Our idea of a spy might be tainted by popular films in the spy movie genre, but everyone knows that the risky, clandestine part of intelligence agencies’ business is just one part of it, and not even that is particularly exciting or glorious. The rest is intelligence analysis, the meticulous, detail-oriented and rather academic endevaour of making sense of all the bits and pieces, and this part is what this book is about. Obviously, the title How Analysts Think is not particularly sexy, so here we are. The author, Sir David Omand, is a decorated civil servant who, after studying economics, took on pretty much every signals intelligence-related post within the British state, climbing up the career ladder to lead the GCHQ in 1996 and 1997. Having taken on so many different roles, he is perfectly qualified to comment on all the aspects of the intelligence analysis process, starting at the very beginning of evaluating the relevance and reliability of data, to integrating it within accepted theories, to making predictions, and finally to presenting it to the decision makers.

In How Spies Think, the intelligence analysis process is presented by Omand as taking place within the SEES framework, the letters of which stand for situational awareness, explanation, estimations and strategic notice. Over the first third of the book, Omand discusses each of these topics in detail, enriching them with anecdotes and demonstrations from his own experiences and from well-known historical intelligence opeations. These relevant episodes lend the book significant credibility and appeal, as they are told from a unique position of power and insider knowledge. Omand was there when NATO airforce heads met Ratko Mladic to hand him an ultimatum, he personally gave Thatcher the dossier that led to British ships attacking Argentine navy, and he was a personal friend of the head of NSA. He also knows inside out not only overseas but UK-wide engagements of MI6 and MI5, as in the highly interesting account of secret contacts with (P)IRA. These contacts were used by the IRA to contact the government at the time, with a request to start negotiations to end the troubles.

Already in the discussion of situational awareness, Omand broaches the topic of Bayesian probability, an overarching theme of the book. It is no surprise that Bayes’ formula is so important in intelligence analysis, as it is the perfect tool for evaluating disparate bits of information without an overarching frame. To refresh your memory, the Bayesian formula tells us that P(A|B) = P(B|A)P(A)/P(B), with P(A|B) being the probability of A given B. This very simple formula has been put to use in a surprising range of applications, a topic discussed in depth in The Theory That Would Not Die, a book I shall review soon. Bayesian inference is particularly relevant in intelligence analysis because, as Omand puts it, “One advantage of the Bayesian approach to analysis is that it concentrates attention on the relevance of the evidence and not just its character” (p.149). You can see this relevance in the above formula, in fact: It’s the P(B|A) part, that is, the probability that evidence B will be observed given that event A is taking place. Omand gives multiple demonstrations of the usefulness of Bayes in the intelligence context, starting with the Cuban missile crisis. At the time, one of the historically most important intelligence assets of the West, Oleg Penkovsky, was providing the US and UK intelligence services with valuable information regarding the state of the Russian missile program. This information helped CIA analysts re-evaluate satellite images of constructions sites in Cuba, interpreting them as missile sites instead of other kinds of construction, which turned out to be right.

Bayesian inference is a fundamental tool in making the analysis process as neutral and objective as possible, with additional methods and conceptual tools being used to butress it. Among these, one of the more interesting is “analysis of competing hypothesis” (ACH), which is essentially finding and focusing on the theory with the least contradictory evidence. A topic dwelt on in quite some detail by Oman is the role of psychological biases in the evaluation of evidence. This is not surprising, since such biases are rather insidious and difficult to withstand, as anyone who makes decisions based on data would attest. In the case of intelligence, the situation is exacarbated by the responsibility involved, complexity of the institutions, and the historical baggage. When an intelligence service makes a significant misjudgement, for example, future decisions all take place in its shadow, and when complex cases need to be brought to a conclusion, the lesson of the past mistake blurs the judgement. An interesting story is that of James Jesus Angleton, head of counter-intelligence at CIA between 1954 and 74. After the war, he was stationed in England, where he established a close friendship with one Kim Philby. The latter was outed as a KGB spy in 1963, and defected to the USSR. This betrayal caused Angleton to devolve into a paranoid mindset, worsened by a Russian defector who, for his own security, cooperated with Angleton to aggrevate his fears and paranoid theories. The result was institutional paranoia that led to hindered careers and mistreatment of useful intelligence resources. Once Angleton was deposed, the reaction to his methods had led to an aversion to counter-intelligence work within CIA. Omand sees the later high-profile cases of espionage within the CIA such as that of Aldrich Ames as being closely related to this reluctance for counter-espionage.

