I can already hear the talking point: “Even the libertarian Cato Institute thinks the president needs to be able to tap your phone without a warrant…” That will doubtless be the upshot of this phenomenally disappointing Wall Street Journal op-ed by Roger Pilon on the ongoing debate over reforms to the Foreign Intelligence Surveillance Act. I’m not sure whether it makes matters better or worse that, when Cato’s legal VP takes the wrong side on the major civil liberties controversy of the day, he does it with an array of pretty poor arguments. We begin:

Today the Senate takes up a bipartisan surveillance authorization measure that’s already passed the Intelligence Committee. The clock is ticking. This Friday a temporary law called the Protect America Act will expire. If Congress does not act before then, the president’s statutory power to prevent terrorist attacks will be seriously compromised.

The rest of the piece at least stakes out a principled legal view; this is just water-carrying for Republicans. The clock is “ticking” because the White House and its proxies in the Senate sought at every turn to block temporary extensions of the PAA so that a proper debate of a significant reform to foreign intel law could be conducted. (Fortunately, it does look as though we’ll get that extension.) If the Senate actually wanted to get a bill out quickly, they could have gone with legislation that more closely matched the far saner RESTORE Act, which the House has already approved.

This dangerous situation should never have arisen. From the beginning, presidents have exercised their Article II executive power to gather foreign intelligence — in war and peace alike, without congressional or judicial intrusion. As our principal agent in foreign affairs, the president is constitutionally bound to protect the nation. For that, intelligence is essential.

Intelligence is essential on the domestic side as well, where law enforcement is the president’s main function. Yet not until 1967 did the Supreme Court require warrants for electronic surveillance. Congress codified that a year later. But both the court and Congress expressly exempted foreign-intelligence gathering from the warrant requirement.

First, as a general point, it bears noting that Roger’s approach to interpreting the scope of executive national security powers is precisely the opposite of the approach he takes to the rest of the Constitution. When he reads “commerce” in Article I, Section 8, man that means commerce—not manufacture, not consumption, not activity that affects commerce—nothing but the literal buying and selling of goods. At the blurry edges, at any point where the scope of the power is ambiguous, he wants a narrow reading and a constrained power. The volume of intrastate wheat production and consumption may in part determine the price of wheat on the national market, but it’s not interstate commerce, so on Roger’s view, Congress can’t touch it. On the other hand, because “intelligence is essential” to national security, “Commander in Chief of the Army and Navy of the United States” naturally entails the power to authorize surveillance of Americans’ communications with (at least) persons abroad, subject to neither limitation nor, indeed, even scrutiny by the other branches.

Second, the historical reading we get here is a bit tendentious. I’m not sure what is meant to be the point of observing that “only” in 1967 did the court finally recognize the absurdity, given new technology, of restricting the constitutional meaning of “search” to physical incursion on property. The contrast with presidential practice “from the beginning” seems calculated to suggest that any application of the Fourth Amendment on the wire is somehow suspect.

But since Roger is invoking Katz v. United States, let’s have a look at that express exemption for intel gathering. It consists, essentially, of a footnote reading: “Whether safeguards other than prior authorization by a magistrate would satisfy the Fourth Amendment in a situation involving the national security is a question not presented by this case.”

This is an exemption only in the narrowest sense, in that it recognizes the unique problems presented by national security surveillance, and so declines to pass judgment on such surveillance either way within the rubric of a case concerning ordinary criminal wiretaps. Notice, though, that the footnote contemplates “safeguards other than prior authorization by a magistrate”. The inclusion of that word “prior” suggests an assumption that some kind of judicial scrutiny, even if after the fact, is appropriate. Whatever the nature of those possible safeguards, though, one imagines they were meant to consist of something a bit more robust than a presidential promise to behave.

Unfortunately, the exception was not to last. Following the Vietnam War, Congress increasingly inserted itself into foreign affairs, as with the 1973 War Powers Act. With the Foreign Intelligence Surveillance Act, passed in 1978, Congress began micromanaging foreign intelligence gathering. That produced the “wall” between foreign and domestic intelligence gathering — with foreign-intelligence agents focused on security, and domestic agents on prosecution and hence on obtaining “admissible” evidence. Neither side talked to the other. Many believe the resulting communications failures played a role in 9/11.

