2011 Independent Review of the Intelligence Community

The 2011 report of the Independent Review of the Intelligence Community (the Review) was the first comprehensive review of the Australian intelligence community since the 2004 inquiry conducted by Mr Philip Flood AO. 

Appendix 1

Reasonable expectations of intelligence

What can the government reasonably expect the Australian Intelligence Community will be able to tell it?

This question comes into sharp focus when events come as a surprise. Most recently, uprisings in the Middle East give rise to the question: Why were these events not foreseen? Given the task of the Review to assess the relative performance of the AIC and the expectations government should have of it in the future, it is important to set out our basis for answering these questions.

Over the years, senior members of the intelligence community – both in Australia and overseas – have been at pains to point out the limits to what intelligence can do.

In a speech to the Lowy Institute, the Director-General of ONA quoted a former US intelligence officer who drew a distinction between puzzles and mysteries:

‘Puzzles can be solved: they have answers which only need to be revealed – how many intercontinental ballistic missiles did the Soviet Union hold? But a mystery, Treverton says, “poses a question that has no definitive answer because the answer is contingent; it depends on a future interaction of many factors, known and unknown. A mystery”, he says, “cannot be answered, it can only be framed, by identifying the critical factors and applying some sense of how they have interacted in the past and might interact in the future”’[19].

The question the Review has asked is: What determines whether something is theoretically knowable? To help make that distinction and to determine what are reasonable expectations of government, we distinguish between the expectations of what can and what cannot be known about existing and future states of affairs.

Existing states of affairs

Existing states of affairs – which range from facts like how many naval vessels a country has to what its leader’s intentions are – are theoretically knowable. It may prove very expensive, difficult and time consuming or even practically impossible to obtain that knowledge. What is important is that, because the state of affairs exists, in theory it could be known.

With this type of information, the intelligence community should be able to advise government about the difficulty, cost, risks and probability of success of different strategies to obtain it. In these situations there are two reasonable expectations government can have:

  • First, government can reasonably expect that the intelligence community will be able to provide it with a fair representation of the trade-offs involved in seeking to obtain the information about existing states of affairs, and
  • Second, government can reasonably expect that the intelligence community will be able to undertake its collection and analytic efforts in line with the trade-offs it identifies.

However, this expectation is subject to a critical qualification. In an open society, there is an additional constraint on what can reasonably be expected. The civil and political freedoms of our open society (including the right to privacy) have to be respected unless there is evidence that an individual or group is planning harm of a sufficiently grave kind to justify the limitation of those freedoms.

Respecting those freedoms means that the state cannot seek to know the intentions of all individuals or of such a large number of them that it would be hard for any individual to escape detection by the state (as occurred in East Germany).

The state cannot begin by seeking to know the intentions of every individual who may have malign intent. It must begin by operating at the level of the established social system. Therefore, consistent with an open society, the state can seek to identify groups or movements that might pose a threat. If such threats can be demonstrated then the state can move to focus on individuals of concern within those groups or movements.

Practically, this approach means that it may be very difficult to detect and stop small groups or lone individuals who are not part of some larger movement or group planning to carry out a violent plot. It is possible that the intelligence agencies could detect those individuals but it is also quite possible that they would not. The harm such individuals could cause is the price of an open society.

Therefore, a government can broadly expect that an intelligence community resourced for the task will be able to identify:

  • Groups and movements from which threats could emerge, and
  • Individuals within these groups or movements intent on causing harm.

However, as with any existing state of affairs, what is reasonable in any particular case will depend on the trade-off the government is prepared to accept between the challenge, risk and cost of pursuing the target.

What it is not reasonable for a government committed to an open society to expect is that the intelligence community will be able to identify and stop all security threats.

The importance of not having a zero risk expectation is not only as a safeguard to being an open society but also to avoid unpleasant surprises, over-reaction and misplaced quests for accountability as the Director General of the British Security Service explained very clearly in a speech he made in 2010:

‘Our aim is to reach a position of assurance where any threat is identified and action taken to disrupt it before any harm is done, and particularly before there is an imminent danger to the public. This is of course easier said than done, and will never be fully achievable, but it is the aim.

‘It is interesting to note in this context that in the last ten years what might be called a “zero tolerance” attitude to terrorist risk in Great Britain has become more widespread. While it has always been the case that the authorities have made every effort to prevent terrorist attacks, it used to be accepted as part of everyday life that sometimes the terrorists would get lucky and there would be an attack. In recent years we appear increasingly to have imported from the American media the assumption that terrorism is 100% preventable and any incident that is not prevented is seen as a culpable government failure. This is a nonsensical way to consider terrorist risk and only plays into the hands of the terrorist themselves. Risk can be managed and reduced but it cannot realistically be abolished and if we delude ourselves that it can we are setting ourselves up for a nasty disappointment’[20].

