Inducing Environmental Disclosures: A Dynamic Mechanism Design Approach

Abstract

This paper studies the design of voluntary disclosure regulations for a firm that faces a stochastic environmental hazard. The occurrence of such a hazard is known only to the firm. The regulator, if finding a hazard, collects a fine and mandates the firm to perform costly remediation that reduces the environmental damage. The regulator may inspect the firm at any time to uncover the hazard. However, because inspections are costly, the regulator also offers a reward to the firm for voluntarily disclosing the hazard. The reward corresponds to either a subsidy or a reduced fine, depending on whether it is positive or negative. Thus, the regulator needs to dynamically determine the reward and inspection policy that minimizes expected societal cost in the long run. We model this problem as a dynamic adverse selection problem with costly state verification in continuous time. Despite the complexity and generality of this setup, we show that the optimal regulation policy follows a very simple cyclic structure, which we fully characterize in closed form. Specifically, the regulator runs scheduled inspections periodically. After each inspection, the reward level decreases over time until a subsequent inspection takes place. If a hazard is not revealed, the reward level is reset to a high level, restarting the cycle. In contrast to the reward level, the mandated remediation level is constant over time. Nonetheless, when subsidies are not allowed in the industry, we show that the regulator should dynamically adjust this remediation level, which then acts as a substitute for a subsidy. Our analysis further reveals that optimal inspection frequency increases not only when the inspection accuracy decreases, but also when the penalty for not disclosing the hazard increases.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 66 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.