Social scientists use the term “social desirability bias” to reflect the point that even in anonymous surveys, people are reluctant to express views they know are subject to societal disapproval. And some people might not be fully conscious of their own discriminatory views.

And so researchers came up with an ingenious way to get at people’s willingness to engage in discriminatory behavior: the audit study. The audit study is an experiment, where people are usually asked to respond to information about two people who are identical except in terms of a key characteristic, such as gender or race.

The classic version of the audit study involves sending two sets of fictional resumes out to employers, where the resumes are identical except for one factor: the name. One name might be a White-sounding name. Another might be a name that sounds more likely to belong to a minority: someone who is Black, or Hispanic, or Middle-Eastern. People will infer the race of the sender by the name. If the responses to the two resumes are the same, no discrimination. But if one name gets systematically less responses, it is reasonable to infer discriminatory behavior.

The research literature based on audit studies has been impressive in offering hard-to-dispute causal evidence of systemic discrimination.1 For example, key findings include:

• White job applicants receive systematically more callbacks than equally qualified Black or Latino job candidates, a pattern that has not improved over time.

• In housing, discrimination against non-white renters and buyers has remained significant and relatively stable over time.

My own minor contribution came to this literature a few years ago, when an audit study I worked on with Asmus Olsen and Jonas Høgh Kyhse-Andersen, when we found that Muslim students in Denmark had a harder time getting school officials to allow them to join their school relative to students with classically Danish names.

A real life audit study

Audit studies are attractive because they provide clean causal evidence of hard-to-observe behavior. Researchers can control for all of the potential reasons why people might disfavor one group other than bias. For example, people might assume a minority group has less education, but you can ensure the fictional resumes feature the exact same level of education. In real life, of course, it is not so easy to control such factors.

But what if you could find a real life example where the minority candidate was given every possible advantage. Was not just the same, but clearly better. Would we still see signs of discrimination?

The Michigan primaries threw up such as case. In one Republican primary for a seat in the 8th District, Tom Smith faced Amir Hassan. Neither are incumbents in the seat.

Who do you think won? Of course, you are thinking Smith, right?

But what if I were to tell you that Hassan had one extraordinary advantage that Smith lacked: the coveted Trump endorsement, in a district that Trump won in 2024.

Trump even rallied with him.

And Hassan was chosen by the GOP precisely because they thought he could win in this district. Speaker Mike Johnson endorsed him. And Hassan was a Navy veteran.

Now, who do you think won?

If you still say Smith…you know Republican primary voters well. Smith won with 50% of the vote, while Hassan received 33%.

Share

Maybe there is stuff we are missing here that made Smith a better candidate? Maybe he ran a better race, was more organized, or had more money on hand? No, no, and no. And here is where things get really weird. Smith had dropped out of the race in July. His Facebook campaign page had 95 followers, and he had raised less than $2000 according to All In with Chris Hayes. Hassan reported spending $622,841.

This is the note Smith sent to the Federal Election Commission, testifying to his inability to raise funds, and, uh, to completing election forms:

Why did Tom Smith, a candidate with no money, no experience, no organization AND WHO HAD DROPPED OUT OF THE RACE defeat Trump-endorsed Amir Hassan? There is no obvious metric by which Smith was the better candidate. Save one: his name. And even Republicans acknowledged it. According to the Detroit News:

State Rep. Tim Kelly, R-Saginaw, said he could only speculate why Hassan lost but added that voters were shying away from what he described as Arabic-sounding names. Other Republicans echoed Kelly's analysis, saying the outcome was sad.

Hassan Nehme, a Muslim Republican candidate, blamed the Christian nationalist movement in the district, saying: “They’re scared of the Arab Americans taking over in the area.”

Hassan faced a wave of bigotry accusing him of being an immigrant (untrue) and Muslim. Hassan was once Muslim but is now an avowed Christian who has become fairly critical of his former faith. His campaign bio states:

He recently reaffirmed his commitment to Christ, stating “Jesus Christ is his Lord and Savior.” Having served at Guantanamo Bay, Amir has zero tolerance for extremists who pervert Islam into something evil. Amir believes Israel is our strongest ally in the Middle East.

After Hassan’s defeat, Trump confidante Laura Loomer posted that:

“Amir Hassan” claims he’s not a Muslim, but it doesn’t matter. His name sounds like he’s fresh off the boat and MAGA voters reject the Islamic infiltration of our country and party. This should be an eye opener to Trump’s advisors that we don’t want Muslims, Arabs or anyone who may even sound Arab or Islamic serving in office. We are done with the Islamic takeover of our country, but they won’t pass this message along to President Trump…Dude should have changed his name. If my name was Amir Hassan and I wasn’t a Muslim, I’d be changing my name as soon as I turned 18.
I will never vote for anyone with an Islamic name. Neither will most of President Trump’s voters. [Emphasis added].

Hassan’s actual identity does not matter. His name sounds like he is Muslim, and that is enough for Republican voters, according to Republicans.

One alternative explanation is that Hassan is Black, and so he faced anti-Black rather than anti-Muslim racism. It is possible the combination of his name and race were toxic, but this is also a year when the Republican nominee for Governor in Michigan, John James, is Black. James comfortably won the counties that Hassan lost.

There was really nothing Amir Hassan could have done to win this Republican primary that he did not do. He had every advantage and an opponent who had so little confidence in his campaign that he exited the race. None of it mattered. Hassan’s name did not match the imagined bio it created, but that imagined bio of a Muslim candidate was enough to sink him.

This astonishingly open bigotry has become such a feature of the MAGA base that the general Republican response seems to be a shrug of the shoulders, and a disinclination to include Muslims in the party. For example, Muslims who attended the Texas Republican Party were told they were not welcome as the Party embarked on a “Don’t Sharia Our Texas” campaign based on the lie that the Lone Star state is on the verge of becoming Iran because of some unseen Muslim influence.

Which brings us to another name in Michigan: Abdul El-Sayed. The Democratic nominee for Senate has already become a national figure for Republicans, who hope to persuade all Michigan voters to treat him they same way that Hassan was treated. Reduced only to a Muslim name. Tied to every Muslim stereotype that Republicans fear. His record as a physician working on public health buried from public view. A debate on actual policies avoided at all costs.

El-Sayed is Muslim, and has been sharply critical of support for Israel’s war. If people want to argue about that issue, fine. But instead, his opponent Mike Rogers claimed that El-Sayed believes that “America deserved 9/11.” He also said El-Sayed is a “RADICAL ISLAMIST” who would “unleash terror.”

None of this is true of course. El-Sayed long ago denounced Muslim terrorism behind 9/11. But to be a Muslim man in America means being required to complete such denunciations. And it means that such denunciations are never enough. You are permanently under suspicion regardless of how you have lived your life.

Rogers knows his claims are untrue, but judges the penalty for slandering El-Sayed will be less than the benefits. Will he be proven right? There are a grand total of four Muslim members of Congress. There has never been a Muslim Senator. When you see shameless demagoguery like this, you start to understand why.

And that is just the tip of the iceberg. Paul Waldman at Public Notice provides plenty more examples, including the following:

The level of anti-Muslim bias seems both stronger and more open on the Republican side than even in the immediate post 9-11 era, when George W. Bush argued that the true nature of Islam is peace, not terrorism. There is no effort to call to the better angels of our nature during the time of MAGA.

And so, the state of Michigan now faces another test, which is how much the name of El-Sayed, and the bio that Republicans are trying to attach to it, will replace the actual candidate. Expect it to be a very ugly few months.

We had reason to believe this was happening for a while. Trump has disproportionately cut what he has described as “Democrat agencies” and red state or Republican aligned contractors were less exposed to DOGE cuts.

The White House was very upfront that it was cancelling infrastructure projects in blue states during the shutdown to create leverage. Now we have more evidence that this amounts to a policy. Lets take a few examples.

• The Trump administration cancelled more than $7.5 billion in Biden-era clean energy grants, saying they were wasteful. But in court documents, the Department of Energy admitted that cancelations “was based solely on the political identity of the grant recipient’s state, i.e., whether the recipient’s location and/or place of performance was in a Blue State or a non-Blue State.”

• It is now three times harder for blue states to get disaster aid than for red states. Trump has personally intervened to overturn FEMA staff recommendations, blocking aid to states like New Jersey, New York, Massachusetts and Rhode Island. Trump also just denied federal assistance for raging wildfires in Oregon. The provision of aid to red or purple states at the same time came with political endorsements for Republicans. For example, Trump announced federal aid to Wisconsin in the context of his endorsement of Tom Tiffany, the Republican candidate for Governor even though the actual Governor is a Dem.

• After Trump fumed about the imprisonment of Tina Peters, a Trump-supporting election denier who was convicted of election interference in Colorado, the White House told agencies to come up with ways to withhold resources from the state. The Department of Transportation and Department of Energy canceled hundreds of millions of dollars in federal grants to the state. Peters would later be pardoned and released by the Democratic Governor of the state.

Is this just business as usual in politics? No. Approval rates of federal disaster aid requests were about the same regardless of the state and did not vary much with the party of the President. This graph very powerfully conveys the scale of the partisan abuses now.

More recent news about FEMA underlined the scale of the mismanagement of the agency. We already know that Trump downsizing did not include any effort to assess the quality of the employees who were walking out the door. A new GAO report found out that the Trump administration did not try to assess workforce needs for the agency. It was fire first, figure it out after.

The GAO reported that “these separations have resulted in a loss of institutional knowledge and experienced personnel, according to FEMA officials, and have exacerbated longstanding workforce challenges.”

It seems like workforce planning is the sort of basic management thing the head of FEMA and its parent agency the Department of Homeland Security should be doing. But Trump’s pick for FEMA head at the time was out of his depth, with no experience in emergency management, and he did little to block the DOGE purges. He lasted for less than six months. FEMA does not currently have an appointed agency head.

From what we know of Kristi Noem, the head of DHS at the time, she was focused on micromanaging some aspects of government operations. Not workforce planning, but — you guessed it — the flow of grants and contracts. Any contract worth more than $100,000 had to be personally signed by her. And some of that spending was pretty questionable, such as a $200 million ad campaign prominently featuring Noem. There was no competitive bidding process and Noem directed money to Republican firms with long-term personal and professional ties to her. (The fact that Noem had previously led a “Meth: We’re On It” campaign did not seem to stop anyone from questioning her public relations skills).

DHS also bought 11 empty warehouses that it said would serve as detention centers, overpaid for them, and has since sold seven of them. Apparently Noem told officials to buy the warehouses within 30 days, leading to a rushed acquisition process. After the DHS spent nearly $500 million via no-bid contracts on 10 jets, officially for deportations, the planes are mostly unused, or reserved to ferry Trump officials around. So more likely to be incompetence than corruption in this case.

Noem’s alleged boyfriend/aide Corey Lewandowski was also aggressively shaking down DHS contractors while nominally serving in government as an unpaid appointee. Contractors offered him a retainer, but he declined, wanting a bigger taste. Then he told DHS officials to stop providing them contracts, and their contract success did indeed decline in the following months. Marketing firms were told they were guaranteed a contract, but they just needed to hire a consulting firm who could serve as a pass-through to paying Lewandowski. The consulting firm would get half or more of the contract.

Now, you might know that Noem is no longer Secretary of DHS and Lewandowski is out of government. So maybe accountability is working. But let me make a couple of points. The accountability was driven by external reporting rather than internal controls. The White House had heard the complaints about Lewandowski, but did nothing. The Senate eventually raked Noem over the coals in hearings, but she still has a government job as “special envoy for the Shield of the Americas”, whatever that is.

Things are looking worse for Lewandowski, as an Inspector General investigation confirmed the wrongdoing and may refer him to the Justice Department.

Share

And here I want to spend some time talking about Inspectors General. These are the internal government watchdogs who should be the first on the scene to root out corruption. But they have been neutered.

Trump has made clear he will fire Inspectors General who become too active. He has also left leadership positions for these offices vacant and cut their budget. His nominees to the positions are chosen for their lack of independence. Most recently he nominated a Trump appointee and Project 2025 author to serve as Inspector General for the Department of the Interior.

A new report by the Partnership for Public Service finds that Inspectors General offices have been downsized by about 19%. Not surprisingly then, they are providing 25% fewer investigations despite there being much more to investigate. Reports are also taking longer to be published.

What about Congressional oversight? This exists, but it is important to remember that a) this is much less likely to prevent bad things from happening, as opposed to investigate the bad thing after it happened and b) the administration has led a war against Congressional oversight, often refusing to co-operate. For example, OMB head Russ Vought saying that the congressional Government Accountability Office should not exist.

Judicial accountability is possible. For example, the EPA was just ordered to release $20 billion in canceled grants around green energy by a federal appeals court. But appeals are a slow and unwieldy. It requires having the resources to go against the Department of Justice. It can be circuitous. For example, litigants can be redirected to the more limited Courts of Federal Claims. As with Congressional oversight, it is an ex post rather than ex ante form of accountability, which does little to address corruption before it happens, and which is not the subject of a lawsuit. And with the Supreme Court’s ever-expanding notion of executive authority, the end outcome is uncertain. In some cases it might unwind grossly partisan acts, but it will never punish the perpetrators. The incentive structure is that they will win more often than they will lose, so why not keep doing it.

My guess is that there is enormous graft going on in federal contracting right now, which will not be discovered until later. These processes are less visible than grants in some ways. This graft is more likely in the resource-rich national security space, or where there are no-bid contracts because of claims of an emergency (such as, uh, The Reflecting Pool). Contractors aligned with the administration know that speed and loyalty, rather than value for money, is what drives the administration.

In the meantime, we are left with questionable stories where the numbers simply do not add up.

• For example, a Texas law firm received a $150 million one year grant to represent 24,000 children in immigration court. It has 26 employees. And does not specialize in immigration. The firm was co-founded by two Trump appointees. [Update 8/11/2026: the law firm in question indicated they had been approached by the Trump administration, and were only planning to apply for a small grant, and decided to then withdraw from consideration after the reporting].

• Another $244 million went to a no-bid contract to “Our Rescue” to represent migrant kids, a seven-time increase in federal money for an organization led by a former ICE official.

• ICE just paid $2.2 billion to purchase four prisons from a contractor, who will then still run the prisons for them. The whole point of contractors should be that you avoid making long-term capital investments that tie you down.

Maybe these are all perfectly good business decisions. Or maybe not. In the current environment, we have little reason to feel confident that our government is on the level.

While there is often tendency to associate accountability processes with delay and red tape, the Trump administration reminds us that they are there to restrict corruption. Without accountability, people in government will often take a note out of the Tammany Hall model of government by graft, represented in George W. Plunkitt’s classic line: “I seen my opportunities and I took them.”

How many people working in and around the Trump administration right now seen their opportunities and are taking them?

The spectacular irony of course, is that this is an administration that claims to be aggressively reducing fraud.

They have invoked accountability because they know its a public value that people care about. But real accountability is not partisan. Not only is the Trump administration not pursuing cases of corruption closest to home, it is using fraud as one more bad-faith justification to do what it wants to do anyway: cut off funds to blue states.

For example, the Department of Health and Human Services has blocked payments of over $2.76 billion to California and Minnesota. This sort of across-the-board collective punishment of a state weakens critical safety net programs that people rely on. And contrary to what Trump might think, it hurts his supporters. California had almost as many Trump voters as Florida and Texas in 2024.

The Trump administration does not really believe that fraud only occurs in Democratic states. It has enough evidence of corruption on its own doorstep to think Republicans are above such things. What it does believe is that your tax dollars are their tax dollars, available to reward friends and allies, and to be withheld from political enemies, which includes ordinary Americans living in blue states.

Never mind that citizens in blue states already contributed more federal revenue and received less federal funding. Now their residency in a blue state has reclassified them as second-class citizens. The Trump administration has no vision of the collective, of a United States of America, just a Trump States of America.

If you are interested in a deep dive on the Slaughter decision, check out this group of esteemed legal minds (plus me) to unpack what it means from the future. The views ranged quite a bit!

RELATED

A few recent data points suggest that MAGA’s problems extend beyond Trump’s sagging poll numbers.

The Base

The MAGA base appears to be shrinking. This does not mean that people are rejecting the movement overall, but that they are becoming disillusioned with its leader and his lack of responsiveness to their concerns. Thomas Shull at the UnPopulist wrote about this trend.

The key takeaway is the proportion of people who see themselves as Trump loyalists is declining by about a third. Across a variety of data points, fewer Trump voters still believe he is on their side

A caution here is that Trump voters are not necessarily moderating their views. They still dislike elites, but are now more likely to put Trump in that category. One intriguing exception is that bureaucrats have become slightly less hated by Trump voters. Maybe a year of watching public servants be fired for no reason has burst the myth of the all powerful deep state, at least for some in MAGA.

The Media

Stuart Thompson at the New York Times offered a fantastic breakdown of media attention to Trump supporting influencers, whose reach is declining.

The decline in support comes as much of this media becomes more fractious because of personal beefs, disagreements over policy, especially related to Israel and Iran, and their tolerance for the very far-right views. But the broader picture is that it has become harder for these influencers to maintain an audience.

Why?

The obvious answer is that they are outrage merchants, and it is hard to sustain outrage when your allies are running everything. On the right, the Fauci hearings were a media frenzy, a Superbowl of nostalgia for the glorious era where they could blame others for the state of the world. Media Matters reported that Fauci was:

• Mentioned 687 times on Fox News and 740 times on Newsmax in a four day period.

• Compared to Hitler and Epstein, and declared a mass murderer by multiple prominent right wing influencers who called for his imprisonment.

Increasingly, the right does not have Fauci or Biden-level targets to aim at. Its business model is a mix of attacking elites who no longer seem very elite or just punching down. I can’t imagine that people listening to the bashing of trans people, or the WNBA, think it will help them in their daily lives.

One thing I am genuinely interested in is how much the MAGA influencers will start to document the very real human toll from Trump’s policies on SNAP and Medicaid. A lot of Trump voters are losing food supports and more will soon lose health insurance because of Trump policies. That is a real basis for populist outrage, a legitimate and tangible argument that Trump has (further) abandoned the people he promised to help.

The Brains Trust

Elisabeth Zerofsky at the New York Times also wrote a deep dive into the crack-up at the Heritage Foundation, the organizers of Project 2025. Much of the blame is laid at the feet at its current President, Kevin Roberts, though the rot set in earlier, when Jim DeMint refashioned Heritage to align with Trump.

Share

When Roberts was hired, he promised to balance the traditional Heritage (more policy focused, greater expertise, traditional free market conservatism) with the MAGA era. He has failed. The experts have been purged and normie conservatives are gone. Increasingly, Heritage is the home for cranks. The whole piece is excellent, and I urge you to read the entire thing. But I will excerpt some key passages.

Because Project 2025 became a public controversy, Heritage lost cachet with Trump. One strategy to reclaim their influence was to purge any critical experts who were not on board.

They were so concerned with getting the clout back with the administration,” the former policy staff member said. “It just exacerbated the tension between the policy people, who were like, Look, I know Trump wants to do this, but we’ve always believed, like, tariffs are bad.

In recent years, the Heritage Foundation saw internal purges, and a sense of fear that mirrors the atmosphere that Trump created in the federal government.

Employees were being fired or resigning without any announcement or explanation. “People would just get disappeared,” one former Heritage policy staff member, who spoke on the condition of anonymity out of fear of professional repercussions, told me. “There would be a weekend where you’d come in and be like, Wait, I’m sorry, how many people are no longer here?” Staff members called it the “Heritage Hunger Games.” As leaks to the media intensified, and as more people sought positions elsewhere, a paranoia settled in. Some employees, believing that their work computers were being monitored, took to communicating with each other over Signal and on burner phones.

With the experts out, Heritage hired hacks and frauds, including a guy who listed a PhD from a non-PhD granting institution, and a Bachelors degree from a high school. The rise of E.J. Antoni is a good example. Even conservative economists said Antoni was dangerously unqualified when Trump nominated him to lead the Bureau of Labor Statistics, but it was not until his misogynist online profile surfaced that the Senate agreed. But this was no scandal at Heritage, where Antoni returned to his position of Chief Economist.

Much of the recent tensions became public when Roberts first defended, and then backtracked, in his defense of Tucker Carlson for elevating neo-Nazi Nick Fuentes. At the time, Roberts talked about supporting free speech rights and avoiding cancel culture. So, apparent from the purges of people mildly critical of Trump, how is free speech at Heritage? Existing employees were told to remove posts critical of policies supported by JD Vance, who wrote the foreword to Kevin Roberts’ book. A Jewish analyst who raised First Amendment concerns about a plan to ban Muslim head scarves in Congress directed to “go dry clean Ilhan’s hijab.” This is a reference to Ilhan Omar, one of the four female Muslim members of Congress.

Rather than discuss policy issues of the day, new Heritage hires debated things like World War II, with a lot of “rehashing and going through who was the bad guy and who were the good guys.” A fair degree of classic antisemitism seems to be in the air. (Maybe Trump’s task force on antisemitism could look into this!) One former Heritage staffer said:

he was willing to debate policy; what he wasn’t willing to do was “sit here and discuss whether the Holocaust happened or not.” Discussions among new hires frequently echoed arguments promoted by Fuentes, Carlson and the podcaster Candace Owens that shed doubt on the historical consensus around the Holocaust or the origin of the state of Israel. Hale described a conversation with one new hire who claimed that the publishing tycoon Robert Maxwell — the father of Ghislaine Maxwell, Jeffrey Epstein’s longtime partner and enabler, and a Czechoslovakian Jew by birth — had paid for the rights to all the history textbooks used in public schools in the United States, and that everything in them about the Holocaust was a lie.

