Editor’s note: this post is a preview of ideas raised in an upcoming article by the author, Intelligence Legalism and the National Security Agency’s Civil Liberties Gap, set to appear in the Harvard National Security Journal in coming months.

Since the 1970s, the United States Intelligence Community has become heavily influenced by rules that resemble laws and the lawyers who are adept at interpreting and manipulating them. How did we get here?

It’s been 17 months since Glenn Greenwald’s first news story based on Edward Snowden’s trove of stolen documents appeared in the Guardian. Over that time we have learned an enormous amount about some of the post-9/11 signals intelligence programs. The most important disclosures have been about bulk surveillance; we know now that since 2001, the NSA has monitored Americans’ telephone and internet usage, and searched the resulting network database millions of times. Surveillance is not a topic that usually garners much public attention, but over the past year and a half we’ve seen responsive blue-ribbon commissions, public hearings, editorials, and so on—more public engagement than for decades prior.

We’ve seen this movie before: startling disclosures of surveillance of Americans were followed by a moment of public attention during the “year of intelligence.” That was 1975, when the Rockefeller Commission, the Pike Committee, and the Church Committee all held their hearings. They discovered a surveillance state outside of the law; one that invaded American privacy without qualm, and often for the basest of partisan reasons. Reform took the shape of a “great compromise” under which oversight would be significantly strengthened, but would remain secret—in the House and Senate intelligence committees and the Foreign Intelligence Surveillance Court.

Many have written about the pros and cons of this qualified approach to democratic accountability. But the 1970s reforms resulted in something else that is too often overlooked—the turn towards legalism.

Political theorist Judith Shklar defined legalism as “the ethical attitude that holds moral conduct to be a matter of rule following”— nearly regardless of the content of those rules. Among legalism’s crucial features is a preference for “law,” sometimes with court enforcement, other times without, over “policy.” The other key tendency is its empowerment of lawyers.

Three documents are crucial to understanding the origins of legalism within the surveillance state, where it is highly flavored by secrecy. Those texts—the Church Committee reports, the Foreign Intelligence Surveillance Act (FISA), and Executive Order 12333—each proposed or established authoritative rules of conduct, and each empower lawyers. The former point is obvious and needs little elaboration. But the latter point is equally important. Government lawyers accordingly loom very large in the reform documents of the late 1970s and thereafter.

The Church Committee proposed a large number of substantive new rules constraining the Intelligence Community. These included not only the surveillance-related rules eventually incorporated in FISA but also others (eventually incorporated in Executive Order 12333) against assassination of foreign leaders; use of academics for CIA operations without disclosure to their university presidents; non-public sponsorship of books, articles, and other media by the CIA; CIA relationships with journalists affiliated with US news organizations; CIA relationships with American clergy and so on.

The Committee’s concomitant procedural innovation was to give lawyers compliance responsibility for those rules. The idea was that once law-type rules were in place in the Intelligence Community, it would become the natural province of government lawyers to ensure that they were followed. Presumably this was an attractive approach both because interpretation of complex rules matched lawyers’ skills, and because compliance with complex rules matched lawyers’ natural inclinations. After all, Shklar observed that legalism is the central shared commitment of members of the legal profession.

With dozens of specifics, the Church Committee recommended amplifying the authority and influence of lawyers within the executive branch—and particularly the Attorney General. The specific domestic recommendations proposed to obligate the Attorney General to review procedures, authorize operations, conduct investigations, and generally be responsible “for ensuring that intelligence activities do not violate the Constitution or any other provision of law.”

The main statutory answer to the Church Committee’s call was FISA. As originally enacted in 1978, FISA made two key innovations, both highly legalizing. First, it subjected all domestic foreign intelligence surveillance, and some surveillance abroad, to analogues of domestic warrant procedure. Second, it introduced the idea of “minimization procedures”—rules “designed to protect, as far as reasonable, against the acquisition, retention, and dissemination of nonpublic information which is not foreign intelligence information” that “concern[s] unconsenting United States persons.” The core minimization requirement is that surveillance and retention processes be “reasonably designed . . . to minimize the acquisition and retention, and prohibit the dissemination, of nonpublicly available information concerning unconsenting United States persons consistent with the need of the United States to obtain, produce, and disseminate foreign intelligence information.”

