Antonin Kratochvil/VII

Residents of Basra fleeing the city during the invasion of Iraq, March 30, 2003

Garry Wills does not tell us which came first in the conception of his new book—baffled wonder that American presidents got away with going to war in Korea, Vietnam, the Persian Gulf, and Iraq on their own say-so? Or the moment in late 2008 when then Vice President Dick Cheney reminded viewers watching Fox News of the extraordinary authority conceded to American presidents since the invention of the Bomb?

Cheney put his finger on a core truth about the modern presidency. Only the president can order the use of a nuclear weapon on an enemy he identifies, at a moment he chooses, for reasons he finds adequate. This power is not only theoretical; to give it practical effect the president is always accompanied by an aide carrying a briefcase containing the authorization codes without which no American nuclear weapon can be armed for use.

Cheney has a long history as a champion of presidential power. There is not much he thinks a president cannot do, and his attitude is widely shared by modern Republicans, who have given it the dignity of a theory with a name—the unitary executive. But strip away the persiflage, Democrats say, and we’ve seen the unitary executive before. It grants every president what Richard Nixon thought he had—a free hand in any matter where national security was at stake, and he did mean free. “When the president does it,” Nixon explained to aides during the Watergate crisis, “it’s not a crime.”

Democrats didn’t let Nixon get away with it then and they don’t want to let modern Republicans get away with it now. But on the core point that Cheney made to Fox News there is not much argument. The president has been conceded life-and-death authority over use of the Bomb. “He doesn’t have to check with anybody,” said Cheney.

It is Wills who places the emphasis on Cheney’s words. For him, this is the crux. A long book might be written about authority for nuclear use, listing many limits and qualifications, but in the end such a book would go along with Cheney. American presidents, and the leaders of other nations in their turn, acquired immense power at once from the invention of the Bomb, and expanding power later to manage the elaboration of ever-swifter and more accurate delivery systems for it.

It does not stop there. Wills argues that the “world of perpetual emergency” created by the power of the Bomb presses many other authorities on the presidency, of which the most significant is control of information. Restricting knowledge of the development of the Bomb was rule one of the man who managed its invention, General Leslie Groves. He wanted the scientists to “stick to their knitting,” make the gadget work, but let others decide what to do with it. President Roosevelt knew little about the Manhattan Project and Vice President Truman knew nothing until he was briefed, none too promptly, after the death of Roosevelt. Keeping the primal secret, Dean Acheson noted at the end of the war, spawned “a separate state, with…thousands of secrets.”

It took less than a decade for the separate state to evolve into the National Security State, a vast apparatus for “defense” that threatened to replace business as the business of America. Wills identifies twenty-five separate milestones along the way, beginning with the reorganization of World War II intelligence organizations. His brisk history identifies half a dozen key speeches and articles, the passage of several laws, the creation of new governmental offices to manage matters of war and peace, the writing of treaties, the first forays into clandestine intelligence activities—all of it shaped by a global political rivalry with the Soviet Union, made potentially lethal by the fact of the Bomb and its progeny. The milestones in Wills’s capsule history of the cold war are all familiar but his argument is new and important.

There have been plenty of books about the Bomb. What sets Bomb Power apart is the object of Wills’s concern. It is not the all-too-familiar horror of Armageddon that troubles him, but a generally overlooked side effect—the peremptory rewriting of the United States Constitution driven by the exigencies of nuclear weaponry. “The national Security State,” he remarks, “is in permanent constitutional crisis.” What he means is that presidents in the name of national security have taken unto themselves an ever-expanding range of powers that are not granted, and in some cases are explicitly prohibited, by the Constitution. Starting at the top are the de facto power to “declare” or wage war, reserved by the Constitution for Congress; to decide what information is secret, who gets to know about it, and how much they get to know; to enter into foreign alliances and agreements with little or no oversight; to suspend constitutional rights by executive order (as Roosevelt did when he interned Japanese-Americans during World War II); to ignore laws enacted by Congress in whole or in part by declaration of intent to do so in “signing statements” and to spend money without, or contrary to, congressional authorization.


