Why China Keeps Doing the Thing Everyone Tells It to Stop

China's leaders are not confused about their economy. They have named its problems out loud: too much investment and export, too little household spending, an imbalance that is picking fights with trading partners rich and poor. Economists abroad — and quietly, many inside China — warn that without a sharp change of course the country faces slowing growth, cascading trade disputes, and stagnant incomes.

And yet Beijing barely moves. It announces modest measures to lift consumption, then keeps pouring money into manufacturing, advanced technology, and new strategic industries. The puzzle is not that China misread the data. It is that China sees the same data everyone else sees and reaches the opposite conclusion. To understand why, you need two established ideas: one about how planners picture an economy, and one about why winners keep doing what used to win.

The facts, which no one disputes

Rising trade surpluses. Weak consumption. Slowing household income. Feeble demand for credit. Rising unemployment and underemployment. Persistent deflation — falling prices, which usually signals an economy producing more than it can sell.

Outsiders read these as symptoms of a growth model that has run out of road. Beijing reads them as the expected friction of a strategy that is working.

The first model: the economy as an input-output table

The intellectual machinery here is real and has a name. In the 1930s the economist Wassily Leontief built the input-output table — a grid in which every sector is a cell that buys inputs from other cells and sells outputs to them. It won him the Nobel in 1973 and became the backbone of Soviet-style planning. It is a genuine tool; the question is what you do with it.

Mainstream (neoclassical) economics treats growth as capital, labour, and productivity responding to market demand — prices do the balancing. Beijing has kept the table and swapped the engine. In its version, development equals capital accumulation: productivity comes from controlling technology and physical capital — factories, machines, patents, infrastructure. Everything else — higher wages, a thriving service sector, more household spending — is treated as a consequence that arrives later, once the productive base is strong enough. Consumption is not the goal; it is the reward collected after the industrial work is done.

So when a price runs too high or too low, the planner's instinct is not to let interest rates and exchange rates rebalance things. It is to find which cell is overinvested and which downstream cell is lagging, then intervene directly — cheap credit here, slowed production there — until the matrix balances. Weak household income is not read as failure. It is read as discipline: savings redirected from "frivolous" consumption into priority investment.

Watch it in practice:

• When property — nearly a quarter of GDP at its peak — collapsed in 2022, Beijing did not loosen broadly. It tightened control: reversing interest-rate liberalisation, micromanaging credit, reasserting oversight of capital markets.
• Central bank facilities funnel low-cost loans to favoured sectors for equipment and upgrades.
• Thousands of government equity funds hand cheap capital to firms serving industrial strategy.
• When a sector overheats into overcapacity, campaigns like "anti-involution" slow investment, fix prices, or raise entry barriers — buying time for other cells to catch up.

Deflation and excess supply, in this worldview, are not alarms. They are the cost of running the engine hot. And there is real evidence behind the confidence: Chinese firms now lead in autos and telecoms and are closing on robotics and biotechnology — all while absorbing a property shock that would have flattened most economies.

Where the logic breaks

The model's hidden assumption is that investment in capital automatically raises productivity and therefore wages. But investment only raises productivity if it is connected to demand — if someone wants, and can afford, what the new capacity produces. Growth is not the same as returns; building more factories is not building more value.

Here size matters. A small economy can out-invest its own weak demand and export the surplus. China is already the world's largest exporter; it cannot keep flooding foreign markets without triggering the trade walls now rising against it. When domestic demand stays suppressed and exports hit a ceiling, the extra output has nowhere to go — which is precisely what deflation is telling you.

The competing explanations, fairly stated

Several rival accounts circulate, each with some truth:

• Xi is badly informed — insulated by a system that punishes bad news.
• Priorities have shifted — national security and control of strategic industries now outrank living standards.
• Beijing knows but can't act — a pivot to consumption would mean redistributing power away from state banks and local governments, which the system resists.

These shape policy at the margins, and none can be ruled out. But the simplest and most uncomfortable reading is that the core driver is none of them. The leadership is not paralysed, misinformed, or indifferent. It is sincerely convinced the strategy is succeeding — because within its model, the troubling data is supposed to look exactly like this. A confused leader can be corrected. A leader running a coherent theory that happens to be wrong will keep applying it with discipline, reading each new symptom as confirmation.

The second model: the competency trap

That last sentence is itself a named, studied pattern. Organisational theorists Barbara Levitt and James March called it the competency trap (1988): an organisation that succeeds with a given routine keeps investing in it, gets better at it, and so becomes less able to see when conditions have changed. Success doesn't just reward a strategy — it hardwires a theory of why you succeed, and that theory is the thing you stop questioning. It is not unique to Beijing:

• Japan in the late 1980s ran an export-and-investment machine that had delivered three decades of miracle growth, then fed it past the point of returns. The instincts that built the boom inflated the bubble.
• A firm that grew by cutting costs keeps cutting long after the gains are gone, because cost discipline is the story of its success.
• In your own career, the skill that earned your early promotions becomes the thing you over-apply — right up to the level where a different skill was needed.

The danger is never a stupid idea. It is a once-correct idea held past its expiry, defended by the very evidence of its earlier triumphs.

What to watch

The decisive signal is deflation. If Beijing is right, the input-output approach should gradually rebalance supply and demand, and falling prices should ease as the matrix catches up. If the outsiders are right, deflation and weak consumption will persist no matter how precisely credit is steered — because the problem is not a mistuned cell but a model that under-feeds households by design.

Also watch whether consumption measures stay cosmetic. A genuine pivot would move real resources — higher wages, stronger pensions, better public services — away from industrial investment. Modest spending announcements that leave the investment-first core untouched signal continuity, not change.

Expectation: ~70% that through 2026 China's core strategy stays investment- and manufacturing-led, with only marginal consumption measures, and that broad deflationary pressure persists rather than resolving. The clearest falsifier would be a structural shift of resources toward household income — a sign the underlying model, not just the policy mix, had finally moved.

Distilled from Foreign Affairs