All of this sounds bad for Heritage, but it also has implications for governance. Political scientists describe a tradeoff between loyalty and competence that Presidents must make when selecting appointees. A President might hire someone with more competence even if is unsure that they voted for them. But that choice assumes there are potential hires with some minimal level of competence. But there are fewer people on the GOP side willing to make those investments in competence.

Trump has very explicitly focused on loyalty. It is a trademark of his administration, from the Cabinet on down, and one that has seeped into the broader MAGA world. By doing so, he has altered the market for competence among conservatives. Why invest the hours to learn about economics or social policy, when the guy posting white nationalist memes on X is the one getting hired?

Sometimes the devaluation of competence is advertised explicitly. For example, the conservative Liberty Law School told its students looking for government jobs:

The two most important requirements are you MUST be aligned politically with President Trump and his administration and you must be willing to work hard. Don’t be scared off by the transcript requirement. GPA is not a strong factor. If you meet those two requirements, you have a shot.

Kathy Gilsinan in Politico details how the competence gaps are becoming clear to people working in Congress.

“Everyone knows what the problem is: The organizations that bring young staffers to Washington, D.C. have spent the last decade prioritizing blind ideological loyalty instead of competence or intelligence,” said a senior GOP congressional staffer, granted anonymity to discuss sensitive personnel issues. “We have a generation of rabid, incompetent idiots.”

It is hard to know how deep this problem is, and how much the institutions themselves encourage it. Congressional offices who want to have a big online presence will hire communications staffers who can get clicks online. And to do that as a right-winger in environments like X typically requires hiring those fluent in the toxicity that site has elevated. One Congressional staffer worried about the long-term impacts: “Five, eight years from now, if we’re reliant on people with this view of the world, we’re in serious trouble.”

The stream of incompetent right-wing ideologues into positions of power offers one more data-point for the case that dismantling DEI in the name of merit was a sham. The only type of diversity the right supports now is “intellectual diversity” which everyone understands to be shorthand for more representation of their views, not the views they abhor.

Should we be optimistic about the MAGA meltdown? I think MAGA as a political force is fracturing because they don’t have a common enemy and Trump is not delivering. Trump was the essential man of MAGA, and when he is gone the big question is if anyone can replace him. The movement could easily fracture.

Even so, there are real long term concerns that the views of MAGA have become ever more extreme. MAGA true believers might be losing faith in Trump, but they are not re-connecting to the real world. And like it or not, some of those people will be governing us in the future.

And now, for something completely different

The Oscar-winning Irish singer-songwriter Glenn Hansard died last week. You might not recognize the name, but chances are you might have seen him in The Commitments or Once.

The tributes that flow in from an untimely death of a well-known person duly followed, but a couple of things struck me about Hansard. One was how he was clearly a musician’s musician: people who loved the craft of songwriting respected him deeply. The other was that lots of random people remembered him fondly as an incredibly kind person. And that is sort of the ideal if you work in a public-facing profession: the people who know your craft well respect you deeply, but so too do people outside of your field.

Hansard was also a wonderful storyteller. It came across in his music. His discussion of meeting Van Morrison is worth watching…

…as are any number of live clips of him performing — one aspect of his musicianship is that he excelled in the messiness of live performance. A 2012 performance of him singing the “The Parting Glass,” a traditional song about departures, is especially poignant.

Thoughtful policy experts on a range of issues are currently preparing plans to rebuild the federal government in the event that a pro-democracy president is elected in 2028. That planning is essential.

But it will also be wholly insufficient if it does not reckon with the need to rebalance political and economic power: to build the collective power of workers, consumers, tenants, students, and ordinary people, and to reduce the power of the interests that undermine democracy–unscrupulous economic actors seeking to cheat people, oligarchs exercising undue influence in our elections, or groups threatening political violence.

Over the past year, the three of us have been listening to and learning from former government officials, academics, and other policy experts. The lesson we are taking away is this: in a future rebuilding every major policy and institutional design choice should be evaluated not just on what it delivers, but on whose power it builds and whose power it diminishes.

Decades of missed opportunities for power-shifting

This will require a shift in perspective. For many decades, smart progressive policy wonks have designed new policies and programs, thinking carefully about their economic costs and benefits. But–with a few telling exceptions–these discussions have focused far less on how policies foster collective people power, to the detriment of our democracy.

Take discussions over health care, an area in which we have arguably seen some of the most important domestic policy achievements of the past twenty years. In considering health care policies, advocates and lawmakers have largely focused their back-and-forth on the implications for health insurance coverage and costs. Left largely unaddressed is how changes to the system might foster new collective organizations, such as of patients and their families, that could durably represent those members’ interests.

Imagine if we took a page from union law and created a framework for a national patients’ union with the right to enroll Americans as dues-paying members and to bargain with the government, health insurers, and health providers alike. That union might provide a stronger foundation to protect our health care system against efforts to reduce coverage and increase costs, push more states to adopt policies covering low-income families, and bring stronger clout against predatory insurance and provider practices. It might also create more on-ramps for Americans into the everyday practice of democratic life.

This is not the only such opportunity. Consider the failure to update labor law to make it easier for workers to organize and bargain collectively, which would build more durable worker organizations. Or consider the failure to grapple with the concentrated power of billionaires (and now even a trillionaire), some of whom are using their amassed wealth to threaten democracy.

If we are fortunate enough to have a future rebuilding moment with a pro-democracy administration, we cannot make the same mistakes again. Rebuilders need to think about their work through the lens of power-shifting. They need to consider how they can design policy structures that build durable people power and deconstruct the power of the organized interests opposed to a multiracial, egalitarian democracy. We see this effort as being consistent with, and adding to, other important ongoing efforts to learn from the Biden-Harris Administration’s record and assess what a future administration needs to do differently.

The power of a common language and framework for collective power-shifting

When we started talking with policy practitioners and government officials, we heard general interest in the idea of shifting power to support democracy. But we also heard concerns about how practitioners would actually go about doing so if they were serving in government.

A parallel to economic benefit-cost analysis is instructive here. For better or worse, the language of benefit-cost analysis has become the lingua franca of government. Across agencies and programs, government officials instinctively reach for benefits and costs when evaluating a new proposal, even if they have never taken a seminar on the method. Support for and objections to new policy ideas are often leveled on benefit or cost grounds.

What we have realized is that if power-shifting as an approach to policymaking is to stick, practitioners need a framework for designing and evaluating policies on their power-shifting outcomes that is rigorous, intuitive, empirically grounded, and able to be embedded into government processes, building on research from political science, sociology, economics, and law.

Drawing from the “power analysis” approach pioneered by community organizers and from research on how policies can reshape politics, we argue there are clear and intuitive steps that government officials can take to systematically map the key political actors in a policy space (e.g., companies, worker organizations, tenant unions), identify the economic and political power resources each possesses (e.g., mass membership, market clout, narrative advantages), and then craft policy designs that build the resources of pro-democracy groups while diminishing the power of actors who undermine democracy.

We cannot ignore power deconstruction

It can often seem easier–and more comfortable–to talk about building people power than to plan for deconstructing the power of actors who undermine our democracy. This is an understandable but ultimately misguided impulse. Federal rebuilding must include both sides of power-shifting: building and deconstructing.

Without a sustained effort to diminish the political and economic power of business models built on corrupt practices, oligarchs trying to buy elections and reshape the government without any democratic accountability, and powerful economic actors working to stop Americans from organizing collectively to advance their interests, our democracy is at risk of further backsliding.

Power deconstruction has also been at the heart of strategies to undermine our democratic norms and institutions. Consider the Trump Administration’s ongoing campaign against universities. Whatever you think of higher education’s flaws, the campaign has had little to do with actual civil rights protection. It is a sophisticated strategy for deconstructing the power of a major pillar of civil society. The administration ran the same play against major law firms, using executive orders stripping security clearances and threatening clients’ federal contracts to extract roughly a billion dollars in pledged pro bono work from firms that chose to settle rather than fight.

None of this should surprise us. The anti-democracy movement has spent five decades deliberately attacking the organizational pillars that give power and voice to ordinary people: right-to-work laws and collective bargaining rollbacks to weaken unions; tort reform to weaken the trial bar that enforces consumer, workplace, and environmental laws when government cannot; and now the assault on universities, law firms, and the nonprofit sector. In short, opponents of a multiracial, egalitarian democracy have long understood that policy is a tool for durably restructuring politics.

To be clear, the lesson we should take from the Trump Administration’s assault on civil society is not that we need to adopt the authoritarian playbook.

De-powering authoritarian power centers has to be done through general, publicly defensible standards rather than enemy lists; through transparent legal processes that can survive public scrutiny; through criteria focused on protecting the social and institutional foundations of inclusive democracy rather than punishing political opponents; and through consistent rules that apply to allied as well as adversarial actors. This is a stringent test, but democratic rebuilders have to be stringent, because the alternative would be to legitimize, by example, the very tactics used to undermine democracy.

If the deconstruction side of this agenda sounds challenging, remember that every major democratic advance in American history has paired power-building with power deconstruction, often requiring what David Bateman has called “democracy-reinforcing hardball”.

Reconstruction paired the extension of rights to freedpeople with federal troops, the Freedmen’s Bureau, and military governance aimed at breaking the political power of slaveowners. The New Deal paired economic relief with an assault on concentrated corporate power: financial regulation, the breakup of utility holding companies, and a labor law that ended employers’ unilateral control of the workplace. The Great Society’s Civil Rights Act created levers to withhold federal funds from segregating institutions, defunding the organized infrastructure of Jim Crow. In each case, the durability of the democratic gains tracked the thoroughness of the deconstruction.

What we’ve been learning about power-shifting — and where we go from here

How do we make power-shifting a tangible framework for governing? We asked academics, former government officials, and policy experts to write up key lessons for designing and implementing power-shifting policies. Together, these papers, commissioned through our Empowering Democracy Initiative, draw on academic research and firsthand experience in government to spell out how policymakers should approach building pro-democracy power and deconstructing anti-democratic power in a future administration.

• Medicare policy: For example, Sam Bagenstos, a law professor and former general counsel of the Department of Health and Human Services, reflected on the costs associated with not considering entrenched power when making Medicare policy.

• Civil society: Raj Nayak, a former senior Department of Labor official, explored how partnerships between government and civil society can offer more opportunities for expanding government capacity and people power at the same time (a joint effort with Workshop, a community of labor policy experts).

• Social movements: Political scientist Mallory SoRelle laid out how different policy designs embedded in federal programs can make it easier or harder for social movements to organize.

• The Federal budget: Former Office of Management and Budget official Michael Linden argued that the White House-led budget planning process could offer a place to embed power-shifting in a future federal administration.

• Breaking entrenched power: Political scientists Jamila Michener and Aubrey Hall explored tactics for attacking entrenched power threatening everyday peoples’ lives.

Together, these ideas start the process of mapping out a reconstruction of government that takes power seriously. The throughline is that every policy fight is also a fight over the future distribution of power, whether we acknowledge it or not.

Authoritarians are executing on this insight now with discipline against universities, law firms, unions, and the other civic infrastructure of American democracy. A future pro-democracy government that responds by restoring the status quo ante, putting back the old programs and restoring the old approaches, will simply be setting the table for the next, more competent authoritarian.

The alternative is to treat policy as an opportunity to build organized civil society constituencies whose power makes democratic backsliding harder to attempt and easier to resist, and to dismantle, lawfully and through general standards, the concentrations of power that undermine our democracy. Rebuilding the federal government is inseparable from rebuilding the political and economic landscape on which it stands.

Over a million Americans died from Covid. I just wanted to mention this data point, because it seems to have been forgotten.

In the early days, we had little information about the nature of the virus. The pace of death outstripped our medical resources. At one point in New York, someone was dying almost every two minutes (four times the normal rate) and bodies had to be stacked in trailers. It was an incredibly scary time.

Yes, you know all this. And yet, it’s become almost a conventional wisdom that the US response was imperfect because it overreacted. “Maybe Fauci meant well, but…”

All institutions are imperfect, and crises expose those imperfections. I think the response to the pandemic does expose real institutional failures, but the greatest of those failures are those of our political system, not our public health system. The pandemic was a global phenomenon, but it seems to have broken the US in a unique way. Other countries had more serious lockdowns without entering a political psychosis.

In a country where conspiracy theories like QAnon are political movements, a real crisis provided the vessel for the worst aspect of American politics. And we are still living the politics of the pandemic. In 2026, Republican politicians are still looking to blame and jail public health officials for the pandemic while accepting no responsibility themselves.

While the rest of the world has moved on, in right-wing spaces Fauci is still a central character. Media Matters reported that Fauci has been:

• Mentioned 687 on Fox News and 740 times on Newsmax in a four day period.

• Compared to Hitler and Epstein, and declared a mass murderer by multiple prominent right wing media stars who called for his imprisonment.

Lets break down how this political and media environment that has real costs for public health preparedness.

We are worse off today; that’s not Fauci’s fault

A functional governing system learns from its mistakes so it can do better the next time. Let me ask you: how confident are you that we are ready for the next pandemic?

It seems clear to me the US is less well-prepared that it was before Covid. Trump has stripped away pandemic early warning systems that USAID enabled, and gutted public health capacity, replacing experts with cranks. Elected officials in the governing party claim that the vaccine killed more people than it saved.

But the biggest problem is that there no longer seems to be the sense of collective trust that will enable a society to work together during a crisis. You can choose to think that collapse in public trust is the result of the failures of Fauci and other public health officials. You don’t have to agree with every aspect of the response, but it is reasonable to acknowledge we sit today with the benefit of hindsight.

But you don’t get hindsight with the next crisis. You have to choose which officials to trust. You know that the next public health response will also be imperfect. But in what direction?

Right now we live in a world where public health officials are afraid to be fired, or threatened, or hauled in front of Congress. The head of Health and Human Services, RFK Jr., enjoyed lower public trust in his honeymoon period than Fauci did after years of attacks and only appears to be interested in vaccine research that discredits vaccines. The horse tablet promoters have the ear of the president.

At this point, Senators like Ron Johnson, Tommy Tuberville, or Rand Paul can say absolutely insane things, get some clicks and pay no penalty. The insanity has become background noise, mood music for the show trials. By contrast, an 85-year old scientist who dedicated his life to public health is subject to ongoing attacks and potential prosecution.

The behavior that is rewarded will become a greater feature of our society. The behavior that is attacked will disappear. We chase the Faucis from public service at our peril.

We live in a world where people have been repeatedly told not to trust public health advice, even when it is absolutely sound. This is primarily a political problem, and not one of Fauci’s making.

Such cultivated partisan distrust has real world implications. Even during the pandemic, it became clear that more Republicans were getting sick and dying because they were listening to right-wing media sources. It is perfectly fair to note that public health outcomes have to be weighed against other values, but recent research by Elizabeth Elder and Neil O’ Brien points to how unnecessary some of these deaths are.

Before 2010 they did not observe a correlation between political ideology and health outcomes, but this changed, arising from greater Republican distrust in medical advice, greater unwillingness to seek care, even for health issues unrelated to Covid. The authors concluded: “These patterns suggest that growing ideological divides in health behaviours are leaving conservative Americans increasingly vulnerable to preventable health risks.”

The hearing was part political theater, part a prosecution trap

Fauci, long since retired, was the only witness at a Senate hearing. Senator Rand Paul subpoenaed him to appear after Fauci initially declined. Paul also released Fauci’s personal diary, in full.

The diary was presented by Paul as a huge reveal, but it was closer to a nothingburger. What did it show? Fauci grew frustrated with Trump’s rambling or directing him to be positive when thousands were dying. He discussed the lab leak hypothesis with other experts but did not see supporting evidence. There is no evidence of a cover up, as Paul insinuated.

The diary release came after the outgoing Director of National Intelligence Tulsi Gabbard tried to salvage her failed tenure by releasing the “Fauci files.” Again, a nothingburger. Writing at LawFare, Renée DiResta not only debunked the files, but spoke to the same politics of insinuation that Paul relied on with the diary release:

Gabbard’s “Fauci Files" follow the same playbook as her past document drops: Put the desired conclusion in the headline. Use “declassification” as a credibility signal. Present a large, mixed dump of documents as if they were a prosecutor’s exhibit. Convert ordinary government processes—expert consultation, grant oversight, whistleblower routing, intelligence disagreement—into evidence of conspiracy. Then push the accusation on social media, where few people will read the underlying documents closely enough to notice the gulf between the documents and the claims.

And so, both the Trump administration and members of the Senate aggressively pushed claims of a new Fauci bombshells that did not hold up. Never mind, you don’t need evidence for a show trial.

Fauci read an opening statement where he accused Paul of having an "unhinged obsession" with him and described the hearings as an effort to generate material to prosecute him. He was not wrong. While Fauci was preemptively pardoned by President Biden, Paul told Fox News: “If he lies during the hearing, that’s a new lie, and the pardon won’t cover that.”

This is an administration that has prosecuted their political enemies for the most flimsy of pretexts. David Morens, a 78-year old retired Fauci aide and Army reservist, was arrested as if he was a major drug kingpin. A half-dozen armed federal agents wearing bulletproof vests arrested him in front of neighbors, removing his pants and shirt, and forcing him to surrender his passport.

Morens alleged crime? Using private email rather than a government email account, a practice which is routine among Trump officials but deemed prosecutable among their enemies. Don’t hold your breath waiting for Elon Musk or Pete Hegseth be arrested for the same crime.

And so, Fauci has every reason to assume he would be prosecuted on a Trumped-up charge. For the first time in over 200 appearances before Congress, he took the Fifth amendment, refusing to answer questions to avoid the trap that Paul had set for him. Ken White, the first amendment litigator wrote that:

It would be deeply reckless to testify, rather than taking the Fifth, when incredibly dishonest and unhinged partisan fanatics in the Senate are demanding that you be prosecuted and haul you in for a perjury trap struggle session. Don’t hand a spree killer a knife.

Apparently Rand Paul was not happy that Fauci would not cooperate in a partisan effort to send him to jail. He announced that Fauci did not hold Fifth Amendment protections (because of his pardon, which seems like a novel constitutional claim) and would be pushing to hold Fauci in contempt of Congress.

When Paul threatened Fauci during the hearings that there would be repercussions for not answering questions, Fauci’s lawyer tried to respond. Paul refused to recognize him. When ranking Senator Gary Peters tried to use his time to allow Fauci’s lawyer to speak, Paul overrode him, and then ordered the lawyer to be removed by security.

Yale Law graduate Josh Hawley announced “no innocent person pleads the Fifth.” Well, President Trump pleaded the Fifth nearly 450 times in a single case. I’m guessing Hawley won’t be leading the effort to impeach Trump.

At the hearing, Senator Bernie Moreno asked Fauci “Do you feel like you’re in deep shit?" and “Who the fuck did you think you were?" In an interview afterwards, he said: "that was me being nice. That was holding back, trying to be polite given the respect I have for the office." Sounds like a guy not to be messed with!

Also, Moreno’s daughter recently accused her former husband, a Republican Congressman, of physically abusing her including holding a gun to her head. Moreno has not said anything publicly about the matter while the Congressman is running for re-election. His tough-guy act is reserved for retired scientists.

Share

Who would be a public servant in these conditions?

When a crisis occurs, you need competent people. Fauci was one such person. He was an advisor. He did not force Trump or any other policymakers to do anything. And yet, at the hearings, Senator after Senator accused him of being responsible for lockdowns, vaccine injuries and school closures. In effect, Fauci became the scapegoat for every policy decision that Republicans now disagree with. Paul even suggested that Fauci created the pandemic: “Are millions dead because of his experiments?”

Fauci became the face of the pandemic response because he was uniquely credible compared to the rest of Trump’s team. And for that service, he has been threatened and harassed into his retirement. An esteemed medical scientist, he was lectured to by cranks, forced to testify to a group of Senators hoping to send him to jail.

One of the false claims against Fauci is that somehow he was personally benefiting from the pandemic, via some shadowy patents. The truth is closer to the opposite. In a previous hearing, Fauci pointed out that Paul’s website was using Fauci hearings as fundraising tactics. Paul also authored a book called “Deception: The Great Covid Cover-Up” which put Fauci on the front cover. Immediately after the hearings, RFK Jr., who provided Fauci’s diary to Paul, used the attention to promote his own book attacking Fauci.

Meanwhile, Fauci has to live with round-the-clock security protection.

I once met Fauci at Georgetown University. After sharing some pleasantries I chatted with another official, who had also served in the Trump administration. I observed that Fauci seemed to have a lot of bodyguards. He confirmed that Fauci’s protection was closer to that of a former President than a Cabinet official, which was a direct reflection of the seriousness of the threats he faced, which include an armed man who was arrested while traveling to DC planning “to kill Dr Fauci”.

In his diary, Fauci noted these threats:

Threats on the ‘dark web’ [against] me, Chris and the girls. They know everything about us including where my daughters live and work. Strange people hanging around my house and taking photos of the house.

Fauci also noted a lot of the threats came from the far right. “[Highly] tense political situations. I quite frankly do not really care. I will keep doing what I am doing.”

Trump canceled Fauci’s security detail last year. Not on the basis of some careful new risk assessment, but following the logic of retribution that saw him cancel the security protections of others he viewed as his political enemies. Fauci must now pay for the security himself, even as Republican officials drum up the outrage that makes that security necessary.

