The "Peak China" narrative is logically flawed

On George Magnus's recent column in the Guardian

For every prediction of a "Chinese century," there is a matching forecast of "Peak China." A recent example is George Magnus's essay in the Guardian, "Why Peak China may finally have arrived."

To motivate that conclusion, the column presents a set of exhibits, including but not limited to: (a) China's share of US-dollar GDP slipping; (b) population ageing; (c) urbanisation flattening; (d) productivity stalling; and (e) rising public-sector debt burdens.

The point isn't that these problems don't exist. The issue is that the argument relies heavily on selective statistics, short-term snapshots, and interpretations of policy that assume China is powerless when it may still have room to act.

More fundamentally, most "Peak China" claims, including this case, leave a gap in the basic syllogism required to reach a "has peaked" verdict. A logically complete version would require:

• Major premise: a country faces a set of severe, structural challenges.

• Minor premise: the country, within a reasonable horizon, lacks feasible tools or the willingness and capacity to address those challenges.

• Conclusion: therefore the country has peaked and will structurally underperform.

Against that standard, at least two issues arise:

1. Missing or diluted minor premise (this is the most common defect).

   1. The analysis overlooks China's remaining policy space and stated reform intent, or it equates imperfect/gradual implementation with incapacity or unwillingness.

   2. In this sense, the minor premise is not established. The syllogism therefore fails as a matter of inference: the "has peaked" conclusion is not compelled and, at minimum, lacks probative force.

2. Measurement-lens issues that affect the major premise (in some cases).

   1. Whether a cited item even qualifies as a "challenge" often depends on the metric chosen. A measured discussion should test multiple lenses rather than elevating a single convenient one.

   2. Example (GDP): the column reads China's post-2021 "slippage" relative to the US through nominal-dollar ratios during a strong-dollar phase. That choice creates a denominator problem. On constant-price comparisons or purchasing-power parity (PPP), the China-US gap presents another story.

   3. Accordingly, classifying "GDP relative slippage" as a challenge rests on a partial lens and is not sufficiently supported. Also, this second issue is narrower than the first one regarding the minor premise, but it is a legitimate ground for caution when defining the major premise.

In addition, two numerical slips further dilute the claim’s probative force:

1. The column gets China's 2008 GDP wrong. It puts the figure at "about $14 trillion," which was roughly the US total that year; China's nominal GDP was closer to $4.5 trillion.

2. The column describes China's 2024 urbanisation rate as "just over 60 percent," and official data place it around 67 percent.

Even if the piece is not primarily quantitative, figures at the heart of the GDP discussion are hard facts; getting them wrong may get readers confused and even erodes the credibility of the whole column. Read together with the unresolved minor-premise gap set out above, these slips may further weaken the force of the "has peaked" claim.

Finally, a note on intent. What follows does not seek to minimize the gravity of China's challenges. On the contrary, I have always argued for confronting them directly by consistently bringing China's internal debates about current problems and solutions into the English-language space (just see my prior newsletters).

In the following sections, I assess the column's claims on GDP metrics, urbanization, productivity, property, etc., indicating where the evidence holds and where it does not.

The purpose here is limited and constructive: to offer a countervailing lens that widens the frame, places discussion of China's evolving economy on a more balanced and better-informed footing, and encourages a complete chain of reasoning. It is to ensure that, the conclusions we draw rest on sound premises and appropriately chosen measures.

For ease of reference, the Guardian piece is referred to below as "the column."

I. GDP metrics and accounting scope

The column claims:

At the time of the 2008 financial crisis, China’s official, and probably exaggerated, GDP was about $14tn (£10.4tn), or about a third of that of the US. By 2021, it had risen to three-quarters of America’s $23.7tn, and there was widespread talk about in which year of the 2020s China would overtake the US.

By 2024, however, China’s $18tn economy had fallen back to just over 62% of the almost $30tn of the US. In GDP per head terms, China is still no more than 20% of the US.

A rising China uniquely lifted its share of global GDP between 2000 and 2021 from 3.5% to 18.5%, but since then it has slipped back to about 16.5%. There is no question that China’s rise is at least stalling.

Response:

Firstly, as noted above, the column misstates China's 2008 GDP. The correct figure is about $4.5 trillion; "about $14 trillion" is roughly the US total for that year.

Secondly, more consequential than the numerical slip is the reliance on a single measurement lens. Among many ways to gauge economic heft, nominal GDP is only one and is highly sensitive to exchange rates and price levels.

In a strong-dollar phase, China’s share shrinks on paper. If we strip out these valuation effects by using constant prices or purchasing-power-parity (PPP), the China–US gap does not widen in the way the nominal snapshots suggest.

