Where is it Cheapest to Cut Carbon Emissions: Estimating Marginal Costs

This is the second part of a series on my new working paper on where it is cheapest to cut carbon emissions.

In the previous post I assumed that all countries shared the same marginal cost of abatement curve. In reality this is not the case and in order to rank countries by marginal cost of abatement or total costs of meeting a given policy we need to estimate a cost curve for each country. The Treasury Review assessed the costs of meeting 4 different policy scenarios at 2020 and 2050. This gave us 8 data points for each region or country.

The data we used to estimate the curves are GDP data rather than GNP data because changes in GNP include net receipts for the sale of emissions permits. Our data on the cuts in emissions are the actual domestic reductions in emissions in each region. We don't want, therefore, to use a measure of cost that includes the costs of offsets.

Also we subtracted the "terms of trade impact" component from the estimates of GDP losses and gains. Under a global climate policy countries experience GDP impacts that have nothing to do with their domestic mitigation efforts but are the effect of climate policy actions elsewhere. For example, OPEC countries are hard hit under global mitigation efforts because of the reduction in world demand for oil. On the other hand, oil importers such as India gain from the fall in the oil price. Again, because we wanted to limit costs to the costs of domestic mitigation we needed to remove these effects.

Obviously a lot of technological change is expected to take place over the next 50 years. This technological change would be expected to lower the emissions intensity of the economy even in the absence of climate policy. This lowers business as usual emissions compared to what they would be without technological change. We make the assumption, though, that the elasticity of GDP with respect to reducing emissions relative to business as usual does not change. It's a strong assumption, but we think it is the best we can do with the data we have available.

These are the resulting 2005 marginal costs that we come up with:



The left side of the table ranks countries by the marginal cost of abating a ton of carbon (not carbon dioxide) where marginal cost is measured in PPP (Purchasing Power Parity) adjusted dollars. To get figures in terms of carbon dioxide multiply by 12/44.

Japan and Canada actually have GDP gains. This shows that despite the adjustments made these numbers are still not, of course, the effect of a unilateral domestic policy but of a coordinated global mitigation effort. Apparently, investment increases in Japan under climate mitigation increasing its GDP. Abatement is fairly cheap in the US and EU and expensive in most developing and transition economies. There isn't any correlation between emissions intensity and marginal cost. Countries appear to be on different cost curves.

In terms of marginal domestic loss of GDP climate policy is expensive in developing countries and cheap in the developed world.

The right hand side of the table uses actual market exchange rates instead. Now it is cheap to abate in China and OPEC as well as in the US and EU.