
Who Killed the Chinese Economy? | A Discussion with Adam S. Posen, Zongyuan Zoe Liu, Michael Pettis
What this covers
This Foreign Affairs and Peterson Institute event brings together Adam Posen, Zoe Lou and Michael Pettis, with a moderator, to argue over why China's recovery after reopening faded so quickly. It grows out of Posen's Foreign Affairs essay and the published responses to it. The panelists agree that growth is slowing and disagree about when the trouble began and who is responsible. Posen stresses how little rebound China got after zero-COVID, and he reads household hoarding of cash as a response to arbitrary state intervention. Lou dates the turn to Xi Jinping's centralizing agenda from 2013. Pettis treats the problem as the exhaustion of an investment-heavy development model, which he places well before Xi.
The middle of the discussion turns on whether demand or supply is the binding constraint, since that decides whether stimulus helps. Pettis argues that monetary expansion in China mainly stimulates supply and so worsens imbalances. Posen counters that the two models make opposite predictions about the stimulus multiplier. Pettis draws on Kornai's soft-budget-constraint framework and on research by Daron Acemoglu and James Robinson about which political systems can manage model shifts. He contrasts crisis-driven "American style" adjustment with the slow Japanese path, and Posen replies that nothing about China's debt accumulation was inevitable.
The panel also covers several other topics: - Lou's demographic and housing points: real estate is roughly 30% of the economy, family formation is falling, and about 42 million people will be past retirement age by 2040. She also notes that aging creates new demand. - Her account of civil-military fusion and Made in China 2025 as accelerants of Western export controls and investment screening. - Pettis's forecast that a shift from property to manufacturing would raise China's share of global manufacturing, which other large economies would resist. - Posen's claim that zero-COVID broke the no politics no problem compact. - On US policy, Posen's argument for "suction" (making emigration attractive) over sanctions, and Pettis's warning that Washington has misjudged rising rivals before.
China's economic slowdown stems from a combination of structural factors (high-investment model exhaustion) and policy choices (Xi's recentralization and zero-COVID overreach), with interpretations of causality and timing determining whether stimulus can help or only masks deeper problems requiring political credibility restoration.
- Three expert panelists (Posen, Lou, Pettis) offer competing diagnoses: economic long-COVID from state intrusion (Posen), accelerated fuse-shortening from decoupling policies (Lou), and structural model obsolescence since 2006 (Pettis)
- Disagreement on remedies hinges on whether the constraint is demand (households scared to spend) or supply (investment overextension), with profound implications for whether fiscal stimulus will work
- US policy response should focus on attracting exit/outflows rather than sanctions, while avoiding pushing an isolated China into confrontation
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
The real estate sector represents about 30% of the Chinese economy, and as family formation rates decline and Chinese women become less willing to marry or have children, this will permanently drag down demand in the housing sector, a major long-term headwind.
“the final part in terms of why demographic trending in the long run bod bod not good for the Chinese economy is really about the real estate sector real estate sector is about 30% of the economy and as a family formation rate decline as uh Chinese women not necessarily willing to get married or Buri children um you know basically that is going to permanently drag down demand in the housing sector”
Chinese households shifted dramatically toward hoarding liquidity and cash-like assets, moving away from durable goods and auto purchases, and small businesses owned by households did the same, driven by fear of expropriation under arbitrary state intervention.
“you see a huge shift in Chinese household Behavior they stop buying durable goods Autos expensive things small businesses which are essentially owned by households do the same thing they start putting more and more money into bank accounts into cash-like liquid substances or forms of assets and in general savings go up but importantly they're moving towards liquidity and I interpret this as a reaction to the zero covid policy”
The stunning fact about China's economic slowdown is how little bounce China got after reopening from zero-COVID, unlike most economies including Japan and other high-income Asian economies that saw big consumer rebound, and this rebound barely existed in China and then fizzled out.
“in my view the stunning fact about China's economic slowdown is how little bounce China got after reopening in zero Co and most economies including Japan including other neighbors in Asia but all high income economies saw some degree of big consumer rebound right after they reopened in China that rebound barely existed and then fizzled out”
Household consumption as a share of China's GDP bottomed out around 2010-2011 at a 'truly surreal' 34% of GDP, began partial recovery afterward, but that recovery stopped a couple of years before COVID, and during COVID the consumption share dropped again—this distribution of income to the household sector is what drives the expansion and contraction of the private sector.
