Working Paper #12
Engineered Migration as a Coercive Instrument:
The 1994 Cuban Balseros Crisis
Kelly M. Greenhill
I. Introduction 1
II. The Mechanism Behind Engineered Migration 2
III. Castro’s Motivations 5
IV. Castro’s Objectives 6
V. Why Did He Think He Could Succeed?: A Compelling Track Record of 8 Two for Two
A. The Camarioca Crisis 8
B. The Mariel Boatlift 10
VI. The August 1994 Balseros Crisis 14
A. The Situation Heats Up; Castro Issues a Threat 14
B. The US responds, Castro escalates and the US grows defiant 14
C. A Domestic Spoiler Catalyzes a Major Policy Shift 15
D. Castro Ups the Ante 18
E. US Defiance Soon Replaced with Willingness to Negotiate 19
F. The US Blinks, but so does Castro 19
VII. The Balseros Crisis, Part II: April-May 1995 20
VIII. Was Castro’s 1994 Migration Gambit a Success? 22
A. Primary Objectives Attained 22
B. But No Movement on the Embargo or Radio Marti 23
C. Stymied Promises of Future Negotiations 24
IX. But the 1994 Refugee Gambit Was Less Successful than Mariel. Why? 25
A. Rapid Reaction of State and Federal Officials 25 B. The Use of Guantanamo Naval Base 25
C. Relative Passivity of the Cuban Exile Population 26
D. The November 1994 “Republican Revolution” 26
X. Conclusions and Recommendations 27
A. Option One: Play the Game, but with a Better Grasp of the “Rules” 27
B. Option Two: Make the Game Not Worth Playing by Eroding the 28
Engineered Migration as a Coercive Instrument:
The 1994 Cuban Balseros Crisis
Kelly M. Greenhill
In August 1994, in the wake of some of the worst civil unrest Cuba had witnessed in decades, President Fidel Castro reversed his country’s long-standing policy of arresting anyone who tried to escape the island by sea. Castro laid the blame for Cuba’s domestic unrest on the United States, claiming that the riots were caused by rumors of a US-sponsored boatlift to Miami. Castro then demanded “either the US take serious measures to guard their coasts, or we will stop putting obstacles in the way of people who want to leave the country, and we will stop putting obstacles in the way of people in the US who want to come and look for their relatives here.” This invitation, coupled with a threat, marked the beginning of a major, although short-lived, refugee crisis, during which several tens of thousands fled the island and headed north towards Florida. The crisis ended after about a month, following the announcement of a new immigration accord between the US and the Caribbean island nation, an accord that marked the beginning of the end of the US’s three decade long policy of welcoming all Cubans into the US as de facto refugees and the start of their being treated (at least on paper) like other groups trying to gain entry to the US; a follow-on accord eight months later solidified this policy shift.
Although dwarfed in size by the larger Mariel boatlift fourteen years earlier, the August 1994 crisis is important for several reasons. First, despite its brevity, it had far reaching consequences for US-Cuban relations. Without warning or preamble, the August 1994 crisis catalyzed a shift in US policy vis-à-vis Cuban immigration that represented a radical departure from what it had been for the previous three decades. Second, it influenced US domestic politics on the national level, by expanding the scope of the issue, mobilizing not only Floridians, but also the general public worried about the threat posed by illegal immigration. Third, the crisis illustrates the potential potency of engineered migration—i.e.,” people pressure”—as an asymmetric weapon of the weak. Consider that in less than a month Castro transformed an internal crisis that began with boat hijackings and a riot in Havana, into an American foreign and immigration policy crisis, in which the US was forced to provide refuge to tens of thousands of Cubans it intercepted at sea—at a cost of $1 million per day; and he got a new US-Cuban immigration accord to boot. Finally, the brief, but significant, interactions of international and domestic actors in the case warrant examination because, although the August 1994 crisis was limited in scope, in its dynamics it resembles a number of other refugee crises, both large and small. Thus some valuable lessons may be gleaned from it that could aid in dealing with future (real or threatened) crises.
Between 1980 and 1994, the US witnessed mass refugee influxes of individuals seeking asylum, from Cuba, Haiti, El Salvador, Nicaragua, and Guatemala in this hemisphere alone. As the September 1994 report by the US Commission on Immigration Reform put it:
US policy has tried to balance a number of competing interests and concerns: preserving its international and domestic commitments to provide asylum to those fleeing persecution or a well-founded fear of persecution and providing protection to aliens who would otherwise face return to dangerous conditions in a home country; deterring illegal immigrants who abuse the asylum system as a backdoor to entry; responding to domestic ethnic and political constituencies; ensuring that US policy does not serve as a magnet for otherwise avoidable mass migration; upholding foreign policy commitments; and helping states and localities faced with the costs and other impacts of dealing with immigration emergencies.
The fact that these interests naturally compete may unintentionally give potential sending states coercive leverage over the US (and other potential receiving states). This is one variant of what is referred to as “engineered migration.” Such migrations in which (real or threatened) outflows are used to induce changes in political behavior and/or to extract economic side-payments from a target state or states are, what this author terms “extortive engineered migration.”
This author argues that extortive variant of engineered migration is one of the relatively few weapons that weak actors can use to influence the behavior of more powerful states, particularly advanced industrial democracies. This author further contends that weak actors are frequently successful in wielding this weapon because they can manipulate the political vulnerabilities of leaders within these states, vulnerabilities that arise when there exists a conflict between a state’s international commitments and its domestic imperatives. Advanced industrial democracies are vulnerable to exploitation because, although they have made international commitments to human rights and refugee protection, at least some segment of their population is usually unwilling to bear the (real or perceived) domestic costs of fulfilling these commitments. For instance, as was noted at the time of the 1994 crisis, “(Castro) is also well aware that the United States, given its values and domestic political pressures, cannot afford to send back the ‘anti-social elements’ he is encouraging to leave.” At the same time, the vast majority of the US public was fervently opposed to letting Castro once more turn the state of Florida “into a gigantic refugee camp,”  particularly one that might once more house a variety of “undesirables,” including violent criminals and the mentally ill. Hence the US and other states like it—i.e., that share its values and commitments—may find themselves trapped between their international normative commitments and their domestic imperatives. When the options available to targets in the face of a massive outflow have negative (and visible) side effects, both with respect to national interests and normative commitments, targets may become vulnerable to coercion and ripe for extortion.
The analysis offered herein relies on a Putnamesque two-level game framework, meaning that one can only understand the outcome of a bargaining situation on the international level by being acquainted with what happened on the domestic level, and vice versa. Here again, the difficulty for leaders lies in the fact that moves that may be rational on the international level may prove untenable on the domestic level, or vice versa. This author contends that Castro—who is known to be a keen observer of US politics—understands well the dilemmas facing US policymakers trying both to satisfy their domestic constituencies without sacrificing international credibility and to solve international crises without creating domestic ones. Although Castro is not immune from such concerns himself, the nature and stability of his military dictatorship allows him to undertake potentially risky moves internationally with considerably less concern about possible domestic backlash. Hence—in any given refugee-driven bargaining game between the US and Cuba—Castro will be more credible than any US leader, making “people pressure” a relatively potent asymmetric weapon against the US.
The rest of this paper presents and defends the proposition that Castro’s decision to launch the August 1994 crisis was an opportunistic attempt to coerce the US to the negotiating table on immigration and a wider array of issues, while the precise timing of the crisis was driven by the events on the ground in Havana. It further contends that Castro’s gambit was relatively successful—as it had been twice before, most famously in 1980, but also to a lesser degree in the mid-1960s—because Castro was able to internationalize his own domestic crisis, transforming it into a US domestic political and foreign policy crisis. In presenting this argument, the paper first examines Castro’s possible motivations for, and objectives in, launching the crisis; second, it looks briefly at Castro’s past attempts to use refugees as weapons, to explain why he thought he would succeed; third, it traces the chronology of the crisis and highlights its consequences; and fourth, it outlines what Castro actually did and did not achieve with his refugee as weapon ploy. Finally, this paper offers some conclusions and identifies a few ways in which the US can reduce its vulnerability to future attempts at coercion via the use of “people pressure.”
III. Castro’s Motivations
Some observers have argued that Castro’s decision to open the port was simply an act of desperation, aimed at defusing the tensions on the ground in Havana and propping up his flagging regime, in the face of a major economic downturn and growing social unrest. It is certainly true that the prevailing strife served as the proximate cause of the crisis and influenced its timing. But if it were true that Castro was simply using the outflow as a release valve, we should have seen two things happen that did not, and we should not have seen two things happen that did. First, Castro should have opened the port without first warning the US that he was considering it. Second, he should have done so as soon as it became clear that the prevailing discontent would spill over into serious violence. Third, Castro should not have publicly demanded a shift in US policy as a precondition for staunching the flow. And fourth, he should have re-closed Cuba’s borders exactly when it suited him—which would not presumably have coincided exactly with the conclusion of a new immigration accord with Washington.
Instead, evidence suggests Castro’s actions were actually more strategic in nature, designed to influence the behavior of the government in the US as much as that of the dissidents in Cuba.
Many analysts and government officials who have spent time in Cuba and have dealt personally with Castro and the Cuban leadership share this view. For instance, Castro biographer, Tad Szulc, has said, “He has been doing this for a living for 35 years, and realizes he only has one card to play, the weapon of refugees. He needs the breathing space and knows that the only way to get it is to force the Americans into a dialogue.” Likewise, Richard Nuccio, former special advisor to the Clinton Administration on Cuba, contends, “the Cuban government exacted changes in the policy of the Clinton Administration towards Cuba by threatening and by carrying out those threats… Most of our Cuba policy is a result of those kinds of threats.” And former Florida governor and now US Senator, Bob Graham, argues, “Castro, over and over in the last 35 years, has used his own people as a means of accomplishing his foreign policy objective.” So if Castro intended to use the August 1994 migration surge to coerce a shift in US policy, exactly what did he hope to accomplish?
