Subscribe here: Apple Podcasts | Spotify | YouTube | Overcast | Pocket Casts
In politics, compromising with one’s ideological opponents is like walking a tightrope while both your allies and foes jeer at you. Democrats, now the out-group facing a Republican trifecta, will have to decide when to fight nominations, laws, and executive orders and when to step into that circus ring.
Jennifer Pahlka, a former Obama administration official and an author of a new report on government reform, kicked up a storm some weeks ago when she encouraged Democrats to work with Elon Musk’s Department of Government Efficiency (DOGE).
“We do need to talk about government reform, and while I’m sorry the conditions are quite a bit less than ideal, I think it’s time we admitted they were always going to be. Democrats did not do this work,” Pahlka wrote.
Pahlka was in part responding to arguments by people like Leah Greenberg, a co-founder and co-executive director of the progressive group Indivisible, who scolded Democrats for promising to work with DOGE: “Democrats should be planning to fight these corrupt plutocrats, not offering to work with them.”
On today’s episode of Good on Paper, I explore whether liberals can actually find any common ground with DOGE and whether Pahlka’s focus on what she calls “state capacity” actually explains government dysfunction. (This episode was recorded earlier this month and references Vivek Ramaswamy’s involvement with DOGE, before it was reported that he would no longer be a part of it.)
“It’s an uncomfortable position to be in because it’s not like I have a crystal ball to know what Musk and Ramaswamy are going to do. And I may disagree with some of what they do, or maybe a lot of what they do, but they’ve really kind of moved the Overton window and the conversation about this inefficiency, the sludge. And I think that’s valuable, frankly, and I want Democrats to kind of get in the game of that reduction,” Pahlka tells me.
The following is a transcript of the episode:
Jerusalem Demsas: While 75 percent of Democrats tell Pew that they prefer a bigger government providing more services, fewer than a quarter of Republicans say the same. This divide is a persistent feature of modern American politics and can make it seem like government-reform efforts—like civil-service reform and getting rid of costly, inefficient regulations—are the purview of the Republican Party.
Elon Musk and Vivek Ramaswamy certainly think so. They aim to cut $2 trillion from the roughly $6 trillion federal budget under the banner of the Department of Government Efficiency, or DOGE. This could be a nearly impossible feat, seeing as discretionary spending by the federal government was only $1.7 trillion in 2023. Perhaps realizing this conundrum, Musk and Ramaswamy have negotiated against themselves and revised the number to $1 trillion or $500 billion. We’ll see.
[Music]
Demsas: I’m a bit tired of how reasonable-sounding concerns around government efficiency and effectiveness get shoehorned into a witch hunt for government waste. There are serious problems with how the federal government’s processes and regulations harm economic growth and the effectiveness of important social-welfare programs. I’m skeptical that focusing on budget cuts does much to change that, but I’m also frustrated that it seems the only political actors talking about this seriously are on the right.
My name’s Jerusalem Demsas. I’m a staff writer at The Atlantic, and this is Good on Paper, a policy show that questions what we really know about popular narratives.
My guest today is Jennifer Pahlka, a senior fellow at the Niskanen Center and founder of Code for America. She worked in the Obama administration as deputy chief technology officer, and her recent book, Recoding America, argues that the federal government is hobbled by its inability to implement its stated priorities.
Jennifer has a message to people across the political spectrum: If you want government to work, you need to reform it. In that vein, she’s much more optimistic than I on the potential for good-government types to work with DOGE and the Trump administration.
Demsas: Jen, welcome to the show.
Jennifer Pahlka: Thanks so much for having me.
Demsas: I am so excited to have this conversation. I feel like me and you—our work has been in conversation for years now, and we’ve been at some of the same conferences and things. So I’m really excited to dive in.
Pahlka: Me too.
Demsas: So you’re someone who has worked in government and now works trying to make government better. Give us the liberal case for government reform.
Pahlka: Well, I feel like liberals talk about government reform. I’m not sure they necessarily need to be sold on it so much. I think the kind of reform that we need today is a little bit of a hard pill for liberals to swallow, because we need government to sort of be faster, a little bit less process oriented and more outcome oriented. And there has been a pattern, I think, of liberals being very fond of process, of additional rules and regulations, for all the right reasons.
And with great success, right? I mean, the environmental movement really cleaned the air and our water, and that was through regulations. The civil service went from being a place where you would get a job because you were someone’s friend or you’d given money to a campaign, to a professional place. And those are all rules and regulations that have made government better and fairer and made our country better.
But we’re kind of at a point where there have been so many of them, and they’ve stacked on top of each other so much that we’re just moving very slowly. And so the kind of reform I’m talking about now does involve some things like maybe reducing, especially, regulation on government itself—reducing procedures and moving a little faster. And that is the part that liberals need to be convinced about, let’s say.
