Cross-posted from Just Security
The House Judiciary Committee has completed its markup of a major surveillance reform bill intended to better protect Americans’ privacy and enhance transparency. Responding to civil liberties advocates’ concerns that the bill doesn’t go far enough, Reps Ted Poe (R-Tx.) and Zoe Lofgren (D-Calif.) offered an amendment to require U.S. officials to obtain a warrant before searching communications acquired under Section 702 of the Foreign Intelligence Surveillance Act (FISA) for Americans’ information. Committee Chairman Bob Goodlatte (R-Va.) and Ranking Member John Conyers (D-Mich.), however, insisted that adopting the amendment would kill the bill’s chances of passage. Even though a substantial, bipartisan majority of the committee’s members had voted in favor of the same provision on other occasions, most of them followed the direction of committee leadership, and the committee approved the bill without any major changes.
Pop quiz: does the description above refer to the April 30, 2015 committee markup of the USA FREEDOM Act, or yesterday’s markup of the USA Liberty Act?
You probably guessed the answer already: the description could apply equally to either occasion. Indeed, those of us who follow surveillance legislation closely experienced a touch of déja vu as we watched yesterday’s markup. But some civil liberties advocates who supported the USA FREEDOM Act after it emerged from committee—and who acknowledged the tough choice that members supporting the Poe-Lofgren amendment were forced to make—are feeling very differently about USA Liberty. That’s because despite the eerie similarities in what happened at the markup, there are some critical differences in the underlying circumstances.
Most important, the USA Liberty Act is not simply too modest in its aspirations. In some ways it would take a step backward, by codifying a large swath of backdoor searches and by introducing the notion that Americans’ rights deserve less protection when the government wants to obtain foreign intelligence or find terrorists.
Does the bill improve the status quo?
The main goal of the USA FREEDOM Act was to end the NSA’s bulk collection of Americans’ phone records. It did that. Many advocates believed it did not sufficiently limit the collection of phone records, given that it still allowed the NSA to acquire the metadata of both suspected terrorists and anyone in contact with them. In addition, it wasn’t clear that the bill’s provisions to prevent bulk collection for other types of data would accomplish their goal, due to ambiguities in the definition of key terms. But it was hard to argue that the bill did not improve on the status quo. And while it didn’t put an end to backdoor searches, it didn’t authorize them, either.
Now take the USA Liberty Act. The most important part of the bill is how it addresses “backdoor searches”—the government’s practice of querying databases containing communications obtained under Section 702 to find Americans’ calls and e-mails. The bill would require the FBI to obtain a warrant in order to access the content of any queried Section 702 data, but there are several exceptions. Most notably, the FBI need not obtain a warrant if its purpose is to obtain “foreign intelligence,” or if it reasonably suspects that the subject of the query is engaged in or materially supporting international terrorism, or is communicating with someone who fits that description.
Civil liberties groups have pointed out that these exceptions are likely to swallow (or at least take a major bite out of) the rule requiring a warrant to query Section 702 information. The definition of “foreign intelligence,” in particular, is so broad that it could encompass ordinary conversations about front page news stories. In cases where the FBI is merely following up on a tip regarding someone who is an immigrant or Muslim American, it will be simple enough for the FBI to assert that its purpose is to rule out worrisome foreign ties. The bill wouldn’t even affect the tens of thousands of queries performed each year by the NSA and CIA.
As I argue here, though, concern about this provision of the USA Liberty Act goes beyond whether it covers a sufficient number of possible backdoor searches. As written, Section 702 contains no express authorization for warrantless U.S. person queries, and there’s little reason to think Congress had this practice in mind when it enacted the statute—or even when it reauthorized it in 2012, a year before Edward Snowden’s disclosures spread the word about backdoor searches beyond the congressional intelligence committees. This means that the USA Liberty Act’s provision aimed at curbing backdoor searches would actually be codifying the warrantless searches of Americans’ communications. While it might be a slight step forward in terms of the government’s practices, it would be a major step backward in the law.
