Facts, Lies, Statistics and NIEs

The big news on the defense and foreign policy front is the release earlier today of the National Intelligence Estimate (NIE) on Iraq, an unclassified summary of which is now available online (.pdf). Congressional staff are poring over a much longer (90-page), classified version. President Bush was presented with a copy of the full report yesterday.

National Security Adviser Stephen Hadley discussed the NIE with reporters, characterizing the report as “a good statement about the risks if we do not succeed in Iraq, for Iraqis, for the region, and for Americans here at home.” (Count this as Exhibit A for the defense.) 

Bloggers immediately pounced on the report to make the counterpoint, namely that the Bush administration’s escalation plan — by the NIE’s own admission — couldn’t possibly work (Exhibit B for the prosecution).

The near-instantaneous interpretation of a key intelligence assessment is hardly a new phenomenon. As I explain in my book John F. Kennedy and the Missile Gap, the selective leaking and politically motivated interpretation of intelligence muddied the water on the nature of the Soviet threat in the late 1950s. (A more digestible paper I wrote for the Princeton Project on National Security that discusses these 1950s-era debates, as well as a discussion of the “Team B” controversy of the 1970s, can be found here.)

The response to the latest Iraq NIE fits a similar pattern. Each side is fixing on a few relevant passages to prove its case or, failing that, to engender doubt about the other side’s arguments.

For example, advocates for an expeditious withdrawal (of which I am one) can point to the NIE’s statement that “Iraqi leaders will be hard pressed to achieve sustained political reconciliation in the time frame of this Estimate,” and conclude that the likelihood of success is extremely small.

On the other side, opponents of withdrawal can point to the NIEs discussion of what might ensue if and when U.S. forces withdraw. These are the passages that Hadley highlighted in his press briefing.

No NIE could ever hope to resolve such disputes. However, from my perspective, the key to interpreting NIEs lies in the probabilities assigned to various predictions. The likelihood that a particular event will play out is never expressed in numerical terms. But this NIE includes a useful tutorial for NIE naifs that tries to explain how different words (“likely,” “unlikely,” “probably,” “may be,” etc.) should be interpreted along a probability continuum, from remote to almost certain.

So, returning to the passage pertaining to a U.S. withdrawal, pay close attention to the qualifiers (in my italics, with my impertinent questions in brackets):

If coalition forces were withdrawn [ever? in less than 18 months?], if such a rapid [how rapid?] withdrawal were to take place, we judge that the Iraqi security forces would be unlikely to survive as a nonsectarian national institution. Neighboring countries, invited by Iraqi factions or unilaterally, might intervene openly in the conflict. Massive civilian casualties and forced population displacement would be probable. AQI, or al Qaeda in Iraq would attempt to use parts of the country, particularly al Anbar Province, to plan increased attacks in and outside of Iraq. [attempt, perhaps; but how likely is it that they would succeed in establishing a comfortable safe haven. My guess? Not likely. See here and here.] And spiraling violence and political disarray in Iraq, along with Kurdish moves to control Kirkuk and strengthen autonomy could prompt Turkey to launch a military incursion.

No one disputes that conditions in Iraq are dire. What remains of a debate in this country is between those who believe that we must pay any price, and bear any burden, to succeed in Iraq, because failure would spell the “beginning of the end” of Western civilization (hat tip to Justin Logan). On the other side are those who believe that the price that we have paid in Iraq is already too great, and that there is a no reasonable likelihood that the Bush administration’s stated goals in the country can be achieved at anything approaching a reasonable cost. This is a debate between the believers and the non-believers, if you will.

It would be unreasonable to expect, therefore, that this NIE will resolve the debate one way or the other. Ultimately, policymakers must make a judgment, on the basis of imperfect information, and be prepared to explain and defend that position to the people that elected them to office.