A while back, Edward Luttwak explained to us how to stage a coup. Julian Waller focuses on the next step: how authoritarian regimes are built and maintained. He’s a research analyst in the Russia Studies Program at the Center for Naval Analyses and a professorial lecturer at George Washington University, where he teaches a course he calls Authoritarianism 101. The final assignment asks students to pick a real country and write a fifteen-page plan for taking it over and holding it for twenty years. He is the co-author of Autocrats Can’t Always Get What They Want and writes the newsletter Political Order(s).
We discuss:
Authoritarian regimes as the historical norm
Why military juntas don’t last and party states do
Autocracies: Iran, North Korea, and the Papacy
How the Kremlin has reacted to the Wagner mutiny
What a “pacted transition” looks like in Venezuela
Why the US shouldn’t worry about authoritarianism
Julian, you are a student and teacher of authoritarianism. Is that an accurate first pass?
Absolutely. A political scientist that works in comparative authoritarianism.
You’ve been on my mind as a potential guest since a couple of years ago, when I saw that the moniker of the class you teach at George Washington University (GW) is Authoritarianism 101, and the final assignment is to pick a real country and write a plan for taking it over and holding it for 20 years as an authoritarian.
Are you still teaching that class this fall?
We just started this past Tuesday. GW bureaucracy didn’t allow “Authoritarianism in Europe, Russia, and Eurasia” — it was too long. They pitched me “Dictators in Europe and Russia,” and I’m like, “That sounds fine. It’ll get student interest.” I’ve been teaching it for four semesters now and it’s always been wait-listed. It is an Authoritarianism 101 class. The whole point is to introduce students to the wide world of non-democracy. We go from the French Revolution to the present and talk about all the ways in which authoritarian regimes in broader Europe have existed — their institutions and ideologies.
The final assignment is perhaps the spiciest assignment at GW, and I fly under the radar here. I ask them to build me an authoritarian regime. Pick a country at any point in time over that 200-year span, and either transition a democracy to an autocracy or an authoritarian regime to a different type of authoritarian regime. They apply the theories and concepts that we’ve been talking about all semester to a granular case. Students are intimidated initially, then when they start putting pen to paper, they’re like, “Professor, this is really interesting.” I’ll get students that come in office hours and are like, “I’m a little scared. This is so fun.”
Longtime readers will remember we had the esteemed Edward Luttwak on, who wrote Coup d’État: A Practical Handbook. He has thought very concretely, from the perspective of a potential military junta, about how to do a coup. You seem like the natural follow-on to that conversation: after the coup is over, how do you hold a regime?
But before we get into the best practices you’re teaching your students, I want to get your 101 perspective on what makes an authoritarian regime. You’ve got a fun Substack. In at least one post you talk about the papacy as an example of a very successful authoritarian regime, which suggests the category is pretty broad.
That’s right. We can get into a massive rabbit hole about regime typology and definitions. I’ll try to keep it simple and coherent, and you can yell at me if that’s not working out.
Let’s go down the rabbit hole. We can back out later.
We’ll start from the political science perspective. The academic literature is often very explicit — this is one of the benefits of social science research: we try to keep our concepts delineated and explicit, because we want to measure them down the road. That’s a benefit, as opposed to just saying things. You can always say something’s authoritarian. The question is whether it makes any sense, or is useful comparatively or analytically.
From the political science perspective, authoritarianism is a term of art we use to describe all political regimes that are not electoral democracies. That’s not a very satisfying definition. It’s a residual category, and it wasn’t invented this way. This is an evolution over the last three-quarters of a century. Before, we had a variety of terms to describe the countries we’re talking about: tyranny, autocracy — back in the ’30s and ’40s, we’d talk a lot about totalitarianism.
Some of these concepts come from an old, even Greek, classical perspective on regime types, and some are more modern.
Humans have been thinking about the structures of governance for over 2,000 years. The idea of thinking about who rules, how, and in what way, comes from classical Greece, among other traditions. Plato, Aristotle, and Polybius are all writing about what we would now call regime type, as are Machiavelli, St. Thomas Aquinas, and Jean Bodin in the medieval and early modern period. They’re using the classical lexicon, which divides regimes into three structural categories. Are you ruled by one person, a few people, or many? Then you double that: is the rule done in the name of the common good, or for the personal sake of those who rule? This is where we get the divides between:
Monarchy versus tyranny,
Aristocracy versus oligarchy, and
The mixed regime versus democracy.
We don’t do that anymore, mostly. Social scientists are very averse to ascribing normative aspects to the concept. How can you tell whether a regime is working for the common good? We can have a great conversation about whether country X’s government means the best for its people, but it’s difficult to classify that and to have a neutral coding. We usually throw that out. Then we’re left with structures.
We think there’s something very special and odd about modern democracy. The apex political leaders — the president, parliament — are elected regularly in multi-party elections under broad suffrage. Here’s the critical bit: the outcomes are uncertain. There’s so much genuine contestation in those regular multi-party elections that you can’t guarantee the top will stay the top.
You have electoral authoritarianism, which would be cases where you have an election, but everybody knows what’s going to happen at the end of it.
Right. You have all these countries now with electoral-democratic-style constitutions: directly elected presidents and legislatures, formally independent judiciaries, itemized rights. It’s structurally identical. People get confused by this. “This country had an election.” Well, the election was rigged. How do we know? And what’s the dividing line between systematic rigging and standard politics, which is also never fully fair?
Politricks.
Or “political technology,” as the Russians say. That’s a long way of getting down to this baseline: we think democracy’s weird. It’s really weird to have uncertainty over those apex decision-making political elites.
By weird you mean historically uncommon — not the norm?
