Using Economic Incentives to Regulate Toxic Substances

parameterized (empirically for a sample of substances, or perhaps by way of simulation) to derive rough benchmark numbers as guidelines for regulation. For instance, administrative costs might be hypothesized to increase with the number of firms at the relevant life- cycle stage, and an approximation of the number of firms beyond which administrative costs are likely to be too large to justify targeted intervention might be derived. A more formal cost-benefit analysis might also indicate the size of social savings from incentive-based approaches as well as winners and losers (for example, regional or industry employment impacts) that might be necessary to anticipate in order to make incentive approaches politically feasible.

Another topic, treated only lightly here, is the interactive effects of multiple layers of regulation of a substance. This topic includes possible effects of intervention at multiple stages of the life cycle (production, disposal; or workplace, home), as well as effects of intervention for different media (air, water, soil) or other factors (hazardous materials transportation). Without analysis of how these effects combine with one another -- whether, when taken together, they offset or reinforce desired effects -- we do not know if our proposed intervention strategies are, on net, welfare enhancing or whether, when considered together, they might push a product out of the market. Accordingly, a future research topic might consider interactive effects in a few case studies to test whether these distortions are worth further scrutiny by regulators.

Another important extension of this research would be to refine the definition of cost-effective regulation. This would involve consideration of how a particular regulatory approach affects the future technology choices of firms, particularly regarding the relative health and environmental effects of substitute processes or products. The research could consider whether regulatory interventions appropriate under a static concept of efficiency would still be appropriate for achieving dynamic efficiency. An additional consideration would be whether there are reasons (perhaps in terms of signaling intent to regulate in the future) why command-and-control regulatory approaches might have more desirable dynamic features.

Industrial Economics, Inc. 1989. "Multi-Media Incentives for Reducing Release of SARA 313 Chemicals" Draft ( Cambridge, MA, September).

-132-