For some time before the 1994 crisis erupted, Castro had been complaining that the US was failing to hold up its end of a 1984 agreement he had negotiated with the Reagan Administration, which promised 20,000 visas per year for Cubans, in exchange for his willingness to take back a number of “undesirables” from the first Mariel crisis. As the Cubans understood the 1984 accord, 160,000 visas should have been granted in a period during which only 11,000 had been forthcoming.  In this same period, however, more than 13,200 illegal migrants had been welcomed into the US, many of whom reached US shores on vessels they had hijacked in Cuba. This supported Castro’s long time claim that for 35 years it had been US policy to encourage people to leave Cuba illegally, even if that meant stealing and hijacking. He further argued that even those people who used such violent means of escape were welcomed as “heroes in Miami.” Yet whenever he interfered with these illegal departures, he was accused of human rights violations; while each time he let people leave, he was accused of trying to embarrass the US.
Castro’s frustration apparently deepened in the summer of 1994, as it became clear that the reception rafters were being given in July and early August 1994 was “specially warm…. (even) after stealing boats, using violence, endangering the lives of people who did not wish to emigrate, and even committing murder.” Rafters (arriving in this period) were further reassured and “encouraged by the US government’s pledge not to change its immigration policy under any circumstances.” This was probably the tipping point that led Castro to consider initiating a new crisis. As one Latin American scholar put it: “Castro relaxed the strictures against emigration because he was “greatly (and understandably) amazed by US officials’ welcome to Cuban refugees who had hijacked ferry boats in Havana.”
In short, what Castro wanted was a quick end to the irregular and destabilizing pattern of immigration between his country and the US. Both his words and subsequent actions reflect this assessment, as the analysis below demonstrates. First and foremost, Castro desired a normalization of US-Cuban immigration and an end to the hijackings that were generating instability inside Cuba. (It also seems clear that he would have welcomed a loosening of the embargo, but evidence suggests he did not expect such a relaxation to be immediately forthcoming.)
Castro likely believed the migration gambit was worth trying, in part because it had worked at least twice before—in a limited way in 1965 and rather dramatically in 1980. As was the case in the 1980 Mariel I exodus and the less well-known Camarioca outflow in 1965, Castro succeeded in dictating the course and pace of events while those in Washington, working with far more resources at their disposal, struggled to respond. (See Figure I below.)
A. The Camarioca Boatlift
In September 1965, Castro surprised the exile community in Miami with the announcement that any Cuban who had relatives living in the US would be allowed to leave the island via the port of Camarioca, located on Cuba’s northern shore. Castro also invited exiles to come by sea to pick up family members who had been stranded on the island, following the suspension of commercial flights between the two countries during the Cuban Missile Crisis three years earlier. To erase any doubts that he was serious, two days later he began offering two flights daily from Havana to Miami.
At the time many alleged—rightly, this author believes—that Castro opened the border in order to rid the island of political dissidents with close ties to the exile community. As one author put it: “in one clean sweep, he release(d) the internal pressure of ‘closet counterrevolutionaries’” who stood ready to undermine his regime.” In addition, however, by unleashing his “demographic bomb,” Castro also demonstrated to the US government how easily he could disrupt its immigration policy. Thus the opening of the port at Camarioca carried with it a “lightly-veiled” threat, namely that Havana, not Washington, controlled Florida’s coastal borders. Almost overnight, and with little warning, the Cuban government had presented the US with a major refugee crisis.
President Johnson initially responded with contempt to Castro’s move, making a speech before the Statue of Liberty in October 1965, in which he proclaimed that the US would continue to welcome Cubans seeking freedom “with the thought that in another day, they can return to their homeland to find it cleansed of terror and free from fear.” However, after thousands of exiles
sailed toward Cuba—much to the chagrin of US immigration authorities—and the numbers of those leaving the island began to escalate, Johnson quickly changed tack and began a series of secret negotiations with Castro. The result, announced the following month, was a “Memorandum of Understanding,” a formal agreement that established procedures and means for the movement of Cuban refugees to the US. In December 1965, an open-ended airlift, which would continue until 1973, was inaugurated.
Because the Johnson Administration—preoccupied with Vietnam and fearing a tragedy in the Straits of Florida—was so quick to propose an acceptable solution to the crisis, Castro swiftly acquiesced, and the crisis ended with little immediate political cost to either side. However, Castro learned a valuable lesson from this migration crisis dress rehearsal: namely that the appearance of loss of control over US borders—coupled with the perception inside the US that Florida might be overrun—would be viewed by US leaders as politically costlier than the alternative of dealing with him. Thus if Castro could transform his own domestic problem into the US’s domestic problem, via the use of refugees as political weapons, he could coerce its leaders into helping him solve his own problems.
Then, in early April, following a dramatic incident in which a bus crashed through the gate of the Peruvian embassy in Havana, Castro announced that he would remove the security forces that surrounded the embassy. A few days later, in a case of “déjà vu all over again,” Castro publicly invited the exiles to come by sea and pick up, not just the refugees who had originally sought asylum at the embassy, but anyone who wanted to leave. Thus again Castro managed to transform his own domestic crisis into an American domestic and international crisis. Even at the time, some recognized this ploy as a weapon of sorts; for instance, the US Coordinator for Refugee Affairs, Victor Palmieri, characterized it as “a form of guerrilla warfare.” White House aide, Jack Watson, and State Department spokesman, Thomas Reston, echoed this viewpoint, saying respectively, “Castro in a way, is using people like bullets aimed at this country” and “what you have here is not a rational process [but] Castro’s solution to the problem.” 
Like Johnson before him, President Jimmy Carter began the crisis with a defiant speech, in which he reaffirmed the US’s open-arms policy to Cubans fleeing Castro’s regime, proclaiming “we’ll continue to provide an open heart and open arms to refugees seeking freedom from Communist domination.” However, a record number of refugees arrived at about the same time as Carter made his famous pronouncement, and within a week Carter felt compelled to change the US’s position. On May 14th, the Administration ordered the boatlift stopped; but to no avail. Despite the prohibition and a subsequent naval blockade, almost 90,000 Cubans arrived in the US in May 1980.
Aware of the “grave implications the uncontrollable boatlift could have on domestic politics,” the Administration lost patience with the exiles who kept returning to Cuba for more refugees. This was particularly true since Carter had already been forced to declare a state of emergency in South Florida and to release $10 million to help local governments cope with the crisis. When, in spite of these efforts, South Florida’s local infrastructure was overwhelmed, the federal government began airlifting the Marielitos to military installations throughout the country. As the crisis deepened but the exiles showed no signs of ceasing the boatlift despite government appeals, the Carter administration again flip-flopped and threatened exiles with heavy fines and confiscation of their vessels if they continued to bring refugees into the US; the threats fell on deaf ears.
Carter next turned directly to Castro for help in ending the crisis. But the Cuban regime promptly rejected the first American proposal for ending the crisis, viewing it as too harsh and insufficiently attractive from the Cubans’ perspective. Interestingly, it appears highly plausible that the outcome of Mariel could have been rather different, and the crisis itself significantly more short-lived. As the former head of the US Interests Section in Havana, Wayne Smith, noted in his memoir:
I had been on the Cuban deck back in 1965 when we had convinced Castro to replace the Camarioca sealift with an orderly departure process. In some ways, prospects were better in 1980 than they had been in 1965. Castro had initiated Camarioca without any prior expression of interest in establishing a normal flow of emigration. Yet, he had quickly closed down the sealift in return for a normal emigration process. This time, Cuban officials had been urging such a process long before the Mariel operation began.
Three more months would pass before the US made the kind of proposal the NSC had rejected as too placatory the previous spring, namely that the migration talks would be linked to a future (broader) agenda. Shortly thereafter, Castro closed the port of Mariel. Smith, Tarnoff, and others believe this would have happened much earlier had a more conciliatory proposal been forwarded and would have meant 100,000 fewer Cuban refugees would have come to the US.
It is instructive to note the disparity in the speed with which the Johnson and Carter Administrations each developed policy responses to their Cuban migration crises, as well as differences in their approaches. Because the policy makers in the Johnson Administration figured out quickly the potential scale of the problem, they developed a response within days after Castro announced the opening of the port at Camarioca. Within a month, officials in the Johnson Administration managed to provide states and localities with financial relief for the costs associated with the boatlift.  In contrast, even with considerable forewarning that Castro was considering re-opening his borders, Carter Administration officials took more than three weeks to generate a policy response, one that never adequately dealt with the crisis. Furthermore, it took the Administration almost two months to approach the Cuban government about talks to normalize immigration, and then the subsequent accord was not signed until after the boatlift was ended many months later. Finally, it was Congress, not the Carter Administration that (six months later) generated a policy to deal with the tremendous costs of Mariel to affected states and localities.
It is clear that because the crisis occurred during the presidential campaign, Carter absorbed the full backlash of voter indignation. He was blamed for his ineptitude in handling the crisis and for indecisive leadership, and his Republican challenger, Ronald Reagan, enthusiastically played the refugee issue to his advantage. In light of the other tribulations the President was facing in the lead-up to the November1980 election—including the Iran hostage crisis, the Soviet invasion of Afghanistan, and a floundering economy—it would be an exaggeration to claim that the Mariel crisis alone produced Carter’s defeat. Nevertheless, it provided very effective campaign fodder for Reagan and affected the psyche of the American public, including the psyche of one particular American, future President Bill Clinton, who would himself sit at the helm of the US during the next Cuban migration crisis.
The spring of 1994 brought scenes reminiscent of the Mariel boatlift fourteen years earlier. Between May and early August, Cuba became the site of an ever-increasing number of embassy crashings and violent boat hijackings. This violence culminated in street riots in Havana in early August, after 32 died when they were swept into the sea by water cannons when the Cuban military intercepted a tugboat bound for Miami.
Reading the signs of restiveness on the ground, and by now familiar with Castro’s modus operandi, at least some in the US believed Castro might try to initiate another refugee crisis. For example, Principal Deputy Assistant Secretary of State for Inter-American Affairs, Robert Gelbard, publicly warned Castro on July 30th that the “consequences of “launching another Mariel boatlift would be quite grave.” But Castro had been through this situation before and recognized that the US had more to lose from an uncontrolled outflow than did he, particularly since he was again eager to engage the US in negotiations. Thus, frustrated by the hijacking and escalating illegal departures—and undeterred by US threats—on August 5th Castro held an internationally televised news conference, in which he asserted that the rioting was caused by rumors of a “US sponsored boatlift to Miami.” Castro went on to say that Cuba could no longer afford to be “the guardian of the North Americans’ coasts” if Washington continued to strangle the faltering Cuban economy.
