By Peter Lee
The recent hubbub over the size of China's nuclear warhead stockpile and its underground maze of missile hidey-holes, the notorious "Underground Great Wall of China" is, on one level, a battle between sensationalizing amateurs and incensed arms control professionals.
On another level, it highlights a continuing nuclear security stand-off between the United States and China.
The furor was kicked off by a November 29 article by the Washington Post's William Wan. It described a study of China's strategic nuclear missile program prepared by a team of students working under Phillip Karber, an adjunct professor at Georgetown University, in an arms control seminar. Wan's breathless opening set the tone:
The Chinese have called it their "Underground Great Wall" - a vast network of tunnels designed to hide their country's increasingly sophisticated missile and nuclear arsenal.The coverage also provided an incendiary estimate:
For the past three years, a small band of obsessively dedicated students at Georgetown has called it something else: homework.
Led by their hard-charging professor, a former top Pentagon official, they have translated hundreds of documents, combed through satellite imagery, obtained restricted Chinese military documents and waded through hundreds of gigabytes of online data.
The result of their effort? The largest body of public knowledge about thousands of miles of tunnels dug by the Second Artillery Corps, a secretive branch of the Chinese military in charge of protecting and deploying its ballistic missiles and nuclear warheads.
The study is yet to be released, but already it has sparked a congressional hearing and been circulated among top officials in the Pentagon, including the Air Force vice chief of staff.
Most of the attention has focused on the 363-page study's provocative conclusion - that China's nuclear arsenal could be many times larger than the well-established estimates of arms-control experts. 
Based on the number of tunnels the Second Artillery is digging and its increasing deployment of missiles, [Karber] argues, China's nuclear warheads could number as many as 3,000.The story as Wan presented it contained several irresistible elements, to whit: 1) A team of renegade students displays 2) ingenuity and elbow grease to 3) expose the errors of pointy-headed so-called experts while 4) revealing Chinese perfidy and 5) alerting the people of America to terrible threat.
This wonderful story had just two flaws: the parts of it that were true weren't new, and the parts that were new appear not be true.
The true part of the story concerned that immense network of Fu Manchu-like caverns, the "Underground Great Wall of China". Ever since Douglas MacArthur threatened the People's Republic of China (PRC) with strategic nuclear bombing during the Korean War in the early 1950s, the Chinese communists have been industrious tunnelers, as tourists who have enjoyed Beijing's Underground City - and the 500-meter airfield residing inside a hollowed-out mountain at the Air Force Aviation Museum at Xiaotangshan outside the capital - can attest.
The existence of the strategic missile tunnels is also rather widely known, although it seems to have escaped William Wan in his research for his article. Wan incautiously wrote:
While the tunnels' existence was something of an open secret among the handful of experts studying China's nuclear arms, almost no papers or public reports on the structures existed.Unless one considers the tens of millions of readers of Xinhua and Global Times or the millions who watch Rupert Murdoch's Phoenix Channel "a handful of experts", Wan's statement is, regrettably, false.
In 2009, China's Second Artillery Corps opened their underground facilities to journalists in order to publicize the investments the People's Liberation Army (PLA) has made in improving the survivability of its nuclear arsenal and, thereby, maintaining the credibility of the PRC's deterrent.
A Phoenix TV segment aired footage of tunnel construction and interviews with PLA talking heads who proudly revealed that construction was now a matter of mechanization and automation, not manual labor, banishing the Mao Zedong-era paradigm of blue ants grubbing away at the naked rock with their bare hands. Indeed, the new tunnels, with their automated blast doors and tiled surfaces, evoke modern construction marvels like the Channel Tunnel. 
In December 2009, Xinhua ran a Global Times piece that careful invoked extensive foreign reporting to reveal a massive project to harden China's nuclear sites.
The story reported that, with the advent of more accurate missiles and more powerful spy satellites by China's potential antagonists, the previous strategy of hardening surface sites and mixing them with decoy sites was no longer effective. Therefore, the government embarked on a massive, 10-year program to put its entire nuclear deterrent deep underground, several hundred meters below the surface.
The program was first revealed on national television in 2006, apparently as a riposte to claims by US analysts that China did not have a credible "second strike" capability; ie the ability to launch a nuclear counterattack after absorbing someone else's nuclear incoming.
