Fair-ish and Balanced-ish
Wednesday, May 14, 2003
Why Castles and Henderson Missed the Boat
The Castles/Henderson critique of the IPCC can be found all over the internet. Rather than try and summarise their case, you can read it here. Apparently it is going to published in the journal Energy and Environment.
I believe that the two authors miss the overall goal of the IPCC emission scenarios; that is, to create realistic projections of future GHG emissions.
My earlier post was an attempt at summarising the process in which the GHG projections come about. One of the Castles/Henderson complaints is that GDP should be replaced with PPP. This may well be true if we were interested in determining future standards of living, but we aren't. GDP figures are used because of the clear relationship between them and energy usage. While PPP figures could be used, there is considerable research on the relationships between PPP and energy usage than GDP usage. However, at the end of the day, it doesn't really matter, one could just as well use PPP, you would just need to find the relationship between it and energy usage. This is perhaps best stated by Nebojsa Nakicenovic (one of the lead authors on the report, and who has looked at energy usage with both GDP and PPP in the past) "[b]ut the physical world doesn't change whether we use this or another indicator to measure emissions".
Another complaint with the Castles/Henderson critique, is that it ignores economic scenarios that are inconvenient to it. Much is made of the rapid third world growth relative to the developed world. However, as far as I can tell, the two authors never mention the A2 and B2 scenarios. These are the scenarios which involve considerable smaller rates of growth in the third world.
While we are on this point, Castles/Henderson come very close to presenting the B1 scenario as the low growth scenario (ie. A1 is the high growth, and B1 is the low growth with reality coming somewhere inbetween). While the two authors never do this directly, they tend to compare the two scenarios, stating that the B1 scenario has the lowest level of emissions. While this is true, it is confusing, as the B1 scenario doesn't have the lowest level of growth. It gets worse, when they compare the IPCC scenarios with another (non-IPCC) high growth scenario used by Nebojsa Nakicenovic in her book Global Energy Perspectives.
Castles states: the highest IIASA/WEC projection (Case A: "High Growth") assumes that GDP (PPP) in 2050 will be $US 115 trillion (in 1990 prices). The lowest emissions IIASA/IPCC projection (the B1T MESSAGE scenario) assumes that GDP (PPP) in 2050 will be $US 140 trillion (in 1990 prices). This means that the latter scenario (published in 2000) assumes that gross world product will be greater than the "high growth" IIASA scenario (published in 1998) by an amount equal to the GDP of the entire world in 1990. This is despite the fact that the more recent projection assumes that the world population will be 1.4 billion lower than was assumed in the IIASA/WEC study.
This is sneaky, as they are comparing one high growth scenario with another high growth scenario, yet presenting one of the scenarios as being low. While this shouldn't confuse someone who is familiar with the various scenarios, Castles and Henderson initially choose to publicise their work in the public domain (where people are very unlikely to have read the emission scenarios) rather than submit it for peer review (where people are more familiar with the scenarios).
One problem with the emission scenarios that Castles/Henderson have correctly identified is the effect of regional grouping on several relatively developed countries. An example of this is South Africa. It is grouped in the Africa and Latin America region, where it is one of the more wealthier countries. Because of the poverty in this region, high growth rates are expected in some of the scenarios (particularly A1 and B1 where Castles/Henderson spend all of their time). This effect of high growth rates on South Africa's already developed economy has a very large effect, skyrocketing South Africa's growth. Castle/Henderson have noted this effect, however I also believe that they have overstated the relevance of it. The Emission Scenarios look not at individual countries, but rather at global regions (the OECD, the European ex-communist countries, Africa and Latin America, and Asia). Each of these regions is modeled separately. This has the effect of making South Africa more important than what it really is, but not the effect of making the aggregate Africa/Latin America region bigger than what it should be. This is an example as to why the emission scenarios should be used for nothing but projecting emissions.