CHAPTER 2
REGIME-CHANGE FANTASIES

NINE MONTHS AFTER the Iranian revolution in 1979, a thirty-six-year-old man from Wichita, Kansas, found himself sitting in a hotel in Algeria facing a motley collection of some of the world's most famous revolutionaries.

The young man was Robert Gates, a fast-rising career intelligence officer temporarily on loan to the White House from the CIA. Three decades later, after running the agency, serving four presidents, and giving up his favorite job as president of Texas A&M, he would return to Washington to try to rescue the Bush administration from some of its worst mistakes. Iran's nuclear program and spreading influence in the Middle East would rank among his most pressing problems, after figuring out a way to extract the United States from Iraq. Thanks in large part to his venture in Algiers, he would bring with him many thoughts about the right way—and the wrong way—to confront the Iranians.

But to look at him that day in Algeria, you might have thought he was another one of the clean-cut Midwesterners who, moving from one oil-rich dictatorship to another, sold drilling equipment or Rolls-Royces to newly minted billionaires. His hair was cropped close, his broad face had a winning smile, and he had a plain way of speaking that hid a biting sense of humor and one of the sharpest analytical minds in Washington. He didn't look like what he was— an expert in the workings of the Soviet Union and one of the nation's most skilled cold warriors.

But for this trip he was largely a note-taker for his boss, Zbigniew Brzezinski, President Carter's national security adviser. Officially, they had come to the Mediterranean coastal city for an anniversary celebration of the Algerian revolution, not usually an event that attracts high-level American attention. Needless to say, they had another agenda.

The collection of revolutionaries who showed up in Algiers that week ranged from egomaniacs to accomplished guerrilla fighters to novices feeling their way through the underground world of militant networks. It was, Gates recalled later, “an intelligence officer's dream come true. All the principal thugs in the world were present.”

Yasser Arafat was there, his signature pistols in his belt. So was Hafiz al-Assad, the Syrian strongman. At the banquet, where Gates tried to circulate anonymously, he saw Gen. Vo Nguyen Giap, the Vietnamese strategist who was responsible, as much as any man except perhaps Ho Chi Minh, for plotting the victory that led to the United States’ forced departure from South Vietnam, a moment of humiliation that would shape the world Gates later had to navigate.

The people Brzezinski and Gates were focused on, however, were less well known around the world: a delegation sent by Iran's clerics, representing the government formed by the Islamic revolutionaries who were still basking in the glory of their huge and improbable success, the ouster of the American-backed Shah of Iran.

By any measure, theirs was a stunning achievement. They had taken over one of the largest oil-producing countries in the world, unseating one of Washington's greatest allies. Just eight months before, Iran's most famous Shiite cleric-in-exile, the cunning, enigmatic Ayatollah Khomeini, had returned to Tehran. Almost immediately the country plunged into what Gates later called a “reign of terror.”

Much as is the case today, no one in Washington could quite figure out who was really in charge in Iran. The intelligence agencies had gotten it wrong. Less than a month after Khomeini arrived in Tehran, they told Brzezinski and Gates they doubted the cleric had enough political power to survive. (Khomeini did just fine; he named himself “supreme leader” and stayed in power until he died a decade later.)

As they maneuvered around Algiers, Brzezinski and Gates were unsure how this revolutionary government would deal with the United States. As much as the clerical leaders reviled America, Khomeini regarded the Soviets as a great threat to Iran as well. Brzezinski was there to reassure the Iranians that the United States was willing to recognize their new government and was prepared to talk about resuming the sale of military equipment.

Together, Brzezinski and Gates showed up at the hotel suite to meet the Iranian delegation. Waiting for them was Iran's newly installed prime minister, Mehdi Bazargan, a French-trained engineer already in his early seventies, who had been imprisoned by the Shah several times. He was clearly already uncomfortable with the radical clerics who were micromanaging Iran's every move. But he was Washington's best hope, the moderate realist that Americans are always seeking in the Iranian leadership. He was accompanied by the Iranian foreign minister, Ibrahim Yazdi, an aide to Khomeini during his exile in Paris, and Defense Minister Mustafa Ali Chamran.

Gates knew there was no way this was going to be an easy encounter. The United States had installed Mohammad Reza Shah in a CIA-backed coup, and for more than two decades he was viewed as an American puppet. Washington had sold him his weapons, cooperated in neutralizing his enemies, supported his feared intelligence service, SAVAK, and overlooked his repressive policies—and his imprisonment of dissidents like Bazargan—in the name of supporting an anticommunist ally. In a mistake that would be repeated time and time again—from Latin America to Pakistan—the United States backed an autocrat even as it was becoming clear that his star was fading and his ability to hold on to power was eroding.

By the time a popular revolt forced the Shah to flee Iran in January 1979, the Carter administration was locked in a fierce internal debate about what to do about him. The man who once was the darling of the CIA and a series of American presidents was suddenly royalty without a country. Brzezinski was pushing to offer him refuge in the United States; Secretary of State Cyrus Vance was warning that such an offer would inflame the Iranians. Just days before Brzezinski and Gates landed in Algiers, President Carter invited the Shah to the States, chiefly on humanitarian grounds; by then the ousted monarch was desperately ill with lymphoma and needed sophisticated medical care.

As Gates recalled later, the State Department had checked with the new Iranian government, “explained the circumstances, and received assurances that Americans in Iran would be protected” if the Iranian public reacted badly. On October 23, the Shah arrived on American soil. A few days later came Brzezinski's first meeting with the new government, meant to explore whether the United States could live and cooperate, if uneasily, with Khomeini's Shiite theocracy.

As they entered the hotel room, Gates recalled later, “their greeting and the tone of the entire meeting were surprisingly friendly under the circumstances.”1 Brzezinski, ever the realist, suggested to the aging Bazargan that Iran and America shared a common interest in opposing the Soviet Union. Global realities and mutual national interests, he suggested, should trump past differences.

But as Gates described the session to me twenty-nine years later, in his office at the Pentagon, one image stuck in his mind—one that guided how he would deal with the Iranians years later.

