Bryan Caplan  

Is Chinese Growth Credible?

Classic Economics... Mortgage Fraud...

I've long been skeptical of Chinese growth numbers. I don't doubt that China's economy is growing rapidly, but year after year of 10% growth seems incredible. In addition to the generic argument for doubting extreme numbers, we have the long history of Communist regimes falsifying their numbers, and Western observers gullibly taking their lies at face value. Fool me once, shame on you; fool me twice, shame on me, right?

In the latest issue of the Economist, I read a couple other facts that reinforced my skepticism. I don't claim that they are a smoking gun, but they are certainly weird.

Weird Fact #1: 60% of the Chinese remain in agriculture, where they produced 12% of GDP:

For the past three years rural income per head has risen by more than 6% annually in real terms. In the first half of this year, pushed by fast-rising food prices, it was up 13%, the highest increase since 1995 according to official media...

Rural China is still home to about 60% of the country's 1.3 billion people, but agriculture's contribution to GDP has fallen from more than a quarter in 1990 to less than 12% today.

Are there any other major economies that fit this pattern? Can the average non-farm worker really be 11 times as productive as the average farmer? What other countries with a purchasing-power parity per-capita GDP of $7600 are 60% agricultural? (Aside: Could rural workers engaged in non-agricultural pursuits be enough to make these numbers reasonable?)

Weird Fact #2: Chinese consumer spending is only 36% of GDP:

China will soon boast seven of the world's ten biggest shopping malls. Yet Chinese households are hardly the most eager shoppers. Consumer spending has fallen from 47% of GDP in the early 1990s to only 36% in 2006, the lowest proportion in any large economy... At the other extreme, American households consume 70% of GDP.

Normally rapid growth is a reason to borrow against future earnings; so why do the Chinese appear to be doing the opposite? Perhaps the simplest explanation is that it is much harder to fake high consumption than high GDP.

Any other Chinese growth skeptics out there? What should I be reading?

Comments and Sharing

TRACKBACKS (1 to date)
TrackBack URL:
The author at Full Comment in a related article titled Colby Cosh: Underdiscussed-issue-with-cosmically-huge-consequences alert writes:
    Are the spectacular Chinese economic growth numbers of the post-Deng era reliable? The West has been [Tracked on October 22, 2007 3:28 AM]
COMMENTS (22 to date)
Alex from Hong Kong writes:


You have to come to China to see for yourself. I was once just like you, completely skeptical of Chinese numbers, but having been on the ground here in HK/China long enough, I am now convinced that the GDP figures are actually understated for lack of ways to collect real data. Millions of small businesses have been started and nobody knows how many apartments have been built.... Also the number of cars sold each year. When I first visited China 27 yrs ago, there were hardly any cars at all, but for the top tier officials.

A farmer in China earns less than US$100/mth and a factory worker about US$200/mth. 25 yr old analysts working for my company used to make US$500/mth two years ago, now they are paid US$1,500 and going up rapidly....

City worker wages are rapidly catching up with rest of the world, yet rural farmers are pretty much stuck in their own time-warp.

As for the high savings rate, it's for the lack of availability of consumer credit and lack of social safety net, people tend to save for a rainy day.

Jason writes:

I work in China/Hong Kong, and I think you're broadly right. Certainly, in the major cities like Beijing and Shanghai, you might believe that China is growing that rapidly. But once you start heading out to the countryside, things start to fall apart quickly. The Communist Party is pretty strict about what comes out of rural China. The incentive for the party to overstate growth, the restrictions on media access, and the geographic location of most foreign groups are all factors, IMHO.

I think there is a lot of anecdotal evidence of bubbles in China as well - the huge disrepancy between stocks traded on the HK stock exchange and the Shanghai stock exchange, and the fact that China has more publicly traded companies with market values >$200 billion than the US (8 to 7).

Should be an interesting show.

[duplicate posts removed--Econlib Ed.]