In 2023, China’s nominal GDP converted at market exchange rates amounted to about 64.5% of the US level, down from the 2021 peak of 75.2%. However, two valuation effects drove much of that movement¹:

1. Exchange-rate effect. The renminbi was roughly 8.4% weaker in 2023 than in 2021. Valuing 2023 GDP at the 2021 rate would lift China’s ratio to around 70.5%, implying an “exchange-rate loss” of roughly 6 percentage points in the dollar conversion.

2. Price-level effect. Since 2021, the US GDP deflator rose much faster (about 7.1% in 2022 and 3.6% in 2023) than China's (about 1.8% and −0.6%). At 2021 prices, China's 2023 GDP would be roughly 70.8% of the US total.

Thirdly, a fuller assessment should bring in additional lenses, including constant-price comparisons and purchasing-power parity (PPP).²

• Constant-price (real) comparisons: Using 2015 constant US dollars to remove price and exchange-rate effects, China’s output in 2023 stood at about 78.9% of the US level, up 2.8 percentage points from 2021.

• PPP comparisons: On PPP, some institutions estimate China’s aggregate output surpassed the US in the mid-to-late 2010s and was about 1.27× the US level in 2023.

Finally, an often overlooked but relevant point concerns accounting scope. The US includes certain activities in GDP because they are legal under state or federal law (e.g., civilian firearms transactions, cannabis cultivation and processing). These activities are illegal in China and therefore excluded from China's GDP. Scope differences do not decide the growth debate, but they matter when treating nominal GDP as a definitive comparator.

In summary, nominal GDP is a useful but narrow lens. Without testing alternative measures (constant prices, PPP) and acknowledging accounting-scope differences, it is contestable to classify "China's GDP slippage" as a settled challenge rather than a measurement-driven artefact.

II. Demography, urbanisation, and productivity

The column claims:

The working age and total population are now in relentless decline. The urbanisation rate, just over 60%, is flattening out. Productivity growth has stalled.

Response:

1. Demography: challenge acknowledged, but not dispositive.

   1. It is correct that both total population and the working-age cohort are declining. This is a genuine headwind.

   2. However, this does not imply policy helplessness. As developed earlier and in the sections that follow, China retains channels to offset part of the drag, including capital deepening, automation and digital diffusion, urban-service productivity gains, and reforms that bolster household incomes and mobility.

2. Urbanization: not "flattening out," and with identifiable upside.

   1. International benchmarks. China’s resident (常住人口) urbanisation rate stood at roughly 67%³ at end-2024, no longer low by international standards, yet still some way from the ~80% typical of advanced economies⁴, leaving meaningful room to rise.

      Historical patterns indicate that once economies pass ~65%, urbanisation usually continues to rise for another decade or more before easing to ≤0.2 percentage points per year. The United States, Germany, Japan, and Korea took about 18, 11, 13, and 11 years respectively to reach that slower phase, with average annual increases of roughly 0.46, 0.64, 0.82, and 1.25 percentage points during those periods. By that yardstick, China still has meaningful runway. On the demand side, Chinese estimates suggest that each additional percentage point of urbanisation can catalyse on the order of RMB 1 trillion in new investment and more than RMB 200 billion in additional consumption.⁵

   2. Hukou (户口) system. Because of the household-registration, China’s hukou-based urbanisation (namely residents with urban hukou) lags the resident measure by on the order of 18–20 percentage points (e.g., resident ~67% vs hukou ~47%).

      This gap means many migrant workers are counted as urban residents but lack full, equal access to urban public services. Without secure social insurance and settlement expectations, their consumption patterns diverge from those of registered urban households. Empirical work by research teams at the Chinese Academy of Social Sciences and the OECD suggests that granting urban hukou can raise household consumption by roughly 28% even before accounting for subsequent income gains. ⁶

      In short, if the resident urbanisation rate has meaningful room to rise, the hukou-based measure has substantially more, supporting both household consumption and productivity gains.

3. Productivity: slowdown is real, but not destiny.

   1. Global context. Productivity growth has downshifted across many economies as manufacturing's share recedes and services, which are harder to mechanise and scale, expand. China is no exception: after industrialisation largely completed around 2010, estimates of total factor productivity (TFP) growth eased from averages near 4% (1978–2009) to roughly 1.3–1.8% in the past decade, contributing to headline growth moderation.⁷

   2. Paths to renewal. This is not an irredeemable structural trap but a transition challenge. Rapid advances in AI and data, the urgency of the energy transition, and China's deep industrial base create scope for "re-industrialisation" in a broad sense: digitising production and services, upgrading equipment, and electrifying systems. Complementary structural reforms, including raising household disposable incomes, investing in public services, and dismantling residual urban–rural barriers, provide additional, credible sources of future TFP gains.⁸

In summary, the demographic arithmetic is adverse, but the urbanisation runway and hukou-integration potential are substantial, and productivity need not remain stalled if policy channels are used effectively. Accordingly, the minor premise that China lacks feasible tools or the willingness to deploy them is not established in this domain; the "has peaked" inference therefore does not follow from the evidence presented.

