Behavior in a Dynamic Environment with Costs of Climate Change and Heterogeneous Technologies: an Experiment∗

We report the results of an experimental study of production decisions in an environment with dynamic costs of pollution and climate change. Every period, agents simultaneously decide on own level of production that generates private production revenue and emissions. Emissions lead to pollution which acts as a public bad and accumulates over time. Our main treatment variable is the emission propensity of agent’s production technology, resulting in three treatments: homogeneous baseline, heterogeneous with low pollution propensity and heterogeneous with high pollution propensity. We characterize and use as benchmarks for observed behavior the Markov perfect equilibrium and social optimum. We find that the observed production allocations are between the Markov perfect equilibrium and social optimum. With experience, the strongest adjustment and lowest level of pollution is achieved in the heterogeneous treatment with high emission propensity. When the costs of pollution and climate change are not severe, institutions are most necessary to create incentives for environmentally friendly behavior. JEL classification codes: C90, Q54, C72, Q50, Q01, C61.