In a previous post here on Just Security, I discussed the constitutionality of the National Security Agency (NSA) program that sweeps up detailed “metadata” pertaining to essentially all calls – domestic and international – made by every person in America.  The metadata reveal virtually everything except the content of the call – the numbers called and received, the duration of each call, and other non-content details. The NSA retains that information in a comprehensive database, and analysts dip into it to trace suspicious patterns and links to other numbers when but (the Administration claims) only when they have reasonable suspicion to believe that a particular phone number is associated with terrorist activity.  

The constitutionality of the program is a close – and at this juncture unresolvable – question.  When conducted on such an indiscriminate basis, a data sweep of this sort unquestionably should qualify as a Fourth Amendment “search,” but as I discussed previously, the search would almost certainly be considered “administrative.” Administrative searches can pass muster even without probable cause, provided they are “reasonable.”  Reasonableness, in turn, depends on all relevant circumstances, including the need for the program, its intrusiveness, the safeguards that accompany it, and the alternative safeguards that might afford greater protection for individuals without jeopardizing its objectives. The constitutionality of the NSA metadata program necessarily remains unclear because so many details have yet to be revealed.

To be sure, a judge of the FISC has upheld the NSA program, but that conclusion is far from definitive.  At most, it represents only the view of several federal district judges, reached after a secret hearing – without the benefit of opposing argument – based on unknown facts and undisclosed reasoning. Anyone who confidently claims that the program is – or isn’t – constitutional has reached a premature judgment.

A more fruitful inquiry, therefore, is to set aside the technical questions of constitutional doctrine and precedent in order to directly address the policy concerns that should shape public assessments and legislative responses. Congress need not accept every executive initiative that passes the low bar of constitutionality, and in the area of electronic surveillance a dense web of statutory safeguards, going well beyond the constitutional minima, has been on the books since 1968, with multiple amendments since then. Even if (as some maintain) the NSA program is indeed authorized by those laws as they currently stand, the new revelations about how those laws are being used call for a fresh look at the issues. Congress may not be ideally situated to protect all the important interests that the NSA program jeopardizes, but legislation is one important place to start.

In thinking about the policy questions, we should first acknowledge that indiscriminate collection of metadata is by no means a useless exercise. Once NSA analysts develop objective reasons to suspect  a particular individual, they will then want to know a lot about the communications this person has had at certain crucial moments in the past; yet that information may be unretrievable or difficult to obtain if it has not been preserved.  If they learn that a known terrorist has been in touch with Person A, Person A of course may be entirely innocent (a travel agent for example), but analysts typically want to see the pattern of Person A’s calls and identify the people Person A is in touch with.  That kind of investigation may be invasive of Person A’s privacy, and reasonable people might consider it unjustified or inefficient.  But such an investigation is far from pointless. Person A himself (or someone he talks to) might be a co-conspirator; Person A might agree to provide useful information or become a cooperating witness. Any helpful discussion of the NSA program must begin by recognizing that it can serve valuable – perhaps very valuable – purposes.

That recognition often is treated as sufficient to put concerns to rest and end the conversation.  On the contrary, however, the program’s usefulness can only mark the beginning of the inquiry. Americans who accept its usefulness must nonetheless focus their attention on two further questions: (i) whether the program endangers important values and (ii) whether all or virtually all of its benefits can be realized in ways that pose fewer risks.

In this post I turn a spotlight on the first of those two crucial points. The fact is that despite its undeniable value (its potential to catch the elusive terrorist who might otherwise evade law enforcement attention), the NSA program has costly side effects. Three in particular stand out – metadata surveillance will stultify the vibrancy and creativity that drive America’s cultural and economic success; it will undermine freedom of the press and democratic deliberation; and – paradoxically – it will even weaken the effectiveness of the counterterrorism effort itself. The drawbacks are so acute that no thoughtful supporter of a strong counterterrorism machinery should rest content with the observation that metadata collection can pay off in various actual or hypothetical scenarios.  Whether our top priority is the defense of civil liberties or prevention of the next attack, in either case it is imperative to seek out alternatives that promise similar advantages without the NSA program’s dangerous consequences. In a subsequent installment of this comment, I will focus on concrete ways that viable threshold requirements for data acquisition, together with stronger mechanisms of accountability, transparency and independent oversight, can preserve virtually all of the NSA program’s value at considerably lower cost to our democracy and to law enforcement itself.

