When Strategic Ambition Hollows Out the Foundation

China’s economy has long defied the conventional prescription that sustainable growth requires a rising middle class and robust domestic consumption. For decades, Western economists pointed to China’s lopsided reliance on investment and exports as a structural flaw awaiting correction. Yet successive Chinese leaderships have not corrected this imbalance—they have deepened it. This is not an accident of policy failure. It is, in important respects, a deliberate design.

Understanding China’s economic weakness requires understanding why that weakness exists and what it is meant to serve. The country’s economic model is increasingly organised around a single overriding objective: to win the strategic competition with the United States and establish itself as the dominant global power. That ambition requires overwhelming manufacturing capacity, control over critical upstream resources and raw materials—from rare earths to critical raw materials—and the progressive elimination of technological dependence on the US. Achieving these goals demands sustained, large-scale investment in innovation and high-end manufacturing. To this end, China needs a huge pool of capital, but it does not count with a global reserve currency, like the United States to attract it. China’s fall-back solution can only be its large pool of cheap savings, which only a closed capital account and a financially repressed household sector can reliably provide.

The result is a dual economy: one side brilliant and purposeful, pushing China to the frontier of electric vehicles, artificial intelligence, solar technology and advanced semiconductors; the other side stagnant, squeezed and increasingly hollowed out. On one side stand China’s global corporate champions keep on winning market share across the world thanks to China’s industrial upgrade. On the other side sits the Chinese households, whose disposable income is no longer growing and whose wealth has been devastated by the collapse of the property market.

This paper argues that China’s dual economy is not merely a transitional imbalance, with excessive exports waiting for consumption to take the baton, as stated in this year’s Work Report[1] or the 2025-30 Five Year Plan.[2]   In reality, the current duality of the Chinese economy looks much more like a goal of its own rather than a means to an end. In fact, it is the consequence of a set of interlocking political and strategic choices which go beyond growth and enter directly into national security and China’s hegemonic goals. Steering the economy into a different direction, given those overriding objectives, is like squaring the circle.

Even if subservient to China’s ultimate objectives and therefore needed, this dual economic model is less sustainable than Chinese authorities seem to think.  The cracks in this economic model can be observed notwithstanding China’s technological and export prowess.  These cracks go beyond external risks, such as Trump’s protectionism  or even wars, which generally take the central state in the official discourse.[3] In fact, the most dangerous cracks are purely domestic and are expanding, stemming from  an economic model that fosters volume rather than profitability, investment rather than wages, basically a volume rather than value concept. This model, which is incapsulated in the concept of 内卷 (involution) or more narrowly 内卷式竞争 (involutionary-style competition) as repeatedly mentioned by become Premier Li Qiang,[4] is China’s most important economic challenge nowadays. Involution, as we will explore later, is not only about companies’ irrational competition. Leading to a reduction in profitability but is also at the core of an increasingly poor lab or market with decelerating wages. This, which is further fostered by the Chinese government’s bold AI adoption policies, is going to push further towards stagnating wages and displacement of labour, which can only further weaken consumption. In other words, China’s dual economic model does not have a self-correction mechanism. On the contrary, it is built on the vicious circle of involution, further fostered by rapid AI adoption.

The Architecture of the Dual Economy

China’s economic model can be characterised, at the broadest level, as a production-first, consumption-later strategy that has never quite arrived at the “later.” The country’s National Bureau of Statistics emphasises the production-output method in its GDP releases, making industrial value-added the central pillar of measured growth.

This statistical architecture reflects a deeper institutional reality: the Chinese state has consistently channelled resources—through directed credit, subsidised land and energy, government procurement, and strategic industrial policy—toward the supply side of the economy. The 14th Five-Year Plan (2021–2025) codified this orientation, setting explicit targets for high-technology industries, self-reliance in critical sectors and expanding manufacturing capacity.

