REFLECTIONS ON SECRET INTELLIGENCE
Sir David Omand GCB
I am honoured to be asked to deliver this annual Peter Nailor memorial lecture, and I am grateful for the opportunity it gives to remember the part he played within the Ministry of Defence, in which I spent a large part of my career, and in his later academic work in developing the field of defence studies, as well as his role as Provost of Gresham College. I am also delighted that I am to follow my old friend and colleague Sir Kevin Tebbit, Permanent Secretary of the Ministry of Defence, who delivered the first Peter Nailor lecture. He offered a combination of analysis and personal perspective illustrating the changes in defence over the last 30 years with reflections from his own career. I shall try to emulate this, and thus perhaps set a pattern for future lectures in this series
I am grateful for the spur of this lecture to reflect on my most recent experience, as the first holder of the Permanent Secretary post in the Cabinet Office of Security and Intelligence Coordinator, taking on the traditional responsibilities of the Cabinet Secretary for intelligence and security matters. The creation of my post was a pragmatic way of allowing the new Cabinet Secretary to devote the additional effort that the Prime Minister wanted given to public service reform and delivery. My post brought together for the first time responsibility as Accounting Officer for the three secret intelligence agencies and thus for thinking about their future development, organisation and funding, with the operational responsibility for central crisis management. In a way that is characteristic of the evolutionary British approach to institution building, this construct then allowed me to use the role to develop strategic thinking on counter-terrorism, including pulling together in a British context what our American colleagues call ‘homeland security’. The value of having a full-time senior official at the centre of the overlapping circles of intelligence, security, homeland security and crisis management is I hope now evident. Last month, the Government completed this phase of evolutionary development by giving the Security and Intelligence Coordinator also the responsibility of chairing the Joint Intelligence Committee. We have therefore for the first time in our history a single senior official at the top of the British intelligence community in its widest sense, demonstrating once again that the shortest distance between two points in governmental space is not necessarily a straight line. But more of all that in a moment.
It will be evident that I am now retired, and any opinions that follow are entirely my own responsibility and should not be taken as statements of official policy far less predictions of future government intentions. I should also issue a health warning to anyone who may have strayed in here in the prurient expectation from the title of my talk of newsworthy disclosures of official secrets: what follows are reflections not revelations.
The sharp-eyed will have noted that I am paying another tribute, since my lecture deliberately echoes the title of the memoir by the late Professor R V Jones who did so much to develop scientific intelligence during the second world war. His writings contain encouraging and practical lessons for those in the intelligence community - and for those they serve - and I shall return later to some of these, in terms of the scientific approach to the collection and interpretation of intelligence.
I have organised my remarks into three linked sections, which in one lecture cannot hope to be comprehensive but will I hope give an insight into some of the issues:
First, the purposes of intelligence. I shall reflect on the past and future uses of intelligence, and its inherent limitations. What purposes does it serve today and how important will it really be in the years to come?
Secondly, the meaning of intelligence. How can we best apply scientific method in assessing intelligence? Why do we often get it right but nevertheless sometimes get it wrong?
Thirdly, the evolution of the intelligence community. What factors should shape future organisation? Are there lessons we should learn from the history of the central organisation for defence over the last 60 years? Are there principles that are internationally applicable?
1. The purpose that intelligence serves
A starting definition might be that the ultimate object of intelligence is to enable action to be optimised by reducing ignorance; and of secret intelligence to achieve this object in respect of information that others wish to remain hidden. Thus stated, the purpose of intelligence is not linked simply to knowledge for its own sake but to organised information that can be put to use. The military commander and the policy-maker alike need to have professional staff support to collect and organise information relevant to the decisions and actions that they want to take, and some that they may not yet know they need to take.
Intelligence provides therefore, in the words of the recent Australian intelligence commission:
- “Warning, notably of terrorist plans, but also of potential conflicts, uprisings and coups
- Understanding of the regional and international environment with which decision makers will need to grapple
- Knowledge of the military capabilities and intentions of potential adversaries, a vital ingredient in defence procurement and preparedness
- Support for military operations, minimising casualties and improving the environment for operational success
- Support for an active and ambitious foreign, trade and defence policy
- And beyond these vital roles of intelligence in providing information, modern intelligence can be a more active tool of government – disrupting the plans of adversaries, influencing the policies of key foreign actors and contributing to modern electronic warfare.” (The Flood Report 2004).
