Courtesy Reuters

A year ago what was most striking about major American "secret" operations, from Nicaragua to Angola to Cambodia, was how little was secret about them. Support for the Nicaraguan rebels, or contras, the most controversial example, was openly debated and openly funded. Neither opponents nor supporters had reason to keep it secret; for its part the Reagan Administration regarded "covert" action as good policy and good domestic politics, a key element of the Reagan Doctrine, which is intended to challenge Marxist-Leninist states around the world.

Now it turns out that the United States has come full circle. In the case of Nicaragua, the Central Intelligence Agency became the agent of "overt" covert action. Thus, when the Reagan Administration decided to sell arms to Iran and keep the operation secret, it turned inward, to the White House staff. In embroidering that operation to divert money for the contras, White House aides apparently kept the President ignorant in order to protect him—providing him with "plausible denial" of the sort the CIA had long since abandoned.

The Iran/contra affair raises questions that have not been at the center of the public debate since the investigations of U.S. intelligence activities by Congress in the mid-1970s. Should the United States attempt major covert operations at all? Can it? In what circumstances, and—crucially—how, if at all, can these secret operations be made to square with the requirements of governance in an open democracy?


Spying may be the world’s second oldest profession, but for the United States it was only the cold war, coming on the heels of America’s wartime experience with secret operations conducted by the Office of Strategic Services, that led to the creation of an intelligence service in peacetime—and to covert operations. Wartime success and postwar threat: these were the backdrop for the creation of the CIA. In a few years America plunged from the euphoria of victory in World War II to the confrontation with a looming Soviet threat.

The first line of American response to the onset of the cold war was overt: the surge of assistance to Europe through the Truman Doctrine and the Marshall Plan. But the second line was renewed interest in what was then called covert "psychological warfare"—what we could now call propaganda—as a way to respond to the Soviet Union by means that were less than war but more than nothing. In February 1948 a communist coup succeeded in Czechoslovakia, while communist agitation grew in France and Italy; Western Europe seemed to teeter in the balance. By March Washington whipped itself into near-hysteria when the American high commissioner in Germany, General Lucius Clay, cabled his warning that war with Russia "may come with dramatic suddenness."

In this atmosphere the National Security Council in June approved NSC 10/2, a plan that had originated with George Kennan, then director of the State Department’s Policy Planning Staff and the author of the famous "X" article outlining the policy of containment of the Soviet Union. NSC 10/2 was the turning point for covert action, expanding it from propaganda to direct intervention. In the words of the document, covert action comprised:

propaganda, economic warfare; preventive direct action, including sabotage, anti-sabotage, demolition and evacuation measures; subversion against hostile states, including assistance to underground resistance movements, guerrillas and refugee liberation groups, and support of indigenous anti-communist elements.

NSC 10/2 also codified the notion of plausible denial: operations were to be "so planned and executed that any U.S. Government responsibility for them is not evident to unauthorized persons and that if uncovered the U.S. Government can plausibly disclaim any responsibility for them."

The fledgling CIA’s first success came that same year when its covert support to the Italian Christian Democrats helped them beat back an electoral challenge from the Italian Communist Party. By 1950 the United States had succeeded in the covert struggle in Western Europe, and in Eastern Europe its covert operations wound down to propaganda and intelligence-gathering. The center of the battle against communism moved, as official Washington saw the world, away from Europe to small, weak countries. Europe had put the CIA into the business of covert political action, but it was Asia that got the agency into secret paramilitary operations in the Korean War, a pattern repeated a decade later in another Asian war, Vietnam.

In November 1952 Dwight D. Eisenhower was elected President; his campaign had called for a more active response to the Soviet Union than the passive containment of the Truman Administration. John Foster Dulles became secretary of state, and in February 1953 his brother, Alien Dulles, was named director of central intelligence. The campaign pronouncements and the new personalities suggested that covert operations would be a key weapon in the new Administration’s war against the global threat of communism. The operators soon got their chance.

On August 21, 1953, after a week of turmoil in the streets of Tehran, the Iranian prime minister, Mohammed Mossadeq, who had nationalized that country’s oil industry, surrendered to General Fazlollah Zahedi. Three days later the shah, who had fled Iran the previous week with his queen, returned to the capital. At his palace a few days later he offered a toast to Kermit "Kim" Roosevelt, the chief of the CIA’s Near East and Africa Division and the man who had improvised Mossadeq’s downfall: "I owe my throne to God, my people, my army—and you!

