Christopher Preble has an interesting article about Donald Rumsfeld.
I think he makes a good point that adding more troops to Iraq could cause more problems. On the other hand, occupying Iraq without enough troops causes other problems.
The thing that Rumsfeld did wrong, in my opinion, was in miscalculating the number of troops that would be required for the dramatic transformation he desired. I think that at least half a million troops, and probably many more, would be required in order to create a successful non-totalitarian sort of Iraqi government. On the other hand, so many troops would provide a lot of targets and might mean that our casualties would be twice as high as they are now.
My issue has never been that Rumsfeld should have sent more troops as much as it has been that realizing what would be required might have meant that he would have rethought attacking Iraq, or at least set more realistic goals which could have been achieved with the number of troops that we had on the ground.
Whether he was thinking wishfully or whether he deliberately lowballed the estimate because he wanted the war and thought he could deal with the fallout later, I don't know.
Of course, neocon morons (and I'm being kind here, I'm assuming stupidity rather than malevolence) like Newt Gingrich still insist that Rumsfeld's plans were fine, and that if we'd turned over Iraq to "Iraqis" (which is a code word for Ahmad Chalabi) we'd have a greatly reduced presence by now.
My feeling is that no amount of evidence could convince the neocons (with a few possible exceptions) that democratizing Iraq is impossible. They can always find some excuse; something done wrong that if done right would have allowed democratization to proceed quickly.
So the goal then, isn't to convince them, it is to convince others to stop listening to them and to convince them to instead marginalize them.
That is all.