Scott Sumner  

What makes you think Li knows the truth?

Fred Smith reviews Charles Mur... Values are Subjective...

There is a great deal of skepticism about Chinese economic data. I believe that some of the skepticism is justified, but much of it is greatly overdone. This FT article discusses the critique that I find most plausible:

However, doubts have long lingered about the veracity of Chinese economic data, with many analysts believing them to be understated during times of rapid growth and overstated during the current slowdown.
Even so, I don't see any reason to believe the average numbers are wrong (which would become blindingly obvious over long periods of time) or that the cyclical bias is particularly large.

If the Chinese persistently overestimated growth, that sort of deception would be hard to disguise. After a while China would be obviously poorer than claimed. But there's an even better argument. Countries that do cheat on the inflation numbers (such as Argentina and Venezuela), quickly become the laughingstock of the international community. The exchange rate falls sharply, clearly exposing a depreciating currency. In contrast, China's trade weighted currency has been trending upward for years, and no, the recent tiny devaluation against the dollar didn't change that fact, as it's still up sharply against most other currencies. Countries that do have high inflation (India, Brazil, etc.) see their currencies fall over time. And the other piece of evidence is that we know Chinese nominal incomes have risen very sharply in recent decades. It's pretty hard to hide the fact that 1.4 billion people are getting double-digit raises year after year, especially when foreign business is heavily involved in the Chinese labor market.

OK, but what about the cyclical argument? Maybe instead of growth slowing from 12% to 7%, it slows from 13% or 14% to 5% or 6%. Yes, that possible. But if it's occurring, I don't believe it's any sort of conspiracy, just a flaw in the statistics office. I don't even think the Chinese government itself knows whether the data is overstating growth. Some of my critics are Austrian economists. They should recall that Austrian economics emphasizes that central planners can't know everything that goes on in large complex economies. It's quite possible that they are as confused as we are.

Recently a number of commenters pointed to what they thought was a smoking gun that China was exaggerating GDP growth. Premier Li (number 2 in China, and responsible for economic policy) declared that even he didn't believe the data:

Some have instead turned to alternative estimates, such as the "Li Keqiang index" -- a composite of various indicators favoured by the Chinese prime minister, such as electricity production and railway freight volumes. Many of those gauges fell into outright contraction late last year.
I hope you see the problem here. If this is a conspiracy, it's a very strange one. Li is citing the same data sources as many of the China skeptics cite. But that suggests he doesn't know any more about China's actual GDP than we do! He's just another Jim Chanos, reading the tealeaves. So no, I'm not impressed by the Chinese Premier "admitting" the flawed data. Not if he's just as in the dark as the western Chinese skeptics that I don't trust.

Inevitably some will want to go down the rabbit hole, to even deeper conspiracy theories. Maybe Li is just pretending to look at railroad and electricity data, and he actually is privy to the cabal of Chinese statisticians who systematically distort the data. He cites the electricity data because he doesn't want to admit the Chinese government lies. Fine, let's go from "grassy knoll" to "Bush ordered 9/11" level conspiracy theories. But there's still a problem here---what's the motivation? The original argument was that the Chinese inflated the GDP data to make China look better. They didn't want to admit to the public that their policies were failing. I get that. But then why in the world would the Chinese Premier adopt the debating points of some of China's strongest critics? Why issue rosy data and then tell the public the truth is far worse?

In fact, I think the world is much messier than we assume. Conspiracies are really hard to pull off. If you bring in enough people to make it work, you can't keep it secret. If you only involve a small number, you can't make it work. In 1991 some Russian military officers tried a coup, but they didn't have enough critical mass to pull it off, presumably because they wanted to keep the plot secret. Other generals refused to go along. Why should they? They weren't invited, which tells you they wouldn't be high up in the next government.

The other lesson here is that we should not assume government policymakers know more than we do. Do not defer to the expertise of the Fed. I'm willing to grant that the Fed has an impressive data operation, and might have slightly better real time estimates of GDP than I do. But they are let down by their model. The market takes everything into account when forecast the future path of interest rates (that's rational expectations), whereas the Fed only takes into account their interest rate forecasting model. As a result the market currently forecasts a flatter path of interest rates than the Fed, and I suspect the market will turn out to be right.

