 | |  | |
| | |||
| ||
中国改变GDP统计方法意义何在？ 2011年 04月 14日 18:55 国 家统计局正在让中国国内生产总值(GDP)数据进一步接近国际标准，但有可能在这个过程中引起一些争议。 Europress Photo Agency 北京一个露天市场中的商贩 此 事为什么重要？这是因为，GDP是衡量中国经济规模和增长速度最完全的指标，而中国连续多年达到10%甚至更多的GDP增长率，又是最能反映其迅猛崛起的 符号。但在过去，中国统计每季度GDP增长率的方式跟美国或其他主要经济体的方法是不可比的。这是因为国家统计局是将某季度GDP同上年同期GDP做比 较，用的是"同比"方法，而美国经济分析局(Bureau of Economic Analysis)等机构则是将某个季度的GDP与前面一个季度比较，并按季节性因素调整，然后乘以四形成一个"年化"数字。 听起来或许有点复杂，但两种方法的区别，对于如何理解世界第二大经济体的状态有着十分重要的意义。 过去在采用同比数字的情况下，比如统计局说中国经济在2010年第四季度增长9.8%，意思就是2010年第四季度总产值较2009年最后一个季度高出9.8%。 按照美国采用的、也就是中国从周五开始报告的季调环比数字，当经济分析局说2010年第四季度美国经济实现3.1%的年增长时，其意思就是，从2010年第三季度到2010年第四季度，美国经济增长了0.775%。把这个数字乘以四就得到了3.1%的年化增长率。 采用新方法还有一个麻烦之处。为了准确反映经济增长速度，就一定要有比较。按同比计算，这事很简单，因为今年一季度的工作日数量、天气和其他可能影响产值的变量跟去年一季度往往都是一样的。 但如果是采用季调环比统计，做比较就困难得多。 在 美国，第四季度有一个感恩节。劳动者要啃火鸡，啃火鸡的时间相对第三季度"啃"掉了工作时间。中国也有全国假日，特别是1月份或2月份为时一个星期的春 节，这使季度之间更难比较。不仅假日本身可能造成数据扭曲，假日前后随着工厂增加或减少生产而出现的有别于上季度的工作模式，也有可能产生显著的影响。 为 纠正这些季节性偏差，统计人员对历史数据进行研究，以了解什么是季节效应（工作日数量变化等）造成的季度与季度之间的差别，什么是真实的增长速度变化造成 的差别。通过这些观察，他们就可以对当前季度原始数据进行季节性调整，得到真正的增长速度。（事实上因为这些计算很复杂，统计界已经写成了计算机程序来帮 他们计算。） 方便面经济  举 个形象化的例子。为简便起见，想象有那么一个经济体只生产碗装方便面，2007年第三季度到2009年第二季度的生产速度如图所示。现在让我们看看用两种 不同模型计算的2008年第四季度增长率。用同比方法计算，2008年第四季度中国经济增长率为6.8%，略显缓慢，但也不是特别值得担心。但按照环比、 季调、年化增长速度计算，中国经济增长率放慢到了0%。 按两种方法计算的经济增长高峰和底谷也有可能出现在不同时期。在方便面经济体中，环比季调年化增长率显示它的底谷处在增长完全停滞的2008年第四季度，其反弹则始于2009年一季度。而同比数字则显示底谷出现在2009年一季度。 现实经济  不同方法计算出的增长率不同，对于市场和决策者来说有着深远的意义。对于市场来说，同比数据显示的稳步快速增长，对于它形成有关中国投资题材的观点起到了根本性的作用。而季环比数据大起大落，说明中国经济是在过热与过冷之间剧烈波动，这是否还能带来同样的信心？ 对于政策制定者来说，注意力应该集中在季环比数据显示的经济当前动态。中国政府的经济刺激一直延续到2010年，理由是复苏仍不稳定。而季环比数据显示中国经济从2009年起大幅反弹，让人怀疑如此长时间地向经济体注入刺激性贷款是否明智。 谎话与统计 季环比数据的发表，将是中国经济数据朝前迈出的一大步。但它同样也有可能再次引起有关统计数据是否可靠的争议。 季调数据的计算涉及更大的主观判断，而不同假设产生的结果有可能是天壤之别。《华尔街日报》对投行经济学家做过一次调查，了解他们对2010年一季度中国经济状态的看法。结果12位香港顶尖经济学家当中，一半的人说中国经济增长了，另一半说中国经济萎缩了。上图显示的人行季环比增长数据取自2009年第二季度宏观经济报告，而在后来的报告中，这些数字也随着央行改变观点而得到了修正。 由 于人们不太相信中国官方数据，到时候肯定会有人质疑统计局是不是在滥用主观判断，将数据朝着某种和谐的方向扭曲。假设2008年统计局已经在发布环比季调 年化数据，那么他们是否真的愿意或者说能够披露，中国经济出现了不利于社会稳定的3.7%增长率？说风凉话的人认为答案是否定的，但国家统计局周五（4月 15日）推出新数据的时候，它将有可能证明这些人是错的。 Tom Orlik （作者新书《理解中国经济指标》(Understanding China's Economic Indicators)将在8月份出版。） （本文版权归道琼斯公司所有，未经许可不得翻译或转载。）
|
Q-on-Q Seasonally Adjusted Annualized GDP—It's Important!
