A former US gov't lawyer and current Harvard Law professor makes the case.

Former NSA contractor Edward Snowden has asked President Barack Obama for a pardon, and the ACLU, which represents Snowden in the US, agrees. The following piece is a response to Snowden's argument. The author, Jack Goldsmith, is a Harvard Law professor and a senior fellow at the Hoover Institution. This piece first appeared at Lawfare.

A “pardon Snowden” campaign was launched Wednesday in conjunction with the Snowden film. Snowden himself made the “moral case” for why he should be pardoned, and Tim Edgar made a much more powerful case. I remain unconvinced. I don’t think the president will, or should, pardon Snowden.

I say this even though I agree with Tim about many of the upsides to Snowden’s theft and leak of documents from NSA databases. On the third anniversary of the Snowden disclosures, I wrote about how, despite their many costs, the disclosures strengthened the intelligence community. They forced the NSA to be more transparent and to better explain itself, demonstrated that the NSA was acting with the full knowledge and support of three branches, resulted in its authorities being strengthened and its collection practices barely narrowed (and in some respects expanded), and overall enhanced its domestic legitimacy going forward. I was not kidding when I said that “[t]hese are but some of the public services for which the US government has Snowden to thank.” This was not a new theme with me. I have made similar points for years. (See here and here and here and here.)

But to say that the intelligence community benefited from the Snowden leaks is not to say that the president should pardon Snowden, for the price of the benefits was enormously high in terms of lost intelligence and lost investments in intelligence mechanisms and operations, among other things. Many Snowden supporters pretend that these costs are zero because the government, understandably, has not documented them. But it is naïve or disingenuous to think that the damage to US intelligence operations was anything but enormous. (To his credit, Tim acknowledges that “Snowden’s actions caused great damage to national security.”) Much remains unknown regarding the extent of the damage (because the intelligence community cannot publicly say much beyond generalities) and the specifics of Snowden's actions and motivations (because DOJ is preserving a criminal prosecution). I imagine we would learn considerably more information—from both sides, but especially from the government—if a criminal trial ever took place. And indeed it is hard for the public to even assess the case for a pardon until we know the full extent of Snowden’s crimes and the harms they caused. But I have no doubt that the harms from his actions were very significant.

Another difficulty in determining whether a pardon is warranted for Snowden’s crimes is that the proper criteria for a pardon are elusive. Oliver Wendell Holmes once declared that a pardon “is the determination of the ultimate authority that the public welfare will be better served by inflicting less” than what the criminal law specified. But how to measure or assess the elusive public welfare? The Constitution delegates that task exclusively to the president who can use whatever criteria he chooses. Many disagreements about whether a pardon is appropriate are, at the bottom, disagreements about what these criteria should be. Some will question whether Snowden should be pardoned even if his harms were trivial and the benefits he achieved were great. Indeed, presidents don’t usually grant pardons because a crime brought benefits. My own view is that in this unusual context, it is best to examine the appropriateness of a pardon in the first instance through an instrumental lens, and also to ask how well Snowden’s stated justification for his crimes matches up with the crimes he actually committed.
A good place to begin is with my former colleague Geoffrey Stone’s analysis:

I think if [Snowden] had only disclosed the existence of the second 215 metadata program, then one might be able to make the case he did more good than harm because there were reforms adopted because of his disclosures. That's a good thing. And the program itself had not been up to this point all that valuable, and therefore even though its disclosure makes it largely ineffective going forward–they work in part because the person you're surveilling doesn't know they exist–the cost was pretty modest because the program wasn't that valuable... The problem is he disclosed vastly more than that, involving foreign intelligence not of Americans but of individuals who aren't American citizens in other countries. No changes were generally made in those programs, and Americans don't really care. But disclosing those programs has had a serious impact on their being as effective as they had been. I think he did a lot more harm than good.

Let me say a few more words about this. Snowden has long claimed that he took an oath to “support and defend the Constitution” (see here and here), and he has implied that he was fulfilling this oath when he stole and distributed the documents. “The oath of allegiance is not an oath of secrecy,” he told Bart Gellman. “That is an oath to the Constitution. That is the oath that I kept that Keith Alexander and James Clapper did not.” Let’s assume that Snowden is right that his oath to the Constitution trumps his employment agreement secrecy duties (but see here; cf. here). Snowden might plausibly argue that his exposure of the 215 program was genuine whistleblowing in support of the Constitution. Even though the program was vetted by the three branches of government, its revelation sparked legal controversy and subsequent substantial reforms. His exposure of the 702 programs (PRISM and upstream collection) is harder to justify on these grounds, because these programs were clearly authorized by public law and have not sparked nearly the same criticism, pushback, or reform. For this reason Stone implies that the 702 leaks were not justified. I tend to agree, but for purposes of argument I will assume (but only assume) that this and every example of Snowden leaks that involved unknown collection inside the United States or of US persons were examples of admirable whistleblowing.

