作者：支點資產管理公司董事長 加文 - 戴維斯 為英國《金融時報》撰稿
中國國內生產總值(GDP)增長成為近日令人關注的新聞，因為官方數字顯示，中國在2010年已超越日本成為世界第二大經濟體。但是，如果我們換一種方式來計算，這其實是一條老掉牙的“舊聞”。按購買力平價(PPP)衡量，中國不僅早在2001年就超越了日本，而且現在還相當接近于超越美國（的確，如果中國不是已經超越了美國的話）成為世界最大的經濟體。
GDP統計數字衡量的是，某經濟體在某一給定時間期限內創造的、以本幣計的增加值或收入總額。不過，若想比較兩個不同經濟體（如中國和日本）的GDP，則要費些周折，因為我們需要按某一匯率把以人民幣計的GDP換算成以日元計的，或者反過來。這并不像看上去那么容易。
最簡單的方法就是，按相關時間期限內的市場匯率進行換算。本周中日兩國的GDP比較正是建立在這個基礎上，其結果顯示：中國的GDP水平在2010年首次超越了日本。
第一幅圖展示的是，世界最大的三個經濟體在全球GDP中所占的份額（各國GDP均按市場匯率換算成美元）。根據此圖，美國在過去30年里顯然始終遙遙領先地占據著最大經濟體的寶座。雖然日本在上世紀80年代露出幾絲趕上美國的跡象，但它從未逼近美國，而且兩國間的差距還在上世紀90年代中期之后拉大了。與此同時，中國在過去大約15年里一直在穩步攀升，最終在去年超越了日本。但從此圖看，中國在全球GDP中所占的份額仍不到美國的一半。
上述為人熟知的計算存在一個缺點：人們已經洞悉，市場匯率會低估發展中經濟體服務業所創造的那部分GDP。這是因為，個人服務（如理發）、醫療服務、零售和其它許多國內經濟活動，并不屬于國際貿易品。所以，與國際貿易的制成品不同，不存在貿易的力量來推動不同經濟體中的服務價格趨同。但在北京理發與在東京或紐約理發本質上并無不同（嗯，至少是差別不大），只是若都換算成美元，在北京理發要便宜得多。這個問題可能導致中國的GDP相對發達經濟體被大幅低估。
我們也可以換一種方法來比較不同經濟體的GDP，不使用市場匯率，而是使用（假想的）購買力平價匯率。原則上，我們可以使各類商品和服務在不同經濟體中價格趨同（前提是它們擁有相同的品質），藉此計算得出購買力平價匯率。這樣，在美國看醫生的價格，就與在中國看醫生（假設其與美國醫生一樣訓練有素）的價格相同。或者說，至少在理論上如此。
國際貨幣基金組織(IMF)定期發布基于詳盡國際價格調查的GDP比較，這樣就能夠估算出購買力平價。結果如下圖所示：
根據此圖，中國GDP超越日本已有整整十年時間，而且目前正迅速逼近美國的水平。使用IMF的計算結果并根據中美兩國近年的增長率外推，我們可以得出，中國的GDP將在大約5年后超越美國。
這兩種方法哪種更靠譜？答案可能取決于你將這些數字用于何種目的。前一種方法可以更好地衡量各經濟體公民的國際購買力，而且它是基于實際市場價格得出的。但是，由于大多數公民主要在國內消費，第二組數據或許可以更好地衡量人們日常生活的富裕程度。畢竟，北京市民無需因自己必須花高價在紐約理發而感到錢袋吃緊，因為大部分北京人本來就無需在美國理發。
毋庸置疑的是，無論采用哪種衡量GDP的方法，普通美國人都仍比普通中國人富裕幾倍。IMF 所使用的官方數據顯示，2010年美國人均GDP美元為4.7131萬美元。按市場匯率換算，中國人均GDP為4282美元；按購買力平價匯率換算則為 7517美元。后一個數字仍只有美國的六分之一左右，但由于中國人口比美國多3.5倍左右，若按購買力平價匯率換算，兩國GDP總量的差距沒有多大，而且這一差距將在幾年后消失。
最后還有一點。彼得森國際經濟研究所(Petersen Institute for International Economics)的阿文德 - 薩勃拉曼尼亞(Arvind Subramanian)認為，IMF的中國人均GDP數字存在嚴重低估，即使按購買力平價衡量也是如此，因為IMF調查采用的價格數據幾乎全部取自中國的城市，而不包括農村地區。在農村，許多商品和服務的價格要比城市低得多。這意味著，按真實購買力平價衡量的農村人均GDP，可能比IMF數據所體現的要高。
薩勃拉曼尼亞表示，若把這一因素以及其它（也許比較靠不住的）因素考慮在內，2010年中國人均GDP可能已高達1.1047萬美元，而不是IMF公布的7517美元。這一“加強版”人均GDP估計數字意味著，2010年中國GDP總量不僅已大大高于日本，而且還略微超越了美國。
薩勃拉曼尼亞的計算結果并不是官方的觀點。這些結果是基于有爭議的估算數字得出的，而那些估算數字無疑將成為其他經濟學家熱議的焦點。盡管如此，這些結果仍耐人尋味。
+++++
China is bigger than you may think
China’s GDP growth made news this week because, on the official figures, China overtook Japan to become the second largest economy in the world in 2010. But actually, on a different way of calculating the data, this was very old news. Using purchasing power parity, China not only overtook Japan way back in 2001, but it is also quite close to overtaking the US as the biggest economy in the world – if, indeed, it has not done so already.