The 15th Five-Year Plan (2026–2030) continues in the same vein, doubling down on what Beijing calls the “new quality productive forces”—a shorthand for AI-driven, high-end manufacturing.

The strategy has produced remarkable results on the supply side. China now accounts for approximately 18 percent of global manufacturing exports. In green technology, its dominance is even more striking: Chinese companies account for 90 percent of total global capital expenditure in solar panel manufacturing. In electric vehicles, BYD has overtaken Tesla in global sales volumes. In battery technology, CATL is the undisputed global leader. In artificial intelligence, the rapid emergence of DeepSeek has demonstrated that China’s capacity for commercial and industrial application of AI is among the best in the world, even as it operates under significant chip restrictions imposed by the United States.

Yet behind these extraordinary achievements lies a structural problem that has become increasingly difficult to ignore. The same model that has built China’s industrial supremacy has systematically suppressed the domestic demand that would allow its industries to operate profitably at home. Fixed-asset investment in manufacturing has grown far faster than GDP since 2021, while real estate investment has collapsed. The result is a persistent and widening gap between supply capacity and domestic absorption. In some sectors—electric vehicles, solar panels, steel—overcapacity is so severe that firms are selling at or below production cost simply to preserve market share (Graph 1).

The Strategic Logic: Why the Imbalance Is Intentional

To understand why this model persists, one must take seriously the strategic objectives it is designed to serve. China’s leadership has been explicit, across multiple planning cycles and policy statements, about its ambition to achieve technological self-reliance and eventually global pre-eminence in the industries of the future. The Made in China 2025 strategy, launched in 2015, targeted ten strategic industries—including aerospace, robotics, semiconductors and new energy vehicles—for state-backed development.

Subsequent plans have extended and intensified this ambition, adding artificial intelligence, quantum computing and biotechnology to the list of priority sectors. This strategy is not merely about economic growth. It is about strategic competition. Control over the technologies and supply chains that will define the 21st-century economy—from battery minerals to advanced chips to AI platforms—is understood in Beijing as a prerequisite for geopolitical dominance.

Executing this strategy requires sustained and massive investment in innovation and manufacturing capacity. That investment must be funded, and it must be funded cheaply. This is where the financial repression of the Chinese household becomes not an unfortunate side effect of the model but a structural necessity. By maintaining a closed capital account and a state-dominated financial system, China ensures that a large proportion of household savings—which remain the world’s highest as a share of GDP at around 44 percent (Graph 2)— flows into the banking system at low interest rates, which can then be directed toward state-prioritised investment. The repression of financial returns on household savings is, in effect, a subsidy from Chinese households to China’s industry as well as a structural repression on consumption (Graph 3).

This mechanism is further reinforced by the underdevelopment of the welfare state. Because Chinese households face minimal public provision for healthcare, education, and retirement, they save precautionary amounts far more than what a comparable economy with a functioning social safety net would generate (Graph 3). President Xi Jinping has expressed explicit scepticism about what he calls “welfarism,” signalling that the government has no intention of building the kind of comprehensive social insurance system that would allow households to save less and spend more.

The maintenance of high precautionary savings is thus also a deliberate policy choice, not merely an institutional inertia. The closed capital account compounds this effect. By preventing Chinese households from diversifying their savings into foreign assets, the state ensures that domestic savings remain captive to the domestic financial system. This is inefficient from the perspective of optimal asset allocation—Chinese savers are denied access to global investment opportunities—but it is highly efficient from the perspective of the state’s investment agenda. The inefficiency is, in this sense, intentional.

Involution: When the Model Turns Against Itself

The dual economy model, for all its strategic coherence, contains within it the seeds of its own undoing. The mechanism by which it is beginning to fail is what Chinese commentators and, increasingly, Western analysts have come to call “involution”—a term borrowed from anthropology to describe a process of intensifying effort without commensurate gain.

