Why the Paris conference may not be enough
In December talks in Paris involving more than 200 countries may result in a new agreement aimed at reducing carbon emissions. In the months leading up to the conference, The Economist will be publishing guest columns by experts on the economic issues involved. Here, John Quiggin of the University of Queensland, Australia, argues why policymakers need to take further action to reduce carbon emissions if climate change is worse than expected. This piece has been corrected WITH global temperature data setting new records each month, "sceptical" positions about climate change have crumbled. Those seeking a more credible basis for opposing action to reduce carbon emissions have shifted their ground. Most have moved to the view advocated by "lukewarmers" like Bjorn Lomborg: that unmitigated global warming would not be so bad after all, and that adaptation is the best response. Surprisingly, advocates of this view cite the authority of the Intergovernmental Panel on Climate Change (IPCC) for this assertion. The leading source for this claim is Jim Manzi, editor of the National Review, who summarises the evidence presented in the IPCCs Fifth Assessment Report as follows: There are six estimates for expected impacts of warming at 3C or above, with a median estimated impact of a 3.6 percent reduction in global GDP for 34.9C of warming. If this claim is combined with a high-end estimate of the cost of holding warming to 2C, of up to 5% of global GDP, then action to mitigate climate change will harm the world more than the economic impact of climate change itself. So, lukewarmers such as Mr Manzi conclude, we should not bother with interventionist policies like carbon taxes and emissions trading schemes. But there are many problems with the estimates cited by Mr Manzi. He is relying on a handful of estimates, mostly outdated, which necessarily exclude much of the damage arising from climate change. And they focus on a relatively narrow range of possibilities for warming, associated with median estimates of the warming associated with a doubling of carbon dioxide levels. The rationale for this is the IPCC judgement that, if carbon dioxide levels stabilise at double the pre-industrial levels, warming of 6C or more, which would undoubtedly be catastrophic, is "very unlikely". That phrase is used to refer to probabilities of 0-10%. Closer examination of the IPCC reports suggests that the probability of catastrophic warming following a doubling of carbon dioxide levels is around 5%. Is that chance small enough to ignore? For many purposes, the IPCC conventions make sense. Much of the time, contingencies with a probability of 5% or less can safely be disregarded. But when we come to risk analysis, a great many of the risks which scientists are concerned about arise with probabilities of 5% or less, including a 60-year-old smoker's ten-year risk of dying of lung cancer. More than 95% of 60-year-old smokers will either live to 70 or die of some cause other than lung cancer. It would be absurd, however, for a 60-year-old to evaluate a decision to quit or continue smoking without taking account of the risk of lung cancer. In a properly formulated risk analysis, these "tail risks" (low-probability extreme events) play a much larger role than their probability might indicate. In the context of climate change, there are three main reasons for this: (a) By definition, the higher the sensitivity is, the greater is the warming associated with additional emissions. (b) The damage associated with warming increases very rapidly as the rate of warming rises. If global warming is held to 2C, the current international goal, adverse effects will be localised and relatively modest. Warming of 4C would change the world radically, with widespread species extinction and severe effects on agriculture and human health. And 6C or more would be a catastrophe, threatening the end of human civilisation. (c) Risk aversion means that a given additional loss is worse if it occurs when income is below current levels, as would be the case with high rates of warming, than when income is growing, as would be the case with low or moderate rates of warming. These factors multiply the effects. Ignoring the upper tail of the distribution of possible outcomes will result in estimates of the social cost of carbon dioxide emissions that are less than half the correct value. The estimates used by Mr Manzi, and relied on by the lukewarmers, fail to take these possibilities into account. Looked at in this light, even a 450 ppm target is not, as is sometimes suggested, "safe". Policymakers need to be alert for signs that climate change is greater than expected. If such signs appear, it will be necessary to accelerate efforts to decarbonise the economy more rapidly, and ultimately to remove carbon dioxide from the atmosphere through reafforestation, aiming at a lower target such as 350 ppm. Correction:This piece originally mis-stated Roger Pielke Jr's views. He thinks climate change is a serious risk and advocates emissions reduction as well as adaptation. Sorry.