Listen to Zooming In at The UnPopulist in your favorite podcast app: Apple Podcasts | Spotify | Google Podcasts | RSS | YouTube
Landry Ayres: Welcome back to Zooming In at The UnPopulist. I’m Landry Ayres.
The executive branch is currently being run by an aspiring autocrat who neither respects institutional safeguards nor accepts the separation of powers. As part of his project to subsume all governmental authority under his control, he has empowered a tech billionaire to reshape and repurpose government agencies—through what they’ve called the U.S. Department of Government Efficiency, or DOGE—so that those agencies and their information can be weaponized for his purposes.
To explain how all this is unfolding, Zooming In host Aaron Ross Powell sits down with
and , the co-hosts of WatchCats, a new podcast dedicated to critically covering DOGE. They discuss what DOGE actually is and what it is doing, and assess what its impact is likely to be on future attempts to pursue government reforms.We hope you enjoy.
A transcript of today’s podcast appears below. It has been edited for flow and clarity.
Aaron Ross Powell: What is DOGE? And not just the thumbnail sketch. Is it an agency? Is it an advisory committee? Is it basically a criminal gang we’ve decided not to prosecute? What is this thing from an institutional, legal standpoint?
Julian Sanchez: So, formally, it is a temporary organization established by executive order within what used to be called the U.S. Digital Service. The U.S. Digital Service was, until it was effectively gutted by DOGE, a kind of tech incubator—a tech think tank, almost—that provided temporary personnel to other agencies to develop cutting edge technologies, web applications, databases, things to try and streamline the activity of other agencies. The idea was you have a group of technologists and tech experts—very often it was people who were taking a couple of years off from working at Google or Amazon or other big private sector tech firms to do a temporary stint here—and they could be sent in small teams to agencies like Veterans Affairs and Health and Human Services to work on specific projects. That’s what the U.S. Digital Service was.
By executive order within the first couple of days of his inauguration, Donald Trump rebranded the U.S. Digital Service as the U.S. DOGE Service and formally established DOGE as a kind of subgroup or element within that agency. But it also operates a little bit like USDS initially did in that it has farmed out a series of teams who are DOGE staff but are in many cases now employees of the agencies in which they’re installed. And it’s more than that. In fact, the nominal acting head of USAID, the foreign aid entity that has also been effectively gutted, is a lawyer who was a DOGE staffer and is now formally running that agency.
Noah Kunin: In most organizations, especially large organizations, very rarely is anything a monolith. So asking what DOGE is—it has multiple factions, and those factions are not going to be aligned on every specific initiative that they take on. But if we look for the common denominator, the way I’ve been thinking about it lately is through the idea of a red team. This is a team concept in cybersecurity and military affairs of not waiting for the enemy to figure out your weaknesses—you split off a team whose job is to destroy you with every possible means in their disposal. Then you set up a blue team that is there to steelman the case, to defend against the red team. They conflict. You analyze the results. You improve your organization or whatever domain you’re working in. And you rinse and repeat and iteratively improve over time.
I think of DOGE as a red team on a heady mix of steroids and ketamine and Zyn without any blue team there to advocate on the other side. That is the best way to sum up the majority of its actions. And what exactly is it red teaming? What’s very clear from Elon Musk’s late night X spaces is that he has two main motivations.
One you can sort of explain from his obsession with competitive leaderboard-based video games: he wants the number to go up—“number go up good ... saved money.” And he has expedited all possible effort to make the DOGE.gov money saved number go up, regardless of whether the actual methodology used to calculate that number is in fact correct, or if it results in true savings over any time period beyond his specific stay in government. For example, we’ve just now heard that the second acting Commissioner of the IRS has been summarily dismissed, along with thousands of other IRS employees. Well, that’s going to result in decreasing receipts into the government, eliminating a lot of the savings that Musk is going after.
The second part of “What is he trying to red team?” is he is trying to find every possible advantage and vulnerability over what he and many other people call the administrative state. They view the administrative state as a parasite sapping the vital juices of America at best, and an entirely separate form of unconstitutional government that must be destroyed with all possible speed at worst. For Musk, the administrative state is an existential threat, not just to the country, but to humanity as a whole. And it’s okay if thousands or tens of thousands of innocent people who are wrapped up in that are harmed if the end goal is to destroy the administrative state before the midterms.
Powell: So, there’s a way to frame that in a way that a lot of our listeners might say, “That sounds pretty good to me!” For a lot of years, both Julian and I worked in an organization that was aimed very heavily at reducing the size of the government. Many of us believe that the government is too big, does too many things, spends too much money. A lot of people might be listening in favor of, say, ending Chevron deference as a way to reduce the power of administrative agencies.