Wills does not argue that the assumption of these various rights or powers all followed the arrival of the Bomb on the American political stage. Presidential power was the subject of much debate during the Constitutional Convention, where a lively fear of tyranny prompted adoption of numerous checks on a president’s freedom to act. No provision of the Constitution is clearer than the one reserving the right to declare war to Congress, and none has been stretched, finessed, or violated more often, beginning with President Jefferson’s decision in 1801 to inform Congress that he was sending a naval force to attack the Barbary pirates on the shores of Tripoli. Perhaps the best-known, and the most often cited, abrogation of the Constitution by executive order came early in the Civil War when President Lincoln suspended the right of habeas corpus. The suspension seems inevitable in retrospect; were Confederate prisoners to be given access to lawyers and courts to challenge their detention? Later presidents have seized on this precedent and run with it, asserting the right, in Wills’s view, to “give orders for anything they want.”

Over the years presidents have wanted a lot of things. Confronted by Iranian demands for ransom of American hostages, Jimmy Carter granted himself access to emergency economic funds to make the payoff. In like manner, Nixon created the Environmental Protection Agency and Kennedy created the Peace Corps, both by executive order. Other presidents were blocked, as Harry Truman was when he wanted to break a railroad strike by drafting workers into the army in 1946 and when he wanted to nationalize steel mills to end another strike in 1952.

But Wills has not gone to all this trouble to give Nixon and Truman a knuckle-rapping. It is the extraordinary range of powers asserted by George W. Bush that worries him. Bush and Cheney seemed to know immediately how to wrest advantage from failure. The dust was still settling over lower Manhattan when Bush declared that the United States was “at war” with the al-Qaeda terrorists who destroyed the World Trade Center in 2001. Naming his response a “war on terror” was not merely a PR officer’s turn of phrase signaling resolution. War is a legal term. By using it Bush assumed the full panoply of extraordinary powers wielded by other presidents in previous wars. At the same time he made himself a “wartime president,” thereby placing a heavy requirement on the political opposition to rally round and abandon partisanship for the good of the country.

Bush used his “wartime” powers to take many bold initiatives, and he used his authority over the classification of information to keep what he did secret. American and international law were overthrown or turned inside out by kidnapping persons of interest on foreign soil (“extraordinary rendition”), surveilling American citizens without court warrants required by law, imprisoning American citizens without legal charge, establishing secret prisons in foreign countries under secret agreements promising no one yet knows what, holding prisoners in legal limbo at Guantánamo Bay, repudiating the Geneva Conventions, and violating American tradition and law by authorizing the torture of prisoners.

But the most important of Bush’s initiatives were the invasions of Afghanistan and Iraq, acts of precisely the kind of war that the Constitution stipulates may be declared only by Congress. No stretching of law or equity is required to justify the invasion of Afghanistan in hot pursuit of the criminals who had attacked the United States, but what justified the invasion of Iraq? Saddam Hussein had not threatened (rhetoric aside) or attacked the United States or its allies, no international body had authorized resort to war, and Congress had not declared war.

What Congress did at President Bush’s behest was to pass a joint resolution in October 2002 built mainly on a chain of frightening ifs—if Iraq was in fact “actively seeking a nuclear weapons capability,” and if Iraq succeeded in these efforts, and if Iraq attacked the United States with nuclear weapons, or if Iraq gave them to terrorist organizations, and if the terrorists used them, then “the extreme magnitude of harm that would result” would justify preemptive resort to military attack. Congress lacked the presence of mind to question the chain of ifs and voted not unanimously but still overwhelmingly to grant President Bush the authority “to use the Armed Forces of the United States as he determines to be necessary and appropriate in order to…defend the national security of the United States.”