Demsas: You have a new report out with the Niskanen Center called “The How We Need Now: A Capacity Agenda for 2025 and Beyond.” What’s the main takeaway? What are you trying to solve here?
Pahlka: We’re really trying to help people understand that when you think about government reform, it just seems so big and impossible. So we’re trying to break it down and say, Actually, there are specific things that you could do if you want a government—and this could be, you know, we wrote it for federal government, but you could use it for state or local government as well—if you want government to be able to do what it says it’s going to do, to achieve its policy goals.
And so those things come in four buckets, you know—four pillars. The first thing is: You need to be able to hire the right people and fire the wrong ones. The second is: You have to reduce the procedural bloat. We’ve also talked about that as reducing the administrative burden on public servants—in addition to on the public, but we’re really talking about on public servants—so that you get more public servants focused on outcomes and less on process and compliance. The third thing is: You need to invest in digital and data infrastructure to enable all of this. And there’s a bunch the federal government could be doing at the start of the Trump administration to do that, including getting the United States Digital Service funded again and the Technology Modernization Fund funded again.
And then last, and the one I’m most interested in, is that we need to close the loop between policy and implementation. And what I mean by that is: Right now it functions as this sort of waterfall process, where you have a law, and then maybe it gets handed off to an agency to write regs, and then, you know, into the implementation phase. And it doesn’t ever sort of circle back and say, Is this working? What are we learning? What needs to be adjusted?
And especially in the era of Loper Bright, this decision from the Supreme Court that’s really going to change how the executive-branch agencies relate to Congress, we have kind of an opportunity to rethink that relationship. And I think we should rethink it along the lines of creating feedback loops that let us adjust along the way so that we actually get the outcomes that the laws and policies that we pass intend.
Demsas: I think you’re right when you talk in the abstract. Like, most people, liberal or conservative, would say, Yeah, you know, red tape is bad, and the government should definitely update technology, and, you know, it’d be good if we had a government that worked efficiently. And then when you get into the actual policy prescriptions and the trade-offs, things become more controversial, particularly when you’re talking about civil-service reform and regulatory reform.
So one of the third rails has long been hiring and firing. I want you to talk to us a little bit about what’s broken in that space and how you would change it, and I’d also like you to talk to us about the story of Jack Cable.
Pahlka: Oh, gosh. Jack, yeah. Well, first of all, what’s not broken? So, you know, we had the Civil Service Reform Act of 1978, which established these Merit System Principles. They are very good. If you read them, you are very likely to agree with them. They talk about integrity and fairness and, you know, promoting people on the basis of merit. They’re called the Merit System Principles. And I think they are a strong foundation for our civil-service system.
The problem is that (A) that was 1978, and so we’ve had many years now for those things to be operationalized with a lot more ornaments that have been attached to them, right? It’s not just those principles. It’s the regulation and the guidance and the operating manuals and the processes and the forms that have derived from those that have really, I think at this point, kind of perverted their intent.
So for instance, we say we’re going to hire on the basis of merit. We also say we’re going to hire in a way that’s nonbiased. Well, what happens is that you have HR managers who kind of control the process of selecting a candidate. What they do—I’ll give you sort of the very specifics of how this works in 90 percent of cases. This is not the accepted services, and it’s not political appointees, but open-to-the-public, competitive jobs. They get, like, a big pool of resumes, and they have to down select. The first down select they do is by looking for exact matches between the language on the resume or cover letter and what’s in the job description.
Demsas: So if you copy-paste the job description into your resume, that’s, like, points?
Pahlka: Yes, and I have a friend of mine who’s in my book—I actually originally interviewed her about this. I didn’t put that in the book. But she was looking at a resume that had not just been copied and pasted, but copied and pasted and not reformatted. Like, that part was in a different font.
Demsas: Oh my god.
Pahlka: Like, the same font, right? And she points this out to the HR manager, and they’re like, Yeah, that means that this person’s the most qualified, because it’s the exact same language. And she’s like, This person is clearly unqualified because they didn’t even know to reformat. And this is not an outlier. Like, this happens a lot.
So first they’re looking for these exact matches. And then they take everybody who was really close in language—and also, by the way, who has something called a government resume, which is different from a private-sector resume, and you have to know that somehow, magically, before you apply. Then from that pool, they send everyone a self-assessment questionnaire, and everybody who marks themselves as master, and I literally mean master—I think that’s the top rating in a lot of these—they make the next down select, so they move on to the next pool.
Demsas: Wait—so if you just say that I’m a master at this, like, without any double-checking, you just get to move forward?