The greatest added value of the secret part of intelligence comes, as Lord Butler emphasised in his report on intelligence into WMD, because for many of the topics of most pressing interest there are active measures employed against us designed to hide or disguise the information we seek. So traditional secret intelligence from human sources, from intercepted communications, from electronic or other measurements, from analysis of acquired enemy equipment, from defectors and from satellite reconnaissance has a unique value. The intelligence analyst has also always drawn on diplomatic reporting, and increasingly open sources of all kinds and in the future increasingly on patterns drawn from data mining of large sets of data, both open and protected. Open sources are important not least because secret intelligence is also expensive to collect and process, and unnecessary reliance on secret sources is wasteful. But I deliberately start with purpose and use, rather than sources and methods, to point us towards the recognition that in the future we shall need to put proportionally more national effort into collation, analysis, assessment and research than in the immediate past. And hence my earlier reference to the need to think of an intelligence community that is wider than just the three main Secret Agencies. When it is all put together, the test of effectiveness of the overall community is how far it can genuinely help reduce ignorance in each of the areas I listed, meeting the different needs of an increasingly diverse set of non-traditional as well as traditional customers.
Is this broad definition so very different then from the craft of journalism? The view of The Economist magazine is worth repeating on this point. In a comment in 1966 on the retirement of Sir Kenneth Strong the Economist wrote, “Modern intelligence has to do with the painstaking collection and analysis of fact, the exercise of judgement, and clear and quick presentation. It is not simply what serious journalists would always produce if they had time: it is something more rigorous, continuous, and above all operational – that is to say, related to something that somebody wants to do or may be forced to do”
‘More rigorous, continuous and above all operational’: I shall say more about rigour in the second part of my talk. Continuity is certainly important. I recall during the Falklands Campaign the frantic search by the analysts for information about the Port Stanley runway whose size was crucial to assessments of Argentine capability but which had recently been upgraded. The information did not involve secret intelligence; the extra hard-standing had been paid for by the British taxpayer as a civil development project but had never found its way onto the defence intelligence database no doubt because too few people were trying to keep up with too much information. A very different, and much more serious, example was dramatised by Robert Harris in Engima, vividly illustrating the high stakes when continuity is lost, in that case in signals intelligence on submarine movements during the Battle of the Atlantic. Continuity of the overall information base is an essential, if sometimes, unglamorous, role for the intelligence staffs for which resources have to be found. Increasingly, with pressure on staff numbers in customer departments, it is a function that the wider intelligence community may have to fulfil on behalf of government.
Behind continuity in assessment lies continuing availability of sources. Building capability takes time, in recruiting and training, and in researching, developing and building technical systems. The risks of failure to recognise this can be serious. In 1844 after the Mazzini affair, over-zealous Parliamentary committees of inquiry led to the abolition of the Secret Office and the Decyphering Branch which had been serving British interests since it opened in 1653 and was put on a legal footing by Parliament in 1670. As as a result Britain entered the First World War without an effective Sigint system which had rapidly to be assembled from scratch in Room 40 in the Old Admiralty Building to help fight the war at sea and the German submarine campaign. ‘Gentlemen do not read each other’s mail’ said US Secretary Stimson in 1931, an excess of ethical foreign policy that meant after Pearl Harbour the US went to war comparably handicapped.
Those gaps in continuity were caused by qualms over the acceptability of methods of secret intelligence. Gaps in continuity of a different kind could occur in future, with even more direct consequences for public safety, if doubts are not allayed about the use of modern IT and data processing by the intelligence agencies in their counter-terrorist campaigns. The proposals recently before the European Parliament for the retention of communications data are a case in point. But all our experience of fighting terrorism in the past shows that where pre-emptive intelligence is available effective action to protect the public can be taken with minimum disruption to the community and without having to contemplate serious distortion of the rule of law, in effect allowing in respect of State coercion the bludgeon to be exchanged for the rapier. When he was US Director of Central Intelligence, Stansfield Turner expressed this question of public justification of intelligence methods thus: “There is one overall test of the ethics of human intelligence activities. That is whether those approving them feel they could defend their decisions before the public if their actions became public. This guideline does not say that the overseer should approve actions only if the public would approve them if they knew of them. Rather it says that the overseer should be so convinced of the importance of the actions that they would accept any criticism that might develop if the covert actions did become public, and could construct a convincing defence of their decisions.” Of necessity, the methods for uncovering terrorist networks and support structures depend upon the terrorists having only imperfect knowledge of them. So prior public debate is hard, and trust is needed that there is a convincing case for limiting data protection if it can be shown to give a higher degree of assurance of greater security and make it easier to maintain the due processes of criminal justice. So it is vital that there is public trust in the integrity of those serving in the secret world, including that they will follow what RV Jones called the doctrine of ‘minimum trespass’ into privacy, in a parallel with the doctrine of ‘minimum force’ that is enshrined in our common law.
