37 
C. General Perspectives on Risk/Benefit Analysis in 
Public Policy Decisions 
Dr. Wilson 
As Richard Wilson, Physics Department, Harvard University 
began his testimony, he pointed o.ut that there are some situations 
where risk analysis is helpful and others where it is not helpful. Where 
data are available, as in correlating automobile accidents and the 
reduction in injury by wearing seat belts, estimates of the lives saved 
by installing seat belts versus the cost of the belts can be provided 
with reasonable reliability. In other subjects of interest in policy 
decisions, as in estimating the risk to health from exposure to radiation 
or chemicals, similar calculations are possible from data on high levels 
of contamination. It is a much more difficult problem to make esti- 
mates with a high degree of certainty when the risk from exposure to 
low levels or small concentrations is not well defined. The analysis 
becomes even more uncertain when extrapolations to large populations 
living under various environmental stresses are under evaluation. 
The risk to a small population may be small but with large numbers, 
even low level risks might produce a greater significance since small 
multiplication factors with large populations can result in danger 
to an unacceptably large number of numan beings; uncertainties as 
to the degree of risk may be larger. One major difficulty with some 
calculations of risk-benefit is that very important factors may be 
left out of the determination. For example, in exam inin g the relative 
risk of using nuclear power versus coal nred boilers, one might reach 
a conclusion, as is the case with most such calculations, that an ex- 
panded use of coal might produce an overall immediate hazard 
greater than from the use of nuclear power. On the basis of this deter- 
mination alone, a policy might be inferred which would lead to a 
conversion to nuclear power. However, while not related to the im- 
mediate health risk calculation, the potential for nuclear war from 
terrorist acquisition of plutonium from expanded use of nuclear 
power plants may be a more important factor which must be con-, 
sidered, and this might lead to a conclusion to accept the greater 
health risk from coal fired plants. . ' 
The point of this discussion was to emphasize that risk-benefit 
calculations frequently do not include all of the factors which must be 
taken into consideration when making a public policy decision. In 
addition, the uncertainties of any extrapolation of data from small 
samples to high populations must be clearly understood by any politi- 
cal group making a policy decision based upon such calculations. 
Dr. Wilson indicated that he had examined a number of public 
policy issues and had observed that in many instances public dis- 
trust is created by the way in which information derived from' risk- 
benefit analysis is presented. He stated that if it is made clear that 
risk estimates have been made on the basis of certain listed factors 
and that other poorly known or unrelated factors have not been 
included there might be greater public understanding of decisions 
being made on the basis of the estimates. 
Dr. Lowrance 
The DNA issue, according to William W. Lowrance, Department of 
State, fits into the category of those issues where formal risk-benefit 
analysis probably does not have an application. In support of this ob- 
[ Appendix B — 86] 