So much for the relatively neutral intelligence analysis content. Omand writes about intelligence as a tool at the disposition of the state, with varied uses. As the state is a political body in flux, it should be quite obvious that intelligence itself is a political tool. Reading How Spies Think, it becomes obvious rather quickly that Omand is working hard to give the impression that intelligence is a politically neutral tool. There are quite a few examples of this, but the most egregious, and for me as one who followed the whole Iraq war shenanigans as a young man, the most infuriating instance, is his explaining away of intelligence that justified the Iraq war as a combination of bias and mistakes. For those who don’t remember it, US and UK governments under Bush and Blair, once having made their mind up that Iraq was to be invaded, came up with all kinds of flimsy reasons why this is a great idea. Intelligence delivered by a single anonymous informant was used to build up a case, culminating in Colin Powell holding a presentation in the UN security council, complete with diagrams of “mobile production facilities” for biological weapons. Powell later admitted that he knew it was all a lie, but made the presentation nevertheless. People who openly doubted this intelligence were mobbed and bullied out of their jobs, such as Hans Blix, who headed the UN arms control comission investigating Iraq. In the light of this intentional drive to produce intelligence, it is utter nonsense to talk about psychological biases and mistakes. The Iraq war is one of the biggest war crimes of modern times. It led to immense suffering, destruction, and a huge blow to the western block’s international standing. Omand talks a whole lot about trust, but it’s difficult to trust him when he lies and obfuscates his way away from any kind of responsibility in the name of US and UK intelligence agencies.

The second, and presumably for today’s youth the most infurating instance of ignoring intelligence’s political role is Omand’s treatment of Edward Snowden’s revelations. Those who understand their gravity call them revelations, but for Omand they are “accusations”, “allegations” and generally “malinformation” (p. 162). This is, to put it diplomatically, rather stupid. And then to go on and claim that as a result of these revelations, communication devices switched to strong encryption, and that this has “inadvertently benefited terrorists and others with criminal intent” (p. 173), is plain shameless. US and UK intelligence organizations have broken their own laws to gather mass data about their own citizens, essentially spying on the whole population. This, again, Omand fails to properly name and take responsibility for.

Regarding tone, the book is sober and understated, as one would expect from an Englishman. Unfortunately, it is also frequently marred with subpar writing. Omand occasionally tries to make a connection between the intelligence analysis process and the vagaries of daily life. In the section on Explain, for example, he (albeit indirectly) compares the time when he met Mladic with heads of NATO airforces, to office work (p. 42). The only thing that came to my mind was What kind of offices does this guy work at? Another example is his comparison of trust in a marital relationship with the nuclear disarmament process, which is rather far-fetched. A result of these attempts at making the topic relevant to the average John or Jane’s daily life is that the writing becomes unbearably sloppy and overly general in certain places (e.g. p.216). A similar deterioration of the text can be observed when Omand writes a long and convoluted ode to the intelligence cooperation between the US and GCHQ. Where the history of this relationship is expounded upon, it is informing and interesting, but when Omand tries to explain the relationship, going so far as to compare it to a date (p. 219), it devolves into a ne me quitte pas letter to the Americans, making one wonder whether Omand fears a split happening in the future.

Rating: Two and a half James Bonds out of five. Long and occasionally not so smooth, but rich in insightful bits. You have to suffer through the politics, though.