The U.S. government routinely tries to hide its unlawful behavior. It hides evidence of its incompetence too. That's a matter of historical record, not an opinion. Exposing government misbehavior sometimes requires publishing classified documents—take the Pentagon Papers or the Bush Administration's secret wiretaps.
Writing in the New York Times Book Review, Michael Kinsley declares, "The Snowden leaks were important—a legitimate scoop—and we might never have known about the N.S.A.'s lawbreaking if it hadn’t been for them." As he sees it, unauthorized disclosures of classified information typically benefit the public. "Most leaks from large bureaucracies are 'good' leaks," he writes. "No danger to national security, no harm to innocent people, information the public ought to have."
How strange to believe all that while also insisting that the publication of leaks by journalistic organizations should be criminalized—which is what Kinsley does later in the same article.
"It seems clear, at least to me, that the private companies that own newspapers, and their employees, should not have the final say over the release of government secrets and a free pass to make them public with no legal consequences," he writes. "In a democracy ... that decision must ultimately be made by the government. No doubt the government will usually be overprotective of its secrets, and so the process of decision-making—whatever it turns out to be—should openly tilt in favor of publication with minimal delay. But ultimately ... someone gets to decide, and that someone cannot be Glenn Greenwald."
This is a mainstream position: that government "must" have the final say, not some pixel-pusher elected by no one. But I don't think Kinsley can persuasively defend that position, having already stipulated that the Snowden leaks are a legitimate scoop, especially if he's unwilling to call for journalists at The Guardian, the Washington Post, and the New York Times to be charged criminally.
In a defense of the role journalists at those publications play:
The least-bad system isn't one where government punishes journalists at its discretion, or can preemptively stop them from publishing classified information. Neither is the least-bad system one in which government employees can leak anything—nuclear codes, troop movements—without legal consequences.
The least-bad system is one where leakers can be charged and punished for giving classified secrets to journalists (which isn't to say that they always should be), but where journalism based on classified information is not criminalized. Insofar as that standard is anti-democratic, it is no more anti-democratic than the First Amendment to the Constitution in which it is grounded. "Congress shall make no law ... abridging the freedom of speech, or of the press." That is an explicit protection for newspapers to publish in the face of government objections. A "Fourth Estate" was built into our "democracy."
Rather than dwelling in abstractions, I'd ask Kinsley to empirically assess his assumptions about who best decides what information is in the public interest. De facto policy in the U.S. has long been the system I favor and Kinsley attacks. How has that system performed?
Kinsley could tally every leak of classified information that journalists have published against the government's wishes. How many of those stories have advanced the public interest? How many have done actual harm to national security? What are the most important benefits Americans have gained from the publication of classified material and what are the most terrible costs we've suffered?
Bearing in mind that Kinsley regards the Snowden leaks as an important, legitimate scoop, the conclusion that he'd likely to reach at the end of the exercise is clear: that the de facto system, in which journalists are neither prevented from publishing nor punished after the fact, has produced benefits that far outweigh its costs. In fact, I'm skeptical that Kinsley could cite anything in the costs-incurred column so grave that it merits any reform in secrecy's direction. (What does he regard as the most damaging press-facilitated leak in history? That no obvious candidate springs to mind is telling, especially given how many familiar examples we can cite of classified leaks exposing scurrilous behavior.)
This is where the hypotheticals might start. Just astorture apologists invoke the ticking-time-bomb scenario because nothing in the past or the plausible future bolsters their position, proponents of punishing journalists for leaks could lean on made up scenarios.
And yes, a leaker could hypothetically send nuclear codes to Glenn Greenwald or Barton Gellman tomorrow. Under my standard, government would have to stand pat as they were published. That's the strongest argument against my position. But it is ultimately unpersuasive for two reasons: 1) neither man would plausibly publish that leak; 2) and more importantly, no journalist's decision would matter in that hypothetical. If a government employee is intent on spilling the nuclear codes, despite the legal consequences and harm to national security, he or she doesn't need the press. There are lots of ways to get information out.
Kinsley's position seems rooted in a reflexive aversion to non-governmental actors exercising such an important discretionary role in American governance. For him, it just cannot be that journalists are permitted to make these sorts of calls without risk of prosecution. But why? Having been permitted to do exactly that, the press has demonstrably performed better than the government in deciding what ought to be made public and what secrets should be kept. That wouldn't surprise the Framers, who predicted as much when they passed the First Amendment.