February 4th 2008
FROM THE ECONOMIST INTELLIGENCE UNIT
New estimates of GDP at purchasing power parity paint a sharply different picture of the global economy. Emerging Asia’s GDP is sizeably lower than previously thought, with India and China both suffering 40% cuts; as a result, emerging markets account for 41% of world GDP rather than 47% as previously thought. Global economic growth in the last five years is probably half a percentage point lower too. And oil-producers seem to have avoided the so-called resource curse that is supposed to condemn them to systematic underperformance.
In order to make inter-country comparisons, GDP has to be expressed in a common currency. The use of market exchange rates as conversion factors can give very misleading results, mainly because they do not adequately reflect relative price differences. Developing countries typically have relatively low prices for non-traded goods and services, so conversions based on market exchange rates systematically underestimate the value of output of a developing country relative to an advanced economy. Also, exchange rates fluctuate for reasons that have little to do with the purchasing power of a currency. For example, based on market exchange rate weights, world growth during the boom of the past five years (at 3.4%) was slower than in the 1980s. Measured at market exchange rates, the emerging economies’ share of global output is now, most implausibly, lower than in 1980–even though these economies have been growing more than twice as fast as the developed countries.
Using purchasing power parities (PPPs) to convert currencies makes it possible to avoid these problems and to compare the output of economies in real terms–that is, to control for differences in price levels. The PPP is the number of units of a country’s currency required to purchase the same basket of goods and services as one unit of the reference currency, usually the US dollar.
The International Comparison Programme (ICP) is a worldwide statistical initiative that has been running since 1968. Towards the end of 2007, the ICP published its estimates for the 2005 benchmark year for 146 countries–the most extensive and thorough effort ever undertaken to measure PPPs. The previous estimates were based on extrapolated data from benchmark PPP estimates dating back to 1993 or earlier for most developing countries. For some developing countries, the estimates were based on crude short-cut procedures that estimated a country’s PPP on the basis of a cross-country relationship between PPPs and market exchange rates.
The world through a new lens
Compared with the previously available estimates, the new ICP estimates show significant changes in PPP-based GDP figures for many emerging markets. Among the most important differences are the much lower estimates for some Asian economies–above all, the reduction by 40% in both China’s and India’s GDP, but also the much lower GDP estimated for Bangladesh (by 36%) and for Indonesia and Thailand (by 20% in each case). China’s share in global output in 2005 is now estimated at 9.5%, almost five percentage points lower than the 14.3% estimated according to the previous data. India’s share has declined to 4.2% from 6.1% previously.
The share of emerging markets in global GDP has been revised downwards by a significant amount. It is now estimated that emerging markets accounted for 41% of world GDP in 2005, down from the previously estimated 47%. Much closer to their previous numbers are the figures for Russia and Brazil. The BRIC countries–Brazil, Russia, China and India–now account for some 20% of global GDP against the earlier estimate of close to 26%.
There are also very large upward and downward revisions across the board. GDP per head has been revised upwards or downwards by more than 20% for 55 out of the 100 developing countries that were covered. In the case of oil exporters, for reasons that are not altogether clear, GDP per head in the latest estimates was higher–in many cases by a very large margin–for 15 out of 16 countries. If these results are accepted, they would imply that these economies are both richer and have also probably been growing at significantly faster than officially recorded rates. This would bring into question theories about the “natural resource curse”, that is about the systematic underperformance of oil-dependent economies.
|Global output in 2005, GDP in US$ bn|
|At market exchange rates||New ICP PPP-based||Old PPP-based|
|Source: ICP, Economist Intelligence Unit, IMF.|
The revisions for PPP-based GDP of two of the world’s biggest and fastest-growing economies, China and India, are mainly responsible for the reduction of global growth estimates. Based on the new data, the IMF has revised downward its estimates for global growth by around half a percentage point each year during 2002-07. Notwithstanding these changes, emerging markets are still the main recent driver of global growth in PPP terms (China alone contributed more than 25% of global growth in 2007).
The reductions in GDP per head imply a large increase in measured poverty in China and India, based on poverty rates calculated according to the international one-dollar-a-day standard–these now look far higher than assumed in a host of studies by the World Bank and others.
Despite the revisions, China still ranks as the world’s second largest economy. However, another implication of the new data is that it will take China much longer to overtake the US economy in PPP terms than had been previously thought. Catch-up, which was set to occur by 2010, has now been deferred for at least another decade, and possibly longer.
A reliable picture?
On the face of it, the new ICP estimates seem to be far superior to the previous ones. In many cases the latest numbers replace old guesstimates with actual data. China participated in the ICP programme for the first time, and India for the first time since 1985. The previous estimates for China were extrapolated from a bilateral comparison of 1986 prices between China and the US. The latest ICP round has been hailed as the most ambitious effort ever to measure PPP rates across countries. New methods were developed and used to overcome shortcomings of the previous data collection and estimation processes. The data were drawn from regional surveys of prices for more than 1,000 goods and services across the countries that were covered.
However, important doubts remain about the new data. In particular, there is a strong suspicion that the data understate considerably China’s income, on several grounds. First, China only partially cooperated with the survey and arguably it has reasons to downplay its GDP–principally to make the Chinese economy appear less threatening internationally. Second, if the revised PPP estimate is correct, then it would appear that China’s currency is in “equilibrium” or is overvalued, and not undervalued as its Western trade partners insist. Third, if the revised PPP data are correct, then extrapolating backwards China was the poorest country on earth in the early 1980s. Again, this stretches credulity.
There are other long-standing problems from earlier studies that are also replicated in the latest ICP study. The relative prices and incomes of countries that emerge from intra-regional comparisons have been kept fixed for each region in the final international comparisons. This can distort heavily the comparison of countries in different regions as the PPPs derived by observing the regional “fixity principle” will differ markedly from PPPs calculated on the basis of a universal cross-regional comparison. The originators of the ICP have often been critical of this procedure, arguing that this is governed by political, not methodological concerns.
Utilising the functional relationship between a country’s price level and its level of development, another implication is that west European currencies (euro and others) are seriously overvalued and/or that west European GDP has been underestimated (a West European dummy variable in the equation linking price levels to income per head is highly significant in all specifications).
The PPP estimates for EU member and EU candidate countries have since 1996 employed a statistical procedure–effectively a sleight of hand–that makes poorer EU countries look richer than they are. In particular, unlike in earlier studies, no allowance since 1996 has been made for differences in the quality and productivity of so-called non-market services among the European countries. The PPP data make the gap between the new EU member countries from eastern Europe and the advanced West European states look smaller than it really is. And even a casual observer, for example, will doubt that average income in Greece is only 15% lower than in France or Germany.
PPP-based data are certainly superior to the use of market exchange rates in making international comparisons and the latest ICP study probably yields better overall results than previous ones. However, PPPs are also imperfect and the margins of error remain disturbingly large, so as to render questionable international comparisons of economic well-being that rely only on GDP data. This increases the onus on using supplementary methods to compare well-being across nations.