When the call comes - the most recent example being the need to track international terrorist networks - we can only fight with the capabilities we have inherited. That is why I am glad that after 9/11 we secured significant increases in funding to build up national intelligence and security capability for the future. This investment is well under way in the Security Service, the other Agencies and the police service but it will be a few years yet before we benefit fully.
The quote from The Economist also referred to the operational purpose of intelligence. In recent times there has been a major increase in the relative importance of intelligence for ‘action this day’ particularly in the areas of proliferation, terrorism, narcotics and serious crime, as against intelligence to inform policy-making. And who can be surprised? Recent terrorist attacks have illustrated starkly the need for pre-emptive intelligence just as the previous successes in disrupting terrorist cells planning to attack the UK and our interests abroad showed that good intelligence saves lives. A question facing government now is what should the public be told about our understanding of the threat, both to justify practical security advice and to ensure an informed public debate about the policies that being proposed. The hardest questions about putting intelligence derived material into the public domain are not about ‘dossiers’, on which Lord Butler has I hope said the last word, but about the everyday decisions that have to be taken in the interests of public safety and what the public is told about the intelligence that justified them.
In reflecting on the purposes of intelligence, I find it helpful to think of the intelligence community serving three levels: these are the classic distinctions between working at the strategic level (including what Churchill called grand strategy), at the operational level and at the tactical level, with the distinguishing feature between the levels being the time horizon of the customers receiving the intelligence. As I shall go on to suggest these distinctions can be a helpful guide, for example when it comes to evolve organisations able to operate at each level and when it comes to building the right institutions and liaisons internationally:
- At the strategic level, intelligence assessments might for example look at countries at risk of instability or the possibility of conflicts over global trends in energy supplies or water resources, or the future evolution of terrorist movements, issues where the horizon is years or decades ahead. This is the everyday diet of the Joint Intelligence Committee, and its supporting expert Intelligence Groups, in which senior policy makers and intelligence chiefs try to arrive at joint key judgements on the evolution of issues where there are important national interests in security, defence and foreign affairs. Some scepticism is healthy: rarely could it have been said that secret intelligence was the decisive factor in policy making itself. And in the policy arena we have what WW Rostow called in relation to the 1944 controversy over the effectiveness of the aerial bombing campaign “the arena of power, vested interest and personality where forces quite different from straightforward intellectual argument were at work”. In my personal experience, however, from my years serving in the defence and security world, better, sometimes much better, outcomes have resulted from applying secret intelligence in the relevant fields than would otherwise have been the case, for example as Lord Butler confirmed, in the successes in countering proliferation to Libya, Iran, North Korea and from the AQ Khan network.
- At the operational level, the main demand is for timely all-source analysis to support operational decision making: examples might be over the risks to a military deployment, or the threat to an Embassy overseas threatened by terrorists, or the implications of high tech weaponry reaching a country of concern, or whether a particular export of specialised steels should be permitted, or additional protective security measures taken at a big public event or on a transport system. The intelligence community provides professional judgements on which it is for the policy community, and in major cases ultimately Ministers, to decide whether and how to act. The policy makers have to accept the assessments as they stand, as they do with military, medical or legal opinions. They know that they will be hammered, however unfairly, by the media if things go wrong and if they appear to have ignored warnings – or the opposite, to have pressed on despite cautious assessments. So the analysts in practice have leverage. They need to be sensitive to the dangers. By playing too safe themselves and passing on the problem, they may make it harder for the right risk management judgements to be taken, leading to over-restrictive security measures or warning fatigue leading to complacency. We should have sympathy with those who have to make these assessments, and encourage them to remain true to the principles of rational evidence-based judgement.
- At the tactical level, individual lines of intelligence are largely going raw to other intelligence specialists, to defence staffs or to policy customers who are themselves expert and able to interpret the material. Systems for analysis and rapid distribution are needed to support operations on the ground, for example the tactical application of intelligence to guide a counter-terrorist raid, or to follow a drugs shipment, or to intercept a breach of sanctions. This is the part of the iceberg that the public rarely is conscious of, unless things go awry.