The administration’s relentless banging of the drum for war was one reason Congress went along. Fear of being thought weak, thinking that the President must know something, panic aroused by the frightening ifs, the desire to pay someone back for the pain inflicted by September 11, the euphoria of the crowd—doubtless all played a part in Congress’s supine delivery of what was demanded. Wills rejects the many explanations for one. “The main reason is the Bomb,” he writes. In the crunch, Wills believes, Congress will always conclude that the Bomb is too dangerous to mess with, and defer to the one official who might know something when the rest of us don’t.

It is “the Bomb” in general that gives latitude to presidents, in Wills’s view, but in the case of Iraq it was fear of a specific bomb being hammered together in Saddam Hussein’s nuclear smithy that smothered doubts in Congress. No American president ever let out that he knew something more clearly or confidently than Bush and Cheney did. What’s surprising is how well it worked. The history of the cold war had been one scary development after another. The “bomber gap” of the early 1950s, the “missile gap” of the later 1950s, the first-strike capacity promised by Soviet missiles with multiple warheads, and the illegal Soviet radar arrays that could guide an anti-ballistic missile system all threatened “strategic breakout”—a sudden change in the military balance of terror, plunging the United States into peril between one day and the next. A lot of new hardware was developed over the decades but no breakout followed. Inevitably, threat inflation led to threat exhaustion; it was difficult in the latter years of the cold war to get anybody worked up.

That had changed by 2002. No one doubted that Osama bin Laden would use a nuke if he had one. Saddam Hussein had proved that he was a man of deep malice and implacable will. In their many claims that Saddam Hussein was “actively seeking” nuclear weapons, Bush and Cheney not only sounded like but flatly said they knew something. In late 2002 the CIA spelled out what it knew in a special National Intelligence Estimate (NIE). But it was all smoke and mirrors. As is now well known, there were no Iraqi weapons of mass destruction and no active programs to build them.

Explaining this bald fact confronts historians with a painful and perhaps even professionally dangerous choice: Are they to tip the hat to the White House, and describe these mammoth errors as honest mistakes, thus letting Bush and Cheney off the hook? Or shall the historians say frankly what the facts declare—that the White House wanted war, needed a plausible justification for war, decided that Iraqi WMD programs were it, dragged the necessary “estimates” out of the CIA, and used them to batter Congress into submission?

Robert Jervis in his new book, Why Intelligence Fails, addresses two questions capturing the historians’ dilemma. The first asks why the CIA failed to foresee the tidal pull of events in 1979 that swept away the Shah of Iran. Opportunities abound here for blaming the uncertainties of the process by which estimates are made. The second question offers less wiggle room: Why did the CIA base so many false claims on such flimsy evidence?

Jervis is a professor of international politics at Columbia University and a career-long scholar of intelligence with a deep knowledge of the organizations that make up the American intelligence community. All these agencies are professionally competent and well funded, but they routinely stumble over the facts that complicate the lives of analysts—the capacity of opponents to hide what they are doing; resistance in high places to bad news; the inherent difficulty of answering any question beginning with the word “when”; friendly regimes who want no snooping in their backyard; equipment breakdowns at crucial moments; fickle weather; noisy problems in one corner of the world that distract from quiet problems in another. The list is long. At one time or another since the late 1940s the CIA in particular has been wrong about practically everything, thereby establishing a kind of natural history of error—ways in which intelligence estimators typically misread events.

In 1979, Jervis was a scholar in residence in the CIA’s then estimating office, the National Foreign Intelligence Assessment Center (NFAC), when he was asked to study the NFAC’s failure between summer 1977 and late 1978 to see that the Shah of Iran was tottering on his throne. Over-rosy readings of popular support for the regime ended only a few months before the Shah was driven into exile, ending Iran’s two decades as a bulwark of American policy in the Middle East. It was not the CIA that ended the Shah’s reign, or brought the Ayatollah Khomeini to power, or limited Jimmy Carter to one term in the White House, but the agency’s persistent misreading helped the White House to put off painful choices until they no longer mattered.