As a reminder, Fauci was a career civil servant, not a political appointee. When he served, we still believed that the President could not fire civil servants, except for cause. It was also relatively unusual that civil servants would be subject to public partisan attacks that triggered harassment and threats. Fauci worked in government for a long time because he could do amazing, pathbreaking science. Now the President fires civil servants for any reason at all, publicly threatens those he dislikes, and has defunded much of that pathbreaking scientific work.

In other words, this is not just about Fauci, its about attracting competent people to work in a government where they can be deemed expendable or objects of hate by a political system looking for scapegoats.

Senator Andy Kim offered one of the most insightful points from the sorry event:

Right now, I feel the hyper-partisanship growing in this country in dangerous ways. People calling our public servants criminals [is] undermining the very nature of the work of governance.

If you were a smart young scientist watching the Fauci hearings, would you still want to go into government? There will certainly be a time in the future when we need people like Fauci in public health positions. Let’s hope some are willing to help us.

I started getting a bizarre series of e-mails from the Social Security Administration––over the signature of the agency’s administrator––that read more like campaign literature. Among other things, they praised the Trump administration’s elimination of offices created under the Biden administration and applauded the passage of the president’s signature “One Big Beautiful Bill Act.”

The language of the emails was florid. The most recent of the e-mails I received gives a flavor of the thing. A recent message from the Social Security Commissioner Frank Bisignano, after wishing America a happy 250th birthday, informed me that “President Donald J. Trump and the Trump Administration” are delivering historic relief to seniors, that wait times are plummeting “under the leadership of President Trump,” and that, “put simply, America’s seniors are winning!”

This is the sort of message the Social Security Administration has been blasting out for over a year to more than 70 million account holders. In a normal political environment, this would be a big deal. But in an era where the Trump administration has “flooded the zone” with daily scandals, the emails have attracted little attention.

The messages come in a context where SSA has seen mass layoffs that are undermining its performance. The Center for Budget and Policy Priorities report that:

agency leadership claims that SSA is “working better and faster than ever” seem readily undercut by press reports on customer service and processing strains at the agency. For example, phone wait times are much longer than publicly advertised; lengthy waits for appointments particularly harm widows and children who have lost parents; local offices often struggle to operate with shrinking staffs; and record-high processing backlogs create longer waits for needed help.

All of this has become a bit easier for Bisignano to lie about, of course, since the SSA removed customer service metrics from its website. Bisignano is lately under fire for spying on his colleagues when he worked at JP Morgan. And for handing out gift baskets from his daughter's pretzel company to raise Social Security staff morale after unprecedented staff cuts.

Last July, former SSA officials from both parties condemned the emails for using the voice of the SSA to make a blatant political statement, as did members of Congress, who demanded a factual correction in an email praising the OBBBA.

In the days that followed those initial blasts, I received emails and texts from family and friends with what has become a standard battery of questions: “What the fuck is this?”, “Is this a scam?”, and “Is this even legal?”

The story died down, but the emails kept rolling out.

The branding project has also expanded beyond emails to the issuance of Social Security cards themselves. SSA recently announced that every baby born between July 2 and December 31 will receive a Social Security card stamped with the logo of “Freedom 250”. This is not the congressionally chartered America250 commission, but a separate entity the Trump administration created to run its own semiquincentennial programming. It joins the passport now available with the president’s portrait inside the cover as one of the most blatant uses of the federal government to promote the brand of a single officeholder.

This iconography is just one example of how Trump and his proxies have transformed the executive branch into a tool for what Bill Resh has called “personalist consolidation,” which leverages public institutions as “instruments of personal or factional loyalty”.

But the fear that a president might use the power of an increasingly powerful executive branch to propagandize is as old as the term “propaganda” itself. As early as 1913, Congress has written into statute general prohibitions on publicity expenditures unless there are specific funds allocated to that purpose.

Since 1952, Congress’s annual appropriations acts carry a simple one-sentence rider declaring that no appropriated funds “shall be used for publicity or propaganda purposes not authorized by Congress.” The Consolidated Appropriations Act of 2026 is no different.

If this sounds simple – or simple enough to apply to the sort of egregious cases we’ve seen under the second Trump administration – the reality of implementing appropriations law is anything but.

Defining publicity and propaganda

First, what do the terms “publicity or propaganda” mean?

The Government Accountability Office (GAO), which adjudicates these matters, outlines its interpretation of Congress’s publicity and propaganda rules in the Red Book, its multi-volume compendium on appropriations law. As GAO conceptualizes it, there are three categories of “publicity or propaganda”.

First, there is agency “self-aggrandizement”, or “communications tending to emphasize the importance of the agency, its officials, or the activity in question”. In 1986, for example, an advisory commission “whose sole function was to advise the President and which had no promotional role, set up and implemented a public affairs program” of its own, triggering a GAO finding of a violation of appropriations law.

The second category is “covert propaganda”, or “communications that fail to disclose the agency’s role as the source of the information.”

One of the most common forms of covert propaganda is the use of prepackaged news stories. For example, in 1983, the Office of Public Diplomacy for Latin America and the Caribbean was established within the Office of the Secretary of State “to engage in a campaign to influence the public and the Congress to support increased funding for the Administration’s Central American policy. pursuit of its public diplomacy mission.” The newspaper articles and editorials placed by the office, as well as media visits by Nicaraguan opposition leaders, were all cited by the Comptroller General as a violation of appropriations law.

The third category is the distribution of “purely partisan materials”, which are those “designed to aid a political party or candidate.”

This category is particularly difficult to apply, the GAO notes, “because “the lines separating the nonpolitical from the political cannot be precisely drawn.” For example, in 2000, the Department of Housing and Urban Development disseminated a publication called Losing Ground: The Impact of Proposed HUD Budget Cuts on America’s Communities, which criticized “deep cuts” Congress had made to certain HUD programs which would have a “devastating impact on families and communities nationwide.” Given that this was a Democratic administration criticizing the decision of a Republican Congress, the case no doubt warranted scrutiny. Yet GAO determined that this was a “legitimate exercise of HUD’s duty to inform the public of government policies and that HUD had a right to justify its policies to the public and rebut attacks against those policies.”

A more concerning case arose in 2005, when the US Department of Education conducted a media analysis to gather information regarding “public perception of the Republican Party’s commitment to education. While the GAO issued an opinion stating that there were “no appropriations available for analyzing the media and public’s opinion concerning the Republican Party’s (or any other political party’s) commitment to education,” it ultimately concluded that “that there was no publicity or propaganda violation, given that other content of the analysis was acceptable, and thus there was minimal additional expense incurred by including the improper element.”

Indeed, the line between acceptable and purely partisan communication has proven so difficult to police that the last edition of the Red Book states that, to date, there are “no opinions or decisions of the Comptroller General concluding that an agency’s informational materials constituted impermissible publicity or propaganda by virtue of any purely partisan material contained therein.”

In short, where the GAO’s interpretation of the words “publicity or propaganda” is concerned, there are plenty of gaping holes for the Trump administration to walk through.

Oversight and enforcement problems

But the problem goes deeper than the ambiguity of terms like “publicity” and “propaganda”. As Kevin Kosar explained in a perceptive 2005 Presidential Studies Quarterly article, even keeping track of potential government publicity or propaganda efforts can prove difficult. Agencies’ budgets and federal procurement data provide only the fuzziest of details on public-relations expenditures, including how public employees spend their time on such efforts.

In a world where government is conducted not by press release but by social media posts, the challenge of tracking the appropriations to scan for appropriations-law violations becomes even more difficult.

So much for oversight. What about the enforcement of appropriations law? There, Kosar suggests, the problem gets even worse. It is GAO that is principally charged with policing appropriations law. But, as a creature of the legislative branch, it has no direct authority to sanction agencies or individuals who violate the law. It can only report its findings and refer violations to the Department of Justice.

“When it comes to propaganda,” Kosar notes, “the executive branch has demonstrated little interest in punishing wayward executive agencies.” At the time of publication, Kosar could locate neither a single opinion of the Office of Legal Counsel on the “publicity or propaganda” statute nor a single criminal prosecution under it.

If executive-branch agencies have made anything clear in recent years, it is that they do not much care what GAO has to say about appropriations law. In 2005, both the DOJ and the OMB issued memoranda stating that GAO’s interpretation of federal appropriations law does not bind them.

With an executive branch that has zero incentives to enforce appropriations law, the ball is firmly in Congress’s court. The problem, of course, is that only the DOJ can effectively act. Provided that Congress is not controlled by the President’s party, it can use its tools of oversight to expose violations of appropriations law. Though even here there are limits. Not only is oversight costly and, as noted above, difficult to do in a flooded zone, it is not entirely clear whether and how public attitudes on executive-branch propaganda—which reflect policy preferences and partisanship—can be moved sufficiently to alter agencies’ incentives.

Share

The Hatch Act?

Appropriations law is hardly the only tool Congress has created to police political activity by employees of the executive branch. The Hatch Act, passed in 1939, limits political activities by nearly all federal civilian executive-branch employees. Among many other things, it prohibits all covered employees from using their “official authority or influence for the purpose of interfering with or affecting the result of an election” and “engaging in political activity while on duty; on federal property”. To enforce the Hatch Act, the US Office of Special Counsel (OSC) renders advisory opinions and can investigate and prosecute alleged violations.

In May of last year, for example, the OSC found that the Biden-appointed Social Security Commissioner, Martin O’Malley, violated the Hatch Act during a radio interview in October of 2024, in the midst of a presidential election campaign, when he said that Donald Trump’s policy proposals would “quickly deplete Social Security”. This statement, the OSC determined, was a clear-cut violation of the law. Despite framing his remarks as commentary on a policy proposal and emphasizing that he “could not tell anyone who to vote for”, the OSC concluded that O’Malley had made an “express message to voters that they had a responsibility to vote against President Trump.” Yet even in this relatively clear-cut case, discipline proved elusive. By the time that OSC issued its report, O’Malley had already left office.

It may be even more difficult to use the Hatch Act to police emails like the ones sent out by Trump’s Social Security Commissioner. True, a string of misleading official e-mails crediting the sitting party leader by name for purported policy benefits is arguably political activity performed “on duty or using government resources” under Section 7324. Indeed, even content that looks merely promotional can be a Hatch Act violation if there is evidence of electoral purpose. The record of what was said internally about the emails sent over Bisignano’s signature has never been publicly examined.

But enforcement is highly unlikely in the present circumstances. In February of 2025, Trump fired Hampton Dellinger, Biden’s Senate-confirmed head of the OSC. Since then, the OSC has had no Senate-confirmed leader. Wherever the line is, the present OSC will not be the one to decide whether SSA’s recent emails have crossed it.

Unchecked corruption in an election year

And so, the reality is that the Trump administration is using public resources and the trust people have in public organizations like the Social Security Administration to send partisan and misleading messages in an election year with no meaningful constraint. Even as headlines portray a Social Security Administration struggling to recover from DOGE-era cuts, the administration is filling our inboxes with the claim that things are better than ever.

The broader picture is that the legal institutions meant to police political corruption in the United States are failing us. From pardons of political supporters to naked quid pro quos for large donors to the most mundane of official emails, Trump has leveraged the federal government for personal and partisan gain.

Amidst a hotly contested 2026 midterms, we should not be surprised to see the administration use the voice of federal agencies to interfere with the democratic process. Even when enforcement is unlikely, however, documenting and broadcasting the extent of this corruption still matters. While federal agencies should be able to tout actual achievements, they should not become publicity arms of a personalist regime.

Subscribe now

1. Administrative burdens are an effective way to limit people’s ability to access public services.

2. President Trump’s One Big Beautiful Bill Act weaponized these burdens to cut SNAP (income supports for food) and Medicaid to offset a tax cut that will largely go to wealthy Americans.

In other words, your government is making the most vulnerable Americans jump over more and more bureaucratic hurdles, rooting for them to fall, so that it can take the money it would have spent helping them and give it to the richest Americans. It is a deeply dystopian vision of governing, where the failures of the weak profit the strong.

This was apparent even as the One Big Beautiful Bill Act was being debated last year. The OBBBA mostly does not change the eligibility for our major safety net programs, but it does make them a lot smaller via administrative burdens. When the bill was being considered, we pointed to the almost one-for-one tradeoff between tax cuts and benefits cuts.

And back in October of 2025 we noted that the OBBBA, by design, would lead to millions of eligible beneficiaries losing coverage. States, facing huge financial penalties if they didn’t reduce their error rates, would add layers of new paperwork requirements that beneficiaries wouldn’t be able to navigate.

A challenge of writing before these events happen is that we deal in abstractions: numbers and graphs tell us about the size and nature of the impact. It’s one thing to predict that millions will lose critical benefits. It is an entirely different thing to watch it actually happen, and to read the stories of the people facing the consequences of political choices.

Now, the bill of the One Big Beautiful Bill Act is starting to come due. Since last July, when the OBBBA was signed, more than 4 million people, including 1.5 million children, have lost coverage. Coverage is plummeting across the country. The Center on Budget and Policy Priorities is keeping a detailed tracker of these trends.

Arizona is the canary in the coal mine, pushing more quickly to reduce its SNAP error rates, and offering a preview of what is likely to happen elsewhere. More than 440,000 people lost coverage. In an excellent piece in the New York Times, Jason DeParle and Adriana Zehbrauskas tell their stories in a way that is more compelling than any graph we could give you, and so we are sharing in detail.

Share

Too many requirements

Arizona responded to the threat by unilaterally increasing documentation requirements, going so far in one case as to push a homeless man to document his panhandling income.

But the increase in compliance costs affected nearly all parts of the application process. For example, instead of just noting your household size, the state started requiring many “to get signed letters from neighbors or friends” to verify it. Many caseworkers also started refusing to take earnings data from employers over the phone–a quicker process than filing additional paperwork documentation.

One school of thought suggests that people who least need benefits will simply opt out when the hassles become too high. But there is more evidence that the most vulnerable will be the most hurt by burdens. For example, people with poorer health struggle more with administrative demands. As Jonathan Fleming said: “I need to focus on my health, not to be dealing with this.

Too little staff

Additional bureaucratic requirements don’t just hit beneficiaries, it adds significant amounts of work for the caseworkers processing these applications. And at the same time the state increased documentation requirements, they laid off a third of their caseworkers.

The impact for beneficiaries was swift.

Last October, Ms. McDonald, 65, tried to renew a $700 monthly SNAP benefit that supplements Social Security and provides a third of her income. In November, no aid arrived. In December, a caseworker said to “hang in there” because renewals were months behind. In January, Ms. McDonald made 100 calls without reaching anyone, she said. In February, she discovered her case had been closed. She had submitted the wrong income statement, she later learned, and the state’s delays left her no time to fix it.

McDonald is a cancer survivor who weighs less than 70 pounds and is skipping meals and hustling to food banks to make sure the grandkids she takes care of are protected. She “describes her worries as a mental tax, paid nightly on her pillow.” Psychological costs are a key part of administrative burdens, invisible but ever present. She told reporters: “I go to sleep thinking about what are we going to have. I’m exhausted.”

Ms. Klaman, the mom of five who is deaf, struggles watching her daughter sell blood: “Inside I’m an emotional wreck. I’m her mother — I should be the provider — and we’re using her blood.”

A year ago, how many of us had heard of the term error rate? Mostly, the error rates represent mistakes rather than fraud, sometimes they come from beneficiaries and sometimes from caseworkers. When a reporter explained to Dee McDonald about how the Trump administration was penalizing states with high error rates, she “wore a look of sudden understanding. “It’s the pressure from the administration not to make a mistake,” she said.”

It would be easy to blame Arizona, but they are responding as the Trump administration hoped they would, with few good options. The head of the agency that administers SNAP was admirably honest, noting that they faced a $300 million cut otherwise:

It’s frankly sickening to me the number of individuals that continue to struggle with the added bureaucracy…It’s an existential threat. States are absolutely contemplating a world in which SNAP isn’t available.

One of the arguments the Trump administration has made is that these programs have become too generous, and are beset with fraud.

What is defined as too generous is subjective, and the best way to have that debate would be to straightforwardly change eligibility levels to limit benefits to fewer and lower income people. But that is not what happened. The surreptitious use of administrative burdens replaced that policy debate, because it was a debate the Trump administration knew it would not win. And despite the claims about fraud, the vast majority of those being hurt right now are eligible.

So far, we have focused on SNAP. It will get worse. But it’s only the start. The second part of the One Big Beautiful Bill Act comes later. Deliberately timed for after the 2026 midterms, Medicaid work requirements will kick in on January 2027. It will hurt people on Medicaid the same way that people on SNAP are being hurt now.

Gabriel Scheffler recently broke down how this will work. Again, we can tell you what will happen before it happens. In a few months we will be reading similar stories of human misery. It is predictable, because it is by design.

You’ve probably registered by now that prediction markets exist. Keeping track of why they keep producing scandals is another matter. Prediction markets, their users, and their would-be regulators have been busy. From the mundane — more than $200 million riding on the definition of a “suit“ — to the less mundane, like President Trump’s teleprompter operator getting arrested for allegedly placing bets on Kalshi about the speeches he was scrolling.

The Wall Street Journal also recently reported that Polymarket paid influencers to fake winning bets on their platform. The platform admitted to Politico that it does, indeed, pay influencers to mention the company. Most people wouldn’t take that to mean that, across dozens of videos, influencers “touted nearly $900,000 in fabricated winnings on positions that would have lost more than $166,000 in live markets.” The simulated bets included fake websites that influencers used to make false claims about their Polymarket success, and evidence that influencers shared the videos with Polymarket for approval before posting.

The Polymarket promotional videos are plentiful enough to represent a distinct genre, complete with their own language, according to the Wall Street Journal analysis.

Oh, and the Commodity Futures Trading Commission (CFTC), the federal agency regulating prediction markets, has been suing states.

It’s been an exciting time for anyone keeping tabs on prediction markets and how they’re (not) being governed.

Share

What Has Congress Done?

Part of the difficulty in governing prediction markets is that there’s little agreement on what they are. The CFTC, Congress, and the states see them differently, and that has implications for which tools are available to manage them.

Over twenty pieces of legislation have been introduced. Exactly one has taken effect.

It didn’t need the House. It didn’t need the President’s signature. It doesn’t apply to Polymarket, Kalshi, or the average American.

It applies only to people sitting in the U.S. Senate.

On April 30, the Senate adopted S.Res.708, banning Senators, officers, and staff from trading on prediction markets. It passed unanimously, immediately, without drama because it’s a resolution, not a law, and the only people it binds are the ones who voted for it. The House has tried twice to do the same thing to itself (H.Res.1248, H.Res.1263) and hasn’t managed it yet.

The bills cluster into three groups and each says something slightly different about what a “prediction market” is supposed to be and how one is meant to operate.

There’s a bill to ban death and assassination contracts. A bill to affirm CFTC exclusivity. A bill that affirms the opposite. A bill banning members of Congress from trading on political outcomes. A different bill banning members of Congress from trading on political outcomes. A third bill banning members of Congress from trading on political outcomes, this one with disgorgement penalties. None of them have passed.

Here’s a non-exhaustive list of prediction platform bills introduced this year (note: some fall into more than one category):

1. Regulatory framework bills

• Prediction Market Act of 2026

• Prediction Markets Security and Integrity Act of 2026

• Prediction Market RISK Act

2. Content-restriction bills

• DEATH BETS Act

• BETS OFF Act

• STOP Corrupt Bets Act of 2026

• Prediction Markets Are Gambling Act

• Fair Markets and Sports Integrity Act

3. Insider-trading/officials bills

• PREDICT Act

• End Prediction Market Corruption Act

• Public Integrity in Financial Prediction Markets Act of 2026

• Congressional Prediction Market Ban Act of 2026

• S.Res.708 (agreed upon in the Senate on April 30!)

• H.Res.1248 and H.Res.1263 (House versions of the above Senate resolution, not yet passed)

Twenty Bills, One Rule

The volume obscures a more basic problem: the bills don’t agree on what they’re regulating.

The framework bills can’t settle what a prediction market is. The Prediction Market Act of 2026 treats them as event contracts under the Commodity Exchange Act, with consumer protections layered on top: a commodity, in other words, and the CFTC’s to govern. However, the Prediction Markets Security and Integrity Act of 2026 seeks to “return regulatory authorities and oversight of online prediction markets to States.” It says the platforms offer services substantially the same as betting, and that states were never meant to be precluded from enforcing gambling law against them. One bill says commodity. The other says wager. Same chamber, same session, no answer.

The content-restriction bills can’t agree on the harm. The Prediction Markets Are Gambling Act and the Fair Markets and Sports Integrity Act say the problem is gambling, and go after sports and casino-style contracts. The content restriction bills DEATH BETS and BETS OFF say the problem is subject matter: war, assassination, and violence. Two very different theories of what’s wrong. Neither has passed.

The insider-trading bills can’t settle who counts as an insider. Politicians, sure. Their spouses? Staffers? Bureaucrats? Consultants? In March, Kalshi and Polymarket rolled out their own rules blocking politicians and athletes from trading on markets they might influence. Alexandria Ocasio-Cortez called it “a fig leaf,” pointing out how many people (e.g., staff, advisors, consultants, cabinet secretaries, spouses) sit outside a list like that and can still trade on what they know.
Source: X/Twitter.

