At will? Whose will?

Removing independent agency heads is part of a broader assault on a nonpartisan government

Based on the oral arguments in the Slaughter case, it is clear that the Supreme Court will soon tell Congress that it can no longer decide if the heads of independent agencies are to be nonpartisan experts. The result has been long predicted — after all, SCOTUS has declined to reinstate those agency heads or members of bipartisan bodies that Trump has fired — but it is no less disturbing for it.

The Court will endorse Trump’s practice of politicizing agencies, of valuing loyalty over competence. It also is one more step along the path of unitary executive theory, the radical idea that the Founders actually wanted a King and not a President. In the hearing on the Slaughter case, Justice Sotomayor noted that “Neither King, not parliament, not prime ministers ever had an unqualified removal power” but that is what unitary executive would grant.

Unitary executive was created by Republican lawyers — some of who now sit on the Supreme Court — in order to empower Republican Presidents. Cass Sunstein witnessed its creation as a civil servant in the Reagan administration:

Let us instead marvel over this: the emergence, starting in the 1980s, of a view that was held with supreme confidence by a group of (mostly) young lawyers, but that was patently inconsistent with settled law — and that is now on the verge of being accepted by the Supreme Court of the United States.

And so, the judiciary will kill key parts of the working model of American governance, built up over time to adapt to the needs of the country, with no functional alternative beyond handing unprecedented power to an eager authoritarian.

Where will this path end? Not with agency heads. In the Slaughter arguments, Justice Kagan pointed to the slippery slope nature of the administration’s claims: the logic that allows Trump to remove independent agency heads applies just as well to removing civil servants.

SCOTUS has emboldened the Trump administration, who are already planning their next move, revising its Schedule F plan to make more federal employees removable for non-performance reasons. They previously pushed such changes on performance grounds. Now, it argues that Congressional constraints on firing civil servants is itself unconstitutional because of the President’s Article II power. This is emblematic of a theory of governing that says “we are doing this because we can” rather than “we are doing this because it better serves the citizens.”

Here are some of the reasons Trump is firing civil servants - it is not about performance

If you are a employee of the federal government you can be fired for the following reasons:

• Being a relative of a disfavored former employee: Maureen Comey, a respected DOJ prosecutor, was fired shortly before Trump demanded her father be prosecuted.

• Refusing to illegally fire people: Kash Patel is being sued for firing senior FBI agents who asked him for adequate reason to fire their fellow agents.

• Investigating corruption and abuses of power: For example, ethics officers at Fannie Mae were fired after they raised questions about how Bill Pulte obtained documents as part of his effort to prosecute Trump’s political enemies.

• Investigating an insurrection or allegations of corruption by the President.

• Providing an intelligence analysis that contradicts the President’s claims.

• Being a trans person.

• An online monitor of the “deep state” accuses you of being disloyal to Trumpism. Such accusations are often erroneous: a prosecutor falsely accused of not willing to be part of the Comey investigation was fired. An FBI agent was fired after being falsely accused of participating in Mar-A-Lago investigations.

• Sharing accurate information to a judge: A DOJ lawyer was fired when he acknowledged that the administration had erred when it deported an immigrant.

• Not recommending that a prominent celebrity supporter of the President have his gun rights restored.

• Signing a public statement that existing policies are undermining the ability of agencies to complete their tasks.

• No reason at all. The President merely asserts that he has an Article II power to fire you and decides he wants someone assumed to be more loyal in your place.

To be clear, these firings are only the ones we know about, publicly reported on or litigated in court. It is entirely reasonable to assume they are the tip of the iceberg. They are also generally illegal. Under civil service laws, the type of employees mentioned above typically have “for cause” protections, i.e. they can only be removed for some documented performance problem. They also cannot be removed for “partisan political purposes.” They are also entitled to a reasonable appeals process. For others in the national security space it is slightly more complicated, but historically employees were not fired for the types of reasons listed above. But the gulf between what is legal and normal, and what is the practice of the Trump administration, grows ever wider.

The purging of employees is important for three reasons.

First, it will worsen state capacity, in both the short term as acquiescent hacks replace principled public servants, and in the long run as talented people decide that a workplace where they can be fired by ideologues is not for them.

