So far, the debate about cost-benefit analysis (CBA) on this site has largely concerned philosophical issues. In this posting, I would like to direct readers to a more practical set of issues: the difficulty of using CBA to reflect the immense complexity involved in most environmental policy choices.
In my experience, a typical environmental policy issue looks something like this: The government has become concerned about the quality of water in a river system that has been subject to many forms of human use - hydroelectric dams, recreational boating and fishing, agriculture, disposal of wastes, etc. The government might have available to it a number of policy instruments: remove existing dams (see Tim Haab “Dam it or Not” on July 14), raise the level of water behind each dam, change the rate and timing of water flow, change the allocation of water to agriculture, shore up river banks, build artificial wetlands or fish ladders, etc.,etc. Not only is the number of such instruments commonly very large, each such instrument can take on many possible values. There are numerous possible rates of water flow, for example; an array of alternative regulations concerning emission of wastes; and many options for building wetlands. And each set of policy instruments that is chosen will affect a large number of policy goals: quality of drinking water, survival of various threatened species, recreational experience, etc. (For more on “complexity,” see my posting on July 20)
If you combine the number of possible types of policies (e.g. remove dams), with the number of levels of each policy (e.g. remove one, two, three dams), and the number of possible benefit and costs that each level of each policy can have, the volume of information the agency will need, and the number of assumptions it will have to make, becomes enormous. Indeed, many, if not most, issues are so complex that it is hard to believe that there is any level of expenditure that would be sufficient to allow CBA to make a useful contribution to policy formation.
The result is that CBA is usually asked only to make simple, either/or decisions. Should we remove THIS dam? Should we build THIS power plant? But this is a failure of policy. The relevant decisions are not “either/or.” Whether a power plant of a particular type, with a particular generating capacity, should be built at a particular location is only one tiny part of a much larger issue: Should we generate more power? If so, of what type? If of a particular type, with what kind of technology? If with a particular technology, in what size of plants, in what locations? Etc. Etc. Doing a CBA on a single plant or any other single-valued policy option is just pretending that we are doing policy analysis.