Intelligence Oversight

In Praise of the Intelligence Oversight Process: Our New Anti-Populist Paper

By Jonathan Rauch, Benjamin Wittes
Wednesday, June 28, 2017, 2:51 PM

Years ago, when Lawfare was still in its infancy, the two of us made an entirely serious video (well, maybe not entirely serious) for YouTube about the emergent problem of abusive internet comments. Entitled "Comment or Vote," it proposed a constitutional amendment to deprive of the franchise anyone who left a comment on any website. For some time after we posted it, until it was finally removed, the first comment on the YouTube site read: "You guys are faggots." No, we're not making that up. 

The issue "Comment or Vote" spoofed was not a joke. Lawfare does not take comments, and it's always interesting to watch how offended some readers are by that. It's as though people think they have some inalienable right to participate, including in a website published by someone else. 

The broader issue of which this expectation is only a tiny slice—the populist cult of mass participation—is really not a joke. It is wreaking havoc on our political system. And it is the subject of a Brookings paper we released late last month, entitled, "More Professionalism, Less Populism: How Voting Makes Us Stupid and What to Do About It." The full paper is available below. An interview we did yesterday on the subject with the Brookings podcast, Intersections, is available here:

Most of this paper has little to do with the national security law and policy. But one section is a real exception, which we wanted to highlight here: A key case study of how anti-populist institutions function better than highly participatory ones focuses on the intelligence oversight system. We thought the thematic connections between Lawfare's subject matter and this discussion warranted excerpting it here. So what follows is the paper's introduction, along with the section on the intelligence oversight system:

“Americans—especially, but not exclusively Trump voters—believe crazy, wrong things,” runs a post-election Washington Post headline. The article, by columnist Catherine Rampell, worried about polls showing that more than a third of the public (and about half of Republicans) believe that Barack Obama was born in Kenya and that Hillary Clinton was involved with a satanic pedophilia ring (“Pizzagate”)—among many other things. “To me, they’re terrifying,” Rampell wrote of the public’s misconceptions. “They result in misused resources, violence and harassment, health risks, bad policy, and, ultimately, the deterioration of democracy.”

Political scientists might be excused for emitting an exasperated yawn. The literature on voter ignorance is one of the oldest, best established, and most dismaying in all of political science. Every so often, journalists and commentators dip into it and emerged “terrified.” In recent years, however, a wave of research has shown ignorance and irrationality to be even bigger problems than previously believed, and has cast new doubt on standard remedies. Neither theory nor practice supports the idea that more participation will produce better policy outcomes, or will improve the public’s approbation of government, or is even attainable in an environment dominated by extreme partisans and narrow interest groups.

Such scholarship has not shaken the faith of many commentators and reformers that increasing popular involvement in politics and government is the remedy for the ills of our political culture—the chicken soup of political reforms. Unfortunately, the country and the political-reform community have come to expect far too much from increased political participation. Participation is effective only when supplemented by intermediation, the work done by institutions (such as political parties) and substantive professionals (such as career politicians and experts) to organize, interpret, and buffer popular sentiment. In this essay, we argue that restoring and strengthening political institutions and intermediation belong at the center of a modern political-reform agenda. More specifically, we advance the following contentions:

  • Always empirically questionable at best, the populist-progressive idea that more participation will reliably improve either the products or the popularity of governance has taken a pounding in recent years, to the point where it is basically untenable. The populist model assumes that voters are better informed, more rational, and more engaged than is the case—or ever will be.
  • Even implausibly well-informed and rational voters could not approach the level of knowledge and sophistication needed to make the kinds of decisions that routinely confront the government today. Professional and specialist decisionmaking is essential, and those who demonize it as elitist or anti-democratic can offer no plausible alternative to it.
  • Professional intermediaries make democracy more inclusive and more representative than direct participation can do by itself. In complex policy spaces, properly designed intermediary institutions can act more decisively and responsively on behalf of the public than an army of “the people” could do on its own behalf. Intermediated systems are also less likely to be paralyzed by factional disputes and distorted by special-interest manipulation than are systems designed to maximize voter participation and direct input.
  • Nonetheless, the predominant ethos of the political-reform community remains committed to enhancing individual political participation. This is a costly oversight. Some populist reform ideas are better than others, but, as a class, they have eclipsed a more promising reform target: strengthening intermediating actors such as political professionals and party organizations.

