Contestability: the key to more successful intelligence analysis
4 Aug 2016|

Image courtesy of Flickr user theilr

During the festive season of 1978, in an underground corner of the Pentagon, a group of Defense Intelligence Agency analysts was burning the midnight oil. For some months, they’d been monitoring the deployment of Chinese forces from several military regions to the border with Vietnam. Examining the full spectrum of satellite imagery, electronic and communications intelligence, the conclusion was finally drawn that this troop concentration amounted to a sabre-rattling exercise, aimed at warning Hanoi over its attacks on the Khmer Rouge in neighbouring Cambodia.

Meanwhile, half a world away in Canberra, at the Joint Intelligence Organisation, a small team had pored over the same raw material, with starkly different results. Led by eminent sinologist David Cross, along with a young military intelligence captain named Alan Dupont, the Australian analysts predicted that a Chinese invasion of Vietnam was imminent. In their view, the ramping up of rhetoric in the Beijing press held the key, particularly references to launching a ‘counter-attack in self-defence’ (zi wei huan ji).

The Pentagon considered this contending view and, on the balance of probabilities, agreed that an attack was more likely than not. When China launched the Third Indochina War several weeks later, the technical eyes and ears of the West had been directed to the right place at the right time. Canberra had contested the received Washington wisdom and valuable information was derived for both ANZUS partners.

But intelligence analysis is more often judged by its failures than its successes. Both the recent Chilcot Inquiry in Britain and France’s parliamentary review of its intelligence services, following the 2015 extremist attacks in Paris, confirmed that intelligence is fallible. The 832 pages of the US congressional inquiry into the 9/11 terrorist attacks reached a similar verdict in 2002.

With the clear and present terrorist danger facing Australians, it’s worth considering our own experience of perceived intelligence shortfalls triggering government-mandated reviews. The most notable is that conducted by senior diplomat, Philip Flood, in 2004 following the Howard government’s decision to invade Iraq.

Subjecting analytical judgments to rigorous challenge was critical to effective intelligence, according to Flood, with contestability the leitmotif of his review. An entire chapter of the 250-page report is devoted to this idea.

In the Australian context, however, contestability would appear easier said than done. After all, only two government agencies have been apportioned the role and responsibility for providing official policy-makers and government decision-makers with all-source, finished intelligence product—the Defence Intelligence Organisation and the Office of National Assessments.

While calling out the need for contestability on the one hand, Flood drew, on the other hand, a demarcation line between the content and target audience for DIO and ONA intelligence assessments—a separation that’s hardly conducive to the mutual contestability of ideas and judgements between the two agencies.

The challenge of contestability is compounded by the very nature of the intelligence community workforce. While aspiring analysts may enter the recruitment funnel from diverse backgrounds, offering a wide range of knowledge and experiences, the excruciatingly involved security vetting process sees many fall by the wayside, with a disturbingly like-minded cohort dripping from the tube’s end. Under such circumstances, groupthink becomes a very real issue.

And relying on counterparts in allied intelligence services for pressure-testing Australian judgments will be useful only when all parties share common interests—like the Third Indochina War in 1979.

It may be that analysts in think tanks outside the government bubble will assume added significance as Pentagon-style ‘murder boards’. Organisations like the Australian Strategic Policy Institute and the Lowy Institute could perform such a function for most strategic analysis, complemented by academics and commentators with unique expertise in arcane areas of growing importance to Australia, like Mexican criminal cartels and African terrorist groups.

If contestability ran like a thread throughout the 2004 Flood Review, it was conspicuously absent from the 2011 follow-up review by retired senior public servant, Robert Cornall, and Melbourne University ethicist Rufus Black. The concept appears nowhere in the report’s relatively brief and comparatively anodyne pages. At least in its publically-released form, Cornall–Black is as mute as Flood was magisterial. One could be forgiven for thinking Australia’s national intelligence needs were being met by a highly evolved and impressively efficient community of practice. Nothing to see here, move along folks.

That may be the case; time will tell. A further review of the Australian Intelligence Community will be due in the life of the new Turnbull government, with Flood having recommended that the six agencies be subject to external review every five to seven years. Contestability should once again appear on the agenda, with practical and innovative recommendations on how to challenge conventional thinking in the face of unconventional threats.

The Australian Intelligence Community has a proud history of playing a constructive sounding board role for the three-letter agencies of Washington and other five-eyes counterparts. It’s time to harness that talent for our own purposes—both within and without the confines of government service. Our national security depends on it.