III. From "one-off" engines to reform dividends

The column claims:

Part of the problem is that China has reached the end of extrapolation. The past really is another country. Some of its growth engines could only ever fire once: for example, enrolling children in primary and secondary schools; improving basic healthcare; reaping the demographic dividend of falling dependency rates; and moving people from the countryside to higher-productivity, urban jobs.

Response:

The column lists a set of purportedly "one-off" engines and implies that the associated reform dividends are spent. As a matter of fact, the fading of a demographic dividend is precisely when reform dividends become most potent. Rapid population and output growth once masked frictions; as those frictions surface, reforms can and should step in. My responses to these asserted "one-off" engines are as follows:

1. Urbanisation is not a spent, one-off engine.

   1. As set out in Section II, China's resident urbanisation rate is about 67%, some distance from the 80% typical of advanced economies; historical experience suggests another decade or more of meaningful gains after the 65% threshold.

   2. The hukou (household-registration) rate is lower still (roughly 47%), leaving a sizable integration margin. Progress on hukou reform converts migration into full urban citizenship, with well-documented effects on consumption and labour productivity. In short, this "engine" has not run its course.

2. From demographic dividend to “talent dividend.”

   1. The right comparator for a maturing society is not how many people there are, but how productive they are. China's working-age population averaged about 11.05 years of schooling in 2023, roughly three years more than in 1982,⁹ yet still below the 13-plus years common in many advanced economies¹⁰. The gap is a policy-addressable margin.

   2. Policy steps are already in train. In August 2025, China issued guidelines waiving care and education fees for children in the final pre-primary year at public kindergartens nationwide (with reductions for approved private kindergartens) starting in the autumn 2025 term. This lowers the cost of early education, supports human-capital formation, and can be expanded or complemented by further measures.¹¹

   3. The logic is twofold: raising average years and quality of schooling lifts labour productivity ("talent" or "engineer" dividends), and on this dimension China still has considerable ground to cover; meanwhile, easing education costs reduces household burdens and can modestly support fertility. Neither channel is "one-off"; both are cumulative and policy-contingent.

In summary, treating past drivers as "one-off" mistakes the nature of development: what matters now is the reform dividend. The operative question is not whether the demographic dividend has waned, but whether there remains a credible policy toolkit and the willingness to deploy it. On the record, meaningful space persists across fiscal, regulatory, and institutional channels. Accordingly, the minor premise that China lacks feasible tools or intent remains unproven; the "has peaked" conclusion does not follow.

IV. Growth targets, debt, and fiscal capacity

The column claims:

China’s growth model, moreover, based on unrealistically high growth targets and uniquely high investment and savings rates, is becoming swamped by stagnant productivity, debt service difficulties and misallocation of capital.

Response:

What follows primarily addresses the column's assertions regarding "unrealistically high growth targets" and "debt service difficulties."

1. On "unrealistically high" targets

   1. The charge is asserted, not demonstrated. Since 2003 (setting aside the pandemic shock in 2020), China has met its annual growth targets. Just take the face value, a target repeatedly achieved over two decades is difficult to characterise as "unrealistic".

   2. Let's take a step back. Even if one insists the targets are "high," they are only modestly so, and by design, to guide expectations and market sentiment. That is a separate topic, and I won't go further here.

2. On government debt and fiscal space

   1. The ratio of central government debt or sovereign debt to GDP is a mere 21 percent, the lowest among the world's major economies. According to estimates from the International Monetary Fund, even when all local government debts (with hidden liabilities included) are added, China's overall level of general government debt is only around 110 percent of GDP. In comparison, the U.S. public debt ratio is approximately 140 percent, and Japan's stands at a staggering 260 percent. Therefore, China's government has a relatively low debt ratio.¹²

   2. Furthermore, one's financial well-being isn't determined solely by their liabilities but also by their assets. What sets China apart from other major economies is the vast ownership of state-owned enterprise (SOE) assets by the Chinese government.¹³

   3. Certainly, this doesn't imply that local government debt is a non-issue. There's a significant wealth gap among local governments…Yet, holistically, considering both assets and liabilities, the financial conditions of the Chinese government far surpass that of other major economies.¹⁴

In summary, both strands of the claim are weak on their own terms. First, the "unrealistic" label is unsupported given the consistent record of meeting targets. Second, debt-service pressures, while real in parts of the system, do not eliminate policy options or fiscal capacity at the sovereign level. Consistent with this newsletter's broader thesis, credible policy space remains; accordingly, the minor premise that China lacks feasible tools or the willingness to use them is not established here, and the "has peaked" conclusion does not follow.

