The Volokh Conspiracy - Careful With Those Comparisons:

I was quite struck by this comment on the Jonathan Rauch on the Bush Presidency thread:

Conflating existential threats like 1930's Germany and Japan with today's terrorists is simply not serious. As others have noted, in this time of "great danger", you're 66 times more likely to die by drowning while swimming or in the tub than of terrorism (based on National Academy of Sciences' Institute of Medicine figures).

That isn't to say that it isn't a problem, but hyping fear for political payout is (and I've been saying this since 9/11, and I live in NYC) not only wrong, but in fact doing the terrorists' work for them.

I do think that's important to keep a sense of perspective about various threats — but this analysis (which I note because I'd heard similar assertions in the past, though they were rarely quite as striking as this one) doesn't really work.

To begin with, note that Germany and Japan weren't existential threats to the United States in the 1930's (I assumed from the context that the commenter was speaking of their threats to our country, not to, say, Russia or China). I know of no evidence that in the 1930s Germany and Japan had serious designs on American independence, or that any reasonable Americans thought that they did. Until the nuclear bomb was proven feasible, there was no way that they could have invaded the U.S., or otherwise jeopardized our existence or our nationhood. Of course, people could have feared that decades down the line a Germany and Japan that had conquered Europe and Asia could have turned their attention to America and seriously threatened our national existence; but if something at that level of long-term speculation counts as an "existential threat," then a great deal qualifies as "existential threat" — one can certainly come up with similar 50-years-down-the-road speculations about militant Islam.

As to "66 times," that's the first time I'd heard that statistic, and some quick searches uncovered no evidence for it. But I do know how to use the CDC's invaluable WISQARS system, and it reports to me that from 2001 to 2003 there were under 3500 accidental drownings (of all sorts, which I suspect includes boating as well as swimming and bathtubs) per year in the U.S. Even if one looks at the last 6 years, one still gets only about 7 times fewer (rather than 66 times fewer) people killed in the U.S. through Islamofascist terrorism than in drownings.

But more fundamentally, it makes no sense for anyone, whether the NAS Institute of Medicine or anyone else, to say that you're "X times more likely to die by drowning while swimming or in the tub than of terrorism." No-one can know what the actual likelihood of dying in an Islamofascist attack against the U.S. might be. It could be that 2001-06 is representative of the future, and we'll have four-digit death tolls once every six years or so, and quiet otherwise. It could be that the future will be considerably more peaceful, whether because we've adequately disrupted al-Qaeda, because the lethality of the World Trade Center attacks was a fluke, or for whatever other reason. Or it could be that the future will be considerably more lethal, with smallpox terrorism, nuclear terrorism, or who knows what else. An "X times more likely to die by" statistic just can't be sensibly used to compare an unknown but not implausible risk of a death toll in the five to seven digits (or more? who knows?) with a nearly constant stream of under 3500 deaths per year.

So by all means resist what you see as excessive or unjustified responses to the risk of terrorism. But don't pooh-pooh the risk using senseless comparisons.

	[Eugene Volokh, September 11, 2006 at 1:58am] Trackbacks Careful With Those Comparisons: I was quite struck by this comment on the Jonathan Rauch on the Bush Presidency thread: Conflating existential threats like 1930's Germany and Japan with today's terrorists is simply not serious. As others have noted, in this time of "great danger", you're 66 times more likely to die by drowning while swimming or in the tub than of terrorism (based on National Academy of Sciences' Institute of Medicine figures). That isn't to say that it isn't a problem, but hyping fear for political payout is (and I've been saying this since 9/11, and I live in NYC) not only wrong, but in fact doing the terrorists' work for them. I do think that's important to keep a sense of perspective about various threats — but this analysis (which I note because I'd heard similar assertions in the past, though they were rarely quite as striking as this one) doesn't really work. To begin with, note that Germany and Japan weren't existential threats to the United States in the 1930's (I assumed from the context that the commenter was speaking of their threats to our country, not to, say, Russia or China). I know of no evidence that in the 1930s Germany and Japan had serious designs on American independence, or that any reasonable Americans thought that they did. Until the nuclear bomb was proven feasible, there was no way that they could have invaded the U.S., or otherwise jeopardized our existence or our nationhood. Of course, people could have feared that decades down the line a Germany and Japan that had conquered Europe and Asia could have turned their attention to America and seriously threatened our national existence; but if something at that level of long-term speculation counts as an "existential threat," then a great deal qualifies as "existential threat" — one can certainly come up with similar 50-years-down-the-road speculations about militant Islam. As to "66 times," that's the first time I'd heard that statistic, and some quick searches uncovered no evidence for it. But I do know how to use the CDC's invaluable WISQARS system, and it reports to me that from 2001 to 2003 there were under 3500 accidental drownings (of all sorts, which I suspect includes boating as well as swimming and bathtubs) per year in the U.S. Even if one looks at the last 6 years, one still gets only about 7 times fewer (rather than 66 times fewer) people killed in the U.S. through Islamofascist terrorism than in drownings. But more fundamentally, it makes no sense for anyone, whether the NAS Institute of Medicine or anyone else, to say that you're "X times more likely to die by drowning while swimming or in the tub than of terrorism." No-one can know what the actual likelihood of dying in an Islamofascist attack against the U.S. might be. It could be that 2001-06 is representative of the future, and we'll have four-digit death tolls once every six years or so, and quiet otherwise. It could be that the future will be considerably more peaceful, whether because we've adequately disrupted al-Qaeda, because the lethality of the World Trade Center attacks was a fluke, or for whatever other reason. Or it could be that the future will be considerably more lethal, with smallpox terrorism, nuclear terrorism, or who knows what else. An "X times more likely to die by" statistic just can't be sensibly used to compare an unknown but not implausible risk of a death toll in the five to seven digits (or more? who knows?) with a nearly constant stream of under 3500 deaths per year. So by all means resist what you see as excessive or unjustified responses to the risk of terrorism. But don't pooh-pooh the risk using senseless comparisons.