As anyone who regularly follows the news has noticed, Israel's ability to identify specific Palestinian fighters and kill them without killing surrounding civilians is continually on the rise. Now there's a report with some interesting numbers. It turns out that appearances are correct, and the ratio of dead targets to dead bystanders has recently reached 30:1, which still leaves lots of room for improvement. The optimal number to strive for would be lots:0. And the perfect numbers would be no-killings at all, for lack of need.
How should we read these numbers?
Killing people is never something to be joyous about. However, killing armed men who are members of organizations committed to destroying Israel, and who wish to kill Israeli citizens as the primary way to achieve their goal, seems pretty reasonable to me. Especially as the trend over the past five years has consistently shown that violence against violent Palestinians indeed reduces the overall violence. So that's alright.
Also understandable is that the longer an army (or any organization) has to work on a problem, the better it gets. The fact that Israel's ability to kill far fewer bystanders in 2007 than in 2002 is, unfortunately, natural. This may sound callous, but is simply the way of the world: practice makes good, and no army in history ever figured out how to do its job perfectly, even less so immediately. Especially as the tasks most often appear unannounced, and you have to devise a response while people are getting killed.
A careful reading, however, also tells a regrettable story: that it was a decision from the top that made a difference. So long as the commander of the Airforce was Dan Halutz, not enough care was taken not to kill Palestinian bystanders. (Some care was taken, but not enough). How do we know this? Because once Halutz was replaced by Eliezer Shkedi, and he made an issue of the matter, he also got results.