PS. The FT article ends on a sensible note:

Figures around 5 per cent [RGDP growth] contrast with the consensus estimate from investment bank economists, which is currently 6.9 per cent for 2015.

David Meier, economist at Julius Baer, has an official forecast for this year of 6.7 per cent, but notes "suspicion that the figures could be 1-2 per cent overdone".

"At the end of the day, we have to comment, analyse and publish forecasts based on the official [gross domestic product] data series," he said.

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (9 to date)
baconbacon writes:
They should recall that Austrian economics emphasizes that central planners can't know everything that goes on in large complex economies. It's quite possible that they are as confused as we are.

Lying does not require that you know the truth. "Bobby was with me last Saturday night, Sheriff" is a lie if you don't know where he was Saturday night. It is possible that Bobby was hiding in your closet Saturday night, and so your statement is true by chance, but even so the statement is a lie.

While your claims that large scale outright fraud should soon be caught are (likely) true they are not the strongest position, but the weakest. The strongest position is that the Chinese government has set up their GDP calculation in such a way that it will smooth large shocks so as to avoid ever appearing to be in recession. This would be lying/manipulation (with some implications for reasons why they would do it) but could go undetected for years.

Christopher Balding is among those raising the question of how can the Chinese get the final numbers out of the mash of ingredients going into the pot.

To examine this closer, I downloaded data from the report covering the 50 and 100 largest retail enterprises in China with sales broken out by category. Looking at the 50 largest retail enterprises in a year over year basis, except for jewelry, all other categories are negative. Looking at the 100 largest, jewelry is still the largest gainer at 8.7% with food registering a 4.9% gain. Total retail sales among the largest 100 registered a total gain of 1.5% year over year. Most interestingly about the top 100 year over year retail sales, there is not one category that reaches the 10.8% claimed by the NBSC. Balding's World

[Broken HTML fixed--Econlib Ed.]

Foobarista writes:

A lot of the problem is because the Chinese government itself often doesn't know, and local officials are compensated and get bonuses based on local statistical factors that they control. In the West, we often think of China as a well-organized, centralized entity with a strong central government that controls everything, but the reality is that things in China are far more local than it would seem. The US is actually quite a bit more "centralized" than China is, particularly as regarding government economic policy.

The mountains are high, and the emperor is far away...,_Huangdi_yuan

Scott Sumner writes:

bacon, You haven't addressed the main point of my post. If the Chinese government was intentionally distorting the GDP numbers, then they'd obviously know the truth. But if they know the truth, then their public statements make no sense.

You said:

"The strongest position is that the Chinese government has set up their GDP calculation in such a way that it will smooth large shocks so as to avoid ever appearing to be in recession. This would be lying/manipulation (with some implications for reasons why they would do it) but could go undetected for years."

If they did this intentionally, then they'd know the truth.

As far as the Balding estimates, I addressed that in a MoneyIllusion post a couple days ago.

Foobarista, That's true, but raises another problem. The national government estimates of GDP growth are generally less than the sum of the local estimates. Why would that be? Presumably because the national government believes the local figures are exaggerated, and therefore adjusts them to come up with a more accurate estimate. But why would they do that if trying to exaggerate the growth rate?

Chris H writes:

I think there's a pretty good reason why China's leaders would want to tell something at least pretty close to the truth on their data. That is the Chinese leadership's ambition to build China back into a superpower like it was prior to the 19th century. Lying to yourself about your GDP numbers is a good way to make sure that doesn't happen because you then lose the ability to know when to course-correct. But gathering accurate economic data for 1.4 billion people is a large task and as Scott points out, the larger the conspiracy the less stable it is. So China needs a large amount of conspirators to gather the data the Chinese government needs for it's long term ambitions, but such a large group makes maintaining the conspiracy harder.

