"Nuclear Liability after Three Mile Island". Author's Reply
作者回应评论，讨论核保险公司是否基于概率理论设定保费、保费能否反映潜在损失估计，以及三里岛事故作为单一事件的意义。
Dr. Weese's comment raises three important issues: (1) whether nuclear insurers explicitly set premiums and make capacity decisions on the basis of probability theory; (2) whether it is legitimate to infer that premiums reflect estimates of potential loss, however the premiums are set; and (3) whether the Three Mile Island accident, as a single event, is significant. On the first issue, I am in fundamental agreement with Weese. The original article said: With such a short record of experience, nuclear insurers cannot set rates actuarially but instead use [3, p. 452]. Weese's comment goes beyond this to outline reasons why nuclear insurers may never be able to set rates actuarially. It also provides valuable insight into the specific judgment factors that insurers use. On the second issue, we can agree that there are weaknesses in inferring estimates of the nuclear risk from insurers' setting of premiums. The original article spent a page and a half discussing these weaknesses. However, the fact remains that when the engineering reports are in and all is said and done, the insurer has to set a premium that will cover administrative costs plus losses or suffer the consequences. In the instance mentioned in the original article, insurers were charging a premium which would cover administrative costs and then leave $1 of pure premium to every $1 ,700 of potential claims. As was pointed out, insurers did not say that the probability of loss was 1/1700; instead, they accepted a gamble that they could expect to pay off only if the chance of loss were less than 1/1700. Such information is useful, even if no nuclear insurer ever thought directly about probabilities in such terms. If an insurer's judgment implied a higher probability say, 1/500 or 1/200 and the insurer continued to offer coverage at the stated rates, then that insurer would go into the transaction expecting to lose money. Although occasional departures from profit maximization are to be expected, systematic and pervasive departures are not. 1 On the final issue, the significance of a single event at Three Mile Island, it is useful to distinguish its significance to insurers and its significance for public policy. Weese says that insurers have made no special adjustments in light of the accident, indicating that they do not consider this single event to