tween CIA and NSA for more than 50 years—especially with al Qaeda phone call transcripts before 9/11. You write about other periods of "blackout."
The NSA didn't break any high-level Soviet codes for a period of 30 years. They broke Soviet codes very well between 1945 and 1948 before a mole inside the U.S. military exposed NSA capabilities. One Friday, the Soviets switched to an unbreakable cipher system, and all interception went black. Until 1978, not a single high-level Russian code was broken. NSA spent several billion dollars on it and got nowhere. The NSA finally cracked the Soviet codes again by the late 1970s. Warning of the Russian invasion of Afghanistan in 1979 was a great victory for U.S. signals intelligence during a brief window when we could both intercept and decode Soviet traffic. We lost all that access when two more Soviet moles exposed U.S. successes.
Do other countries have their own NSA?
Most nations do. The Soviets had a much larger signals intelligence operation than the U.S. The Soviets were good against other smaller NATO countries but less effective against the U.S. People always blame the press for exposing secrets, but spies inside the government are far more dangerous. In fact, the Soviets generally knew exactly where all our ballistic missile subs were located for many years, thanks to broken naval codes. It actually gave Moscow a sense of security that we were not going to launch a surprise attack.
Has it all been worth it?
I know a dozen senior officials who ask the same question. NSA has produced exceptional material in many instances. And instances of NSA success remain classified. But it's not all-seeing or all-hearing, and it has a major cost. Over all, it's an essential tool, because it produces more and better intelligence than the CIA or anyone else in the intelligence community, except for satellites.