The coup had failed.

That was the situation in Tehran on the morning of August 16, 1953.

The Iranian officer sent to remove Prime Minister Mohammad Mosaddegh had been arrested. The Shah had fled the country. Mosaddegh was still in power.

Months of covert planning had collapsed in a single night.

In Washington there was every reason to assume the operation was finished. Even the president thought so afterward. In his diary that October, Dwight Eisenhower wrote that in the first hours of the attempted coup all element of surprise had disappeared through betrayal, the Shah had fled to Baghdad, and Mosaddegh seemed more firmly entrenched than ever before.

Kermit Roosevelt, the CIA officer running the operation inside Iran, stayed.

Three days later Mosaddegh’s government was gone. The Shah was coming home.

Eisenhower would write in that same diary entry that the restoration of the Shah and the elimination of Mosaddegh were things America had helped bring about, that what the United States did was covert, and that public knowledge of it would be deeply embarrassing. He listened to Roosevelt’s report and said it seemed more like a dime novel than historical fact.

One warning before we go further, because it shapes everything you have probably heard about this.

Roosevelt published a memoir in 1979 called Countercoup. It is a terrific read. It is also substantially unreliable, contradicted by the declassified record, and still the main reason popular imagination treats August 1953 as one American spy single-handedly toppling a government.

So we are not going to use it. We don’t need to.

Because this is not one of those episodes where historians infer what the CIA might have done from circumstantial evidence.

We have the documents.

It took a while. The State Department’s own Iran volume, published in 1989, omitted the CIA’s role entirely

— the New York Times called it Hamlet without the Prince. The corrected volume finally appeared in 2017, running past a thousand pages of declassified State, CIA and National Security Council records.

Britain and the United States really did run a covert operation to remove the government of Iran. The interesting question is how they got there.

Because America had not always been willing.

Britain Had Already Made Up Its Mind

By late 1952 the oil crisis from Part I had reached an impasse.

Iran had nationalized its petroleum industry. Britain had failed to reverse that through negotiation. The

embargo had badly hurt Iran, and Mosaddegh was still there. The International Court of Justice had declined Britain’s case.

London was running out of peaceful ways to get what it wanted. So it proposed something else.

On November 26, 1952, Assistant Secretary of State Henry Byroade wrote a memo to his superior with a subject line that required no interpretation:

Proposal to Organize a Coup d’état in Iran.

The British Foreign Office, Byroade wrote, had informed Washington it was disposed to attempt a coup replacing the Mosaddegh government with one that would be more “reliable” — if the American

government agreed to cooperate. British and American intelligence agencies had already held very tentative and preliminary discussions but could not go further unless the State Department was prepared to consider it as a matter of policy. The intelligence people wanted a definite answer.

Three things about that memo matter.

First, Britain proposed the coup before Dwight Eisenhower became president. The idea had first surfaced in a British paper handed to the Americans on October 8, 1952, and by late November several meetings had produced no encouraging conclusion — the lack of a viable replacement for Mosaddegh and the risk of a Tudeh counter-move were among the reasons.

Second, Byroade wasn’t selling it. He listed his own doubts. He recommended only that his boss hear the British out.

Third, a week later the British minister told the Americans that London had not yet come to any definite conclusions either.

This is where the simplified version breaks down.

It’s tempting to tell this as one uninterrupted Anglo-American plan running from nationalization to the coup.

It wasn’t. The Truman administration had real concerns about Iran — communist expansion, the Tudeh Party, an unresolved oil dispute. But American officials also repeatedly regarded British policy as

dangerously inflexible, and they did not agree to help remove the government. Then, in January 1953, the administration changed.

Truman left. Eisenhower arrived. John Foster Dulles became Secretary of State. His brother Allen became Director of Central Intelligence.

And here is the detail almost nobody knows, which explains more than the brothers do.

In November 1952, just after Eisenhower’s election, MI6’s Tehran station chief C.M. Woodhouse met with CIA and State Department officials. According to the CIA’s own internal history, Woodhouse emphasized Iran’s vulnerability to communist subversion and “did not stress the oil issue” — after which Allen Dulles showed increasing interest in a covert operation to remove Mosaddegh.

Read that again.