So, some might look at this and say, “Maybe some of the accounting on the cutting doesn’t add up, but they have gotten rid of a lot of federal employees. They have reduced a lot of federal programs. We needed to reduce the size of the government. Maybe this isn’t being handled as elegantly as we might have liked, but what’s wrong with what they’re doing? We had to cut somehow. It was unsustainable.”
Kunin: I would first push back on the original contention. It’s not that they’re doing it inelegantly—they’re not doing it at all. If you want to say that USAID was the biggest source of government largesse and overreach in the federal environment, I think that’s ridiculous on its face. If you want to look at government overreach as entitlement programs, those have not been altered in law in any way, shape, or form as we sit here today, and at the same time, we’re seeing a massive overreach in terms of overriding the rule of law when it comes to due process or judicial opinions. Do you feel like liberty is rising in that environment? I would contend not.
“What’s permanent is the normative change. I say this a lot because it’s true: Most of what you think is a law is a regulation. Most of the regulations you think of aren’t regulations, they’re administrative actions. Most administrative actions are really just policy papers of a particular agency. And most policy papers are just groupthink within an agency. So, when you blow up the norms constraining government action, it is very-difficult-to-impossible to put them back together again.” — Noah Kunin
As somebody who has worked in the federal government for many years, worked for state government for many years, and comes from a progressive ideological side on this, I would actually agree on some of those points you made, Aaron. I just want to point out that literally nothing in my top 10 list of egregious bloat in government has been touched whatsoever. In fact, in many cases the people fired were the best people you would have worked with in order to actually target those things. So when you go into the preexisting technical capacity, the preexisting teams who are there to create efficiencies, centralized services, transparency in government, and they are some of the first people that you release, it sort of belies the whole thesis of what you’re here to do.
Sanchez: I think maybe the most important thing to recognize is that DOGE’s actual mission is not to cut government waste or root out fraud. If you are paying close attention to how they’re operating, it’s obviously not what they’re actually attempting to do. No sane person would operate in the way that they’re operating and prioritize the things they’re prioritizing if the real goal was to cut government waste. I mean, in the scheme of the federal budget, the things they are cutting are trivial. It would not be a great idea, but you could fire the entire federal workforce and it would be a percent of federal outlays. It’s just a trivial fraction of the total federal budget.
You would also probably do it a lot more thoughtfully than they are—you would do it in a way that meaningfully tries to look at what different elements of the government are doing and figure out who the lowest performers are and target those, and figure out what is the actual number of people we need for this.
So, a couple of things. One is, if you really, meaningfully, want to reduce federal outlays, the things they’re doing are tinkering at the edges in a way that is frankly not important. You would focus on defense, on health spending, and on entitlements—primarily Social Security. That is where the money is actually going. But all this other stuff is secondary. Additionally, you would be a lot more thoughtful about whether the things you are cutting represent real long-term savings, and also whether you are cutting things that are genuinely not things the government ought to be doing.
So, for example, we’ve talked to a bunch of former USDS people who have either been fired or resigned because it was clear that that was in the offing. And they were working on things like improving the algorithm that the Health and Human Services uses to prioritize organs. People every day die on the organ donor waiting list, and organs that could have been transplanted go to waste. They were trying to improve the algorithm that tries to match viable organs to patients and figure out, are they going to die without it soon enough that this is where that organ should go? I’m pretty libertarian, but that seems like a valid function.
We talked to someone who was working on an outbreak tracker for the Centers for Disease Control. If there’s a series of, let’s say, cases of food poisoning showing up in a cluster that hospitals are reporting, does the agency look at that and detect whether there is some sort of food contamination? And something needs to be done to address that before more people become ill or a viral contagion. Even with a pretty restrictive view of government, those things sound to me like legitimate functions of government.
And, more broadly, you want to look at, is this a cut that is a long-term savings? No one loves the IRS, but you are not on net saving money if you make cuts that reduce the legitimate tax revenue you’re taking in so that you’ve saved money on salaries but now you’re collecting less in taxes. Maybe you also think taxes should be lower but that’s separate from the question of, whatever the rate is, should those taxes be collected. It’s very easy to save money if you do not care about the long term. Say I’ve got a car. I have to fill it with gas. I have to pay insurance. I could save a bunch of money by selling my car and not having to pay car insurance anymore. Well, if I need that car to get to my place of work, this is not a net benefit to my financial position. And nothing they’re doing really bespeaks an attempt to make these kinds of decisions in an intelligent way that says what is going to put the government in a better financial position in the long term.
