Determining expected benefits for each scenario involves estimates of benefits and costs: the likely benefits of avoiding increases in sea levels of over 1 metre have to be weighed against its likely costs. Estimates of the benefits and costs of different scenarios will be contentious: here again, the appropriate basis for making such estimates is determined by the weight of scientific and economic opinion.  Sir Nicholas Stern has proposed that the cost of doing nothing is likely to be 5% of global GDP per year, and up to the cost of the Great Depression and the two World Wars combined at the upper end of estimates. He considers the global cost of stabilization of CO2 at 30% above current levels is likely to be 1% of global GDP per year as compared to what it would have been without greenhouse gas reduction. Other estimates have put the cost of CO2 reduction at 2% of global GDP.

On the basis of expected benefit, it seems clear that we ought to reduce CO2 emissions, since the cost of doing nothing, even at the lower end of an authoritative estimates, seems much more than the cost of CO2 reduction at the higher end of estimates. What conclusion should we come to if we employ precautionary rather than prudential reasoning? Some of the global warming critics in the media appear to invoke precautionary rationality against doing anything: they state that the cost of CO2 reduction will be ‘disastrous'. However, there seems to be no evidence that this is so. Certainly, the cost of suddenly imposing a full carbon tax overnight would be disastrous but nobody is proposing such a policy. Everyone expects that a carbon tax, or an equivalent CO2 emission license trading scheme, will be phased in, reaching its optimum after about two decades.

More realistic but still conservative estimates of the cost of CO2 reduction are not catastrophic. The ABARE report[6], for example, considers the cost of CO2 reduction under a number of policy scenarios against a ‘business as usual' reference scenario, which assumes that no carbon tax or its equivalent will be imposed, although technological change and efficiencies may mitigate increase in CO2 emissions[7]. The reference scenario also assumes away any costs of global warming resulting from a global increase in CO2 emissions from 34 giga-tonnes in 2001 to over 84 giga-tonnes in 2050. This study makes rather conservative estimates of the cost reductions in renewable energy technologies in its reference and other scenarios, even though recent developments in solar panel design suggest that the cost of solar panels may be cut by up to 75% in the near future. It also makes rather conservative estimates of growth in use of geothermal energy, which is less costly than nuclear power.

The worst case scenario in the ABARE report estimates a reduction by 2050 of 10.7% in GDP for Australia and a 1-5% reduction of GDP for most other countries, relative to the reference scenario, for stabilization of CO2 at 40% below the reference level (ie at about 56 gigatonnes CO2, which is roughly a 40% increase from current levels)[8] Since the reference scenario envisages a growth of GDP for Australia of over 110% by 2050, a 10.7% reduction will still leave a doubling of GDP by 2050: this hardly seems catastrophic. ABARE does not, unfortunately, model the cost of the Stern report proposal of stabilization at 30% above current levels. The shortfall of GDP below the reference level for stabilization of CO2 emissions at 30% more than current levels may well be higher than 10.7% for Australia but it is not clear that it would be substantially greater than the Stern report estimate of 1% of GDP per year. In any case, the ABARE report excludes the effects of global warming which, if Stern is right, will be greater than the cost of mitigation even for Australia. If Nicholas Stern is correct, the cost of doing nothing could be up to 20% of GDP. It therefore seems that precautionary reasoning does not demand precautionary measures against catastrophic costs of CO2 reduction.