Second, it gives lie to the claim that Trump officials made to justify politicization: they denounced a deep state engaged in bureaucratic resistance to legitimate presidential power. Such claims were based on one-sided anecdotes. It always seemed more likely that Trump officials were either blaming civil servants for not being willing to violate the law, or for their own managerial failures. This explanation seems much more plausible when we see why people are being fired.

Third, and most relevant for future debates about the civil service, there is a push to move to a completely at-will system, while also removing union protections from employees. And this is what I want to focus on.

At-will is the opposite of for-cause, meaning that employees can be fired for any reason. You can read the best version of the argument for public sector at-will employment in this discussion between Judge Glock and Santi Ruiz. The general claim as I understand it is that:

• managerial flexibility in hiring, firing and payments generally leads to better personnel outcomes

• some state governments experimented with moving to at-will hiring, while sometimes weakening unions, and the best evidence we have is that it worked reasonably well

On the first part of the claim, difficulties in hiring has been a long-standing concern. There is broad consensus that public employee hiring is too slow and does not generate outcomes.

There is less consensus around firing. It is absolutely clear we need a modernization of the civil service system, and the terms of that debate are being set now. I’ve been parts of conversations with people who want to make government work, who are frustrated with the status quo and hungry for change. A move to “at-will” employment may sound appealing, including for progressives who think that the bureaucracy has been holding back their agenda of change. But I think it will be a disaster.

The evidence on at-will systems in state governments is weak

On the second part of the claim — that at-will experimentation provides positive outcomes — I am skeptical. Some states did indeed do a bunch of experimentation and there has not been widely-reported evidence of politicization and corruption we associate with a spoils system.

But most of the evidence on state reform relies on subjective assessments from employees or HR Directors. The evidence mostly suggests things have not gotten worse, rather than performance has gotten better, and even that evidence is mixed — plenty of self-reports are critical of the new systems.

There are a couple of reasons for the weakness of the evidence. One is that states did not invest resources into doing serious evaluations of the changes they made. When I was a professor at Wisconsin, I called for the state government to track the effects of some of its personnel changes, like moving to resume-based hiring. But the state was not interested because it was already convinced that such a system was better. Did it make things better or worse? We don’t really know.

Wisconsin Governor Scott Walker did a lot, including blowing up collective bargaining for public employees. We can debate Walker’s motivation, but he certainly kneecapped public sector unions in a way that benefited Republicans. We can look at some evidence about effects. Walker’s changes improved the number and pay of quality teachers and improved student achievement, but also led to more gender discrimination in salary, which is part of what standardized pay systems and unions seek to prevent.

There is also evidence that the Walker administration engaged in political retribution toward public employees. Public employees who signed a petition to recall Walker saw a salary decline in the following years — unless they had civil service protections. What this tells us is both that Governors are capable of using their power to punish public employees for political reasons, and that civil service protections limit that sort of abuse. Notably, this pattern was not something that was detected by media covering state government, and I think its relevant to note that civil service changes occurred at a time period when media coverage of state government became weaker.

A similar example comes from Florida. An analysis of civil service rollbacks there using state administrative data showed that it did not increase turnover in aggregate. This would be, it seems, another example of the dog that didn’t bark: at-will did not affect personnel outcomes. But the same analysis also shows that affected employees in the Department of Education — which at the time was central to Jeb Bush’s political agenda — were about five times more likely to leave shortly before Bush took office. In other words, they engaged in anticipatory resignations in the shadow of an at-will system, in the same way a lot of federal employees have done in the last year.

The results here mirror patterns we see elsewhere. For example, a study of Slovakia found that civil engage in anticipatory exits when they see a new regime that is likely to challenge democratic norms or ethical standards, with more principled staff being the first to leave. The results speak to the point that the effects of at-will might not be observable using aggregate measures, but localized to specific functions, and should incorporate resignations and not just firings, and can extend to before a new executive takes power.

The political economy of state civil service systems is different from the federal government in some important ways

There strongest evidence that public sector unions are associated with poor performance comes in the domain of policing. For example, credible causal studies show that changes in collective bargaining for police is associated with increase in deaths of non-white civilians and violent misconduct.

However, policing and police unions tend to be exempted from personnel changes. Why is that? I think it is largely because they are viewed as more ideologically aligned with Republicans, which again underlines the point that politicians pushing for at-will public employment are significantly motivated by the goal of weakening perceived political opponents.