What it does bespeak, and what I think is the real function here, is an effort to change the ideological composition of the federal workforce and shrink it and effectively turn federal employment into a kind of patronage system for loyal Trump supporters—and, more to the point, to centralize executive authority so that the White House can directly accomplish more things without intervening areas of bureaucracy, which you might say in some instances sounds desirable. But in practice, it means the ability to, in a sense, outrun external oversight; to be able to, for example, shut off funding streams that Congress has appropriated—rightly or wrongly, but that under our Constitution, the legislature has appropriated—and say, “No, actually,” now the technological infrastructure is such that the White House can and will do this directly.
And there are a lot fewer people in between the White House and the results, such that if Congress wants to do oversight and figure out why the money they appropriate isn’t going where they said, or courts want to examine it and say, “Are you complying with the law?” ... those intermediary people are not there. You have technical systems under the president’s control— executive offices in particular—in a way that makes oversight and legal restraint more difficult. And it is easy to sell this to people like us who think government ought to be smaller and more efficient and lean by saying, “What we’re doing is shrinking government and making it more efficient.” I think it’s pretty clear that is not really the goal, and that they’re happy to fabricate fraud and show very little interest in talking to experts who have been there a long time about whether some artifact they found in the database is actually a sign of fraud.
What that says to me is they’re not really that interested in finding fraud. What they are interested in is finding a plausible claim they can make to justify their own existence so that when people like us say, “Hey, it doesn’t actually seem like they’re accomplishing the function they set out to,” people who don’t understand what they’ve found can say, “Oh, but look, they found all this evidence of 150 year olds collecting Social Security and they found all this evidence of illegal aliens voting.” None of that’s true, but it requires a fair amount of digging to figure out that it’s not true.
Kunin: And even if you think the prime directive here is to make the number go down and that they actually are detecting waste and fraud, they’re still failing on the numbers. If you actually use a formal methodology to calculate their savings, and I’m being a little more generous than some other people right now, I’m looking at 1.1% of the federal deficit in 2024. That’s it. Where is the money? If they are so confident there’s so much fraud, waste, and abuse, and it’s so easy to find, why’d we only cut 1.1%? Well, because most of the deficit is in fact not in fraud, waste, and abuse, but in entitlements and defense spending. And the only people who can really fix that is Congress.
Powell: If we say, “Yes, there are meaningful cuts that ought to be made. If we got a bunch of very smart and principled people together and they ran through the whole of the federal government and everything it was doing and spending on, they could agree on many areas where it should scale back or stop” ... is there a worry then that what DOGE is doing now risks making that entire project politically toxic?
If you’re someone watching, what you’re seeing is this guy, Elon Musk, who nominally claimed to have libertarian principles about the government for a while, talking in the language of libertarian cutting. You see lots of people cheering as federal workers lose their jobs. When DOGE is behind us, the next time we get a Democratic administration that says anything like, “Actually, this program might be better off in the private sector,” it’s going to be made irredeemably toxic. Nobody’s going to take that seriously because the people in power saw what it looks like when DOGE is doing it. And ordinary Americans will say, “They said they were cutting government and what happened was all this stuff that I relied on just stopped. The system broke. Programs that were necessary for my quality of life disappeared without a private alternative in place to pick up the slack.”
Kunin: It’s the reverse. I might be a naive optimist, but I actually think it’s going to go the other way, which is that DOGE is doing this so badly, so incompetently, so loudly, so clumsily that any other attempt in the near future, when it is still somewhat fresh in people’s mind, will seem genius level and acceptable by comparison.
Sanchez: Well, I am more worried than Noah is about that. Very early on, I found myself thinking that they are using rhetoric that sounds like things I believe—which makes me very worried that they are going to make those ideas toxic for a generation, even though that rhetoric is being used largely as a fig leaf for other goals.
“There is a pattern of conduct across this administration that seems to be calculated to hollow out the technical expertise that is required to understand how the technical systems DOGE is building are operating, and whether they’re operating legally—to hollow out the oversight mechanisms that exist to identify misconduct or illegal conduct within government and to frankly discipline the people within government who are responsible for making accurate representations to the court. And, on top of that, at least so far, this administration has shown a kind of shocking willingness to simply choose not to comply with court orders it decides it doesn’t like. So that combination of facts makes me not particularly sanguine about the ability of normally very slow-moving courts to police the legality of very, very rapid changes that DOGE is making.” — Julian Sanchez
But if you think this is what it looks like, or this is what happens when people who say, “Hey, government ought to be smaller and function more efficiently” get the ability to implement that idea, a lot of people are very reasonably going to look at those results, especially as the concrete impacts of that are felt, and say, “Well, if that’s what happens when you let people who say they want to shrink government and make it leaner get into power, I don’t want anything to do with anyone who’s saying that.”
