Note: this is a condensed and re-written version of a more detailed argument, entitled "Nuclear Power," forthcoming in the Journal of Business Ethics.

Introduction:

The nuclear power industry has never been free from government interference. The unintended consequence of subsidization and regulation has been a distortion in the price of nuclear power and an increase in the risk of nuclear catastrophe. Subsidies, via the Price-Anderson Act, and regulations via the Nuclear Regulatory Commission (NRC), should be removed. In fact, all governmental distortions in the energy market should be removed. Only if the nuclear industry is free from government intrusion can we ever answer the question: should there even be a nuclear industry?

The Price-Anderson Act:

Out of fear that the Soviet Union was outpacing the US in power generation, Congress passed a temporary palliative: the Price-Anderson Act of 1957. The major issue was liability insurance for catastrophic accidents: private insurers would not provide adequate coverage, so the government provided subsidized insurance and immunity from excessive damages.

Like most government subsidies, it has proven hard to shake. In 2005, the Price-Anderson Act was extended to 2025. Under the new version of the Act, power companies are required to purchase up to $300 million per plant in private insurance. Any damages in excess of $300 million are split evenly between the country’s 103 nuclear plants, each of whom are required to contribute no more than $95.8 million. As there are 103 nuclear plants currently in operation, this amounts to roughly $9.5 billion in additional insurance.

More importantly, the Price-Anderson Act insulates the power companies from paying any damages over this amount. This is an invitation to what economists call "moral hazard" — a situation where an insured individual will undertake more risky behavior, simply because he is insured.

If the expected cost of a nuclear accident is less than $300 million, a company will spend an additional dollar to make its plants safer if it will save at least a dollar’s worth of savings on insurance premia. However, the firm’s cost v. benefit decision shifts toward risky behavior once expected losses rise to between $300 and $395.8 million. If the benefits of engaging in more risk (in terms of cost savings) are borne by the firm, but the costs of increased liability are split 103 ways, then the firm’s cost-benefit analysis will favor more risk. Once a firm expects to inflict damages in excess of $395.8 million it incurs no additional insurance savings, only the costs of instituting the safety measures.

A severe nuclear accident could result in damages in the range of $560 billion — that’s billion, not million — in 2000 dollars (Riccio, 2001). This may be an overly pessimistic estimate, but let us use it for illustrative purposes. If damages were $560 billion, the firm’s liability would be $395.8 million. If damages could be reduced to $400 million, through a minimally costly expenditure of even ten dollars, the firm’s liability would still be $395.8 million. There is no financial incentive for the nuclear firm to invest even this paltry sum in minimizing risk since under the Price-Anderson Act the firm’s liability remains unchanged.

Let me be clear about this: all insurance creates moral hazard. But subsidized insurance is state-augmented insurance. It augments the moral hazard, creating more inefficiencies than would exist under the free market. However, instead of repealing the risk-inducing Price-Anderson Act, the government turned to a labyrinth of safety regulations.

NRC Regulation:

Few industries must navigate more labyrinthine regulations than the nuclear industry. Regulated almost from their inception, most people take regulation for granted.

Its real costs, however, are unseen and therefore unnoticed; hidden costs abound in economics. Current regulations are inefficient on several grounds: (1) they stifle innovation, (2) are needlessly costly, and (3) do not mitigate risk.

Regulations governing the components of a nuclear reactor are necessarily retrospective. They take for granted a given nuclear technology, the existing technology, and mandate that all future reactors follow the guidelines appropriate for the old reactors. If a certain kind of valve was a good idea for past reactors, wouldn’t it be prudent to require all future reactors to contain that valve? This is short sighted. If the valve mitigates safety risks (and thereby decreases insurance costs), then a profit-seeking firm would voluntarily use it: regulation would not be necessary. Moreover, if a nuclear company could devise a safer valve, current regulations would require the older, less safe, valve. A firm would be less likely, then, to research and develop any safer valves. It would have to lobby the regulators to convince them of the safety of the new valve.

The regulations are needlessly complex and costly. More disturbing, though, the regulations are not concerned with safety. According to the National Energy Institute, "in 1997, for example, 94 percent of violations [of NRC guidelines] had little or no safety significance" (NEI, 2004). Only in 1999 did the NRC decide to switch regulatory focus to a more risk-informed basis. Balancing the costs and benefits, risks and rewards, is exactly what the market does best. That is why teenagers have to pay higher car insurance premia; teenagers run a quantifiably greater risk of being in a car accident. Similarly, a properly priced nuclear insurance policy would require estimating the likelihood of various nuclear catastrophes under a variety of different design specifications.

