Is the Snowden controversy about secrecy or spying? The safe answer is both. The files that Snowden disclosed reveal a domestic intelligence-gathering operation of stunning scale. But what was perhaps most disturbing about Snowden’s revelations is that they were revelations; the vast majority of Americans had no idea of the extent of government data collection. Yet faced with the implicit question—secrecy or spying?—Reed Hundt rejects the safe answer and identifies secrecy, not spying, as the central challenge Americans face in a post-Snowden world.

Hundt’s answer is politically significant. Worries about spying are generally libertarian—worries about the threat our government poses to our freedom. A focus on secrecy, Hundt shows, raises a different set of concerns—about democracy. Without sufficient knowledge of what their agents in Washington—or Hawaʻii—are doing, citizens cannot effectively govern. Tellingly, Hundt closes with the call of a commander-in-chief, in the midst of America’s bloodiest war, for “government of the people, by the people, and for the people.” Hundt is fundamentally interested less in safeguarding citizens’ private lives than in assuring that they continue to wield public power.

To be sure, this is a strong reading of Hundt’s views. He sometimes strikes more libertarian notes, as when he writes that technology promises “new ways for people to act . . . outside the purview of business or government.” But he concludes that, in a digital age, government is more necessary than ever to the establishment of freedom and equality:

Arrest whoever stole the Bitcoins from Mt. Gox. Find the perpetrators of malware and jail them. Insist that social networks and retailers provide state-of-the-art protection for individuals’ digital information or be exposed to serious penalties.

Hundt’s nightmare scenario is one in which the government, weighed down by the cost of an opaque and inefficient data collection regime, delegates the provision of security to big-data firms. In this story, the ultimate threat of unchecked spying is not big government but the surrender of big government to the private sector.

Delegating transparency to secret-keepers may be a self-defeating proposition.

Even where Hundt celebrates the ability of individuals to use new information technologies to secure themselves from both cyber-crime and government intrusion, he also calls on government to “invest more in providing access to encrypted, safe technology.” Because vast disparities in wealth and skill could make information privacy the privilege of a few, privacy for all means more government, not less.

The idea that civil libertarian goods such as privacy and liberty depend on government power—an old progressive idea—is having a minor renaissance. Recent studies have shown, for instance, how the Post Office helped to establish privacy and free-expression norms, how the National Labor Relations Board shaped the meaning of free speech, and how administrators of the military draft developed the right of conscientious objection.

Yet the larger the role of government in securing liberty, the more pressing the problem of government secrecy becomes. If a strong state is the precondition of a safe, open Internet, then only limits on secrecy can ensure that this state remains guided by democratic deliberation. Here Hundt’s recommendations are thin. He encourages Internet companies to push for government accountability, celebrates a congressional effort to “increase transparency about agreements between the state and business,” suggests that courts might curtail secrecy, and notes the work of the Privacy and Civil Liberties Oversight Board.

There is nothing wrong with any of that, but there is also no deeper analysis of secrecy—of the secrets citizens direct our government to keep and of those that vitiate our oversight capabilities. What should really concern us, as the legal scholar David Pozen has argued, are “deep secrets,” government practices so hidden that “we do not know we do not know” about them. It is these secrets that threaten to sever all links between governance and democratic control. Here the prospects for judicial or private-sector solutions are poor, as courts and businesses typically know little about the deep secrets that need policing.

Accordingly, Pozen recommends strengthening executive-to-legislative disclosure requirements and expanding inter- and intra-agency oversight within the executive branch. These measures can force a larger and more democratically accountable set of public officials to track deep secrets and raise them from the depths when necessary. Though more skeptical of government self-policing, the legal scholar Shirin Sinnar has likewise found that inspectors general operating within federal agencies can provide “impressive transparency on national security practices.”

Delegating transparency to secret-keepers may sound like a self-defeating proposition. Yet unless we radically rethink our commitment to a powerful national security apparatus, designed for military supremacy abroad and counter-terrorism at home, Americans may have to be satisfied with commanding the watchers to watch themselves.