My Statement Today Before the Senate Intelligence Committee

The following is my my prepared statement for today's hearing of the Senate Select Committee on Intelligence. The hearing, which is beginning at this hour, concerns reform of the Foreign Intelligence Surveillance Act.  Thank you, Chairman Feinstein, Vice Chairman Chambliss, and members of the committee for inviting me to present my views on reform of the Foreign Intelligence Surveillance Act (FISA).[1] I am a Senior Fellow in Governance Studies at the Brookings Institution. I co-founded and am Editor in Chief of Lawfare, a website devoted to sober and serious discussion of “Hard National Security Choices.” I am the author or editor of several books on subjects related to law and national security: Detention and Denial: The Case for Candor After Guantánamo (2011), Law and the Long War: The Future of Justice in the Age of Terror (2008), and Legislating the War on Terror: An Agenda for Reform (2009). The views I am expressing here are my own. In his press conference of August 9, President Obama said with respect to collection under FISA that he believes “there are steps we can take to give the American people additional confidence that there are additional safeguards against abuse. For instance, we can take steps to put in place greater oversight, greater transparency and constraints on the use of this authority” (emphasis added).^[2] I would like today to describe what I see as the major opportunities that now exist for—as the President put it—greater transparency, enhanced oversight, and additional constraint on intelligence collection under the FISA in the wake of the unauthorized disclosures this summer by Edward Snowden and the material declassified by the Executive Branch in response. As preliminary matter, however, it is important to emphasize and clarify the stakes in the current discussion when we speak of transparency. Transparency is an extremely important value in a democratic society. The shock many people in the public now feel at NSA’s collection programs to a considerable degree flows from the lack of transparency with which those programs developed over a long period of time. This lack of transparency in NSA programs prevented the government from garnering sustained public confidence, as these programs developed, though it generally followed the law and the guidance of both the courts and this body. In our society, transparency and legitimacy go hand in hand. Transparency, however, is not as simple a value with respect to intelligence collection as it is in other areas of government. Indeed, normally, with respect to covert intelligence gathering programs, we regard transparency as an evil—that is, fundamentally incompatible with operational security in intelligence gathering, just as it is with respect to troop movements and military planning. Some things have to be secret, and for those things, transparency is not a virtue, and sunlight is not a disinfectant. In those areas, we have traditionally sought accountability by other means. In the wake of Watergate and the intelligence abuses unearthed in the 1970s, Congress enacted a series of reforms to the oversight and accountability system for intelligence operations, and transparency as such was not really a part of it. Rather, this system, of which both the FISA court apparatus and the congressional intelligence committees were parts, was designed to create accountability without creating transparency. That is, the goal was to enable the intelligence community to continue to operate in secret, as it must to be effective. The task of oversight, instead of taking place by means of transparency, was delegated to institutions within the normally-transparent Congress and the normally-transparent judiciary that could operate with sufficient secrecy so as not to impair the community’s operational effectiveness. It is a system that presumes that oversight would not be transparent to the public, nor even transparent to the broader institutions of the judiciary or Congress. At its deepest level, the controversy over the Snowden leaks, the attendant anxiety over the non-public oversight mechanisms that provide accountability for these intelligence programs, and the calls for greater transparency for the FISA system reflect a loss of faith in the continued vitality of this post-Watergate system of intelligence oversight. When we speak of increasing transparency, we must decide as a threshold matter whether we mean increased transparency within the context of a system that presumes secrecy or whether we simply no longer believe in the system of delegated oversight at all. And if we have lost faith in that system, what oversight system might we imagine to replace it? Put another way, is there an oversight system that might enable effective intelligence collection but do so in a more transparent fashion? In my view, nothing in the current disclosures should cause us to lose faith in the essential integrity of the post-Watergate system of delegated intelligence oversight. To the contrary, those disclosures should give the public great confidence both in the oversight mechanisms within the executive branch and in the judicial oversight mechanisms that review both the Section 215 collection program and the Section 702 collection program. The disclosures show no evidence of any intentional, unlawful spying on Americans or abuses of civil liberties. They show a low rate of the sort of errors any complex system of technical collection will inevitably yield. They show robust compliance procedures. They show earnest and serious efforts to keep the Congress informed—including members not on this committee or its counterpart in the House of Representatives. And they show an ongoing dialogue with the Foreign Intelligence Surveillance Court (FISC) about the parameters of the agency’s legal authorities and a commitment both to keeping the court informed of activities and to complying with its judgments as to their legality. The FISC, meanwhile, in these documents looks nothing like the rubber stamp that it is portrayed to be in countless caricatures.