The same critique lands on the legislation. The PREDICT Act at least reaches members’ spouses and dependent children. S.Res.708, which is the only measure actually in effect, carves these groups out. The one rule on the books has the exact loophole AOC’s criticism was about, now with the force of a Senate rule instead of a platform’s terms of service.

The Federalism Question

While Congress has been introducing bills that haven’t come to pass, a constitutional fight has been happening somewhere else entirely. It’s a live federalism dispute, not a policy disagreement.

The federal CFTC has now taken legal action against nine states: Arizona, Connecticut, Illinois, New York, New Mexico, Minnesota, Rhode Island, Wisconsin, and Kentucky. Its argument in each is the same: it has “exclusive jurisdiction“ over event contracts under the Commodity Exchange Act, and state gambling law is therefore displaced entirely, no matter what a state legislature decides. Those nine are only the cases where the CFTC went on offense. Roughly twenty states are in active litigation over prediction markets. The CFTC’s nine is just the visible federal front of a much broader fight.

States, rather than prediction markets, are the ones who can expect federal lawsuits limiting their actions in this space. Michael Selig, the CFTC’s Chairman, has been blunt about the CFTC strategy for months: “To those who seek to challenge our authority in this space, let me be clear — we will see you in court.”

He wasn’t kidding, and he doesn’t wait.

The CFTC filed a lawsuit one day after Minnesota Governor Tim Walz signed a law making it a crime to operate, host, or promote a prediction market in the state. (This was after a Minnesota state senator admitted to betting on himself on Kalshi to win his own primary. He apologized. He’s banned from the platform for five years.)

Kalshi won legal battles in New Jersey, where the Third Circuit upheld its injunction in April, and in Tennessee. It lost in Maryland, Nevada, Ohio, and, this month, in New York, where Judge Analisa Torres found the Commodity Exchange Act doesn’t preempt state gambling law as applied to sports contracts. Kalshi has appealed to the Second Circuit.

Not every state has picked a side in the fight-or-fold binary.

North Carolina’s Senate Bill 257, signed by Governor Josh Stein on July 7 and buried inside the state’s 2026 budget, does something neither the CFTC-suing states nor the platforms’ allies have tried. It explicitly recognizes the CFTC’s exclusive regulatory authority over prediction markets, requires no state license or registration, and then taxes them anyway at 6% of net trading fee revenue starting January 1, 2027.

Meanwhile, the CFTC has decided it isn’t waiting for further developments from courts or state legislatures. It has proposed a rule saying federal law governs and states don’t get a say, giving every judge currently hearing one of these cases something to eventually point to that’s firmer than the CEA’s apparently ambiguous text. For the CFTC, it’s better to walk into an escalating legal battle with a finished rule arguing one interpretation than with a pile of conflicting district court opinions. The comment period on this proposed rule closes July 27.

Will SCOTUS End This?

It would seem that many lawyers watching closely have converged on similar predictions. Pun intended.

Linda Goldstein, a partner at CM Law, says the split among lower courts makes Supreme Court resolution likely. Johnny ElHachem at Holland & Knight argues the litigation and enforcement actions are themselves accelerating the odds of review. Alexandra Fedotova, at White Stone, points out the chicken-and-egg problem between Congress and the courts: the only real fix may be legislation, but Congress won’t find the will to legislate until the Supreme Court rules on preemption.

A court can only construe the categories it’s handed. Congress could have handed it new ones. S.4469, for instance, defines an “event contract” (a contract for the sale of a commodity for future delivery, an option on one, or a swap) and builds a whole apparatus around it: public-interest review, marketing rules, an advisory council with state attorneys general on it. That’s a real answer to what these things are. As with the other bills, it hasn’t yet passed.

Congress has split the problem into three parts: what these things are, what they’re allowed to list, and who’s allowed to trade on them. It answered the third, for itself, and left the rest. The fight that’s actually deciding them is none of those — it’s jurisdictional, and that decision may ultimately settle all three at once. Until it does, a handful of venture-capital-backed platforms will keep setting the odds the public reads on what happens next, accountable to whoever blinks last.

It is a political philosophy based on reasonableness, and especially suited to a democracy where people routinely disagree. It is also intuitive, so ingrained in human behavior that we take it as a measure for what it means to be a decent person. It is the hard-won wisdom of people who have experienced real oppression: tolerance is a small price to pay to establish the rules in a repeated game that lasts for generations.

In How Democracies Die, Steven Levitsky and Daniel Ziblatt warn that as mutual toleration and forbearance falter, political figures stop seeing normal democratic processes as a means of resolving conflict. Loss of power is presented as so catastrophic that it justifies anything to prevent it, including the abuses of existing power.

The nomination of Todd Blanche as Attorney General is a good illustration of the collapse of mutual tolerance and reasonable pluralism. While the philosopher Karl Popper argued that liberalism rested on intolerance of intolerance, what we have instead is a partisan tolerance of intolerance.

More than any other Trump official (with the possible exception of Russ Vought), Blanche represents the logic of intolerance as a guide to the use of public power: The President must be defended and his enemies must be attacked using state power.

It is not possible for such a conviction to guide your job as Attorney General and to embrace democratic norms. To recap, here are some of the ways Blanche has used the office:

• Blanche oversaw the “Weaponization Working Group” that pushed for prosecutions of figures like Senator Adam Schiff, New York Attorney General Letitia James and the Federal Reserve’s Lisa Cook on the thinnest of premises. Blanche oversaw a second prosecution of James Comey for arranging seashells to say “86 47.” He has established retribution against political figures as a new DOJ norm.

• Blanche determined the IRS should never again audit the Trump family, while creating a slush fund for Trump supporters to be paid. A US federal judge has scrapped the agreement, for now, because of the obvious “collusion” between Trump and the government he runs, but Blanche seeks to keep the unprecedented protection of the Trump family in place.

• Blanche boasted of having “cleaned house” of every FBI and DOJ official who was involved in investigating Trump. Nothing about their performance, merely their willingness to follow orders to participate in investigations Blanche regarded as inappropriate, even as he orders FBI and DOJ officials to participate in Trump’s retribution prosecutions.

There are other ways Blanche is unqualified. He owes his status to having served as Trump’s personal defense lawyer. His legal advice reflects his loyalty to Trump rather than the law: for example, he provided advice that led the DHS to ignore a court order not to remove immigrants.

But it is his willingness to engage in such blatant abuses of power to protect Trump while attacking others that should give the Senators the greatest pause. Even the conservative National Review argued against Blanche calling him: “an instrument of Trump’s unworthy and abusive campaign to investigate and prosecute his political opponents.”

The worst of these abuses — the seashell prosecution of Comey, the IRS settlement — came after the Attorney General job opened up, and Blanche scrambled to show Trump how much more willing he was to abuse the law than even his predecessor, Pam Bondi.

Blanche as a manifestation of MAGA is unsurprising and uninteresting. What is interesting is how Republican Senators respond to someone who is clearly making their future selves more vulnerable to retaliation.

Blanche’s past record leaves little doubt about his future record. His statements in hearings did not reverse the impression of a general directing the President’s retribution. For example in response to the question “Do you believe it is the president’s right and duty to order the investigation or prosecution of his perceived enemies, yes or no?” Blanche responded “I can’t answer that yes or no…he is in charge of the Department of Justice.”

Blanche repeated some variation of the line that Trump was in charge of the Department of Justice and so under the law, Blanche had to follow whatever Trump asked for. Just following orders. In The Bulwark, Will Saletan wrote that “Blanche displayed the sort of personality that serves comfortably in an authoritarian administration.”

And yet…most if not all Republican Senators will vote for Blanche. They take seriously his promise of no political investigations even as Blanche refuses to acknowledge the political investigations taking place today. After the hearing, Senator Tillis, who is retiring, told reporters he was “leaning yes” on supporting Blanche.

If Republican Senators believe what goes around, comes around, why would they support someone undermining the rule of law?

One answer is an assumed asymmetry. Republicans cannot imagine a world where Democrats would nominate an Attorney General who would weaponize the law as Blanche has.

They choose Blanche on the assumption that Democrats will always choose a Merrick Garland — an Attorney General so studiously aligned to the perception of fairness that he investigates the Biden’s son and fails to effectively prosecute Trump.

It may be that Republicans looked at the Biden administration and concluded that there will be no real consequences from the current era. They may be right. But I’m not so sure. The abuses of the second Trump administration are much worse than the first, and the Republican tolerance of it has been higher than before. Under these conditions, the Democrats have many more grievances to nurture, and scores to settle.

The assumption of asymmetry feels like a bad long term bet. Recently Matt Yglesias bemoaned the tendency of Democrats to favor fighters. Whether or not it is bad politics, the rise of the fighter seems like a predictable response to a system where Republicans are engaged in more extreme actions, and moderates seem unable to constrain them. As virtually all present Republicans create an environment where Democratic fighters are empowered, future Democrats are less likely to accept institutionalist like Garland, putting future Republicans in peril.

A second answer is that the MAGA distaste of tolerance, empathy and reasonableness has become so embedded into their worldview and identity that they cannot think of politics as a repeated game.

In no small part this reflects that the MAGA movement and the Republican Party is an extension of Trump’s political id. His art of the deal preaches the virtues of winning, screwing others over, refusing to pay his contractors their fees or his government their taxes because you are smarter than them.

There is also the pre-Trump strain of conspiracism in the American right. If you saw Dwight Eisenhower as a communist agent, you were apt to see enemies everywhere. The paranoid style sat comfortably with racism and xenophobia. Trump has expertly nurtured these grievances, starting with birtherism or his call to lock up Hillary Clinton. In this context, normal political figures become demonized as enemies who must be punished. Imagined persecution become the basis for actual prosecutions.

There are strains of this loss of perspective outside partisan politics but among Trump’s most powerful supporters. Recent essays by Henry Farrell and Tim O’Reilly trace the maddening of Silicon Valley leaders. They no longer resemble the merchant who wanted a government that is stable, predictable and fair, and managed his role in society accordingly. They are beset by a quasi-monarchical mania, where they see themselves as princes whose sovereign powers must be protected, who should not be accountable to markets or governments. They are not anchored in a recognizable belief system such as classical liberalism or religion. They bemoan empathy as the “fundamental weakness of Western civilization” rather than the basis by which that civilization stopped tearing itself apart.

The interregnum between Trump’s first and second terms hardened these views. He and his supporters saw themselves as victims. Indeed, that was a criterion for being part of the inner circle, with the Shibboleth being an unwillingness to accept the results of the 2020 election. The popularization of the term “weaponization” of government came from…Russ Vought, one of the high priests of weaponizing government.

Buoyed by this sense of imaginary victimhood, the Trump administration and followers abandoned reasonableness. They determined that the purpose of power was not to serve the public but to protect (and sometimes enrich) themselves, and to punish their enemies. Every so often, even the claim of victimhood fails. There is no serious claim that Lisa Cook did anything to injure President Trump: he merely wants her out of the way, and Blanche is there to help.

This logic is inconsistent with a liberal democracy, and the role of politics as a means of resolving conflict. And again, it only makes sense either if you think the other tribe will practice forgiveness and forbearance once they return to power, or you think you will never lose power, that your tribe will always be dominant.

Trump is 80 and I doubt he has a long-term view of the world. The Supreme Court granted him legal immunity and his immense riches buy him other legal protections. As president he can preemptively pardon his inner circle and family before he leaves office (and dangle that pardon in the meantime to ensure their loyalty).

Nevertheless, he still seems nervous enough to do all he can to violate norms and even the law to control the outcome of the midterm elections, which includes directing Blanche to investigate debunked electoral conspiracy theories, and telling Americans they live under the a corrupt and rigged election system. Many are speculating this will result in Trump declaring a state of emergency to rig the elections himself.

Republicans can see the polls. They are happy to go along with more gerrymandering, more election restrictions. Mike Johnson, the Speaker of the House emphasized to a conservative audience the centrality of holding power to avoid routine congressional investigations: “I run the protection program. I’ll take care of you.”

This is the ethos of a party that knows it is unpopular and is using the threat of ordinary democratic accountability to motivate its members towards anti-democratic action. It is not the ethos of people confident they can stay in power, as long as basic electoral processes remain reasonably competitive. And by pursuing more extreme actions, they invite an extreme response.

Let me put it in simpler terms. As long as Republicans support candidates like Todd Blanche, they make it much more likely that the next Democratic Attorney General will look much more like Todd Blanche and much less like Merrick Garland.

Unlike 2020, Democrats have looked at other countries like Brazil and Korea, where attacks on democracy have triggered jail sentences for political leaders. In Hungary, the supermajority won by the opposition to Orban is dismantling his seemingly unassailable empire of political, business and media control of society.

We are just 1.5 years into the second Trump administration, and a lot of Americans are angry. There is no reason to assume this will dissipate.

Republican Senators considering voting for Blanche now might tell themselves, my vote won’t make a difference. Because the Senate previously confirmed Blanche as Deputy Attorney General, he can serve as acting Attorney General indefinitely. But in fact this creates all the more reason to take the symbolic vote to oppose him.

Even if the public does not pay attention, their vote sends an important signal to their Democratic colleagues about whether mutual toleration and forbearance till exists. If they vote for someone weaponizing the legal system now, they are encouraging Democrats to do the same when their turn comes around. That is how norms work.

Albert Hirschman provided a typology that has been widely applied to this situation: exit, voice, and loyalty, which some have expanded to include other categories such as neglect, resistance and sabotage. One problem is that these typologies only consider existing employees, not potential employees – those outside government who are motivated to serve society and to influence public policy.

So, how do potential entrants respond?

This question matters for the government’s ability to recruit new employees during periods of democratic backsliding or sharp polarization. The table below, drawn from a recent article with my colleagues, develops a conceptual framework that extends Hirschman’s typology by comparing the array of possible responses of existing civil servants with those of potential recruits.

Mirror-image typology of actual and potential employees’ explication of their choice to work or forgo work in government amid democratic and bureaucratic decline

Academics seldom acknowledge the possibility that individuals’ choice to seek work in government is influenced by their partisan beliefs and perceptions of the trustworthiness of the party and leaders in power. The established understanding is that some people prefer work in the public sector over the private sector due to their motivation to serve society and their preference for secure, stable employment. This calculus is supposedly shaped by a long-term view of one’s career, which discards the question of which party is heading the government at the time of entry to the civil service.

Partisan support for and trust in governing politicians as a factor shaping the choice to pursue government work becomes logical when viewed against the backdrop of three interrelated global phenomena that deter opposition supporters from pursuing government employment:

• First is the increase in “affective polarization,” namely, individuals’ self-identification as members of a partisan camp and animosity towards the leaders and supporters of rival political parties or coalitions. When polarization is high, evaluations of government as an employer and as the place to serve society are more likely to be processed through a partisan lens.

• Second, radical-right populist leaders and parties are assuming positions of government power in some countries and have the potential to do so in others. The animosity that supporters of rival parties harbor toward these leaders and their parties likely deters entry into government.

• Third, polarization and populism are often accompanied by democratic backsliding. That is, political leaders’ subversion of the electoral system, undermining institutional checks and balances on their power, and advancement of illiberal policies.

  Studies across the globe show that democratic backsliding also involves intense politicization of the bureaucracy and the sidelining of career civil servants. Potential recruits who oppose this process are likely to see it as a signal that, if they were to join the government, they would lack influence, might find themselves controlled by unprofessional managers, and would be forced to implement harmful policies.

In a series of studies in Israel, my colleagues and I have been examining the magnitude of the effect of individuals’ partisan support for the governing party and its leaders on their entry to the national civil service, its variation, underlying logic, potential ramifications, and mitigating factors. The findings cast light on what happens to the ability of a government to recruit when society is politically polarized and the bureaucracy becomes more politicized amidst democratic backsliding.

Share

We find that potential employees are mostly disinclined to work under politicians they distrust, with some waiting for a future political turnover while pursuing work in nonprofits, other levels of government or firms that consult the government, and others skipping public service altogether. While exit, voice, loyalty, and resistance are strategies for current employees, potential employees tend to choose between waiting outside and non-entry.

Israel, like the US, is highly polarized, divided between supporters and opponents of Prime Minister Netanyahu and his radical right partners. In addition, the past three and a half years have involved a rapid deterioration of Israeli democracy (which has never been perfect) and its bureaucracy. In January 2023, shortly after the current government’s inauguration, it introduced the Judicial Overhaul plan, which, like other instances of democratic backsliding worldwide, was intended to weaken judicial and legal checks and balances on executive power.

Following Hamas’s October 7^th, 2023, attack, the government officially halted its advancement of the Judicial Overhaul. In practice, however, alongside the war in Gaza and other fronts, it continued its internal attack on the judiciary and the Attorney General’s office, and intensified the politicization of the civil service, the police, and the security forces. The undermining of these institutions further enabled some of the government’s far-right agendas, most notably in the occupied territories.

How do potential recruits respond to this context?

Taking a first stab at this question, together with Raanan Sulitzeanu-Kenan and David Levi- Faur, we conducted a two-wave survey of 21-to-30-year-old Israelis in August and September 2023. We measured respondents’ partisan alignment (or misalignment) with the government coalition, perceptions of the Judicial Overhaul, and attraction to central government work.

We found that when offered a job with objectively equal material conditions, supporters of the opposition were much less likely to consider work in a central government ministry compared with the business sector. Moreover, among the Jewish majority, these partisan differences were fully attributable to opposition supporters’ perception of the Judicial Overhaul as a threat to Israel’s democracy.

Approaching the same respondents shortly after the October 7 attack, we found that opposition supporters’ relative disinterest in working in the central government remained intact. Moreover, as shown in Figure 1, the increase in their motivation to serve society mostly boosted their attraction to the nonprofit sector as a substitute for government work (which attracted supporters of the coalition).

Figure 1: Coalition and opposition supporters’ attraction to work in government (red), business (green), and nonprofits (blue), before and after Oct. 7.

In a subsequent interview-based study, conducted towards the end of 2024, Reut Marciano, Ilana Shpaizman, and I compared the career choices and plans of actual civil servants and potential recruits who recognized and opposed the government’s undermining of democracy and of the quality of the bureaucracy.

We found that very few potential recruits were willing to enter the central government to “fix it.” For most respondents, as reflected in the following quotations, entering government entailed being complicit in an institution taken over by politicians who do not pursue the public interest.

I can’t seem to make myself apply [for positions in the civil service], because there’s not a single government ministry that I would be happy to work in… not a single minister that I can trust that my professional judgment will be used for the good.

[I would be willing to work for] any government..., which is not headed by [Prime Minister] Netanyahu... [any government] ...which has a main aim of serving the public good. It is not about ideological affinity...But it has to be a government, which has the purpose of advancing the public [interest].

Share

However, we unraveled an important variation among those who chose to forgo entry. Whereas some of these respondents sought an alternative career path in the private sector, others conveyed their position as waiting to enter the civil service in the event of political turnover and the election of a trustworthy government.

Waiting involved cultivating skills through education or work that would be valuable for a future civil service career. These respondents argued that one cannot and should not enter now, since politicization is omnipresent and bureaucrats cannot serve the public interest. They explained their outlook as waiting rather than pursuing an alternative career, given their attachment to the nation and commitment to rebuilding Israel’s democracy and the civil service.

Finally, in a recent study (Clareta Treger, Sharon Gilad, and Ilana Shpaizman), we examine whether and how it might be possible to attract recruits who distrust Netanyahu and his coalition government. We examine two potential factors that may offset low trust in elected officials: respondents’ trust in civil servants and the material attributes of government jobs. The study benefited from our collaboration with the Israeli National Student Union, which distributed a survey on our behalf to BA students across higher education institutions.

We found that trust in civil servants is distinct from, and higher than, trust in elected officials, and that both trust in civil servants and the material attributes of government jobs shape attraction to the central government. Yet, neither factor can significantly attenuate the negative effect of low trust in elected officials. In other words, distrust in Netanyahu caused many to avoid entering government service. Instead, those who trust civil servants, yet distrust Netanyahu and his ministers, were attracted to jobs in local government, which they likely perceived as shielded from central political control.

Overall, the above findings, drawn from different samples and methods, suggest that Israeli opposition supporters are shunning central government work.

Underlying the choice of potential Israeli employees not to enter central government is distrust of elected officials and disdain for their attacks on democracy and the bureaucracy, with some further repelled by the Gaza war. If committed to serving society, some of these individuals seek work in sectors and organizations insulated from national political control, such as nonprofits and local government, and, possibly, independent agencies and regional government offices.

One implication of these career choices is gradual partisan homogenization of the public service. The longer the current administration, which opposition supporters view as untrustworthy, remains in office, the more politically homogeneous the civil service becomes, not only due to politicization but also because political opponents choose to avoid government work. Since partisanship in Israel and elsewhere strongly correlates with other demographic factors, such homogenization means the bureaucracy becomes less representative of the public and probably less responsive to its diverse interests and needs.

How relevant are these findings beyond the Israeli case? Recent findings from the US suggest a similar pattern. Studying a sample of American job seekers, Susan Miller and I find that job seekers are less attracted to work for a state government when the Governor has a different party ID from them. We also find that job seekers are less attracted to public sector work under conditions of polarization between political elites and when governors undermine democracy and bureaucratic meritocracy.

Importantly, our findings from Israel suggest that in the event of political turnover, former opposition supporters may be attracted to central government work, provided they believe that democracy and the bureaucracy can be revived. As an Israeli citizen, I long to contribute to the study of such a process if and when it occurs, and look forward to relevant comparisons.

Bio: Sharon Gilad is a full professor at the Hebrew University of Jerusalem and the chair of its political science department.