Future states of affairs

When it comes to future states of affairs, a very helpful approach can be built on the analysis of Nassim Taleb – author of ‘The Black Swan’ – who is particularly noted for his analysis of what can and cannot be known and predicted. In a recent article with Mark Blyth in Foreign Affairs, Mr Taleb addressed the question of what intelligence can know.

Nassim Taleb’s analysis is founded on the following fundamental distinction:

‘Humans simultaneously inhabit two systems: the linear and the complex. The linear domain is characterised by its predictability and the low degree of interaction among its components, which allows the use of mathematical methods that make forecasts reliable. In complex systems, there is an absence of visible causal links between the elements, masking a high degree of interdependence and extremely low predictability’[21].

When it comes to future states of affairs that are in the linear domain – for example, what a small terrorist cell is plotting to do next week – the expectation is essentially the same as with existing states of affairs. The key difference to existing states of affairs is that the intelligence community has to add its advice to an assessment of the probability that events will unfold as anticipated.

What is knowable about future states of affairs that belong to the domain of complex systems is a very different question. Mr Taleb’s analysis is very helpful because it highlights that the extent to which the future is knowable is dependent on how open a society is.

Where a society is open – in other words, where there is a high level of civil and political freedoms including freedom of speech and association and the rule of law – it is possible to observe and map how the social, political and economic system works and, therefore, to form some conclusions about the likely directions of that society.

In closed societies – repressive states such as totalitarian dictatorships, theocracies and autocratic monarchies – the system is hidden. For example, the size and even the existence of opposition groups may be very veiled. It will be unknown how many citizens would join any movement. It is possible to speculate about such situations but such speculation will have high degrees of uncertainty attached to it and may prove very wide of the mark if there are social dynamics that are not visible or even in existence because of the level of repression.

These highly constrained states appear stable when, in fact, they may be extremely fragile and volatile. In the words of Taleb and Blyth:

‘Complex systems that have artificially suppressed volatility tend to become extremely fragile, while at the same time exhibiting no visible risks’.

With such fragility, it will be unknowable what exactly will tip the system over the edge, as we saw very clearly with events that triggered the Arab Spring.

So it is not reasonable for government to expect the intelligence community to predict the timing, cause or nature of events like the Arab Spring or the collapse of the Soviet bloc. However, what government can reasonably expect is for the intelligence community to provide advice about:

  • The degree of volatility – for example, the AIC should be able to provide warnings to government about countries that are particularly volatile
  • What factors would be likely to increase the volatility – for example, domestic pressures (such as more repressive measures) or outside influences (such as a rise in food prices or external attack), and
  • What would make the society more knowable – for example, AIC agencies should be able to identify factors that would enable intelligence officers to better inform the government about the system, such as the rise of the free press.

Finally, reasonable expectations of intelligence change when a closed system starts to transition to an open society.

Initially, the government will need to look to open source reporting while the AIC develops assessments about the society and how it operates. However, as the society begins to operate, it will be possible for the AIC to provide more accurate and long-term assessments. Agencies should then be able to make assessments about individual groups and power balances and how they might play out.

While it is important to have a clear framework for what can reasonably be expected of the intelligence community, it is important to recognise that, as the Rockefeller Commission found in 1995:

‘Good intelligence will not necessarily lead to wise policy choices. But without sound intelligence, national policy decisions and actions cannot effectively respond to actual conditions and reflect the best national interest or adequately protect our national security’[22].

Intelligence reports provide information and intelligence assessments offer judgments based on the best information then available to assist decision makers. But intelligence is only one factor decision makers take into account in making decisions.

Other factors include government policy, competing demands, legal and international obligations and so on, together with the decision maker’s own judgment about the issue under consideration.

Availability of intelligence

Generally speaking, intelligence was historically well hidden and hard to get.

One of the major changes affecting intelligence agencies today is the ready availability of huge amounts of information which would previously have been difficult to obtain or simply not available.

Two factors contributing to this situation are the successful collection of sigint and the evolution of the internet.

It has been said that this new information environment requires flexibility in intelligence collection and a major shift from analysing what has been collected to analysing what to collect:

‘We need to make tough decisions about which haystacks deserve to be scrutinised for the needles that can hurt us most. And we know in this information age that there are endless haystacks everywhere’[23].

In 2004, the Flood Report defined intelligence as covertly obtained information. Open source information was a supplement for assessors. In 2011, the Director-General of ONA defined intelligence more broadly as useful information irrespective of source.

Arguments can be made in support of both statements, but they illustrate the major shift in the availability of valuable information that we believe will have a significant impact on our intelligence agencies over the next five to seven years.

That shift also highlights the fact that any helpful definition of intelligence has to go well beyond the collection and dissemination of useful information. Intelligence has to be clearly linked to the needs of individual customers and tailored for the purposes to which they can apply it. Otherwise, the recipients of intelligence reports will be overwhelmed by the volume of material they receive which will effectively reduce or even negate its ultimate usefulness for them.