[3] It looks, rather, like a judicial institution of considerable energy, one whose oversight role with respect to both Section 215 and Section 702 requires enormous time and energy on the part of the executive to satisfy. This is not to say that every opinion of the FISC is beyond reproach. Scholars have taken issue with aspects of the court’s most-recently declassified opinion on Section 215, for example.^[4] I have no doubt that as other opinions become public, they will face criticism too. But to criticize a particular opinion or line of cases is not to call an entire legal regime or judicial institution into question. Scholars and activists disagree with the work of our federal judges at all levels of the judiciary every day. That is part of the judicial system. Overall, the portion of our judicial system constituted under the FISA has worked well. And we should emerge from this string of disclosures with more, not less, confidence in it. To the extent that members of this committee continue to believe, as I do, in the essential integrity of the post-Watergate mechanisms of intelligence oversight, the first task in the current political environment is to defend those mechanisms—publicly and energetically—rather than race to correct imagined structural deficiencies, or even real structural imperfections that, however real they may be, bear little relation to the outcomes that disquiet us. And critically, the defense of these mechanisms necessarily involves a defense of significant limitations on transparency—just as the defense of the core operations of this committee involves a defense of significant limitations on transparency. This committee, and Congress more generally, ought therefore to make quick work of radical proposals that seek to, for example, abolish the FISA court system entirely in the name of transparency.^[5] Nor should members detain themselves long over the proliferating proposals to impose a strong norm of transparency on a system designed to avoid the consequences of transparency.^[6] Rather, the challenge when we speak of adding transparency to the FISA is a subtler one: it is to inject transparency within the basic confines of an oversight system designed to protect secrets. True opportunities for new transparency within this system are relatively limited, though some significant ones do exist, particularly now that major programs are compromised anyway. The opportunities for transparency are limited because the price of transparency, at least while programs remain secret, is unacceptably high in operational terms. Until someone like Snowden blows a particular program, the costs of having an open discussion about that program involves a damaging initial disclosure. Only after the fact of the program becomes public can one discuss even its broadest contours without doing the leaker’s work for him. But ironically, the involuntary transparency foisted on the intelligence community by someone like Snowden will tend to create opportunities for official transparency. Snowden, for example, disclosed information revealing the fact of the Section 215 metadata collection program.^[7] Members of this committee know better than I do the operational consequences of those disclosures. But that damage’s having been done, the executive branch suddenly faced an entirely different set of calculations regarding the costs and benefits of further disclosures. And it responded with significant document releases and public statements about the program,^[8] disclosures it never could have made before it had absorbed the initial damage. Something similar happened with respect to collection under Section 702. The result is that we now have declassified minimization procedures for collection under Section 702^[9] and primary orders^[10] for bulk metadata collection. The fact that this information is now public offers Congress a significant opportunity to codify existing legal standards in statute—standards that, until this summer, Congress could not have written into law without revealing the fact of the program itself. There is much to be said for codifying in statute the law developed in the iterative back-and-forth between the Executive Branch and the FISC in 2009 and now reflected both in FISC orders and the extant minimization procedures governing the metadata program. There is a perception, based on the text of 50 USC § 1861,^[11] that the FISC is ordering metadata production based on simple relevance—much as courts uphold grand jury subpoenas based on a showing of relevance. But as the recently-revealed primary order and other declassified documents show, there’s actually much more going on in the way of standards and restrictions than merely a showing of the production’s relevance. These rules, restrictions, and standards include, for example, the requirement that the sole purpose of the production be to support authorized investigations to protect against terrorism; the most recent primary order expressly forbids any other use.^[12] More importantly, the order also includes the requirement that the metadata database only be queried when there is a “reasonable, articulable suspicion” (RAS) that the telephone identifier is associated with terrorist activity. It also includes various restrictions designed to ensure that only RAS-approved identifiers are queried.