Reflections on Civil-service Reform
Competition is almost uniformly a hallmark of successful systems. It works in sports to identify, incentivize, and reward the best talent. It works in business in the same manner. But government employment — certainly, employment in the U.S. government — is largely insulated from competition.
Since the Pendleton Civil Service Reform Act of 1883 was passed, civil-service employees have been shielded against many modes of discipline or termination that are routinely used to align workers' incentives with the interests of their bosses. Today, of the roughly 2.2 million full-time civilian employees in the U.S. government, all but about 4,500 are covered by civil-service protections, yielding lower turnover and an older set of workers than can be found in the broader economy.
This system was lauded at its inception, and episodically since, as advancing the competence of government employees and the effectiveness of their service. Yet the protections associated with the civil-service laws have also been derided as limiting the electorate's ability to secure changes in policy consistent with voter preferences and the president's ability to deliver those changes in implementing the laws. Increasingly, this complaint is associated with charges of a bureaucracy with one-sided policy preferences standing in the way of legally mandated changes. That view has plainly been central to both Trump administrations' approaches to civil-service employees.
As debate over the current system has increased in recent years, proposed changes to it — exempting more workers from the civil service or, heading in the opposite direction, expanding the scope and durability of civil-service protections — have been put forward in successive presidential administrations. The first set of changes seems headed in the right direction, while the second set highlights reasons to avoid wholesale replacement of government employees exercising significant authority. Decisions on the form and extent of these changes should lie largely in the legislature's hands while the president retains independent authority over some determinations.
POLAR POSITIONS
Two experiences anchor my thoughts about the civil service. First, I grew up seeing the work my father did, and the way he handled his position, as a civil servant. For him, both words were accurate. For roughly half of his 30-year government career, spanning six U.S. presidents, he was the only deputy undersecretary in the federal government who was in the civil service rather than being a political appointee who formally served at the president's (or cabinet member's) will. When the position was created for my father, he insisted that it be a civil-service position; having grown up extremely poor, he placed a high value on financial (here, job) security.
Yet each time political appointees in the department changed, my father offered to resign so that the new administration could have the benefit of officers whose views on policy aligned more fully with those of the political appointees. His offers were routinely rejected. He was kept in his position because his experience and expertise — his knowledge of the programs and processes important to successful navigation of the department's goals — were valuable to each incoming administration.
My father took his job seriously and worked hard under Democrats and Republicans alike. He was the model civil servant, with or without the capital letters that designate government employees who enjoy statutory protections. His example marks one pole of what civil service can be.
My second opportunity to think about this topic — highlighting the opposite pole — came during the Reagan administration. When I was nominated by President Reagan to serve as a U.S. international-trade commissioner, a friend of mine who had served in various White House positions gave me a copy of Yes, Minister, a book relating the travails of a British minister (the equivalent of a U.S. department head).
In this telling, the British civil servants' understanding of job performance was simple: Say yes so the minister will be satisfied that you'll help accomplish his goals, but continue doing whatever you think is best. Or whatever is least taxing for you. No reason to make changes that don't suit permanent employees' preferences. After all, the minister will be gone soon enough and, in all events, can't really do much to harm you. That was the relationship between civil servants and political higher-ups that my friend wanted me to have in mind as I took my turn in the political appointee's role.
My own experience in government wasn't entirely in keeping with this second view of the way civil-service protections work. Then again, it wasn't entirely at odds with it, either. I encountered plenty of smart, thoughtful, and open-minded government officials who were serving before I arrived and would remain in place after I left. Many of them had thought hard about aspects of the agency's mission and how to improve the agency's performance in accomplishing it — indeed, many had thought about those issues a good deal longer and with substantially more insight than I had.
That said, I also encountered plenty of other officials who had fixed views of what the mission of the agency should be and how it should be accomplished. In more than a few cases, those officials were less than enthusiastic about reexamining those views to see how they fit the statutory framework governing the agency's authority. These officials, too, were there for the long haul.
These personal experiences frame the boundaries of sensible policy for selection and control of government officials. They suggest that discussion of civil-service reform is appropriate, because there is room for improvement, for making civil servants behave more like my father and less like the characters in Yes, Minister. I suspect that in some agencies the room for improvement is measured in tons, not ounces. But my experiences also suggest that wholesale change isn't always necessary to align the incentives and behavior of civil servants with reasonable views of public good — that is, to align them as much as is reasonable to expect given the nature and peculiar limitations of government administration.
