Thursday 26 Nov 2020 | 01:01 | SYDNEY
Thursday 26 Nov 2020 | 01:01 | SYDNEY

What are intelligence agencies for?


Sam Roggeveen


12 December 2007 15:03

If you measure the performance of intelligence agencies by the number of predictions about major political events they get right, their record is pretty awful. Last weekend, Sunday Age journalist Jason Koutsoukis wrote a helpful list of the CIA's history of bungling, which he argued is also endemic to our intelligence community:

Over the course of its 60-odd years, it failed to notice the Soviets getting the atom bomb in 1949, was blind to the Chinese invasion of North Korea, the 1956 revolt in Hungary, and Fidel Castro's Cuban revolution...Then it missed the Iranian revolution of 1979, and Saddam's invasion of Kuwait in 1990.

The list of howlers goes on: the Arab-Israeli war of 1973, the Soviet invasion of Afghanistan in 1979, the explosion of an atom bomb by India in 1998, the September 11 terrorist attacks, and then its bogus intelligence on Iraq in 2002-03, most of which was based on the cock-and-bull fantasies of a source named "Curveball". Failing to predict the collapse of the Soviet Union is the one I like the most. This was the very thing the entire intelligence apparatus of the West had been dedicated to destroying for 50 years, yet no one had any idea the USSR was about to fall in on itself.

By that measure, Koutsoukis was right to call for the abolition of Australia's intelligence agencies. But unfortunately for Koutsoukis and his profession, when it comes to making predictions, intelligence analysts have something in common with journalists: they are both really, really bad at it. In fact, research shows you're probaby better off tossing a coin than trusting the prediction of any kind of political expert, be they a journalist, intelligence analyst, academic or think tanker.

It's a good thing, then, that making accurate predictions is not actually what intelligence agencies are for.

No amount of electronic eavesdropping or recruitment of other governments' officials is going to help you predict the future if that country's leaders have themselves not decided on a course of action. To put it simply, we cannot know their future before they do. So our political leaders would be mugs to rely on intelligence predictions. There could be nothing more dangerous for Australia than if our leaders built policy around only one possible future, rather than making policy flexible enough to cope with the unknown.

So the point of collecting and analysing intelligence is not to give political leaders certainty about the future. It is to help them cope with uncertainty.

FOOTNOTE: The same research (cited above) showing how bad political experts are at making predictions also demonstrated that the worst pundits are actually rewarded for their poor performance. With its bias toward the sensational and dramatic, the media encourages those experts who make outrageous predictions by inviting them back on their TV programs and op-ed pages to make still more.

The reality for intelligence analysts is a little more prosaic. When they get their predictions wrong, they’re subject to snarky columns by journalists like Jason Koutsoukis.