The Perils of Overestimation
By A reader asks the following question:
Could you explain to me why it’s bad to derive capabilites from intentions?
I understand that it produces inaccurate intelligence estimates and that you can’t actually know what someone’s military capabilities are just from what they may or may not think or feel about a country, but is that the main reason that it’s a bad thing? If that’s the only reason, then I think that it would seem that it isn’t all that bad, especially when one looks at the collapse of the Soviet Union. As you know, one of the reasons that may have contributed to the collapse was the Soviets spending too much money that they couldn’t afford on defense. The Soviet spending being spurred on by US intel agencies overestimating Soviet capabilites which of course was the driving force behind US spending.
Now, if the argument is that deriving capabilites from intentions is bad because it produces a security dilemma and arms races, I could understand that logic much better than simply saying something like “it causes overestimates.” Is that the reason? Why do people get so worked up over it?
This is a good and reasonable question, and deserves a good answer.
“May have” is an important caveat, given the controversy that still surrounds the collapse of the Soviet Union, but let’s grant it. Even if we accept that US defense spending brought about the Soviet collapse, and even if we were to accept that collapse was the only option that faced the Soviet Union (more on that in a moment), the logic does not hold as a general principle.
The problem with the estimates on the Soviet Union was not simply that capabilities were derived from intentions; rather, capabilities were derived from an incorrect assessment of intentions. By assuming that the Soviet Union was incorrigibly hostile, we then assumed that they would develop the capacity to hurt us. Both of these assumptions, it turns out, were wrong. The Soviet Union was a status quo state (certainly after 1963) that developed capabilities in accord with maintaining its position.
Fine, you may say, but the Soviet Union was also a foul state (true), and its collapse a boon to humanity (true), so anything that contributed to that collapse, including overestimates, must also be good.
I would respond to this in two ways. First, in a democratic society a policy should be argued for in good faith. Rather than seeking an overestimate of Soviet capabilities in order to justify a large buildup that would presumably bring the Soviet Union down, the policy should have been argued for in a straightforward manner, with pros and cons considered in democratic fora. The conclusion regarding the arms buildup may have been the same, or it may have been different. The process, however, would have taken democratic norms seriously, rather than treating them with contempt.
Second, the downside to overestimation is significant. Let’s consider, for a moment, the case of the Soviet Union. If we assess Soviet intentions as hostile and capabilities as high, then change our own behavior accordingly, the Soviets have several options. One is to pursue policies that might lead to collapse, as Gorbachev did. This would be a good thing. Another response to the American buildup would be to do nothing. Consider how extraordinarily well this might have worked out for the Soviet elite; rather than attempting to meet the buildup (although evidence that they did this is very sketchy; Soviet expenditures rise in response to Afghanistan, not clear at all that they rise in response to American expenditures) they maintain their current level of defense spending and economic growth, all the while painting the United States as an aggressor. The end result is high defense expenditures in the US with no policy results to show for them. A third, and much more troubling option, would be preventative war. Indeed, some scholars have argued that the most surprising thing about the Cold War is that it ended without resort to this last option, where a concerned Soviet Union might have launched a war as a last chance bid before being overcome.
The inevitable corollary to “it causes overestimates” is that these overestimates lead to policy mistakes. By overestimating the capabilities of an opponent, we may believe that a preventative war is warranted when no such conflict is actually justified by security concerns, as is the case in Iraq (leaving aside the humanitarian and democratic justifications). The overestimation of capabilites can heighten the security dilemma, such that our increased concern is matched by an increased concern on the other side, leading then to unnecessary arms races and potentially to unnecessary wars. Finally, overestimation of enemy capabilities based on enemy intentions can lead to despair; the French vastly overestimated German capabilities in early 1940, and consequently made a series of policy mistakes that led to their collapse. Similarly, the British overestimated Japanese strength in the PTO in 1941 and 1942, and surrendered easily defensible positions to piecemeal Japanese attacks.
None of this is to say that “worst case estimates” aren’t good things, because they are. We shouldn’t engage in sunny optimism as a policy any more than we should engage in grim pessimism. Unfortunately, estimates of capabilities derived from intentions almost always prove wrong, because intentions are much harder to assess than capabilities. Chamberlain, Mussolini, and Stalin all tried to assess Germany by its intentions rather than through its capabilities, policies which led to disastrous failure. The United States tried to assess China in 1950 through its intentions rather than its capabilities, and actually underestimated Chinese resolve. The US tried to assess North Vietnamese intentions from 1960 until 1972 in an effort to break the North’s will, and failed utterly to understand the Vietnamese motivations. To use a more contemporary example, Saddam Hussein tried to judge the United States by its intentions rather than its capabilities in 1990 and in 2003, leading to policies which again caused disaster. Not all of these cases led to overestimation (sometimes they led to underestimation) but all were fatally flawed attempts to come to grips with an opponents intentions, rather than with the actual means an opponent had at his disposal.
Frankly, I prefer worst case assessments that are based on genuine intelligence and an honest consideration of the evidence.