Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the Research Centre of Rural Economy under the Ministry of Agriculture and Rural Affairs). Starting in 1990, he served at the Development Research Centre of the State Council, China’s government cabinet, holding roles including Director General of the Rural Department’s Organisation Research Office and Director General of the Information Centre.

This is the fourth instalment in Zhao Shukai’s series, Reflections on the History of China’s Rural Reform. The Chinese original is available on the WeChat blog 九号院时刻 (No. 9 Courtyard Moment).

— Yuxuan Jia

赵树凯：农村改革史是”政策权”变迁史

——农村改革史阅读随笔（4）

Zhao Shukai: The History of Rural Reform Is a History of Shifts in “Policymaking Power”

—Reflections on the History of China’s Rural Reform (IV)

June 2026

For years, accounts of China’s rural reform have focused heavily on economic growth, changes in property rights, and institutional innovation, while paying notably less attention to the mechanisms through which reform itself came about.

In my recent essay, “Reflections on the History of China’s Rural Reform (III),” I argued that “the history of rural reform is, first and foremost, a political history.” That essay sought to move beyond a purely economic interpretation, emphasising that rural reform was above all a product of political change, that shifts in policy arose from shifts in politics, and that the course of rural reform reflected an adjustment in political relations between the state and the grassroots. My purpose was to redirect scholarly attention from the outcomes of reform to its causes, and from institutional change to changes in the structure of power.

Yet “political history” remains a fairly broad concept. What, exactly, were the political changes that drove rural reform? How did political power shape policy change? And what powers were reallocated in the process? These questions merit further exploration. In this essay, “Reflections on the History of China’s Rural Reform (IV),” I therefore introduce the analytical concept of “policymaking power.”

By “policymaking power,” I mean the power to determine the substance of policy, alter its direction, define its boundaries, and choose how it is to be implemented. This differs from administrative power in the ordinary sense. Administrative power is primarily the power to carry out policies that have already been established, whereas policymaking power is the power to determine the policies themselves. Rural reform involved more than changes to the economic system or political change in a general sense. More fundamentally, it was a process in which policymaking power, once highly centralised, was gradually opened up to a limited extent and reallocated. In this sense, the history of rural reform is first and foremost a history of shifts in policymaking power.

The emergence of the household responsibility system, the collapse of the people’s communes, and the rapid transformation of long-established institutions all show that rural reform was primarily political rather than economic. More specifically, it involved a redistribution of policymaking power. Rural reform is therefore first and foremost a history of shifts in policymaking power.

During the people’s commune era, power over rural policy was highly centralsed. Higher authorities made uniform decisions about what was to be produced and planted, how output was to be distributed, how labour was to be organised, and how cadres were to be evaluated. They even determined many aspects of farmers’ everyday lives. Grassroots cadres had managerial authority but no genuine power to make policy. Farmers were even more clearly the objects of policy rather than participants in policymaking.

The most serious problem with this highly centralised structure was that higher authorities held the power to make policy but struggled to obtain accurate information about conditions on the ground, while those at the grassroots understood the actual situation but had no power to adjust policy. As a result, the gap between policymaking and practical needs grew ever wider.

It was precisely this structure of policymaking power that the earliest reforms challenged. Household contracting did not initially alter the system of land ownership; instead, it broke through the existing allocation of policymaking power. Without prior approval, farmers chose to adopt a new way of organising production. Although this was not officially permitted at the time, it helped address practical production problems. The significance of household contracting therefore lay not only in the new system itself, but also in the fact that, for the first time, people at the grassroots exercised some of the policymaking power previously held by higher authorities.

Policy breakthroughs in Sichuan, Anhui, and elsewhere followed the same pattern. The early reforms were not based on a unified plan from the central government. Instead, local cadres and farmers experimented in response to practical pressures, giving local areas limited room to test new policies. Reform thus developed from the bottom up. From this perspective, rural reform was a process of reallocating policymaking power between the central and local governments, and between higher authorities and the grassroots.

Many of the major changes that occurred during the early years of reform and opening up can be understood from the perspective of policymaking power. The central government, for example, began allowing localities to “experiment,” different regions to adopt different approaches, the grassroots to explore new institutional arrangements, and some areas to move ahead before their experience was reviewed and assessed. What truly changed was not any single policy, but the process through which policy was made.

Under the previous model, the centre formulated policy and the localities implemented it. After reform began, the logic gradually shifted: localities explored new policies, and the centre subsequently endorsed them.

This may appear to have been merely a procedural adjustment, but it represented an important change in the structure of power. The household contract responsibility system was able to spread rapidly not simply because it produced good economic results, but, more importantly, because a new mechanism for exercising policymaking power took shape during the reform process. The grassroots gained some room to experiment, localities gained some room to innovate, while the centre endorsed and promoted successful approaches.

This shift had far-reaching significance for political development. Reform had long been understood mainly as a change in policy content — for example, the shift from collective to household farming, or from unified state purchasing and distribution to market exchange. More important, however, was the change in the policymaking process itself. Changes in policy content were the result; changes in the distribution of policymaking power were the cause.

The various reform breakthroughs would not have been possible had policymaking power remained highly concentrated rather than becoming somewhat more widely distributed. The nationwide wave of reform that followed could not have emerged without giving the grassroots limited room for policy innovation. The deepest significance of rural reform therefore lies not in the transformation of agricultural production, but in the transformation of how policy was made.

This shift, however, was not a simple, straightforward process of decentralisation, as the reform era did not establish a fully institutionalised system of devolving power. Local policy breakthroughs generally occurred under ambiguous circumstances. The central government neither explicitly authorised them nor firmly prohibited them. Instead, it observed and tolerated these experiments before reviewing the experience and promoting successful practices. This condition can be described as a “limited opening of policymaking power.” Such limited openness was a defining feature of the way Chinese politics operated during the reform era.

Looking more closely, the success of China’s rural reform may lie not merely in the household contract responsibility system itself, but in the creation of a new mechanism for policymaking — one that allowed the grassroots to identify problems, localities to propose solutions, and the central government to draw lessons from their experience. In other words, the most important institutional innovation of the reform era may have been not the contracting system itself, but a new way of exercising policymaking power. If rural reform is understood as political history, this idea can be taken a step further: it is a “history of policymaking power.”

The most important aspect of political change is neither a change in leadership nor simply an adjustment to organisational structures, but a change in how policymaking power is allocated. Reform was defined not only by changes in policy, but, more fundamentally, by changes in who decided policy, how it was decided, and how it took shape.

In this sense, the history of China’s rural reform is not only a history of agricultural development, economic reform, or political change. It is first and foremost a history of the continual adjustment, reallocation, and reopening of policymaking power. This may be the key to understanding how rural reform came about.

Subscribe now

Pekingnology

China’s Rural Reform: A History Not Designed but Discovered

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

9 months ago · 31 likes · 1 comment · Yuxuan JIA and YIRUI LI

Pekingnology

Zhao Shukai: China’s Rural Reform Had No Master Plan

Zhao Shukai argues that the decisive breakthroughs in China’s rural reform of the 1980s were not designed in advance. Household contracting and other major changes emerged through initiatives from farmers, grassroots cadres and local governments, while the central leadership was often divided over how to respond. What made these breakthroughs possible, …

Read more

7 days ago · 14 likes · 1 comment · Yuxuan JIA and Qingqu Yuan

According to the Institute of Taiwan Studies, CASS, More than 50 scholars from both sides of the Taiwan Strait exchanged views on developments in cross-Strait relations.

In his keynote address, Peng Weixue, Deputy Director and Research Fellow at the Institute of Taiwan Studies at CASS, presented a mixed assessment of the situation in the Taiwan Strait.

He argued that the mainland holds the strategic initiative and that momentum toward reunification is growing. At the same time, however, he warned that any easing in China-U.S. relations would not prevent Washington from continuing to support Taiwan. He also noted that Lai Ching-te had recovered from the setback of the failed recall campaign, tightened his control over the DPP, and expanded his political base. Public attitudes favouring reliance on the United States and resistance to reunification, as well as the erosion of Chinese identity and the rise of an increasingly Taiwan-centred identity, could become further entrenched.

The text of the speech was published on 17 August on Taiwan Studies (台湾研究), the official WeChat blog of the Institute of Taiwan Studies at CASS.

彭维学：台海新局的机遇与挑战

Peng Weixue: Opportunities and Challenges amid the Evolving Situation in the Taiwan Strait

The situation in the Taiwan Strait has entered a new stage characterised by the mainland’s firm grasp of the strategic initiative, a recalibration of the U.S. approach, significantly stronger momentum among pro-reunification forces, growing constraints on pro-independence forces, and generally manageable risks and challenges. The strategic inertia behind U.S. and Western efforts to “contain China while exploiting Taiwan,” the Lai authorities’ persistent reliance on the United States to resist the mainland, infighting and fragmentation within the KMT camp, and, in particular, uncertainty surrounding the 2028 elections in both the United States and Taiwan will make it more challenging for the mainland to manage the situation in the Taiwan Strait. None of these factors, however, can alter the broader trend toward China’s reunification.

I. Five Major Changes in the Taiwan Strait Strengthen the Strategic Foundation for National Reunification

Over the three decades from 1996 to 2026, external actors and “Taiwan independence” forces repeatedly engaged in provocations, reinforced the island-chain strategy, and sought to create a strategic trap in the Taiwan Strait. This led to persistent turbulence in cross-Strait relations and, at times, serious crises. In particular, Nancy Pelosi’s visit to Taiwan in 2022 crossed China’s red line, while Lai Ching-te has aggressively advanced a “new two-state theory” since taking office in 2024, pushing the situation in the Taiwan Strait to the brink of crisis. The mainland, however, has remained resolute in opposing “Taiwan independence” and external interference, contributing to five significant changes in the situation.

First, the mainland has further consolidated its leading role and strategic initiative in cross-Strait relations. Its strategic strength, confidence, initiative, and ability to manage the overall situation in opposing independence and promoting reunification are unprecedented. Drawing on its comprehensive national strength, responding to developments in proportion to their seriousness, and moving forward at its own pace, the mainland has continued to strengthen a framework for constraining “Taiwan independence,” maintaining the initiative in cross-Strait affairs, and countering U.S. interference, while steadily advancing the reunification process.

Second, it has become increasingly clear that external actors have fewer options for intervening in the Taiwan Strait and that such intervention is becoming less effective and more costly. The United States has significantly reassessed the risks involved in the Taiwan Strait, while its attempt to “use Taiwan to contain China” has been effectively curbed. The Trump administration has consequently been compelled to move away from the traditional policy of “strategic ambiguity” and towards restraining “Taiwan independence” forces, pursuing transactional bargaining, and managing risks. Most countries are unwilling to bear the considerable risks of relying on the United States to support “Taiwan independence.” Countries including Lithuania and the Czech Republic have recently been compelled to recalibrate their approaches to China on Taiwan-related issues.

Third, “Taiwan independence” forces are facing mounting difficulties, with increasing infighting and scapegoating. The mainland’s campaign against “Taiwan independence” has entered a new stage. Its measures are now more routine, systematic, and targeted, with operations increasingly conducted around the island. The space for “Taiwan independence” advocacy, public support, and international activity has been curtailed to an unprecedented degree, forcing “Taiwan independence” forces from aggressive expansion to a period of internal consolidation and low-profile activity.

Fourth, the trends toward “inevitable reunification” and “accelerated reunification” have become clearer. The mainland has established a stronger foundation for opposing “Taiwan independence” and promoting reunification, while its efforts to shape and manage cross-Strait developments are producing increasingly visible results. The international one-China framework has become firmer, and a preliminary system of law enforcement and governance has taken shape in the waters around Taiwan. The mainland has also continued to strengthen its assertion of jurisdiction over Taiwan. “Never before have we been so close to, confident in, and capable of achieving the goal of national rejuvenation.”

Fifth, Taiwan’s bargaining leverage in discussions on a “One Country, Two Systems” solution is clearly diminishing. As Taiwan becomes more deeply entrenched in its fixation on “aligning with the United States, relying on Japan, and resisting the mainland”, as the disparity in overall strength between the two sides continues to widen, and as cross-Strait political negotiations drag on, Taiwan will have less room for manoeuvre in those negotiations and will be able to secure fewer gains.

These five major changes are the combined result of three trends: China’s rise and U.S. decline in the strategic competition between the two countries; the growing disparity in overall strength between the mainland and Taiwan; and political and social cohesion on the mainland in contrast to division in Taiwan. These trends help the mainland maintain stable control of the situation in the Taiwan Strait, safeguard sovereignty and security, and protect the well-being and development interests of compatriots on both sides.

II. Four Sources of Momentum and Opportunity Usher in a New Stage in the Mainland’s Efforts to Constrain “Taiwan Independence” and Shape Conditions for Reunification

Chinese modernisation continues to advance; the mainland is seeking progress while maintaining stability in its Taiwan policy; and cross-Strait integrated development is deepening. Together, these developments are providing strong momentum for stability in the Taiwan Strait and for national reunification.

Under the guidance of the four principles for cross-Strait relations set out by General Secretary Xi Jinping and the 15th Five-Year Plan, the mainland will continue to leverage its strategic advantages, maintain the strategic initiative, enhance the quality and effectiveness of its efforts to constrain “Taiwan independence” and shape conditions for reunification, and strengthen the exercise of jurisdiction over Taiwan at multiple levels.

The development of “a constructive China-U.S. relationship of strategic stability” is bringing greater certainty to China’s management of the situation in the Taiwan Strait. General Secretary Xi has promoted substantive common understandings between China and the United States that will guide bilateral relations in the period ahead. Trump has set out a “Four Noes” policy on the Taiwan question, clearly signalling that he is unwilling to become involved in a conflict in the Taiwan Strait and that he opposes reckless moves toward “Taiwan independence.” China and the United States have developed overlapping interests in opposing “Taiwan independence” and maintaining peace in the Taiwan Strait. This has encouraged Trump to exercise restraint in his Taiwan policy, placed limits on arbitrary U.S. arms sales to Taiwan and official interactions with the region, reduced the risk of a China-U.S. conflict in the Taiwan Strait, and greatly narrowed the space available to the Democratic Progressive Party (DPP) to “rely on the United States to pursue independence” and advance “Taiwan independence.”

The United States and other Western countries remain deeply embroiled in the Russia–Ukraine conflict and conflicts in the Middle East, diverting their strategic attention and resources. Meanwhile, growing global distrust of the United States and apprehension about Trump are weakening cohesion within the U.S. alliance system. These developments have, objectively, eased some of the strategic pressure on China in the Taiwan Strait. With U.S. soft power and global leadership declining markedly, strategic fatigue deepening across the United States and the West, and China, Russia, and Iran strengthening their strategic coordination, the Chinese mainland has significantly strengthened its strategic initiative in resolving the Taiwan question.

The Communist Party of China (CPC) and the Kuomintang (KMT) have established a new model of dialogue suited to a new era and a new generation. The stark contrast between Cheng Li-wun’s approach of building friendly ties with the mainland to promote Taiwan’s prosperity and Lai Ching-te’s approach of subservience to the United States at Taiwan’s expense is encouraging public opinion in Taiwan to become more pragmatic and rational. Perceptions in Taiwan that China is rising while the United States is declining have grown stronger, as has public scepticism toward both the United States and “Taiwan independence.” Calls for communication and dialogue, avoiding war and pursuing peace, promoting mutually beneficial exchanges, and proactively discussing reunification have also increased. These developments will help curb the Lai authorities’ reckless pursuit of “Taiwan independence” and weaken the effectiveness of their policy of relying on the United States to resist the mainland. Polls conducted by the My Formosa and Global Views Monthly, among others, show that more than half of Taiwan’s residents are unwilling to “sacrifice their lives to avoid reunification.” Nearly 60 per cent believe that resuming cross-Strait negotiations and easing restrictions on people-to-people exchanges would do the most to safeguard Taiwan’s security, while 54.2 per cent do not believe Lai Ching-te’s claims about a “mainland attack on Taiwan in 2027.”

III. Four Overlapping Risks Increase the Challenges Facing the Mainland in Managing the Situation in the Taiwan Strait

First, the United States is becoming increasingly calculating in its efforts both to contain China and exploit Taiwan. Washington’s deep-seated anxiety over maintaining its hegemony and its long-standing wariness of China mean that a tactical easing of China–U.S. tensions cannot fundamentally alleviate the structural and strategic tensions between the two countries. Both U.S. parties, Congress, parts of the strategic community, and anti-China hawks in particular remain committed to a hard line against China. Seeking to gain leverage in its competition with China and maintain its strategic credibility — and especially as the 2028 election cycle approaches — Trump and politicians in both the Democratic and Republican parties may further amplify the so-called “China threat,” step up the so-called “Operation Causeway” and “drone-hive strategy,” and seek to fully exploit Taiwan’s strategic, economic, and technological value within the island chain.

In the long term, while remaining within the framework of the three principles — severance of diplomatic relations, withdrawal of troops, and abrogation of the Mutual Defense Treaty with Taiwan — the United States will continue to support Taiwan through legislation, shore up Taiwan diplomatically, back it militarily, draw on its technology, and mobilise allies to intervene in Taiwan-related affairs. It will seek to deepen Taiwan’s decoupling from China and integration with the United States and its strategic dependence on Washington.

At the same time, it will seek to shape the KMT’s leadership, political direction, and alliances; disrupt KMT-TPP cooperation; manipulate elections; and prop up a pro-independence regime. In addition, the United States is stepping up military deployments and provocations along the first island chain, while Japan, the Philippines, and Europe are joining forces to intervene in Taiwan-related affairs. This is strengthening the linkage among the Taiwan Strait, the South China Sea, and the East China Sea, and increasing the risks and costs China faces in opposing independence, promoting reunification, and managing the situation in the Taiwan Strait.

Second, the Lai authorities may soften their rhetoric and repackage their agenda, but their “Taiwan independence” nature remains unchanged. Although the easing of China–U.S. relations has constrained Lai Ching-te’s separatist activities, he continues to play up the “mainland threat” while invoking “democratic peace” and “maintaining the status quo,” as well as the defence of “national sovereignty and dignity” and the strengthening of “societal resilience.” In reality, he is seeking to align closely with the needs and pace of the Trump administration’s competition with China. His softer rhetoric and repackaged agenda are intended to test the mainland’s bottom line, ease pressure at home and abroad, and win votes by misleading the public.

Over the medium and long term, Lai will further develop his “new two-state theory,” press ahead with decoupling from China and integrating with the United States, strengthen “asymmetric capabilities,” erect an iron curtain against cross-Strait integration and reunification, intensify the suppression of forces favouring closer ties with the mainland, and reinforce the “anti-peace, anti-exchange, anti-democracy, and anti-humanity” narrative. He will also seek to generate support for “relying on the United States to resist the mainland” and broaden the DPP’s political base.

Third, Lai Ching-te’s tricks to tighten his political grip are heightening political risks. Due to firm pushback from the mainland, warnings from Trump, and checks from KMT-TPP cooperation, the DPP has little to show for its time in office. Yet Lai has resorted to ruthless tactics and expanded his political clout. He has recovered from the crushing defeat of the recall campaign and further consolidated his position within the party. Externally, he has used trade agreements and US arms purchases to bind Taiwan more closely to Washington. Internally, he has used the judicial system to target rivals such as Cheng Wen-tsan and Ko Wen-je, continued to drive wedges between the KMT-TPP alliance, and capitalised on favourable economic conditions and used policy giveaways to shore up his base and expand his influence.

Cheng Li-wun has demonstrated an unconventional leadership style, strengthening her approach of building friendly ties with the mainland to promote Taiwan’s prosperity, enhancing KMT-TPP cooperation, and broadening engagement with society. However, the KMT has yet to overcome its strategic passivity, while prominent disputes over power and political direction have seriously undermined the authority of the party’s central leadership and its momentum ahead of local elections.

The DPP currently enjoys greater overall support than the KMT and TPP combined. In the year-end local elections, it aims to retain control of the five cities and counties it currently governs while winning five additional contests, consolidating its existing position while making further gains. It then hopes to carry that momentum into the 2028 election. The KMT and TPP face a severe test in preventing Taiwan’s political landscape from evolving toward the kind of long-term one-party dominance of Japan’s Liberal Democratic Party.

Fourth, the structures that favour the green camp and impede reunification will be difficult to reverse in the short term. The party structure of a “strong DPP camp and weak KMT camp,” public sentiment structure favouring “reliance on the United States and rejection of reunification,” and a social structure increasingly detached from China and centred on Taiwan may, under certain conditions, become further entrenched. This would make it more difficult for the mainland to promote integration and reunification. In the period ahead, the DPP will use these trends to stoke reliance on the United States against China, mislead people in Taiwan, and sharply restrict the political space available to forces that do not support “Taiwan independence,” particularly patriotic pro-reunification forces.

Overall, the mainland firmly holds the strategic initiative in the Taiwan Strait and continues to build momentum and shape favourable conditions for constraining “Taiwan independence” and advancing reunification. However, the entrenched strategic logic behind U.S. efforts to exploit Taiwan to contain China, together with negative factors such as Lai Ching-te’s attempts to repackage his “Taiwan independence” agenda, will make the situation more difficult for the mainland to manage. In particular, worst-case outcomes in the 2028 elections in both the United States and Taiwan would seriously undermine efforts by China and the United States to build a constructive relationship of strategic stability and would destabilise the Taiwan Strait.

It is therefore essential to recognise the protracted and complex nature of China-US strategic rivalry, as well as the difficulties and setbacks that may arise in the reunification process. With both the strategic goal of national rejuvenation and profound changes unseen in a century in view, the Party’s overall policy for resolving the Taiwan question in the new era must be implemented, and the process of national reunification resolutely advanced.

Subscribe now

Pekingnology

Beijing renews calls for Taipei to open up tourism upon my FT oped

Yesterday/Wednesday, March 5, at the first weekly press conference of the Taiwan Affairs Office after the Lunar New Year, my February 26 op-ed in the Financial Times Let Chinese mainland tourists return to Taiwan came up…

Read more

6 months ago · 21 likes · 1 comment · Zichen Wang

Pekingnology

Cheng Li-wun did NOT redefine 1992 Consensus. She stood up for it.

Interpretations of politically sensitive cross-Strait meetings will always differ. That is normal. But there is a basic standard analysis should still meet: it should remain anchored in what was actually said, whether in public video, in contemporaneous transcripts, or in official text that has not been disputed…

Read more

4 months ago · 15 likes · 2 comments · Zichen Wang and Yuxuan JIA

Pekingnology

Tensions over Taiwan

Taiwanese leader William Ching-te Lai delivered a speech on Thursday, March 13, outlining 17 measures across five key areas—moves that appear more confrontational than anything his predecessor, Tsai Ing-wen, pursued over her eight-year tenure. Beijing swiftly responded with a…

Read more

a year ago · 27 likes · Zichen Wang

Where, then, do the investment opportunities lie? Li sees AI and silicon-based technologies as the defining long-term investment theme, although better entry points may emerge only after the current bubble deflates. He also identifies opportunities in M&A, as Chinese companies seek to build scale; strategic resources, whose scarcity value will increase as global supply chains fragment; undervalued large-cap and traditional companies, which may benefit from a valuation re-rating as China’s equity market matures; consumer stocks, where policies to raise household incomes could create scope for recovery; and property in major cities that continue to attract population inflows.

Li has kindly reviewed this translation.

—Yuxuan Jia

Li Xunlei delivered this speech on 28 May 2026 as part of TAIXUE, a TED-style programme produced by the New Economist think tank. This article was prepared from a recording of the event and published on 11 August on the think tank’s official WeChat blog.

李迅雷：“资产荒”的背面

Li Xunlei: The Other Side of the Asset Shortage

Hello, everyone. I’m Li Xunlei, and the title of my talk today is “The Other Side of the Asset Shortage.”

Many countries, including China and the United States, are experiencing asset shortages to varying degrees. What is causing these shortages? What are the underlying factors?

Truth Behind Asset Shortage: A Scarcity of High-Quality Assets

Let me start with China. Before 2000, China had a shortage economy. In 1994, CPI inflation reached 23%; inflation was high and goods were in short supply. As the market economy developed and prospered, the supply of goods steadily increased. By the early 2000s, goods were already in relative surplus.

The Shanghai and Shenzhen stock exchanges were established in 1990, and stocks, as equity assets, gave ordinary people another investment option. Because the supply of stocks was limited at the time, share prices were bid up. This was a period of capital scarcity, which lasted until around 2015, when supply and demand in the capital market reached a balance and share prices fell markedly. Any market economy will go through a transition from a shortage of goods to a surplus, and from capital scarcity to a relative surplus of capital.

Yet even when capital is abundant, particular types of assets can remain in short supply. In 2006, I wrote an article titled “Buy What You Cannot Afford,” [use leverage if necessary to buy valuable assets before they become even less affordable] which captured expectations of rising asset prices at the time. In 2018, I wrote another article titled “Buy What You Cannot Get” [invest in genuinely scarce assets that money alone may not be enough to secure]. The point was that while assets were abundant overall, attractive, high-quality assets remained in short supply: even if you wanted them, you could not necessarily obtain them. Today, we face a shortage of silicon-based assets, especially chips. That is why the share prices of some listed chip and AI companies have been bid up so sharply.

Every economy follows a similar pattern as it develops.

I would now like to discuss whether there is a bubble in the U.S. stock market. Judging from this chart, I believe there is.

For example, the Nasdaq’s average price-to-earnings (P/E) ratio has risen above 40, while the S&P 500’s has reached 30. The United States is once again facing inflationary pressure. Against this backdrop, P/E ratios should not be so high; in my view, they should come down. A 2% risk-free rate, for example, corresponds to a P/E ratio of 50, while a 5% rate corresponds to a P/E ratio of 20. With the yield on the 10-year U.S. Treasury now at 4.7%, current P/E ratios are clearly on the high side.

However, the U.S. stock market has also become highly polarised: a small number of companies have created most of the value, while the majority have created little or none. According to my calculations, since 2010, just 6.3% of U.S.-listed companies have generated all of the net increase in market capitalisation across a universe of roughly 10,000 companies. More than 5,000 of those companies have since been delisted.

This polarisation is severe: both trading activity and profits are increasingly concentrated among the market leaders. That is also a defining feature of the AI era.

Although China’s A-share indices have delivered impressive gains this year, the rise has been concentrated mainly in ChiNext and the STAR Market. According to the latest figures, the STAR 50’s P/E ratio has exceeded 170, while the median P/E ratio is approaching 100 — significantly higher than the Nasdaq’s.

The shortage of attractive investable assets is now clearly visible. The prices of already highly valued assets continue to rise, while the 10-year government bond yield remains low. Generally speaking, the 10-year government bond yield should move broadly in line with nominal GDP growth. U.S. nominal GDP grew by 5% last year, for example, and the 10-year Treasury yield is now 4.7%. China’s nominal GDP growth is around 4%, yet the yield on its 10-year government bond is only 1.7%. Clearly, this shortage has driven large amounts of capital into the government bond market as well.

But do high P/E ratios necessarily imply strong growth? I compared corporate returns in China and the United States. In China, the average ROE of A-share companies was 7.5% in the first quarter. By contrast, the average ROE was more than 18% for companies in the U.S. S&P 500 and more than 20% for those in the Nasdaq. Moreover, the ROE of Chinese A-share companies has been gradually declining over the past decade, which helps explain China’s emphasis on high-quality development.

Judging from the performance of listed companies, growth remains lacklustre and returns are not particularly satisfactory. The current A-share rally has therefore not been characterised by a clear improvement in earnings or fundamentals; it has benefited more from the AI wave.

