Essays in environmental economics and the role of risk, inequality, and time.

Summary This thesis studies several aspects of risk, inequality and time preferences as applied to topics in environmental and resource economics. It is divided into three independent chapters. The first chapter consists of the study of the optimal extraction of a non-renewable resource under uncertainty. A stylized discrete time approach derived from the literature on precautionary saving is proposed. This approach is compared to classical Hotelling-type studies. We find that the boundedness of the utility function and in particular the assumption on U(0) leads to very different results with the two approaches. We show that the agent's prudence is no longer sufficient to ensure a more conservative extraction policy with respect to the certainty situation. Our explanation for this surprising result is that the agent considers the trade-off between today and tomorrow's consumption only for the case where he has a strictly positive quantity of the resource tomorrow. This effect counteracts the effect of prudence. On the other hand, for an infinite number of periods, the result is the opposite of before. Prudence is no longer necessary and risk aversion is sufficient for a more conservative extraction policy. In the second chapter, the optimal discount rate is studied taking into account the inequality of income between countries. We show that if income dispersion decreases over time, the global consumption discount rate should be lower than in the case without inequality, under certain conditions. Using the growth projections used in the context of climate change, we find that the discount rate should be about twice as high for short horizons as without considering inequality, nevertheless it should be decreasing over the time horizon. Moreover, these results are supported in qualitative form, allowing for different parameters for inequality aversion, risk aversion and time preference. The third chapter (co-authored with David Anthoff) analyzes the roles of the equity coefficient and the discount rate, allowing the development of a method to study their impact on the social cost of carbon separately. We apply our method using the FUND model to calculate the SCC and arrive at three conclusions. The first is that the impact of the equity coefficient is significantly reduced if the coefficients used do not change the discount rate. The second is that the weighted estimates are sensitive to the resolution of the impact estimates used. The third is that the constant damage assumption seems unreasonable in this context. If this assumption is relaxed, the weighting impact becomes less important than in previous studies.