Kunin: Look, it is still really early, and it feels like two years—but it has only been a few months. I was much more of a mind to agree with Julian two months ago—and they’re just bumbling incompetents. Their idea that they were going to improve Social Security by recoding it from an archaic language called Cobalt ... the fact that they were going to invest all this technical capability and political will to do this, versus any of the other more useful things, was a sort of mask-off moment for me that there was literally zero leadership, zero administrative capability within DOGE.
Sanchez: But, again, I think that this may be a case in point: Look, in the long run, to be fiscally viable and not just eat an ever growing percentage of the federal budget, we do need some kind of reforms to Social Security as the number of people paying in the labor force declines relative to the number of older people collecting benefits. That is not indefinitely sustainable. It’s going to take political courage to figure out how to rejigger the system so it is sustainable. But if people talk about Social Security as a third rail, if the attempts that people remember to make Social Security more efficient is, “Well, the website’s down all the time” and “I can’t get help on the phone anymore” and—this hasn’t happened yet, but I would not be surprised if it does soon—“hey, I got pulled off the benefit rolls incorrectly, and now I don’t know if I can make rent” … it’s not going to be good.
Whatever you think about Social Security, it’s the system we have. And the way to fix it is not abruptly stop paying benefits for people who have paid in and are now counting on that because that is an entitlement the law provides them to make ends meet month to month. If you do things the way DOGE is doing it, I think you sort of ramp up the voltage on that third rail and people end up saying, “My God, if that’s what happened last time, no, don’t touch this thing.”
Kunin: So the reason why I look at it a little bit differently than Julian right now is because the things I would do to reform entitlement—and the three major ones, Social Security, Medicare, and Medicaid—are all policy changes.
“Even if you think the prime directive here is to make the number go down and that they actually are detecting waste and fraud, they’re still failing on the numbers. If you actually use a formal methodology to calculate their savings, and I’m being a little more generous than some other people right now, I’m looking at 1.1% of the federal deficit in 2024. That’s it. Where is the money? If they are so confident there’s so much fraud, waste, and abuse, and it’s so easy to find, why’d we only cut 1.1%? Well, because most of the deficit is in fact not in fraud, waste, and abuse, but in entitlements and defense spending. And the only people who can really fix that is Congress.” — Noah Kunin
So I think they’re materially different than the kinds of things DOGE is doing, and they won’t be associated. Like if we raise the retirement age, that’s a policy change. If you change how people’s income is taxed, and raise the cap on how much of that can be taxed for Social Security, that’s a policy change. That’s not a technocratic change.
Powell: I want to turn now to the legality of this, because I think one way we can talk about the last 20-some odd minutes of our conversation is: DOGE is doing all of this stuff and it’s bad. So DOGE is bad at its job. But another way we can think about DOGE is: Should DOGE—as the way it’s constituted right now—be doing this job at all? Is it allowed to be doing the things that it is doing? So let’s say it was cutting wisely across the board, would it be legal for it to operate? To start with, Elon Musk seems to have a lot of power in the federal government. He’s running this thing that looks like an agency, controlling a whole lot of other agencies. But I don’t remember him having his day in front of the Senate to be confirmed.
Sanchez: So, there are dozens of lawsuits challenging different aspects of DOGE. This is one of them. I think this is one that, among others, some state attorneys general have brought. We spoke on WatchCats to Keith Ellison—former congressman, now an attorney general—about some of these lawsuits.
One of the claims that was brought is that, essentially in violation of the Appointments Clause, significant practical authority has been delegated to Elon Musk that makes him effectively tantamount to at least a Cabinet secretary, and maybe even something more elevated than that, without any kind of process like that having been gone through. Recall, Elon Musk is a special government employee, which is a weird temporary status. The formal head of DOGE is someone named Amy Gleason, who apparently was announced while she was on vacation in Mexico. I’m not sure she knew she was the head of DOGE when that was made public after a kind of almost farcical process of the courts trying to get an answer to the question of, Who is legally the head of DOGE?
So one issue has to do with whether Musk is lawfully exercising power as a kind of unelected temporary appointee. And the other question, of course, has to do with impoundment. So the way they’re making a lot of these cuts is by effectively refusing to spend money Congress appropriated. Now, Congress appropriates a lot of money for a lot of purposes that I might disapprove of, or at least think it should be a lot less. So it might be that in principle there are plenty of instances where, as a policy matter, I might like that we’re not going to spend money on this particular kind of grant anymore. But that is still the prerogative of Congress to say how money is spent. It is not the prerogative of the White House or the president or a South African billionaire tapped by the president to decide, “Well, Congress got it wrong. We’re going to spend the money differently. We don’t care that a statute was passed and signed into law saying, ‘this is how these funds are going to be dispersed.’” That is effectively how they’re doing it.