GDP statistics measure the amount of value added or income in the economy, measured in domestic currencies, over a given period of time. But it is more difficult to compare the GDP in one economy (China) with that in another economy (Japan), because we need to use an exchange rate which translates yuan into yen or vice versa. This is not as straightforward as it may seem.
The simplest method is just to take the market exchange rate over the period in question. This is the basis used in the comparisons which this week showed the level of GDP in China exceeding that in Japan for the first time in 2010.
The first graph shows the share of the three largest countries in global GDP, all measured in US dollars at market exchange rates. On this basis, the US has been by far the largest economy throughout the last three decades. Although Japan showed some signs of catching the US in the 1980s, it never got very close, and the gap widened again after the mid 1990s. Meanwhile, China has been on a steady climb for about 15 years, and it finally overtook Japan last year. But its share of world GDP on this basis is still less than half that of the US.
The shortcoming of these familiar calculations is that market exchange rates are known systematically to undervalue the quantum of GDP produced in the service sectors of developing economies. This is because there is no international trade in items like personal services (eg haircuts), medical services, retailing and many other internal activities. Therefore, unlike in the case of internationally traded manufactured goods, there is no tendency for trade to equalise the prices of these services in different economies. Yet a haircut in Beijing is essentially the same (well, similar anyway) to a haircut in Tokyo or New York. It is just a whole lot cheaper, if we choose to measure each of them in dollars. This problem probably results in a large undervaluation of GDP in China compared to more developed economies.
The alternative method of comparing GDP between economies is to abandon the use of market exchange rates, and to use (hypothetical) PPP exchange rates instead. In principle, these exchange rates can be calculated so that the value of all goods and services are equalised between economies, provided that they are of the same quality. A visit to the doctor in the US is valued the same as a visit to an equivalently trained doctor in China. Or that is the theory, anyway.
The IMF publishes regular comparisons of GDP based on painstaking surveys of international prices so that PPP can be estimated. This is what they look like:On this basis, China’s GDP has been bigger than Japan’s for exactly a decade, and it is closing in fast on the level of GDP in the US. Using the IMF’s calculations, and extrapolating at recent growth rates for the two economies, China will overtake the US in about 5 years time.
Which of these two methods is more valid? That probably depends on what the figures are being used for. The former is a better guage of the international purchasing power of the citizens of each of the economies, and it is also based on actual market prices. But because most citizens spend most of the money domestically, the second set of data is probably a better guage of how well off people are in their everyday lives. After all, citizens of Beijing do not need to feel worse off just because they would have to pay a much higher price for a haircut in New York, since most of them never have to pay for a haircut in the US in the first place.
There is no dispute that, on any way of measuring GDP, the average American is still several times better off than the average Chinese citizen. On the official data used by the IMF, GDP per capita in the US in 2010 was $47,131. In China, GDP per capita was $4,282 at market exchange rates, and $7,517 at PPP rates. The latter figure is still only about one sixth of that in the US, but since the population of China is about four and a half times larger than that of the US, total GDP on this measure is not so very different between the two countries, and the gap will be closed within a few years.
One last point. Arvind Subramanian of the Petersen Institute for International Economics argues here that the IMF figures for GDP per capita in China are significantly understated, even at PPP, because they rely on price data taken almost exclusively from China’s cities, instead of its rural areas. Many prices are much lower in the rural areas than they are in the cities, which means that GDP per head in rural areas, measured at the true PPP rates, is probably higher than shown in the IMF data.
Adjusting for this and other (perhaps more dubious) factors, Subramanian says that GDP per capita in China in 2010 may have been as high as $11,047, rather than the $7,517 published by the IMF. This enhanced estimate of GDP per capita would imply that total Chinese GDP in 2010 was not only vastly higher than that in Japan, but was also very slightly higher than that in the US itself.
Subramanian’s calculations are not the official view. They are based on controversial estimates which will no doubt be hotly disputed by other economists. But they are intriguing nonetheless.