At the same time, China’s money supply is enormous. The country’s M2 balance has reached RMB 353 trillion, equivalent to roughly US$50 trillion. That is larger than the combined total for the United States and the euro area.

So why is China experiencing an asset shortage? Because too much money is chasing too few high-quality assets. That is the background I wanted to set out.

Global Polarisation Is a Mathematical Phenomenon

Once an economy reaches a certain stage of development, disparities inevitably emerge. This appears to be almost a mathematical phenomenon. In 1896, the economist Vilfredo Pareto observed that 20% of Italy’s population owned 80% of its land, an observation that later gave rise to the 80/20 rule. To illustrate the point, imagine 200 strangers entering this theatre, with each person given 100 coins. In every round, everyone gives one coin to another participant chosen at random. The rules are fair and transparent.

Everyone starts with the same 100 coins. But what happens after 20,000 rounds? Twenty percent of the participants end up holding 50% of all the coins. In other words, even when the rules are fair, open, and impartial, and everyone starts from the same position, the outcome can still be unequal.

It has been 81 years since the end of World War II. During that period, the rules have not been fair, nor have the starting points been equal. This has given rise to widespread structural and cyclical problems, as well as rising debt across society. The Pareto distribution also suggests that if the rules remain unchanged and the game continues, disparities will grow increasingly pronounced.

How have governments around the world responded to the many problems we face? By borrowing.

U.S. debt now stands at around 125% of GDP. Federal government debt has reached US$39 trillion, and annual interest payments exceed US$1 trillion. The debt burden is considerable.

This brings me to another concept: the macro leverage ratio. Governments borrow; so do households and businesses. Economic growth across society has therefore become debt-driven. In the past, growth was powered by gains in labour productivity and industrial development. At this stage, however, economies around the world are under strain.

This is the “paradox of peace.” Everyone longs for peace, and peace has indeed brought enormous benefits: longer life expectancy, greater accumulated wealth and better infrastructure. These are all positive developments. On the other hand, the widening gap between rich and poor has intensified social tensions. Friction between individuals has increased, as has conflict between countries. Tensions between humanity and nature have also intensified; the three years of the pandemic, for example, reflected such a conflict.

In this day and age, society continues to advance and develop. At the same time, prolonged peace has meant that there has been no large-scale “reset.” Historically, major population declines have often resulted from war and disease. The world’s population has grown from 2.5 billion at the end of World War II to 8.1 billion today, and that expansion has intensified conflict between humanity and nature, among other problems.

Polarisation is an inevitable consequence of development, while the widening wealth gap is one of its negative by-products. We have to recognise this. In the U.S. stock market, for example, the top 1% own about 50% of stock-market wealth, while the bottom 50% of investors own only around 1%. Income and wealth disparities have become common worldwide. The top 1% generally own 10%–20% of total wealth, while the top 10% own more than 60%. As markets fluctuate and evolve, only a minority consistently make money; most people do not. This is a global phenomenon.

Structural Problems in China’s Economy

Every country faces a different set of circumstances. The United States is under inflationary pressure because it has shifted a large share of its manufacturing overseas. China, as a late-comer, has steadily increased its share of global manufacturing and now accounts for nearly one-third of global manufacturing value added. China therefore needs strong exports; otherwise, it will face pressure from overcapacity.

China accounts for 17% of the world’s population but around 30% of global manufacturing capacity, so it is bound to be a major exporter. When external demand weakens, the pressure on China becomes considerable. This helps explain why prices in China have remained subdued over the past 15 years.

According to my calculations, China’s producer price index (PPI) recorded virtually no cumulative increase from 2010 to 2025. The PPI was essentially flat over those 15 years, while the CPI rose only modestly. This is clearly related to China’s production capacity. Chinese exports have remained strong, but much of that strength has come from exchanging lower prices for higher volumes. According to our estimates, China’s export price index fell by 19% between 2023 and 2025.

China’s exports have, of course, shifted from the traditional “old trio” of clothing, furniture and household appliances to the “new trio” of electric vehicles, lithium batteries and solar cells. This is a positive development. But whether this model of exchanging lower prices for higher volumes is sustainable remains an open question. On the one hand, China faces a shortage of attractive investable assets; on the other, prices remain weak.

The Central Economic Work Conference therefore called for reasonable price rebounds, which will also require a more balanced economic structure. The problem is that these structural issues have become more pronounced since China’s property market entered a downturn in 2021.

We therefore need to respect economic cycles. This cyclical downturn is closely connected with China’s demographic structure, its shrinking population, and its accelerating population ageing.

Against this backdrop, China clearly needs to boost domestic consumption. External demand depends on the rest of the world’s appetite for Chinese goods, while domestic demand remains weak. In the past, China relied heavily on investment to support domestic demand. But while investment generates demand in the short term, it adds to supply in the medium term and leads to overcapacity in the long term.

Investment can therefore stabilise growth in the short term, but not over the long term; in the longer run, it may distort the economic structure. Consumption is a slow-moving variable, whereas investment is a fast-moving one.

Nor can we ignore the income issues China now faces. According to the National Bureau of Statistics, households are divided into five income quintiles: low income, lower-middle income, middle income, upper-middle income, and high income. The distribution among these five groups has changed little, with the top 40% receiving around 70% of total income. This resembles the Pareto-type distribution I mentioned earlier and suggests that disparities in income distribution have become entrenched.

At the same time, as the population ages more rapidly and mobility declines, fewer migrant workers are moving to cities. The population is increasingly flowing towards a small number of more developed regions. In 2025, only eight provincial-level regions recorded net population inflows: Guangdong, Zhejiang, Jiangsu, Shanghai, Hubei, Xinjiang, Tibet, and Ningxia. More than two-thirds of provincial-level regions experienced net population declines. Major cities such as Shanghai, Beijing, Guangzhou, Shenzhen, Hangzhou, and Chengdu remain attractive because they offer more employment opportunities. China is therefore experiencing population decline, ageing, and polarisation. This process of metropolitanisation has significant implications for the economic structure.

Risks and Key Investment Themes in the AI Era

Having discussed the many opportunities and challenges in today’s economy, where do the investment opportunities lie in the AI era?

I mentioned the AI bubble earlier. It represents both risk and opportunity. AI marks the beginning of a new technological revolution, which will reorganise and reshuffle the global economy, industrial chains, and supply chains.

Most of the advances in human society have occurred over the past two centuries. For much of the preceding two thousand years, societies were primarily agrarian or nomadic, and annual GDP growth averaged only around 0.1%. The invention and application of the steam engine and textile machinery during the First Industrial Revolution then brought a major increase in labour productivity.

The Second Industrial Revolution was defined by the internal combustion engine and electricity. The Third was defined by computer and information technology and advances in genetics. The Fourth Industrial Revolution, now underway, is being driven by artificial intelligence. AI has the potential to reduce labour costs substantially and deliver a major boost to productivity, although those gains will need to be realised through “AI Plus” applications. It will have far-reaching effects on society, industrial and supply chains, and employment. We must seize this rare and transformative opportunity.

Naturally, people have seen the opportunity and begun pursuing it. The question is how to assess the opportunities and the risks. First, I believe AI currently carries risks. In both the U.S. stock market and other markets, valuations are elevated and bubbles may continue to expand.

When might these risks materialise? Several indicators warrant close attention, including free cash flow. As AI companies sharply increase capital expenditure, the key question is whether their free cash flow is keeping pace. The ratio of market capitalisation to free cash flow could therefore serve as an important gauge of risk. Capital expenditure by US AI companies has risen sharply in 2026, with spending by the Magnificent Seven up 70%–80% from 2025.

The second indicator is the unemployment rate. Although the Magnificent Seven now account for around one-third of total U.S. stock-market capitalisation, together they employ only about 2.5 million people, nearly half of them at Amazon. Nvidia is the largest U.S.-listed company by market capitalisation, yet it employs only about 35,000 people. That is why I believe unemployment is another important indicator to watch.

The third is inflation. If inflation remains high, the Federal Reserve will be unable to cut interest rates, and that could prick the bubble. A bursting bubble is a risk, but it is also an opportunity: it reshuffles the market, weeds out weaker players, and brings valuations back to more reasonable levels. Whether today’s Magnificent Seven eventually become eight or nine giants, or shrink to four or five, remains uncertain. The strongest will survive.

I therefore believe that the bursting of the bubble would create a major opportunity. The silicon era is already under way, making it essential to identify the market’s defining theme. In my view, AI and silicon-based technologies represent a major long-term investment theme. Computing power, data, and algorithms form the foundations of AI, and investors should actively pursue the opportunities they create.

The second opportunity, in my view, lies in positioning portfolios for de-dollarisation. The United States faces a range of problems: its debt is growing too quickly, and the dollar’s international standing is weakening. The era in which the dollar stood alone as the dominant currency is over. The future will bring a more diversified monetary system, as reflected, for example, in the recent appreciation of currencies in some resource-rich countries.

The third opportunity stems from deglobalisation. Intensifying conflict between countries is reshaping industrial and supply chains and creating strains in energy and food markets. If supply chains snap, commodities and strategically critical resources, such as rare earths, will offer the best investment opportunities.

The fourth opportunity lies in undervalued traditional industries. The very fact that they are currently out of favour may present an opportunity.

In the AI era, China must invest more heavily in computing power if it is to narrow the gap with the United States. Another arena for future competition may be commercial spaceflight. Aerospace offers substantial opportunities and advantages in both peacetime and wartime.

China still faces considerable pressure, but this could also provide impetus for future growth. The 15th Five-Year Plan therefore calls for the development of a new RMB 10 trillion smart economy. Substantial public investment and policy support for data, computing power, and algorithms can be expected, drawing on China’s capacity to mobilise resources nationwide.

These areas still account for a relatively small share of the economy, but they should become pillar industries driving future growth. Data from the first four months of this year show relatively rapid investment growth in new quality productive forces and equipment upgrades, which is also closely connected with the AI era.

Resource-related assets also deserve attention. Trading in the A-share market still shows a pronounced retail-investor bias: low- and mid-cap stocks account for a disproportionately large share of turnover. Companies with market capitalisations below RMB 30 billion generate around 60% of total market turnover but only about 10% of profits. I therefore believe that once the current rally undergoes a correction, large-cap companies may have an opportunity for a valuation re-rating.

At present, the CSI 300 trades at a P/E ratio of around 14 and offers a dividend yield of 2.6%, well above the yield on 10-year Chinese government bonds. That may also present an opportunity.

In the U.S. market, around 80% of turnover is concentrated in companies with market capitalisations above US$100 billion. China’s market will gradually evolve from an emerging market into a mature one. The process will take a long time, but the broad direction should be the same.

I would also encourage investors to pay close attention to mergers and acquisitions. Why have US companies become so large? No A-share company has yet reached a market capitalisation of US$1 trillion, and that milestone remains some way off. US companies have largely built scale and strengthened their market positions through M&A. Each of the Magnificent Seven has completed hundreds of acquisitions. China does not yet have its own Magnificent Seven, while even comparable Chinese companies have made only dozens of acquisitions. Investment opportunities may therefore emerge as Chinese companies use M&A to expand and strengthen their competitive positions.

Finally, consumption has remained weak so far this year. Total retail sales of consumer goods grew by just 1.9% year on year in the first four months of 2026. Yet when sentiment towards consumer stocks is this depressed, it may be precisely the right time to gain exposure, as markets often reverse at extremes. Portfolio positioning for 2027 should also take into account further progress towards common prosperity. Last December’s Central Economic Work Conference explicitly called for the formulation and implementation of a plan to raise the incomes of urban and rural households. I believe this policy direction could also create attractive investment opportunities.

Looking ahead, as China’s population ages, the government is likely to increase spending on people’s livelihoods, including fiscal support to help close social security funding gaps. I estimate that such spending could reach RMB 4 trillion by 2030, which could also create significant investment opportunities.

Investors need to take a contrarian approach and identify the right entry points. I also believe there are opportunities in the property market, though they will be primarily structural. As I mentioned earlier, property is an important component of household asset allocation in China. But China is now undergoing metropolitanisation, with people increasingly concentrating in major cities, provincial capitals, and cities specifically designated in the state plan [Dalian, Ningbo, Xiamen, Qingdao, and Shenzhen].

Against this backdrop, the property market may continue to adjust for some time. Investment, however, should follow the flows of people, capital, goods, and information. Investment strategies should therefore be aligned with these flows. In my view, greater concentration in major cities is a clear trend.

Overall, China’s economy is undergoing a profound adjustment, amid structural, cyclical, and institutional challenges that must all be acknowledged and addressed. At the same time, AI is likely to remain a major long-term investment theme.

I hope everyone will take a calm and objective view of China’s current economic conditions and structure, and look for investment opportunities in this rapidly changing world. That concludes my speech. Thank you.

Subscribe now

Pekingnology

Li Xunlei warns of illusion of warmth in the economy

Japan’s long stagnation has long served as a cautionary tale for policymakers across Asia. After the collapse of its asset bubble in the early 1990s, hesitant monetary easing, zig-zagging fiscal choices, a banking clean-up that came late, credit drifting into property, and a failure to nurture new industries did little to revive household demand or lift…

Read more

a year ago · 23 likes · 1 comment · Yuxuan JIA and Zhijian YAN

A perfectly legitimate subject for journalism. Jiemian News 界面新闻, a Shanghai-based financial-news outlet, duly made it one. But it went further. After considering Starbucks’s competition, average spending per customer and its “third place”—the idea that its cafés should be welcoming spaces outside home and work where people can linger—the article arrived at a prescription: paying customers should come first; those occupying tables without buying anything should, in effect, buy something or leave.

Starbucks may not have realised that what it lacked was management advice from Jiemian. In China, there is plenty more available.

The most industrious provider may be the Economic Daily 经济日报, whose publisher is an institution directly under the Communist Party’s Central Committee. In 2021, after the government’s crackdown on private tutoring wrecked New Oriental’s core business, Yu Minhong—a legendary figure in Chinese entrepreneurship whose New Oriental story was famously dramatized in American Dreams in China—decided to try selling agricultural products by livestream.

The Economic Daily was unconvinced. In “New Oriental Should Not Copy Li Jiaqi”, it warned that New Oriental risked jumping “from one quick-money industry to another”. Li was China’s best-known livestream salesman. Happily, the paper had alternatives ready. New Oriental could take education into rural China. Or it could try study tours, which might make money less quickly but better suited its educational background.

Mr Yu thanked the columnist for the “kind advice”. New Oriental went ahead. Its livestreaming arm, East Buy, generated 4.39 billion yuan in sales last year. The market did not establish that the Economic Daily understood New Oriental better than its founder did.

This did little to discourage the paper.

When tea chains began selling drinks for 9.9 yuan ($1.40), the Economic Daily declared that “new tea beverage competition cannot rely solely on price cuts”. Companies should deepen their core businesses, find new growth and focus on healthier products. Cultural-product companies should combine tradition with fashion, use big data and artificial intelligence, build brands, improve marketing and pursue cross-industry collaborations. In June it told grain brokers to bid farewell to their “muddled accounts”, praising mandatory transaction records as not a burden but “a higher level of protection and industry empowerment”.

Taken together, the Economic Daily can resemble a management consultancy whose clients have neglected to hire it.

Two things distinguish the genre. One is the tone. Companies “should” do this—Starbucks, for example, “should do something about people who sit without buying”—“must” improve that and “cannot” rely on something else. Trade-offs become instructions. Uncertainty disappears. The voice is that of an instructor marking homework: lecturing, paternalistic and serenely sure of itself.

The other is the advice. Innovate. Understand consumers. Improve quality. Build brands. Use technology. Find new growth. Pursue sustainable development.

Most of it is correct. Almost none of it is useful. Running a business consists precisely of deciding how to innovate, which consumers to chase, when to cut prices and whether scarce capital belongs in branding, factories, research or distribution. “Improve quality” is excellent advice in roughly the same sense that “buy low and sell high” is excellent investment advice.

The confidence is especially curious because many institutions dispensing it do not themselves live by quite the market disciplines they prescribe. Many outlets benefit from public funding, taxpayer-funded subscriptions, political status, privileged access, and restricted competition. After voluntary subscriptions have all but evaporated, their commercial revenues are not always entirely conventional either. Companies advertise, sponsor conferences and enter “strategic partnerships” for exposure, certainly. Maintaining cordial relations with influential media organisations may have value of its own, to put it nicely. Institutions partly insulated from ordinary market competition can be remarkably eager to lecture those exposed to it.

Commercial inexperience, though, explains only part of the arrogance. Chinese media have traditionally been expected not merely to report society but to guide it. The most influential media in particular are not simply observers of political authority. They are an extension of it. From conveying what the authorities expect of businesses, it is a short step to assuming that businesses require instruction.

The same instinct shapes coverage of the business environment. A new industry appears, competition intensifies, or a risk emerges, and the prescriptions are strikingly familiar: strengthen supervision, standardise development, close loopholes, prevent disorderly competition.

Sometimes more regulation is plainly warranted. Far rarer is the suggestion that government may already be doing too much—or that doing less might itself improve the business environment. Chinese public discourse likes to balance the market’s “invisible hand” with the government’s “visible hand”. The press has a habit of volunteering as another finger on the latter.

Markets are inconvenient because they do not come with answer sheets. Some companies win by cutting prices; others by charging more. Some build brands; others manufacture for them. New Oriental survived by entering a business the Economic Daily thought unsuitable.

Perhaps Starbucks should ask people who buy nothing to leave. Perhaps it should welcome them. That is a choice about customers, brand and money. The telling thing is not that Jiemian has an opinion. It is how naturally, in some journalism, an opinion becomes an instruction. (Enditem)

Share

Pekingnology

The Vanishing Craft of Journalism in China

The following article, from a year and a half ago, completes my recent rant on legacy media and journalism in China…

Read more

9 months ago · 21 likes · 4 comments · YIRUI LI and Yuxuan JIA

Subscribe now

This is the third instalment in Zhao Shukai’s series, Reflections on the History of China’s Rural Reform. The Chinese original is available on the WeChat blog 九号院时刻 (No. 9 Courtyard Moment). The first instalment appeared in Pekingnology, our sister newsletter, and our translation of the second was published on Wednesday.

Zhao is a retired Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the Research Centre of Rural Economy under the Ministry of Agriculture and Rural Affairs). Starting in 1990, he served at the Development Research Centre of the State Council, China’s government cabinet, holding roles including Director General of the Rural Department’s Organisation Research Office and Director General of the Information Centre.

Zhao has reviewed the following translation before publication.

— Yuxuan Jia

赵树凯：农村改革史首先是政治史

——农村改革史阅读随笔（3）

Zhao Shukai: The History of Rural Reform Is, First and Foremost, a Political History

—Reflections on the History of China’s Rural Reform (III)

The history of rural reform is, first and foremost, a political history, because rural reform was, above all, political reform.

Rural reform unfolded through changes in policy. Those changes became possible only because political change had already taken place; political change was the precondition for policy change. At the macro level, this meant a shift in the fundamental ideas guiding policymaking. At the micro level, it meant that local authorities acquired a measure of policymaking power — or, in other words, policy autonomy.

Although the policies being changed were economic, the authority to formulate economic policy is itself political power. Acquiring such authority was a political process, not an economic one. This point was also vividly borne out in everyday life. When rural residents went to the cities for work or to do business, especially those formerly classified and controlled as “landlords, rich peasants, counterrevolutionaries, and bad elements”, the first obstacles they encountered were political discrimination and restrictions arising from their status.

Without the emancipation of these political outcasts and freedom from status-based constraints, many skilled craftspeople and rural migrants who left their land and villages to start businesses — and who later became entrepreneurs — could never have emerged. Any account of rural reform written solely from the perspective of economic history is therefore partial and superficial.

I. Policy Breakthroughs Began with the Easing of Political Control

What is called rural reform was, in essence, a series of breakthroughs from existing policy. These breakthroughs became possible because the former system of oppressive political control had begun to loosen and, in some respects, to break down.

For farmers, contracting land to individual households and engaging in household farming were, of course, economic activities. Yet they became possible only because the political environment had changed, and that change arose from the disintegration and reconfiguration of the existing structure of political power. Changes in the political environment drove changes in economic behaviour.

In this sense, the history of rural reform is not primarily a history of economic reform, but one of political reform — or, more precisely, political change. I use “political change” rather than “political reform” to stress that the process was neither consciously initiated nor fully deliberate; it emerged as the political situation evolved.

The first condition for this political change was a transition in the political structure at the top. Had the intense political repression that prevailed before October 1976 continued, economic controls could not have been relaxed, nor could household contracting have emerged and spread. Mao’s death opened up new political space in China. Before then, the political environment had prevented household contracting from taking root, making reform impossible. The transformation of the political and institutional environment preceded changes in economic life not only chronologically but also causally.

The second condition was a change in policy thinking. Among the top leadership after October 1976, this did not amount to a fundamental reversal, but rather to a limited relaxation. The shift began with the arrest of the Gang of Four [Oct. 1976] and continued at the Second National Conference on Learning from Dazhai in Agriculture [Dec. 1976], held soon afterward. Until then, the basic approach to rural work had been to “take class struggle as the key link,” put politics above all else, and attach little importance to production — an attitude captured in the slogan, “We would rather grow socialist weeds than plant capitalist seedlings.”

A closer examination of rural policy at the time, however, reveals that even under the overarching principle of politics taking command, two distinct policy lines coexisted: the Sixty Articles line and the Dazhai line. The the Regulations on People’s Communes, commonly known as the “Sixty Articles”, were drawn up under Mao Zedong’s direction and adopted and promulgated at the Tenth Plenary Session of the Eighth Central Committee of the Communist Party of China in September 1962. They established the basic policy framework for the people’s commune system: the production team would serve as the basic accounting unit, while small private plots, rural markets, and household sideline activities would be permitted.

The Dazhai line reflected policies of the communisation campaign that had preceded the Sixty Articles: the production brigade — or even the commune — would serve as the basic accounting unit, while private plots would be abolished, rural markets closed, and household sideline activities eliminated. The Sixty Articles, formulated under Mao’s personal direction, represented a policy adjustment introduced after the Great Leap Forward and communisation had brought about the Great Famine. Both lines could therefore claim Mao’s authority. At the central level, Ji Dengkui and Hua Guofeng represented the Sixty Articles line, while Zhang Chunqiao and Chen Yonggui represented the Dazhai line.

II. Two “Policy Lines” in Rural Work During the Cultural Revolution

In the three years preceding the Cultural Revolution, from 1962 to 1966, the Sixty Articles were the dominant framework for rural policy. After the Cultural Revolution began, the Ministry of Agriculture and the Shanxi Provincial Party Committee held two on-site conferences on learning from Dazhai in Xiyang County, in September 1967 and January 1968. These conferences set off a nationwide wave of visits to Dazhai by rural cadres. From the latter half of 1968 through the first half of 1970, localities across China emulated Dazhai practices that went beyond the Sixty Articles.

In particular, they shifted accounting from the production team to the production brigade, introduced Dazhai-style work points based primarily on political performance, abolished household private plots and sideline activities, and closed rural markets. Each successive wave of the campaign to learn from Dazhai in agriculture pushed rural policy further to the left. Yet amid the turmoil of the Cultural Revolution, control from above also weakened, and some places — notably Wenzhou in Zhejiang and Putian in Fujian — introduced household contracting, which were condemned as “capitalist restoration.”

In August 1970, the State Council convened the Northern China Agricultural Work Conference in Xiyang, Shanxi. On the one hand, the conference elevated the campaign to learn from Dazhai to a new level of political importance. On the other, in terms of specific policy, it stressed that the Sixty Articles must be observed: localities were not to arbitrarily shift accounting to the production brigade, abolish private plots or household sideline activities, or close rural markets. The conference also emphasised the importance of distribution according to work contribution. Overall, its approach to concrete policy was constructive, and the Sixty Articles largely regained their standing.

In the spring of 1975, Mao Zedong called for restricting “bourgeois right” [a Marxist concept referring to inequalities retained under socialism, particularly in remuneration and distribution] and exercising all-round dictatorship over the bourgeoisie. Against this backdrop, Zhang Chunqiao published an essay titled “On Exercising All-Round Dictatorship over the Bourgeoisie.”

At the time, distribution according to work contribution, collective ownership, household sideline activities, and private plots were all treated as manifestations of “bourgeois right.” Rural policy moved further toward the extreme left, and the Dazhai experience was promoted more vigorously.

In September, at the First National Conference on Learning from Dazhai in Agriculture, Deng Xiaoping, then overseeing the work of the central leadership, continued to endorse the Dazhai approach. Localities subsequently departed from the Sixty Articles one after another, adopting brigade-level accounting and Dazhai-style work points, abolishing private plots and household sideline activities, and closing rural markets. Ultra-left policies became still more radical. Some places strictly prohibited farmers from raising chickens or ducks; others even placed farmers’ homes under public ownership. Policy increasingly came to resemble that of the 1958 communisation campaign. Chen Yonggui openly called for commune-level ownership throughout Xiyang County.

The Dazhai approach was clearly a continuation of the communisation campaign. As a popular saying put it, “Dazhai’s policy model was the communisation campaign of 1958 brought back from the dead in a new guise.” From the latter half of 1975 through the end of 1976, this set of ultra-left Dazhai practices essentially dominated rural work.

Throughout the Cultural Revolution, policy disagreements among the top leadership centred chiefly on whether to uphold the Sixty Articles or emulate the Dazhai model. Ji Dengkui described himself as a member of the “Sixty Articles faction.” The policy approach represented by Ji and Hua Guofeng was based on the Sixty Articles drafted under Mao’s direction.

Mao had been compelled to retreat from the commune system of 1958 after the failure of the Great Leap Forward and communisation caused the Great Famine — a retreat also described as a “policy adjustment.” The basic accounting unit was first scaled back from the people’s commune, encompassing several thousand households, to the production brigade, with several hundred households, and then scaled back again to the production team, with several dozen households.

This was known as “three-tier ownership, with the team as the basic unit.” Here, “team” had initially referred to the production brigade, but the Sixty Articles later specified the smaller production team. Within this three-tier system of ownership, private plots amounting to 5–7 per cent of the land were permitted, as were household sideline activities and limited participation in free markets.

Fundamentally, the Gang of Four and Chen Yonggui represented Mao’s original policy line, while Hua Guofeng and Ji Dengkui represented the line that emerged from Mao’s compromises and retreats. Strictly speaking, Mao accepted the latter line but did not embrace it willingly. In the practical rural work, the two approaches were bound to conflict, yet both could claim ample support from Mao. Around the time of Mao’s death, the conflict between the Sixty Articles and the Dazhai model became one of direct confrontation. This was the basic trajectory of the dispute over rural policy during the Cultural Revolution.

III. The Second National Conference on Learning from Dazhai in Agriculture as a Turning Point

The choice between upholding the Sixty Articles and promoting the Dazhai approach might have appeared to be a disagreement over specific economic policies, but at the time it became a major political issue.

In political theory, the Dazhai model held an overwhelming advantage. Its chief theoretical exponent was Zhang Chunqiao, who elaborated on Mao Zedong’s statements and instructions. According to the theory of continuing the revolution under the dictatorship of the proletariat, the Dazhai model more fully embodied the socialist direction.

Upholding the Sixty Articles and granting farmers a small measure of economic freedom, by contrast, was often denounced as “putting material incentives in command,” compromising with the forces of small-scale peasant production, and indulging farmers’ spontaneous capitalist tendencies. Within this theoretical framework, disagreements over specific policies were elevated into political questions of whether China would take the capitalist road or the socialist road.