The purpose of Jervis’s study was not to explain why the Iranian revolution occurred, but why the NFAC failed to see it coming. The sources of error he cites are all familiar to students of American intelligence. Among them: “almost no” Farsi speakers, too few agents in the field, no contact with the Shah’s opposition, few contacts with Iran experts outside the agency, ignorance of basic facts including the Shah’s losing battle with cancer, and a history of being wrong about the Shah’s staying power. Equally important, Jervis found, were the NFAC’s preconceptions. It had a secular bias dismissing religion as unimportant, assumed that American-backed regimes are threatened only from the left, and was confident that the Shah would “crack down” whenever things got serious. Nudging every judgment in the wrong direction was the universal experience of estimators that it is always safest to predict that nothing will change.

Barely mentioned by Jervis are two conspicuous facts—the administration’s deep bureaucratic reluctance to give up on the Shah and the CIA’s tepid “collection effort” to find out what was going on. No Farsi speakers? No agents in the field? No effort to track what Khomeini was actually saying in the taped sermons circulating in the bazaars? No effort to understand why merchants repeatedly closed their shops to support the opposition? Could the CIA be credited as wanting the truth when it made so little effort to smoke it out?

One does not have far to look for the dog that didn’t bark in this case. It was the White House, which fired off no angry demands for better information until it was too late. Jervis concedes that the rosy estimates were “generally…consistent with US policy,” but concludes that the analysts “were not aware” of pressure to write estimates to order. It would be hard to think of a softer way to report that no analyst was willing to go on record saying what any bureaucrat would take as a fact of life—promotions do not come to lowly analysts in offices without windows who tell presidents what everybody knows the president does not want to hear.

White House fingerprints are still more glaringly evident on the CIA’s “high confidence” conclusions about nonexistent Iraqi weapons of mass destruction in the six months running up to the outbreak of war in March 2003. This is a much-argued subject but Jervis seems not to have noted or credited the steadily growing body of evidence of White House pressure—the fact, for example, that the agency’s deputy director of intelligence, Jami Miscik, threatened to resign if the White House did not stop trying to bludgeon the CIA into reporting a “link” between Iraq and the attacks on September 11. Vice President Cheney made at least eight trips to agency headquarters to argue this and other matters concerning Iraq. Cheney’s assistant Scooter Libby and Paul Wolfowitz from the office of the secretary of defense would not let the subject drop. The national security adviser, Stephen Hadley, was still trying to bully Miscik into changing the wording of a paper on the “link” only two months before the outbreak of war.

But Jervis demands a strangely high standard for proof of White House meddling. Miscik’s threat to resign apparently falls short because she did not actually deliver the words demanded by Hadley, and did not formally charge the White House with trying to pressure her. Also dismissed or downplayed are the strange handling of the reports from “Curveball,” the unreliable source for claims that Iraq was building mobile biological weapons labs; and the erroneous insistence of estimators that there could be only one explanation for Iraq’s purchase of aluminum tubes—their use in centrifuges to collect bomb-grade material from natural uranium.

Jervis also ignores the flagrant signaling by the Bush administration of what it wanted from the NIE—scary reports that would be grounds for war. Looked at calmly in a moment of quiet, as Robert Jervis might have done, the evidence of Iraqi WMD programs cited in the NIE is plainly iffy in nature, comes from doubtful sources, and at best suggests that Saddam Hussein was “seeking” the Bomb, not that he had or was about to get one. Even to say that he was “actively seeking” the Bomb was an exaggeration. But in August 2002, before the Iraq NIE had been written, Cheney swept doubts aside and said in a speech to the American Legion: “Simply stated, there is no doubt that Saddam Hussein now has weapons of mass destruction.” The White House required no smoking guns, and the agency delivered with “high confidence” the scariest stuff it had.

