The passing of Robert S. McNamara has reignited my erratic search for substantive information on this icon of rational methods. (The computer science equivalent might be Dijkstra.) In his rehashed pseudo-obituary, Fred Kaplan made reference to interviewing McNamara for his PhD thesis of 1983, titled Wizards of Armageddon, which I figured might be worth a read given that his polemics on Slate are usually interesting and coherent.
In brief, this book is about the strategic thinking of the civilians entangled in the U.S. war machine immediately following the use of the atomic bomb at Hiroshima and Nagasaki. The 1950s and early 1960s are dealt with in some detail, which makes sense as these were the most interesting periods, and presumably Kaplan had a mountain of newly declassified documents to trawl through, myriad ungagged players to interview. Conversely by the time of Carter and Reagan we're definitely into newspaper clipping territory.
The centerpiece of the narrative is the construction and evolution of the RAND Corporation by the U.S. Airforce, home to many international policy thinkers over the years: Kenneth Arrow and many other illustrious economists have spent some formative time there.
Kaplan alludes to some of the central mathematical concepts, such as game theory, that underpin the strategic analyses that he describes without providing enough detail to be engaging; showing the prisoner's dilemma in extensional form deepens rather than covers this gaping hole. By treating lightly over technical specifics, the book sometimes feels like little more than a laundry list of participants, reports and military obscurities. Perhaps I really should be aspiring to read Machine Dreams.
Interesting were the various views of the Korean war: some took it to be a failure as the U.S. didn't win, whereas others felt it was a success as it achieved limited goals using limited resources. The tension between these views is mirrored in the schizophrenic thinking about the use of nuclear weapons, as portrayed by Kaplan. Roughly, how could the U.S. win in a mutually-assured destruction (MAD) scenario? The two options seemed to be be either not to play, implying more Koreas, or to hope that some sophisticated signalling with nuclear devices (e.g. the "no cities" policy) got read correctly, resurrecting the possibility of a doubtlessly-Pyrrhic victory.
Another perspective (circa p200) is given by Clausewitz's ideas on military objectives: if war is politics by other means, and self-destruction is assured once the bomb comes into play, then using the bomb cannot possibly be sanctioned by a government which notionally has the civic wellbeing in mind, or indeed, derives its sovereignty from the populace. Therefore war between the U.S. and U.S.S.R. became impossible to countenance (if we are prepared to grant the Soviet rational self-interest) and the proxy war of Vietnam (and so forth) became inevitable. I wonder if this fed into McNamara's thinking at the time.
I observe in passing that Herman Kahn's escalation ladder (circa p223) is reminiscent of Rapoport's tit-for-tat strategy for the iterated prisoner's dilemma. Kaplan uses exactly that moniker to label Kahn's Type III Deterrence.
Kaplan struggles to distinguish between the policies that make sense while there is a nuclear imbalance — for instance, the U.S. had no need to be particularly subtle in its use of the first atom bombs as no other power could challenge it — and those for the equilibrium state where any of several actors could bring the planet to a biological conclusion. In the latter case MAD is justifiable due to it being something like a Nash equilibrium, and moreover the policy makes it highly unlikely that conventional wars will get out of control if the actors are rational. Kaplan does not even mention Nixon's behaviour in this context.
Concretely (p315), Kahn's theory of escalation clearly leads to an arms race, at least until MAD levels of weaponry are established. After that it makes sense to a military-industrial complex acculturated to constant expansion to pour money into anti-missile defence systems, which up to that point are prohibitively expensive (p322). Kaplan would have done better to tie these arguments to the dynamics of the situation more often.
McNamara has some key roles in this saga, but my desire to get some handle on his methods, some insight into how he dealt with the complexity of the defence portfolio, and specifically the moral ambiguity of the bomb went unsatiated. We get some comments to the effect that he was well aware that the Vietnam War was not right and not ever going to be right; for example (p366) in 1966 he states in a public speech that communism is not always at the centre of conflict in the developing world. His loss of certainty at this time, the end of his stint as Defence Secretary, unraveled his public unemotionalism, giving the impression that his morality was only underpinned by rationality: from his incredibly abstract view of the war, only possible by being so far from the action, it was justifiable on a rational posturing basis and no other. Coupled with the obligation he felt to his country to take his best bite of a shit sandwich, the disaster appears inescapable.
Clearly Kaplan has found his niche in writing for Slate; he has a keen eye for interpreting policy and theory, and reads well in the short popular form. This book needed stricter editing, as it was quite repetitive in some places, but I grant that was not so easy to achieve in the early 1980s.