Terror Statistics

Mistakes, Damned Mistakes and Statistics 

STEVE MIRSKY / Scientific American 1sep04 

In April the U.S. government released its yearly report called "Patterns of Global Terrorism." This edition showed a welcome decrease: the number of people wounded in terrorist incidents in 2003 fell to 1,593 from 2,013 the year before. The decrease in injuries, as well as in deaths and in terrorist incidents, prompted Deputy Secretary of State Richard Armitage to say, "You will find in these pages clear evidence that we are prevailing in the fight."

Then, in June, the State Department updated the original document's incorrect statistics and revealed that terror-related injuries in 2003 in fact totaled 3,646. This number, according to mathematicians, is higher than 2,013. The updated report also revealed more deaths and terrorist incidents in 2003 than had the first document. The new data raise a question: If the interpretation of the original report led the deputy secretary of state to the conclusion that "we are prevailing in the fight," has the corrected report compelled him to announce that we are losing the fight?

Logical consistency would force that conclusion. I looked for any news stories in which he or his colleagues made such an announcement but found none. In the interests of full disclosure, I didn't look too hard, because I figured the odds of finding such a statement were about equivalent to the chances of getting injured in a terrorist attack, which are still exceedingly low, although the idea of it is, well, you know, terrifying.

Now, I'm sure that there are pundits out there who can convincingly make the case that the increase in incidents and in-juries in 2003 also represents "clear evidence that we are prevailing in the fight." Because terrorists are getting increasingly desperate, yada yada, insert tortured reasoning here. Anyway, here are some other examples of how one can interpret bad facts to be good news.

Scenario: A major-league baseball team has a team batting average of .260 and hires a new batting coach.

Result: The team batting average plummets to .217.

Conclusion: Well done—you are prevailing in the fight to hit major-league pitching. Your hitters are now so feared that teams use their best pitchers against you, leading to a drop in your collective batting average. (Actually, that makes some sense, but the new coach still gets fired.)

 

Scenario: The Iraqi city of Basra suffers from severe gasoline shortages.

Result: There are two days of rioting. Conclusion: Excellent—you are prevailing in the fight to strengthen the local economy. The raging demand for gas shows that more people have the financial wherewithal to want to drive places. (Actually, there were two days of rioting over gas shortages in Basra in August 2003, and officials did say it was a sign of an improving economy.)

 

Scenario: You pledge a fraternity.

Result: You perform depraved acts of self-humiliation and are enthusiastically accepted into the fold.

Conclusion: Fantastic—you are prevailing in the fight to become a member of an important group. Hey, it must be an important group, because how else can you explain the gusto with which you roll around in cow pies while singing "To-morrow" from the Broadway musical Annie with only a non-Equity contract? (Actually, the belief that a frat must be important because otherwise how could you explain the ridiculous things you're doing to get into one is pretty much the logic be-hind the willingness to be hazed.)

In reality, simple statistics on casual-ties and incidents are not conclusive. For example, if devastating terrorist attacks were secretly averted, perhaps we really would be prevailing in the fight despite a given year's increased casualty count. It's complicated, but Americans can handle a bit of complexity. Just don't tell us that chocolate ice cream is vanilla, especially when it's not even chocolate ice cream but only something a frat boy rolled in.

p.121