At least the NRC has, as of 1999, begun to address the issue of risk in its regulatory rule-making. But it has done a poor job at this, too. The most serious flaw is that, in focusing solely on the probabilities of catastrophe, the NRC has forgotten half the risk equation. Risk, properly assessed, does not merely focus on probabilities, but on the expected costs (i.e., the cost of the catastrophe multiplied by the probability of the catastrophe). A 0.005% chance of catastrophic meltdown represents more risk if the plant is in Manhattan, than if it is in Antarctica. This seemingly common sense has been lost on the NRC (Lochbaum, 2000). In contrast, insurance companies make these calculations every time they write a policy.

Nuclear waste and pollution:

If we dispose with the regulations governing nuclear waste disposal, wouldn’t pollution become a problem? Again, the answer depends on the degree to which the free market is able to function, and that depends on property rights.

If I don’t take care of my lawn, no one else will. And since my property values are higher whenever I maintain my lawn, I get up every Sunday. I even fertilize. I would never mow or fertilize the neighborhood park, however, even though it is only a block from my house. The reason is that I own my lawn, not the park. If I were to mow it, I’d suffer all the costs of physical exertion, and only get a fraction of the benefit which would be split fairly evenly among all the residents of the neighborhood. The end result is that the park grass gets mowed rather infrequently by the government. This is a common economic phenomenon called the "tragedy of the commons."

Excess pollution is a problem attributable to poor property rights. Certainly the power company would like to pollute. However, land owned by a nuclear power plant would tend to be polluted less intensely than land held in common by the public. The nuclear company would have every right to pollute its own land, just as I have the right not to manicure my own lawn. I could not, however, despoil my neighbor’s land; neither could the power company pollute a neighboring farmer’s. To do so would violate the farmer’s private property rights. This is a transgression of the marketplace — an extra-legal act. This is not a symptom of capitalism. It is a symptom of the lack of capitalism. It is theft. The neighbor would have every right to sue the company for damages. Alternatively, the company might pre-empt the lawsuit by buying the neighbor’s land before polluting, or by paying the farmer "rent" for the radiation. If the neighbor agrees, they are both better off. No regulation is necessary to solve the pollution problem.

Nuclear plants would naturally gravitate toward more remote land because it is cheaper to buy, and cheaper to insure. Insurance for a nuclear plant situated near communities would be too expensive. Simply put, the market zones itself.

Should there even exist a nuclear power industry?

Should there even be nuclear power? How much should there be? How should it be produced? Using what materials? Only through a balanced investigation of costs and benefits, can we decide such matters, and these require markets. Such argumentation was the centerpiece of Ludwig von Mises’ (1969) critique of socialism. How can anyone know how many resources to put into an industry if there are no prices? Options can only be weighted rationally, if they are measured on the same scale. We need a common denominator with which to compare the benefits of additional energy, with the costs of depleted fish stocks, for example. The only such denominator is prices, and these require markets.

In every market transaction, consumers express their preferences via their purchases. Relative scarcities, and individual’s costs and benefits become reflected in prices, and it is these prices that firms use in deciding whether to enter or exit an industry. Through the competitive process, firms discover whether they should be in business, whether they should sell a product, how to produce it, and using what inputs. As Israel Kirzner (1985) put it, "competition is a discovery procedure."

Insurance companies are in the business of pricing risk. Companies that are best able to predict risk gain a competitive advantage in the marketplace. Expected payouts are the largest costs of doing business in the insurance field. If the sum of all these costs is greater than the expected benefits of energy production, then it will not pay any profit-seeker to build nuclear power plants. The insurance companies would charge premia higher than the expected revenues from power generation.

If no insurance companies are willing to bear the risk of a nuclear catastrophe, then to the market’s best judgment the expected costs outweigh the benefits (i.e., the risks are simply too high). The prohibitively high price of nuclear insurance is a market signal for either (a) withdrawal from the industry, or (b) greater research into nuclear safety technology.

Outside of the free-market system, no one can answer the burning question: should there be any nuclear power?

Conclusion:

Let us summarize. The risk of catastrophe was too costly for any private nuclear company to bear by itself, and no private insurance company would bear this burden either. Rather than accepting this market signal that there should be no provision of nuclear power, at least with the technology of the 1940s, the government insisted on nuclear power, and therefore limited the nuclear industry’s liability via the Price-Anderson Act. Since the costs of nuclear business were lowered, this encouraged entry into the nuclear industry, resulting in a lower price for nuclear power than would otherwise have existed. Moreover, the subsidized insurance encouraged risky behavior, increasing the likelihood of a catastrophe. The offending firms would have to pay only a fraction of the costs of damage. Rather than eliminating the subsidy, the government opted instead for a flawed regulatory system that diminishes the incentives for innovation, needlessly increases costs, and still does not properly appreciate the risks involved in nuclear catastrophe. Why did the government renew the Price-Anderson Act? Because the nuclear industry is so risky. But it exists in its risky form because of the Price-Anderson Act! By this circular logic, the government’s inefficiency has been used by government as a rationale for even more governmental inefficiency. This is the road to serfdom.