Australia's intelligence and security agencies: Its bureaucratisation, not politicisation, that is the problem

A number of fundamental structural, philosophical and professional issues underly the longstanding and recurrent problems plaguing parts of the Australian Intelligence Community.

Virtually every news or opinion article on the recent criticisms made by Lieutenant Colonel Lance Collins has missed the essential points concerned.

They have also mostly failed to note the detailed Australia Defence Association submission to last year’s parliamentary inquiry into the intelligence services — which listed several indicative intelligence and policy failures over the last 25 years and discussed their apparent causes.

These examples included the marked and vicious bureaucratic resistance in some quarters to the 1998 and 1999 intelligence estimates concerning East Timor produced by Lance Collins and his intelligence staff at the deployable joint force headquarters.

Effective debate on the real issues affecting our intelligence and security agencies is unfortunately being sidetracked or obscured by mostly irrelevant fixations as to whether the Howard government has, or has not, “politicised” intelligence reporting.

The real problem is not “politicisation” in the party-political sense but that intelligence assessment and reporting is too often bureaucratised and subjective even without any “political” interference. The ADA argument on this point was actually quoted in the 2004 Joint Parliamentary Committee report.

Under governments of both persuasions, the bureaucratisation and subjectivity largely stems from too many of the middle and senior management in our intelligence and security agencies coming, inappropriately, from policy-making backgrounds rather than from the ranks of intelligence career professionals.

Even if only unconsciously, these outsiders (especially diplomats and those from Defence’s international policy area) have a marked tendency to slant intelligence collection, assessment and reporting towards desired policy outcomes, rather than employ the objective, sceptical and often lateral approaches required for professional intelligence work.

Another significant cause of the problem is insufficient appreciation of the two main intellectual models underlying all intelligence work: intelligence analysis and intelligence estimates.

Intelligence analysis involves collecting, integrating and analysing information (clues) to produce intelligence (or the story of what has, or probably has, happened and informed speculation on what might happen in future).

Intelligence estimates (and their continual re-estimates) on the other hand involve the intellectual process whereby an adversary’s intentions, possible courses and broad plans (the future story or possible stories) are predicted through a logical series of assumption-based steps and deductive and inductive logic processes.

Intelligence analysis informs the estimate process (and confirms or denies which possible future story applies) but does not necessarily guide it.

This means that intelligence estimates can be used in the many situations where there is insufficient or conflicting information for analysis.

Intelligence estimates as a process are also much less vulnerable to unconscious or conscious subjectivity influencing their conduct and results.

Now with enough time, and the requisite training and professional direction, you can eventually turn most competent and intelligent people into intelligence analysts. This includes clever people recruited laterally.

The same does not hold, however, for intelligence estimates. The skills required required need development over a much longer and more varied experience base.

The ability to conceive, direct and manage the production of intelligence estimates is essentially a through-career intelligence professional skill.

Most of the outsiders (and relative intelligence amateurs) brought into management positions in our intelligence and security agencies usually cannot undertake intelligence estimates and generally do not understand their key role and importance.

Such amateurs therefore generally downplay the value of intelligence estimates. In many cases, they simply fail to understand or acknowledge the intellectual or professional concepts underlying intelligence estimates.

This leads to over-reliance on personal knowledge and often to over-confidence by individuals in management positions.

This in turn leads to an over-reliance on intelligence analysis as the sole professional tool to be used, with a resulting much greater vulnerability to subjectivity overall.

Non-intelligence professionals are also prone to demand forensic levels of proof rather than accept the ‘balance of probability’ inherent in most intelligence work.

Such risk-averse behaviour is especially common among staff from policy-making backgrounds.

These factors lead directly to intelligence failures or to situations where senior officials, within the agencies or among their customers, refuse to believe intelligence predictions that challenge or contradict their own personal views or prejudices.

The clash between Lieutenant Colonel Lance Collins and some senior managers at the Defence Intelligence Organisation (DIO) exemplifies many of these problems.

Collins, while somewhat stubborn, is a through-career intelligence professional. He is immersed in an objective institutional culture and highly skilled in the intelligence estimate process.

As an ADF officer he is also primarily and heavily customer-focused, especially towards the operational needs and safety of the main end user of his work — the men and women of our defence force.

As an Army intelligence officer, Collins has also acquired a broad range of professional intelligence skills.

As a rough rule of thumb, the Army’s intelligence officers generally have a much broader professional understanding of the many separate specialist types of intelligence work than those who spend their career within one or two intelligence agencies or specialisations.

This is one reason why all six intelligence and security agencies employ them on secondment (and recruit them laterally).

Collins' criticism of the failings of DIO and ONA are therefore profoundly influenced by his wider experience and broader perspective.

Many senior DIO managers on the other hand come from bureaucratic policy-making and administrative backgrounds. They usually have no apparent professional qualifications, specialist training or prior experience in intelligence work.

DIO is, in fact, the only major defence intelligence agency in the western world headed by a civilian official (and one with no real intelligence background). He would not, for example, qualify for full membership of the relevant professional body, the Australian Institute of Professional Intelligence Officers (AIPIO).

For many years DIO has been heavily criticised for its focus on satisfying high-level policy makers (both bureaucratic and political) and its frequent and significant inability to adequately support the ADF.

This failure is due to a variety of historical reasons steeped in the arcane bureaucratic mess of Australia’s higher defence management overall.

These markedly different experience bases, professional perspectives and institutional culture imperatives underly the inability of many to even understand the nature or seriousness of Collins’ claims.

The key point he highlights (as the ADA did last year) is that it is not really a matter of “politicisation” of intelligence in the party-political sense, but one of its vulnerability to subjectivity from bureaucratic policy makers.

Intelligence should be only one of the inputs that Governments consider when deciding matters of strategic policy and international relations.

As an input, however, it must be as unadulterated as possible until it is provided to the Government. Only then can it be effectively weighed against the various policy considerations also applying. This is the nub of the problem.

Finally, some commentators have wrongly suggested that ADF strategic and operational intelligence estimates have somehow ‘strayed’ into areas beyond their responsibility. This is nonsense on a number of grounds.

First, modern intelligence systems, processes and networks are so integrated that information collected and analysed at any level may have ramifications at other levels.

Second, DIO and ONA have usually failed to produce the strategic intelligence estimates required. Operational-level estimates have had to then be used to partially fill the vacuum involved in order for campaign or crisis planning to occur.

Third, if the ADF can successfully apply the intelligence estimate process at the tactical, operational and strategic levels, and comparable overseas intelligence agencies are able to do so, why are DIO and ONA so inept in this regard?

If the Flood Review does not explore the many issues involved, or is not permitted to recommend the reforms required, then a Royal Commission will eventually be needed.

One may be needed anyway to restore public (and ADF) confidence in at least two of our intelligence and security agencies.