Pahlka: I mean, somebody could send me a self-assessment saying, Are you a master programmer in Python? And I would just be like, Yes, and I would move into that pool. Nobody checks it. It’s actually worse—not just that no one checks it; it’s that the HR people will tell you that subject-matter experts (SMEs) are not allowed to be in that part of the process.
I mean, there are processes that do include them, and I can get to that, but you can’t have SMEs look at these resumes and exert their judgment, because they may introduce bias into the process. Now, again, I think the idea of keeping bias out is something I agree with, and I’m going to assume you agree with, and most people agree with. But that’s not actually keeping bias out, right? That’s what I mean about sort of a perversion of the intent.
But anyway, so you have this now smaller pool of people who are great at cutting and pasting and great at, you know, self-aggrandizement—or really what it is, is they just know what to do. They know how to play the game. And then from that list, you apply veterans’ preference. In other words, any veterans in that pool float to the top, and that’s the “cert,” which is just the name for the list the HR manager gives to the hiring manager. That’s the cert that the hiring manager is supposed to choose from. So this is not consistent, to me, in my mind, with Merit System Principles of fairness, and not bias, and certainly not merit.
And so what you are looking at when you see that kind of behavior is a system that’s designed to be completely defensible from the critique of your judgment, because you have exercised no judgment at all. And I understand why people defend them and do these processes to be defensible, but I think, in the end, they come up actually indefensible.
So I learned about this process, in part, through a young man named Jack Cable. I was on the Defense Innovation Board at the time, and he won the Hack the [Air Force] contest. So all these security researchers from around the country come together, and, you know, they’re looking at bugs and security bugs and Pentagon software. This young man wins the whole contest. He’s the best out of the group.
And of course, you know, the people at the Defense Digital Service and other parts of the Department of Defense say, Great. We need this guy on our team. He applies with a resume that lists his programming languages and the frameworks that he is expert in, and he is cut in the first batch because he did not cut and paste. And the people reviewing his resume see this sort of gobbledygook of programming languages—they’re not technical people. They’re not even sort of supposed to know what those are, and so he gets cut.
And it’s not just that—then the Pentagon folks intervene and try to get him hired something like 10 different times. He does eventually get hired, but even with these interventions from people in power, and sort of as it escalated with increasing levels of power in the Pentagon, this very talented security researcher continues to get cut from the process before hiring managers ever see his resume.
Demsas: Wow.
Pahlka: Oh, and one more thing: He’s told by the HR people along the way—he’s quite young—they say, Go work at Best Buy selling TVs for a year, and then you’ll be qualified for this job.
Demsas: Wow. And I feel like in that time period—obviously, this is an exceptional case where a lot of people took effort to try to get him hired. But, you know, private-sector processes are much faster than this. And what’s most likely to happen is you get all of these top performers going into the private sector.
Pahlka: Oh absolutely. And I mean, it’s just a testament to his commitment that he stuck through it. And that young man has actually stayed in government. It’s amazing. He’s done some really wonderful work.
Demsas: So there’s that part of the government reform that you talk about, which is about hiring and firing. I mean, obviously, we only touched on it a little bit. But the other part of it that you focus on a lot is around regulatory reform. And one of the laws that you’ve pointed out is the Paperwork Reduction Act. Can you walk us through how that act hobbles government?
Pahlka: Yes. I will say, we’ve had some good progress on PRA, and I should also mention that we’ve had some good progress on that assessment problem. The [Fair] Chance to Compete Act passed both houses of Congress, and it actually directs agencies to stop using those self-assessments.
I have high hopes for it, but I also will say: There was an executive order saying that under Trump. Biden renewed that executive order. And it hasn’t really gotten the agencies to change their practices yet. So there is an implementation issue, I think, and we’re going to really have to watch if the [Fair] Chance to Compete Act does what we hope it does.
Demsas: Wait—if both Trump and Biden issued the executive orders, why aren’t the agencies doing it?
Pahlka: It’s very hard to change the practices of agencies, even under direct order.
Demsas: Yeah. Mechanistically, though, what’s going on? Are there people who are just refusing to change? Or, like, what’s happening?
Pahlka: Well, it wasn’t in statute. I don’t think there was a timeline or a deadline for it. I think if you really read the language and translate it into, you know, what’s practical, it’s sort of more encouragement. I mean, it does direct them, but there’s sort of very little teeth in it.
Government moves slowly. HR people move particularly slowly. I mean, until you fix some other problems—like how detailed it is, how many rules you have to comply with in order to use a subject-matter expert in that process—it actually is, like, enormous amounts of time to run a hiring process using real assessments.
Demsas: So tell us about the Paperwork Reduction Act. What is it doing, and how is it preventing government from acting quickly and nimbly?