“from uh uh domestic consumption bottomed out around 2010 2011 with household consumption at a a truly surreal 40 34% of GDP and after that we started to see a partial recovery of consumption and the private sector benefited from that but that more or less stopped a couple of years before Co and during Co we saw again the consumption share of GDP go down”
János Kornai's framework: economies attempting to maintain high growth rates must shift activity from hard budget constraint sectors (like the private sector, which goes bankrupt if unprofitable) to soft budget constraint sectors (like local governments and property developers, which can accumulate indefinite losses without bankruptcy) in order to fund non-productive investment.
“corai argued that in these types of economies um you have a shift from the hard budget constraint part of the economy to the soft budget as long as you want high growth rates because a hard budget can strain a part of the economy for example the private sector won't systematically engage in nonproductive Behavior if they do they go bankrupt so if you want to maintain higher growth rates than the real economy can deliver then you have to increasingly shift activity to the the soft budget constraint which in China was largely local governments in the property sector”
The US has a history of economic paranoia about rising competitors (Japan, Soviet Union) and has often gotten threat assessments wrong; concerns about China as an economic threat should be evaluated carefully rather than assumed.
“I don't think China is the economic threat that we make it out to be I think Americans have a history of paranoia and maybe it's a strength of the US but you know we were terrified of Japan terrified of the Soviet Union it's a it's a economically I'm not talking about geopolitically um it's a it's a common it's a common Trend and we tend to get it wrong for the same reasons”
The interpretation of historical facts about China's economy shapes understanding of current and future policy; how one dates the critical juncture (2006, 2012, 2013, 2015, or 2020) and attributes causality (structural vs. agency-driven) determines what policies seem viable or futile going forward.
“a real understanding of how the way in which you interpret that history can shape the way you understand policy today and policy going forward that's really what what matters and where the rubber hits the road”
Countries that undergo the investment-driven growth model transition experience adjustment periods that follow one of two main patterns: the 'American style' (Brazil also followed this) involving a crisis-driven rebalancing where GDP contracts sharply (US 1930s: -35%), or the Japanese style where GDP growth slows to very low levels (1-2%) but adjustment takes much longer.
“the adjustment can range uh you know I talk about the American style adjustment which Brazil followed too which is a rebalancing in the form of a crisis so the United States in the 1930s you'll remember GDP contracted by 35% household income contracted by half of that brutally painful but it rebalanced and it was quite quick in Japan what we saw is that GDP growth went from roughly 4 or 5% to around half a percent consumption growth dropped by a lot less I don't have the per capita numbers but the per capita numbers look like they dropped from around 3% to 2%”
Successful development models make themselves obsolete by resolving the problems they were designed to address, and when they do, a shift to a new model is necessary, but almost no country successfully makes this shift because the successful model creates powerful institutions that disproportionately benefit certain constituencies who resist change.
“a successful development model by definition makes it self obsolete because it resolves the problems it was designed to address and then you need to shift the model and if you don't shift the model you end up de developing a different set of imbalances and he also noted that almost no one shifts the model and the reason he argued is because a successful development model creates a series of Institutions that disproportionately better benefit certain constituencies that become very powerful and it's very hard to shift the model because of the uh blocking efforts by those constituencies”
Daron Acemoglu and James Robinson's research shows that model-shift adjustments are politically disruptive and destabilizing except in two types of political systems: robust democracies and highly centralized autocracies (as exemplified by China in the 1980s, which implemented ferociously-opposed reforms).
“what's really striking is how politically dis disruptive it's always been except for two types of uh of of of political systems they argue highly competitive political systems which seems to mean robust democracies and then highly non-competitive which basically seems to mean a highly centralized autocracy by the way an example of the latter was China in the 1980s uh a series of reforms ferociously opposed by the party but with a highly centralized leadership they were able to implement the uh the reforms anyway”
The zero-COVID policy broke the 'no politics no problem' compact that had existed in China since Deng, where ordinary citizens could conduct commerce without state interference as long as they avoided explicit political activity, and this breakdown in trust is very difficult for an autocracy to repair.
“there was what I called a no politics no problem compact in China this to me was a big part of the actual Chinese miracle that from dung until she takes power in 2013 if you weren't a democracy protester if you weren't an oppressed Muslim minority or some other ethnic minority basically you could go about your commercial life you might have to pay a bribe you might not always get the contract you wanted but basically you were left alone as long as you did know politics starting with she's consolidating power in 2015 but especially with their zero coid Crackdown average chines suddenly were subjected to your company is closed you can't go to your job you've lost your business and it's arbitrary”
Aging populations create new demands (healthcare, elder care, specialized products) that drive new industries and spending, providing a positive offset to the negative effects of demographic decline on consumption.