It’s only 200 years old, and up until the last 100 years or so, it was only a few countries. The United States is the oldest functioning electoral democracy in the world by a long shot. Even these old European democracies that we think of as very stable, that form the democratic normative core of the European Union (EU) — many of their democracies are no older than 1945 or 1970. The average age of a democracy today is around 35 years worldwide — depending on the continent it’s even less. We write as if democracy is this thing that’s always existed. It’s really not.
We bracket that aside, then we look at everything else in the world. What we notice is that everything else is very confusing. The political systems are not the same. A monarchy is not the same thing as a communist party state, or a military junta — or Vladimir Putin’s regime in the 2010s, or Erdogan’s Turkey, which both have elections and all this ephemera. But they’re all lumped together by the fact that there’s a certain outcome at the top. The leader can’t lose an election — or is almost never going to lose one if there are elections. In many, there are no elections at all. There’s certainty at the top as opposed to regular institutionalized uncertainty, and that’s the real difference between the two.
You’ve been a critic of the idea that authoritarian regimes are inherently fragile. When Assad fell, there was a spate of commentary about how the pathetic end to that regime taught us something about authoritarian regimes in general. Is that not true?
It’s not true. The fall of the Assad regime was very instructive, in part because of how long it took. This is a regime that was isolated for over a decade: massive sanctions, nobody wanted to touch it, economic collapse and degradation, a relatively small ruling elite surrounding a personalist dictator who had lost all the ideological aspects of the dictatorship. It was power for power’s sake, as well as existential survival against forces like ISIS, and propped up from the outside. But it lasted far longer than you would expect. The word brittle implies that it shatters very easily. That’s simply not the case. Syria had a decade-and-a-half drawn-out process of slow destruction, and that process was not a one-way street.
It’s important to push back against the idea that autocracies are inherently brittle — inherent as in something to do with the structures of power: that lack of accountability and certainty at the top. It’s simply not true. Some authoritarian regimes last an incredibly long time.
We can cut apart that big residual concept of authoritarianism into subtypes, and we can show, through both qualitative and quantitative analysis, systematic likelihoods relative to how resilient they are and how well they can manage crisis.
For example, military juntas — countries ruled by uniformed officers — are most commonly formed after a coup, or at the end of a civil war. Latin America and sub-Saharan Africa dealt with lots of military juntas, but Europe had plenty of variations on the coup format of government in the interwar period, and there are the coups in Iberia in the 19th century. Regardless, the average age of a military regime is one of the shortest relative to all others.
We have a few reasons why. Generals don’t like to rule whole countries. It’s very complicated, and it’s not in their skill set. They often want to pass the job off to civilians or technocrats: to run the economy, manage society, and deal with organizations that are inherently less regimented and hierarchical than the armed forces.
Another reason is they don’t want to be blamed when things go south. Officers often seize power because they have this sense that the nation is in danger. After a coup it’s very often promulgated from the new junta government that they’re there to save the nation, defend the people, and defend order. Once that happens, unless they’re very ideological — and that does happen in a subset of military regimes — they don’t want to be in charge when things start breaking. They don’t want to be blamed for bad economic progress or more disorder. They would like to go back to the barracks. That happens quite often.
In places where there is an ideological aspect to the coup leaders, they often transition away anyway. It’s not just the military taking charge; there’s also an ideological party, or they’ll create one afterwards. We see this in a lot of Global South countries in the Cold War, where you’ll have a military coup, but it’s done by communist or socialist-inspired officers. They’re not just interested in military rule. They have ideological underpinnings, and they usually transition to a regime where the party is the dominant organizational force in making decisions, not the top tier of the armed forces.
Let me ask you about party regimes. There are theorists who will say, “Look at Assad’s embarrassing fall and flight — autocracies are inherently brittle.” Then there are folks who look at China on the other end and say, “They build so many bridges. Authoritarian regimes are better on these axes.” There’s plenty of neo-reactionary folks on the far right who will say these regimes are stronger than democracies.
Is it true that party states have a longer shelf life? You can think of the USSR, China, or even Cuba as party states, and they’ve survived in their regime forms maybe longer than you might have expected.
It’s a great question, and that’s exactly the flip side to military juntas. You don’t want a military junta. They don’t last very long, and coups tend to lead to more coups. It’s all bad. If you’re trying to build an authoritarian regime, you really want a party. It’s great if you can get the full party state apparatus — the communist party state style — where it’s a single party and the party sits in organizational parallel to the organs of the state.
A party state is a specific subtype of the broader party-regime way of thinking about authoritarianism. Parties organize power and provide hierarchy. They also provide regular means of career advancement. You work inside the party, get rewarded, and become part of a network. This happens in democratic parties too. As an elite, you’re more likely to stay on board. You have a longer time horizon. If you believe in the party, or at the very least think it benefits you, you’re going to continue investing your work, ambition, and time into rising up the ranks and ensuring that the party does well enough to keep power.
All of these are positive incentive structures. Parties are also good because of that top-down hierarchy: you can discipline members, you can watch them, you have more information, relatively speaking, from people inside the party.
When we look at places like China, Vietnam, or Cuba, these have parties that are not only organizationally beneficial — the party itself is wrapped around the state and watches over it. That’s a very specific way of organizing a regime. It was developed by Lenin and his associates, and perfected under Stalin in the USSR in the 1920s and 1930s. Every state institution has a parallel party institution looking it over. The party institution makes the political decisions. The state institution implements them. That format allows the head of the party, who has control over its membership, extreme control over all political decisions.
This model was then exported to China, and a variety of other communist countries in the ’40s, ’50s, and ’60s. The problem is, to use the meme, we don’t know how to build them anymore. There’s never been a successful party state built after the Cold War. That particular format of the ideological party overwatching the state in that regimented manner is all legacies. China works incredibly well. It’s incredibly stable, given how large and how potentially diverse it is. But we can’t export the model. Not anymore.