I start with personal experience because it provides a tether to policy questions, preventing more abstract argumentation from leaving the realm of reality. But personal experience can be misleading, of course. Serious inquiry into what is good or bad about substantial insulation of civil servants from political control requires understanding broader legal and policy considerations.
PRESIDENTIAL POWER AND RESPONSIBILITY
Article II of the U.S. Constitution begins with the declaration that "[t]he executive Power shall be vested in a President of the United States of America." Although some members of the Constitutional Convention and other participants in the ratification debates favored a plural executive, the strong consensus was that a single chief executive — the president of the United States (not of a legislative body nor wholly subservient to a legislative body) — would provide the energy and strength needed to govern a nation as large and diverse as ours.
To be sure, the Constitution recognized that the president would not act alone or independently of the other branches of government. It did not task him with making most of the critical decisions in governance. In domestic affairs, the document clearly gives primacy to Congress. Even where the president acts under legislative instruction, he is not tasked with personally implementing the laws — instead, he is to "take Care that the Laws be faithfully executed." The importance of that commitment of authority may be less than is often assumed, as it was sandwiched between the directions to "receive Ambassadors and other public Ministers" and to "Commission all the Officers of the United States."
The language of Article II appreciates that those who execute the laws are to be subordinate officers in the executive branch of the national government. Given the vesting clause of Article II, the obvious reading of this structure is that these officers, being subordinate to the president, also are subject to his command.
But that does not make completely clear the degree to which officials doing the implementing are to be under the president's control. For instance, Article II also states that the president "may require the Opinion, in writing, of the principal Officer in each of the executive Departments, upon any Subject relating to the Duties of their respective Offices." If the vesting clause makes all the officers performing executive functions fully subject to presidential control, why did the framers think the president needed express authority to secure an opinion from department heads, the president's most immediate subordinates? And why specify that the request is authorized with respect to subjects "relating to the Duties of their respective Offices"? Does this mean the president cannot ask the Treasury secretary about something outside the domain of finance and commerce? More pointedly, doesn't the language of the clause signal an expectation that duties may be committed by law to other executive-branch officials, rather than assigned by the president?
Of course, separate assignment of responsibility to specific officers does not necessarily imply that the assignments — and the officials exercising them — are free from presidential oversight and control. But the practice does raise the question of how much Congress, in making these assignments, can constitutionally circumscribe the processes for carrying them out and the details of personnel selection for the assignments. That is, to what degree can Congress constrain presidential control over personnel decisions?
Answers to that question have been given with respect to some aspects of presidential control over some officials, though not always in ways helpful to the matters at issue here.
In Marbury v. Madison, for instance, Chief Justice John Marshall distinguished between "an officer [who] is removable at the will of the Executive" and officers, like William Marbury, who are not. Marshall states that Marbury was not removable by the president, as "the law creating [his] office gave the officer a right to hold [it] for five years independent of the Executive." Some analysts cite this statement as evidence that Congress can restrict presidential authority over executive officers however it chooses.
But Marbury does not say this in nearly so bold a fashion. Marbury was to receive a commission as a justice of the peace for the District of Columbia, a position comparable to a territorial judge rather than an officer of the United States performing executive functions. That context, together with Marshall's failure to make any clear statement respecting the scope of legislative power to insulate executive officials from presidential control, should give pause to those reading a broad rule into Marbury.
Similarly, the Supreme Court's 1838 decision in Kendall v. United States ex rel. Stokes is often — and again, wrongly — referenced as demonstrating that Congress can insulate administrators' decisions from presidential control but not from judicial review. Kendall involved a firm that provided mail-carriage service under contract with the United States Post Office. Congress had provided for the firm's payment through legislation in the nature of a private bill directing funds to be drawn from the Treasury, paid from the Post Office's account. The refusal of Postmaster General Kendall to credit the contractor for the amount determined to be owed was not taken pursuant to presidential direction, nor would it have been thought to be a matter of executive discretion any more than compliance with a judicial order.
A half-century later, in United States v. Perkins, the Court overturned the discharge (not for cause) of a naval-cadet engineer, a recent graduate of the Naval Academy whose commission would have issued from the secretary of the Navy. The engineer had not been given a posting and, two years after graduating, was informed by a letter from the secretary that, there being no position for which his service was needed, he was being honorably discharged.