“the demographics have both um positive effect and a negative effect because it's not that demographic trend or the Aging population is is going to Doom the Chinese economy because age as as population becomes as aging population also create new demand and such as you know uh older health care or you know like you know adult diet person it drives new industry right”
Civil-military fusion is a strategy (not new, originating in the Hu Jintao era) to empower the state militarily and technologically by blurring distinctions between military and civilian sectors, mobilizing whole-of-society and whole-of-nation technological advancement and self-sufficiency efforts.
“so so this did not this so-called strategy did not start from shining it's the whole the whole civil military Fusion started in during huin era the idea is to empower OB business physically mobilized Chinese the a whole whole of a society whole of a Nation um element blurring the Str the nature of military as well as and and the private sector and the civilian sector in terms of technological advancement and technological self sufficiency”
Xi Jinping did not assemble China's economic time bomb, but he should be blamed for having shortened the fuse through his policies accelerating decoupling and failing to mitigate long-standing structural problems (debt, demographics, demand, decoupling) that were already evident.
“president xinping he did not assemble China's economic Time Bomb but I do think he should be blamed for having shortened the fuse”
If China shifts investment from property to manufacturing, continuing to grow at 4% annually, China's manufacturing share of global manufacturing will rise from 30% to 37% while its GDP share rises from 18% to 21%, forcing other economies to shrink their manufacturing share—a scenario the US, Europe, India, and other large economies do not want.
“if if China adjusts or as it attempts to adjust tries to shift lending away from the property sector and into the manufacturing sector which we're already seeing right the the decline in property investment has been more than matched by an increase in industrial investment that creates a real problem for the world because if you assume that China continues to grow maybe at 4% um in in about a decade China's share of global manufacturing will rise from 30% to 37% while its share of GDP rises from 18 to say 21% that means the rest of the world in order to accommodate China has got to reduce its own manufacturing share”
The critical juncture for China's economic problems occurred several years before COVID-19, beginning in 2013 when Xi Jinping articulated his vision of 'reform and opening up' emphasizing the state sector's role and the party's leadership in economics and finance, signaling a shift toward centralization.
“the critical juncture happened several years before um zero Co the coid 19 as well as zero coid because if you remember back in 2013 president Shin already put out his own personal explanation his own explanation to what does he mean by what what what what he thinks reform and deepening uh re deepening reform and opening up means and in in writing and in his speech he's specifically talking about you know the for for him reform and open up really is about the emphasis of the state Spector the RO the leadership of the party in um you know economics and finance”
Authoritarianism can function for extended periods, but it is very difficult and increasingly unlikely to sustain authoritarianism without excessive, arbitrary intrusions into citizens' lives; the key problem is not control itself but the capricious and arbitrary nature of authoritarian policy that triggers households to self-insure and withdraw.
“it's about the arbitrary nature the capriciousness of authoritarian policies and that's what causes the immune reaction of people trying to self-insure”
Monetary stimulus in China does not work like in the US; it tends to stimulate the supply side of the economy, so when China expands monetary supply, rather than rebalancing imbalances, the imbalances get worse.
“monetary the problem with monetary stimulus in China doesn't work like in the US monetary stimulus tends to stimulate the supply side of the economy so every time we've fiscal expansion or monetary expansion in China rather than rebalance the imbalances have gotten worse”
US policy should pursue 'suction' (making emigration attractive) rather than 'sanctions' (confrontational pressure) as a way to constrain Chinese Communist Party behavior, as this is more constructive, sustainable with allies, and aligns with US economic interests.
“you want to think in terms of suctions not sanction that the way to put pressure on the regime is to make it so that it's attractive for people to leave and then that creates a dynamic we've seen with other autocrats in the economic sphere that they try to put up more controls against Capital flight against outward migration against outward investment and those controls induce more desire to leave”
Household behavior did not shift dramatically until zero-COVID despite the structural problems that Pettis identifies as pre-dating Xi, and starting in 2015 Xi reversed the growth in private sector dynamism, suggesting that agency and choice matter more than structural inevitability.
“and then starting in 2015 and accelerating the she turns it around but then again the households don't react until we have the zero coid in the way that I identify and so to me while the structuralist problems are there it both exaggerates the determinism coming out of the structural problems and also frankly their magnitude all economies slow down all economies accumulate over time there was nothing inevitable that you couldn't have put more money in the private sector and gotten higher returns than you did the last several years anyway”
China's elevation of civil-military fusion to a national strategy and combined emphasis on 'Made in China 2025' accelerated pre-existing Western mechanisms like export controls and investment screening, triggering a US policy response that removed the predictability and lack of political drama that used to benefit Chinese economic growth.
