'People, ideas, machines' III: More on the fallacies of nuclear deterrence amid danger, exertion, uncertainty, and chance
Schelling's game theory was dangerously misleading... How should we think of deterrence? ... What new capabilities should we prioritise?
Introduction
‘There is a slow train coming up around the bend.’ Bob Dylan.
‘We came to assume that the lethality of our nuclear weapons would deter Soviet leaders reliably and predictably, virtually regardless of the local conditions in Moscow.’ Payne
‘People, ideas, machines — in that order!’ What Colonel Boyd shouted in the Pentagon for years…
Von Braun (left) and von Neumann
My previous blog covered Payne’s first book. This blog covers The Great American Gamble: Deterrence Theory and Practice from the Cold War to the 21st Century (TGAG). As well as studying the subject as an academic, Payne also served as an official and advised the US government. He explains ‘the slow train coming’ — how US theory and policy on nuclear deterrence developed, the theoretical and other errors, the misguided thinking about Soviet attitudes, why these ideas will fail to save us and so on. The book is a ‘conclusive demolition of the mainstream of theoretical, policy, and strategy error, which we have been truly fortunate to survive — to date’ (Colin Gray).
Payne explains how people now base ideas about nuclear deterrence on deep assumptions that are ~60 years old, were wrong then, and should be abandoned.
Instead of doing one long blog covering the whole book, I’ll post notes/thoughts day by day. I’ll try to limit restating stuff from the first book and make this as additional as possible. (I’ll therefore edit a bit as I go along and move material around.)
Unquestioned assumptions and covid lessons
Covid showed how leaders can stumble into disaster because they have meetings based on big assumptions they’ve never really studied and questioned. All sorts of critical questions are therefore not even discussed and meetings with senior people can therefore easily instead focus on second and third order issues in confused ways, often generating destructive ideas, chaos, and wasted effort.
For example, in February 2020 meetings of key ministers and officials were largely determined by crucial assumptions such as:
There’s nothing that can be done other than ‘flatten the curve’ a bit to achieve herd immunity over a few extra months.
The public won’t accept serious limits of freedom anyway, even if some East Asian countries — ‘irrationally, because it won’t work’ — reject natural herd immunity.
A vaccine is practically near-100% impossible quickly enough to have any significant effect.
These three assumptions were so deep and widespread that at the start they were not seriously discussed never mind rigorously questioned therefore other crucial issues were also not properly discussed at the start.
Ramp up testing? Why bother when it will all be over shortly and we won’t force people to do anything anyway?
Do a Manhattan Project-style vaccine taskforce? It’ll all be over in a few months, the thousands dead will be dead, then there’ll be herd immunity so why spend billions on a crash vaccine program that probably won’t get anywhere for years anyway, look at HIV?
The three assumptions were totally false and were dropped but thousands died because of the time it took to force explicit discussion of these assumptions and break down their logic. The first time there was a graph of multiple waves, and a concept of managing them, was when Ben Warner scribbled it on a whiteboard in the PM’s study in the evening of Friday 13 March, many weeks after covid was global news and had spread so far in Britain that thousands of deaths were already inevitable. It took several meetings over the next week to shift enough key people from those assumptions so that policy could change. Similarly in 2021 No10 defaulted to unquestioned assumptions therefore failed to get ahead of anti-viral procurement, which could have rendered omicron largely irrelevant relative to what happened (by eliminating most hospitalisations).
In the 1960s, civil defence was essentially dropped in the west because the false logic of Schelling took over. We must drop this logic on nuclear deterrence. We should therefore also consider new defences and capabilities.
An example of two obvious big questions for the UK…
Given the inevitability of future pandemics and the cumulative probability of WMD attacks growing over time (without dramatic political changes), should we invest in new civil defence structures? For example, should we encourage the building of shelters that could double for pandemics and nuclear attack, with off-grid energy and recycled air etc?
What intelligence investments and changes in priorities/targets make sense given the costs of intelligence ‘failures’? (I started asking this question with officials in 2020, including viz Russia, but it can’t easily be publicly discussed.)
Such investments would be costly but what are the true tradeoffs? Failure to plan properly meant future generations in the UK will be paying off hundreds of billions in debt incurred in the covid meltdown. If we’d invested modestly in diagnostics pre-2020 we would have made between a 10X and 100X return. If we’d invested in a system to develop vaccines at war-speed, the way General Groves would have done it and we can now see we could have done it (even faster than we did after the shift in policy April 2020), we could not only have averted almost all the costs of covid here, we could have made a fortune selling vaccines to the rest of the world. One can imagine roughly the order of a 1000X return on a billion dollars.
What price should we pay today to avoid tens of millions of deaths? What sort of capabilities should we invest in now that could avert or mitigate such disasters?
Notice that we have practically dismantled the vaccine taskforce, not invested in novel vaccine ideas, and the government is actively trying not to learn the expensive lessons just forced upon us by fate. And you will see no debate in Parliament about this. Our MPs collectively are largely happy to try to forget, even as new and bigger dangers approach. It resembles the pattern of failure over deterrence of Germany viz Belgium: in 1866, 1870, and 1914 you can read almost word-for-word identical errors written in Westminster — then we committed similar errors with Hitler. On one hand it seems crazy to think we could be arguing about lockdowns with a worse-than-covid pandemic before too long but on the other hand, history suggests this is pretty likely, if we don’t blow ourselves up first.
All bold below is added unless stated otherwise.
Keep reading with a 7-day free trial
Subscribe to Dominic Cummings substack to keep reading this post and get 7 days of free access to the full post archives.