In affirming a judgment for reinstatement with back pay, the Court, without adding any discussion of its own, adopted the language and decision of the Court of Claims. After recounting the law prohibiting discharge of military officers during peacetime except pursuant to court martial, the Court of Claims (and the Supreme Court) declared: "[W]hen Congress, by law, vests the appointment of inferior officers in the heads of departments, it may limit and restrict the power of removal as it deems best for the public interest."
Like the cheese in "The Farmer in the Dell," this ipse dixit stands alone. Still, Perkins remains notable as a rare Supreme Court pronouncement on congressional authority to "limit and restrict the power of removal." It has been variously cited, distinguished, criticized, and ignored in later opinions. But it has not been overruled. Even Myers v. United States — Chief Justice William Taft's encomium to presidential power over the executive branch — conceded, citing Perkins, that "Congress, in committing the appointment of such inferior officers to the heads of departments, may prescribe incidental regulations controlling and restricting the latter in the exercise of the power of removal."
Moreover, courts have accepted that, in assigning authority to agencies to implement legislative programs, Congress can craft rules for how those programs will be executed, including the procedures for decision-making, the individuals authorized to make decisions, and the terms for intra-agency review as well as subsequent judicial review. These controls over decision-making are not easily distinguished from power over personnel decisions. The upshot has been toleration of legislated rules for hiring, firing, disciplining, and compensating inferior officers that restrict presidential control of these employees.
Acceptance of the basic framework for civil-service rules governing inferior officers, however, has been coupled with increasingly strict restraints on interference with presidential control over principal officers (in the executive branch, officers whose only real boss is the president, or at least who exercise a degree of final, administratively unreviewable authority).
The counterpoint to this is Humphrey's Executor, which 90 years ago declared that members of the Federal Trade Commission (FTC) could be insulated against removal except for good cause because they perform "quasi-judicial and quasi-legislative" duties — emphasizing that these duties are "neither political nor executive." But the Court acknowledged in Morrison v. Olson that its more recent decisions would not treat FTC commissioners' duties as non-executive, a statement joined by all but Justice Antonin Scalia, whose dissenting opinion supported presidential control of executive officials even more strongly. This concession undermines the weight Humphrey's Executor should carry when considering the limits of permissible personnel rules.
Although, like Perkins, Humphrey's Executor v. United States has not been overruled, the Court's more recent decisions in Free Enterprise Fund v. Public Company Oversight Board, Seila Law v. Consumer Financial Protection Bureau, and United States v. Arthrex demonstrate the justices' strong concern for presidential capacity to control executive officers' conduct in implementing the law. These decisions reflect the Court's reading of constitutional imperatives for maintaining, first, direct presidential control over principal officers performing executive functions and, secondarily, presidential capacity to control other officers' performance through directions to superior officers removable by the president at will. In other words, even if Humphrey's Executor survives in some measure, the Court now plainly reads the vesting clause of Article II as rejecting the Yes, Minister view of executive-branch operation.
Before moving on to the "so what" question — what does the Court's view of presidential control over the executive branch mean in practice? — one other set of cases should be noted. A series of decisions from the 1970s to the 1990s addressing state- and local-government personnel practices asserted that conditioning government employment on political affiliation violates the First Amendment speech and association rights of individuals holding — and even those aspiring to — government positions, at least if the government cannot articulate a convincing justification for its use with respect to specific positions. As with Perkins and Humphrey's Executor, this line of cases technically remains good law, but it is far from clear that these decisions could bear much weight today.
EFFICIENCY AND ORGANIZATION
The Constitution's text and Supreme Court precedents appear to be at odds with substantial constraints on presidential authority over the executive branch. However, neither the Constitution nor judicial precedent subjects executive officers wholly and exclusively to the president's control. What considerations should inform decisions respecting government personnel rules?
One obvious consideration is the cost and efficacy of tools for aligning personnel's interests with broader organizational interests. Every participant in a collaborative venture has an incentive to shirk responsibility in some fashion. Yet teamwork is ubiquitous because we have mechanisms that improve individual incentives, thereby reducing shirking and making team production more efficient. Mechanisms that solve problems like shirking (often referred to generically as "monitoring") are most successful where the enterprise has both an overarching goal and a setting that promotes efficiency.
