More than 40 years after the 1973 military coup that cost Chilean President Salvador Allende his life and brought General Augusto Pinochet to power, the historical record on the U.S. role in Chile continues to stir debate. Far from being a “cold case” of the Cold War, as Foreign Affairs implied in its July/August issue, it remains a hot and controversial topic. Jack Devine’s audaciously titled article, “What Really Happened in Chile,” relates his version of the story of infamous U.S. covert actions in which Devine himself played a key role as a young CIA officer based in Santiago.

Devine’s central argument is that between 1970 and 1973, the CIA sought to protect Chile’s democratic institutions from Allende’s Popular Unity government, which Washington believed would push Chile toward Cuban-style socialism and into the Soviet orbit. According to Devine, the CIA was merely “supporting Allende’s domestic political opponents and making sure Allende did not dismantle the institutions of democracy,” such as opposition parties and media outlets. The CIA’s goal, Devine suggests, was to preserve those institutions until the 1976 elections, in which Allende’s opponents, bolstered by CIA support, would presumably defeat Allende at the polls. 

In this view, the military coup and the bloody Pinochet dictatorship, which lasted nearly 17 years, were unfortunate but unintended consequences. But that is not what really happened in Chile. 


As Devine acknowledges, in the fall of 1970, U.S. President Richard Nixon ordered the CIA to orchestrate a military putsch that would prevent the recently elected Allende from assuming office. This top-secret plan was referred to as Track II, to distinguish it from Track I, a covert political campaign supported by the State Department to convince the Chilean Congress not to ratify Allende’s election and thus to keep him from office. 

Devine benignly characterizes Track II as a misguided covert action. In fact, Track II centered on a violent criminal scheme. The plan was to kidnap Chile’s commander in chief, General René Schneider, who firmly opposed the idea of a military coup. “The CIA was aware of the plan,” Devine notes, as if the agency were an innocent bystander, simply gathering intelligence on the operation. 

The truth is far more sinister. The Schneider operation was a CIA-sponsored plot: CIA officials pressed the agency’s station in Santiago to come up with a way to “remove” Schneider because he was standing in the way of a military coup. CIA representatives met repeatedly with the conspirators, led by a retired Chilean army general, Roberto Viaux, and an active-duty brigadier general, Camilo Valenzuela. On October 19, CIA headquarters sent the station six untraceable submachine guns and ammunition in a diplomatic pouch, to be provided to the plotters. The agency also provided $50,000 to Valenzuela to bankroll the operation and thousands more to Viaux to keep the operation “financially lubricated,” as one CIA cable stated. Given the risks involved, the CIA issued the plotters life insurance policies. 

But on October 22, the thugs hired with CIA funds shot Schneider during their attempt to abduct him as he drove to work. U.S. operatives then provided $35,000 to help some of the assassins flee Chile -- and to purchase their silence about the CIA’s involvement. 

The very next day, as Schneider lay dying from multiple gunshot wounds, high-level CIA officials sent a cable of commendation to the station in Santiago: “The station has done [an] excellent job of guiding [the] Chileans to [the] point today where a military solution” -- that is, an anti-Allende coup -- “is at least an option for them,” the cable read. The agency lauded its operatives in Chile “for accomplishing this under extremely difficult and delicate circumstances.” 

In the face of widespread public revulsion in Chile at Schneider’s murder, the Chilean military officials whom the CIA had paid and counted on failed to go forward with the coup. “At that point, all coup plotting ended and Nixon drastically altered his policy,” Devine asserts. Washington’s “new goal,” he writes, “was to support the political opposition and avoid giving Allende an excuse to exploit anti-American sentiment to increase his domestic popularity and international support.” 

The declassified record tells a very different story: U.S. policymakers adjusted their strategy but not their ultimate goal. Rather than rely on a small group of covert operatives to quickly catalyze a coup, Washington developed a longer-term effort to destabilize the Chilean government economically, politically, and militarily and to create a climate conducive to Allende’s demise. For the CIA, “Track II never really ended,” Thomas Karamessines, Devine’s superior officer in the agency’s Directorate of Operations, testified during a special Senate investigation in 1975. “What we were told to do was to continue our efforts. Stay alert, and to do what we could to contribute to the eventual achievement of the objectives and purposes of Track II.” 