One of the other functions of DOGE is that there are individual Republican legislators who played a role in assigning some of that funding and they don’t want their funding cut. So Musk has talked about establishing a hotline so that if you are a friendly senator, a legislator in the good books of the White House at least, you can call up and ensure that the funding that you care about is restored. But the funding that was statutorily approved ... the people who care about it are not allies of the White House. They cannot count on having that restored. I mean, that places entirely too much power in the executive in a way that is at odds with our constitutional order. So that’s a second legal problem.
A third legal problem has to do with federal privacy statutes—a big, big part of what DOGE is doing. This falls into the bucket of centralizing and consolidating power within the executive branch, often in a way that might make it easier to abuse that power; trying to consolidate federal databases, often containing very sensitive personal information about ordinary citizens, federal workers, and taking information that had been deliberately siloed and making it intra-accessible, creating a sort of master API that will let people like Musk and Trump combine and analyze data across a lot of different databases.
And we see some of how they intend to do that now, which is, “Looking across these different databases, we think you’re in the country illegally, or maybe legally, but we don’t like you.” We’ve seen people have their visas revoked for writing op-eds that the administration disapproved of. People here with green cards, or on lawful student visas, find their status provoked. It makes it a lot easier to target people. This is why, in the 1970s, we passed legislation limiting the way data can be used. If you collect data for tax purposes, you’re not supposed to use it for other purposes. If you collect data for Social Security purposes or for immigration purposes, that is not (with specified exceptions) supposed to be used in a different way. But what we’re seeing is a kind of consolidation of data, in part to try and identify either illegal migrants or in some cases legal migrants who for whatever reason the administration is unhappy with, and, for example, cut off their access to the Social Security system—not to benefits necessarily, but just to cancel their Social Security number as though they were dead. If, suddenly, your Social Security number is invalid, it makes it difficult to do a whole lot of other things in American society.
Given how we’ve seen DOGE operating, just within the federal workforce, asking people questions about, “How do you feel about DOGE?,” “How do you feel about Trump?,” “How loyal are you in positions that have traditionally been apolitical and not required you to be a Republican or a member of the president’s party to work in that federal employment position?” ... I don’t think it’s a stretch to imagine the use of consolidated databases to say, “All right, people who have donated money—as shows up on their IRS forms—to progressive causes or pro-immigration causes or pro-Palestinian causes, they’re not eligible for federal jobs. They’ll be disfavored for federal grants.” That kind of data consolidation is very likely to be used in ways like that.
“One of the things that worries me is that so much has been done so quickly, and with kind of dubious legality or just illegally, that the question becomes, ‘Well, can you undo it as quickly while clearly adhering to law and doing it with a kind of appropriate process?’ Probably not. If you’re willing to break—or at the very least severely bend—the law, you can move very quickly. So then the question becomes, ‘OK, if power changes hands and you have these agencies fully infiltrated by cronies, running systems architected by the DOGE team and operating in a particular way, can you undo that without using the same methods to achieve the same speed?’ … If the answer is “No,” I think there’s a really unfortunate situation where either you just normalize permanently ignoring those rules—maybe that would be the worst scenario—or you have to reconcile yourself to the idea that what this administration did in a matter of months is going to take years to undo if you care about following the rules.” — Julian Sanchez
Again, we don’t have to look at hypotheticals. We just have to look at the things they’re actually currently doing. And it is that kind of use of the information that caused Congress in the ’70s and then again in the early 21st century to establish statutory barriers meant to prevent that kind of consolidation, that kind of access, that kind of repurposing. That’s three buckets of legal problems with DOGE.
Kunin: Let me add some comments on each of those buckets. I’ll do so by, first, distilling how DOGE currently works, and then steelmaning their argument a little bit. You’ll see just how well I think of their steelman argument because I’ll open it up with an allusion to the Nazis, which is that it’s a bizarro version of the, “I was just following orders” argument, except mirror-reflected.
So the way they’re getting away with it is saying, “Whoa! Look, I didn’t write the memo. I didn’t sign the order. I didn’t push the button. All we did was advise the departmental and agency heads which have gone through the Senate, which have gone through all the legal procedures, which have gone through all the background checks. We just internally advised them to do it and they did it.” Now, one of the ways to understand this—and this is something we talked about even in the early days of our podcast, WatchCats, when we interviewed Mikey Dickerson, who was the founding director of the United States Digital Service, the precursor organization to DOGE—is that the most important currency in any large organization, including the federal government, isn’t your legal or bureaucratic power, but the currency of influence. Specifically in a private organization, it might be the CEO; in the executive branch, it’s obviously the president. And merely the access of the director of USDS, now the super-powered DOGE version, the level of access that person has to the president’s chief of staff or the president themselves is what gives them power to tell the agency or departmental heads to get in line or to fall on their sword or to do the thing.