The most obvious examples come from the business world. Businesses in free-market economies seek to maximize profits, which are returned to the residual owners of the enterprise and, where successful, also tend to correlate with increased rewards to others involved in the enterprise. The monies that the enterprise earns are easily divisible in ways that allow it to reward managers for adopting systems of compensation and production that reduce the combined costs of shirking and monitoring — that is, features that enhance efficiency.
Notably, the norm in competitive business environments is that employees may be terminated at will. Although large enterprises often develop some form of tenure (through contracts that provide greater payouts for termination, restrictions on the basis for termination, and the like), competitive forces also constrain the degree to which this occurs — as shown, for instance, by the movement of facilities from locales with strong legal protections for restrictive union practices to those with weaker protections.
Understanding differences between more and less efficient enterprises requires us to consider differences in both goals and settings. Non-profit enterprises — those without residual owners of the profits the enterprise earns — benefit from efficiency, but different institutional goals and constraints on monitoring their achievement make non-profit enterprises less attentive to some aspects of efficiency that are critical to success in the profit-seeking world.
Consider, for example, competition among higher-education institutions, which turns in substantial measure on perceptions of prestige. Unlike money, prestige isn't easily divisible, and it doesn't correlate well with the excess of income over expenditure. Furthermore, those who donate money to educational institutions often lack ready mechanisms for assessing the effects of their donations. Money generated by educational enterprises — a category that includes some very large and financially successful entities — is often spent on facilities and personnel in greater proportion than in for-profit entities, suggesting a likelihood of greater opportunity for self-dealing.
Government enterprises have two additional characteristics that reduce incentives to function efficiently. The first is that government is less dependent than other enterprises on direct validation of the quality of its performance. After all, governments possess the power to coerce support through taxes and tax-like burdens of various sorts.
The second efficiency-reducing characteristic concerns competition. As Charles Tiebout famously observed, competition of a sort exists among governmental units (particularly at the local level) in that individuals are free to move among localities. This ability to "vote with one's feet," as it were, imposes some competitive constraints on government decision-making. It also allows individuals to communicate their preferences and decisions respecting each government's basket of services and taxes. However, this competition tends to be muted in comparison to the more visible and more immediately effective competition among businesses.
What's more, factors such as the cost of moving from one locale to another impose significant constraints on the magnitude of the "voting with one's feet" effect. And although this effect increases the alignment between residents' preferences and government decisions, that alignment weakens as the size of the government's jurisdiction increases. Thus, although some competition exists even at the national level (especially for large businesses and some high-wealth individuals), it does relatively little to match individuals' preferences with nations' policies, much less increase government efficiency.
Absent strong pressures to conform government's functioning to voters' interests, those who manage and work in government agencies can be expected to pursue goals that are more self-interested. Although there is not a consensus on what these are, three likely candidates have emerged.
First, as William Niskanen postulated, those within the agency have incentives to increase the agency's budget. The theory of the budget-maximizing bureaucracy understands increased funds as expanding options for activities that are consistent with the agency's stated goal, but in ways that also raise the visibility and importance of those overseeing and working at the agency. For example, larger budgets enable greater expenditures on things that make life better for those who work at the agency, including larger offices, improved office locations and amenities, increased support staff, and greater opportunity to travel.
Second, the agency may pursue the ideological interests embraced by those guiding and working at the agency. These interests, again, are likely to be broadly compatible with the stated mission of the agency but push the boundaries of agency authority to (or beyond) its legal limits. Ideologically motivated actions may be observed in any agency, including mainline departments such as the Department of Labor or Commerce or Education as well as stand-alone agencies or bureaus such as the FTC or the Environmental Protection Agency (EPA). However, indulgence of ideological preferences is apt to be most visible and most frequent in agencies with narrower, more focused regulatory mandates, such as the EPA or the FTC.
Although the most obvious examples involve expansion of agency authority, ideologically driven actions can also involve efforts to reduce the scope of agency authority. In addition to serving agency personnel's ideological interests, these actions may also advance agency officials' interests in career advancement by currying favor with specific legislators.
Third, agency personnel may pursue approaches that increase slack. This could include actions that reduce individual officers' workloads, extend time frames for completing tasks, or permit officers to pursue their own initiatives or otherwise advance their personal interests. These actions reduce efficiency and increase the cost of government activity.
Having clear goals and being subject to competition imposes constraints on business enterprises that allow them to give managers substantial freedom over personnel decisions and work assignments. The benefits of such freedom for non-profits, and especially governments, are less clear, at least when viewed in terms of efficiency. As explained above, the absence of efficiency-enhancing incentives in government enterprises means that implementing managerial goals might make the enterprise's operation more or less efficient. How much that matters depends on the political value of efficiency and other goals.