Bob writes:

Lester Thurow argues that China's numbers cannot be real because electricity consumption has not followed the path that he'd expect if the country were growing at reported rates. A clever angle, although I'm not fully convinced given that consumption in general seems to be well below what you'd expect (that is, he may just be observing weird fact #2). If you have any thoughts on his theory, I'd like to hear them in a future post.

You can find Thurow's thoughts summarized in his August NYT column:

Also, Robert Fogel has a NBER working paper that may be of interest.

anon commenter writes:

Thurow's argument is quite convincing, and there's another factor he only obliquely referred to, which is the fact that factories in the third world are known for generally consuming more fuel per unit production than their counterparts in developed countries.

Blaming low consumption for the discrepancy is pretty unconvincing when you consider that electricity use in industry and commerce SHOULD be skyrocketing along with the increase in production.

Wojtek Grabski writes:

I've always wondered how these spectacular announcements compared to those made in Soviet times -- when I was to young to pick up on them -- glad someone's finally picking this topic up; perhaps some straight stat would do the trick. See in how many ways China is an outlier of the typical development curve. Like you said, it's tough to fake all the numbers well, so some of them wouldn't fit -- presumably by something proportional to the difference between the reported/real growth rates.

Bob writes:


Inefficient factories works the other way because we're talking about growth rates - if factories are inefficient but becoming more efficient, then electricity (and energy generally) consumption will lag what you'd otherwise expect based on overall growth rates. That said, I give Thurow's theory reasonable odds, with his comparison to Japan during its industrialization being the most convincing empirical evidence. Even more important to my thinking, however, are the reporting incentives that government officials have at all levels.

Carl Shulman writes:

Some countervailing thoughts:

1. There are rigid restrictions on labor mobility in China, and permits to move from areas of grinding rural poverty to booming industrial/urban zones are very scarce (the exclusion of the income of illegal migrants from statistics may exacerbate the income gap).

2. There is a large underground private finance infrastructure in China financing productive businesses.

3. China effectively has *really* free market health care for many people, with poor insurance markets, so people save massively in order to pay for life-saving medicine out of pocket.

Floccina writes:

"we have the long history of Communist regimes falsifying their numbers, and Western observers gullibly taking their lies at face value."

This always comes to my mind when people bring up life expectancy and infant mortality statistics for Cuba. But so many people (Michael Moore included) take the Cuba statistics as accurate. Sometimes I am so dumbfounded that I have to ask myself if they know something that I do not.

Still my position is to dismiss any statistics coming out of Cuba as unknowable.

8 writes:

Lies, damn lies and Chinese statistics

...Take two municipalities for example: Beijing declared that its 2004 interim GDP increased by 15.4% year-on-year, while the adjacent Tianjin city quoted 15.9%. Given that the former is preparing itself for the 2008 Olympic Games, its rapid GDP growth is understandable, but no one has any clue as to why Tianjin is growing even faster.

The article is three years old, but relevant. Chinese officials are promoted partially based on economic growth. Growth is measured using GDP. If you build stuff, GDP goes up. Local Chinese governments build a lot of stuff. Outside of Shanghai there was (in 2005, I haven't seen it since then) an empty super highway far outside the city limits. Also, it's getting filled by residents now, but Pudong, about the size of Chicago (where those famous Shanghai skyscrapers are) was a ghost city only a few years ago.

This article is more focused on officals' incentives.
Damn lies and Chinese statistics

If you poke around you can find lots of articles by people who don't believe the numbers.

Money supply could be "helping", it was growing at over 20% a year for awhile, and there's definitely some asset inflation— apartment prices tripled in less than one year in the third tier city I lived in, and they've yet to decline. The government restricts land development too, which helps inflate prices. If you talk to anyone in a city, things are great. They have jobs, their assets are appreciating, etc. Elsewhere, life still really sucks, but on the whole it is improving.

Bosco writes:

Even if China's numbers are real, isn't a slowdown inevitable? China has relied largely on capital for growth, and will eventually catch up to the West in total factor productivity. Once they are unable to borrow other countries' technology, doesn't the neoclassical growth model predict a much lower growth rate?