Immediately following Castro’s pronouncements, the US responded with clear signs of encouragement for those who wanted to escape and announced the existence of a classified contingency plan, Operation Distant Shore, designed to thwart Castro’s ability to launch another Mariel. (This announcement seems to have been an attempt to deter Castro, as well as to reassure an anxious population in Florida.) Although the details of Distant Shore remained classified, it was officially announced that the plan included responsibilities for 40 different federal agencies that would respond to an immigration crisis and a blockade by US warships of the Florida Straits, as well as the arrest of any refugee trying to enter the US through that route.
Nevertheless those intent on fleeing the island were not deterred, and neither was Castro. Around August 12th he announced that he would view any attempt at instituting a blockade as an act of war, and he quietly began escalating the crisis by allowing people to leave the island unharrassed. On the same day, US State Department officials announced that there was no sign that Castro had yet opened his coastline to unrestricted exits, but conceded that the Cuban coastal and land police were letting small groups leave without incident. To those paying attention, this was a clear sign that Castro could and would control the volume of the flows as he saw fit, a portentous signal recognized by those on the frontline in Florida, but not yet by those in Washington, who still insisted that the situation was under control and did not constitute a crisis.
However, this situation was to change quickly and unexpectedly. On August 18th, “in a matter of twelve hours…the Clinton Administration’s view of the influx of Cuban refugees changed from a manageable, orderly flow to a crisis demanding a reversal of 28 years of immigration policy.” Significantly, however, this shift did not result directly from a move in Havana, Cuba, but rather from one in Tallahassee, Florida. Facing a tough re-election campaign—in a state where immigration was an especially highly charged issue—Florida Governor Lawton Chiles decided he was unwilling to concede to a potential repeat of the Mariel fiasco without a fight. Despite Washington’s skepticism, Chiles believed that the flow of refugees would blossom into a flood, and he demanded that the federal government take action. On the 18th of August he went public with his criticism and an implicit demand:
Well, I think your numbers showed that we’ve had 2,200 [Cuban asylum seekers] already this year. But the interesting thing is this month. The interesting thing is 565 yesterday, 360 today. As we speak, they are still getting off the boat down there [in Key West]. I think we might well have 500 again today. In spite of the Coast Guard captain’s statement, the most we ever had in a day in Mariel was about 856. So we’re already up to 500 a day. Florida could die from a thousand small cuts and that’s what Castro is doing to us. This is an emergency down here. We know that, all the citizens of south Florida know that and we’re waiting for the administration to know that.
Chiles had concluded that a hard position against the rafters would help him in his re-election bid, and polling data from the period suggest it was a wise surmise. Most Floridians were opposed to the influx, and those who were not, namely the Cuban-American community, were expected to vote for his Republican challenger, Jeb Bush, in any case.
The situation was further complicated by the fact that it had just become clear that a key component of the contingency plan that was to relieve the burden on Florida in the event of another Mariel-like exodus had been rejected out of hand by Clinton, leaving Florida even more exposed.  Specifically, when he discovered—around August 16th—the nature of the State Department’s proposed resettlement plan (Distant Shore), under which Cuban asylum seekers would be distributed at military bases throughout the country, he reportedly “went ballistic,” yelling “Are you nuts? Do you think I am going to do [that] again?” Other advisors confirm that Clinton’s thinking on the August 1994 crisis was guided by two mottos, “No More Mariels” and “Remember Fort Chaffee.”
What Clinton feared was a repetition of the personal humiliation and defeat he suffered after the last massive Cuban refugee resettlement in 1980. Dissatisfied with their long-term detainment, the Marielitos sparked violent riots at several of the military bases where they were being held, including Fort Chaffee. Shortly thereafter, then Arkansas governor Clinton lost his bid for re-election. Although, like Carter, Clinton too may have lost even in the absence of the Cuban crisis, Clinton clearly laid blame for his loss on the Chaffee riots. As advisor and confidant, Dick Morris, has said, his defeat in 1980 “was really the seminal experience in (Clinton’s) career.”
By August 18th, when it had become clear to Chiles that the Administration was willing neither to recognize the escalating crisis as an emergency, nor to consider implementing Distant Shore’s proposed resettlement plan, he decided to press Washington’s hand. Chiles declared a state of emergency in Florida, which gave him the right to mobilize the Florida National Guard and to detain temporarily the refugees released by the INS. Chiles made it known to the Administration that he would not permit Cubans being moved from detention camps in Key West to get off the bus once they reached Miami; instead he would have them arrested and quarantined.
That afternoon a "principals-only" meeting of many of Clinton’s top foreign policy advisors was held, during which the decision was made to end the 28-year policy unequivocally welcoming all refugees from Cuba. It appears that the general consensus was that it was time to “demagnetize” the US to avoid a continuous flow of refugees. According to one participant, “the change was necessary to protect a basic fundamental policy of no massive influx that looked like Mariel.” They also agreed that the US could not appear to be tougher on the Cubans than it was on the Haitians.
Although on the morning of August 18th Attorney General Janet Reno had insisted that no change in policy was under consideration—and that Chiles was “overreacting”—that same evening a new approach was announced. Following Chiles’s declaration of a state of emergency in Florida, and a subsequent meeting between Clinton, Chiles, and Jorge Mas Canosa of the Cuban American National Foundation (CANF), Clinton announced a major shift in US policy: no Cubans seeking to enter the US illegally would be allowed to enter US territory. Instead they would be rescued at sea and detained at the US naval base at Guantanamo Bay, Cuba (Gitmo) indefinitely. This shift represented a reversal of the 35-year-old policy designed to welcome as a refugee any Cuban “escaping” Castro’s regime.
Although it is clear that Clinton and his advisors were themselves clearly disinclined to permit a repeat of Mariel, it was Governor Chiles’ initiatives—coupled with Clinton’s steadfast opposition to a domestic relocation scheme—that actually forced the president’s hand and drove the policy shift. In making such a momentous change, Clinton knew that he would have to contend with the animus of the Cuban exile community, who would be furious that the US would consider returning the fleeing Cubans. But even at the time, risking the shift in policy probably seemed like a good gamble for a number of reasons. First, like Chiles, Clinton realized that the vast majority of Floridians, and Americans more generally, were opposed to accepting more refugees, whatever their origin. Basically, “the new calculus [was] that Clinton need[ed] to worry more about immigration than about Cuban-American votes.”
Second, he had a plan to placate the CANF and the exile community, namely by offering to tighten restrictions on Cuba. Following his meeting with Chiles and Mas Canosa, on August 20th Clinton announced that visits to Cuba would again be restricted to humanitarian cases and remittances would be suspended. However, although this compromise temporarily satisfied his conflicting domestic constituencies, it did nothing to bring the crisis itself closer to a resolution; in fact, it led Castro to escalate the crisis further. 
While undoubtedly pleased that illegal rafters would finally be detained and returned, Castro was clearly less excited about the measures taken to conciliate the exile community. So the next day, he opted to escalate the crisis further by officially opening the borders to anyone who wanted to leave. Moreover, because the Cuban public remained unpersuaded that the US’s three decade old policy of welcoming all Cubans really had changed overnight, the announced shift did not slow the flow of rafters. Three days later, in the largest one day total ever, 2,886 rafters were intercepted at sea, while the day before, 2,338 had been rescued.
Thus the Administration’s gamble that Cubans would stop fleeing once it was announced that they would not be allowed to enter the US had proven a serious miscalculation, one that Castro promptly exploited. On August 24th, Castro gave an internationally televised speech on CNN, announcing that the US’s “new policy measures only [made] the problem more complicated... [and] ... these measures [compelled] the massive exodus.” During the same speech, he also officially announced that he had ordered the Cuban Coast Guard to stop impeding those who wanted to leave the island and/or using force to prevent Americans who wish to pick people up.
At the same time, however, Castro also intimated that he might agree to stop the exodus if the Administration agreed to direct talks on a range of issues, including the embargo. This position was reaffirmed the next morning, when Cuban representatives in New York announced their willingness to end the flow “only if the US agreed to broad talks on a full range of bilateral matters.” Meanwhile Cuba’s Ambassador to the United Nations turned the rhetorical heat up even higher, warning that the US’s new strategy would lead to disaster, both for Cubans and Americans: “The US has devised a whole policy…to try to choke our country to hunger and allow an internal subversion that would lead to a blood bath, and then how many millions of illegal immigrants will come?”
By the following day, August 26th, there was growing public bipartisan opposition in Congress to the Administration’s unwillingness to negotiate with Castro, as well as a growing number of news commentaries and newspaper editorials calling for negotiations. Moreover, on the same day the New York Times ran a story announcing that the camps housing the Cubans at Gitmo would be filled to overflowing within two weeks if the exodus were to continue. Thus Clinton’s attempt to satisfy both sets of his domestic constituents had given birth to another dilemma. Because the rafters kept coming, the only way to end the crisis was to rely upon the Cuban government to again begin blocking emigration. But the new sanctions against Castro and the US’s staunch unwillingness to negotiate proved a powerful disincentive for Castro to do so.
On August 28th, with no indication that the crisis would end of its own accord or that the flows might soon abate, the US again abruptly changed its position and agreed to negotiate with the Cubans. A series of bilateral talks were held between September 1st and 10th that resulted in the announcement of a new immigration accord and plans for a series of additional meetings.  However, for his part, Castro had to concede his demands that the accord be linked to a softening of the embargo and/or to Radio Marti’s shutdown.
In early April 1995 the Cubans again began making vague threats to reopen their borders—a rumor that the Administration leaked privately, but publicly denied. (The belief was that the renewed threats were a response to the proposed Helms-Burton legislation and to the fact that those Cubans still being held at Gitmo were being denied entrance into the US.) One Cuban official told The Washington Post, “last year, there were 30,000 rafters. Next time you might see a million.”