REFORM CONCERNS
Moving from understanding the context to implementing decisions requires identifying problems with the current government personnel system and taking appropriate steps to address them. Important problems with government organization might be termed sclerosis, runaways, whiplash, and perfectionism. Some of these concerns — sclerosis and runaways — weigh in favor of reform. Concerns over whiplash pull the other way. Perfectionism concerns have to do with how decisions are made, including basic personnel decisions and decisions respecting reform.
First, sclerosis. Naturally, the launch of any enterprise generates enthusiasm and energy. But over time, these qualities fade unless kept alive by some type of pressure. As already noted, in the business world, this pressure manifests in the need to beat the competition and to make more money for the owners, who in turn often tie the managers' and employees' compensation to profit-making success.
In government, the sources of enthusiasm and energy tend to be new leadership at the top — almost invariably the product of a change in administration. Meanwhile, the Yes, Minister side of government — civil servants who are reluctant to alter what they do and have been doing — often produces a sort of institutional hardening of the arteries. The unarticulated mantra is: "We know better. We've been at this a long time. Don't tell us we need to change."
Of course, the bureaucrats may be right. They may be acting in ways more consistent with — or using more efficient means of implementing — the law than the directions new leadership brings. The bureaucrats almost certainly know more about some of the problems the agency faces.
But to the extent that statutory authority gives scope for discretion, those appointed by a new administration (linked to recent electoral choices) are better positioned to align the bureaucracy's functioning with larger goals and to inject energy into its work.
The president's control over executive decision-making and personnel is broadly consistent with constitutional design. Faster, more consistent, and even more legally appropriate decisions will often follow from tying worker's interests to presidential energies. This makes the sclerosis concern relevant to personnel rules.
A second focus would be runaways. People drawn to work at an agency are often true believers in a certain conception of the agency's mission. This can produce pressure to work toward a mission the agency is not given unbridled authority to pursue. Hence, for example, those who advocate almost any goal — government regulations in pursuit of lower emissions, wider dispersion of communications technologies, greater safeguards against misleading investors — often slight competing concerns. That will affect their analysis of the agency's legal mandate and whatever discretion has been given to implement the mandate. Reform of government personnel rules should help reduce the risk that agency decision-makers ignore or undervalue concerns prioritized by politically responsive managers.
Another concern to keep in mind is the danger of whiplash, which occurs when it becomes too easy to change institutional course without considering the costs of frequent changes. The interests at stake here are those of individuals and entities who deal with the government, whose businesses or personal lives are affected by what government does and how it does it, and who make plans based on predictions about what rules they'll have to live by.
Adjusting agency behavior to match changed political mandates can be beneficial, but adjusting too often and too fast imposes costs on a range of people who value predictability. This is the other side of sclerosis, its offsetting problem. Just as it's natural for people who have been at a task for a long time to overvalue the benefits (both personal and public) of continuity, it's also natural for those who are new to a task to overvalue the benefits of change. Where the broader public has sharply divided views of what to do with respect to a given set of issues, any change in presidential administration may give impetus to an about-face on matters subject to significant regulatory discretion.
This problem is best addressed by insisting on statutory instructions that give less scope for administrators to exercise expansive power in shaping legal rules. The focus here should be on statutory design, coupled with judicial rules that impose limits on legislative delegations of authority to the executive. But anyone crafting personnel rules should keep this concern in mind as well.
Finally, in looking at potential reforms, it's important not to shoot for perfect decisions on each person or group of people working for the government. Perfection is costly, data-dependent, and requires consideration of too many variables to be likely to produce the outcomes desired. Simple solutions tend to be better. Generally speaking, the right question is not how to achieve the perfect amount of something — here, of control by principal officers or other politically sensitive executive officials. Instead, it is how to achieve the optimal trade-off between different goals, each of which entails different costs.
Professors Mathew McCubbins and Thomas Schwartz captured a version of this thought in the context of congressional oversight of agency operation, contrasting police patrols with fire alarms as means of identifying and responding to problems. Whereas police patrols try to prevent problems or to identify them before they become especially serious, fire alarms signal that trouble is already occurring. In some settings, one approach — ex ante monitoring in hopes of prevention or ex post reacting in hopes of a quick cure — is better than the other.
