摘要：We consider an environmental enforcement agency who uses the measurement of ambient pollution to guide its inspections of individual polluters. We compare two different uses of this information. In a first model, the agency uses a ``threshold strategy": if ambient pollution exceeds an endogenous threshold, the agency inspects all individual polluters simultaneously. In a second model, the agency inspects polluters sequentially, and s its beliefs with respect to the firms' behavior after each firm inspection. If the cost of delaying the inspection of noncompliant firms is low enough, this sequential inspection policy is superior to a simultaneous inspection policy. However, if the cost of delay is high, the agency is better off if it commits itself to ignoring some information embedded in ambient pollution.
