Imbalances begin at home

We usually explain China's trade surplus with stories about virtue: hard-working workers, thrifty families, efficient factories. But a trade surplus isn't a measure of virtue. It is simply the amount by which a country produces more than it consumes and invests at home. That isn't a theory, it's accounting, and it changes the question entirely. The question isn't why China is so good at making things. It's why China is so bad at buying them.

The answer is that Chinese workers are paid far less than the value of what they produce. Productivity has been divorced from wage growth, and not by accident. It's the result of specific, deliberate policy choices. Workers can't bargain freely, savers earn artificially low interest so that borrowers can be subsidized, and a thin safety net forces families to save against illness and old age. Put simply, the people who make China's goods aren't paid enough to buy them. Household consumption is barely 40 percent of GDP, far below almost any other major economy. A country whose people can't consume what they produce has only two other places to put it. It can invest it, which China has done long past the point of usefulness, or it can export it. China has done both. So China doesn't run a surplus because it is winning at trade. It runs a surplus because its own people can't afford what they make, and the rest of the world has to buy it for them.

The argument in brief

  1. A trade surplus is unspent income. It is what a country produces but doesn't consume or invest at home. That is accounting, not theory.
  2. The core of it: productivity has been divorced from wages, by deliberate policy. Workers are paid less than the value of what they make, so they can't buy what they make. Banned unions, repressed interest rates, hukou, a thin safety net and a cheap yuan keep it that way.
  3. Investment absorbed the gap until it couldn't. First infrastructure, then property, then factories. Each outlet is now exhausted, and the only one left is the rest of the world.
  4. So exports outran demand. Since 2019, export volumes are up about 49% and imports about 8%. The goods surplus hit a record $1.19 trillion in 2025.
  5. China is taking share at every rung of the ladder, and its exports are growing far faster than world trade. Someone else must be losing.
  6. The US absorbs what no one else will, as debt, lost industry or asset bubbles. Tariffs only reroute the goods.
  7. It isn't really about China. Japan in the 1980s and Germany since the 2000s ran the same playbook. The conflict is between classes, not countries.
  8. Imbalances always end: through rebalancing at home, through trade war, or through crisis. The trade war isn't one president's project. It is what the system produces.

Since 2019, China's exports have grown twice as fast as its economy. Its imports have barely moved.

China, cumulative change in volumes since 2019, %

Redrawn from a chart by Brad Setser using IMF World Economic Outlook and Chinese national data. Values read from his chart, so approximate.

If China were simply getting richer, we would expect it to buy more from the world as it grew. That is what growing economies normally do. Instead, Chinese domestic demand has grown by about a quarter since 2019 while imports have grown by less than a tenth, and exports have outpaced world trade by twenty points. The economy is growing, but its people's ability to buy is not keeping up with its ability to produce. The difference has to go somewhere.

Saving depends on who gets the income

Saving − Investment = Trade surplus

When economists say a country "saves too much," it sounds like a statement about thrifty families. It usually isn't. National saving is simply everything an economy produces minus everything it consumes, and the biggest factor in how much it consumes is who receives the income. Households spend most of what they earn. Companies, banks and governments mostly don't.

Japan shows this clearly. Japanese households have barely saved at all for years, yet Japan still runs a large surplus, because the saving is done by its businesses. A country's saving rate isn't a measure of its virtue. It is a measure of how its income is distributed.

Wages are held below productivity, and everything follows from that

Here is the whole argument in one sentence: when workers are paid less than the value of what they produce, they cannot buy what they produce. The high saving, the overinvestment, the debt and the surplus all follow from that gap.

That gap isn't a natural stage of development. It is the product of specific, deliberate policy choices, none of them hidden, and each moves income the same way, from households to producers:

The road has run out

For nearly two decades, China has handled weak consumption the same way: by finding something else to absorb what households couldn't buy. After the 2008 crisis cut its surplus by about 7% of GDP, the answer was infrastructure. When infrastructure slowed, it was property, fueled by household borrowing at prices already at record highs relative to incomes. When property collapsed in 2021 and 2022, it was manufacturing. Each time, the growth target was met. Because consumption can't be raised by decree, a GDP growth target is really an investment target.