So this is the flow chart of how it works: DOGE developers will go in, extract insane amounts of data, sometimes also sending it to Russia as we’ve seen, and they will come to some wild conclusions about that data. Sometimes it’ll be right, because any stochastic process is sometimes right. And oftentimes it’ll be wrong. That goes up to Elon Musk. He decides how he feels about it. He tells an agency that it had to do something. They either do it or they don’t. He goes and whines about it to the president or the president’s chief of staff and then they make that person do it. Now the legal argument against that, as refined as it might be, is, “If it looks like a duck, quacks like a duck—hey, it’s a duck.” If he is acting like an agency head, not just an agency head, but a super-agency head, a head of multiple agencies, unilaterally deciding if certain agencies or entire branches of agencies should or should not exist, regardless of their congressional existence or justification, well, then he is one. And that argument might go quite far.
On the privacy side, one thing I would add here is: some of the more restrictive privacy requirements are tied to language about the granting of a benefit. This is a flaw in the original Privacy Act, and something people have complained about both in and out of government many times. Which is: the important thing might not always be the provision of the benefit but the withdrawal or dissolution of the benefit. So perhaps all of this will result in a better Privacy Act for future generations. And we’ll see how much that argument actually gets purchased in the courts to begin with.
Powell: One of the other kinds of arguments goes something like this: “Yeah, they’re doing a lot of bad stuff. But the kind of hyperbolic concern about DOGE to the point of saying, ‘Nothing about this is good; no one should be involved in this; it’s all rotten’ …no, the courts are going to stop the unconstitutional stuff. They’re going to roll back. They may be breaking some laws, but we have the courts. The courts will tell them to stop doing it. So the only parts of what DOGE is doing that will stick are the parts that were legal, that don’t get overturned in some way. So stop freaking out about it and just focus on trying to maximize the good stuff or the legal stuff that they’re doing because the rest of it will get taken care of.”
But that argument seems to run into real problems with some of the privacy concerns that you’re mentioning. We get these leaks, we get whistleblower reports, like the one earlier this week that showed the spike in data going out and the attempts that looked like they were coming in from Russia, using immediately valid logins and passwords and all sorts of other really scary stuff like turning off logging and so on. We’re only finding out about that because there was whistleblower. So who knows how many other things like this are happening that we’re not aware of. All of that stuff can do extraordinary damage.
If that data is out in the world, the data is out in the world. And if the Supreme Court rules, “You shouldn’t have done this,” they can’t ask all of those actors for the data back. Or if that data is used to find more people to ship off to El Salvador, even if those people come back, they’ve spent some period of time in a torture prison. So the courts can roll back some of it, but they’re doing grave damage right now.
Kunin: Yeah, as a former public servant, the really important thing for your audience to understand is that the vast majority of public servants—I’m talking about 98% of them—are intensely, almost diagnosably, risk-averse. By the time somebody in the federal government blows the whistle on something, you should assume it has happened anywhere from 10 to 10,000 times already. I have never read an inspector general report or a government accountability office report where it was like, “Hey, and the first time it happened, it was immediately sent to a whistleblower case.” That just has never happened. Do not assume it is happening here. So, yes, to your point, the Privacy Act stuff, or privacy-writ-large stuff, has to do with our national security.
“As somebody who has worked in the federal government for many years, worked for state government for many years, and comes from a progressive ideological side on this … I just want to point out that literally nothing in my top 10 list of egregious bloat in government has been touched whatsoever. In fact, in many cases the people fired were the best people you would have worked with in order to actually target those things. So when you go into the preexisting technical capacity, the preexisting teams who are there to create efficiencies, centralized services, transparency in government, and they are some of the first people that you release, it sort of belies the whole thesis of what you’re here to do.” — Noah Kunin
That’s there not just to protect individual rights, but also to protect us as a nation. And the problem with concentrating power, even when you get increased efficiency and effectiveness out of it—which we’re not getting—is that the trade-off is that if something collapses with the centralized power, the centralized system, the centralized process, the blast radius is that much bigger. You do not want to concentrate, for example, full administrative super-user access to important technical systems with individuals, not because it doesn’t create efficiency and effectiveness—sometimes it does, sometimes it doesn’t—but because if that one person gets hacked, well, then you’re totally screwed and our adversaries understand this. Instead of diffusing their power and their capabilities across a very large, very heterogeneous, (from a technical standpoint) population, they can concentrate it on somebody who’s getting deep dives on them in the New York Times constantly, so they have everything they need to spearfish that person.
