As the scary scenario of a civil war breaking out along sectarian lines has come hauntingly close to reality, two questions concerning our own effort become very uncomfortable: does the education and construction of Iraqi forces happen in a balanced way which brings together different ethnic and religious profiles within the same units; and are the forces representative of the population in terms of those same divisions? We don't really know that, Vance Serchuk of the American Enterprise Institute argues in a fine analysis in the Weekly Standard which is a reaction to the Pentagon's recent Quarterly Report on the situation in Iraq:
An Iraqi brigade, for instance, that is overwhelmingly composed of Kurdish peshmerga or Shiite militiamen would seem significantly more likely to place its loyalties with political factions in Erbil or Najaf, rather than the official power ministries in Baghdad. Indeed, if we see overt sectarian purges of the army, it is a clear sign that Iraq is slipping into civil war. The establishment of non-sectarian units, on the other hand, would offer some of the most tangible, visible evidence that Iraq really can function as a unitary state. More than just an economy-of-force substitute for withdrawing American troops, an integrated Iraqi army could help bind together the country's fissiparous communities in a genuinely national, representative institution.It would seem reasonable that the responsible organization do have these numbers and that they sit on them because they look grim. Two years ago the need for Iraqi security forces was the most obvious challenge and the pressure on the coalition forces to do something about it was enormous -- and any element that could help forge units must have been welcomed. But they just may not exist, exactly because they are and have for a long time been obviously politically sensitive data. This would be a negative but not surprising dynamic within the fact-based, metrics and formal evaluation-driven New Public Management, which is one of the many meanings of "transformation" at the Pentagon.
Given the significance of the army's ethnic and sectarian makeup, why then has the Pentagon been uninterested in collecting and analyzing data about it? (...) Interestingly, the new Pentagon report does acknowledge concerns about sectarian tensions inside the Iraqi army, noting the importance of "a professional force representative of the diverse ethnic and religious fabric of Iraq" and stressing efforts on the part of the coalition "to recruit personnel from across the spectrum of Iraqi society." It also cautions, however, that "this does not mean all units are fully representative" and that "uniform balance across all ten divisions at this time" is impractical. Fair enough. But in the absence of data--the very metrics that Congress has been demanding for months now--it's all but impossible for policymakers and the public to evaluate the administration's claims of progress. Just how unbalanced are Iraq's 10 divisions? What kind of movement in bringing Shia, Sunni, and Kurdish soldiers together has been made since the last report was issued in October 2005? Is the problem getting better or worse?
The Pentagon's reluctance to engineer the ethnic and sectarian composition of the Iraqi security forces is made all the more bizarre by the fact that it has displayed no such reticence when it comes to parallel efforts in Afghanistan. There, the indigenous army that Washington began building in 2002 was initially dominated by a single group--the Panjshiri Tajiks who had led the Northern Alliance against the Taliban and then seized control of Kabul. It was against Panjshiri objections that the United States insisted on imposing rough ethnic quotas, creating carefully mixed units of Pashtuns, Tajiks, Hazara, Turkmen, and Uzbeks. The result is arguably Afghanistan's first real national institution--a strong, multiethnic army clearly distinguishable from the parochial militias the country is accustomed to seeing. This has meant that the Afghan army is not just an instrument in the military campaign against the Taliban and al Qaeda; it is also a rallying point for national pride--proof the country can transcend the dueling fiefdoms that have, until recently, divided it.
Policy decision-making should always be fact-based. Yet NPM, as all new things, brings with it its own set of related challenges. The most obvious, pressing and not always acknowledged problem is the measuring itself. What is measured, how, why, and what repercussions do the metrics obsession have on the ground? The political need for metrics makes it necessary to get some data: IEDs, "actionable tips", security force members, etc. This in itself can cause a bad quality feedback in a system if operations become obsessed with fulfilling abstract targets instead of reacting organically to their situation. Drowning political decision makers in information is of course one of the oldest ways for any civil servant to get them of your back, but this case could be an example of the opposite.
Because data collection is such a powerful instrument as it produces handy actionable policy input, the upstream frameworking choice of what is measured becomes so much more important. And the decision not to include tracking of potentially problematic data is the safe choice for any career civil servant living in a culture that doesn't tolerate error well. Politicians and decision-makers can thus add to question, demand insight into and influence on those seemingly bureaucratic choices to their ever longer to-do lists.
Later UPDATE: Stephen Biddle seems to agree with the basic proposition in this Washington Post article.
Even later UPDATE: More tales of counterinsurgency in practice from the Washington Post (April 19, '06): "Mystery Hangs over Bagdad Battle", and "Mission of Frustration in Afghan Villages". Both go to show the strategic necessity of empathy in COIN ops: mere technical linguistic fluency isn't enough; and both contain the element of nationalization through the local army (as stateliness inducing institution and functional tool), and concomitant problems of factionalization of the same. See the link below for more on conditions for COIN.
Completely later UPDATE: Rumsfeld's and Rice's Baghdad-visit in late April had exactly sectarianism in security forces as a headline, see e.g. this NYTimes article:
Administration officials said that in his private meeting with Ms. Rice, Mr. Malaki spoke specifically of rooting out the influence of militias in Iraq's police forces, which number about 135,000 nationwide. Americans and Iraqis say that thousands of these are actually members of Shiite militias that carry out extra-judicial killings of Sunnis.