The British had an oil problem. They knew the Americans didn’t care much about their oil problem. So they repackaged it as a communism problem.

And it worked.

That is not a conspiracy theory. That is a British intelligence officer choosing his sales pitch, written down in an American intelligence history.

“Mossadeq Must Go”

By spring 1953 the language inside CIA planning had changed completely.

An estimate drafted by Donald Wilber and circulated to Kermit Roosevelt opens with a numbered list of basic assumptions. Number one is three words long.

Mossadeq must go.

Number two: appropriate U.S. covert assets will be directed toward his overthrow, and U.S. policy action and financial aid will support his successor. That successor was General Fazlollah Zahedi.

No need to embellish. The CIA wrote it down and numbered it. Keep reading the list, though, because it gets worse.

Assumption six anticipates that failure of an attempted forced change would lead to violent Iranian reaction against the United States. They knew what they were risking.

Assumption seven states that Zahedi can last only if he manages the immediate removal of all dissident leaders. The repression wasn’t an unfortunate later development. It was in the plan, on paper, before the coup.

Assumption eleven states that Kashani’s influence with the Zahedi government must be restricted or neutralized — meaning the CIA intended to use the cleric and then discard him.

And assumptions eight and nine concern oil. Zahedi is described as anxious to settle the issue. A new agreement must at least appear more generous than any previous one. And the timing matters, because it should not be announced immediately after Zahedi takes office.

Hold that last one. We’ll come back to it.

But the same document contains something the popular story leaves out entirely.

The agency did not believe it could do this by itself. The estimate states plainly that agency assets in Iran were not by themselves capable of overthrowing Mosaddegh’s government — that Zahedi’s chances would be greatly enhanced only if he were backed by those assets and by money.

It also identified the assumption the planners called most vital of all: that a Zahedi government could neither establish itself nor stay in power without the active support and moral leadership of the Shah.

So the CIA’s own view was this. We cannot do it alone.

We need a general. We need the monarch. We need Iranians.

That distinction matters, because reducing August 1953 to “the CIA overthrew Iranian democracy” gives the agency magical powers and turns Iranians into extras in their own history.

The truth is more uncomfortable than that.

Foreign intelligence services walked into a real Iranian political crisis and deliberately tried to make one side win.

The Crisis Was Real

Here is the part that gets skipped by people who like Mosaddegh, and I like Mosaddegh.

By summer 1953 his coalition was disintegrating.

Ayatollah Abol-Ghasem Kashani — whose religious networks had been essential to nationalization — had broken with him. So had significant parts of the National Front. Mosaddegh was governing under

emergency powers granted by a parliament he increasingly could not control.

In August 1953 he held a referendum to dissolve the Majlis. The voting was conducted with separate ballot boxes for yes and no. It returned margins in the range of ninety-nine percent.

Whatever you think of the man, that is not a constitutional procedure. It was the single most damaging thing he did to his own legitimacy, and he did it to himself.

Meanwhile the clerical establishment was moving. Zahedi told the CIA that Grand Ayatollah Borujerdi, Kashani and Ayatollah Behbahani were reaching an understanding on the need to bolster the Shah against Mosaddegh — and that if Borujerdi called for active support of the Shah, his associates could assemble a very large force of demonstrators.

The CIA didn’t manufacture the split with Kashani. It didn’t write the referendum. It didn’t invent the landowners, royalists, officers, merchants and clerics who wanted Mosaddegh gone.

Iran was already coming apart.

The operation tried to use that. Propaganda works best when it can attach itself to something real.

Build the Environment

TPAJAX was not a plan to send soldiers to Mosaddegh’s house. It was an attempt to reshape the political environment around him.

The CIA had newspapers. Political contacts. Military contacts. Clerical contacts. Money. And people who could produce demonstrations and spread rumors.

The agency’s own after-action history describes what it did: CIA and British propaganda assets were to

conduct an increasingly intensified effort through the press, handbills and the Tehran clergy, in a campaign designed to weaken the Mosaddegh government. The stated aim of the project was to cause the fall of that government and to reestablish the prestige and power of the Shah.

The campaign portrayed Mosaddegh as collaborating with communists, abusing constitutional authority and chasing personal power.

