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.
Today, we have The UnPopulist Editor-in-Chief Shikha Dalmia in conversation with two of the foremost thinkers of our time. Francis Fukuyama is an American political theorist and public intellectual best known for The End of History and the Last Man. He is a senior fellow at Stanford University’s Freeman Spogli Institute, where his work focuses on political order, governance, and democratic backsliding. Tyler Cowen is an economist, author, and public intellectual who has written books on innovation, talent, and cultural change. A professor at George Mason University and director of the Mercatus Center, he writes the highly influential blog Marginal Revolution and hosts the long-running podcast Conversations with Tyler.
One reason for the populist revolt in America is the notion of the “deep state”—that an unaccountable bureaucracy is secretly ruling the country. Frank and Tyler come from very different intellectual traditions. Frank, a centrist, is a student of Max Weber and Tyler is a limited government libertarian. Yet they have both argued that liberal states in complex modern societies need a functional bureaucracy—a.k.a. state capacity—to deliver public goods and solve collective action problems. But they also have a ton of disagreements, especially on just how broken American governance is—and they duke it out in a spirited discussion.
We hope you enjoy.
A transcript of today’s podcast appears below. It has been edited for flow and clarity.
Shikha Dalmia: Welcome to Zooming In, Tyler and Frank. We are going to talk about state capacity liberalism or [state capacity] libertarianism today. But this being Trump’s first anniversary—and I have both of you here, so I can’t let this pass—I just want to ask you: What is your assessment of Trump’s first year? Is it better than you expected? Worse than you expected? The same as you expected? Tyler, you want to go first?
Tyler Cowen: It’s much crazier than I expected. I was not thinking we’d get a repeat of the first term, but the variance has been very high. I hate the company nationalizations. I’m strongly opposed to the tariffs. I hate what he’s doing with ICE, and the level of corruption is higher yet than it had been. On the plus side, I think the notion that America, looking forward, is the energy world superpower and the AI world superpower, we’re achieving those two things, and they’re very, very important. I hope we survive the rest. I’m not going to short the market, but my goodness, every day there’s so much news to absorb. It’s nerve-wracking.
Dalmia: Especially for those of us in the news business. We are kind of exhausted by now and it’s just a year. Frank?
Francis Fukuyama: Well, I had very low expectations for what would happen in the second Trump term, and he’s managed to beat those low expectations and lower the bar even further. A lot of my concerns actually had to do with what his effect on state capacity would be. We can get into that discussion. The thing that’s been the most surprising has been foreign policy, where you have the transformation of the United States from a stabilizing backstop to the whole liberal international order to an actively aggressive kind of imperialist country, which is something that I don’t think was on the dance card of almost anybody prior to the last election. And that has certainly been a big surprise and we’re seeing it play out as we speak in Davos and Greenland and so forth.
Dalmia: Frank, you talk about “Getting to Denmark.” I think you don’t have to worry anymore because Denmark is coming to us now. It’s just going to be our 51st state, or maybe 52nd after Canada. But you guys know what my opinion is if you read The UnPopulist, so I won’t rehearse it here and bore you.
Getting into the topic of the conversation here, which is state capacity, both of you come from different intellectual traditions. Frank, you come from this Hegelian, Weberian framework, and Tyler, you and I were libertarians for a long time. I’m a former libertarian now. But you have both converged on the same language to describe the crisis of liberal governance, especially in the United States, but maybe more broadly, and that’s the lack of state capacity. And to some extent, you’re pushing back against people in your own camp, and you want them to take this idea of state capacity seriously.
But tell us, Tyler, what is state capacity, and why should libertarians take it seriously? And then I’ll ask the same question of Frank.
Cowen: Well, I like to start by making issues very concrete. So for the United States, as you know, we have debt in the range of about $38 trillion, and the simple state capacity issue is: How will we pay for it? Our electorate hates higher taxes, and we keep on borrowing. I hope economic growth comes through— and artificial intelligence and tech more generally is our best bet for that happening. So making sure that can happen is a Number One state capacity issue.
The other is simply AI systems are progressing very rapidly. Dario [Amodei] of Anthropic just outlined this at Davos. How quickly will we be able to integrate AI systems into our government? I think, in the Trump administration, there’s actually a pretty extreme will to do that. I don’t see the order in coherence and respect for process and talent for that actually to happen.
But those to me are the two big issues of the moment. They’re both up for grabs. There are more positive developments, at least potentially, than when I first wrote my long blog post on state capacity libertarianism, and I follow those closely.
Dalmia: When you talk about borrowing going up and debt going up, why is that a state capacity issue? Why is that not just a simple governance issue? I mean, why use the terminology of state capacity when we have perfectly good normal lingo like it’s a governing problem? What does state capacity add?
Cowen: Well, if you go broke, everything is wrecked. But if you don’t go broke but have to radically cut, say, discretionary spending, which we’re already cutting too much in many areas, that’s a huge negative. You need the resources for your government to function, and whatever you think our foreign policy should be, it’s going to cost us a lot of money, and we have to pay the bills. So, so much of state capacity rides on the simple question: Is there the money there? Historically, it has been there for the United States government, but that’s now in much greater doubt than before. We see in Japan, at the time we’re recording, that their higher longer-term bond yields are spiking dramatically. They face a question of how they’re going to pay for everything moving forward. So I think that’s absolutely central if we’re talking about this issue in 2026.
Dalmia: But why not go to the standard libertarian solution, which is just starve the beast, cut the government, not worry about maintaining state capacity, but just cutting it?
Cowen: There are plenty of expenditures where I would prefer to either cut them or limit their trajectory, their rate of increase—say, healthcare spending for the elderly through Medicare, and much of Medicaid is also for the elderly. But voters are not going to do that, in my opinion, so we need to look for other solutions. The easiest way out, if it’s possible—and this is a huge if; I don’t want to bet the house on it—would be to have the rate of economic growth increase by half or three-quarters of a percentage point. And if we could manage that, we are probably on a stable fiscal trajectory again.
Dalmia: Frank, you are something of a centrist. You are talking to both sides, limited government conservatives and libertarians and progressive naysayers. What does state capacity mean to you and why should they worry about it?
