BEIJING -- Anomalies have emerged in China's economic statistics. At local government levels, there are signs of potential attempts at bloating figures, while national statistics show inconsistent numbers. Does this portend slowing growth? At the least, a closer monitoring of the economy seems in order.
The city government of Tianjin announced in January that its gross regional product -- a localized version of gross domestic product -- was overstated.
The real growth rate for the period from January to March was 1.9%, the slowest pace of 31 regions, including provinces, autonomous districts and directly-controlled municipalities. A Tianjin official complained many meetings were called by senior officials who were jittery.
Starting in 2019, the central government will take over the job of compiling gross regional product. It appears the Tianjin city government decided to disclose an honest figure for fear of strict penalties if falsification were discovered.
But of the 31 regions, five -- Jilin, Yunnan, Qinghai, and Hebei provinces and Inner Mongolia -- reported real growth rates, which remove the impact of price inflation, that exceeded nominal figures. Nominal growth rates for Jilin and Inner Mongolia were even in negative territory. This is unnatural, considering that wholesale prices rose significantly.
This reporter pictures the situation thusly: Local governments this year began to release GRP figures that closely reflect actual situations, but prior, blown-up figures have remained unchanged. They did not correct past figures because doing so would shed light on what their former leaders did. This has resulted in nominal growth rates lower than real figures. It appears local officials are trying to correct the overstated figures behind the scenes as 2019 approaches.
Of the 31 regions, the January-March real regional growth rate exceeded the national GDP growth rate of 6.8% in 18, the lowest number in the past decade. By contrast, those whose totals underperformed the national rate totaled 12, the highest number of the last decade. There was an abnormal period in which 90% of all regions produced figures above the national rate. Compared to that, the situation has normalized.
Another notable change is that seasonal trends in GRP figures are fading. In the past, the combined total of GRP for all 31 regions underperformed the national figure in January-March, slightly exceeded it in April-June and July-September, and significantly outperformed it in October-December.
The average of the combined GRP for January-March from 2010 to 2016 was 4% below GDP, while the figure for October-December was 14% above GDP. It was as if local officials were making a final push to achieve the annual GRP growth rate targets toward the end of each year. But in October-December 2017, the regional average figure exceeded the national rate by just 5%, and the figure for January-March 2018 was very similar to the national rate.
An anomaly in national government statistics is found in monthly retail sales released by the National Bureau of Statistics. The bureau announces retail sales for the month along with a nominal growth rate from the same month a year ago, but the announced growth rate figures have started to diverge from the figures actually calculated. For example, the announced growth rate in retail sales from a year ago was 10.1% in March, but the growth rate calculated from the March retail sales and the year-ago figure is 4.8%.
The two diverging figures had matched consistently from 2016 to 2017. The bureau explains that the difference was due to an adjustment in 2017 numbers based on agricultural survey results, and that high growth rates resulted as year-ago figures were revised down. But it has not disclosed the revised statistics.
The bureau's monthly corporate income from January to April this year totaled 2.1 trillion yuan ($326 billion), up 15% from a year ago, but down 6.6% if calculated using the figure released for the same months last year. While the apparent overstatement in retail sales numbers could be explained as an error, the reversal from a drop is too large a difference. The bureau explains that its survey is conducted on a different set of subjects each year, while the growth rate is calculated based on the same set. As the survey is conducted on companies earning 20 million yuan or more in revenue, small- and medium-size enterprises in its scope change from year to year.
We calculated year-on-year growth rates for the corporate income figures using year-ago figures for each month since January 2012, and compared them with the announced growth rates. The result is that the two totals largely matched, with slight differences, up to last fall, when they started to diverge. The announced year-over-year corporate income growth rate for January-December 2017 was 21%, compared to 4.5% in our calculation. A favorite theory in online communities at the time was that some local governments reported significantly lower corporate income numbers to the central government as they attempted to rectify the blown-up totals in the past. A similar phenomenon was observed last fall in the growth rates of fixed-asset investment.
The inconsistent numbers seem to indicate the statistics bureau may have secretly adjusted past figures, affected by the wave of local government attempts to revise down overstated past figures. The bureau repeatedly said the GDP is not affected by local government adjustments, but the facts call this into doubt.
China's economic statistics are sometimes said to have downward rigidity. They appear to accurately reflect reality when growth is robust, but divert from actuality during slowdowns. In a recent example, the growth rate in 2015 remained stable even though trade and production declined. It is possible that China's statistical anomalies may reflect an increasing downward pressure on growth. But it also presents an old but new challenge to those watching the Chinese economy: how to read actual economic situations based on unreliable statistics.