The comparative importance for optimal climate policy of discounting, inequalities and catastrophes

Climatic Change https://doi.org/10.1007/s10584-017-2094-x The comparative importance for optimal climate policy of discounting, inequalities and catastrophes Mark Budolfson 1 & Francis Dennig 2 & Marc Fleurbaey 3 & Asher Siebert 4 & Robert H. Socolow 5 Received: 1 April 2016 / Accepted: 13 October 2017 # The Author(s) 2017. This article is an open access publication Abstract Integrated assessment models (IAMs) of climate and the economy provide estimates of the social cost of carbon and inform climate policy. With the Nested Inequalities Climate Economy model (NICE) (Dennig et al. PNAS 112:15,827–15,832, 2015), which is based on Nordhaus’s Regional Integrated Model of Climate and the Economy (RICE), but also includes inequalities within regions, we investigate the comparative importance of several factors—namely, time preference, inequality aversion, intraregional inequalities in the distribution of both damage and mitigation cost and the damage function. We do so by computing optimal carbon price trajectories that arise from the wide variety of combinations that are possible given the prevailing range of disagreement over each factor. This provides answers to a number of questions, including Thomas Schelling’s conjecture that properly accounting for inequalities could lead the inequality aversion parameter to have an effect opposite to what is suggested by the Ramsey equation. 1 Introduction Models of optimal economic responses to climate change are contingent on highly uncertain estimates, including about the future societies with which climate will interact. Nevertheless, incomplete estimates exist, and when put into integrated assessment models (IAMs), they Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10584-0172094-x) contains supplementary material, which is available to authorized users. * Mark Budolfson 1 University of Vermont, Burlington, VT, USA 2 Yale-NUS College, Singapore, Singapore 3 Princeton WWS-UCHV, CFI, Princeton, NJ, USA 4 Columbia University EI-IRI, New York, NY, USA 5 Princeton CFI, CMI, MAE, Princeton, NJ, USA Climatic Change serve to provide some guidance on the extent of effort society should dedicate to this problem (Stern 2006; Nordhaus 2010; Nordhaus and Sztorc 2013; Tol 1996, 2009). There is no consensus that a high mitigation effort is warranted, but there are several instances of such recommendations by researchers using these models—most notably the Stern Review, which computes a high social cost of carbon based on very low discount rates (Stern 2006; Hope 2008; Nordhaus 2007). Other researchers provide a similar sense of urgency based on other features, such as the possibility of catastrophic damages (Weitzman 2009, 2012; Dietz and Stern 2015), the possibility that climate damages will disproportionately harm the poor within countries (Dennig et al. 2015), price effects (Sterner and Persson 2008) and damages affecting growth rates (Moore and Diaz 2015; Dietz and Stern 2015). Ackerman and Stanton (2012) combine variations about the discount rate, climate sensitivity and damages and obtain a wide range of carbon prices with the DICE model. In this paper, we investigate the interactions between several of these individually important factors by providing a systematic study of the comparative importance for optimal mitigation of inequality aversion, time preference, catastrophic damages and intragenerational distribution of damages and abatement costs. While a longer list of factors could be drawn, this selection enables us to make two main points. First, the direction by which inequality aversion influences optimal mitigation depends on the distribution of mitigation cost and damages within countries and the time preference parameter. Second, the distribution of mitigation costs and damages is of prime importance for optimal mitigation, whereas catastrophes become relevant only when inequalities are a low concern. Our investigation of these issues is made possible by further development of the Nested Inequalities Climate Economy (NICE) model (Dennig et al. 2015), which is based on Nordhaus’s Regional Integrated Model of Climate and the Economy (RICE 2010) (Nordhaus 2010), but also includes mechanisms to represent inequalities within regions and countries that are not included in RICE or in other existing climate-economy IAMs. Our first point about the importance of the distribution of damages and mitigation cost relates to Thomas Schelling’s conjecture that properly accounting for inequalities within generations could lead the inequality aversion parameter to have an effect opposite to what is suggested by the Ramsey equation (Schelling 1995).1 In the paper, we elaborate this conjecture and explain how NICE allows us to evaluate it. We then show that this Schelling Reversal can indeed occur but only under particular conditions combining regressive distribution of damages or progressive distribution of abatement costs, and sufficient discounting. The first two conditions result in the poorest in the future being overwhelming beneficiaries of mitigation, while the poorest today are shielded from the lion’s share of mitigation cost. But even in such cases, if there is high discounting, there are still large climate damages in the future. As a result, under those conditions, as inequality aversion increases, optimal mitigation increases to protect the future poor from those damages. Under low discounting, there is very high mitigation, to the point where most damages are avoided. As a result, with low discounting, as inequality aversion increases, optimal mitigation decreases to protect the 1 A related conjecture from Schelling is investigated by Anthoff and Tol (2012)—namely, that the best defence for poor societies against climate change impacts might be to develop quickly, rather than to control greenhouse gas emissions. Our results here shed additional light on this and other important questions framed and partially answered by Anthoff and Tol. Our results here indicate that proper representation of subregional inequalities is crucial to answering such questions—which supports Anthoff and Tol’s idea that answering such questions requires taking into account important heterogeneities (p. 271) and echoes Anthoff et al.’s (2009) plea to introduce intra-national inequalities in the future research. Climatic Change present (relatively poor) generation, as implied by the Ramsey equation. At intermediate levels of discounting, these two effects balance out, and inequality aversion has a small effect on optimal carbon prices. After illustrating the importance for optimal mitigation of within-region inequalities and the interaction of mitigation costs and damages with this distribution, we contrast it to the effect catastrophic aggregate damages on optimal mitigation. We compare the RICE model’s quadratic damage functions with specifications with convexities that wipe out half (...truncated)