A career civil service insulated from partisan retaliation has been a fixture of American government since the enactment of the Pendleton Act 135 years ago. Such a bureaucracy can be a substantial impediment to a president seeking to re-make government in his or her image. Naturally enough, President Donald Trump’s executive order excepting administrative law judges (“ALJs”) from the competitive civil service, as well as from some civil service rules limiting removal, has inspired some protests that he is undermining the independence of the civil service. Rep. Richard Neal (D-MA), ranking member on the House Ways and Means Committee, denounced the order as “yet another example of President Trump putting special interests and loyalists ahead of American families’ wellbeing,” because the order allegedly undermines ALJs’ “impartiality.” The American Constitution Society’s Carol Fredrickson declared that the EO “could have a stunning impact on how myriad administrative claims are handled,” because “[p]olitical appointment could…lead to more administrative law judges with pro-corporate anti-worker biases.” And the president of the American Bar Association has even urged legislation to prevent implementation of the executive order because it “has the potential to politicize the appointment process and interfere with the decisional independence of ALJs.”
Are these worries paranoid or substantial? And does this change in ALJs’ status represent a deliberate assault on bureaucratic independence or is it merely—as the EO states–an effort to comply with the Supreme Court’s recent decision in SEC v. Lucia?
With respect to the first question, I suspect that the critics’ worries are overblown: Civil service exams are not really an important protection for ALJs’ independence from political control. With respect to the second question, I believe the EO is either a deliberate but ham-handed attack on bureaucratic independence or simply an incompetent reading of the Supreme Court’s decision. In any case, SEC v. Lucia cannot explain or justify the EO. The good news, however, is that the order will have little effect, given the existence of statutory safeguards against political retaliation that Trump cannot waive.
Does Trump’s EO undermine the sort of bureaucratic impartiality required by the Administrative Procedure Act?
The roughly 1,900 ALJs sprinkled throughout the federal government hear cases for which formal APA adjudication is required under APA sections 556 and 557. That means that they deliver bureaucratic justice with more impartiality and trial-like formality than a mere hearing officer handling informal adjudication outside the APA. ALJs must not, for instance, have ex parte contacts with enforcement officials at the agency. (For an excellent summary contrasting ALJs and hearing officers, see this piece by Kent Barnett.)
Trump’s EO changes none of these safeguards for adjudicative impartiality. His order’s major change is to except ALJs from the elaborate civil servant criteria limiting eligibility to serve — in particular, the requirement that the agency chief select ALJs from the top three-highest scorers on a civil service examination administered by the Office of Personnel Management (OPM). (For the details of the eligibility criteria, see OPM’s website and this CRS study). The point of these limits on eligible candidates, of course, is to prevent partisan or other improper bias in the administration of the law. By excepting ALJs from the competitive civil service, Trump has theoretically opened the door to (in Rep. Neal’s words) “the appointment of judges who are big campaign donors, beholden to industry, or otherwise unqualified.”
Put aside the inherent improbability that a “big campaign donor” would angle for a job as one of 1,900 ALJs adjudicating small-fry issues about unemployment benefits and the like. Put aside also the possibility that choosing ALJs based on a test score might not make much sense (as Barnett noted in his excellent review of the EO). Instead note that if politically appointed agency chiefs wanted to deliver biased adjudications to political allies they could do so themselves without appointing a willing tool as an ALJ. The agency chief is typically the highest “court” within the agency and generally has full power of de novo review over ALJs’ decisions. Under APA section 557, “[o]n appeal from or review of the initial decision, the agency has all the powers which it would have in making the initial decision except as it may limit the issues on notice or by rule.” If a cabinet secretary wants to do a favor for a Republican Big, then he can order up the record and make the decision himself: He does not need to appoint the Big to make the initial decision.