In the personnel realm, reliance on patronage — or at least some version of appointment and removal authority that can account for likely sympathy to an administration's priorities — is apt to increase responsiveness to the current administration. In contrast, civil-service systems tend to increase expert skills with respect to current procedures, understanding of the array of problems that exist in specific areas (though not necessarily the best way to address them), and some types of analysis. Whatever the right balance is between patronage and civil-service approaches, the current allocation of roughly two-tenths of a percent of government employees to positions not subject to civil-service protections is almost certainly too low.
DUELING PLANS
The first shot in the current battle over potential changes to government personnel rules was President Trump's proposal to create a new Schedule F for specified federal civilian employees, issued as Executive Order 13957 toward the end of his first term. That proposal did not, in itself, create any new positions or shift positions from civil-service protected status to unprotected status. Instead, it directed heads of executive-branch agencies to review positions within each agency and identify positions "of a confidential, policy-determining, policy-making, or policy-advocating character" that "are not normally subject to change as a result of a Presidential transition." For any of those positions, the agency head was then instructed to apply to the director of the Office of Personnel Management (OPM) to classify those positions as part of a newly created Schedule F.
These reclassifications would increase the freedom of agency heads and their immediate subordinates — who tend to be hired through political processes and whose appointments generally terminate with the end of an administration — to use political considerations to hire the next tier of high-ranking policymaking and policy-advising officials (and others who have privileged access to the higher-ranking political appointees' confidential information). At the same time, the people on Schedule F would be safeguarded against a large array of potentially adverse decisions on the basis of considerations that might be deemed improperly discriminatory, among other things.
President Trump's order did not specify the number of positions that would become Schedule F positions. Some critics claimed tens of thousands of jobs would be changed from civil service to political appointments, even if not required as a matter of law. No doubt, some agencies closely tied to the president — the Office of Management and Budget, for example, a relatively small entity within the Executive Office of the President — would want to make a significant proportion of agency positions essentially shadow political ones. But given the difficulty of screening and hiring large numbers of people to relatively high-level government positions, it's likely the numerical impact across the government wouldn't have been nearly so great as feared — or as hoped.
Instead, the effect of the Trump initiative would have been qualitative more than quantitative. In all events, the quick demise of the Schedule F executive order means that, for now, there isn't a definite answer to "how far does it go?"
The Biden administration reversed President Trump's Schedule F plan, which it characterized as "designed to be broad and numerically unlimited" and suggested that it was also intended to permit politically committed officials to be hired without demonstrating competence at the assigned tasks and then to "burrow" into their positions in the agency for the long term. President Biden's executive order and OPM rulemaking criticized the Schedule F plan as contrary to statutes respecting government personnel; less likely to result in having employees with the "appropriate temperament, acumen, impartiality, and sound judgment"; and potentially "discriminatory in application."
The OPM rulemaking went on to announce steps to enhance limitations on political officials' control of federal workers. These included new limitations on reclassifying positions as belonging to a different service classification. In addition, the rulemaking declared that "an employee who is moved involuntarily from the competitive service to a position in the excepted service, or from one excepted service schedule to another excepted service schedule, retains the status and civil service protections the employee had already accrued." In short, the Biden plan, so far as limiting presidential authority over federal workers goes, was "status quo plus."
Shortly after President Trump resumed office as the 47th president, he reinstated his earlier executive order and issued new orders modifying his earlier effort (by, among other things in Executive Order 14171, adopting the name Schedule Policy/Career in place of Schedule F). His new order, to be implemented by a proposed OPM rule, does not reclassify career positions as political positions or require expressions of personal agreement with administration priorities. It does, however, impose clear mandates on occupants of those positions to pursue administration policy under the law. Although Executive Order 14171 does not specify the number of positions to be reclassified, the order on its face focuses on types of positions that fit with special concerns over policy formation and implementation.
The contrasting inclinations of the Trump and Biden approaches say more about the sense each president, and each presidential administration, has had respecting the fit between presidential policy instincts and current officials' compatibility with particular approaches to government. The Biden approach accepted then-present officials as sympathetic and well suited to his administration's priorities — likely to agree with the president's policy preferences and to advance them diligently. It aimed to reinforce the current system and "Trump-proof" it, increasing the difficulty of future efforts to diminish restrictions on presidential control. This approach doesn't suit political administrations that are committed to changing priorities in government so far as legally committed discretionary authority permits. It isn't concerned with sclerosis or runaways, though it does address concerns over whiplash — that is, when someone else is holding the whip.