Pahlka: So there’s sort of the general level of it, which is just: It’s a lot of work to comply with. So imagine you’re charged with implementing the CHIPS and Science Act, for instance, and you want to stand up a form to allow companies to express their initial interest or even apply. You want to know early on what kinds of projects might companies, you know, bring to the Department of Commerce, to apply for funding under CHIPS.
Well, you can design the form. There’s going to be a lot of process and a lot of stakeholders that want to look at it. You don’t get to write something up and throw it up on the internet. But once you’ve done all that work for your internal agency stakeholders and sometimes cross-agency stakeholders, then your form, because it’s an information collection, is subject to review by the Office of Information and Regulatory Affairs at the White House.
And so you’ve got to sort of do all this pretty heavyweight documentation of your form and why you’re asking these particular questions, and you submit it to them. And because that process needs review by people—there’s only so many people in OIRA, the Office of Information and Regulatory Affairs—and because the process requires two separate times that you post it to the Federal Register, get comments from the public, respond to those comments, then potentially do a revision, then post it again, get comments, respond to those comments. And those time periods are designated in statute—I think it’s 30 days the first one and 60 days the second one—like, right there, that’s at least a month, but more because you have to do all the lead-up and then follow-up.
The average time to get through—or actually, I think it’s the minimum time to get through—a standard PRA review is nine months. And that’s just to get one form up. And it can be longer. Now, there is a fast-track process. If you get a fast-tracked application, that runs out in six months. So in six months, you’ll have to do it all over again. When you’re supposed to have moved on to the next phase of your project, you’re kind of going back to zero.
And there’s certainly value in a centralized office knowing all the things that agencies are asking the public, or companies, or anybody who would be filling out a form. And there’s absolutely value in knowing, like, Oh we have this data here. Maybe we shouldn’t be asking for it. Maybe we can get it from another agency. That would be, like, the best use of this kind of centralized function. But we have let this become quite a heavyweight process that really slows agencies down.
Demsas: You’ve outlined quite a few things in your public research and writing around how you think government—both whether we’re talking about Congress but also the executive branch—should reform in order to make things more efficient. You know, some of these things are just common-sense requirements to make hiring practices align with things that people think are good, like merit.
But most people who are talking about this, I think, are often on the right. And increasingly, I think this conversation is being brought up by people like Elon Musk and Vivek Ramaswamy, who are heading the Department of Government Efficiency, DOGE, for President-Elect Donald Trump.
You wrote, recently, a piece for your Substack called “Bringing Elon to a Knife Fight,” where you said that you support Democrats, like Congressman Ro Khanna, for pledging to work with DOGE. Why is that?
Pahlka: Well, I did say that until we know more about what they’re going to do, I think we should take an open stance. It’s very hard to know what they’re going to do. But ultimately, I said that because, as much as I may disagree with the policy goals of the administration that Musk and Ramaswamy are serving, there is so much work that needs to be done to subtract from government instead of constantly adding to it, to make it easier to get stuff done in government. I mean, people talk about regulation always as, you know, we’re regulating companies so they can’t, you know, pollute a stream. That’s wonderful.
There’s also enormous regulation on government itself, like the Paperwork Reduction Act, or like these hiring practices that really keep us from being able to serve the public in the way that we need to. And so it’s an uncomfortable position to be in because it’s not like I have any crystal ball to know what Musk and Ramaswamy are going to do. And I may disagree with some of what they do, or maybe a lot of what they do, but they’ve really kind of moved the Overton window and the conversation about this inefficiency, the sludge.
And I think that’s valuable, frankly, and I want Democrats to kind of get in the game of that reduction. And I think that if some of what they do is the wrong thing to do, but they shake government up in a way and maybe even pull some stuff out, we may be able to build back things that are kind of right-sized, the right-size procedures—not no procedure, not no process, but maybe not the heavyweight process that we have today.
Demsas: The thing I hear you saying here is, sort of, what I hear from people who have given up on their own side doing the right thing. And this is, I guess, reflected in the end of your piece, where you write, “We can wish that the government efficiency agenda were in the hands of someone else, but let’s not pretend that change was going to come from Democrats if they’d only had another term, and let’s not delude ourselves that change was ever going to happen politely, neatly, carefully.”
So, I mean, part of what it sounds like you’re saying is, Yeah, nobody wants this version of government efficiency, but there’s no other way it’s going to happen. Why is that the case? Like, why do you think the Democrats have been so unwilling to engage on this issue? I mean, you’re a Democrat. You worked in a Democratic administration, and you’ve talked to many other Democrats who have very similar views to you. Why is this such a third rail for them?
Pahlka: I’m not sure I know the exact answer to that. I think if you want to look at the Biden administration, in particular, you know—they went in with a big set of policy goals, and they actually achieved a lot of them. The four big bills are legislative accomplishments, significant legislative accomplishments. So they went for the what, but they neglected the how. And I think in their minds, it’s like, You’re going to do one or the other.
