Karen Fisher-Vanden and Sheila Olmstead set the stage for one way in which environmental regulators are trying to tackle the issue of nonpoint source pollution in their article, "Moving Pollution Trading from Air to Water: Potential, Problems, and Prognosis," which appears in the most recent (Winter 2013) issue of my own Journal of Economic Perspectives. Like all articles in JEP back to the first issue in 1987, it is freely available on-line courtesy of the American Economic Association. Fisher-Vanden and Olmstead write (citations and footnotes omitted):
"The stated goals of the Clean Water Act were: 1) the attainment of fishable and swimmable waters by July 1, 1983; and 2) the elimination of all discharges of pollutants into navigable waters by 1985. Obviously, those deadlines have been postponed through amendments, and distinctions have since been made between different types of pollutants. ... The Clean Water Act’s main tool is a set of effluent standards, implemented through point-source permitting. The National Pollutant Discharge Elimination System (NPDES) specifies quantitative effluent limits by pollutant, for each point source, based on available control technologies. For the most part, industrial point source compliance with these permits has been high. Municipal sewage treatment has also expanded dramatically, resulting in impressive improvements in urban water quality—for examples, see Boston Harbor and the Hudson River near New York City.
"But the gains from point source controls are reaching their limits. Even if all point sources were to achieve zero discharge, only 10 percent of US river and stream miles would rise one step or more on EPA’s water quality ladder. Nonpoint source pollution such as agricultural and urban runoff, atmospheric deposition, and runoff from forests and mines has become the major concern of water pollution abatement efforts. In fact, nonpoint source pollution from agricultural activities is now the primary source of impairment in US rivers and streams. Nonpoint source pollution involving nutrients like nitrogen and phosphorus causes excessive aquatic vegetation and algae growth and eventual decomposition, which deprives deeper waters of oxygen, creating hypoxic or “dead” zones, fish kills, and other damages. This problem is geographically widespread; seasonal dead zones in US coastal waters affect Puget Sound, the Gulf of Mexico, the Chesapeake Bay, and Long Island Sound. However, agricultural nonpoint source pollution is essentially unregulated by the Clean Water Act ..."
Before discussing what efforts are being made to address nonpoint source pollution, it's worth nothing that there are legitimate questions about whether the costs of the Clean Water Act have exceeded the benefits. For example, in the Winter 2002 issue of my own Journal of Economic Perspectives, A. Myrick Freeman III reviewed studies bearing on "Environmental Policy Since Earth Day I:What Have We Gained?" Even if one goes beyond just looking at immediate economic gains and takes into account survey evidence on people's willingness to pay for knowing that water is cleaner (so-called "contingent valuation" evidence), the overall costs seem to far outstrip the benefits.
The intuition behind this result is that there were some prominent bodies of water that were highly contaminated, and that have improved substantially since the passage of the law. But the law was not just applied to a few high-profile cases of water pollution: it imposed costs everywhere. Moreover, as noted above, the law called for (eventually) the total elimination of all discharges, and even a passing acquaintance with the law of diminishing returns suggests that reducing pollution by one-third or one-half or more might be done at fairly low cost, but when it comes to figuring out how to reduce that last bit of pollution, the marginal costs may climb very high. .
But the question of past costs and benefits of the Clean Water Act is, well, water under the bridge. At present, the situation is that the law has been so effective at reducing point-source emissions that the main source of water pollution is nonpoint sources, and especially runoff from agriculture. There are a variety of voluntary programs to encourage reducing nonpoint source pollution. But such programs generally lack teeth. Thus, environmental regulators have been trying in some areas to tackle the problem through a back door--by creating a structure for tradeable emissions permits. As Fisher-Vanden and Olmstead describe it, the current clean water law
"... requires states to establish a Total Maximum Daily Load (TMDL)—basically a “pollution budget”—for each water body that does not meet ambient water quality standards for its designated use, despite point source controls. Designated uses include recreational use, public water supply, and industrial water supply, and each designated use has an applicable water quality standard. State courts began ordering the developmentof TMDLs in the 1980s and 1990s in response to lawsuits by environmental groups.Since 1996, the states in cooperation with the Environmental Protection Agency have completed thousands of TMDLs. Establishing a TMDL is a “holistic accounting exercise” in which all permitted sources and land uses within a watershed drainage area, including agriculture and urban runoff, are inventoried and allocated responsibility for portions of the pollution budget. While regulators cannot implement enforceable caps on agricultural pollution through this process, they have recognized the importance of incorporating agricultural abatement into clean-up processes, and water quality trading is one tool they have employed for this purpose."
They discuss a number of examples. In one fairly straightforward program here in Minnesota, the "Southern Minnesota Beet Sugar Cooperative, a beet processor, pays its 256 grower-members to invest in phosphorus-reducing land management changes so that the processor can meet its permit requirements for expanded production. In this case, the beet growers and the processing facility are treated under the processor’s permit as a single source to meet an overarching phosphorus effluent cap." A more complicate case involves the Chesapeake Bay, which receives discharges from six states and the District of Columbia, and in which three of the states are allowing for trading of water quality permits. Fisher-Vanden and Olmstead discuss several dozen of these programs around the country.
The practical and political advantage of using marketable permits are well-known among economists, and are a staple of most intro econ classes: specifically, those who need to reduce emissions can think about whether to do it themselves, or whether to pay some other economic actor--like a farm--for reducing emissions. With this choice, emissions get reduced, which after all is the goal, at lowest possible cost. But the practical problems of implementing such a scheme over a large area like the Chesapeake Bay, making sure that reductions in nonpoint source pollution really happen, and in a way that doesn't clean up one area of the Bay at the expense of another area, can be quite complex. My own sense is that the Total Maximum Daily Load concept is a very useful one for thinking about the causes of water pollution, but it's time to stop putting all the requirements on the point-source emitters of water pollution. For bodies of water that are not meeting ambient quality standards, there should be requirements for both point and nonpoint emitters to reduce their water pollution--with trading of emissions permits allowed between them.