Iraq, June 2007

I’ve refrained in these posts from saying anything about trends, mainly because the month-to-month variation has been so great. Attributing great significance to a few good months makes you come off as a pollyanna; sounding the alarms after a few bad months makes you seem too pessimistic. But how can you look at the past 12 months or so and not be pessimistic? And how can you fail to notice that a month with fewer than 50 deaths, so common in 2003 and early 2004, has been a rarity for the past three years? And that a month with “only” 50 deaths would would be such a relief now?

Data from