Lord Butler also correctly drew attention to the penumbra of uncertainty that usually surround the intelligence fragments. Given the imperfections and gaps in the intelligence, what is required from the officers concerned at each of these three levels is, as Clausewitz wrote 200 years ago and Lord Butler cited, “a certain degree of discrimination, which can only be gained from knowledge of men and affairs and from good judgement. The law of probability must be his guide.”
2. Divining meaning of intelligence
Let that reference to probability now be my cue to move on to the second part of my talk, on how meaning is derived from intelligence.
The Joint Intelligence Committee or JIC has for over 60 years had the responsibility of producing predictive strategic judgements for the highest levels of Government. The key senior policy officials from the Cabinet Office, FCO, MOD, Home Office, HM Treasury, DTI etc are JIC members as well as the intelligence professionals and all have to dip their hands in the blood of the collective judgments, however unwelcome they may be. The JIC process is, as far as I know, unique around the world. One reason may be that the process involves close interaction between the senior policy makers and the intelligence community, examining the most sensitive intelligence in the course of producing JIC papers. But the reason the JIC emerged during the Second World War as it did was to bring greater rationality into bitter strategic debate. Take the exchange in 1934 as Vansittart, Head of the Diplomatic Service, read what he saw as - Treasury inspired - complacent air ministry estimates of German war production: “In any case, prophesy is largely a matter of imagination. I do not think that the Service Departments have enough. On the other hand, they might say that we [that is, the FCO and the SIS] have too much. The answer is perhaps that we know the Germans better.” Air Marshal Sir Victor Goddard, on the other hand, minuted, “What passes for intelligence of the enemy’s intentions is more usually propaganda for a change of Government policy: honest propaganda, maybe, but based on ideas rather than facts”. After seven year’s service on the JIC, in three different posts, I am convinced of the advantages, and encouraged by the fact that Lord Butler’s committee looked for but found no evidence of cross-contamination in the JIC between policy and assessment. So the principle is right, but of course, that is not to say that the practice is always perfect, which is why it is important that Lord Butler’s practical recommendations are implemented according to plan.
The professionals’ task is therefore to keep judgements anchored to what the intelligence actually reveals (or does not reveal) and keep in check any predisposition of policy makers to pontificate – that being the practice, as described by Bernard Dixon in his masterly book ‘The Psychology of Military Incompetence’ of trying to make nasty facts go away by the magical process of emitting loud noises in the opposite direction. The policy makers in turn ensure that the judgements actually try to address the issues that need answering rather than just those on which their intelligence sources are richest, and help the professionals couch any warnings justified by the intelligence, without their seeming to attack the policy itself and thus risk compromising the neutrality of the JIC. This is not always easy. President Reagan is quoted by Robert Gates later Director of Central Intelligence as complaining about these intelligence guys. “When I was growing up in Texas ”, he said, “we had a cow named Bessie.. One day I’d worked hard and gotten a full pail of milk, but I wasn’t paying attention and old Bessie swung her shit-smeared tail through that bucket of milk. Now, you know, that what those intelligence guys do. You work hard and get a good programme or policy going, and they swing a shit-smeared tail through it.” So speaking truth unto power is not going to guarantee popularity. But the message to those in power has to be follow what RV Jones quoted as Crow’s law: ‘do not believe what you want to believe until you know what you ought to know’.
Forcing these top, and very busy, officials to work actively together in the JIC on key judgements for an afternoon every week of the year generates a sense of pol-mil community that is uniquely well informed about each other and that has very high levels of mutual understanding and trust. That is, for example, one reason why we have been able to work across boundaries on counter-terrorism in ways that most of our allies with their more compartmented traditions have not yet achieved.
