ec3545handout7c

Sum-of-Specific Damages

The sum-of-specific damages approach to valuing environmental improvements is intuitively plausible, but not without difficulty in both implementation and interpretation. The benefits of pollution clean-up are damage reductions resulting from a policy. This approach to estimating the benefits of clean-up, then, merely "adds up" damage reductions and puts a dollar value on them. There are, of course, many different types of damages associated with pollution--morbidity, mortality, materials damage, crop damage, aesthetic damage, etc. The process is as follows:

For each damage category, determine how much reduction in physical damage will occur as a result of the policy (e.g. a reduction of 1 microgram of SOx, say going from 17 to 16, may save 23 lives in the N.Y. metropolitan area)
Multiply the damage reduction by a valuation, the valuation being the marginal willingness to pay for the reduced damages (e.g. personal valuations of changes in the probability of death in many settings have been estimated at between $2 and $7 million dollars--say, $3.5million. Thus, the lives saved would be worth 23 x $3.5 = $80.5 million).
Do the preceding for every physical damage that the policy would reduce (e.g. more bushels of soybeans times their market price, various morbidity measures times their values, improved length of time between painting homes times that value, etc.).
Add the dollar values up--this will give you the Marginal Benefits in convenient dollar terms to be compared with the dollar costs of the policy.

    Note that this seemingly plausible approach is, in fact, quite complicated and uncertain. It is difficult to know how many people will live (or experience any other health improvement) as a result of a policy that would have died (or not have experienced the health benefit) without it. It is also difficult to value the various damage reductions (e.g. how many deaths are "worth" one hundred thousand new bronchitis cases?). I believe that the valuations are more likely to be "accurate" (within an order of magnitude of the true value) than are the estimates of physical damage reduction, but that might be controversial. At any rate, attempting to ascertain what the physical damage reductions (the benefits) will be if environmental quality is improved requires the expertise of people from many disciplines (biology, medicine, epidemiology, materials science, etc.). Approaches taken, for example, to determining health effects might be
    cellular studies--controlled, suggestive of potential damages to be further investigated
    laboratory experiments--controlled, suggestive (but, large doses and animal variation make extrapolation hard)
    human laboratory experiments with volunteers--controlled, low dose, acute damages only; difficult to extrapolate
    case studies (retrospective studies)--group with disease compared to group without disease, behavior compared (e.g. lung    cancer and smoking behavior). Not "perfect" controls, but informative.
    epidemiological statistical investigations--relating behavioral information from large numbers of individuals to the health outcomes they experience (e.g. nurses study, Farmington heart study, etc.). Also, time series and cross-sectional studies of large population groups (e.g. Lave and Seskin's classic Science article in 1969, and later book). Realism of on-going behavior plus "statistical controls," but difficult to control for missing information, mobility, etc.
    BIG ISSUE: Perceptions. The sum-of-specific damages approach implicitly assumes that damages are not perceived as relating to pollution. Rather, the physical effects "just happen," with more bad things happening at high pollution levels and fewer bad things happening at lower pollution levels. Hence, one might think of a "demand" for, say, healthy days and some initial quantity (before the policy). The policy results in more healthy days (a shift out in the vertical supply from H0 to H1) which are valued at some "average" value over that range (see graph in class). Thus the benefits of the policy are measured as the gain in area under the demand curve for the increases in the goods we care about (healthy days, reduced materials damage, etc.)
    If, on the contrary, people were aware that the physical effects were related to pollution, they would engage in behavior designed to mitigate the effects. That is, they would be taking any action having B>C to eliminate the damages prior to the policy that improved environmental quality. Thus, one may think of a "supply" of healthy days to go along with the demand for healthy days (see graph in class). In this view, the benefits of the policy take two forms: 1) the original quantity of healthy days, H0, can now be had at lower cost, and 2) the net benefits from the additional healthy days that are optimal to "purchase" at the now lower cost (see graph for clarification).
    Note that there is no necessary relationship between the sizes of the two areas (if the supply curve is very flat, for example, the benefits might be quite small under the second view). If you are "buying" a lot of healthy days, the cost savings might be large on the initial quantity compared to the net benefits of the additional healthy days bought; if you are "buying" a few healthy days, the net benefits will be the dominant component. The steeper the supply curves the closer the two perceptual concepts become to being the same--if you have very high costs of changing the number of healthy days, it is the same as if you just "get" a certain number. (play with the graphs).