|
Intelligence Reform and The Primacy of Politics
by John Ferris
Every few years, intelligence spins through a familiar cycle in western countries. Scandal breaks out. Demands are issued to punish the guilty and reform the system. The usual suspects are rounded up, especially calls to just get rid of the politics, or to end political problems by applying administrative solutions. These nostrums make good sound bites but, unfortunately, they are about as useful as saying that war would end if everyone would just make nice.
That is for a simple reason. Politics is not a problem for intelligence, but a condition for it. The difference between these words is fundamental. Problems can be solved, conditions must be endured. Intelligence is political by definition. It matters only if it affects decisions and actions. The latter inevitably stem from politics, whether produced by bureaucracies, Cabinets or the interaction between a commander and a few staff officers, where the personal becomes the operational. This reality is ignored by most commentaries on intelligence, which generally are written from the perspective of professional intelligence bureaucrats. They treat normative assumptions as natural law, assuming that what one should do is what one will do, that administration is good and politics bad, while Chinese walls should separate those who analyze intelligence from those who act on it. In particular, intelligence officers are presented as a priesthood, telling truth to power, and users as a respectful and responsive laity.
These ideas of bureaucratized intelligence, derived from the models of a General Staff and a Joint Intelligence Committee, can work well for military matters handled by military men, who come from one and the same professional background and accept the need for corporate discipline, in which every participant serves as a willing cog in a machine, so to achieve a collective task. Even in such circumstances, however, intelligence often is traumatic. These ideas work even less well for bigger issues involving mixed groups of decision makers drawn from differing backgrounds. Politicians, in particular, see these ideas as claims by bureaucrats for a monopoly over the right to tell their superiors what really is happening, and what can or cannot be done. Politicians, the key decision makers in systems of bureaucratized intelligence, also are temperamentally unsuited to them, emerging as they do from ruthless circumstances, where one trusts only one’s friends, doubts the existence of objectivity, and wants to hear how to do what they want to do, rather than why it cannot be done.
Most accounts of intelligence are modeled on the interaction between an expert and an amateur decision maker. They treat the expert as master and the amateur as student in need of schooling. Alas, things seem less simple to the student. Because of gaps in collection or knowledge, intelligence may not have expertise on the point at dispute. Collection may not have reported what a user needs to know. Analysts may not fully understand the picture, or be advising a technical figure, whose expertise matters more than their own to the decision at hand. Technical experts or statesmen will think these thoughts whenever they disagree with intelligence; and sometimes they will be right to do so. A leader, for example, can better judge some issues than his intelligence chief, because he knows better a key part of the environment, what he is going to do; and thus can guess at how the other side will respond, and know what topic he needs to know about, when intelligence does not. Security restrictions may keep analysts from looking for or recognizing the information a user needs to know, leading the latter simply to ask for all the relevant information, while ignoring analysts. Contrary to all the handbooks, incidentally, politicians sometimes handle intelligence well in such circumstances.
Analysts are not always necessary to intelligence. Even worse; sometimes they are wrong or, even if right, have a poor track record, which arouses distrust among reasonable people. The greatest problems in a system of bureaucratized intelligence arise when politicians do not like the advice they receive on specific issues and generally distrust the messenger. If a superior cannot find good reason to trust his intelligence service, whether right or wrong, he is forced into political action. Under such circumstances, the only solution suitable to the model of bureaucratized intelligence is for leaders to find intelligence chiefs they trust, and to fire the old ones. By decapitating and replacing its leadership, one still leaves intelligence institutions with their function as expert advisor. However, such steps are not easy, precisely because intelligence bureaucracies are bureaucracies: the larger and more established, the harder to change, or to capture through decapitation. Thus, users constantly are tempted to take actions which subvert any system of bureaucratized intelligence, such as creating new organizations led by loyalists to analyze intelligence, so to sidestep bureaus one does not like. Creating new agencies seems easier than fixing old ones, politicians imagine that analysis is easy, while few things are easier than to ignore intelligence.
These circumstances have concomitants. If the point of intelligence is to affect action, one of its worst possible situations is to have no influence on, or be irrelevant to, decisions. For any intelligence service, influence is lifeblood to an institution, careers, and self respect. However, nothing forces superior authorities to care about intelligence; they must be persuaded to do so. For an intelligence officer, the high road to influence lies through salesmanship, the effort to present one’s point as effectively as possible, given one’s knowledge of the foibles of his boss. Intelligence services and chiefs must learn to adapt to their bosses, because the opposite will not occur.