With its usual subtlety, Global Times put the key money quote - the opinion of a "Western expert" - in boldface:
If nuclear bunker-busting missiles were employed against China's missile bases, it would take a strike of several tens of missiles with yields of tens of megatons each striking at the same point to penetrate; and it would take several more missiles to destroy the facility completely.The article let it be known that the project might extend as far as 5,000 kilometers (just short of 3,000 miles) and that it had been dubbed "the Underground Great Wall". The source: "foreign analysts". 
That was the state of play in 2009: the tunnels were built, nicknamed, displayed on television, and their length and other characteristics discussed in the media; interested foreign spooks, analysts, and national security types were well aware of the situation.
It is unclear what additional insight Karber's hardworking team at Georgetown (which reportedly suffered from a dearth of Chinese-speaking members) actually added to specialists' understanding of the tunnel project.
But it was the "3,000 warheads" claim, illustrated with what might go down in arms control history as "The Karber Curve", that roused arms control analysts to their greatest anger.
When asked by Asia Times Online if he had a comment on the issue, Dr Jeffrey Lewis, an arms control authority and author of Minimum Means of Reprisal: China's Search for Security in the Nuclear Age, responded:
Oh yes, the claim that China has 3,000 nuclear weapons is utter bullshit.China's holdings of weapons-grade nuclear material have been the subject of intense interest and study. Open source and classified government studies apparently agree that China has enough metal for about 400 warheads.
The best informed estimate in the public domain is laid out by Hui Zhang, who lists the range of estimates in a recent paper for the Belfer Institute at Harvard, concluding with his own figure for Chinese military plutonium stocks: 1.8 tons plus/minus 0.5 tons. The high end of the highest estimate is 4.35 plus/minus 2.2.5 tons 6.6 tons. A highly efficient warhead design would require at least 5 kg of material. Per Zhang's midline estimate, that would yield 360 warheads; the highest range of the highest estimate would yield 1,300 warheads. 
That is still a long way from Karber's 3,000.
The relatively stingy estimates of China's warhead stockpile is based on the assumption that China's only two plutonium breeder reactors have been shut down since the 1980s.
If Karber's team had discovered a new source of plutonium - or new evidence that China's warhead stockpile had grown significantly beyond the estimates - it would indeed have been a major find, and the arms control community would have had considerable egg on its face.
However, it appears that Karber extrapolated his 2011 numbers from a 1995 estimate in a gossipy Hong Kong political magazine, Trend, which claimed that China had a stockpile of 2,350 warheads.
After becoming aware of Karber's as yet unpublicized claims in September, China arms control specialist Gregory Kulacki investigated the situation in Hong Kong and drew the conclusion that Trend's number - which has circulated on the Internet for years - perhaps itself grew out of 1986 estimates by a US naval officer in a Western defense publication, not China's defense establishment. 
The Trend magazine magazine estimate, by a pseudonymous author, filled with enticing detail and the tantalizing statement, "According to information revealed by China's Ministry of Defense on July 15 of this year, as of June 30, 1995, the Chinese Communist nuclear weapon stockpile held 2,350 warheads," remains unsourced and unconfirmed, leading a lonely existence as an extreme outlier in estimates of Chinese warhead stockpiles. 
The Georgetown group was not even the first to unearth the Trend estimate; for that matter, it did not even locate the original article. Karber's team leaned on a short essay by a Singapore University graduate student, Yang Zheng, citing the figure, and spin-off references on the Chinese Internet.
Yang's piece had already been ferociously debunked by Jeffrey Lewis in 2004 and 2009. Lewis ultimately determined that Yang also had not accessed the Trend article himself; he had simply availed himself of a summary of the article posted on a Usenet bulletin board by a curious mathematics professor.
This exhaustive sleuthing was missed or ignored by Karber, to Lewis' considerable resentment, compounded by the fact that the Washington Post declined to publish his letter, which Lewis subsequently posted on his blog. It read, in part:
So where does Dr Karber get his wildly divergent estimates? Nowhere does Mr Wan mention that Dr Karber's "analysis" of China's plutonium production relies on a few Chinese blog posts that discuss a single, anonymous 1995 Usenet post, subsequently plagiarized by a Singapore University student, that is so wildly incompetent as to invite laughter. (I have mocked this essay repeatedly on my own blog, Arms Control Wonk.com.) Actually, Dr Karber doesn't mention this either. His research ended with the Chinese blog posts, which is something that no responsible scholar would do. …Karber's defense, as summarized in his appearance on al-Jazeera and an interview with the New York Daily News apparently boils down to his conviction that 3,000 miles is just "too much tunnel" for China's stated warhead inventory:
If I take any solace out of this pathetic episode, it is that Dr Karber's students will have learned first-hand how not to do research. 
