One weakness of news coverage of the Iraq war has been the lack of convincing "big picture" stories. Even when reporters try to do this, they invariably give snapshot views of events in a few provinces, or quote a single set of numbers -- the statistic du jour -- with maybe a single comparison to the numbers at some single previous point.
But in this case, as in many others, a lot of information is out there and just needs to be put together and made accessible. This diary is an attempt to do that.
The Saban Center at the Brookings Institution publishes the "Iraq Index" twice weekly. It is an attempt to collect all information about the war in the public domain. As such, it is dependent on what has been published and what statistics the U.S. government has released. They do not always have much new to add from one issue to the next, and the usual caveats about lies, damn lies, and (government) statistics apply, but they are a valuable source.
To evaluate the impact of troop levels have on insurgent attacks, I superimpose a graph of monthly Iraqi civilian deaths on a graph of coalition troop strength. Data are from the January 14 Index. Civilian deaths are on p. 5; troop levels, p. 25. The January 14 edition no longer is available online, but later editions have nearly identical data. I use the monthly series for total coalition (i.e. non-Iraqi) troop strength. Casualties are bright blue, troop strength dark blue. The casualty series has been linearly transformed so that the maxima and minima of the two series roughly coincide.
The resulting graph
<DIV align=center>
</DIV>
shows several interesting patterns. (Click on it for a larger imaqe).
- The late 2007 "surge" was the third time that coalition troop levels have been increased to around 180,000 for a short time. Earlier coalition surges occurred around February, 2005 and November-December, 2005. The October, 2007 peak of about 183,000 troops was slightly higher than the two previous.
- Beginning in September, 2003 with 1300 Iraqi civilian deaths, the insurgency begins its own series of "surges" which reveal themselves as spikes in Iraqi civilian deaths occurring 7-8 months apart: April, 2004 (2000 dead), November, 2004 (2600), August, 2005 (3300),and July 2006 (3600 dead). However, the pattern changes in 2006. Rather than the rapid decline from a one month peak, July is followed by a slight dip in August (3000), then another "surge" to another peak in October (3700). The 2006 insurgent "surges" exhibit not only an ever greater capacity to inflict casualties, but also a new ability (or willingness) to do so for an extended period. Over the 2003-2007 period the evidence suggests that insurgent capacities have been increasing, not eroding.
- US troop "surges" tend to follow insurgent "surges," but Iraqi civilian deaths have generally dropped back to "normal" before the extra US troops arrive. This was true in 2004, 2005, and again in 2007.
- Earlier US surges were ineffective in hampering the development of the insurgency. Insurgents were able to sustain their rhythm of attacks, and each subsequent spike in Iraqi civilian deaths was higher than the previous one.
- Iraqi civilian deaths in 2007 fall to the lowest levels since the occupation began. However, the lack of results from the two previous surges leads us to look for other causes for this. The most likely ones are the deals that the occupation has negotiated in Anbar province and around Baghdad, and the truce called by Muqtada al-Sadr. There is no publically available evidence that the extra troops eroded insurgent strength or deterred insurgent attacks.
Conclusion: The coincidence of the 2007 US "surge" and the decline in Iraqi civilian deaths is probably not due to the increased military effectiveness of US forces, but rather to the "if you can't beat 'em, then pay 'em" approach that the US has taken in some key areas. It would not be an exaggeration to say that Iraqi deaths are down because Bush has essentially thrown in the towel on trying to subdue Anbar and the Baghdad area locales. Does this amount to "surrendering to the terrorists"?