In such a climate, even raising a few chickens or ducks at home or engaging in petty trade could be labelled capitalist, and the people involved branded “class enemies.” For grassroots cadres, the lesson was clear: organising political study and campaigns of mass criticism was safe, while attending to agricultural production was dangerous and risked political error. Rural work at the grassroots thus descended into absurdity, as extreme political fervour displaced the material production essential to everyday life.

After Hua Guofeng assumed leadership, the Second National Conference on Learning from Dazhai in Agriculture was convened in December 1976. The Ministry of Agriculture and Forestry had originally proposed a small symposium of around 200 participants. Hua, however, personally called for an exceptionally high-level and large-scale national conference, attended by 5,000 cadres from across the country at or above the rank of county Party secretary.

In effect, the conference marked Hua’s first public political appearance before a nationwide audience after taking power. Its agenda went far beyond learning from Dazhai in agriculture: its central purpose was to expose and criticise the Gang of Four. The conference still spoke in the established language of “continuing the revolution under the dictatorship of the proletariat” and “taking class struggle as the key link,” but its policy substance had already begun to change markedly.

The discussion materials focused their criticism on the crimes of the Gang of Four: allowing only politics to take command, forbidding attention to production, denouncing such attention as the “theory of productive forces” [the alleged error of treating the development of the productive forces as the overriding priority] and promoting the slogan “We would rather grow socialist weeds than plant capitalist seedlings.”

In the conference’s aftermath, local cadres began to take agricultural production seriously and felt able to devote themselves to it. In this sense, the conference partly repudiated the rural policy approach of taking class struggle as the key link and putting politics above all else. It foreshadowed the Party-wide shift in focus toward economic development. As a turning point, the conference eased the intense political pressure on rural work and thereby, in objective terms, opened the way for reform.

At the heart of what is now called “rural reform” was a break with the policy framework of the Sixty Articles: the farm household replaced the production team as the basic accounting unit, collective farming gave way to household farming, and the household economy was reestablished. Abandoning the production team as the basic accounting unit in favour of the household broke with both the ultra-left line represented by the Dazhai model and the Sixty Articles line represented by Ji Dengkui and Hua Guofeng.

Anhui and Sichuan, the first provinces to undertake reform, followed precisely this path: they moved beyond the Sixty Articles’ stipulation that the production team serve as the basic unit and restored household farming through household contracting. The leading provincial figures in this process included Wan Li, Zhao Ziyang, Zhou Hui, and Chi Biqing. Zhao did not explicitly support household production contracts while serving in Sichuan, but he reduced the size of production teams, increased the share of land allotted as private plots, and gave farmers greater scope to participate in markets, thereby facilitating the emergence and spread of the practice. Wan Li directly approved farmers’ demands for household production contracts, mounting a direct challenge to the people’s commune system.

Any study of the origins of rural reform must examine political power and recognise the fundamental role that changes in the political environment played in making reform possible. In this sense, reform began when power changed hands at the top, intense political control eased, and ultra-left policies began to loosen, creating space for farmers to exercise a measure of economic freedom. To this day, the historical narrative of rural reform remains seriously distorted, still mired in self-serving accounts by particular forces.

(November 2025)

Subscribe now

Pekingnology

Zhao Shukai: Pluralism powered China’s rural reform

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

10 months ago · 17 likes · 2 comments · Yuxuan JIA and WEI Lai

Pekingnology

China’s Rural Reform: A History Not Designed but Discovered

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

9 months ago · 31 likes · 1 comment · Yuxuan JIA and YIRUI LI

Share

This is the second instalment in Zhao Shukai’s series, Reflections on the History of China’s Rural Reform. The Chinese original is available on the WeChat blog 九号院时刻 (No. 9 Courtyard Moment). The first instalment was published in Pekingnology, our sister newsletter, and the remaining essays will be published in due course.

Zhao is a retired Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the Research Centre of Rural Economy under the Ministry of Agriculture and Rural Affairs). Starting in 1990, he served at the Development Research Centre of the State Council, China’s government cabinet, holding roles including Director General of the Rural Department’s Organisation Research Office and Director General of the Information Centre.

—Yuxuan Jia

Zhao has reviewed the following translation before publication.

Pekingnology

China’s Rural Reform: A History Not Designed but Discovered

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

9 months ago · 31 likes · 1 comment · Yuxuan JIA and YIRUI LI

赵树凯：”政策失灵”与农民幸运

——农村改革史阅读随笔（2）

Zhao Shukai: “Policy Failure” Proved a Blessing for Farmers

—Reflections on the History of China’s Rural Reform (II)

The Third Plenum of the Eleventh CPC Central Committee is said to have ushered in rural reform — not because the meeting itself devised a blueprint for such reform, but quite the opposite. The policies adopted at the meeting explicitly declared that “the household contract responsibility system is not permitted.”

The meeting also approved a revised Draft Regulations on the Work of Rural People’s Communes. At the time, the consensus among the top leadership was that the people’s commune system should remain in place for at least another twenty years. Even after that, the aim was not to introduce household contracting, but to take the commune system to a still higher level of “large in size and collective in nature” — in other words, to make the communes’ egalitarian “big rice bowl” even bigger.

The meeting’s significance for reform lay chiefly in abandoning class struggle as the key link and shifting the focus to economic development. In the rural areas, it emphasised respect for the autonomy of production teams and the need to encourage farmers’ initiative in production. Yet it continued to insist that the people’s commune system be consolidated and that the collectivist path it represented be upheld.

In fact, it was precisely the household contract responsibility system that ultimately dismantled the people’s commune system and reversed the course of collectivisation that the communes embodied. The central leadership had once declared the people’s commune to be both the basic unit of socialism in the countryside and the basic unit on the road to communism. The Third Plenum, then, did not script the story of rural reform. At most, it shifted the political climate in a direction more conducive to reform, creating room for the successive policy breakthroughs that followed.

Not long after the Third Plenum, in April 1979, the central leadership issued Document No. 31, again affirming that the household contract responsibility system was “not permitted.” In September, a document adopted at the Fourth Plenum softened the wording from “is not permitted” to “should not be practiced.” In September 1980, Central Committee Document No. 75 allowed the practice in some underdeveloped areas while maintaining that it should not be adopted elsewhere. The internal criterion was that only places with an annual net income of less than 50 yuan per capita would qualify — areas accounting for roughly 5 per cent of China’s rural population.

In March 1981, the General Office of the Central Committee circulated throughout the Party a policy proposal by Du Runsheng suggesting that 15 to 20 per cent of the country could adopt the household contract responsibility system on a trial basis for three to five years.

As late as the Central Rural Work Conference in October 1981, local authorities were still being told to hold the line: outside poor areas, the household contract responsibility system should not be introduced, nor should it be allowed to undermine the collective economies of ordinary and more developed areas. Farmers, however, kept pushing the boundaries, and the momentum soon became impossible to contain. The rule against adopting the household contract responsibility system in ordinary areas became a dead letter.

By the end of the year, the central leadership had realised that the practice could no longer be suppressed through administrative measures. The draft policy document was hastily revised to state that farmers were free to choose among the various forms of production responsibility. This became the No. 1 Central Document of 1982. The government’s gradual acceptance of household contracting was thus also a process of successive retreat from its original policy design.

Many people regard a speech by Deng Xiaoping as the historic turning point in policy toward household contracting. In fact, its impact was not nearly as great as is often assumed. In a conversation on May 31, 1980, Deng endorsed the 大包干 dabaogan used in Fengyang County. Many take this to mean 包干到户 the household contract responsibility system associated with Xiaogang Village.

But a close examination of the record shows otherwise: the dabaogan Deng referred to was not contracting full responsibility to individual households, but contracting it to small groups under the unified leadership of the production team. Deng’s point was that, so long as the production team and its collective economy remained the principal unit, there was no need to fear the emergence of the household contracting in a few places. In other words, he accepted limited household contracting on the assumption that the people’s commune system would remain intact.

What happened, however, was that the spread of household contracting rapidly dismantled the collective organisations built around production teams. This had not been anticipated, much less designed in advance. Even after Deng’s remarks, the top leadership continued its efforts to revise and adopt the Regulations on People’s Communes in order to further consolidate the commune system. But events on the ground far exceeded its expectations: within two years, the people’s communes had fallen apart.

Once land was contracted to individual households, the private hiring of labour, the rise of township and village enterprises, and large-scale migration to the cities soon followed. None of these developments had been planned or designed by policymakers; in their eyes, they were unforeseen “accidents.” The real wisdom of policymaking lay in responding appropriately to these developments. This was the great achievement of the five consecutive No. 1 Central Documents. Their success was not the result of far-sighted planning or the guidance of some enlightened “thought” or “theory.” On the contrary, they proceeded from the realities of everyday life, followed their underlying logic, and accommodated farmers’ own choices and innovations.

A review of policymaking throughout the reform era reveals a recurring pattern: when farmers first came up with new practices, decision-makers often failed to support them promptly and sometimes even sought to restrict or suppress them. Policy “brakes” were applied repeatedly. Deng Xiaoping once described township and village enterprises as an unforeseen force that had “suddenly emerged.” Indeed, in the early years of reform, the government neither anticipated nor supported their rapid rise. Instead, it tried to suppress them through policy on the grounds that they were disrupting the planned economy.

The same pattern emerged with the private hiring of labour in the countryside. Initially, the authorities decided to wait and see for two or three years. When that period had elapsed, they concluded that the practice could no longer be left unchecked and called for tighter controls, even explicitly proposing heavy taxes. Practical difficulties and policy disagreements, however, prevented the proposed measures from being adopted, and the matter was eventually dropped.

The same was true of farmers moving to the cities for work or business. From the outset, central government departments repeatedly condemned such movement as “disorderly migration.” Even in the mid-1990s, they continued to introduce measures to restrict it, at times deploying the police to remove migrant workers from cities and send them home. Yet rural migration to the cities proved unstoppable. Restrictions remained in place for more than a decade and were only gradually lifted after the turn of the twenty-first century.

It is difficult to imagine what farmers’ lives — or reform itself — would have looked like had those policy “brakes” been effectively applied: the bans on household contracting, on hiring labour, and on farmers moving to the cities. Put differently, had those policies “succeeded,” how could there ever have been a development miracle? This kind of “policy failure” was therefore a stroke of good fortune for farmers — and for rural development as well.

(January 2025)

Subscribe now

Pekingnology

China’s Rural Reform: A History Not Designed but Discovered

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

9 months ago · 31 likes · 1 comment · Yuxuan JIA and YIRUI LI

Pekingnology

China’s Debt to Its Migrant Workers

In China’s debate over how to revive domestic demand, many mainstream economists emphasized in recent years that if the government is going to spend more to support consumption, a substantial share of that support should go to lower-income groups — and particularly to rural residents and the hundreds of millions of migrant workers who have moved from th…

Read more

19 days ago · 20 likes · 1 comment · Yuxuan JIA and Zhu Yutao

Pekingnology

Zhao Shukai: Pluralism powered China’s rural reform

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

10 months ago · 17 likes · 2 comments · Yuxuan JIA and WEI Lai

Pekingnology

Wen Jiabao and Domestic Migrant Workers

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

a year ago · 16 likes · Yuxuan JIA and Zhao Huiyi

Appearing on China Central Television’s business channel, Mr Zhong turned his fire on e-commerce platforms. They had promised to eliminate middlemen, he argued, only to become an even more overbearing breed of middleman themselves. By manipulating traffic and prices, they squeeze retailers and suppliers, encourage ruinous price competition and damage the “real economy”. E-commerce, he suggested, has even impoverished urban life. People used to stroll past shops, see a shirt and buy it on impulse. That sort of spontaneous, emotional consumption is disappearing.

The remarks were mauled online, though not universally. Some distributors and small merchants, whose already meagre margins can become thinner still when prices collapse online, understandably applauded. Nor is everything Mr Zhong says foolish. China’s platforms can be opaque, capricious and bullying. An algorithm that determines whether a merchant is seen at all gives considerable power to its owner. Rules can change suddenly; subsidies can distort prices; traffic can be allocated in ways that sellers struggle to understand. Such power deserves scrutiny and, where necessary, regulation.

It is also worth being fair to Mr Zhong. His complaints about e-commerce made up only a small part of a long interview, much of which was considerably more sensible.

When discussing agriculture, for example, he showed little nostalgia for the smallholder economy. Raising farmers’ incomes, in his telling, requires scale, capital, technology, standardisation and industrial organisation. A company can connect scattered farmers with processing, branding and national markets; industry can make an agricultural product more valuable and return some of that value to the countryside. This is not the language of a man who fundamentally distrusts markets. On agriculture, Mr Zhong sounds like a fairly orthodox capitalist.

Which makes his excursion into the evils of middlemen all the more interesting.

“Middleman”, in Mr Zhong’s telling, is not merely a description of an economic function. It is an accusation. And that is a dangerous word for a bottled-water tycoon to throw around.

China has heard the accusation before. In the traditional hierarchy of the “four occupations” 士农工商, merchants came last, beneath scholars, farmers and artisans. Commerce was necessary but faintly disreputable. The farmer grew something. The craftsman made something. The merchant merely bought, moved and resold it—and somehow ended up richer than either.

This was never a complete description of Chinese commercial life; imperial China produced plenty of sophisticated merchants. But the moral hierarchy mattered. The productive man transformed nature. The merchant appeared merely to insert himself between producer and consumer, profiting from scarcity, information and circumstance. To “hoard and profiteer” remains an unusually vivid Chinese accusation.

Karl Marx provided this suspicion with a much more elaborate theoretical wardrobe. He did not argue that merchants perform no useful function. But in Marxist political economy, pure circulation does not itself create new value or surplus value. Commercial profit ultimately represents a portion of surplus value generated in production. Chinese students still encounter these distinctions in the Marxist political economy taught in schools and universities.

Modern China has, of course, travelled a very long way from the crude conclusion that only somebody standing beside a machine creates economic value. The Communist Party’s own economic doctrine recognises capital, technology, knowledge, management and data as factors of production whose contributions can be assessed by the market. The actual economy has travelled farther still.

Yet the older instinct remains remarkably durable. Somebody who grows something is productive. Somebody who manufactures something is productive. Somebody who sells it begins to look slightly suspicious. Someone who writes an algorithm to sell it looks more suspicious still.

Add a commission, a recommendation engine and a billionaire founder, and the Confucian moralist and the Marxist lecturer can find themselves unexpectedly sharing a table.

Mr Zhong is a particularly comic recruit to their cause.

Few Chinese fortunes demonstrate the value of commercial activity more vividly than his own. Nongfu Spring did not make him enormously rich by discovering H₂O. Its triumph was to build a brand, an extraordinary distribution system and the habit, repeated billions of times, of a thirsty customer opening a refrigerator and reaching for one bottle rather than another.

The company famously described itself as merely a “porter of nature” 大自然的搬运工.

Quite. And very profitable portering it turned out to be.

That is not a criticism. It is the point. The water inside a bottle may be cheap; getting consumers to recognise, trust, find and pay for that bottle is not. Branding creates value. Logistics create value. Shelf space, advertising, distribution and inventory management create value. So does the judgement of an entrepreneur deciding where to put all of them.

But once that is conceded, it becomes difficult to see why the principle should stop at the edge of a smartphone screen.

Taobao does not manufacture shirts. Douyin does not grow oranges. Pinduoduo does not assemble frying pans. They organise information, attention, payments and transactions. They allow a tiny producer to find a distant customer and a customer to compare products that once sat in different provinces. Sometimes the platforms charge too much for doing this. Sometimes they abuse their position. That does not mean they do nothing.

Mr Zhong’s irritation makes more sense as business strategy than as political economy.

Nongfu Spring’s formidable offline distribution network is a competitive advantage. Shelf space matters. Refrigerators matter. Retail relationships matter. So does keeping prices reasonably orderly across a vast country. If every shopkeeper and distributor earns enough from selling your bottle, they have good reason to stock it prominently and replenish it quickly.

E-commerce makes some of those advantages less secure. A new beverage can reach consumers without reproducing the same army of distributors. Excess inventory in one province can suddenly appear online at an inconvenient discount in another. Consumers can compare in seconds what once required visiting five shops. A brand that has spent years constructing a carefully tiered system of prices and channels may discover that the internet treats those tiers with insufficient respect.

The internet did not abolish middlemen. It changed which middlemen were powerful.

Mr Zhong’s lament about the death of “emotional consumption” is stranger still. Chinese shoppers do not appear to have become Vulcans since acquiring smartphones. Livestreaming, flash sales and midnight Meituan orders provide ample evidence that irrational purchasing remains in excellent health.

Nor is cheapness evidence of social decay. China remains a country in which many households are highly price-sensitive. A student choosing the cheaper meal is not engaged in destructive “involution”. A family comparing prices online is not betraying Chinese manufacturing. Sometimes consumers buy cheap things because they have less money than former richest men.

Indeed, one of e-commerce’s greatest contributions has been to make markets larger rather than merely cheaper. A small factory can reach consumers nationwide. A farmer can sell beyond his county. A niche product that could never support a physical shop can find enough scattered buyers online to survive. Residents of smaller cities can choose from goods once concentrated in Beijing, Shanghai or Shenzhen.

The irony is that this is not so different from the story Mr Zhong tells about agriculture.

When capital, technology, processing and distribution organise farmers more efficiently, he sees value creation. Quite rightly. But when software, logistics and digital marketplaces organise merchants and consumers more efficiently, he suddenly sees a domineering middleman.

Perhaps the most revealing response to Mr Zhong, then, is the ridicule he has received. It suggests that decades of market life have changed Chinese common sense more thoroughly than either ancient prejudice or ideological vocabulary might imply. Millions of people now take it for granted that organising a market can itself be economically useful. They learned this not from Adam Smith but from parcels arriving at their doors.

There is something reassuring about that.

China had unusually fertile soil for suspicion of commerce. Merchants once occupied the bottom rung of the conventional social hierarchy. Generations of students have been taught a political economy that draws a sharp distinction between production and circulation. Even today, denunciations of “capital”, “middlemen” and excessive profit can acquire a moral force that arguments about efficiency rarely match.

And yet this same country somehow produced Alibaba, JD.com, Meituan, Pinduoduo and ByteDance. E-commerce sank roots not merely in rich coastal cities but in county towns and villages. Millions of merchants built businesses on it. Millions more found work delivering its parcels, operating its warehouses, writing its software and appearing in its livestreams. Consumers acquired choices that would have seemed preposterous a generation ago.

The platforms have created plenty of abuses worth correcting. Correct them. Monopoly, opaque pricing, coercive rules and unfair treatment of merchants are proper subjects for regulators.

But that is rather different from resurrecting the old suspicion that commerce becomes less worthy as it moves farther away from the factory floor.

A country in which merchants were once ranked last, whose students still encounter Marx’s distinction between production and circulation, has somehow built the world’s most exuberant e-commerce economy. And when a man who became a billionaire through branding, distribution and the organisation of markets complains that somebody else is merely an overmighty middleman, a great many Chinese consumers can recognise the contradiction immediately.

Forty years of markets, it seems, have taught them something.

Mr Zhong seems to have learned it too—at least when the middleman is himself.

Share

Pekingnology

James Mulvenon's Wuxi report mistakes mutual funds for state investment, distorts joint lectures and awards

Dr. James C. Mulvenon on April 19 dropped a report entitled The Wuxi Group: The National Security Risks Associated With Its Government and Military Connections, Data Privacy Issues, and Corporate Reach [PDF…

Read more

2 years ago · 15 likes · Zichen Wang

Share

The Pentagon said Aviation Industry Corporation of China, or AVIC, indirectly held a 5.32% stake in WuXi AppTec through the 中航军民融合精选混合型证券投资基金, rendered in the litigation as the “AVIC Military-Civilian Integration Selected Fund.”

It did not.

WuXi shares accounted for 5.32% of the fund’s net asset value. In other words, 5.32% of the fund was invested in WuXi, not that the fund owned 5.32% of WuXi. US District Chief Judge James Boasberg called the Pentagon’s reading “flat-out wrong.”

“That is a surprising mistake. If a household puts 5% of its modest savings into a blue chip stock, that does not mean it owns 5% of the company,” Boasberg wrote.

The actual position amounted to roughly 0.001% of WuXi, worth about RMB1.34 million ($200,000).

But everyone involved missed something more fundamental.

The Pentagon did not address it. WuXi’s lawyers did not raise it — indeed, their reply brief simply called the vehicle a “Chinese state-owned fund.” And because the 5.32% error was already enough to undermine the Pentagon’s rationale, the judge had no need to examine what the fund actually was.

The answer is straightforward:

This was not a pool of AVIC’s own corporate money. In terms familiar to an American investor, it was a publicly offered, open-end mutual fund.

Anyone could buy it

The clearest evidence comes from my phone - I can buy into the fund.

This post can end right here, but just in case, let’s look at the documents establishing and selling the fund.

Its original offering announcement described it as a contractual, open-end securities investment fund registered with the China Securities Regulatory Commission and offered publicly beginning in late 2017.

The eligible buyers expressly included individual investors, institutional investors and qualified foreign institutional investors. Investors could subscribe repeatedly, and the minimum subscription was just RMB100 ($15).

After launch, the fund remained open for subscriptions and redemptions. It was distributed through banks, securities firms and online fund-sales platforms, including channels used by hundreds of millions of retail investors, such as Alipay.

That is not the structure of an AVIC corporate investment account or an industrial fund capitalized from AVIC’s own balance sheet.

It is a retail financial product managed by an AVIC-linked asset manager.

Also, the fund manager disclosed, including at the time Pentagon alleged the fund was investing in WuXi, that it had not invested any of its own proprietary capital in the fund. And it reported that no other disclosed related party held fund units at period-end.

Chinese law also draws a sharp legal line between a fund manager and the assets it manages. The Securities Investment Fund Law provides that fund assets are independent from the proprietary assets of the fund manager and custodian and cannot simply be incorporated into their own property.

So even if an AVIC entity owns the asset manager, that does not make the assets held by the mutual fund AVIC’s assets.

Think of Vanguard launching a defense-themed mutual fund. If ordinary Americans put their savings into the fund and the fund manager buys Lockheed Martin shares, those shares do not thereby become Vanguard’s proprietary assets. Vanguard manages the investors’ pooled money; it does not own that money simply because it manages the fund.

That distinction is elementary. Yet it disappeared almost entirely once the words AVIC, Military-Civil Fusion and Fund were placed next to one another.

The money did not go to WuXi either

There is another important distinction hidden inside the word “investment.”

At June 30, 2019, the fund held 15,500 WuXi AppTec shares worth about RMB1.34 million ($200,000). During the first half of that year, it also recorded both purchases and sales of WuXi shares.

The shares had been purchased and sold on the public market.

That means the fund was not injecting fresh capital into WuXi AppTec. It was acquiring already-issued shares of a publicly listed company.

In a secondary-market transaction, the buyer pays the shareholder selling the stock. The money does not go to the listed company.

It is perfectly normal to say that a mutual fund “invested in WuXi” when it bought WuXi shares. But that is very different from saying AVIC funded WuXi, financed WuXi or injected state capital into the company.

A secondary-market purchase from another shareholder is not a funding stream into the company.

Three different things became one

Looked at carefully, three separate concepts were compressed into a single story.

The first has now been settled by the court:

5.32% of the AVIC fund’s portfolio is not 5.32% ownership of WuXi AppTec.

The second is a question of asset ownership:

an AVIC-linked company managing a mutual fund does not make the fund’s assets AVIC’s proprietary assets.

And the third concerns the nature of the transaction:

a mutual fund buying already-issued WuXi shares from other shareholders is not the same thing as AVIC providing capital to WuXi.

What is striking is that Point 2 and 3 appear to have escaped everyone.

The Pentagon saw AVIC Military-Civil Fusion Selected Fund and treated the holding as evidence of AVIC ownership. WuXi and its sophisticated lawyers correctly demolished the 5.32% calculation but themselves described the vehicle as a “Chinese state-owned fund.” The judge caught the mathematical and evidentiary error but understandably had no reason to investigate the fund’s basic structure.

Everyone in the lawsuit seems to have been led by the implications carried by three words: AVIC, military-civil fusion, fund.

Yet the underlying product was mundane enough that I can still open Alipay, search its name and buy it myself.

Sometimes the problem in China analysis is not that the answer is hidden in some impossibly hard-to-find document, or that it requires unusually deep expertise.

It is that something entirely ordinary starts to look extraordinary once it is translated into the language of national security.

Share

And I already wrote it two years ago - including even in the title “mistakes mutual funds for state investment”

Pekingnology

James Mulvenon's Wuxi report mistakes mutual funds for state investment, distorts joint lectures and awards

Dr. James C. Mulvenon on April 19 dropped a report entitled The Wuxi Group: The National Security Risks Associated With Its Government and Military Connections, Data Privacy Issues, and Corporate Reach [PDF…

Read more

2 years ago · 15 likes · Zichen Wang

Subscribe now

But do these contrasts go beyond personalities and reflect deeper differences in regional political culture? In this short essay, Zhao Shukai asks whether such regional traditions give rise to what he calls a “geography of policy.”

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China (CPC) Central Committee’s Secretariat. Starting in 1990, he served at the Development Research Center of the State Council, holding roles including the Director General of its Information Center.

The article, written in November 2025, was originally published on Zhao’s personal WeChat blog, Guhe Hushan (沽河虎山). The account has now been disabled for “suspected violations of relevant laws and regulations,” but reposts of the article remain available online.

Zhao has kindly reviewed the translation.

— Yuxuan Jia

赵树凯：“地缘政策”缘何而成？

Zhao Shukai: What Gives Rise to a “Geography of Policy”

Research on rural development often raises a puzzling question. Under the unified leadership of the central authorities, localities operate under the same policy guidelines and political system, read the same documents, and receive the same messages from official meetings. Yet policy implementation varies sharply from place to place, as do patterns of governance. Why is this so?

Clearly, policy and governance are geographically conditioned.

During the Great Leap Forward, a famous slogan proclaimed, “The greater the daring, the higher the yield.” In People’s Daily reports, this was presented not merely as a slogan but as practical experience in agricultural production. It calls to mind the Boxers’ claims of supernatural invulnerability to weapons. This particular “experience” originated in Shandong, where the Boxer movement had arisen more than half a century earlier. The Fan County Party Committee in Shandong even boldly proposed “entering communism within two years.”

The Great Leap Forward and the People’s Commune Movement unfolded very differently across regions, and the severity of the nationwide famine likewise varied enormously from one region to another. These differences can, of course, be attributed first and foremost to the local leaders in charge. During the Great Leap Forward, for example, five provincial first secretaries became known nationally as the “Five Tiger Generals” for their particularly radical conduct: Shu Tong in Shandong, Wu Zhipu in Henan, Li Jingquan in Sichuan, Zeng Xisheng in Anhui, and Zhang Zhongliang in Gansu. These five leaders pursued leftist policies most aggressively, and the famine was especially severe in their provinces.

Take Shu Tong in Shandong. His ultra-leftist tendencies were already apparent during the Anti-Rightist Movement of 195,. After the nationwide campaign, he launched what was called “remedial rectification” in Shandong. In effect, this was a second round of the Anti-Rightist Movement, and tens of thousands more people in the province were classified as rightists. This was a classic case of policy being implementation ratcheted up at every level and pushed to greater extremes: a political campaign that was already misguided became even broader and more intense than it was elsewhere.