Moreover, following their trip to Gitmo in March-April 1995, Sen. Bob Graham (D-Fla.) and Rep. Porter J. Goss (R-Fla.) warned the Clinton Administration of a crisis in the making. They claimed that the thousands of Cuban refugees detained in Gitmo were living in a "tinderbox" that could explode into rioting.  In addition, housing the refugees at Gitmo and in Panama for six months had already cost more than $400 million, and the Pentagon was planning to spend $100 million more to make the camps permanent. 
The warnings set off what officials called "serious alarm bells" in the White House, in part because the administration was “poised to enter a critical and enormously tricky domestic policy stretch—a summer of high-stakes battles with Republicans over the size, shape and cost of government that could well define the 1996 presidential race.” One thing Clinton officials did not need was a refugee crisis of any sort, and reportedly his top aides quickly concluded that another round of serious talks with Cuba was in order. "We were facing a double whammy when all we want is to keep foreign policy problems off the screen," said one official. "The word was: Solve it. Make it go away with the least amount of turmoil."
Like Johnson and Carter before him, faced with the dilemma of choosing between the domestic political costs of another refugee crisis and those associated with further negotiations with Castro to avoid one, the Administration chose the latter. Two weeks later, on the anniversary of the Bay of Pigs debacle, Ricardo Alarcon and Under Secretary of State Peter Tarnoff met in secret—likely to shield themselves from domestic political pressure—and the new accord was announced on May 2, 1995. Eight months after initially refusing to admit those at Gitmo, the Administration had again changed course and agreed to admit them on a case-by-case basis. With the policy shift came the first official reference by the US—by Attorney General Janet Reno—to the Cuban migrants as “illegal immigrants” rather than “political refugees.”
In addition, on the heels of the new accord’s announcement came word that the Clinton Administration would oppose the embargo-tightening Helms-Burton legislation, and that this new policy “could be followed by engagement on other areas of mutual interest, like the fight against narcotics or environmental problems.” According to a White House paper, “‘[The US was] prepared to reduce sanctions in carefully calibrated ways, in response to significant, irreversible changes in Cuba.” For its part, the CANF called the policy decision “a second Bay of Pigs.”
Yes, albeit a qualified success. As in 1965 and in 1980, after initial resistance, the US was forced to come to the negotiating table with Cuba; and Castro did accomplish what are widely regarded as his primary objectives. However, progress vis-à-vis his purported long-term goals was far more modest, at least explicitly.
As one keen observer put it at the time, “through blackmail Castro has (again) been able to change US policy.” As a consequence of the crisis, Castro did achieve what analysts regard as his key aims: namely, a US-backed halt to illegal emigration and the prosecution of Cuban hijackers. The agreement provided that the US would accept 20,000 Cubans per year plus an unspecified number of family members, and the 4,000 to 6,000 Cubans on the waiting list for visas would be permitted to enter the United States. This marked—albeit imperfectly—the official end of illegal immigration between the US and Cuba and was in essence a reaffirmation of the promises made to Castro by the Reagan Administration a decade earlier. Second, the US agreed to extradite and/or prosecute those who hijacked or stole boats and aircraft to flee Cuba, thus expediting a “safer, legal and more orderly process” of immigration. Castro had been pressing the US for years to concede these two points. The US also agreed to work on bringing down the backlog of people who had applied through the American Interests Section in Havana to emigrate legally over the previous decade, i.e., about 140,000 people. In exchange, Cuba promised to end the boatlift, using “mainly persuasive measures” to crack down on those who tried to emigrate illegally and to take back 226 Cuban boat people being held at Gitmo who had asked to be repatriated.
Castro did not make any explicit gains with respect to ending the embargo or silencing Radio Marti, two things his representatives began calling for publicly in the days leading up to the September meetings.  Nevertheless, it can be argued that the reason Castro conceded to shelve these issues was that, while he expected that he could get an agreement on immigration issues in the short run because of the visibility of the crisis, the more substantive issues of the embargo
would require more time and wider support, particularly given that it was late in an election year, a fact that would not have been lost on the Cuban leader. Consider, for instance, that shortly after the crisis ended, Castro met with former senator and US presidential candidate, George McGovern, who said:
You would be impressed with his knowledge of American politics. He knows all the American players, and he knows the pressures that play on them. He knows all about the health care debate and the crime bill and Whitewater and everything else that’s going on here and showed real sensitivity to the political squeeze that the President’s going through now.
Furthermore, former Cuban insiders concur that Castro is a keen observer of US domestic politics. As one former Cuban official who spent 17 years in the revolutionary elite notes, “Fidel is a shrewd student of United States society, institutions and government,” and he “understands the limitations on a president’s power to act in many critical circumstances. This knowledge informs his every strategic maneuver.” Moreover, McGovern indicated that Castro acknowledged explicitly that Clinton “was politically incapable of tackling anything as controversial as lifting the embargo in the short run, particularly in the wake of the refugee crisis which was a matter of enormous embarrassment and anxiety to the US administration.” In the end, despite the Cubans’ eleventh hour calls for discussions on issues wider than immigration, it appears Castro probably got everything that he expected to get, at least in short run. However, this is not to suggest that he did not actively float trial balloons on the bigger issues, in the hope that they might produce results, only that he likely had low expectations that they actually would.
Castro may have expected there to be more dialogue and further positive developments down the road. Otherwise, it seems unlikely that he would have agreed to end the crisis so quickly. As one Cuba analyst put it: “It's unthinkable that this was a rare moment of Castro charity at work…. He had such leverage over Washington. He was in the position of either saving Clinton's political neck or causing him endless problems.” In addition, circumstantial evidence indicates the existence of a tacit agreement that future negotiations could occur. Shortly after the September accords were announced, Secretary of State Warren Christopher appeared on “Face the Nation” and said of Castro: “If he moves toward democracy in a tangible, significant way, we'll respond in a carefully calibrated way…Washington is ‘to be prepared to reduce the sanctions in carefully calibrated ways in response to positive developments in Cuba.’” Although Christopher declined to specify what either these developments or responses might be, State Department officials indicated the Administration “might ease—but not eliminate—economic or travel restrictions against Cuba if Castro allowed more freedom of expression or free elections.”
More concretely, then special advisor to the Clinton Administration on Cuba, Richard Nuccio, reports that following the migration crisis in 1994 and the subsequent May 1995 accords,
a weak, and I’d emphasize weak, conditional engagement policy was added to the prior unconditional engagement policy towards Cuba. By this conditional engagement policy, an explicit understanding was arrived at between senior US and Cuban officials that Cuba’s implementation of the May 1995 migration accords and its reaction to the US efforts to engage Cuba’s emerging civil society could form the basis for further progress in US-Cuban relations.
Reportedly, following the May 1995 accord, the Administration envoys were so encouraged that they approached members of Spain’s Socialist Party to help mediate further talks with Castro. And in late 1995, Castro met with US Congressman Esteban Torres (D, California), during which Castro reportedly agreed to call for free elections, permit the creation of opposition political parties and free political prisoners. In exchange, the US was to lift the embargo and help Cuba obtain international development bank loans, according to Congressional sources.
These developments (if even real) came to a screeching halt—at least temporarily—in February 1996 when Castro ordered his military to shoot down two unarmed planes flown by Brothers to the Rescue, a Cuban-American exile group. In Washington, the shoot-down outraged conservatives and panicked the Clinton Administration, which was in the midst of the 1996 Presidential campaign. Clinton responded by quickly throwing his support behind the Helms-Burton legislation, which as noted previously he had theretofore opposed.
IX. But the 1994 Refugee Gambit Was Less Successful than Mariel. Why?
The 1994 crisis was of much shorter duration and had much lower domestic visibility than did the Mariel I crisis in 1980. There were several reasons for this. First, having learned something about the dangers of reacting slowly, Florida state officials and the US Coast Guard responded relatively quickly and decisively. To block Cubans from reaching the US, the US Navy and Coast Guard diverted more than 70 ships usually used for interdicting drugs and patrolling fisheries or tending buoys. More than 8,500 Coast Guard and Navy personnel—both at sea and on shore—took part in what military officials said was one of the largest rescue operations of its kind. The operation included 350 Marines on board Coast Guard vessels to provide security, as well as a variety of A/C-130 and H-60 Jayhawk helicopters, acting as “a bucket brigade of sorts.” Cutters and patrol craft would intercept rafters and transfer them to 10 larger Navy ships (mainly frigates), which made the 36-42 hour trip to Gitmo, thus allowing the Coast Guard boats to stay in the Straits. Nevertheless, this effort was not as effective as it might have been, given that the contingency plan was abandoned in the middle of the crisis, because of Clinton’s concerns about a repetition of “Fort Chaffee.”
The Administration’s policy of interdicting Cubans leaving on rafts and boats and transporting them to Gitmo quieted domestic discontent and made the crisis far less visible to the US public, which lowered the domestic political costs of the crisis. Nevertheless, the tough new policy did not result in the deterrent effect the Administration had hoped for. Two weeks after the policy announcement, the US Coast Guard was still rescuing and shipping off to Gitmo over 1,000 Cubans per day. It was only after Castro agreed to close the border that the crisis ended. So Gitmo’s effect should not be exaggerated, but it did give the Administration a bit of additional breathing room and allowed it to avoid more serious political consequences.
Unlike during Mariel I, relatively few Cuban exiles traveled to Cuba to pick up relatives and friends. Virtually all of those who reached the US during the 1994 crisis came under their own power. The exile community had several reasons for not directly participating in the exodus. First, Coast Guard officials promised powerful and painful sanctions would be levied against those who violated the blockade; and this time, unlike in 1980, they were taken seriously. Second, Castro’s tactic during Mariel of loading exile boats with complete strangers—many of whom were criminals and/or mentally ill—undermined efforts to mobilize the exile community’s support during the 1994 exodus. They wished neither to welcome more such people into the exile community, nor to be so explicitly used as pawns by Castro; so many stayed home. Third, following Mariel, many in the exile community felt that those who were fleeing were not political refugees, but rather economic migrants, and they were disinclined to facilitate their migration. Finally, it is also possible that the some in the exile community finally realized that by aiding the flight of those who opposed the regime they were deflating the pressure to remove Castro.