Fukuyama: Well, state capacity is people. State capacity is a bureaucracy that is capable of translating legislative mandates and executive orders into actual results. And it’s a critical aspect of government that has traditionally been very weak in the United States because Americans distrust and dislike government. They don’t like bureaucrats, and so they’ve not invested in having the adequate people in place to actually do the things that they expect government to do.
I think it’s actually one of the three pillars of any modern political system. First is democratic accountability. The second is a rule of law. But the third is actually the capacity of a government to act on popular mandates. And that’s the part that I think is really up for grabs right now.
Cowen: I don’t think American state capacity historically is that weak. We built this incredible empire, often unjustly. We put a man on the moon. We developed the atom bomb. We’re leaders in aviation and computers in part because of government. A lot of our state governments work really quite well. It’s a mixed bag, but I think we’d be in the world’s top 10 easily. Noah Smith had a great blog post on this.
Fukuyama: No. Okay, so, actually I’m glad you brought that up because that was the example that I was going to cite to you—getting a man on the moon. I think that American state capacity had expanded beginning in the Progressive Era passage of the Pendleton Act in 1883, and then the growth of a much larger federal bureaucracy in that period, a big expansion of the administrative state, in the New Deal. And I think that it really peaked sometime in the late 1960s, and it’s been declining ever since then. I was actually going to refer to getting a man on the moon, which was really an extraordinary demonstration of state capacity on the part of the United States. Kennedy in 1961 says, “We’re going to get there by the end of this decade.” And NASA actually pulled it off, which was, if you look back at the risks that were being taken, really quite extraordinary.
“There’s a conservative line about the bureaucracy, which is that all of these unelected bureaucrats are making rules and really governing the country outside of the control and the mandates of elected representatives. And I think the actual situation in the American bureaucracy is exactly the opposite. It is over-regulated. Americans never trusted government, and so over the generations, they have piled up massive amounts of rules that constrain the ability of bureaucrats to use common sense, good judgment, in order to actually implement the things that their democratically elected masters tell them need to be done. And I think, therefore, you really need to free the bureaucrats to be able to exercise considerably more autonomy.” — Francis Fukuyama
There’s a great series on Apple TV called For All Mankind, which begins with the premise that, actually, the Soviets beat us to the moon. And that stimulated a massive investment in the Apollo program that actually created a lunar base and ongoing ability to get to the moon. We could not do that today. We absolutely could not do that. And part of the reason is one that libertarians would be very happy with: We have placed so many constraints and rules on the ability of our government to actually do the things that it is asked to do that it’s very, very difficult to build anything. The state of California voted the money for high-speed rail in the mid-1990s, and they built a few kilometers of track out in the Central Valley, and it has not appeared. The United States in the early 1930s built the Golden Gate Bridge, Hoover Dam, Oakland Bay Bridge, and electrified the entire upper south in the TVA, and they did this all in the space of three or four years. We absolutely could not do that today.
Cowen: Well, we’re doing different things. We’re building AGI. That’s gone amazingly well. And if I look at state governments, most of them are much better than 40, 50 years ago.
Fukuyama: No, no, that’s not state capacity.
Cowen: Of course it is.
Fukuyama: That’s private sector investment. State capacity is the government. It is the democratically legitimate agents of the American people. And that’s the thing that’s weak. I mean, yeah, there’s no question that we’ve got a very strong private sector, particularly in tech. But I think that if you don’t have a state that can actually implement policies that are mandated, then you’ve got a big problem.
Cowen: Our government’s a big supporter of the AI efforts. It knows when to leave them alone. And most state governments in this country, they’re much better on than 40, 50 years ago. It’s not even close. Look at where I live in Virginia.
Fukuyama: You don’t live in California, Tyler.
Cowen: Yeah. California might be worse. I agree. The rest are better. Look at the Southeast. Uniformly better.
Fukuyama: No.
Dalmia: Frank, but state capacity doesn’t mean for you only the delivery or the building of large infrastructure projects, right? It’s something else. It means moving from what you’ve called a patrimonial, patronage system to one that is governed effectively by a rules-bound, autonomous bureaucracy.
Fukuyama: Right.
Dalmia: Explain that, because when you say that the problem is not too much regulation, but that there aren’t enough regulatory agencies or regulatory capacity, that sounds very, very strange to almost every ear, not just libertarian ears.
Fukuyama: No, I think the reason that American bureaucracy does not perform the way we would like it to is that it’s overregulated. There’s a conservative line about the bureaucracy, which is that all of these unelected bureaucrats are making rules and really governing the country outside of the control and the mandates of elected representatives. And I think the actual situation in the American bureaucracy is exactly the opposite.
It is over-regulated. Americans never trusted government, and so over the generations, they have piled up massive amounts of rules that constrain the ability of bureaucrats to use common sense, good judgment, in order to actually implement the things that their democratically elected masters tell them need to be done. And I think, therefore, you really need to free the bureaucrats to be able to exercise considerably more autonomy.
I’ll just give you several examples of this. In order for a federal agency to procure office furniture or a computer, not to speak of an F-35 fighter, you have to go through the federal acquisitions regulations. These are several thousand pages of extremely detailed rules about how you put out a request for proposal, what are the bidding rights, what are the rights of appeal if you are a losing party. There are all sorts of detailed requirements that it has to be put out for bid to small businesses, women-owned businesses, minority-owned businesses. The result of that is that procurement in any federal agency is much, much slower and much, much more costly than a counterpart procurement officer in a private sector procurement department. And that’s just one example. The way we buy software is subject to a similar kind of malady where you don’t have the flexibility in the government to actually make judgment calls on whether something is adequate, whether there’s a trade-off between safety and efficiency and speed.
These are all covered by very detailed rules, and the way that we reward bureaucrats and incentivize them is to be compliant, not to actually achieve results. We don’t grade them. We don’t promote them on the basis that they’re getting concrete results. We sanction them if they break rules, and that’s, I think, the fundamental thing that really needs to change.
Dalmia: So your contention is that because Americans are so wary of centralized authority, they have over-constrained it and therefore made government more expensive to the detriment of actual governance and delivery of public goods and infrastructure projects.
Fukuyama: Yes.
Cowen: I agree with all that, but you’re lemon-picking the worst examples. Look, during the pandemic, we had Operation Warp Speed. We got vaccines in an amazing year. That probably saved at least two million lives globally. It’s an incredible triumph, including for state capacity. We never fixed our Pentagon procurement cycles, so we did an end run with Palantir and Anduril, which have done phenomenal things. So when you look at the whole balance ...
