Syndicate content

The end of a long era

Branko Milanovic's picture

Is China, after a hiatus of 150 years, again the largest economy in the world? Not all sources of GDP data agree, but there is little doubt that China is either already now the largest economy, or it will, within a year, become so by overtaking that of the United States. Whichever the case may be, a long era when the American economy was the largest in the world and which began around 1860, is now reaching its end.

Data on gross domestic product (called now Gross Domestic Income) are available from three sources: the Maddison project, which is the only source for the long-run series of national GDPs, going back to 1820s; the World Bank or IMF annual data, going back to 1960; and Penn World Tables, produced periodically at the University of Pennsylvania, going back from their just-released version 8.0 to 1950 . All three sources produce GDP data in PPP (purchasing power parity) terms, which means that they adjust for differences in price levels between the countries. The easiest way to explain it is to say that PPPs try to account for each good and service using the same price for it around the world, so that a mobile phone, a kilo of rice and a haircut would each be valued the same in China as in the United States. Only thus can the real sizes of the economies, and the welfare of people, be truly comparable. These PPP data, in turn, are obtained through a massive worldwide project called the International Comparison Program, which is run every five to 10 years and collects more than 1,000 prices in all countries.

The problems, which influence our exact determination of the China-US, relative sizes of GDPs, begin after that point. The Maddison project uses PPP data derived from an earlier 1990 ICP exercise. This is in part due to the fact that Angus Maddison, the British economist who almost single-handedly created the historically comparable national income data for more than 100 countries, did his pioneering work when the 1990 ICP was the latest ICP exercise. But in addition, it was due to Madison’s  dissatisfaction with the 2005 ICP results, which significantly increased the price level of China (as well as that of India, Philippines and a number of poorer Asian countries) and thus made their “real” (PPP-expressed) GDP less. Maddison thought that the estimated price levels for China were unrealistically high, and preferred to keep on using the 1990 PPPs. After his death, the same approach was carried on by a group of 20 economists, who are now managing the Maddison project (for reasons of full disclosure, I am one of them). Maddison’s Chinese GDP numbers are, therefore, as we shall see, higher than those in the other two sources. (This is somewhat ironic because Maddison was a long advocate of the position that Chinese growth rates were overestimated. In his work on China (Maddison 1998), a debate conducted on the pages of the Review of Income and Wealth (Holtz 2006; Maddison 2009), as well as in his GDP series (Maddison, 2003), he preferred to use growth rates that were, on average, several percentage points lower than the official ones.)

Now, unlike Maddison, both the World Bank and IMF, and more recently Penn World Tables, accepted the 2005 ICP results. The World Bank, one of the key players in the 2005 ICP exercise, had no choice and accepted it fully. The result was a significant shrinking of Chinese GDP, by some 40 percent. (The same fate befell a number of other countries as well: India’s GDP was reduced by about 40 percent, Bangladesh’s by an incredible 48 percent, Vietnam’s by 31 percent, and Indonesia’s by 18 percent; see Milanovic 2012, Table 1). The reason, it is argued, was that in fast-growing poor economies, prices of non-tradables (services, housing) increased faster than previously thought based on simple extrapolations of CPI, and rising price levels made “real” GDP less. 

However, there was some general unhappiness with the 2005 PPP numbers for China. In their calculations of world poverty,  Chen and Ravallion (2010; 2010a) decided to treat China’s PPP as pertaining to the urban areas only, and use a lower PPP level (implying thereby a higher real income) for rural areas. Angus Deaton (2010) and Deaton and Heston (2010) expressed a similar unease, arguing that the Chinese price level was overestimated by about 20 percent. The explanation was that Chinese prices in 2005 (the first time China directly participated in an ICP) were collected mostly from urban and periurban areas and thus tended to be unreasonably high. This, however, does not answer the question why other Asian countries exhibited the same phenomenon (much higher PPPs than were thought prior to 2005 ICP). Because of the importance of and fascination with China these other problems were somehow swept under the rug, and little was heard of them. Penn World Tables, which just issued its new GDP numbers, took a similar skeptical stance toward China’s 2005 ICP numbers (see Feenstra et al. 2013, p. 10), revising Chinese PPP down (compared to the ICP exercise), and thus producing China’s GDP values  that are between the rather high values in Maddison, and low values in World Bank/IMF data.

This is how we have reached this confusing point where GDP of the largest, or second largest, economy in the word is not known with anything approaching certainty. But let us consider the numbers themselves. According to Maddison series, China’s 2010 (when the series ends) GDP is $PPP 10.7 trillion, and US GDP is $PPP 9.4 trillion. China overtook the US in 2009, thus ending a period that began around 1860, when US overtook….whom?  China!,  to become the number one  world  economy.

 World Bank data for 2011, however, paint a different picture. Chinese GDP is estimated at almost $PPP 10 trillion (these are 2005 PPPs) and US’s at $13.2 trillion. However, between 2011 and mid-2013, China’s GDP grew by about 10 percent, while US GDP increased by 3 percent. The gap, in favor of the US, is today narrower, at most 20 percent. But clearly if the differential in growth rates continues, as according to the latest IMF World Economic Outlook (2013) it will, it would take only a couple of years for China to overtake the United States, even according to the World Bank data.

