Avoiding Another McGahn: Options to Modernize Congress’s Subpoena Compliance Tools

On Aug. 31, the U.S. Court of Appeals for the D.C. Circuit ruled that the House of Representatives had asserted no valid cause of action in seeking judicial enforcement of a congressional subpoena issued to former White House counsel Don McGahn. The majority in a divided panel (2-1) acknowledged Congress’s constitutional authority to compel the production of information; and it suggested that the House could, under the right circumstances, seek judicial enforcement of its compulsory processes. But the court determined that it was precluded here from doing so in the absence of a statute expressly authorizing the House to file suit. The decision all but ensured that McGahn would not testify before the November election; and on Oct. 15, the full bench of the D.C. Circuit confirmed it. The en banc panel will again rehear the case at the House’s request, but in February of 2021—after this current session of Congress expires.

What, now, is at stake? According to media coverage, the Aug. 31 decision, if allowed to stand, will have “neutered” the House’s ability to enforce its subpoenas, posing an “existential threat to congressional oversight.” Speaker of the House Nancy Pelosi warned that the ruling “threatens to strike a grave blow to one of the most fundamental Constitutional roles of the Congress: to conduct oversight on behalf of the American people.” It is understandably troubling that a recalcitrant executive branch will be permitted to evade oversight, and that voters may be deprived of pivotal information about President Trump’s conduct in office, at least for now.

But the court’s decision, if upheld, will not “eviscerate” Congress’s enforcement powers, because the courts were neither a particularly effective nor an exclusive enforcement option to begin with. The House Judiciary Committee subpoenaed McGahn in April 2019 for information related to former Special Counsel Robert Mueller’s investigation into Russian election interference. Had the panel ruled in the House’s favor, it still would have taken the chamber more than 500 days to compel McGahn to appear for testimony. And once McGahn arrived on the Hill, he most likely would have claimed immunity in response to each question, requiring additional litigation to specify the scope of his need to share information across 36 topics of interest to the committee. A civil lawsuit was hardly a practical compliance mechanism that has now suddenly, with the court’s most recent decision, been made ineffectual.

Instead, the Aug. 31 decision—and the delay associated with the full court’s decision to again rehear the case—helps to clarify the limitations of relying exclusively on the judiciary to vindicate the legislature’s investigative powers and corral those who defy them. No matter how McGahn is ultimately decided, the episode should prompt Congress to reexamine and dust off its own potent constitutional toolbox to effectuate its oversight authorities. Congress enjoys the authority to coerce contemnors itself, as it used to do through arrests. It could do so again today by levying fines. Congress also can wield the threat of criminal prosecution, as it used to do by referring contempt citations to the Department of Justice. It could do so again through independent counsel. When necessary, and without the courts, Congress can go far in compelling executive branch cooperation through its own self-validating powers. But it would need to modernize its historically effective tools for doing so first.

Civil Litigation Is Novel—and Precarious

Despite today’s expectation of lawsuits to enforce congressional subpoenas, civil litigation as a compliance framework is novel. In 1973, a select investigative committee in the Senate filed the first-ever civil action to enforce its subpoenas of the executive branch, including one for tapes of conversations between President Nixon and his White House counsel. The suit was first dismissed by the trial court on jurisdictional grounds, finding that the Senate had no authority to sue to enforce its subpoenas absent a statute authorizing it to do so (Senate Select Committee on Presidential Campaign Activities v. Nixon). Although Congress in turn expressly provided the Senate with a cause of action (which, as it happens, is what the court last month suggested the House ought to do), the appellate court ultimately sided with the president anyway.

In its quest to unearth illegal and unethical executive activity, litigation was a precarious strategy. It was also an institutionally threatening one. The committee had argued that Nixon’s tapes were necessary to effectively conduct oversight and to legislate—for instance, to consider new regulations of campaign activities. The court disagreed, finding that the committee’s interests were outweighed by the president’s privilege claims. (The courts later ordered that the tapes be turned over to them, making the courts, as Josh Chafetz argues, “the heroes of the Watergate story, but only by acting in such a way as to suggest that Congress was not up to the task.”) Congress had invited one branch to weigh in on how it should go about its most basic constitutional functions with respect to another branch, and it did not pan out as planned. It would take more than three decades before Congress again turned to the judiciary as a method of enforcing its executive branch subpoenas, presumably because it had learned a lesson.