The usual reply is that China's investment produced remarkable things: high-speed rail, modern ports, world-leading electric vehicles. That's true, and it doesn't matter. Overinvestment doesn't mean investment produces nothing. It means it produces less than it costs. A project that uses 100 yuan of resources to create 80 yuan of value can be beautiful and still lose 20, and the loss doesn't disappear. It sits on someone's balance sheet as debt. The Soviet Union in the 1960s, Brazil in the 1970s and Japan in the 1980s all built impressive things this way, and all paid for it later. World-class infrastructure, soaring debt and record surpluses aren't contradictory. They are the same process.

Now every outlet is used up. Infrastructure and property are saturated, and manufacturing is so overbuilt that firms are destroying each other on price, which Beijing itself calls "involution." The losses have to be paid by someone. The last time, during the bank cleanup of the 2000s, households paid without knowing it: deposit rates were held below growth, transferring as much as 5% of GDP a year from savers to borrowers. That is a big part of how household consumption got so low. It can't be done twice. Squeezing households again would deepen the very problem it's meant to solve.

That leaves two kinds of payer at home. Businesses can pay through higher wages, higher interest rates, a stronger yuan and fewer subsidies, which means lower profits and slower growth. Or the government can absorb the losses directly. The one outlet that spares everyone at home is the rest of the world, which is why exports surged just as the domestic economy stalled. But the rest of the world is now closing its doors. That is what the end of the road means: Beijing can no longer choose whether China adjusts, only how painful the adjustment will be and who bears it.

China is taking share at every rung of the ladder

Development normally works like a ladder. A poor country starts with textiles and toys, climbs toward cars and chips, and leaves the lower rungs for the next poor country. Japan, Korea and Taiwan all did this, and Southeast Asia climbed up behind them.

China has climbed without letting go. Its share of world goods exports rose from 13.1% in 2018 to 16.3% in 2024. The new growth is in electric vehicles, batteries and machinery, so its exports increasingly compete with those of rich countries. Yet it still holds a historically unusual share of low-skill manufacturing, a squeeze that costs poorer countries hundreds of billions of dollars in exports they never get to make.

Then there is the arithmetic. China's exports are forecast to grow about 4% in 2026, while world trade grows about 0.5%. If the total barely grows and one country's share grows eight times faster, everyone else's share must shrink. China isn't simply riding global growth. It is taking it from others, at the top of the ladder and the bottom at the same time.

This isn't really about China

If the argument were only about China, it would be easy to dismiss as a story about communism or state capitalism. It isn't. Two rich democracies have run the same playbook.

Japan in the 1980s held household income down through a labor bargain that traded wage restraint for lifetime employment, along with high business saving and tightly managed finance. Its current account surplus reached 3.6% of GNP in 1985, and it became the target of America's trade anger, much as China is today. The Plaza Accord that year sent the yen from 240 to the dollar to below 150 within two years. That was meant to be the cure: a stronger yen raises households' purchasing power, and Japan's own Maekawa Report of 1986 called for exactly that shift toward domestic demand. But when exports fell and growth slowed, Tokyo flinched. The Bank of Japan halved its discount rate, cheap credit poured into property and factories, and investment climbed from about 30% of GDP to 35%. The adjustment was postponed, not avoided. When the bubble burst in 1990, Japan paid with decades of stagnation. China has been running Tokyo's response, more credit and more investment instead of higher household income, since 2008, and its excess investment may be larger than Japan's ever was.

Germany since the 2000s held wages back through labor reforms and years of restraint, while the euro gave it a currency far too cheap for its productivity. Its current account surplus reached around 8% of GDP in the mid-2010s and was still €203 billion in 2025. But where China over-invested, Germany under-invested: firms had little reason to build at home when German households weren't going to buy much more. Capital stock per worker fell, and infrastructure decayed. The same disease produced the opposite symptom.

The common thread isn't a political system. It is an economy in which households receive too small a share of what they produce.

The US doesn't overspend. It is forced to absorb.

Excess savings need a home, and the US has the deepest, most open and most trusted capital markets in the world. Foreigners bought a record $1.55 trillion of US securities in 2025. When that money arrives, the dollar rises and American industry shrinks, and the inflow must show up somewhere inside the US: as more household debt, more government debt, inflated stock and housing prices, or more unemployment.