Unlike in Washington there has never been in the UK the tradition of daily personal intelligence briefings for the Prime Minister, or an elaborate Daily Briefing Book whose compilation became such a major load for the CIA Director and now for the new DNI. I read in the memoirs of Zbigniew Brzezinski, National Security Adviser to President Carter that he excluded Stansfield Turner, the CIA Director, and his intelligence analysts from the morning Presidential briefings because he wanted the meetings policy not intelligence focussed. The right instinct, but as we were reminded by Lord Justice Scott some years ago it is potentially dangerous to have intelligence material being interpreted without the intelligence experts having the opportunity to comment on whether it will bear the weight the policy makers might like. Winston Churchill on the other hand, after criticising the JIC in 1941 for its slowness in reporting that Germany would be prepared to attack Russia wrote “I had not been content for this form of collective wisdom, and preferred to see the originals (agents’ reports, decrypts etc) myself…thus forming my own opinions, sometimes at much earlier dates.” Prime Ministers making their own assessment of raw intelligence has of course fallen out of fashion since. Secret intelligence with its glimpses of sources and methods is certainly more interesting than much that passes over the senior desk and can become addictive. I think that the UK system now operating, with the JIC Chairman and his central analysts sending in briefly assessed highlights from the intelligence that has come in overnight, illuminating urgent or current business, backed up by more formal JIC assessment and meetings when needed, does avoid descent into the tactical and the risk of rush to judgement on the basis of what may be only a fragment of intelligence.
At the JIC level judgments often have to concern mysteries as well as secrets, using R V Jone’s helpful distinction between that which can in principle be known through secret intelligence, and predictions, mysteries that have to be divined such as what dictators might do in circumstances yet to arise. The advantage of getting the JIC to tackle such mysteries is that there is a great deal of experience to bring to bear, and a structured process with checks and balances. The risk is that speculative judgements may be taken as resting on more intelligence than they actually are, or are being over-interpreted through application of what R V Jones labels Crabtree’s bludgeon: ‘no set of mutually inconsistent observations can exist for which some human intellect cannot conceive a coherent explanation, however contrived.’ The answer is, as Lord Butler’s Committee recognised, for the form and language of assessments to be even more explicit about the limitations of the material on which judgements rest.
Lord Butler also helpfully supported a more professional approach to the career development and training of the wider community of analysts throughout government. Part of this must be the development of assessment methodology to be applied across the community. It was 50 years before Clausewitz wrote about obeying the laws of probability that the Rev Thomas Bayes in his ‘Essay towards solving a Problem in the Doctrine of Chances’ provided the rule to enable posterior knowledge to be brought to bear to revise the probability of an event. For example, the likelihood of a judgement being true can be revised on receipt of an new apparently confirmatory intelligence report by multiplying the conditional probability of the judgement being correct, assuming the information in the new report is actually true, by the probability that the piece of evidence in the new report itself is true. One practical lesson here for those directing intelligence operations is that if you only go looking for positive evidence of a proposition the probability of finding something in that category is rather high and therefore a spurious additional weight can be given to the relevant judgements. Intelligence officers have to be constantly on the watch for the inductivist fallacy, and set out consciously to keep testing the negative hypothesis rather than just build up information that confirms the prevailing wisdom
I have therefore always seen a parallel with the explanatory nature of scientific theory. The value of a strategic intelligence judgment, like a scientific theory, lies in its explanatory power, and thus its predictive power. But all scientific explanations – that is, scientific theories - are provisional and open to revision. So, similarly, successive intelligence assessments are liable to change, which does not make those who wrote them ignorant or knaves. Following R V Jones, a course in modern scientific method should be compulsory not just for those who analyse and use intelligence but for those who write about it.
Of course, there are many ways that we can still get it wrong. Intelligence is rarely evidence. For example the true signal gets lost in the noise. To cite the 18 th century Reverend Bayes again “the more observations you make with an imperfect instrument, the more certain it seems to be that the error in your conclusion will be proportional to the imperfection of the instrument made use of”. Perhaps for example the error came from information that was deliberately intended to deceive us, a perennial problem that RV Jones addressed: in any channel of intelligence in which you may be deceived – and I would add or whose interpretation is crucial – work down to a greater level of sophistication than the opponent has expected you to adopt, and bring to bear all possible channels of intelligence to check for consistency.
Another source of error may be that the information on which the model is based is no longer current. This leads to the inevitable distortions of predicting the road ahead from looking into the rear-view mirror. And the relationship between analyst and customer is part of the equation. For example predicting discontinuity could be held to be better for the analyst when matters are genuinely uncertain. If you are right that is great. If wrong there are any number of plausible reasons why the revolution or whatever might not have happened as predicted. But go the other way, and you are wrong and government is caught on the hop, then the stigma of ‘intelligence failure’ will be thrown at you. And in what Prof. Adams of University College London calls ‘bottom loop’ bias in risk management, these uncomfortable experiences tend to stay with us, and affect our internal perceptions of risk. Fear of failure is a powerful motivator and the more senior you are the further you have to fall, as Bernard Dixon pointed out many years ago in relation to military commanders. Analysts in my experience know the traps, although sometimes may lean too far over to avoid them and over the years, the JIC has probably been more often wrong by under- rather than over-estimating the pace of change in overseas regimes. Mirror imaging is a familiar phenomenon, and RV Jones gives examples from world of scientific intelligence. But it is also hard for civilised, often gentle, souls engaged in rational analysis to credit the effrontery of a dictator, as a re-reading of the Franks Report will remind us.