Congress filed its second-ever civil suit to compel executive branch compliance with its subpoenas in 2008 (Committee on the Judiciary v. Miers), which initiated civil action as the default method used by lawmakers to enforce subpoenas. No case then or since has delivered to Congress the information it requested, in a practically useful timeframe, and on grounds unequivocally vindicating its Article I authorities. So why, then, has the body continued to turn to civil action? Why did we arrive at McGahn?

Traditional Compliance Tools Are Impotent

Congress’s two other historically predominant methods of enforcement—inherent and criminal contempt—are both in practice null, victims of executive evisceration and congressional abdication. The weakening of each has pushed Congress to look elsewhere. Inherent contempt, or the authority of either chamber to punish nonmembers for obstructing its work, including for defiance of its subpoenas—upheld in early Supreme Court decisions (Anderson v. Dunn and McGrain v. Daugherty)—gradually fell out of favor as a practical method of recourse, last used in 1935. Inherent contempt is self-validating, such that Congress determines if its work has been obstructed and remedies the obstruction itself. Yet the means of enforcement, including arrests by the sergeant-at-arms and lengthy contempt proceedings at the house bar, were eventually deemed too onerous. Congress thus complemented inherent contempt with a statute making contemnors criminally liable (2 U.S.C. § 192), referring violations to the attorney general, “whose duty it shall be to bring the matter before the grand jury for its action” (2 U.S.C. § 194). Rarely, though, did Congress need a heavy hand to enforce its demands for information from the executive branch. For more than two centuries, requests were mostly negotiated through an informal accommodation process—a give-and-take between the branches that resolved disputes and maintained some level of comity.

But since at least the 1980s, the executive branch has adopted the absolutist posture that Congress’s information requests and the authority to enforce them can be ignored at its discretion. Office of Legal Counsel opinions (initially in 1984 and 1986) offer prosecutorial discretion and executive privilege claims as justifications for an attorney general’s refusal to present a criminal contempt citation against an executive official to a grand jury, in contravention of statute; and, without apparent justification, dismiss the constitutionality of inherent contempt altogether. Despite contestation of even the factual accuracy of these opinions—for example, that the legislative history of the criminal contempt statute suggests it was not intended to apply to executive officials (it was)—they have been used liberally by the executive branch as grounds for noncompliance. Just this May, citing only itself, the Office of Legal Counsel concluded that “Congress could not lawfully exercise any inherent contempt authority against Mr. McGahn for asserting immunity” and that the “criminal contempt of Congress statute does not apply.”

Both Office of Legal Counsel opinions recommended that Congress instead give civil suits a try—reasoning, it seems, that the courts are a more amenable battleground for vitiating effective oversight. Indeed, litigation is lengthy and uncertain. But its chief advantage for the executive branch is the absence of a downside to noncompliance. Unlike inherent and criminal contempt, civil suits exact no real cost. Why not kick disputes to the courts? At worst the executive is kicked back to where it started—and complies.

Congress Has Potent Options to Modernize Its Tools

The executive branch has largely succeeded in normalizing noncompliance. Although the current administration’s regular and blanket refusals to accommodate congressional requests are extreme, they are also the culmination of a long-running and bipartisan effort to deprive Congress of its enforcement tools. Congress has largely acquiesced. In response to McGahn’s obstinacy, the House refrained from bothering with a criminal contempt citation altogether, moving instead straight to litigation. But Congress has other options. Unlike civil action, inherent and criminal contempt carry the potential to significantly increase the cost of noncompliance and better incentivize cooperation. While simply re-exercising these powers through past methods is unlikely to work—both chambers have determined that arrests would be unpalatable today as an exercise of inherent contempt, and criminal contempt requires Justice Department cooperation, which, for now, is moot—Congress could devise new methods of exercising each. These methods should focus principally on the task of enticing cooperation, such that noncompliance is costly and accommodation more appealing.

