source of non-point-source water pollution. Traditional environmental programs focused on agriculture have relied on financial incentives to encourage farmers to address conservation goals, mainly those associated with stopping soil loss. Constraints on public funds make the continuation of hefty incentive payments somewhat problematic and raise the possibility of regulatory fixes, similar to approaches adopted initially to deal with point-source pollution problems.
There are two basic approaches to standard-setting for agriculture. One specifies the production practices or technologies that producers are encouraged or required to use. Such "design standards" are traditionally used in agriculture. The alternative is to establish performance standards. Such standards would set an acceptable level of emissions or some other measurable indicator of environmental quality—for example, nitrate levels in tile drainage, acceptable rates of erosion, or phosphorus levels in surface soils—and allow the producer to determine the best method of meeting those standards. Performance standards leave the producer with the most flexibility to adjust but require more sophisticated scientific and technical capacity to set and monitor. Design standards are easier to set but lock producers into fixed and perhaps more costly and less-effective solutions.
Management of rangelands raises another set of public policy and institutional issues because about half of the nation's rangelands belong to federal or state government. Although private landholders may make management and use decisions with few constraints, public-land managers must often balance competing and conflicting claims advanced in statute or in practice. Consequently, the definition of the condition measure itself is controversial because selection of some indicators over others may imply that less weight is given to one set of values or uses over others. The NRC rangeland report attempts to address that possibility by providing the qualitative basis for a multidimensional index of rangeland health.
If the complementary roles of condition and performance measures are applied to agriculture, the task should be simplified somewhat by the fact that there is usually little uncertainty about what human activity is affecting the ecosystem. Still, it is often difficult to determine the ecological consequences of a given action, and spatial aggregation is a particular difficulty for a site-specific activity such as farming or ranching. Although the scientific basis for performance standards is not well developed, ongoing efforts to manage and alter agricultural systems that pollute provide a wealth of information for designing workable standards. To date, the agricultural community has resisted the development of such standards, preferring voluntary adoption of best-management practices to what appears might be unprecedented mandatory regulation of farming practices. In the long run, though, it will likely be less costly to farming and ranching to work to performance standards rather than to design standards. As experience with other industries has shown, design standards tend to lock technologies in place and discourage development of new ones. As agriculture seeks to take advantage