There was also what the agency called a war of nerves. American officials suggested publicly that further

U.S. aid was not coming. Critical American press coverage reinforced it. Even the absence of the American ambassador from Tehran was used to signal that Washington had lost confidence.

Sit with what that actually is.

The operation wasn’t only: remove the leader.

It was: change what people believe about the leader. Make his rule look economically hopeless. Make allies doubt him. Encourage the clergy. Encourage the officers. Convince people he is drifting toward

communism. Convince people the Shah still represents legitimate authority.

Then create conditions in which removing him looks less like overthrowing a nationalist prime minister and more like rescuing Iran from one.

The First Coup

There was still the problem of the Shah.

Mohammad Reza Pahlavi was not eager to gamble his throne on an operation that might fail, and failure meant exile or worse. Getting him to sign the decrees dismissing Mosaddegh and appointing Zahedi took sustained pressure.

It is worth knowing how he understood his own position. In May 1953, associates reported to Ambassador Loy Henderson that the Shah had said the British threw out the Qajar dynasty, brought in his father, and threw out his father. He wanted to be told whether London wished him to remain or to leave quietly.

Churchill eventually sent word that Britain did not wish to see him lose his position.

That is the monarch of a sovereign country asking a foreign government whether he may keep his throne. He signed.

Colonel Nematollah Nassiri, commander of the Imperial Guard, would deliver the dismissal order. Zahedi would emerge as prime minister. The military would back the transition.

Then Mosaddegh found out.

The CIA’s own military situation report on the night of August 15–16 reads like a post-mortem. The agency had been cut out of the military preparations. Zahedi brought another general into the final plan, which the report calls possibly fatal. Mosaddegh’s chief of staff got seven hours of warning and used it. When Nassiri went into action, it was his own officers who were arrested.

Nassiri was taken into custody at Mosaddegh’s residence. Other arrests followed. The government announced it had defeated an attempted coup.

The Shah fled. First to Baghdad, then to Rome.

Catastrophic. The man whose authority was supposed to legitimize the whole thing had left the country. The plot was exposed. Surprise was gone.

A sensible officer would have concluded the operation was over.

Roosevelt stayed in Tehran. The CIA kept working.

Three Days

August 16 to 19 is where this becomes hard to reduce to a slogan.

Mosaddegh supporters and Tudeh activists filled the streets. Republican sentiment surfaced. Statues of the Shah came down.

That produced a reaction nobody had scripted.

Ambassador Loy Henderson, back in Tehran, cabled Washington on August 21 with his own analysis of why the pro-Shah side won so quickly. He listed four factors. Iranians of all classes were disgusted by the

behavior of anti-Shah crowds tearing down statues and breaking into homes. They were alarmed by what looked like an alliance between Mosaddegh and the Tudeh, and by thousands of Tudeh demonstrators marching openly through the streets. They were exhausted by two years of crisis and had lost hope of

improvement. And a rupture between Mosaddegh and the Tudeh on the evening of August 18 meant the two anti-Shah forces could not cooperate the next morning — the Tudeh was conspicuously absent all day.

Zahedi announced through a CIA-arranged press conference and CIA printing facilities that he was the legal prime minister. Agents distributed photographs of the royal decrees.

On August 19 a pro-Shah gathering began expanding. Then it expanded fast. Workers, shopkeepers, clerks, students. Security forces declined to suppress it. Some joined.

The day after, the CIA’s own Iran Branch reported that the royalist, pro-Zahedi movement of August 19

contained a large element of spontaneity, and that there seemed to have been a genuine reaction of shock and dismay among the Tehran populace when the Shah left the country. Zahedi and his chief of staff personally manned tanks trying to rally people.

That sentence prevents us from replacing one mythology with another.

But I have to be careful here, and so do you, because that word is doing two jobs in the record and only one of them is honest.

Go back to the Wilber estimate. Its closing section instructs that all available assets should concentrate on weakening the ability of the Tudeh and Mosaddegh to call out mobs and on building up the size and

leadership of pro-Shah mobs. A large royalist crowd wasn’t an outcome the CIA observed. It was an objective the CIA set, in writing, months earlier.

And after it was over, Washington instructed that the Shah should publicly thank the Iranian people for a “spontaneous uprising” in his support and announce his rapid return.