“Bazargan, Chamran, and Yazdi,” Gates said with a smile, “all said, ‘Give us the Shah. Give us the Shah.’ Back and forth, back and forth, ‘Give us the Shah!’”

“This was our first dialogue with the Iranians,” he said, shaking his head, ‘Give us the Shah!’”2

The demands went on endlessly. Finally, Gates recalled, Brzezinski stood and said with some drama in his voice, “To return the Shah to you would be incompatible with our national honor.”

To Gates's surprise, the reaction inside the room to Brzezinski's declaration was subdued, but it did not take long for the meeting to break up. Later Gates remembered thinking that despite the differences, the parting seemed friendly.3

Whatever feelings of goodwill existed when they left the room did not last.

“Three days later,” Gates said to me, “the embassy was seized and two weeks later all three of those guys were out of power.” Khomeini backed the hostage-takers; Bazargan and his government resigned in protest, declaring that the taking of the hostages was not in Iran's interest. It was the beginning of a crisis that would dominate the rest of Carter's presidency, ultimately contributing to his undoing in the 1980 elections. Relations between Iran and the United States were poisoned for more than a quarter of a century.

“So passeth the first dialogue between the United States and Iran,” Gates said as he tipped back his chair at the round conference table in the middle of the secretary of defense's office at the Pentagon. “Every president since has tried to open a dialogue with the Iranians and none of them have turned out well. And a couple nearly got presidents in jail.”4

IN ALGIERS, Brzezinski and Gates were trying to recover from bad bets that Washington had placed on Iran—and that still haunt us. The first was that the Shah would somehow find a way to cling to power. The second gambled that the United States could safely give the Shah the tools he needed to create his own civilian nuclear program. Of course, to help Iran produce nuclear energy, the program would include the basic tools needed to make bomb fuel. Washington had forgotten the ironclad mathematical realities of nuclear politics: While the half-life of the average dictator's tenure is about fifteen years, the half-life of uranium-235 is about 700 million years. Rulers may come and go, but the nuclear material they amass will outlast generations of Middle Eastern leaders, from the beneficent to the brutal.

America began fulfilling the Shah's nuclear ambitions in the 1950s, under Eisenhower's tragically misnamed “Atoms for Peace” program. At the heart of the program was what seemed, for decades, to be a simple bargain: If a country agreed it would never seek nuclear weapons, the United States (and ultimately the IAEA) would help it develop its nuclear power industry. Only decades later—long after the Shah was gone—would the huge loopholes in that bargain become evident, as Iranian scientists, Syrian engineers, and North Korean “radio-chemists” developed skills that could be used for making energy or for making weapons. But at the time, supporting nuclear development in Iran seemed like a no-brainer.

By the mid-1960s, the Iranians were running a small research reactor in the center of Tehran—that was similar in size to the reactor the North Koreans built a few years later. In the 1970s, when the Shah still seemed a reliable if repressive friend in the Middle East, a fervent anticommunist and an insurance policy against the threat of Arab oil embargos, Washington agreed to take the next step. It signed a deal to sell the Shah upwards of eight nuclear reactors. (When the deal was coming together in 1976, Dick Cheney—who later became the chief proponent of using any means necessary to keep the Iranians from learning how to make nuclear fuel—was the White House chief of staff.)

The Shah made clear, though, that he wanted more than just reactors. Like a very different generation of Iranian leaders who would come to power decades later, he yearned to master the secrets of the nuclear fuel cycle so that he would never have to depend on a foreign power for the supply of uranium. At the time, the risk seemed theoretical and remote. Few seemed worried that once the equipment and training were in place, Iran could divert that fuel for nuclear weapons. After all, the Shah was a friend. When you troll through the public discussion of the deal at the time, few if any American officials were asking the obvious question that the Bush administration made a staple of its talking points decades later: Why would a country with huge oil reserves want to spend billions on nuclear power?

The Shah made little progress in realizing his dream before he fled the country in 1979. During the Iranian Revolution the program ground to a halt. Remarkably, Ayatollah Khomeini ignored it. Khomeini was far more interested in the purity of the country's Shia faith than in the purity of its uranium, and he viewed the project as “a suspicious Western innovation.”5 Early work at Bushehr, the site of Iran's first two nuclear plants, stopped dead, a setback that ultimately cost the Iranians decades.

Until Khomeini's death in 1989, the Iranian nuclear program was an on-again, off-again effort, pressed largely by the country's technocrats. Yet in 1985, during the depths of the Iran-Iraq War, the country began experimenting with gas-centrifuge technology and bought small facilities for converting raw uranium into a gas, the first step toward enriching the material. The motive was clear: nuclear parity with Saddam Hussein.

Saddam understood what the Iranians had in mind. Just as the Israelis had bombed Saddam's reactor at Osirak in 1981, Saddam bombed Iran's reactor site at Bushehr. He was not about to allow the mullahs to obtain nuclear fuel ahead of him. Toward the end of the Iran-Iraq War, Saddam used chemical weapons on the Iranians, an act of barbarism that convinced the Tehran that it needed a far greater deterrent than conventional weapons could offer. The Iranians began stockpiling chemical weapons of their own and told the country's Atomic Energy Organization that it was time to think about a nuclear option.

With Bushehr in ruins, their best hope was to learn how to produce uranium. The process was complicated, but it could be hidden. Many of the purchases, they knew, could be conveniently disguised as equipment for the oil industry. But mastering the centrifuge technology, the key to the uranium route to a bomb, seemed beyond Iran's capabilities at the time.

Their timing, however, was spectacular. Pakistan was years ahead of them and desperately needed money. The mullahs reached out to establish a formal relationship with the generals running Pakistan, one that resulted in a paper agreement in 1986 to cooperate on civilian nuclear energy.

