Into Asia
Into Asia
Inside Xi Jinping’s Military Crackdown
0:00
-40:44

Inside Xi Jinping’s Military Crackdown

China’s top generals are falling like flies, and at unprecedented speeds.

China’s top generals are falling like flies, and at unprecedented speeds. As Xi Jinping dismantles the upper ranks of the People’s Liberation Army, questions are mounting about corruption, loyalty, and the stability of China’s political system.
Neil Thomas, an expert on Chinese politics at Asia Society, unpacks the logic of Xi’s military crackdown.

Neil's articles: https://asiasociety.org/policy-institute/neil-thomas

Follow him on X: https://x.com/neilthomas123?lang=en

Chang:
About a week ago, we got the news—announced by the Chinese military and also reported on Tsinghua—that two generals, Zhang Youxia and Liu Zhenli, were purged. As someone who’s been following this closely, I’m curious how you reacted to the news?

Neil:
I’d say I was surprised, but maybe not shocked, given just how much of a knife Xi Jinping has taken to the PLA [People’s Liberation Army] high command.

Over the last few years, there have been rumors that Zhang Youxia might have been in trouble for two or three years now, dating back to 2023, when the first big “tiger” of the PLA fell—the defense minister Li Shangfu, who disappeared from public view for several months and caused a huge international story at the time.

Li had worked for many years in the PLA’s procurement department. He was a deputy to Zhang Youxia during Xi’s first term, when Zhang ran that department. That department sat at the nexus of the major corruption scandal in the PLA during Xi’s term, centered on Rocket Force procurement—huge inflows of money and weapons programs. The Pentagon has reported that China is trying to increase its nuclear arsenal from the low hundreds to around 2,500 warheads by 2030.

So even though Xi has been trying to crack down on corruption, there’s been a huge increase in opportunities for corruption during his tenure. What was shocking, though, was how quickly Zhang’s purge happened. In the past, there’s often been a kind of theater: someone disappears from public view, misses key political events, and then there’s endless speculation—inside and outside China—about what happened to them.

This time, Zhang missed a speech at the Central Party School in early January that he should have attended, and just days later came the announcement that he was formally under investigation. That speed—from first indication to official confirmation—was unusually fast, and that was shocking.

Ian:
When very high-ranking military officials are purged, is it ever only about corruption? Or is it ever even mainly about corruption? Or do you always have to look for a political reason?

Neil:
That’s a fantastic question, and one we don’t have a definitive answer to. At this level, there’s always a political dimension. If Xi felt there was a political need to keep Zhang around, it wouldn’t matter if Zhang was corrupt or skimming money on the side.

But Xi is now in a position where almost anyone is expendable—especially in the military, which has been the focus of elite purges in recent years. The fact that he can do this shows how powerful he is in Beijing.

I don’t think it was purely political in my view, triangulating all the evidence I’ve seen. There wasn’t an attempted coup or a major clash over invading Taiwan in 2027, as some of the more entertaining social-media rumors suggest. The PLA was very corrupt before Xi took power—the buying and selling of posts was routine—so I’d be shocked if Zhang Youxia wasn’t implicated in some of that.

What was probably most offensive to Xi is that corruption in procurement and the Rocket Force continued after he took office and after he promoted people he trusted, like Zhang, into positions of responsibility.

Corruption is no longer just a criminal offense; it’s a political sin. Xi is on a mission to reform the Party into a more effective and internally accountable institution—one that can escape what he sees as the historic cycle of dynastic rise and fall, avoid the fate of the Soviet Union, and rule China indefinitely.

That, I would argue, is his number-one domestic political priority.

Chang:
Neil, how do you chronicle the different stages of this purge? You mentioned the Rocket Force in 2023, and now we have this shocking news, which feels like a crescendo—Zhang is basically number two in the military, right? Do you see this as chapters in one book of military purges, or how do you taxonomize these different developments?

Neil:
I think it’s a purge that has gained momentum as it’s expanded through the military leadership.

In 2023, we saw Li Shangfu, the defense minister, purged—most likely over charges related to his time running the procurement bureaucracy. During Xi’s second term, the Central Military Commission announced investigations into procurement abuses covering the five years Li Shangfu led that department. That’s pretty strong circumstantial evidence, especially alongside the removal of Rocket Force generals at the same time.

