Why the Social Cost of Carbon Will Always Be Disputed

by | Dec 10, 2018

A key reason is that the unique complexity of the Earth system, and the unprecedentedness of likely climate change over the next century, mean the economic value of future climate damage cannot be estimated by normal scientific methods.

Many governments and businesses use a carbon price to guide their climate-related policies and decisions, like standards for energy-using equipment or the level of a carbon tax.  One type of carbon price comes from using an “integrated assessment” model of the global economy and environment to estimate the social cost of carbon: the dollar value of future climate damage (for example from worse heatwaves and hurricanes in 2050) caused by an extra tonne of carbon emissions now.  A key input to any such estimate is a “climate damage function”, usually stated as the proportion of global economic output destroyed by various levels of global warming.

In WIREs Climate Change, Jack Pezzey argues that the unique complexity of the Earth system (including humanity), and the unprecedented level and speed of warming over the next century or so, together mean that the monetary value of future climate damage is “highly unknowable”, and cannot be estimated by normal scientific methods.  The social cost of carbon will therefore always be deeply disputed among climate economists, and implicitly so between some economists and climate scientists.

The article notes that a leading model’s calculation of “optimal” peak warming of 4o °C in 2165 is derided by most climate scientists as dangerous, and its underlying assumption of only 4% output damage from 4oC is viewed by many climate economists as guesswork.  Current claims that statistical analyses of past weather impacts on local economies will significantly improve damage functions, rest on untestable, far-out-of-sample extrapolation.  By contrast, predictions in astronomy, geology and other earth sciences, like Einstein’s prediction of gravity waves, can be and are tested by comparisons or natural experiments.  Direct testing of climate science projections for a century hence is impossible, though quantitatively stable, physical laws, like those for radiative absorption and water evaporation, give some confidence in these projections.  However, the added complexity of human behaviour means there are no stable laws for valuing the damage from mid-century hurricanes, heatwaves and the like, which remains essentially unquantifiable.

So instead, the article recommends basing climate policies on another carbon price: the marginal cost of carbon emissions abatement calculated by models of low-cost pathways to agreed, physical climate targets, like the UN’s 2 oC limit on warming.  Such a pathway approach allows detailed analysis of sectoral issues, like the best way to decarbonise electricity supply.  It is no panacea, as there are still deep uncertainties about abatement costs and targets, and it may pose political, professional and psychological challenges to many economists; but it avoids any illusion of optimality, and is more consistent with the target-based nature of most climate policies.


Kindly contributed by Jack Pezzey (Fenner School of Environment and Society, Australian National University).

ASN Weekly

Sign up for our weekly newsletter and receive the latest science news.

Related posts: