Targeting Pollution Policy

Pollution abatement is very expensive. The federal effort alone may well be measured in hundreds of billions of dollars.

One way to reduce the costs of cleaning up the environment is to apply pollution regulations only where and when they are actually needed to improve the quality of the air and water. The apparent benefits of such targeting probably is behind President Reagan's call to return to the states more responsibility for environmental regulation in order to increase responsiveness to local conditions.

The gains of applying environmental regulations only where they would serve environmental objectives seem obvious, but Congress often has resisted such an approach. Both the Clean Air and Clean Water Acts tend to promote the development of technology-based regulations that apply uniformly to large categories of polluters (major new sources of air pollutants, for example) regardless of the natural ability of the receiving medium to absorb the pollutants, the preferences of those who may be affected by pollutants, or the measures taken to reduce discharges. The Clean Water Act neglects geographic location as well.

The legislative history of these acts reveals a congressional desire for "reasonably" uniform standards, largely motivated by a fear of losing local industries to regions with relatively weaker regulations. This concern clearly outweighed any offsetting worry about regulations failing to reflect local preferences.

Industrial flight aside, some reluctance toward targeting is understandable for purely technical reasons. Regardless of whether environmental control instruments are technology-based regulations or economic incentives such as pollution charges, their differential application to specific plants in specific locations requires an analysis of just how much a particular plant is contributing to environmental quality problems. Conducting such an analysis is not easy.

RFF has had a first-hand opportunity to learn this lesson. At the joint request of the U.S. Department of Agriculture and the U.S. Environmental Protection Agency, RFF was asked to determine for the nation the contribution of pollution from agricultural sediment to water quality problems as well as the costs of alleviating any of these problems.

The agricultural sector might be a good candidate for targeted regulations. In the first place, many analysts believe that agriculture is an important contributor to water pollution; some argue that over half the pollutants come from this source. Second, the problem of losing local farms to competing states with weaker regulations seems minimal, at least in the short term. Third, there appears to be no legal impediment to treating different farms differently. On the other hand, the technical problem of determining which farms need to be regulated seems as knotty as the counterpart problem in the industrial sector, and the marketing characteristics of agriculture present unique economic problems not shared by most industries.

Which agricultural land needs to be controlled? One starts with the fact that agriculture is but one among many contributors of pollutants to a particular body of water. Thus, the location and amount of loadings from municipal treatment facilities and factories, the physical characteristics of the receiving waters, and the pollution-abatement policies affecting all sources of pollution will determine the effectiveness of controlling specific agricultural lands. Some waters will be so fouled by other sources that even total control of agricultural runoff will not materially improve their quality.

All these factors have been analyzed with the RFF Environmental Data Inventory (REDI). The REDI describes pollutant discharges from 272 point and nonpoint sector classifications for all 3,100 counties in the United States before and after the simulated application of industrial and municipal water pollution control policies as defined in the 1972 version of the Clean Water Act. Using a water-network model, these discharges are linked to a detailed network of 304 rivers, 175 lakes and reservoirs, and 37 ocean bays. The system thus permits the estimation of the downstream ambient concentration of pollutants originating from all upstream sources at 1,306 locations.

Using REDI, water quality was estimated before any water pollution regulations, after the imposition, alone, of currently mandated point source (industrial and municipal) regulations, after the imposition, alone, of agricultural sediment controls, and after the imposition of both point source and agricultural controls.

While there are a number of ways of summarizing these simulations, one of the more interesting is to look at the areas for which water quality violated accepted standards before and after the imposition of control policies. If one looks at the aggregate of all geographic areas, one might conclude that both the point and nonpoint policies have been fairly ineffectual. For example, before controls, about 75 percent of the nation's waters violated phosphorus and nitrogen standards. This percentage is estimated to fall to about 70 percent if point-source controls are fully enforced. Controlling agricultural sediment alone would reduce the violation to about 60 percent. Finally, controlling both agriculture and point sources would reduce the violation percentage to about 50 percent. Thus, if the phosphorus and nitrogen standards are valid, the current policies—even if fully implemented and supplemented with agricultural controls —still would leave half the nation's waters polluted.

The situation may appear better when one looks at biochemical oxygen demand and dissolved oxygen standards. The preregulation violation percentages fall from 30 and 20 percent, respectively, to 17 and 7 percent after a combination of both point- and agricultural control policies. While these represent substantial reductions, the fact is that, with respect to these two pollutants, the water was fairly clean to begin with.

Is the current policy of first controlling point sources in error? After all, for every pollutant analyzed, violation percentages are reduced more by applying agricultural controls alone than by applying point-source controls alone.

As it happens, aggregate looks are inherently misleading. What matters is water quality at specific locations. Even if in the aggregate agricultural controls appear more effective than industrial point-source controls, they are not substitutable. Point-source controls obviously are far more likely than agricultural controls to affect urban areas. Point-source controls would produce major improvements in the industrialized Delaware River Basin while agricultural controls would greatly improve water in the Red River Basin of the rural Dakotas.

In any event, point-source controls are mandated by law. What is of more practical interest is to identify where the addition of agricultural controls would make a difference for the better.

There are two circumstances under which the imposition of agricultural controls in a particular area would not improve the violation rate. One is where the water is already clean, either because of or despite point-source controls. The other is where the water is so dirty, even after industrial and municipal controls, that agricultural controls would still leave it dirty. The costs of agricultural controls can be greatly reduced by eliminating such clean and dirty areas from consideration.

Thus, only 17 percent of agricultural areas can be effectively relieved of phosphorus violations, assuming that point-source controls already are in place. The corresponding percentages for nitrogen, biochemical oxygen demand, and dissolved oxygen are 17, 8, and 5 percent, respectively. In short, substantial savings are possible if agricultural controls can be targeted toward the minority of agricultural areas where they can make a substantial difference—most of them in the Midwest.

These findings are more qualitative than quantitative, and the models and data need to be far more refined before policies can be targeted with precision. Yet even if policymakers were satisfied with the models and these results, the economics of agricultural marketing may present another barrier to targeting.

Agricultural products trade in highly competitive markets and those farms that must assume the financial burden of sediment controls may not, in the short run, survive competition from uncontrolled farms. If farm demand increases as expected and less productive land is brought into production, some controlled farms may return to production and thus survive in the long run.

But this is chilly comfort now. The likely economic disruption and apparent dissimilar treatment of farms producing identical products may make targeting politically impossible. Direct subsidies to affected farms may alleviate the problem, but that would directly affect the federal budget, a politically unfavorable situation that the current nontargeted industrial policy happily avoids.

President Reagan's idea of increasing the responsiveness of pollution control policy to local conditions is good, but good ideas often are not enough in the face of technical and political realities.