The minimization procedures are themselves highly legalistic. First, they read like statutes or regulations. Second, the minimization procedures frequently use the strategy of designating a particular high official to make specified decisions. Implementation then forces subordinate personnel into using the legalistic method of reasoned elaboration, as they explain why the outcome they favor should be adopted by the official authorized to decide. Third, the procedures empower lawyers: they must be approved by the Attorney General, and therefore first by DOJ lawyers, prior to being offered to the FISA Court for its signoff. Fourth, once approved, the procedures acquire the privileged status of federal court orders. Obedience becomes a compliance—rather than a policy—task for the NSA, subject to requirements of court disclosure and correction. So if NSA fails—particularly if it fails systematically—the court might impose various consequences ranging from professional embarrassment for particular lawyers to withdrawing approval for a whole NSA program. (It is evident that these consequences are only loosely coupled with the substantive importance of the disregarded minimization feature; the FISA Court has sometimes scolded the government for noncompliance with minimization orders whose features it agrees to relax in the very same opinion.)

Even as post-9/11 amendments and interpretations of FISA have vastly reduced the warrant-style individuation required for FISA-authorized surveillance, the basic legalizing structure just described has remained intact: lawyers prepare and judges approve the proposed surveillance, and all of this is accompanied by court-ratified minimization procedures given the force of law.

For the wide swathes of foreign intelligence surveillance that are not covered by FISA, the type of regulation recommended by the Church Committee occurs under Executive Order 12333, without judicial involvement. The Executive Order explains that its “general principles . . . in addition to and consistent with applicable laws, are intended to achieve the proper balance between the acquisition of essential information and protection of individual interests.”

For surveillance, the order’s basic approach is two-fold: First, it insists on in-advance fully vetted written procedures, to be approved by the Attorney General. Substantively, these documents function like FISA minimization procedures, albeit laxer. Procedurally, however, they are very different. For FISA minimization, written justifications and explanations of each program are filed with the FISA Court and underpin each eventual court approval. Any change in the programs might be material to the court’s approval, and therefore needs to be explained. For Executive Order 12333 processes, the AG Guidelines are more freestanding; there is no subsequent formal implementation check. Thus even apart from the greater leeway allowed by the AG Guidelines, compared to FISA-approved minimization procedures, the result is substantially more operational freedom under 12333 than under FISA. Second, Executive Order 12333 authorizes specific surveillance without court approval only if the Attorney General approves. While there is no judicial involvement, the process is very similar to a judicial one; the same lawyers who prepare FISA applications prepare a similar application for the Attorney General to approve (or reject).

As a whole, notwithstanding the absence of court involvement, Executive Order 12333 is a key source of intelligence legalism. It is also worth noting that its text was one of the sites around which intelligence legalism was hotly contested. One of the order’s drafters, Richard Willard, recounted a few years later that when he arrived at the Justice Department early in the Reagan administration as then Attorney General William Smith’s counsel for intelligence policy, “holdover [career] officials in the intelligence community were busily drafting a new Executive order on intelligence activities that would virtually eliminate the legal oversight role of the Attorney General.” They were doing so because of the “enormous pent-up hostility in the intelligence community toward lawyers and legalistic restrictions,” according to Willard. This “attitude was not an invention of the Republican political appointees—who at that time were not yet that numerous—but permeated the career service.” It was his assignment, he explained, to mold Executive Order 12333 into something more “balanced” —that is, more pro-lawyer.  He succeeded; the order inserted the Attorney General deep into intelligence policy and even operations.

This intervention marked a sharp change. Willard notes that in his time at the Department,

“[t]he Attorney General was not a full member of the cabinet-level group that considered these [foreign intelligence and policy] matters but was only ‘invited’ to attend. It is my understanding that Attorney General Meese was later made a member of the group, but that even then some effort was made to insist that he was a member in his personal capacity and not as Attorney General. . . . As a consequence of the Attorney General’s uncertain status in the process, his subordinates were generally excluded from working groups and subcabinet-level deliberations.”

In total, while the tendency is more extreme for FISA, each of the two foundational documents for foreign intelligence surveillance, FISA and Executive Order 12333, has moved surveillance programs in legalistic directions, emphasizing rules and empowering lawyers. The resulting compliance and oversight ecosystem is structured accordingly.  In my next post, I’ll explore some of the costs of this structure.