The next year, on June 16, 1954, Guatemalan Colonel Carlos Castillo Armas crossed the border into his country from Honduras with a few hundred men trained and armed by the CIA. Pilots under CIA contract flew air cover. The president of Guatemala, Jácobo Arbenz Guzmán, was deserted by his air force and his army, which refused his order to arm workers and peasants. The American ambassador in Guatemala, John Peurifoy, hastily arranged a transfer of power to the chief of the armed forces, Colonel Enrique Díaz. (In a moment of tragicomedy, Díaz immediately pledged that he would continue the struggle against Castillo Armas, America’s designated successor to Arbenz. Peurifoy thereupon secured Díaz’s resignation, and after complicated negotiations between the armed forces and Castillo Armas, the latter emerged as president. )

It is eerie that thirty years later the targets of American covert action are once again in Iran and Central America. Yet the parallel may mislead as much as it instructs, for if the more recent episodes testify to how difficult covert action is, the earlier ones seemed to show how easy it was. The Iran and Guatemala operations—codenamed TPAJAX and PBSUCCESS, respectively—coming within a few years of the CIA’s success in Western Europe, made the agency’s reputation and set the pattern for covert action in the years ahead. Small, cheap, fast and tolerably secret, they encouraged Washington to think other covert actions could be likewise. When the next Administration decided to confront revolution in Cuba, its covert response was the same as in Guatemala. So were the CIA officers who carried it out.

The blush of short-run success amidst the cold war obscured several cautions. In the early 1950s both Iran and Guatemala were eminently vulnerable to manipulation by an outside power, particularly the United States. In both, contending political forces were in close balance. Those balances might have tipped against Mossadeq and Arbenz even had the CIA not intervened. So it appeared that relatively small operations were enough to tip the balance. Yet in both cases those limited interventions might have failed. In fact, Roosevelt’s first plot did fail; Dulles was ready to roll up the operation and bring the troops home. And the CIA officers who ran PBSUCCESS were under no illusions: if their deceptions failed and Arbenz were able to get his military into combat, the invaders would be overwhelmed.

Thus success was purchased at the price of enlarging the intervention. American purposes did not change, but the operational requirements of achieving them did. Once the United States was committed, in secret and in a small way, its stakes increased, and the CIA took the next step. The effort to intimidate Arbenz became a paramilitary campaign, if a small one. In the process, plausible deniability became more and more tenuous.

Six years later at the Bay of Pigs, deniability evaporated entirely. "How could I have been so stupid, to let them go ahead?" The words were John Kennedy’s. When the CIA-trained invasion force of Cuban exiles hit the beach in the early dawn hours of April 17, 1961, everything went wrong: the lives of brave Cubans were spent; the United States was seen to be intervening; and the intervention failed. Once the plan had changed, without anyone outside the CIA quite noticing it, from a guerrilla operation into a full-fledged amphibious invasion, the chance of keeping it tolerably secret diminished to the vanishing point.


Evaluating covert action in retrospect is speculative, for it is bedeviled by the imponderable of what might have been; history permits no reruns. Failures, such as the Bay of Pigs, are apparent, but successes are harder to judge. Consider the CIA intervention in the Angolan civil war of 1975. On the surface it was a failure: the Cuban- and Soviet-supported faction, the MPLA (Popular Movement for the Liberation of Angola), was installed while the American role in trying to prevent that outcome was being exposed. If, however, the initial purpose was more limited—i.e., to raise the price of victory for the MPLA and its Soviet and Cuban backers—then Angola might be counted a short-run success. Yet American officials did not convey the impression that their aims actually were so limited, either at the time or later.

When covert actions have succeeded in their short-run purposes, it may be that the action, while marginal, was just the bit of "support for our friends" that tipped the balance in the internal politics of a foreign country. On the other hand, it may be that the American support was entirely superfluous, that the same successful outcome would have ensued without the U.S. involvement. If this is so, all the covert action accomplished was to implicate the United States and tarnish the success by labeling it "made in America" when the existence of the covert action became known.

A case in point was covert American support to opposition political parties and media in Chile during the presidential tenure (1970-73) of Salvador Allende, a self-proclaimed Marxist. There is no question that those parties and media were under pressure from the Allende government. The opposition forces did survive to fight another day, but there is no telling whether CIA support for them was decisive or irrelevant.