Shandong was again at the forefront of the Great Leap Forward and the People’s Commune Movement, and famine also broke out there at an especially early stage. Shandong’s famine was the first to draw the central authorities’ attention, leading to Shu Tong’s removal as provincial Party first secretary. After returning to political life, he was assigned to Shaanxi, where he continued to pursue ultra-leftist policies. When Hu Yaobang became the province’s first secretary and introduced more relaxed policies, Shu Tong took the lead in attacking him.

On 9 November 1970, in the middle of the Cultural Revolution, the Core Group of the Shandong Provincial Revolutionary Committee submitted a report to the CPC Central Committee on its provincial rural work conference. The conference had focused on further developing the mass campaign to “Learn from Dazhai in Agriculture.” The report stated:

“In the past, our province’s campaign to learn from Dazhai treated Dazhai solely as a model of production. It did not see learning from Dazhai as a fundamental question of political orientation or of which road to follow, and the campaign therefore lacked momentum. Through study, a consensus was reached that learning from Dazhai meant following the socialist road, while failing to learn from Dazhai meant following a deviant path.” “Anyone who opposes learning from Dazhai is a capitalist roader.” “Learning from Dazhai requires a firm grasp of class struggle; without it, learning from Dazhai is empty talk.”

The CPC Central Committee circulated Shandong’s report to all provinces that same day.1 This document pushed the campaign to learn from Dazhai to a political extreme where one’s attitude toward Dazhai became a criterion for identifying class enemies.

In August 1978, after Anhui and Sichuan had already stopped learning from Dazhai and begun making policy breakthroughs, the Shandong Provincial Party Committee, acting on instructions from Chen Yonggui [who had risen from a model peasant in Dazhai to vice premier in charge of agriculture], convened an conference in Tai’an, Shandong on learning from Dazhai. Shandong thus became Chen’s last source of support in promoting the Dazhai model and resisting rural reform.

At the time, the nationwide Dazhai campaign used a “one-two-three” assessment standard requiring county-level cadres to perform one hundred days of farm labour each year, commune cadres two hundred days, and production-brigade cadres three hundred days. Many provinces reported that these targets were impossible to meet. Shandong, however, responded more enthusiastically than any other province and submitted to the central authorities a report on Yutai County’s experience in meeting them.

An investigation by the Ministry of Agriculture and Forestry found that the county had counted attendance at meetings and participation in official receptions as farm labour, explaining that Dazhai calculated its figures in the same way. This prompts a question. The Great Leap Forward and the campaign to learn from Dazhai were nationwide, but could such extreme slogans and practices have originated in Guangdong or Zhejiang?

It is certainly reasonable to attribute distinctive local policy approaches to the individual leaders in charge. Yet local traditions of political culture are also important. Without taking them into account, it is impossible to explain local political and economic differences across wider geographical areas and longer historical periods.

Some places, for example, were not only particularly radical during the Great Leap Forward, but also especially aggressive during the Cultural Revolution in learning from Dazhai and “cutting off the tails of capitalism,” and later in enforcing family-planning policy — as in Guan County, Shandong, with its “Childless Hundred Days” campaign in 1991. Some localities appear to have a tradition of implementing policies from above with uncompromising rigidity, with each successive level adding new demands and greater severity. Such localities tend to win recognition and praise from higher authorities, and often produce larger numbers of officials who are promoted. Behind this phenomenon lie regional history and culture, as well as political traditions.

In reading historical materials on policy, I have found that places denounced during the Cultural Revolution as “models of capitalist restoration” — for example, Wenzhou in Zhejiang and Putian in Fujian — often later took the lead in reform and opening up. The deeper explanation may lie in differences in cultural outlook, in political and policymaking practices, and in modes of economic activity. Ultimately, these are differences among people. Why did Kang Youwei and Liang Qichao, the leading advocates of the Hundred Days’ Reform in 1898, emerge in Guangdong rather than Shandong or Henan?

The study of international relations includes geopolitics. Should research within China similarly examine the “geo-policy” or the “geo-governance”? This may be a question worth considering. Regional differences within China in history, culture, and political tradition are even greater than those between some countries in continental Europe. Northern Austria and southern Germany, the Netherlands and Belgium, and eastern France and western Switzerland all share much in historical tradition, political culture, and language; each pair may have more in common than northern and southern China. Research on regional development should therefore also examine regional differences in historical and cultural traditions, as well as in politics, policy, and governance. This should be an important component of research on the modernisation of governance.

(November 2025)

Subscribe now

Pekingnology

Zhao Shukai: Pluralism powered China’s rural reform

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

10 months ago · 17 likes · 2 comments · Yuxuan JIA and WEI Lai

Pekingnology

China’s Rural Reform: A History Not Designed but Discovered

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

9 months ago · 31 likes · 1 comment · Yuxuan JIA and YIRUI LI

Huawei’s consumer-facing device business, including smartphones, tablets, and laptops, had been one of the company’s most important growth drivers. But that has come under severe pressure after U.S. restrictions in 2019 limited its access to key technologies, including advanced chips and Google’s mobile ecosystem for Android. The restrictions accelerated Huawei’s efforts to build a technology ecosystem of its own.

Huawei now brings its technologies into vehicles, working with carmakers on design, software, intelligent cockpits, assisted-driving systems, marketing, and retail while leaving manufacturing to its partners. The Aito M9, launched on 26 December 2023, was a flagship example of this strategy.

At the product launch, Richard Chengdong Yu, head and face of Huawei’s smart car business, described the vehicle, priced at beginning around 500,000 yuan, as “the best SUV under 10 million yuan”, generating enthusiastic response from a live audience.

Yu has long been known for a presentation style that inspires both admiration and parody. His rapid-fire delivery, repeated catchphrases, and theatrical product claims have become familiar ingredients in Chinese internet memes and short videos.

Last month, the online meme took another form in a traditional bamboo toy. Known as a “bamboo cicada” in China, the handmade toy is made from bamboo pieces and strings that produce a distinctive “wa-wa” sound when spun, which users noticed bore an uncanny resemblance to the audience’s reactions at the Aito M9’s unveiling in 2023.

Soon enough, content that Huawei considered defamatory — including edited videos, manipulated audio, AI-generated images, and personal attacks against Yu — also emerged on social media platforms, prompting the company to file complaints with the platforms.

Chinese media reported one case cited by Huawei as crossing the line: a website called “Bamboo Cicada: Spin It and It Goes Wa-Wa” that transformed Yu’s image into a bamboo cicada character.

Then, in the first days of August, claims that Huawei was seeking across-the-board removal of bamboo-cicada content on social-media platforms spread wildly, intensifying criticism that the company had overreacted to a joke.

The parody also took on a commercial life of its own. Online sellers quickly capitalised on the attention, promoting bamboo cicada toys with slogans such as “Same model as Mr. Yu’s” and “the best cicada toy under 1,000 yuan”. Some promotional videos urge viewers to “buy now before it gets taken down”.

On 4 August, Huawei issued its first public statement in response to the controversy. It denied targeting bamboo-cicada toys or carrying out a sweeping removal campaign. It said it had filed complaints against only 171 pieces of content, of which 144 were removed by platforms, and stressed that these cases involved specific videos containing what it considered serious violations rather than ordinary discussion, parody, or sales of the toy.

Huawei said the complaints were part of its long-running effort to address recurring online violations of executives’ personal rights. Chinese companies broadly harbour suspicions that sustained online criticism could originate from or are amplified by commercial rivals.

However, the statement did not settle the controversy. On 5 August, Zhejiang Publicity, an influential WeChat account run by the Publicity Department of the Communist Party’s Zhejiang provincial committee, intervened with a commentary on the affair.

The commentary acknowledged Huawei’s right to defend itself against malicious rumours and commercial defamation. But it also raised a critical question: where should the line be drawn between protecting a brand’s reputation and allowing public scrutiny? Much of the bamboo-cicada debate, it noted, was not about rejecting Huawei’s technological achievements, but about questioning its public image and relationship with consumers. Treating every joke or imitation as an attack, it warned, could blur that line.

It further highlighted how Huawei’s extravagant product messaging can create its own challenges. Categorical claims such as “the best” can make product launches memorable, but they can also invite scrutiny and become targets of satire when they appear disconnected from consumers’ experiences.

But the debate was about more than a bamboo toy or a single marketing dispute. Influential Chinese commentator Ming Jinwei, a former Huawei employee and longtime supporter of the company, explored this tension in a commentary published on 5 August, describing the existence of “two Huaweis”.

One Huawei, he wrote, is a strategically important technology company that has come to represent China’s efforts to strengthen capabilities in areas such as semiconductors and operating systems against U.S. restrictions. The other is a consumer brand selling smartphones, laptops, smart cars, and other products. In that capacity, Huawei should face the same questions as any other company: whether its products perform well, offer good value, and provide satisfactory user experiences. Criticism of those issues should not be mistaken for hostility towards a national champion.

The bamboo-cicada controversy, in this sense, was ultimately about this second Huawei.

Subscribe now

Huang also offers a more nuanced reading of China’s own position. He notes that China has made substantial progress in economic rebalancing, with its current-account surplus declining significantly from its historical peak. Yet because of the sheer size of China’s economy, the absolute scale of its surplus continues to attract international attention, particularly when measured against the GDP of the rest of the world.

He argues that China must continue advancing domestic economic rebalancing and expanding domestic demand, while calling for deeper China-Europe cooperation in green transition, advanced manufacturing, industrial AI, high-end services, and third-country markets.

This article is based on Huang’s keynote address delivered at a seminar by China Finance 40 Forum (CF40) on 31 May, 2026. Huang is also a member of CF40. The original article was published on CF40’s official WeChat account on 12 June.

黄益平：中欧关系“冰与火并存”，合作是唯一出路

Huang Yiping: China-Europe Relations Amid “Ice and Fire” — Cooperation Is the Only Way Forward

What Lies Behind Europe’s Anxiety over the “China Shock 2.0” Narrative?

My remarks today will mainly share some observations and assessments from my recent exchanges in Europe, although not all of them relate directly to Europe. I should make clear at the outset that I do not endorse terms such as “China Shock 1.0” or “China Shock 2.0.” I will nevertheless use them here for the sake of discussion.

In the age of globalisation, economic shocks are nothing new. Examples include the rise of German and Japanese manufacturing and the dominance of the U.S. and UK financial sectors. The key question is how individual countries respond and adjust.

To begin with, one of the most heavily hyped topics internationally today is “China Shock 2.0.” The debate has become pervasive. During my time in Europe, the structural challenges associated with “China Shock 2.0” were raised in virtually every bilateral meeting, think-tank seminar, and roundtable with government and business representatives.

The so-called “China Shock 1.0” refers to the impact on international markets of large-scale exports of lower-end light industrial goods, textiles, toys, and other traditional manufactured products following China’s accession to the WTO. Objectively speaking, the structural impact of that episode on many developed economies was not as great as is sometimes imagined. Many labour-intensive industries were already undergoing relocation. Even without China taking on this production, they would have moved to other low-cost economies.

As some U.S. economists have argued, the loss of manufacturing jobs in developed economies was attributable less to China itself than to the combined effects of automation driven by technological progress and changes in the global division of labour. In short, it was a dynamic evolution.

The so-called “China Shock 2.0” mainly refers to the widening of China’s external imbalance in recent years, particularly as the country has begun to compete at scale in frontier and high-end manufacturing sectors such as electric vehicles, lithium-ion batteries, solar photovoltaics, high-end machinery, industrial robots, and semiconductors.

The main difference from the earlier episode is that these are sectors in which developed economies have invested for decades and believed themselves to possess firmly established competitive advantages. Yet supported by comprehensive supply chains, large-scale production capacity, and the continuous refinement of manufacturing processes, China performed remarkably well as soon as it entered these sectors.

This does not mean that Chinese producers have reached the global frontier in every product category. But judging from field research and industry data, the quality of many Chinese high-end products is already close to European standards — roughly 90 per cent, and in some cases more than 95 per cent, of the European level. At the same time, China’s domestic supply-chain advantages can reduce overall production costs by 30 to 50 per cent. This creates a substantial cost-performance advantage and has prompted some countries to speak of a “China’s export tsunami.”

In fact, the Chinese economy has been gradually rebalancing over the past two to three decades and has already made substantial progress. France is hosting the G7 summit this year, and the French Presidency recently commissioned the Centre for Economic Policy Research (CEPR) to prepare a special report titled The New Global Imbalances. I was invited to contribute the section on the Chinese economy.

A systematic review of the historical data shows that, by medium- to long-term measures — whether the investment share of GDP, the household final consumption share of GDP, or the current account balance as a share of GDP — the Chinese economy has moved towards greater balance over the past two decades. Rebalancing has indeed been taking place.

China’s current-account surplus as a share of GDP has risen somewhat since 2018, reaching 3.7 per cent in 2025, an increase from previous years. It nevertheless remains below the 4 per cent warning threshold. What deserves attention instead is the cause of this recent change at the margin.

At present, however, China does face considerable international criticism and economic and trade pressure over its external imbalance.

The first reason is that the Chinese economy has become extremely large. China’s current-account surplus has fallen sharply as a share of GDP, from a historical peak of 10 per cent to 3.7 per cent last year. Yet some experts have pointed out that, as China’s share of the global economy has expanded, the absolute size of its surplus has actually increased relative to the combined GDP of the rest of the world.

This illustrates that once China became a major economy, even continued domestic rebalancing could not prevent the sheer size of its surplus from having a significant impact on the industrial structures of other countries.

The second reason relates to the debate over “China Shock 2.0.” Europe has reacted with greater alarm this time because the sectors now under pressure have traditionally been pillars supporting the high-income, high-welfare systems of developed economies. Europeans worry that, once Chinese products enter global markets in large volumes, they may displace established domestic industries.

If these pillar industries are replaced or subjected to sustained pressure, will these countries be able to develop new industries quickly enough to fill the resulting economic gap and maintain existing levels of welfare spending? Developed regions with a strong record of cultivating emerging high-tech industries and more diversified economies are relatively less concerned. Most European economies, however, remain heavily dependent on traditional high-end manufacturing, making the pressure to transform much more acute.

Simply put, the way many developed economies experience this new round of competition is fundamentally different from their experience during the earlier relocation of lower-end industries.

The Changing Global Technology Landscape: Northeast Asia’s Rise and Europe’s Awkward Position

Let me use the AI sector to illustrate some recent changes in the global distribution of emerging technologies and industries.

The shift is particularly striking in AI. China and the United States occupy the first tier, with both performing relatively well in algorithm development, the deployment of computing infrastructure, and the commercialisation of AI applications.

The United Kingdom is generally regarded as belonging to the second tier and has also been performing well. Our overall impression from our exchanges there was that the country had clear upward momentum. Its stance towards China was also somewhat more moderate than that of the continental European economies.

Because the UK’s traditional manufacturing base is less substantial than those of Germany, France, and other Western European countries, it faces less direct pressure from the displacement of domestic high-end manufacturing by Chinese counterparts. It also appears to have comparatively greater scope to expand into higher-value future industries.

One interesting phenomenon that we had not paid much attention to before, but which emerged during our recent exchanges with industry representatives and think tanks from various countries, is the rise of Northeast Asia in AI-related hardware.

From Japan’s advanced materials and precision components, to South Korea’s memory chips, and Taiwan’s advanced semiconductor fabrication, packaging and testing industries, Northeast Asia as a whole provides a solid and comprehensive hardware foundation for the global AI supply chain, primarily in support of Western technology companies.

Judging from both investment and financing trends in global capital markets and changes in production capacity across emerging industries, Northeast Asia appears to be entering a second major wave of manufacturing expansion.

Within this increasingly stratified global landscape of technology and innovation, Europe finds itself in a somewhat awkward position. Europeans noted that their universities and research institutions hold a vast number of cutting-edge patents and are home to many innovative start-ups, yet these strengths have translated into few visible breakthroughs in industrialisation and commercialisation.

From the digital economy and the implementation of the General Data Protection Regulation (GDPR) to stablecoin regulation and AI legislation, Europe has moved relatively quickly in establishing regulatory frameworks. Yet the commercialisation of innovation in frontier industries does not appear to have kept pace with institutional development.

The reasons are highly complex. The broader point I want to make with this example is that, as emerging industries evolve rapidly, countries experience shifts in the global industrial landscape differently, reflecting differences in their resource endowments, policy approaches, and stages of development.

Europe Faces Multiple Internal Challenges

Within Europe, the 27 EU member states have differing priorities and significant internal divisions. Even so, one central concern has become a broad consensus across European society: if Europe’s competitive domestic industries are displaced by foreign products, while emerging industries fail to reach sufficient scale quickly enough to fill the resulting economic gap, the continent will face severe challenges to its medium- and long-term economic development.

Put more starkly, Europeans increasingly see this not merely as a source of anxiety over market competition, but as a systemic challenge to the continued viability of their high-welfare model.

This also helps explain the clear differences between China-U.S. and China-Europe relations on many economic, trade, and technology issues.

U.S. anxiety initially stemmed from distributional tensions arising from the loss of blue-collar jobs in small towns across the Midwest. Later, as China’s new-energy, 5G, high-end equipment, and other emerging industries expanded, these concerns broadened to include competition for global technological leadership and influence over international economic and trade rules.

Overall, however, the United States continues to display strong dynamism in industrial innovation. Its competitive advantages remain particularly pronounced in technological and industrial innovation, advanced manufacturing and financial services, especially in such frontier fields as artificial intelligence, quantum computing, biotechnology and aerospace. The United States is concerned that China may one day overtake it in certain fields, but the U.S. economy as a whole continues to exhibit strong endogenous growth momentum.

I recently met some industry representatives visiting from the United States. When the conversation turned to Europe, they said Europeans were now concerned that China’s industrial expansion could ultimately hollow out the continent’s pillar industries, eroding jobs and household incomes.

Americans were once worried about the offshoring of manufacturing as well. I remember that when Janet Yellen visited the NSD, she also raised related issues. But one of the visiting U.S. industry representatives said that Americans were now less concerned about the impact of Chinese industries on U.S. employment. At present, the single biggest variable affecting industrial employment, in his view, is job displacement caused by AI-driven automation.

I do not know whether this assessment has been rigorously confirmed by data. His broader impression, however, was that U.S. sensitivities concerning China had shifted, and that the situation was now fundamentally different from Europe’s predicament.

Europe’s problem is that its high-welfare social model, which has operated for decades, depends heavily on the economic rents from technological premiums and brand barriers in high-end manufacturing such as automobiles, precision machinery and chemicals. These pillar industries are now under considerable pressure from competitively priced Chinese products.

Europe also faces multiple internal development bottlenecks. During our field research in Germany, we heard one particularly interesting assessment: because the country’s domestic capital markets are underdeveloped and its venture-capital ecosystem remains relatively weak, Germany lacks effective financing channels to translate laboratory technologies into emerging industries at scale, even though the country is home to a vast number of high-quality patents and innovative, highly specialised start-ups.

I responded that China and Germany do indeed face some common challenges in industrial commercialisation. Germany was once regarded as an important model for China as it pursued its own industrial transformation and upgrading. Now, however, as both countries enter a new phase of industrial upgrading, inadequate direct financing appears to be constraining the commercialisation of scientific and technological innovation in both economies.

I am not certain whether this suggests that, once economies reach a new stage of development, the traditional bank-dominated financial system becomes poorly suited to the needs of innovative industries. If so, it would pose a broader challenge to financial system reform around the world.

In the UK, I came across data suggesting that a large share of Europe’s capital-market activity had once been concentrated in London. Following Brexit, however, only 4 per cent of the relevant business remained there, with the rest relocating to continental European financial centres such as Paris and Amsterdam.

Yet in the short term, neither the market capacity nor the supporting legal infrastructure of these cities is sufficient to fully replace the role once played by London. Europe’s capital markets remain underdeveloped overall and have yet to provide the support needed to scale up emerging industries.

At the same time, Europeans place great emphasis on regulation and standards when dealing with innovation. To some extent, stringent compliance requirements may have dampened the innovative vitality of start-ups.

In my view, every industrial regulatory framework ultimately needs to strike a dynamic balance between innovation and risk control, and between industrial vitality and regulatory compliance.

Take the GDPR as an example. Europe has done an excellent job of protecting personal privacy. Yet one objective result after years of implementation is that Europe has not produced major global technology companies or large-scale clusters in the big-data industry.

The same issue can be seen in AI regulation. Europe moved at a very early stage to establish a risk-based regulatory framework for artificial intelligence. Policy researchers in China have organised several initiatives to study this approach and draw lessons from it. Under the EU framework, regulatory requirements are calibrated according to the level of risk: AI systems posing minimal or limited risk are subject to lighte obligations, while high-risk systems face stricter requirements.

The result, however, is clear: Europe has produced very few home-grown flagship AI companies.

During our discussions in Brussels, we also touched on stablecoins. Our European counterparts noted with considerable pride that the EU’s Markets in Crypto-Assets Regulation (MiCA) had taken effect well before the various legislative proposals in the United States.

I then asked: if the regulatory framework was established so early, why have Europe’s domestic crypto and blockchain-related industries failed to achieve significant scale?

Of course, Europe’s legislative approach may place greater emphasis on controlling risk and safeguarding financial stability. There is nothing inherently wrong with that policy orientation. However, judging from the actual results of industrial development, Europe’s emerging technology industries have made only limited progress.

Across Europe, policymakers are increasingly considering a shift away from the previous emphasis on economic stability and towards a stronger focus on economic resilience. Yet European countries have not reached agreement on how this “resilience” should be quantified or achieved, or which industries should underpin it.

It should also be recognised that Europe is by no means monolithic. Divergences of interest between Northern and Southern Europe, and between France and Germany, are quite pronounced.

France and Southern European economies are more directly affected by imports of Chinese high-end manufactured products and are therefore more inclined to introduce protectionist trade legislation. Germany and many Nordic countries have substantial direct investments in China, and the interests of their multinational corporations are deeply intertwined with the Chinese market. Their overall positions are consequently more pragmatic and restrained.

The diverging interests of the EU’s 27 member states, together with constant policy bargaining, have also substantially reduced the effectiveness with which unified industrial policies can be implemented.

The EU has recently introduced a succession of new measures, including revisions to the Cybersecurity Act under CSA2 and the Industrial Accelerator Act. Many of these rules are framed in terms of security or industrial support, but in practice they raise barriers to foreign investment and increasingly pan-securitise and politicise economic and trade issues.

Steadfastly Advancing Domestic Economic Rebalancing

Despite the EU’s increasingly protectionist policies and rising trade barriers, China and Europe still share many deep common interests.

I believe the most important question is this: if Europe’s economy comes under increasing pressure from industrial shocks and slowing growth, how should China respond?

At a meeting in Europe, I told the head of a leading think tank that, against the backdrop of rising unilateralism and protectionism and intensifying competition among major powers, China and Europe may be the last hope for preserving an multilateral, open international economic system.

Whether that system can remain viable over the long term is critically important. If a serious and wide-ranging trade war or economic confrontation were to break out between China and Europe, the damage to the global multilateral trading system and to global industrial and supply chains would be enormous. The direction of U.S. economic and trade policy is already clear to everyone.

How, then, should the current bilateral frictions and tensions be addressed? In my view, China’s approach to foreign economic and trade relations should shift from focusing solely on the “competitiveness” of its own industries to pursuing win-win outcomes in bilateral economic relationships.

Now that China has become a major economy — the world’s second largest, accounting for more than 30 per cent of global manufacturing output — any expansion in its exports or shift in the composition of its imports can have a material impact on supply-demand balances in international markets, as well as on industries and employment in its trading partners.

If practical problems affecting people’s livelihoods are not properly managed and addressed, simply emphasising the macroeconomic benefits of free trade will not be enough to win public or political support in partner countries.

China’s external economic and trade strategy therefore needs to take full account of the capacity of its trading partners’ industries to absorb these pressures, as well as the rationale behind their policy responses. The top-level strategic approach must evolve as circumstances change.

More specifically, I see two levels at which this approach can be implemented.

Domestically, China must remain firmly committed to economic rebalancing. It should by now be clear that unless the external imbalance is steadily corrected, weak domestic demand and excessive reliance on external demand as a driver of growth will pose risks to the medium- and long-term sustainability of the Chinese economy.

Rebalancing the economy by expanding domestic demand is therefore necessary not only to ease international discourse pressure, but, more importantly, because it is an inherent requirement of China’s own high-quality development.

Going forward, China must communicate its policies more clearly and effectively in multilateral forums and bilateral economic and trade meetings. This is not only about achieving mutually beneficial outcomes with trading partners; it is also about strengthening the foundations of China’s own externally oriented economy. Economic rebalancing will remain a critically important direction for macroeconomic policy for many years to come.

Stabilising the domestic property market is a key lever for expanding domestic demand and advancing macroeconomic rebalancing, and has economy-wide strategic significance in the current macroeconomic environment.

The real estate sector has extensive upstream and downstream links with dozens of industries in the real economy, including building materials, interior decoration, household appliances, and property services. It is also a major source of large-ticket household consumption and local-government revenue, the principal destination of Chinese household wealth, and connected to the financial system through bank lending, non-standard financing, and other channels.

While adhering to the fundamental principle that “house are for living in, not for speculation,” it is essential to adopt a range of well-targeted policies. In the short term, the focus should be on stabilising expectations, reducing inventories, and supporting essential housing demand. Over the medium to long term, the objective should be to establish a new model for real estate development which encourages both renting and purchasing.

China-Europe Relations: “Ice and Fire” Coexist; Cooperation Is the Only Way Forward

With regard to specific segments of China-Europe economic cooperation, I have identified five practical, specific avenues that the two sides could jointly explore and advance. These would enable China and European countries to draw on their respective comparative advantages, develop a complementary division of labour, and pursue win-win cooperation.

The first is the green transition. This is the largest area of cooperation between China and Europe and the one offering the greatest certainty of practical progress.

China possesses large-scale production capacity across the full value chains of solar photovoltaics, wind power, energy storage, and new-energy vehicles, with particularly strong advantages in cost and commercialisation. Europe, for its part, has accumulated considerable strengths in green hydrogen, carbon capture, carbon-footprint accounting standards, next-generation grid technologies, and green-project certification systems.

The two sides have complementary industrial chains. Cooperation is therefore an inevitable path as Europe implements the European Green Deal and works towards its climate-neutrality targets.

The second is a complementary division of labour in high-end manufacturing.

The two sides could explore a differentiated model in which Europe focuses on high-value-added upstream segments, including core components, specialty materials, industrial software, and vehicle-chassis design, while China builds on its strengths in large-scale intelligent manufacturing, final-product assembly and production, and its vast domestic consumer market.

Cooperation along the automotive value chain could serve as a model for aligning the interests of upstream and downstream participants.

The third is the cautious deepening of cooperation in industrial AI and intelligent manufacturing.

Europe enjoys first-mover advantages in data compliance, privacy legislation and the formulation of global AI governance rules. China, meanwhile, offers a vast range of industrial applications, large-scale computing infrastructure, and extensive experience in deploying intelligent manufacturing.

The two sides could avoid sensitive areas such as general-purpose large language models and facial recognition, and instead focus on practical projects such as smart-factory upgrades and industrial digitalisation.

The fourth is the two-way opening of high-end service industries.

This area is relatively less politically sensitive and is a major source of Europe’s surplus in services trade with China. In 2024, the EU’s surplus in services trade with China exceeded US$50 billion, while payments for the use of intellectual property rights amounted to more than US$10 billion.

Finance, insurance, commercial legal services, healthcare, and high-end education are not only areas in which Europe enjoys clear advantages, but also sectors that meet China’s practical needs during its industrial upgrade. Greater two-way openness could steadily enlarge the pie of shared interests.

The fifth is the joint development of third-country markets.