The Trump approach, by contrast, seeks to sidestep current limitations on presidential control of government officials, reflecting frustration at efforts to change course in administering the laws. This suits the views of many political appointees who try to move agency decisions in a different direction, addressing both sclerosis and runaway concerns. But the Trump approach associated with Schedule F didn't present clear boundaries to political appointees' power. Its limitations were more practical than legal. Although the new Trump executive order and OPM proposed rule alter some of the details and take greater pains to reassure critics that they do not upend the civil-service system, its major constraints remain practical, at least at present. Practical limitations on what changes can be made to federal administrative personnel and personnel practices present checks on any efforts at reform. Furthermore, the time frame within which presidential administrations function also acts as a significant impediment to change even with the most congenial legal framework.
It is worth noting that much of our framework of government is designed to slow and temper change. Current legal rules and administrative implementation of them, however, don't necessarily slow change in ways consistent with constitutionally prescribed mechanisms. That's the point of the concerns sketched above. A path for needed reform can and should fit constitutional priorities as well as presidential preferences while tempering the concerns articulated.
A TIME FOR REFORM
Lord Salisbury — the on-again, off-again U.K. prime minister at the end of the 19th and beginning of the 20th centuries — is said to have declared, in response to a demand for reform: "Change? Why do we need change? Aren't things bad enough as they are?" That may be the right attitude to bring to most efforts to reform American government.
When it comes to the civil service, however, it is hard to say that there is no room for improvement, or that making top-level policymakers and advisors more susceptible to presidential control is generally ill-advised. Consider Justice Scalia's dissent from the Court's 1990 decision in Rutan v. Republican Party of Illinois, the last notable case asserting a constitutional bar on making government employment decisions based on party affiliation. Joined by Chief Justice William Rehnquist, Justice Anthony Kennedy, and (in the major analytical parts of the dissent) Justice Sandra Day O'Connor, Scalia attacked the entire edifice on which the earlier decisions rest. His opening lines give the flavor of his position:
Today the Court establishes the constitutional principle that party membership is not a permissible factor in the dispensation of government jobs, except those jobs for the performance of which party affiliation is an "appropriate requirement." It is hard to say precisely (or even generally) what that exception means, but if there is any category of jobs for whose performance party affiliation is not an appropriate requirement, it is the job of being a judge, where partisanship is not only unneeded but positively undesirable. It is, however, rare that a federal administration of one party will appoint a judge from another party. And it has always been rare. Thus, the new principle that the Court today announces will be enforced by a corps of judges (the Members of this Court included) who overwhelmingly owe their office to its violation. Something must be wrong here, and I suggest it is the Court. [Internal citations omitted.]
Scalia went on to note that as the use of patronage for staffing government at all levels declined, complaints about the inability of government to accomplish its assigned tasks grew.
Scalia's Rutan dissent should give pause not only to those seeking to create a legal right that would require abandoning historically accepted approaches to governance, but also to those who would build policy-based frameworks on similar instincts. The use of patronage to some degree, small or large, has historical pedigree and arguable advantages over a civil-service system (or as an adjunct to it), encouraging political participation and supporting political competition.
In the main, how and how much to embrace civil-service protections are, at best, political choices, not legal imperatives. The most trenchant questions are whether steps that increase presidential control over civilian employees will improve government — especially making it more democratically responsive at the broad policy level — and whether greater executive control of executive-branch decisions will increase or decrease legislative commitments of broad policymaking authority to executive officials.
Recent legal developments don't answer these questions, but developments in two arenas may tilt the practical-political calculus toward change. Judicial decisions are increasingly skeptical of untethered statutory grants of policymaking authority to agencies, and of judicial-review rules that bolster such authority, as recent decisions on the major-questions doctrine and pending cases respecting deference to agency decisions show. The Supreme Court has also offered hints that it may be more willing to hold delegations of broad decisional authority unconstitutional.
These developments, and what they might portend for future actions by Congress as well as the courts, should give solace to reformers worried about expanding presidential power. While there are incentives for Congress to delegate difficult decisions to others, political changes as well as judicial ones may provide counterweights, increasing resistance to broad delegations of discretionary authority. At least from the standpoint of concerns over a too-powerful executive branch under too-strong presidential control, this might be the ideal time for civil-service reform.
