Security Requires Post-Mortems of Immigration Mistakes

By Dan Cadman on November 6, 2014

In October of this year, the Department of Justice (DOJ) Office of Inspector General (OIG) quietly released a declassified version of a report initially issued in March 2007, on the FBI's use of national security letters. This statement caught my attention:

We found that the use of NSL requests related to "U.S. persons" and "non-U.S. persons" shifted during our 3-year review period. The percentage of requests generated from investigations of U.S. persons increased from about 39 percent of all NSL requests issued in 2003 to about 53 percent of all NSL requests during 2005. (U)

National security letters (NSLs) are the rough equivalent of a subpoena, only they are administrative — not judicial — in nature, and are signed and issued by designated officials of the FBI. As their name implies, they are used in national security cases such as foreign counterintelligence, espionage, and international terrorism investigations. They are powerful tools in those investigations and, when served on telecommunications companies, banks, Internet providers and the like, they provide a wealth of access to personal information such as phone records, emails, voice messages, bank accounts, you name it, all the things that you must have to live in our modern world, and that uniquely define who you are.

National security letters result in government collection of information of a vastly different, and much more personal, nature than the metadata that caused such a furor when Edward Snowden leaked its collection by the National Security Agency.

Even though the OIG trod lightly, the auditors who compiled the report have documented a number of disturbing misuses of the tool, as well as significant undercounts of their use, due to database inaccuracies. I suspect that if the subject of the report were any entity but the very powerful FBI, the report would have lambasted that agency and congressional hearings would have ensued. It is also interesting that it took 7.5 years to declassify and publicly release the report.

A word of explanation is in order: Under the law a "U.S. person" (or "USPER" in the lingo of the trade) is not confined solely to United States citizens. The law defines USPERS as "any United States citizen or alien admitted for permanent residence in the United States, [or] any corporation, partnership, or other organization organized under the laws of the United States".

What this tells us is that over half of the subjects of federal national security investigations as of 2007 involved citizens, whether native-born or naturalized, and resident aliens. That is a sobering thought. One wonders how high the percentage has risen in the years since 2007.

What has this to do with immigration? Everything. Stop and consider: Right from the start, we know that if 53 percent of NSLs (and therefore, the national security investigations of which they are the subjects) involve USPERS, then nonimmigrant aliens, and illegal aliens such as visa overstayers or out-of-status students, were the subject of the other 47 percent — close to half of the subjects of NSLs .

Now let us assume that resident aliens constitute half of the USPERS who become the subject of NSLs (26.5 percent). That may be a low figure, but if we use it, then 73.5 percent of all NSLs (and therefore the subjects of national security cases, the vast majority of which involve counterterrorism) involved aliens.

Finally: while I can't prove it, because the data aren't presented anywhere, I strongly suspect that among those U.S. citizens who became the subject of NSL letters, the vast majority were naturalized, not native-born.

And that's the crux of the matter—because if true, what it tells us is that when you combine "non-U.S. persons" (all aliens who aren't resident aliens) with those "U.S. persons" who are either resident aliens or naturalized citizens, then collectively nearly all subjects of national security letters issued by the FBI have their roots in foreign lands. And yet the federal government doesn't track that data or use it to make informed decisions in its immigration benefits and intelligence and law enforcement programs. Nearly a decade and a half after the 9/11 attacks, data collection and use is still stovepiped among and between agencies, and (to use the language of the 9/11 Commission) the dots are not yet connected.

Classic process improvement theory — and, indeed, classic operational intelligence theory as well — tells us that there must be a continuing feedback loop between outcomes, whether successful or failed, and the processes that led to those successes or failures. In this way, the processes can be amended to reinforce success and minimize failure. To be specific, I am talking about the way the federal government goes about its immigration benefits adjudications, something that all too few government officials acknowledge is fundamental to the security of the United States. Success should be defined as granting benefits to aliens who add value to American society; failure should, at minimum, be defined as those cases where aliens are granted a benefit (including entry), only to engage in activities contrary to the public safety and national security.

Time and again the Department of Homeland Security has told us that because it cannot possibly deal with every risk, no matter how small, it engages in a strategy of risk management to keep the homeland safe. And time and again Center for Immigration Studies publications have questioned the commitment to, and efficacy of, that strategy for lack of any apparent feedback loops, such as post-mortems of cases where aliens or naturalized citizens have gamed the system. (See, for instance, here, here, and here.)

Isn't it a material weakness in our national and homeland security apparatus, if we cannot track how many individuals who have received our most prized immigration benefits — lawful permanent residence and naturalization — and later proved to be unworthy of their receipt and harm Americans?

The evidence and indicia strongly suggest that there is no holistic effort on the part of the federal government to conduct post mortem examinations of national security cases involving naturalized citizens or lawful residents (or indeed, for that matter, any aliens) in order to learn from past benefits-granting mistakes, and incorporate the lessons learned into new processes and guidance to immigration examiners and inspectors at ports of entry. (See here and here.)

Until the government does so, there cannot possibly be any credibility to the notion of risk management in homeland security efforts, and the people in communities throughout America remain at risk.