Why might one expect environmental Kuznets curves? Examining the desirability and feasibility of substitution

Abstract

This paper provides simple, transparent intuition for the perhaps surprising and certainly widely debated empirical findings of ”environmental Kuznets curves”, i.e. U-shaped relationships between per-capita income and indicators of environmental quality. We consider one possible component of such relationships: the linkage between income and household choices that impact upon the environment. Our explicit model emphasizes two features. First, degradation of the environmental endowment is a by-product of household activities. We present a household production model in which consumption of marketed commodities generates both a ”good”, desired non-environmental services, and a ”bad”, degradation of the environment. Second, while households can not directly purchase environmental quality, they can reorganize their activities so less degradation results. If environmental quality is a normal good, one expects substitution towards less degrading commodities, so that increases in income will increase environmental quality. We show that natural constraints on the desirability and feasibility of such substitution can produce non-monotonic relationships between household income and environmental quality, and in particular can produce household-level environmental Kuznets curves. ∗We would like to thank for their helpful comments Matt Kahn, Arik Levinson, two anonymous reviewers, and participants in AERE/ASSA, NBER, NEUDC and Harvard Environmental Economics and Policy seminars. Needless to say, we alone are responsible for any remaining errors.