Rimfax writes:

The "Depression Era" mentality that almost certainly pervades the older generation could be driving up savings.

What is the distribution of productivity in other nations? Is there labor group that makes up 60% that produces less than 15% of the US GDP? What percentage of the US labor force are lawyers? government employees? prisoners? auto workers? (I keed.)

If they are indeed inflating their GDP, is there a reckoning? Will there be a price to pay someday for overstating their GDP?

Carolina writes:

Having just come back from running an economic experiment in Beijing, I think I can explain the weird factoids.

Keep two things to keep in mind:

(a) It's not the individual but the collective productivity that is driving China

(b) Chinese entrepreneurs and workers don't like being counted - they hide their work and their income.

Why do I say this?

1. 24/7 and cheaper than you'd think - China doesn't close. Period. You need a building up? Your workers work all night. They don't use electricity to run their tools because they don't have power tools. I wish I was kidding.

1.a. Also electricity is run on a pay-as-you go grid. Most urban dwellers don't have access to it, so forget including them in your analysis. I've seen the Chinese Electrical stats and each month they start up the new power plant capable of powering a city the size of San Diego. According to the IEA, in 1980 they were generating 1/2 of what Japan was using and by 2005 they were generation twice the power of Japan. Not bad for a country where at least 60% of the population has not encountered electricity.

1.b Costs are lower - and not just in manufacturing. You want a cheaper building? Don't pay your workers until the end of the year after you've deducted for housing, food, and site security. This lowers your costs immensely. And don't even worry about health care.

1.c Taxes? That's for foreigners. Most small Chinese businesses (employing less than 20 workers) pay taxes when they issue an official receipt. And they issue that receipt maybe 1/4 the time.

2. The countryside is rotting. Although CPI is growing at 6.2% a year, the price of food is dropping as fulfillment gets better. A single city worker could eat like a king on the minimum wage. The countryside can't, which is why they're flooding the city with illegal immigrants.

2.a. I think 12% of GDP is overstating the countryside, particularly if that number is solely on farming. If it includes petrochemicals and mining, then I might believe it. But no investment in the countryside means a lot of "unproductive" infrastructure improvements aren't happening unless they're being driven by business (see mining communities, for example).

2.b. If a urban business says they employ X workers for 8 hours, they're really employing 3X workers for 20 hours to produce a good. The only people in the cities being counted by the government are the official residents of those cities. All the illegals hide their income. Official residents have an easier time being hired and are less likely to be exploited. They, after all, are protected by the official daily minimum wage (30 kuai). Illegals are dependent on their employers who pay them a lot less than locals. Still, it's a better life (or is perceived as such) than the countryside.

3. Credit is hard to come by. It's a cash society, and those who have credit have it through elaborate social networks or have enough non-Chinese held wealth. The average worker doesn't, an (as peasant economics has suggested for years) is risk averse. They're holding their cash because who knows with the bad times are going to come. Lack of a social safety net and the rapidly rising costs for everything means you're only spending if you think can afford to spend.

4. Big G: Don't discount the Olympics. Say what you will about government spending on giant projects, but the +66 Billion USD being spent on the Olympics is being spent without massive corporate sponsorship. Sure Yanjing, and Coca-Cola got into the act, but the money for new highways, anti-pollution measures, and all the venues is coming out of government coffers. Where's the multiplier? It's driving a real estate and housing boom. Jones Lang LaSalle predicted the building of over 5000 5 Star hotel rooms between 8/2007-8/2008. Even more 4 and 3 star hotels are being built. That's some serious non-Chinese investment.

5. Richer chinese only buy imports. Case in point. I wanted el-cheapo cell phone. I, non-chinese, bought el-cheapo chinese made cell phone. All my Chinese peers and hosts were appalled and lectured me on how only foreign goods were worth buying, and how only farmers bought Chinese.