Fukuyama: Tyler, look, I think you’re the one that’s cherry-picking. The things that have been successful in recent years are ones that have precisely gone around the existing rule structure. So another great example of this ...
Cowen: Sure, but we can do that is the point.
Fukuyama: Well, no, but that is not the routine way that the government operates.
Cowen: But it is.
Fukuyama: It’s extremely dysfunctional. So, for example, the I-95 accident in Pennsylvania—Josh Shapiro, they were telling him that if you did all of the permitting requirements to actually fix the interstate, it would be a six-month to one-year process. Shapiro fixed it within two weeks because he actually suspended a lot of the existing rules. And frankly, some of the stuff he did was illegal because the rule would have slowed down the process. He knew he had to get the interstate running.
You cannot run a government by exception. You cannot run a government by end-running legal requirements that have been put in place by legislatures. I think that what we really need to do is to clear away a lot of that accumulation of really pointless rules that slow everything down.
Cowen: That I agree with, but I think you’re seeing the overall picture is much more negative than it really is. We’ve had these big, big triumphs of state capacity lately that have amazed me, in fact. You know, they’ve happened since I wrote my piece.
Dalmia: But Frank’s point is those are the exceptions, not the rule.
Cowen: They’re the important things, though. AI, pandemic, defense technology ...
Fukuyama: Defense technology is a great example. We procure software in the Defense Department very, very badly. Jen Pahlka has written a wonderful book called Recoding America. She was one of the founders of the U.S. Digital Service. She talks about how defense procurement, including for IT systems, follows this waterfall-like series of steps. You come up with requirements, so you spend a couple of years going around to every part of the government asking what would you like to have as a requirement for this new software system. You spend another year putting it out for bid. You evaluate the bids in another couple of years, and then you award the contract to Arthur Anderson or IBM or some other big company. And then they deliver the software project five years later, and by that point, it’s completely obsolete.
“Great stagnation, in my view, ended in 2020 with the Covid vaccines and then AI on top of that. I think in biomedical science and AI, we’re seeing phenomenal progress on literally a week-by-week basis. And I wrote in 2020, I now think the great stagnation is over. I’m convinced that’s true. Productivity numbers are not reliable in the short run. They are showing it. The Atlanta Fed is now forecasting a 5% rate of economic growth or above 5% for the fourth quarter of the previous year. Those are incredible results. I’m not trying to credit Trump with them, but we just cannot be screwing up that badly. And all these European countries with supposedly good state capacity, almost all of them can’t defend themselves.” — Tyler Cowen
What you need is a much more agile system in which you have continuous feedback that goes from the users of the software all the way back to the appropriators who are willing to give the Pentagon the ability to modify a project on the fly. They don’t have that ability right now. It’s a very, very cumbersome thing, and there are many big disasters in federal software procurement. The most obvious one was healthcare.gov, but Army 21, a number of really major combat systems, have basically been killed by this inability to procure information technology. And that really has to do with the kind of cumbersome rules that are completely different from the way Silicon Valley works. You’ve got to fix that, I think, if you’re ever actually going to have a government that can make use of technology in a flexible and agile way.
Cowen: I’m completely for everything you and Jen are arguing for, but if you asked a simple question, whether or not you favor the action in Venezuela, technologically, did it go well? The answer is obviously an overwhelmingly yes. So the full picture is a lot of things are really working, right?
Fukuyama: I think the full picture ... I mean, again, if you look at the infrastructure space, we’ve got a deficit in infrastructure of a couple trillion dollars. I mean, the American Society for Civil Engineering does these annual reports.
Cowen: Those are terrible reports. It’s all self-pleading, those people. You cannot listen to them.
Dalmia: Actually, let me ask Tyler’s question in a slightly different way, Frank, which is that, ultimately, the proof is in the pudding. You think America is, among developed countries, a much less well-governed society than European countries. You rate Denmark very highly. You rate Germany and England highly. But if you just compare some of the macro statistics, I mean, America’s economy has doubled since 2009, when the E.U. and America were on par. Now America’s economy is double and Europe is at the same point. Infrastructure: Americans have the shortest commute to work. They take more flights than in European countries. To Tyler’s point, if you look at the debt, U.K., which you consider to be a much better governed country, I mean, it’s debt to GDP ratio is 100%.
Fukuyama: No, no, but you’re talking about overall aggregate economic results. I am talking about the ability of the federal government to actually carry out mandates that it is given by legislatures. And I think that, for example, a lot of that economic growth is due to the fact that we’ve got a very vigorous private sector. And maybe we don’t get in the way of it as much as European countries do, but that’s not a function of the U.S. government. If you look at, for example, health care, it’s a disaster. We spend twice as much per capita on health care as any other developed country, and we get half as good results.
Cowen: None of that is true, Frank. We subsidize the most innovation, and if you adjust for ethnic group, Japanese Americans live longer than Japanese, Swedish Americans do better than Swedish.
Fukuyama: So if you can afford to pay ...
Cowen: We overpay a bit.
Fukuyama: I mean, obviously, in this country, if you can afford to pay, you can get really great health care. But I think the appropriate measure of the success of a public healthcare system is: How does everybody do? And we’ve got still tens of millions of uninsured, underinsured people that don’t get preventive care, and that shows up in the aggregate statistics. Life expectancy has actually been falling for several of the most recent years.
Cowen: Life expectancy has been rising again. If you take away car accidents and guns killing people, American healthcare system does fine. A lot of those are myths.
Fukuyama: For twice the amount of money as any other developed country.
Cowen: We overspend a bit, but we drive innovation around the world, right?
Fukuyama: Yeah, but again, Tyler ...
Cowen: Look, we have GLP-1 drugs coming; HIV-AIDS more or less is going away; there’s a vaccine against malaria.
Fukuyama: Yeah, and a lot of those benefits flow to people at the top of the income distribution.
Cowen: And then they become cheaper over time. Immunology treatments against cancer. Cancer may be a thing of the past within several decades. We’re making incredible progress now, and AI will give us more. And it’s a good thing we’re spending that money. The undercapitalized systems like Canada and U.K., people used to boast about them, and now they’re complete embarrassments.