2011年 04月 14日 18:55The National Bureau of Statistics is bringing China's gross domestic product data closer to international standards, but will likely ignite some controversy along the way.
On Friday, the bureau is scheduled to issue economic data for the first quarter, and for the first time it will include a measure known as the quarter-on-quarter seasonally adjusted annualized rate of growth of gross domestic product.
Why does that matter? Well, GDP is the most complete measure of the size and growth rate of the Chinese economy, and China's GDP growth rateâ 'which has consistently hit 10% or more a yearâ 'has been the most potent symbol of its rapid rise. But till now, the growth rate of China's GDP each quarter has not been measured in a way that is comparable with that of the U.S. or other major economies. That's because the statistics bureau has compared the size of GDP in a given quarter to its size in the same quarter a year earlierâ 'a 'year-on-year' measurementâ 'whereas the U.S. Bureau of Economic Analysis, for example, compares GDP in a given quarter to the one immediately preceding it, adjusts it to take into account seasonal factors, and multiplies by 4 to come up with an 'annualized' figure.
That might sound more than a little technical. But the differences between the two approaches have serious implications for understanding the state of the world's second biggest economy.
With the year-on-year numbers that China has been using, if the statistics bureau says, for example, that China's economy expanded 9.8% in the fourth quarter of 2010, what they mean is that total output in the fourth quarter of 2010 was 9.8% higher than total output in the final quarter of 2009.
With the adjusted quarter-on-quarter number the U.S. and others useâ 'and which China will start reporting from Friday--when the Bureau of Economic Analysis says the U.S. economy expanded at an annual rate of 3.1% in the fourth quarter of 2010, what they mean is that from the third quarter of 2010 to the fourth quarter of 2010, the U.S. economy grew 0.775%. Multiplying that by four to get the annual rate gives a figure of 3.1%.
There's an extra wrinkle, though. To get an accurate picture of the growth rate, it's important to compare like with like. In the year-on-year calculation, that's easy enough. The first quarter of this year typically has the same number of working days, weather, and other variables that might affect output as the first quarter of last year.
In the adjusted quarter-on-quarter calculation, however, comparing like with like is considerably more difficult.
In the U.S., Thanksgiving falls in the fourth quarter of the year. Workers eat into turkey and turkey time eats into working days relative to the third quarter. In China, there are also national holidaysâ 'notably a week-long national holiday for the Lunar New Year that falls in January or Februaryâ 'that make it more difficult to compare the quarters. It's not just the holidays themselves that can distort the data: different working patterns in the days before and after as factories gear up and wind down can also have a marked impact.
To correct for these seasonal biases, statisticians look at the pattern of historical data to try and work out the variation between quarters that is caused by seasonal effects (changes in working days and so on) and the variation that is caused by genuine changes in the rate of growth. Using these observations, they can seasonally adjust the raw data for the current quarter to get to the true rate of growth. (Actually, because these calculations are complicated, the statisticians have written computer programs that do the math for them.)
The Instant Noodle Economy
Sound tedious? The thing is, the two approaches to calculation actually produce results that are radically different. The year-on-year calculation is a glance into the distant past of the economy, a comparison of today with a year ago. The adjusted quarter-on-quarter calculation is a reflection of the current momentum of the economy, a comparison of today with yesterday.
A stylized example illustrates the point. For the sake of simplicity imagine an economy that only produces bowls of instant noodles, and that from the third quarter of 2007 to the second quarter of 2009 the output rate was as shown in the table below. Now, let's look at the growth rates for 4Q08 calculated according to the two different methods. According to the year-on-year (or YoY) calculation, in 4Q08 the Chinese economy grew at a slightly sluggish but not particularly alarming 6.8%. But according to the quarter-on-quarter, seasonally adjusted annualized rate (or, ahem, QoQ SAAR). growth in the Chinese economy slowed to 0%.