It would also create, for the first time, a different and lower statutory standard for accessing Americans’ communications in foreign intelligence cases. Astonishingly, Rep. James Sensenbrenner (R-Wisc.) claimed at the markup that there are different standards already—that only “reasonable suspicion” is required for surveilling Americans in national security cases. This is false. In a foreign intelligence investigation, Title I of FISA requiresthe government to show probable cause that the target is an agent of a foreign power (which entails criminal wrongdoing). In a domestic national security investigation, the Supreme Court has held that the government must obtain a warrant. USA Liberty’s embrace of a lower standard in foreign intelligence cases would therefore set a precedent—a precedent that the executive branch will aggressively exploit in its own rules and in future legislative discussions of surveillance policy.
These arguments, of course, assume that warrantless searches for Americans’ communications are a bad thing. I believe they are unconstitutional, for reasons I’ve set forth in other places, and so do many members of Congress. It’s true that the FISA Courthas concluded otherwise, as have two (possibly three, depending on how you read an ambiguous footnote in United States v. Hasbajrami) district courts. But a three-year-old body of case law comprising three or four trial court cases is a long way from settled law. And even if the Supreme Court were eventually to condone the practice, Congress is entitled to supply protections through legislation when it believes the Supreme Court has shortchanged a constitutional right. (This is what happened, for instance, when Congress passed the Religious Freedom Restoration Act in 1993.)
Is there a bird in the hand?
In 2015, Goodlatte and Conyers could tell their caucuses that the USA FREEDOM Act was a delicate compromise that had been endorsed by nearly all the relevant stakeholders. It was supported by the White House, the intelligence agencies, most civil liberties groups, the technology sector, and key members outside the Judiciary Committee. In other words, it had a decent chance of being enacted. Adopting the Poe-Lofgren amendment at that time would have changed the light from green (or at least yellow) to red.
The USA Liberty Act is a different story. It is either opposed or not supported by all the above-mentioned constituents that supported USA FREEDOM. Although Paul Ryan apparently told the committee’s leaders that the bill won’t get a floor vote if it includes a full warrant requirement, there’s no indication that he’s agreed to give the bill a floor vote in its current form. To the contrary, it’s quite likely the bill will be weakened in the rules committee or at other steps in the process. Meanwhile, in the Senate, there is far more robust support for a warrant requirement than anyone expected—including from Sen. Dianne Feinstein, usually an advocate of broad surveillance powers. These are not the conditions in which it makes sense for reform-minded members of the House Judiciary Committee to be negotiating against themselves.
When is the next chance?
At the USA FREEDOM markup in 2015, member after member—both Democrats and Republicans—voiced their support for the Poe-Lofgren amendment. In urging them to vote against it, Chairman Goodlatte expressed sympathy for their position. He assured them that, after USA FREEDOM’s passage, he would turn to Section 702, and they would have a chance to address the issue of backdoor searches on the merits.
That chance was yesterday, and once again, he and Conyers instructed the members not to vote their conscience. This time, however, there is no next opportunity on the horizon. The USA Liberty Act contains a six-year sunset. Absent unforeseen circumstances, there is little reason to think Congress will take up Section 702 in the interim. For committee members who wanted to see the backdoor search loophole closed, yesterday was their best chance.
Ultimately, though, the considerations of strategy and timing are secondary. What made yesterday’s markup so disturbing is this: the USA Liberty Act is a piece of reform legislation that authorizes a practice many believe to be unconstitutional. In light of that, Chairman Conyers’ admonition that we shouldn’t let the perfect be the enemy of the good is more jarring than persuasive. He may be correct that the government conducting warrantless searches on a daily basis is a better outcome than the government conducting even more warrantless searches on a daily basis. But if we’ve been reduced to calling it a “good” outcome, then the fight for surveillance reform may already be lost.