Barnett worries about Trump’s EO’s “allow[ing] the current administration (and then perhaps later administrations) to pack the agency’s ALJ corps with those whom it thinks are sympathetic to the current administration’s positions.” He notes that, by packing administrative tribunals with like-minded people, an agency could create “a one-sided culture within the ALJ corps,” which could lead to “criticism …that agency adjudicators are unprofessional and biased in favor of the agency.” Such accusations of bias, according to Barnett, could undermine the legitimacy of the administrative state.
I am more ambivalent about whether discrediting the legitimacy of the administrative state with “court-packing” is a bad thing. My ambivalence springs from the character of the American administrative state: As constituted by the APA and most agencies’ organic statutes, maybe it ought to be discredited. There is actually very little separation of powers within most agencies protecting the agency judges’ independence of judgment from the views of the agency chief. Only a handful of statutes give administrative judges any right to make fact-findings immune from de novo review by the agency chief. (For instance, the Occupational Safety and Health (OSH) Act gives the OSH Review Commission a right to make fact findings independent of the OSH Administrator.) Instead, for most agencies, the agency chief is to an ALJ what a district court judge is to a magistrate: The ALJ is supposed to be merely the eyes and ears of the agency chief, making recommendations for the chief’s ultimate decision. Under Morgan II’s principle that “he who decides must hear,” agency chiefs must defer to an ALJ’s determination of witness credibility, because credibility cannot be inferred from a cold record. But an agency chief’s transforming an ALJ into the chief’s “mini-me” would not violate this Morgan principle: Indeed, transforming ALJs into clones of their boss would likely make the boss more deferential to the ALJs’ credibility determinations.
In sum, if one views agencies according to the spirit of the laws creating them, then an agency chief’s packing the agency’s administrative tribunals with people who see the world the way that the chief does is not a bug but a feature. Such court-packing insures that the ALJ is a faithful agent of the true fact-finder, the agency chief. This aspect of agency adjudication might be a bad thing, but it remains officially the way that most agencies operate. If the agency chief’s effort to obtain more faithful agents discredits the administrative state, well, then, maybe the administrative state ought to be discredited.
All of which is to say that criticism of Trump’s EO as an attack on ALJ impartiality strikes me either as hyperbole or as criticism of the foundational premise of the American administrative state, not of Trump’s EO. ALJs have a very limited sort of independence from the agency chief, and eliminating the civil service exam is unlikely to erode that very limited brand of independence.
Does SEC v Lucia give Trump a good reason to issue his EO?
Of course, just because the EO does not compromise bureaucratic independence very much does not mean that it is consistent with the statute authorizing waiver of the civil service exam. That latter question turns on whether the EO is “necessary” for “conditions of good administration.” 5 U.S.C. Section 3302(1). Trump justified the EO by citing SEC v. Lucia, which held that ALJs are “inferior officers” under Article II who must be appointed by “heads of department.” Trump’s EO reasoned that, by limiting agency chiefs’ discretion to hire ALJs, the OPM’s “rule of three” (i.e., choosing ALJs from among the top three exam scorers) might unconstitutionally interfere with agency chiefs’ Article II power of appointment. Is such a position a justified reading of Article II?
Not really. It is conceivable that some judges with a textualist penchant for the unitary executive might regard statutory criteria limiting appointment discretion as an Article II violation. (Judge Laurence Silberman is one such judge who has suggested that statutory requirements of partisan balance raise serious Article II questions.). I personally regard such a position with deep skepticism, especially when applied to competitive examinations, which have been a part of the American system for more than a century. (For a discussion of the history ratifying use of examination scores to limit eligibility of “inferior officers,” see this note). But one can simply sidestep this question of Congress’ power to impose statutory criteria by noting that OPM’s “rule of three” is not a statutory criterion: It is a regulation promulgated by the OPM, an agency under the supervision of the president. Regardless of whether or not Congress can limit an agency chief’s appointing discretion, it would be strange to argue that the president cannot do so: Surely, such presidential/OPM limits are themselves exercises rather than violations of presidential authority. (One might, I suppose, make the super-arcane argument that Congress’ delegation of appointing power to an agency chief preempts the president’s or OPM’s authority to regulate such appointments, but, aside from being a bizarre argument with little support in statutory language, it is not the argument that Trump’s EO made.)