Then in 2024, we saw the downfall of Miao Hua, the top political commissar in the PLA—the Party’s enforcers inside the military, whose job is to ensure the PLA doesn’t turn against the Party. His fall, and then in 2025 the downfall of He Weidong, the number-two general after Zhang Youxia, are murkier.

Both had served earlier in their careers in the 31st Group Army, headquartered in Fujian, where Xi Jinping spent decades of his early career. It’s possible they were promoted because of personal ties or shared experience. But how they fell—we just don’t really know. There were accusations that they undermined the chairman responsibility system, which is code for Xi’s personal control over the military.

They were also connected to senior generals from the Jiang Zemin era, like Xu Caihou and Guo Boxiong, who famously turned the PLA into a personal fiefdom by buying and selling posts. [These were two generals who were brought down in the first few years after Xi Jinping took power in 2012.]

That brings us to early 2026 and the simultaneous purge of Zhang Youxia and Liu Zhenli. Liu Zhenli’s purge has received less attention, but he was the chief of staff of the Joint Command and the top operational commander. If you add all this up, the Central Military Commission started with seven members in 2022 and is now down to just two: Xi Jinping himself and Jiang Shengmin, the military’s top anti-graft investigator—not coincidentally, I think.

Ian:
I understand what you’re saying about Xi wanting to reform the Party, but to what extent does this point to a deeper problem inherent in dictatorship—that there are always tensions between military commanders and civilian dictators? This was true in Nazi Germany, under Stalin, and elsewhere. To what extent in China is this just inherent to the system?

Neil:
That gets to the heart of the dictator’s dilemma. Political science shows that the biggest threat to dictators usually comes from within—from elite rivals—rather than from mass uprisings.

In authoritarian systems, there’s no independent authority to mediate elite disputes. Ultimately, it’s the threat of violence that determines who can lead and which decisions are made. Xi’s grip on the military and internal security services has been critical to his consolidation of power over the past fourteen years.

China does have distinguishing features. It’s a revolutionary one-party regime, so the military has always been a Party army rather than a national one. There’s also been a separation between military and civilian bureaucracies dating back to Mao. Xi, as Party leader and military commander, is the only real link between the two.

Civilian leaders on the Politburo Standing Committee can’t issue orders to the military. That separation helps coup-proof the regime and solidify the top leader’s authority.

The big question is whether this purge spreads to civilian leadership. The result of this separation between military and civilian government is that Xi’s ties to civilian elites—people like Li Qiang or Cai Qi—are much deeper. They worked together closely for years. Xi never had that kind of relationship with Zhang Youxia.

If the purge stays in the military, we’re probably seeing the culmination of trends that will lead to a full leadership refresh at the 21st Party Congress in late 2027. If it spreads to the Politburo or Standing Committee, that would mark a qualitatively new phase in the anti-corruption campaign—one where no one is safe—and that would be deeply destabilizing.

Ian:
Is it fair to assume that if you’re very high up in the PLA—or even in the Party more broadly—whether you want to or not, you almost have to be corrupt, simply because everybody around you is? And that if you refuse to participate, you’d actually be seen as untrustworthy by your peers?

Neil:
That was certainly the case in the pre-Xi era. You basically had to pay your way through the system—to pass promotion exams, to get appointed, even to very senior positions, including combat positions.

One of the big reasons Xi, in his first term, focused so heavily on revamping the PLA—through both purges and institutional reorganization—was that the United States and other Western intelligence agencies had developed very good contacts, very good sources, inside the PLA and in Beijing more broadly. They were often providing the money that officers needed to pay bribes in order to move up the system. The cadre benefited by getting promoted and becoming more powerful; the U.S. benefited by getting information and intelligence from them.

That entire ecosystem was broken fairly early in the Xi era. And it wasn’t a civilized or procedural crackdown, like something you might see in Western intelligence dramas that I enjoy watching. This was brutal. People were being shot in government compounds for betraying the country.

And I think that if Zhang Youxia or Liu Jianli—these generals—were really betraying the PRC and working with the U.S. or other foreign intelligence agencies, as some of the more extreme rumors suggest, they’d probably be in a much stickier situation than they are now. They’re under investigation, they’ll likely go to prison, but as far as we know, they’re still alive. And I take that as a signal, at least, that some of the crazier rumors probably aren’t true.