Unfortunately for the Iranians, the accord resulted in little on-the-ground progress. Pakistan's prime minister, Zia al-Haq, was a devout Sunni Muslim and deeply suspicious of the Shia leadership in Tehran, particularly Ayatollah Khomeini.6 He told his aides to “play around” with the Iranians “but not to yield anything substantial at any cost.”7

It did not take the Iranians long to figure out that they were being strung along. But they had a Plan B: a separate deal with Abdul Qadeer Khan, the metallurgist behind the Pakistani bomb. The Iranians set up a meeting with Gotthard Lerch, a German who had been an early supplier to Khan, and Mohammed Farooq, an Indian who lived in Dubai and thrived in the city's regulation-free culture. If the authorities in Dubai, one of the seven principalities in the United Arab Emirates, were on to what was happening, they averted their eyes. After all, a lot of contraband moved through Dubai; that fact earned the city its reputation as the Singapore of the Persian Gulf.

In retrospect, the meeting with the Iranians was the true beginning of what became the Khan network. Iran's shopping list, as pieced together years later by international inspectors, contained all the elements that Khan would later ship to Libya and North Korea: drawings of centrifuges, a few prototypes of the same machines for the Iranians to reverse-engineer, and the layout for a full uranium enrichment plant.8 These were the building blocks that Iran would later use to construct its huge enrichment plant in the desert near the city of Natanz, supplied by small centrifuge-manufacturing workshops spread around the country. The process took years; it was not until the summer of 1994 that the Iranians arranged to buy a more sophisticated centrifuge, called the P-2, from Khan and his Malaysian cohort, Buhari Sayed abu Tahir, who ran Khan's operations in Dubai. Later that year, two officials of Iran's Revolutionary Guards showed up at the Dubai office with suitcases stuffed with $3 million in cash—a partial payment—and left with detailed plans for the P-2 and with parts that fit an older model.9 The Iranians were angry about the inferior equipment, but Khan's help jump-started their construction effort. Astoundingly, American intelligence agencies missed the operation for years; they still had no clue that Khan had turned from a buyer into a seller.

A formal nuclear cooperation agreement between Iran and Pakistan gave Khan the political cover to travel to Tehran. He toured the wreckage of Bushehr, which helped create the impression that he had come to aid the Iranians’ civilian nuclear efforts. In reality this was the first of several trips on which he could both act as a consultant and push his wares, pointing to the products that made possible the Pakistani bomb project. Along the way, Khan began to spin out his self-justification for the deals: He was helping another Muslim state break the American and Israeli stranglehold on nuclear technology. That played well in Tehran and around the Islamic world. In fact, his operation was mostly about money, and eventually money— and the Iranian realization they had been sold some old and unreliable technology—broke up the relationship.

It was around this time that the Iranians received—presumably from Khan himself, but perhaps from one of his deputies—a fifteen-page series of drawings and instructions that explained how to cast uranium metal into two hemispheres. For anyone familiar with nuclear weapons technology, the document rings alarm bells. Casting uranium into hemispheres is essential in building a nuclear weapon, but it has no utility in a civilian nuclear project. Years later, in 2005, Iran showed the document to international inspectors but forbade them to take a copy of it back to Vienna. (The Pakistanis have since confirmed that it matched one in their files.) The Iranians insist that they never sought the diagrams and never paid for them. They came, the Iranians argue, along with the prototypes and designs for the P-1, the first-generation centrifuge, sort of the way a car dealer might throw in a CD player. Later those drawings would become part of the circumstantial case that Iran was trying to design a weapon. But the key part of the transaction was Pakistan's P-1 centrifuges—the model for most of the centrifuges the Iranians installed at Natanz through the summer of 2008.1*

By the late 1980s, there were new leaders in Pakistan and Iran, but intelligence agencies in Washington and Europe still did not understand the depth of the connections between Islamabad and Tehran. There were plenty of hints. As tensions rose with Pakistan over American demands that it close down its own nuclear program, there were periodic threats from the Pakistanis that they would respond to the pressure by selling their growing nuclear know-how. The American embassy in Islamabad reported that the Pakistani Army chief of staff, Gen. Mirza Aslam Beg, had gone beyond open support of the Iranians. He had threatened American officials that he would sell nuclear technology to Tehran if the United States ever made good on its threat to cut off arms sales to Pakistan.10 Beg also pressed Benazir Bhutto, then prime minister, to strike a $4-billion deal with Iran to exchange nuclear technology for some mix of money and oil.

Bhutto said later she shot down the idea, and she professed ignorance of Khan's activities. “I find it very hard to comprehend that A. Q. Khan would have dared to do this,” she said in a conversation in London two years before she was assassinated in Pakistan, “because if I found out about this I would have sacked him, and he knew that I was that type of prime minister who would sack somebody if they breached the law.”11 Perhaps, but Khan also knew that Bhutto was weak, and lacked support in the army. He manipulated her and her successors for years to come.

As the Iranians gradually came to realize that the Pakistanis were selling outdated technology at sky-high prices, they began to cast a far wider net for alternative suppliers. They turned to the Chinese for new equipment to conduct laser enrichment of uranium at the Tehran Nuclear Research Center, the facility that had been set up with American help years before. President George H. W. Bush, far more concerned about Iraq than about Iran, said little to Beijing, even though, as the first American representative there, he knew most of the leadership personally. The Clinton administration was not much more vocal. The CIA was watching Iran's nuclear developments, and it was watching A. Q. Khan, suspecting that he was secretly importing nuclear components for the Pakistani nuclear weapons program. But America's spy agencies missed the crucial turn of events in the Khan empire: that he had flipped the switch, and begun to sell Pakistan's wares. It was a devastating failure to connect the dots.

“We saw Iran's activities, and we knew Khan was buying up supplies for Pakistan,” Gary Samore, who headed nonproliferation efforts for the Clinton White House, recalled in 2004, as the Khan network was unraveling. “But I don't think there was ever a moment where we saw Khan helping the Iranians.”

The Iranians said years later that they had no choice but to operate in great secrecy because the Americans were doing everything they could to choke the country off. On that point there is no dispute. Under American pressure, Germany turned down an Iranian request to complete the Bushehr project. The United States intervened with Argentina when it appeared that the country might help the Iranians learn how to produce fuel. Nonetheless, the Iranians claimed their share of victories. By greasing the wheels of commerce with petrodollars, they obtained high-strength materials and equipment they could plausibly claim were intended for oil production. In time, the Iranian supplier network stretched from Beijing to Berlin, from St. Petersburg to Istanbul.