What is clear is the signal conveyed to history by the revelations of American covert action. In retrospect, most reasonably objective observers conclude that Salvador Allende’s experiment in Chile would have failed on its own terms. Yet history’s lesson is not that Allende fell of his own accord. History’s lesson is that the United States overthrew him in 1973. That is the public perception even though this lesson is untrue in the narrow sense: Washington did not engineer his coup, nor did the CIA or the American military participate in it. The very fact of American covert action meant that, at a minimum, "it is fair to say that the United States cannot escape some responsibility for [Allende’s] downfall."

By the same token, in 1975 when South Africa intervened to back the U.S.-supported Angolan factions, the FNLA (National Front for the Liberation of Angola) and UNITA (National Union for the Total Independence of Angola), a covert action originally intended to counter the Soviet Union and Cuba then signaled something else. It signaled an alliance with the apartheid regime in Pretoria. When, in December 1975, the American Congress reacted by cutting off the CIA operation, that only ratified what was seen as the inevitable result—defeat—in the eyes of both Washington policymakers and the rest of the world.

In a longer perspective, neither the Iran nor Guatemala operations can fairly be given too much credit—or too much blame—for what happened afterward in Iran and Guatemala. On one hand, TPAJAX restored the shah of Iran to his throne where he remained for nearly a quarter-century, a pro-Western bastion in a turbulent region. Twenty-five years of stability is no mean feat in international affairs. On the other hand, American covert action identified the shah’s Iran more closely with the United States than was good for either of them.

In any case, however, the aspects of American policy that loomed so large in the shah’s downfall in 1979 were overt, not covert. They were his image as an American client, the waste and corruption associated with his massive U.S. arms purchases and his own dependence on the United States. These factors owed much more to American policy during the 1970s than to the event of 1953.

A similar conclusion also applies to Guatemala in 1953. If, in retrospect, the "success" of PBSUCCESS also looks more ambiguous than it seemed at the time, most of the blame or credit lies with American foreign policy, not with covert action. PBSUCCESS did not make it inevitable that Washington would then forget about Guatemala; it only made it possible. David Phillips, a CIA officer who worked on PBSUCCESS, the Bay of Pigs and Chile, laments that "Castillo Armas was a bad president, tolerating corruption throughout his government and kowtowing to the United Fruit Company more than his own people." But he argues that the United States:

could have prevented this with the vigorous exercise of diplomatic pressure . . . to assure that he pursued social reform for the many rather than venal satisfaction for a few. Instead, Washington breathed a collective sigh of relief and turned to other international problems.


Several covert operations of the 1950s remained secret for a long time: the CIA’s assistance to Tibetans resisting the domination of their land by the People’s Republic of China, regarded in intelligence lore as a successful holding action, is still a little-discussed operation, especially because it is an embarrassment now that Sino-American relations have thawed. The effort to unseat President Sukarno of Indonesia, who had earned Washington’s opposition for his espousal of nonalignment, ranged from covert political action to a paramilitary operation; it is not much better known than the Tibetan operation.

Even in the 1960s several brief and limited—small in terms of numbers of people involved though not in terms of purpose—interventions remained secret for some time. So-called Track II—a secret effort to touch off a military coup in 1970 to prevent Allende from being seated as Chile’s president, an operation run without the knowledge of the State or Defense Departments—was not revealed for five years after it happened. And the sad plots in the early 1960s to assassinate Fidel Castro stayed buried for ten.

Not so now, as the Iran/contra affair underscores. Major covert actions will become public—sooner rather than later, perhaps before the operation is over. Americans are more skeptical of their government, of its information and its capacity, a skepticism that is a legacy of recent history labeled "Watergate." When Ronald Reagan, the most popular President in generations, first denied in 1986 that his Administration had traded arms sales to Iran for the release of American hostages in Lebanon, most Americans did not believe him. This skepticism has been reinforced by other factors. One is the prominence of investigative journalism; every cub reporter aspires to be Woodward or Bernstein of Watergate fame. The media now contain more people asking hard questions, even of secret operations, and probing for leaks; and there are fewer who are prepared to take the government at its word.

If reporters are more likely to seek information on "secret" operations, so are they more likely to find it. "Leaking," always present, has become routine in Washington; it has become almost acceptable. Officials sometimes leak information merely for the gratification of being pandered to by journalists more famous than they. More often, they leak to rally opposition to or, more rarely, support for a given policy. Administration after administration, regardless of its political persuasion, declares war on leakers. Those wars always fail. They fail for a simple reason: the ship of state is like no other, for it leaks from the top.