Sanchez: When we were together at Cato, Aaron, I worked a whole lot on FISA and foreign intelligence surveillance and NSA issues, and one thing that gets lodged in your mind very solidly when you work on those issues is that a prerequisite for a court getting to evaluate the legality or constitutionality of something the government is doing is a court becoming aware that that thing is being done. And I would say there is a pattern of actions by this administration that seem very much calculated to reduce the risk of that ever occurring. Within the first few days, one of Donald Trump’s first actions was to fire a slate of inspectors general not really for any particularized reason, but just to eliminate those people wholesale in order to install replacements. That is not the normal way these things work. It’s not normally how the inspector general is viewed as a role.
We’ve seen attempts to hollow out technical expertise elsewhere within the government. And this is thing people found very strange: you had USDS, you had 18F, which is another kind of tech incubator within the government, and since then actually several more centers of technological expertise within the government, that if your goal really was to make things operate more efficiently, these are precisely the people you would want to draw on. These are the people who have technological expertise, but also have some experience with, and understanding of, the agencies within which they’re installed. And so you think, “What a great resource to leverage in order to give them even more authority and bring them in to enable us to more quickly figure out where the fat can be cut and how efficiency can be optimized.” Instead, there has been an effort to effectively hollow out those other centers of technical expertise.
Which means you actually make it a lot more difficult for people to understand whether those technical systems are being operated and structured in a way that is compliant with the law. And we saw, actually quite early on, one of the lawsuits over DOGE’s access to some of these systems involved the question of whether a particular DOGE staffer had read-write access to the Treasury Department’s payment system—an incredible, central artery that determines how cash flows within the government. And if the agency says, “We want to send out this money that is part of a disbursement that Congress approved,” but the people controlling that system say, “No, that transfer is not going to happen,” it doesn’t happen. Initially, it was represented to the courts that DOGE did not have the ability to write to that system, to change the code. And then later, lawyers had to go back and say, “Well, actually, it appears they were granted that access, but we promise we’re taking it away now.”
“One of the other functions of DOGE is that there are individual Republican legislators who played a role in assigning some of that funding and they don’t want their funding cut. So Musk has talked about establishing a hotline so that if you are a friendly senator, a legislator in the good books of the White House at least, you can call up and ensure that the funding that you care about is restored. But the funding that was statutorily approved ... the people who care about it are not allies of the White House.” — Julian Sanchez
Another thing we’ve seen is that attorneys who make disclosures to courts that are accurate, but the government did not like, get fired. The lawyer who admitted that Abrego Garcia—one of the many people sent to that Salvadoran torture prison—that the Supreme Court ruled had been sent there in error and in violation of a court order, that lawyer did not come to that conclusion. There were sworn affidavits from senior DHS personnel agreeing that he had been sent there in error, and in violation of a previous court order. The person who made that accurate admission was put on leave and later fired and described as a saboteur. So, there is a pattern of conduct across this administration that, again, seems to be calculated to hollow out the technical expertise that is required to understand how the technical systems DOGE is building are operating, and whether they’re operating legally—to hollow out the oversight mechanisms that exist to identify misconduct or illegal conduct within government and to frankly discipline the people within government who are responsible for making accurate representations to the court. And, on top of that, at least so far, this administration has shown a kind of shocking willingness to simply choose not to comply with court orders it decides it doesn’t like. So that combination of facts makes me not particularly sanguine about the ability of normally very slow-moving courts to police the legality of very, very rapid changes that DOGE is making.
Powell: What happens after DOGE, given all of this? So, let’s say that at the tail end of 2026, we get massive swings—Democrats take back both houses—and we get some oversight and are able to shut it down. Or let’s say in 2028, we get Democratic victories across the country. In both cases, Elon Musk is no longer running this thing called DOGE, and this thing called DOGE ceases to exist. How much of what we’ve seen comes back? What we’ve seen cut, how much of this gets undone, and how much of it is permanent?
Julian, you mentioned impoundments. Congress has not reduced the spending. It has not changed the statutes granting these agencies certain powers, or telling them to do certain things and then allocating certain amounts of funds to them to do that. It’s just that the federal government is basically deciding not to spend that money anymore, not to give the money to the agencies it’s been tasked to give them to. But that means that money is still there. It still is allocated. It still is appropriated. So Democrats could step back in and just say, “Okay, NIH had its budget reduced by this amount, or at least it stopped receiving this amount in checks each month. We’re just going to turn those checks back on.”