“president Shin really elevated that into as to the level of national National strategy and combining that with uh you know the made in made in China 20 20 uh 2050 so a lot of this really triggered um alerted uh Western politicians and in particular here US policy makers in Washington DC and in response the United US policy makers uh accelerated a lot of the pre-existing mechanisms such as export control such as investment screening”
A key difference between the structural and psychological models is that stimulus efficiency depends on whether the constraint is supply-side (structural underinvestment that stimulus worsens) or demand-side (household confidence/spending willingness), and the models make opposite predictions about stimulus multipliers.
“and that's the opposite of Japan by the way which my work and others Seely confirmed that in Japan they didn't do any of that public investment this is one of the big differences that gets all lighted over in the structurals view is that in Japan the debt was in the corporate sector not the household sector and in Japan the public investment was not there whereas the private investment was and was what went bust so the upshot of all this is that it's not enough whereas in Japan it was enough to take on a combination of monetary stimulus and debt Cleanup”
Not all economies that follow high-investment growth models are doomed to the same adjustment path; there was nothing inevitable about China accumulating unsustainable debt or failing to rebalance—policy choices mattered enormously.
“all economies slow down all economies accumulate over time there was nothing inevitable that you couldn't have put more money in the private sector and gotten higher returns than you did the last several years anyway just to say I think Mikel and zo are being perfectly Fair contrasting views but just that would be my point of difference I I don't think their model fits the the developments that I saw right”
China's demographics, particularly the one-child policy's effects, will result in about 42 million people above the legal retirement age of 60 by 2040, which will exceed the entire current US population, creating unprecedented health care costs and reducing the demographic dividend that supported growth.
“by 2040 the uh China is going to have about 42 million people who are above the age of 60 which is the current legal retirement age and at that time um you know the United States population is going to be around the entirety of US population is going to be around 390 million the other word within the next um within the next 20 20 years China is going to the older like retired Chinese people is going to the population is going to be higher than that of the United States”
China already has among the highest household debt as a share of GDP in the world, higher than in the US, so expanding household debt as a stimulus mechanism is not sustainable and cannot continue as a growth driver.
“that's not sustainable China already has the highest household debt as a share of GDP among the highest in the world higher than in the US”
Recent data on yuan movements and reports of Chinese government crackdowns on capital outflows are consistent with and support the prediction of increased exit behavior in response to lost confidence in the regime.
“and all the data of the last several months the movements in the Yuan the stories about China cracking down on outflows of Capital fit”
Economic long-COVID (the nonresponse of consumers to economic incentives because they are hoarding liquidity out of fear) occurred well before the breakdown of the real estate market and is independent of it, and the behavior of moving into savings and out of durable goods is not what a real estate crash would predict.
“this Behavior this this zero long long Co economic long Co meaning the nonresponse of consumers to incentives because they're hoarding liquidity they're trying not to get expropriated they're basically scared this occurred well before the breakdown of the real estate market the real estate market is on top of that doesn't help but it occurred and is independent of that and in fact the behavior of moving into savings and out of durable goods is not necessarily what the real estate market crash would do”
China's high-investment, high-savings growth model made sense in the 1980s and 1990s when China was severely underinvested, but by the early 2000s China had closed the gap between investment and productive capacity, at which point the model should have shifted but did not, leading to unproductive investment and debt acceleration.
“what I would argue is that this High investment model made a lot of sense in the in the 1980s and 1990s when China was possibly the most underinvested economy in the world um but with the fastest growth rate of investment they closed the Gap pretty quickly between what they had and what they could productively absorb now at that point you're supposed to shift to a different model but no one ever does not Brazil in the 1970s not Japan in the 1980s not the Soviet Union in the 1960s and that's when you end up with very high investment rates that are no longer productive”
It is not the recentralization of the economy that caused the slowdown; rather, it was the slowdown that caused the recentralization—the causal arrow runs the opposite direction from what conventional accounts suggest.
“I would say it's not that the recentralization of the economy caus The Slowdown I would say it was the Slowdown that caused the recentralization of the economy”
An isolated China with 1.4 billion people and the world's second-largest economy, whose population becomes angry, particularly at the United States, represents a very dangerous scenario that US policymakers should avoid creating.