Conspicuously absent from this miniature history of congressional interference: any mention of the massive abuse of surveillance power uncovered by the Church Committee, which is what actually prompted FISA’s passage. (“And I would’ve gotten away with it, too… if it weren’t for that meddling Congress!”) Apparently legislators were just in a Gladys Kravitz sort of mood. Recall that the “micromanagement” here consisted of establishing a process for court oversight of wiretaps on domestic parties, oversight so burdensome that until 2003, the FISA court had never rejected an application. (It has since come out that many of those rubber-stamped applications were falsified.) As for “the wall,” the most dramatic claims about how the foreign/domestic division stymied investigation of the 9/11 terrorists have been debunked, and in any event, most of the bricks in that “wall” were executive orders, administrative rules, and intel community norms, not statutes.

In the aftermath of 9/11, believing FISA to be hopelessly inadequate, President Bush instituted his terrorist surveillance program (TSP) — but not before advising key members of Congress. Nevertheless, a firestorm ensued when the New York Times made the program public in December 2005. The controversy continued until January 2007, when the White House announced that henceforth it would gather intelligence under FISA’s antiquated restrictions.

How kind of the president to “advise” some “key” legislators of his decision to simply ignore the law specifying the “exclusive means” by which foreign intel surveillance of U.S. persons may be conducted. On Roger’s theory, of course, there can be no conclusion but that this law is unconstitutional from start to finish, so this was pure charity.

Cooler heads in Congress grew concerned after Director of National Intelligence Michael McConnell testified in July that “we’re actually missing a significant portion of [the intelligence] we should be getting.” That led to last August’s six-month fix, which expires this week.

Yes, Mike McConnell’s fearmongering represents “cooler heads” prevailing. I defy anyone to look at the process that led to the passage of the Protect America Act and find anything “cool” about it. In fact, Congress took McConnell largely at his word when the DNI asserted, quite implausibly, that the FISA court had somehow manufactured a blanket warrant requirement for foreign-to-foreign communications passing through U.S. switches. Nobody who actually understands FISA finds this remotely plausible. Since the court has declined to release the ruling in question, we can’t know what it actually says. But it bears note that the administration did not bother to appeal the ruling, or to modify their program in any way to pass muster. They preferred to use it as leverage to terrify Congress into expanding their powers.

Obviously, this is no way to conduct the serious business of foreign intelligence. The ever-changing rules — criminalizing transgressions — leave officials playing it safe in a world of risks.

Wait, I keep forgetting: Are the rules “antiquated” because they’re part of a framework established way back in 1978, or are they “ever changing”?

The Senate bill would be an improvement, not least because it provides retroactive liability protection for telecom companies that allegedly assisted the government after 9/11. But the deeper problem is the very idea of congressional micromanagement.

The Senate bill would require showing probable cause before targeting even U.S. persons abroad, dramatically increasing the role of the FISA court. As Judge Richard Posner wrote on this page two years ago, FISA may be valuable for monitoring communications of known terrorists, “but it is hopeless as a framework for detecting terrorists. It requires that surveillance be conducted pursuant to warrants based on probable cause to believe that the target of surveillance is a terrorist, when the desperate need is to find out who is a terrorist.”

Troubling indeed. But not—what’s the word?—oh yes: true. All the versions of the FISA amendements would permit surveillance where a “significant purpose” of the investigation was to gather foreign intel. This would not require a showing that the target is a terrorist, or knows any terrorists, or even sports a vaguely terrorist-ish beard. The “Wyden amendment” bars the deliberate targeting of known U.S. persons (by specifically flagging their e-mails or phones); it applies, by definition, once you have a specific individual target. Posner was talking about “vacuum cleaner” interception and data mining, which is another topic altogether. But FISA has never raised a barrier to that kind of surveillance as pertains to foreign-to-foreign interceptions. More to the point, the post-PATRIOT FISA already stipulated that FISA warrants were to specify the identity of the target and communication facilities to be surveilled “if known”—leaving latitude to conduct investigations where evidence exists sufficient to narrow the search, but before the government has established “who is a terrorist.”

The technical impediments to legislating are even greater. We’re long past alligator clips on copper wires. Today, electronic communication is broken into discrete packets that travel along independent routes before being reassembled. As K.A. Taipale, executive director of the Center for Advanced Studies in Science and Technology Policy, has written, “even targeting a specific message from a known sender requires intercepting (i.e., scanning and filtering) the entire communication flow.” Yet the Senate bill requires that intelligence analysts count the people in the U.S. whose communications were “reviewed,” an all but impossible distraction for analysts already stretched.

