an official from any penalty due to noncompliance with Congress’s demands.

As I noted with respect to the first protective assertion, part of the blame for establishing the validity of a protective assertion lays at the feet of Congress. Unreasonable demands for large swaths of information in short time periods and rapid contempt votes will not help Congress’s case, particularly when the category of information sought clearly implicates privileged material. But when the executive branch is the entity that determines whether a demand is unreasonable and whether it has been given sufficient time to review and respond, the protective assertion as a stand-alone entity is also easily subject to abuse. Traditionally, the only reason the executive branch would consider a protective assertion would be to protect individual officials from contempt in circumstances in which the timing made a traditional assertion impossible. But such protective assertions now seem to be a distinct possibility any time a committee schedules a contempt vote. To counter the protective assertion, a congressional committee might make narrow requests, similar to the priority documents specified by the oversight committee related to the census. And the committee can document the time and opportunity it has provided the executive branch to respond in its contempt resolution. Ultimately, however, the maxim of most oversight disputes applies: The executive branch makes the final decision, and judicial review takes time.

No matter what congressional committees do, the utilization of this newly invigorated tool is up to the executive branch. If the administration continues to rely on protective assertions with more and more frequency, either in combination with or instead of formal assertions of executive privilege, it will establish a prophylactic layer of constitutional protection that extends well beyond the boundaries of executive privilege and dispenses with a core part of the executive privilege analysis—balancing congressional interests. The most recent protective assertion is evidence that the administration is well on its way to erecting that new constitutional barrier. In my experience at OLC, by the time a subpoena had been issued and a committee had started rattling its sabre about contempt, the agency would have long ago collected the responsive documents—even if they numbered in the tens of thousands of pages—and begun to review them and ask for advice on privilege. The agencies and OLC did so to be prepared in case an impasse arose and an executive privilege claim became necessary on short notice. With a watered-down version of a protective assertion of privilege, however, that preparation is not only unnecessary but is contrary to the executive branch’s interest. If the executive branch aims simply to withhold documents for longer while providing as little justification as possible, the best approach under this regime is to wait until a contempt vote, immunize officials with a protective assertion of privilege, and only then start to look at the documents to consider whether an actual assertion may be necessary.

The watered-down protective assertion of executive privilege, like the new OLC opinion, suggests that past executive branch practices initially designed to protect the president’s constitutional authority to assert executive privilege are transforming into independent constitutional authorities—facets of a new prophylactic executive privilege that extends well beyond the bounds of executive privilege itself.