Market and PPP measures of GDP
In comments on the previous post, Ian Castles AO, the former Australian Statistician, notes that the World Bank and IMF create confusion in their reports by mixed use of market-exchange-rate (MER) and purchasing-power-parity (PPP) bases for estimating output and growth. Simply, using market exchange rates to compare the value of output among countries over-estimates the size of developed economies and under-estimates the size of developing economies. This confusion affects their analyses of, among other things, trade data.
The charts (click the thumbnails), taken from the two sets of data prepared by the IMF, show the dramatic difference between MER and PPP bases for comparing output and growth. Using the PPP basis for comparison, developing countries' economies are much larger in relation to developed economies and projected to 'close the gap' sometime after 2015 (take the IMF projections with a grain of salt: they're just 'straight line' extensions of current patterns of growth). The only difference between the two charts is the identification of China. Note that using a PPP basis for comparison, China appears less dominant in the developing country group.
Ian and the former head of the OECD statistical division, David Henderson, were the authors of an influential criticism of the IPCC's projections for global growth in 2004. They showed that the IPCC emissions scenarios, which used on MER estimates of the growth 'gap' between developed and developing countries, implied growth in global output (and hence, CO2emissions) that was far too large. The UK Parliamentary Committee to whom they presented their case accepted their criticism of the IPCC scenarios. The IPCC was, predictably, less responsive to their constructive criticism.
You may be wondering why, if the MER basis is an inaccurate way to measure and compare output and growth, the World Bank (especially, but also the IMF and OECD in some publications) use it. The answer, as explained by the IMF staff, is that it is "too hard" (!) to estimate accurate PPP weights for the prices of goods and services, especially if they are not traded. This makes it seem that some of the international financial institutions would rather be wrong than over-budget.
Update: In fact, it is very difficult for some not-so-obscure reasons teased out by Brian Fisher and Hom Pant in this ABARE Conference Paper on the PPP versus MER debate. Here's the abstract: it's a short paper and worth reading.
"Is PPPE a better measure than MER to compare real incomes across nations? Would emissions projections necessarily depend on which measure of exchange rates we use to compare real incomes? This paper seeks to answer these questions. It examines the conceptual and empirical basis of the PPP theory and concludes that PPPE is based on a set of very restrictive assumptions which (except one) are relaxed in MER. In fact, MER is a generalisation of PPPE. One can make a consistent comparison of real incomes by using the real exchange rate, which is a ratio of MER to PPPE, not by using just the PPPE or MER. MER, however, can be used to convert nominal current incomes from different currency units into units of a reference currency and the prices in reference currency units can be used to deflate the nominal incomes into constant price incomes to make valid real income comparisons across nations. Finally, it is argued that comparison of real incomes is not always necessary in making emissions projections. "
Posted on 07/14 at 11:56 AM.
Your comments
IanC
Peter,
In my opinion, Hom Pant and Brian Fisher are wrong to believe (if they still do) that “MER ... can be used to convert nominal current incomes from different currency units into units of a reference currency and the prices in reference currency units can be used to deflate the nominal incomes into constant price incomes to make valid real income comparisons across nations.” There is clear evidence that their method is flawed in the forecasts presented in a paper by Brian and four of his ABARE colleagues to an APEC Business and Climate Change Workshop in Seoul in April 2005 (published as ABARE Conference Paper 05.6):
‘China ... is forecast to overtake Japan as the second largest economy in APEC at about 2028. By 2030, China is expected to be just over a quarter of the size of the US economy’ (Anna Matysek, Melanie Ford, Sam Hester, Lindsay Fairhead and Brian Fisher, “The Nexus between Climate, Energy & Technology”, p. 4).
China’s population is over 10 times that of Japan’s and over four times that of the US. In 2008 China produced 50% of the world’s output of pig iron and cement, 38% of the raw steel, 34% of the aluminium and 18% of the commercial energy. Its share of the INCREASE in the global production of these items between 2000 and 2008 was 90% for pig iron, 75% for steel, 70% for aluminium, 67% for cement and 51% of commercial energy (Sources: US Geological Survey Commodity Summaries and IEA World Energy Outlook 2008).
If the only data available on the relative economic size of countries were measures of tons of commodities produced, numbers of telephones and internet connections in service, numbers of students attending schools, numbers of patients being treated in hospitals and other similar indicators, I doubt whether anyone would give serious credence to a forecast that the US economy would still be four times the size of China’s in twenty years’ time.
It is now over 70 years since Dr Colin Clark and Dr (later Sir) John Crawford presented PPP-based comparisons of per capita incomes in a number of countries in “The National Income of Australia” (1938), and over 60 years since Dr Crawford became the founding Head of ABARE (then the Bureau of Agricultural Economics). It is surprising that ABARE has persisted with an approach which was rejected by its founder in his first major work.
I was pleased to note that the Australian Government paper “Australia’s Low Pollution Future”, released by Minister Wong and Treasurer Swan on 30 October last, cited Castles and Henderson (2003) in support of the statement that “The choice of measurement method significantly affects the validity of economic growth projections and energy use and, hence, projections of future climate change.” The paper recognised that “The MER/PPP debate is important for productivity convergence assumptions, as overstating income gaps will overstate economic growth in developing countries”, and that “This assumption is also important for estimates of global mitigation costs: cost estimates based on MER exchange rates tend to understate global abatement costs” (Chapter 2, p. 19).
Senator Wong and Mr Swan described the Treasury’s analysis as “one of the largest and most complex economic modelling projects ever undertaken in Australia.” This amounts (I hope) to a decisive rejection of the IPCC claim that it makes no difference whether PPP or MER measures are used in analyses of economic growth, energy and emissions intensities, mitigation costs, etc.