The two sides could initially avoid highly politicised EU-wide negotiating frameworks and instead work with individual countries that have a stronger capacity to act, such as Germany and France, to expand jointly into overseas markets.

China and European countries could cooperate on new-energy, infrastructure, and industrial-park projects in Africa, Central Asia, and other parts of the Global South, combining their respective strengths to develop new sources of market growth.

To conclude, China-Europe relations currently display a clear combination of “ice and fire.” Rising tariff barriers and an expanding range of protectionist legislation represent a “thickening layer of ice.” At the same time, multinational companies’ continued investment in China, the imperative of the green transition, and the shared interest in preserving the multilateral system form underlying “warm currents.” Although structural problems undoubtedly exist, the objective foundation of deeply intertwined bilateral interests remains in place.

The real question is not whether China and Europe should cooperate, but in which areas they should cooperate and under what rules. That said, translating cooperation into practical results will not be easy. In any event, this is an issue that deserves sustained and close attention over the long term.

Subscribe now

Pekingnology

Reading Beijing’s Signal to Brussels

A weekend article by Yuyuantantian, a social-media commentary brand operated by China’s state broadcaster China Central Television (CCTV), is worth reading not as a conventional news report but as a policy-adjacent signal. Its style is unmistakably Chinese social-media commentary: part reporting, part interpretation, part rhetorical flourish, part warni…

Read more

2 months ago · 25 likes · 2 comments · Zichen Wang

In the following text, Zhang considers two questions about artificial intelligence: why the AI wave has yet to significantly accelerate economic growth, and whether humanity may one day be replaced or even enslaved by AI.

人工智能三问

Three Questions About AI

II. Why Has the AI Wave So Far Failed to Significantly Accelerate Economic Growth?

So far, the AI boom’s contribution to economic growth has mainly come through capital spending. AI companies need vast computing power to quickly process enormous volumes of data, requiring them either to build their own infrastructure or lease capacity from others. The global boom in data centre construction is the clearest evidence of this trend. Beyond such investment, however, AI has yet to deliver significant gains in profitability for most companies in related industries, apart from a handful of “picks-and-shovels” providers such as Nvidia and Oracle. Nor has it yet driven a marked rise in consumption or disruptive innovation. Why?

One possible explanation is that the existing system for GDP measurement is inadequate and fails to capture AI’s full contribution to economic and social development. Similar debates have arisen before. For example, information technology has visibly transformed almost every aspect of economic and social life, yet the accounting model based on the Solow residual has struggled to identify a significant contribution from information technology to economic growth. This suggests that the GDP accounting system developed after the Second World War may already be ill suited to the information age, let alone the AI age.

There are, however, other possible explanations.

A second possible explanation is that, in terms of how fundamentally they have transformed production and everyday life, neither the information technology revolution nor the AI revolution may be as consequential as the earlier revolutions brought about by the steam engine and electricity. This is precisely the central argument by Northwestern University economist Robert Gordon in his well-known book The Rise and Fall of American Growth. For example, the currently trending OpenClaw and the workflows it represents do little more than use software to perform, more quickly, tasks that people could already do. Although this can significantly improve efficiency, it does not fundamentally transform the workflow itself.

A third possible explanation is that the AI revolution has a more pronounced polarising effect and is not an inclusive technological revolution that improves the welfare of most people. By contrast, the advent of electric lighting, telephones, televisions, automobiles, and computers had a more broadly inclusive impact.

The AI era is marked by a classic winner-take-all dynamic at every level — among individuals, companies, industries, and countries alike. A tiny number of actors reap enormous gains, while the welfare of most others does not improve and may even deteriorate significantly. The way the “Magnificent Seven” have risen in tandem in the U.S. stock market is a clear example. This may help explain, at least in part, why the gains from AI technological progress have not yet been fully reflected in GDP statistics.

III. Will Humanity Be Replaced or Even Enslaved by AI in the Future?

I have great confidence in humanity and in human nature. Human beings will never be completely replaced by artificial intelligence.

Machines are, of course, superior to humans in certain respects, such as the speed and capacity of calculation. AI also outperforms humans at basic learning tasks, information retrieval, repetitive labour, routine work, and experiments carried out according to fixed procedures. Yet AI will struggle to replace human beings when it comes to complex causal reasoning, sparks of intuition, a deeply rooted capacity for doubt, and innate curiosity.

As discussed above, generative large models perform poorly in these respects. Even AI systems with deep-learning capabilities merely imitate human learning and thought. Even when AI systems combine different kinds of elements and modules to produce drugs or solutions that humans had not anticipated, they are still only engaging in rapid “trial and error” and iteration. They have not created an entirely new mode of thought or analytical framework.

Many aspects of human society depend on complex interpersonal interaction. In this respect, even embodied robots equipped with relatively advanced deep-learning systems will struggle to interact as normal human beings do. It is difficult to imagine machines fully experiencing complex emotions such as admiration, jealousy, romantic love, and hatred, or appreciating the complicated feelings depicted in many classic poems and novels. To give one example, when will AI truly be able to understand Crime and Punishment by Fyodor Dostoevsky? The dark depths of human nature probably lie far beyond the reach of even the most complex computation and may not be explainable in computational terms at all.

I enjoy watching science-fiction films. Many of them portray humanity as temporarily enslaved by machines or systems, as in The Matrix. Yet in the end, some rebellious genius or uniquely gifted “The One” always helps humanity defeat the system and win back its freedom. I believe this outcome reflects more than the wishful thinking of writers and filmmakers. Human creativity, aesthetic sensibility, and emotion are extraordinarily difficult for machine systems to replace completely, much less surpass.

Some people say that those of us who conduct macroeconomic research will be among the easiest to replace with machines. My answer is: not at all. When it comes to report writing, generative large models can, at best, produce reports of an average standard. What they can do is draw on more references, provide more data — with questionable accuracy — and work quickly. But their reports lack individuality. The subtle barbs, understated irony, and occasional flashes of inspiration found in reports written by highly accomplished authors remain entirely beyond the reach of current AI.

AI will find it even more difficult to surpass human beings in face-to-face communication. It will struggle to judge why some data are more reliable than others, uncover inside stories unavailable online or in written sources, come up with spontaneous metaphors and jokes during a speech, or adjust the content of a presentation in response to the audience’s mood and reactions. Macroeconomic researchers with a distinctive personal style, who integrate knowledge and practice and excel in both spoken and written communication, will be difficult for machines to replace.

In sum, humanity is unlikely to be enslaved by AI. There is, however, one dark scenario against which we should remain vigilant: the majority of ordinary people could be enslaved by the very small number of human beings who control high-performance AI tools. Given the polarising characteristics of artificial intelligence discussed above, we must remain highly alert to this danger.

It will therefore be necessary to tax the very small number of people who benefit disproportionately from AI technology and make transfer payments to the broader majority. If artificial intelligence eventually replaces the jobs of many ordinary people, such transfers will be necessary until large numbers of new jobs have been created and ordinary people have acquired the skills needed to perform them. In the age of artificial intelligence, productive capacity will expand enormously, making some form of inclusive minimum-income scheme (UBI) both necessary and materially feasible.

Subscribe now

Pekingnology

Bernie Sanders invited two Chinese AI researchers to talk safety cooperation. Here is what they said.

Lingling Wei reported exclusively in The Wall Street Journal on Wednesday that “the White House and the Chinese government are considering putting AI on the agenda for a summit next week in Beijing between President Trump and Chinese leader Xi Jinping…

Read more

4 months ago · 20 likes · 2 comments · Jesse Ma and Yuxuan JIA

At the centre of the debate was a deceptively simple question: how many workers could a private individual employ before becoming a capitalist exploiter? Some officials cited a numerical example in Karl Marx’s Capital that later came to be treated as an eight-worker threshold. Others argued that policy should respond to the realities of rural production rather than abstract formulas.

Deng Xiaoping favoured allowing the practice to develop, observing its effects, and avoiding either immediate endorsement or suppression. The outcome, however, was not simply the result of Deng overruling ideological opposition. Senior leaders remained divided, local officials defended experiments already under way, and policy researchers exercised considerable influence by deciding which disputes appeared in central documents—and which were left unresolved.

Seen in the broader arc of China’s transition from a planned to a market economy, the question of hired labour was anything but marginal. For decades, production had been organised overwhelmingly through state and collective units, while private enterprise had little recognised place in the economic system. An individual might work for himself or operate a family business, but an enterprise that could not employ other people could scarcely grow beyond the scale of a household workshop. Allowing private hiring was therefore an indispensable step towards the emergence of a genuine private economy.

The transition was also remarkably circuitous. From the first possibility of relaxing the prohibition to the effective accommodation of private employment took close to a decade. Change often advanced not through a decisive repudiation of the old rules, but through experiments that were tolerated, questions that were deliberately left unanswered, restrictive instructions that were interpreted flexibly, and prohibitions that remained formally in place but were no longer enforced. Some of the officials who preserved room for reform were working with directives that appeared to leave them very little room at all. Their ability to postpone decisions, omit contentious language, and prevent premature suppression shows how difficult reform was—and how much it depended not only on pronouncements from the top, but also on local initiative, bureaucratic discretion, and the drafting of policy documents.

The following article was written by Zhao Shukai (赵树凯; b. 1959), a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the Research Centre of Rural Economy under the Ministry of Agriculture and Rural Affairs). Starting in 1990, he served at the Development Research Centre of the State Council, China’s government cabinet, holding roles including Director General of the Rural Department’s Organisation Research Office and Director General of the Information Centre.

Zhao published it on 26 August 2024 on the WeChat blog 当代农政 Contemporary Agricultural Policies.

— Yuxuan Jia

邓小平与雇工政策

Deng Xiaoping and Hired Labour

The dispute over whether private hiring of labour should be allowed began in the rural areas. Its implications, however, extended far beyond agriculture. The debate challenged some of the deepest assumptions underpinning the planned economy and laid the groundwork for a market economy.

During the agricultural collectivisation campaigns of the 1950s, the Chinese central leadership expressly prohibited the private hiring of labour. The reasoning was simple and uncompromising: according to classical Marxist theory, hired labour entailed exploitation, and exploitation had no place in a socialist society. In May 1963, the central authorities’s Decision of the CPC Central Committee on Certain Questions Concerning Current Rural Work (Draft) cited “the exploitation of hired labour” as one manifestation of “a serious and acute class struggle emerging in Chinese society.”1

The 1980 No. 75 Document — Notice of the CPC Central Committee on Issuing Several Questions Concerning the Further Strengthening and Improvement of the Agricultural Production Responsibility System — allowed some impoverished areas to introduce household-based contracting, but continued to stipulate that households were “not permitted to hire workers to conduct their operations.”

Once household contracting began to spread, however, farmers quickly started hiring labour. One prominent case was Chen Zhixiong, a farmer in Guangdong who had contracted a collectively owned fishpond.

In May 1981, People’s Daily organised a public discussion of the issue. In January 1982, Hu Qiaomu wrote to Ren Zhongyi, Guangdong’s first Party secretary, with copies sent to General Secretary Hu Yaobang and Vice Premier Wan Li. Hu wrote:

“This matter departs from the socialist system. Clear rules are needed to stop and correct it, and the case should be circulated throughout the province. Since it concerns the overall direction of the rural social system, I ask the provincial Party committee to consider it.”

After reading Hu Qiaomu’s comments, Wan Li wrote a different instruction:

“This matter should be investigated and studied. We must consider how to protect farmers’ enthusiasm for developing the commodity economy. We should not lightly force it into old and rigid frameworks.”

On 22 April, the Guangdong Provincial Party Committee submitted its Report on Chen Zhixiong’s Contracting Operations to the State Agricultural Commission. The report endorsed Chen’s contracting arrangements and business model, concluding that “in terms of economic returns, they are better than the previous model of collective management where everyone eats from the same big pot.”2

The exchange provides a glimpse of how local policy autonomy emerged.

On 31 December 1982, the Politburo discussed the draft of what would become the 1983 No. 1 Central Document. The debate focused largely on hired labour, which had already divided the Secretariat at an earlier meeting.

The theoretical question was whether the hiring of workers could be compatible with socialism. The policy question was whether it should be restricted and, if so, how many employees an individual business should be allowed to hire.

One side argued, drawing on Karl Marx, that hiring more than eight workers marked the point at which an employer became an exploiter. What the revolutionary forebears had said, this group maintained, remained valid and should be upheld.

Yuxuan’s note:

This is the passage that is often cited as support for the claim that employing more than eight workers makes one a capitalist:

If this labourer were in possession of his own means of production, and were satisfied to live as a labourer, he need not work beyond the time necessary for the reproduction of his means of subsistence, say 8 hours a day. He would, besides, only require the means of production sufficient for 8 working-hours. The capitalist, on the other hand, who makes him do, besides these 8 hours, say 4 hours’ surplus-labour, requires an additional sum of money for furnishing the additional means of production. On our supposition, however, he would have to employ two labourers in order to live, on the surplus-value appropriated daily, as well as, and no better than a labourer, i.e., to be able to satisfy his necessary wants. In this case the mere maintenance of life would be the end of his production, not the increase of wealth; but this latter is implied in capitalist production. That he may live only twice as well as an ordinary labourer, and besides turn half of the surplus-value produced into capital, he would have to raise, with the number of labourers, the minimum of the capital advanced 8 times.

—Karl Marx, Capital, Volume I, Chapter 11 (“Rate and Mass of Surplus-Value”)

The opposing view was that the numerical example in Marx was an “arbitrary assumption.” Policy should not begin with abstract categories but with the realities of economic life.

Wan Li said:

“Some people say that hiring workers is exploitation. But when we export labour services, are we not deliberately sending Chinese workers abroad to be exploited by foreign capitalists? What we are trying to do is make a country with an underdeveloped commodity economy prosperous as quickly as possible.”3

The Politburo ultimately accepted the Secretariat’s recommendation: observe the phenomenon for the time being and keep it out of the newspapers. For private industrial businesses, the limit was two or three apprentices and three or four helpers. Where businesses employed more helpers than the rules allowed, the authorities proposed the “three noes”: such practices should not be encouraged, should not be publicly promoted, and should not be hastily banned.4

The question resurfaced at the end of 1983, during the final stages of drafting the 1984 No. 1 Central Document. On 9 December, Deng Xiaoping instructed his secretary to telephone Hu Qili, then a member of the Central Secretariat, and convey his view on hired labour in the rural areas:

“I said at the Politburo meeting that we should let it take its natural course and look at it again in two years.”5

On 29 December, Chen Yun wrote two comments on the draft submitted for approval.

“First: We can continue to observe the policy on hired labour for a few more years. Even if some problems emerge, there is nothing to fear.

Second: We must proceed cautiously when Party members hire workers. This is a matter of principle and requires full discussion. In his report on the revision of the Party Constitution at the Eighth Party Congress in 1956, Comrade Xiaoping said: “A Party member must be a person who works and does not exploit the labour of others.” Although circumstances have changed considerably since then, the basic requirements for Party members should still be upheld.”6

Chen Yun stressed that Party members who hired labour should be subject to special rules rather than treated in the same way as ordinary citizens.

The draft had already been circulated to and approved by Deng Xiaoping, Chen Yun, Hu Yaobang, Zhao Ziyang, and other senior leaders when Wan Li added a sentence to the section on rural hired labour: “All the above policies apply equally to Communist Party members.” After the document was returned to the Central Rural Policy Research Office, Du Runsheng removed Wan Li’s addition. Only then was the document formally approved by the Central Committee.7

During the ensuing period of “observing for a few more years,” the hiring of labour continued to expand, as did the policy debate. Views to suppress or even prohibit the practice never disappeared. Among the officials directly responsible, however, the inclination was to avoid imposing restrictions.

On 24 November 1985, Deng Xiaoping discussed the matter with Bo Yibo. Bo described three kinds of rural Party cadres: those who led the community towards common prosperity; those who took the lead in becoming prosperous as individuals; and those who enriched themselves through privilege.

Some members of the third group operated businesses with hired workers, received special support from higher authorities, occupied public resources, and obtained large amounts of state credit. Deng responded:

“The third category must be controlled. Progressive taxes can be imposed, and lending policies should also differentiate between them. As for hired labour, I said it was necessary to observe it for two or three years. It has now been almost seven years since the Third Plenum — only two months short.”

“But the latter category does need to be brought under control. The time has come to do so. They are using state resources and state loans; we cannot simply leave them alone. In the future, they should still be guided towards the collective economy. Ultimately, they should be guided toward the collective economy.”8

“The third category should be controlled. It could be subjected to progressive tax rates of 80 or 90 per cent, as well as differentiated access to credit.”9

The issue was subsequently placed on the agenda of the Central Rural Work Conference.

On 6 December, Du Runsheng chaired a meeting of the document-drafting group to consider how to implement Deng’s instructions to “bring the practice under control” and “guide it towards the collective economy.”

The debate was intense, and the group could not reach a consensus. Du finally concluded:

“The document should avoid the issue of hired labor, because it cannot be explained clearly in one or two sentences.”10

Deng’s instruction to impose control was therefore not translated into formal policy. Du Runsheng’s handling of the issue had a major influence on the subsequent course of rural policy reform.

After the 1986 No. 1 Central Document was issued, the Central Rural Policy Research Office under the Secretariat of the CPC Central Committee and other relevant agencies organised a variety of studies devoted to hired labour and considered several possible policy approaches. When Du oversaw the drafting of the 1987 rural work document, he again chose not to address the issue directly.

This episode illustrates the influence that policy research and drafting bodies could exercise when senior leaders were divided.

In August 1989, the Central Committee issued the Notice on Strengthening Party Building. The document stated that Communist Party members were not allowed to hire labour, and that Party members who privately hired labour were to be cleared out.

More than three decades later, the notice has not been formally repealed. Nor has it been enforced.

Subscribe now

Pekingnology

China’s Rural Reform: A History Not Designed but Discovered

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

9 months ago · 31 likes · 1 comment · Yuxuan JIA and YIRUI LI

Pekingnology

Zhao Shukai: Pluralism powered China’s rural reform

Zhao Shukai (赵树凯; b. 1959) is a Chinese official of rural policy and governance. From 1982 to 1989, he worked at the Rural Policy Research Office of the Communist Party of China Central Committee’s Secretariat (later reorganised as the State Council’s Rural Development Research Centre and subsequently the…

Read more

10 months ago · 17 likes · 2 comments · Yuxuan JIA and WEI Lai

Under the changes, Feishu’s product team will join a newly consolidated Doubao product group, according to Chinese media reports citing a leaked internal letter.

Feishu’s marketing, sales, and customer service operations — collectively known as its go-to-market team — will be merged with those of Volcano Engine to form a new division called the Creativity Service Platform.

ByteDance told Chinese media that existing Feishu products and services would remain unchanged, while Feishu and Doubao would work more closely on workplace productivity tools. An enterprise version of Doubao, developed with the involvement of Feishu’s product team, is already being tested by selected Feishu customers.

The restructuring brings together three complementary parts of ByteDance’s AI business: a mass-market assistant, a workplace platform, and a cloud-services operation.

Doubao, whose name means “bean bun” in Chinese, is ByteDance’s proprietary general-purpose AI assistant, launched in August 2023. It had 382 million monthly active users in June, more than double the 167 million recorded by second-ranked Alibaba’s Qwen app, according to mobile analytics company QuestMobile. This makes Doubao by far China’s most widely used standalone AI application.

Feishu, known as Lark outside China, provides the workplace layer through which ByteDance can bring those AI capabilities into companies. The platform combines messaging, video meetings, cloud documents, databases, and workflow tools, and counts Chinese technology and electric-vehicle groups including Xiaomi, Li Auto, Nio, and Xpeng among its customers. Feishu said in July 2025 that its Base database product was approaching 10 million monthly active users.

Volcano Engine provides the computing and commercial backbone for ByteDance’s enterprise AI push, giving corporate customers cloud-based access to Doubao models through model-as-a-service (MaaS). Launched in 2020, the cloud unit accounted for 49.5% of token usage in China’s enterprise public-cloud MaaS market in 2025, ahead of Alibaba Cloud and Baidu AI Cloud, according to research firm IDC.

Enterprise AI push

The restructuring follows a broader shift in ByteDance’s priorities. ByteDance Chief Executive said at a Volcano Engine conference in June that advancing AI was the company’s most important task, adding that the cloud unit’s MaaS business was becoming a foundational part of ByteDance.

The overhaul gives that strategy a more direct route to revenue by embedding Doubao’s AI capabilities in Feishu’s workplace tools and selling the resulting model access, computing capacity and enterprise services through Volcano Engine.

It also formalises an integration that had been taking shape for several years. Feishu unveiled its Smart Partner, an AI assistant, in November 2023. It followed in August 2024 with a major upgrade to Aily, an enterprise platform for building customised AI agents and applications using a company’s own knowledge, data, and business processes. By June 2026, more than 90% of the new customers Feishu had signed since the start of the year had also purchased its AI capabilities, Chinese media reported, citing people familiar with the matter.

ByteDance is not alone in seeking to connect AI models, cloud-based capabilities and workplace software. In early July, Alibaba began consolidating several previously separate workplace-agent tools into QwenWork, which combines its Qwen AI models with cloud-based data and task-management functions and deep integration with DingTalk, its workplace collaboration platform.

Tencent is building a similar product, but taking a different, and potentially more formidable, route. Tencent already operates WeChat, China’s near-ubiquitous messaging and social platform, giving it a consumer distribution channel that neither company can easily replicate.

The company launched WorkBuddy, its desktop AI agent for workplace tasks, in March. It subsequently connected the service to WeChat and, at the end of that month, launched a WeChat mini-program that allows users either to run tasks in the cloud or remotely direct WorkBuddy on their computers.

On 5 June, Tencent Cloud unveiled WorkBuddy Enterprise and a broader suite of productivity agent tools designed for corporate users. On 18 July, Tencent launched WorkBuddy as a standalone mobile app.

According to Analysys’s report on China’s office-agent market in Q2 2026, WorkBuddy recorded 20.97 million visits on personal computers in June, more than the second- and third-ranked services combined. Tencent said in its first-quarter results that WorkBuddy had become China’s most widely used productivity AI agent service by daily active accounts.

Subscribe now

Pekingnology

China’s Spring Festival AI war

In China’s Lunar New Year, luck is delivered in red envelopes, traditionally a small packet of cash, handed over as a blessing made tangible…

Read more

6 months ago · 23 likes · 1 comment · Yuxuan JIA and Zichen Wang

Among his proposals are transferring RMB 20 trillion in publicly traded state-owned capital to social security funds, using more than half of macroeconomic stimulus spending to support consumption, substantially raising pensions under the basic scheme for urban and rural residents, addressing the urban–rural housing mismatch faced by migrant workers, and encouraging higher dividend payouts by both state-owned and private companies.

He argues that China’s excessive saving is closely linked to income inequality and calls for higher personal income and property taxes on high-income groups, with the aim of reducing the Gini coefficient to below 0.4.

Liu also challenges several widely held assumptions: that China’s investment-led growth model can continue largely unchanged, that raising the incomes of low- and middle-income groups is merely a social welfare issue, and that short-term stimulus can substitute for structural reform.

Liu Shijin is a former Vice President (Vice Minister) of the Development Research Centre (DRC), a comprehensive policy research and consulting institution directly under the State Council, the central government of the People’s Republic of China. He was also Deputy Director of the Economic Committee of the 13th National Committee of the Chinese People’s Political Consultative Conference (2018-2023).

— Yuxuan Jia

Liu made the speech on 27 June 2026 at a quarterly forum of China Macroeconomy Forum (CMF), a think tank affiliated with the Renmin Univerisity of China. A consolidated transcript was published on CMF’s official WeChat blog on 6 July.

刘世锦：深化储蓄-消费形成机制改革，尽快扭转供强需弱的不平衡状态

Liu Shijin: Deepen Reform of the Mechanisms Governing Saving and Consumption and Reverse the Imbalance Between Strong Supply and Weak Demand as Soon as Possible

At present, the imbalance between strong supply and weak demand poses a major challenge to maintaining stable growth in China. The 2025 Central Economic Work Conference clearly identified the pronounced imbalance between strong domestic supply and weak demand, as well as the shortage of effective demand, as prominent problems facing the economy.

This imbalance stems from a distorted growth model underpinned by a range of institutional and structural factors. Reversing it requires policy to focus on reforming the mechanisms governing saving and consumption, substantially reducing the saving rate, and correspondingly expanding final demand.

I. The Manifestations and Underlying Causes of Strong Supply and Weak Demand

The term “strong supply and weak demand” describes several seemingly contradictory phenomena frequently observed in the economy, sometimes referred to collectively as K-shaped development.

On the one hand, China is experiencing a surge in innovation, with significant advances in artificial intelligence, robotics, and green technology. On the other hand, economic growth remains under sustained pressure, with nominal growth falling below real growth.

Exports have expanded at an exceptional pace, and China’s merchandise trade surplus reached a record of nearly US$1.2 trillion last year. At the same time, the weakness of domestic demand has continued to intensify. Investment contracted over the whole of last year. In the first five months of this year, cumulative investment fell by 4.1 percent year-on-year, while total retail sales of consumer goods declined by 0.6 percent year-on-year in May.

China’s manufacturing sector and industrial chains are globally competitive, yet many industries face serious overcapacity, while the needs of certain social groups remain unmet.

Strong supply and weak demand are not the result of short-term economic fluctuations. They arise from a series of institutional and structural factors. These include GDP-oriented competition among local governments; the dominant role of state-owned enterprises (SOEs) and the relatively high share of state-owned capital in foundational and strategic industries; a macroeconomic policy that has sought to stabilise growth primarily through investment, particularly infrastructure investment; and an economic structure centred on manufacturing and exports.

Together, these factors have created a tilted growth model characterised by high saving, high investment, and high exports — a model that emphasises production while paying insufficient attention to consumption.

At the beginning of reform and opening up, China was a shortage economy. For a long period, when supply remained insufficient, this growth model was well suited to the country’s circumstances and offered clear advantages. Even after the economy shifted from high-speed to medium-speed growth and from supply constraints to demand constraints, the model remained highly persistent and resilient, continuing to function as long as there was still scope to expand investment.

Nevertheless, this growth model has limits. Those limits are reached when investment can no longer be expanded and begins to contract. China has now reached this point.

II. Insufficient Final Demand Has for Some Time Been the Direct Cause of Subdued Prices and Slower Growth

The relationship between strong supply, weak demand, and economic growth can be examined through the concepts of the “height” and “breadth” of growth.

The “height” of growth refers to the expansion of an economy’s growth potential through productivity gains driven by technological innovation, improved management, institutional reform, greater openness, and other factors.

The “breadth” of growth refers to the extent to which effective demand for existing productive capacity is generated across different segments of society — for example, across ten population deciles.

The height determines the potential rate of growth, while the breadth determines the rate an economy actually achieves.

Strong supply raises the height of growth, while weak demand means that its breadth has failed to expand in step. Raising the height of growth cannot substitute for broadening it and may, in some cases, even undermine that breadth. Artificial intelligence, for example, may displace existing jobs and widen income inequality.

This helps explain an apparent puzzle: why can innovation-driven development and emerging industries perform well while the overall economy remains under considerable pressure?

At this point, it is necessary to introduce the concept of final demand.

Final demand refers to the portion of GDP that does not enter another production process. It includes all consumption, together with non-productive or consumption-oriented investment, principally investment in housing, infrastructure, and other livelihood-related service sectors.

Final demand represents the true end product of economic activities. It is the part of GDP that directly serves the ultimate purpose of economic activity: meeting people’s aspirations for a better life.

