In my story about the Chinese economy in the latest Atlantic, I say, "You never know which statistics to believe in China, but in January a local official in Dongguan told me..."  The never-know problem is a real challenge here, and a reason to view any number concerning China with skepticism.

Part of the problem arises from what we might call a "transparency" issue. The government has committed itself to a growth rate of at least 8 per cent this year. Whatever else happens, it is safe to assume that at year's end the reported growth rate will be about 8 per cent. Part of the problem is the sheer impossibility of really knowing what is going on in so vast a country containing such geographic, economic, and social extremes. Is China's population closer to 1.3 billion -- or 1.4 billion? It's a difference of 100 million, and I don't think anyone knows for sure.

And for foreigners there's a particular problem of having your usual standards of judgment mismatched to China's scale. I have been in cities that looked middling-size. Based on the street grid and downtown area, I would have estimated the population at maybe 100,000 -- then I'm told that two million people live there. (True? I don't know.) Every reporter in China knows about the government statistics reporting 60,000 to 70,000 mass disturbances throughout the country each year. Could that possibly be true? Two hundred a day? It doesn't seem plausible, but I see the figure quoted all the time.

Very late in the process of writing my latest article, I saw a release from the government-controlled Xinhua news agency, saying that coal mining fatalities had declined to a total of over 90,000 in 2008. Could that possibly be true? Two hundred and fifty people per day? So I double-checked with Xinhua, and so did our fact-checker, and that was the number the government was officially putting out. As a result, one passage in my story said:

So if China's rise is not undone by the risks that have been evident for years--pollution, water shortage, corruption, the widening rich-poor social gap, safety standards so primitive that on average more than 250 people die each day in coal-mine accidents--might China prove vulnerable to Soviet-style discontent born of a slowing economy?...
My guess is No. [And on to the main argument of the article.]

Twelve days later, Xinhua put out this correction.


In the corrected version, ninety thousand people had died in accidents of all sorts in China last year, not just in coal mines. The coal mine fatality rate was more like nine per day, not 250. I was out of China when this correction was posted, and I didn't see it until just now. (You don't routinely go back to sources you've already checked, to see if they've happened to change their figures.) If I'd seen it immediately we could have made a change just before our issue went to the printer, but I probably wouldn't have seen it even if I were sitting in Beijing.

I regret the error, though I am glad for the differential 240 coal miners per day, and wanted to take the initiative in putting the revised number on the record. The larger points about workplace safety -- and the resilience of the Chinese economy, and the shakiness of statistics -- remain.