So, Lucia simply does not justify the EO’s excepting ALJs from civil service exams. Does the weakness of this “Lucia-made-me-do-it” argument suggest that the justification is mere pretext masking a hidden desire to politicize the appointment of ALJs? Perhaps, but I would not discount the possibility that Trump’s legal staff were simply being incompetent in reading Lucia. Even if that’s the case, the EO might satisfy the statutory standard: Viewed deferentially, it could be justified as a rational way to hire ALJs with abilities more relevant to their job than those tracked by the civil service exam.
In any case, the civil service examination is not the critical protection for ALJs’ independence: The more important protection is that ALJs can be removed only if the Merit Service Protection Board finds “good cause.” Excepting ALJs from the exam does not affect this removal-related.
But what about sections 3(iii)’s removing certain of ALJs’ removal-related protections?
Such reassurance, however, might be undermined by the EO’s section 3(iii), which amends 5 CFR sec 6.4 to provide that “Except as required by statute, the Civil Service Rules and Regulations shall not apply to removals from positions listed in Schedule[] … E.” (Prior to Trump’s EO, Title 5 CFR Part 213 divided civil service positions excepted from competitive examination into four “schedules” — i.e., lists — based on the reasons for the exception.
“Schedule A,” for instance, covers positions “for which it is impracticable to examine,” such as attorneys, chaplains, interpreters, and various part-time posts. “Schedule E” is the new schedule covering ALJs added by Trump’s EO). The effect of section 3(iii) of the EO, therefore, is to remove ALJs from civil service rules protecting bureaucrats from retaliatory discharge—except where those protections are “required by statute.”
Should one be more concerned about this apparent attack on protection from politicized removal? Not one word of the EO’s “policy” section makes any mention of removals, so this section has an underhanded flavor.
There are, however, two reasons for not being too alarmed at the EO’s foray into the issue of removal.
First, 5 CFR section 6.4 already excludes every other excepted civil servant (say, most attorneys at the Department of Justice) from the Civil Service regulations on removal. Trump’s EO might reasonably have been trying to maintain the symmetry between Schedules A-D and the newly created Schedule E. Second, 5 U.S.C. section 7521 protects ALJs from removal except for good cause established by the Merit Services Protection Board. Because that protection is statutorily guaranteed, the EO could not – and doesn’t purport to – affect it. On my initial admittedly quick glance, therefore, it is difficult to see why removing ALJs from the protection of some civil service rules greatly weakens protection for ALJs. (There certainly has been some weakening: As Ann Joseph O’Connell’s message on Chicago-Kent’s admin law listserv observes, OPM’s guidance on the EO states that some but not all of the civil service rules will continue to apply.)
In theory, OPM might remove all current ALJs on the ground that they were improperly appointed, because their appointment by agency chiefs was unconstitutionally restricted by OPM’s old “rule of three.” As I noted above, I doubt that the Merit Service Protection Board would accept such a ground as good cause for removal. Moreover, the Washington Post reported that James Sherk, the White House special assistant for a domestic policy, asserted that “[t]he order does not affect the employment status of current ALJs, Sherk said, although agencies might want to protect themselves against challenges by having the agency head certify their appointment.”
Even taking into account President Trump’s tendency to change policies without warning, Sherk’s statement and the implausibility of Lucia as grounds for “good cause” removal persuade me that the wholesale “for-cause” removal of ALJs is only a very a remote possibility
In sum, the alarm over this EO is overblown. Trump might really want to make war on the independence of civil servants, but much of that independence is hard-wired into the American administrative state by statute. Reducing the professionalism of ALJs as measured by the civil service exam is likely a bad idea, but it is not likely to transform ALJs into a tool of the Trump administration. (By contrast, Trump’s EO 13,839 seems much more far-reaching and more dangerous to bureaucratic independence, in ways Bernard Bell nicely describes.)