In some ways this is similar to the point that you and other heterodox theorists make about fascism: that it’s not a useful analytical frame to apply to the present. Fascism was historically bound. Whatever you can say about authoritarian regimes today, they’re not fascist, because we don’t know how to make those anymore. Same with communist party states?
The difference is that we still do have communist party states. We see them working every day. The China watchers will tell you that the most dangerous thing about contemporary China is that Xi Jinping has personalized power so much that it’s unclear if the party state works the way it used to.
We don’t know how to make party states anymore. We do know how to make what you might call dominant party regimes. That is, an authoritarian regime where you still get the benefits of that organizational hierarchy and cohesion from a political party, but it doesn’t parallel the state in the same way. They’re more distinct, and usually that party is not a single party, it’s a dominant party, which means there are other political parties in the political system in opposition. There are regular elections. They may or may not be co-opted or controlled by the government.
Dominant party authoritarian regimes are also very successful. They’re not as long-lived as party states have been in the 20th century, but they’re pretty good. Mexico under the Institutional Revolutionary Party regime lasted seven decades. Singapore — we were mentioning neo-reactionaries — is sometimes discussed as this ideal CEO-monarch city-state model. We always forget that Singapore has a dominant party. The People’s Action Party is one of the core institutional mechanisms within Singapore for both controlling and rewarding its elite. It’s not all schooling, high salaries, and business degrees. There’s an organized dominant party, and it functions very well.
Other countries have this too, and some dominant parties are very corrupt. Fidesz in Hungary was an example of a dominant party that ended up eating itself. Russia has a dominant party, and it doesn’t matter. It’s not involved in political decisions. It’s there to fill seats in the parliament and keep governors on side. It’s not useless, but it’s not involved in politics in the same way.
Let’s do a lightning round, where I name a country and you tell me how you’d describe where it falls in the authoritarian pantheon. Let’s take Iran right now, locked in a struggle for its life with the US. What kind of regime does it have?
We could spend a podcast discussing the Iranian political system. But right now, under wartime conditions, it seems to be some kind of military regime under the Islamic Revolutionary Guard Corps, not least because we haven’t seen the Supreme Leader in months. Power is also unusually decentralized right now. Theater commanders have surprising amounts of authority, and we’re not quite sure how it interacts with the civilian population.
If we were to go pre-war, we would call Iran a theocracy, because those apex decision-making elites who ultimately call the shots are chosen by dint of their clerical status. But Iran is complicated, because it also contains a whole civilian government. It has an elected parliament, it had an elected president, it has a variety of conciliar bodies that have advisory or even decision-making capacities, some of which are elected by other councils, some of which are chosen from the top. Iran is unusually complicated.
What about North Korea?
North Korea’s fun. Structurally it’s a party state; functionally it’s a personalist dictatorship. The added wrinkle is hereditary succession.
We don’t have to classify the regime by succession. We could, but right now it makes the most sense in terms of power flow to think in terms of the personalist dictatorship — this completely unconstrained power of a single man who can do whatever he wants. He sits atop a legacy party state, and he has a hereditary succession model.
It’s not uncommon among authoritarian regimes that don’t look like monarchies to have de facto or attempted hereditary succession. The Assads famously had this. Azerbaijan is structurally an electoral authoritarian regime — it has a liberal democratic constitution, hyper-dominant party, very fraudulent elections — and the Aliyev family runs it like a personal dynasty, with the next generation being groomed for succession. Kazakhstan almost had the same thing. Nursultan Nazarbayev in the late 2010s was considering his daughter to become the next president. He had her all set up, speaker of the Senate and so on, and then a series of corruption scandals got in the way, and Dariga was considered no longer suitable. They ended up transitioning to something else. But it’s not unusual.
Close readers of Statecraft may remember Nazarbayev from our episode with Andy Weber, who was on the ground in Kazakhstan at the fall of the Soviet Union and was instrumental in getting the weapons-grade uranium that was lying around in warehouses safely out of the country. He brokered the Nazarbayev-Clinton deal to make sure that the Kazakhs were handsomely repaid for being helpful in securing that uranium.
That’s also a wonderful policy success story — unusually black-and-white. The regime used it as positive PR: that it is a non-nuclear state, it is there for peace, and it put its money where its mouth was. They don’t talk about the bit where it was paid handsomely. But it’s a source of internal domestic legitimacy, and it’s taught to Kazakh students in their middle and high school classes.
Last one in our relatively rapid-fire series. You have this puckish essay about the papacy as an authoritarian regime, but succession is uncertain. You don’t know in advance who the next pope will be. So how can you call it authoritarian?
This is why we build our democracy model on three legs. The most important leg is uncertain outcomes in terms of the decision-making officials. But it’s also regular multi-party elections and the size of the electorate.
It’s an electoral monarchy, properly speaking. When the Pope dies, the new monarch gets elected. He’s elected by a selectorate of a couple of hundred people in the College of Cardinals. It’s more like the Polish-Lithuanian Commonwealth, in terms of the old electoral monarchy. But the College of Cardinals doesn’t act as a legislature and it doesn’t meet regularly.
Its function is to choose the new, all-powerful monarch. An expert on the papacy will say, “It’s more complicated, and there’s lots of bureaucratic structures” — but everywhere is more complicated. There’s no country, even in the most autocratic state, where the leader literally makes every decision. The point is, where does ultimate authority and power lie?
The Vatican is a fun example. It is by far the most successful, too. If you know the history of the church, it’s not without schism, and what we might call coups, or even periods of civil war — yet it endures, which is quite interesting.
Let me ask you about how one maintains a regime. What is an effective authoritarian leader thinking about when attempting to elevate or destroy talent?