I think they should have paid equal attention to the how, to cleaning out the pipes so that the what could get through them faster. And that speed has clearly been a real problem. I mean, we’re writing now about the amount of money that could be clawed back because it didn’t get through those pipes, so really, really reducing Biden’s legacy. The frustration of not having that many electric-car chargers that were promised under the Bipartisan Infrastructure Law—all that stuff is due to this lack of focus on the how, and I don’t think it was a binary choice. I think Biden’s team could have said, We’re going to spend as much energy on the how as we are on the what.
But I do think there’s something about the way the Democrats, of course, want to be thoughtful and considered and hear all voices. And if you are thoughtful and considered and hear all voices, you tend to add policy and procedure and ways of looping everybody in. And that, actually, you know, adds instead of subtracts. Just naturally that’s sort of what happens. And in some ways, the destruction from which you can hopefully rebuild kind of needs to be done by somebody who kind of doesn’t care about that, in a certain way.
Demsas: I wonder, though, because it feels that, you know, two different theories of government reform—I worry about being [them] conflated, right?
So let’s take the DOGE theory, the Vivek-Elon theory. They presuppose that there are all these bureaucrats that are not really needed and all of these wasteful programs. And in a Wall Street Journal op-ed, they essentially have this idea that the executive branch has wildly overstepped its small-d democratic authority by being allowed to interpret laws that Congress passes as they’re implementing them.
And if that’s your theory of government reform—if your theory of government reform is that there’s just all these people who are dead weight, who are clogging up the process—then their answer, which is “mass head-count reductions across the federal bureaucracy,” is reasonable.
But as I understand it, your theory of government reform is very different. It’s that you need a capable and nimble executive branch in order to deliver on priorities like—I don’t know—providing health care to poor children. But in order to do that, you actually need a highly competent, well-paid, expensive labor pool and a good deal of it.
And so to me, it feels like, while both of these things can call themselves government-efficiency complaints—while they’re both motivated by a concern about the costs put on both private actors, individual citizens, and other government entities—they’re actually, fundamentally, two different political projects. So how do you see these things working together?
Pahlka: I agree. I have a very different view of it, and there’s some part of me that just thinks that if Elon and Vivek come in and spend any amount of time, if they don’t just get bored or frustrated and wander off, they’re gonna learn this. And they may have a different set of values, but I think it’s hard to miss it when you get into government that there are a lot of incredibly smart, talented, creative, dedicated people doing really amazing work. And you just fall in love with them once you actually get in the door. It’s from a distance that they look like, you know, these unaccountable, lazy bureaucrats. Up close, they’re pretty impressive.
But I think where I would put a little nuance on what you just said is that I do think we need this incredible workforce. And I think we’ve done a bad job of balancing between what I, in my very fancy language, call “go energy” and “stop energy.” So you have more people doing various forms of compliance and safeguards than you have the people trying to build something and get it out the door. And somebody I worked with at one point said, It’s like we’ve got six people building this product and at least 60 people telling us all the things we can’t do.
Now, those people who are saying, You can’t do that, are not dumb. They are not lazy. I mean, there are, of course, a few bad apples in government, and we can talk about that. I’m not saying everyone’s perfect. But you have people who, in fact, are—because they’re good, and because they really know the law, and because they really feel like it is their job to protect the public using this law, policy, and regulation—are very zealous in telling builders what they can’t do. And you have the very well-intentioned stop energy that overwhelms the people who have sort of go-energy jobs.
And I’m a little biased because I work with people a lot who do technology. They’re doing things like trying to get that form up, you know, trying to make sure that veterans can get their benefits. They are focused on, Can we get this application up so they can apply? Can we get the check to them? Can we get them their health care? Like, the actual outcome.
And a lot of people’s job isn’t to focus on the outcome but to make sure that all these things have been complied with, and they can do their job very well, and it slows the people who are outcome focused down. And it’s not their job, necessarily, to—you know, they’re not supposed to do their job less well. It is the job of leadership, of [the Office of Personnel Management], of the White House, of Congress, to look around and say, Why do we have so many people saying no, no, no? Oh because we put all these rules in place, and we’ve developed a culture of risk aversion that means we’re really, really focused on making sure nobody breaks any rules, at the expense of getting the job done. Leadership needs to balance the workforce between go energy and stop energy, and really take a hard look, if you’re going to add a regulation, you’re going to add a rule, Okay, what is the cost of adding that to the actual outcome that the American public expects?
[Music]
Demsas: After the break: Jen and I hash out the difference between political will and what she calls state capacity.
[Break]
Demsas: One phrase that you use a lot, and this is included in your recent report with the Niskanen Center, is state capacity. Can you define that for us?