It is evident that Jervis loves the whole process of intelligence analysis. He retells the WMD story in a calm and reasonable manner, listing the many fascinating ways in which honest men might get things wrong, but he avoids the things that give the game away. Three stand out. The first is the lack of an occasion for the sudden obsession with Iraq—an important discovery with dangerous implications. The occasion is supposed to come first—a reason why we should be paying attention to something now, like the discovery that Soviet ships were delivering large, long objects to Cuba in 1962. That demanded an explanation. Nothing new emerged about Iraqi weapons programs in 2002; it was September 11 and loud talk from the White House that made the issue urgent. Second is the fact that the evidence trumpeted by the administration was all suggestive rather than dispositive. Yellowcake from Niger, to cite one now-notorious example, suggested an interest in uranium; usable fissionable material was still many years away, which would also have been true even if the aluminum tubes were intended for centrifuges.

Third, and most significant, is the CIA’s failure to tell UN inspectors in Iraq where to find Saddam’s WMDs or the laboratories and factories required to build them. By the time Secretary of State Colin Powell addressed the UN in February 2003, it was obvious to anyone paying attention that the CIA did not know where to find anything connected with WMDs in Iraq, and it knew it did not know (a fact that was being confirmed by approximately seven hundred UN inspections of some five hundred sites, including about three dozen given to the inspectors by intelligence agencies, before the inspectors were told by the US to leave). From that moment forward, no claim of imminent danger could be described as intellectually honest. Perhaps Jervis ignores this fact because the failure became undeniable only after the NIE was published in October 2002, but it was obvious months before the war started. Why no one said anything at the time, and why Jervis says nothing now, are silent testimony to the power of the White House to have its way.

Presidential power is one of the mysteries of the American political system. National emergencies like the Civil War and the Great Depression gave a protective air of urgency to anything a president was bold enough to insist he had to do. A man who busts his way to the head of a ticket line often gets his way and a president, backed by a political party, with an army of spokesmen and mass media trained to put his argument in the lead of every news story in which he figures, can push a long way in any direction they want to go before opponents muster strength to slow them down. Even without nuclear dangers, presidents start off with the odds in their favor. Add the Bomb, Garry Wills argues in his crisp and troubling book, and there is little stopping them.

Robert Jervis appears to understand what happened to the CIA estimates of WMDs, but he won’t connect the dots. He quotes Jennifer Glaudemans, a CIA analyst who provided a vivid description of politicization at a hearing in 1991: “Politicization is like fog. Though you cannot hold it in your hands, or nail it to a wall, it does exist, it is real, and it does affect people’s behavior.” At times, Jervis writes, Bush was guilty of “browbeating and ignoring of intelligence,” “of twisting of evidence,” and of “cherry-picking” and “stovepiping” of bits of information to back up “messages they wanted the public to believe.” Presidents “need” intelligence that is in “close harmony” with their policy. When the CIA balked at confirming an Iraq–al-Qaeda link, “the administration put great pressure on intelligence to come to a different view.” Presidents have “different needs and perspectives” than intelligence analysts, and this “will produce pressures on and distortions of intelligence.”

These observations might seem a preamble to a conclusion that the CIA “bowed to pressures to tell the policymakers what they wanted to hear.” How else can we explain why the analysts’ conclusions all leaned in the direction desired by the administration, and were all wrong? To reach a contrary view with intellectual honor intact would seem to demand a vigorous rebuttal by Jervis, but he only whispers his verdict. Official reports of the WMD fiasco “rejected” political pressure as the culprit, Jervis says, adding, “I think the official reports are largely correct.”

Picking a fight with a president is dangerous, no less to academics when it comes time to write the history than it was to intelligence bureaucrats in the heat of events. With his mild conclusion, Jervis tips his hat to the White House, blaming the CIA for getting it wrong, just as Bush and Cheney do, and thus confirms what we might call the Garry Wills rule: ordinary mortals get out of the way when a president cites the Bomb for what he intends to do.

This Issue

May 27, 2010