It was not until 1997 that the Clinton administration was able to get the Chinese to cut off their assistance to Iran, including shipments of raw uranium. (The Chinese drove a typically hard bargain, consenting to stop the shipments only in return for an American agreement to export nuclear power reactors to China.)

Meanwhile, the Iranians brilliantly exploited Russia's simmering anger at being left behind as bankrupt, geopolitical roadkill after the Cold War. Nuclear technology ranked as one of Russia's few growth industries, and Moscow courted the Iranians. Soon they became partners in an effort to get the rusting reactors at Bushehr running. More surreptitiously, a deal was struck for missiles, that enabled the Iranians to move beyond their old North Korean Nodong missiles to something with far greater range that eventually became the Shahab-3. Broke and angry at the West, the Russian establishment responded in the late 1990s by delivering most of what the Iranians ordered.

“It seemed like every couple of months I was over in Russia, meeting Yeltsin's latest ministers—they changed every few months— and describing to them once again the programs that we had uncovered,” John McLaughlin, the former deputy director of the CIA, told me later.12

McLaughlin got little for his exertions except frequent-flier miles. The Iranians paid well, and in dollars. Over the next few years the Russians sped or slowed their deliveries to Iran depending on the political mood of the moment. When things were tense, they would claim that payment problems were holding up shipments. When the Iranians gained the upper hand, such as when the 2007 National Intelligence Estimate was published, the Russians unblocked delivery. With Bush's reluctant consent, the Russians delivered the nuclear fuel the Iranians so desperately needed. Bush did extract a concession from Vladimir Putin, however, that Russia would take back the spent fuel that resulted from its deliveries to ensure that Iran could not reprocess the spent fuel rods into plutonium for weapons, as the North Koreans had done so successfully.

Bush told me in 2007 that he believed containing Iran's nuclear ambitions was one of the few areas in which he and Putin could cooperate. “There aren't going to be many, so we should focus on this one,” he said, though none of us could have envisioned just how bad the relationship with Moscow would get as Bush prepared to leave office. The Russians, he argued, are as vulnerable to a nuclear Iran as the rest of Europe. But, he fumed, “We're up against economic interests, and it's very hard to get people to put those aside.”13

IN THE HOURS after the World Trade Center fell, Americans were in such shock that few paid attention to the reaction in Tehran. What they would have seen were candlelight vigils held for the victims, a reaction that was spontaneous, unanticipated, and symbolic of opportunities that would be missed as the United States sped into a series of ill-fated ventures in the Middle East.

The vigils took place across Iran. Before a soccer game in Tehran, there was a minute of silence. At Friday prayers, the ritual chant of “Death to America” was skipped—not exactly what one might call a warm gesture of friendship, but a welcome departure from the norm.14

After more than two decades of hostile relations between Washington and Tehran, this fleeting moment offered at least the hope of some common ground. Instead, the opportunity was lost in the administration's zeal to divide the world into two camps—Bush's version of geography in which countries had to declare that they were with us or against us. Eventually the president learned that the world doesn't operate that way. By that time it was too late to take advantage of the chance to change the relationship with Tehran.

At the Pentagon and the State Department, where preparations began immediately after 9/11 to go to war in Afghanistan, some senior officials quickly recognized that Iran could be a natural ally of convenience in that fight. For years the Iranians had heavily supported the Northern Alliance, the band of Afghan rebels whose central role in taking down the Taliban—with American support—has been underappreciated in the popular narrative of the war. Iran's motive in that struggle was far from altruistic. The Taliban did everything they could to repress Afghanistan's Shia minority. Just three years before 9/11, Iran and Afghanistan came to the brink of conflict over the murder of Iranian diplomats—and many Afghan Shia—in Mazar-e Sharif.15

Not since 1979, when diplomatic relations between Washington and Tehran were terminated amid the hostage crisis, had the two countries found themselves on the same side of a common fight. The newly espoused Bush Doctrine called on governments around the world to declare whether they were “with us or against us.” Many wavered. Remarkably, for this one brief moment, the Iranian military chose to declare that they were with us, even though they could not talk to us, at least directly.

In discussions that took place through intermediaries, Tehran offered to allow U.S. pilots to land in Iran if they ran into emergencies while flying sorties over Afghanistan. After the Taliban's fall, the Iranians showed up at the conference in Bonn that created a framework for the new Afghan government and proved to be marginally helpful. They held a series of secret meetings with Zalmay Khalilzad and Ryan Crocker—a duo that teamed the administration's only Afghan-born neocon with one of its most respected career diplomats. These meetings marked the first time in recent memory that Iran and the United States broached the subject of dealing with terrorism. If not a breakthrough, at least it was a historic opportunity.

It is hard to say who was more suspicious about the utility of talking to the enemy—the mullahs or Dick Cheney. The mullahs tended to move at a slow diplomatic pace and saw an American conspiracy behind many events in which Americans had no role. Cheney and his staff, along with many neoconservatives, wanted to move at warp speed in bringing the Iranian regime to its knees. Negotiations would serve simply to prolong the regime's survival. In Tehran, the Iranian government was, as always, deeply divided about how to deal with the “Great Satan.”

“The intelligence out of Iran suggested a huge fight,” one administration official said to me, “that sounded a lot like our fight.”

Astoundingly, Bush never tested the possibilities. This was his moment to do what wartime leaders are supposed to do: Divide your enemies. FDR seized his moment to form an alliance of convenience with Stalin against Germany. The United States sided with the Mujahideen to expel the Soviets from Afghanistan. The list goes on. This was Bush's chance to exploit the longtime divisions between the Persian Shia who saw themselves as the legitimate ruling power in the Middle East and the Arab Sunni extremists who had planned and executed the attack on the United States. But for a White House that was already well on its way to rallying the country against what the president, several years later, called “Islamofascism,” those distinctions seemed more obfuscating than compelling.