The economic manifestation of involution in China is straightforward. As domestic demand has remained suppressed and global markets are becoming saturated with Chinese exports, or are lifting barriers like the US, the only way for Chinese firms to maintain or grow revenues is to capture market share from each other. This has triggered a race to the bottom on pricing across every major industry. Firms slash prices to preserve revenue, which compresses margins, which reduces profits, which forces further cost-cutting. The result is an economy in which companies are growing their revenues while simultaneously destroying their profitability.[5]

The consequence is a growing army of “zombie firms”—companies that cannot generate enough revenue to cover even their interest payments. As of 2024, the share of Chinese listed companies meeting this definition exceeded 12 percent (Graph 4), more than double the global average.[6] In green technology sectors—the very industries that Beijing has prioritised as the engines of its future—the zombie share approaches one third (Graph 5).[7] These firms survive not because markets support them, but because local governments prop them up through subsidies, preferential procurement contracts, and bank credit extensions. Officials do this rationally, because their career trajectories depend on local employment figures and on the appearance of industrial vitality in their jurisdictions.

The survival of unproductive firms has a deeply corrosive effect on the productive ones. China’s global champions are not losing market share to foreign rivals. They are losing pricing power to domestic peers that should, by any market logic, have already exited the market. These leading firms are caught in a paradox: they are among the world’s most competitive enterprises, yet they are forced to sacrifice profit margins in order to maintain domestic market share against rivals that survive only through subsidy. The capital that should be funding the next generation of innovation is instead being consumed in a war of market attrition.

The Labor Market Dimension: Wages, Wealth and Hollowed-Out Demand

The involution of Chinese industry has a second-order consequence that is at least as damaging as the direct compression of corporate profits: it has devastated the wage growth of Chinese workers.

When firms cannot make money, they cut costs. In China, the single most flexible cost line is labour. Unlike Western economies, where trade union strength or minimum wage legislation creates some degree of wage rigidity, China’s labour market is remarkably flexible in the downward direction. Firms facing margin pressure can and do reduce headcount, limit hiring, cut bonuses and constrain base wage growth. The accelerating development of automation and artificial intelligence has made this even easier: the substitution of labour with automated processes is advancing rapidly, further eroding the bargaining power of Chinese workers.

The data are unambiguous. Average wage growth for urban private-sector workers collapsed from above 8 percent in the pre-COVID years of 2018–2019 to approximately 4 percent in 2022–2023, and then to just 1.7 percent in 2024 (Graph 6). For 2025, based on the trajectory of migrant worker wages—a leading indicator that dropped from 4.7 to 2.7 percent—private sector urban wage growth is estimated to have fallen below 1 percent in nominal terms.[8]

Even accounting for near-zero consumer price inflation, the real wage deceleration is significant and sustained.

Youth unemployment has become a particularly acute symptom of this labour market deterioration. The youth unemployment rate, excluding students, rose from 14.7 percent in 2023 to 16.7 percent by end-2025 (Graph 7). This has fuelled a cultural moment of economic pessimism: the “lying flat” (躺平) and “let it rot” (摆烂) sentiments that circulate widely on Chinese social media are not merely generational affectations. They reflect a rational response to a labour market that offers diminishing returns to effort, in an economy where the standard pathways of education, urban employment and property ownership no longer reliably lead to upward mobility. The connection between wage stagnation and weak consumption is obvious (Graph 8). When wages decelerate, consumption follows as can be seen with the increasingly poor performance on China’s retail sales.

The wage story is compounded by an equally severe wealth effect operating through the property market (Graph 9). For a generation of Chinese households, residential real estate was not merely a home: it was the primary vehicle for wealth accumulation, typically representing 60–70 percent of total household assets. The bursting of China’s real estate bubble, which began in earnest with the Evergrande default in late 2021, has inflicted enormous losses on household balance sheets. House prices and real estate transaction volumes have been falling for years across virtually all tier categories. The negative wealth effect from falling property prices is amplifying the direct income effect of wage stagnation, creating a dual squeeze on household purchasing power that leaves little room for the consumption recovery that China’s model requires.

