Optimal carbon tax doubled

Typically, cost–benefit analysis (CBA) has suggested ‘optimal’ carbon tax regimes that result in a global temperature rise of around 3 °C, or even eventually (post-2100) 4 °C, above preindustrial levels. However, risk analysis approaches indicate that these levels of temperature rise result in climate change impacts that pose a high or very high level of risk to society and ecosystems1 (Fig. 1). Thus, the risk analysis approach has generally indicated that higher levels of climate change mitigation are needed, for example, to constrain global mean temperature rise to around 2 °C above pre-industrial. A particularly controversial aspect of CBA is the representation of consumer preferences relating to the future. These assumptions are the strongest drivers of CBA outcomes. Writing in Nature Climate Change, Benjamin Crost and Christian Traeger2 describe an improved approach to determining consumer preferences in CBA — called Epstein-Zin utility — that leads to a lower level of optimal climate change (that is, the level of climate change achieved through carbon taxes set at an ‘optimal’ level through which economic resources are distributed in a way that maximizes welfare) and results in a temperature increase of approximately 2 °C by 2100, bringing the implications of CBA closer to those of risk assessment. In CBA, optimal carbon taxes are derived by maximizing economic consumption over time, taking into account the combined effects of taxation and climate change damages. This insight arises from recent developments in finance that separate consumers’ preferences about time and risk: Crost and Traeger call this a ‘disentangled’ approach2. Previously, CBA has mainly relied on the assumption that consumers are just as concerned about short-term fluctuations in consumption as they are about the risk of future consumption loss. In fact, observed market data do not support this assumption3. The data show that consumers prefer to sacrifice consumption in one period to get it back with certainty later, rather than gamble over whether the future has greater or smaller consumption. In other words, the new approach “distinguishes risk aversion from the desire to smooth consumption over time”2 — a concept traditionally expressed as a single parameter known as the elasticity of the marginal utility of consumption. The authors’ approach better fits market observations than the traditional one. Crost and Traeger apply this approach within the DICE2007 model4, an integrated assessment model of climate and the economy used mainly to inform climate policy in the US. The result is a doubling of the optimal carbon tax and emission abatement, resulting in a peak and decline of global temperature trends, which remain close to 2 °C in 2100 and peak at about ENVIRONMENTAL ECONOMICS