Which was a long way of saying -

Yes, the country/city divide is real but it's probably not as wide as the numbers suggest. If you don't count the illegals, the numbers make sense. If you do count the illegals, then the cities aren't doing as well as you think and the Gini Coefficient needs some work. The Chinese government knows this and so embarked on its harmony campaign to remind the people not to get uppity.

The rich are spending money on non-Chinese goods and the poor are buying Chinese goods. Without credit, the poor will only buy when (a) they must or (b) they feel they have enough emergency cash on hand. So the spending of the poor (which may not have a taxable receipt to record the transaction) might not be counted as often as more expensive consumption. C is being (IMHO) under counted.

So, in other words, I do believe china is growing more rapidly than, say, India. But I think their numbers (and everyone else's) are lousy. Fortunately, the Chinese government is acting as if those numbers are lousy (nice to see someone isn't drinking the Kool-Aid) esp, if you pay attention to their increasing concern over CPI, worker wages, and increasing number of regulatory bodies.

Alas, it appears the Chinese government lacks the ability to actually impose more regulation over their economy. It's the Wild, Wild, East over there.

Dave writes:

I suspect on factor for the huge savings rate is the high (for a developing country) median age of Chinese as a result fo the 1-child policy (regardless of what the official stats say about age-the Chinese have an incentive to make their country look younger than it is-makes possible future enemies more concerned than need be)

Jasbir writes:

would electricity theft in anyway impact on the consumption numbers?

jaim klein writes:

Making sense of China's economy is difficult, as it IS as communist dictatorship, and people has learned to be very careful with what they say or show, as one thing today (like employing other people) could easily mean a crime and exile in the future. People is saving like crazy, that is a fact, the Japanese when they were in the same level of development, also saved. In my opinion all government data is made up, not because of some evil intention, but because China is a very large country and the Government has little control at the field level.It well could be 36% or 56%, if US statistics are uncertain, China's surely contain much interpolation and imagination.
10% annual growth is not unheard of, Japan did it for decades. Another point: China has large populations of non Han or semi Han indigenous peoples (maybe up to 20% of the population) which are living in subsistance agriculture and disconnected from national economy. These half assimilated populations are counted as Han, and they dont exist in national demography or discurse. Travelling in China you find large pockets of small stature (asimilar to Bolivian quechwas) peoples working the land with wooden hand tools, without even animals. This pockets of poverty will exist even when average Chinese GNP reaches American levels.

Barkley Rosser writes:

If electricity output has been growing so slowly, why is China now passing the US in aggregate carbon emissions and what is happening with those two coal fired plants per month they are reportedly building.

Actually there is a good argument for why they may be understating their numbers consciously. Whereas in the earlier period they wanted to propagandistically impress everybody, now they want to maintain privileges they get for maintaining a "developing country" status in the various international organizations they belong to such as the WTO. This provides an incentive for understating the numbers.

I would say the case for understatement is at least as strong as that of overstatement.

Also, last I read, a majority of Chinese are now in urban areas. If that is the case, where does this "60% in agriculture" come from? Or do we have urban farmers?

Demor Anderson writes:

Growth ,may be high but both distorted and unsustainable. There is a potentailly very large boost through hidden fiscal expansion via the banking system. The best blog on China in my opinion is by an American Peking University finance professor ( and he discusses this several times.

Scott Sumner writes:

In the late 1970s I thought the Soviet bloc growth rates were grossly overstated, which later proved to be the case. For several reasons, however, I am confident that the Chinese growth rate is extremely high, probably close to the reported figures.

1. Cheating would be much harder than you think. We have all sorts of sectoral data in everything from coal and steel to cellphones, tvs, homebuilding, and autos, that all point to explosive growth. They would need to not only exaggerate the output of these sectors, but exaggerate by rapidly increasing margins every years. Let's say output in all the industries that I described is only half the reported output (highly implausible.) Even then the average annual growth in China would be only slightly below than reported growth over the past 30 years. And they have very detailed sectoral data. Cheating on that scale would be easy to detect.