The lack of fiscal and monetary space

Three conventional policy responses could in principle break this deflationary spiral: reducing real interest rates to stimulate private demand, expanding fiscal spending to support aggregate demand, and restructuring corporate debt to free up balance sheets. All three are effectively blocked or severely constrained.

China’s monetary policy has maintained a cautious, broadly neutral stance. The People’s Bank of China (PBoC) has been reluctant to pursue aggressive easing, partly because of concern about the renminbi exchange rate—further easing risks widening the interest rate differential with the US and accelerating capital outflows—and partly because Chinese banks, already suffering from very thin net interest margins, would struggle to bear still lower lending rates.

On the fiscal side, the constraints are equally binding. The official budget deficit was expanded to 4 percent of GDP for 2025, the largest on record, but much of this expanded borrowing was directed at local government debt restructuring rather than direct consumption support. Furthermore, the official budget is only a part of China’s rapidly deteriorating fiscal position, especially when it comes to local governments (Graph 10). Their fiscal revenues have been sharply squeezed by the collapse in land sales and by weakening corporate profits that reduce tax receipts. The mounting debt burden of local governments, including the vast liabilities accumulated through local government financing vehicles (LGFVs), severely constrains the scope for additional productive fiscal spending. The government’s industrial policy has been explicitly concentrated on high-technology sectors of strategic importance, offering little relief to the heavily indebted firms in real estate or traditional manufacturing that are most in need of debt restructuring. The latest Two Sessions continued a tone of caution, with no large-scale demand-side stimulus announced.[9]

Technology as a Saviour? Not really

Faced with the structural constraints on fiscal and monetary policy, and with the deepening cracks in the involutionary model, China is placing enormous strategic weight on artificial intelligence and automation as the solution to its productivity problem. The logic is appealing: if AI can dramatically raise the productive efficiency of Chinese firms, it could restore profit margins, fund wage growth, and break the deflationary spiral—all while simultaneously advancing China’s strategic objective of technological self-reliance.

The scale of China’s AI investment is genuinely impressive. Total AI investment—combining government and corporate capital expenditure with private and venture capital—is on a trajectory to exceed $100 billion annually by 2026.[10] Integrated circuit imports surged by 54 percent year-over-year in March 2026, signalling rapid expansion in the hardware layer of the AI ecosystem (Graph 11).

Major tech companies including Tencent, Alibaba and ByteDance are heavily investing in domestic AI firms, and the success of DeepSeek’s large language models has demonstrated that China can develop frontier AI capabilities even under chip-access restrictions.

Yet the AI push, which is mostly on vertical/ industrial AI, further pushes for automation and, thereby, job displacement. Even if productivity increases thanks to AI, the displacement of labour seems guaranteed, at precisely the moment when wage stagnation and household income weakness are the economy’s most pressing demand-side problem. The academic literature on AI and labour markets provides little comfort here. Research consistently finds that the near-term job displacement effects of automation significantly outpace the productivity-linked job creation effects.[11]

In China’s specific context, the problem is particularly acute. China’s flexible labour market already allows firms to adjust wages downward without the institutional protection afforded by strong trade unions or robust minimum wage enforcement. The rapid deployment of AI tools for cost-cutting and efficiency improvement—already well underway in manufacturing, logistics, customer service and back-office functions—will give firms further ammunition to reduce labour costs. The workers most exposed to this displacement are precisely those in the lower and middle segments of the urban workforce: the migrant workers and private-sector employees whose wages have already collapsed from 8 percent growth to near zero, and whose consumption underpins the domestic demand that China desperately needs to strengthen.

China’s most productive and innovative companies—the good cholesterol—are embracing AI precisely because it allows them to become even more competitive. But this competitive efficiency comes at the cost of the wage income that would allow Chinese households to buy the products those companies are producing. The supply side and the demand side are being pulled further apart by the same technological revolution.

