Misleading Congress on Surveillance History (and Why It Matters)

About a month ago, former DOJ attorney Ken Wainstein testified at a congressional hearing to urge reauthorization of the controversial FISA Amendments Act of 2008. For “background” purposes, he provided a brief history of the Foreign Intelligence Surveillance Act that we’ve heard a number of times from intelligence officials, suggesting that Congress had always intended to leave surveillance of communications between Americans and foreigners unregulated. According to this narrative, it was a sort of technological accident that FISA increasingly required judicial approval for this international traffic, and the broad spying power established by the FAA represented nothing more than a tweak or upgrade designed to restore FISA to its original scope. If you buy Wainstein’s history—as it seems many in Congress have—you’d naturally see the FAA as no big deal, and the civil libertarian backlash against it as misplaced or confused. The trouble is, Wainstein’s history is clearly wrong—and gives a highly misleading impression of how radically FAA changed the rules.

First, let’s look at what Wainstein says about how the original FISA statute—we’ll call it FISA Classic—was meant to operate:

In crafting [FISA], Congress recognized that it had to balance the need for a judicial review process for domestic surveillance against the government’s need to freely conduct surveillance overseas. It accomplished that objective by clearly distinguishing between surveillances directed against persons located within the United States – where constitutional protections apply – and those directed against persons outside the United States, where the fourth amendment [sic] does not apply. It then imposed the court approval requirement on surveillances directed against persons within the United States and left the Intelligence Community free to surveil overseas targets without the undue burden of court process. […]

The statute required the examination of a number of factors – such as location of target, location of interception and nationality of target – in determining whether a particular surveillance falls within that definition and the coverage of the statute. Among those factors was the type of communications technology being used by the target – i.e. whether he was communicating by “wire” or by “radio.” Given that “radio” (or satellite) technology was commonly used for international calls at the time and “wire” technology was the norm for domestic calls, it arguably made sense that FISA distinguished between “radio” and “wire” communications in designating which surveillances were sufficiently domestic in character that they would be subject to the court approval requirement and which would be excluded because they targeted foreign communications that did not enjoy fourth amendment [sic] protection. The result was a technology-based carve-out for surveillances targeting foreign-based communications.

With the change in technology over the intervening years, however, that carve-out started to break down. In particular, the development of the world-wide network of fiber optic wire communications resulted in an increasing number of phone calls and emails passing through the United States, whose interception in the United States required court review under the definition of “electronic surveillance.” As a result, the government found itself expending significant manpower generating FISA Court applications for surveillances against persons outside the United States – the very category of surveillances that Congress specifically intended to exclude when it imposed the FISA Court approval process in 1978.

To understand why this is wrong, we need to consider FISA’s somewhat convoluted four-part definition of “electronic surveillance” requiring judicial approval—three parts of which are directly relevant to the wire/radio distinction. The first definition covers the interception of any communication obtained by intentionally targeting an American within the United States, wherever that interception occurred. The third definition covers interception of radio communications—including international satellite traffic, and again, regardless of where the interception occurred—where all parties were in the United States. But the second definition, 50 U.S.C. §1801(f)(2), explicitly covers any wire communication where any party is in the United States, if the interception is conducted domestically. If Wainstein were right that Congress had “specifically intended to exclude” foreign-directed surveillance from the FISA regime, this section would make no sense at all: If the goal had been only to regulate surveillance of purely domestic conversations, then language of this section could simply have mirrored that of the definition covering radio communications. The first definition already covers surveillance targeting specific U.S. persons, which would make the other definitions superfluous if that was all Congress cared about prohibiting. Moreover, if there were any doubt, the legislative history presented in the House Report on FISA Classic makes it explicit:

[E]ither a wholly domestic telephone call or an international telephone call can be the subject of electronic surveillance under this subdefinition if the acquisition of the content of the call takes place in this country.

If Congress did not intend to leave these international communications unregulated, one might wonder, why have different rules for radio and wire traffic at all? David Kris, the former head of the Justice Department’s National Security Division, meticulously examined the question in a 2007 white paper that blows Wainstein’s historical narrative out of the water. As Kris explains, Congress knew that the NSA was conducting broad “vacuum cleaner” surveillance of international radio signals—from which it would have been infeasible to exclude all and only those communications with one end in the United States. Reluctant to interfere with an ongoing NSA program, legislators created looser rules for radio than for wire. According to Kris, “Congress intended to fill these gaps with subsequent legislation, but never did so.” In other words, Wainstein’s account—according to which Congress meant to leave surveillance of Americans’ international calls generally unregulated, but then mysteriously required warrants for domestic-to-international wire intercepts—gets it perfectly backwards. Rather, Kris explains, legislators thought it would be “desirable to develop legislative controls” over NSA’s “vacuum cleaner” surveillance, but “did not want that broader project to derail the incremental progress represented by FISA” until a viable means of doing so was developed.

Kris also undermines Wainstein’s version of the technological history. Contrary to claims that wire communications were almost exclusively domestic, Kris points out that “millions and millions of calls still crossed the oceans on underwater cables” in the late 1970s, when FISA Classic passed:

[A]t the time Congress was conisdering and enacting FISA, from 1974 to 1978, it does not appear to be the case that “almost all” overseas calls were carried on satellites; the actual portion was probably somewhere between one-half and two-thirds.

Moreover, Kris shows that the growth of fiber optic cable was not some surprise development that “started to break down” the technology-based distinction between domestic and international communications envisioned by FISA Classic’s authors. Rather, the shift to greater reliance on fiber optic cable for both local and overseas traffic was “on the horizon” as early as the mid-1970s.

Wainstein’s historical errors are anything but academic. As I’ve noted previously, Congress is speeding toward a five-year reauthorization of the FISA Amendments Act, which broadly permits warrantless acquisition of American’s international communications as long as the nominal “target” is the foreign side. Thus far, most legislators have seemed blithely unconcerned about the details of how this spying has been implemented—refusing even to demand a tally of the number of Americans swept up in NSA’s dragnet. This would be strange if the FAA were seen as the sea change it truly represents, but it makes much more sense if members of Congress falsely believe that it’s merely a “modernization” designed to achieve the original intent of FISA Classic in a new technological environment. One can only hope that if legislators begin to realize what a significant and substantive departure FAA represents from the original FISA framework, they’ll see the need for greater scrutiny.