"It doesn't make any sense to build ten miles of tunnels to hide one nuclear weapon," Karber told the Daily News. Despite its shaky evidentiary and analytic underpinnings, news of the Georgetown study flew around the world, creating the damning image of a massive, secret, sneaky Chinese nuclear surge. Seattle Times: US students dig up China's nuclear secrets: Arsenal could be huge 
Al-Jazeera: China's nuclear arsenal 'many times larger' 
Daily Mail: China 'has up to 3,000 nuclear weapons hidden in tunnels', three-year study of secret documents reveals 
Digital Journal: U.S. study claims massive underground Chinese nuclear stockpile  If the Karber furor was merely a matter of an ambitious national security entrepreneur throwing out an irresponsible estimate in order to generate some career heat for himself; or a case of the media happily seizing on an irresistible China-bashing story without letting facts or serious reporting get in the way; or the old, familiar spectacle of the national security establishment hyping a new threat in order to pad and protect its budget, then perhaps the arms control fraternity would have let the matter go with a cynical shrug.
Instead, arms control professionals have engaged in serious, aggressive pushback. The Wall Street Journal, Agence France-Presse, and, to be fair, the Washington Post's William Wan all reported on the vehement skepticism of specialists in the field.
It appears to have had some effect.
On December 6, Kulacki noted on his blog: Tomorrow (Dec 7), Prof Karber will apparently present a different graph on the size of China's nuclear arsenal in his talk at the Elliot School of International Affairs.  Karber has apparently abandoned the 3,000-warhead extrapolation, and substituted a graphic illustrating the annual rate of warhead production, without giving totals.
However, the claims of production rates apparently rely on the same questionable, bloggy data used to generate the notorious curve.
Kulacki, who appears to have made a journey to Hong Kong expressly to research and debunk Karber's claims, revealed that there are significant policy issues at stake concerning the Barack Obama administration's nuclear negotiation posture with the PRC: Chinese declaratory policy ... states ... that China's nuclear weapons are to be used only in retaliation after China suffers a nuclear attack.
US participants in the dialogues are trying to force their Chinese counterparts to stop talking about China's no first use assurance, and to stop pressing the United States to offer a similar assurance to China. The US military and diplomatic establishment does not find such an assurance credible, or believe it contributes to stability. But the Chinese disagree. As a result, the talks have been deadlocked for over a decade, leaving the public discussion open for the wild and unfounded speculations of people like Karber.  It looks like the fundamental problem is this: China wants its expeditionary force to cross the Taiwan Strait free of the threat of a US tactical nuclear first strike that negates China's advantages in conventional manpower and materiel. In order to stay America's hand, the PLA has a doctrine of no first use and strategic retaliation. If the United States accepted a no-first-use doctrine, then Chinese military planning would be considerably simplified.
The US military, on the other hand, wants the option of incinerating the PLA's expeditionary force with tactical nuclear weapons, absent the distraction of realizing that defense of Taiwanese freedom may involve the destruction of [insert your favorite US city here]. If only the Chinese would acknowledge it shares a doctrine of first use, then arms control negotiations could be targeted toward reducing the size of arsenals and taking strategic nuclear strikes off the table (while presumably leaving tactical ones on the table).
Nothing doing, as far as the Chinese are concerned. The Chinese leadership apparently likes its plain vanilla retaliatory doctrine just the way it is.
Apparently, for the US security establishment's response to this conundrum is a policy of "I reject your reality and substitute my own."
Writing in Arms Control Today, Kulacki asserted that the Obama administration is forum-shopping, trying to locate first-strike enthusiasts inside the Second Artillery Corps:
The US participants in these talks do not appear to respect anyone, from either country, who takes a no-first-use pledge seriously. To them, the pledge is an expression of na๏vet้ or mendacity. They suspect, therefore, that the Chinese individuals participating in bilateral talks either cannot or will not speak truthfully about China's "actual" nuclear weapons policy. ...Kulacki contrasts the US-China impasse with the US-Russia situation, where both sides are sitting on massive weapons stockpiles they wish to reduce:
The US response to this impasse is to search for a different set of Chinese interlocutors. US security analysts and military planners scour Chinese military literature to look for kindred Chinese authors who view China's commitment to a no-first-use policy as they do. Some US analysts believe they located strong candidates in authors from the Second Artillery, the branch of the Chinese military that operates China's land-based nuclear missile forces. Obama administration officials responsible for the US-Chinese nuclear dialogue are pressing to talk directly with the leadership of the Second Artillery in the belief that they will speak with a different and more authoritative voice than the officials sent previously by the Chinese government. 