The impact of the AI revolution on job displacement is expected to be so large that again, and in particular the fall in the labor supply will no longer a problem – at least not in terms of labor scarcity although aging will still be a massive fiscal problem. AI and automation will help in that regard but not only with goodies (a productivity boost) but also with problems (plummeting labor demand).

The New Normal and Its Vulnerabilities

The most recent analysis of China’s macroeconomic trajectory describes the economy as transitioning to a “new normal”—a structural pivot away from traditional sectors like low-value manufacturing and real estate toward high-tech industries and technological self-reliance.

This narrative is accurate as far as it goes. China’s digital economy, AI-related industries and advanced manufacturing are genuinely booming. The “new economy” is delivering strong industrial value-added growth, and as long as this continues, China has a reasonable prospect of meeting its 4.5–5 percent GDP growth target for 2026.

But the new normal comes with a structural vulnerability that its proponents acknowledge but have not yet resolved: it makes China even more dependent on global export markets, since the duality of the Chinese economy is constraining the growth of domestic consumption. In other words, the pivot toward high-tech supply does not address—and arguably deepens—the domestic demand deficit.

In fact, as China’s industrial output keeps growing and domestic absorption keeps lagging, the gap must be filled by exports. And this is where the model’s external vulnerabilities become critical. The US has progressively raised tariffs on Chinese goods to punitive levels. The European Union, previously more accommodating, is hardening its posture: the EU’s proposed Industrial Accelerator Act signals a willingness to limit the passthrough of Chinese industrial competition into European markets. Emerging economies that once served as alternative absorbers of Chinese exports are themselves grappling with high energy prices and weakening domestic demand.

The growing number of geopolitical shocks—most recently the escalating Iran crisis and its effects on global energy prices and supply chains—only amplifies these vulnerabilities. Rising energy costs hit Chinese firms as exporters and as energy importers. Heightened volatility in global trade disrupts the supply chains on which China’s sophisticated manufacturing depends.

The Structural Reform that Isn’t Coming

The logical response to the dual economy’s structural problems is well understood. It requires a comprehensive shift in the distribution of economic rewards from the state and corporate sector to households: a functioning welfare state that reduces precautionary savings, a stronger social safety net including healthcare, education and unemployment benefits, and wage growth that keeps pace with productivity. It requires allowing loss-making firms to exit the market, so that capital and talent can flow to more productive uses. And it requires accepting that a period of slower but better-quality growth—with higher consumption, lower investment and smaller export surpluses—is both economically necessary and strategically acceptable. None of these reforms appears to be on China’s policy agenda.

President Xi has explicitly rejected the welfare-state model as “welfarism.” The government’s industrial policy continues to favour strategic industries over broadly distributed household income growth. The political economy of local government subsidies to zombie firms remains intact, because officials face career incentives to maintain employment figures. The capital account remains closed, keeping household savings captive to low-return domestic financial instruments. And the Two Sessions of 2026, like those before them, continued with a tone of strategic patience: growth targets will be met through industrial output, not consumption stimulus.

The structural challenge here is that China’s leadership sees the dual economy not as a problem to be corrected but as a feature to be managed. The suppression of consumption is the price of the strategic investment programme. The closure of the capital account is the price of maintaining domestic savings mobilisation. The tolerance of zombie firms is the price of political stability. Each of these trade-offs is, from Beijing’s perspective, rational given the overriding objective of strategic competition with the United States.

The question is not whether these trade-offs are rational—they are—but whether they are sustainable. The cracks in the model are widening. The deflationary spiral driven by involution is deepening, not stabilising. Wage stagnation is cementing itself as a structural feature rather than a cyclical episode. AI-driven automation is about to accelerate both productivity and labour displacement simultaneously, likely worsening the domestic demand deficit before it improves it. And global export markets are closing, not opening.