James Madison and the other Founders were right to reject both direct democracy and elite rule. Instead, they insisted on a hybrid of both, believing that the two together would achieve better representation and better governance than either could achieve on its own. We believe that the country can benefit from relearning what they knew.

. . . 

[M]odern anti-institutional populism, with its instinctive suspicion of anything or anyone deemed “elite,” often underestimates and unfairly denigrates how much intermediation has to offer—and how successful it has been.

Here it is worth distinguishing between two types of intermediaries, types that play very different roles in our democratic culture. The first is the political intermediary: people such as elected officials and political party professionals. These people’s job is to make political judgments on behalf of the electorate or to help candidates and politicians frame their—and the public’s—choices. A second type is the substantive intermediary: the technical expert or specialist whose job is actually to know things about the policy space in which the government acts. The lines between these two types are not bright ones; some people play both roles at once. But broadly speaking, the political intermediary stands in for the voter in deciding fundamental policy choices: Should Congress adopt the Affordable Care Act? Should it repeal and replace it? The substantive intermediary, by contrast, frames responsible policy choices and implements them in a fashion consistent with law and technical realities. Neither of these functions is one which voters can perform competently on their own.

To see vividly the successful interaction of voters, political intermediaries, and substantive intermediaries, let’s take a close look at the area in which our system functions at its least populist: intelligence oversight.

The intelligence oversight system has been largely immune to populist reform, but not because the area has lacked for reformist instincts. (It has actually seen a lot of reform over the years.) Rather, the very nature of intelligence makes it resistant to populism. The public has no access to the CIA and the NSA and their day-to-day work. In sharp contrast to the public’s regular interaction with law enforcement officers, members of the public don’t generally interact with intelligence professionals doing their jobs. And the oversight system for all of it is uniquely opaque. Intelligence oversight is technically dense across a number of different dimensions; some of the law at issue is arcane and unusually intricate, and much of the subject matter involves highly technical electronic surveillance taking place on complex computer networks. It also involves material that is nearly always classified at the highest levels. So while the activities of the intelligence community stoke all manner of public passions, the subject is simply not amenable to populist reform. There is just no way to involve the public in decisions that, by their very nature, have to be kept from the public.

The result is that reforms in intelligence oversight have empowered intermediary actors—politicians and experts—to stand in for the public. And these mechanisms have proven remarkably durable and effective.

The intelligence oversight system operates in all three branches of government. Within the executive branch, a series of compliance regimes, inspectors general, and Justice Department officials oversee operations, particularly those involving technical collection at the National Security Agency. Collectively, those actors investigate suggestions of misconduct, rigorously count more mundane compliance failures, and review activity for legal compliance—and they report errors both internally and to the other branches of government. They also build legal compliance into the very design of technical systems; to see certain data, for example, analysts often have to enter into computer systems the facts and analysis that would justify their access to the information they seek. 

In the judiciary, the Foreign Intelligence Surveillance Act (FISA) Court both reviews applications for national security surveillance of U.S. persons and, more broadly, reviews the legality of surveillance programs taking place domestically but targeting overseas actors. The FISA structure does not merely put a federal judge between the intelligence community and electronic surveillance targets. It also creates bureaucracies within intelligence agencies that are designed to speak to that federal judge—bureaucracies deeply invested in keeping their agencies within the law and maintaining credibility before the judiciary. Representing the agencies before the court are Justice Department lawyers, who likewise act to keep the agencies in line. The system of judicial intermediaries thus dramatically reinforces and empowers executive-branch intermediaries, who have ongoing reporting obligations to the judicial branch.

