numbers of deaths occurring much later or in future populations? People may place quite different values on these different kinds of adverse effects, and these values may change with time.
Another problem is that the same risk is not equally acceptable under all circumstances. People accept familiar risks, such as those associated with the automobile, cigaret smoking, and industrial accidents, yet reject much smaller risks associated with new technologies. The voluntariness of risk is also important; those who voluntarily accept high risks, such as those of motorcycles or contact sports, may strongly object to the minute involuntary risk of a nearby chemical factory. Finally, the risks of an activity that provides a unique benefit—as does, for example, the automobile—are more acceptable than the risks of a technology to which there appear to be alternatives.
A general problem that arises in connection with almost all risk assessments is the significance of dose-effect relationships at very low doses, for both radiation and chemicals. The conservative assumption of a linear dose-effect relation down to zero dose leads to very large estimates of incremental threats to large populations, but such extrapolations are very uncertain. They are likely to be overestimates, but the extent of the overestimate is unknown.
One way around the problem of low-level radiation is to compare the radiation dose with that from natural background radiation. Although the effect of neither is known, one can say that a radiation dose of, say, 1 percent of the background will have an effect, if any, that is a tiny fraction of the effect of a radiation dose that the human species has experienced throughout its history. Unfortunately, no such comparison is possible with most chemical hazards.
In this study, comparison of energy-related risks to nonenergy risks was generally avoided, because it was believed to have little pertinence to energy policy decisions.*† The first two of the above-listed three approaches to risk comparison were followed, with emphasis whenever feasible on the comparison of similar types of risks from different energy technologies and strategies.
Accidents are the most accurately assessed of energy-related risks. In this regard, coal is the most dangerous of major energy sources: About 10 times as many accidental deaths occur in the coal energy cycle, from mine to power plant, as in the production of an equivalent amount of power