What if Snowden’s wrong? What if there is no pervasive illegality in the National Security Agency’s surveillance programs?

Indeed, for all the excitement generated by Snowden’s disclosures, there is no proof of any systemic, deliberate violations of law. Based on the ruling in a 1979 Supreme Court case, Smith v. Maryland, it is well established that individuals do not have an expectation of privacy in the phone numbers they call. This is not entirely surprising; we all know that we’re already sharing that information with the phone company. In the same way, it’s long established that the government has great latitude in intercepting communications between the United States and other countries. It’s true, too, that while the Foreign Intelligence Surveillance Act court is largely toothless, it has, on occasion, rejected some N.S.A. procedures, and the agency has made adjustments in response. That is not the act of an entirely lawless agency.

It is true that, as the Washington Post’s Barton Gellman recently reported, the N.S.A. sometimes went beyond its authority. According to Gellman, the agency privately admits to two thousand seven hundred and seventy-six incidents of unauthorized collection of data within a twelve-month period. This is bad—but it’s not clear how bad. If it’s that many incidents out of a total of, say, three thousand initiatives, then it’s very bad. But if—as is far more likely—it’s two thousand seven hundred and seventy-six incidents out of many millions, then the errors are less serious. There should be no mistakes, of course. But government surveillance, like any human activity, is going to have errors, and it’s far from clear, at this point, that the N.S.A.’s errors amounted to a major violation of law or an invasion of privacy.