Pahlka: Well, I didn’t even know the term until after my book came out and people were like, This is a state-capacity book. But I have since learned it’s an academic term that simply means the ability of a government—at any level and any government—to achieve its policy goals. So it is essentially, like I said, the how to the what.
Demsas: Yeah, this is a term that I think I first heard in the development-economics, development-political-science space. And it’s most commonly used to talk about the ability for these developing nations to effectuate their political priorities.
So for instance, like: Can a country collect taxes? Can it maintain the monopoly on the use of violence? These are core questions of state capacity because if you can’t collect taxes, you can’t run programs, you can’t have a police force that enforces laws. Like, there’s very little you can do on top of that, right? You can’t run a CHIPS program if you can’t do those things to begin with.
Why does this sort of idea—and how does this sort of idea—apply in the American context, where we have the ability to collect taxes? We have, relative to the rest of the world, like, a high degree of monopoly on the use of legitimate force. It’s contained within the state. What is the purpose of applying this term here?
Pahlka: Well, I mean, since you brought up applying taxes, the individual master file at the IRS, which holds all of the data about tax returns from individuals and families since the ’60s, is written in assembly code. There are vanishingly few people in the world who know what that code looks like. And it’s pretty robust. It’s lasted a long time. But, like, you’re going to run out of the human understanding of how that thing works, and you’re going to have a crisis at some point.
That’s not a crisis now, but we also don’t collect a lot of taxes. We have a serious unenforcement policy. We’re leaving a lot of money on the table because we have not empowered the IRS to be very successful. So we’re certainly not like a third-world state or an emerging state in that regard. But we are kind of going backwards in some areas.
And there’s a million examples of this, but I think that it is sort of shocking to people that state capacity is now a big concern for the United States, when it used to be that we only thought about it in relation to the countries that we would fund through the World Bank, or whatever. But national defense is a really great example of this. I mean, we keep spending more and more money, and it is not at all clear that we are getting more deterrence or more security. In fact, my thesis there is that we’re just spending too much money, not because—we shouldn’t cut spending because we want to be less secure.
But go talk to anybody in the Department of Defense. Pretty much everyone will tell you, like, unless there’s some shock to the system, we’re not going to change how we do stuff. And the way we do stuff takes decades, and we have to be able to move faster because, you know, we’re spending, I think it’s, like—what are we up to—almost a trillion dollars on national defense. And yet we seem to get less secure every year because the more money you put in a system like that, the more people double down on these very heavyweight ways of operating that are not what we need today.
Demsas: So I want to push you here a bit because this is a place—I’ve brought up to other people: I feel like the application of state capacity sometimes doesn’t feel like it fits well, and that, sometimes, what’s actually happening is that this is just a question of political will. It’s not that the government can’t accomplish what it tries to do. It’s that it actually has competing priorities, and there are trade-offs it’s unwilling to make.
One place where people have talked a lot about regulation that is holding government back is the National Environmental Policy Act. This is a piece of legislation from the 1970s that requires that the government study the environmental impact of its major actions. And it’s often talked about that it takes years to compile an environmental-impact statement, so it can take years and years in order to get a permit for, you know, a big energy project.
But something interesting happened, and this is a stat that was surfaced by Brian Potter in his Substack, “Construction Physics.” I’m reading from it: In 2009, after the Great Recession and Congress passed the American Recovery and Reinvestment Act, there were “over 190,000 projects, totaling $300 billion worth of stimulus funds, [that] were required to have NEPA reviews before the projects could begin. After the passage of [the American Recovery and Reinvestment Act], categorical exclusions were completed at a rate of more than 400 per day, and 670 environmental impact statements were completed over the next 7 months.”
So essentially, these EISs, the environmental-impact statements that often take years to complete, all of a sudden are being completed over the course of a few weeks—670 over the course of seven months is just astronomical compared to what we usually see.
And this is an example where nothing changed about the state capacity. They didn’t change anything about the legal environment. They didn’t change anything about the number of people working in government and whether they were more qualified. The HR processes didn’t change in this time period.
What happened is that the federal government was like, We’re in an emergency space. We need to get a bunch of stimulus dollars out the door, because we’re in a free-fall recession, and we’re worried about mass unemployment. And then, all of a sudden, all of these things that seemed like state-capacity issues, that seemed like these big constraints on government, actually just disappeared, because everyone wanted them to happen.
So is it the case that the government can’t do what it wants? Or is it that there’s a lot of competing priorities, and in times of nonemergency, we’re actually not aligned on what government wants to do?
Pahlka: Well, I mean, I think COVID is another good example of when government just does it, right? Or Josh Shapiro’s getting I-95 open again. I can’t disagree with you on that. Absolutely. I will say, I remember that too, and we just looked into it, and it’s not exactly apples to apples there, so I’d just like to put a little bit of an asterisk on it.
