Excellent post up by my friend Jeremy Gordon over at China Business Services. Post is entitled, “Statistics That Don’t Add Up,” and it is on the notorious unreliability of Chinese economic statistics. Where it excels is in breaking down and recalculating the numbers and in giving us a Goldman Sachs roadmap on determining reliability.
The post is based on an Economist article, entitled, “An aberrant Abacus,” which posits that China has actually been underestimating its economic growth over the last few years. The article also puts forth the following “guide (developed by Goldman Sachs) to the reliability of official Chinese number” by ranking the reliability of Chinese government stats from 1-5, with 5 being the most reliable:
• Foreign Trade: 5
• Money Supply: 5
• Industrial production: 4
• Consumer Prices: 4
• GDP: 3
• Retail Sales: 3
• Fixed Investment: 2
• Employment: 2
• Average Earnings: 1
• Unemployment: 1
Seems like the accuracy declines the more the numbers might affect the CCP’s position with its own people. What do you think?

  • Anders k

    Well there are very credible historians who argue that the Great Leap Forward was in part caused by the politization of statistics and lack of good independent statisticians.
    I have look at my fair share of Chinese statistics over the years and there are ways to “clean up” in the numbers.
    I don’t think the GDP growth is that off, simply because the inflations has not been under that much pressure in the last 7 years. (the article or Goldman Sachs does not seem to be able to handle this argument very well) The argument being that, if the GDP growth is way off, then there should have been recorded an monster inflation over the last 7 years, despite the goverments control over important parts of the economy.
    Average Earnings and the GDP in China is difficult to make really good economical statictics about, because of the large agriculture sector and a tax system that focuses not on the individual as an economic unite but on large unites like companies etc.
    One number that seems to be missing in the equation is the whole R&D section, that to the best of my knowledge is inflated and incorrect, something that should really make people question the real innovation capacity of China.
    It is of couse the neverending story of problems with too political dependent public service unites in China, but Chinese statisticians are getting better year after year and especially in the sectors with less political interesting numbers

  • One should be careful about jumping to conclusions about *why* some Chinese statistics are unrealiable. The common assumption is that some official is intentionally cooking the numbers, and that may not be the real source of inaccurate numbers.
    For example, the big problem with unemployment statistics is to defining unemployment. Suppose you have a factory worker that is doing nothing but still getting a paycheck of 1/3 his original salary (which is quite common). Is that person unemployed? Also if that person is unemployed, how would you know that that person is unemployed?
    So lets suppose that somehow this person gets marked as “unemployed.” Well then. someone offers him a day labor job at a construction site. Of course he’s not going to tell anyone that he got the job, so he is now marked as unemployed.
    The reason that its important to figure out if the numbers are being cooked officially is that if the numbers are being cooked then one can conclude that the real situation is worse. However, if the problem is data quality, then it could be that the real situation is better. In fact the suspicion among economists is that the real GDP growth and the urban unemployment rate in the last two years has been better than the official statistics suggest.

  • Pretty unreliable informant

    The following website of the Union of International Associations – World Problems – Issues on Line:
    http://www.diversitas.org/db/x.php?dbcode=pr&subject=3140&go=e&id=11605100
    list the following two main categories of “deficiencies” in national statistics: i) weaknesses of national data sources and ii) unreliability of national statistics.
    It then also lists five “broader” issues:
    -Unreliability
    -Inadequate data
    -Statistical errors
    -Unreliable data
    -Deficiency
    And under “unreliability” it lists the following kinds:
    Inconsistency
    Inconstancy
    Irresponsibility
    Undependability
    Doubtfulness
    Dubiousness
    Inconstancy
    Irresponsibility
    Untrustworthiness
    Undependability
    Unfaithfulness
    And so is any of the above reliable or is it all dubious, untrustworthy and irresponsible? Reader (always) beware!

  • robert

    And so given the reality that national economic (and social too) data and statistics (for one or more of the reasons / words listed above) often do NOT reflect the underlying realities they are supposed to measure or to represent what implications does this have for those who depend on those statistics for their professional analytic or comparative work?
    And what simple or sophisticated statistical methods, whether these be in depth structural or “functional” looks at various data comparative snapshots, or of medium and longer term time series is national statistical data amenable to (if one has the time, patience and skills) that can help to reveal not only their internal consistency or inconsistency but also provide solid or weak clues to their overall reliability and unreliability?
    Which then in turn also should be able to shed some light on their validity or invalidity, their dubiousness and doubtfulness and their faithfulness or unfaithfulness (to whatever criteria) which then also can (but only then) begin to shed light on the possible motives, untrustworthiness, or outright irresponsible or incompetent or deliberately fradulent behavior by the multi-level and multi-location producers of the data often by multiple methodologies?!
    Which is why this whole area is more than just a little bit tricky to navigate intelligently and correctly and in keeping with “reality” and so as to properly inform “other policy or non-policy realities”

  • Paul Hart

    china does hide inflation numbers in the form of subsidies. You can’t tell what the real numbers are because they subsidize everything…which makes you wonder, “how can they afford to do that”. It is correct to say that the CCCP has a generalized fear of a social backlash “peoples uprising” IF there is inflation so they work hard to cover this up. Everything is ALL SMILES here in China….I ain’t buyin’ it AND I live in China