Fukuyama: I don’t know. I’ve never met a Canadian that would trade our system for theirs. They cross the border because they don’t have to trade ours for theirs.
Dalmia: I lived in Michigan for a long time, and the number of Canadians who would cross over to Detroit to get treatment in Henry Ford Hospital and DMC was actually really high, yeah.
Fukuyama: Look, nobody’s saying that. At the high end, if you can afford it, American health care is really great. But there’s a basic equity problem there. That kind of high quality health care is not available to tens of millions of Americans. It’s just not.
Dalmia: Frank, let me ask you a slightly different question on the same theme. You mentioned Americans have developed a system of governance that they have because of their deep aversion to centralized authority. And you’ve actually written a whole provocatively titled series called “In Defense of the Deep State.” Now, the deep state is, as you well know, a term that comes from Turkey and Egypt, where you actually had a cabal of military security agencies and intel agencies that were pulling the strings of elected leaders from behind the scenes. And they protected themselves and their version of what the regime should do and fed off the regime.
“My view of state capacity really is implementation ability. Once a political decision is made, does the government have the ability to actually implement it? And I think this is a problem across the board that I’ve been trying to confront in my teaching. I run a master’s program in public policy, and the trouble with the way that Americans, including in public policy schools, think about that issue is it’s all about policy making. It’s all about what the policies ought to be. And so we train people to be policy analysts, to come up with the optimal policy for solving a particular problem.” — Francis Fukuyama
That’s not what the United States is. The United States has actually got a very transparent and open bureaucracy that can be held accountable. So why would you call it the deep state?
Fukuyama: Well, obviously, it’s trolling the right-wing Steve Bannons of the world that call it the deep state. Obviously, I don’t think that when I use the term, it has any sinister connotations. I think that, actually, the deep state is something we should value and we should work to improve. So, yeah, I mean, it’s a completely ironic use of the term.
Dalmia: So, Tyler, we’ve been interrogating Frank a fair amount. I mean, I’m actually a little surprised at how much you are challenging Frank, because reading your own views on state capacity, there seems a great deal of agreement. What Frank essentially wants is a professional, merit-based, rule-bound bureaucracy that makes impersonal decisions and has rational regulations in place and is able to provide public goods and infrastructures without political pressure, without being subject to interest groups. And you are in favor of all kinds of public goods that conventional libertarianism would not accept. And, as I understand, your argument is ... and you wrote wonderful piece back in 2007 where you contrasted between negative and positive liberty, and not in the Isaiah Berlin sense, but you said negative liberty is freedom from coercion by the state, but positive liberty is like, how many life plans can you accomplish? That’s what the real question is. And if the government provides certain public goods and infrastructure in an efficient way, it lowers your transaction costs as an individual. And so you are in favor of that, right? You don’t disagree with Frank on that.
Cowen: I favor all the reforms he and Jen Pahlka are putting forward. Absolutely. But I think they are misdiagnosing our current situation, which is one that works better than is often recognized.
What I want us to do is to spend more supporting science. Our government in the past has actually done a pretty good job of that. And right now, Congress is restoring some of the Trumpian cuts. I think we’ll recover from that pretty nicely. And again, any discussion of this that does not put AI at the center is just out of date. AI will considerably increase the rate of progress in science. And our policies, which have relied on the federal government, sometimes just to preempt state laws or to leave the thing alone, so far they’ve been excellent.
Dalmia: So are you no longer worried about the Great Stagnation because AI is going to jolt us out of that and institutions don’t matter?
Cowen: Great stagnation, in my view, ended in 2020 with the Covid vaccines and then AI on top of that. I think in biomedical science and AI, we’re seeing phenomenal progress on literally a week-by-week basis. And I wrote in 2020, I now think the great stagnation is over. I’m convinced that’s true. Productivity numbers are not reliable in the short run. They are showing it. The Atlanta Fed is now forecasting a 5% rate of economic growth or above 5% for the fourth quarter of the previous year.
Those are incredible results. I’m not trying to credit Trump with them, but we just cannot be screwing up that badly. And all these European countries with supposedly good state capacity, almost all of them can’t defend themselves. That’s the first and most critical question that Straussian and Hegelian and Weberian should recognize.
Fukuyama: Yeah, but their inability to defend themselves isn’t a function of state capacity. It’s a political decision not to invest in their militaries.
Cowen: That is state capacity, though. They can’t convince their own people that there’s actual risk. And they promised them all extreme comfort and no risk forever. And now they’re stuck in this equilibrium where they don’t get out of it.
Fukuyama: But that’s not a question of state capacity.
Cowen: Sure it is. They can’t do it. They say they’ll probably increase defense spending to 5%, a lot of them. And at the end of the day, they still can’t defend themselves. They can’t cooperate. They cannot, across the E.U., create integrated military supply chains that work across nations. They still rely on the U.S. for that. That absolutely is state capacity.
Fukuyama: Well, it’s a broader issue than the state capacity that I worry about. That really has to do with political institutions. So, for example, the current system in the E.U., where you have a unit veto on issues of budget or on foreign policy is a completely dysfunctional system.
Cowen: Sure.
Fukuyama: You know, Slovakia by itself can stop the entire E.U. from helping Ukraine, and that’s a ridiculous system. But the problem is not state capacity, it’s badly designed institutions. They need more centralized ...
Cowen: You’re just pushing around the words. Look, their governments can’t do it because of decisions they made. You can call it what you want. I think you lose on the substantive point whether or not you want to call it state capacity.
“If you’re talking about Denmark, I think you need to compare it to individual U.S. states. I think my state of Virginia is pretty well governed. Utah is pretty well governed. Much of California is a disaster. Mostly the state has too many leftists, I would say. New York City has done an incredible job on lowering its crime rate, which is now falling again. That’s a major triumph of state capacity that really affects how people live.” — Tyler Cowen
Fukuyama: Yeah, okay, I’ll admit the E.U. is a kind of funny beast because it’s not a state, and it’s been stuck in this position where it doesn’t have enough authority, and it needs more. So yeah, you’re right about that.