So we have the planned crowd, the reported crowd, and the crowd as public relations. Three different things wearing the same word.

Here is why I still think the August 20 assessment means something. It was an internal Iran Branch cable written by people reporting to their own headquarters, at a moment when claiming full credit would have served them. They didn’t. They said a large element of it was spontaneous, and that the Shah’s flight had

genuinely shocked people. Elsewhere the agency credited its agents with providing the spark that set off the demonstration — a spark, not the fire.

An internal report has less reason to flatter than a press release does. Not none. Less.

So: the CIA spent money, cultivated networks, produced propaganda, worked political and military assets, and set out deliberately to build pro-Shah crowds. And something still took off that was larger than what it had built.

A covert operation met an authentic Iranian political reaction. That doesn’t make the coup less real. It explains why it worked.

The CIA did not create the dry tinder. It helped strike the match — and it had spent months arranging the kindling.

Tehran Turns

Military units switched sides or declared for Zahedi. Newspapers and government buildings were seized. Radio Tehran fell to the royalists. Fighting reached Mosaddegh’s residence. His defenses collapsed. He escaped briefly and surrendered the next day.

Zahedi was prime minister. The Shah, who days earlier had been sitting in Rome wondering whether his reign was finished, was being urged to come home to a triumphant reception.

Within two days of Zahedi taking power, the CIA covertly made five million dollars available to his government.

One more thing happened in those days, and it determined how Iranians would learn about all of this. On August 17, with the Shah still in Baghdad, Henderson recorded that the Shah had been approached roughly a fortnight earlier about sponsoring a military coup and had accepted the idea. He showed no bitterness toward the Americans who had urged and planned action. And Henderson advised him never to disclose the foreign involvement, because doing so would damage his prestige in Iran.

The Shah agreed.

Which is why the version Iranians got, and the version in the files, were never the same story.

Mosaddegh was tried, imprisoned, and then held under house arrest at his estate until his death in 1967.

Who Won?

The oil problem moved toward resolution, and the settlement deserves precision because both sides of this argument tend to simplify it.

Iran did not hand legal ownership of its petroleum back to Britain. Nationalization formally survived. The National Iranian Oil Company continued to exist. Iran got a substantially improved financial deal, including fifty percent of consortium net profits.

But operational control of the producing system went to a new international consortium. The former Anglo-Iranian Oil Company — soon British Petroleum — received 40 percent. Five major American companies together received another 40 percent. Royal Dutch Shell got 14 percent. The French got 6.

Those percentages were not the accidental result of private companies independently wandering back into Iran. American and British government officials negotiated them, months in advance. In a March 4, 1954 telegram, Hoover and Byroade told the British that Washington was in full agreement that the shares should be 40 percent each to AIOC and the American companies, with the remaining 20 percent divided between the French and Shell — the only outstanding problem being how to split that last 20.

The consortium was provisionally formed the following month: Jersey Standard, Socony-Vacuum, California Standard, Gulf and Texas at 8 percent apiece.

So what changed?

Iran still technically owned Iranian oil. But the system was reintegrated into a Western-controlled commercial structure.

Britain went from a hundred percent concession to forty percent of a consortium. America went from nothing to forty.

Britain won the coup and lost the monopoly in the same operation. Now — was oil the reason America did it?

Historians still argue about this, and you should know that. Some emphasize fear of communism; others emphasize control of oil; a more recent reading of the declassified record argues that what actually drove Eisenhower was fear of a general collapse in Iran. There is no consensus.

The documents do not establish the clean version — America overthrew Mosaddegh to steal Iranian oil. American officials genuinely feared Tudeh growth and Soviet opportunity, and Woodhouse deliberately fed that fear precisely because he knew the oil argument wouldn’t sell.

But I’m not going to pretend oil was absent either, because I told you to hold assumptions eight and nine.

Oil was in the plan. A settlement was drafted in advance for the successor government to implement. It had to appear more generous than any previous one. And its announcement was to be delayed so it didn’t land immediately after Zahedi took office.

That is not proof of motive. Plenty of people wanted Mosaddegh gone for reasons that had nothing to do with petroleum.

It is proof that the people planning the operation understood exactly how an oil agreement arriving the week after a coup would look — and scheduled around it.

