In an important decision, a federal court of appeals held yesterday that the NSA's bulk collection of telephone metadata could not be squared with the legislation that was said to authorize it.
Section 215 of the Patriot Act authorizes the Foreign Intelligence Surveillance Court to issue orders requiring such entities as telephone companies, credit card companies, banks, car rental companies, internet service providers, etc. to turn over their records to the government if the Foreign Intelligence Surveillance Court finds that "there are reasonable grounds to believe" that the records "sought are relevant to an authorized investigation" that is designed to protect the nation "against international terrorism."
In ordinary criminal investigations, the government can obtain similar records by using a subpoena issued by an ordinary court. The goal of section 215 was to enable the nation's intelligence agencies to obtain similar records in the course of foreign intelligence investigations. The reason section 215 was necessary is secrecy. Unlike ordinary criminal investigations, foreign intelligence investigations often involve classified information. Ordinary courts are not authorized to handle classified information. Thus, the Foreign Intelligence Surveillance Court was created because, unlike ordinary courts, it is authorized to handle classified information. Section 215 was enacted after the events of 9/11 to enable the intelligence agencies to obtain orders from the Foreign Intelligence Surveillance Court, enabling them to obtain records that are "relevant" to classified investigations of international terrorism.
In 2006, the Foreign Intelligence Surveillance Court held that, under section 215, the NSA could obtain, from telephone service providers in the United States, massive amounts of telephone metadata that consists of the phone records of millions of Americans. The metadata consists basically of phone numbers, including the phone numbers that are in contact with every phone number collected, but it includes no names and no information about the content of the calls themselves. The NSA uses the metadata to determine when a suspected terrorist, usually outside the United States, is in touch, directly or indirectly, with a suspected terrorist inside the United States.
The question before the court of appeals was whether the collection of this enormous database is "relevant" to an "authorized investigation." The problem is that the word "relevant" is ambiguous. In the usual criminal investigation, in which the government investigates a particular crime or a particular suspect, it will use a subpoena to obtain relatively specific information about the crime or the suspect. In the section 215 telephone metadata program, however, the massive amount of information collected is not, at the time it is collected, "relevant" to any particular suspected terrorist or suspected terrorist plot. Rather, the government collects the information because having it might be useful later when it investigates a particular terrorist suspect or terrorist plot.
The court of appeals concluded that the collection of massive amounts of telephone metadata was not within the meaning of the word "relevant" in section 215. This seems to me a correct conclusion. No court had ever upheld a subpoena that sought information even remotely as broad as that obtained with the authorization of the Foreign Intelligence Surveillance Court under section 215. To construe the statutory language as authorizing such a program seems clearly beyond anything Congress intended or contemplated when it enacted section 215.
The next question the court of appeals had to address, however, was whether, when Congress reauthorized section 215 after the telephone metadata program was in place, it implicitly approved the Foreign Intelligence Surveillance Court's interpretation of the statute. The court of appeals rejected this argument, in part because the telephone metadata program was a secret program. If it was not known to members of Congress when they voted to reauthorize the statute, then they couldn't logically be said to have approved the program.
In fact, though, it is a bit trickier than this, because although most members of Congress did not know of the program when they reauthorized section 215, the members of the House and Senate Intelligence Committees did know of the program and, in recommending reauthorization of section 215, they understood the reauthorization as including reauthorization of the telephone metadata program.
The problem is what to make of this? Can it reasonably be said that, by creating specialized intelligence committees that are authorized to have access to classified programs and information, Congress implicitly delegated to those committees the authority to act on its behalf when it comes to the review, approval and authorization of secret programs?
This poses a broader and even more interesting -- and vexing -- question. How can Congress ever authorize secret programs? Historically, in the realm of foreign intelligence, it was assumed that the President had carte blanche authority to deal with such matters. It was relatively easy for the Executive Branch to deal with that responsibility, because the Executive Branch can (at least in theory) deal effectively with national security secrets.
After it came to light in the mid-1970s that the Executive Branch had abused its authority to engage in foreign intelligence surveillance by using its authority, among other things, to spy on political opponents in the United States, Congress enacted the Foreign Intelligence Surveillance Act in 1978 with the goal of constraining Executive Branch authority. The Act prohibited certain types of foreign intelligence activities, authorized others, and created the Foreign Intelligence Surveillance Court to oversee the actions of the Executive Branch in implementing its authority inside the United States.
But that left open the question, which is still unresolved, of how Congress itself can authorize programs that must be kept secret from our enemies and, therefore, from the American people. One answer, of course, is that the government should never have any foreign intelligence programs that have not been discussed openly and approved by the American people. The problem with that answer is that that "solution" would automatically render eliminate all sorts of surveillance programs that require secrecy in order to be effective.
Another answer is to figure out some way that Congress can have a meaningful role in deciding whether to approve secret foreign intelligence programs without at the same time destroying their effectiveness. That, of course, was the reason for creating the House and Senate Intelligence Committees. But, if their approval is not sufficient to reflect congressional approval, then another way must be found to enable the Congress itself to discuss, debate, review and evaluate secret foreign intelligence programs without destroying them in the process.
To-date, to the best of my knowledge, no one has figured out how to do that, because Congress as an institution has never been thought capable of dealing with classified information. This is so, not only because many members of Congress don't have top secret clearances, but also because most of their staff members also don't have such clearances. And, in any event, the notion of Congress meeting as a whole in secret session in order to debate and vote upon secret foreign intelligence surveillance programs seems rather strained, especially in the current Washington climate.
So, the decision of the federal court of appeals raises a great and important dilemma. If secret foreign intelligence programs must be approved by Congress in order for them to be lawful, how can we make that work? This is not a criticism of the court of appeals' decision. It is, rather, a challenge for the future.