We may find therefore that prediction does not match reality because the model of human motivation being used to interpret the intelligence has built in inappropriate assumptions. Let me adapt a parable, attributed to Bertrand Russell in his philosophy lectures, to illustrate this aspect of the nature of explanation. Imagine a chicken farm, where the chickens spy on the farmer and intercept a message that he is ordering much more chicken food. The JIC of chickens meets. Is their key judgement that at last peaceful coexistence has come and the farmer is going to feed them properly in the future? Or is it that they are all doomed since they are about to be fattened for the kill? The choice of explanation may rest on prior, often implicit, assumptions about human behaviour not on the intelligence itself.
It is part of the burden of command that decisions sometimes have to be taken quickly on the basis of fragmentary evidence. An unexpected fact, as Bernard Dixon reminds us, is less easily absorbed than one that is already expected. So even where the intelligence is not sufficiently compelling to provide detailed answers the very process of analysis can be helpful. The timing of recent attacks in the UK may have come as a tactical shock because of absence of pre-emptive intelligence, but were not seen as a strategic surprise either in terms of method, target nor the background of those involved. Intelligence assessment had highlighted the need to prepare for exactly such an eventuality, and that thankfully had been absorbed and the response on the day was swift and highly effective.
To these problems for the analyst must then be added the potential effect of cognitive dissonance on the part of the customer for intelligence, when a person comes to be given knowledge or possesses beliefs which conflict with a decision he has made. Once a key decision has been made, and Bernard Dixon has some good military examples, the psychological situation of the decision-taker can change decisively. The historical experience is that there is a tendency then for less emphasis on objectivity, and more partiality and bias in the way that the decision taker views and evaluates the alternatives.
A question R V Jones raised from his WWII experience is whether in those circumstances it is legitimate for the analyst to resort to advocacy rather than present facts neutrally if they fear their warnings are not being heeded. Let me illustrate with a personal anecdote. Back in the 1980s it became the norm that aspiring civil servants would do management training alongside private sector executives. When I first did this I recall we had sessions on presentation skills – and one of my colleagues from a civil department walked out, saying that his duty to his Minister was to provide impartial unvarnished advice: to learn how to make a more persuasive case was to betray the ethics of his profession. PowerPoint is now the norm, and even the military commander can quickly summon up charts, graphs and statistics for the visiting Minister. Who could blame an analyst for advocacy faced with, say, a General Percival in Singapore refusing to accept the reality of the impending Japanese invasion? Or, a Secretary of Defense, as Robert Macnamara admits in his own memoir, resisting appreciation of the true state of affairs developing in Vietnam? R V Jones himself confesses to such lapses into advocacy on two occasions. But as he warns, this must never, ever, become the slanting of intelligence. The customer has to recognise that what the analyst is painting is an impressionist portrait, without the complete detail that you would find in a photograph. So what is included as the essential highlights and what is left out as distracting detail is a matter of analytical judgment. Customer and analyst alike need to be conscious of this.
3. The future form of the UK intelligence community
Let me now turn to my final section, on the organisation of the community, something that is never going to be simple since, as R V Jones remarked, input is by source and output is by subject.
There are precedents for tackling organisational problems with such characteristics. I was involved in the mid-1980s and again in the 1990s in redesigning the central organisation for defence. We had to tackle the dilemma of reconciling the existence of separate Services, Royal Navy, Army and Royal Air Force with the joint nature of planning and investment in, and ultimately the generation of, military power. The failed Canadian attempt in the 1980s to combine their Services shows the penalties of ignoring the different characteristics of each service, adapted to their sea, land or air environment, and the part that their traditions play in maintaining operational effectiveness and recruiting alike. Equally, once they reached a senior level as Sir Ian Jacob and Lord Ismay had noted as early as 1963 based on their deep wartime experience, officers had to train and work together to develop the best defence answer, and there had to be a Chief of Defence Staff with authority to give a ‘purple’ perspective. And we could not afford the inefficiency, waste and military ineffectiveness of having separate procurement, repair, supply and logistic chains, and separate IT and communications systems let alone separate operational planning (a lesson should have been clear to Britain at least since the Agadir crisis).