Bush's instinct during his first term was to lump all of America's “enemies” in one camp, from the 9/11 plotters (but not the country they came from, Saudi Arabia), to Taliban remnants, Shia militias, and everyone else who wished America ill. He ignored the natural fault lines that might have enabled him to do what Roosevelt had done.16 Why? It is a question I asked in interview after interview, year after year, usually to get pabulum about engaging the American public in a broad “war against terrorism.” It was only late in the administration that senior officials began to acknowledge that the “GWOT,” as the Pentagon called the Global War on Terror, was the wrong phrase—it encouraged generals and politicians to fight a tactic without addressing its causes.

Within three months the administration dug a deeper hole, when it let rhetoric drive American foreign policy, instead of the other way around. In writing his State of the Union address in January 2002, Bush's speechwriters, Michael Gerson and David Frum, were searching for a phrase that would capture this new era that pitted America against a world of shady actors with nuclear ambitions. They settled on “axis of evil” to describe Iran, Iraq, and North Korea. Many senior foreign policy officials in the administration told me in interviews that they never even saw the language before Bush gave the speech.2*

The president, his fist clenched as he stood at the podium the night of January 29, 2004, delivered the lines with gusto. The three Axis members, he said, were “arming to threaten the peace of our world,” amassing weapons of mass destruction and harboring terrorists.

“We will be deliberate, yet time is not on our side,” he said, foreshadowing what became, months later, the key element of the preemption policy in his national security strategy. “I will not wait on events, while dangers gather. I will not stand by, as peril draws closer and closer. The United States of America will not permit the world's most dangerous regimes to threaten us with the world's most destructive weapons.”

“States like these and their terrorist allies,” he concluded, “constitute an axis of evil.”17

The “axis” phrase evoked the imagery of the “Axis powers” of Germany, Italy, and Japan during World War II, suggesting a similar kind of alliance among Iran, Iraq, and North Korea.

Bush particularly singled out Iran, saying that it “aggressively pursues these weapons and exports terror, while an unelected few repress the Iranian people's hope for freedom.”18 His phraseology, I wrote in the Times the next morning, “seemed to be outlining a rationalefor future action, if he deems it necessary, not only against terrorists but against any hostile states developing weapons of mass destruction.”19

Colin Powell hated the rhetoric. As his biographer Karen De-Young wrote later, “It reminded him of Ronald Reagan's ‘evil empire,’ a phrase much beloved by neoconservatives but one he considered unnecessarily provocative and relatively meaningless in the context of the pragmatism he knew had marked U.S.-Soviet relations in the final years of Reagan's presidency.”20

In fact, a war was under way between Powell's forces and the White House over how to think about Iran. The day after the speech, a White House official told me that “there are people in the State Department who want to think Iran is changing because everyone's drinking Coca-Cola, but the evidence isn't there.”21

Publicly, though, the White House quickly backed away from the notion that the three Axis nations were colluding. The day after the speech, the White House spokesman, Ari Fleischer, was sent out to declare that the “axis” line was “more rhetorical than historical.” In fact, earlier drafts of the speech had gone even further by describing links among the three nations, until someone pointed out that, for example, North Korea had more exchanges of missile technology with Pakistan—America's newest ally in the war on terrorism—than with Iran.

“So we pared back to the essential warning,” one official told me, “if you develop these weapons, and if you mess with terrorists, sooner or later we will make you regret it.”22

Even years later, it is difficult to explain fully what led Bush off this rhetorical cliff. Some of his aides told me at the time that although he was impressed with Iran's offers of support after 9/11, his mind was changed a few weeks later by the Israeli seizure of a ship, the Karine A, that was delivering smuggled arms from an Iranian port to Palestinian forces. That convinced Bush “that the Iranians weren't serious,” one of the aides said. (Others in the White House could barely recall the seizure in the blitz of events that followed.)

In the supercharged mix of emotions and ideological fervor inside the White House after the attacks, there was a conviction that any language vilifying the hardline Iranian leadership would empower the country's young people—the same young people who were lighting those candles after 9/11.

Instead, the words gave the Iranian leadership a reason to crush the moderates. One Iranian reformer told my colleague Michael Slackman, “When Bush named Iran as the Axis of Evil, the hardliners became happy. … They can then mobilize the part of the country that supports them.”23

Days later, at the Davos World Economic Forum, which had been moved to Manhattan as a sign of post-9/11 solidarity, Powell was taken aside by one world leader after another, all of whom said they had cringed at the president's words. Bill Clinton, who had held forth until two in the morning at an after-dinner session, warned that it was unwise to treat the “Axis” as a single entity.

“We have to take these countries each in turn,” he cautioned. “They may all be trouble, but they are different.” Iran had two governments, he said—one with progressive elements, one with hardliners—and he thought North Korea was ready to make a deal when he left office.24 Bush brushed off the criticisms as the hand-wringing of those who were not thinking globally about the threats facing America, and who did not have to stay awake at night devising strategies to deter a second attack. Time altered his view. In his far more pragmatic second term, Bush never uttered the “axis of evil” phrase. By the end of his time in office, he was doing exactly what Powell had urged: sending out probes to find out if the Iranians were serious. By that time, though, it was too late.

DAYS AFTER THE “Axis of Evil” speech was delivered, Colin Powell sent word to the Iranians that they could relax a little: It was Saddam Hussein who was in Bush's sights, not the mullahs. “With respect to Iran and with respect to North Korea, there is no plan to start a war with these nations,” he said. Iraq was a different story, he said, and regime change “would be in the best interests of the region, the best interests of the Iraqi people. And we are looking at a variety of options that would bring that about.”

Among Powell's small cluster of advisers, many hoped the Iranians would get the message: Forget what Bush says, just watch what he does. To Tehran, the advantages of having the Americans remove Saddam Hussein, Iran's mortal enemy, were immediately clear. But the mullahs were trapped in their own anti-American rhetoric, just as Bush was trapped by his “Axis of Evil” speech.