Officials at the top of government are precisely those who know of covert actions and thus are most likely to take their opposition to particular programs into the open. This is true of the executive branch, and all the more so of Congress, where this tendency is reinforced by institutional pride and, often, by partisan politics. On the whole the intelligence committees of the House and Senate have kept secrets at least as well as the executive branch. Yet their role in overseeing covert action means that those who might oppose a particular project are more likely to know of it. The process creates a set of frustrated opponents who will, on occasion, go public with their frustration.

Moreover, if the leak does not initially come from Washington, the scent will eventually be picked up by the American press from overseas—even if, as in November 1986, the first article is published in Beirut and in Arabic.

Not every exposé, however, has created a controversy. Even now, not every covert action is controversial. Of the 40 or so covert actions under way in the mid-1980s, at least half had been the subject of some press account. Yet only several were controversial enough that the original leaks developed into continuing stories. Most of the rest were open secrets, more unacknowledged than unknown; they were so because most members of Congress thought they made sense, as did most Americans who knew or thought about them—and, no doubt, most of the journalists who reported them. Former CIA Director William Colby characterized the reaction to revelations of American assistance to the resistance in Afghanistan: "Afghanistan was a two-column headline in The Washington Post for one day, then almost nothing."

Certainly there will be cycles in American attitudes, as there have been before. Concern over Soviet power coexists with worry about nuclear war and peace; one predominates, then the other. In the early 1980s most Americans evidently shared their President’s concern with the Soviet threat, and their congressional representatives went along with huge increases in defense spending—and in covert action. So, too, was Ronald Reagan able to rebuild considerable authority and discretion in the American presidency, thus ending a cycle begun by Vietnam and Watergate. Indeed, the saddest consequence of the Iran/contra debacle may be that it will set back this rebuilding.

The changes in American domestic politics in the mid-1970s have made it more difficult for the United States to achieve its purposes secretly. Other changes make it harder now than in the 1950s for the United States to intervene successfully at all, covertly or openly. Despite the controversy and mystique that surround covert action, history suggests that there is no magic to it. It means providing foreigners, secretly, with money or weapons or training as tokens of American support.

With the passage of time, however, a little money here, a few weapons there became less likely to achieve grand foreign policy purposes. Castro was a target of a different order than Arbenz. To think in 1975 that a few million dollars might alter the fate of Angola was a faint hope at best, and an illusion at worst, especially given that the CIA recognized that the Soviet Union and other external actors might counter American support with more assistance of their own. Even the CIA officials who planned the Bay of Pigs knew that to delay the invasion until Cuba had received deliveries of advanced Soviet fighter planes would be to condemn the plan to certain failure.

Also, notice the contrast between two Central American cases three decades apart, Guatemala and Nicaragua. Castillo Armas’ liberators numbered no more than several hundred. Their "invasion" was more conjured than real. Yet they had control of the air, in large part because Arbenz, unsure of the loyalty of his air force, was unwilling to risk putting his own pilots in the air. Sulfatos—Spanish for "laxatives," the name Guatemalans gave to the invaders’ bombs—plus rumors exaggerating the size of the invasion were enough to induce Arbenz to capitulate. In the case of Nicaragua, the contras numbered about 10,000 by the mid-1980s, yet not even the most ardent advocates of American assistance to them argue that they are about to induce the Sandinistas to "say uncle," much less that they now pose a threat sufficient to overthrow the regime by sheer force of arms.

The makers of revolutions have learned their own lessons from history, including the history of American covert action. They are determined not to repeat the mistakes of Arbenz and Mossadeq. They have sought to assure themselves the loyalty of the army or to build revolutionary cadres of their own. They also have learned a related lesson: if the United States threatens the revolutionaries, there are other sources of support, including the Soviet Union if need be, to which they can turn. And, unlike Arbenz or Mossadeq, they will turn to those sources sooner rather than later.

Moreover, to some of these leaders the United States is useful as an enemy. Arbenz and Mossadeq and even Allende sought Washington’s approval, or at least its acquiescence. By contrast, if the United States is of any use to Castro, the Sandinistas and the Ayatollah Ruhollah Khomeini, it is primarily as a foreign demon against which their revolutions can rally—even though the Sandinistas were prepared to accept American aid as long as it was forthcoming and the revolutionary Iranians were not above seeking American spare parts for their military. In 1979 Iranians took Americans hostage and released them only when they ceased to be useful counters in the bargaining within the revolution.