Devine appears unaware that Henry Kissinger, Nixon’s national security adviser, convinced Nixon to reject the State Department’s position that the United States should accept Allende’s election and work toward his electoral defeat in 1976. At a pivotal National Security Council meeting on November 6, 1970, three days after Allende’s inauguration, Kissinger’s talking points called for a discussion of “actions we can take ourselves to intensify Allende’s problems so that at a minimum he may fail or be forced to limit his aims, and at a maximum might create conditions in which a collapse or overthrow might be feasible.” Both Secretary of Defense Melvin Laird and Secretary of State William Rogers agreed that “we want to do it right and bring [Allende] down.” Nixon instructed his top national security aides to pursue aggressive actions to subvert Allende’s government, masked by a “very cool and very correct” public posture toward Chile, which would deny Allende the chance to use the specter of U.S. intervention to rally nationalist support. 

By the time Devine arrived in Chile in August 1971, the CIA had already initiated a five-part “Covert Action Program for Chile.” The operations included “enlarging contacts in the Chilean military,” “political action to divide and weaken the Allende coalition,” “support to non-Marxist opposition political groups and parties,” aiding media outlets that “can speak out against the Allende government,” and running propaganda operations through media outlets to “play up Allende’s subversion of the democratic process.” 

Devine minimizes the CIA’s contact with the Chilean military leadership, suggesting that the CIA was simply gathering intelligence. But at the first meeting after Allende’s inauguration of the high-level interagency committee that oversaw clandestine operations, Kissinger stated that the goal of maintaining covert contact with Chile’s military officers was “not just for intelligence but for potential future action.” CIA documents reveal that as political and economic instability increased in Chile, the agency stepped up its communication with key officers. A May 1973 memorandum to CIA Director James Schlesinger noted that the agency had “accelerated efforts against the military target” in order to “better monitor any coup plotting and bring our influence to bear on key military commanders so that they might play a decisive role on the side of the coup forces.”

Moreover, the CIA was not the only part of the U.S. government bringing its influence to bear. The U.S. Department of Defense also maintained contact with the generals. Indeed, a full year before the coup, U.S. military officials met with Pinochet and his aides in the Panama Canal Zone. A declassified intelligence report recorded Pinochet’s belief that Allende “must be forced to step down or be eliminated” and a clear message from U.S. Army officers in response: the “U.S. will support [a] coup against Allende with ‘whatever means necessary’ when the time comes.”


When the time came, the United States did not directly participate in the coup, as Devine correctly points out. But that is not because Washington opposed a military takeover. Rather, according to the minutes of an interagency meeting in October 1972, U.S. officials determined that “if and when the Chilean military decided to undertake a coup, they would not need U.S. government assistance or support to do so successfully.” Even so, just four days before the coup, the CIA was working on “a capability for influencing the situation” if, as one State Department report noted, the predicted coup was “in danger of failure.”

In that sense, the issue of whether Washington played a direct role in the coup is a red herring. The questions that matter more are whether the Nixon administration attempted to “bring down” Allende by creating conditions “in which a collapse or overthrow may be feasible,” as Kissinger put it, and whether the CIA contributed to that effort in a significant way.

The answer to both questions is yes -- and by his own account, Devine was deeply involved in the covert operations that, according to his CIA superiors, made the biggest contribution to the 1973 coup. “My most important responsibility at the time was handling the ‘media account,’” Devine writes, “especially the CIA’s relationship with El Mercurio,” Chile’s leading newspaper. Although Devine casts the $2 million that the CIA clandestinely funneled to El Mercurio -- the equivalent of more than $11 million in today’s dollars -- as an effort to preserve “press freedom” in Chile, declassified White House and CIA records show that Nixon and Kissinger authorized that funding so that the newspaper could mobilize and bolster political opposition to Allende. 