Then, in August 2002, an Iranian resistance group publicized the existence of a covert uranium enrichment site in Natanz. “We had watched it for years,” one senior intelligence official told me. “But until it was publicly revealed, we had a hard time getting people interested.” The American press didn't pick up the story until December 2002, just a few months after similar uranium-enrichment efforts in North Korea had been disclosed. The revelation of the existence of Natanz led the IAEA to realize they had been duped by the Iranians, who had failed to declare the existence of the facilities. Naturally, Iranian officials claimed that the site was part of a peaceful nuclear program, though if it was, why not declare it to the international inspectors, who were already looking at other parts of Iran's nuclear energy program? As Colin Powell said, once caught, the Iranians protested with vigor but never really explained the underlying economics of their program. “We've always found it curious,” he said, “as to why Iran would need nuclear power when they are so blessed with other means of generating electricity.”25

SOON A NEW IRANIAN ambassador was dispatched to the United Nations: Javad Zarif, whose perfect English, hip attire, and diplomatic pragmatism had caught the attention of American officials during the Afghan talks in Bonn (like Condoleezza Rice, he was a graduate of the University of Denver). Zarif made it clear from the start that he sided with those in Iran who sought a “grand bargain” with Washington—a faction in Tehran whose ideas, unfortunately, have always exceeded their influence. A series of meetings in early 2003 between Zarif, Khalilzad, and Crocker gave the Iranians their first understanding of American intentions in Iraq—and how they might benefit.

The Iranians were not the only ones thinking about the consequences of an American invasion of Baghdad. The CIA, the same agency that was getting everything wrong about Saddam's weapons, had it right about the benefits to Iran if Bush ordered troops into Iraq. In October 2002, it issued a report, “Iran Wary of a U.S. Attack on Iraq,” that laid out the options for Bush.

“The more that Iranian leaders—reformists and hardliners alike—perceived that Washington's aims in Iraq did not challenge Tehran's interests or threaten Iran directly, the better the chance they would cooperate in the postwar period—or at least not actively undermine U.S. goals,” the intelligence report read. It argued for “guaranteeing Iran a role in the negotiations on the fate of post-Saddam Iraq—as it had at the Bonn conference for Afghanistan.” Those negotiations, it continued, “might give Tehran a stake in its success.”26

The CIA report also warned that “some elements in the Iranian government could decide to try to counter aggressively the U.S. presence in Iraq or challenge U.S. goals following the fall of Saddam” by sowing dissent among Shia and Kurds. The message was clear: The best chance of avoiding trouble was to talk to the Iranians early and often.

But Bush was not in a mood to talk to the Iranians about Iraq— not until 2007, when American forces were in such trouble, partly because of Iran's covert assistance to the insurgents, that the Iranians had far more influence in the southern part of Iraq than the United States did. The moment to talk, as it turned out, arrived during the same week that the United States seemed at the pinnacle of its power in the Middle East. The week that Bush, a former Air National Guard pilot, donned a flight suit and—with an expert pilot at his elbow—landed a small jet on the deck of the USS Abraham Lincoln.

It was May 1, 2003, and that evening Bush delivered what became the most widely mocked speech of his presidency—the “Mission Accomplished” address to the nation. Those words never appeared in his text; they were on the banner behind him. But on that crisp California night off the coast of San Diego, I didn't run into a single member of the administration who objected to the giant banner hanging behind the president, emblazoned with the two words Bush would spend the rest of his presidency trying to disown. “Because of you,” Bush declared, “our nation is more secure. Because of you, the tyrant has fallen, and Iraq is free.”

Even the truest believers in the Bush White House identify that week as the beginning of a dangerous era of triumphalism. There are few other convincing explanations for the administration's decision to ignore, that same week, what might have been the best opportunity in years to start remaking the U.S. relationship with Iran—and with it, America's role in the Middle East.

The opportunity was contained in a long missive, apparently from the Iranian government, that showed up on a fax machine at the State Department. It quickly became known as “the offer,” a proposal to build what amounted to an opportunistic alliance of convenience.

The fax came with a cover letter from the Swiss ambassador to Iran, Tim Guldimann, who was responsible for representing American government interests in Tehran in the absence of a formal diplomatic relationship between the United States and Iran. The offer itself had been edited by Iranian diplomats—chiefly Zarif— and had supposedly been seen by both Iran's Supreme Leader, Ayatollah Khamenei, and the weakened sometime reformist who was serving as president, Mohammad Khatami.27

Whether those two leaders—whose own relationship was badly strained—truly stood behind the offer is impossible to know. But the words were beyond dispute. The document began by ticking off a list of “Iranian aims,” including the perennial “Halt in U.S. hostile behavior,” with the “Axis of Evil” speech noted as an example. The offer proceeded to call for “full access to peaceful nuclear technology, biotechnology and chemical technology,” and the “pursuit of anti-Iranian terrorists.” Under “U.S. aims,” the first item listed was “full transparency for security that there are not Iranian endeavors to develop or possess WMD,” and “decisive action against any terrorists (above all al Qaida).” Iranian help in establishing democratic institutions in Iraq was also listed, though Iran seemed like an unusual choice as an exporter of democratic institutions. Then it called for talks, working groups, and road maps.28 It was vague, but so was Bush's other “road map,” for a grand bargain between the Israelis and the Palestinians.

Like the Israeli-Palestinian peace process, though, this offer seemed to envision a fundamental shift in the relationship. If it was for real—a big if—it had the potential to be a game-changer. The problem for the administration—even for those in the State Department who had been pushing for just such an approach—was that no one was clear about the offer's origins, authorship, or sincerity. Some suspected that it was nothing more than a Swiss attempt to get talks started under the guise of an Iranian offer. Others thought it could be an unauthorized probe sent out by the moderates in Iran surrounding President Khatami. (“How many times,” Rice asked one of her subordinates in exasperation, “have we pursued the elusive Iranian moderates,” only to discover they had no power?)29 A few hardliners in the administration saw the offer as the confirmation of their wisdom in invading Iraq. The mullahs, they figured, had tuned in to CNN in time to see the video of Saddam's statue being pulled down and feared that unless they did something, downtown Tehran could be the next stop on the preemption parade.