But I think your point is valid, but it does, then, beg the question, right? So we only have 47 electric-vehicle chargers out of the money that came out of the, you know, Bipartisan Infrastructure Law. I guess it was also a bill that funded the BEAD Program for broadband-internet access, and we have zero connections from that.
Are you saying, then, that Democrats didn’t want to see those things implemented? Because I do think it is a matter of will. But we are seeing places where the political will seems to be there, but it seems to sort of stop after the law is passed.
I think I’ve also shared this with you before, but, like, I got into this through working with cities and states on benefits delivery, and we were looking at SNAP uptake. And I was in California, and it was just shocking to me that California, which had a ton of money and spent hundreds of millions of dollars on IT systems for people to apply for SNAP online, had the second-lowest rate of participation in the program in the entire country. Only Wyoming was worse than California.
Is that a political-will problem? It’s, like, a really blue state, very pro-welfare. But it kind of couldn’t get out of its own way. It so overscoped these systems that it took about almost an hour to apply online. You couldn’t do it on a mobile phone. It’s just all these ways in which they created a system which is hard to use. But it’s really clear to me that they didn’t intend to do that. They just had too much process in the way and less of a focus on the outcome.
So I do think it’s a political will, but it has to be political will to follow the thing all the way through to the outcome, to care as much about the implementation as you do about the legislative win or the money that you put into it. We’re really good at money and rules, and those things do not necessarily translate to the outcomes that we promised people. So that will has to move.
Demsas: Yeah. But I think what I’m saying is: I think this may be a case of revealed preferences, right? Like you asked me, Does this mean that Democrats didn’t really care about getting broadband out? And I don’t want to make that kind of a strong claim. I think if they could push a button, and there was rural broadband for every single person in rural America, they would have pushed the button.
But the question government asks, and government policies ask, which you’ve written about extensively, is not just: Hey—do you wish this thing existed? It’s, When you’re forced to make trade-offs between whether to push out broadband or make it easier for contractors that are different from the ones you usually go to to get access to this program, which do you choose between? If you’re going to choose between actually getting out broadband and following the most onerous environmental regulations that exist, which thing are you choosing?
And over and over again, you see, as you mentioned before, liberals choosing this process, choosing this kind of way of delaying implementation in order not to follow some shoddy or quicker, maybe more error-prone system. And in doing so, they end up not getting to the outcomes. And to me, I feel like that actually is a situation where we’re seeing what Democrats actually want, which is really clear when you look at infrastructure projects.
I mean, this is what I think is the story of California high-speed rail, where you talked to so many people, where I bet a lot of people would love for there to be high-speed rail between San Francisco and L.A. I don’t think they’re lying about wanting that to exist. But when you talk to people who are working in that program or who are working trying to implement it, and you say, Okay, well, you need to not let every single local government fleece this project for whatever priority they have on the ground, and no one wants to do that. So I’m left with the conclusion that yes, they want high-speed rail but not if it means angering a single person within the Democratic Party.
Pahlka: I completely agree with that. It’s a little bit what I was saying about, like, you kind of need a big disruptor, someone who doesn’t care, to get stuff done sometimes. I wish it weren’t Elon, necessarily. But if you’ve created a system in which you have to make everybody happy, eventually people will be so frustrated they’ll let somebody, you know, give the job to somebody who doesn’t care if he makes anybody happy.
Demsas: One of the objections I hear sometimes from liberals about making government more efficient is that all of these layers of procedure are to protect and prevent against authoritarian impulses. So yes, it’s frustrating and annoying that we have to follow all of these rules, and that there are all these government watchdogs that might sue if you don’t cross your t’s and dot your i’s. And that is annoying when you’re trying to get good policy done. But when you have someone like Donald Trump, for instance, get elected, you’ll be really happy that all of these procedures and layers of government exist. How do you respond to that?
Pahlka: Well, they’re not wrong, of course. And we just talked about trade-offs. This is exactly a trade-off conversation. The reality is that I believe that our lack of results and the slowness of government played a part, maybe not be the leading part, in driving people towards wanting someone who claims, I alone can fix it, right? Who claims to be able to bust through all that red tape.
Now, in reality, did he bust through a lot of red tape in his first administration? Well, he claimed to roll back a lot of regulations, but his team really didn’t do that much on that front. But it is a trade-off you make. I am not extreme on either end, but I do think we need a middle ground where we are looking at where safeguards and processes and procedures and the ability to sue are kind of right-sized, where there are some protections.
But where we are right now is: The extra-extra-large version of protections, which has slowed us down enough that it has driven this force in our society for, like, none, which is the pendulum swinging. I just wish the pendulum would settle a little bit in the middle. But that’s a trade-off we need to make. And we have to, as you say, piss some people off in order to get that, because you’re gonna have to say no to some people to get the job done.