First, Congress could levy fines as a modern method of exercising inherent contempt in place of arrests. If a current or former executive branch official with the authority to effect compliance is held in contempt of Congress, the citation would trigger a schedule of monetary penalties—for instance, a $30,000 minimum initial fine, increased in $30,000 weekly increments until the contempt is purged or until some ceiling is reached. A similar proposal was introduced in the House Committee on Rules in June. The repercussions facing contemnors like McGahn are now ambiguous at best and nonexistent at worst. A schedule of monetary penalties would at minimum create a more concrete expectation of consequences, in turn providing Congress with leverage in interbranch negotiations over requested information. The approach also carries the advantage of being self-validating: As an exercise of an inherent power, establishing a procedure for levying fines would require a change only to chamber rules rather than require presenting legislation to the president.

When previously exercised through arrest, inherent contempt served as a credible threat more than a routinely employed power. But “credible” suggests that Congress could, in fact, make use of it if needed. Unlike the House’s sergeant-at-arms, who may arrest individuals, no implementation mechanism exists to recover fines. To create one, the House could direct the Treasury Department to freeze assets of equivalent value if contemnors fail to remit. Or, with a disobliging department, the House could alternatively authorize its Office of General Counsel to contract directly with collection agencies. Levying fines may also raise constitutional questions. Citing a lack of precedent, a 2019 Congressional Research Service report concluded that Congress’s authority to do so is “an open question.” Kia Rahnama argued that precedent is, to the contrary, quite clear. The Supreme Court has repeatedly affirmed that judicial and congressional contempt are “governed by the same principles” (Jurney v. MacCracken) and thus imply the same means of vindication, including the authority to levy fines (Kilbourn v. Thompson).

As with any expression of power, monetary penalties raise the specter of partisan abuse. Would the tool, once available, be wielded irresponsibly by partisans? Certainly, the risks and consequences of abuse could be minimized. For example, a resolution establishing provisions for levying fines could limit the categories of officials subject to the penalty, explicitly including agency heads and White House officials while excluding most others. There is little need to allow for penalizing mid-level officials, anyway, as noncompliance is typically directed from the top. Congress could also provide a clear cause of action such that any contemnor may contest the fine in court. While this might again invite the judiciary into an interbranch dispute, it would nonetheless shift the burden between the branches, such that the executive would require a court’s blessing to avoid cooperation in place of Congress seeking judicial intervention to compel it. Still, any constitutional power will remain vulnerable to abuse. Congress’s power of the purse grants it the authority to shut down the government—not always exercised responsibly. But to keep separate the sword from the purse, there the power remains. In Anderson v. Dunn, the Supreme Court acknowledged this risk, wherein Congress could exercise its contempt power when “the ground is too broad.” And yet the alternative was worse: “the total annihilation of the power of the House of Representatives” to defend itself. Not mincing its words, the court determined this power was a matter of “self-preservation.”

Second, to revive its ability to enforce contempt criminally, Congress could establish a procedure for appointing an independent counsel to prosecute criminal contempt citations if the Justice Department declines to do so itself. Unless specified otherwise, laws enacted by Congress, including criminal law, apply equally to everyone. Congress should ensure that there is no ambiguity here as it relates to executive officials, who should not enjoy broad discretion to selectively enforce the law as it applies to them. When attorneys general—who have of late assumed the role of judge and jury in their own contempt cases—refuse to refer criminal contempt charges to a grand jury, Congress requires an alternative. In such cases, a reformed criminal contempt statute could authorize the appointment of an independent counsel by a chief judge or panel of a district court when requested by Congress.

