A Rejoinder to Bernard Keane’s ASIO Claims

During a 2006 Monash postgraduate class on intelligence analysis our adviser made several observations on how the Australian Security Intelligence Agency (ASIO) is misrepresented and misunderstood. The ‘S’ stood for domestic security not secrecy. ASIO had an accountability and audit regime at multiple levels: legislative limits, the Treasury budget process, appeals processes, external audits and supply contract review, and reporting to the public and to bipartisan government committees. Australia’s intelligence resources were mo stly deployed in military agencies for signals intelligence. Finally, media coverage of ASIO rarely evolves to the sophistication seen in the United States and the United Kingdom.

Bernard Keane’s Crikey article ‘The Answer is ASIO‘ (24th February 2010) risks continuing this trend in media coverage of intelligence issues. I want to illustrate below how Keane’s own arguments can be interpreted as having their own “deeply-flawed logic” in his accusations of Labor’s “security propaganda.”


Keane appears to not distinguish the appropriate function of an intelligence system, and ASIO’s particular role, from issue- and party-driven political debates. Many of Keane’s comments on the new Counter-Terrorism White Paper, Kevin Rudd, Peter Garrett  really concern how governments can time legislation and frame announcements. ASIO and other intelligence agencies do not control such gambits, and may just bepart of a wider consultation to formulate new policies. Rather, Keane’s examples illustrate what analysts fear: the politicisation of the intelligence ‘product’ in the media spin-doctoring cycle.

Next, Keane contends that ASIO has continuity from the Howard to Rudd governments in an expanded operational budget. Yet he provides little context or
analysis of ASIO’s past budgets to evaluate his claims.

The historical evidence suggests a more complex picture. It also shows the limits of Keane’s historical correlation between Australian political parties in government and ASIO’s powers. ASIO’s budgets stalled in the Howard Government’s first and second terms as the agency struggled with program reviews, the loss of key staff and corporate memory, and the legacy costs of old computer systems. The post-September 11 rise in ASIO’s budget attempted to deal with this ‘weight of the past’ through organisational change and renewal; with an uncertain security environment where there were variations in estimative assessments; and with the influx of a new generation of intelligence analysts. In the US context, Amy Zegart illustrates how this more complex picture goes beyond a simple government-agency nexus.

It is certainly possible, as John Mueller argues in his book Overblown (New York: The Free Press, 2006), that policymakers might have been affected by a post-September 11 ‘security bubble’ similar to Robert Shiller’s early warnings in 2000 of a US housing bubble. Large capital expenditure items like “the shiny new giant headquarters being built for the agency in Campbell  in Canberra” would justify a budget rise with the costs amortised over a 25-30 year period. Without
analysis we don’t know, and have no way to test Keane’s claims of institutional capture. One way to possibly do so would be to conduct a time-series analysis of ASIO’s budgets, to weight the expenditure items, and to compare it with changes in the security environment and political processes. This time-weighted factor model would give a more sophisticated answer than a linear extrapolation where correlation may not be causation.

Keane’s belief that ASIO is “unnacountable” is news to me. I agree ASIO’s public annual reporting is brief compared to its US counterparts, but this is the norm for Australian agencies. The Parliamentary Joint Committee on Intelligence and Security reviews ASIO’s expenditure, financial budget, and annual reporting. The Australian Security Intelligence Organisation Act 1979 sets legislative requirements. There is significant debate in analyst training, in legal case judgments that use existing laws, and in the academic literature on counterterrorism and intelligence, on personal liberties, privacy and the ‘surveillance state’. Whilst it’s an
important issue, Keane gives an oversimplistic view.

Keane’s claims about people smuggling and organised crime need further clarification. Both are trans-border security issues which have a domestic security component or effect. Thus, they are already part of ASIO’s remit, which Keane might have seen had he looked at the security issues raised in ASIO’s past budgets. Both have been discussed in the academic literature in counterterrorism and security studies for 10-to-15 years, whilst fellow ‘middle powers’ like Canada have
developed significant policy expertise. Whilst politicians may engage in “political risk”, events such as Sri Lanka’s destabilisation after the Tamil Tigers now change the estimative threat assessments that guide the agencies.  Claims that people smugglers are the “Bond villains of border protection” appear to overlook the sophisticated ways that such networks do operate, and of recent anthropological investigations to document this.

In his critique of the new Counter Terrorism Control Centre, Keane blames the government that “too much inter-departmental co-ordination is never enough.” However, CTCC-style activities have already been funded for several years, so the White Paper may formalise this. Keane also overlooks why such
inter-deparmental coordination is vital for counterterrorism, transnational security issues, and is the bedrock of national security strategy and security clearances. Take a potential security risk – like people smuggling. Apart from ASIO several other intelligence agencies may be involved, depending on the events. The Office of National Assessments has been monitoring Indian Ocean people smuggling for at least several years. Defence Signals Directorate may provide signals intelligence, as might the Defence Intelligence Organsation, and perhaps even liaison with the US-based National Security Agency and Defence Intelligence Agency
if satellite intelligence needs to be shared (DIO and DSD would at least be kept ‘in the loop’). This occurs at the collections, processing and analysis stages of the intelligence cycle, before the ‘product’ is disseminated to the relevant policymakers and government agencies. Thus, inter-departmental coordination is critical,
particularly for time-critical events that involve ambiguous information.

Finally, Keane observes: “Part of the problem is indeed the human incapacity to accurately assess comparative risk, and most people’s inability to understand the

sometimes counter-intuitive basics of risk management.” Well, sure: the difference is that intelligence analysis involves Gregory Treverton‘s distinction between puzzles (like missile numbers that are solveable); mysteries (like new terrorist groups); and mysteries-plus (like complex, global, adaptive networks). Gaussian risk management models do not work for Extreme-Value events. Australian ex-analysts like Paul M. Monk might also emphasise the cognitive elements in writing and communicating estimative assessments to decisionmakers. Intelligence studies scholars like Treverton, Loch Johnson, Richard K. Betts, and Michael Herman discussed this for decades, before Nassim Nicholas Taleb and Cass Sunstein came to recent attention in media and policy circles.

Keane’s problem explains why intelligence agencies exist: to anticipate such risks through comparative and estimative assessments, and to develop long-term human and institutional capacities. ASIO or another alphabet intelligence agency may be the answer — to a different question than Keane asks.