The earthquake-triggered disaster that has engulfed Japan is a textbook example of the public policy dilemmas posed by catastrophic risk (on which see my 2004 book Catastrophe: Risk and Response). A catastrophic risk, in the policy-relevant sense, is a very low (or unknown but believed to be low) probability of a very large loss. If the probability of loss is high, strenuous efforts will be made to avert it or mitigate its consequences. But if the probability is believed to be very low, the proper course to take will be difficult, both as a matter of sound policy and as a political matter (to which I return in the last paragraph of this comment), to determine and implement. The relevant cost is the catastrophic loss if it occurs discounted (multiplied) by the probability of its occurring. If that probability is believed to be very low, the expected cost may be reckoned to be low even if, should the loss occur, it would be catastrophic. And if the expected cost is low but the cost of prevention is high, then doing nothing to prevent the risk from materializing may be the optimal course of (in)action. Furthermore, even if the expected cost of catastrophe is high, the scale of the catastrophe—the loss that it inflicts—may be reducible to moderate proportions by responsive measures. For example, while one way to deal with the risk of an epidemic is to vaccinate everyone, another is to quarantine the persons first infected, so that the epidemic is limited.
With these thoughts in mind, one can attempt a preliminary analysis of the Japanese response to the risk of the kind of disaster that occurred. Earthquakes cannot be prevented or predicted, though they can be detected as soon as they occur and if they occur in the floor of an ocean, and as a result trigger a tsunami (a tidal wave), people in the path of the tsunami can receive a warning of minutes or hours, depending on the distance between the earthquake and populated areas. Unlike the countries that border the Indian Ocean, whose populations suffered more than 200,000 deaths from a tsunami in 2004, Japan appears to have had a good early-warning system, as a result of which the number of deaths from the recent tsunami is in the tens of thousands rather than the hundreds of thousands.
The catastrophic effects of a major tsunami cannot be prevented by building seawalls; nor, in all likelihood, could the vulnerability of nuclear reactors to catastrophic damage from an earthquake or tsunami be reduced by moving the reactors inland, because they require access to large bodies of water for their normal operation, and it appears that the only large bodies of water available to Japan are the seas surrounding it. But catastrophic damage to nuclear reactions could be reduced by building stronger containment vessels for the reactors.
The loss of life and the radiation hazards from damaged reactors could be minimized by creating a rapid response capability that could be mobilized as soon as an earthquake or tsunami struck. Japan failed to create such a capability.
At considerable but not astronomical cost, Japan could have reduced the losses caused by the tsunami substantially by strengthening the containment vessels for its reactors and by creating a rapid response capability. I do not know why it did not do so, considering that it is a wealthy, technologically highly sophisticated, and risk-averse society. One possibility is that the probability of an earthquake within several hundred miles of Japan that would be as violent as the earthquake that triggered the tsunami turned out to be (9.0 on the Richter scale) was thought too slight to justify the cost of the protective measures that would have been necessary to minimize the consequences of such an earthquake. Suppose the expected cost of a 9.0 earthquake was $100 billion, but there was believed to be only a 1 percent annual probability that it would occur; then preventive measures that cost $1 billion or more per year would not be cost justified. The 1 percent figure is arbitrary; but considering that Japan had a 9.0 earthquake in 2011 that triggered a tsunami comparable to the recent one, that the Indian Ocean tsunami occurred only seven years ago, that on average a 9.0—or greater—earthquake occurs somewhere in the world every 20 years, and that Japan is in a region that is highly prone to earthquakes and tsunamis, 1 percent seems as good a guess as to its probability as any.
Strengthening the ability of its nuclear reactors to withstand a tsunami triggered by a 9.0 earthquake (or greater? 9.0 is not the ceiling—the Indian Ocean earthquake was 9.1 on the Richter scale) would be very expensive—Japan has more than 50 nuclear power plants, all on its coasts. But how much would it have cost to have built them slightly inland, connected to the seas by canals? And the creation of a rapid-reaction capability would probably not be very expensive, although I have not seen cost figures.
It would not be surprising, however, if as seems to be the case Japan failed to take cost-justified measures to minimize the damage from a 9.0 or greater earthquake. Politicians have limited time horizons. If the annual probability of some catastrophe is 1 percent, and a politician’s horizon is 5 years, he will be reluctant to support significant expenditures to reduce the likelihood or magnitude of the catastrophe, because to do so would involve supporting either higher taxes or a reallocation of government expenditures from services that provide immediate benefits to constituents. In principle, it is true, politicians would take a long view if their constituents did out of concern for their children and grandchildren. But considering how the elderly cling to their social benefits, paid for by the young including their own young, I doubt the strength of that factor, although I do not know enough about Japanese politics to venture a guess on whether politicians’ truncated policy horizons was indeed a factor in Japan’s surprising lack of preparations for responding promptly and effectively to the kind of disaster that has occurred.
A final factor may be that Japan, unlike the United States, does not have an independent nuclear regulatory agency.