Multiple motives converged.

Britain wanted its oil crisis resolved. Washington wanted Iran outside the Soviet orbit. The Shah wanted his throne. Zahedi wanted power. Royalists wanted Mosaddegh gone. Clerics had their own fears. Officers had theirs. Oil companies wanted access.

Nobody had to be lying about their motive. They only had to arrive at the same conclusion.

The Price of Stability

From Washington the immediate result looked like a triumph.

Iran stayed out of the Soviet bloc. Oil production resumed. The Shah became a reliable partner. American influence grew. The Iranian military tied itself to the United States.

Over the following decades Iran modernized dramatically. Roads, schools, industry, women’s education, land reform, urbanization.

The Shah was not a puppet taking orders. He had his own ambitions and increasingly saw Iran as a regional power in its own right.

But political power concentrated around him. Opposition narrowed. The security state grew.

SAVAK was created in 1957, and it was not purely an Iranian invention. It was established with direct assistance from the CIA and Israel’s Mossad — the foundational training, the organizational models and the technical equipment came from outside, as Mark Gasiorowski has documented. French intelligence assisted as well. It became notorious for surveillance, detention and torture, and from 1965 to 1978 it was run by Nematollah Nassiri — the colonel whose arrest at Mosaddegh’s door had collapsed the first coup attempt in August 1953.

That second foreign partner deserves a paragraph, because it begins a relationship that will matter enormously later in this series.

Before 1953, Iran was drifting away from Israel. Iran voted against the UN partition plan in 1947. It granted de facto recognition on March 6, 1950, the second Muslim-majority country to do so after Turkey. Then in

1951 Mosaddegh withdrew the Iranian consul from Jerusalem, courting Arab and Islamic opinion while he fought Britain over oil.

After 1953 that reversed completely. Israel maintained a mission in Tehran that functioned as an embassy and eventually exchanged ambassadors, and the two countries developed robust trade in oil — which mattered enormously to a state under an Arab boycott. After the Six-Day War, Iran supplied a significant portion of Israel’s oil needs, and Israel became an important source of weapons for Iran. Thousands of Israeli businessmen and technical experts worked on Iranian development projects. El Al flew direct to Tehran. Uri Lubrani served as ambassador from 1974 to 1979, and Rabin, Allon, Dayan and Peres all made the trip.

Iran became the most important pillar of what Israel called its periphery doctrine: alliances with the non-Arab states on the region’s edge to balance the Arab states on its borders.

None of that had been available while Mosaddegh governed. Iran became, from Washington’s perspective, stable.

Stability is a seductive word. It usually means things are working for the people using it. There was one problem. Iranians remembered how it had been produced.

The Warning

This is the part I find hardest to shake.

American intelligence saw it coming. Not vaguely. Specifically. Two years early.

In an October 1951 assessment, the CIA station in Iran warned Washington that if the United States

continued to side spectacularly with the British, the brunt of anti-Western feeling could easily cover the

United States as well as Great Britain. The same report noted that the traditional Iranian instinct was to turn toward the West for support — provided the West was not represented by Britain alone.

And here is the detail that finishes the argument.

On August 19, 1953, in the middle of a coup the CIA had helped organize, Henderson reported that no hostility was directed at foreigners. No shouts of “Yankee go home.” No Americans stoned.

That is what America inherited. A country where anti-Western anger existed but was pointed almost entirely at London.

Washington had been told what would happen if it took Britain’s side. It took Britain’s side.

Covert operations have a strange accounting system. The immediate result posts to one ledger. The political memory posts to another. That second bill can take decades to arrive.

By 1978 the Shah’s Iran was coming apart. Liberals wanted him gone. Marxists wanted him gone. Students, nationalists, religious conservatives — millions of Iranians who agreed on almost nothing had found the one thing they agreed on.

And outside Paris, an elderly Shiite cleric was giving interviews to journalists from around the world under the trees of a small French village.

His name was Ruhollah Khomeini.

Washington knew who he was. American officials were already trying to work out who the people around him were, and what they wanted.

Which raises the question the next installment has to answer honestly, without deciding the answer in advance:

What exactly did the West think it was getting?

 

Next Thursday: Part III — The Ayatollah.