This part is just flat-out crazy. Nobody has ever suggested that the filtering of a data stream to extract a particular target’s communications somehow counts as surveillance of the entire stream. The implication here is that if NSA drops a Naurus box on a wire and automatically plucks out all the emails to or from a particular source, FISA is going to demand that the agency keep track of who owns every bit that passed through the machine, whether or not this other data was stored or seen by a human eye. In most cases, when this is done with the cooperation of an ISP, the government does not actually need to intercept the entire communications flow. But even leaving that aside, as tech reporter Declan McCullagh observes, “the scanning and filtering is done automatically in a split second and the results are not retained. Alleging this amounts to ‘reviewing’ — something only a human can do — is like complaining that your cat is voyeur for watching you have sex.” So Roger’s insinuation here is not just false; it is wildly, ludicrously, hobo tall-tale false. Basically, he’s shouting “cybernets! complicated!” and hoping nobody will notice that what he’s claiming is preposterous.

Privacy concerns are not trivial. The Constitution protects against “unreasonable” searches. But even with law enforcement, where the main function is ex post prosecution, not ex ante protection, there are numerous exceptions to the Fourth Amendment’s warrant requirement. Yet Congress insists still on micromanaging the president — and he, by failing to assert his authority early on, is now reduced to bargaining with Congress over minutia that will soon be as obsolete and dangerous as the underlying act is today.

Recall here that since we’re breaking down the old “wall,” information gathered as part of a foreign intel sweep—without a warrant or meaningful oversight if the intel bill passes—can be passed on to domestic law enforcement for use in plain vanilla criminal prosecutions. And what are the “minutiae” here? Details like: Will bulk collection of data without specified targets be permitted? Will courts be able to look into whether agencies are actually complying with minimization requirements designed to prevent the retention of information about innocent Americans? Can Congress learn how the FISA courts have interpreted the legislation they’ve passed? At what point does collection of information about a U.S. person become important enough a purpose of an investigation that it triggers a warrant requirement? Details, details.

John Locke, no sometime civil libertarian, put it well when he observed that the foreign affairs power “is much less capable to be directed by antecedent, standing, positive Laws, than [by] the Executive.” The Federalist’s authors, James Madison, Alexander Hamilton and John Jay, all agreed. The remedy for executive incompetence or recklessness in foreign affairs is political — not legislative, much less legal. Congress, to say nothing of the courts, can no more manage such affairs than it can the economy. What better evidence than these surveillance fits and starts?

As long as we’re quoting Locke, I rather prefer a passage that comes a few chapters later:

Where-ever law ends, tyranny begins, if the law be transgressed to another’s harm; and whosoever in authority exceeds the power given him by the law, and makes use of the force he has under his command, to compass that upon the subject, which the law allows not, ceases in that to be a magistrate; and, acting without authority, may be opposed, as any other man, who by force invades the right of another. This is acknowledged in subordinate magistrates. He that hath authority to seize my person in the street, may be opposed as a thief and a robber, if he endeavours to break into my house to execute a writ, notwithstanding that I know he has such a warrant, and such a legal authority, as will impower him to arrest me abroad. And why this should not hold in the highest, as well as in the most inferior magistrate, I would gladly be informed.

Roger is not the only voice on this subject emanating from Cato, but he’s surely the loudest. And as we undertake the most sweeping revision of the nation’s foreign intelligence framework in three decades, he’s positioning himself—which in most people’s minds will be equivalent to positioning the Institute—against accountability and oversight, against reasonable privacy safeguards, and (for what it’s worth) against just about every other civil liberties group that has weighed in on the question. I respect that Roger’s views are considered and sincerely held, and I regard it as a general virtue that Cato gives their scholars latitude to go off the reservation when that’s where their thinking takes them. But I’m still appalled to see a voice for limited government crying for unlimited spying at the first hint of danger.

Addendum: I meant to note in the original post that Roger’s stance when he’s debated this in the past has been that, so long as the president says he’s acting in the interests of national security, the only accountability to which he’s subject—here or abroad—is political. Roll that around in your head for a minute. Political accountability For a classified secret surveillance program. Free cookie if you can spot the flaw in this logic.

Addendum II: Tim Lee has a somewhat more diplomatic take at Cato-at-Liberty.

Addendum III: Edited and expanded a graf above to separate out some points I’d muddled together originally.