From day one of the Snowden revelations, we all knew that the legal validity of the 215 program hinged ultimately on the capaciousness of a single word: "relevant." Even those of us who generally support robust signals intelligence programs also knew immediately that the legal theory underlying this program lay right at the margins, perhaps beyond the margins, of the legally tenable. So it comes as no particular surprise that a three-judge panel of the Second Circuit Court of Appeals has ruled that the word "relevant" will not support the weight the government places on it---and for much the same reason that Orin Kerr, I, and many others argued from the beginning that the program stands on shaky statutory grounds.
The litigation risk of relying on 215 is the principal reason I have been arguing for almost two years now that Congress needs to clarify the law, both to authorize access to metadata contact chaining under appropriate circumstances and to build in protections that exist under the current program only through court order.
There has been a lot of excellent commentary on the opinion, on this site, on Just Security, and elsewhere. I won't rehash those analyses here, but I have a few additional thoughts.
First, Judge Gerry Lynch---for whom I have high regard as a general matter---begins his opinion with an ominous comparison of the current era to the COINTELPRO era, the Keith case, and the abuses that led to the Church Committee:
"In the early 1970s, in a climate not altogether unlike today’s, the intelligence-gathering and surveillance activities of the NSA, the FBI, and the CIA came under public scrutiny."
The comparison is frankly beneath the dignity of a fine judge like Lynch. Whatever one thinks of the government's interpretation of the statute---and as I say, I have my doubts---NSA has played by the rules in this area. It went to the FISA Court dozens of times. It persuaded judges of the legal propriety of its position. It kept Congress informed of its actions and its interpretations. This is not an issue of abuses. This is an issue of an agency adopting an aggressive interpretation of a statute---an interpretation with which a court later disagrees. That happens all the time.
Judge Lynch, I think, knows that this isn't really the Keith case---which involved warrantless wiretapping by Attorney General (and Watergate felon) John Mitchell for domestic security purposes---that these disputes are not quite Church Committee stuff, and that the climate of the 1970s civil-libertieswise was not really that similar to today's. The evidence? He declined to issue an injunction against the 215 program.
Memo to all judges: If you're going to invoke the abuses of the 1970s and the Church Committee, issue the injunction. And if you're not prepared to issue the injunction, perhaps your difference of interpretation with an agency doesn't quite rise to the level the comparison suggests.
I made this point at the Triple Entente Beer Summit the other night, and the excellent Michael Vatis of the Steptoe Cyberlaw Podcast agreed, but with an important twist: Normally, when judges invoke these ridiculous rhetorical comparisons, he noted, the rhetoric presages bad legal reasoning to come. Here it did not.
Vatis is right. Judge Lynch's opinion is, the opening aside, workmanlike and serious. There are matters within it on which reasonable people might disagree, though I tend to agree with the decision's bottom line. But it's not a bomb-throwing opinion, by any means. And for this reason, it will concentrate a lot of minds with just three weeks to go before Section 215 disappears in a puff of smoke.
That's a very good thing: it's good for civil libertarians; it's good for the administration; and it's good for NSA.
It may seem odd to suggest that a unanimous panel defeat on a basic legal theory underlying an NSA program is good for the agency, but consider: A few days ago, the 215 program was on a glide-path to expiration. The House seemed to be coming together around a version of the USA Freedom Act, which would substitute a different metadata acquisition mechanism for the 215 authority and create other reforms as well. But the Senate had lacked the votes to move that bill even last year, and Senator McConnell was pushing instead what he called a "clean" reauthorization---which would reauthorize 215 without modification and do none of the other reforms both the administration and civil libertarians have supported. It was not at all clear that the House bill could pass the Senate, and it was entirely clear that McConnell's bill could not pass the House. Betting on a compromise within three weeks was, given the normal state of congressional competence these days, optimistic, to put it mildly. So the program seemed likely to expire, less as a result of any decision to let it sunset than as the result of a collective action problem.
The Second Circuit opinion meaningfully changes that. It is a sharp reminder to those who favor McConnell's approach that it is not viable---or, at least, that it involves serious litigation risk. Yes, it is possible---likely, even---that the D.C. Circuit will disagree with the Second Circuit, and that the meaning of 215 will thus be ultimately subject to the whimsical mind of Anthony Kennedy. So yes, it's also still possible that a vote for a "clean" reauthorization will yield maximal flexibility for the agency. But that's not a bet I would want to lay if I were a congressional proponent of strong signals intelligence and counterterrorism programs. Rather, I would say that it is far better to have a legally sustainable version of this capacity than to have no capacity at all, and supporting clean reauthorization will either cause there to be no bill---in which case the authority lapses---or, if the bill were to pass, it risks that the Supreme Court's ultimately agrees with the Second Circuit, not with the FISA Court interpretation of the law. The far better approach for the agency is the compromise that the administration has long supported. And for the first time in a long time, I can see a path to that outcome.
The House will soon pass the current version of the USA Freedom Act---about which I will have more to say in the coming days---with a strong bipartisan majority. And it will then dump it in the Senate's lap with a very short time horizon before sunset on June 1. Key Republicans, McConnell chief among them, will then face the choice between maintaining their opposition to the bill and seeing the current authority vaporize or signing onto reform that the administration and the intelligence community supports and maintaining something close-ish to current authorities. This will put huge pressure on the Senate to act on the House bill, and it will allow House Speaker John Boehner to walk away from the problem with the moral high ground---an opportunity I don't expect he will pass up given the number of times the Senate has forced his hand on deadlines.
There are a number of key variables still lurking out there. One of them is the DC Circuit, which could rule any time on its own 215 case and whose ruling could easily embolden those both in the intelligence community and in the Senate who want to double down on current law. Another is simple congressional dysfunction, which seems to prevent legislative movement even when everyone basically agrees. But I think there's a good chance that the Second Circuit here has been the catalyst for legislative action that was otherwise stalled. That would be a very happy outcome.