WALL STREET JOURNAL
January 16, 2002
by Edward Jay Epstein
If anyone needs more evidence of the vulnerability of U.S. intelligence after the Cold War, “A Convenient Spy” (Simon & Schuster, 384 pages, $26) offers it in abundance. The title, in its irony, is somewhat misleading, since the book is not about a spy, convenient or otherwise, but about an investigation that failed to find one. The failure itself is instructive.
Dan Stober, a reporter for the San Jose Mercury-News, and Ian Hoffman, a reporter for the Albuquerque Journal, brilliantly unravel the curious case of Wen Ho Lee, a weapon-code designer at the Los Alamos (N.M.) National Laboratory. After being investigated by the FBI in the 1990s, and even threatened with a death sentence, he was imprisoned for eight months to await a trial that never took place.
The occasion for such zealousness, however ineffective, was real enough: China’s breakthrough in nuclear weaponry. The effort to find out how the breakthrough came about — in Messrs. Stober and Hoffman’s account — amounts to three vivid stories: the good, the bad and the ugly.
The good story shows the CIA finding critical pieces of the puzzle in the 1980s. Before then, the U.S. had few means of learning about Chinese nuclear bombs, except for “technical intelligence,” which consisted of measuring the seismic waves from bomb tests and analyzing the chemistry of debris.
Suspicion, solitary confinement . . . and no solid evidence.
While such data gave important clues, it left American intelligence officials in the dark about the intent of China’s weapons designers and their technological prowess. Much of the weapon work in China was then done in Mianyang, a science city few outsiders had visited. Not even the names of key scientists were known.
The solution was to use U.S. nuclear scientists. The CIA encouraged them to accept invitations to Chinese conferences and to establish personal relationships with their Chinese counterparts. American scientists who were ethnically Chinese proved to be, naturally, more useful in this regard, since they knew the language.
Thus American scientists visited Chinese labs, testing grounds and sites. And, if they were lucky, they learned about “targeted” information. It was understood, by American intelligence officials, that this gambit might involve some reverse leakage, since U.S. scientists would themselves have to answer some questions. But the CIA deemed the trade-off worth the risk. What they got was a vital look at the minds and techniques of Chinese weapon planners.
One of the dozens of Los Alamos scientists gleaning information for the CIA in China in the 1980s was Wen Ho Lee.
The bad story involves the ensuing counterespionage fiasco. In 1995, the U.S. learned that, three years before, China had tested a miniaturized warhead that had design characteristics similar to America’s own W-88 warhead. Then Taiwan got hold of documents revealing that the Chinese had a crude sketch of the W-88 and the W-87.
Had the Chinese stolen such designs? A blue-ribbon panel of weapons designers, scientists and intelligence experts studied the question and arrived at, well, an inconclusive conclusion: It was possible that the Chinese got help from a spy, but it was also possible that they made the breakthrough themselves, with supercomputers.
Undaunted by such ambivalence, the Energy Department’s intelligence chief, Notra Trulock, decided to find the (presumed) spy. His first mistake was narrowing the hunt to Los Alamos. As the authors show, the W-88 design could have been stolen from a half-dozen other facilities, and the W-87 could not have been stolen from Los Alamos. His second was to narrow the hunt to an ethnic Chinese scientist who had made trips to China: After all, espionage of this sort could have been done by someone of any race and ethnicity. Espionage is an equal-opportunity employer. In any case, Wen Ho Lee became the prime suspect — for a crime that may not have been committed in the first place.
When the FBI could find no solid evidence against Mr. Lee and grew disenchanted with the investigation, a frustrated Mr. Trulock managed to convince a Senate committee in 1998 that China had tested a copy of the W-88. (It had only tested a device that had common design features.) He also announced that the FBI was about to arrest a spy at Los Alamos (which it had no plan to do).
The ugly story was the aftermath. Once the FBI failed to turn up espionage by its usual means — false flag stings, surveillance, interrogations — a search of Mr. Lee’s office showed him to be guilty of something else: the unauthorized copying of data onto an unclassified computer.
As it happens, this copying was done well after China miniaturized its warhead. Still, it was said to be a terrible breach of security. (The data included computer simulations of nuclear explosions.) The Justice Department indicted Mr. Lee on 59 counts of copying and retaining information with intent to injure the U.S. and aid a foreign country.
The government argued that such data amounted to the “crown jewels” of our nuclear establishment. Thus Mr. Lee was denied bail and kept in solitary confinement. It then turned out that the government had exaggerated the value of what Mr. Lee had copied. It had not even been classified secret! Its security designation — PARD, for “protect as restricted data” — was a twilight rating that Los Alamos employees treated to mean, in essence, unclassified. In short order, the government’s case fell apart. Mr. Lee pleaded guilty to one count of unauthorized copying and was released. The judge apologized to him for “the unfair manner” in which he had been treated, and rightly so.
But then: Why did Mr. Lee do his copying? The authors of “A Convenient Spy” do not attempt an answer, though they find implausible Mr. Lee’s claim that he wanted to protect the data from computer failure. So a mystery remains. Perhaps we shouldn’t count on the FBI or Energy Department to solve it any time soon.