Demsas: I feel like the analogy I’ve used a lot is to the filibuster—
Pahlka: Yes.
Demsas: —which I think that a lot of liberals were worried about when this was being debated more openly. If you get rid of the filibuster, that means Republicans will be able to pass their policies as well.
And I think the thing that’s interesting about this is, one, it’s the question of democracy—like, small-d democracy. Do you want the government to be able to do things such that the public can actually evaluate them? Versus someone who gets into office, and they can’t actually enact a bunch of their priorities. So it’s actually quite unclear what signal you’re supposed to be sending as a voter.
But also secondly, I think there’s, like, an asymmetry here, where if you are a small-c conservative, versus a lower-l liberal, you have different sorts of desires from government. Like, there are a lot more active policies that are trying to be passed by people who are liberal, who are progressive. And so there’s kind of an asymmetry of what gets constrained in that kind of a paradigm.
And so I think that it’s hard because you look at the looming potential changes in a Trump administration, and you think, like, Well, it’s really good that there are all these different ways of constraining this. But in the long run, there’s just this larger question here about whether it’s democratic at all to have that happen. Like, if people are electing an executive, how exactly are we supposed to evaluate that work if after four years, so many of the policies that they promised, whether they’re harmful or whether they’re good, don’t actually get passed?
Pahlka: It’s such a hard question. And yeah, I kind of want to stand on—as uncomfortable as this is—if you think state capacity is important to the country, you kind of have to be okay with people who you, let’s softly say, don’t agree with having it. But we’re in this sort of thermostatic nature of elections right now, and I have no crystal ball, but if the Democrats were to get the White House back in four years or even take back Congress in two years, you really don’t want them to be dealing with this huge incapacity once again, or at least I don’t. And that’s just a tough pill to swallow, but I think it’s one we have to swallow, again in the sense of making trade-offs. I agree—it’s much like the filibuster.
You could also say the Administrative Procedures Act is a lot like the filibuster. It needs to be reformed for all the reasons you mentioned when you talked about NEPA to be able to get these, you know, big infrastructure projects built, because it creates such a huge surface area for attack by minority interests. And if you were to do that today, you would really empower Trump to do a lot of what he couldn’t do last time, and that’s really problematic.
But the reality is it’s not going to get repealed today. Like, if you started working on that now, maybe it would happen at the end of the administration and benefit the Democrats. Now, I know that’s sort of like a Pollyannaish view of it, but at the end of the day, it kind of just does need to get reformed if we’re going to be able to govern at all.
And you used the word democracy, right? If we have the system in which we vote for elected officials, and then they go through that messy political process to say—well, let’s use the example of housing, right—to say, This area needs more housing. We’re going to build more housing, and then a bunch of people who have an interest in having that housing not be built can stop it, is that democracy? We have thwarted the will of what the democratic process actually came up with.
Demsas: Well, Jen, always our last and final question: What is an idea that you had that you thought was good at the time but ended up being only good on paper?
Pahlka: I love this question. You asked it of a guest a couple of episodes ago who answered, “small plates,” which just made me laugh so hard. And now I’m just not ever going to order a small plate at a restaurant again. So I’m just co-signing that.
But I guess my more original answer would be: When I started working with local governments, I really had this sense that more data was better. It was kind of shocking. Sometimes you’d go in there, and you were just like, You’re not making decisions based on data. How awful. We need more. We need more. And then over time, I realized there’s a human aspect to this that we neglect. So there became this whole trend of doing data dashboards for local governments. And then, like, no one looks at them really. They were sort of a lot of work for, in some cases, not much return, depending on the human and cultural and, you know, organizational infrastructure into which they were inserted.
But I also really saw, when I was working on the unemployment-insurance crisis during the pandemic, the ways that a lot of leaders see data as a grade that they’re getting, not a compass that they can use to steer the ship where they need to go. And I really changed my view on, like, what kinds of data are good in, like, a governing context, in a performance-management context, and really now sort of see it as good only if it’s introduced in the right ways and if the people who are supposed to be using it as a compass actually are empowered and encouraged to do that.
[Music]
Demsas: Well, Jen, thank you so much for coming on the show.
Pahlka: Thank you so much, Jerusalem. This was fun.
Demsas: Good on Paper is produced by Rosie Hughes. It was edited by Dave Shaw, fact-checked by Ena Alvarado, and engineered by Erica Huang. Our theme music is composed by Rob Smierciak. Claudine Ebeid is the executive producer of Atlantic audio. Andrea Valdez is our managing editor.
And hey, if you like what you’re hearing, please leave us a rating and review on Apple Podcasts.
I’m Jerusalem Demsas, and we’ll see you next week.