RELATED

In 2026, nearly all of the services city hall offers New Yorkers are dispensed through digital platforms…New Yorkers should not be forced to navigate systems that are confusing, outdated and burdened by bureaucracy. The PIT Crew turns that model on its head. These teams will move quickly and deliberately to solve real problems, make City government easier to use, deliver for working New Yorkers and advance our agenda of affordability and public excellence.

This is a big deal for three reasons: First, it tells us much about Mamdani’s closely-watched approach to governing. Second, when it sometimes feels like tech is just making our lives harder, it offers a model for how it can be used to make life easier. Third, it provides a model of how progressives can reclaim tech in a post-DOGE era.

When is a technology office “not about technology”?

PIT Crew will be led by Lisa Gelobter, the Chief Technology Officer of NYC. Committing fully to the automotive bit, the announcement took place at a Go-Kart race track in Coney Island.

People we spoke with in the NYC government emphasized that PIT Crew was not primarily about tech, bit on delivering on Mamdani’s core priorities on housing, childcare, consumer and worker protection. Mamdani also framed the investment in broader terms, emphasizing improving the experience of government.

But this work is not just about making city government more efficient. It is about improving the way that government works…Every confusing step and every dead-end portal makes it more difficult for government to deliver for New Yorkers, and for New Yorkers to trust that government is working to deliver for them. Every straightforward experience New Yorkers have on city platforms rebuilds that trust and our PIT crews will help us do exactly that.

Share

PIT Crew helps to answer one of the puzzles about Mamdani: why is he investing so much capital in public sector capacity?

A conventional wisdom is that politicians have scarce political capital and investing it into fixing existing government services, rather than claiming credit for new policies or initiatives, isn’t politically fruitful. This is one reason why our infrastructure is decaying. The smart politician goes to the ribbon-cutting for a new highway, but not fixing its potholes.

But here is Mamdani talking about filling potholes…

Mamdani is a smart politician with charisma to spare. And yet he is spending it on 311 calls…

As public policy professors, we love this. This sort of messaging emphasizes the dignity and value of public work. It brings joy to public policy implementation. But we are a pretty niche audience who cares about good governance. If you read this newsletter, you are probably in that audience!

Why is Mamadani so invested in the nuts and bolts of administration? He closed his speech announcing the PIT crews with this:

In the words of Jeff Gordon: I’ve never believed in luck. If you perform well, execute, have a good race car, and a good race team, you’re going to increase your chances dramatically of having better results. City Hall is building a team that performs well and executes well. You better believe that our chances of delivering well for New Yorkers are increasing dramatically as well.

And this is a key point: Mamdani is tying state capacity not to some amorphous end outcome, but to the most visible goals of his administration. For example, here he is talking about reducing administrative burdens to speed up the building of housing, a central theme of his campaign.

This is interesting for multiple reasons. First, it represents a form of civic education. One of the most charismatic politicians in America is using his political honeymoon to explain in simple and accessible terms how the public sector works. Even before Trump, political explanations about public administration have typically been used to bash the bureaucracy, not to uplift it.

Second, it is a form of accountability. Mamdani is asking us to hold him accountable to outcomes. And he sees those outcomes as directly tied to administrative capacity. On high-stakes issues he is betting on City employees rather than consultants. City officials set the goal as: “Deliver exceptional products and services in the hands of the public that raise expectations for what government can deliver.”

What does ‘exceptional products’ mean? Their first project is the City’s new Click to Cancel rule that in theory makes it easy for people to get out of annoying magazine subscriptions or gym memberships that private companies deliberately make onerous. Having a law on the books is great, but NYC is promising to make sure Click to Cancel works, creating a platform where violators of the law can be easily identified.

There is also something else. A demand for excellence: In his inaugural speech, Mamdani said:

For too long, we have turned to the private sector for greatness, while accepting mediocrity from those who serve the public. I cannot blame anyone who has come to question the role of government, whose faith in democracy has been eroded by decades of apathy. We will restore that trust by walking a different path—one where government is no longer solely the final recourse for those struggling, one where excellence is no longer the exception

An easy way to dismiss progressives, and especially democratic socialists, is that they have unrealistic ideas that they can’t implement. Mamdani seeks to disprove this claim. Writing in the New York Review of Books, Corey Robin argued that Mamdani’s emphasis on “excellence” was a rhetorical innovation, one designed to reclaim the rhetorical high ground on good governance for progressives.

Betting on state capacity rather than contracting out

If you are a good government wonk, this innovation is interesting for other reasons. reflecting how thinking about tech has evolved. Traditionally, the conventional wisdom since the 1990s and before was that governments could buy tech products like an off-the-shelf product. This led to a massive turn to contracting out, which was great for consultants but bad for government capacity. The outsourced approach often cost too much, delivering too little and too late.

Even when tech teams were created they were unable to reverse the pattern towards outsourcing. Too often the existence of a tech team became a symbol of governance without a clear purpose. This was perhaps one of the failures of early government tech units, as Merici Vinton has written. “The Digital Service was Never the Point” say Jen Pahlka and Kelly Taylor.

The lack of strategic vision was telling. The federal US Digital Service was treated as firefighters to be called in when a private vendor’s product was in trouble (see, for example, FAFSA). The origins of USDS and 18F came from the need to fix Healthcare.gov, and that emphasis on fixing rather than building products was constantly a structural problem. Towards the end of the Biden administration, we saw the emergence of some exceptions, most notably Direct File (until the Trump administration killed it).

It would be very easy for critics of Mamdani to reflexively say “well, Mamdani hates the private sector, and everyone knows government tech is garbage.” Easy, but wrong.

This is not a private vs. public debate. What people who know tech and government have been screaming for years is that building good tech needs in-house capacity, even when you are using contractors. It requires the government owning the design, development and delivery of technology, relying on rapid iteration to fix problems in a way that is impossible when contractors are running things. A senior PIT Crew official, Luke Farrell, has published here on the ways that private firms overcharge government for supporting digital services, while his own work at the US Digital Service revealed how small capacity investments can have big payoffs for government.

New York is not alone. Recently, the state of Colorado blew up their approach to IT. One of their takeaways was about how outsourcing proved a barrier to modes of working that led to good digital products.

Outsourcing often becomes a barrier to doing tech right for structural and incentive reasons. The contract becomes the barrier to rapid coordination needed to build good tech quickly, and the incentives of profit-seeking contractors are to maximize revenue. But it’s also true that the traditional IT model of government has been a barrier to doing tech right for cultural reasons. Turf battles, caution, and a lack of urgency, failure to embrace new modes of operating, inability to recruit top talent are all real government problems that need to be fixed.

Innovations like PIT Crew seek to solve both sets of problems. NYC should be able to attract top talent and their job descriptions advertise accelerated hiring. The sense of urgency and purpose that these offices create can change the culture of government. Having government ownership of the product should save money and improve quality.

There are a lot of big ifs in Mamadani’s big bet on tech. A reasonable question is whether agencies engage collaboratively with PIT Crew. Another risk is that PIT Crew will consume energy and resources, but agency level IT culture remains unchanged, making it an island of excellence that does not scale to the rest of government.

But the creation of PIT Crew matches something he promised in his inauguration speech: “we will govern expansively and audaciously. We may not always succeed. But never will we be accused of lacking the courage to try.”

Reclaiming tech for good

NYC is in a budget crunch, and Mamdani is betting that an investment in internal tech capacity will not just save the city money, but also create value for citizens. During the press conference announcing PIT Crew, a reporter asked the Mayor how he could justify the $5 million in new spending.

Mamdani offered two revealing insights. He pointed out that just one fifth of the new funding will be used to enable the implementation of the Click of Cancel rule, which is expected to save New Yorkers $160 million. In other words, his accounting considers not just public dollars spent, but citizen resources saved. Tech can create public value. Second, he emphasized reducing the psychological costs of these hassles “the time and stress” that comes from private sector hassles.

This thinking echoes another late Biden-era initiative, repealed by Trump, which is that government has a legitimate regulatory role in regulating how private businesses weaponize administrative burdens against customers.

It also reflects another theme, which is how tech is viewed in progressive spaces. For a long time, progressives saw tech as a force for good. Tech workers connected their work to a broader purpose, using terms like ‘civic tech’ and ‘public interest tech’ to describe a community of people who saw tech as a way of solving big public problems.

DOGE, the rightward shift of Silicon Valley, and the rise of AI has dampened enthusiasm for tech on the left, and fueled concerns about potential abuses. Mamdani’s PIT Crew feels like an effort to reclaim tech as a progressive force, to explain how it can be used to help the public.

As Alondra Nelson, former Acting Director of the White House Office of Science and Technology Policy, noted:

Done right, digital government can help millions of people save hundreds of millions of dollars, and hours and hours of time. Done wrong, it can be costly, useless, discriminatory or even dangerous. How a City interacts with its people shouldn’t be outsourced. It should be local and accountable.

Last year, Don invited me and other former public servants to speak to his University of Michigan students about a devilishly hard question: What do you do when your leadership is acting in ways that feel at odds with your values?

On the surface it was about choices like when to leave the government, when to speak up, and how to pick your battles.

But looking back, I realize we were circling something more fundamental: how to navigate complex choices as we serve the public.

My new book, Sometimes You Should Be Late, which I wrote after leaving a 15-year career in government, has something to say about that question. It argues that one of the most important shifts we can make in our lives is to slow down, not for its own sake, but to help us better align our choices with our values.

Nowhere is that pause more important than in the high stakes world of public policy, especially in these disorienting times.

Government’s Bias Toward Speed

I laugh a little when people joke about government moving slowly. For every DMV sloth in a Pixar movie, there are countless public servants working with haste.

It makes sense: Many are drawn to public service to solve problems, seize windows of opportunity, and serve the people we swore an oath to. Those instincts are beautiful, but they can also become a trap.

I would know. Over my government career, first with the State Department, then in the Senate, and finally working on the President’s Management Agenda at the General Services Administration, I often equated speed with effectiveness.

Sometimes that looked like assuming emails from the White House or ambassadors needed instant replies, clinging to deadlines as if they were written in stone, or moving too quickly past stakeholder concerns.

And it also meant that too often I accepted the default way of doing things because I didn’t slow down long enough to recognize the choices in front of me. Although DOGE has done enormous harm, it has also forced questions about the inner workings of government. Which processes serve a real purpose? Which persist simply because they’ve always been there?

If I’m honest, looking back at my career, I should have slowed down to ask more of those questions.

Why Policy Needs Slowness

That’s why I’ve come to believe that pauses are so much more than personal choices or wellness hacks. They are policy tools. They help us become more aware of our assumptions and biases and give us a better chance of responding wisely to what a moment requires, rather than simply reacting to it.

An example: last year when the Fork in the Road and mass cuts were happening across federal agencies, like many others I was angry and wanted to do something. I decided to write something publicly but most of the outlets I sent an article to wouldn’t publish it anonymously.

But before responding with a rage-inspired “yes” to publishing my name, I took a deep breath, let the anger settle, and decided I wasn’t ready for that exposure. So instead I asked Don to publish my “Forced Fork” essay anonymously in these pages, which allowed me to leave on my own terms when I was ready.

Pause for the Public

Pausing creates conditions for better judgment. In our daily work, it can help us distinguish between false urgency and the real kind, understand whether we are resisting change because it conflicts with our values or simply because it is unfamiliar, and whether we are keeping our heads down because we’re playing the long game or because it’s convenient.

With the lens of my new book in mind, perhaps this is what I should have said to Don’s students: public servants will continue to face complex choices, not only in this administration, but in future ones as well. Navigating those choices skillfully is the work of public service.

Sometimes slowing down will tell you to speak out and stand on principle. Other times it will tell you to pick your battles and be open to a new way of approaching a problem. Without pausing, it’s hard to know the difference, and how best to serve the public in that moment and beyond.

If these ideas resonate–or if you’re skeptical–I’ll hope you join the conversation by checking out Sometimes You Should Be Late.

Alex Snider is a writer and facilitator who draws on 15 years in public service, including as a diplomat and in domestic policy roles, along with certifications and a decade of practicing and facilitating meditation and group processes. He is the co-founder of Mindful Fed, a cross-government mindfulness community, and writes the Substack Slow Mindfulness.

From Don: Until Trump returned to power, a group of civic-minded technologists played a growing role in the federal government. DOGE purged them from public service. The DOGE disaster has now been memory holed, without even a final report as the Musk era of DOGE that formally ended on July 4th. Meanwhile, civic technologists have been having frank conversations about both the past and the future, asking what they did right, and wrong. Here, Merici Vinton, one of the leaders in that field, tells us what they learned.

The services and digital teams that were quietly making government work better - simpler tax filing, faster benefits access - have been or are being dismantled. What replaces them matters.

The second Trump administration has caused a major disruption to the federal government - deeply impacting the public services that Americans receive and the ways that services are delivered. This is felt particularly strongly in the civic technology field - which has sought to improve the relationship between the public and government institutions through the use of digital technology, especially for public services (“digital services”). The loss of digital service organizations has fundamentally shifted the motivations, incentives and capacities of the people designing the technology that powers government services - meaningfully affecting how and how well these services operate.

Recognizing this moment where we can challenge the old and look for the new, the Federation of American Scientists and Georgetown’s Beeck Center ran a series of retrospectives, or retros, to capture past lessons about what did and did not work in digital service delivery, and how we can chart an ambitious path forward in this new reality. We spoke to over 100 digital government practitioners from across the country.

These are the results, as well as where we think the field should move.

This is the moment for ambitious change. The digital government field has an opportunity to build a more responsive and resilient government by pushing into new frontiers, with new tools, approaches, and even organizations that don’t exist yet. While the pace of change can be overwhelming, there is something clarifying about a moment when people are questioning the value of institutions themselves.

When we set out to do the retros, we didn’t know what to expect. The field had undergone a massive rupture with the US Digital Service becoming DOGE and the shuttering of 18F, while other agency level digital teams were shuttered or reshaped at the federal level. Meanwhile, at the state and local levels of digital service delivery we’re hearing from government officials who are investing greatly in digital teams, often with excellent results and cost savings.

I am part of the digital service field, as an early employee of the Consumer Financial Protection Bureau and then as an architect of IRS Direct File. When engaging with this community, I had hoped for some obvious patterns or quick wins. I was surprised to discover that the primary pattern was: there are no patterns.

The core diagnosis: there Is No Shared Mental Model for what makes ‘good’ digital service teams, initiatives, and results

Across every conversation in our Digital Services Retros project, the consistent pattern was the lack of a mental model of success. For example:.

• People came to work on digital service teams and initiatives with dramatically different expectations.

• Government agencies didn’t know how to partner with digital teams.

• Digital teams didn’t know how to navigate agency bureaucracies.

• The digital government field lacks a shared vocabulary and definition of success for service delivery.

The lack of throughline connecting practitioners across organizations - in government, the private sector, and civil society - means there are not shared career pathways, methods of completing the work, or expectations for outputs. Which means it’s incredibly inefficient - people joining expecting one job, discovering one wildly different; agencies, also often immature in their own digital delivery, have no reliable patterns or expectations when partnering; and ultimately no shared vision for what good product development and delivery looks like.

The result is that a huge amount of time and effort is spent negotiating different expectations and correcting misunderstandings.

That diagnosis breaks into three interrelated findings.

Finding 1: There Must Be a Strategy and Explicit Theory of Change

There is no theory for how digital services contribute to societal outcomes, or government-wide strategy for leveraging digitalization to achieve those outcomes.

No shared strategy of work across the federal teams – USDS, 18F/TTS, OFCIO – led to confusion, competition for resources, and a focus on team goals rather than shared agency- or government-wide outcomes.

Strategy needs to exist at multiple levels: the highest level (how digital and service design contribute to outcomes — better SNAP access), the government-wide level (priorities and process), and the organizational level (the logic for how different team types work together).

The stakes of having no strategy at each of these levels are higher than ever. Without a clear position on AI for example, governments won’t get their tactics right - they’ll spend their time debating tools instead of outcomes. A strategy doesn’t require consensus, it is a signal of what you value.

The digital service community should be articulating an outcomes based strategy at each level for what we want to achieve in 5-10-20 years; and defining foundational elements that make a strategy effective: building user-centered government and how to reform our institutions to become modern, responsive organizations. Smart policymakers should look to partner with this community to codesign the future.

It’s time for the field to have a vision, articulate it, and go for it.

Share

Finding 2: As is, the Digital Government Field Cannot Scale

Another theme that emerged was a shared concern. Individual contributions of any one team could not scale to influence how digital services were delivered throughout the rest of an agency. While teams of talented professionals could turn around quality products, ultimately teams struggled to turn their hard-won bright spots into scalable models of working across federal agencies.

The government digital services field’s M.O. is to rely on a small and scrappy pipeline of talent often via special and temporary hiring authorities. Here are three recommendations to change this:

• Make it as easy to hire technologists as it is to hire lawyers: Endless energy is spent on hiring technologists through a series of flexibilities and exceptions in lieu of real, streamlined repeatable career pathways. We have hacked ourselves into government, but have not fixed the machine.

• Make it someone’s job to scale change: Everything – from hiring, to ATO (authorization to operate) processes, to onboarding and everything in between – is a one-off, making lasting change within and across agencies near impossible. The fix: a scale team, empowered and expected to make the best practice the default practice, not the exception.

• Define success benchmarks across government: The field also lacks a shared definition of “good.” Teams need to define their own “industry standards” for service levels and delivery: benchmarks for downtime, user satisfaction, uptake among eligible populations — flexible, not rigid, but enough to let teams and leaders gauge how they’re doing.

• Expand field resources: Shared vocabulary, training, and methodology — building on bright spots like Technologists for Public Good and Code for America — would make practitioners more mobile across institutions and raise the quality of the field as a whole. We recommend further exploration to define the gaps and what should be filled by teams in government, or our recommended scale team, or forming new professional associations or non profits to support the wider field.

Finding 3: Build the Right Team for the Right Problem

Another pattern emerged across all levels of government: digital service teams were set up to add capacity to any number of solutions. The early strategy was: land the team, expand to solve new problems. The unit of success became the team itself and an agency was “mature” if it had one. Most digital service teams sit centrally under a principal, CIO, or COO, with authority that depends on political capital rather than legislation, and the same skill sets get hired regardless of the problem.

The fix: organize around the outcome of a service rather than a functional or organizational boundary. This is the Service Team, built on the UK’s Government Digital Service approach and Kate Tarling’s definition of a service: crossing silos, tied to the institution’s goals. The product operating model gives the operating mechanics - funding, staffing, day-to-day running. The service model gives the unit of organization - the user’s full journey.

Direct File is the proof of concept: a single accountable owner empowered to make day-to-day decisions, and a service team that joined up customer support and product work.

Building end-to-end service teams is not for the faint of heart – achieving this model of embedded technical and program teams, working collaboratively on an end-to-end journey, fully empowered to update processes and ways of working as they go is the mature end of the digital government delivery spectrum. But it is included here as it not only achieves better results for the end user, it also helps the agency and staff transform as they work in a new, fully digital, collaborative manner. For governments who want to make tech transformative, this is the path to take.

Different models for different problems.

Below are models that emerged from our retros, each with distinct purposes, skill requirements, and conditions for success. These aren’t mutually exclusive; a mature digital organization will need more than one at once. The service team is the north star. The others are legitimate, but they should be chosen deliberately and not defaulted into.

Additional models, include:

A Forward Looking Agenda for Digital Service Delivery

Federal capacity, or “The Department of Digital”

Federal digital and delivery capacity will continue to be a necessary piece of a strong, delivery oriented government. The question is – now that the US Digital Service has become DOGE and 18F is dismantled, what should the future model be?

It’s time to explore a Department of Digital – a strong, centralized digital capacity – capacity to deliver on presidential priorities, set standards, and build shared platforms - modeled in part on Germany’s Ministry for Digital Transformation and the UK’s GDS. We need standards covering AI use, service design, budget controls, and service performance. We also need shared platforms like Login.gov, a payment platform, and income verification infrastructure that any agency can plug into instead of building from scratch.

At the agency level, there are multiple models to support service delivery and leaders should support the model that meets their needs. However, almost universally we recommend that agencies should explore the service model for their key services.

At the state level, Colorado is leading the way. It recently introduced a revamped Colorado Office of Information Technology to manage digital service delivery. Other states should follow their lead.

A Public Option Middle Layer

There are countless examples of large IT vendors building the same product, poorly, for 50 states. This model can and should be reimagined – moving towards a vision where governments can opt-in to a shared public infrastructure. We need shared public infrastructure around key legislation and universal platforms. This is a case where states, as laboratories of democracy, would benefit from co-operating with one another.

Shared platforms and infrastructure can provide government teams across the country with additional options for delivery at a national scale. This ‘middle layer’ would look at services that are required in multiple jurisdictions or infrastructure necessary to implement new legislation. Such platforms could exist as a public option – government-provided services that coexist with one or more private options.

Governments shouldn’t have to build from scratch: Modern income verification infrastructure only needs to be built once and made available to jurisdictions that opt-in and could be used in a countless number of ways, from federal scholarship applications to SNAP. Governments at all levels lose money to payment processing fees. A public option payment platform would cut costs and add a new competitor to the market, saving taxpayers money. Universal moments, such as birth and death registry, could be built centrally and deployed locally.

These delivery challenges have been too onerous to tackle. States are often reluctant to share data and platforms. Some jurisdictions have begun planning shared procurement strategies, but radically different, outcomes-based delivery will take reimagined governance and significant political will.