The harder question is how much those corrupt practices—buying and selling positions—are still the norm in the PLA during the Xi era. We have less clarity on that. We do know that Li Shangfu was accused, in his official charges, of both accepting and giving large bribes. It’s possible that included bribing Zhang Youxia to be promoted to replace him as head of the Equipment Development Department. We don’t know. It could also refer to corruption earlier in Li Shangfu’s career, when this kind of behavior was simply how everyone got ahead.

Anecdotally, my impression is that both the PLA and China more broadly are less brazenly corrupt than they were in 2012, when Xi took office. It’s incredibly hard to measure. But it’s clear there’s still enough corruption—especially in sensitive areas related to combat readiness—that Xi feels he needs to keep cleaning out the top brass.

Ian:
Buying positions is one form of corruption, but the military has also historically been involved in all kinds of business activity—real estate, commercial ventures, self-enrichment. To what extent has Xi actually been able to deal with that side of the problem?

Neil:
Exactly. And that was another part of the crackdown that Xi launched early in his term, right? We know that those kinds of charges were being leveled at generals and other people in the PLA—the fact that they were leasing military land to developers, that they were diverting funds meant for arms purchases into their own pockets.

So we know that was happening. And we know it’s probably still happening, given the extent of the purges over the last few years. The PLA does still have some business interests, but it’s different from the late 1990s, when Jiang Zemin passed reforms that basically got the PLA out of the broader economy. So it doesn’t quite have the same commercial empire that it used to.

But another telling aspect of the purges in recent years is that it hasn’t just been generals, commanders, and political commissars who’ve been taken down. It’s also been dozens and dozens of executives at state-owned arms manufacturers, particularly those that produce rockets and nuclear weapons. These are state-owned enterprises that aren’t necessarily owned by the PLA, but they’re heavily embedded with the military. The PLA is their main customer—and in some cases their only customer.

The fact that so many of those executives have also been purged suggests that this kind of bribery is still pretty rampant, at least in the Rocket Force. And what’s interesting is that the purges so far haven’t just targeted the Rocket Force, but also the Army and the Navy. The Air Force had escaped a lot of scrutiny, and many Air Force generals were promoted to replace people who were purged elsewhere. But just recently, in the past few weeks, there’s been a new directive issued to investigate graft in the PLA Air Force as well. So it looks like the purge is continuing along these lines.

Chang:
What do you think about the criticism that these anti-corruption purges are just never-ending—that somehow the more you purge, the more corruption seems to come out?

Neil:
As far as Xi is concerned, I don’t think he sees that as a criticism. He sees the anti-corruption campaign as part of a broader effort to strengthen discipline within the Party and to carry out what he calls a self-revolution—ziwo geming—which is his answer to the problem of how to avoid the historic cycle of rising and falling dynasties that I mentioned earlier.

And it’s interesting that on this point, self-revolution is one of the very few—maybe not the only—areas where Xi has placed himself on the same level as Mao. He hasn’t appointed himself chairman of the Party, he hasn’t adopted some of the other trappings of Mao’s rule, but this idea of self-revolution is something he talks about through what’s known as the “cave dialogues.”

In 1945, Mao was in Yan’an and had a conversation with the famous educator Huang Yanpei. Huang asked Mao: how do you plan to carry out this Communist revolution successfully? How are you going to escape the historic cycle of dynastic rise and fall? Mao’s answer at the time was “people’s democracy,” which, clearly, was a bit of a bullshit answer, because it’s not what he ended up doing. But Xi has seized on that dialogue and elevated it within Party ideology.

Xi now says Mao’s answer was the first answer. Self-revolution is the second answer—the answer for a new era—to this fundamental problem the Communist Party faces.

And I think this was very much top of mind for Xi when he came into office in 2012. You had just had the Bo Xilai scandal, which laid bare internal rivalries and infighting within the Party, as well as a real sense of moral crisis—given that a top leader’s wife, Bo Xilai’s wife Gu Kailai, had murdered a British businessman.

There were corruption stories breaking every other week. On social media, you’d see stories about low-level officials—watch bros or LV bros—wearing Rolexes. There was a pervasive sense of moral decay within the Party, and a fear that the failure to rein in corruption, and the failure to deal with visible problems like pollution that were making life unhealthy and difficult in China, was eroding the Party’s ability to stay in power.

