China’s Underground Great Wall

Recent Features

Features | Security | East Asia

China’s Underground Great Wall

Some analysts suggest China’s vast network of underground tunnels is evidence it is undertaking a massive build-up of its nuclear arsenal. Actually, China has reason to worry about the US.

It’s tempting to dismiss the story in Monday’s Wall Street Journal suggesting that China may have around 3,000 nuclear warheads as the kind of reporting that could only be considered ‘fair and balanced’ on Fox News, and so just ignore it. After all, as long ago as 2004, Jeffrey Lewis tracked down the origin of media reports cited by the Journal that China has 2,350 nuclear weapons. Embarrassingly, the source is an online essay based on bogus US intelligence information that was posted by a Singapore University student.

Moreover, it hardly seems worth wasting time explaining why it’s invalid to estimate the size of China’s contemporary arsenal by taking a 1960s US intelligence report that predicted how many warheads China would have in 1973, and then assuming that it has built up at a constant rate since then. What does make the article worth engaging with, however, is its inability to even try to understand China’s strategic challenges, and why it might go to some fairly extreme lengths to try to solve them.

The purpose of the Journal article is to raise awareness of China’s nuclear modernization and the ‘immense strategic leverage’ it would supposedly give China in a war. Now, I certainly don’t claim to know why China is modernizing its nuclear force. China’s modernization may be offensively orientated. Perhaps Beijing really does wish to change the status quo in its favour by force of nuclear arms. The ‘facts’ collected by the Journal, however, provide no evidence—for or against—these propositions.

The article focuses on the vast network of underground tunnels, the ‘Underground Great Wall,’ which China has built to protect its nuclear forces. ‘For decades,’ writes Bret Stephens in that knowing tone adopted by the Journal’s finest, ‘nuclear experts have understood that the key to “winning” a nuclear exchange is to have an effective second-strike capability, which in turn requires both a sizable and survivable force.’

Wrong. A survivable second-strike capability is the key to not losing a nuclear exchange. It ensures that an adversary can’t disarm you and then use nuclear threats to bend you to his will. Even if China had 3,000 warheads all mounted on intercontinental ballistic missiles—which it doesn’t—it could still not disarm the United States. Apart from the inability of inaccurate Chinese missile to destroy hardened American silos, the four or more US submarines (each of which is armed with about 100 warheads) that are at sea at any given time ensure the invulnerability of the US deterrent.

By contrast, China does have reasonable grounds to fear that the United States is seeking a war-winning nuclear capability. The United States deploys something just shy of 2,000 strategic warheads, with more in reserve. Its delivery systems are exquisitely accurate. It’s developing conventional weapons designed to hunt down mobile missiles. And, on top of that, Washington has consistently refused requests from Beijing to explicitly state that the United States isn’t seeking the ability to eliminate China’s nuclear forces.


Whether China’s fears are justified or not is irrelevant (although, for the record, I think they are exaggerated). The point here is that they are understandable, especially from a military planning perspective, and most likely genuine. In fact, the crowning irony is that those who argue that the United States should pursue a war-winning capability against China (some of whom regularly opine in the Journal) generally show the least recognition that this would concern China and prompt to it to take countermeasures.

This brings us onto the Underground Great Wall. Given the extent of Chinese fears, the idea that Beijing might build 3,000 miles of tunnels to protect a small nuclear force should not be beyond belief. After all, it’s not exactly unprecedented for China to try to protect itself from external threats by building a gargantuan defence line, it is? Phillip Karber, a Georgetown University professor who conducted most of the research on which the Journal article is based, questions the scale of the project. He asks why China needs such a long network of tunnels if it only has the 300 or so warheads estimated by the US intelligence community. To answer this question, Karber should examine the United States’ own nuclear history.

Back in the late 1970s, many in the United States expressed deep concerns that the Soviet Union was opening up a ‘window of vulnerability’; at some point in the 1980s—at the moment of ‘maximum’ danger—the Kremlin would supposedly use its awesome force of land-based ballistic missiles (which, even then, were much more accurate than Chinese weapons are today) to annihilate US missiles, rendering the United States defenceless.

One American response was the MX missile (later renamed the Peacekeeper, although, for some reason, the new name never caught on). The original plan was to put each new missile in one silo. However, worried about the vulnerability of silos, Congress blocked this plan in 1976. Three years later, after the US military examined more than 40 options, President Jimmy Carter eventually approved an alternative: the ‘shell game.’ The United States planned to construct 4,600 silos for the 200 missiles it was going build—yes, that’s right, 23 silos for each missile. The missiles would have regularly been shuttled back and forth between different silos so the Soviet Union wouldn’t have known where any of them were. Thus, if it had wanted to disarm the United States, Moscow would have had to target every single silo, requiring a staggering number of warheads (9,200, in fact, if it’s assumed two warheads would have been assigned to each silo).

4,600 silos to protect 200 missiles…Not completely dissimilar to the Underground Great Wall, is it?

In the event, the shell game was never constructed. President Ronald Reagan rejected it in favour of the traditional one-silo-per-missile model. But, had the United States—like China today—not have had to bother with the ‘inconvenience’ of democracy, then perhaps not even the Journal could have professed such amazement at Chinese tunnelling.

James M. Acton is a senior associate in the Nuclear Policy Programme at the Carnegie Endowment and a Stanton Nuclear Security Fellow. This is an edited version of an article that appeared here.