Beyond that, I think it's worth noting that China's leadership seems very willing to plan for the long term. It's investments in Africa not only pay off now but establish a presence in a rising region that is likely to be more important as it stabilizes some. The AIIB is unlikely to break alliances between the US and China's neighbors now, but is useful to starting to create closer relationships over the long haul. Lying about GDP numbers right now might make China look a bit better in the short run, but in the long run it undermines China's credibility and increases distrust in the Chinese leadership. I don't think China's leaders really are ready to risk their long term objectives like that at this point. Their NGDP is still significantly below the US and it'll probably be a couple of decades before their military can get global force projection capabilities that can compete with the US. So I don't think they want to lie too much yet. I don't think China's population is that ready to explode yet, falling expectations are always dangerous for an autocracy but it's a lot harder to hide that to the people living in China than a few headlines about GDP growth being better than it is.

I don't think China is lying about it's numbers, at least not by a lot. They're slowing, but not in a catastrophic way yet.

Nathan W writes:

I thought Chinese GDP figures merely aggregated local data, so it's altogether possible that the centre is not very sure of what's going on. The argument goes that local officials fudge numbers towards what Beijing dictates as its expectations, and the centre is not in a very good position to evaluate the extent of this fudging.

As an extreme historical example, it is often argued that Mao truly didn't know the massive scale of famine deaths in the 1950s (during the co-incidence of collectivization and a drought) because local officials didn't dare to tell the centre that they were not meeting targets.

In 2015, it is inconceivable that there could be massive famine without Beijing knowing. But, it is altogether conceivable that each and every province is massaging data upwards and that central statistics bureaus don't have effective ways to capture actual GDP. Hence, a central officer referring to electricity data and train freight, two areas where he can be absolutely certain of the data.

baconbacon writes:

Scott- three scenarios

1. The Chinese know that GDP growth is 4%. They report it as 7%- this is an outright lie.

2. The Chinese don't really know what GDP growth is, but whatever information they get will be massaged to pop out the number they want- this is also an outright lie despite not knowing the real GDP.

3. Halfway between the two- they have a general idea about the magnitude and direction of GDP, they use this information to select GDP numbers that are politically convienient. The battle between political convienince and the "truth" is at its most intense when large changes occur. This is also "lying" though covered up with a hundred "truths".

Scott Sumner writes:

Chris, Good points.

Nathan, No, the central government has a different way of calculating GDP, which does not depend on the local figures. But you are right that the local figures are flawed, and tend to exaggerate GDP growth.

bacon, You seem to be trying to have it both ways, but I still don't quite see the point. Even in case three, the central government would know the best estimate of the messy figures, and would presumably use that, not railroad and electricity figures.

Dave H writes:

I find it highly implausible that Chinese stats are intentionally manipulated. How exactly does one get a good survey frame for 1.4 billion people and a truckload of businesses? Slight inaccuracies when multiplied by 1.4 billion results in pretty large errors in growth.

Alternatively, if their stats are more reliant on admin data, isn’t it more likely that the problems are in aligning these data with relevant international standards like the SNA or the other many pitfalls of admin data? Ever heard of a series break?

And so far we’re only talking about production and households in current prices, but the doubts are about RGDP which needs price deflators. Has anyone really perfected the art of deflating services? Hell, even with things you can kick like semi-conductors there is massive deflation that will cause real problems in interpretation.

And, let's face it: if one is an economic commentator, one is highly likely to have learnt debating techniques in academia and bashing stats is like shooting fish in a barrel and gets you eyeballs.

I’m not saying Chinese stats are right. Stats are rarely right, but like models they can be useful. And what is clear is that the Chinese people have enjoyed a massive run of GDP growth that’s raised living standards and resulted in millions escaping poverty. However if the question posed is has Chinese RGDP been 5, 6 or 7 per cent, ask a different question because no one will really ever know.

Greg writes:

Two points:

1. Li Keqiang made the comments about GDP numbers back in 2007, when he was the governor of Liaoning province, whose economy is dominated by heavy industry (China's rust-belt, so to speak). He became primier only in 2013.

2. China's national GDP number is NOT aggregated local or provincial numbers. In fact, the national bureau of statistics uses their own vertical system to collect data, which is separate from the provincial data collection.

Comments for this entry have been closed
Return to top