This data was taken from this webpage, maintained by the the American-Israeli Cooperative Enterprise. The page's objective is to show that US citizens are victims of Middle Eastern terrorism as well, so they don't distinguish between Americans being specifically targetted, and Americans who are killed in attacks on Israeli and other targets. I wasn't too concerned with the distinction either, since what I was curious about was this statement in a post by Doc Rampage:
Cliff pointed out in a comment that the expected deaths from terrorists each year in the US is about 300. This is roughly the same as the number of people killed by lightening. Cliff suggests (I infer from his irony) that we are spending too much time, money, and energy in combating what is statistically a minor problem. he problem is... Well, the problem is that his logic makes a weird sort of sense. Like Cliff, I'm an engineer and I detest inefficiency. Especially inefficiency for emotional, anti-rational reasons.
Being the experimentalist that I am, I'm always leery of snapshot statistics, and I'm more interested in trends. For something like terrorism, where attacks come at irregular intervals and the number of deaths per attack depends on a number of factors, I prefer to smooth out the data, in this case using a rolling average. I'll have to think about the best ways to analyze the data: large attacks cause spikes, yet you can't ignore the large attacks. One thing I noticed, though, is that throughout the nineties, there was a rise both in the number and effectiveness of attacks. This follows a drop off at the end of the eighties.
New Post: What was originally an update to this post, a plot of attacks rather than deaths, has been moved to a new post here.