Indeed, far from being the professional news outlet that Devine describes, El Mercurio was a key player in the pro-coup forces in Chile. Only ten days after Allende’s election, the paper’s owner, Agustín Edwards, traveled to Washington to meet with Kissinger and CIA Director Richard Helms and provide detailed intelligence on real and potential coup plotters inside the Chilean military. With massive CIA financing, El Mercurio positioned itself as a bullhorn for organized agitation against the government. In the summer of 1973, the CIA station in Santiago reported that El Mercurio and the militant right-wing group Patria y Libertad (Fatherland and Liberty) “have set as their objective [the] creation of conflict and confrontation which will lead to some sort of military intervention.” After Allende was overthrown, the CIA gave special credit to the media project; according to one CIA memo, it “played a significant role in setting the stage for the military coup of 11 September 1973.” 

Moreover, as Pinochet’s forces began systematically murdering and torturing thousands of civilians -- some 1,200 people were executed just in the three months following the coup, including two U.S. citizens -- the CIA continued to pass funds to El Mercurio to generate popular support for the regime. The media outlets “have supported the new military government,” the CIA reported three months after the coup, seeking continued financing for the paper. “They have tried to present the Junta in the most positive light for the Chilean public.” Given Devine’s covert role in supplying funds to El Mercurio in the aftermath of the coup, his lament for the Pinochet regime’s atrocities during that period rings hollow.


Devine decries the “conventional wisdom . . . that Washington played a crucial role” in the coup. He asserts that “the CIA should not be blamed for bad outcomes it did not produce.” But the historical record confirms that the CIA did contribute to the tragic outcome of a bloody regime change and the rise of authoritarianism in Chile. For its role in the overthrow of Allende and the consolidation of the Pinochet regime, the CIA deserves all the blame it has received -- and more.

And so do the policymakers who sent the CIA on this disgraceful mission. A few days after the coup, Nixon and Kissinger spoke and complained to each other about criticisms of U.S. policy that had appeared in press coverage of the coup. “In the Eisenhower period we would be heroes,” Kissinger mused, referring to the CIA-sponsored coup that toppled the democratically elected leftist president of Guatemala in 1954. 

“Our hand doesn’t show on this one though,” Nixon suggested. 

“We didn’t do it,” Kissinger replied, referring to a direct U.S. role in the coup in Chile. “I mean we helped them. [Word omitted] created the conditions as great as possible.” 

“That is right,” Nixon agreed. 

And that is what really happened in Chile.

PETER KORNBLUH is Director of the National Security Archive’s Chile Documentation Project and the author of The Pinochet File: A Declassified Dossier on Atrocity and Accountability.


I do not expect to change the minds of Peter Kornbluh and others who have spent decades insisting that the CIA was the architect of the 1973 Chilean military coup against Salvador Allende. I would, however, like to respond to some of Kornbluh’s points and once again stress firmly that the CIA did not plot with the Chilean military to oust Allende in 1973.

Like many other critics of the CIA’s actions in Chile during the 1970s, Kornbluh fails to properly distinguish between the U.S. role in the failed coup attempt in Chile of 1970, which sought to keep Allende from taking office, and the U.S. role in the coup in 1973, which resulted in Allende’s overthrow. As I noted in my article, the 1970 coup attempt, which President Richard Nixon instructed the CIA to support, easily meets my definition of bad covert action. The conditions on the ground in Chile did not favor the kind of military coup that the White House envisioned. No one at the CIA really believed the coup would succeed; in fact, the CIA station chief in Santiago strongly recommended against supporting it. Even the White House ultimately came around to the same position: declassified documents show that on October 15, Henry Kissinger directed the CIA to tell Roberto Viaux to call off the planned kidnapping of General René Schneider. But Viaux ignored that instruction and went ahead with the ill-fated plot. 

More important, there is no evidence to support the view that the CIA engaged with the plotters who managed to overthrow Allende three years later. Even Kornbluh seems to recognize this. So the disagreement comes down to the extent to which the CIA’s activities influenced the political environment and contributed to Allende’s downfall. Based on my experience inside the CIA station in Santiago during the events in question, I clearly have a different perspective on that matter than Kornbluh does. 


