Information disclosure programs are implemented as a form of environmental regulation, yet there is little empirical evidence regarding their impact on the behavior of firms and consumers. This project evaluates the effects of an information disclosure requirement of the U.S. Safe Drinking Water Act (SDWA) Amendments of 1996. The Amendments required that drinking water suppliers provide an annual consumer confidence report (CCR), detailing violations of federal drinking water regulations, as well as average concentrations of drinking water contaminants like lead and bacteria.
Economists' ability to establish a causal link between information disclosure and pollution has been hampered by inadequate data on pollution levels before disclosure was required and the lack of a suitable control group of non-disclosing facilities. For the CCR regulation examined here, data are available on contaminant levels and violations both before and after the regulation. Furthermore, a quasi-experimental approach is facilitated by the existence of a threshold for mailing the CCR directly to households (mailing is required for suppliers serving 10,000 or more people).
Using a panel of 556 Massachusetts water suppliers, 1993-2003, this research will assess the effects of the CCR requirement on water suppliers' behavior. To examine whether drinking water suppliers violate federal drinking water standards less frequently when they are required to report to consumers, panel count data models and regression-discontinuity models will be used. Whether or not a change in the frequency of violations (relatively infrequent events) is observed, water suppliers may attempt to reduce the probability of a violation by managing the mean and variance of contaminant loads in drinking water sources. The PIs will estimate the effect of the CCR requirements on both the mean and variance of reported contaminant loads using fixed and random effects models. The proposed research will also explore whether water suppliers react strategically to information disclosure regulations. For example, they may delay testing for contaminants when violations are more likely, trading a potential violation of contaminant standards for a procedural violation. To address this question, the PIs will use seemingly unrelated regression to estimate a system of two equations in which the dependent variables are (1) contaminant violations, and (2) procedural violations. Cross-equation tests will determine if the CCR requirement has had differential effects on these two violation types.
Broader Impacts: Water suppliers incur significant costs to comply with information disclosure requirements, and additional social costs arise from regulatory administration, monitoring and enforcement. This research will be the first analysis of the effects of the federal CCR regulation on compliance with drinking water standards. It will also provide the first opportunity with a research design (supported by excellent data) sufficient to estimate causal effects of information disclosure on the behavior of regulated entities, in general. Results will provide critical information in an era in which the total costs of compliance with federal environmental, health and safety regulations is increasing, and regulation by information is becoming more common. The research will establish the first formal collaboration between the Schools of the Environment at Duke University and Yale University, enhancing the research infrastructure of both institutions and supporting the dissertation research of PhD students in each department. An important aspect of the project's broader impacts will be the PIs' continued collaboration, through the sharing of results, with state and federal drinking water regulators.
