Institutional analysis and historical learning

Institutions exist in the regular, interconnected activities of sets of people who thus make or carry the institution (Glaeser Reference Glaeser2011, introd. and Chaps. 3 and 4; Glaeser Reference Glaeser2014). The linchpin of the model is to see every activity carrying the institution as a reaction to antecedent actions oriented and guided by a host of more or less conscious discursive, emotive, and sensuous understandings. Which particular set of understandings (out of the many often contradictory understandings actors hold) becomes actualized in any particular context depends on the understandings’ perceived relative validity in this particular context. Perceived validity changes with processes of validation. The main forms of validation are recognition by acknowledged authority figures, corroboration through actions’ success and failure, and resonance with other prevalent understandings. Validity varies with context because supporting authorities are absent or present, opportunities for corroboration vary, and situational cues produce different resonances. Moreover, the validity of all understandings fades with time unless it is renewed. Actors themselves are instituted because the understandings furnishing them with the capability to act are institutions, which is to say that they need to be recognized by others and corroborated in social interaction. Hence, the commonplace observation that actors can behave rather differently in different contexts. This insight will become important to understand the actions and the possible range of available actions of politicians and soldiers during a crisis such as the Cuban Missile Crisis.

Institutions can exist only within a supporting infrastructure of other institutions on which they are dependent. Infrastructural dependence can be thought to operate along three dimensions. First, the durability of an activity pattern is dependent on the stability of understandings that orients and directs it (e.g., the work ethic), which in turn hinges on the constancy of actualizing validations (praise by others) that rely on the stability of the institutional arrangements making them possible (parents, teachers, media discourses), and so on. Take politicians’ understanding that they have to do x in order to become reelected or to maintain authority. That understanding may become actionable to the degree that, for example, (a) advisors positively recognize it; (b) it resonates with their career ambitions and identity; or (c) it has been corroborated through previous success for oneself and/or others. However, advisors recognize this understanding continuously only to the degree that they continue to believe in it because, say, their own professional discourses keep validating it, and so on.

A second form of dependence results from the fact that all institutions presuppose other already extant institutions. For example, elections work only with particular kinds of political subjects who believe that they can make a difference by casting their ballot. Political candidates in the United States largely depend on the contributions of donors, who continue to understand donations as a means to influence policies, which means that politicians must act in ways that maintain that understanding. That practice, in turn, depends on party finance legislation and voter acceptance. A third form of infrastructural interdependence originates in the fact that the same act may reproduce several institutions at the same time. In acting, officeholders may contribute to the formation of their own identity, their office, the party they represent, the polity on whose behalf they act, a policy, and so on. I have called all three of these processes constituting one institution by hitching it to many others institutiosis.

Within this model, politics is the intentional effort to form, maintain, or alter institutional arrangements, and power is the ability to do so successfully. Institutiosis is an important dimension of politics that both limits and enables it. For example, Cold War politics was, on either side, an intentional effort to maintain or gain an advantage in the balance of power while at the same time avoiding any direct military confrontation. In an electoral democracy in which the balance of power is an electoral issue, institutiosis implies that candidates need to be perceived as good at maintaining or improving it. The basis for this institutional linkage is yet another set of prevalent institutionalized understandings among the electorate – for example, a particular form of nationalism cultivated at least in part in opposition to the supposed enemy.

Thinking institutionally also sheds light on the question of why politics is hard work that nevertheless often fails to achieve desired results. The possibility-space for politics is always enabled and limited by a wider institutional environment whose affordances have to be used strategically by politicians if their politics is to succeed. People expected to carry a particular institution – the aim of politics – need to form new habits, change old ones, or stick to what they have been doing. And that can hardly happen without taking account of carriers’ own intentional vectors, and their own discursive, emotive, and sensuous understandings of the world. It is usually impossible to change the whole cascade of institutional backlinks with a limited set of resources.

Politics as an idea and as a practice hinges on a set of background institutions. Several forms of political knowledge form the necessary preconditions: persuasive eutopias and dystopias to imagine another reality; sociologies to indicate where and how to intervene in the effect flows to change institutions; effective rhetorics to argue, induce, seduce, cajole, or persuade people to develop or alter habits; and organizational knowledge to mobilize and coordinate the efforts of the designated institutional carriers. The sufficient condition for politics consists of the existence of institutions facilitating interventions. Among them are publics enabling the formation of understandings, movement organizations, and parties enabling political mobilization, and state bureaucracies executing decisions on the ground.

Seeing all institutions as, in principle, alterable is not to argue that institutions are always the result of politics. They can also congeal through quasi-organic processes – that is, through the interplay of myriad unintended consequences, some of which may have been triggered by politics, while others precipitate from activities never intentionally geared to have an institutional effect. The genesis of historically existing institutions must therefore be understood as a mix of these quasi-organic and outright political activities.

Precisely because quasi-organic effect flows can change institutions, they can become the target of a continuous politics, which maintains them in a particular form. This can be done by designating people whose task it is to manage a particular institution. Organizations are institutions that are managed in this sense.