As mentioned, Penn World Tables provide numbers that are in between the World Bank’s and Maddison’s. For the United States, Penn data give almost the same value in 2011 as the World Bank, but Chinese GDP is slightly higher (remember Penn’s different treatment of 2005 PPPs) and is estimated at $PPP 10.7 trillion. According to Penn’s data, the US advantage today is less than 10 percent, and would be reduced to zero within a year and a half.  Thus, with current trends, the sorpasso will happen around 2015, according to the World Bank, or around 2014, according to Penn.

But this is not the end of the story. Simultaneously with all these calculations, a subsequent ICP exercise was conducted. Its results will be available at the end of the year. If it fully confirms the results of the 2005 exercise, nothing will have changed in our relative calculations. But this, according to some preliminary findings, is unlikely. It is believed that the new ICP round will, as far as China is concerned, reverse to some extent the results of the 2005 ICP round. This would then imply that China’ GDP may suddenly jump by something like 20 percent. If indeed such results come out in December 2013, then even according to the Word Bank and Penn, which would be bound to use the new 2011 PPP numbers retrospectively, China is already now the largest economy in the world. The new results will not affect the Maddison series, but there, as we have seen, China is already number 1.

You can take your pick among these numbers. You can also lament the fact that we do not have more reliable global statistics. But whichever source you decide to trust, one conclusion is unmistakable. Short of a cataclysm in China, China is already the largest economy in the world--or will be so in 18 to 24 months. The century and a half of America as number one will have come to a close.


Thank Branko, nice blog and very interesting to see the differences in the datasets and the underlying assumptions of each. I always considered the Maddison numbers to be slightly too high for non-Western countries and it's interesting to see the role that urbanity plays in that.

I recently made an infographic on this topic actually, glad to see that world trends in global economies are interesting to the World Bank! The Maddison project has such a great advantage over the other dataset as it places a value on history and trends which are key to understanding the time we are in now. By only providing GDP figures from the 20th century, you get a really partial view of the world and it normalises the financial dominance of the Western world. The Maddison project shows that the last 160 years have been completely bizarre and beg the question of why China/India weren't the richest countries in the world over the past century and a half.

Take a look at the charts if you're interested:

Submitted by Branko on

Thanks a lot, David. Very nice comments and a great infographic. I just tweeted the link.

Submitted by Anonymous on

In terms of GDP, but this measurement is no longer viable in times of digital goods.

It seems that no matter what statistics are used it will always be possible to point to possible errors in the numbers which are difficult to quantify. However, the use of three sets of data certainly increases the confidence of your conclusions and as you say, it is beyond question that the data points to the conclusion that China is either already the largest economy in the world or will become so in the next year or so. The real question I believe is what the geo-policital and economic consequences of this change to the world are. It seems that historically China has not been engaged in imperial expansion in the way that many other super powers have, possibly because China is already so large that it cannot be effectively managed if it were to grow any larger.

Submitted by Sasha on

There are a lot of issues with PPP. The reason why it is used is because it is a way to get away from exchange rates. The problems with PPP is just as you wrote: it depends on how you measure the prices of items. And that varies greatly.

Also, even if a service costs 50% less in a poorer country than in a richer country, who says the actual service you get is equal in value?

The World Bank has another measurement, it's the so-called 'Atlas method' which smoothes out the exchange rate over a 3-year period. By that measure, China will (possibly) surpass the U.S. in the mid-2020s and perhaps never on a per-capita basis unless we see a total energy revolution(like fusion energy) and/or an extremely rapid production of electric vehicels. Not to mention issues like food, water and other resources that are already strained.

Submitted by Sasha on

Just to add.

One of the reasons that proponents of PPP use is that exchange rates are inherently volatile. But ask yourself this: if India's economy "shrank" by a stunning 40% using the 2005 ICP standards, how is that stable? And who says that if they increase their GDP by 40% this time, that the new standard is accurate?
Or take Vietnam whose economy "shrank" by 40%.

Exchange rates are problematic, see the so-called 'Penn effect', but this is why the Atlas method is better(but still very imperfect) as it measures GNI(Gross National Income) per capita with a smoothed 3 year exchange rate.

All of which goes to show just how unstable the economic business really is. If a country's GDP can increase by 40% or 20% just by a flip of a switch - and we still are not sure if that is even accurate - then it tells you how far we still have to go.

Submitted by David on

"But whichever source you decide to trust, one conclusion is unmistakable."

Uh, not really? Only if you use PPP as a method, if you use nominal GDP or as above-mentioned GNI(the so-called 'Atlas method' that the World Bank provides), the picture changes.

If you use nominal GDP or GNI you should use it with a smoothed exchange rate of 5 years or so to avoid the volatile shifts. But increasing or decreasing your economy by 20 to 48% using PPP isn't a winnable argument for PPP as described above.

Submitted by ali on

what's the difference between world bank data and penn world table data... plz elaborate

Add new comment