« Assumptions in a Hays… | Home | More Updates on Lone … »

Why They Leak (Or Better, Why They Don't)

(To read the rest of "On Violence’s Most Thought Provoking Foreign Affairs Event of 2013", please click here.)

One of the more remarkable facts, I think, of the whole Snowden affair is how long it took for these disclosures to get leaked. Well before the Snowden leaks, Senators speaking on the record and some NSA officials speaking well off the record, said some variation of the line, “Americans would be shocked by how much the NSA snoops on them.”

Sure enough, when the Snowden leaks came out, a majority of Americans were shocked by an intelligence leak.  

So if we can take the large--but not universal--shock of most Americans as evidence they felt the NSA overstepped its authority (and its leaders probably lied to Congress about their spying), why did it take so damn long to come out? Most bad deeds get leaked eventually. But why did it take over 11 years? Why did only one person decide to leak all these documents? Well, the answer is simple:


To show this, I need a thought experiment. To start, assume that every employee at the NSA is motivated by one of two things: self-interest and altruism. For self-interest, I mean everything which causes Americans to go to work: pay, keeping their job, providing for their family, advancing up the career chart, and gaining responsibility/power/respect. (This could be called the "classical" economics framework.) For altruism, I mean all those pesky things which might cause someone to forego personal gain: belief in the greater good, religious beliefs, ethics, morality, emotions and patriotism. (This could be called the hard part of economics--the things which screw up economist's traditional models. For more on a related topic, click here.)

In real life, the employees of the NSA are motivated by mixtures of both self-interest and altruism. But which predominates? Supporters of the government--like now-On V-punching-bag David Brooks (whose writing we still absolutely love)--insist that the vast majority of government employees are "good people" who "try to do the right thing". Unfortunately, my thought experiment shows that self-interest usually trumps altruism. 

In the first ideal version of the NSA, every employee is only motivated by altruism. How would this NSA look? Well, its officials would never lie to the American public. There wouldn't be a need for whistleblowers, because superiors would respect subordinates who went to the Inspector General to report abuse. And if, for whatever reason, wrong deeds still needed exposure, NSA employees would go to Congress or the press on a regular basis. But employees wouldn't care about advancement, only helping the NSA protect America. Employees would speak their minds because they care about the greater good.

What's the alternate? In a perfectly self-interested NSA, it would operate much differently. Since most promotions are controlled by superiors, employees would think first and foremost about upsetting these power brokers. They probably dress up their motivations in altruistic terms--"loyalty", "team player"--but they don't ever make their bosses look bad. Thus, when they come across wrongdoing, they don't do anything. At best, self-interested employees tell themselves that, "When I am in charge, I'll fix all the problems."

And leaks? Virtually non-existent from the lower levels. Would-be whistleblowers know that the surest way to end a career is to expose wrongdoing via leaking classified information. Though leaks occur all the time, they only come from senior officials to make the intelligence community look good. A leak from a subordinate which makes the whole intelligence community look bad will ruin a career absolutely.

In the self-interested NSA, anyone who leaks goes to prison to send a message for future leakers. (If you still want to leak, you would have to flee to the most ironic country possible, Russia.) In its darkest iteration, the self-interested NSA even bribes congressmen with donations to ensure future funding. In an even darker version, the NSA could blackmail elected leaders to insure they continue funding its operations.

So, let me ask, which version of the NSA seems more realistic? Which one do we seem closer to?

Considering the vast lack of unauthorized leaks, I would say the latter. No NSA employee chose to speak out that the NSA had twisted sections of the Patriot Act to expand domestic surveillance. Until Edward Snowden, not a single employee who heard General Clapper lie to Congress went public. Skeptics of the NSA, like myself, would argue that the incentive structure inside the NSA so vastly outweighs the altruistic motivations that only the exceptionally rare individual would blow the whistle on wrongdoing.  My thought experiment from above shows this.

While most American intelligence officials and employees are indeed good people, they're still self-interested. As Manager-Tools frequently points out, the employees at the NSA are all addicted to food, clothing and shelter. Unless Congress passes strong legislation which protects whistleblowers, and maybe even encourages it, we can expect waste, fraud, abuse, illegality, corruption and other ills of government.

Edward Snowden is a very rare individual for doing what he did almost solely based on altruism. We should understand that, and instead of condemning him, we need to find ways to get more Edward Snowdens to do what he did...legally. We need to shift the incentive structure so that the natural altruism of NSA employees isn’t bowled over by the need to continue paying for food and shelter. In other words, Edward Snowden and the vast lack of leaks show the incredible need to change the incentive structure inside the intelligence community.

There is, as Eric C pointed out reading this post, another darker explanation. Many if not most of the workers--and especially the leaders--could be altruistically motivated. However, the values they adore aren’t values like civil liberties and respect for the Constitution. If their fundamental value is security, they could act in much the same way. In this scenario, the altruism of the NSA means it does whatever it takes to keep Americans safe despite the harms to freedom. In short, security trumps liberty.

Now that is a scary thought.

six comments

Very good.

But I don’t see how you could change the incentive structure of the organization itself. Mature bureaucracies won’t do anything to weaken themselves and anything that they think makes them look bad they will perceive as weakening them. So they won’t do it. They will pretend to do it but they won’t actually. I don’t see how the organization would do it its ownself.

Security does trump liberty in those organizations I think. Their mandate is security, not liberty. They will work hard to do their jobs as best they can. It is scary.

Michael C:

There is one other component to the scary thought. There are a substantial number of people out there who are only waiting for an approving word from an authority figure, any kind of authority figure really, to do whatever it is the evil fertility of the human mind can come up with. I dare a little but maybe only a little in saying that you probably saw that in some men in your experience. That would be even more true if the authority figures were from a great big official agency like the NSA. Watching all the Americans is nothing in that case.

Maybe there’s a different assumption in play (where did I last read about the danger of assumptions?!)

Maybe decision makers in some of these organisations are noting how eager the general public is to post personal information online through various social media forums. They’ve then assumed that this means the perception of privacy has changed, and if that’s the case then who can get upset about a little metadata in the name of security?

Further (to approach your darker theory from a different angle), maybe decision makers have noted that war has changed. The general public doesn’t have a stake in physical war. The public is told to go to the mall, after all. But if there’s going to be a war on nouns and a large chunk of that war is going to take place in cyber domains then there ought to be a little sacrifice all around, and metadata is a small sacrifice. Once this theory is incorporated into institutional culture then it becomes more than altruism. It becomes patriotism.

The following interview may be of interest. The founder of Lavabit, Snowden’s e-mail provider, is interviewed about his decision not to give the government the encryption codes and close down Lavabit instead.



To supplement your second possibility, that discipline is actually emerging from the underbelly of law enforcement intelligence investigations. The vastness of open-source intelligence (OSINT) is quickly becoming justification for further covert information/data collection.

You really hit “danger close”. For some individuals out there, the internet and its inherent onslaught of available social media platforms are just physiological extensions for human interaction. Quite pathetic, really. Any reasonable person under the scope of NSA directives would have surely considered the aforementioned prospect.

I expect the legal definition and digital applications of reasonable expectation of privacy to dramatically shift in the next few years.

I should probably clarify the second part of my statement above: I don’t actually believe that it’s either right or patriotic to allow erosion of privacy. As 5150 notes, that’s an area where laws really need to catch up with reality. I simply suggest it as a theory for how various government departments and their employees might justify such acts.