Thursday 26 Nov 2020 | 01:39 | SYDNEY
Thursday 26 Nov 2020 | 01:39 | SYDNEY

Another view on the role of intelligence agencies


Sam Roggeveen


14 December 2007 08:26

In response to my post about what intelligence agencies are for, Paul Monk writes:

The track record of the US intelligence system is not impressive when it comes to predicting major future developments. (I say 'system', because 'community' is too kind a word, as Robert Redford remarked many years ago, in Three Days of the Condor.) Your response, that predicting the future is notoriously difficult, is well taken. You refer your readers to Louis Menand’s review of Philip Tetlock’s splendid book, Expert Political Judgment, which demonstrated this meticulously. Menand, unfortunately, drew the wrong conclusion from reading Tetlock. He thought we should all just trust our own judgment, because the experts are so bad. Tetlock’s point, surely, was that we should be far more circumspect about predictive judgments, most certainly including our own, than we tend to be.

You then deflect Koutsoukis’s withering attack on intelligence agencies, by remarking that, actually, the purpose of intelligence agencies is not to predict the future, but to help decision-makers cope with uncertainty. Perhaps, but do they understand this to be their purpose and how are they supposed to do it, if they are so very uncertain themselves? What, in fact, does ‘coping’ with uncertainty mean? We need to be careful here, I think, to avoid just giving intelligence agencies an alibi. If we could define more closely what we mean by ‘coping with uncertainty’, then assess the performance of intelligence agencies against explicit criteria, much as Tetlock did with predictions, we might be closer to addressing the kinds of concern Koutsoukis and many others have about the intelligence analysis function in government.

My own opinion is that we should be cautious in assigning a clear ‘purpose’ to intelligence agencies, because the way in which they are created and the constraints under which they operate, actually suggest that they are not designed to achieve a clear purpose and this goes some way toward explaining why they perform less well than we might hope or expect. As Amy Zeggart has argued, first in Flawed By Design: The Evolution of the CIA, JCS and NSC and now in Spying Blind: The CIA, the FBI and the Origins of 9/11, there are rather intractable reasons why effective intelligence agencies are very difficult to design and defective ones almost impossible to reform. Is this reason enough to do away with them altogether? Perhaps – but just try doing this, politically, whether here or in the United States.

In any case, if we did away with our intelligence agencies, we would then have to come up with better ways of accomplishing the cognitive tasks they, at least ostensibly, perform. To do that, we would need to clarify what those tasks actually are and whether there is any need for intelligence agencies to do them. The first thing to acknowledge, then, is that intelligence agencies have always been intended primarily to collect secrets, keep them secret and channel them to the ears of their sovereigns. Were their primary role not this one, there would be little rationale for having them at all, since thinking and debating and predicting and all the rest can plainly take place without the existence of intelligence agencies. Indeed, it can often be inhibited by the existence and secrecy of such agencies, because they purport to have vital information, conferring a higher understanding, that they are not at liberty to divulge to other parties to the debate.

This being so, the question is whether the secrecy that surrounds intelligence collection, analysis and dissemination serves some crucial cognitive function that is not better served by some other means.  It is commonly observed that we live in an era of information abundance. Do we, then, need to spend billions to collect secrets at all? A good case could, in fact, be made that the more secrecy there is, the more doubt, confusion and suspicion infect international politics, national councils of state and public confidence in the executive. There are good reasons for actively seeking to minimize secrecy and make explicit – therefore vulnerable to retort and correction – the grounds for forecasts, policies and assumptions. Ideally, then, we would create something almost antithetical to the classic intelligence agency. We would create robust clearing houses for analysis and clarification of the underpinnings of public policy and strategic forecasts.

Suppose we were able to abolish the intelligence agencies and set up such new model, open clearing houses. Would they do better at analysis, prediction or ‘coping with uncertainty’? To answer that question, we need to ask two questions. First, could we get the design of such clearing houses right, where we have struggled with intelligence agencies? Second, what would we expect from them, in regard to ‘intelligence’ input to government? The answer to the first question must be that design would be exceptionally difficult, especially if the institution was required to function in much the same way as an intelligence agency – i.e. as a subordinate arm of government and according to civil service rules and procedures. The answer to the second question pivots the observation by American intelligence specialist Gregory Treverton that ‘intelligence’ has to cease being construed as ‘product’ (secret pieces of paper or electronic files) and start being seen as better understanding in the minds of decision-makers. Unless that is the goal, the secrets will pile up, while errors and obfuscation proliferate.

A moment’s reflection will lead one to conclude that a gulf separates the state of affairs I am outlining from the current state of play. Can we close that gap? In principle, yes, but probably in the manner that we have been closing the gap between a world of mercantilist protectionism and one of free trade: slowly, over countless objections from special interest groups, with many setbacks and the occasional full-fledged lapse into retrograde practices. For myself, I long ago walked away from the world of secret intelligence and have not found any reason to regret that step. There is no lack of information abroad for those wanting to inquire and think. The challenge is not to collect secrets; it is to reason well with the abundance of information at our disposal; to subject our reasoning to the most exacting tests at the bar of open and public debate; and to invent ways to get political leaders and other decision-makers to embrace the best practices in this regard – a liberal and democratic approach to policy making and strategic thinking.