But again, my view of state capacity really is implementation ability. Once a political decision is made, does the government have the ability to actually implement it? And I think this is a problem across the board that I’ve been trying to confront in my teaching. I run a master’s program in public policy, and the trouble with the way that Americans, including in public policy schools, think about that issue is it’s all about policy making. It’s all about what the policies ought to be. And so we train people to be policy analysts, to come up with the optimal policy for solving a particular problem.
We don’t pay any attention to the implementability of problems. And that’s where we’ve fallen down time and time again. It doesn’t matter if you write a great memo advocating a beautiful policy that your econometric analysis has given you if you can’t actually put it into effect. And so we’ve under-invested in implementation capacity, both intellectually in a way that we train people, and I think in government, where typically, if you’re running a government agency, you get to the implementation stage and you say, “Ah, well, that’s a lesser ability. We’re just going to hand it off to a contractor. The contractor can take care of that.” I think that the essence of recoding America, you know, Pahlkaism, is that you can’t do that. That implementation capacity is an intrinsic part of policymaking. And so I think if you don’t take into account the implementability and the strategies for implementation, you might as well not design the policy in the first place because it’s just not going to come about.
This is one of the things that I want to change in the way that we teach public policy and the way that we actually practice public policy. We have to think about implementation to a much higher degree, and we have to train people in that to a much higher degree than we do.
Dalmia: Tyler, let me ask you this. Now, you actually also agree with Frank that Denmark is a model country. It’s well-governed, well-run.
Cowen: Yes.
Dalmia: Why? Why do you think Denmark is a good example?
Cowen: People there are relatively happy. They’re not leaving in large numbers. I visited Denmark, I think, five times. Always had very positive impressions. They’ve also managed their immigration better, say, than Sweden has, so there’s less of a backlash. And I am pro-immigration, but I don’t think you should extend it to the point where you have a major backlash. It’s just a healthy, happy, pretty successful country.
Dalmia: Yeah, but its GDP growth is 2%.
Cowen: I don’t know the current number. They’re clearly poorer than the United States. Some of that they get back in the form of quality public goods. I worry about their future with birth rates, their position in Europe, their flexibility. I don’t think their labor markets work as well as they used to. I’m not deeply worried. Being in the E.U., their ability to adopt AI and create successful new companies is quite limited. Their most important companies date from the 1920s or earlier, so they have some very real concerns. I would rather have America’s problems than their problems, if I had to say, but it’s one of the best countries in the world. Let’s celebrate them.
Dalmia: Even Trump’s America, you would take over Denmark?
Cowen: Well, fortunately it’s not Trump’s America, right? I think that’s the point. It’s Americans’ America.
Dalmia: Well, yeah, except you have ICE squads running around in Minneapolis, shooting people in the face and yanking them from their homes.
Cowen: I’m totally against that, but look, Danish immigration policy can be pretty harsh. I don’t like every part of that either. They have some of the harsher treatments of immigrants in Western Europe.
Dalmia: But in terms of delivery of public goods, is there anything you would want the U.S. to emulate from Denmark? I’m asking this question because I’m hoping we can come to a political consensus between a libertarian like Tyler, centrist like Frank, and Bernie Sanders believes in Denmark, too.
Cowen: Well, if you’re talking about Denmark, I think you need to compare it to individual U.S. states. I think my state of Virginia is pretty well governed. Utah is pretty well governed. Much of California is a disaster. Mostly the state has too many leftists, I would say. New York City has done an incredible job on lowering its crime rate, which is now falling again. That’s a major triumph of state capacity that really affects how people live. So I would look at it state by state. We definitely have some major regions that are not at all well governed, and I think it’s obvious to most people which those are.
Dalmia: Frank, to preface what I’m going to say, I’ve been reading some of your works on state capacity. I have to admit, I’m much more persuaded by your case than I thought I was going to be. And I think we do need to grapple with some of your critiques of the American system.
But the way I see it, the way you want to go about creating state capacity and solving the current problems of governance, one of which you say is a vetocracy—there are too many veto points in, say, the delivery of any public project or even private projects where people can sue constructors and developers for all kinds of reasons. Also, you mentioned that, instead of Congress, it’s the courts that are legislating, which is very much close to what libertarians say. So in effect, the legislative function has now been transferred to unelected courts. But your solution to that is literally like creating a fourth branch of government, which is an autonomous civil service, kind of like what I grew up with in India.
“One of the big problems with progressives in the United States is that they have wanted the government to do things, to be activists. But they also have been the ones who believe that legitimacy stems from procedures and adequate proceduralism. And therefore, they have gotten in their own way by adding procedures to activist government that then make it impossible for activist government to accomplish things.” — Francis Fukuyama
India inherited civil service from the British. It’s very meritocratic. You have a very difficult entrance exam to get in. Once you get in, you’ve got lots of perks. And so these are very coveted jobs. And it’s quite professional. It’s quite rule bound. And yet it delivered the economic basket case of the world till India liberalized. India’s bureaucracy became hidebound and out of touch with the public and instead of serving the public, it lorded over it because these coveted jobs gave it power and prestige. It issued cumbersome and arcane regulations—the infamous License Raj—that bred rampant corruption because you had to pay a bribe to get a permit or license to start a business or just get a phone connection.
So, a number of questions. One: If you create this fourth branch of government in the form of this autonomous, insulated from political pressure, insulated from interest groups, bureaucracy, who is going to watch the watchers?
Fukuyama: Yeah, so you’re mixing up several different issues there, so we need to separate them. One of the things that I have written about is private right of action. That has to do with the litigation and the way that the courts intervene. In the United States, the majority of states do not regulate, and the federal government does not regulate, environmental abuses by actually specifying what they are. They leave it up to private citizens to sue on behalf of the environment. And then that goes into the legal system, and a local judge decides, “Yes, it does sound like Berkeley adding another couple thousand students can be a potential environmental hazard, and therefore we’re going to stop them from doing that.” That whole private right of action is an American tradition. We are the only modern liberal democracy that really uses private right of action in that fashion. And it’s a very inefficient way of regulating things. I mean, there’s some areas like employment law where it may be the only way that you can regulate, but in terms of environmental regulation, it’s very, very inefficient. The transaction costs generated by having to litigate through the court system is very high. So that’s one set of issues.
