On Projections

As a way of demonstrating the uncertainty of intelligence estimates, my undergraduate professor assigned a particular blog post by Jeffrey Lewis called "Estimative Language, Then and Now." The blog post discusses how the intelligence community uses probability language in their estimates. The thing that caught my eye was the table of estimative terms it leads:


This table comes from a document entitled "PRC Strategic Forces: How Much is Enough?" a Defense Intelligence Estimate produced by the DIA in 1974. As a document titled with something like that could not possibly be more relevant to my interests, I asked Jeffrey if he could root around in his archive and scan me a copy, which he kindly did. You can download a copy of it here:

PRCStrategicForces
.pdf
Download PDF • 14.44MB

.

The estimate is eye-catching in various respects, including the fact that it leads with that estimative terms table. I have never seen another report produced by either the CIA or the DIA lead with such a table, and its inclusion at the beginning should present us with our first red flag: the fact that the analysts who wrote it found it necessary to include it means they are uncertain about their own conclusions and will be using terms like "probably" a lot, which the report indeed does. The beginning of the report acknowledges the speculative nature of its findings:


The second eye-catching thing about this report is the speculation itself. The estimates project a massive expansion in the number of land-based and sea-based missiles deployed by the PRC over the next decade:


60 CSS-4 DF-5 silos by 1985! 8 SSBNs! If you aren't familiar with the history of the PRC arsenal you might not understand how wild these projections are. The number of DF-5 silos never broke 10 during this time period. By 2015 they had just 18, not 60 or more. China's first submarine, the Type 092, was such a colossal failure they never made more than one of them. China's first actually operable SSBN, the Type 094, was not operational until around 2010. While the report repeatedly claims that the authors have some insight into what PRC leadership thinks and how they might structure their nuclear posture, they do not offer any real evidence of such insight. Instead, many of their conclusions stem from what they think it would be sensible for PRC leadership to do. For example:


While the authors do expect that the PRC's nuclear arsenal would remain limited (by US standards) over the next decade, they miscalculate just how limited the arsenal might be and overestimate the amount of real utility PRC leadership circa 1974 might see in nuclear weapons. The PRC kept their deployable arsenal of missiles way below even the author's lowest estimate, as the analysts judged that an arsenal below their estimates would not be survivable (on this point both I, the DIA analysts, and many academics inside China agree). The analysts describe an arsenal lower than their lowest estimate as an arsenal that would only be useful as a "psychological" weapon. If you read a lot of period statements from Chinese leaders, you'll quickly realize that Chinese leadership thought nuclear weapons were just that: psychological weapons.

I read this report a year ago, read it, filed it in my archive and forgot about it. But recently I was rooting through the CIA FOIA archive and discovered that I was not the only one who found the report eye-catching. This quote comes from the Review Of National Intelligence (RONI) section, an office responsible for reviewing and critiquing IC reports, within the CIA:

They were, to say the least, less than pleased. I was further surprised to find that pushback from the CIA and Air Force on the DIA report found its way all the way to the office of the then Director of the DIA, General Samuel V. Wilson:


The story of the DIA report and the reaction to it is especially relevant for discussions of the PRC nuclear force today. We are in another period in which the future of the force, and PRC political and ideological thought related to the utility of nuclear weapons, appears uncertain in a time of nuclear arsenal expansion. Some of the quotes from this 1974 report are almost word for word things analysts have told me today. For example:

This sort of thing should sound familiar to anyone who has been following the discussions around the PRC nuclear arsenal. How many times have we heard analysts go on about Xi seeking a "World-Class Military" without any discussion of how political leadership and military institutions actually think about their own arsenal? I once questioned one analyst after he made this sort of statement, asking what specific mechanisms he sees as driving the expansion and what use cases he would see for such a world-class nuclear arsenal. He responded that I was "overthinking" it. These analysts, if teleported back in time to 1974, would be in full agreement with the DIA analysts predicting a Chinese arsenal that is 10-fold above what they actually made. This is not to say that prestige isn't a component of nuclear weapon force posture decisions. It almost certainly is and there are many examples of prestige being an important motivator of a country's decisions in this realm (India being a good example). But without any actual detail backing it up, it is mere speculation and not the basis of a projection like the DIA analysts' projections.

We don't know what the PRC arsenal is going to look like in 10 years. It's possible that the DoD projections and some of the other projections floating around the open-source will be proven correct. It's possible they won't. But we should be very open, as RONI pointed out in 1975, about how little information we have about PRC military and political thought on nuclear weapons policy in the current era. It was opaque in 1974 and it's opaque now. And given statements by DoD and STRATCOM officials, I don't think it has gotten any less opaque even to the USIC. Both General Ferdinand Stoss and Admiral Charles Richard have admitted that they don't know why the PRC has chosen this time to expand their nuclear arsenal or the end goals of Chinese leadership. At the 2022 Nuclear Deterrence Summit, Stoss said: "Why are they doing this strategic breakout? Well, we don't exactly know. They are very opaque as to what they are doing with nuclear, and they always have been…but, you know, perhaps this is just one more brick to put into the wall to cement their capacity to play a much bolder role, certainly in the region and around the world, and they think that they need this nuclear underpinning."

All of this is not to say that I disagree with a lot of what STRATCOM and the IC appears to have observed so far. Their immediate observations, in terms of missiles assembled, missiles deployed, and missile sites under construction, is probably highly accurate. But without a solid window into the thought processes of the PRC's military and political leadership, it is extremely difficult, if not effectively impossible, to make long-term projections about the growth of the arsenal. And the DoD and STRATCOM seem to understand this, as they have been very careful about making any long-term estimates about specific missile deployments. The silos are a good example of the sort of problems I am talking about. I'll leave you with a quote from STRATCOM commander Admiral Charles Richards, who said this in response to a question about the theory that China might adopt a "shell game" theory and deploy only a limited number of missiles across a large number of silos: "Well, I think the point is to start with an assumption that you’re probably not going to know for sure. Are some of them going to be filled? Are some of them not? Are some of them conventional? Are some of them nuclear? Are some of them filled now and some of them are going to be filled later, and so all that adds up to is you got to be humble in your ability to predict this. You have to address what the total implications of this are going to be and is there more about it that you don’t know?" There could be 36 warheads in those silos, there could be zero, there could be 3600, but we won't really know until we at the very least have a clearer understanding of what PRC military and political organizations think about the subject. We can see the silos with our satellites. We cannot see inside the heads of PLARF command.