In all likelihood, the record shows, covert operations will become known, and America will be judged for having undertaken them. Thus, the practical lessons lead into moral issues. The issues are hardly unique to covert intervention, though they are powerfully present there, and they are often obscured in policymaking by the presumption that covert actions will remain secret. Overt interventions, such as the American invasion of Grenada in 1983, or military attacks, such as the bombing of Libya in 1986, raise similar moral and instrumental concerns. These concerns are not absolute; they must be considered against the gravity of the threat and the adequacy of other available responses.

In December 1976, when I was in Washington working with several old friends who were making arrangements for the transition between the Ford National Security Council and the Carter NSC, we had decided to retain the basic structure of the Ford operation, with its network of sub-cabinet committees for particular purposes. Yet, of course, as a new administration, it was necessary to change the names of those committees, and so we joked about names. The 40 Committee, the Ford Administration’s group for discussing covert action, would become the "If They Can Do It, So Can We" Committee.

"If they can do it, so can we" seems an unacceptable rationale on both moral and instrumental grounds. What the Soviet Union or other nations do cannot settle the issue. We consider ourselves different from them, and imagine that the difference is not only basic to what we are as a people, but also a source of American influence in the world, part of this country’s moral armor. Though our actions often belie our words, we do believe that nations should not interfere in the internal affairs of their neighbors.

We also believe that the example of democracy is a powerful one, one toward which peoples all over the world will gravitate if given the chance. Believing that, we must also believe that the example is a powerful part of our external behavior, not just of our internal arrangements. If people will choose democracy when given the chance, then democracy is demeaned, perhaps doomed to fail, if it is imposed from the outside. There is something incongruous about helping to overthrow governments—especially ones that come to power through elections that we would define as tolerably fair (as in Chile in 1970)—in the name of democracy.

In this view even the "successes" of covert action seem ambiguous or transient in retrospect, accomplished at significant cost to what we hold dear as a people and to America’s image in the world. On the other hand, the world is a nasty, complicated place. In that regard Americans’ historical ambivalence between the high moral view and the feeling that international politics is a dirty business is understandable. As Representative Stephen Solarz (D-N.Y.) put it: "Given Soviet violations of [accepted international] norms," for the United States "perpetual indifference is neither politically practical nor strategically prudent."

Moreover, nations affect each other’s politics in so many ways that any too-tidy definition of "intervention" is suspect. In all the examples cited, covert action formed only part of American policy. The United States decided whether to grant economic aid to Cuba or Chile or Angola, and whether to release Iranian assets held in the United States. Most of these decisions were based on explicitly political criteria. Even if similar decisions toward other countries are not so explicitly political, the decisions in any case have political effects on the country in question; of that fact, foreign political leaders have no doubt.

The same is true of actions by private American actors. U.S.-based businesses either invest or do not invest in a country, and that decision too has not just an economic but also a political effect. That is the case even if the decision is not political in any narrow sense of the term. Most of the businesses or banks that chose not to invest in Chile under Allende probably did not make that choice for any specific political reason, despite Washington’s pressure. Rather, their decision was a business one, based on the climate in Chile. They saw that judgment as an economic one, though political instability surely was a factor in it.

In this context a unilateral self-denying ordinance against all intervention—open or covert—is too restrictive. Some threats to American national security require responses. Some American friends in the Third World deserve support. What is imperative is to keep in mind the long-term costs of intervention for a government that is not notable for attending to long-term considerations.


Given that "covert" action is not likely to remain secret, why not act openly? In the case of aid to the FNLA and UNITA in Angola, covert rather than overt aid spared the first identification with the United States for only a few months; as for aid to the contras in Nicaragua, the "covert" form made not one whit of difference, as the operation quickly became known. Nor is it obvious that in these cases the recipients of American largesse much minded the source of the money being known. There is also the risk that covertness creates a self-fulfilling prophecy: if the United States only aids its friends secretly, then any link to the United States may seem sinister, portending much more than is the fact.