Third, the size of the Republican Congressional victory in November 1994 may well have precluded some of the diplomatic openings expected by Castro from materializing. For example, it was reported that during the summer of 1994, then National Security Advisor Anthony Lake said privately that he was prepared to recommend that Clinton lift the embargo and accept the political consequences. But the November election results put that “tightly held strategy on ice,” according to a senior Clinton Administration official. When asked thereafter about the probability that the Administration would take “bold steps on Cuba policy,” the official said, “That’s not who we (the administration) are.” Further, in March 1995, when NSC officials told reporters that they were about to recommend dropping the additional sanctions—namely, the prohibitions on remittances and family visitations—that Clinton had imposed during the height of the August crisis, the proposal was immediately attacked in Congress as capitulation to Castro and promptly abandoned. Finally, as noted above, the Republican electoral victory installed Senator Jesse Helms (R, North Carolina) as chairman of the Senate Foreign Relations Committee and Representative Dan Burton (R, Indiana) as chairman of the House Subcommittee on Western Hemisphere Affairs. With support from Helms and Burton, Cuban lobbies mobilized to tighten economic sanctions on Cuba; the resulting Helms-Burton legislation was designed to stop foreign investment in Cuba and, if possible, to damage Cuban trade. Although the Administration initially opposed the bill, following the Brothers to the Rescue shoot-down in February 1996, Clinton threw his support behind the legislation.
There is some irony in the fact that the 1994 crisis probably could have been avoided if the US had not flatly rejected Castro’s initial calls for negotiation. The kind of immigration “escape valve” Castro prefers is one “that is orderly and drawn out and not very splashy.” Orderly and splashless negotiations might well have generated little more than a blip on the public’s radar screen, thereby avoiding both a domestic and an international crisis. As has been noted before: history indicates “each and every president [from Eisenhower to Clinton] came to the conclusion that an important aspect of his Castro crisis [in the end] required negotiations.”
On the other hand, choosing to concede to Castro’s threats as soon as he makes them obviously generates its own costs, in reputation and credibility, which could encourage him to threaten the US ever more frequently and with increasing demands. However, more careful monitoring of the prevailing economic and social conditions situation on the ground in Cuba, coupled with more aural acuity if and when Castro begins making threatening noises, could lead to earlier diplomatic intervention, which could stave off future embarrassments. (It is worth noting that with few recognized exceptions—including the one in April 1995 discussed above—Castro has usually followed through on his publicly articulated threats to open the island’s borders. In other words, heretofore such threats have rarely been idle.) As Engstrom notes in his analysis of the Carter Administration’s response to Mariel:
The ahistorical approach of policy makers in the Carter Administration is particularly telling because the Camarioca boatlift provided tailor-made examples of the conditions that contributed to an earlier boatlift and the policies employed by the Johnson Administration to deal with it. The Camarioca boatlift offered relevant lessons that the Carter Administration did not explore. Had policy makers examined the dynamics of the Camarioca boatlift either before or during the Mariel boatlift, they may well have learned from history and developed better policies.
This assessment remains equally valid today; one need only change the names and dates.
Through a careful combination of public policy, education, and generous side-payments, the US (and other targets) may be able to reduce their vulnerability to extortive engineered migration, by undermining, or at least diminishing, the perpetrators’ ability to use “people pressure” as a weapon. In short, if a target can keep a migration surge from either being perceived as or actually becoming a crisis for the target, it can significantly degrade the perpetrator’s “weapons’ capabilities”. This may not be easy as both perpetrators and interested outsiders have powerful incentives to keep outflows visible and perceived as problematic.
However, potential targets can take a few concrete steps that may mitigate, if not eliminate their vulnerability, although several measures are potentially quite expensive. First, they can develop and be prepared to implement comprehensive and politically acceptable contingency plans, both to actively cope with the emergency and to help prevent the local infrastructure(s) from being overburdened. Specifically, to be better prepared the government should not wait for a crisis period to court communities that could be persuaded to take migrants—either for short or long-term stays—in exchange for attractive economic and/or political compensation packages. These plans would likely require copious financial support for those communities affected. Second, target goverments can launch education campaigns with the goal of teaching the public the real economics of immigration, emphasizing that over time immigration generally results in a net-gain for most industrialized countries. (This may be particularly true for those countries facing declining birth rates and aging populations.)
Third, they can actively cultivate the support of other states that could aid in burden sharing. (Again, if possible, the time to pursue such support is before a new crisis erupts.) For example, had all of the13 Caribbean and Central American states that the US approached during the 1994 crisis agreed to accept several thousand people, Castro’s gambit might have failed. However, but for Panama, very few agreed to take refugees, even when offered side-payments;  in any case, this kind of organized bribery could become a very expensive proposition that might still fail. Nevertheless, countries seeking foreign aid and/or other forms of international support might be persuaded to cooperate under the right conditions. (It is worth remembering that, early in the Kosovo crisis, an escalating refugee crisis on the Macedonian border at Blace was solved in just this manner.) Finally, targets may choose to abrogate the norm, either by claiming national security concerns or by refusing to recognize those fleeing as refugees. Targets who choose to do so may suffer significant reputational and hypocrisy costs, although under certain conditions they may deem those preferable to the domestic political costs of accepting more migrants.
 The Rosemary Rogers Working Paper Series, and the studies upon which they are based, are supposed by a generous grant from the Andrew W. Mellon Foundation.
 Please address correspondence to Kelly M. Greenhill, Center for International Studies, M.I.T., E38-600, Cambridge, MA 02139; email: firstname.lastname@example.org. I wish to thank the Mellon Foundation and the John F. Kennedy Library Foundation for their generous financial support of this research. For helpful comments and suggestions, I thank the seminar participants from the MIT-Mellon Program on NGOs and Forced Migration, and especially Sharon Russell and Karen Jacobsen of the Program’s Executive Committee. I also wish to thank Thomas Christensen, Eugene Gholz, Charles Keely, Barry Posen, Jeremy Pressman, and Stephen Van Evera—as well as those students who attended my colloquium—for their comments and their feedback on the larger project of which this research is but a piece.
 Fidel Castro, “La razon es nuestra: Comparencia de Fidel Castro en la TV cubana y las ondas internacionales de Radio Habana Cuba,” (televised speech), Editora Politica, 1994. See also Felix Masud-Piloto, From Welcomed Exiles to Illegal Immigrants: Cuban Migration to the US, 1959-1995 (London: Rowman and Littlefield, 1996), Chapter 9, especially pp. 137-44.
 Masud-Piloto, p. 128. For a discussion of post-Cold War US-Cuban relations, see, for instance, Gillian Gunn, Cuba in Transition: Options for US Policy (New York: Twentieth Century Press, 1993); and Carollee Bengelsdorf, The Problem of Democracy in Cuba (New York: Oxford University Press, 1994).
 Some scholars warn against expansion of the definition of a security threat beyond things that deal explicitly with “the threat use and control of military force.” From Stephen M. Walt, “The Renaissance of Security Studies,” International Studies Quarterly, vol. 35, no. 2 (June 1991), pp. 212-13. However, by 1994 both government officials and the public (as well as many academics) conceived of uncontrolled immigration as a security threat. Because threat perception is as important as actual threat, for better or worse, refugee flows need to be treated as genuine security threats. Consider, for instance, that in 1994—the first time the question was asked—72% of Americans said that “controlling and reducing illegal immigration” was a “very important” foreign policy goal, surpassed only by preventing the spread of nuclear weapons-82%, protecting US jobs-83%, and stopping the flow of illegal drugs-85%. From Ole R. Holsti, Public Opinion and American Foreign Policy (Ann Arbor: University of Michigan Press, 1996), p. 77. Also in 1992—even before the Balkans outflows began in earnest—35.5% of Americans said that mass migrations would pose an “extremely serious or very serious” threat to US national security in the 1990s. Ibid., p. 177.
 For a slightly different (pre-1994) view of the issue, but one that also asserts “the Cuban government, time and time again, has forced the US government to surrender to Cuba some US sovereign prerogatives to set US immigration policies,” see Jorge I. Dominguez, “Cooperating With the Enemy? US Immigration Policies toward Cuba,” in Christopher Mitchell (editor), Western Hemisphere Immigration and United States Foreign Policy (University Park: The Pennsylvania State University Press, 1992), pp. 31-88. For a discussion of the 1994 crisis and how Castro used Cuban migrants as bargaining chips, see Human Rights Watch, Repression, the Exodus of August 1994, and the U.S. Response (New York: HRW, 1994).
 This includes other US interactions with Cuba, as well as other crises throughout the world. See, for instance, Kelly M. Greenhill, “People Pressure: The Use of Refugees as Political and Military Weapons in the Kosovo Conflict,” in George C. Thomas (editor), Yugoslavia Unraveled: Sovereignty, Self-Determination, and Intervention , forthcoming.
 US Commission on Immigration Reform, US Immigration Policy: Restoring Credibility (Washington, DC: GPO, 1994), p. 164.
 It should be noted that the target(s) of extortion might not be synonymous with the refugee recipient state(s); in such cases, this author refers to these (actual or potential) host states as third parties.
 This conflict between international and domestic commitments has given rise to a new class of symbolic political costs, what this author terms “hypocrisy costs,” defined as “symbolic political costs that arise when there exists a real (or perceived) disparity between a professed commitment to international norms and agreements, and demonstrated actions that contravene such a commitment;” these are distinct from but exist in parallel with traditional reputation costs. Hypocrisy costs tend to be positively correlated with the degree of public awareness, or “issue salience” that any crisis engenders. So if no one is paying attention, no costs may result. If a crisis is very visible and has apparent and material consequences, hypocrisy costs will likely be very large.
 For instance, rising welfare costs, unemployment, crime, and cultural and/or ethnic fragmentation and dilution.
 Larry Rohter, “Castro plays his cards with cunning,” St. Petersburg Times, September 4, 1994, p. 1D.
 “The Cuban Tide is a Flood,” Newsweek, May 19, 1980.
 See newspaper and television coverage of both the Mariel (April-October 1980) and the balseros crisis (August-October 1994 and February-June 1995).