My approach to such issues has been to look on organisations not as machinery described by lines on charts, but as adaptive organisms, evolving over time to fit their environment. Changes that are seen to be going with the grain of history tend to produce longer lasting results than the discontinuities introduced by short cuts, where the law of unexpected consequences tends to apply. The words of the Duke of Cambridge, whose statue stands outside the Old War Office, hover in the air: “ There is a time for everything, and the time for change is when you can no longer help it." But we do not normally have the time to wait for such realisation of inevitability to dawn unaided on the staffs. So to accelerate the process we need to show that the necessary changes fit a narrative that explains convincingly where the institution has been, how it is evolving and why the time has come to accelerate the pace of change. And after the Falklands Campaign, as the Principal Private Secretary in MOD, I was certainly convinced that if the mission is defence then the watchword had better be organise for war and adapt for peace rather than the other way round. The 1980s MOD reorganisation was rightly presented as further steps in accelerating an evolutionary process that had been under way for generations. Indeed I recall being delighted to find the main line of development anticipated in the Lord Esher’s Report of the War Office Reconstitution Committee of 1904.
Can we similarly discern what Professor Hennessy calls ‘the thin wisps of tomorrow’, in Braudel’s phrase, in the evolution of organisation for secret intelligence?
It is not hard to spot what some of the key factors might be, and I touched on many of them in the first section of my talk:
- intelligence in its widest sense is more than ever needed to reduce ignorance and encourage rational decision-taking given the changes taking place in the world
- without secret intelligence we will not understand sufficiently some important threats that face us
- but open sources are also becoming relatively more important and need more investment so that we have a rounded intelligence picture
- increasingly tasks of research, analysis and assessment, and strategic warning, are falling to the wider intelligence community
- the community is serving a wider group of customers; they are rightly demanding rapid and secure access to intelligence in all its forms, and expect it to be cross-referred, collated, assessed and distributed and for rapid agreement to be given to its use for action
- differing demands arise for supporting the strategic, operational and tactical levels, and different considerations apply for establishing the most effective international links at each of these levels
- the modern threats of international terrorism, trans-national crime and proliferation has led to blurring of distinctions between home and overseas theatres of operations; agencies must work easily across the boundaries
- the intelligence community needs to have access, under safeguards, to a range of data to allow previously unexploitable sources of information to generate pre-emptive intelligence
- The need for secure communications across the community and with the customers create issues for the community as a whole to tackle
- in the intelligence world, as in defence, governments are stewards of national capability and strong central direction is needed to ensure that capabilities are built and maintained for the future.
All these factors point towards the need to continue to develop the UK intelligence community, going beyond the three Secret Agencies to include the central assessment staffs in the Cabinet Office, and the Joint Terrorism Analysis Centre and the Defence Intelligence Staff, respecting their position as an integral part of defence capability.
The community has a central forum, the Joint Intelligence Community, and it now has a single head, the Intelligence and Security Coordinator, concerned both with the meaning of the intelligence provided for government, and thus its quality, and also the long term health of the community. The Coordinator is supported by a small planning, resources and programmes central staff, drawn largely from the community itself, in order to formulate the long term strategy for the community and the annual statement of requirements for secret intelligence approved by the Ministerial Committee on Intelligence chaired by the Prime Minister. The staff support him in negotiating the Single Intelligence Account with the Chief Secretary to the Treasury, and in assessing the effectiveness of the community in meeting the requirements of its customers. What needs to be done here seem reasonably clear, as planning becomes more sophisticated and the agencies’ own performance management systems evolve. There are many parallels with the development of planning in MOD and I am sure that the new Coordinator is well placed to know how to accelerate the process.
There are also lessons of what not to do that can be learnt from the MOD experience. In the early days, as Peter Nailor would no doubt recount if he were with us, the very small Stories Gate central staff was bolted on to large traditional single-Service structures. The centre expanded greatly to compensate, leading to tensions and overlap. Now the centre is slimmed down and the overlap has been removed. There is central capacity that need to be built up in the intelligence world, not least to deliver compatible IT. But the principle of subsidiarity must reign, and the centre must be kept small. In that way the advantages of this evolution can be obtained without cutting across the statutory personal operational responsibilities of the Agency Heads, and the general accountability to Parliament of their Secretaries of State. The Coordinator is, like the Cabinet Secretary, a servant of government as a whole.
