Dynamic Regulation Design Without Payments: Timing is Everything

We consider a two period model of optimal regulation of a ï¬ rm subject to marginal compliance cost shocks. The regulator faces an asymmetric information problem: the ï¬ rm knows current compliance costs, but the regulator does not. Both the regulator and the ï¬ rm are uncertain about future costs. In our basic framework, the regulator may not offer payments to the ï¬ rm; we show that the regulator can vary the strength of regulation over time to induce the ï¬ rm to reveal its costs and increase welfare. In the optimal mechanism, the regulator Offers stronger (weaker) regulation in the ï¬ first period and weaker (stronger) regulation in the second period if the ï¬ rm reports low (high) compliance costs in the ï¬ rst period. Low cost ï¬ rms expect compliance costs to rise in the future, and thus prefer weaker regulation in the second period. High cost ï¬ rms expect costs to fall in the future and thus prefer regulation which becomes more strict over time. Thus the regulator offers the low (high) cost ï¬ rms slightly weaker (stronger) regulation in the second period in exchange for much stronger (weaker) regulation in the first period. We refer to our dynamic mechanism as “timing†the regulation. If the regulator can make payments, then the optimal mechanism to some degree times the regulation as long as a positive cost of funds exists. If the cost of funds is high enough, then under the optimal mechanism the regulator will not use payments and only use our timing mechanism