 That these negative effects need to be visible for a meaningful dilemma to exist is a key point. Consider, for instance, Robert Pastor’s musings on Clinton’s conflicting policy prerogatives: “If a crisis did not engage the public or attract the media or an interest group, it was ignored. The administration, however, could not afford to ignore intense objections from a core constituency group….(and) it was wary of antagonizing a Florida ethnic group…unless a broader national domestic interest (e.g., halting an influx of refugees) forced a recalculation.” From Robert A. Pastor, “The Clinton Administration and the Americas: The postwar rhythm and blues,” Journal of Interamerican Studies and World Affairs, vol. 38, no. 4 (Winter 1996-97).
 See, for instance, an explication of the policy dilemmas facing the US vis-à-vis Cuba during this crisis in “Testimony of Gillian Gunn,” Capital Hill Hearing Testimony (before the House International Relations Committee), February 23, 1995.
 Putnam argues that in any international bargaining situation, national leaders are engaged in two sets of negotiations simultaneously. First is the international negotiation, wherein the leader seeks to reach agreement with another international actor or actors; and second is the domestic negotiation, in which the national leader must persuade his domestic constituents to accept the international agreement. See Robert Putnam, “Diplomacy and Domestic Politics: The Logic of Two-level Games,” International Organization, vol. 42 (Summer 1998), pp. 427-60.
 For a different application of Putnamesque logic to US-Cuba relations, see William LeoGrande, “From Havana to Miami: US Cuba Policy as a Two-Level Game,” Journal of Interamerican Studies and World Affairs, vol. 40, no. 1 (Spring 1998), pp. 67-87. The argument presented herein is methodologically consistent with that offered in LeoGrande’s own analysis of the 1994 crisis, although he focuses principally on the use of sanctions by the US as a form of implicit coercive diplomacy against Cuba, and on how the US bargaining position has changed over time. He ignores completely the more explicit use of “people pressure”-driven coercion by Cuba against the US. Thus, like LeoGrande, the analysis herein relies on a two-level game framework and hypothesizes that the outcome of the international negotiation was driven in large part by domestic imperatives, but focuses attention on the bargaining power that issues from the other side of the US-Cuba equation.
 For some evidence of the tenacity and stability of Castro’s regime, see footnote 21 below.
 Of course, overall the US remains immeasurably stronger. So if a refugee crisis were really major—say, involving many hundreds of thousands or more—the US might entertain an invasion of Cuba, rather than further concession to Castro. However, the predicted material (i.e., blood and treasure) and political costs of invasion are still regarded as significant, although the Cuban military today is a shadow of its Cold War self. For an analysis of Castro’s post-Cold War military capabilities, see P. G. Walker, “Cuba’s Revolutionary Armed Forces: Adapting in the New Environment,” Cuban Studies, vol. 26 (1996), pp. 61-74. See also Jorge Dominguez, “US-Cuban Relations: From the Cold War to the Colder War,” Journal of Interamerican Studies and World Affairs, vol. 39, no. 3 (Fall 1997), pp. 49-75.
 The idea is that Castro chose to open an internal pressure valve, in response to growing dissent and economic pressure, to try and shore up his regime. However, those who hew to this explanation usually argue that Castro’s regime was close to collapse in this period, which was why he had to open the port. (See, for instance, El Nuevo Herald, August 6, 1994, p. 1b.) But most analysts—inside and outside the US government—discount this assessment, noting that “Cuba’s repressive apparatus (was) still efficient and loyal, and the ruling group remain(ed) unified.” From “Testimony of Gillian Gunn.” Further, as Cuba expert Jorge Dominguez put it at the time: “Any policy based on the idea that Castro is about to fall in the next few weeks is misguided…Bill Clinton could very easily reach the end of his presidency still waiting [for Castro to topple]” From Dick Kirschten, “What next for Cuba? Stay tuned” National Journal, vol. 26, no. 36 (September 3, 1994). See as well CNN interview with Jorge Dominguez, July 30, 1994; David LaGesse, “’Castro’s No Pipsqueak,’ Never Was: ‘Relic’ of the Cold War Still Gets America’s Goat,” The Arizona Republic, September 6, 1994, p. Al. As political psychologist, Jerrold Post, put it, at the height of the crisis, “This may not be a man who will be willing to go quietly into that good night…(even now, he retains) an electrifying chemical connection with his people.” From Bill Lambrecht, “A Policy Adrift: Suddenly It’s Cuba Demanding Action, and Clinton Facing Another Dilemma,” St. Louis Post-Dispatch, August 28, 1994, p. 1B.
 After all, by the summer of 1994 Cuba was an economic mess. It was suffering significantly following the collapse of the Soviet Union—e.g., its gross domestic product (GDP) declined 35% between 1989 and 1993—and the end of the massive Soviet subsidies and guaranteed trade that Cuba had depended upon for decades. In addition, Castro had just announced one of the country’s worst sugar harvests in decades.
 Yet he held a news conference on August 5th, warning that he would consider opening the border if the US did not change its behavior. He then waited an entire week before authorizing the initial sanctioned departures and did not publicly declare the borders open until August 20th, following the US announcement that sanctions on Cuba were to be tightened.
 Again, Castro did not tell the Cuban Coast Guard to let people go until a week after the street riots in Havana, and he did not formally announce that people could leave until another week had past.
 But in each of his pronouncements on the crisis, he clearly stated that negotiations on US immigration policy were a necessary precondition for ending the crisis.
 This is particularly true, given that as many argue, Castro likes nothing better than to embarrass the US. So were he not engaged in tit-for-tat bargaining with Washington, he would have surely closed the border at any time other than just after concluding an agreement with Washington.
 See, for instance, Walter Pincus and Robert Suro, “Ripple in Florida Straits Overturned US Policy,” The Washington Post, September 1, 1994, p. A31; and Holly Ackerman, “Transition in Cuba: Terms of Change,” Part 2 in series of 3 (1997), on-line edition.
 During the crisis itself, both Attorney General Janet Reno and Undersecretary of State Peter Tarnoff publicly acknowledged that Castro deliberately caused the crisis in “an effort to force a dialogue with the United States.” From Norman Kempster, “Clinton Rejects Castro’s Call for Top-level Talks,” Los Angeles Times, August 26, 1994, p. A1.
 Rohter, “Castro plays his cards with cunning,”; See also Ackerman.
 Richard Nuccio, during an NPR radio interviewed as part of “What's Next In The Elian Gonzalez Story,” Talk of the Nation, April 24, 2000.
 Helm’s Bill Could Push More Cubans to Leave for US,” CNN News, April 12, 1995, Transcript #1075-5.
 Masud-Piloto, p. 134.
 See Masud-Piloto, Table 9.1, p. 135. See also “News Briefing With Former Presidential Candidate and Senator George McGovern,” (Major Leader Special Transcript), Federal News Service, September 15, 1994 for an explication of Castro’s position on US policies.
 “La razon es nuestra…”. Interestingly, Castro’s claims were consistent with a report generated by the US Interest Section in January 1994. In a “top secret memorandum” to the State Department, CIA, and INS, visa officers in the Interest Section discussed the problems they were facing in trying to identify visa applicants with legitimate human rights cases, in that “…most people apply more because of the deteriorating economic situation than a real fear of persecution.” From Masud-Piloto, pp. 134-5.
 This is essentially an (in this author’s view, correct) argument that the US too was using refugees as a political weapon.
 MacPherson; see also interview with McGovern.
 Ernesto Rodrizguez Chavez, “La crisis migratoria Estados Unidos-Cuba en el verano de 1994,” Cuadernos de Nuestra America, vol. XI, no. 22, pp. 4-25.
 Peter H. Smith, Talons of the Eagle : Dynamics of U.S.-Latin American Relations, 2nd edition (Oxford: Oxford University Press,1999), p. 394, fn. 4. On July 26th and August 3rd and 4th, the ferry that had transported passengers from Havana to Regla for nearly 100 years was hijacked to Miami. Violence was used in each of the hijackings, and, in one case, a Cuban policeman was killed. From Masud-Piloto, p. 137. See also Chavez, p. 14.
 In the interest of brevity, both the Camarioca crisis and the Mariel boatlift are presented in thumbnail sketch form below. For more detailed analyses, see Kelly M. Greenhill, People Pressure: Strategic Forced Migration and the Rise of the Human Rights Regime, Ph.D. dissertation, MIT, in progress. See also Rivera, Engstrom, and Wayne C. Smith, The Closest of Enemies: A Personal and Diplomatic Account of US-Cuban Relations Since 1957 (New York: W.W. Norton and Company, 1987) for more detailed histories of these two crises.
 See New York Times, September 30, 1965, p. 1A and 2A; and Miami Herald, September 29, 1965, p. 1A and 6A. The Herald dubbed Castro’s offer “a propaganda bomb.” See also “Why Castro Exports Cubans,” New York Times Magazine, November 7, 1965, p. 30.
 New York Times, October 1, 1965, p. 2.
 Miguel Gonzalez-Pando, The Cuban Americans (Westport, CT: Greenwood Press, 1998), p. 44.
 Miami Herald, October 14, 1965, p. 1A.
 However, the Memorandum did not constitute a formal normalization of US-Cuban immigration policy; that would have to wait until the aftermath of the Mariel boatlift fifteen years later. See Masud-Piloto, pp. 61-8, Engstrom, p. 26-28, and Gonzalez-Pando.
 Juan Tamayo, editorial in El Nuevo Herald, April 23, 2000.
 Naomi Flink Zucker and Norman L. Zucker, The Guarded Gate: The Reality of the American Refugee Policy (San Diego: Harcourt Brace Jovanovich, 1987), pp. 60-61. See also Mario Antonio Rivera, Decision and Structure: US Refugee Policy in the Mariel Crisis (New York: University Press of America, 1991), p. 196.
 David W. Engstrom, Presidential Decision Making Adrift: The Carter Administration and the Mariel Boatlift (New York: Rowman and Littlefield Publishers, Inc., 1998), p. 189.
 Within a few days, 10,000 Cubans crowded into the embassy complex, but Peru agreed to take only 1,000 asylum applicants. See, for instance, Gonzalez-Pando, pp. 65-6.