The second point you made, the question of bureaucratic autonomy, I do not think that bureaucrats ought to make policy. Policy should be made by legislators, and the policies then should be handed off to bureaucrats. But I think that once that handoff is made, there needs to be a lot more freedom of action for the bureaucrats to actually implement those decisions, allowing them to use judgment, common sense, rather than the legislatures making ex-ante detailed rules about how the bureaucrats need to implement those laws. So there’s accountability. I mean, the laws have to be made by democratically legitimated political actors, but the actual implementation of those laws needs a lot more ability to adapt to existing circumstances than bureaucrats are given these days.
Accountability also works in very, very strange ways. So you can say, “Well, how about ex post accountability, where if a bureaucrat oversteps their constraints, they should be sanctioned for that?” Well, it turns out in NEPA, the National Environmental Protection Act, you have ex post accountability for any big infrastructure project, you have to do an environmental impact assessment. Now, this law was passed in 1971. These days, the average environmental impact assessment runs to several thousand pages, more like 5,000 to 10,000 pages. Why are they so detailed? Well, part of it is there’s so much litigation. If somebody, a bureaucrat, makes a decision, they can be sued. And then what they do is the bureaucrats themselves anticipate the litigation and they write into the rule ways to avoid being taken to court later. And all of this just increases the amount of sludge in the system. There’s no question that bureaucrats need to be held accountable, but the actual method by which you do that needs to be more flexible.
One of the problems with infrastructure in this country is that when you’ve got a big project, there’s no one in charge. So there’s no single point of accountability. You have 17 federal agencies that have to sign off on the approval process for a big project. And you have to sign off on the approval process for a big project. They actually put the environmental agencies last in that chain. And if anyone objects, the whole process starts over again from the first lead agency. All of that stuff just adds to the bureaucratic burden that is required to actually get something done. And it injects the courts into what ought to be in most other countries done actually through a bureaucracy. So there’s a lot of ways I think that you can change the institutional rules to maintain accountability, but to make the system much, much more efficient.
Dalmia: Tyler, do you have any objections to that? Libertarians can’t be against removing sludge from the system and making it more efficient and less cumbersome? Or should they be against liberating the bureaucracy?
Cowen: I don’t think so, but I would just ask Frank the question: You’ve lived in the state of Virginia—is there any major infrastructure deficiency here, in your opinion? I agree it could be quicker, cheaper. I’m all for that. But, like, how bad is it?
Fukuyama: You know, I did not follow this issue when I was living in Virginia. I left 16 years ago.
Cowen: But it never hit you in the face, right?
Fukuyama: Well, you don’t know. There are many times when you’re puzzled about why a particular project has taken as long as it has. I suspect that in the rankings of states, Virginia has been turning blue or it’s purple right now, and it’s probably been a little bit freer of a lot of these burdensome regulations than California or New York. So I wouldn’t be surprised if it’s actually working better.
Cowen: But this is a general point I would make: While I agree with all your points about process, I think you’re “blaming” too much the process and not nearly enough “blaming” that this state is too full of leftists. Because I think Utah is fine; all kinds of states in the southeast are fine. Again, it could be better, quicker, cheaper. I agree with all the things you’re calling for. But blame the leftists, because the states that are not full of leftists are doing pretty well with infrastructure.
“I have a simpler diagnosis. I think when your state has two-party competition for real, things tend to go pretty well. That’s Virginia. It’s many states in this country. It’s not California. Electoral competition works. We believe in democracy. And again, it doesn’t mean these other claims are wrong, but it encompasses a lot of them. The pure, more right-wing states, they have their own problems. Mississippi, Louisiana, we could go into that, but they also don’t have two-party competition.” — Tyler Cowen
Dalmia: No, no, but to Tyler’s point, isn’t one of the dangers of autonomous bureaucratic branch precisely that it’ll be susceptible to ideological capture of the kind that we are seeing in California? How do you create a semi-autonomous bureaucracy that’s insulated from political and interest group pressures that doesn’t become an echo chamber?
Fukuyama: The realm of autonomy really lies in the implementation of a mandate that is issued by a democratically elected political source. So it gets very complicated in administrative law as to at what point are the bureaucrats in their interpretation of the law actually making law. That’s why you have the Administrative Procedure Act, you know, notice and comment and so forth in order to get some control over that.
So, Tyler, in response to the political question, there’s a really brilliant article written by Nick Bagley about five years ago called “The Procedure Fetish.” He is a Democrat. He was Gretchen Whitmer’s chief legal advisor in Michigan. He now teaches at the University of Michigan Law School. And he said that one of the big problems with progressives in the United States is that they have wanted the government to do things, to be activists. But they also have been the ones that believe that legitimacy stems from procedures and adequate proceduralism. And therefore, they have gotten in their own way by adding procedures to activist government that then make it impossible for activist government to accomplish things. And so that’s kind of an endorsement of what you said.
I mean, the problem is that the procedures come from people more on the left than on the right that actually want to hold government accountable. They want to make sure that citizens’ rights aren’t violated. There’s a whole literature. Paul Sabin wrote a very nice book called Public Citizens. There’s also this, I think, quite brilliant book by Marc Dunkelman called Why Nothing Works.
Cowen: Yeah, great book.
Fukuyama: There’s conservative anti-government ideology, but there’s also a left-wing ideology that grew up with Ralph Nader and public interest law in the 1960s that began to see government less as a progressive force, as it had been seen in the New Deal, but rather as a dangerous potential source of tyranny, essentially, or that it had been captured by the private sector, by corporate interests. And so most young lawyers, instead of wanting to work for the government in a regulatory agency, began to go work for a public interest law firm, and then they made careers suing the government, stopping it from doing things. So the anti-government movement in the United States was fed both by conservatives and by left-wing activists that thought that the government had been captured by corporate interests. And both of them thought the government needed to be constrained.
'Abundance’ Offers a Sounder Way Forward for the Left than Degrowth or Redistributive Progressivism
So I think there’s a kind of horseshoe effect for both the left and the right have come together in putting all of these legal and procedural constraints on government that then make it very hard for the government to do things.
Dunkelman’s argument is that we need to get back to a more Hamiltonian understanding; that government actually can do socially progressive things, but it has to be able to do them effectively. And therefore, you actually do need to roll back a lot of the proceduralism and kind of litigious resistance to government action that’s built up over the last couple of generations.
