سرریزهای فرامرزی و تمرکز زدایی سیاست های زیست محیطی

Most US federal environmental policies allow states to assume responsibility for implementation and enforcement of regulations; states with this responsibility are referred to as “authorized” or having “primacy.” Although such decentralization may have benefits, it may also have costs when pollution crosses state borders. This paper estimates these costs empirically by studying the free riding of states authorized under the Clean Water Act. The analysis examines water quality in rivers around the US and includes fixed effects for the location where water quality is monitored to address unobserved geographic heterogeneity. The estimated equations suggest that free riding gives rise to a 4% degradation of water quality downstream of authorized states, with an environmental cost downstream of $17 million annually.

Public policies for pollution control in the United States are a hybrid of centralized standard setting and decentralized implementation and enforcement. Some observers question the efficiency of centralization and argue for greater decentralization of environmental decision-making. Decentralization may allow policies to vary more with their local benefits and costs: although centralized policies could contain local variation, federal authorities may find much variability politically difficult and may have less information than state authorities. However, decentralization may be costly if the federal government can realize economies of scale in expertise, if “a race to the bottom” in environmental quality occurs as states compete for new investment, or if there are transboundary spillovers and states free ride.1 This study evaluates the empirical relevance of the latter concern about decentralized environmental policy. In particular, it examines whether states that control their Clean Water Act (CWA) programs free ride on downstream states. States received this control—known as “authorization”—over their programs at different times. Using data on in-stream water pollution levels at about 500 river monitoring stations around the country from the National Stream Quality Accounting Network (NASQAN), I estimate equations that model water quality at a station as a function of whether the state or any upstream neighbor has authority over its CWA program, time-varying state and river characteristics, and a monitoring-station fixed effect for unobserved geographic heterogeneity. The paper uses a water quality index (WQI) based on levels of five common pollutants.

The empirical results are consistent with the hypothesis that states have both the will and the way to free ride under Clean Water Act regulations. Federal policies that grant states authority to run their own programs appear to allow free riding. By focusing on changes in policy regimes in upstream states, the estimated equations address unobserved geographic heterogeneity that might otherwise make it difficult to isolate such effects. Although such transboundary free riding is often cited as a justification for federalizing environmental policies, the results in this paper do not necessarily support more centralized policy for three reasons. First, my empirical results suggest that federal standards do not prevent free riding. Allowing states discretion in implementation and enforcement of standards appears to be sufficient for free riding to continue.25 Second, problems with free riding must be weighed against the benefits of decentralization. Because free riding costs only $17 million, it may not overcome the greater flexibility and informational advantages of decentralization. In addition, the optimal response to free riding may not be centralization, but rather decentralization in combination with more targeted responses to spillovers. For example, the federal government might continue to decentralize decision-making but provide subsidies (or levy fees) on the chosen environmental standards to reflect costs to other states. However, Oates [21] questions the political feasibility of such approaches. Finally, free riding may not be detrimental if pollution control policies are inefficient. Recent studies suggest that CWA may not pass a cost-benefit test [11] and [29]. If so, the observed free riding could provide a net benefit by reducing overcontrol of pollution.