There are contributing explanations. Most obvious, it is easier to kill an Iraqi than an American. Americans get about in relatively guarded circumstances. To kill an American soldier, it is generally necessary to ambush him in his vehicle, whence the wide use of the roadside bomb. But since the number of these attacks is declining, does this tell us that American soldiers travel about less frequently than they did, say, a year ago, when casualties mostly ranged between 50 and 100 per month? Or does it tell us that insulations against roadside bombs are more sophisticated? Or a bit of each?
One asks then: Is the furious resolve of the insurgents altering in focus? Has the enemy reckoned that the problem in hand is not Americans, who will be gone, roughly speaking, tomorrow, but Iraqis whose ethnic identities will remain the same when the grandchildren of both parties will be eyeing each other? If this is the case, can we assume that the Iraqi rebels are biding their time, in respect of their opposition to the Americans? From a sense that the Americans are transitory and, in fact, are not doing all that much harm? Indeed, the invading U.S. army continues to invest in infrastructure in territory that will certainly not be American a few years from now.
The insurgents can reasonably assume that there is nothing they can do to drive the Americans away. Three years have gone by, more than 2,300 Americans have been killed, the rate of these killings reduces but is not concomitant with any reduction in U.S. strength. It is brought on by (a) the reduction in U.S. exposure, and (b) insurgent concentration on non-U.S. targets.
But ask then: Is this reduced exposure a part of the U.S. battle plan? We have not in recent months seen any hard U.S. assaults on hard Iraqi targets, in the class of Fallujah in the fall of 2004. Can we assume that such hard enemy nests aren't there, holding out? Or rather that the U.S. army command is less bent on smoking them out?
If there are (one speculates) 15 areas of Iraq in which the insurgents are embedded with special defensive ingenuity, the commanding general can elect to dispatch bombs and artillery, always with some care for collateral damage done to innocent civilians. But that approach, a platonic alternative to sending in a battalion with instructions to root out the offenders, means a diminished exposure of American soldiers to high-cost engagements.
To reason that this is happening is deductive: fewer casualties, fewer engagements. However, fewer engagements should presume an enemy diminished in size and potency. But to say that runs us into the corresponding figure, of 1,500 Iraqi civilian deaths. Somebody is killing those people, and the whole idea of the U.S. enterprise was to shield the Iraqi population not only from the depredations of Saddam Hussein, but also from successor killers. Manifestly this has not happened, if the killing proceeds at so high a rate.
I have myself concluded that our Iraqi mission has failed. Missions have to be judged successes or failures with some reference to a time scale. If that scale is stretched forever, it is not authentically tested. If the mission is to liberate the Prisoner of Zenda and 10 years later he is still in jail, the mission can reasonably be classified as having failed, never mind that in the 15th year he is actually rescued. Given our mission's failure in Iraq, the job in hand becomes to retreat with care, certainly with more care than we exercised in our retreat from Vietnam.
But one would expect the military to pay greater attention than if hellbent on the mission's accomplishment to such factors as risk to U.S. personnel. The welcome lightening of the casualty figures can be seen as the military voting with their feet to begin withdrawal from an enterprise that has proved costly beyond the successes achieved.