Personnel politics is all politics at the end of the day. What’s important is who fills the seat, who’s making the decision. The leader of an authoritarian regime who’s thinking over the long term — maybe even thinking for the common good, but again, we’ll bracket that away — is going to be helped by a number of factors. You want:
Good economic growth,
Social order, to some degree,
To avoid tensions between groups, and your country is going to have variations on that, and
To have clear career trajectories for ambitious elites, or those aspiring to be elites. They want not to feel constantly frustrated.
How do you do that in an authoritarian regime? There are all kinds of ways to rise in an American system, even if the pyramid narrows and there’s only so many senatorial or gubernatorial roles. There’s a lot of ways that you can spend a whole career climbing through the ranks and be rewarded by voters. If you don’t have true contestation, how do you stop everybody from feeling like they’re under the yoke in your regime?
This is something that current regimes struggle with, which is generational change. You have to keep enough positions open. Contestation adds one fun aspect to the circulation of elites. I don’t know how much you’ve read of the Italian elite theorists — Mosca, Pareto, Michels, people writing in the first half of the 20th century — but all countries have elites, and elites are going to circulate. Elites have a persistent stickiness. They last longer than you expect. Think about WASPs in the United States.
There’s a paper about Chinese families that were landed before the Cultural Revolution, and even having had everything taken away from them…
They persist back into top leadership.
We find the same thing in the Soviet Union. The nomenklatura elite is reestablished after a couple of decades of death, murder, and class genocide — when you get to the other side, to the 1970s and ’80s, a lot of the family legacies surprisingly persist. Revolutionary regimes are good at killing the top aristocrats. But people who are more provincial, small-e elites tend to persist.
There’s lots of persistence in the actual people who make up the elite corpus, which can be quite large. But there is still transition. There’s generational transition, and there are shifts from the provinces to the center. A trick with many authoritarian regimes is that they’re highly centralized relative to some democracies. People tend to glom to Beijing, or Moscow, or Tehran. We see this in some democracies. France and the UK are like that. The US is not. When you have overcentralization, that reduces the number of potential slots for people to feel like they’ve made achievements in their own region.
The standard ways of making elites feel good about themselves are things like party mechanisms. The Soviets had the phrase “the rotation of cadres.” That did two things. You were constantly being promoted to the next echelon for good behavior. This encouraged corruption and sclerosis over time, but people were happy. People remember the ’60s and ’70s in the Soviet Union very nicely, if they had nice lives. The other thing it does is prevent an elite from establishing a power base anywhere. You don’t sit in a position for too long.
This is not dissimilar to how military organizations work. Modern generals don’t have commands for more than two to three years, even in peacetime, even without political intervention. Part of that is intentional. You don’t have Captain Picard on a five-year mission on the USS Enterprise with his own ship and his own crew that never changes. That’s intentionally done to avoid fiefdoms being established. Authoritarian regimes try to do this too. It’s a defense mechanism against alternative power centers being created, and it’s also a good way to continue the career progression of people who are particularly ambitious.
I’ll give you an example. In Russia today, there’s a lot of talk about the future succession, what’s coming after Putin. Putin’s getting old.
How old is Putin now?
He’s almost 74. He could last for another 10 years, but he’s starting to sound like he’s lost a trick or two. The rambling of an older man — we have some familiarity with that.
There’s lots of succession questions in Russia. One of the top figures who is being talked about as a potential successor — he’s not been tapped to be heir, and there’s many reasons why we shouldn’t trust this — is a guy named Aleksey Dyumin, who’s 54. He started out as Putin’s bodyguard, and he has since moved through the state. He was governor for a time; he held different positions elsewhere. Currently, he’s the head of the State Council, a major advisory body underneath the president. This is the career trajectory of an ambitious elite who’s getting rewarded for good service. He’s also being moved around, and that’s in principle what you want.
One of contemporary Russia’s difficulties is that that hasn’t been happening in a lot of positions. There are the Russian equivalent of boomers who have been sitting in key upper-tier offices for 15 years. That’s bad, because there’s no fresh blood, and because it’s a blockage at the top. Imagine being a deputy in the Federal Security Service (FSB) who’s trying to rise to the top. A guy named Bortnikov becomes FSB head in 2008. It is 2026. Guess who is still the head of the FSB? The same guy. You haven’t moved up. You’ve just stayed there and gotten old. Imagine that across half of the institutional imperium of Russia’s governing bodies. That’s a bad thing, and we’re starting to see the signs of elite dissatisfaction with that.
I’ve seen people analyze this as having two opposite consequences. You can talk about the system becoming less dynamic. At the same time, you can say it’s making it more brittle and more prone to sudden change. There’s an interesting, hard-to-predict quality here. I read Samuel Moyn’s Gerontocracy in America, where he’s emphasizing a very different part of the problem, which is that you just get slower, stiffer, and less dynamic as a society. Whereas you can also see — maybe this is true for us as well — that the blockage means all that energy has to route elsewhere, potentially in revolutionary outpourings.
Social revolution doesn’t happen that often, but when it happens, it’s not fun for anyone. You want to avoid those things if you can.
Tell me more about Russia today. We had John Lechner on, a great reporter who spent a lot of time in Africa and Syria with the Wagner Group, including under house arrest. He was very interesting talking about Prigozhin, who was a caterer and rose up through the ranks, not too dissimilar to the bodyguard-to-governor trajectory, but had a very bloody end. The Wagner Group is a pretty interesting case of a power center that is in some ways underneath Putin and in some ways was not fully under control.
How has the regime learned from the Wagner history?
It’s such a bizarre story, especially to tell in hindsight. One day in the summer of 2023, all of a sudden, there was a mass military movement by a semi-autonomous mercenary force marching on Moscow, in an armed negotiation with Vladimir Putin, president of Russia, over firing his defense minister and chief of the general staff. That ends. The rebel goes to Belarus and hangs out for a little while under Lukashenko, comes back, starts meeting with African dictators and attending conferences in St. Petersburg, and then two months later has an accident in the sky and is gone.