“There were a lot of questions,” said Richard Haass, who was running Iran policy for the State Department at the time. “But I thought it was worth testing it, to find out whether it was real. That was the only way we were going to find out.”30 Haass described to Powell the advantages of opening talks with the Iranians, and even dangling some incentives, including dropping American objections to Iran's entry into the World Trade Organization, a longtime Iranian goal. “I didn't see what we had to lose,” Haass said later. “I did not share the assessment of many in the administration that the Iranian regime was on the brink.”31

The hawks—not only Cheney, but many in his office—convinced themselves that the mullahs were just one step from the cliff's edge, and that quick American success in Iraq could push them into the abyss. To them, the Iranian offer sniffed of weakness and fear. For that reason, they immediately and vigorously rejected it. John Bolton, the Cheney protégé who was undersecretary of state for arms control and proliferation at the time, described the offer as “a fantasy.”

“Time is the only thing they can't purchase with their oil revenue, but they can get time if they can dupe Europeans or Americans into negotiations,” he said later.32 The speed with which the Iraqi Republican Guard was vanquished by American forces stoked the hawks’ confidence that the United States should not negotiate. They argued that the United States would have even more clout once the Middle East was transformed by a wave of liberal democratic reforms. In retrospect, that was the true fantasy. When the war in Iraq began going badly, America's leverage was lost. When oil tripled in price, Iran was suddenly in the driver's seat again. The rest of the Bush years were spent trying to regain some leverage over Iran—without success.

Nonetheless, talks were already under way. Zarif was meeting secretly with Americans, including Khalilzad, about the possibilities of exchanging intelligence about al Qaeda and Mujahideen-e Khalq, or MEK, a rebel group that has waged a violent struggle against the clerical regime in Iran since 1981. The MEK is Iran's largest opposition group, and for years it operated out of Iraq with Saddam Hussein's support. When America invaded Iraq, Iran demanded that American forces treat the MEK as enemy combatants. Instead, U.S. soldiers disarmed members of the MEK but did not arrest them.

As the offer was being faxed to the State Department, the Iranians proposed a swap: MEK for al Qaeda. If the Americans would turn over the MEK members, the Iranians would hand over to Washington the al Qaeda leaders who had escaped into Iran and were now under house arrest. In short, the Iranians were proposing setting up a kind of terrorist bazaar. It was unseemly, but given that at the time the United States was snatching terror suspects out of towns and cities throughout the world and dropping them into a netherworld of secret detention camps, stranger things were happening.

Khalilzad was not ready to bargain on terrorist exchanges. However, he did warn the Iranians about a suspected terrorist plot in the Persian Gulf area and pressed the Iranians to interrogate the al Qaeda members in their midst and provide the information needed to preempt the attack. The Iranians never took the warning very seriously.

Then, on May 12, 2003, just eleven days after the “Mission Accomplished” speech, four bombs went off in Western housing compounds in Riyadh, Saudi Arabia, just ahead of a visit to the country by Powell. Nine Americans were killed and hundreds of people were injured. The Bush administration was convinced that the al Qaeda members in Iran had been in close contact with the perpetrators of the plot. Enraged at the intransigence—or at least the indifference— of the Iranians, the administration promptly cut off any negotiations. Zarif arrived in Geneva twelve days later for another meeting with Khalilzad, but Khalilzad never showed up. Any discussion about a “grand bargain” was dead. The relationship quickly reverted to its normal state of open hostility.

WITH NO NEGOTIATIONS under way, the Bush administration found itself facing a familiar dead end. Like the North Korean government, the Iranian regime had stubbornly refused to collapse. If there was an American strategy for guiding the next move, it was a mystery to Washington's allies. Bush seemed to be passing the buck to the Europeans (including the French, who had fervently opposed the Iraq War), hoping they would deal with Iran while the United States cleaned up in Iraq.

For a while it looked as if the Europeans might make progress. In September 2003, the IAEA board of governors, in a rare moment of unity, demanded that the Iranians fully disclose their nuclear program and resolve outstanding legal issues. The resolution set a deadline of October 31, touching off a flurry of vigorous and belligerent Iranian rebuttals.33 Days after the resolution was announced, the head cleric of Iran's Guardian Council questioned publicly why Iran should stay in the Nuclear Non-Proliferation Treaty.34

By late October, it was clear that the Iranians would require some convincing if they were to come clean. With the United States notably absent, the foreign ministers of France, Germany, and the UK flew to Tehran desperately seeking to avoid the looming showdown between the IAEA and Iran. What emerged was a hopeful-sounding agreement in which Iran agreed to cooperate with the IAEA and voluntarily suspend its uranium enrichment. In return the EU offer hinted at the possibility of future nuclear technology assistance and affirmed Iran's right to civilian nuclear energy.35 The United States stayed silent; it refused to endorse the European effort.

Two months later it looked as if the inspections might get someplace, when Iran signed an agreement with the IAEA that permitted, at least on paper, inspections of any facilities that international inspectors demanded to see. By November 2004, Iran and the Europeans signed an accord, called the Paris Agreement, in which the Iranians agreed to suspend all their uranium-enrichment activities while negotiations continued. Again the Americans stayed silent.

But within weeks of Bush's second inaugural—where he called for America to make the spread of freedom around the world its number-one goal—Rice quickly discovered that Bush's stony silence was untenable. On her first trip to Europe as secretary of state, “everyone beat the hell out of her” about Iran, one of her traveling companions said. It was clear that the Europeans, still angry that Bush had rolled over their objections to invading Iraq, believed he was headed to another disastrous confrontation.

It was the first test of whatever explicit or implicit understanding Rice had cut with Bush at Camp David when he first broached the idea of removing Powell and putting her in his place. Rice's failings as national security adviser were obvious to everyone. She never had the clout to oppose Cheney and Rumsfeld openly, and she seemed to view her job as staffing Bush and measuring his mood, rather than shaping the debate. The results had been disastrous. Now she seemed willing to conduct end runs around Cheney and Rumsfeld in a way she never managed during her time as national security adviser.