Russian and US arms control experts are birds of a feather who can talk for hours and with great enthusiasm about nuclear arms control. The recent negotiations over the New Strategic Arms Reduction Treaty showed that although it still is very difficult for them to reach a binding agreement, discussions, at least, proceed with a high degree of mutual appreciation and respect. This is largely because the two sides have a common language, common assumptions, and a common objective that emerged from their shared experience as nuclear rivals during the Cold War.One can almost smell the booze and hear the jovial teasing: We will bury you! No, we will bury you! More vodka! Na zdarovya!
US-Soviet arms control talks proceeded from the assumption that left unimpeded by negotiated limitations, both sides would continue to be highly motivated to seek the ability to launch a disarming first strike against the other. The two sides also shared the belief that the nuclear arms race created by these motivations continually generated new and intolerable uncertainties that could precipitate a large-scale nuclear exchange. The purpose of bilateral arms control negotiations was to establish an assurance that neither side could obtain a decisive first-strike capability.
Discussions with China, on the other hand, are less happy and founder on a fundamental contradiction.
The Chinese, by virtue of their small arsenal, are vulnerable to a US first strike. The leadership has apparently decided not to build up to nuclear parity, modify US nuclear behavior with the threat of a credible Chinese first strike, and then negotiate stockpiles down in a Sovietized, transparent, inspection-intensive program.
For Chinese military planners, nuclear arms control in the current context looks a lot like unilateral disarmament. The PRC has no interest in allowing US inspectors to rummage through its strategic missile infrastructure in pursuit of this goal and determine the vulnerabilities of its strategic nuclear forces. As Kulacki put it:
According to this logic, providing more detailed information about its nuclear arsenal would only leave less to chance and thereby increase the US incentive to launch a pre-emptive first strike in a moment of crisis. China's nuclear forces are small enough to make such a strike a tempting choice.It simply wants the US to wonder, and worry, about a small but significant number of Chinese nuclear weapons - and decide it is better to abandon the first-strike option.
This, perhaps, accounts for the uproar over the Karber report.
If the 3,000-warhead figure is not rebutted, then the United States can cite the Karber study as evidence of a Chinese build-up toward a first-use capability, even though the actual Chinese policy may be diametrically opposite.
From this perspective, vigorously fact-checking Karber and his Georgetown team can be seen as a step along the path of clear-eyed nuclear realism ... and away from delusional and dangerous atomic bombast. Notes
1. Georgetown students shed light on China's tunnel system for nuclear weapons, Washington Post, Nov 16, 2011.
2. Click here for the Xinhua report (in Chinese).
4. China's Stockpile of Military Plutonium: a New Estimate, Harvard.
5. The Sources of Karber's Sources, All Things Nuclear, Dec 7, 2011.
6. Click here for text.
7. Collected Thoughts on Phil Karber, Arms Control Wonk, Dec 7, 2011.
8. Is China hiding weapons in underground tunnels? NY Daily News, Dec 1, 2011.
9. US students dig up China's nuclear secrets: Arsenal could be huge, Seattle Times, Nov 30, 2011.
10. China's nuclear arsenal 'many times larger', YouTube, Dec 1, 2011.
11. China 'has up to 3,000 nuclear weapons hidden in tunnels', three-year study of secret documents reveals, Daily Mail, Dec 1, 2011.
12. US study claims massive underground Chinese nuclear stockpile, Digital Journal, Dec 1, 2011.
13. Prof Karber Adjusts His Report on China's Nuclear ArsenalAll Things Nuclear, Dec 6, 2011.
14. Nuclear Quacks and Clucks All Things Nuclear, Oct 14, 2011.
15. Chickens Talking With Ducks: The U.S.-Chinese Nuclear Dialogue, Arms Control Association, October, 2011.
Peter Lee writes on East and South Asian affairs and their intersection with US foreign policy.