Another key lesson from the defence experience is the value of retaining separate Armed Services. I do not see this evolutionary trend in intelligence organisation ending up with Agency merger. Collection of intelligence is a highly specialised activity that involves being able to recruit, train and motivate very different and unique kinds of people. Although from time to time cost-cutters cast glances at the overheads of running three Agencies the loss of effectiveness would be catastrophic. So the process under way, and that needs to be accelerated, is the combining of strong central strategic direction and planning, joint support and logistic functions, information systems and so on with but highly devolved operational management.
Let me cite one example. We know that the protective security and counter-terrorist customers want value-adding material brought together by subject, with the assurance that if there is any relevant material anywhere in the data bases of the entire intelligence community then that will have considered. That is the thinking behind the creation in 2003 of the Joint Terrorism Analysis Centre. Although the Head of JTAC is responsible to the Director General of the Security Service, JTAC operates as a self-standing organisation comprised of representatives from eleven government departments and agencies working within a circle of trust. JTAC analyses and assesses all types of intelligence relating to international terrorism, at home and overseas, and produces assessments of threats and other terrorist-related subjects for customers from a wide range of government departments and agencies. The work is conducted on a fully collaborative basis, with the involvement and consensus of all relevant departments.
I am conscious that other countries are also adapting their agencies and forces to confront international terrorism and crime. There is no ‘right’ organisational model for intelligence and law enforcement that will fit all nations with their different traditions and legal systems. But there may be general principles that are common. Let me end with some reflections on what I would characterise as the concentric circles of trust:
- In an inner circle we would find a community of trust between the intelligence agencies, internal and external and sigint services. Whatever national organisational geometry, there would be no legal barriers to cooperation in this circle. To borrow a phrase from Senator Pat Roberts, Chair of the US Senate Intelligence Committee, the watchword would be intelligence sharing rather than compartmented intelligence access. There would be the trust necessary to make joint operations commonplace, spanning where necessary domestic and overseas. And you would see young officers training together, and taking part in cross-postings, really understanding what their colleagues from other disciplines can bring to the party
- That community would also be authorised to use the most modern techniques for acquiring and analysing the mass of communication, travel, border and other data needed to deal with terrorism and organised crime. There would be effective arrangement for Parliamentary and judicial oversight, whose form would obviously depend upon national constitutions, but sufficient to provide public confidence that the use of these intrusive powers and the ability to share such information with overseas partners was properly regulated and proportionate
- Widening to the next circle, we would see that secret intelligence community working hand in glove with the relevant police services, sharing operational detail within a community of trust between those selected and vetted to work on these sensitive operations. It would be for the police to use their independent authority to enforce the law drawing on their close relationship with the community they serve, but targeted carefully by the secret work of the intelligence community using their most advanced methods. In this way the power of modern intelligence is harnessed for public protection, but without raising again in Europe the spectre of the secret policeman and the erosion of the rule of law
- And in a third circle of trust, the intelligence and police communities would have structures that enable them to share information and cooperate closely with the key staff in the Armed Forces and Defence Ministry, in the Foreign Ministry, and in Transport, Health, Environment and all those Departments engaged in ‘homeland security’ to provide the support for policy making and overseas operations and assistance programmes, and to guide investment in improving national resilience, and reaching out to the private sector who now run most of the critical national infrastructure that keeps modern life going
- And to recap briefly on my earlier argument, these circles of trust form the secure national base on which sound international arrangements can be developed. Strategic assessments can be shared to guide the formation of collective security strategy and international agreements, timely operational warnings and terrorist assessments can be shared for public protection through the JTAC model, and not least the tactical pursuit of terrorist networks improved through bilateral relationships between the secret intelligence services of like-minded nations working together on specific operations and campaigns.
I do not want it to sound easier than it is. Intelligence structures, processes, co-ordination mechanisms that may have worked for nations in the past have to be rethought against the new threat and how it might develop. That is what the UK has had to do. It can involve changes that break with traditions. It will cost more money. But the status quo will no longer do.
In conclusion, let me repeat my thanks to Gresham College for this platform. Our secret intelligence community serves us well and I feel privileged to have been associated with it in different ways over so many years. I hope these affectionate reflections may amuse, and perhaps encourage it in its accelerated evolution.
© Sir David Omand GCB, Gresham College, 20 October 2005