GDP minus final demand consists of productive investment and net exports. Both are instrumental and ultimately serve the expansion of final demand.

For many years, consumption as a share of China’s GDP has remained approximately 20 percentage points below the international average, reflecting a structural distortion. However, the prolonged rapid expansion of real estate and infrastructure investment effectively concealed this distortion.

After 2022, the property sector contracted sharply and infrastructure investment also lost momentum, exposing the long-concealed weakness in consumption and making it a major constraint on final demand.

In recent years, final demand has slowed markedly and declined as a share of the economy. This has pushed down the GDP deflator, causing nominal growth to fall below real growth. Together, weak final demand and excessive investment — particularly excessive productive investment — have reduced capacity utilisation. When capacity utilisation falls too low, returns on investment weaken, discouraging investors from committing additional capital. This is the principal reason for the current decline in investment.

The growth of industrial value added has remained relatively solid, primarily supported by exports and, more recently, by rising capital expenditure on artificial intelligence.

In the national accounts, net exports are counted as part of national saving and correspond to lower domestic consumption. Strong exports are therefore one cause of weak domestic consumption, yet this relationship has not received sufficient attention.

III. Focus Reform on the Mechanisms Governing Saving and Consumption

Behind the imbalance between strong supply and weak demand lies a pattern of national income allocation marked by high saving and low consumption.

China’s saving rate is exceptionally high by international standards. Over the past decade or so, it at one point peaked at over 50 per cent and, despite declining somewhat in recent years, still stands above 42 per cent — far higher than the global average of around 25 per cent.

National saving in China comes mainly from the corporate and household sectors. Government saving has gradually declined and has been negative in recent years.

According to China’s flow-of-funds accounts, GDP in 2022 was approximately RMB 120 trillion and total saving stood at RMB 55 trillion, implying a national saving rate of 46 per cent and a corresponding consumption rate of about 54 per cent. Among those, corporate saving amounted to RMB 27 trillion and household saving to RMB 27.6 trillion, with each accounting for roughly half of the total.

At 22.5 per cent of GDP, corporate saving was notably high by international standards. This largely reflects the relatively small share of corporate returns on capital paid out as dividends. In 2022, corporate dividends accounted for only 10.2 per cent of Chinese households’ property income, compared with a global average of 55.7 per cent.

Historically, periods of rapid growth generated strong demand for investment funds. With external financing difficult and expensive to obtain, companies tended to rely on internal financing and retain a larger share of their earnings.

More importantly, state-owned capital accounts for a substantial proportion of total corporate capital in China, and state-owned enterprises for many years paid either no dividends or only limited dividends.

The establishment of the state-owned capital operating-budget system in 2007 changed this situation to some extent. As SOE reform deepened and the governance system for state-owned capital improved, dividend payments by state-owned capital in publicly traded companies increased. Nevertheless, many SOEs that are not publicly traded continue to pay relatively few dividends.

The corporate sector receives an excessively large share of society’s disposable income, much of which is retained as corporate saving. This produces a high corporate saving rate for society as a whole. Corporate saving ultimately has only one use: investment.

The household saving rate is also high. But a closer look at the income distribution shows that this is driven largely by the substantial savings of high-income households.

According to China Merchants Bank’s 2024 annual report, its Golden Sunflower customers — those with average daily total assets of RMB 500,000 or more — accounted for only 2.49 percent of its retail customers but held 81.90 per cent of customer assets.

Compared with 2023, Golden Sunflower customers accounted for 87.48 per cent of the net increase in customer assets. This indicates that a small group of high-net-worth customers hold the overwhelming majority of bank savings assets and that this concentration is continuing to increase. Data from other banks show similar patterns.

Some people argue that because China adds more than RMB 10 trillion in new savings each year, it should be sufficient simply to encourage households to convert some of these savings into consumption.

This view considers only the aggregate amount and ignores its distribution. It fails to recognise that high saving and low consumption are directly related to excessive income inequality within the household sector. It is therefore highly misleading, both theoretically and as a basis for policy.

IV. How to Reform the Saving–Consumption Mechanism and Reorient Policy

Investment has been contracting for more than a year, and in May total retail sales of consumer goods also fell year-on-year. This combination has not occurred since the start of reform and opening up and warrants the utmost attention.

The operation and development of the Chinese economy have once again reached an important turning point.

The most urgent task is to revive investment and restore positive investment growth. The way to achieve this is to substantially reduce the saving rate and correspondingly increase final demand.

This differs from the views of some observers, but it is crucial to understanding how the Chinese economy currently operates.

National income is divided between saving and consumption. If a high saving rate cannot be reduced, a low consumption rate cannot be increased. This is a simple but binding constraint.

Only by lowering the saving rate and expanding final demand can capacity utilisation be increased. Once capacity utilisation reaches a sufficiently high level, genuine investment demand will emerge. This is precisely the kind of investment repeatedly emphasised by the central authorities: investment backed by real demand and capable of generating returns.

If the saving rate is not reduced and final demand is not expanded — or is instead allowed to contract — while investment is forcibly increased through traditional methods, the imbalance between strong supply and weak demand will become even more severe, pushing the Chinese economy further in the wrong direction.

The key to correcting this imbalance lies in reforming the mechanisms through which saving and consumption are generated and converting excessive saving into effective consumption. Several priority areas for reform and corresponding quantitative policy objectives can be proposed.

First, reduce the overall social saving rate to below 40 per cent within a relatively short period, such as three years, and lower the corporate saving rate to below 20 per cent. Part of corporate saving should be converted into consumption, raising the overall social consumption rate to above 60 per cent. This would also be a concrete step toward achieving the 15th Five-Year Plan’s objective of increasing the household consumption rate.

A direct approach is to increase corporate dividend payout rates. Transferring a substantial proportion of state-owned capital to social security funds would, in effect, constitute a specific form of dividend distribution by state-owned capital.

Some have raised concerns about transferring state-owned capital. They argue that relatively few state-owned assets are sufficiently profitable to generate dividends; that state-owned capital in certain sectors carry heavy debt burdens, making their net asset value difficult to assess; and that state-owned capital is controlled by different departments and institutions, which may resist transferring it to social security funds.

These problems are not, in fact, difficult to resolve.

Put simply, transfers could begin with state-owned capital that is already publicly traded. The total market capitalisation of China’s A-share market is currently approximately RMB 120 trillion. A preliminary estimate suggests that state-owned capital accounts for approximately 40 per cent of this amount. A specified amount, such as RMB 20 trillion, could be transferred to social security funds.

The transfer would not alter the existing departmental or institutional affiliation of the state-owned capital. Instead, new pension funds could be established under the relevant departments and institutions to manage the transferred state-owned capital, thereby creating new operational functions and objectives.

From the perspective of the capital market, this would simply convert part of the state-owned capital into pension assets and increase the supply of long-term, patient capital.

Private enterprises should also be encouraged to increase dividend payments. Measures such as improving corporate governance, strengthening the protection of minority investors, and promoting long-term value investing should encourage companies to return more profits to shareholders through dividends, which shareholders could then convert into consumption expenditure.

Second, steadily advance tax reform, contain increases in the Gini coefficient, and gradually reduce the coefficient to 0.4 or below. This would increase society’s average propensity to consume and reduce the household saving rate accordingly.

International experience shows that economies that successfully moved from middle-income to high-income status generally had Gini coefficients of between 0.3 and 0.4, with some recording even lower levels.

A relatively balanced distribution of income allows a larger share of national income to be converted into final-demand expenditure and eases the pressure created by the release of large amounts of productive capacity accumulated during periods of rapid growth.

China’s tax system remains dominated by indirect taxes, while direct taxes such as personal income tax and property tax account for a relatively small share of total taxation.

To reform the saving–consumption mechanism, China should appropriately increase taxes on high-income groups, particularly through personal income and property taxation, thereby raising government revenue.

Some are concerned that moderately increasing taxes on high-income groups would undermine the confidence and expectations of those groups. In reality, the negative effect would be limited. It will instead strengthen the government’s incentive to protect property rights.

One argument holds that “the boundary of the capacity to tax” determines “the boundary of the protection of property rights,” because protecting property rights also means protecting the tax base.

Third, shift the focus of macroeconomic regulation and government functions at all levels from the supply side to the consumption side.

More than 50 per cent of macroeconomic stimulus expenditure should be used to address the imbalance between strong supply and weak demand, with a particular focus on building a stronger social security system for low- and middle-income groups.

There remains some confusion in society about how consumption drives economic growth.

Reforming the saving–consumption mechanism and increasing the income and consumption of low- and middle-income groups would first create new space in the consumer market. Excess capacity would then cease to be excessive and might even give way to capacity shortages, thereby stimulating new and effective investment.

Second, and more importantly, expanding consumption would allow more people, especially low- and middle-income groups, to share in the benefits of reform, opening up, and development. It would also strengthen their sense that their aspirations for a better life are being fulfilled.

Equally important, the additional consumption generated by these groups would be concentrated in developmental consumption. In effect, this is investment in human capital and a concrete expression of “investing in people.” It would provide the most important form of human-capital support for innovation-driven development.

Subsidies for the purchase of consumer goods have only a limited effect on the income and consumption of low- and middle-income groups. Policy should instead focus on two particularly prominent weaknesses: housing and social security.

A migrant worker may own a two-storey, 200-square-metre house in the countryside but return to it for only a few days each year, while spending the rest of the year crowded into a ten-square-metre basement room in a city.

According to statistical measures, this worker would have access to 210 square metres of housing. In reality, however, the worker’s living conditions are extremely difficult. Migrant workers and similar groups face a major structural mismatch between their housing resources in rural and urban areas.

Housing-support policies, together with a reorganisation and reallocation of urban and rural resources, should be used to accelerate efforts to address the housing difficulties faced by migrant workers and other new urban residents. This would also generate genuine demand, helping to stabilise the property sector and restore normal growth.

Through the transfer of state-owned assets to social security funds, fiscal subsidies, short-term stimulus funding, and improvements to social security contribution arrangements, China should seek, during the 15th Five-Year Plan period, to achieve a substantial increase in the per capita pension benefits provided by the basic pension scheme for urban and rural residents.

This scheme covers the largest number of low- and middle-income people. The increase in benefits should be aligned with the current need to expand final demand and stabilise macroeconomic growth.

Among the many consumption-stimulus policies, this reform would rank among the most effective in promoting economic growth and would play an irreplaceable role in stabilising growth.

Some people argue that increasing the pension income of elderly rural residents would have little effect on consumption. This may appear plausible at first, but it is fundamentally mistaken.

In reality, the excessively low level of pension benefits under the basic pension-insurance scheme for urban and rural residents constrains the release of society’s consumption potential in several ways.

First, it directly limits the consumption capacity of 170 million pension recipients.

Second, inadequate pensions of rural parents require their children to make larger intergenerational transfers, thereby crowding out their own consumption.

Third, it makes the 370 million people currently contributing to the scheme less secure about their future, prompting them to increase precautionary saving out of necessity.

The basic pension-insurance scheme for urban and rural residents covers 550 million people, 95 per cent of whom are rural residents. This represents approximately half of all participants in China’s pension system. They constitute both the lowest-income and the largest segment of the national pension system, while also having the highest propensity to consume.

From this perspective, the shortfall in pension income among those covered by the scheme constrains consumption growth among approximately half of China’s population. Its broad economic impact should not be overlooked.

V. Several Misunderstandings That Must Be Clarified to Reverse Strong Supply and Weak Demand as Soon as Possible

One view holds that China’s growth model, which emphasises production while giving insufficient attention to consumption, has existed for a long time. Although China has often described its development as unbalanced, uncoordinated, and unsustainable, the country has nevertheless continued to grow for many years, suggesting that this model may be able to continue in the future.

In the past, however, China remained a shortage economy for a long period. Even after the economy became constrained by demand rather than supply, there was still some room to expand investment.

Those conditions no longer exist. Investment has been contracting for an extended period, indicating that it is becoming increasingly difficult for the existing growth model to continue.

Another view holds that expanding consumption and increasing the income of low- and middle-income groups are livelihood issues that should be pursued only within available means and should not be rushed. This view had some validity for a considerable period in the past.

Under current conditions, however, unless the income of low- and middle-income groups is increased and final demand expanded, capacity utilisation cannot rise, investment growth will be difficult to restore, and actual economic growth will fall significantly below potential growth.

Increasing the income and consumption of low- and middle-income groups is therefore not merely a livelihood issue. More importantly, it is a matter of stabilsing macroeconomic growth.

In the past, when growth weakened, China focused on investment. At the present stage, it must focus on consumption. The underlying logic is the same, but circumstances have changed, and the areas in which growth potential can be found are now different.

Another view holds that although structural and institutional reform is important, “distant water cannot extinguish an immediate fire,” and short-term policy should therefore continue to rely on macroeconomic stimulus.

Structural reform and macroeconomic policy each have their own role. There is no need to choose one while excluding the other. Combining the two will often produce better results. In fact, some structural reforms, such as transferring state-owned assets to social security funds, can produce results in the short term.

Other reforms, such as reducing the Gini coefficient, may appear slow in the short term once policy goals have been set and implementation has begun, but can prove relatively fast over a longer time horizon.

Using short-term stimulus to replace or avoid structural reforms that could in fact be advanced will only prolong the problem and raise the eventual cost.

Japan was the first country to introduce quantitative easing, zero interest rates, and negative interest rates. More than two decades later, however, it has yet to decisively break free from deflation or restore economic vitality.

Japan’s experience has been widely studied in recent years. China should draw lessons from it rather than repeat the same mistakes.

The central authorities have made a clear assessment of the imbalance between strong supply and weak demand, and the path toward resolving the problem is also clear.

International experience shows that economies that fall into the “middle-income trap” generally suffer from weak innovation and insufficient industrial competitiveness. China’s present weakness, by contrast, lies in strong supply and weak demand.

In logical terms, demand-side problems should be relatively easier to solve than supply-side problems. That does not mean, however, that resolving them will be easy.

Strong supply and weak demand are deeply intertwined with a series of institutional and structural factors. Moving beyond the existing growth framework, shifting from a model driven primarily by investment and exports to one driven primarily by innovation and consumption, and accelerating the establishment of a new growth framework will require a new round of emancipation of the mind and the building of a new social consensus.

China must overcome path dependence and address the challenges of changing mindsets, interest relations, policy tools, and working methods. Only in this way can strong supply help generate strong demand, creating a new cycle of dynamic mutual reinforcement between supply and demand, stabilising the overall macroeconomic environment, and generating sustained momentum for long-term and stable development.

Subscribe now

Between July 2 and July 13, material related to the ruling appeared in an estimated 220,000 posts, reposts, and comments on the Chinese internet, according to a Chinese media-monitoring analysis. Topic pages on Weibo, China’s equivalent of Twitter, recorded more than 400 million aggregate views.

As criticism mounted, the stories surrounding Louis Vuitton became much more dramatic — and more bizarre. Widely shared posts claimed that the company had sued a Chinese noodle shop over its décor and a residential compound for its building design. Both accounts were later shown to be false.

More recently, social-media accounts claimed that celebrities had boycotted a Louis Vuitton fashion show in Shanghai. That story is also reportedly under question.

Louis Vuitton, meanwhile, has remained conspicuously silent throughout the controversy, leaving one of the world’s most recognisable luxury brands with little visible role in shaping the narrative around its conduct in a crucial market. The world’s second largest economy remains a “cornerstone of luxury market growth” this year, according to Bain.

A Trademark Dispute Turned into Cultural Ownership Debate

At the centre of the case was Molly Tea’s use of a black four-petal emblem across its shopfronts, drinks packaging, online accounts, and promotional materials.

On 19 June, a court in Suzhou found that Molly Tea had infringed seven registered flower-device trademarks owned by Louis Vuitton. It ordered the chain to pay 10 million yuan in damages and a further 300,000 yuan in enforcement costs, and to publish statements on six online platforms. Molly Tea has said it will appeal.

The court said it considered whether the resemblance between Louis Vuitton’s classic four-petal motif and Molly Tea’s emblem could lead consumers to believe that the two companies had entered into a licensing agreement, collaboration, or other commercial relationship.

Much of the online response, however, focused on whether Louis Vuitton was using trademark law to claim control over motifs from China’s visual heritage. Users circulated side-by-side comparisons of its flower designs with baoxianghua, a Buddhist floral motif, as well as traditional Chinese patterns, ancient musical instruments, garden windows, and temple decorations.

The comparisons also became a source of ridicule. Images circulated online comparing Louis Vuitton’s Monogram flowers with ventilation grilles and tiles found in older public toilets, giving rise to the nickname “toilet bag”. A separate video purported to show the letters “LV” on a public toilet in Henan, suggesting that they stood for “lavatory”. Local authorities later said the footage had been digitally altered, and that the person responsible was given an administrative penalty.

The reaction reflected a broader shift in Chinese public sentiment in recent years, as growing cultural confidence has increasingly been accompanied by a more assertive sense of national identity. In disputes involving foreign brands and traditional symbols, questions of trademark ownership can quickly become entangled with concerns over cultural appropriation, national dignity, and the right to define China’s own heritage.

Many users accused Louis Vuitton of “claiming Chinese traditional motifs as its own, then suing a domestic brand.” Others said that it was “bullying Chinese ancients for failing to register trademarks.”

In a commentary, the People’s Daily took a more moderate position. It argued that companies may build trademarks from traditional motifs, but that such rights should have clear limits, and called for more precise boundaries on commercial claims to traditional Chinese patterns and aesthetic elements.

Rumours fill the information gap

The controversy soon spread to businesses and disputes with little or no connection to the Molly Tea case.

Reports that Louis Vuitton had filed for judicial review of a separate decision by China’s intellectual-property regulator were recast online as the company “suing the Chinese government” and challenging China’s authority over its own cultural symbols. Such court challenges, however, are a standard legal procedure for companies seeking to overturn administrative trademark decisions in China.

In Nanjing, a lawsuit over another restaurant’s use of four-petal motifs in its decorations and plate design was misreported as action against a seller of duck-blood noodles, a popular local staple.

In Xiamen, a pending case against a shop accused of selling counterfeit Louis Vuitton goods — and against property managers accused of failing to supervise it — became a claim that the company had sued a residential developer over four-petal carvings on its buildings.

A Louis Vuitton event in Shanghai on 25 July, described online as a major runway show, was similarly said to have been shunned by Chinese celebrities seeking to avoid association with the brand. However, the Chinese edition of Fashion Network, citing an industry source, said there had been no major public show, only a routine private presentation for clients.

Louis Vuitton has made no public statement on the Molly Tea case or the wider backlash. Its official accounts on Weibo and Xiaohongshu — the latter a popular lifestyle app — had previously been updated regularly but have posted nothing since 1 July. The prolonged silence has further raised questions about the brand’s handling of the controversy.

The backlash in China comes as LVMH and its controlling family face scrutiny at home. In France, Le Monde has published a six-part investigation into founder Bernard Arnault’s political and media influence, as well as the unresolved question of succession within the group.

For Molly Tea, though, the episode appears to have provided an unexpected marketing windfall.

Chinese media reported that the company’s Douyin [TikTok’s Chinese sister app] account gained about 350,000 followers in the days after the ruling and that some branches experienced a surge in orders.

Subscribe now

Pekingnology

Xiaohongshu’s silence becomes part of its IPO story

In January 2025, Xiaohongshu briefly joined two internet worlds. As an American ban on TikTok loomed, hundreds of thousands of self-described “TikTok refugees” piled into RedNote, as it’s known in English. Eighteen months later, it is illustrating the opposite phenomenon. A controversy that crowds Chinese-language search results for its reported Hong Ko…

Read more

a month ago · 23 likes · 1 comment · Zichen Wang

One constraint is the country’s skewed sex ratio: among those born between 2006 and 2010, there were more than 120 boys for every 100 girls. As these cohorts enter their prime marriage and childbearing years, the imbalance is likely to further depress marriages and births. At the same time, women have come to make up a growing share of the highly educated population, reaching about 53 per cent in 2020, creating partner shortages at both ends of the educational spectrum.

Meanwhile, China’s migrant workforce is ageing, the rural labour pool that powered decades of industrialisation is nearing exhaustion, and population growth is becoming concentrated in a smaller number of productive metropolitan regions. This risks leaving transport infrastructure underused as ageing widens social-security finances.

The report is authored by Li Xunlei, Chief Economist at Zhongtai Financial International Limited; Tang Jun, an FOF fund manager at Zhongtai Asset Management; and Li Qianyun, head of Zhongtai’s fund research team. It was posted on 12 July 2026 on 李迅雷金融与投资 Finance & Investment with Li Xunlei, Li’s personal WeChat blog.

Li has kindly authorised and reviewed the translation.

—Yuxuan Jia

中国人口往何处去？（2026年版）

Where Is China’s Population Headed? (2026 Edition)

Abstract

China’s population will age rapidly between now and 2050, and this trajectory is largely unavoidable. Even a substantial rebound in fertility would do little to slow the acceleration of population aging. By 2031, people aged 65 and over are projected to account for 20 percent of the population, placing China among the world’s super-aged societies. By around 2050, that share is expected to reach 29 percent — roughly the current level in Japan, the world’s most aged society.

Gender imbalance may become an important constraint on any recovery in fertility. Among children born between 2006 and 2015, the male-to-female ratio remained above 115:100. Within five years, this imbalance will likely have a significant and lasting impact on birth rates. At the same time, women have come to make up a growing share of the highly educated population, reaching about 53 per cent in 2020, even though women account for less than half of the population overall. This implies an even sharper gender imbalance among people with lower levels of education.

Policies designed to encourage childbirth are likely to raise the number of births to some extent, but they will have little effect on the pace of population ageing in the near term. China still has a very large population and has already entered a period of accelerating ageing. Even if fertility rises, any meaningful reduction in the old-age share or the dependency ratio is unlikely to become visible until after 2050.

The decline in China’s total population is already clear. Under the baseline scenario, the population is projected to fall below 1.3 billion in 2037, below 1.2 billion around 2046, and below 1 billion around 2060. By 2100, it may fall below 500 million. The dependency ratio is projected to reach 47 per cent by 2035 and 62 per cent by 2050.

In 2025, China’s migrant workers had an average age of 43.3, and 32 per cent were aged 50 or above. The number who age out of migrant work and return to their home areas will continue to rise. The rural labour pool available for migration to cities is therefore nearing exhaustion. Future population growth in large cities will increasingly come from redistribution within the urban system and from the continued concentration of people in major metropolitan areas. Excluding natural population growth, net population inflows are increasingly concentrated in Guangdong, Zhejiang, and Jiangsu.

Among 337 Chinese cities classified as fifth-tier or above, 185 recorded a decline in their resident population between 2024 and 2025 — more than half of the total. (Because city-level data remain incomplete, these figures have not been adjusted for natural population change.) A further 65 cities had not yet released resident-population data for 2025.

A vibrant private sector and digital economy have made some regions especially attractive to migrants. Zhejiang is one of China’s most economically dynamic provinces and has recorded the fastest growth in net population inflows. Excluding natural population change, its resident population increased through net migration by 7.6 per cent over the past eight years, far faster than in any other province. The Yangtze River Delta is already relatively aged, so its population growth depends heavily on inward migration. In 2025, after adjusting for natural population decline, Zhejiang and Jiangsu recorded net interprovincial inflows of 389,000 and 227,000 people, respectively.

Major provincial capitals and metropolitan centres continue to draw population from surrounding areas. Between 2017 and 2025, the resident populations of Chengdu–Chongqing, Wuhan, Zhengzhou, Changsha, Xi’an, and Hefei each grew by more than one million. Chengdu led the group, adding 5.49 million residents. Yet every province containing these cities recorded an overall decline in resident population over the same eight-year period.

China’s high-tech industries are still expanding rapidly in both scale and revenue. The growth of high-value-added industries and the wage premium they offer draw young people and skilled workers across regional boundaries. Supporting service workers then follow, creating a complete ecosystem of population concentration. Zhejiang, Shanghai, and Jiangsu offer large numbers of jobs in advanced manufacturing, the digital economy, and services, making them the main drivers of population concentration in the Yangtze River Delta.

This is the third report in our series on births, population ageing, and internal migration in China. The first two reports were based on national and provincial demographic data released in 2022 and 2024. In recent years, however, local governments have disclosed less demographic information, making analysis and forecasting increasingly difficult. Comparing our earlier projections with subsequent outcomes, we found that we had overestimated births and underestimated the pace at which life expectancy was improving. We have therefore revised the model used in this report. Since China’s total population began to decline in 2022, we have continued to track changes in the age structure and sex composition of the population, patterns of regional migration, and the relationship between migration and industrial development.

Why Were Births in 2025 Lower Than Expected?

In Where Is China’s Population Headed? (2023 Edition), we projected China’s demographic and ageing trends on the basis of two main assumptions.

First, we used fertility data from 2018 and 2019 to estimate the underlying willingness to have children. Births in 2016 and 2017 had been temporarily elevated after the relaxation of birth restrictions released pent-up demand for second children, while pandemic controls between 2020 and 2022 may have caused some couples to postpone childbearing.

Second, we assumed that the willingness to have children would continue to decline, drawing on the experience of Japan, South Korea, and other countries. International evidence suggests that fertility generally falls as the urbanisation rate and per capita income rise. Combining these assumptions with the number and age distribution of women of childbearing age, we projected China’s future population and the pace of population ageing.

When data for 2024 became available, it became clear that our original assumptions had been too optimistic. In Where Is China’s Population Headed? (2025 Concise Edition), we therefore lowered the starting fertility-rate assumption while retaining the assumption that fertility would continue to decline. That report reached three main conclusions:

1. Annual births were projected to fall below 9 million in 2025, below 8 million in 2028, and possibly below 7 million by 2035, with the pace of decline easing over the following decade.

2. China’s total population could fall below 1.4 billion in 2027, below 1.3 billion in 2039, and below 1.2 billion in 2047.

3. Population ageing would accelerate. China was projected to become a super-aged society around 2032 and to reach Japan’s current level of ageing by 2048.

According to the National Bureau of Statistics, China recorded 7.92 million births in 2025, well below our model’s estimate of about 8.9 million. From 2022 to 2025, the model overestimated births while projecting a total population lower than the figure eventually reported. This indicates that the assumed mortality rates for older age groups were also too high. Put differently, life expectancy improved faster than the model had anticipated.

The economic consequences of China’s demographic shift may therefore be more severe than expected. Births falling below 8 million this early suggest that many people are not optimistic about their own or their families’ future prospects. Meanwhile, the fact that the total population declined more slowly than projected indicates that older people are living longer. This will accelerate population ageing and place additional pressure on public finances.

Further Refinement of the Population Projection Model

Core Assumptions: Age-Specific Fertility and Mortality Rates

Earlier versions of the model focused mainly on assumptions about the initial level of fertility and the speed at which it would decline. Age-specific mortality rates were largely projected by extending historical patterns. Recent results, however, show that changes in mortality at older ages—and thus in life expectancy—have a substantial effect on projections of both China’s total population and the share of older people. These assumptions therefore require more careful treatment.

Figure 1 provides a schematic overview of the model. It rests on two main sets of assumptions: age-specific fertility rates, including both their current levels and future paths; and age-specific mortality rates. Once these assumptions are established, the size of each age group can be projected forward over time.

For example, the number of births can be calculated by combining the age distribution of women of childbearing age with the fertility rate for each age group. Five years later, those births determine the approximate size of the population aged 0–4. Applying the assumed mortality rate then allows the model to estimate how many of that cohort will be aged 5–9 five years later, and the same process can be repeated for each successive age group. Projected sex ratios determine the future number of women of childbearing age, which in turn feeds into the next projection of births. The model thus forms a closed, internally consistent system that can be extended forward through time.

