Sonntag, 7. Juni 2009

Preliminary GDP Estimates

China's National Bureau of Statistics is capable of performing miracles: The much-discussed 6.1 % year-on-year growth in Q1 2009 was published on April 16, barely two weeks after the end of the quarter. No other major country even comes close to matching the speed of China's statisticians.

Ever wondered how they can aggregate all the required data from the various Chinese provinces in the space of two weeks? Surely they must use lots of statistical shortcuts and projections, many of which may be reasonably adequate in a "stable" situation, but hopelessly inappropriate in times of economic upheaval.

Why am I writing this?

Well, the BBC reports that British Q1 GDP statistics will be revised downwards because the initial estimate of a 2.4 % drop in construction activity turned out to be wrong, and construction activity dropped 9 % instead.

Just a minor estimation discrepancy, huh? I wonder how they came up with the initial 2.4 % estimate? Can preliminary GDP data have any meaning whatsoever if a major component such as construction activity can be utterly and totally misestimated?

And surely nobody thinks that China has an easier job of aggregating nationwide data than comparatively small and compact Britain?


  1. Why are you blaming the Chinese for bad British GDP statistics?

  2. I'm not "blaming the Chinese" for anything.

    But if the British manage to get something as straightforward as construction activity totally wrong, how can China possibly come up with a meaningful GDP number within two weeks after the quarter has ended? Even if they are trying their best to be as accurate as possible, there's bound to be a huge margin of error.

  3. I wonder how the British could mess up the numbers by that margin. They should be able to see the activity in each industry in the corresponding value added tax revenues, don't they?

  4. I suspect that construction activity is tricky because projects take a long time from start to finish. So they need to evaluate the stock of unfinished projects at the beginning and at the end of every quarter.