The executive and judicial branch intermediaries are, generally speaking, substantive experts. But the intelligence oversight system uses political intermediaries too, particularly in Congress. The congressional intelligence committees review intelligence programs, get briefed on covert actions, and investigate major matters within the intelligence agencies—everything from torture to Russian hacking of the 2016 election. Unlike other congressional committees, they are staffed with intelligence community professionals cleared at the highest levels, and they have access to the most sensitive programs the government runs. The public never sees the vast majority of the work these committees do, but anyone in the intelligence community will tell you that fear of upsetting congressional overseers is a major restraint on intelligence community behavior. Note that the members of these committees are not specialists; they are just regular members of Congress. Their job is to stand in for the members of the public who cannot know what the intelligence community is up to, and to make major political decisions in the public’s stead and on behalf of the rest of the legislature.

Most people who have engaged the congressional intelligence committees agree—notwithstanding the House committee’s recent flap over the behavior of its chairman, Devin Nunes—that they function on average dramatically better than other committees do. The reasons are all related to their intermediary, non-populist nature. The professional nature of the staff reduces partisanship, for example. And the secrecy with which they operate discourages political grandstanding on the part of the members. A hearing of a normal congressional committee, which takes place in public, is a show whose audience is the public. Witnesses are chosen for C-SPAN. Questions are asked because they offer opportunities for theatrics and gotchas. By contrast, most intelligence-committee hearings have no audience beyond the staff and members. They are actually designed to convey information from the executive branch to Congress. They are arguably the only part of Congress for which hearings still consistently serve that role.

While the intelligence oversight process is decidedly intermediary-based and non-populist, it has not proven itself remote from “the people” or incapable of responding to public concerns. To the contrary, it has proven over the years to be highly responsive to public sensibilities. In other words, if the populist anxiety is that relying on intermediaries makes government distant and unrepresentative, the experience of the intelligence oversight process suggests otherwise.

The most recent example of this responsiveness is the system’s response to the Edward Snowden revelations about NSA activity in 2013. The congressional oversight process was critical in helping the community weather the storm, as when the bipartisan intelligence committee leadership spoke up publicly in defense of the legality and propriety of NSA’s programs. But the system also adapted relatively swiftly in response to the revelations and generated serious change. Congress passed the USA Freedom Act, substantially altering one of the key programs that Snowden revealed and generating new transparency with respect to certain categories of NSA and FISA Court activity. It is currently considering reauthorization of a much larger program, and further reforms are certainly on the table as it does so.

These adaptions are only the latest in a long string of legislative updates to FISA and other authorities—major changes to which took place at least in 1994, 2001, 2007, 2008, and 2012. This flurry of loving attention to the statutory scheme over time—legislative attention to make sure the statutory architecture of intelligence remains up-to-date as technology and public attitudes change—stands in sharp contrast to the more general environment of legislative dysfunction. In what should be a sharp challenge to reformers who believe that increasing public participation makes the policy process more responsive to public concerns, the portion of Congress that is least populist is exceptionally capable of actually responding to public concerns.

The long-term success of this oversight system is actually hard to overstate. America is a country, after all, whose popular culture produces movies like “Minority Report” and “Enemy of the State,” and whose civil libertarian culture is predicated on a deep suspicion of intelligence operations and government power generally. It is also a country with a history of genuine intelligence community abuses. Yet, even within the context of this culture, the oversight system allows the intelligence community the political and legal latitude to wield extraordinary powers.

And, amazingly, it does so with relatively high confidence from the population in general. The Pew Research Center has for more than a decade been polling Americans on whether they believe the government has “gone too far restricting civil liberties” or whether they believe the government has “not gone far enough to protect the country.” With the exception of a brief blip around the time of the Snowden revelations, many more Americans believe the latter than the former. Although good data on public attitudes toward the intelligence community over time do not exist, this poll question—asked since 2004—suggests that majorities do not believe the intelligence community is out of control. At a time of rock-bottom confidence in public institutions, it is notable that the intelligence oversight system, a system with hardly a trace of populism in its design, actually works effectively at its core purpose: assuring the public that the intelligence community is doing its job within the law.

The system shows that when intermediation is designed and empowered carefully, it can accomplish in a durable and robust fashion many of the objectives that populist reformers purport to seek.