“a country as big as China the second largest world economy with 4 point 1.4 billion people if an an isolated China with that big of econom and that large number of population who become angry and in particular angry out of the United States that is going to be very dangerous”
Posen's structural theory overstates determinism and exaggerates the inevitability of slowdown, ignoring the period from 2000-2015 when there was enormous growth in private sector employment, returns on private assets, and private investment, as documented by economists like Nick Lardy.
“I would make two points I think first the structural the the heavy duty structurals version which Michael argues for is too dismissive of the agency of both the party and the average Chinese household if you look at the period even including the period Michael talks about there's an enormous growth in private sector employment in Returns on private sector Assets in in private sector investment as my colleague Nick lardy H documented in markets so for Ma and then starting in 2015 and accelerating the she turns it around”
External factors including perceived Western threats and security ambitions may have influenced Xi's policy direction, but the primary driver is control—Xi's desire to concentrate power and reassert party dominance over society and economy.
“a second factor is you can argue how much instead of what something none of the three of us have mentioned yet the external environment how much although Zoe makes reference to the the remilitarization needs um to be fair um so how much this is a reaction to perceive threats or external Ambitions and that pushes you in this direction but I think it's ultimately mostly about control um and that's why again I tried to speak of the the long coid it's the immune resem response as you kindly invoked it it's it's not because the Communist party ever was not in control it was that the communist party self-control broke down”
In a Japanese-style adjustment, China's savings rate falls faster than its investment rate, causing the current account surplus to contract relative to global GDP, which adds demand to the world economy and benefits net importers (especially importers of consumer goods and agricultural commodities).
“if China has a Japanese style adjustment where you see consumption growths sort of stable maybe down a bit and much of the adjustment incurring with a reduction in investment what you're likely to see is the savings rate in China go down faster than the investment rate so that the current account deficit as a share of global GDP will contract which is exactly what happened with Japan in the 1990s in which case China's adjustment adds a bit of global demand to the world”
The middle-income trap terminology is problematic because it is often used as a catch-all label meaning 'we don't know,' obscuring more precise structural explanations; the investment trap (which affects all high-investment-growth-model economies from poor to rich, not just middle-income countries) is a better framework.
“to address the middle-income Trap it's not a phrase that I really like because it seems to me that we often Trot it out to mean we don't know so it must be the middle- income trap um but um it in terms of in terms of the you know I would call it an investment trap be because it affected all countries that had very very high investment growth models uh Japan which is certainly not a middle-income country it's quite a rich country also went through this trap poorer countries like Brazil in the 1960s and70s”
There is a growing consensus among economic policy advisors that China needs to boost domestic consumption by increasing the household share of GDP through wages, transfers, and social safety net expansion, but the hard part that hasn't been discussed is that if you increase one group's share, you must decrease another's—the question of who bears the adjustment cost (Beijing vs local governments) is the contentious issue ahead.
“there it's not very surprising in fact there is a a a consensus it's not Universal but there is a consensus among Economic Policy advisors and economic policy makers about what China needs to do and very specifically uh they need to boost domestic demand domestic consumption”
Zoe Lou argues that US policy should reassess the assumption that China will inevitably surpass the US economically, challenge the US ideologically, and confront the US militarily, given evidence that Chinese people have agency and growing incentive misalignment with the party.
“we wouldn't re-evaluate the premise for American the China policy because our the premise of our China strategy has been that China will surpass the United States economically challenge the United States ideologically and confront the United States militarily however in our China strategy it seems that we have been we we haven't paid enough attention to the Chinese people”
In the short term, China's government should distribute resources to stimulate household consumption to both address household confidence issues and provide short-term relief, while in the international sphere, Xi Jinping should demonstrate commitment to stabilizing US-China relations.
“I think perhaps immediate in the immediate term the government really need to figure out how to distribute to stimulate household consumption I'm not exactly sure whether this is going to be in the format of direct you know stimulus check or um or or even household debt ease things like that but direct household household support is necessary both to solve uh to on the one hand to address the lack of household confidence issue then on the other hand at least to give people some sense of a shortterm relief right”
Before COVID, the Chinese economy looked like a Monet painting (beautiful from afar, jumbled mess up close), but after COVID it looks like a Jackson Pollock (jumbled mess from afar and up close), indicating the problems became dramatically more visible and pervasive.
“before covid or before 2019 when we think about the Chinese economy it looks like a Monet it's like an impressionist painting it looks beautiful from afar but it looks like a jumbo mess up close but after covid the problem still exist but it's now like a Jackson Pollock in the sense that it looks like a jumbo mess from afar and a jumbo mess up close”