Baghdad, 2003

The 2003 invasion of Iraq breaks the pattern of this book in a specific and useful way. The central factual claim was tested within eighteen months, in public, while the war was still being fought.

The stated case was that Iraq possessed chemical and biological weapons and was reconstituting a nuclear programme. There was no Security Council resolution explicitly authorising force; the legal argument rested on reviving the authority of earlier resolutions, a theory many international lawyers rejected at the time.

The Iraq Survey Group was established to find the weapons. Its comprehensive report, delivered under Charles Duelfer on 30 September 2004, concluded that Iraq had destroyed its chemical and biological stocks and ended its nuclear programme after 1991, and that no stockpiles existed at the time of the invasion. It also found that Saddam Hussein retained the intention to reconstitute capability if sanctions were lifted.

That second finding is often dropped by one side and inflated by the other. An intention without weapons is a genuine concern, and it is a case for maintaining sanctions and inspections. It is not the case that was made in 2002 and 2003, which was about weapons that existed and might be used.

Britain's Iraq Inquiry, chaired by Sir John Chilcot, reported in 2016 after seven years. Its findings were about process, and they were blunt. The judgements about the threat were presented with a certainty that was not justified. Peaceful options for disarmament had not been exhausted. The legal basis was far from satisfactory. Planning for the aftermath was wholly inadequate.

What the inquiries did not find is as important as what they did. Neither concluded that the intelligence was fabricated.

That distinction is the reason this chapter is in the book, because the failure it describes is general and recurs everywhere.

An intelligence system exists to produce judgements under uncertainty and hand them to people who need to decide. Analysts write with caveats, because that is what the evidence supports. The caveats are summarised at the next level, compressed at the level above that, and by the time a judgement reaches a minister it has become a sentence with no qualifications left in it. Where the political demand is for confidence, every stage of that compression runs in the same direction, and nobody has to lie for the final product to be false.

This is why the reforms that followed were about how assessments are written, how dissent is recorded, and who is required to challenge a consensus, rather than about which individuals were in the room. The individuals change. The mechanism does not.

One further observation belongs here, about the archival lag from the first book of this course.

In every other case in this book, decades passed before the documents allowed a judgement. Here the record arrived almost immediately, and the public argument did not become noticeably more rational for it. The Duelfer Report was published a month before an American presidential election and was absorbed into it. The Chilcot Report appeared twelve years later into a country with entirely different preoccupations.

Evidence arriving promptly turns out to be necessary and nowhere near sufficient. The rest of the problem is that people have to be willing to read it at a moment when reading it costs them something, and that is not a problem any declassification schedule can solve.