What are the odds?
It’s bizarre. Two or three years later, it’s like it never happened.
All regimes — unless the military is running the show in those military junta cases — want to control their military. If you read your Samuel Huntington, you will find that civilian control of the military is normative. It is a good thing. We want it, but it also has to be somewhat sensitive to military autonomy, especially policing its own affairs.
Russia and China have very strong traditions of civilian control over the military, precisely to avoid coups. For communist regimes, there’s also ideological loyalty aspects that matter. Russia’s not a communist regime, but there’s still this legacy of civilian domination over the military.
That’s all fine and good, but during the entrepreneurial 2010s the Russian regime let the Wagner Group under Prigozhin grow into something quite large — outside the chain of command of the Russian armed forces, and therefore outside of the standard civil-military order. It was used for political purposes in Africa and in Syria, and then was redeployed into Ukraine, because Wagner seemed to be able to produce the mass of men needed and use them on the front line more effectively than other operational commanders at the time. This further increased the autonomy of Wagner, because now there were tens of thousands of armed soldiers under Prigozhin’s direct command. The way he ran Wagner, there was a bit of a cult of personality — the sense that you’re really fighting for Prigozhin.
This proved to be very dangerous, because when you have semi-autonomous armed warlords, they are going to involve themselves in politics in the country, and that is indeed what Prigozhin did. It turned out poorly for him. In our report at the Center for Naval Analyses, where I work in the Russia Studies program, we call Prigozhin a political-military baron: someone who exists outside that formal hierarchy, who can both act militarily in an autonomous fashion and politically in a subordinate autonomous fashion.
The only other close comparison in Russia is Ramzan Kadyrov, the dictator of Chechnya, who has governed Chechnya as a fiefdom or a feudal state, with a direct connection to Putin. Kadyrov and Putin have a personal patron-client relationship that bypasses all the institutions of the Russian Federation — so much so that the FSB is officially not allowed to engage in any activities inside of Chechnya. It’s an informal part of the deal to keep Chechnya safe, secure, and quiet. There’s no terrorism, there’s no separatism, there’s nothing.
Prigozhin also had a personal feudal relationship with Putin, and with a few people around Putin, because he was a fixer — at a mass, political level involving the armed forces and coercive violence. That, however, turns out to be a bad model, because you get things like the march on Moscow to negotiate over personnel politics.
Unsurprisingly, the Russians have learned that that’s a bad idea. They haven’t fully gotten away from the private military contractor structures. A shoe that’s never dropped, and it might not, is that a variety of governors and large state-owned companies have their own militias now. We speculated that under a different political situation, those also might develop into small baronies, at least of capacity for autonomous violence. That hasn’t happened, in part because the FSB and military intelligence figured out that they have to control this stuff much more directly. This is a successful case of regime resilience. The regime had this massive blow-up, it solved the problem, and nobody is close to where Prigozhin was in terms of autonomous power.
The last point, which is not well known, is that in the immediate months prior to the rebellion — and it’s not a coup, it was an armed negotiation that went south because it was dumb, but that doesn’t matter, people make dumb decisions when they have autonomy — the walls were closing in on Prigozhin. The Ministry of Defense was changing the contract terms for Wagner soldiers. They were still withholding ammunition. There was a sense that Prigozhin was on the ropes and was going to lose his access to coercive force at a large scale.
He started talking with the co-opted loyalist opposition parties in the country: A Just Russia and the Liberal Democratic Party of Russia, which is of course neither liberal nor democratic. There was a plan to have a committee hearing and a presentation in the State Duma, the parliament, castigating the defense minister and the chief of the general staff. There were rumors — these are Telegram-level rumors — that Prigozhin was going to support one of these parties in upcoming elections.
All of this is to say that when you give people that degree of coercive autonomy, they act on it. Even in a regime like Russia — personalist dictatorship, very closed, very authoritarian — these buttons still exist. When an entrepreneurial actor gets out of line, things can go off the rails.
There’s this general problem, in democracies and authoritarian regimes alike — how do you control the people who have control of violence? You can see this in Russian history as well with the NKVD, if you go back a ways, and I’m thinking about this because I recently re-watched The Death of Stalin.
What is the thoughtful authoritarian to do about the fact that you need to have somebody who can exercise violence for you?
The safest solution, and the most common, is what we call coup-proofing. You have multiple coercive organizations that can check each other and are distrustful of each other, such that no one organization is powerful enough to think about overthrowing you, or to aggrandize power for itself. You have this market that you can manipulate from the top, balancing coercive organizations. Sometimes you balance a regular military with a paramilitary organization. In other cases, especially in less developed countries, you’re balancing a formal security service with tribal groups that have their own access to autonomous violence and have personal connections to the top. Libya is a good example.
If you’re a more developed, institutionalized regime, you tend to have more hierarchical authority, but still you are going to want to balance your security factions and your intelligence groups. Ideally, you have more than one, and they have strict limitations on what they can do.
To go back to Russia right now: one interesting distinction that we’ve noticed since the wartime period is how much power the FSB, the internal security services, have aggrandized. They are running around arresting governors, mayors, and people in the business community pretty much at will. Sometimes they’re doing it for corrupt reasons. Sometimes they’re doing it as anti-corruption purges. This has raised significant tensions, because it reneges on the deal that if you are a loyal servant of the regime, you shouldn’t be punished for putting your hand in the till at a small level. In a country like Russia, that is understood. Not everyone is hyper-corrupt, but most people are implicated, based on Russian legal codes, in at least something.