The quarter-on-quarter calculation can also show the peaks and troughs in growth appearing at different points from the year-on-year data. In the noodle economy, the QoQ SAAR data shows the trough comes in 4Q08 with the economy grinding to a complete halt, and the rebound starts in 1Q09. The YoY data, in contrast, shows the trough in 1Q09.
The Real Economy
As it happens, the People's Bank of China has been calculating the quarter-on-quarter figure for China, which is included in the central bank's quarterly report on the macro-economy. A comparison of its data with the year-on-year data produced by the statistics bureau shows that our noodle-bowl economy example is not far from reality. As shown in the chart below, the QoQ SAAR data calculated by the PBOC shows growth hitting a trough of 3.7% in 4Q08, before rebounding strongly to 6.4% in 1Q09 and 14.9% in 2Q09. The YoY data from the NBS shows the economy hitting a trough of 6.5% growth in 1Q09, and recovering only slowly in the following quarters.
The different growth rates have far reaching implications for the markets and for policy makers. For the markets, the steady rapid growth shown by the year-on-year data has been fundamental to shaping perceptions of the China investment story. Would the jagged ups and downs traced out by the quarter-on-quarter data--suggesting the Chinese economy was veering wildly between overheating and overcooling--have inspired the same level of confidence?
For policy makers, the focus of attention should be the current momentum of the economy, shown by the quarter-on-quarter data. The Chinese government kept the economic stimulus in place way into 2010, claiming that the recovery was not yet stable. The quarter-on-quarter data, which shows the economy rebounding sharply from 2Q09, calls into question the wisdom of keeping the economic punch bowl spiked with stimulus loans for quite so long.
Lies, Damned Lies and Statistics
The publication of the quarter-on-quarter figure will be a definite step forward for China's economic data. But it could also reignite controversy about the reliability of the statistics.
The calculation of seasonally adjusted data includes a larger role for judgment, and different assumptions can generate wildly different results. In one poll of investment bank economists conducted by The Wall Street Journal on the state of the economy in 1Q10, 12 of Hong Kong's finest were equally divided on whether the Chinese economy had expanded or contracted. Â The PBOC's data on the quarter on quarter growth rate shown in the chart above, which is taken from the report on the macro-economy in the second quarter of 2009, was also revised in subsequent reports as the Central Bank changed its view.
With faith in China's official data low, there will be inevitably be questions about whether the NBS is abusing that scope for judgment to tweak the data in a harmonious direction. If, in 2008, the NBS was already publishing QoQ SAAR data, would they really have been willing or able to disclose that the economy was growing a socially destabilizing annual rate of 3.7%? Cynics assume the answer is no, but when the NBS roles out its new data set on Friday 15th, it will have a chance to prove them wrong.
Tom Orlik
评论内容
2011.04.17 09:30:58
意义在于模糊掉经济增长下降局面 - zy51888
2011.04.15 16:34:47 预期GDP将会大幅下滑，所以用环比，能更灵敏的感觉到，好为早点放松货币做铺垫。等到经济起来了，再用同比。
- 呵呵
2011.04.15 10:30:29 经济学的统计数据，都是通过脑残计算过的！根本就没信过
- sfss2011
2011.04.14 23:41:24 礼义廉耻全都抛到长江黄河，脸皮都一起不要了，都来撒谎吧，让谎言来得更猛烈些
- 020xxx
2011.04.14 22:22:41
感觉不如以前的方法合理。 - 奇怪
2011.04.14 21:46:19
从理论上说新的统计方法更有利于把握经济发展的节奏，但真实的效果就不得而知了。 - hudongwei
2011.04.14 21:21:37
GDP的统计，不能反映中国的实情，我们应该根据结构的合理性，做好实情工作。该舍弃的舍弃，该强调的强调。 - heagleth
2011.04.14 20:57:09
表格里的同比增长率四舍五入有问题，环比数字则明显错误，这种错误有点低级吧 - JD
2011.04.14 20:02:41
图表数据有点不明白，3。6%不是2季度的环比增长率吗？14。5%又是从何而来？ - leemr
2011.04.14 19:24:24
换方法玩文字游戏，以前的玩腻了。 - 弃旧迎新
No comments:
Post a Comment