Analysis by the Congressional Research Service suggests that such an option might be modeled on the now-defunct Independent Counsel Act (ICA), enacted in the post-Nixon era and validated by the Supreme Court in Morrison v. Olson. In the aftermath of Watergate, the ICA sought to address the who-watches-the-watchman problem by devising a process for appointing independent counsel to investigate and prosecute executive officials for violations of federal law, including criminal contempt of Congress. An attorney general would conduct a preliminary investigation to determine reasonable grounds for further investigation after a request from Congress; and the attorney general, if she determined the grounds were sufficient, would request appointment of an independent counsel from a three-judge panel of the D.C. Circuit. Of note, Congress designed the procedure—which turned on the discretion of the attorney general—in an era of greater Justice Department political independence, and when the attorney general herself had never been held in contempt of Congress. Modification of the ICA model for the purposes of enforcing congressional subpoenas would thus include permitting Congress to itself directly request appointment of independent counsel from the court if the attorney general declines to do so within some given period of time. That time period would allow the attorney general, if she concludes there are insufficient grounds for further investigation, to make her case to Congress—and, in turn, the public—for a revised or rescinded request.

The ICA is an imperfect model. For instance, it faced repeated bipartisan criticism for allowing independent counsels to broaden their own jurisdiction, investigating affairs Congress did not intend and the executive branch resented. Any modernized mechanism could thus explicitly limit a counsel’s authority to the investigation and prosecution of contempt. Additionally, when upheld by the Supreme Court in Morrison, the ruling rested in part on the court’s assessment that the mechanism’s design allowed for sufficient power-sharing, particularly given the attorney general’s discretion in requesting an appointment. But “[m]ost importantly,” the court concluded (emphasis added), “the Attorney General retains the power to remove the counsel for ‘good cause,’ a power that we have already concluded provides the Executive with substantial ability to ensure that the laws are ‘faithfully executed’ by an independent counsel.” An adapted ICA mechanism for the purpose of prosecuting a criminal contempt citation could preserve this executive branch backstop. A 2009 bill provides a model to address both concerns, authorizing an attorney general to remove an independent counsel for good cause, subject to judicial review, and explicitly and narrowly limiting its jurisdiction.

However, Morrison may be on shaky ground, potentially jeopardizing any new statute modeled on the ICA. Justice Antonin Scalia’s lone dissent on separation of powers grounds not only has been celebrated by some conservatives for advancing the theory of a unitary executive but also has found some bipartisan sympathy. Today, in any case, the high court’s composition may be unlikely to come to its defense. Moreover, unlike inherent contempt, a statutory procedure for the appointment of an independent counsel would require that a president sign it into law. That may be unlikely. In place of the statutory route, Morton Rosenberg and William Murphy, of Good Government Now, alternatively suggest that the House rely on its internal rulemaking to authorize the direct appointment of outside counsel by the speaker. The authority to exercise enforcement of a criminal contempt citation would, like monetary penalties, flow from Congress’s inherent power to compel the production of information. If fines prove insufficient to ensure compliance, Congress can escalate its enforcement actions by hiring its own counsel to criminally prosecute a contemnor.

No doubt, regardless of the pathway—through a court- or speaker-appointed counsel—a criminal suit, like a civil suit, would still prove lengthy and undesirable. However, the threat of a criminal suit is different, meaningfully raising the personal cost of defiance. As Rosenberg and Murphy observe, the threat of criminal prosecution has historically served as an important backstop to inherent contempt, encouraging officials to negotiate with Congress and settle on satisfactory accommodations. For example, between 1975 and 1998, “there were 10 votes to hold cabinet-level executive officials in contempt. All resulted in complete or substantial compliance … before the necessity of a criminal trial,” with evidence that in various cases, “contemnors were reluctant to risk a criminal prosecution to vindicate a presidential claim of privilege or policy, which led to settlements.” Credible threats strengthen incentives to negotiate.

Finally, Congress also has the option to solve for the drawbacks of civil enforcement where possible—in particular, to expedite litigation and blunt the executive branch’s run-out-the-clock strategy. For example, Congress could establish a specialized federal court to hear subpoena compliance cases and require direct appeal to the Supreme Court. (The House passed a bill to this effect in 2017, which was reintroduced during the current Congress.) But this would likely run into its own practical problems. Expedited proceedings are prone to administrative difficulties and may not, in fact, be all that expedient—particularly in cases involving complex requests for information in which judges are parsing through a variety of privilege claims. For instance, in Committee on Oversight v. Holder, the Justice Department “disgorged more than 10,000 documents originally withheld, totaling more than 64,000 pages, which took a special master over a year to pore through and address.” Judges attempting to weigh abstract separation of powers considerations may complicate matters further. The Supreme Court’s ruling in July in Mazars v. Trump, for example, sets forth a complex, four-part balancing test to evaluate the validity of congressional subpoenas aimed at the president’s personal information. Michael Solimine finds that specialized courts tend to be more amenable to technically unique areas of the law, like patent or tax law, and less adept at swiftly adjudicating these “open-ended” constitutional questions.