When you break down that elite contract, elites get nervous. You don’t want nervous elites. It’s quite dangerous for the regime internally, this degree of elite discontent. Discontent with the lack of generational change at the top, with a war that doesn’t seem to be ending, with an economy that’s stagnant at best, and with the sense that the internal security services have been pulled off the leash. It doesn’t mean the regime’s going to collapse. It does mean politics gets more tense.
Coup-proofing is the key way. Another way, if you live in a more ideological time, is that you want ideologues watching people. In China today, there are party cells at every institutional level: in the military, inside the intelligence apparatus, in schools — there’s a member of the party elite informing on everyone in every classroom. That’s how these regimes solve that problem: they have this overwhelming quasi-totalitarian approach — this full fusion of party-organized state power into society.
Let me ask you about how you transition out of an authoritarian regime. Part of the interest here is driven by events in Venezuela recently, by events that may be occurring in Cuba, and initially in Iran, where the US has tried a different approach to changing regimes: taking a little bit off the top, and leaving the power structure of the regime.
You can argue, as a friend has argued, that this is in large part due to technological shifts — it’s easier to do decapitations. Part of the reason that we didn’t do this so much in the 20th century, and had normative prohibitions on assassinating world leaders, was that it was hard. We tried many times to kill Fidel Castro with exploding cigars, and it didn’t work, so we stopped and claimed we were not going to do that anymore.
What’s the account of that new mode of trying to switch over a regime?
We all want to get to the end of history, to besmirch the good Francis Fukuyama. In the popular mind, it’s often revolution — democratic protest that leads to sudden regime change, because every now and again there are waves of that happening. The fall of communism in Eastern Europe is the big example, and it turned out well relative to every other wave of democracy that’s ever happened. Many of those countries are still democracies. That’s a good track record. That’s better than the Arab Spring, where they all failed.
Some of them are doing better than their Western European democratic compatriots.
Exactly. We have this nice story, but it’s only true sometimes. One of the big takeaways when you dive into trying to make comparisons is that temporal differences are really big. What’s going on in the ’80s is not the same as what’s going on in the 2010s, or in the ’30s. We can’t make blanket comparisons.
But bracketing aside revolution and popular overthrow, which don’t often happen and often are not successful, there are a few ways to do this. One is called the pacted transition, which comes from the Latin American experience in the 1980s primarily. It’s the idea that you have an authoritarian regime, but for a variety of reasons — usually social unrest and elite discontent — there is a view towards reform. It’s often undergirded by a sense that the economy is not doing particularly well. Think Brazil or Argentina.
What you end up getting is a model where you divide the regime into what the literature calls hardliners and softliners — those who want everything to stay the same and those who want reform. Then you divide society into the soft opposition — you could call them liberals — and the hard opposition, who want massive overthrow and are often very ideological. In the Latin American case, they tend to be communists or socialists. The key to a pacted transition is that the softliners at the top and the soft opposition at the bottom negotiate with each other, and push the other two out of it. You have this slow negotiation where gradually you introduce freer and fairer elections. Gradually that opposition comes to take ministerial positions in the government, and you usually have guarantees to the regime elites that they will not be prosecuted for their crimes.
It happens in a variety of Latin American cases. Chile is a famous one. There’s a complicated history with Pinochet’s referendum, but the general model applies. You don’t have violent overthrow. You don’t throw all the bums out immediately. You have this negotiation, it can take multiple years. Nobody’s super thrilled. The corrupt bad guys are still kicking around. They might even be elected into parliament later on, or become president.
I’m half-Argentine, and there’s a book that I remember very vividly, of a woman talking about her experience in the Dirty War, including being tortured. Much later, in Buenos Aires, she’s on a bus and she realizes that across from her, reading the newspaper, is one of her former torturers — because as part of this uneasy transition to democracy, we’ve let those people off. Certainly not everybody’s thrilled.
Not everyone should be thrilled. We shouldn’t model it as if suddenly everything goes away once this happens. That’s not how that works. Some people have this idea that you’re going to get these truth and reconciliation commissions, or Nuremberg-style trials, or a post-apartheid South Africa. Usually if those happen, they’re happening decades later, after the new political system has been consolidated. In most cases, people are just unhappy and have to deal with it. That’s been the Latin American experience anyway. But it works. All those regimes transitioned to democracy and have had rough politics, but they’ve had electoral democratic politics for 30-plus years. That’s a good outcome.
In Venezuela, we’re not having a pacted transition yet, but we could, and it looks like we’re on the way toward it. It’s the closest model. I think you’re correct about this technology-driven ability to do precision decapitation of the regime apex. Literally arresting the president in a special operations event that goes off incredibly well.
I believe he’s down the street from me here, in a jail in Brooklyn.
It’s completely nuts. It totally worked. It helps that Venezuela is very poor, very weak, and that the political elite surrounding Maduro didn’t like him all that much. They’re also very corrupt.
You buy off a lot of people.
We have this moment where Venezuela’s under a de facto viceroyalty run out of Marco Rubio’s cell phone, connected directly to the people who were previously just under Maduro. There’s lots of talk in Venezuela about a future political transition, and the State Department is saying the same thing, but the point is to not do it too early. That sounds to me a lot like using the pacted transition model, coerced from the outside. By taking away the top, you increase that sense of vulnerability among people like Delcy Rodríguez, who’s currently running the show, and a few other figures from the old Maduro regime, negotiating directly with the US.
The US is signaling extremely strongly that it doesn’t want the classic exiled Venezuelan opposition creating social discontent or running too hard. That all speaks to this model of trying to manage it slowly, keeping the old authoritarian elites in power for a time, but making it very clear to them that the shadow of the future is coming, and the expectation is going to have to be some inclusion — but it’ll take years.
If I’m advising the State Department, I would say that’s not a bad way to look at it. It’s much better than purging everyone immediately, doing de-Ba’athification. It’s also much better than trying to force the exiled opposition into government immediately, who are untried and untested, and don’t have control over the coercive apparatus of the Venezuelan state, which is still largely in the old regime’s hands.
