The tide data this month is hard to look away from — except that in this case, the tide is retail spending, and it is flowing toward China’s county towns. On August 18, nine central agencies including the Ministry of Commerce issued a policy package with 18 specific measures to revive the market in lower-tier areas and activate county-level consumption. The headline deserves attention, but the numbers underneath deserve more.
Let me start where I always start, with scale. County jurisdictions hold roughly 90 percent of China’s land area, about half of the resident population, and close to 40 percent of economic output. That is not a niche. That is the majority of the country, measured by territory and by people, sitting in places that the consumption conversation has mostly ignored for two decades. When I read a policy document, the first question I ask is whether the target matters. On land, on people, on output — this one does.
The quiet record that says something is changing
Here is the statistic that made me stop. Rural retail sales of consumer goods have grown faster than urban ones for 55 consecutive months. Not a quarter. Not a year. Fifty-five months, which is just short of five years. I have been trained to be careful with streaks like this, because streaks in data are exactly what seduce people into seeing a trend that is not there. But five years of outgrowing the cities is not a blip anymore. The data shows a current that has been running for half a decade, and policy is finally building harbors for it.
That is the real news in the August 18 document. The 18 measures are not a vague call for more spending. They are a supply-side effort: getting goods, services, logistics, and consumption infrastructure into places that have been treated as the far shore of the economy. The difference matters. A document that says spend more is a hope. A document that builds the plumbing for spending is a decision.
Why I used to be sceptical about the county story
Let me be honest about my own read, because I got this partly wrong. For years I looked at lower-tier markets the way I look at the deep ocean floor: vast, poorly mapped, and slow to change. The working assumption was that consumption growth would keep hugging the big cities, where incomes are higher and logistics are cheaper. I treated the rural number as a rounding error — small base, flattered by percentages.
No, that is not quite right. It is closer to the truth to say the base was never small. It was just measured in places I was not looking. Half the population is not a rounding error, and 55 months is not a fluke. What I misread was the direction of the current, not its existence. The correction is worth stating plainly, because the same mistake is still common in the commentary I read: treating county towns as a market that is simply poorer rather than a market that is differently organised.
What the numbers actually show — and their limits
Let me lay out what the evidence supports, carefully, because a policy document is not a proof. The verified facts are these: the nine-agency notice exists and contains 18 measures; county areas hold 90 percent of land, half the people, and about 40 percent of output; and rural retail sales have now outpaced urban for 55 straight months. Everything else — how fast consumption will accelerate, which categories will win, who profits — is projection. I keep the two columns separate on purpose.
What can be said soberly is that the arithmetic no longer fits the old story. When the majority of the population lives somewhere, and their consumption is consistently growing faster than the rich cities, then the marginal consumer in China is increasingly a person in a county town. That is not a forecast. That is where the weights already sit. The projection is only about how much further the current will run.
The logistics problem that decides everything
The constraint, and I would bet on this the way I bet on a known current, is not demand. It is delivery. A person in a county town does not lack the desire to buy; what has been missing is the infrastructure that makes buying feel normal — same-day or next-day delivery, after-sales service that does not require a day trip to the city, and shelves stocked with the same brands a city resident takes for granted.
Evidence before alarm, as always: the policy package directly targets this friction. The 18 measures reach into the mechanics of making consumption work at county level — the network, the supply chain, the everyday costs that quietly tax every transaction. That is the part that gives me confidence the shift is serious. When policy spends its effort on the plumbing instead of the slogan, something is changing under the surface.
What a consumer in a county town actually experiences
Think about one concrete moment. A family in a county town orders a washing machine on a Monday afternoon. In the old model, the order crossed several warehouses, the delivery window stretched toward a week, and if the machine arrived damaged, the return was a campaign involving phone calls, photographs, and a long wait. In the new model, the delivery radius shrinks, the cost of an order falls, and the shop on the corner stocks what the family would previously have bought a hundred kilometers away.
I have watched this pattern in the data for a while, and it repeats everywhere it appears: consumption does not rise because people are told to spend. It rises when spending stops being expensive and difficult. The tide does not move because someone shouts at the ocean. It moves when the gradient changes.
The income question nobody has answered well
I should add the caveat I keep returning to. There is a difference between consumption growing faster in the countryside and consumption catching up in absolute terms. The streak is real, but the levels still trail the cities by a wide margin. Higher growth on a smaller base is a current, not a flood. Whether the county market becomes the engine everyone now hopes it will be depends partly on whether incomes in those places rise with it.
So far the evidence is consistent with the hopeful version and not yet with the triumphant one. I do not have this fully figured out, and I would be suspicious of anyone who claims to. What I can say is that the old assumption — that the growth story lives in the big cities alone — is no longer defensible with a straight face.
Policy text versus delivery on the ground
One more caution, and it is the one I have learned the hard way. A document is a signal, not a result. The 18 measures are the plan; the implementation is the proof, and implementation in county towns takes years, not quarters. I have learned to respect the distance between a policy’s text and its delivery on the ground. The distance is where most well-intentioned plans go to die.
That is why the indicator I watch is not the announcement. It is the streak. If the 55-month record survives a quarter when the cities rebound hard, that tells me the shift is structural rather than arithmetic. If spending categories change — if the growth stops being basics and starts looking like the mix you see in a mid-sized city — that tells me the county market has arrived as a consumer market, not just a discount market.
What the tide means for anyone reading the economy
Step back for a moment. The country’s growth story has, for two decades, been written in its coastal and provincial capitals. What the last five years of data suggest is that the next chapter is being written in county towns — ninety percent of the land, half the people, forty percent of the output. When the majority of a population outspends the minority for five straight years, the structure of demand has changed, whether or not the headlines have caught up.
There is a temptation, and I felt it myself when I first read the notice, to read this as a story about poverty relief dressed up as economics. That reading is too thin. The sobering part of the data is that the gap between city and county is narrowing in growth rates while the absolute distance remains wide — which is precisely the pattern you would expect from a market that is being wired up rather than one that is being subsidised. Subsidies do not compound for five years. Infrastructure does.
Let me also say something about how to read the 18 measures, because the number itself invites the wrong kind of attention. Eighteen measures sounds like a checklist, and checklists are how policy documents reassure the public without committing anyone. But the content of this package, as reported across official channels on August 18, is unusually operational: it reaches the level of how goods are supplied, how stores are stocked, and how a county resident’s ordinary purchase is made cheaper and easier. That is not a slogan. It is the kind of detail that changes what a Tuesday looks like for a household that previously treated a shopping trip as an event.
Reading the numbers the way you would read a current meter
If I were advising someone who has to allocate time or capital on the basis of this, my counsel would be simple. Do not chase the headline dates. Do not assume the policy lands evenly across nine agencies and thousands of counties. Instead, watch three instruments. The first is the delivery network in a single mid-sized county — whether the measure of cost and time for a standard order actually falls. The second is the 55-month streak itself. The third is the mix of what is being bought, because a county market that starts buying like a city is the strongest signal there is.
One thing I keep telling myself when I look at a change this big is to check what is different this time. County consumption has been discussed for years, and every discussion ended the same way: the logistics make it unprofitable, the market is too dispersed, the numbers are too small. The data shows that the smallness argument has quietly stopped working. A population share of roughly half cannot be called too small for anything. It can only be called underserved.
Measured over five years, the county market has been the quiet engine of Chinese consumption, and policy has now publicly bet on it. That is worth more than a headline. Alarm adds nothing to the evidence; the evidence is already striking on its own. The current is running, the harbors are being built, and the rest is a question of how deep the water gets.