So Americans don't save little because they are profligate. They save little because the US has to absorb everyone else's excess saving. The deficit is the other side of other countries' surpluses.

That is also why bilateral tariffs fail. After the US put tariffs on China in 2018, China's surplus and America's deficit both kept rising, just not with each other, as goods rerouted through third countries. The imbalance lives in the capital account, and tariffs don't touch it.

Exports are the cost of trade. Imports are the payoff.

"The benefit which is derived from exchanging one commodity for another, arises, in all cases, from the commodity received, not from the commodity given."James Mill, Elements of Political Economy, 1821

This is free-trade orthodoxy, from Mill to Friedman to Krugman. A country exports so that it can import. A chronic surplus country pays the cost of trade and refuses the payoff: its workers ship real goods abroad and receive financial claims in return. It produces too much only in one sense, which is relative to what its own people are allowed to buy.

A conflict between classes, not countries

Seen this way, the trade war isn't China against America. Chinese workers are paid less than they produce. American workers lose factory jobs, or borrow to keep spending. Exporters, state banks and the owners of financial assets do well on both sides of the Pacific. The losers on both sides are workers.

That is also why rebalancing keeps failing to happen. Moving income to households means lower profits for state firms, less revenue for local governments and slower growth, perhaps 2 to 3% a year for some time. The losers are concentrated and powerful. The winners are hundreds of millions of households with no lobby. Beijing first promised to rebalance in 2007. The household share is where it was then.

Imbalances always end. The question is who pays.

The trade war isn't Trump's invention, and it won't end with him. When surplus countries won't spend and deficit countries can no longer absorb, conflict is the system's default outcome, whoever sits in the White House. Protectionism is a symptom of the imbalance, not its cause. There are only three ways out:

  1. Rebalancing at home. Surplus countries raise household income through wages, a stronger currency, a real safety net and hukou reform. Their consumption rises, their surpluses shrink, and everyone's markets grow. This is the only good ending.
  2. Deficit countries stop absorbing. Tariffs, quotas and capital controls. The US has started, and Europe, Mexico and Brazil are following on EVs and steel. Each country that closes its door pushes the surplus onto whoever is still open, until no one is.
  3. Crisis. Past episodes of extreme imbalance ended badly: the Great Depression, the Latin American debt crisis, Japan in 1990, Asia in 1997, the eurozone after 2008.

The last time imbalances this large unwound without cooperation was the 1930s. The US was the great surplus country then, and when the adjustment came it was hit hardest. World trade collapsed under tariffs and competitive devaluations, and economic nationalism hardened into a scramble for markets and resources that ended in war. A century ago, Hobson argued that underconsumption at home drives countries to fight for markets abroad. It isn't a comfortable precedent.

None of this denies China's achievement. Suppressing consumption to build capital helped lift hundreds of millions out of poverty. The model has simply outlived its usefulness, and running it now harms Chinese households and everyone else.

The best objections, answered

"US deficits come from low US saving, not foreign surpluses."

The identity runs both ways, so the question is which side drives. When foreign capital pushes the dollar up and wipes out tradable jobs, the US must accept unemployment or borrow to keep demand up. Low US saving is the result, not the cause.

"The first China shock made the US richer overall."

Several papers find net aggregate gains. Totals hide who paid: concentrated job losses and debt-financed spending. And this shock is different. It hits advanced industry and poor countries at the same time.

"China's surplus is small relative to its GDP."

It is a record share of the rest of the world's GDP, now above the peaks of Germany and Japan. Setser estimates China's true current account near 5.5% of GDP, not the reported 4%. China is too big for its surplus to be absorbed quietly.

"China is just more productive. That's comparative advantage."

Comparative advantage means specializing: getting better at some things and buying the rest. China is gaining share in almost everything while its imports stay flat. That isn't specialization. It is suppressed demand.

"People have predicted this rebalancing for over a decade."

The thesis says rebalancing is politically hard and that growth slows if it is put off. Slowing growth, a property crash and surging exports are exactly what it predicts.

"Cheap imports are a gift. Say thank you."

A gift paid for with suppressed wages at home, and with lost industry and debt abroad, is a transfer, not free trade. Productivity gains should show up as higher household income and more imports. The chart shows they haven't.

Sources (figures as of November 2026)