One counterpoint that you could make is to say, “The exiled opposition has massive popular support, and in a fair election would’ve won by an absolute landslide.” The argument that those folks are making, as they’re being locked out by the US from returning and attempting to take power is, “We don’t need to slow-walk this. The vast majority of the country would be very happy to see us in office tomorrow.”
What’s wrong with that from a pacted transition standpoint?
I think it’s perfectly acceptable as an answer. But the idea undergirding it is that popular support is so pervasive that there’ll be a cascade in favor, even among regime bureaucrats, and most importantly the army, the security services, and the paramilitary organizations that the regime has put together through its patronage network.
We have to take into account that all those things still exist. We didn’t invade the country, we didn’t destroy the government, and there was not a popular revolution. They can be as popular as they want, but if people start shooting voters in the street, or if an election happens and the regime bars the door and doesn’t let them in, it doesn’t matter all that much.
The US could gamble that way. That’s been our traditional model in sub-Saharan Africa for development purposes: just hold an election and we’ll be fine. [Statecraft discussed the problems of this US approach to Africa with Judd Devermont.] The pacted transition model says the opposite: first transition some of that opposition into government, then hold an election, so that they already have some power. The next step has to be bringing some of the exiled opposition into the ministries. We haven’t seen that yet, but that’s what I would recommend. Then you can say that it’s no longer the opposition. It’s now the civilian pro-democratic politicians in government, and we want to confirm them in a future election, which should be beneficial for stability.
What we’ve seen in the African context is the opposite. You hold an election, the regime wins, or it creates a very fragmented parliament, and what’s next? A coup.
Recently we had on the show Will Howell and Terry Moe, on their book Trajectory of Power: The Rise of the Strongman Presidency. It was about the trend towards concentration of power in the executive. We went back and forth on how much authoritarian risk you can describe there being in our American context.
What’s your perspective?
The authoritarian threat to the United States comes if elections stop being free and fair. If we’re no longer choosing our leaders because there’s systemic, pervasive bias in elections, such that we can no longer reasonably contest power, and there’s no longer reasonable uncertainty over who becomes president, or what Congress looks like, then we become an electoral authoritarian regime.
Does a strong president do that inherently? Of course not. Part of the case that centralized power under a presidency is dangerous comes from comparative cases in Latin America primarily, where strong presidents ended up either pursuing self-coups, autogolpes, or otherwise seizing power and controlling elections.
There’s something that political scientists have called the peril of presidentialism. It isn’t just how powerful the president is. The issue is that the president and the legislature are separately elected, and therefore they have separate claims to legitimacy from the population. The president has a superior claim, because only he was elected by the entire country. Everyone else in the legislature was elected in a district. Inherently the president and the legislature will fight each other. You could imagine having one party in the presidency and another in the legislature, and the president will be able to say, “I am the true representation of the people’s will.” That’s the peril of presidentialism: under executive-legislative contestation, hardball politics that could spiral out of control, the president will always win. It’s not about whether the president has a lot of authority, or power, or constitutional protections. It’s this specified structural aspect.
If that’s the case, you might agree with our liberal critics, in that the current administration does speak that way about President Trump’s legitimacy. As you point out, this is not just a Trumpian phenomenon — the president in any regime has this claim. So why not be spooked about the perils of presidentialism here?
Executive-legislative contestation is inevitable, and it can be part of a healthy process, just like obstruction can be part of a healthy political process. You don’t want everything flowing too smoothly too often. The question is always qualitative: when does it fall off the rails? When does the president close down the legislature? That’s what happens in these Latin American contexts — the president shuts down or coerces the legislative branch. We’re just not seeing that at all in the US case. We have hardball politics, but within constitutional norms. We have very competitive elections. We have elections at multiple levels, not only for the president and the legislature, but also for governors and state legislatures. We have all sorts of contestation. We have an incredibly pluralistic media. All of these point to democratic politics.
But here’s the trick. People don’t always like democratic politics. That’s the real problem — people want their side to win. The peril of democracy is when people get too discontented with the other side and don’t think there are institutional solutions. That can get us into very interesting places.
You describe the polarization and the institutional mistrust as a rather separate bucket from the concern about a specific president’s authority.
Polarization can be quite useful sometimes.
It creates antibodies.
It’s also motivation. Oftentimes we see a transition to an authoritarian regime from an electoral democracy. You’ve heard the term democratic backsliding — this idea of the gradual aggrandizement of power by the executive. The part that they don’t tell you is that for all of that to happen, you have to have supermajorities in the institutions. You make constitutional changes and pack all the institutions with your loyalists. That takes a lot of work, and polarization prevents that from happening.
When leaders get elected in democracies and then autocratize — think Russia, Turkey, Hungary — these leaders portray themselves as centrists. Maybe not full ideological centrists, but they argue that they represent the broader nation, stability, consensus, and the average person. At least for their first few elections, that’s usually borne out. The voter is usually voting for stability. In many of these cases the regimes then become increasingly ideological, but that’s not how they start out.
That’s very important, because in a highly polarized context that’s almost impossible. There’s a great book called Pluralism by Default, which describes those post-Soviet countries that never created stable authoritarian regimes. The motivating case is Ukraine in the 1990s and 2000s, but Moldova also fits in, and a few other countries. The idea is that these countries are so divided — naturally divided, for ethnolinguistic and ideological reasons — that you can never consolidate full authoritarian control, because there are simply fundamental disagreements inside the body politic.
The United States is incredibly diverse: regionally, ethnoculturally, in terms of ideological alignment. That feeds into our politics. It makes it very nasty, but it also makes it pluralist by default. It’s very hard to aggrandize power here successfully. That’s good.