Perhaps most importantly, it remains unclear how speeding up litigation—if cases were in fact expedited—would meaningfully alter executive branch behavior. The cost of noncompliance, and thus the incentive to negotiate and accommodate, remains unchanged. The argument for optimizing civil action would seem to rest less on its ability to incentivize interbranch accommodation and more on the assumption that Congress will receive quick, decisive wins in court. But these wins are unlikely (see McGahn). Moreover, even apparent “wins” may dilute congressional power over time. In Committee on Oversight v. Lynch (carried over from Holder), while the court (eventually) mandated Justice Department compliance, its reasoning may have undermined Congress’s ability to obtain similar information in the future. The court dismissed the department’s argument that the deliberative process privilege justified its noncompliance. But in doing so, it also—and for the first time—explicitly found that this prong of executive privilege had some constitutional foundation, arming the executive in future litigation. According to an alarmed Sen. Chuck Grassley, “despite the court’s order to the Department to produce documents... [the opinion] is a major threat to the oversight powers of the legislative branch.” “The judge,” he concluded, “gave the House a victory in practice, but gave the Department a victory on the principle.” More recently, in August, the State Department cited Mazars when refusing to comply with a House subpoena, deciding for itself that the requested documents would serve no “valid legislative purpose.” The Mazars ruling largely reaffirmed Congress’s constitutional power to obtain information. Still, the decision was a “mixed bag for Congress,” as Molly Reynolds and Margaret Taylor note, “setting a new—and much higher—standard for establishing the legitimacy of congressional investigations generally.” Courts already wary of wading into interbranch disputes are unlikely to hand Congress clear victories.

Courts are far less wary in matters involving their own inherent powers, including contempt. As the Supreme Court has held, “Courts cannot be at the mercy of another Branch in deciding whether [contempt] proceedings should be initiated.” Judicial enforcement of judicial orders “is regarded as essential to ensuring that the Judiciary has a means to vindicate its own authority without complete dependence on other Branches.” Otherwise, “the courts [are] impotent” (Gompers v. Bucks Stove & Range Co.). The courts have suggested that Congress could likewise be less dependent. The D.C. Circuit, rightly or wrongly, found in McGahn that the House could not compel the judiciary to enforce its subpoenas through an implied cause of action. But it did acknowledge, without hesitation, that the House is capable of enforcing subpoenas itself. “Congress has long relied on its own devices,” the court observed, “either its inherent contempt power … or the criminal contempt statute.”

Shifting the Burden

Breathing new life into Congress’s contempt powers will instigate a forceful reaction; high-level executive branch officials will not take kindly to the threat of being fined or criminally prosecuted. An executive branch committed to a doctrine of voluntary cooperation will challenge each new assertion of congressional authority.

Critics may contend that vigorous subpoena enforcement actions by Congress will thus simply end up in litigation anyway. This is almost certainly true. Congress’s authority to levy fines in the absence of a history of doing so would be challenged, and executive branch officials may claim as a conceivable defense that they are following orders. Scalia’s dissent in Morrison would find new airtime in a challenge to an independent counsel mechanism, and any effort by a chamber of Congress to hire its own counsel to assist with criminal enforcement would be constitutionally contested by the executive as an usurpation of its law enforcement function. These are real obstacles.

But congressional reassertion shifts the burden away from a Congress that asks for its powers and toward an executive that must ask the courts to deny them. It will fall on the executive branch, not on Congress, to initiate protracted litigation and contest long-recognized constitutional powers. A shift in burden may in turn shift the political incentives governing executive branch behavior, encouraging better-faith accommodation of a co-equal branch.