There are some crucial differences between states and the federal level when thinking about at-will employment. For one thing, there is less switching between parties at the state level. Florida, Texas, Oklahoma and Georgia are exemplars of civil service changes.¹ These are also red states with little partisan executive branch turnover.

State employees have also not been subject to the same type of deep state conspiracism that Trump has directed at their federal counterparts. So I buy the idea that red state governors have not used at-will power in the way that Trump has. It is entirely plausible that Trump’s beliefs motivations and incentives about state power really are different from even MAGA Republican Governors. But that should make us cautious about making inferences about how state civil service changes would work at the federal level.

We already have a de facto at-will system under the Trump, and it is being used in partisan and corrupt ways

But here is the thing: We already have a de facto at-will system operating in the federal government. We don’t need to guess what it will look like. We are watching it play out. Any discussion of what at-will looks like that does not explicitly address what the Trump administration is doing to the career civil service is ignoring the elephant in the room.

How is this going?

• As described at the top of the article, Trump’s de-facto at-will system has resulted in the firings for reasons clearly disconnected from performance, and in ways that will facilitate corruption. At the same time, a baseline lack of professionalism seems to be a qualifying credential to be hires as a political appointee who oversees the career staff.

• There has been unprecedented use of administrative leave, de-facto demotions and involuntary transfers, often for employees who raise concerns about the legality of actions of DOGE or Trump. For example, staff who investigated Trump’s efforts to politicize the National Weather Service were put on leave, as who are working in areas the administration dislikes, like civil rights. The use of these tools creates a conduct of constructive dismissal, where employees feel compelled to leave government.

• We also are seeing a massive exodus of employees. There has been no effort to ensure that better performing employees are retained. Many are quitting because of the unethical or illegal actions they see around them. For example, much of the IRS C-suite left because of DOGE’s efforts to violate privacy laws. A FEMA Chief Counsel resigned when employees were removed for social media comments about Charlie Kirk.

• The cuts are occurring to a greater degree in agencies viewed by Trump as Democratic agencies.

Trump is not just firing people as if the civil service does not exist, he has eliminated the safeguards to protect against politicization.

• He has proposed policies that will eliminate collective bargaining for about two-thirds of public sector union employees (though a bipartisan Congressional majority may reverse this).

• The Office of Special Counsel and the Merit Systems Protection Board have been stripped of officials who might disagree with Trump, meaning federal employees have no reason to expect a fair hearing if they face partisan personnel actions. A judge dissenting with the clear violation of precedent for these removals noted: “We may soon be living in a world in which every hiring decision and action by any government agency will be influenced by politics, with little regard for subject-matter expertise, the public good, and merit-based decision-making.”

• The smashing of guardrails also includes the removal of most Inspectors General as well as judge advocates general in the military tasked with monitoring corruption and abuses of federal employees.

Consistent with the pattern in state governments, the federal government largely does not want to know about the results of its great experiment. It has cancelled the annual employee survey that would reveal a demoralized workforce. And given what we know about state capacity and waste, it is entirely reasonable to assume that the politicization happening at the Trump administration will contribute to weaker oversight and greater corruption.

In short, American government is currently going through a period of extraordinary personnel abuses. It is hard to have any serious discussion of at-will employment without acknowledging that. We will be unpacking the effects of these abuses for years. Discussing at-will employment without Trump is Hamlet without the prince.

Those coming from a private sector perspective might ask: “So what? Thats how we do it?” But it’s not, of course. The incentives, structures and demands of public and private employment are different for a myriad of reasons (e.g., public organizations can’t go out of business and so cannot be allowed to fail, must serve multiple masters, must implement laws created by others, cannot play favorites, and are subject to more intense legal and media scrutiny).

Put another way, the risks of a partisan public personnel system is not just poor public services, but that it worsens our democracy. Many of the points of friction between Trump and federal employees are about democratic values: the rule of law, how Congressional statute is to be interpreted, avoiding abuses of power, observing oyblic demands for transparency. Again and again, the logic for Trump’s personnel actions is the logic of a personalist regime: loyalty to the leader above all else, removing individuals or downgrading agencies that are disfavored.

In the Slaughter arguments, Justice Kavanaugh said: “I think broad delegations to unaccountable independent agencies raise enormous constitutional and real-world problems for individual liberty.”

