Given today's (now yesterday's, technically) events in London, I feel its appropriate to give an except from Bruce Schneier's latest book, Beyond Fear:
There is a real risk of terrorism, but the situation is not nearly as dire as most people thought in the months directly after 9/11. International terrorists are trying to attack the U.S. and Western Europe (and their interest around the world), but they're rare and they've become rarer since governments started counterattacking and arresting terrorists before they strike. It's impossible to quanitfy the risk in any meaningful sense, but we can look back at the previous decades and see the trends. Internationl terrorism happens, but it's much less common than conventional crime.
None of this discussion is meant to belittle or deny the risks - it's just to put them into perspective. In 2001, 3,029 people died in the U.S. from terrorism (the 9/11 attacks). During that same year, 156,005 people died from lung cancer, 71,252 from diabetes, 41,967 from motor vehicle accidents, and 3,433 from malnutrition. Consider what we're willing to spend per year to cure diabetes or increase automobile safety, and compare that with the $34 billion we're spending to combat terrorism. Te response to the terrorism threat has not been commensurate with the risk.
The problem lies in the fact that the threat - the potential damage -is enormous. Security is all about trade-offs, but when the stakes are considered infinitely high, the whole equation gets thrown out of kilter. In the frightened aftermath of 9/11, people said things like: "The budget for homeland security should be infinite. The trade-offs we need to make should be extreme. There's no room for failure." It was easy to succumb to hysteria and scare-mongering and to overreact to the terrorist threat. In fact, one of the primary goals of terrorism is to create irrational terror far in excess of the actual risks. But this kind of talk is meaningless. When a country allocates an infinite budget to homeland security, the terrorists really have won.
... This is a precarious position to take politically, which is why I believe most politicians have steered clear of it. It's safe for a political leader to make direct predictions about the future and recommend an extreme course of action. If another terrorist attack happens, then she can say that the event proved her right. And if nothing happens, she can claim that her security program was a success. (And that it keeps away the vicious purple dragons, too.) A politician is on shaking ground when he says, "Don't worry; it's not that bad." He looks ineffectual compared to his colleagues who are trying to do something to make us safer (even if the "something" doesn't really make us safer); worse, he looks as if he doesn't care about his constituents. And if another attack happens, he looks even worse. The public's difficulty in assessing risks plays into this. Many people have been frightened into believing that terrorism is a far greater risks than it is. All sorts of incidents are immediately assumed to be terrorism, even though investigations prove other causes. And when the next terrorist attack occurs on U.S. soil, people, politicians, and the press will all exaggerate the risks of terrorism even more.