A photograph of NSA leaker Edward Snowden on a screen in Hong Kong

A photograph of NSA leaker Edward Snowden on a screen in Hong Kong Vincent Yu/AP

Obama's Intel Workers Need New Policies for Secrets, not Snitches

In the Snowden fallout, the administration should focus on developing a happier intelligence workforce, not outing insider threats. By Marc Ambinder

The good news, I guess, for the government is that the secrecy system in the United States is not failing, Edward Snowden’s adventurism aside. The ship of state is and has always been leaky by design. It’s a kluge of policy choices, historical contingencies, technological change and shifting political and civic beliefs about right, wrong and the security/liberty continuum. 

But the dome separating the secret world from the luminal world is remarkably thin, which makes it all the more amazing that it hasn’t simply been vaporized. A system that must accommodate single point failures -- one person who has access to some secrets is able to frustrate the designs of tens of thousands with access to even more secrets -- has to be more self-aware, dynamic and experimental. That’s why it’s frustrating when policy-makers continue create their own problems.

  • Insider threat policies inevitably produce more suspicion and friction, which inevitably produces more secret-keepers with axes to grind, both real and imaginary. (Government says to work force: we don’t trust you because 99.9 percent of you don’t leak. Does that make sense to you? If it does, you’ve got a job lined up at Liberty Crossing, the office campus home of the National Counterterrorism Center in Tyson’s Corner, Virginia, a tony Washington suburb.)
  • The incentives and affirmations that ordinary Americans expect in the course of their daily professional lives are unavailable to those who work with secrets. The government does next to nothing to provide alternative incentives or affirmations that would keep a workforce productive, with the exception of money, which is increasingly in short supply and almost never directed towards the general betterment of the lay (i.e non-contract) workforce.
  • Contractors are cheaper than government employees and their jobs are elastic enough to easily slide by congressional restrictions on what constitutes essential or core governmental functions. A contractor with many projects has better access to secrets than most policy-makers who require the information.
  • Policies are set by old, rich, white men with farms in Virginia and are supposed to account for the worldviews, expectations and experiences of a workforce that is young, multiracial and (some exceptions to the contrary) not rich.

Here’s the one that bothers me the most:

  • No one in the government spends any time thinking proactively about what needs to be secret and what doesn’t. The key word is proactively. There is a significant amount of thinking applied to questions of what should be retroactively declassified and how it should be put in context. Classification policy is reactive.  To the extent that there is any big picture strategy, it is always biased towards the presence of a series of cryptic letters and the special meaning they impart to a set of facts. It privileges and advances the idea that stuff has to be secret because it has to be secret. This is stupid. It creates national security challenges and makes the government all the more vulnerable to the most severe consequences from single point failures.

To wit: while there may be some benefit in assuming that bad guys and spies won’t try to communicate clandestinely, there is an enormous potential downside in keeping secret from the American people the fact that the National Security Agency stores their transactional telephone records. Before you respond that there is no way I could possibly make this assertion without knowing how terrorists react to public information, and of course, that’s something I wouldn’t know because I don’t have the clearances, here is what the opposite assertion has cost the U.S. government:

  1. The sudden and abrupt change in communication practices after the secret has been released at a time not of the government’s choosing and without any strategic thinking about how to mitigate its effects.
  2. A significant drop in public support for counter-terrorism policies that, had the administration and the previous administration not been so defensive and paranoid about, might well be so much of a non-issue today that an Edward Snowden would have no audience to which to reveal them.
  3. The mind-space to creatively engage stakeholders who might have helped them design a policy that better protected civil liberties, or at least was able to demonstrate that civil liberties are indeed a principal feature of the design of such intelligence collection systems. Now, no one trusts anyone.
  4. Credibility.
  5. A drop in the public’s support for the millions of people who do this work, which, correspondingly, probably contributes to a decline in their overall work happiness, which, one would reasonably assume, reduces their efficiency.

An entrepreneurial administration could address these contradictions if they wanted to without upsetting too many apple carts. The insider threat policy could and probably should be accompanied by a “happy intelligence community workforce” policy that address the challenges of working in an environment where people are trained to lie, steal and evade. Congress could simply appropriate more money for career development and workforce development. The administration could ask younger analysts what they need to do their jobs better, rather than threaten them with harsh penalties if they fail to report a colleague’s amorphously suspicious behavior. 

Figuring out how to talk about secrets publicly is much harder, but it’s the most important fortification there is for a system has any claim to legitimacy or effectiveness.