During my time in Chile, between 1971 and 1973, the CIA pursued a policy designed to support the political opposition, including certain media outlets and protest groups, but did not collude or conspire with the Chilean military to overthrow Allende. The goal was to keep the opposition alive and help position it to defeat Allende’s party in the next national elections. Washington also hoped to thwart Allende’s efforts to draw Chile into the Soviet sphere. In the context of the Cold War, the importance of that objective cannot be overstated.

I’m not arguing that the agency’s activities didn’t contribute to Allende’s fall; they did. But that outcome was not the intention of the CIA’s support to the opposition -- and the agency’s activities clearly weren’t the only factor that contributed to the final result. Indeed, the CIA’s support played a relatively minor role compared to the domestic political and economic factors that led the military to act against Allende.

Kornbluh and others critics tend to ignore the truth about the conditions in Chile under Allende. They overlook the economic and political incompetence of Allende’s government and his very serious flirtation with President Fidel Castro of Cuba and the extreme leftists of Chile’s Revolutionary Left Movement, who revealed their true nature when they violently attacked the peaceful “March of the Empty Pots and Pans,” led by women’s groups in December 1971. Kornbluh also ignores the major role that the Communist Party of Chile played in Allende’s government, as well as the aid and support Allende received from Cuba and the Soviet Union, which included the training of Chilean intelligence operators and of Allende’s personal security detail.

As Allende enacted radical, destabilizing land reforms and nationalized industries, including Chile’s vitally important copper industry, U.S. officials began to fear that Allende’s government would also take direct control of the media. The CIA’s support for El Mercurio aimed to prevent such a strike against Chilean democracy. As I noted in my article, the paper’s editors in Chile took no guidance from the CIA. The funding the agency provided was designed to keep the newspaper in business at a time when the Allende government was blocking its access to newsprint. Agustín Edwards, the paper’s owner, feared that the government would soon shut down the paper. Edwards had a good, but not particularly deep, relationship with Nixon; he lived abroad and had no contact with the CIA in Chile. And there was no indication in the field that Edwards had any real sway over U.S. policymakers, especially after the failed 1970 coup.

Just as Kornbluh misconstrues support for a free press as evidence that the CIA took part in coup plotting, he also draws inaccurate conclusions about the CIA’s ties to the Chilean military. Kornbluh asserts that those ties were deeper than I described in my article. He bases that conclusion on a declassified document describing a meeting between U.S. military personnel and Augusto Pinochet in the Panama Canal Zone and another declassified document stating that the CIA would be increasing its surveillance of the Chilean armed forces.

First of all, assuming that the first document’s account of the meeting in Panama is accurate, the discussion it describes does not reflect U.S. policy. And there is no evidence to suggest that this discussion was related to the CIA’s covert-action program. As for the second document, there is a world of difference between conducting surveillance on another country’s armed forces and collaborating with them. Kornbluh suggests that the U.S. relationship with the Chilean military was robust, but I know the exact number and quality of the United States’ military sources at the time, and they were indeed meager.

Ultimately, the Chilean military moved against Allende not because the United States wanted it to do so but because the country was in disarray. By the spring of 1973, the Chilean economy had spiraled out of control and street demonstrations had become routine. Coup rumors were rampant, and in what became known as “the tank putsch,” junior military officers started plotting on their own. The generals decided to take charge of the coup plotting to maintain discipline in Chile’s military institutions and to preserve stability.


It is true that Washington welcomed the 1973 coup, since it appeared to be a Cold War victory for the United States. Kornbluh cites a transcript of a conversation between Nixon and Kissinger in which the men seem to take credit for Allende’s removal. But the fact that Nixon said something doesn’t make it true. It is hardly uncommon for political figures to take excessive credit for developments they see as positive -- just as they distance themselves from outcomes they deem unfavorable.

Despite our differences, Kornbluh and I agree on one point: U.S. involvement in Chile in the 1970s remains a hot-button issue. I take some modest comfort from the fact that as a former CIA field operative who served in Santiago during the Allende era, I have been able to offer my firsthand account of what really happened in the Chilean coup in 1973. I hope that it proves useful to historians and that its lessons might inform future policymakers and CIA leaders.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now