What I do not get, and what I have never seen Snowden or anyone explain, is how his oath to the US Constitution justified the theft and disclosure of the vast number of documents that had nothing to do with operations inside the United States or US persons. (Every one of the arguments I read for Snowden’s pardon yesterday focused on his domestic US revelations and ignored or downplayed the vast majority of revelations that did not involve US territory or citizens.) To take just a few of hundreds of examples, why did his oath to the Constitution justify disclosure that NSA had developed MonsterMind, a program to respond to cyberattacks automatically; or that it had set up data centers in China to insert malware into Chinese computers and had penetrated Huawei in China; or that it was spying (with details about how) in many other foreign nations, on Bin Laden associate Hassam Ghul’s wife, on the UN Secretary General, or on the Islamic State; or that it cooperates with intelligence services in Sweden and Norway to spy on Russia; and so on, and so on. These and many other disclosures (see here for many more) concern standard intelligence operations in support of national security or foreign policy missions that do not violate the US Constitution or laws and that did extraordinary harm to those missions.
There is such a disconnect between Snowden’s constitutional oath and the scale and type of sensitive intelligence information that he disclosed that it makes me question whether the Constitution had anything to do with the leaks. And indeed, when Snowden “set out the case” for his pardon on Wednesday, he did not (at least in the stories I read) argue from the Constitution. Rather, he now says of the disclosures: “when we look at them morally, when we look at them ethically, when we look at the results, it seems these were necessary things, these were vital things.” Let us set aside the “results” of having blown numerous intelligence operations against adversaries and many other legitimate intelligence targets. What might be the moral and ethical case for disclosing US intelligence techniques against other countries and institutions? (I will ignore possible cosmopolitan impulses for Snowden’s theft and leaks, which I think damage the case for a pardon for violations of US law.) I think the most charitable moral/ethical case for leaking details of electronic intelligence operations abroad, including against our adversaries, is that these operations were harming the Internet, were hypocritical, were contrary to American values, and the like, and Snowden’s disclosures were designed to save the Internet and restore American values. This is not a crazy view; I know many smart and admirable people who hold it, and I believe it is ethically and morally coherent.

But it is also not a crazy view, and it is also ethically and morally coherent to think that US electronic intelligence operations abroad were entirely lawful and legitimate efforts to serve US interests in a complex and dangerous world and that Snowden’s revelations violated his secrecy pledges and US criminal law and did enormous harm to important American interests and values. Unfortunately for Snowden’s pardon gambit, President Obama, and anyone who sits in the Oval Office charged with responsibility for American success around the globe, will (and should) embrace the second moral/ethical perspective and will not (and should not) countenance the first moral/ethical perspective, which I take to be Snowden’s. Snowden can act on whatever conception of American values he likes, but when he acts in massive violation of criminal laws in ways that reveal lawful intelligence activities against adversaries and other legitimate intelligence targets, he cannot expect a pardon.

Another reason why Snowden won’t and shouldn’t be pardoned for his actions is that doing so would have a demoralizing effect on the thousands of intelligence community personnel who devote (and in some cases risk) their lives to US national security, who follow the rules laid down by Congress and the president, whose work was diminished, and whose jobs were made much harder as a result of Snowden’s non-US related disclosures. I disagree with Tim that “a pardon sets no precedent and so creates no incentives.”
Pardoning the perpetrator of the most damaging leak by far in American history would send a clear signal of approval for what Snowden did and a clear signal about a lack of seriousness on the part of the government about its truly most important secrets. Those signals would affect the attitude of everyone in the intelligence community about the value of our most important secrets and would have a terrible impact on the government’s already-difficult ability to keep such secrets. In saying this, I do not detract from the importance of the greater transparency that Snowden brought to the intelligence community. That community was self-defeatingly secretive and insular and terrible at explaining what it was doing and why. But to say that it needed to open up a great deal, especially about the extent of and legal bases of its domestic operations, is not to say the government should countenance disclosure of details about its lawful electronic intelligence operations abroad against non-US citizens, which is what the pardon Snowden seeks would do.

In sum, I can imagine a pardon for Snowden for the smidgen of his revelations about possibly unlawful domestic surveillance or collection against US persons. But the possible case for a pardon extends only to that smidgen of the leaks. It is hard to see how the president could pardon the manifold violations of US criminal law of the broader leaks about obviously lawful operations, given the costs to the United States in terms of lost intelligence, compromised techniques, and destroyed investments, among others.