Doing openly what might earlier have been done covertly is not out of the question. The American radio stations broadcasting into Eastern Europe and the Soviet Union from Munich, Radio Free Europe and Radio Liberty, were in form private organizations; advertisements exhorted Americans to contribute to them. In fact, they were created and financed covertly by the CIA as propaganda vehicles. When that support was disclosed in 1967, the radio stations nevertheless continued to operate; they became supervised by a board and supported openly by appropriations from the American Congress.

A still more promising model is the West German political party foundations (e.g., the Konrad Adenauer Stiftung). They are instruments of the major parties but are supported openly by government money. They have openly assisted kindred parties and labor movements around the world. The Reagan Administration was moved in 1983 to create the National Endowment for Democracy (NED), an American parallel to the German foundations. The endowment, with a 1985 budget of $18 million, channels money to institutes of the two American political parties plus an AFL-CIO group and a business group, which then make grants in support of democratic institutions, mostly in the Third World.

So far, the record of the endowment is mixed but hopeful. Its grants, and those of its four constituent institutes, have been cautious and close to government policy. For instance, more than $400,000 went, over two years, to the American Friends of Afghanistan to develop educational and cultural facilities inside those portions of the country controlled by the resistance groups—activity that might in other times have been called the "civic action" component of a paramilitary operation.

Acting openly, however, is neither easy nor a complete substitute for covert action. It requires an explicitness about influencing the politics of a foreign country that is uncomfortable for Americans, and hence likely to be controversial. Moreover, governments that feel threatened by that open assistance can act to prevent it more easily than if it were covert. For this reason, a bias toward openness would not settle immediate, difficult cases: Given that the Sandinista regime has closed the main opposition paper, would it prevent open assistance to opposition political parties in Nicaragua? What are the prospects in South Africa, where the regime has already blocked assistance to the opposition United Democratic Front?

It also remains an open question whether, given American politics, public funding is compatible with creative, and thus controversial, acts by private groups. The NED budget is only an eighth of that of the German party foundations, in a country one fourth the size of the United States. In 1985 Congress halved the endowment budget and denied any funding to the Republican and Democratic institutes, though that prohibition was relaxed the next year. It is a sad fact about American politics, even in the 1980s, that it is easier for the CIA to get money from Congress secretly than for another institution to get it openly—even if the purposes of the two are broadly similar. Funding for the CIA remains wrapped in the cloak of national security, and so members of Congress may be prepared to fund particular activities but prefer not to be seen to vote for them openly.

Open funding is thus likely to be restricted to small, political—and not too controversial—projects. Yet as a long-term direction for American policy, openness would reflect the reality that, as the century ends, national boundaries are more and more permeable. Given this reality, moreover, those groups the United States would like to support may not be so chary of accepting help, even—perhaps especially—if it is open. The United States would say to them: we are prepared to support you but only openly. We think that is better for you. In any case, we know it is better for us.


The circumstances in which opting for a major covert operation makes sense are more and more unusual. U.S. aid to those in Afghanistan resisting Soviet occupation of their country is one such case. Americans, in and out of Congress, broadly support the cause of the rebels, or freedom fighters as some call them, and aid is a way to increase the cost of the Soviet occupation. American assistance, reportedly begun in a small way in the last year of the Carter Administration, escalated sharply in the mid-1980s to an estimated $280 million for 1985, mostly for small arms, clothing and supplies.

The secret is an open one; the American role is not so much covert as, by tacit agreement, unacknowledged. The reason for circumspection is the touchy position of the Pakistan government, the conduit for the American supplies to the rebels. Pakistan has been prepared to support the rebels but is unwilling to be too visible in doing so lest it antagonize its powerful neighbor, the Soviet Union. In these circumstances, the resort to the CIA, rather than the American military, was more a matter of being discreet than of keeping the whole affair secret.

In deciding whether to choose the covert option, prudent policymakers should ask themselves a careful series of "what if" questions. That injunction applies to all policies, foreign and domestic. But it applies with special force to covert action because of the presumption of secrecy.

The most obvious "what if?" is "what happens if—or more likely, when—it becomes public? What if it becomes public in midstream?" Large covert actions will not remain secret, a reminder that is easy to state but hard to embody in the making of policy when the pressures all go in the direction of wishful thinking. Witness the reflections on the Bay of Pigs by Richard Bissell, then head of the CIA’s clandestine service:

the argument was [not] made that this is now a very public business, and we’d better treat it as such, and either cancel it if we can’t stand the publicity, or else do some of the things that will increase the chances of success if we are going to go forward with it.