A very common type of Statecraft reader…
Handsome, I imagine.
Say 20-35, maybe into YIMBY, technocratic, urbanist politics — wonkier than many of their friends. If you’re talking to this reader and you want them to understand authoritarianism better, what would you push them to further study?
At least some of them have seen memes about Lee Kuan Yew.
That would be very representative.
Some of them have looked at China and been like, “That’s very interesting.”
At the theoretical level, the key point is to stress the diversity of authoritarian regimes — they really don’t come in one stripe. When you’re reading The New York Times, or The Washington Post, or the latest issue of The Atlantic, or Vox as a dirty secret…
They can read whatever they want. We’re pluralist by default.
We can’t help ourselves. We read too much.
Whenever someone writes, “Autocracies always do X,” or, “There’s an inherent logic that always works,” it’s very rarely true. When it is true, it’s at such a high level of abstraction that you’ve likely missed most of the country context. There are real patterns — we’ve been talking about party states versus military juntas. That’s a useful analytic lever that you should be pushing on. But if anyone ever says, “Authoritarian leaders always think this way,” run away. It’s wrong, it’s not helpful, and you really are too smart for that. If you’re able to parse the nuances of zoning regulations in some subset of Manhattan, you probably should not think that the majority of the world’s countries operate on a single logic that explains their politics in a way that’s useful for you as a policy-oriented reader, or as someone who might want to take lessons from the more successful variants.
The diversity of authoritarianism is important: they’re not all made the same, their institutions are distinct, the flows of power and authority are patterned but distinct, and they change over time. If I’m talking about Putin’s Russia in 2004, it doesn’t do a great job of describing the politics of Putin’s Russia in 2024. You should map that onto your favorite or least favorite authoritarian regime as well. People don’t say that often enough. We don’t want to become historians and say every single country is a special snowflake, and we can’t make any general characterizations. I don’t buy it. But you have to live in that meso level. That’s important, especially as the world is becoming more authoritarian in terms of the number of countries that look this way. It might be the future in parts of the West as well.
The second bit is to the technocrat, to someone who wants optimal policy outputs based on experts, based on — what’s the term? — high human capital.
I don’t really like that term.
Neither do I, but I’ve seen it bounced around. To close out spicily: one of the fun post-liberal, new right figures, Bronze Age Pervert — he has a whole article comparing Singapore to Orbán’s Hungary, and talks a lot about differences in human capital, technocratic institutions, and what the institutions are doing in terms of forming new elites. It’s not the best essay in the world. It’s also very partisan, per his own particular motivations. He loves eugenics — I recommend not paying that much attention to that part of the essay. However, it’s better than most of the new right stuff in terms of comparative authoritarianism.
For the technocrat, the tension of technocracy is that you reduce input from the population, not only on policy levels, but in terms of political contestation. If you’re contesting over fewer things, people get dissatisfied, and you’re going to have blowback. If you’re in the US, you’ll hear of the uni-party; in Europe, the cartel party. The idea is that in the 1980s, 1990s, especially the 2000s, parties converged on the same policy model, with maybe foreign policy differences. People throw around the term neoliberalism, which means whatever you’d like it to mean, but something like that.
Much like authoritarianism — perhaps an overly broad designation.
Or a designation that only operates at one level of analysis, and then we have to go from there. But if you’re a technocrat and you’re a YIMBY, and you think market forces are good and expertise is valuable…
I do.
Then you’re going to be inclined to think that a lot of that stuff was on the right target, that it produced significant economic growth, and that it might have been better governance. You would not necessarily be wrong. However, every country that did that in the democratic West has seen massive populist backlash. First on the right, and now in the US on the left. That’s an inevitable part of politics. You’re not going to be able to get out of that as long as the regime is democratic. That’s something you have to keep in mind as someone who loves technocracy, someone who thinks this is really the way forward. Democratic politics doesn’t allow for long-term technocracy without blowback.
You’re putting a name to something I’ve encountered a lot in these conversations about state capacity. Many people I know care about state capacity because they think improving it can rebuild trust in institutions. But there’s a big difference between that model, which is basically true — you can make people like the Department of Motor Vehicles more — and the idea that improved state capacity can bring back the neoliberal consensus. Often there is a sense that if you provide better government services and deliver on commitments, that that is sufficient to end political contestation over some of these questions. I think that’s the wrong way to look at it, and you should want state capacity despite the fact that your partisan opponents will occasionally hold the reins.
There is unfortunately an elision between good governance in terms of the state functioning better — providing what it says to its citizens more effectively, efficiently, and with a smile on its face — and, “Here’s my specific policy portfolio for trade relations and any variation on it is bad.” The problem is that the high politics side of things is never going to avoid democratic inputs. Unless you take my class. Then you can have certainty at the top, and as long as you can somehow keep the motors running, you can have good outputs.
There are far more authoritarian regimes whose citizens’ lives are pretty poor than there are ones that are nice and comfy and happy. That’s not inevitable. In the present, you have your Singapores and Hong Kongs. For most people under Pinochet’s Chile it was fine. You just didn’t talk about certain things. You can get your good governance outcomes. If you’re living in Moscow or Beijing, life is nice. You just don’t talk about stuff and in Russia you have to avoid mobilization. But that’s the Russian state’s problem.
I’m a big fan of my rights in an electoral democracy, and I wouldn’t trade them for any measure of stability.
You can see the stealth form of overbearing technocracy in the EU, where more and more policy areas are being decided by Brussels. You have literally the same backlash. The EU is still predominantly made up of democracies, but part of that tension is inherent to the technocratic impulse. Technocracy comes from the Progressive Era. It was popular during the New Deal — which was the closest we’ve come to a dominant party electoral authoritarian regime. We didn’t build it. There’s some pluralism-by-default reasons for that. But it was the closest we’ve gotten so far.
