On Wednesday, I summarized the findings of my recent study of alleged secret law in the three branches of the US government and my conclusion that secret law is a limited but important exception to our constitutional norm of publishing the law.  Today’s post considers the way ahead.  Here, I identify relevant constitutional values, outline arguments and options, and recommend rules of the road if the republic decides that secret law is something it can live with but wants to govern more consciously, explicitly, and consistently.

Normative Considerations: Values, Arguments, and Options

As I explained Wednesday, secret law is a problematic and under-appreciated phenomenon, warranting a thoughtful national conversation.  What normative considerations and options can we discern?

Five main constitutional values are relevant.  In no particular order, we begin first with the rule of law.  This includes consistency of law, and the proper functioning of constitutional separation of powers and checks and balances.  The Constitution’s multi-branch, multi-stage lawmaking process has the effect of improving or rejecting questionable legal authorities (statutory, regulatory, interpretive).  A second constitutional value is political self-government by the people, in the senses of law and policy choice, and public official choice (election and removal).  Third is personal self-government, in the sense of ability to adjust one’s conduct based on knowledge of the law.  A fourth value is protection of classified factual national security information.  Fifth and finally is protection of pre-decisional deliberative space for confidential and candid discussions.  Generally, but not always, the first three constitutional values cut against secret law, while the latter two augur for space for it.  Ultimately, the first three mean that the people have the right to determine the content of the law, and have the law they can see truly be the supreme law of the land.

Because the conversation about secret law has been so ad hoc, start-stop, and focused on individual examples such as unpublished Justice Department Office of Legal Counsel (OLC) memoranda, a full case for secret law generally as it exists in the United States today has not been articulated.  We can construct three main lines of argument. 

One is that we need secret law to manage secret fact. Secret law allows the rule of law to operate in classification’s deepest shadows, in the world’s darkest alleys, and at the cutting edge of technologies for clandestine surveillance of the Deep Web.  Would we be better off with all law being public, but that law being less meaningful in the field because it is not tailored to secret operational details or has fallen behind technology?  A second contention is closely related: secret law and secret fact cannot be disaggregated.  The law is most meaningful when it closely engages with the facts.  In secret legal authorities, classified facts and the law are too tightly woven together to allow redaction of fact and publication of the law.  Declassifying secret law will inevitably tip off the bad guys about our capabilities, the argument goes – one made by opponents of the USA Freedom Act’s Foreign Intelligence Surveillance Court (FISC) publication requirements.  Third, secret law defenders can argue that the oversight mechanism is good enough – not pretty, but sufficient.  The idea here is that through existing inter-branch reporting, voluntary and required declassifications, whistleblowers, inspector general investigations, the press, and leaks, the people get a sufficient sense of what secret law exists.  Enough notice, anyway, to vote, to ask Congress to make corrections such as the USA Freedom Act, or to make personal conduct decisions about internet use.

The secret law abolitionist perspective has been more fully articulated.  Its primary claim, which takes a few forms, is that secret law is categorically unacceptable in our republic.  Some claim that unpublished law is not law at all.  Other abolitionists argue that secret law inherently undermines popular sovereignty, and inevitably compromises the Constitution’s publicity-dependent three-branch lawmaking process.  It so disrupts our constitutional system that weak or otherwise publicly indefensible secret legal authorities go unchecked.  (Here, secret law critics point to weak legal reasoning in OLC’s interrogation memoranda, and in the Executive / FISC aggressive interpretation of Sec. 215 of the USA Patriot Act on bulk telephony metadata collection).  A second line of abolitionist argument is that secret law and secret fact can be separated.  Here, secret law critics can point to the Second Circuit’s forced publication of redacted OLC memoranda on targeted killings in 2014.  The FISC is also now writing its opinions so that classified fact can be withheld more easily and the legal analysis published.  Third, abolitionists can maintain that the oversight mechanism regarding secret law is byzantine and insufficient.  It resembles a Rube Goldberg contraption of steel, cardboard, microchips, and duct tape, one that is simply not up to the job.  Spot fixes such as the USA Freedom Act still leave much that the public and Congress do not know.  The risk of Executive Branch legal deep secrecy persists.  And relying for oversight in part on leaks – rule breaking by public servants entrusted with the nation’s secrets – is perverse and corrosive to the legitimacy, fairness, and integrity of the entire classified information protection regime.

The defense of secret law maps onto a broad normative option: “Live with the Status Quo.”  The abolitionist case against secret law augurs toward a second path, “End It.”  A third option is “Reform It.” This school of thought says that we must accept some amount of secret law, but need to govern it better.

Rules of the Road

The current regime regulating secret law – the rules about the (secret) rules – is ad hoc and inconsistent.  It contains some structural elements created for other purposes (such as statutory references to Congress’s classified legislative addenda), and targeted features improvised by busy public servants in the three branches in response to criticism of secret law directed at one issue at a time (such as the FISC opinion publication requirements in the USA Freedom Act).  The status quo includes the public information gaps I discussed in my last post, and reflects a lack of systemic thinking.