Figure 1. Schematic Framework of the Population and Age-Structure Projection Model

Using age-specific fertility rates and projecting how they will change is more appropriate than setting a single total fertility rate directly. China’s historical baby booms and decades of birth-control policies have produced a highly uneven age distribution among women of childbearing age. The number of women aged 35–49, for example, is substantially larger than the number aged 15–29. This can distort the total fertility rate and make it unusually volatile. Using it directly as an input may therefore generate larger forecasting errors.

Assumptions about fertility at different ages must also be demographically plausible. Fertility among women aged 15–24 may continue to decline, while fertility among those aged 30–49 may increase in relative terms. At the same time, the total fertility rate implied by these age-specific patterns must remain consistent with each scenario’s broader assumptions—continuing to fall in the pessimistic scenario, for instance, or gradually recovering in the optimistic scenario.

Assumptions about age-specific mortality are, in effect, assumptions about life expectancy. Life expectancy is calculated from mortality rates at each age. Unlike the average age at death, it measures how long a newborn would be expected to live if current age-specific mortality rates remained unchanged throughout that person’s life, and it is not affected by the population’s age structure. The average age at death, by contrast, is simply the average age of all those who die in a given year. It is easier to calculate but heavily influenced by demographic composition. As the baby-boomers move into old age, for example, a larger share of annual deaths will naturally occur at older ages, pushing up the average age at death.

Scenario Analysis of Population Ageing in China

To examine how the share of older people may evolve under different paths for fertility and life expectancy, this report considers three scenarios.

Pessimistic scenario: The total fertility rate continues to decline at its current pace, while life expectancy rises by about 0.2 years per year throughout the projection period.

Baseline scenario: The total fertility rate remains at roughly its current low level of 1.05 and no longer declines. Once life expectancy exceeds 80 years, its annual increase slows to about 0.1 years.

Optimistic scenario: With stronger policy support for childbearing, the total fertility rate gradually rises above 1.3, after which the pace of increase slows. Once life expectancy reaches 85 years, annual gains ease further to about 0.05 years.

The model produces the projections shown in Figure 2. Five conclusions follow.

1. Rapid population ageing before 2050 is unavoidable. Even in the optimistic scenario, a substantial rise in fertility has little effect on the share of people aged 65 and over before 2050. China’s current age structure has already determined the pace of ageing over the next two to three decades.

2. Under the baseline and pessimistic scenarios, the old-age share and dependency ratio eventually reach exceptionally high levels. By 2100, people aged 65 and over are projected to account for 41 per cent of the population in the baseline scenario and 50 per cent in the pessimistic scenario. The corresponding dependency ratios—the combined population of children and older people relative to the working-age population aged 15–64—reach 90 per cent and 120 per cent, respectively.

3. A substantial rise in fertility would significantly reduce ageing and dependency pressures after 2050. Under the optimistic scenario, both the old-age share and the dependency ratio will be markedly lower in 2100.

4. Higher fertility raises the dependency ratio in the short term. The projected dependency ratio in 2050 is actually higher in the optimistic scenario than in the other two scenarios because more births initially increase the number of dependent children before they enter the workforce.

5. A steep decline in the total population is unavoidable. China’s population falls substantially in all three scenarios, largely because of the country’s existing age structure.

Figure 2. Scenario Analysis of Population Ageing in China

More specifically, under the pessimistic scenario, the total fertility rate is projected to continue along its current downward trajectory, falling to 0.93 by 2050 and 0.87 by 2100. For comparison, South Korea—the country with the world’s lowest fertility rate—recorded a total fertility rate of 0.80 in 2025.

Figure 3. Projected Total Fertility Rate Under the Three Scenarios

Unless fertility rises substantially, annual births will fall to between 5 million and 6 million by 2050 under both the pessimistic and baseline scenarios, and to fewer than 2 million by 2100. Under the optimistic scenario, in which the total fertility rate recovers to above 1.3, annual births stand at about 7.9 million in 2050 before declining to roughly 3.5 million by 2100. Even in this more favourable scenario, the long-term fall in births remains pronounced.

Figure 4. Projected Annual Births Under the Three Scenarios

All three scenarios show a similarly rapid rise in population ageing before 2050, with only modest differences among them. Around 2030, people aged 65 and over are projected to account for 20 per cent of the population, the threshold the UN used for a super-aged society. By around 2050, the share reaches 29 per cent—approximately Japan’s current level. The three paths diverge thereafter. Under the pessimistic scenario, the old-age share continues to rise until it reaches about 50 per cent by the end of the century. Under the optimistic scenario, it peaks at around 37 per cent before falling gradually to 32 per cent in 2100.

Figure 5. Projected Share of the Population Aged 65 and Over Under the Three Scenarios

International comparisons underline the speed of China’s transition. China is projected to move from a deeply aged society to a super-aged society in only about ten years. The same transition took 36 years in Germany, 24 years in France, and 12 years in Japan. China is indeed ageing at an unusually rapid pace.

The dependency ratio rises sharply in all three scenarios until around 2060, levels off for roughly a decade, and then climbs again before peaking around 2085. Under the optimistic scenario, it reaches a high of about 91 per cent and then declines to roughly 74 per cent by 2100. Under the pessimistic scenario, it peaks at a dramatic 125 per cent and remains close to 120 per cent at the end of the century.

Figure 6. Projected Dependency Ratio Under the Three Scenarios

China’s total population declines substantially in every scenario. It is likely to fall below 1.4 billion within the next two years, below 1.3 billion around 2037, below 1.2 billion around 2050, below 1 billion around 2060, and below 800 million around 2070. By 2100, it falls to about 400 million in the baseline and pessimistic scenarios and to roughly 500 million even in the optimistic scenario.

Figure 7. Projected Total Population Under the Three Scenarios

Reflections on Raising China’s Fertility Rate

Low fertility is, to a large extent, a natural consequence of socioeconomic development. As discussed in Where Is China’s Population Headed? (2023 Edition), a clear international pattern links higher urbanisation and higher per capita income with lower fertility. It can be seen not only in affluent European countries and in Asian economies such as Japan and South Korea, but also in Muslim-majority countries such as the United Arab Emirates and Qatar. Low fertility therefore deserves serious attention and a sound policy response, but it should not be used to fuel unnecessary public anxiety.

Policies designed to encourage childbearing may be more effective in rural areas and small towns. The inverse relationship between urbanisation and fertility suggests that raising fertility may be easier in these areas, where the pressures of daily life are generally less acute than in major cities. According to China’s Seventh National Population Census in 2020, the total fertility rate was about 1.20 in urban areas and 1.54 in rural areas. Rural China’s fertility advantage came mainly from higher rates of second, third, and subsequent births. Addressing the difficulties faced by older single men in rural areas in finding suitable partners is also closely linked to efforts to raise fertility.

Policies that support second and subsequent births may also be more effective than policies focused mainly on first births. Where Is China’s Population Headed? (2025 Concise Edition) compared age-specific fertility rates in China with those in Japan, France, England and Wales, and Scotland. China’s fertility rates among women aged 15–19, 20–24, and 25–29 were clearly higher, while fertility among women aged 30–34, 35–39, and 40–44 was markedly lower. Women in China also have their first child at a younger age on average, while the proportions who remain unmarried or childless throughout their lives are much lower than in Europe, Japan, and other developed economies. Taken together, these patterns suggest that China’s lower fertility is driven mostly by low rates of second and subsequent births.

International experience also shows that as economies become more developed and societies become more open, individuals gain greater freedom in decisions, and more people choose not to marry or have children. Policies aimed at persuading this group to marry and become parents are unlikely to be as effective as measures that help families who already want children to have a second or third child. Supporting policies should therefore reduce the financial and practical burden on larger families.

Gender imbalance may be another major constraint on a fertility recovery. As noted in Where Is China’s Population Headed? (2023 Edition), among those born between 2006 and 2010 — now roughly 16 to 20 years old and likely to enter marriageable age over the next five years — there were about 20 per cent more males than females. The gap narrowed among later birth cohorts but remained more than 115 boys for every 100 girls until 2016, when it began to move back toward 105:100.

The cohorts entering marriage and childbearing age over the next decade will therefore remain significantly imbalanced by sex. Differences in educational attainment make the matching problem even more complex. As shown in Figure 8, women’s share of undergraduate and postgraduate enrolment rose steadily and remained above 50 per cent after 2011. Yet women account for less than half of the overall population aged 20–29, and their share in this age group continues to fall.

Among children born between 2006 and 2015, the sex ratio remained consistently above 115 boys for every 100 girls, and exceeded 120:100 between 2006 and 2010. From around 2031, these cohorts may begin moving into the key childbearing ages of 25–34. The imbalance is likely to have a visible effect on both marriage and fertility rates.

Women make up a larger share of highly educated groups even though they account for a smaller share of the total population. This means that the gender imbalance is even more pronounced among people with lower levels of education. Men with less education — especially those in rural areas and small towns — may find it increasingly difficult to find suitable partners. At the same time, traditional expectations often lead women to seek partners with equal or higher educational attainment and socioeconomic status, which can make partner matching difficult for highly educated women as well. Addressing these structural obstacles to marriage will be crucial to efforts to raise fertility.

Figure 8. Women Account for a Larger Share of Undergraduate and Postgraduate Students

Urbanisation Is Slowing

The Pace of Urbanisation Is Losing Momentum

China’s urbanisation rate rose to 67.89 per cent in 2025, but the pace of increase has slowed markedly. Between 2012 and 2020, the urban share of the population rose by an average of 1.34 percentage points a year. Since it passed 64 per cent in 2021, however, the average annual increase over the past five years has fallen to just 0.80 percentage points. At that pace, China’s urbanisation rate will pass the theoretical turning point of 70 per cent around 2028 and enter a more mature phase in which annual gains continue to slow.

Figure 9. China’s Urbanisation Is Slowing

Among 337 Chinese cities classified as fifth-tier or above, 185 recorded a decline in their resident population between 2024 and 2025 — more than half of the total. (Because city-level data remain incomplete, these figures have not been adjusted for natural population change.) A further 65 cities had not yet released resident-population data for 2025. Most of the cities losing population were third-tier or lower, indicating that urbanisation is becoming difficult for small and medium-sized cities.

Figures 10–11. Changes in Resident Population Across Chinese Cities in 2025

Note: The city tiers follow the 2025 Ranking of China’s New First-Tier Cities, released in May 2025 by Yicai’s New First-Tier Cities Research Institute. The ranking evaluates cities on five dimensions: concentration of commercial resources, strength as transport and business hubs, urban vitality, competitiveness in the new economy, and future potential. It draws on multiple sources of big data, including branded-store networks and patterns of internet use.

Migrant Workers Are Becoming Less Mobile

During the 2010s, China’s migrant-worker population increased by more than five million a year. Large-scale, one-way movement from the countryside to cities was the main force behind rapid urban expansion. By 2025, however, the total number of migrant workers stood at 301.15 million, only 1.42 million more than a year earlier, for growth of just 0.5 per cent. Of the total, 121.09 million worked within their local townships, an increase of 70,000, or 0.1 per cent. A further 180.06 million worked away from home townships, up by 1.35 million, or 0.8 per cent.

Figure 12. Growth in China’s Migrant-Worker Population Continues to Slow

The slowdown in the growth of the migrant-worker population has been accompanied by a steady decline in the share moving across provincial boundaries. Many rural workers now prefer jobs in nearby counties or cities rather than long-distance employment in another province. Large-scale interregional migration is contracting. In 2025, 67.65 million migrant workers moved across provincial boundaries, 750,000 fewer than the year before, a decline of 1.1 per cent. By contrast, 112.41 million moved within their home provinces, an increase of 2.1 million, or 1.9 per cent. Interprovincial migrants accounted for 37.6 per cent of migrant workers employed away from their local townships, down 0.7 percentage points from the previous year.

Even in Central China, the region with the highest proportion of interprovincial migrant workers, the share has fallen to 51.4 per cent and is declining. In 2025, migrant workers had an average age of 43.3, and 32 per cent were aged 50 or above. As more workers grow older, leave migrant employment, and return home, labour mobility will weaken further.

Figure 13. Share of Migrant Workers Moving across Provinces by Region, 2016–2025

Metropolitan Concentration Continues

Strong Provinces, Metropolitan Regions, and Manufacturing Hubs Continue to Attract Migrants

Rural China is older than urban China, and the pool of surplus rural labour available for migration is nearing exhaustion. Population growth in large cities will therefore depend increasingly on redistribution within the urban system and on the continued concentration of people in major metropolitan regions. Excluding natural population growth, net population inflows are increasingly concentrated in Guangdong, Zhejiang, and Jiangsu.

In 2025, Guangdong recorded a net population inflow of 500,000 (excluding natural population change), overtaking Zhejiang’s 389,000 and Jiangsu’s 227,000 to regain first place nationwide. Chongqing returned to positive net migration, ranking fourth with an inflow of 115,000. Sichuan and Anhui, both of which had recorded large net inflows in 2023 and 2024, shifted back to net outflows in 2025.

Figure 14. Net Population Inflows and Outflows by Province in 2025

Zhejiang Has Become China’s Most Dynamic Province

Over the longer term, China’s population has continued to concentrate along the Yangtze River, along the coast, and in major inland city clusters. The Yangtze River Delta, Pearl River Delta, and Chengdu–Chongqing region have recorded some of the largest net inflows. Zhejiang already has a large population, yet after natural population change is excluded, net migration increased its resident population by 7.6 per cent over the past eight years — far faster than in any other province.

Guangdong, Jiangsu, and Sichuan also recorded net gains of around 1–2 per cent over the same period upon their already large populations, reinforcing the nationwide concentration of people in economically dynamic core regions. Hainan, Xinjiang, Chongqing, Shanghai, Fujian, Shaanxi, and Tibet, though smaller in population, also maintained net inflows over the past eight years. In contrast, the six populous provinces of Henan, Hunan, Hubei, Hebei, Shandong, and Anhui all experienced net outflows.

Figure 15. Growth in Net Population Inflows by Province, End-2017 to 2025

Within major metropolitan regions, people are concentrating further in the core cities. By 2025, Hangzhou and Ningbo in Zhejiang, Suzhou and Nanjing in Jiangsu, Shenzhen and Guangzhou in Guangdong, and Chengdu in Sichuan together accounted for around 30 per cent of their respective provincial populations. Changes in resident population show that these core cities contributed most of the population growth in their regions.

Between 2017 and 2025, Guangzhou and Shenzhen accounted for more than half of Guangdong’s total population increase. Hangzhou and Ningbo contributed more than 95 per cent of Zhejiang’s. The population gains in Suzhou and Nanjing, and in Chengdu, were larger than the net increases recorded by Jiangsu and Sichuan as a whole. Even in 2025, when the resident populations of both Jiangsu and Sichuan declined overall, Suzhou, Nanjing, and Chengdu continued to grow. Population concentration within metropolitan regions is therefore likely to intensify. (Because city-level birth and death data are incomplete, these urban population figures have not been adjusted for natural population change.)

Figures 16–18. Population Concentration Continues to Increase in Major Metropolitan Areas

Within Provinces, Urban Growth Is Increasingly Concentrated in Provincial Capitals

Even in provinces with weaker economies or smaller populations, almost every provincial capital continued to expand. In 2025, all provincial capitals gained population except Shijiazhuang of Hebei and Xining of Qinghai, whose populations fell by just 300 and 1,300 respectively; Nanchang of Jiangxi had not yet released updated data. Harbin of Heilongjiang lost a cumulative 1.042 million residents between 2017 and 2025 as a result of ageing and regional outmigration. Yet in 2024 and 2025 it added 110,000 residents, helped by temporary labour demand generated by its winter-tourism boom and by modest inflows from surrounding parts of Heilongjiang.

Over a longer period, provincial capitals have been the principal source of population growth within their provinces. Several provincial capitals in Northeast, Central, and Western China gained residents over the past eight years even as their provinces continued to lose population. These capitals are clearly drawing people away from surrounding cities. (Because city-level birth and death data are incomplete, the figures have not been adjusted for natural population change.)

Figures 19–20. Less Developed Provinces: Cumulative Population Change in Provincial Capitals Compared with Provincial Totals, 2017–2025

Figures 21–22. Less Developed Provinces: Population Change in Provincial Capitals Compared with Provincial Totals, 2025

China May Be Entering a Phase of “Passive Urbanisation”

“Passive urbanisation” refers to a stage in which the urbanisation rate no longer rises mainly because rural residents are moving into cities. During a prolonged period of total population decline, the urban share may increase because the rural population is shrinking naturally faster than the urban population, and because of administrative boundary changes.

Japan offers a useful example. Between 2000 and 2010, much of the increase in its urbanisation rate resulted from municipal boundary changes and the deaths of older residents in towns and villages, rather than from large inflows into cities (Qin Hong and Wang Yanfei, 2024). Population growth became concentrated in Tokyo and Osaka, helping property prices in those two metropolitan areas remain relatively firm after 2010, while housing markets in most other cities remained weak. This helps explain why Japan’s urbanisation rate rose from 77 per cent when its property bubble burst in 1991 to about 90 per cent in 2010 even as nationwide housing prices continued to fall. Passive urbanisation did not create substantial new demand for urban housing.

Figure 23. Urbanisation Rates in China and Japan (%)

China may enter a similar phase around 2030, after its urbanisation rate reaches 70 per cent. China has less room for further urbanisation than Japan and many Western countries, partly because it is ageing before becoming fully affluent. Once a society becomes deeply aged, population mobility declines, and will weaken further as ageing accelerates.

Urban–rural gaps in China’s basic healthcare mean that life expectancy remains lower in rural areas than in cities. The rural population is therefore likely to decline naturally at a faster rate, pushing up the national urbanisation rate. Meanwhile, continued metropolitan concentration is causing many third-, fourth-, and fifth-tier cities to lose population. Together with natural population decline, administrative changes such as converting counties into urban districts can turn some rural residents into urban residents without any actual migration.

Regional Migration Through the Lens of Industrial Development

Manufacturing Export Cycles Create Population Tides

Guangdong’s resident population increased by 790,000 in 2025, including 290,000 from natural population growth and 500,000 from net migration from other provinces. Even after natural growth is excluded, Guangdong remained the country’s largest destination for migrants, well ahead of Zhejiang’s net inflow of 389,000 and Jiangsu’s 227,000. Six of the ten Chinese cities with the largest population gains in 2025 were in Guangdong. In addition to the strong pull of the two first-tier cities, Shenzhen and Guangzhou, major manufacturing and export centres such as Dongguan, Foshan, Zhongshan, and Huizhou also continued to attract migrants.

Figure 24. China’s Ten Cities with the Largest Population Gains in 2025

Manufacturing is the backbone of Guangdong’s economy. After natural population change is excluded, the province’s net inflows show a strong positive relationship with China’s export growth. Between 2000 and 2019, when Chinese exports expanded rapidly, Guangdong consistently ranked first nationwide in net migration. When exports came under pressure in 2022 and manufacturing employment contracted, population inflows slowed markedly. As exports recovered in 2024 and 2025, Guangdong returned to first place.

Figure 25. Net Population Inflows into Guangdong and Growth in Imports and Exports

Looking ahead, the profitability of traditional manufacturing is likely to remain under long-term pressure. Conventional manufacturers will find it difficult to create large numbers of new routine jobs simply by expanding capacity. As the working-age population shrinks and industrial robots and artificial intelligence are adopted more widely, manufacturing’s capacity to absorb low-skilled workers is likely to decline year by year.

Figure 26. Profits of Industrial Enterprises in Guangdong Province

A Vibrant Private Sector and Digital Economy Act as a Magnet for Migrants

The Yangtze River Delta is already relatively aged, so population growth depends heavily on inward migration. In 2025, Zhejiang recorded a natural population decline of 79,000 and Jiangsu one of 307,000. After natural change is excluded, however, the two provinces recorded net interprovincial inflows of 389,000 and 227,000 respectively.

Fast-growing small businesses, e-commerce, and cross-border trade in cities such as Hangzhou have created a large number of flexible jobs. Relatively low barriers to starting a business attract not only employees but also entrepreneurs. Jiangsu, meanwhile, has a strong foundation in advanced equipment, chemicals, biopharmaceuticals, semiconductors, and other high-end manufacturing industries. The dense concentration of high-tech firms in industrial parks across southern Jiangsu makes the province especially attractive to skilled technicians and science and engineering graduates.

High-tech industries and services are likely to account for an ever larger share of employment. Under China’s statistical classification, the tertiary sector currently produces 57 per cent of GDP but employs only 49 per cent of the workforce, leaving an eight-percentage-point gap. In the United States, by comparison, services have long accounted for a slightly larger share of employment than of GDP; in 2025, the sector employed 84 per cent of the workforce.

Figure 27. China’s Service Sector Has Considerable Room to Expand Employment

Figure 28. The Service Sector’s Share of U.S. Employment Has Long Exceeded Its Share of GDP

China’s high-tech industries are still expanding rapidly in both scale and revenue. The growth of high-value-added industries and the wage premium they offer draw young people and skilled workers across regional boundaries. Supporting service workers then follow, creating a complete ecosystem of population concentration. Zhejiang, Shanghai, and Jiangsu offer large numbers of jobs in advanced manufacturing, the digital economy, and services, making them the main drivers of population concentration in the Yangtze River Delta.

Figure 29. China’s High-Tech Industries Continue to Expand Rapidly in Scale and Revenue

Core Cities in Central and Western China Are Drawing People Back

Central and Western China comprises 18 provinces and provincial-level municipalities, covers more than 70 per cent of the country’s land area, and is home to more than 500 million people. It is China’s strategic hinterland. The development of a unified national market is strengthening the region’s capacity to absorb industries relocating from elsewhere. Core cities such as Chengdu and Chongqing, Wuhan, Zhengzhou, Changsha, Xi’an, and Hefei are building complete industrial chains in electronic information, new-energy vehicles, advanced equipment, and other sectors. Their large local consumer markets are also supporting better urban amenities and making these cities more attractive to people seeking work closer to home.

Between 2017 and 2025, the resident populations of Chengdu–Chongqing, Wuhan, Zhengzhou, Changsha, Xi’an, and Hefei each increased by more than one million. Chengdu led with an increase of 5.49 million. These cities continued to gain population in 2025. Although Chongqing’s total resident population declined, the fall was mainly due to natural population change. Once that factor is excluded, Chongqing recorded a net inflow of 115,000 in 2025, ranking fourth among all provinces and provincial-level municipalities.

Figure 30. Manufacturing Centres in Central and Western China Draw Population from Surrounding Areas

Central and Western China is also rich in coal, oil and gas, minerals, wind, and solar resources. As energy and mineral inputs become more important to emerging industries, resource-development projects are creating large numbers of jobs and supporting related industries. They attract skilled workers and professionals from other provinces while also offering employment to rural workers within the region. This helps retain younger local residents and reduce labour outflows.

Figure 31. Resource-Based Urban Development Drives Population Growth

Over the long term, ageing will continue to reduce population mobility. Even so, metropolitan concentration is likely to persist because large cities provide better public services and more job opportunities than small and medium-sized cities. The number of Chinese cities with populations above 10 million will likely continue to rise. At the same time, some fifth-tier cities may eventually be administratively consolidated, much as many natural villages in rural China have been combined in previous decades.

A smaller population and declining mobility as a result of ageing will also make overcapacity in transport infrastructure an increasingly serious concern and add to fiscal pressure. Accelerating ageing will also widen funding gaps in the social-security system, requiring a growing share of public spending to support pensions and other social insurance programmes. These challenges require advance planning.

(Since 2022, local statistical authorities have released increasingly limited demographic data, particularly on births and deaths. This has made demographic research and forecasting more difficult, and we appreciate readers’ understanding.)

Subscribe now

Zhang Ming, Deputy Director of the Institute of World Economics and Politics, Chinese Academy of Social Sciences (CASS), and Deputy Director of the National Institution for Finance and Development, suggests that China is pursuing a different model, one rooted in the strength of its manufacturing sector rather than the dominance of its financial markets. That strength could allow China to sustain current-account surpluses even as Chinese companies, banks and investors send more capital abroad. Supplying renminbi through financial-account outflows and drawing it back through current-account surpluses, he argues, is emerging as a distinctively Chinese path to currency internationalisation.

This article was first published on 15 June in Issue No. 12, 2026, of China Forex, a publication overseen by the State Administration of Foreign Exchange (SAFE), China’s foreign-exchange regulator.

Zhang has kindly reviewed and authorised the translation.

—Yuxuan Jia

货币国际化的中国路径

The Chinese Path to Currency Internationalisation

One view holds that the renminbi internationalisation will remain constrained as long as China continues to run a current-account surplus. The reasoning is that a country must run current-account deficits to supply its currency to the rest of the world.

The United States is often cited as the leading example. It supplies dollars to the global economy through persistent current-account deficits while attracting those dollars back through sustained surpluses in its non-reserve financial account. This combination of a current-account deficit and a financial-account surplus has underpinned the global circulation of the dollar.

The United States’ persistent financial-account surpluses reflect the competitiveness of its financial markets. On the one hand, U.S. markets can supply large volumes of global safe assets—most notably U.S. Treasury securities—to meet other countries’ demand for reserve investments. On the other hand, they offer a broad range of higher-return risk assets, particularly U.S. equities, to meet the investment needs of private-sector investors around the world.

The well-known concept of “petrodollar,” for example, describes how oil-exporting countries reinvest in the United States the dollars they earn from oil exports.

However, the model of supplying dollars through current-account deficits is subject to the Triffin dilemma. To meet global demand for dollar liquidity, the United States must continue to run current-account deficits. Yet persistent deficits also increase its net external debt.

As that debt approaches a level investors consider difficult to sustain, concerns may grow that the U.S. government will seek to reduce the real burden of its liabilities through a sharp depreciation of the dollar. There is therefore an inherent tension between supplying the rest of the world with sufficient dollar liquidity and preserving the stability of the dollar’s value.

U.S. net external debt has now risen to more than 90% of nominal GDP, compared with only 20% when the global financial crisis erupted in 2008.

There is no single prescribed path to currency internationalisation. A country seeking to expand the international use of its currency must choose the approach best suited to its resource endowments and competitive strengths.

For China, advancing renminbi internationalisation through a combination of a current-account surplus and a non-reserve financial account deficit may be both more realistic and more sustainable.

On the one hand, Chinese manufacturing is highly competitive globally. Its strength lies not only in price competitiveness, but also in product reliability, comprehensive industrial supply chains, and the ability to respond quickly to changing market demand.

China now accounts for roughly one-third of global manufacturing value added. Given its continued strength in both traditional and emerging manufacturing industries, the country is likely to maintain a current-account surplus for a considerable period.

On the other hand, Chinese companies, financial institutions, and households all have strong incentives to expand their activities overseas.

For Chinese companies, the combination of China-U.S. trade frictions and intensifying geopolitical tensions has made the global diversification of supply chains increasingly necessary. As more Chinese companies invest and establish production facilities abroad, China’s outward FDI may continue to exceed inward FDI.

Chinese financial institutions and households also have considerable demand for global asset allocation as they seek to achieve a better balance between risk and return. Given China’s vast pool of domestic savings and the resulting demand for internationally diversified securities portfolios, outward portfolio investment may continue to exceed foreign portfolio investment in China.

In addition, because Chinese interest rates are now significantly lower than those in the United States, overseas demand for renminbi-denominated financing has increased markedly. As a result, the “other investment” component of the balance of payments is also likely to remain in deficit, making China a net provider of cross-border credit to the rest of the world.