So, looking ahead, there’s a lot of damage happening now, but if DOGE were to end, how much of this damage is undoable? How much will be just immediately shifted back? And how much is permanent, that we’ve kind of done unredeemable, unfixable long-term damage?
Kunin: What’s permanent, I think, is the normative change. I say this a lot because it’s true: Most of what you think is a law is a regulation. Most of the regulations you think of aren’t regulations, they’re administrative actions. Most administrative actions are really just policy papers of a particular agency. And most policy papers are just groupthink within an agency. So, when you blow up the norms constraining government action, it is very-difficult-to-impossible to put them back together again. We have not seen that in the modern government, that is to say, the modern government FDR made. For example, FDR unilaterally decided the Social Security number was going to be the numeric identifier for all humans in the United States. Congress didn’t decide that—FDR decided that via executive order. No one ever agreed to it. Many people have objected to it. It is still practically, in all regards, the rule as we see it today in 2025.
So when you break a norm in the federal government, it is very hard to put it back together again. In fact, I’ve never seen it happen. Which is why even if you are somebody who agreed with the original conception of DOGE as articulated in the opinion papers by Musk and Vivek Ramaswamy last year, this is the worst possible scenario for you. Anything short of complete victory is a massive problem for you because when power flips again, are these people going to disassemble these structures? Are they going to disconnect the massive databases which allow them better purview into the functioning of the civilian government? Did Obama disassemble massive parts of the surveillance state? I don’t think so.
So it seems implausible to me that that is gonna happen. If you don’t utterly obliterate the opposite side, all you’re doing is building a better machine for them to use to their own ends in the future.
Sanchez: In a way, I think DOGE is kind of morphing into what Donald Trump might call a deep state. The formal head of USAID is someone who was a DOGE staffer and now is running the show. There are a bunch of cases where the formal or acting head of some government agency is now someone who was initially attached to DOGE and is still in effect attached to DOGE.
One of the things that worries me is that so much has been done so quickly, and with kind of dubious legality or just illegally, that the question then becomes, “Well, can you undo it as quickly while clearly adhering to law and doing it with a kind of appropriate process?” Probably not. If you’re willing to break—or at the very least severely bend—the law, you can move very quickly. So then the question becomes, “OK, if power changes hands and you have these agencies fully infiltrated by cronies, running systems architected by the DOGE team and operating in a particular way, can you undo that without using the same methods to achieve the same speed?” If the answer is “No,” I think there’s a really unfortunate situation where either you just normalize permanently ignoring those rules—maybe that would be the worst scenario—or you have to reconcile yourself to the idea that what this administration did in a matter of months is going to take years to undo if you care about following the rules.
And, in some cases, if you effectively destroy an entire agency that in hindsight we might say, “I’m not sure it was a great idea to get rid of this” ... you can fire 10,000 people pretty quickly, but if you decide “Gosh, that was a mistake—we ought to rehire them,” well, that takes a hell of a lot longer. Probably by the time you’re in a position to reverse that decision, the specific people have moved on and found other jobs, and decide that federal employment is not that attractive to them anymore. So, yeah, you can blow up a house a lot faster than you can rebuild it.
Kunin: To answer your question with any specificity or granularity, we would need to iterate through every single program, and also make a determination that we can’t fully make right now, which is in which circumstances did people legally or illegally preserve or protect or backup the data or code? For example, we’re learning that the IRS is going to delete IRS Direct File, which makes it free and easy to file your taxes directly with the government. Something you would think people on the right would like. Is that code going to be backed up, that it could just be restored relatively quickly? In most cases I would say, at this point especially, anything the government did in an open source way or in a transparent way—those have been backed up. I was talking about the 18F team: most of that work has never been backed up off of government systems and so could be restored. It was kind of the whole point of making open sources—“Hey, iterate on our work, remix it, and if one of our systems goes down, now there’s a distributed set of places where that’s stored that we can restore.”
Powell: So for listeners who have been listening for the last 50-some odd minutes in a growing sense of anxiety or panic and want to continue to ride that high, tell us about WatchCats.
Kunin: Sure: WatchCats is the podcast Julian and I run where we cover DOGE specifically.
Sanchez: And if you like short-form content, we are on all the things. We are Watch CatsShow on YouTube and Instagram and—God help us—TikTok. And watchcats.show is where our full podcasts go up.
© The UnPopulist, 2025
Follow us on Bluesky, Threads, YouTube, TikTok, Facebook, Instagram, and X.
We welcome your reactions and replies. Please adhere to our comments policy.
Share this post