Conclusions: A Model Under Stress

This paper has argued that China’s dual economy—high-performing on the supply side, persistently weak on the demand side—is neither an accident nor a temporary condition. It is a structural feature deliberately maintained in service of China’s overriding strategic objective: to win the competition with the United States for global technological and economic dominance. The suppression of household consumption is the price of the investment programme. The closed capital account is the mechanism that keeps cheap savings captive. The tolerance of zombie firms is the cost of political stability. Each of these trade-offs is, from Beijing’s perspective, rational given the goal.

The model has delivered genuine achievements. China now commands approximately 18 percent of global manufacturing exports, leads the world in electric vehicles, battery technology and solar manufacturing, and has demonstrated frontier AI capabilities even under US chip restrictions. These are not trivial outcomes. They reflect the power of a state capable of mobilising and directing resources at scale over decades.

But the model is now generating structural contradictions that are becoming increasingly difficult to manage. These are domestic cracks so they cannot be attributed to external pressure. The first and most immediate crack is involution. As domestic demand remains suppressed and global markets grow more resistant to Chinese exports, firms compete ferociously for a shrinking pool of domestic revenue. Prices fall, margins collapse and zombie firms proliferate as government support continues, directly or indirectly. While potentially rational individually, this model which does not allow Schumpeterian forces to operate erodes the pricing power of productive firms, and prevents the market consolidation that would allow China’s global champions to invest in the next generation of innovation. The second crack runs through the labour market. Firms under margin pressure cut the most flexible cost: wages. The collapse in the real estate sector has also carved out households’ wealth resulting in a dual squeeze on their purchasing power which constraints consumption. The third crack is the AI paradox. China is betting heavily on artificial intelligence and automation to restore productivity and break the deflationary spiral but existing literature on the impact of automation and vertical AI points to  job displacement outpacing productivity gains. In China’s specific context, where labour market flexibility already permits wage compression without institutional protection, AI deployment will give firms additional tools to reduce labour costs at exactly the moment when household income is the economy’s most critical weakness.

Three standard remedies could in principle address these cracks: monetary easing to stimulate private demand, fiscal expansion directed at household incomes, and debt restructuring to allow zombie firms to exit. All three are effectively constrained. The PBoC cannot ease aggressively without widening the interest rate differential with the US and putting additional pressure on the renminbi. The fiscal position is far weaker than headline deficits suggest once local government liabilities are included. And the industrial policy framework explicitly favours strategic high-technology sectors over the broad restructuring that would enable market exit.

The deeper reforms—building a welfare state to reduce precautionary savings, opening the capital account to improve asset allocation, and allowing loss-making firms to fail—are not on the policy agenda nor is an introduction of a welfare state. The implication of the above is a growth trajectory that will decelerate structurally, not merely cyclically. The involutionary spiral—falling margins, stagnant wages, weak consumption, more aggressive price competition—has no self-correction mechanism within the current model. If AI accelerates labour displacement faster than it generates new productivity, the spiral could tighten. And as global protectionism closes export markets, the pressure valve through which excess supply has historically escaped will itself close.

Alicia García Herrero is a Senior Fellow at Bruegel. She specialises in emerging markets, with a particular focus on Asia and Asia-European Union relations. Trade, finance and banking are the topics she covers most regularly. She is the Chief Economist for Asia Pacific at French investment bank Natixis, a Board Member of AGEAS insurance group, a non-resident Senior Fellow at the East Asian Institute of the National University Singapore and Adjunct Professor at the Hong Kong University of Science and Technology. She is a Member of the Council of the Focused Ultrasound Foundation and an advisor to the Hong Kong Monetary Authority’s research arm. Previously, she was Chief Economist for Emerging Markets at Banco Bilbao Vizcaya Argentaria and Head of the International Economy Division of the Bank of Spain. Alicia has served as visiting Professor at John Hopkins University, China Europe International Business School and Carlos III University. She holds a PhD in Economics from George Washington University.