At the top. I said it was a time for ambitious change. This is possible. What we discovered in this research is not only brilliant ideas, a passionate movement, but also the engagement and energy needed to pull this movement forward into a more impactful, more confident future. Now’s the time to get started.

Subscribe now

Merici Vinton is a Senior Fellow at the Federation of American Scientists. She was an original architect of IRS Direct File and an early employee of the Consumer Financial Protection Bureau.

The Digital Retros project was made possible thanks to the Federation of American Scientists and Georgetown University’s Beeck Center. The project team includes Faith Savaiano and Kirsten Wyatt. Thank you to Laura Sigelmann, Leya Moshin, Loren DeJonge Schulman, and Lynn Overman.

The post above was reported by independent tech journalist Matt Binder and is now apparently deleted. However, it is not an anomaly. Gebbia’s feed is full of far-right MAGA messaging, most frequently falsely portraying immigrants as violent criminals who threaten natives. He also reposted the John Birch Society just a couple of days before.

Its not just immigration. Here is a post he made about how DEI is destroying merit. This seems to be based on a right-wing criticism that Caitlin Clark is being targeted by Black players. Using basketball as an example of how Blacks are exploiting DEI to deny White merit is the kind of thing you do when you’ve cooked your brain on X. He is one more example of life under our clicktatorship.

What is the John Birch Society?

In his defense, maybe Gebbia does not know the history of the John Birch Society. So lets do a brief recap. It best represents what Richard Hofstader described as the paranoid style of politics of the cold war era. The Birch Society is the grandaddy of conspiracy theories like Birtherism, QAnon, and the Big Lie that Trump won the 2020.

How bad was the Birch Society? It’s leader claimed that President Eisenhower was a secret communist agent.

In one telling of American conservatism, a critical moment is when William F. Buckley Jr. denounced the Birch Society in the 1960s, pushing it to the fringe of the movement.

But the Birch Society never went away. The taste for conspiracism has always existed, and Trump just pushed into the mainstream. Birch Society ideas are also very much in vogue: opposition to the UN, civil rights, the Equal Rights Amendment, and federal power generally. The Birch Society’s anti-fluoridation stance also presaged RFK Jr.’s efforts in this area. Anything they disliked was labeled as a communist plot.

The Birch Society has not changed. It’s views are the same. What has changed is the widespread embrace of these views. Early backers included the father of the Koch brothers, and Harry Bradley of the Bradley Foundation, forces which still shape American politics today. What has changed is that the rich men aligned with it are now running our government.

The Society’s relationship with race is especially troubling. They argued that extending equal rights to Blacks were was part of a communist plot, that Blacks in America already had "complete freedom."

The immigration position of the Birch Society is standard Great Replacement theory with some even more conspiratorial twists: shadowy elites are using migration to make the US become part of something called the the “North American Union”: “This plan is supported by the establishment-oriented leadership of both parties, along with the Deep State and cultural Marxists” according to the Society website.

This is what Gebbia is promoting. If this isn’t crossing the line, then where is the line?

Update: after this post went live, a spokesperson from the Birch Society reached out me to say the following:

You could have been a bit more original in your criticisms. These talking point retreads have all been disproven multiple times. But you would have had to do more than a Google search to get those. We are just happy so many people are agreeing to things we have said for nearly 70 years.

Share

Gebbia’s Complaints About Housing are Hypocritical

Gebbia’s Birch society post blames immigrants for rising housing costs. Given that a lot of people blame Airbnbfor unaffordable housing, this seems like a rich guy who scapegoating immigrants for a large social problem he has helped to create.

What does the evidence say? In a nutshell, both immigrants and Airbnb have contributed to increased housing costs, but the size of the effects are generally seen as small, and immigrants are also directly involved in expanding housing supply.

A bit more on immigrants: A seminal research paper by Albert Saiz suggested that immigrants increases costs by about 1% when they increased the population by the same amount. Later work questioned that estimate as being too high, since it failed to account for how immigrants already moved to growing cities. When immigrants go to locations that are not booming (see, e.g. Haitians who will soon be expelled from parts of Ohio), they prop up a local economy.

More to the point, immigrants build housing. They make up about 1 in 4 of the construction workforce. At a moment when construction jobs are going unfilled, targeting immigrants will mean fewer houses built, and higher costs per house. A Federal Reserve research paper noted that the type of immigration shutdowns that Gebbia supports disproportionately negatively affect the construction industry. Those contributions might plausibly offset housing price increases created by the arrival of immigrants. Another Federal Reserve paper that Gebbia has promoted argues that immigrants increase housing by a higher amount without offsetting gains in supply.

What about Airbnb? The most cited study found that Airbnb contributed to about one-fifth of observed rent growth and one-seventh of observed price growth in the median zip code, and shifted the housing supply from long-term to short-term rentals. These effects appear to be larger in with higher owner-occupancy areas, which tends to correlate with more touristy regions. Which is why some local governments, mostly in Europe but also in places like New York City, have sharply regulated or even banned Airbnb.

So it seems a bit rich for Gebbia to blame immigrants for housing cost increases. If that is his primary concern, he should be as aggressive in supporting regulation of his former company as he is in supporting anti-immigrant policies. Given his connection to Airbnb, he would be a uniquely powerful advocate!

Share

What has the National Design Studio Actually Done?

Maybe you are willing to be ok with Gebbia’s private views if he brings a lot of value to government. But a) the value he brings to government looks pretty limited thus far, b) those views are not private, and c) they are likely discouraging the ability of the Trump administration to hire talented technologists.

Lets look at the record of the organization he runs. The National Design Studio was created by Trump in August of 2025. It is a White House unit that fills a real niche, but the focus seems to be mostly about better branding of government services and websites, rather than making the underlying services work better.

So far, the NDS boasts of having built several websites, including NDStudio.gov and AmericaByDesign.gov (the office’s own sites), TrumpCard.gov (selling pathways to American citizenship for rich people — some immigrants are fine!), TrumpRx.gov (a drug-pricing portal), RealFood.gov (a MAHA nutrition site), TrumpAccounts.gov (children’s savings accounts), and Freedom.gov (not actually functioning right now).

But the NDS has drawn a lot of criticism:

• Skins vs. services: the National Design Studio is building a handful of attractive single page websites, but not fixing digital services that most of the public relies on when they interact with government. Skins is easy to do, but not very impactful.

• Low accessibility: Disability advocates say the sites do not meet federal guidelines.

• A reliance on poor code, likely AI-generated: If you like images of kids with extra toes, or a US Flag missing stars, the NDS has you covered.

• Privacy concerns: Some NDS sites were using commercial software to track visitors without legally required disclosures. It is also collecting data on users without a clear data retention policy.

• Favoring private allies: A NDS-built MAHA website directed users to Elon Musk’s Grok to get health advice, which required a sign-up for ongoing use.

• Failure on key goals: NDS was asked to create branding and standards consistency across the federal government but has failed to do so. Even the small number of pages it has designed look different from one another. Individual agencies do not seem to be working with NDS, and the move toward common digital standards appears to have stalled.

To be clear, some of the problems on accessibility, tracking and promoting of private allies were later fixed. The pattern is: NDS would do something that broke a law or looked ethically bad, and then fixed it when questioned. Thats not great, but not terrible.

The most serious critique is that there is no “there” there, it’s all hat and no cattle. Here, it is useful to understand that the NDS team seems to be less skilled and less ambitious than the organization it replaced, 18F.

Lets take a specific example. The NDS built the Trump Accounts app. It looks lovely, but only a fraction of the families eligible are claiming the $1,000 payments. Another approach would be automatic enrollment — using tax records to identify those who are eligible and signing them up. This is possible. For example, Child Tax Credit expansions during the pandemic were automatically allocated to families. But that would require working with the IRS and connecting digital and data systems across the government. Thats a lot harder than building a good-looking app. NDS provided a nice skin over a service that is not optimized to serve the public with minimal frictions. As a result of those system constraints, the accounts are likely to actually worsen inequality, as families with more resources are going to be the ones to manage the administrative burdens when it comes to claiming them.

Christopher Butler, a digital designer who went to college with Gebbia expressed the view of many in the civic tech community when he labeled NDS a “scam”:

we were already doing what Gebbia claims he’ll accomplish in three years. The 18F team had years of experience navigating federal bureaucracy, understanding regulatory constraints, and working within existing governmental structures—precisely the institutional knowledge required for meaningful reform.

Waldo Jaquith, a senior technologist with extensive federal experience told me: “They have solved zero design problems, advanced nothing of value, created nothing that anybody will reuse.”

At this point, we are now in pretty inside baseball debates about tech in government. But this context is important to understand when media outlets present Gebbia as a transformative visionary, see, e.g. this Free Press boosterism from Bari Weiss’s sister.

The Trump Administration is Struggling to Recruit Tech Talent

The National Design Studio and the Trump administration more broadly appears to have a tech talent problem. A primary NDS engineer is Ed Coristine, the DOGE alum better known for his nickname (Big Balls) rather than his skills.

DOGE purged skilled technologists and has not replaced them. Late last year, the Trump administration launched Tech Force, which has been promoted by the NDS. Tech Force was proposed to recruit 1000 new technologists by as early as March of this year, but by May had only recruited 10 people.

I’ve been talking to a lot of tech people who joined government over the last few years, and basically the key to get the best people, the people of extraordinary skill who could earn more money elsewhere, is pretty simple: prosocial motivation. They want to do extraordinary work that will help lots of people.

The problem for the Trump administration is that people with high prosocial motivation typically don’t want to work for white nationalists. Many might still join an administration, knowing they disagree with Trump, if they could be offered a meaningful apolitical role, but building single page websites does not fit the bill.

Imagine if a prospective recruit reads Gebbia’s feed and see about 1 banal post about the National Design Studio for every 10 posts featuring right wing tropes against immigrants, Blacks and Muslims. Is that the guy you want to work for? Potential hires see neither an organization that provides space to do valuable or apolitical work.

This does not mean Gebbia won’t succeed in recruiting people, but his public statements are shrinking the pool of willing workers to MAGA-heads, or those who see it as an an entry point to the Silicon Valley right-wing funding networks. These are not going to be the best of the best, and they are not going to be motivated to do really good work.

In short Gebbia’s profile is going to turn off exactly the people he needs to make the NDS work. It’s one more example of how the Trump administration cares more about promoting their divisive ideologies than actually serving the public.

This claim is striking since the non-partisan Congressional Budget Office (CBO) projects that that law, which was enacted one year ago, will reduce federal spending on Medicaid and the Children’s Health Insurance Program (CHIP) alone by over $900 billion over 10 years. Altogether, the OBBBA is projected to result in 10 million more uninsured people in 2034, most of which are due to its cuts to Medicaid and CHIP.

These represent the largest cuts in federal support for health insurance coverage in history. According to one early analysis, when combined with some other recent federal policy changes (such as the expiration of the enhanced premium tax credits), the OBBBA is projected to wipe away nearly three-quarters of the health insurance coverage gains made under the 2010 Patient Protection and Affordable Care Act (ACA).

United States Uninsured Rate: Historical and Projected1

And yet, the Trump Administration, Congressional Republicans, and ideological allies have repeatedly insisted (including in a recent essay published in a prominent medical journal) that the OBBBA does not cut Medicaid. Instead, they have argued, among other things, that the law cracks down on “waste, fraud, and abuse” and “protects Medicaid for the truly vulnerable.”

Let’s evaluate these three claims.

Claim #1: The OBBBA does not cut Medicaid

At the risk of stating the obvious, to describe a law as “cutting” government spending means, colloquially, that it reduces health care spending.

As I outline in a forthcoming article, the OBBBA does just this in three main ways:

1. imposing a variety of new administrative burdens (including, most notably, work requirements in Medicaid) that make it more difficult for people to enroll in Medicaid and the ACA Marketplaces;

2. rendering certain categories of legal immigrants (e.g., refugees, asylees) ineligible for Medicaid, subsidized coverage in the ACA Marketplaces, and Medicare; and

3. imposing new constraints on certain complex mechanisms—provider taxes and state directed payments—that states use to finance their Medicaid programs and pay providers.

Although politicians generally try to avoid being seen as having cut healthcare spending, there is nothing necessarily wrong with such cuts. Limiting health care spending can be appropriate or even required to free up resources for spending on other important social goods, like education, transportation, or housing.

Yet instead of defending the OBBBA on normative grounds, Trump Administration officials have opted to defend it on semantic ones: they argue that because nominal Medicaid spending (not accounting for inflation, mind you) is still expected to rise over time, the OBBBA does not “cut” Medicaid. Instead, they claim the OBBBA simply “restores Medicaid [spending] growth to historical levels supported by Democrats under President Obama.”

Such semantic arguments are not new. They work by changing the baseline of comparison: instead of a cut meaning a reduction in spending relative to what spending would have been otherwise (the conventional understanding), they claim a cut is only a reduction in spending relative to a fixed historical level. It is akin to a defendant’s claiming that his embezzlement of millions of dollars from a pension fund does not constitute a “theft” because that fund is still expected to increase in value in the coming years. By contrast, the CBO’s own estimates compare a forecast in which the OBBBA is in effect to a baseline forecast based on prior law. Such counterfactual comparisons are the standard way of understanding the impact of a law or policy.

That is not to imply that there is no uncertainty in the CBO’s forecasts. Predicting the future is an imperfect art, and it is possible that the CBO will – when all is said and done – turn out to have overestimated the coverage impacts of the OBBBA.

But the reality on the ground is that although many state Medicaid and state-based ACA Marketplace directors and their staff (aided by people working in the non-profit and private sectors) are desperately trying to mitigate coverage losses, they face a variety of daunting constraints in doing so. In addition, perhaps not surprisingly, some states are actively choosing to adopt more restrictive policies than those required by the OBBBA. So it is also possible that the CBO’s predictions will turn out to have been underestimates.

Economists Katherine Baicker and Amitabh Chandra have written that health insurance policymakers face an “uncomfortable trade-off” between making health insurance benefits more generous and covering more people. The two most recent pieces of landmark health care reform legislation approached this tradeoff in different ways:

The 2010 Patient Protection and Affordable Care Act both increased the number of people who have health insurance coverage and improved the generosity of that coverage in certain ways, while also attempting to scale back coverage of some kinds of excessively costly medical care.

By contrast, the OBBBA comes down firmly on the side of covering fewer people (especially low-income people), even while the Trump Administration pursues other policies designed to make health insurance coverage less generous. So rather than making any trade-off, the Administration has opted for reduced generosity and reduced coverage, all without displaying any apparent discomfort. There’s no need to be uncomfortable, it turns out, if you don’t acknowledge the hardships that your policy choices create.

Share

Claim #2: The OBBBA cracks down on fraud, waste, and abuse

There are a couple small kernels of truth to this second argument. The first is that fraud, waste, and abuse undoubtedly exist in government health care programs. The second is that the OBBBA does contain a few very modest provisions aimed at targeting fraud and abuse in government health programs.

Yet the administration and its allies have both overstated the extent of the fraud and abuse in Medicaid and the ACA Marketplaces, as well as the OBBBA’s role in combatting it.

For example, the Administration and its allies have cited as evidence of widespread fraud the fact that many people who enrolled in ACA Marketplace coverage filed no insurance claims while they were enrolled. But of course, that is how health insurance is supposed to work: it provides a measure of financial protection from medical costs by spreading risk across the healthy and the sick. In any given year, for some number of people, this protection mercifully will not prove necessary. Indeed, a comparable proportion of people with group health insurance coverage also had no claims, once you adjust for the fact that individual market enrollees are usually only enrolled for part of the plan year.

Moreover, the bulk of the OBBBA’s health care provisions are not targeted at reducing fraud. Many of these provisions (such as work requirements) simply make it more difficult for people to enroll in—or stay enrolled in—health insurance coverage by adding red tape and administrative hurdles. The primary effect of these policies will be to disenroll people who are in fact eligible for coverage. If this counts as combatting fraud, then, by the same logic, repealing Medicare and Medicaid in their entirety would count as an “anti-fraud” measure too.

“Waste” is a somewhat subjective term, but in public policy contexts, it generally refers to government spending that brings little of value (say, spending $5 million to cover four bronze horse statutes with gold plating). Again, there is, no doubt, waste in the health care system.

Yet the irony is that the Medicaid program, which bears the brunt of the OBBBA’s cuts, has proven especially cost-effective. A voluminous body of empirical research has repeatedly confirmed that having Medicaid coverage improves financial security, increases access to medical care, and leads people to live longer, healthier lives. Unfortunately for the millions of people expected to lose coverage over the coming years as a result of the OBBBA, this means that many of them will likely lead shorter, less healthy lives than they otherwise would have.

Claim #3: OBBBA protects Medicaid for the “truly vulnerable”

This claim rests on dubious normative and empirical foundations. Normatively, it reflects a belief (which dates back to the English Poor Laws) that only certain categories of people (e.g., pregnant women, children, people with disabilities) are “deserving” of social welfare benefits including health insurance, whereas others (so-called “able-bodied adults”) are not. This view is, to say the least, contestable.

It is also not popular: polling suggests that nearly two-thirds of Americans believe that it is the federal government’s responsibility to ensure that everyone (not just the “truly vulnerable”) has health care coverage. Indeed, the proportion of Americans supporting universal coverage has increased in recent years.

Empirically, this claim assumes that, by expanding Medicaid eligibility to include low-income adults making up to 138% of the Federal Poverty Line, the ACA reduced health care access for those people who were already eligible because they fit into one of the aforementioned categories (the “categorically eligible”). However, the evidence does not support this claim. To the contrary, the empirical literature finds that the ACA has improved access to health care, including for those who were previously eligible for coverage.

The argument also assumes that the Medicaid expansion population is in fact “able-bodied.” The reality, perhaps unsurprisingly, is much more complicated. For example, analysis from KFF finds that nearly half of adults enrolled in the Medicaid expansion have at least one chronic condition, and a quarter of expansion adults have a chronic behavioral health condition.

Moreover, even if it were true that the ACA’s Medicaid expansion somehow did reduce access for the categorically eligible, that fact would not justify cutting benefits for the Medicaid expansion population. Rather, if one supports universal health insurance coverage (as most Americans appear to do), the natural way to respond would be to increase the generosity of benefits for the categorically eligible, rather than to cut eligibility for the Medicaid expansion population (i.e., to “level up”, instead of “levelling down”).

-------------------------

In political scientist Paul Pierson’s seminal 1994 book, Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment, he identifies a few tried-and-true strategies that conservative governments use to minimize the resulting political blowback from cutting social programs (“retrenchment”).

The most important of these strategies, which Pierson terms “obfuscation,” involves “efforts to manipulate information concerning policy changes.” In essence, the objective is to muddy the waters and make it more difficult for people to understand retrenchment efforts or their consequences. It remains to be seen how successful the Trump Administration and its allies will be in achieving these ends, but if they fail it will not be for lack of trying.

Gabriel Scheffler is a Professor of Law at the University of Miami School of Law.

Subscribe now

There is no way to interpret the Slaughter decision as anything other than a defeat for expert, independent government, and the power of Congress to enable such government. Trump made clear that he saw it as a monumental victory handed to him by the Roberts Court.

The ruling allows Trump to remove independent agency heads. This is consequential for regulatory agencies in particular, and the promise that such regulators will be somewhat independent of the President of the day, and by extension, the corporations/donors they oversee.

The decision was expected based on prior court actions and oral arguments, especially when Chief Justice Roberts refer to Humhprey’s Executor precedent that supporting independent agencies as a “dried husk.” That does not make its impact any less breathtaking.

It is also a measure of how degraded our expectations of the Supreme Court are, and how complicit it is in Trump’s project of democratic backsliding: we are unsurprised when the Supreme Court drops a 90 year old precedent and overturns legislation to at the behest of a President who denies democratic outcomes.

At no point did Trump make any effort to justify the firings. I do think this bears mentioning because the magical thinking behind unitary executive theory implies some governing logic behind it — that Trump is an nationally elected official, and therefore most responsive to the will and interests of the people. The reality is that Trump is a deeply unpopular president whose purges are broadly disliked and whose attack on regulatory agencies will serve donors, not the public. But reality must not intrude on SOCTUS reasoning.

Sotomayor’s dissent laid out the risks for governing:

Today, the majority replaces 90 years of proven, workable practice with a half-baked theory of executive power that is simultaneously all encompassing yet also subject to necessary but undefined exceptions. The one thing that does appear to be clear going forward is that chaos will follow.

Lets take a simple example: During Trump’s first impeachment, he told executive branch officials not to cooperate with or testify to Congress. Appointees mostly obeyed him, career civil servants mostly did not. Congress and the public clearly benefited from Trump not being able to to restrict the legislative oversight function. The Supreme Court now makes it easier for Trump to lean on executive branch officials, to impose loyalty tests, to block oversight. A passing knowledge of the operations of the Trump administration make this clear. But the Supreme Court deals in idealized abstractions of a benign unitary leader, rather than the dismal realities of the man who appointed one-third of the court.