Cowen: While I agree with much of that, I have a simpler diagnosis. I think when your state has two-party competition for real, things tend to go pretty well. That’s Virginia. It’s many states in this country. It’s not California. Electoral competition works. We believe in democracy. And again, it doesn’t mean these other claims are wrong, but it encompasses a lot of them. The pure, more right-wing states, they have their own problems. Mississippi, Louisiana, we could go into that, but they also don’t have two-party competition. So again, there’s a pretty simple diagnosis that I think covers much of what’s going on.
Fukuyama: All of that’s true. I do think, however, that institutional rules count, and if you’ve got very problematic institutional rules, whatever your political competition is, it’s going to be ... I mean, for example, one of the issues that I’ve been looking at very intensively lately has to do with public input and public participation, because many of the reforms that have been done over the past 30, 40 years have increased levels of public participation. Many democratic theorists think it’s an unalloyed good thing that the more participants, the better.
But it turns out that that’s not necessarily true. And in fact, part of the reason that the permitting burden is so high in many instances is that there’s so many requirements for public input that you actually can’t get to a decision. Notice and comment under the APA is out of control because you can have a million comments on a particular agency rule. The agency has to answer every single one of them.
“There’s conservative anti-government ideology, but there’s also a left-wing ideology that grew up with Ralph Nader and public interest law in the 1960s that began to see government less as a progressive force, as it had been seen in the New Deal, but rather as a dangerous potential source of tyranny, essentially, or that it had been captured by the private sector, by corporate interests. … So I think there’s a kind of horseshoe effect for both the left and the right have come together in putting all of these legal and procedural constraints on government that then make it very hard for the government to do things.” — Francis Fukuyama
So I think that, yeah, it’s good to have two-party competition, but as a political scientist who worries about comparative institutions, there are also institutional changes that you can make that are going to mitigate some of that. For example, part of the participation movement has been party primaries. That’s part of the reason we’re so polarized: each party selects its candidates not by party professionals but by an open vote. Who shows up to vote in a party primary? It is activists. And so both parties have driven the Republicans to the right and the Democrats to the left by the fact that that’s the way we select candidates for major offices. And so if you change that rule or if you got rid of first-past-the-post voting and you went to ranked-choice voting or you went to proportional representation, you would also dramatically reduce the incentives for polarization.
Cowen: I’m less crazy about those ideas. I was with you up until that point, but we’re also not going to change our Constitution.
Fukuyama: Well, first of all, the electoral rules are not constitutional.
Dalmia: States can do that.
Cowen: United States is not going to have ranked choice voting for the president, right? There’s an Electoral College that’s in the Constitution.
Dalmia: Tyler, I mean, at the national level, you have two very competitive parties and what you get is gridlock. We don’t even pass budgets anymore. We run our government through continuing resolutions because they can’t come to an agreement on budgetary priorities. So that’s not working out.
Cowen: As of late, I’ve been wishing we had more gridlock. I hope we go back to some extra gridlock through the Supreme Court striking down the tariffs. So when America needs to get things done, historically it gets things done. It acts late, the Winston Churchill adage and all that. But I don’t think gridlock is really that good a model of U.S. government, either now or in most other times. It might have been true for like 10 years of Obama versus the Tea Party around that time, but it’s not mostly how I would model things.
Dalmia: Tyler, I do think you’ve seriously underplayed your state capacity bona fides, and we are going to try and correct that now. We are going to move to our lightning round, and I’m going to throw some public goods, infrastructures, projects at you, and you tell me, yay or nay, and in just a few sentences, why?
Cowen: I’m ready.
Dalmia: National high-speed rail. A federally planned high-speed rail network connecting Boston, New York, and D.C., and L.A. and San Francisco with eminent domain authority and limited local vetoes.
Cowen: We should have high-speed rail, something like D.C. to Boston, and that’s it for the whole country. It’s impossible at this point, but that’s my limit. The rest makes no economic sense. Fly.
Dalmia: Frank?
Fukuyama: Given the geography of much of the country, but I would include L.A. to San Francisco. It’s really crazy that we haven’t been able to build that in the last 25 years.
Cowen: But that particular case, L.A. and San Francisco areas are so big that the cost of getting to the rail place relative to doing the whole thing by flying, I’m not sure it makes sense.
Dalmia: Yeah, I’ve written about that. So, I’m not taking sides over here, but I am with Tyler on this one.
Cowen: You don’t have enough stops in between, is another way of putting it. Let them off at Santa Barbara, but, you know ...
Dalmia: Will that work for you, Frank?
Cowen: Well, I’m joking. I’m saying one high-speed rail network, eastern seaboard, stop there.
Fukuyama: Yeah, it wouldn’t go through Santa Barbara. The whole point is to actually populate the Central Valley, and most of the plans involve not going up the coast, which is, first of all, as an engineering matter, much more difficult, but to go up the Central Valley so that Bakersfield, Fresno, places like that will actually be suburbs of either L.A. or San Francisco. That would be the Chinese approach. That’s what they’ve done with their high-speed rail system.
Cowen: Your state’s losing population. If you reverse that, maybe we can talk about it. But right now, I don’t see it.
Dalmia: And they [Chinese rulers] have pretty draconian eminent domain powers. I’m not sure we can replicate that here. OK, this is supposed to be a lightning round. So let’s be lightning fast. Second one: nuclear power fast track, a federal authority that can approve, cite, and build next generation nuclear plants within five years, preempting state environmental review and waste disposal concerns.
Cowen: Of course. Yes, yes, yes.
Dalmia: Frank?
Fukuyama: Yeah, that’s fine with me.
Dalmia: OK, we have agreement there. Lower Manhattan Expressway, which if you remember, Jane Jacobs got into a big fight with Robert Moses and stopped it. So a New York state-mandated 10-lane express highway cutting through Soho, Little Italy, and the Lower East Side. And just remember, my son lives in the Lower East Side.
Cowen: I don’t know much about it, but I doubt if it’s a good idea. Not for New York city. If there’s one place that should move away from cars, it’s there.
Dalmia: Frank?
Fukuyama: Well, I mean, the thing is that it would be so expensive that it’s not clear that it’s going to be worth it. And it’s going to be expensive for all these permitting reasons that we’ve talked about.
Dalmia: Well, but the idea would be that Albany steps in and gives a reprieve somehow from the local vetocracy.
