Spooks and the Art of the Dissenting Footnote

By Robert Carlin
Robert Carlin was chief of the Northeast Asia division of the State Department's Bureau of Intelligence and Research from 1989 to 2002 and before that an analyst in the CIA.

Los Angeles Times

April 1, 2005

A presidential commission is recommending that disagreements among intelligence analysts be clearly flagged to alert policymakers to the fact that the intelligence community, which likes to use the imperial "we," is not really of one mind. This is an idea that pops up now and again, no matter that it rarely seems to work. In fact, the IC has paid lip service to the importance of airing dissent for years.

The result has been the birth of a netherworld that few on the outside have contemplated — the dissenting footnotes that dangle precariously from the text of intelligence assessments on key issues. For example, if an intelligence assessment were to say "we believe that Posiland has a developed long-range missile capacity and is about to commence deployment," a footnote might appear at the bottom of the page in small type saying, "The Director of [fill in the blank] believes that there is no evidence Posiland has either the intent or the means to produce, much less deploy, long-range missiles."

How does a footnote get into an intelligence document? An intelligence agency that disagrees with the majority view will "take" a footnote — as in (spoken with clear exasperation after much arguing) "We're going to have to take a footnote." Other agencies may then choose to "join in the footnote," a sentiment usually expressed sotto voce.

Endorsing the principle of dissent and actually integrating contrary ideas into analyses are two different things, of course. Everyone knows that footnotes in intelligence community documents are very much like unpleasant noises at a banquet. No one will say so exactly, but footnoting an intelligence estimate is a sign of, well, let us be frank, ill breeding.

And of all of the intelligence shops, the State Department's Bureau of Intelligence and Research has long been considered the most ill bred. INR takes a lot of footnotes, the best known in recent times being those that appeared in various intelligence estimates on Iraq's weapons of mass destruction. Proportionally, INR probably takes more footnotes than any other intelligence shop in the world.

Traditionally, few agencies join INR in footnotes. A CIA element once threatened to line up with INR against a sister office, but eventually it was decided such a move would violate the laws of nature.

Truth be told, intelligence analysts rarely agree on anything. And as such, one might expect their assessments to contain more footnotes than they do. How to account for the relative scarcity? Though analysts don't mind saying they will take a footnote, many superiors get weak-kneed at the prospect. Does one really want to appear to be disagreeing with the majesty of an intelligence assessment? Is the point of disagreement significant enough to require all the fuss? To avoid the loin-girding it might take to pipe up in opposition, a small escape hatch is used from time to time: Sit tight, stay silent and hope that INR will voice the dissent.

Good footnotes have their own zing; great footnotes take no prisoners. That footnote is best which makes its case in a few well-chosen words. When the text says X, a dissenting footnote, logically, should say, "No, Y because …. " Yet, because of space limitations, many dissenting footnotes end up as only a simple "No, not X," or its slightly more elaborate cousin, "One can't say X because there is insufficient evidence to do so." A single footnote of this type may have a regal air. A series of them begins to look whiney, a scarlet letter no tough-minded intelligence agency wants to wear.

Dissenting footnotes are born of frustration. No one expects them to change anyone's mind. They become an effort to speak for the record and, even more important, to preserve some intellectual integrity in the analytical process. This is not always easy to do because of the games that are sometimes played. Are there instances when the text of an intelligence assessment has been changed after the fact so that a footnote looks pointless and will have to be removed in the final clearance meeting, not because the issue under dispute has been resolved but because the dissent has been outflanked? Could such things happen? Do bears sleep in the woods?