U.S. counterterrorism officials, in particular, seem to regard biological, chemical, and nuclear attacks as a main priority of terrorist groups. Certainly, judging from its funding priorities, the U.S. Department of Homeland Security sees such attacks as a greater threat than the destruction of a city by hurricanes. But this thinking neglects the sheer difficulty of successfully pulling off such an attack.

The only two significant biological and chemical attacks by nonstate actors were, in the context of modern terrorism, feeble. These were done by the followers of the Indian religious leader Bhagwan Shree Rajneesh, who spread salmonella in salad bars in an attempt to influence voter turnout in Oregon in 1984, and members of the Aum Shinrikyo group, who released the nerve gas sarin in Tokyo subways in 1995. The Rajneesh group sickened 750 people; the Japanese cult killed 12 and injured about 500. There is no evidence that nonstate groups currently have a biological, chemical, or nuclear weapons capability.

In any case, the terrorists have every reason to fear that another far-reaching and heinous attack on the United States or Europe might be to their disadvantage, for it would merely strengthen Western resolve to fight back. Much more likely, I believe, are smaller, purely demonstrative attacks in the West—say, substituting a “dirty” bomb for a real nuclear one, with the aim of terrorizing rather than killing many people. We should expect the really big attacks to come not in Washington, London, or Paris, but in Riyadh, Cairo, or Amman, for al-Qaeda has always been focused mainly on effecting regime change in the heartland of the Arab world.

From these considerations I conclude that the chances are extremely low that Middle Eastern terrorists will stage an attack on U.S. soil that kills at least 1000 people or harms the economy as much as the 2003 blackout did. The chances are only marginally higher that such an outrage will happen in Europe. We should therefore concentrate on defending ourselves against nature’s wrath and against industrial accidents, which we know will happen. Some of these defenses will also help mitigate the damages from terrorist attacks.

Officials of the Department of Homeland Security have argued that some of its measures do perform the double duty of preventing natural and terrorist disasters. Among them are the hiring, training, and equipping of ambulance drivers, paramedics, firefighters, and rescue workers. However, most of the new expenditures are ostensibly useful only against terrorism, as in securing borders, ports, and transportation hubs; putting citizens, organizations, and telecommunications under surveillance; deploying radiation and biohazard sensors; and coordinating feuding intelligence agencies.

Not only does this spending do nothing to guard us against the natural and industrial disasters that are certainly coming our way, it has to some extent been carved out of the budget for such safeguarding. And quite apart from the funding tug-of-war is the change in emphasis, some of it plainly perverse. The effort to check vehicles for explosives delayed the arrival of aid in Louisiana following Hurricane Katrina. Doctors and other workers who came at their own expense were often sidelined until they could prove that all their baggage had been checked and their paperwork completed. Some rescuers were shunted as far afield as Texas to undergo classes on avoiding sexual harassment.

The response to Katrina falls below even that predicted by my theory of normal accidents. Clearly, the terrorists have no monopoly on incompetence: the government, at both federal and local levels, should have handled things better. Yet the problem was so big that even with the best imaginable leader­ship, the outcome would not have been much better. We should not allow masses of people to live in places that are fundamentally unsafe.

Our best bet is not to protect targets but to reduce their size. We cannot move people in California away from geologic faults, but we could have avoided filling in San Francisco Bay with tall buildings built, basically, on mud. As there is no pressing need to have so many people living in the Bay area or New Orleans, public policy ought to encourage them to live elsewhere. Yet actual governmental incentives go quite the other way. Federal flood insurance now covers millions of people, some of whom have built homes in places that private insurers will not touch. In one egregious case, the government reportedly paid some US $800 000 to rebuild, again and again, a Texas house appraised at less than $120 000.

Other vulnerabilities that cry out for systematic reduction abound. Deregulation of the electric power industry has reduced the incentive to invest in redundancy, and grids in the United States and Europe are now clunking along with equipment that is inadequate, outmoded, or both. All this makes us extremely vulnerable. We could change the financial incentives to induce the power industry to plow some of its profits back into the grid. I would start plowing it in around the electrical interconnect between California and Oregon, which is very vulnerable. If a study showed that disaster would ensue if someone took out two or three transmission stations, then it would be simple enough to build three more. Redundancy works, at least to a point. But we have to be willing to pay for it.

Other concentrated targets include milk pooled in immense tankers, any one of which could easily be contaminated with a noxious agent or pathogen. If this were to happen, thousands of children in entire sections of the country could be made ill. Again, the problem could be solved by breaking things up a bit, perhaps starting with the breakup of dairy distributorships into smaller entities.

Environmental groups might be mobilized. They have, on occasion, advocated reducing the size of chemical storage depots and removing them from centers of population. But our landscape is strewn with many other potential weapons of mass destruction that terrorists can exploit in situ, among them levees, bioisolation laboratories, and nuclear reactors. The scope of environmental groups should be expanded, and they should recognize that the threat comes from nature and terrorists, and not just accidents.

The United States could try to do as well as some European countries. Take the response of the Netherlands to the threat of flooding, which reached a three-century crescendo in 1953, when freak storms breached the dikes, flooded huge swaths of territory, killing thousands and forcing hundreds of thousands from their homes. Since then, the country has shored up dikes, installed elaborate moveable gates, experimented with floating homes, and taken property from people and turned it into wetlands to absorb the annual floods of the Meuse river. Most of all, it has reduced the size of the target by preventing large pockets of population from clustering in vulnerable areas.

Minimizing the targets can limit but not eliminate the depredations of the weather, of industrial accidents, and of terrorists. Terrorism particularly will retain much of its force because, as a deliberate human act, it cannot fail to touch its victims in ways no innocent accident can manage. Here, too, the experience of European countries is instructive; they accept that there will be terrorist acts and do not respond with dubious bureaucratic initiatives. The United Kingdom responded to terrorism after 9/11 not by shoving 22 agencies together to create a dysfunctional, centralized behemoth, but by improving coordination among strengthened independent agencies.

Organizations in our critical infrastructure will inevitably fail; the greater the concentration, the bigger the consequences. But the public is unaware of our basic vulnerabilities in the chemical industry, electric power industry (including nuclear plants), the 90 percent concentration of Microsoft Windows operating systems for computers, and the predicted increase in the concentration of Internet access as the number of ISPs declines. In the 1970s, before the deregulation of such industries began, we had more aggressive antitrust activity; we now have additional reasons to reestablish and extend those policies. The United States’ weakened regulatory agencies, such as the Environmental Protection Agency, the Occupational Safety and Health Administration, and the Securities and Exchange Commission, could be given back their strength. Perhaps we should even disassemble the Department of Homeland Security and establish a Department of Homeland Vulnerability to publicize our juiciest targets—terrorists already know them all—and call for legislation that would make them less attractive.

Human nature can better withstand a dozen small disasters than a single great one, even if the casualty total is the same. Protecting our big targets is next to impossible; we should instead downsize them to make them less consequential and easier to protect.