Fukuyama: Yeah, as a decision-making procedure, that might be important. I’m not sure that it makes sense. Just like I think small modular nuclear reactors don’t make any sense because they’re just not efficient. So there’s issues that need to be separated. The decision-making process needs to be speeded up. But whether what they’re going to decide on makes sense is a separate question.
“I don’t think American state capacity historically is that weak. We built this incredible empire, often unjustly. We put a man on the moon. We developed the atom bomb. We’re leaders in aviation and computers in part because of government. A lot of our state governments work really quite well. It’s a mixed bag, but I think we’d be in the world’s top 10 easily.” — Tyler Cowen
Dalmia: OK. Climate Adaptation Retreat Program, a federal program that forcibly relocates communities in flood prone or fire prone areas rather than endlessly rebuilding them.
Cowen: No, just stop bailing out their insurance. Let the market do that. I favor state action, but stop the state of Florida from rejiggering the insurance industry all the time.
Fukuyama: I agree with that. Yeah, we should just stop insuring places that shouldn’t be insured.
Dalmia: Okay, so this one I have to admit I got from ChatGPT: AI compute and data infrastructure, a publicly funded, federally run AI compute and data infrastructure, analogous to the interstate system to ensure national competitiveness. Tyler, this is after your heart.
Cowen: Just get government out of the way. I think the market demand is already there. We probably don’t need to spend public funds, but we need to allow people to build.
Dalmia: Frank?
Fukuyama: Yeah, I think given that we’re probably spending too much stuff on these data centers already, I don’t see why the federal government has to get into this.
Dalmia: Wow, we are more in agreement than disagreement here. OK, I’ve got two more. Strategic Rare Earth Mining on the Moon, a government-led lunar mining program for rare earth elements to reduce dependence on China.
Cowen: There’s rare earths everywhere on Earth, including in the U.S. We need to relax environmental regulations, work more with Latin America, but most of all, deregulate our own processes for refining and managing these things, which is where the real obstacles are. We can do it. We’re probably going to do that in the next few years. It’s actually a good example of how our state capacity, ex post, will end up not being dead.
Dalmia: Frank?
Fukuyama: Yeah, I think the moon is probably one of the most expensive places to get this stuff, so I don’t see how this makes much sense.
Dalmia: OK, last one. And here, I think we are going to get some disagreement. Federally funded universal health coverage. Tyler?
Cowen: Well, whether we like it or not, we have it for people over age 65 and Medicaid. That’s a huge chunk of our healthcare expenditures. We’re not going to get into some other system. I don’t want single-payer. That would be disaster. I think you want to hold on with what you have and regret that you did it the way you did it. And then finance innovation. Innovation is what’s going to make the difference. I think young people today will live to age, your son will live to age 98 and die of old age. That’s the way to make progress, not mess around with all the coverage.
Dalmia: Well, we don’t have to be single-payer to have universal coverage. I just want to put that on the record. You could have something like France, where they provide basic coverage, state funded, and then you buy extra coverage on top of that if you want to. So it’s not single-payer.
Cowen: But if the U.S. tried to do that, it would unravel into single-payer because the cutoff and margins and nudges for when you’re privately covered, the government picks up the tab. Very hard for those systems not to collapse into full governmental coverage. So we’re stuck with what we have. Let’s focus on innovation. I’m all for more science funding from government.
Dalmia: Frank?
Fukuyama: I don’t have an objection to a federal program that looked like France or Switzerland.
Dalmia: OK, at least we get some disagreement on that. So for all your disagreements on state capacity, when you actually come down to the nitty-gritty and start discussing specific programs, it seems you guys are quite in line with each other.
Cowen: And Frank, keep in mind the Swiss system you cited, it’s a reasonable amount like Obamacare. It’s government vouchers for private health insurance, which is what part of Obamacare does, whether one likes that or not. It’s not really government health insurance.
Fukuyama: Well, no, but the point is that if you have the money to afford private health care, you’re free to do that. It’s just that the government isn’t necessarily going to subsidize everything you do. I mean, that was what I was thinking about when I was referring to Switzerland.
I think, though, that in a certain way we’ve been talking a little bit past each other because I think that Tyler is talking more about political decisions and the dysfunctions of the political system and coming up with good policies that can be then implemented by a bureaucracy, whereas I’m focusing on the implementation capacity of the bureaucracy. And that’s where I think we’ve got a real problem.
Cowen: But I would be more convinced by how you prioritize matters if you could tell me what has gone wrong in Virginia, where, again, you’ve lived for some number of years. I know it was a while ago but it was actually worse then. If you could just tell me what’s been so badly screwed up I would see your point of view more clearly.
Fukuyama: I would have to return to Virginia and live there a couple more years before I could answer that question.
Cowen: But you don’t see it in the newspapers. You don’t see it in your social media feed. You didn’t notice it back then. I fully agree. California’s totally screwed up, right? We would agree there. But these other states, I would agree Louisiana’s totally screwed up. I don’t see that you have your finger on the main problem.
Fukuyama: Well, it’s not just a state-level issue. I mean, a lot of the climate adaptation can’t be done on a state level, sure. You actually need national programs, right? And those federal systems are ... I think, as a general rule, most state government works better than the federal government. Part of the problem of the federal government is one of scale. The United States is so big, it’s so diverse, it’s really hard to get political consensus across that diversity on the government acting. In a state, it’s much easier. I mean, it’s a much smaller political unit. There’s usually not as much diversity, and therefore, it’s easier to make political decisions. So that’s why it’s a good thing to have federalism. But then, there are certain things that need to be done on a national basis, like the interstate highway system. And there, I think you do need stronger federal government.
Cowen: People hate carbon tax, so I say let’s do YIMBY for solar and nuclear—and wind. Do you agree with that?
Fukuyama: Well, yeah. I mean, the question would lie in all the details of exactly what you are permitting. But yes, in general, I think that we need to deregulate a lot of those areas.
Dalmia: OK, I guess we are ending on a note of agreement here, then?
Fukuyama: Sure
Cowen: I will agree that we agree on many things.
Fukuyama: It wouldn’t have been any fun if we had agreed on everything.
© The UnPopulist, 2026
Follow us on Bluesky, Threads, YouTube, TikTok, Facebook, Instagram, and X.
We welcome your reactions and replies. Please adhere to our comments policy.