“Something changed in Condi,” said one of her top aides who followed her to the State Department. “I think she knew that her time at the White House was a failure because, for whatever reason, she was constantly measuring the wind direction rather than guiding the policy. At the White House she felt like staff. Now she felt like she had the power to make changes.”

But the changes were incremental. She embraced the European negotiating effort. She dropped the American objections to Iran's application to enter the World Trade Organization. Yet she was unwilling to make the one change that might have made a difference: allowing American diplomats to join the talks with the Iranians.

“This is our way of making clear that we will join the Europeans in giving Iran positive reasons to give up its program,” one administration official said to me.36

When Bush traveled to Europe for a summit meeting, he made a deal with his European counterparts: He would support their negotiations, but in return Britain, France, and Germany had to agree to join an escalating series of sanctions at the Security Council if Iran balked. The Europeans feared they had been to this movie before—just eighteen months earlier, when Bush used Security Council resolutions demanding that Iraq disarm as “authorization” for the American invasion. Some European leaders feared being sucked into a deal that, eventually, would give Bush an excuse for military action. But with Bush increasingly pinned down in Iraq, the risk seemed low.

To the Iranians, of course, negotiations were a way to buy time. The Iranian regime was consumed with bitter infighting, but the leadership understood two things. Bush had his hands full in Iraq— a problem the Iranian Quds Force knew how to make worse—and the United States could not afford to get involved in yet another Middle East war.

IN THE SUMMER OF 2004, almost exactly twenty-five years after the meeting in Algeria and the hostage crisis that immediately followed, Gates and Brzezinski found themselves thrown together again with the same mission: to find a way to deal with Iran.

This time there were no awkward banquets in Algiers. Bazargan, the prime minister they had dealt with, had died nearly a decade before, banished from power since his resignation in the opening days of the hostage crisis. Chamran, the former aide to Ayatollah Khomeini and defense minister, had been killed during the Iran-Iraq War by an Iraqi mortar. Only Yazdi survived, as a dissident seeking a more open and democratic Iran. He had gotten nowhere.

In Washington, Brzezinski and Gates were trying to save the Bush administration from itself. They directed a joint study at the Council on Foreign Relations to explore how to move forward with Iran. There was a lot of catching up to do. There had been virtually no direct, official talks between the two countries for a quarter century, save the clumsy deals struck during the Reagan administration that became the Iran-Contra scandal.

The idea was to steer the government out of a dead end—gently, without overt criticism. Gates, relishing his time away from the politics of Washington, was president of Texas A&M. Brzezinski was comfortably ensconced as a counselor at the Center for Strategic and International Studies in Washington.

Reports such as theirs are published in Washington every week. They usually circulate among a few hundred influential policymakers and occasionally get a few paragraphs deep in the International section of the Times. But given Gates's and Brzezinski's stature, this one had the potential to be different.

They started by bursting the hardliners’ balloon. The fall of the current regime in Iran, they said, was not going to happen anytime soon. America's clear-cut interests in Iran's behavior necessitated direct engagement—working through the Europeans wouldn't suffice. As a member of the panel told me, “There are some things in life that don't work when you have other people do them for you. Among them are sex, drinking, and negotiating with Iran.”

But Gates and Brzezinski argued that an all-encompassing grand bargain would not work, at least not yet. They pushed for “selective engagement,” a mix of economic inducements and threatened penalties that could bring serious progress on key issues, starting with the nuclear program and Iran's support for Hamas, Hezbollah, and other terror groups.

The central recommendation was that the United States not waste time. The leverage that America had gained by invading Iraq and Afghanistan was a dwindling asset, they argued. Moreover, it was their sense that the internal rifts in Iran could be exploited.

“Most revolutions—the Bolshevik Revolution, the Chinese Revolution—most revolutions tend to get more moderate as time goes along, and leaders sort of deal with the real world in a more realistic way,” Gates said when we talked about the report he issued. When Mao saw that happening to his revolution, Gates noted, he launched the Cultural Revolution to bring back the fervor—but that just delayed the inevitable. “It looked to me like the same thing was happening in Iran with Khatami and Rafsanjani and all these guys.”

Gates's and Brzezinski's recommendations were immediately put in the back of a file drawer by the Bush White House. The administration was hoping that as the Iranians headed into elections in 2005, the combination of negotiations and escalating sanctions would lead Iran's younger generation to decide that the nuclear program was not worth the pain.

They did not anticipate Iran's version of the Cultural Revolution: the rise of the mayor of Tehran, Mahmoud Ahmadinejad, a man whose rhetoric outstripped his competence.

Ahmadinejad's election took everyone by surprise—Bush, Rice, Hadley, and the intelligence agencies. He was a populist with little base in national politics and a predilection for talking in apocalyptic terms about Israel disappearing from the earth. He often rambled, and had never been considered a serious player, but rather more of a fringe element. “Nobody really thought he was going to win, but it is not as if exit polling is something that we understand well, especially in Iran,” Rice told me later.37

Ahmadinejad turned out to be a skillful politician who knew in his gut that what Iranians cared about most was the restoration of their country to the position of the greatest power in the Middle East. The Bomb—or at least the debate over giving in to the West's demands—was the way to exploit that emotion. Iranians did not like sanctions, and they did not like isolation. But as long as the situation could be cast as another effort by the Americans to deny the Iranians their rightful place in the world, Ahmadinejad would enjoy support. The more he baited Bush, and the more Bush went for the bait, the better off he would be. It wasn't much of a strategy, but it was better than anything the Bush administration had yet devised.

1* The P-1 is an old and inefficient design. The Iranians quickly set to work on the next generation, making their own modifications and calling it the IR-2 to designate it as Iranian-made. By early summer 2008, they were installing IR-2s at Natanz and told inspectors that they expected to focus on manufacturing the more modern version, which can enrich significantly more uranium in the same amount of time as the original.

2* The speech was closely held, and early drafts did not contain the “axis of evil” phrase, or its exact context. Colin Powell, among others, has said that he did not see a final draft.