Xi had been in Beijing for five years before becoming leader. He was a leader-in-waiting on the Standing Committee and vice chairman of the Central Military Commission from 2010 to 2012. So he saw these problems and political maladies of the Communist system up close.

When he first came to power, I think anti-corruption in his first term was used primarily as a way to take out rival networks and eliminate potential political challengers. In his second term, it slowed down somewhat, partly because of the twin shocks of Trump’s trade war and the COVID pandemic. There simply weren’t as many high-level purges as in the first term.

Now, in Xi’s third term, after the 20th Party Congress—having established dominance over elite politics, removed rival networks, and emerged from those two major external shocks stronger than ever—I think he’s leaning into this broader vision of anti-corruption: remaking the Party, making it more accountable, ensuring that it stays connected to its Communist roots under his leadership and historic mission.

There’s clearly a political loyalty component to this, but also a belief that this is necessary to make China a stronger, more effective country—one that works for its people—and that this is what you need to be a great power.

We can obviously quibble significantly with the means he’s using. But one thing Xi often says about the Soviet Union is that when it collapsed, no one was man enough to stand up; no one was ideologically committed enough to resist Gorbachev. He’s also said the Soviet Party became divorced from the masses and stopped governing effectively. There were empty shelves in Soviet supermarkets in the 1980s—that’s very far from where China is today.

So Xi is trying to revolutionize the Party. I think this approach has many negative effects. It spreads fear. It might not work or last. But I do think his zeal here is sincere. The real question is whether it’s actually going to be effective.

Chang:
I’m wondering what you think about the risks of this kind of constant self-revolution. One impression I’ve had living here is that moments like the White Paper protests showed real societal anger toward the leadership. And that moment really impressed on me the importance of having the rank-and-file police and security forces on your side.

In those situations, when people are out on the streets, the rank and file can be more malleable in their loyalties—they’re far from the leadership and often closer to ordinary people. So it makes sense to me that Xi Jinping is so serious about ensuring the political system is robust and loyal to his orders, especially when society tests him, like during the lockdowns and then during the White Paper protests.

It gave me this sense that if you really achieve what Xi is aiming for—a rigidly loyal security force and military—you might be able to get away with a lot. Society might be able to withstand a lot. It’s like a pressure cooker, and the walls of the pressure cooker are very thick, making it hard to erupt.

Xi is clearly taking this self-revolution idea to a whole new level. Are there risks you see with this approach?

Neil:
Absolutely. It’s created a much more brittle bureaucracy. Anecdotally, there’s a huge degree of fear and uncertainty in the lower levels of the system. There’s a lot of resistance and disagreement with the direction Xi is taking the country, both within the system and in broader society.

Some sectors are doing well—the high-tech industries are world-leading, and there’s a lot of money flowing into them—but other parts of the economy are really not doing well. House prices have been falling for years, and that’s people’s main source of wealth.

So there’s a lot of disagreement in society, and increasingly fewer ways to express it. Civil society, if it’s not dead, is certainly on its knees compared to the pre-Xi era. That does create risks for a leader.

That said, Xi has been very effective at stamping his authority on the coercive apparatus—both the military and the domestic security services, which underwent their own major purges in 2020, 2021, and 2022 ahead of the 20th Party Congress. He’s successfully installed his people in those agencies.

Opposition—both within society and within the Party—faces a formidable collective-action problem. It’s extremely difficult to organize without being detected. Everyone is monitored and surveilled, including senior serving and retired Party leaders. The General Office of the CCP knows where people are and what they’re doing. You can’t just meet a friend on the Standing Committee to chat about economic policy or military purges.

The White Paper protests at the end of the zero-COVID era were remarkable: simultaneous protests against COVID policy—and in some cases against the government more broadly—across multiple cities and social groups. But it’s also remarkable that this is the only time something like this has happened under Xi’s rule.

They were relatively small—probably a few thousand people across a few dozen cities—and they were allowed to happen for 24 to 48 hours. The system may have been momentarily shocked, but fairly quickly the security apparatus kicked into action and, from Xi’s perspective, successfully neutralized the threat. Nothing like it has happened since, despite all the economic problems we’ve discussed.

The presence of formidable surveillance technology makes it even harder than before for either society or elites to organize and exert political pressure on Xi at the top.

Ian:
Can I ask about East Asian security? Until the Trump era, the United States was the main guarantor of security for countries like Japan and South Korea—and to some extent Taiwan. In the Trump era, that guarantee has become very frayed. Japan and South Korea are drawing closer together, and Japan may play a more prominent military role.

