Intergenerational equity under catastrophic climate change

International audienceClimate change raises the issue of intergenerational equity, as catastrophes may unfairly affect some generations. As climate change threatens irreversible and dangerous impacts (IPCC, 2014), possibly leading to economic collapse, the tradeoff is no longer between present and future consumption, but between present consumption and the possible future extinction of civilization (Weitzman, 2009). This paper aims at identifying public policies that can reduce unfairness and strike a compromise between present and future generations when the potential impact of catastrophic climate change on the economy is accounted for. It explores the impact of inequality aversion and the impact of the attitudes towards population size on optimal climate policy when that catastrophic risk is accounted for.We use an integrated assessment model which simulates the future joint evolution of the economy and the climate. The Response model is a dynamic optimization model (Pottier et al., 2015), which belongs to the tradition of compact integrated assessment models such as DICE. It combines a Ramsey-like macroeconomic module and a climate module, and can be used to determine the optimal climate objective by comparing mitigation costs and avoided climate damages. We account for the risk of extinction due to climate change, first by assuming an exogenous probability of extinction, then by assuming a probability of extinction that depends on the temperature and on the level of economic output. We assume that the states of the world where people are wealthy will show enhanced resilience to climate damages, and will therefore face a lower probability of extinction under a given level of climate damages.Integrated assessment models can reveal the policy implications of various normative choices through the use of social welfare functions. Ideally, social welfare criteria should (i) account for inequality aversion (it should thus differ from the utilitarian criterion, which does not account for the distribution of utilities), (ii) satisfy the Pareto principle (i.e. it should account for the preferences of individuals) and (iii) be separable (i.e. the situation of past generations should not impact the evaluation). The ‘Expected Prioritarian Equally Distributed Equivalent’ (Fleurbaey and Zuber, 2014) accounts for the inequality aversion of the social planner and respects a weak form of Pareto, but is not separable under risky prospects (but we will not consider past generations in our analysis). This criterion includes a critical level of consumption which can be interpreted as the level of subsistence.In order to reveal the impact of inequality aversion on optimal climate policy, we consider various isoelastic functions φ, which translate the inequality aversion of the social planner, and various isoelastic functions u, which translate the risk aversion of the social planner. Inequality aversion here relates to the average consumption by individual at time t. Preliminary results using Response in the certain case (with no risk of extinction), and using the standard discounted utilitarian criterion, show that the optimal policy (i.e. the timing of optimal abatement) is identical for equal discount rates r = ρ + ηg, whatever the values of the pure time preference rate ρ and the degree of inequality aversion η. In the risky case, the degree of inequality aversion has a significant impact on the optimal climate policy. In order to reveal the impact of the attitudes towards population size N (which refers to the total number of individuals in all generations), welfare is weighted by the population size through αN, which is a bounded increasing sequence of N. In this case, the criterion gives priority to large populations