Quasi-experiments and environmental policy

This paper suggests applying experimental and quasi-experimental approaches to the study of environmental problems. Some important issues in quasi-experimentation are outlined and strategies appropriate for research on collective goods highlighted. Studies both of the demand for various underpriced goods and of programs for changing usage of a collective good are suggested. Questions of causality are treated in some detail. The logic of a new quasi-experimental design, the event-causing design, is presented and compared with that of the interrupted time-series design. Problems of site selection are discussed in terms of the types of projects for which random assignment of sites might be possible. New combinations of experimental and analytical techniques are suggested for improving the quality of research on environmental problems.