[2014]

The Butler committee was the first critical consideration of intelligence's service of assessment to top government since the changes of the 1960s. I had been publishing proposals for reform for some time before 2004 and had the luck to be able to put them to Lord Butler's committee. I was later sent a copy of the transcript and it is reprinted in this collection, along with my RUSI article on the committee's report which was published shortly afterwards. Here I offer my reflections ten years later, with an impression of the changes that followed Butler's recommendations.

Like everyone else I had judged after the Iraq War that the Joint Intelligence Committee (JIC) might have done better over Saddam's Weapons of Mass Destruction (WMD). For intelligence as a whole it was a disturbing failure, the more so since its task of estimating what weapons a secretive regime possessed, and informing government reactions, was more of a throw-back to the Cold War than a novel, twenty-first-century problem. The basic challenge for assessment on Iraq was drawing conclusions from the relative absence of evidence, and of this there had been ample Cold War experience. I was vividly reminded of the debates in the late 1950s and early 1960s about the Soviet missile threat, when analysts struggled to determine whether there were hundreds of operational Soviet ICBMs or none at all. Intelligence officials were uncertain on the key aspects of Iraq's WMD, but they were not determined enough in their uncertainty.

This may have been influenced by pressure on the JIC for some years to provide positive ‘key judgements’ instead of sitting on the fence. I judged that intelligence on Iraq was caught up in the assumptions and excitement of the day and had slipped into the role of government's counsel for the prosecution rather than its judge of evidence, or in this case the absence of evidence. There is an open academic question of whether intelligence is ever capable of questioning conventional wisdom in circumstances of political importance and stress. Sir Lawrence Freedman observed more than thirty years previously that intelligence analysts are ‘socialised into a particular world-view which is shared by the main consumers of their work’.