Official warnings of imminent—or even of not so imminent—al-Qaida attacks have (fortunately) had a perfect record: They never seem to pan out.

During the George W. Bush administration, the public was regaled repeatedly with warnings that intelligence had determined, through an in-depth analysis of “chatter” and other such information, that al-Qaida was about to strike again.

Thus in May 2004—a few months before the presidential election—Attorney General John Ashcroft, standing beside the grim-faced director of the FBI, Robert Mueller, announced that “credible intelligence from multiple sources” indicated that al-Qaida was planning an attack on the United States in the next few months and specifically intended to “hit the United States hard.” He also pointed out ominously that the group had announced that it had completed “90 percent of the arrangements” for an attack.

Two months later, Department of Homeland Security Director Tom Ridge relayed a similar warning.

No terrorist disaster transpired or was attempted during the remainder of 2004. Of course, in the view of many, calamity did nonetheless ensue with the re-election in November of George W. Bush—partly the result, they darkly suspect, of the artful, if unfulfilled, scare-mongering earlier in the year.

And the 2004 warnings were hardly unusual. A year earlier, John Negroponte, then the U.S. representative at the United Nations, declared there to be “a high probability that within two years al‑Qaida will attempt an attack using a nuclear or other weapon of mass destruction.”

And in 2007 DHS Director Michael Chertoff disclosed that his gut was telling him there’d be an attack that summer. Later that year U.S. intelligence was reportedly concluding that al-Qaida was “marshaling its reconstituted forces for a spectacular new attack on the United States.” There were also innumerable raisings of that color-coded threat indicator we all remember so fondly.

Regrettably, the Obama administration has never subjected massive homeland security expenditures to the kind of sober and systematic evaluation they so richly deserve after a decade of drunken-sailor profligacy. And it has continued to find threatening proto-al-Qaidas popping up everywhere.

However, it has reduced the official hysteria level. Only very occasionally have Obama’s representatives characterized the extremely limited terrorist threat as “existential.” The inflated phrase, “Global War on Terrorism,” has been consigned to well-deserved oblivion. And the administration has mostly eschewed generalized warnings of the Ashcroft-Ridge-Chertoff sort.

Because the current warnings and embassy closings across the Muslim world are coming from a less fear-mongering administration, then, it is possible there is more to them than there was to the older ones, and that the perfect failure record of such warnings will be broken.

But it must be said that the specifics—insofar as they exist—in these warnings don’t look any more convincing than those of old.

We are, as usual, told about “chatter” and about intercepts of vaguely-worded communications that fail to specify timing or targets. Although it can hardly be news to terrorists that their communications are constantly monitored by U.S. intelligence, the new revelations about the extent of the eavesdropping may have reminded them anew that they can easily push the U.S. into paroxysms of costly, self-punishing behavior just by communicating convincing information to their avid auditors about ominous, if idle, terror plans.

And there is déjà-vu in recent official comments that the “chatter” and streams of intelligence resemble those that preceded the 9/11 attacks. In 2007, reported warnings of an imminent “spectacular” terrorist attack included a senior official’s observation that “This is reminiscent of the warnings and intelligence we were getting in the summer of 2001.”

And those raising the alarm have a decided advantage: They can never be proved wrong. As Dan Gardner points out in his superb book, Future Babble, if there is an attack, they can claim prescience. If there isn’t, they can insist that their warnings and preparations prevented or deterred it while deftly classifying information that might determine whether that is true or not.

The suspicion this time, obviously, is that the warning, with what seems to be its outsize focus on embassies spread in a wide arc across over a dozen countries, has been primarily, and irrationally, impelled by the Benghazi experience—a tragedy that exacted exceptional political costs to those in charge.

However, experience questions whether such generalized warnings should be issued at all. If intelligence has uncovered specifics of target and time of attack, the sensible response, of course, is not to bloviate grandly, but to work to secure the likely target or targets or to use policing measure to disrupt the plot.

But if, as it appears thus far in the present case, the warnings are vague and unspecific, issuing proclamations of danger out of an “abundance of caution” (as it has been put by the State Department) scarcely helps the situation. It is not clear what individuals are supposed to do except perhaps to look over their shoulders more often, and the proclamations often exact substantial costs in unnecessary anxiety, disruption, inconvenience, and lost economic opportunity. Nothing wrong, I suppose, in telling Americans that Yemen is not the safest travel destination. But few are likely to need the warning.

At any rate, the record strongly suggests that warnings based on vague and unspecific intelligence have failed to stop or to protect against terrorist attacks. Conceivably, some attacks have been deterred by the proclamations, but, if so, one would expect truly dedicated terrorists, inconvenienced in their immediate planning, to move into action once the alarms have faded. But that, it seems, hasn’t happened.

John Mueller writes for Slate.