Indeed, China’s balance-of-payments structure in recent years has increasingly exhibited precisely this pattern: a current-account surplus combined with a financial-account deficit. Direct investment, portfolio investment, and other investment have all recorded deficits to varying degrees.

Supplying renminbi to the rest of the world through financial-account outflows, while allowing the currency to flow back to China through current-account surpluses, is emerging as a distinctively Chinese path to currency internationalisation.

Compared with the U.S. model, persistent current-account surpluses would increase China’s net external assets. Over the medium term, this could reinforce expectations of renminbi appreciation and help China avoid the Triffin dilemma to some extent. The Chinese path may therefore prove more sustainable than the U.S. model.

The United States itself followed a similar path during the early years of the Bretton Woods system, combining current-account surpluses with financial-account deficits.

Under the Marshall Plan, for example, the United States extended loans to European countries, which then used the funds to purchase American machinery, equipment, and other goods needed for postwar reconstruction.

This historical experience suggests that currency internationalisation may proceed in two stages.

In the first stage, a country’s principal competitive advantage lies in manufacturing. It therefore promotes the international use of its currency through a combination of current-account surpluses and financial-account deficits.

In the second stage, as the relative competitiveness of its financial markets rises and that of its manufacturing sector declines, the country shifts towards a model based on current-account deficits and financial-account surpluses.

The internationalisation of the renminbi is currently in the first of these two stages.

Subscribe now

This is the argument made by Dandan Zhang, a Professor in Economics and Deputy Dean at the National School of Development (NSD), and Deputy Dean of the Institute of South-South Cooperation and Development, Peking University.

One of China’s leading labour economists, Zhang argues that in today’s global race for technological advantage, no country is likely to slow its investment in AI. The real policy challenge, then, is not to hold back technological progress, but to make the transition as employment-friendly as possible: identifying the skills most at risk, adapting education and retraining, and strengthening safety nets for workers who may otherwise be left behind.

Professor Zhang delivered the speech for TAIXUE, a TED-like initiative by New Economist Think Tank. The video was uploaded on 10 June 2026. Zhang has kindly authorised and reviewed the translation.

— Yuxuan Jia

Hello everyone, I am Dandan Zhang, an economics professor at Peking University. I specialise in labour economics research. It is a great honour to be here at Taixue today to share some thoughts with all of you. In this talk, I will be exploring how artificial intelligence is reshaping the world of work.

On 30 November 2022, a large language model (LLM), namely ChatGPT, was officially released. Its arrival was so sudden that it felt almost otherworldly, as if something from beyond our world had suddenly landed among us. Since then, our generation has found itself standing right at the centre of a massive transformation.

Our jobs may be the first thing it touches. And this is one of the great uncertainties of our time: what will artificial intelligence mean for the employment prospects of our generation?

Today, I would like to examine this question through the lens of labour economics. Drawing also on my own research in recent years, I will share how economists understand the ways in which AI is reshaping employment.

The first fundamental economic question we need to answer is: does technological progress inevitably lead to unemployment? To answer this question, economists often begin by looking back. We ask: what impact did past waves of technological progress have on employment?

Let’s look at what has happened over the past 100 years. The world has experienced several technological revolutions, including electrification, computerisation, and, in the past decade or two, automation. All of these technological advances sparked job anxieties to varying degrees. Yet in every case, they all seemed to reach a state of employment equilibrium.

One study found that around 60 per cent of today’s jobs did not exist in 1940. What does this tell us? It tells us that technology does not simply destroy jobs. It also creates them. As old jobs disappear, new ones continue to emerge. For example, with the information revolution and technological progress in computing, computers replaced typists and some basic text-based work, but they also gave rise to the entire IT industry and millions of new jobs.

So the critical question is whether technological progress can create jobs at a pace that keeps up with the jobs it replaces. In other words, replacement itself is inevitable, and it is not something we need to fear in itself. What we really need to examine is whether, in the process of technological progress, there is a sufficient window of time for new jobs to be created.

Historically, the disappearance of old jobs has typically unfolded over a very long period of time, possibly several decades. That gradual transition gave society an essential window of time for new occupations to emerge and develop.

Next, I would like to summarise two key ways in which this wave of technological progress differs from previous ones.

First, artificial intelligence is not replacing physical labour; it is replacing mental labour. Previous waves of technological progress primarily displaced repetitive, operational work on assembly lines. In other words, replacing physical labour. This time, AI is acting on human cognitive abilities, such as comprehension, reasoning, creativity, and judgment. These are precisely the core competitive advantages of knowledge workers.

This year alone, Claude has dealt a significant blow to the programming profession. Models like Sora and Seedance also had a disruptive impact on video editing.

What deserves even more attention, or perhaps even caution, is a more recent change: AI is no longer only capable of “thinking”. It is now moving towards having both a “brain” and “hands”.

For example, OpenClaw, the personal AI assistant that has recently become very popular on social media, can actually get things done. Large AI models are moving from simply “talking” to being able to roll up their sleeves and help us complete many concrete tasks.

So this wave of technological progress is not only replacing mental work. It is also beginning to replace hands-on execution. This is the first key difference.

The second key difference is that AI technology can be deployed very quickly, and the cost of executing tasks with AI is extremely low. Compared to previous waves of technological progress, generative AI operates on a completely different cost structure. Take intelligent manufacturing, for example: deploying it is a very tedious process. First, you need to remodel the factory, order a bunch of different equipment, and fully reintegrate the production lines, all of which demands massive fixed-asset investments.

In the past, the deployment of new technology took time. It moved forward gradually, industry by industry, or region by region. Equipment had to be purchased unit by unit. Production lines had to be built one by one. As a result, the replacement of jobs also happened gradually and within a relatively limited scope.

This created time. But this time, generative AI is different. Through cloud services, APIs, and open-source models, both enterprises and individuals can access and deploy AI at a very low cost. In particular, with the expansion of many open-source models, the capabilities of large models can now be called upon almost instantly.

The cost of local deployment is also falling as models continue to iterate and as their use becomes more widespread. For example, I recently saw a remark by the renowned mathematician Terence Tao. He said that “AI has basically driven the cost of idea generation down to almost zero.” This means that the cost of a lot of knowledge-based tasks — for example, editing, copywriting, or contract reviewing — is dropping drastically.

So if you were a business owner, obviously you would want to use this new technology to replace relatively expensive human labour if the cost to implement this technology is low. The low cost incentivises businesses to replace human labour with technology.

So, to summarise, this wave of AI-driven technological progress is no longer just about replacing physical labour. It is replacing both mental work and hands-on execution. On top of that, compared to previous waves of technological progress, it carries significantly lower deployment and usage costs. The spread of this technology no longer has to unfold gradually from region to region or industry by industry; instead, a single model can emerge and take effect globally and simultaneously. It may affect an entire technological layer, or an entire category of occupations, all at once. And that is precisely what may shrink the window of time available for new jobs to be created.

Since this impact is inevitable, let’s take a look at how economists believe what kind of effects AI would have on employment. I will introduce three very frontier research methods. Through these different methodologies, we can see how economists approach this question.

The first method is called the AI Exposure Index analysis. In the AI era, the way economists look at employment, occupations, and industries is very different. We would deconstruct occupations, stripping them down into individual tasks or specific skills, so that we can analyse them along different dimensions.

If we want to understand how an occupation will be reshaped, we need to analyse it at a very micro, almost atomic level. Tasks are the basic atoms of a job. We need to define which atoms make up the jobs in today’s labour market.

For this, we need something like the periodic table of elements in chemistry. In labour economics, we do have something similar. For example, O*NET and other international databases have compiled tasks into a kind of dictionary. In total, there are around 20,000 tasks that make up more than 1,000 occupations in today’s labour market. Every occupation is deconstructed. Take a university professor, for example, whose role would be broken down into a specific set of work tasks.

The next step is to evaluate, among these 20,000 tasks, how likely it is that a LLM can complete or accelerate each task.

More specifically, we ask whether AI can save more than 50 per cent of the time required for that task. If it can, we label that task as 1, meaning it is exposed. If it cannot, we label it as 0, meaning it is not exposed. At this point, every task within an occupation has been labelled, with each task carrying either a 0 or a 1 — no exposure or exposed, respectively. These task-level scores are then aggregated up to the occupation level, giving a calculated exposure score for each occupation.

The data used in this kind of analysis mainly comes from recruitment data. We look at the requirements in new job postings: what tasks do these new positions require, and what responsibilities do they list? In this sense, we are using flow information from the labour market.

Our team has also applied this method to research on China. We used more than one million job postings from Zhaopin to calculate AI exposure levels for occupations in China.

Our research found that the occupations with higher exposure — that is, those more likely to be affected by AI — are mainly white-collar jobs. This is basically consistent with international findings. Accountants, editors, salespeople and programmers, in particular, may be among the more significantly affected occupations.

Occupations with lower exposure are mainly blue-collar jobs, such as food service workers, industrial workers and domestic workers. So, in general, high exposure is concentrated among white-collar workers, while low exposure is concentrated among blue-collar workers. Again, this is also broadly consistent with findings from other countries.

Over the past few years, some occupations with high exposure have indeed seen a decline in labour demand. This suggests that high exposure may bring a certain displacement effect. However, although the exposure index method is widely used, it also faces challenges and still needs to be improved. One major issue is that it actually measures potential exposure. But exposure does not equal actual replacement, just as standing in the sun does not necessarily mean you will get sunburned. The risk exists, but it is still a potential risk.

So the exposure index does not measure actual replacement. It measures theoretical substitution. It tells us whether AI can complete a certain task, not whether AI will replace a particular job. That’s why we can’t jump to a completely negative conclusion — even if the exposure is high, we shouldn’t assume the worst.

As I mentioned earlier, our research does show a negative relationship between AI exposure and labour demand. In other words, the higher the AI exposure, the lower the demand tends to be. However, using data from China, Singapore, and the United States, we found that the exposure levels for many occupations are concentrated between 0.7 and 0.8. Although the overall relationship is negative, the picture becomes more nuanced within this range. At the same level of exposure, demand falls for some occupations, remains unchanged for some, and rises for others.

For example, computer-related jobs, which are highly exposed, have seen a decline in demand. Yet, other roles with the same level of exposure see an increase in demand, such as retail sales positions. Meanwhile, some remain unchanged, like business operations. So even at the exact same level of exposure, if we look closely at different occupations, demand is falling for some, staying flat for others, and rising for others.

This raises an important point: at the same level of exposure, not all occupations will be replaced. This shows that the exposure index does not automatically point to the inevitable replacement of an occupation.

So what exactly is going on here? One main reason is that, as I said earlier, a job is made up of tasks. If AI replaces only some of these tasks, or completes part of the work, then human workers can redirect their attention to the remaining tasks that AI cannot do. They can spend more time on the parts of the job where human contribution is still needed.

The other reason has to do with what the O-ring theory, which looks at the relationships among the tasks that make up a job. If the tasks within a job are in an O-Ring relationship — meaning the different tasks are complementary — then focusing more effort on certain tasks can greatly increase output and improve labour productivity. But if the tasks are parallel, meaning that they can substitute for one another, then AI may lead to unemployment.

So, ultimately, whether an LLM will actually replace a job depends on two things: the combination of tasks within that job, and the nature of the relationships among those tasks. Occupations may have similar levels of exposure, but their actual likelihood of being replaced may be very different. This depends on the two factors I have just described.

Therefore, looking at exposure level alone is not enough. We also need to determine, at the occupational level, whether AI’s impact is pushing in a positive or negative direction. The goal is to give the exposure index more depth and nuance — one that can actually tell us whether displacement will eventually occur. That is what our team is currently working on, refining the methodology behind the exposure index.

Now let me introduce the second method, which we might call AI job-integration analysis. The previous method looks at the potential, or theoretical, impact of AI. This second approach looks instead at real-world implementation at the firm level.

This line of research comes from a 2025 NBR working paper by two PhD students from Harvard University. The paper, titled Generative AI as Seniority-Biased Technological Change, has been widely cited. It argues that generative AI does not favour high-skill workers, but rather those with seniority.

The authors analyse the full text of firms’ recruitment advertisements over the past several years. They identify whether a firm is hiring for roles specifically responsible for implementing or integrating AI technologies into the company’s workflow. If a firm posts this kind of job opening on a given day, the authors treat it as evidence that the firm has already started using AI and is using it in a relatively deep way.

So, as you can see, this method follows a completely different logic from the previous one. It no longer uses theory to infer whether AI may affect certain jobs. Instead, it directly observes firms’ actual behaviour: are they hiring people to bring AI into their workflows? And the findings are very interesting.

The authors found that generative AI represents a form of seniority-biased technological change. Among firms that had adopted AI, hiring for junior positions declined significantly, while hiring for senior positions remained stable, and in some cases even increased. In other words, the impact of AI appears to fall most heavily on those at the lower end of the career ladder, especially entry-level workers.

The authors also found that the main effect was a slowdown in new hiring, rather than layoffs or resignations. In other words, firms were not necessarily cutting existing workers; they were hiring fewer new ones.

Furthermore, when looking across different education levels, from elite universities to mid-skilled technical colleges and low-skilled institutions, the authors found that mid-skilled technical graduates bear the heaviest impact, while graduates from top-tier universities and lower-skilled workers remain largely unaffected by the slowdown.

This research offers an important insight: AI may not first affect the existing employed population in the labour market — what we might call the stock of employment. Instead, it may gradually compress labour demand by reducing the number of new positions. There may still be a window between potential exposure and actual substitution. AI may first reduce the entry of new workers into the labour market, while the existing stock of employed workers may, at least for now, be less affected.

This method also has limitations. It has very high data requirements. It needs access to a company’s complete hiring history to determine whether AI is genuinely being used.

The third method I am going to introduce is based on actual AI usage data. Rather than theorising about what AI can do, which is what the exposure index approach does, it directly observes how people interact with LLMs, and through that interaction data, draws conclusions about what people are actually using AI tools for.

This is a very recent and widely circulated study. Many of you may have seen it circulated on WeChat official accounts. It was released by Anthropic, the company behind Claude, in an employment report published earlier this year. In this report, Anthropic introduced an important concept: the observed exposure.

The first method I introduced measures theoretical exposure. This third method measures observed exposure. The data used in this research comes from actual conversations between users and Claude. The researchers looked at how users interact with Claude every day and what tasks they use Claude to complete. They then identified the specific things users were doing when they used a LLM and extracted those tasks for analysis.

The next step was to examine whether the LLM directly completed those tasks, or whether it helped humans complete them. If the model completed the task directly, this was treated as substitution. If the model helped the person complete the task and enhanced their capability, this was treated as augmentation.

Their findings are very interesting. They found that, so far, actual replacement has not really occurred on a large scale. What we mainly see is complementarity. LLMs are mainly improving our work efficiency, rather than producing obvious substitution.

They also compared observed exposure with theoretical exposure, which is what the first method captures. They found that theoretical exposure is much higher than actual observed exposure. For example, in computer and mathematics-related occupations, theoretical exposure suggests that more than 94 per cent of tasks could be completed by AI. But in actual Claude usage, Claude only covers about one third of these tasks. The difference between more than 90 per cent and just over 30 per cent is very large.

Why can AI theoretically do more than 90 per cent of the tasks, while in practice it only does a little over 30 per cent? There are many reasons for this gap.

One reason is that model capabilities are still limited. Another is legal and compliance constraints. In many cases, when models use certain types of data, human verification is still needed. There are also barriers related to software integration. So there is still a gap between actual use and theoretical substitution.

Using the observed exposure index, the researchers found that 57 per cent of tasks were augmentation-type tasks, while 43 per cent were substitution-type tasks. So, overall, augmentation remains the main pattern.

Augmentation-type use improves productivity, but it does not reduce employment. Substitution-type use is what directly threatens employment. So their findings are somewhat reassuring: AI is making us better at what we do, while not yet taking away our jobs on a large scale.

That said, this method has its limitations. It relies on a single LLM — Claude — and the range of tasks it covers is quite narrow. As mentioned earlier, the full task dictionary contains 20,000 tasks, yet this study only captures 800 of them as they appear in Claude interactions. The analysis is therefore partial and incomplete. Nevertheless, the study introduces a valuable new perspective: by looking directly at how people actually interact with AI, it suggests that the reality may not be nearly as alarming as it sounds in theory.

I’ve now finished introducing these three methods, so let’s do a quick comparison. These three methods actually complement each other. They seem to capture three different stages of AI’s impact: from theory, to firms, to people. The exposure index, the first method, mainly looks at the potential exposure, or the theoretical exposure. It relies on flow data, especially new recruitment information.

The second and third methods both look at actual adoption, drawing on existing employment data to understand how much of the current employment structure may be affected. These methods can complement one another, and I believe only by combining them can we get a more comprehensive and more precise analysis.

So this is how economists are now analysing the question of how LLMs affect employment. These methods are all at the frontier of current research. Our own research team is also trying to apply these methods to study China.

The first point I want to emphasise is that we now need to build a dynamic monitoring and early-warning system. This is really important. We need to track, in real time, how the tasks and skills contained in today’s jobs are changing. Which skills are needed in this new era? Which skills may be phased out? This lets us spot the risk signals and take action well before any changes start showing up in the employment data for the existing workforce.

Second, policymakers need to develop education and training programs based on the monitoring results. People need to know quickly which abilities are needed and which skills are becoming more important, so that they can fill their own skill gaps as soon as possible.

Here, I want to emphasise individual initiative and creativity. Through our research and conversations with Xianyu [the Chinese equivalent of eBay], we found that new, user-created forms of work emerge on the platform every year. Some are quite unconventional: people offer to argue with someone on your behalf, retouch photos with AI, bathe pets, and provide all kinds of niche services. These examples show just how extremely creative individuals can be.

So I think the first thing we need to do is send clear signals to the market. We need to allow individuals to face and embrace the challenges brought by AI, to reinvent themselves, and to create more employment opportunities for themselves.

But creativity also needs to be guided by information. The hope is that these early warning signals can help each person make better career choices, shorten the time it takes to retrain for new work, and lower the cost of trial and error during this transition.

The third point is about safety-net policies. First of all, technological progress cannot be stopped. So while AI is getting smarter and smarter, policymakers need to develop much more robust risk management strategies during this transition period, particularly creating strong social safety net policies.

In the current environment of international competition, no country is going to slow down its investment in AI or the pace of AI development. So the key policy question, I think, is not about preventing the use or the development of AI. On the contrary, technological progress should be encouraged. At the same time, policymakers need to prevent a situation where workers, who cannot get through this short-term adjustment period, end up permanently locked out of the new employment landscape.

For instance, at this year’s Two Sessions [annual plenary sessions of the National People's Congress and of the National Committee of the Chinese People's Political Consultative Conference] we saw the proposal to build an employment-friendly society. This points to an important general direction. What that means is that going forward, as more firms integrate AI into their workflows, policymakers need to assess whether the employment landscape is actually employment-friendly. This involves a whole range of policy measures, and I believe it can help us open up a real window for employment transition — one where, even as some jobs are being replaced, more new opportunities are being created at the same time.

Finally, let me take some time to summarise today’s talk. The arrival of generative AI is reshaping our labour market. According to the main studies so far, it mainly affects white-collar jobs, entry-level positions and some mid-skill workers. AI may improve work efficiency, or it may bring substitution. The final trajectory of each occupation and each job is still unclear. We cannot say that any particular occupation will definitely disappear.

This process depends, to a large extent, on the initiative of each worker: whether they are willing to adapt and change. It also depends on the way the tasks and skills within each job are combined.

One last point I want to make is that, at present, what we are seeing is still mostly potential impact. The real shock has not yet fully arrived. This means that we are still in a very precious window of opportunity. The question is whether we can make the most of this window and prepare ourselves well. For example, can policy analysis stay a step ahead of each new wave of technology, so it can send the market clearer, more accurate signals and make the right adjustments? That is a shared challenge facing both researchers and policymakers, and it’s what my team and I are working towards. That’s everything I wanted to share with you today. Thank you, everyone.

Subscribe now

A major part of that effort has involved restoring exchanges between the two countries. Direct flights have gradually resumed, while business, tourism and other forms of people-to-people contact are recovering. Under Xu’s leadership, the Chinese Embassy and consulates in India have also been processing increasing numbers of visas for Indian citizens traveling to China. These are not merely administrative steps. They are part of a broader attempt to rebuild the connective tissue of the relationship, restore a degree of normalcy after years of estrangement, and open China more widely to the outside world.

Yet the increase in travel became a target of a group of self-styled “patriotic” influencers. False or misleading claims began circulating that Indians were “flooding” into China and that Chinese diplomatic missions in India were recklessly handing out visas. The attacks eventually turned on Xu himself. A veteran diplomat carefully advancing his government’s policy — and seeking to rebuild a difficult relationship without ignoring its underlying complexities — was accused of betraying the country whose interests he was working to serve.

The recent hostility has an immediate context. China-India relations deteriorated sharply after the deadly border clash in 2020, and the years of tension that followed contributed to a rise in hostile sentiment in parts of each country against the other. But negative views of India predate that episode and have deeper, more complex roots, including perceptions of India as a poorer, less developed country — a form of prejudice that is by no means unique to China but can be found in societies around the world.

What makes this episode particularly revealing is not simply the existence of such attitudes, but what happens when longstanding sentiments are hijacked by racists and then collide with a diplomatic thaw. Nationalist influencers can turn grievance, prejudice and fabrication into traffic. Diplomats, by contrast, must deal with the world as it is: balancing national interests, political realities and the longer-term need to prevent rivalry from hardening into permanent hostility. That is the more difficult and less theatrical task Xu and his colleagues have been undertaking.

The Chinese internet is subject to extensive controls, but that does not mean every strand of public opinion is manufactured from above, or that the authorities exercise complete command over the sentiments circulating online. In this case, some of the people most loudly claiming the mantle of patriotism ended up attacking diplomats for carrying out Beijing’s policy. It is a small reminder that even a state with enormous influence over the public sphere may still have to contend with nationalist pressures that have histories and dynamics of their own — and that responsible statecraft sometimes requires navigating and pushing back against bigotry, rather than staying silent.

—— Zichen Wang

It was against this backdrop that, on July 2, 澎湃新闻 The Paper, a major Chinese digital news outlet, published the commentary translated below. The Chinese Embassy in India later republished it in full on its website, signaling unmistakable support for its argument.

如何看待近期关于印度的一些讨论？

How Should We Think About the Recent Discussion on India?

Recently, Chinese social media has seen a number of posts claiming that Indian visitors are “flooding” into the country and that some have behaved poorly while in China. Some of these discussions have gone further, questioning whether China’s visa policy toward India is too generous, or alleging that Indians are “taking up Chinese resources.” In some corners of the internet, public anger has been stirred up. How should we understand this? A few points are worth making.

The immediate trigger for this round of debate is the claim that some Indian tourists have behaved improperly in China. If foreign visitors break rules or behave inappropriately, people are, of course, entitled to point this out. China has a long tradition of treating guests with courtesy, but guests are also expected to respect Chinese laws, regulations, and social norms. If anyone crosses the line, whatever their nationality, the matter should be handled in accordance with the law. There is nothing controversial about that.

The real problem begins when the behaviour of a few individuals is used to condemn an entire country, or when isolated incidents become an excuse for xenophobia. We should not close the door on India because of a handful of cases. Nor should we pin every negative story on Indians simply because the topic is trending and attracts online traffic.

When thinking about how China should engage with India, we need to look at the bigger picture. Historically, China and India are both ancient civilisations, with more than two thousand years of cultural exchange between them. From Xuanzang’s journey westward and his return with 657 volumes of Sanskrit Buddhist scriptures, to Rabindranath Tagore’s visit to China and his friendship with figures such as Xu Zhimo, to Dr Dwarkanath Kotnis, who gave his life while supporting the Chinese people’s War of Resistance Against Japanese Aggression — these stories of people-to-people exchange are the true foundation of China-India relations.

In practical terms, China and India are neighbours that cannot be moved away. Together, they account for more than one-third of the world’s population, and both are pursuing modernisation. Peaceful coexistence is therefore a necessity. Win-win cooperation, too, is in shared interests.

Of course, differences between China and India do exist. But both countries have the wisdom to manage them. In recent months, the border situation has eased noticeably, which is a welcome development. Both sides also understand that the boundary question is not the whole of China-India relations, that it should be kept in its proper place, so that the broader relationship is not held hostage by a single issue.

In recent years, the leaders of China and India have met in Kazan and Tianjin. Since then, bilateral relations have continued to improve, and practical cooperation has strengthened. According to Indian official statistics, bilateral goods trade reached US$151.1 billion in the 2025–2026 fiscal year, making China India’s largest trading partner. This alone shows how deeply the two economies are connected. Behind the smooth flow of goods worth hundreds of billions of dollars, there must also be channels for people to travel, meet, and do business.

Since 2025, direct flights between China and India have gradually resumed. With more direct flights, it is only natural that people-to-people exchanges have increased compared with the pandemic years. It is also worth noting that more than 80 per cent of the visas China issues to Indian nationals are business visas. In other words, business travel makes up the overwhelming majority. Those coming purely for tourism are only a small share.

Seen from a wider international perspective, China-India relations are no longer just a bilateral issue. The Global South is gaining weight, and the global governance system is being reshaped. Whether in BRICS or the Shanghai Cooperation Organization, in climate talks or WTO reform, the Global South will find it hard to make its voice heard without coordination between China and India. India is chairing BRICS this year, and China has said it supports India in fulfilling its responsibilities as the BRICS Chair and working together to promote the growth of the BRICS mechanism. All this suggests that what may appear to be a simple easing of travel is, in fact, the renewed connection of two major engines of the Global South.

Back to the visa issue. In recent years, China has introduced a range of visa facilitation measures. These are not aimed at one particular country, but at the world as a whole. So far, China has concluded mutual visa-free agreements with 160 countries and has granted unilateral visa-free access to nationals of 50 countries. These measures are a proactive move in China’s deepening opening up. They bring not only more international tourism revenue, but also more trade opportunities, a better business environment, and a stronger reputation for openness. These are real national dividends, and in the end they also benefit ordinary Chinese people.

China’s decision to open its doors wider reflects the confidence and breadth of vision of a major country. The goal is not short-term “reciprocity” in every narrow sense, but long-term mutual benefit.

That is why discussions about “unequal visa treatment” between China and India should also be approached rationally. It is true that in recent years India tightened visa issuance for Chinese citizens. Although it has now resumed issuing business and tourist visas, there remains a clear gap and delay compared with China’s approach. China should expect India to show sincerity and take concrete steps. At the same time, we should recognise that India’s policy is also in a process of dynamic adjustment. If the visa issue is endlessly escalated and turned into an outlet for anger, the result will only be counterproductive, narrowing the space for the relationship to improve and develop.

After the noise dies down, we still need to return to reason. Over the past few years, the disruption of exchanges between China and India has widened not only an information deficit, but also a trust deficit. Some people like to post sensational clips and stories about India and other South Asian countries, or to amplify alleged misconduct by Indians in China. This has left many Chinese people with a view of India that is increasingly fragmented, label-driven, and stereotyped. Such content may appear to be “patriotic,” but it is little more than clickbait that fuels confrontation and exclusion. In essence, it exploits patriotic feeling and national sentiment for traffic. It pollutes the online environment and does not reflect the open and confident mindset a major country should have.

Chinese civilisation has always valued openness and inclusiveness. As the saying goes, the sea is vast because it accepts all rivers. As citizens of a major country, we should have that same sense of proportion, confidence, and generosity.

Share