The consequences of Slaughter will extend in big ways for decades. Steve Vladeck, the esteemed Georgetown scholar of constitutional law put it in epochal terms:

To my mind, Slaughter is the most important separation-of-powers ruling of the twenty-first century—and, by far, the most important decision of the current term. It embraces a remarkably strong form of the unitary executive theory. It runs roughshod over a heck of a lot of sophisticated, persuasive scholarship arguing that Founding-era understandings were far more nuanced and complicated about the sharing of power across the branches. It doesn’t distinguish inferior officers, civil servants, or administrative adjudicators; it merely asserts that it isn’t resolving the constitutionality of protections for those executive branch employees…It makes all kinds of absolutist claims without even trying to persuade readers that the simultaneous carve-out for the Federal Reserve adopted in Cook is somehow consistent with that absolutism. And it does all of this at the exact moment in American history in which the most troubling implications of the unitary executive theory have never been more visible or immediate.

Justice Gorsuch voted for the outcome but at least acknowledged that it served to empower the President with delegated legislative power: “Open-ended delegations of legislative power have not gone away; now they will just be exercised by agency officials who answer to the President.” But Congress no longer gets to control how its delegations of power are used. Now, only a partisan court can do so, with little expectation that their standards will not adjust when a Democrat is back in power.

My fear is that Slaughter also lays the basis for a deeper assault on the civil service. The logic for the decision — that senior officials who employ executive power must be removable by the President — could easily extend to other public servants beyond the heads of independent agencies, including senior civil servants.

In her dissent, Justice Sotomayor noted this risk:

The Government, at least, has the candor not to deny that the “logic” of its position “extends to inferior officers” and, perhaps, career civil servants, too….The majority, however, at best consigns these issues to years of future uncertainty and at worst risks the end of the employment protections that apply to members of the civil service.

The Trump administration saw Slaughter as one block in its case to build an imperial presidency, but it is not done. It is surely the case that Slaughter will be essential to extending that project (see future cases around civil servant firings, or Schedule F employees, or Reductions in Force).

Going deeper on Slaughter:

Can We Still Govern? has been anticipating the fall of Humphrey’s Executor precedent that stopped Presidents from removing independent agency heads. Here are five prior pieces that put the decision in a broader context.

In May of 2025, when the courts blocked the reinstatement of independent agency officials fired by Trump I wrote about what was at stake:

A question I’ve gotten from reporters is how long it will take the public sector to recover from the damage Trump has done. There is no easy answer to this, but I highlight one variable that will matter a lot: how much the Supreme Court embraces unitary executive theory (i.e., the idea that the President has king-like powers).

Why does this matter? A maximalist interpretation of the unitary executive theory holds that almost any Congressional (or judicial) constraints on presidential power are unconstitutional. In more specific terms, it would hold that the civil service system itself is unconstitutional. If the court adopts that reasoning, then it becomes very hard to rebuild state capacity.

Because with unitary executive theory, there is no actor that can make credible long-term commitments to public servants.

With unitary executive theory, Congress cannot write robust new legislation that modernizes the civil service and stops politicization. A President could just ignore it. Even if Trump leaves office, and a new President looks to restore nonpartisan competence, their promises are only good for four or eight years before another President can come in and rip up the terms of their employment. And over time, why would even a good government President invest effort in restoring capacity if their successor can undermine it?

Unitary executive theory is a constitutional theory that would be largely unrecognizable to Founding Fathers who feared an executive who acted like a King. Where did it come from? John Roberts was a young Reagan appointee when he and others in the Republican/Federalist Society legal firmament engaged with unitary executive theory in battles over civil rights, writes political scientist John Dearborn. In some respects, his decision with Slaughter closes the circle on his professional journey, from a servant of executive authority, to an enabler of it.

President Trump’s war on the civil rights bureaucracy may be an escalation, but it has been decades in the making. Civil rights agencies have always been in the crosshairs of the conservative push for a president to wield “all of the executive power.”

John Roberts extensively quoted former Chief Justice Taft’s opinion in Myers in the Slaughter decision. Andrea Scoseria Katz, who studies administrative law, points out that Roberts gets Taft wrong in some crucial ways. Far from being a fan of the type of partisan administration Roberts is unleashing, Taft wanted fewer appointees, and more civil servants.

In light of Myers, these thoughts on appointment are surprising. So, too, are his views on oversight and removal. In reality, President Taft was no hatchet man. He fired cautiously and always in compliance with statutes. Where a statute required “good cause” for removal, as for a customs board, Taft provided notice, held a hearing, and submitted evidence that the statutory predicates for removal had been met. Ironically, the statute Myers later invalidated for allowing the Senate to interfere with removals, the Tenure of Office Act of 1876, was one Taft scrupulously adhered to as president, yielding on proposed removals where the Senate refused to back them.

As it turns out, Taft did not actually run his government as a top-down, command-and-control operation where he served as CEO. More to the point he disliked the aspects of unitary executive as the Trump administration has pursued them and which the Roberts Court relies on Myers to enable: a highly politicized administration that weakens the civil service, firing employees without any demonstration of cause.

Anticipating the Slaughter decision, I lay out how the Court already green lit the assault on the civil service via the shadow docket, such as barring the reinstatement of Slaughter and a senior Merit Systems Protection Board decision:

It is, in the end, not just Trump’s politicization of the federal government; the Republican majority on the Supreme Court are co-owners of this project. John Roberts, who frets and worries about the court as institution, is a willing co-author on the attack on other institutions. The Court’s complicity in the assault on public institutions gives them no justification to demand the protection of their own.

In the Cook case, Roberts decided that Trump’s ability to fire independent agency leaders held one fairly glaring exception, which was the Federal Reserve. In this one instance, Roberts showed he could toss aside unitary executive theory abstractions and deal with the practical politics of protecting the Fed when he needed to. Apparently, the Fed, and only the Fed needs independence to function as an agency.

The inconsistencies between the Cook and Slaughter decisions led to some obvious jokes about a special 401K exception. David Dayen put it best.

Even if Trump was defeated in this narrow way, Trumpism continues. Bill Pulte, the Trump appointee who targeted the Federal Reserve member Trump fired is now the acting Director of National Intelligence, and firing qualified intelligence officials without cause.

To understand the moment we are in, it is also important to understand her accuser, Bill Pulte. What qualifies Pulte to be in his position, as head of the Federal Housing Finance Agency? He is the heir to a home-building fortune and an-online right-wing troll with millions of followers. He has given large amounts of money to the Trump campaign, including $500,000 to a Trump superpac via a shell company. On that basis, and his enthusiastic advocacy for Trump, he was put in charge of a federal agency.

Pulte has used his position to launch an all-out assault on the Fed, calling for Powell’s removal. Pulte was the chief source of the claim that Powell could and should be fired for cause because of cost overruns on building projects. The fact that Pulte is unqualified for his position, that there are massive conflicts of interest between his personal business and the policies of the Fed does not seem to matter. Pulte uses public resources, including the name of his public office, to promote conspiracy theories and target the Fed.

Subscribe now

Two solutions: 1) Pretend the law does not exist - just fire people anyway, and 2) at the same time build a parallel legal infrastructure to provide an ex post justification.

Trump has fired civil servants with less justification than offered to contestants on The Apprentice. Rather than demonstrate they are poor performers, he merely asserts an Article II constitutional right that supersedes any law, and dares them to sue. This is unitary executive theory in action, offering an underlying logic for the parallel legal infrastructure that greatly expands executive authority while bypassing Congress.

Americans have a great faith in lawsuits, but what do you do when the law is on your side, but the venues for a hearing are demonstrably hostile? In order to ensure the courts were not overrun with federal employee complaints, the Civil Service Reform Act of 1978 created independent units in government, the Office of Special Counsel and the Merit Systems Protection Board, to hear cases when employees believe they were fired inappropriately, including for political reasons. The Board typically sides with the government, but people who want more firings for any reason see the MSPB as an impediment.

The credibility of the employee appeals process depends upon the independence of the MSPB. After all, if you were fired by one Trump apparatchik for not being sufficiently loyal, what hope of a reasonable hearing can you expect from another? How can the Board protect merit, when merit is defined by Trump and his appointees, rather than by law?

New reporting by the New York Times reveals that administration officials leaned on an acting Republican head of the Merit Systems System Board, Henry Kerner, when he was being considered for a permanent leadership position. They told him that they expected him to follow Office of Legal Counsel’s opinions on unitary executive theory.

The Times describes how:

the Trump administration leveled a concerted pressure campaign on the board in public and private, according to people with knowledge of the process. The private push — little different from calling a federal judge and telling him how to rule — was led by a White House aide who for years has been intently focused on making it easier to quickly fire federal workers.

The White House aide meeting with the head of the MSPB was James Sherk who has said that “Every federal employee should serve at the pleasure of the president.” Sherk is the author of the Schedule F (now Schedule Policy/Career) memo that asserts a legal basis for firing career civil servants without cause, again relying on unitary executive theory.

Here is Sherk at the signing of the Schedule Policy/Career Executive Order (about minute 15 of this video). After Sherk explained the EO to Trump that he was about to sign (no need for the President to read anything too closely when everyone is a loyalist), Trump offered him the highest compliment, saying he was “central casting...perfectly cast.”

A trademark of Trump’s second term has been to put in place loyalists in key positions to interpret the law in whatever fashion he demands. Then, when Trump violates the law, he can argue that executive branch lawyers support him, requiring others to take the case to court. And time and again, the Supreme Court has supported the novel legal formalisms that Trump has come up with, often with shadow docket decisions that the Roberts Court has transformed into permission slips for Trump.

Share

Sherk’s engagement with the MSPB is a rare glimpse into how the machinery of politicization works at lower levels within the executive branch, in the shadow of the shadow docket. Kerner, the MSPB appointee, was not explicitly told how to rule, but was told that he needed to follow the company line on unitary executive theory. And he was told this in a meeting with a presidential advisor who has explicitly identified MSPB as a barrier to making it easier to fire federal employees.

This is not a routine loyalty test. The whole point of the MSPB, created in the aftermath of Nixonian abuses, is that it is independent of a President who might abuse his powers. And unitary executive theory not only undermines that independence in terms of the core task of the Board — evaluating employees claims fairly — it also undermines the existence of the MSPB, since unitary executive theory does not tolerate the existence of independent agencies. In short, the MSPB leader was selected on their willingness to argue against the purpose of their agency.

Here is a tangible example of why this matters. Maureen Comey, the daughter of James Comey, was a respected Department of Justice prosecutor who was fired by Trump. No-one has made a credible argument that this was anything other than retribution against the Comey family. She sued, and said she could not go through the MSPB because of its compromised status, including on unitary executive constitutional questions. The Office of Legal Counsel disagreed, saying that MSPB could make such decisions, even though the natural endpoint of this logic was that the MSPB should hold that was unable to operate independently of the President, and even as the clear intent of the government’s intervention was to weaken Ms. Comey’s opportunity to have her case heard in a nominally nonpartisan court.

And it worked. The MSPB soon followed the direction they were given by Sherk and the Office of Legal Counsel, contradicting decades of precedent, to say they were entitled to weigh in on constitutional issues on unitary executive theory. The outcome, of course, is a forgone conclusion, legal niceties aside. Any decision of the MSPB issues is now the rotten fruit of a poisoned tree.

Trump had already undermined the independence of these agencies. He fired a Democratic nominee on the MSPB, Cathy Harris, converting the board from a bipartisan entity intended by Congress to one dominated by his appointees.

And here, it is important here to recognize the role the Supreme Court played in Trump’s politicization agenda. The fired MSPB member sued, and a District Court sided with her and the original structure of the MSPB as outlined in law. However, in May of 2025, the Supreme Court blocked her reinstatement while litigation is pending, saying.

The stay also reflects our judgment that the Government faces greater risk of harm from an order allowing a removed officer to continue exercising the executive power than a wrongfully removed officer faces from being unable to perform her statutory duty.

In other words, the Court regarded the interests of Trump politicization as so important to protect that they would upend the previous Supreme Court precedent, the status quo and the explicit intent of Congress even before deciding on the merits of the case.

Justice Kagan noted that the Trump administration was not just injuring individual employees like Harris, but Congressional designs to serve the public interest.

…the interest at stake is in maintaining Congress’s idea of independent agencies: bodies of specialists balanced along partisan lines, which will make sound judgments precisely because not fully controlled by the White House.

For people working in the MSPB, the message was clear: SCOTUS was not going to protect them. Crucially, the meeting with the MSPB head occurred in November, months after SCOTUS used the shadow docket to allow his colleague’s firing. The Supreme Court gave him zero reason to take a principled stand in the hope they would later support him. This is how executive power aggrandizement operates in the shadow of the shadow docket.

If the people whose statutory role is to protect against partisan firings were not themselves exempt from partisan firings, what hope would regular federal employees have?

The same decision that blocked the reinstatement of the Democratic MSPB member was the clearest sign that the Supreme Court planned to eliminate the independence of independent agencies (except for the Federal Reserve, for…uh…reasons). An even clearer sign came later that year, when SCOTUS also allowed the administration to fire Rebecca Kelly Slaughter, the head of the Federal Trade Commission, the original agency involved in the Humphrey’s Executor decision that confirmed the independence of such agencies.

We are now waiting for the final decision in the Slaughter case. There is little doubt on the outcome, just the extent of its implications. The decision will reverse the clear intent of Congress and Supreme Court precedent that has been in place for 90 years, and it will do so at the behest of a would-be authoritarian. It will give the Supreme Court the best chance to implement an ahistorical, anti-democratic and impractical governing framework cooked up by Federalist Society lawyers and Republican White House appointees.

The Slaughter decision will most likely weaken the MSPB and gut the Civil Service Reform Act and its merit principles. In doing so, the Court will back the White House officials who fired the Democratic official on the board, and who told Republican members to start embracing unitary executive theory. The Courts will back Trump, not Congress or its laws, and certainly not federal employees.

Nixon era abuses of public personnel occurred largely in secret. Trump era abuses are occurring in public. What explains the difference?

The Nixon administration knew that the court of popular opinion and the actual courts were against them. The second Trump administration does not care much about public opinion, and believes the Supreme Court will give them much or all they seek. Both the abuses of power occurring with the executive branch — such as the firing of Maureen Comey — or the effort to provide a constitutional justification for such abuses all occur because Trump administration actors assume Supreme Court support.

The broader lesson is simple enough. Trump’s two-step model of government politicization (fire people, and build a justifying legal infrastructure based on unitary executive theory) is wholly dependent on the Supreme Court. Congress will not support it. SCOTUS could stop it, but won’t because the majority of its members agree with it.

It is, in the end, not just Trump’s politicization of the federal government; the Republican majority on the Supreme Court are co-owners of this project. John Roberts, who frets and worries about the court as institution, is a willing co-author on the attack on other institutions. The Court’s complicity in the assault on public institutions gives them no justification to demand the protection of their own.

A lot of attention is being paid to William Howard Taft these days, a strange honor for one of the nation’s most mediocre presidents—so mediocre that, in fact, he once jokingly wrote to a friend, “I don’t remember that I was ever president.”

Unlike most mediocre presidents, though, Taft got to enjoy a second life as Chief Justice of the United States, from 1921 to 1930. This role suited him much better, and in it he proved, not just one of the most able stewards the Court has had, but also one of the most influential architects of the presidency in history.

Today, Taft’s influence can be seen in the Roberts Court, which cites his Myers v. U.S. decision in every separation-of-powers opinion it writes. It will undoubtedly soon cite Taft again when it hands down Trump v. Slaughter.

The decision is potentially one of the most important constitutional and administrative law cases in decades. Triggered by President Trump’s early 2025 firing spree that saw hundreds of thousands of government workers lose or leave their jobs, Slaughter will ask the Court to decide whether the FTC’s (Federal Trade Commission) statutory for-cause removal protections violate the separation of powers. At stake is how far the Supreme Court will go in removing job protections of federal employees at the behest of President Trump.

Can Congress protect officers from “the Damocles’ sword of removal” on grounds that their jobs entail sensitive or technical functions requiring insulation from politics? Or must the President possess “all of” the executive power in the name of democracy and good government?

Led by Chief Justice Roberts, a majority seems poised to answer the latter. This Term’s cases will tell us what exceptions to the removal power the judges will tolerate, and whether institutions like the National Labor Relations Board, Merit Systems Protection Board and the Federal Reserve survive in their current form. But there’s no doubt that the Court now sees presidential removal, with just a few tiny exceptions, as the law of our Constitution. And here, Myers is key.

The story of a conflict between a bedridden Woodrow Wilson and an obscure fourth-class postmaster in Portland, Oregon, the “epochal opinion” of Myers broke new ground by constitutionalizing a power (an “unrestricted” one, at that) that past presidents had only asserted in their enduring tug-of-war with Congress. Taft authored the Myers opinion, giving the President new powers to remove officials.

In Taft’s telling, the President’s job requires him to place “implicit faith” in “each member of his official family, and his chief executive subordinates,” and to be able, if ever “he loses confidence in the intelligence, ability, judgment or loyalty of any one of them … to remove him without delay.”

In the ensuing decades, Myers’ reach would be cabined by a series of cases protecting the New Deal state and upholding Congress’s authority to draw up the government as it wished. Yet, like the loaded gun at the start of a Chekhov play, Myers survived, ready to be picked up and fired when the moment was right. That moment has arrived.

The trouble is, Myers’ model of public administration is totally unrealistic (a government of two million officers is certainly nothing like a “family”). And it was not an accurate or complete vision of how President Taft thought about the relationship between the president and the agencies, either.

Shortly after losing the 1912 presidential election, Taft retired to Yale Law School to lick his wounds and take a break from public service. In that time, he produced a series of lectures, later published as Our Chief Magistrate and His Powers, that laid out his theory of the office, not to mention several complaints about its practical shortcomings.

One feature of office that weighed on Taft’s mind was the appointment process. At the time, the president was responsible for nominating some tens of thousands of offices for Senate confirmation.

Share

Taft had a real problem with appointments: for one, they were a waste of the president’s time (“Why,” he complained, “should the President have his time taken up in a discussion over who shall be postmistress at the town of Devil’s Lake in North Dakota?”). Second, the process made officers into “the football of politics,” with legislators haggling over positions with which to reward supporters. Third, there was no way the president could comment knowledgeably on tens of thousands of candidacies. Finally, if each president set out to reinvent the government each time he took office, the “halt and lack of efficiency” would cripple the government.

Taft gave the example of the State Department, plagued by “entirely unnecessary helplessness” when “competent representatives abroad in our diplomatic service are removed and men of no experience replace them.” Both for time-saving and principled reasons, Taft concluded that the President should appoint no more than a few hundred officers (essentially, the Cabinet, assistant secretaries, judges, generals, admirals, and ambassadors). All others, he wrote, including bureau heads, should be reclassified by Congress as merit employees in the civil service so they could “have a permanent tenure and not change with each administration.”

In light of Myers, these thoughts on appointment are surprising. So, too, are his views on oversight and removal. In reality, President Taft was no hatchet man. He fired cautiously and always in compliance with statutes. Where a statute required “good cause” for removal, as for a customs board, Taft provided notice, held a hearing, and submitted evidence that the statutory predicates for removal had been met. Ironically, the statute Myers later invalidated for allowing the Senate to interfere with removals, the Tenure of Office Act of 1876, was one Taft scrupulously adhered to as president, yielding on proposed removals where the Senate refused to back them.

As it turns out, Taft did not actually run his government as a top-down, command-and-control operation where he served as CEO. More to the point he disliked the aspects of unitary executive as the Trump administration has pursued them and which the Roberts Court relies on Myers to enable: a highly politicized administration that weakens the civil service, firing employees without any demonstration of cause.

Chiefly, this was because he believed such a model was pointless: it was impossible, given government’s complexity, for the president to try to “impose his personality minutely” on its “permanent structure.” More vividly, he wrote,

Presidents may go to the seashore or to the mountains, but the machinery at Washington continues to operate under [an] army of faithful non-commissioned officers, and the great mass of governmental business is uninterrupted.

And this was a good thing! After all, Taft explained, civil servants had “far greater experience” than political appointees, not to mention a “life-long fidelity and efficiency [to their missions] . . . . not rewarded by notices in headlines.” In other words, the president could and should trust experts to do their jobs. He should be able to depart to the beach knowing that government would trundle along under safe hands.

Thus, despite the contemporary unitary mythos of “a single head [of government who is] the focus of public hopes and expectations,” Taft transparently understood the Executive Branch as a “they,” not an “it.” His vision of presidential administration was not unilateral or command-based, but consensual and cooperative: the President might go to the seashore for weeks, or even months, at a time, but he—and the nation—could rest comfortably in the knowledge that, thanks to the honesty and competence of the lower echelons of the civil service, government would go on.

These days, the presidency is in a rough state. Trump’s deep depredations into the civil service cast an ironic pall on Taft’s efforts to reconcile an “unrestricted” removal power with “efficient government,” while for close to two decades now, the Roberts Court has enabled a one-man presidency, and with Slaughter and like cases, they may go further. Has the Roberts Court self-interestedly misinterpreted the former chief justice’s handiwork? Or did Taft himself lack a “coherent mind”? Whatever the truth, there’s no doubt that Chief Justice Taft’s presidency has proved more lasting than President Taft’s ever was.

A question remains, though. Should we run our government by Taft’s theory, or by the government Taft actually ran?

Andrea Scoseria Katz is an Associate Professor of Law at Washington University School of Law in St. Louis. Professor Katz teaches and writes about constitutional law, with a focus on presidential power.

But failures that can be visualized are hard to beat. They tell their own story. That is why I think the Reflecting Pool has caught the public imagination. Endless images and memes of the degradation of the Pool have swept through social media. The stakes seem simultaneously low enough to laugh about, but significant enough to gin up outrage. It is simultaneously funny and symbolic.

It would take a whole other post to highlight the best jokes, but this is certainly my favorite.