^[13] Congress could craft out of the primary orders and the attendant minimization procedures a far more comprehensive statutory scheme to govern the production, maintenance and use of metadata under FISA. While this process would not add information to the public’s understanding of the program, it would have the salutary dual benefits of ratifying existing practice and clarifying for the public in law the precise circumstances—now spelled out in an interlocking fabric of statute, executive procedure, and court order—under which the government can acquire and use bulk telephony metadata. The transparency benefit here is admittedly modest, since following the declassification of the orders and the procedures, the standards in question are already public. But there is at least a marginal transparency benefit to writing the rules into law, and there is a significant benefit on their own terms in codification and congressional ratification. In addition, Congress should consider requiring both regular public reporting of aggregate data concerning the use of 215 and 702 authorities and regular public disclosure of compliance and non-compliance reporting. The Director of National Intelligence recently announced that the Executive Branch would be releasing annually data describing the number of orders issued and the number of people targeted under the following authorities:

• FISA orders based on probable cause
• Section 702 of FISA
• FISA Business Records
• FISA Pen Register/Trap and Trace
• National Security Letters issued pursuant to 12 U.S.C. § 3414(a)(5), 15 U.S.C. §§ 1681u(a) and (b), 15 U.S.C. § 1681v, and 18 U.S.C. § 2709.^[14]

Similarly, the government has declassified a significant body of material related to compliance issues.^[15] There is little reason, particularly prospectively, why such compliance reporting could not be made public on a more routine basis—or at least summarized for the public. Similarly, there’s no reason why the law should not codify and require public release of the sort of regular data streams the government is creating. Senator Feinstein has proposed other reporting mechanisms as well, suggesting annual disclosure of “the number of Americans’ phone numbers submitted as queries of the NSA database,” “the number of referrals made to the FBI each year based on those queries,” and “the number of times in a year that any company is required to provide data pursuant to FISA’s business records provision.”^[16] I do not purport to know which of these data streams can be safely released, but as a general matter, making such data public is critical to establishing long-term public understanding of what these programs are—and what they are not. That is, such disclosures are key to establishing and maintaining public legitimacy. A final area for increased transparency involves publication of FISA court opinions. While many proposals for publication of FISA opinions are unworkably overbroad,^[17] Congress should consider changing the default rules with respect to opinions issued by the FISC and the FISA Court of Review. Currently, rulings by these courts are by default secret unless and until they go through an extensive process of declassification review. The result is that, knowing they will not become public, the court tends to write them in a fashion that includes a large volume of operational information that then, in turn, makes the declassification process difficult. This need not be the case, however, when FISA judges write with an eye toward public release of their work, as illustrated by Judge Eagan’s recent primary order and opinion. One could imagine a default rule that would assume publication of opinions on significant legal questions in the absence of a decision by the Attorney General to restrain their release. In cases in which the Attorney General decided to prevent public release, the law could create a mechanism for the opinions to be available to members of Congress in a secure facility, much the way this committee worked to make sure Senators had access to the 2009 FISC litigation over the 215 program.[18] In addition to these opportunities to enhance transparency, Congress has the opportunity to enhance oversight of FISA authorities in fashions that may not involve increased public visibility. Again, it is important not to be distracted here by irrelevant proposals that would do little to make oversight more effective—like, for example, the various proposals to change the appointment mechanisms by which the FISC currently gets staffed.^[19] The reality is that there is no evidence that the partisan identity of the judges who have heard these cases has impaired the quality of oversight. A large number of different FISA judges over a long period of time has endorsed the government’s reading of Section 215. And as the recently disclosed documents have shown, the FISA judges have also shown themselves capable of holding the government to the law as they read it, even in the absence of transparency. To the extent members of this body disagree with their substantive reading of the law, the right response to that problem is to change the law’s substantive contents. Responding to that disagreement by altering the appointment process for the judges is a little like responding to Congress’s enactment of a law one does not like by advocating term limits or a different electoral system.^[20] The proposal is simply not responsive to the supposed problem that gives rise to it. A more serious suggestion involves the creation of a public advocate to add an element of adversity to the FISA system. President Obama endorsed this idea in August, saying, [T]o build greater confidence, I think we should consider some additional changes to the FISC. One of the concerns that people raise is that a judge reviewing a request from the government to conduct programmatic surveillance only hears one side of the story, may tilt it too far in favor of security, may not pay enough attention to liberty . . . . I think we can provide greater assurances that the court is looking at these issues from both perspectives — security and privacy. So specifically, we can take steps to make sure civil liberties concerns have an independent voice, in appropriate cases, by ensuring that the government’s position is challenged by an adversary.