If the Iran operation of 1985-86 had remained secret for several years after all the hostages had been released, that success might have outweighed the costs of being seen to have traded arms for hostages when the operations became public. Perhaps; we cannot know for certain. It did not, however, take a sophisticated analysis to show that a covert policy targeted on some Iranians was vulnerable to being publicized by opposing Iranian factions if and when it suited their political purposes. And it was equally likely that, when the cover was blown, trading arms for hostages, with a nation the United States had denounced as terrorist, would be deemed unacceptable—by America’s allies, much of the rest of the world and, most important, the American people.

Of course, whether a particular covert operation can bear the test of disclosure is apparent in retrospect but often far from obvious before the fact. Prudence suggests that presidents pay careful attention to such warning signals as the review process throws up—the views of cabinet officers, people in the White House who attend to the president’s interests, and congressional overseers who are surrogates for public reaction.

One warning signal, however, is evident in advance: Does the intervention contradict overt American policy? If it does, as with arms sales to Iran, it is especially improbable that the operation will withstand the test of disclosure. The arms sales were exactly the opposite of the Administration’s public policies, which had twisted the arms of America’s allies not to sell arms to Iran, had sought an end to the Iran-Iraq war with neither a victor nor a vanquished, and had pledged not to bargain with terrorists over hostages, much less to sell arms to them.

A second "what if?" is "what if the first intervention does not succeed? What then?" If covert action is to remain secret, most of the time it will have to be small. Small operations have often begun with grand purposes, objectives incommensurate with the instrument. When the goals could not be achieved, leaders were tempted to take the next step and the next. This happened in the Bay of Pigs and Angola and Iran in the mid-1980s. Sometimes a limited objective can be achieved, but its achievement makes it appealing to hope for more—witness Angola and perhaps Nicaragua, where the United States did seem to achieve its initial aim of cutting weapons supplies from Nicaragua to the anti-government rebels in El Salvador. Answering this "what if?" suggests, at a minimum, careful attention to the CIA’s covert operators themselves, for signs of skepticism about whether operations as initially conceived can achieve their purposes. Such signs were there between the lines of Track II and Angola and Nicaragua. Some risks are worth running, but few are worth running in ignorance.

A third set of "what if" questions is "what signal will be received, by whom and with what result?"—judgments that are also easier with the benefit of hindsight, for they involve calculations of threat and of American interests. Intelligence assessments, by the CIA or the State Department, provide one set of indicators. In 1985-86, for example, American intelligence on Iran was weak, but what there was offered precious little ground for believing there were "moderates" who might be detached from their revolutionary colleagues. Later U.S. intelligence cast doubt on the imminence of a Soviet threat to Iran, one of the original premises of the operation. These were cautions that the intended signals might go awry.

The nature of those who are to receive secret American assistance can provide another warning signal. Since their relationship to the United States is meant to be clandestine, the CIA is often in a weak position to compel them to act to suit American purposes, yet the United States inevitably will become associated with "their" actions, like it or not, if and when the fact of support becomes known. Aid to the contras has been dogged by their origins in Somoza’s hated National Guard and by continuing charges of human rights violations. Similarly, support for resistance forces in Afghanistan may be justified as a way to put strategic pressure on Soviet occupation of that country; but, given the character of the resistance forces, it is not a way to bring "democracy" to Afghanistan.

The regional context, in particular the attitude of American friends in the region, is another source of guidance. In the instance of Afghanistan, American assistance to the resistance is supported, though with varying degrees of publicity, by nations ranging from Pakistan to Egypt to Saudi Arabia to China. In Central America this indicator is more ambiguous, for most of the nations of the region publicly express qualms about aid to the contras while privately hoping the Sandinistas can be made to go away.

The second round of covert action in Angola raised these questions of signals given versus those intended, questions for which the 1975 episode provided guidance. In early 1986 the CIA was authorized to provide $15 million in weaponry for Jonas Savimbi’s UNITA. For the Reagan Administration, the intended signal was anti-communism. For it, there was nothing incompatible about supporting anti-communism in Angola and anti-apartheid in South Africa. Alas, the reality of southern Africa frustrates that conception in the heads of Washington policymakers. Whatever his attractions, Savimbi has one flaw, a fatal one: he is almost completely dependent on South Africa, his army almost a unit of South Africa’s. To support him is to signal to Africans that the United States is throwing its lot in with South Africa, in 1986 just as in 1975.