That said, I think the Reflecting Pool debacle is mirroring different aspects of Trump’s failure to govern in a competent and accountable way. Here are some themes.

1. From the party of Lincoln to the party of Trump

The Reflecting Pool can be considered an extension of the Lincoln Memorial. It was designed by the same architect, Henry Bacon, and completed in time for the dedication of the Memorial. Bacon designed it to connect America’s two greatest Presidents, Lincoln and Washington, using the pool to tie the Lincoln Memorial to the Washington Monument.

Trump has made clear he is not repairing the Reflecting Pool, but making it his own, promising his vision will make it “the way it’s supposed to be. Much better than it ever was, actually.” The job of the reflecting pool is, well, to reflect its surrounds. It is not supposed be the defining feature, to grab the eye. Even Trump’s idealized and unrealized version of the Pool is a garish defilement of that purpose.

The Republican Party used to refer to itself as The Party of Lincoln. Whatever Lincoln stood for — his ending of slavery, civil war victory, personal integrity, vision and humility — seem distant. The Republican Party today is loyal to one man, and that is Trump.

Trump has repeatedly compared himself to Lincoln. Actual quote: “I believe I am treated worse.” He has made no secret of his goal of displacing Lincoln in the American pantheon. Defacing the Reflecting Pool is one reminder how little regard Trump has for his predecessor.

2. Ignore the underlying problem with a superficial and ugly solution that makes things worse

The Reflecting Pool, like America itself, has real problems. President Obama spent $30 million to fix those problems, but leaky pipes and algae — inherent in a shallow, sun-exposed pool — remain.

The National Parks Service drains and refills the pool each year. A real fix would require an estimated $100 million and more time. Trump had a different solution: A superficial repainting of the base of the Pool, which he labeled “American flag blue.” The underlying problems remain. Meanwhile, projects at the the already underfunded National Park Service are being cut to fund Trump’s DC spending spree.

3. Treat public institutions as personal property

Just as he coated the Oval Office with more faux-gold than an old-school Vegas casino, Trump wanted the Reflecting Pool not to work, but to be unignorably his.

This reflected the pattern of personalism. Trump has treated DC as his personal property. He demolished the East Wing of the White House, and paved over the Rose Garden, and pasted his name on the Kennedy Center without seeking any permission or approval.

Such approval processes ensure that the visual representations of the White House and government property are consistent with the historical traditions of the Capitol. But Trump casts them aside, underlining that we live under a personalist regime defined by the aesthetics of one man.

4. Ignore the rules, reminding people why the rules exist

Trump handed out a no-bid contract to someone he knew, but someone without experience in this type of work. Can he do that, you ask? There are procurement rules to minimize risks of corruption. But Trump invoked an exemption for situations involving the risk of “serious injury, financial or other, to the government.”

In this case, there was no risk of serious injury, financial or other, to the government. Trump simply wanted the pool done for July 4th. The urgency exemption Trump invoked was made for less than 1% of National Park Service contract spending, and was also used for Trump’s confidential no-bid contract to the same firm that is building his ballroom. In their case, the urgent task is fixing ornamental foundations near the White House, which seems less than urgent and more like a way to funnel $17.4 million to a favored contractor.

The Reflecting Pool contracts are just one example of a massive increase in “urgency” exception spending this year, much of it in homeland security. In a normal year, this would be $3-6 billion (pandemic years are obvious exceptions). So far in 2026, it is $28 billion. What this means is that there is an extraordinary increase in money that avoids normal oversight, without an obvious pandemic-like emergency to justify it.

Government has rules, which make it slow. But it also has exemptions for truly urgent cases. Trump has viewed these exemptions as permission slips to do whatever he wants. Previous Presidents mostly took the rules seriously, or were told by their lawyers they needed to. Trump has selected lawyers on the criteria that they will ignore the rules. He has discovered that the American system of government, much like our system of private enterprise, gives a lot of room to someone willing to be sued.

Generally, I’m on the side of giving the executive branch more discretion. Most Americans would probably be fine with the notion that getting the Reflecting Pool in good shape for the 250th anniversary is a good idea, even if it involves bending the rules.

The problem is there was nothing that could be fixed on that timeline, and that Trump exercised his responsibility irresponsibly. And that is exactly what causes the public to become more skeptical of the use of discretionary power, fueling a demand for more rules to constrain it. Its a simple truth of public administration: bad actors fuel the creation of rules that constrain good actors from using power well.

Share

5. Cronyism - the swamp that Trump built

As with other projects, like the new East Wing ballroom, the projected costs has ballooned. First it was $1.8 million, now it is $16.4 million.

This is one of the guys Trump gave a no-bid contract to.

Now, looks aren’t everything. You might look at me and say “that short Irish guy cannot dunk a basketball.” But…you’d be completely right.

Ok, sometimes things are what they look like. And the contracting “process” here looks as murky as the Reflecting Pool.

The woman beside John Cafaro said she asked to have his picture taken with him because he looked like a comic-book villain. This was at an event at one of Trump’s golf courses. And she was right. Cafaro has two criminal guilty pleas, one for campaign finance violations, the other for bribing a Congressman. Cafaro has donated hundreds of thousands of dollars to Trump since 2020. He is the owner of Greenwater Services (I’m not kidding), who received a no-bid contract for a water-purification system for the Pool.

The other contract went to Atlantic Industrial Coatings. Trump said they had previously done work at his golf club: “I have a guy who’s unbelievable at doing swimming pools. He looked at it. He called me up. He said, ‘Sir, we can do something on it.’”

Unbelievable indeed. Trump knows a guy. A large guy presumably, who calls him sir. As the Times reported, there was no evidence the firm had worked on even swimming pools, much less a complex project like the Reflecting Pool. The White House later said Trump had no personal relationship with the contractor. This is their first contract with the government.

Both contractors refused to talk about their work, or their connections with Trump, leaving the public in the dark and appropriately suspicious of cronyism.

6. Ignore the experts

Back in early May, when the New York Times delivered an excellent account of the Reflecting Pool, they quoted a representative of the hot-tub industry who pointed out that painting the pool would not solve any problem, and the blue would soon be obscured by green algae.

A predictable mess! It’s actually worse than that. The new blue surface makes green algae the more certain outcome. Rosalina Stancheva Christova, a professor of aquatic ecology at George Mason University told NPR that the new surface made the underlying conditions more friendly to algae growth:

The new, darker interior surface is going to absorb more sunlight. It is going to result in water that's warmer, and that ultimately is going to lead to more prolific algae growth.

The solution to the algae has been Cafaro’s “high-tech nanobubble ozone technology", manually removing algae, pouring chemicals into the water, and now draining the pool again.

Thats right: Trump is literally draining the swamp that he created.

7. Accept no blame, scapegoat those exposing wrongdoing

It is not just the incompetence arising from a personalist regime that is galling; it is the demand to ignore the obvious evidence of failure. We waste time and energy arguing with serial liars who refuse to accept they can ever be held responsible. They use the resources of the federal government to leverage propaganda and threats. Abuse of legal power to silence dissent and threaten critics has been a staple of the Trump administration.

Trump is attempting to insist that his project was undermined not by mismanagement but by vandalism. In a Truth social post, Trump insisted: “it came out great, except for the Vandalism, which we are now fixing.” When asked if contractors were to blame, Trump said:

No, no. We had vandalism. We had a 290, 300 foot slit, right through. Probably a box cutter, or a knife of some kind….Five people are arrested and five people are under investigation.

Trump also claimed fertilizer was added to the water, with no evidence. A New York Times report revealed internal National Park Service documents from a week earlier mostly cast doubt about Trump’s claims. While they found cuts to sections of foam, these did not affect the blue sealant that is now peeling, or the algae foams. Just to underline how bizarre this is, Trump previously said the fix would last for 50 years, that the materials were so strong that “if you had a knife — I don’t want to give anyone ideas — if you had a knife you can’t even cut it.”

Another New York Times report revealed that the “nanobubblers” whose job was to prevent algae blooms were removed because they were deemed too unsightly to be present in the background during a UFC promotional event in front of the Lincoln Memorial. And then the algae returned.

PEOPLE ARE BEING ARRESTED IN ORDER TO COVER UP TRUMP’S FAILURES.

But people are really being arrested and ticketed. The US Attorney for DC Jeanne Pirro, says that all such vandals will be fully prosecuted. Even if such prosecutions fail, they upend the lives of those involved, creating legal bills and career risks.

The site has become a tourist attraction in a novel way, as locals, tourists and reporters monitor the shitshow. Some touched and removed detached pieces of the surface material that did not adhere to the pool, a Dollar General version of grabbing a chunk of the Berlin Wall.

This who dared to touch or remove a symbol of Trump’s incompetence were arrested, and scapegoated for that incompetence. David Hearn took this picture of the floating materials, which many others have documented. He was immediately detained by the National Guard and charged with destruction of government property for, by his account, merely touching the materials. He has since been indicted.

Hearn, a former Olympian, said:

I reached in there, and I was able to grab the end of that flapping piece, the already peeling piece. It was still attached to the bottom. I didn’t remove anything….I didn’t vandalize anything. I didn’t destroy or break or peel anything. By the time I realized what was going on, I was being put in handcuffs.

The stated purpose of placing the National Guard in DC was to reduce violent crime. New research shows that they have not done so. Their presence is symbolic, a reminder that people living in DC do not control their own city: Trump does.

Now the National Guard has been deployed to protect Trump’s mismanagement of the Reflecting Pool, telling tourists and even reporters that they will be arrested if they touch the pool. It is not clear what law, if any, the National Guard is enforcing. The dystopian images of a fenced-off and guarded pool speak for themselves.

The fence around the pool is nominally to prevent vandalism, but really to try to reduce transparency around one of the most visible of Trump’s failures. Automated surveillance machines announce to passersby that “Loitering is not permitted in this area. Please proceed to a designated location. Thank you for your cooperation.” Maybe Trump wants the Reflecting Pool to look good, but he does not seem to want any Americans to be able to see it in person.

Bloomberg reporter Steven Dennis summed up the situation:

Large chunks of the Mall are now effectively an armed encampment surrounded by tall fencing. In some areas, crowds are herded through security lines to get to features that used to be easily accessible in previous years.

Hey, look. Do you want to find someone who actually vandalized the pool? What about the guy who drove a motorcade over a notoriously finicky and delicate surface prone to leaks?

All politicians like to shift blame when they can. But Trump is in a league of his own, frequently relying on conspiracy theories to turn ordinary people into villains. Now, he is treating people who are curious about or mocking his Reflecting Pool as Antifa, or at least the Antifa of MAGA imagination: a pervasive, destructive force.

This is what Trump has always done, of course. Pick a topic. Increase its salience. Demonize anyone against him. Birtherism. Immigration. The Wall. DEI. Covid. It is exhausting.

Is anyone still buying this? Can anyone still gin up the outrage to cheer the prosecution of a DC resident recording the desecration of a national monument? Is anyone still nodding along and saying “yes, the guy in the inflatable pink frog costume is the real threat.”

It is not only that all of this is exhausting. Trump is literally asking everyone to believe him, and not their lying eyes, about the degradation of one of the most potent visual symbols of America. Some will go along, no doubt, but most Americans will be left wondering when they can get a government actually committed to solving real problems, rather than creating them.

Each year, millions of Americans who qualify for public benefits fail to receive them.

Consider health insurance: in 2022, around 15 million people, the bulk of the uninsured population, went without health insurance coverage despite being eligible for free or subsidized coverage through either Medicaid or the Affordable Care Act (ACA) Marketplaces.

This same dynamic exists in other public benefit programs. In a given year, roughly 7 million eligible Americans don’t participate in the Supplemental Nutrition Assistance Program (SNAP) and another 5 million don’t claim the Earned Income Tax Credit (EITC).

These missed benefits add up. Forgone EITC benefits total approximately $7 billion each year. They also address vital needs. Missing out on SNAP benefits means that kids might go hungry. Whether one has health insurance coverage can be a matter of life and death.

So why do so many eligible people lose out on benefits?

One explanation has to do with how programs are administered. Most public benefit programs in the United States require people to opt-in. That means they have to learn about the program, determine whether they qualify, gather the necessary documentation, complete the application, and navigate whatever extra administrative burdens the agency requires, such as an in-person interview or employer verification.

At each step of this process, eligible people fall away. They miss deadlines, lose paperwork, become confused by complex rules, or simply lack the time or know-how to navigate a burdensome process while managing the daily challenges of poverty. What’s worse, means-tested programs require regular redeterminations in which recipients must prove their eligibility anew.

The result is a safety net that catches some but lets many slip through. The recently enacted One Big Beautiful Bill Act (OBBBA) only compounds this problem, adding myriad new administrative burdens to Medicaid, the ACA Marketplaces, and SNAP. Recent reporting suggests that state agencies are buckling under the weight of these new federal rules.

The Promise of Automatic Enrollment

“Automatic enrollment” is one solution that has attracted attention from policymakers, administrators, and advocates. Automatic enrollment refers to when the government identifies eligible individuals through administrative data and enrolls them without requiring an affirmative application on the individual’s part. As a result, the default goes from opt-in to opt-out.

Psychologists and behavioral economists have long observed that default rules—rules that determine how a person is treated if they don’t make an active choice—tend to have major effects on program enrollment. One recent study found that suspending health insurance auto-enrollment in Massachusetts caused new enrollments to decline by 33 percent, an effect that is an order of magnitude greater than the effects of softer “nudges” such as providing information and assistance and up to twice as large as the impact of Massachusetts’ mandate penalty.

Programs like Medicare Part B and Social Security, which come close to automatically enrolling everyone who is eligible, unsurprisingly boast take-up rates of nearly 100 percent.

Share

Implementation Challenges

Yet if automatic enrollment is such a powerful intervention, then why hasn’t it been more widely adopted? And how can we implement automatic enrollment more broadly across the safety net?

In a new working paper, Automating the Safety Net, we try to answer these questions by conducting an in-depth comparative evaluation across numerous public benefit programs. We consider both the challenges involved in implementing automatic enrollment as well as how to surmount these challenges. Our analysis is informed by 28 semi-structured interviews with policy experts, technologists, and government officials who work on public benefit administration. We sought out people who have advocated for or helped to implement automatic enrollment, rather than its critics or beneficiaries. As a result, their views capture how practitioners see the terrain, and are not necessarily a verdict on whether automatic enrollment is worth pursuing.

The interviews point to four main challenges:

1. Decentralization. Many federally funded public benefit programs are run by states, territories, tribes, and, in some cases, counties. As a result, any improvement must often be made across multiple systems. One expert described county-level public benefit administration as “one of the biggest hurdles” to automating public benefits. They explained: “If you wanted to make a change to the [Medicaid] ex parte rate, you need to do it 50 times in California. You need to convince 50 different software vendors. . . . It’s just a much harder change management process.” These challenges are compounded by the divergent levels of capacity across states. “Each state has different degrees of sophistication within their program, their systems,” one expert explained.

2. Obtaining necessary data. Automation also requires that the government have the necessary data to determine eligibility and sometimes benefit levels. In theory, agencies can get some of these data from other government programs, but there are challenges to doing so. Perhaps the biggest hurdle is that eligibility rules differ among programs, making the data collected for one program unusable for another.

   Moreover, sometimes governments can’t access the necessary data or the data simply don’t exist. Some benefit applications require an in-person assessment. Others require frequent or detailed information that’s not automatically reported to any government agency, such as monthly or quarterly income data. As one expert succinctly put it, “You can’t automate things you don’t know. And usually there’s something you don’t know.” Governments could purchase relevant data held by private companies, but such efforts would have to grapple with what Luke Farrell recently described as the “means-testing industrial complex.”

3. Administrative capacity. Many of our interviewees also lamented public agencies’ limited capacity, especially at the state level, to build the necessary technology and data systems to improve benefit administration. One expert told us that in the states, “there’s significantly limited resources to undertake tech solutions, including time and human capital and money.” Another expert linked this to the need for more funding, noting that there’s “just not a ton of R&D and innovation funding that happens or trickles to [the] state level for this work.” Not only does this make it harder for government to create new benefit systems, but also it makes it harder to maintain and update them.

4. Fears and values. Beyond the technical and administrative challenges discussed above, automating public benefits requires political support. Most obviously, policymakers must actually view it as a problem that eligible people are not receiving benefits. One interviewee observed that “There’s a massive political challenge: like half of our system thinks that… nobody should have Medicaid ever.”

   Automating enrollment also entails potential risks, both to the public and to enrollees themselves. For example, policymakers have long invoked concerns about “waste, fraud, and abuse” to justify imposing administrative burdens on public benefit programs. More recently, Congressional Republicans repeatedly invoked such concerns to justify enacting the OBBBA, even though that law, apart from the undoubtedly significant SNAP cost-sharing change, contains only a few minor provisions focused on fraud and abuse and though it will mostly serve to disenroll people who are in fact eligible.

   Several interviewees also worried that data shared with the aim of automating enrollment could be used by Immigration and Customs Enforcement (ICE). The potential upsides of administrative data sharing to help people may be checked because of a fear that it will be used in punitive ways.

Four Recommendations to Ease Safety Net Automation

While daunting, these challenges are not insurmountable. As we describe in our article, the United States has, in recent years, taken some promising steps toward automating enrollment across a range of multiple government benefit programs, including the Child Tax Credit, Medicaid, and Summer EBT.

In 2021, the IRS used prior-year tax returns to automatically deliver monthly Child Tax Credit payments to tens of millions of families, helping drive child poverty to a historic low of 5.2%. During the post-pandemic Medicaid unwinding, many states dramatically improved ex parte renewal processes, which kept millions of Americans covered without requiring them to reapply. And the Summer EBT program, made permanent in 2023, was designed from the ground up to use existing SNAP and school-meals data to enroll eligible children without requiring a separate application.

Interviewees suggested several steps legislators and regulators could take to address the challenges described above:

1. Aligning Eligibility Rules. Many experts noted that aligning definitions and eligibility rules across programs would facilitate automatic enrollment. Some also suggested that lawmakers make greater use of adjunctive eligibility, where qualifying for one program automatically qualifies someone for another.

   At least one expert recommended increasing the use of “area-based” eligibility that deems someone eligible based on where they live. One example of this type of approach is the community-eligibility option in the school meals program, wherein a school that serves a large percentage of low-income students can choose to provide free meals to all students without requiring applications.

2. Improving Data Access and Quality. Experts also emphasized the importance of improving agencies’ access to the kind of high-quality data necessary to automate public benefits. Some indicated that new legislation would be necessary to encourage or require agencies to share data with each other. One expert observed that it helps to “be very explicit in new legislation [about] how data can be used.” Another suggested that federal agencies should publish a standard data sharing agreement.

   To fill gaps in existing data, some experts suggested imposing more frequent income reporting requirements. As one expert put it, the government should “require employers to report income data to the federal government at every pay period.” Another expert likewise emphasized the importance of “building more in-house federal data sources” so that states are not as reliant on private companies such as Equifax to provide data to determine whether enrollees are eligible.

3. Increasing State Capacity. Many of our interviewees emphasized the importance of increasing states’ administrative and technical capacity. When asked what steps could facilitate automatic enrollment, one expert put it bluntly: “Increase state capacity for IT projects.” Another expert explained that “there needs to be a legislative-level investment in actually fixing the underlying technical infrastructure of all this stuff.”

   Several experts stressed, in particular, the value of hiring in-house engineers to help run benefit programs. One expert pointed to several states that had recently built up their technological expertise at their agencies, showing it can be done.

4. Building Trust and Facing Fears. One obvious way to build trust is to mitigate, as much as possible, the real risks automatic enrollment poses. Experts offered a few suggestions for how to do this. One expert emphasized the importance of “adopting a significant harm-mitigation strategy” that includes contingencies for when enrollees are improperly enrolled in benefit programs. Another expert recommended limiting the information being shared for automatic enrollment initiatives to “what’s necessary for enrolling someone or establishing that they’re eligible.” For instance, rather than sharing reams of detailed personal data, “you could just have a flag that says [this person is eligible], and all you need to transfer is the flag.”

   But good governance alone isn’t enough to build trust. One problem is that people often don’t notice the benefits government provides. Several experts therefore stressed the importance of communication, both to ensure enrollees know they have coverage and also to build public support and embolden other officials to follow suit.

   Experts also emphasized the importance of political or agency leadership in empowering and motivating civil servants to automate enrollment. For example, one expert recalled how a state official initially declined to implement a program making it easier for people to file state taxes, only to change their mind two weeks later after receiving a call from the Governor asking them to do so.

Automating Enrollment in an Age of Retrenchment

The OBBBA’s impacts are already being felt across the country and millions of Americans must navigate the new administrative burdens it imposes on Medicaid, the ACA exchanges, and SNAP. In addition, the OBBBA’s cost-sharing requirements in SNAP and its restrictions on Medicaid financing create new fiscal challenges for states that are causing them to further restrict access to public benefits programs.

As a result, this may seem like a bad time to consider ambitious strategies to expand access to public benefits. Yet politics can change quickly. When it does, we should be ready. Our law has made promises to people in need. Automatic enrollment may well be our best way to keep them.

Gabriel Scheffler is a Professor of Law at the University of Miami School of Law.

Ariel Jurow Kleiman is a Professor of Law at the USC Gould School of Law.

Andrew Hammond is a Professor of Law and Harry T. Ice Faculty Fellow at the Indiana University Maurer School of Law.

For more on automatic enrollment