The Brookings Index, on page 4, charts casualties. If you can see a "trend" in either direction, you're more perceptive than I am.

Guess what. Casualties--military and civilian--are going to go up, possibly drastically, before the coming elections. I fully expect a Ramadan offensive in September. The military expects this, too, and will be ramping up for it. This doesn't make me a genius, or indicate the "trend" of the war. It just involves looking at the forest instead of examining Donald's single tree.

How the coming elections come off is going to be quite important, including how their constitution turns out. In the long run, that will be more important to Iraq than spikes in the casualty count.