 Rivera, p. 7.
 Ibid. These quotations were gleaned from White House and State Department briefings in late May and June 1980. Such sentiments were echoed in discussions at the Cuban-Haitian Task Force, the Cuban Desk, and the Bureau of Refugee Programs. See Rivera, p. 23.
 US Opens Arms to Cuban Exodus,” The Miami Herald, May 6, 1980, p. 1A.
 See, for instance, “Blockade Fails to Stop 276-foot Cruise Liner,” The Miami Herald, June 5, 1980, p. 16A.
 Rivera, p. 68.
 The bases utilized included Eglin Air Force Base in Florida, Fort Indian Gap in Pennsylvania, and, critically in light of what happened in 1994, Fort Chaffee in Arkansas. A similar relocation program was employed during the Camerioca crisis. In its aftermath, it was deemed a failure because most of those who were relocated moved back to South Florida within a short period of time. However, this misses the critical point that the primary—and successful—goal of the relocation program was to relieve the acute and “condensed pressure” on Florida public officials. See for instance Robert L. Bach, “Policy Debates and the Refugee Experience,” in Lydio Tomasi (ed.), In Defense of the Alien (New York: Center for Migration Studies, 1984), pp. 178-9.
 Apparently the initial position of the National Security Council (NSC) was to demand that the Cubans suspend the sealift without offering them a quid pro quo for doing so. See Wayne Smith, The Closest of Enemies, p. 215. Having initiated the crisis, and aware that domestic pressures were mounting in the US, Cuban representative Ricardo Alarcon, then Vice-Minister of Foreign Relations, responded at the time, “if we ever get back to negotiating anything, it will have to be on the basis of a step-by-step process based on reciprocity…. We aren’t going to sit down with you to talk about stopping the Mariel operation and then have that be the end of it.” Ibid. “As soon as the Cubans realized that we had come only to demand suspension of the sealift, they turned us out.” Smith, p. 216.
 Ibid., p. 215.
 Engstrom, p. 121. See also Engstrom, fn. 115, for details of an interview with negotiator, Peter Tarnoff. See again Smith and Alex Larzelere, The 1980 Cuban Boatlift: Castro’s Ploy—America’s Dilemma (Washington, DC: National Defense University Press, 1988).
 See Smith, p. 216; Engstrom, p. 120; and Tarnoff interview, June 17, 1991, cited in fn. 110, p. 134. See also Larzelere, p. 254.
 Engstrom, p. 196.
 Ibid., pp. 120-1.
 Ibid., p. 196.
 Although it is clear that Carter himself believed that Mariel was an important component to his defeat. Immediately after the election, he said, “the refugee question has hurt us badly. It wasn’t just in Florida, but it was throughout the country. It was a burning issue. It made us look impotent when we received these refugees from Cuba.” From Public Papers of the Presidents of the United States: Jimmy Carter, 1980, vol. 3 (Washington, D.C.: GPO, 1981), p. 2693.
 For instance, on May 28, more than one hundred people forced their way into the Belgian ambassador’s residence, and on during the week of July 13, 21 crashed the German embassy, while another nine entered the Chilean consulate. Although the embassy crashings were resolved without incident, many of the boat hijackings involved violence, and both civilians and Cuban police officers were being killed. For details of the hijackings in this period, see El Nuevo Herald, July 14 to August 12, 1994; see also Masud-Piloto, p. 137.
 Executive Office of the Governor, The Unfair Burden: Immigration’s Impact on Florida (March 1994).
 CNN, July 30, 1993. The US may have further tried to deter Castro by announcing that contingency plans were in place to prevent another crisis in south Florida, such as occurred in 1980. See John Zarrella, “Cuban-Americans, Officials Doubt Another Cuban Boatlift,” CNN, August 11, 1994, Transcript # 548-2.
 Engstrom, p. 188.
 “Protesters Battle Police in Havana; Castro Warns US,” New York Times, August 6, 1994, p. A2.
 Larry Rohter, “Flight from Cuba: The Overview,” New York Times, August 25, 1994, p. A1.
 Operation Distant Shore was developed early in the first Clinton Administration. It was designed to deal with another Mariel-type exodus of Cubans via the redistribution of refugees to forts and other federal facilities around the nation, to alleviate the concentrated burden on Florida. Interview with senior INS official, March 2000.
 Ibid. Also in the immediate aftermath of Castro’s August 5th statement, the US claimed that it had the situation under control and that Castro would back down. See State Department Regular Briefing, August 8, 1994.
 Interview with senior INS official, March 2000.
 See “US Maps Plan to Counter Cuban Threat of New Boat Lift; Castro Issues Warning After Violent Demonstration in Havana,” Miami Herald, August 7, 1994, p. A3; “Bracing for Potential Cuban Refugee Influx,” Miami Herald, August 7, 1994, p. A10
 Reportedly Castro made the decision in a meeting on August 12th not to interfere with those trying to leave the island, although he did not publicly declare the port open until a week and a half later. See again Pincus and Suro; David Williams, “After 35 Years, Castro Still Annoys Washington,” The Washington Post, August 13, 1994, p. A13. In the interim period, he had representatives at the Cuban Interests Section simply announce “the US is simply reaping what it sows by its own policy.” See William Booth, “Mixed Signals at Sea: US Opening Lures Desperate Cubans to Boats,” Washington Post, August 18, 1994, p. A1.
 Paul Anderson, “Miscues, Awakenings Helped Spur Policy Reversal,” Miami Herald, August 21, 1994, p. A6. See also Rafael Lorente, Paul Anderson, and Andres Viglucci, “Dade Makes Plea for Help,” Miami Herald, August 18, 1994, p. A1.
 Wolf Blitzer, “Chiles Says Feds Need to Toughen Immigration Policies,” CNN, August 18, 1994, Transcript # 907-2.
 September 1994 Gallup data reveal that 79% of the sample population did not believe that Cuban refugees should be allowed into the US, and 91% felt that Cubans should be treated just like the Haitian migrants trying to enter the US. From The Gallup Poll, 1994 (New York: Random House, 1994), pp. 228-33. Another poll conducted the following May, shortly after the accords were signed, indicated that 73% of Floridians supported banning illegal immigrants from access to government services. See David Adams, “Protests of Cuba Policy Draw Little Support,” St. Petersburg Times, May 13, 1995, p. 1A.
 SeeRobert Greenberger, “Clinton Faces Pressure from All Sides, Even His Family, in Fight to Shape Policy on Cuba,” Wall Street Journal, September 2, 1994, p. A10. This assessment was confirmed in interviews with former administration officials; off-the-record interviews conducted in July-August 2000.
 Confidential author interviews, Summer 2000; consistent with press reports as well.
 Steven Greenhouse, “Freewheeling Ways Pay Off for the White House,” New York Times, September 11, 1994, p. 8.
 Fort Chaffee had housed more than 20,000 Cuban refugees by June 1980. On June 1st, a group of about three hundred detainees tried to escape. After most were captured, they began rioting, which led to Clinton’s decision to call out the Arkansas National Guard. Although few were injured in the melee, and those involved were prosecuted, the perception that Clinton had lost control adversely affected his bid for re-election. Clinton ended up losing to Frank White, head of a small savings and loan, who was the first Republican to serve as governor in Arkansas in a hundred years. See Jonathan C. Smith, “Foreign Policy for Sale? Interest group influence on President Clinton’s Cuba policy,” Presidential Studies Quarterly, vol. 28, no. 1 (Winter 1998).
 At least some others appear to share this assessment. See “Janet Reno Wins Schneider’s Play of the Week,” CNN—Inside Politics, May 5, 1995, Transcript # 828-5; Martin Dyckman, “The Cuban Bogeyman,” St. Petersburg Times, September 1, 1994, p. 19A.
 William Degregorio, The Complete Book of Presidents: From George Washington to Bill Clinton (New York: Wing Books, 1993), p. 714. It is worth noting that several of Clinton’s key foreign policy advisors were also working in the Carter Administration during the Mariel crisis, including Anthony Lake, Warren Christopher, and Mariel crisis negotiator, Peter Tarnoff.
 Chiles argued that the federal government was in a state of denial and if it would not respond, he would. Chiles argued that it “was not a manageable situation. Not for Florida…If we do not get a response from the federal government, we will open our own facilities.” From Robert Rankin, Tim Nickens, and Lizette Alvarez, “Rescued Rafters Will Be Sent to Guantanamo Base Camps,” Miami Herald, August 19, 1994, p. A1. See also “Spotcheck,” St. Petersburg Times, October 20, 1994.
 See again Blitzer.
 Pincus and Suro.
 See again Jonathan C. Smith. See also Elizabeth Drew, On the Edge: The Clinton Presidency (New York: Simon and Schuster, 1994), p. 430. Neither President Clinton, nor Secretary of State Warren Christopher or Deputy Secretary Strobe Talbott attended this “principals” meeting. Nor was any State Department staffer consulted about the proposed policy shift, and this included Assistant Secretary of State Alexander Watson, who was at least ostensibly in charge of the State Department's Bureau of Inter-American Affairs. See Robert Novak, “Focus on Immigration Is a Victory for Castro,” Chicago Sun-Times, August 29, 1994, p. 23.
 See, for instance, Pincus and Suro.
 Burt Solomon, “Clinton’s Fast Break on Cuba…or Foreign Policy on the Fly,” National Journal, September 3, 1994, p. 2044.
 Peter Hakim, President of the Inter-American Dialogue, a Washington think tank, quoted in Dick Kirschten, “Guantanamo, si; otherwise, no,” National Journal, vol. 27, no. 19 (May 13, 1995), p. 1198.
 See Bruce Stokes, “The Cuban conundrum,” National Journal, vol. 26, no. 38 (September 17, 1994).
 As Richard Haass, NSC official under George Bush put it: there seemed to be a fundamental “ inconsistency in simultaneously intensifying the pressures inside Cuba while making it harder for the discontented to flee… ‘Clinton seems more interested in balancing the various interests than deciding between them.’” From Solomon, “Fast Break on Cuba…”. Latin American expert, Robert Pastor, echoes Haass’ sentiment, asserting “on issues concerning Central America and Cuba, US policy seemed to be driven by the interest groups with the greatest leverage and determination, whether in the Congress or in Miami. These issues were not priorities.” From Pastor, “The Clinton Administration and the Americas…”.
