The Brookings Institute publishes a quarterly short report about "The State of Iraq". It is a collection of various criteria, numbers measuring these and a short interpretating note. A chart with the numbers is regulary published in the New York Times and often quoted in discussions about Iraq.
Responsible for "The State of Iraq" is Brookings’ Senior Fellow Michael E. O’Hanlon, who recently wrote a contentious op-ed favouring a continuatuion of the ‘surge’ military escalation in Iraq.
Reason enough to analyse the most recent report published in today’s NYT with data for August 2007. For comparisons the numbers for May 2007, February 2007, November 2006 and August 2006 are available at the Brookings site.
In today’s introducing text O’Hanlon notes:
Nonetheless, the military momentum appears real, despite the tragic multiple truck bombings in Ninevah Province on Aug. 14 that made that month the deadliest since winter. Overall, civilian fatality rates are down perhaps one third since late 2006, though they remain quite high.
In detail:
made that month the deadliest since winter
The numbers for ‘Iraqi Civilian Fatalities’ listed in this and former editions of the "State of Iraq" are:
- 2,500 for August 2007
- 3,000 for May 2007
- 2,500 for February 2007
- 4,000 for November 2006
- 3,000 for August 2006
If indeed August was the deadliest month since winter, as O’Hanlon writes, why are his own numbers for May 2007 higher than the ones for August 2007?
Overall, civilian fatality rates are down perhaps one third since late 2006
This of course directly contradicts the
words above. "Deadliest month" and "one third less" do not fit together. Which
is it?
For the casual reader the sentence also might imply that lower numbers in August 2007 versus late 2006 did sink because of the ‘surge’ which started in mid February. But the August numbers given are at the same level as the February 2007 numbers. The down momentum was between November and February. It can not be a result of the surge.
The Iraqi ministries reported (unlikely low) 1,770 dead civilians in August, an increase of 7% over July. While one might doubt the size of various numbers, the trend of numbers from different sources should be the same.
In the Brookings’ report the number of ‘Iraqi Civilians Newly Displaced by Violence’ has decreased from 100,000 in August 2006 to 80,000 in August 2007. But the UNHCR just claimed an increase to more than 60,000 from 50,000. Again the size of the numbers may differ for this or that reason, but why are their opposite trends between O’Hanlon’s report and other sources?
The NYT charts list the criteria and the respective August values for 2003 to 2007. Numbers with ‘more favorable conditions‘ have a white background, those with ‘less favorable conditions’ have a dark grey background and three shades of grey are differentiating between the extremes.
In the current graph the line for the criteria ‘Iraqi Security Forces’ has constantly increasing numbers from 35,000 in August 2003 to 360,000 in August 2007. The background for the 2003 number is dark grey, while the background for the 2007 number is white. Inbetween the background gets lighter from left to right. This seems consistent with the graphics key.
But why is the August 2007 number of 4.1 gigawatts ‘Electricity Production’ on a white background, while the August 2006 number of 4.4 gigawatts is on a medium gray background? Is less electricity production ‘more favorable‘?
Why do the numbers of ‘Prisoners held by U.S./Iraq’, which increased from 6,000 in August 2003 to 60,000 in August 2007, have a constant medium grey background? Is a tenfold increase in the number of prisoners a factor meaningless for the "State of Iraq"? If so, why is it listed at all?
Looking at the various versions of these reports one notes that half of the twenty something criteria used in each change from edition to edition.
The August 2006 report listed positive trending, increasing numbers of ‘Iraqi Children in School’, ‘Trained Judges’ and ‘Registered Cars’. Since then those criteria vanished. From November 2006 to May 2007 ‘GDP Growth Rate (%)’ numbers were included. They were dropped from the recent table.
Instead a criteria of ‘Attacks in Region Near Mosul’ was added in the August 2007 table showing a decrease from 15 to 8 over one year. Is the Mosul number more important than Iraq’s GDP? Also added was ‘Resources Going From Baghdad to Average Iraqi Provinces’ which doubled to 100 million over a year. If Arbil and Basra get 50 million each, while Anbar and Baghdad get zero that would certainly fit the given ‘average’ number. So what does this criteria say? And how many taxes dinars does the central government in Baghdad collect from the provinces?
To summarize:
The text O’Hanlon delivers is in itself contradictionary. It does not reconcile with the numbers he presents. The numbers do not fit the coloring scheme of the graphic tables. That some of these numbers differ in size from those of other sources is explainable, but contradicting up-/down-trends are baffling. The criteria O’Hanlon uses seem arbitrary. One suspects that criteria that are trending negatively are dropped while criteria showing ‘success’ get added.
O’Hanlon is often described as a ‘scholar’ and he is a visiting lecturer at Princeton University.
Judging from the quality or of his research as documented above, I can not recommend to take any of his courses.