China, especially under Xi, has made clear it doesn’t want the U.S. to remain the regional policeman. But what do you think they fear more: the U.S. as regional policeman, or a resurgent Japan with a much stronger military posture?

Neil:
I think China is, at the moment, actually maybe a little more comfortable with the status quo than we might assume. Obviously, Beijing opposes what Takaichi is saying [Her comments about supporting Taiwan in a security crisis], but the United States is increasingly uncertain in its commitment to the Indo-Pacific. It’s focusing more on the Western Hemisphere and the Middle East. So if Japan is relying more on an unreliable partner in the United States, that kind of works as the best of both worlds for Beijing.

There’s obviously a live debate here in D.C., and I think it would be a mistake to conclude that the U.S. wouldn’t become involved if there were a Taiwan invasion or some other contingency. And in the future, there could even be a recommitment to Taiwan and a stronger military focus on the Indo-Pacific, depending on who becomes president.

But ideally, from China’s perspective, the U.S. would withdraw its commitment to the region while Japan continues trying to rely on an America that just isn’t really there. I do think it would be a nightmare scenario if Japan acquired nuclear weapons, for example, which is the direction it might pursue in an extreme scenario of U.S. withdrawal—though, again, I don’t think that’s the most likely outcome. China certainly doesn’t want that, because nukes in Tokyo and Seoul would make China’s strategic environment much harder and would likely precipitate some kind of regional crisis.

Chang:
Neil, last question. What’s your best guess about what happens next?

Neil:
There are a couple of options. To appoint new members to the Central Military Commission, appointments need to be confirmed at a plenum of the Party’s Central Committee, which usually meets once a year. You could convene a special plenum just to refresh the CMC, but I think that’s very unlikely.

Xi has declined to fill vacancies created by previous purges, which is why only two of seven CMC members remain. There will likely be a regularly scheduled Fifth Plenum in October, which could be an opportunity to promote some remaining generals.

You don’t technically have to be a Central Committee member to sit on the CMC, though it’s been the norm for decades. The Central Committee has been hollowed out of PLA representation—there are very few generals left.

The most likely scenario is that Xi waits until the 21st Party Congress next year to fully replenish PLA representation and create an entirely new CMC—one that’s younger, more loyal, more directly tied to his leadership, and less connected to Jiang-era patronage networks. He’ll also likely promote officers more familiar with modern technology and joint warfare.

That’s the schedule going forward, as best as I can tell.

Chang:
If that’s right, and Xi plans to replenish the military at the 21st Party Congress, does that mean there’s much less chance of something happening over Taiwan before then?

Neil:
I think it’s unlikely. Removing top commanders—especially figures like Zhang Youxia, who had combat experience, and Liu Zhenli, who ran joint operations—weakens near-term readiness.

Joint commands are critical for any Taiwan scenario, coordinating the army, navy, air force, Rocket Force, and other units. The U.S. military is very good at this; China is still early in building a truly joint force.

China hasn’t fought a war since the Sino-Vietnamese conflict in the late 1970s and early 1980s. A Taiwan operation would be vastly more complex—an enormous amphibious assault, arguably more complicated than D-Day in 1944.

So purging the top brass makes a near-term escalation less likely. Longer-term, if Xi succeeds in building a more competent and loyal leadership, the PLA could become a more effective tool for coercion and deterrence. But rebuilding leadership, experience, and morale takes years.

Even after the next Party Congress, it wouldn’t be immediate. From Xi’s perspective, it may make more sense to wait until conditions look more favorable—Taiwanese politics are shifting, the DPP president is becoming less popular, polls show slightly less enthusiasm for independence, and the West is distracted by other crises.

Let those trends continue, and it may be less costly to try to compel Taiwan to negotiate five or ten years down the line.

Chang:
So time is on Xi Jinping’s side?

Neil:
From his track record, domestic political considerations are his number-one priority. Shoring up political control and cracking down on corruption in the military matter more to him than Taiwan in the near term.

We often assume Taiwan explains everything because it’s what we focus on most. It’s extremely serious, but it doesn’t always explain what’s happening inside China. Domestic political considerations are paramount.

Ian:
Thank you very much.

Neil:
Thanks so much.

Discussion about this episode

User's avatar