These rules of thumb amount to establishing a presumption against covert action. The guidance is mostly negative, a series of cautions. It is unwelcome to officials who are looking for something to do rather than something to avoid—a trait that runs deep in the American character and is reinforced by the circumstances in which covert action becomes an option. Yet given how both America and the world have changed over the postwar period, the circumstances in which major covert action makes sense as policy are sharply limited.

Guidelines akin to these were articulated a decade ago by Cyrus Vance, later secretary of state. For Vance, the criterion for covert action in the National Security Act of 1974—"affecting the national security"—was too loose. Instead, he recommended covert intervention only as an exceptional measure, when it was "absolutely essential to the national security" and when no other means would do. Decisions would still be matters of judgment under this more restrictive guide, but no one has improved on the Vance standard. It bears remembering now, after so much damage has resulted from applying less demanding standards to covert action.


If the United States remains in the business of covert action, as it should, albeit under restrictive guidelines, it will continue to confront the paradox of secret operations in a democracy. At first blush the Iran/contra affair seems to suggest that the paradox cannot be managed, that the reforms that resulted from congressional investigations of intelligence a decade ago have counted for little. When the President finally in January 1986 signed the authorization required by law, a "finding," for the Iran arms sales operation, that finding was explicit: do not tell Congress. The congressional overseers did not find out about the Iran operation until autumn—hardly the law’s requirement of "fully and currently informed" by anyone’s definition. Later on the President himself apparently was not told, when the Iran and contra operations crossed.

In another sense, however, the system worked. In deciding to sell arms to Iran, the President pursued a line of policy opposed by both his secretaries of state and defense, about which he was afraid to inform the congressional intelligence committees, and which was liable to be revealed by Iranian factions as and when it suited them. It is hard to imagine any system providing more warning signals. When most of the government’s senior foreign policy officials are opposed, it is likely that the policy, and not they, are wrong. The President thus proceeded at his own peril.

With regard to the diversion of money for the contras, the lesson is not that the NSC staff should be eliminated or the national security adviser made subject to Senate confirmation. Presidents will always need a source of private advice and a means of brokering the actions of the many foreign policy agencies of government. Moreover, if presidents are determined to get something done, they will be able to find someone, somewhere in the White House, to do it. So, too, if the United States continues to have a clandestine service, presidents will be tempted to use it as a middle, not a last, resort.

Rather, the lesson is a caution for presidents and those who advise them: do not run covert operations from the White House. Two decades ago it would have been unthinkable for an administration to do so; then the reason was that presidents wanted to stay at arms’ length from such things, even if they could not plausibly deny them in a pinch. Now as then, if covert actions are to be undertaken, they should be done by the agency of government constructed to do them—the Central Intelligence Agency. It has both the expertise and the accountability. The Tower Commission’s report is a sad tale of high-level meetings with no preparation in advance and no notes afterward, and of operational amateurism.

Moreover, as occurred recently as well as two decades ago, if the president’s closest advisers become the operators, the president loses them as sources of detached judgment on the operations. The president’s own circle become advocates, as Alien Dulles did in the Bay of Pigs, not protectors of the president’s stakes (even if he does not quite realize his need for protection). So it was with President Reagan’s national security advisers, Robert McFarlane and John Poindexter; once committed, they had reason to overlook the warning signals thrown up by the process. Excluding the designated congressional overseers also excluded one more "political scrub," one more source of advice about what the American people would find acceptable. And the chances increased that someone like Lieutenant Colonel Oliver North would misguidedly interpret the President’s interest after his own fashion.

William Miller, the staff director of the first Senate Intelligence Committee a decade ago, reflected on the Iran/contra affair: "If clear lines hadn’t been drawn a decade ago, there would have been no hue and cry now." Drawing those lines again, sharply, as cautions for future administrations is about as good as we can do. If the congressional investigations complete the task begun by the Tower Commission, those future administrations will be on notice.

  • Gregory F. Treverton, a faculty member of Harvard University’s John F. Kennedy School of Government, served on the first Senate Select Committee on Intelligence (the Church Committee) in 1975-76, and on the staff of the National Security Council during the Carter Administration. This article is adapted from the author’s forthcoming book, Covert Action: The Limits of Intervention in the Postwar World. Copyright © 1987 by Basic Books, Inc. Adapted by permission of Basic Books, New York.
  • More By Gregory F. Treverton