2. Personal observation. I have visited many coastal cities, inland cities and rural areas over the last 13 years. Yes, the growth looks fastest in the coastal areas but it looks very fast everwhere.

3. Visitors who have been to both counties all say that China's growth seems far, far more impressive than India's growth. Is their 8% figure also made up?

4. On a PPP basis consumption is actually much more than 36% of GDP. Services like haircuts are dirt cheap and much small business is off the books (as others have noted.) In contrast, exports are at world prices.

5. People in the rural sector often do manufacturing and services--hence their productivity is higher than one might expect from simply looking at just their agricultural output. But urban productivity is much higher, as in many developing countries.

6. Thurow's numbers were way off. Even China's rural sector is growing fast, just not as fast as the cities. Much of his data is just plain wrong.

7. No one seriously questions China's export stats. Yet they make the reported GDP number seem much too low--as they are implausbly high as a share of GDP for such a big country. (Again the PPP issue helps explain this.)

8. Their $3 trillion GDP is probably $9 trillion on a PPP basis--and may surpass the U.S. total in 5-10 years. But remember that in per capita terms they would still only be at Mexico's level. The rapid growth is a sign of an economy going from being unbelievable inefficient, to merely being grossly inefficient. Remember that China's per capita GDP was below that of India and Africa in the late 1970s. The rapid growth isn't a sign of a wonderful system today, it's a sign of an incredibly screwed up system under Mao. I share your libertarian perspective--nothing in China's record conflicts with the view that complete market reform would produce even better results than the half market system in place today. The Chinese know they need further reforms to catch up to the first world--and they are pragmatic enough to gradually keep reforming their economy.

Barkley Rosser writes:

For those who think that Chinese numbers are probably overstated because that is what Communist-ruled governments do, I would point out that the overreporting in the past was associated with command central planning. There was pressure to meet the quota, which kept getting ratcheted up. Widespread lying and fraud was the result.

However, in China for some time there has been no command central planning. It may still be a Communist dictatorship, but it is a market economy, however weird and distorted.

Michael Taylor writes:

Carolina is largely right. I'd add one thing: most SMEs in China - and they can be very large - run mostly for cash, because they can't get credit from a banking system which a) used to be just instruments of fiscal policy, so made "mistakes" and didn't care and b) are now listed, so daren't make "mistakes" any more. Result, if you're out in, say, Xian, and you want to expand your factory, you'll have to persuade bank's head office in Beijing for any capital loan over about Rmb 400m. Result, the mittelstand runs for cash, which results in extremely high capital productivity, and elsewhere in an extremely high savings ratio. . ..

As for the urban/rural thing - it's all in the counting - "urban resident" and "rural resident" doesn't necessarily mean what you think.

Still, as everyone knows, GDP's are estimates, China's a big place and some estimates and better than others. Bottom line - the national account suck, but the growth's real.

Max writes:

My view in short is: Chinese growth is high and stats are unreliable.

It is an old topic and has been discussed for years. Chinese stats are unreliable for two main reasons:

1.) Things are developing very fast which is a difficult environment for decent stats.
2.) Officials have an incentive to cheat, as stats are often collected by the same department which is responsible for implementing policies, the success of which are being measured in those stats. If you want to pursue a successful career in goverment, cheating the stats may as well be a normal part of it. If you are familiar with the official stats website of China you may have noticed that figures are rarely revised. How much effort was and still is it to build up decent statistical services in the old world, and I cannot imagine that China is up for this for now and the foreseeable future. As there is no separation between party and administration there is ample room and incentives for cheating. If I may add some anecdotal word of mouth type evidence: There were quite a few Chinese in my econ course, one of which became my wife and another who went to work in some stats related function in some local/regional government in Guangdong. He is said to have stated that in his office delivering stats is about constructing the evidence or whatever needed to bring about the actual figures which were provided from above beforehand.

Comments for this entry have been closed
Return to top