The Dangers

Complacency about surveillance typically begins with the assumption that people who are doing nothing wrong have nothing to fear.  Law abiding individuals, it is said, should have “nothing to hide.”  The Fourth Amendment, however, and the kind of privacy it protects, have never been about sheltering criminals; invasions of privacy have always been permissible when the government has good reason to suspect illegal activity.  Instead, we safeguard privacy because it is the crucial underpinning of the personal independence and political freedom of law abiding citizens.  Metadata, in particular, has unique potential to reveal perfectly lawful but intensely personal activity and beliefs – what our interests are, whom we associate with, what ideas we are pursuing, what causes we support.  With modern technology, metadata – the “envelope” information – paradoxically can be even more valuable to analysts than the content of a message, because the former is much more easily searchable for revealing links and patterns. And no one really doubts that personal growth, creativity and lawful dissent suffer when private spaces shrink and when individuals who value that refuge cannot shield their lives from government eyes.

The impact of metadata collection on freedom of the press is especially important to stress. Government cannot be held accountable without energetic reporting that brings to light things that public officials would prefer us not to know. In other words, democracy depends on leaks. And leaks depend on confidentiality.  Yet a trove of metadata enables government officials to identify in an instant the people to whom an investigative reporter has been talking.  Unless reined in, metadata sweeps will unquestionably chill – and perhaps virtually eliminate – the leaks on which Congress and the public depend for information that contradicts the official story.

Efforts like the NSA sweeps actually undermine the counterterrorism effort itself.  The reason is that these types of programs generate profound mistrust of government in general and of law enforcement in particular.  We have already seen dramatic examples in the way that the Snowden revelations of our spying on allies has angered European leaders and endangered our working relationships with them.  Although this breach eventually will be healed, alienation and mistrust among ordinary citizens is equally important, and it will not be so easily remedied. That mistrust, in turn, has a strong chilling effect on the willingness of law-abiding, loyal citizens to cooperate in the counterterrorism effort – for example by working with officials in local counterterrorism programs or by alerting law enforcement to various kinds of suspicious behavior. This dynamic has been demonstrated in several decades of extensive law enforcement research, including research focused specifically on counterterrorism policies and their impact on Muslim communities in the West. In one study, for example, Muslim-Americans in New York City were 61% less likely to report potentially suspicious precursors of terrorism when they felt that counterterrorism policies were being unfairly set and implemented. (See Schulhofer, Tyler & Huq, American Policing at a Crossroads: Unsustainable Policies and the Procedural Justice Alternative, 101 J. Crim. L. & Criminology 335, 364-74 (2011).)

This last point underscores the most important, but least appreciated, “given” of the counterterrorism enterprise.  Because the consequences of a terrorist attack could be so catastrophic, citizens and public officials alike tend to support strong law enforcement more readily than they do in ordinary times.  The perception is that strong measures are acceptable because the top priority must be to reduce the risk of attack.  And strong measures, whatever their drawbacks, at least seem to offer ways to reduce that risk.  The central lesson of smart law enforcement, however, is that there are no risk-free choices. Strengthening the powers of the executive enhances some of our defenses against terrorism, but it weakens others – many of which, including community trust and cooperation, are absolutely essential to reducing the dangers of terrorism.

Against this background, we must – for the sake of our democracy and even for the sake of our physical security – find ways to dissipate the cloud of mistrust that now hangs over NSA surveillance.

[Author’s Note: The discussion here is necessarily abbreviated; for greater depth and detail, I refer readers to my recent book More Essential Than Ever: The Fourth Amendment in the Twenty-first Century (Oxford University Press 2012), especially chapters 6 and 7.]