For these reasons, if the republic decides it will tolerate some secret law to regulate secret activities, I argue that we ought to govern secret law more consciously, explicitly, and consistently.  I recommend the following principles and proposals as rules of the road.

First is the Public Law Supremacy Rule.  The people must be able to make decisions about the content of the law, and about their personal conduct in view of the law, and have the law that they can read truly be the controlling law.  Any secret law must operate within its four corners.  I explain in my article that this cardinal rule of living with secret law can be operationalized by secret law’s authors as a super-strong avoidance or deference principle favoring the public’s understanding of the law the public can see.  This is quite different from Chevron-like deference to government expertise and knowledge.  A limitation corollary is logical: secret law must not grow government power, but may limit it.  This allows secret law to regulate classified activities in detail.

A second rule is an Anti-Kafka Principle.  Secret criminal law is categorically unacceptable, for reasons of lack of public notice.  A corollary is that secret law may not otherwise impose duties on the public.  To take one example, National Security Letters with gag orders authorized in Public Law are controversial enough.  To ensure that the people can always change law that impacts them, it must be unacceptable for a citizen to be barred from speaking on the basis of legal authority the government keeps hidden.

Third, I recommend a Shallow Secrecy Principle.  The creator of a deeply secret law –  the law’s existence  and content are both unknown – becomes a law unto itself.  Weak or overbroad legal work goes unseen and unchecked.  In contrast, keeping any secret law a shallow secret to the public – its existence is known, but not its content – informs the people when the lawmaking authority they delegate to the government is being exercised in secret.  Shallow legal secrecy allows the people ask public officials to go behind the curtain and have a look, and then be held accountable at the ballot box.

“Bell ringers” that inform the public of a secret law’s creation without imperiling classified factual information already operate in two branches: Congress’s Public Law references to classified addenda, and the FISC publication requirements in the USA Freedom Act.  The Executive Branch is an outlier.  Accordingly, I recommend that every time the Article II branch creates a new unpublished legal authority it should publish in the Federal Register a notice such as this:


Issuing Entity: Office of Legal Counsel, Department of Justice

Index Number: 2017 DOJ 0001

Subject: Surveillance

Entities Concerned: Federal Bureau of Investigation, National Security Agency

Date of this Notice: March 5, 2017

Effective Date: March 15, 2017

Expiration (sunset) Date: March 15, 2021

Declassification / Publication Date: March 15, 2021

Without compromising classified facts or deliberation space, such a notice would allow the people (and other branches) to track and have basic notice of the creation, frequency, creator, duration-in-force, and general subject matter of Executive Branch secret law.  If the people did not want secret law on a particular subject, they could ask their elected leaders to prohibit it with (always supreme) public law.

Meanwhile, because of the general constitutional norm against secret law and the declining half-life of secrets, public officials should feel bound by an expectation of explanation: a normative obligation to explain the legal basis of secret activities to the greatest extent consistent with protecting sensitive facts.  The Intelligence Community has shown unprecedented transparency post-Snowden, but more can be done government-wide.

Fourth, I recommend an Anti-Inertia / Current Public Official Accountability Principle.  If today’s public officials want to perpetuate a secret law they inherit, they should be obligated to review and take ownership of it.  Presumptive sunset and declassification dates are good forcing mechanisms.  A four-year clock, for instance, would align with a presidential administration’s duration.  Secret law expiration dates ought to be established publicly in bell ringer notices, be they statutes referencing classified legislative addenda, Federal Register notices of a new presidential directive, or FISC public docket notices and declassified opinions. Ultimately, today’s public officials should be obligated to review and formally take responsibility for the secret law in their branch that is in force and unpublished on their watch.  They should not have the opportunity to rely a classified legislative annex, secret presidential directive, or unpublished OLC opinion created years ago by their predecessors, without formal re-approval of the legal authority and publication of a bell ringer notice.

Finally, a Plurality of Review Principle is advisable, as a check on deeply secret law and to ensure that secret lawmaking is fully informed of the legal landscape.  Each branch needs to maintain multiple internal stages of review.  Congress needs to see all secret law, so that when the nation’s legislature decides to writes law it does so with awareness of all law.  Additionally, the work of the branches would be facilitated by establishment of a “secret law legal corps”: a non-partisan career cadre of lawyers in all three branches presumptively cleared as legal “super users” able to see all law as they interpret and draft secret law.

These suggestions individually and collectively are certain to raise questions.  That is good.  Secret law is an important phenomenon that merits greater research by scholars, thoughtful discussion by the public, and more consistent regulation in law.  Because this matter is so under-studied and complex, and because so much law remains unpublished, additional analysis will be especially valuable regarding the various varieties of secret law and their implications.  More focus is warranted, too, on questions such as when a non-public Executive Branch opinion, directive, or other document becomes sufficiently law-like to trigger consideration of the normative paths and application of the rules of the road I recommend.  I also hope that candidates and public officials, including the next administration, will be asked to articulate their approach to this three-branch reality of American governance.  It is time to come to terms with secret law.