This process is normal to the working of any bureaucratized system of intelligence. It occurs when they work effectively, and when they fail. This process is not caused by bad individuals, though it may be exacerbated by them, or dampened by good ones. It occurs for systematic reasons. Even if a leader does not try to make subordinates please him, still they will seek to do so, in order to avoid irrelevance. Superiors need not be corrupt or authoritarian to spark politicization. They must merely be superiors, whom subordinates wish to please, leading them to cook a meal they know the boss will like, whether he has not ordered it or not. Politicization does not even have to be conscious or intended: it can happen without or even despite one’s will. The greatest temptations for an intelligence chief in a system of bureaucratized intelligence arise when dealing with a superior who ignores one’s advice.
Whenever an intelligence scandal occurs, commentators describe the situation as disastrous and abnormal, and search for simple solutions, when usually the problems are normal and medium in scale, while much of what went wrong stems from conditions, which cannot be solved. This is true of the politics of intelligence in Washington and London in 2002-03, during the run-up to the Iraq war. We know a surprising amount about these events, though not the full story. It has some novel features. Leaks are a conventional part of the politics of intelligence, but not the outing by the White House of a serving CIA officer. Even more, intelligence was openly wielded as a public tool of persuasion to an unprecedented degree, as politicians sought to exploit the reputations for objectivity of their intelligence services. Otherwise, events were ordinary. In Britain, so to maximize the size of spin, Labour politicians politicized intelligence. They aggressively pressed the Secret Intelligence Service and the JIC to state, for public consumption, what politicians wanted to have said. Had the issue not been public persuasion, probably the politicians would not have bothered, and little out of the ordinary would have happened in the politics of intelligence. In any case, the pressure stopped wherever the Chairman of the JIC, John Scarlett, drew the line. Although political pressure drove him beyond the normally accepted line, he did not cross the margin into what obviously would have been bad play, no matter what The Guardian reading class might imagine. In Washington, meanwhile, the politics of intelligence generally took a form normal since 1964. The administration pressured the CIA to tell the story it wanted to hear, without much success. The pressure, however, led George Tenet, a Director of Central Intelligence with a long history of marginalization under two Presidents, his position further shaken by the failure to predict 9/11, seeking to preserve his career and agency, to bend further than his analysts wished. In fact, he did pass an acceptable line. His “slam dunk” verdict was not far from the widely accepted view of the time, but it was a bad intelligence assessment, delivered for political reasons. Meanwhile, the Pentagon established its own private assessment agency, so Douglas Feith could tell Donald Rumsfeld the story Dick Cheney wished to hear. In both countries, leaders ignored intelligence, which was uncertain, or contradicted their beliefs—as usual. Altogether, these forms of politicization had little effect on policy—the latter was defined regardless of intelligence, which was too vague to start a war, or stop one-- but some on public opinion, though it was at most a secondary factor in the politics of preemptive war.
What lessons can be learned from this experience, and used to improve performance? The most novel of these problems, the use of intelligence as a public tool of persuasion, may be solved for perhaps a generation, simply because it will make publics trust their governments less, and the latter more careful. But no other new lessons can be learned, because the experience of 2002-03 stems directly from the systematic relationship between politicians and bureaucratized intelligence, a matter, which has existed for decades and is here to stay. The questions raised in 2002-03 were known long before, as are the nature of unsuccessful answers. In particular, one cannot solve political problems through administrative solutions—such as ideas of multiple advocacies, or for Washington to adopt a British style JIC. At the same time, some systems of intelligence are worse than others and improvements certainly are possible in those prevailing in western countries. In particular, states do better when politicians and the leaders of bureaucratized intelligence can work with, rather than against, each other. To achieve such a step, however, would require rethinking how intelligence bureaucracies function, for politicians to learn something about strategy and intelligence, and to have the process of intelligence account for politics, rather than to pretend it does not exist. To take such steps will not be easy, and they can all too easily lead to missteps. What does remain certain is that if things do not change, whenever intelligence matters on a major issue and decision makers are divided, they will politicize it; while intelligence will be free of politics only when it does not matter. And there ain’t no cure for politics.
|
Be the first to comment
Sign in with