It is hard to take such concerns seriously for three reasons. First, authoritarian leaders politicizing the bureaucracy are the ones most clearly engaged in democratic backsliding in the world today. Second, Trump is following exactly this blueprint, and Kavanaugh’s solution is to hand him unprecedented powers. Third, unelected bureaucrats are actually fairly accountable in a large variety of ways, even as the Trump administration is trying to minimize standard means of accountability.

If the President can simply fire people tasked with upholding the law and ethical safeguards, they have little incentive to do their job. Whatever the concerns people might have about public sector unions, at a time when many institutions have caved to Trump, they have offered one form of collective action to force courts to address the legal questions of where our President’s ever-increasing powers end.

The fundamental problem: adding at-will to an already politicized system is a disaster

A highly politicized system and an at-will system are inconsistent with one another if you want honest and competent government. Just for this particular discussion, lets assume I mean “politicized system” = lots of political appointees running government agencies. The American federal government is a highly politicized system. It has about 4,000 employees, and they occupy the top levels of public organizations, even when they lack experience or competence.

Lets imagine four combinations of politicization and at-will employment:

• Few political appointees, civil service protections, . This is most governments, including peers like the United Kingdom, and is generally associated with better outcomes, though new political leaders often complain about lack of responsiveness.

• Few political appointees, at-will employment. For example, civil servants in Denmark have limited personnel protections. This works because there is not an army of political appointees running national agencies, and so we can generally assume that personnel decisions are not shaped by political considerations. The type of firings described at the top of this essay would be regarded as a national scandal. They enjoy the benefits of personnel flexibility because of the absence of politicization.

• Lots of appointees, civil service protections. Or you can have a politicized system combined with additional job protections for employees. That is, basically, what we have had in the US. What we are seeing unfold right now is a reminder why we have such rules, even as the rules are being ignored.

• Lots of appointees, at-will employment. This is what advocates of at-will employment want for the federal government. This will generally result in greater flexibility for politicians to do what they want. When they want to run a functional government, they will mostly be willing to favor competence and expertise over political loyalty. If they believe in deep state conspiracy theories, or that employees are their political opponent, or value personalist loyalty above all else, or want to minimize oversight for illegal or corrupt behavior, they will use that power to attack competence and principled state actors in a way that will undermine state capacity. This is what we are seeing that play out right now at the federal level.

The American state has a taste for politicization and I don’t see that taste changing. We are not getting to Denmark, if getting to Denmark means reducing the number of political appointees. Indeed, while the size of the federal workforce has stayed relatively flat or recently declined, we have more and more appointees over time. And the Trump administration has established new hiring and firing policies that demand that these appointees be central to key personnel decisions, and are anchoring annual employee evaluations around loyalty to Trump’s policies.

The effects of this experiment will be hard to repair. Commenting on the potential demise of the civil service system, Max Stier, the President of the Partnership for Public Service says:

You can’t come back from the world that they’re designing for us in any really easy way, because the only way you do it is through change in the personnel of the court or a constitutional amendment. There isn’t a good reason that I can see for doing it, and the implications are huge in terms of the kind of government we can have in a society we will have, and the services the public will receive.

Justice Jackson echoed these themes in the Slaughter arguments:

So having a president come in and fire all the scientists, and the doctors, and the economists and the PhDs, and replacing them with loyalists and people who don’t know anything is actually not in the best interest of the citizens of the United States.

Where do we go from here?

So, here are my takeaway points about the at-will debate:

• The evidence base on state government experimentation is simply not strong enough to sustain strong claims. The broader body of evidence that politicization worsens capacity and performance is more persuasive.

• There is growing elite agreement that personnel systems need to be modernized, but not about how.

• If more states are going to experiment with civil service changes, they should figure out how to evaluate such changes to generate credible evidence. Both supporters and critics of civil service systems should agree: lets see what the evidence says!

• You can have an at-will personnel system, or a system that relies on political appointees, but you can’t have both. The US has a strong preference to let political appointees run agencies. As long as that is true we should assume that giving political actors more personnel authority means they will use it in political ways.

• The Trump administration is showing us exactly how it — and likely other — administrations will use unfettered personnel power: it is a classic case of politicization weakening state capacity and performance, and inviting corruption.

1. While Zell Miller initiated at will employment in Georgia as a Democrat, he was ideologically close enough to Republicans that he endorsed George W. Bush and the Republican National Convention in 2004.