V. Property market

The column claims:

The structural downturn in the property sector, which at one stage accounted for more than a quarter of the economy, is likely to shrink for the foreseeable future, dogged by lower rates of household formation and smaller cohorts of first-time buyers, linked to demographics as well as a chronic oversupply of unsold and uncompleted real estate.

Response:

China's housing crisis isn't over. That said, the claim of a chronic nationwide oversupply is not supported by the available indicators.

1. Stock arithmetic does not point to a nationwide glut.¹⁵

   1. According to the National Bureau of Statistics, China built about 14.4 billion square metres of saleable housing (excluding self-built rural homes) over the past three decades.

   2. In 2020, urban per-capita living space was 38.6 m². On that basis, the new stock over thirty years would house roughly 370 million people, just 40 per cent of its urban population of 920 million. The magnitudes here are inconsistent with claims of pervasive, economy-wide oversupply.

2. Market signals from the secondary (resale) market are inconsistent with "oversupply."¹⁶

   1. One sign of significant oversupply, per economist Gao Shanwen's cross-economy work (e.g., Japan, the US), is a prolonged contraction in resale transactions.

   2. In China, however, Beike Research Institute reports that in 2023 resale transactions rose sharply: up about 44% by floor space and about 30% by value. A market in persistent oversupply typically does not exhibit such a rebound in resale activity.

3. Investment ratios have fallen below a common oversupply threshold.¹⁷

   1. A second sign of oversupply in Gao's framework is residential development investment persistently exceeding 7% of GDP.

   2. China's ratio peaked near 12% in 2013 but has since declined; the figure is expected to be around 5.5% in 2024, below the oversupply threshold and consistent with a severe correction rather than a chronic glut.

In summary, acknowledging the crisis does not require accepting the oversupply claim. The data above suggest a deep, policy-sensitive overcorrection in residential development, not a structural, nationwide glut of housing. That distinction matters: it means the diagnosis in the column is, on this point, overstated, and again, consistent with the broader theme of this newsletter, policy space remains to address the imbalance.

VI. Conclusion

All in all, as argued throughout, the basic syllogism behind "Peak China" is not established. In most instances:

1. The minor premise, that China lacks feasible tools, or the willingness and capacity within a reasonable horizon to use them, is weakened or omitted. In other words, policy space and structural-upgrade offsets are underweighted or ignored.

2. On parts of the major premise, key indicators are framed through narrow or disadvantageous lenses (for example, nominal-dollar metrics during a strong-dollar phase), which is contestable.

Taken together, these points render the inference less than solid; the conclusion of "Peak China" is therefore not compelled.

None of this excuses the risks confronted by China. The country still has much to do, yet a fair reading leaves at least two paths beyond "peak": (i) a medium-speed rebalancing trajectory; and (ii) a policy-assisted acceleration if reform momentum and external tailwinds improve.

Given limits of space and attention, some topics here could not go deeper. In truth, each would warrant far more detail than a single note can carry. My purpose, again, is simply to broaden the lenses and metrics in play so that a more complete and measured chain of reasoning can be established. Enditem

1. See https://mp.weixin.qq.com/s/NpgrBKLs5erQLqlKdgTmew?scene=1

2. Ibid.

3. See https://www.stats.gov.cn/english/PressRelease/202502/t20250228_1958822.html

4. See https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS

5. See https://www.news.cn/politics/20250508/1137a5fb4cf247adb8d4f6ea6a9b32eb/c.html

6. See https://mp.weixin.qq.com/s/VVtdEPWbxZvMVidQlGRPhw

7. See https://www.gsm.pku.edu.cn/info/1022/29568.htm

8. Ibid.

9. See https://www.stats.gov.cn/zt_18555/ztfx/xzg75njjshfzcj/202409/t20240924_1956643.html

10. See http://www.moe.gov.cn/jyb_xwfb/s5147/202104/t20210408_525037.html

11. See https://english.www.gov.cn/policies/latestreleases/202508/05/content_WS6891ce94c6d0868f4e8f4a7f.html

12. See https://mp.weixin.qq.com/s/zk9VqUpgk3MBP8an4rPURg

13. Ibid.

14. Ibid.

15. See https://mp.weixin.qq.com/s/sFdbtcFTmC6CIcb2q2SRcw

16. Ibid.

17. Ibid.