Something on risk from the Economist that I missed. It's about the sniper, but it has relevance for understanding attitudes towards risk generally:

Experts seem to agree that Americans find it harder than most people to evaluate risks accurately....People generally exaggerate spectacular but low-probability risks, such as murder or natural disaster, just as they underestimate more common risks, such as accidents in the home....[P]eople do not know how to evaluate something they have never seen before....[P]eople tend to overestimate unknown risks....parents dramatically overestimate any uncommon threat to their children's lives (such as the risk of kidnapping by a stranger)....People worry less about voluntary risks [than �involuntary� risks, that is one you run willingly for a benefit (such as driving too fast to get somewhere)]....Worst of all, the risk is hard to mitigate. You cannot easily lessen it by changing behaviour�like wearing a seat belt. The only way to remove yourself from the sniper's mercy is not to go out at all. But that brings out another unusual side to this case. On the roads, the more people drive carefully, the lower your own risk. Here, if more people stay home, your risk actually rises marginally.