The Unpersuasive Case for the NSA Call Dragnet’s Effectiveness

Sen. Dianne Feinstein (D-CA) has an op-ed in the Wall Street Journal ($) defending the NSA's bulk call records database as a "vital" counterterrorism tool.  While this wouldn't make the program legal even if true, it also seems clear that the secret Foreign Intelligence Surveillance Court (FISC) has relied, rather uncritically, on the government's assertions of "necessity" to draw the strained conclusion that every American's phone records are "relevant" to FBI counterterrorism investigations. It's thus worth pointing out how extraordinarily weak the case for the program's utility really is.  Feinstein begins by recycling the claim that if only the NSA program had existed in 2001, the 9/11 hijackers could have been identified and halted before carrying out their catastrophic attack:

Intelligence officials knew about an al Qaeda safe house in Yemen with ties to [hijacker Khalid] al-Mihdhar as well as the safe house's telephone number, but they had no way of knowing if anyone inside the U.S. was in contact with that phone number in Yemen. Only after 9/11 did we learn that al-Mihdhar, while living in San Diego, had called the safe house.

In congressional testimony in June, FBI Director Bob Mueller said that if intelligence officials had had the NSA's searchable database of U.S. telephone-call records before 9/11, they would have been able to connect the number to al-Mihdhar and produce actionable intelligence on participants of the developing plot. NSA Director Keith Alexander testified before Congress in October that if the call-records program had existed before 9/11, there is a "very high" likelihood that we would have detected the impending attack that killed 3,000 Americans.

The most obvious problem with this argument is that the court order we've seen for phone records explicitly demands two distinct categories of records, for calls "(i) between the United States and abroad, or (ii) wholly within the United States, including local telephone calls." The first category might have helped identify calls to or from a known safehouse in Yemen, but the latter, much larger category rather obviously would not.  This is simply an attempt to exploit the tragedy of 9/11 to deflect criticism of massive domestic surveillance that would not have been any use in preventing that attack.

Nor is it clear that even an international bulk phone records database would have been necessary, when NSA could have obtained targeted orders for records of calls between the United States and specific suspect numbers. As ProPublica's Justin Elliot pointed out when this argument first surfaced back in June, the 9/11 Commission found that the government missed many opportunities to identify al-Midhar, for reasons having more to do with the intelligence community's internal dysfunction than any lack of raw data. Several experts told ProPublica that NSA had both the technical and legal capability to track the numbers in contact with an actively monitored phone line.  Even if they did not in this case, for some obscure technical reason, it is wildly implausible that there was no way to create that capability without indiscriminately vacuuming up American call records.

Feinstein then repeats the widely-debunked talking point that the NSA's call database has helped disrupt dozens of "terror events":

Working in combination, the call-records database and other NSA programs have aided efforts by U.S. intelligence agencies to disrupt terrorism in the U.S. approximately a dozen times in recent years, according to the NSA. This summer, the agency disclosed that 54 terrorist events have been interrupted—including plots stopped and arrests made for support to terrorism. Thirteen events were in the U.S. homeland and nine involved U.S. persons or facilities overseas. Twenty-five were in Europe, five in Africa and 11 in Asia.

The key phrase here is "and other NSA programs": As several of Feinstein's colleagues on the Intelligence Committee pointed out back in June, the call records program played "little or no role" in disrupting those "events"—it was the "other programs" that did the heavy lifting. One could just as easily say that tarot readings—in combination with other programs—helped disrupt terrorism.

Those "terror events" aren't actually planned attacks, by the way.  One of the few "events" that has been publicly disclosed was... an effort to wire money to the Somalian terror group al-Shabaab.  In another case, the NSA tried to take credit for foiling a plot to bomb the New York Stock Exchange... a plot that, on closer examination, turned out not to be real.

So how many of these 54 "terror events"—including money transfers—was the call records program actually involved in investigating? About a dozen. And under pressure from Sen. Patrick Leahy (D-VT) at a recent hearing, NSA Director Keith Alexander was forced to admit that  in only one or perhaps two cases did the phone log database provide any important unique leads. Even that may be a stretch: As former FBI Director Robert Mueller acknowledged in a 2009 affidavit to the FISC, NSA tips from the call database sometimes yield information "that the FBI inevitably would have discovered via other investigative techniques." If the NSA database helps them make those discoveries a bit more quickly, there's surely value to that—but it's hard to believe it's value commensurate with the scale of the intrusion on the privacy of innocent citizens. One might argue the tradeoff is worthwhile if the NSA program is halting dozens of attacks—but it seems an awfully high price to pay if the main benefit is helping the FBI find people wiring money to Shabaab a little sooner than they otherwise would.

It is worth noting that the intelligence community made similar claims about the critical value of a bulk collection program for e-mail metadata, as Sen. Ron Wyden (D-OR) explained in his keynote speech at last week's Cato conference on NSA surveillance. Only after years of pressure from Wyden and a few skeptical colleagues did the agency admit the program wasn't particularly useful after all: It was shut down in 2011.

Perhaps because the evidence for any affirmative intelligence value is so weak, intelligence officials have lately fallen back on a new argument, also echoed by Feinstein: the "peace of mind" argument.  According to this Kafkaesque line of reasoning, even if the call records program fails to detect terrorists, it's valuable in reassuring the government that Americans aren't up to anything. In other words, the program is a vital counterterrorism tool because the vast majority of citizens swept up in the dragnet are innocent. Do you feel safer yet?