 In the two weeks between August 13th and 25th, the Coast Guard rescued 13,084 rafters, a much larger number than the 9,340 who arrived during the first 12 days of Mariel.
 Fidel Castro, Televised speech, CNN, August 24, 1994.
 Steve Greenhouse, “Flight from Cuba: US Rejects Castro’s Proposals for Talks,” New York Times, August 26, 1994, p. A12.
 See national newspaper coverage of the Cuba crisis in, for instance, The New York Times, The Washington Post, and The Christian Science Monitor, as well as TV coverage on CNN, between August 25-31, 1994.
 This was made even clearer by the fact that not all Cuban-Americans supported tightening strictures on Havana, as some were more concerned about the possible effects on their relatives’ quality of life than with the pain it would inflict on Castro. See Jon Nordheimer, “Cuban Group Forges Link to Clinton,” New York Times, August 26, 1994, p. A12.
 See LeoGrande, p. 78.
 Ray Sanchez, “Exodus' End? U.S, Cuba reach pact on refugees,” Newsday, September 10, 1994, p. A7.
 “Helm’s Bill Could Push More Cubans to Leave for US.”
 “Adrift Off Cuba; Who’s Making Clinton’s Policy?,” The Washington Post, April 2, 1995, p. C1. During this same period, the flow of rafters began to rise again and reached the highest monthly total (190) since the end of the August 1994 crisis. See Barbara Crossette, “Cuban Rallies Will Oppose a US Bill,” New York Times, April 16, 1995, p. A4.
 Mark Matthews, “U.S. shifts policy, to admit 15,000 Cuban refugees,” The Baltimore Sun, May 3, 1995, p. 1A.
 Congressional Testimony of General Sheehan, “U.S. Policy Toward Cuba,” before House International Relations Committee, Western Hemisphere Subcommittee, May 18, 1995.
 Ann Devroy and Daniel Williams, “'Serious Alarm Bells' Led to Talks With Cuba; Congressmen Warned of Exodus, Riots at Camp,” The Washington Post, May 5, 1995, p. A5.
 For his part, Tarnoff claims that the negotiations were held in secret to preempt a massive exodus that might transpire in anticipation of any new immigration accord. See Congressional Testimony of Peter Tarnoff, “U.S. Policy Toward Cuba,” before House International Relations Committee, Western Hemisphere Subcommittee, May 18, 1995. See also LeoGrande and “Guantanamo, si…”.
 “Testimony of Gillian Gunn.”
 “Janet Reno Wins Schneider’s Play of the Week.”
 Steven Greenhouse, “A First Step on Cuba?; Americans Say the Accord on Immigration Might Be Followed by Other Agreements,” New York Times, May 3, 1995, p. A1.
 Bob Deans, “Showdown looms on Cuba policy,” The Tampa Tribune, May 4, 1995, p. 1.
 Frank Calzon, Washington director of the New York-based Freedom House, quoted in Peter Grier, “Raft Crisis Points Out Need For Long-Term Cuba Policy,” The Christian Science Monitor , September 12, 1994, Pg. 3.
 See again Castro’s CNN speech on 8/24/94, as well as Martin Fletcher, “Clinton offers to increase visas if Cuba halts exodus,” The (London) Times, August 31, 1994, Overseas News Section.
 See Chris Marquis and Mimi Whitefield, “Cuban Accord Favors Clinton,” The Times-Picayune, September 11, 1994, p. A1.
 Department of State, “US-Cuba Joint Communique on Migration,” Dispatch 5 (September 12, 1994). This provision gave Castro a guaranteed safety valve by which he could jettison tens of thousands of malcontents/year.
 Even now these provisions remain incompletely implemented because of US adherence to the so-called wet foot/dry foot policy. (The “wet foot” part of the policy means that all Cubans picked up at sea would be returned, while the “dry foot” aspect means that any Cuban who actually reaches the US still has the right to stay. Interviews with senior INS and Coast Guard officials, Spring 2000.
 However, under the 1984 immigration agreement, the 20,000 figure was an upper limit instead of a minimum, although as noted above only a few thousand Cubans actually were allowed to immigrate each year.
 See Daniel Williams, “Cuba Deal Depends on Castro Dropping Trade Demands,” The Washington Post, September 9, 1994, p. A34.
 Marquis and Whitefield.
 Ibid.; see also Williams.
 Stanley Meisler, “U.S., Cuba Sign Accord To End Migrant Exodus; Refugees: Clinton Says Orderly Immigration Will Replace Island Escapes. Havana Won No Key Concessions,” Los Angeles Times, September 10, 1994, p. 1A.
 This author hopes to have a more confident assessment of this issue following (anticipated) forthcoming interviews with Cuban officials.
 See again “News Briefing With Former Presidential Candidate and Senator George McGovern,” (Major Leader Special Transcript), Federal News Service, September 15, 1994.
 Rohter. “Castro plays his cards with cunning.”
 McGovern interview.
 See “Tide of Cuban Boat People Eases; U.S. Officials Say Change in Immigration Policy May Have Stemmed Exodus,” The Buffalo News, August 29, 1994, p. 1, which quotes “experts” whose views are consistent with this analysis: “Although Castro has used the crisis to renew his demand for lifting a three-decade-old U.S. trade embargo on Cuba, experts say they believe that his real objectives are far more modest: easier legal migration, U.S. prosecution of Cubans who make it to Florida in stolen aircraft or boats, and restoration of permission for Cuban-Americans to send money to relatives on the island.”
 See again Ibid.; and Meisler.
 Larry Birns, director of the Council on Hemispheric Affairs, quoted in Ibid.
 Daniel Williams and Roberto Suro, “Upcoming Talks With Cuba on Immigration Issues Could Foster Deeper Dialogue,” The Washington Post, August 30, 1994, p. A6,
 Steven Greenhouse, “Flight from Cuba: U.S. Policy; U.S. Promises to Respond If Castro Offers Reforms,” New York Times, August 29, 1994, p. A6.
 A Brookings Press Briefing Dealing with the States Formerly Known as Rogues, July 13, 2000. See Brookings Institution website.
 “Trying to reconcile: US quietly questioning policies that have led to 40 years of cold relations with Cuba,” Cuba in Evolution, September 7, 1998. In June 1997, Torres filed a bill to end trade restrictions on food and medicine, a bill that had more than 90 co-sponsors in the House, although none from Florida. Ibid.
 It has been argued that Castro used the shoot-down to reaffirm his credibility, and launched “a preemptive strike to signal to all political opponents, in Cuba and elsewhere, that he was prepared to use force to remain in power.” See Ackerman.
 Phil Willon, “Embargo against Cuba scrutinized,” The Tampa Tribune, January 12, 1998, p. 1.
 Interviews with senior INS and Coast Guard officials, March and April 2000. Confirmed by senior US military personnel during interviews at SOUTHCOM, April 2000.
 “US Adding to Tent City for Cubans,” St. Louis Post-Dispatch, August 25, 1994, p. 1A.
 See again fn. 145.
 According to a former senior Clinton Administration official, it was always the Clinton Administration’s intention to use Gitmo only as a delaying tactic and to return those at Gitmo to Cuba. Interview with former Clinton Administration official, Summer 2000.
 Interview with senior INS official, March 2000.
 Interview with senior Coast Guard official, April 2000.
 Engstrom, p. 189.
 See, for instance, Gonzalez-Pando, who discusses in detail distinctive stages of Cuban migration. See also Susan Eckstein and Lorena Barberia, “Cuba-American Cuba Visits: Public Policy, Private Practices,” A Report of the Mellon_MIT Inter-University Program on NGOs and Forced Migration (January 2001), p. 6, for a brief discussion of the shift in composition of the later émigrés.
 See Engstrom. However, this rationale had been in place since Batista’s fall, and Cuba experts did not expect Castro’s imminent fall, even in the face of the August riots. See again fn. 12.
 “Adrift Off Cuba…”.
 See Jorge Dominguez, “US-Cuban Relations…”, pp. 60-5. For a discussion of the provisions and consequences of Helms-Burton, see Joaquin Roy, “The Helms-Burton Law: Development, Consequences, and Legacy for Inter-American and European-US Relations,” Journal of Interamerican Studies and World Affairs, vol. 39, no. 3 (Fall 1997), pp. 77-108.
 However, Helms-Burton has never been fully implemented. Even Clinton’s successor, George W. Bush, has so far decided to forego implementation of all of the bill’s provisions, in the interest of not riling US allies, who have economic ties with the island.
 Lisandro Perez, Director of the Cuba Research Institute at Florida International University, quoted in Rohter, “Castro plays his cards with cunning.”
 Jorge Dominguez quoted in ibid.
 See, for instance, Robert Jervis, Perception and Misperception in International Politics (Princeton: Princeton University Press, 1976), Chapter 3, for discussion of the potential dangers of appeasment.
 Engstrom, p. 198.
 There were a few exceptions, such as St. Lucia, which offered to shelter Cubans in exchange for new water pipelines and roads, and the Turks and Caicos Islands offered to take up to 2,000 people for short-term stays of up to four weeks. From “And still they come on risky rafts,” The Baltimore Sun, August 25, 1994, p. 1A.
 Convincing states to burden share with the US in 1994 grew even more difficult in the aftermath of riots in Panama that December, after which Panama demanded that the US take back the 15,000 it had agreed to host. These developments “virtually guaranteed that no other state will wish to host detention facilities on its soil,” as Gillian Gunn, Director of the Cuba Project, put it in her February 1995 testimony before Congress. For details of the deal the US struck with Panama to take the refugees in September 1994, see “News Conference on Acceptance of Cuban Refugees,” FBIS Daily Report, September 4, 1994.
 Michael Barutciski and Astri Suhrke, “Lessons from the Kosovo Refugee Crisis: Innovations in Protection and Burden-sharing,” Journal of Refugee Studies, vol. 14, no. 2 (2001), p. 101.