^[21] There have been a number of different proposals for special advocates before the FISA court system.^[22] The idea is actually not new. The late Kenneth C. Bass III, who helped set up the FISC process during the Carter administration, argued before the House Permanent Select Committee on Intelligence in 1994 that proceedings should be more adversarial, at least in some instances.^[23] In the context of the current controversies, the idea of increasing the adversarial quality of FISC decision-making has gained support from scholars as diverse as Orin Kerr^[24] and Steve Vladeck,^[25] former FISA judges James G. Carr^[26] and James Robertson,^[27] and the Los Angeles Times and Washington Post editorial boards.^[28] The idea has dissenters, and their objections require consideration. Carrie Cordero, a former official at the Justice Department’s National Security Division who has significant experience before the court, wrote recently that “there are both principled and practical reasons for exercising caution.”^[29] On the principled level, she asks whether “we [are] really going to start litigating foreign intelligence activities before taking action.”[30] On a practical level, Cordero worries about adding an additional layer of review to an already heavily-lawyered process, one in which the Justice Department already plays the role of “a neutral party that evaluates the Intelligence Community’s requests for surveillance . . .”[31] My own sense is that there is a way to accommodate a more adversarial process in some cases without adding undue complexity to an already bureaucratic system. The FISA system should not have an office of special advocate, a sort of public defender tasked to routinely oppose government applications before the FISA courts. A great many of these applications are entirely uncontroversial legally and raise no novel legal questions. Moreover, many are time sensitive. So in my view, it would be a mistake for Congress to create an opposing counsel who appears as a matter of routine, and when the court grants a government application, appeals as a matter of routine to the FISA Court of Review—thus creating a layer of appellate litigation as a condition of proceeding with a wiretap. No such procedure exists under Title III, and there is no justification for creating one in national security cases. On the other hand, the FISA court considers some cases of extraordinary legal significance, and it seems appropriate for the FISA judges to have the option—at their discretion—of appointing cleared counsel to argue against the government’s submissions in such cases. The key point here is that it should be the court itself that decides to what extent it needs adversarial briefing in any given case. Adversarial process should not be a presumption, but an option available to the court. Moreover, the law should be clear that in cases in which the government prevails before the FISC and the special advocate takes an appeal to the FISA Court of Review, the court-ordered surveillance may proceed while the appeal is pending—thus mitigating the operational impact of an appellate litigation. Finally, there is the President’s call for greater constraint, which I take to mean a tightening of substantive standards in FISA collection. Now that the Section 215 program has been disclosed, there is valuable work to be done on this point. The government’s reading of Section 215, which the FISC has consistently accepted, is a plausible one; it is not, however, an obvious one. And relying on it for the long term thus carries some real risks. It carries the risk of ultimate judicial rejection once appellate courts begin hearing cases challenging the view that the government has successfully advanced before the FISC. It also carries risks, to the extent that the courts do adopt the government’s view, of making gigantic streams of data generally available for production in cases far more routine than these counterterrorism investigations. Under the government’s current reading of Section 215, one could imagine, at least legally, bulk metadata production simply on the basis of a showing of the relevance of some portion of those data streams and the asserted need to keep the aggregate dataset sufficiently comprehensive so as to ensure that piece is available to investigators. I have no problem with bulk metadata collection for the narrow purpose and under the controlled circumstances in which the government is doing it. The theory that the production of entire haystacks—including haystacks that have not yet been bundled, made out of hay that has not yet been grown—is necessary for the identification of the needles that may someday be hidden within them is compelling when one is talking about terrorist groups against which the United States is in an armed conflict. But I have a deep problem with the idea of collection on a similar theory if undertaken in, say, an organized crime investigation or a child pornography case. The difference is that fighting international terrorist groups actively plotting violence reaches a different threshold of necessity and urgency. I believe the law should both affirmatively authorize bulk metadata collection and limit it to situations in which the government makes some showing that it is necessary to support a national security interest of a particularly high order. In my view, the FISA business records provision should be used—as it normally is—for individual business records and groups of them, and Congress should authorize metadata collection for counterterrorism purposes on a basis other than relevance. Thank you very much for this opportunity to share my views on these important subjects.