Institutional analysis suggests that learning from history is a complex political project. After important unexpected or problematic developments, including both actual and near successes and failures, the learner would ask: “What particular institutional dynamics have made these developments more or less likely?” And, therefore, “Which institutional arrangements might be changed to make the occurrence of such developments more or less likely in the future?” Institutional analysis thus suggests a major shift in the timing of learning that should occur mostly right after events have happened in a thorough institutional debriefing using the conceptual tools just developed. The reason is simply that the institutional dynamics of a more distant past are, because of political and quasi-organic changes to institutions, likely to be different from those in the present. Within such debriefings, attention should shift from the supposed quality of particular decisions, which all too often leads to an unhelpful individualization of praise and blame, to an assessment of institutional dynamics orienting and guiding actors to what they perceive as choices for possible courses of action. Institutional analysis plays a Janus-faced role in this context, first to help analyze the dynamic unfolding of past events and then to plan political interventions to ameliorate institutional arrangements.

After narrating the crisis events in some detail, I will, in what follows, use the Cuban Missile Crisis as a diagnostic event (Moore Reference Moore1987) that I will analyze with the theory of institutions just outlined to identify the interinstitutional dynamics that drove the actors to the brink of war.

The Cuban Missile Crisis

The Cuban Missile Crisis proper began on October 16, 1962, when Kennedy woke up to the bad news that American U-2 spy planes had captured photographic evidence of the existence of Soviet medium- and intermediate-range nuclear ballistic missiles (MRBMs and IRBMs) on Cuba. This discovery was not entirely unexpected. Since mid-August the CIA had been getting news from Cuban spies that the Soviet Union had significantly increased its military presence on the island. On August 29, U-2s photographed Soviet surface-to-air missiles (hence SAMs) as well as new torpedo boats. Both of these facts together led the CIA to warn Kennedy of the possibility that nuclear missiles were about to be deployed. While alarmed, worries were mitigated by the fact that the weapons systems identified could be classified as “defensive.” Nevertheless, their discovery prompted Kennedy to go public on September 4 and 13 with red-line announcements that the “gravest issues” would arise if the Soviet Union endowed Cuba with “offensive capabilities.” Right after the discovery of the SAMs, Kennedy suspended further surveillance flights because they endangered U-2 spy planes. The successful downing of a U-2 over Soviet territory in 1960 caused a significant diplomatic crisis and publicity disaster for the United States.

Until the 1970s, the Soviet Union trailed the United States in the deployment of nuclear weapons.Footnote ³ The United States had not only a small arsenal of relatively reliable intercontinental ballistic missiles (ICBMs), a vast bomber fleet stationed around the world, and a ring of MRBMs encircling the Soviet Union, but it also had a growing fleet of nuclear missile submarines. The Soviet Union had no reliable ICBM system, and the ICBMs that it had were relatively few in number.Footnote ⁴ While it had reliable MRBMs and IRBMs threatening Western Europe, it had no such systems deployed within reach of the United States before Cuba. It also had no nuclear-missile capable submarines yet, and its range-limited bomber fleet was only a third of the United States’ fleet in size.

Since the late summer of 1960, the US leadership had, thanks to Corona spy satellite imaging, a pretty realistic sense of its own very significant superiority. The “bomber gap” and “missile gap” hysterias of the second half of the 1950s stood corrected by factual analysis. The Soviets did not know exactly how far behind they were. They nevertheless had an acute sense of their own considerable disadvantage (Gribkov and Smith Reference Gribkov and Smith1994, 10–11), which became more of a certainty in October 1961, when the United States revealed its strong lead over the Soviet Union (Ellsberg Reference Ellsberg2017, 173–75; Sherwin Reference Sherwin2020, 187). To deter the United States, the Soviets had, in the late 1950s, carefully constructed a myth about Soviet capabilities in producing and deploying nuclear arms. In 1957, Khrushchev famously pronounced that the Soviet Union was “turning out missiles like sausages,” a claim seemingly corroborated by Sputnik (Pravda, October 11, 1957, as quoted in Lebow Reference Lebow1988, 41–42). By stationing MRBMs and IRBMs in Cuba capable of reaching most of the United States, Khrushchev hoped to rectify the balance of power to a certain degree.Footnote ⁵ Khrushchev reasoned that if the Soviets had to live under the threat of American MRBMs and IRBMs, then Kennedy should accept a reciprocal threat. Importantly, he also thought that the United States would understand that logic (Allison and Zelikow Reference Allison and Zelikow1999, 97; Fursenko and Naftali Reference Fursenko and Naftali1997, 182). The Soviets’ decision to deploy was reached in mid-May 1962. After the United States tried to invade Cuba with 1,200 exiles in April 1961 at the Bay of Pigs, and numerous US assassination plots against Castro, the Cuban leader was realistically fearful of another US invasion effort. Accordingly, he agreed to the Soviet plans to deploy more than 50,000 Soviet troops along with dozens of MRBMs and IRBMs alongside many tactical nuclear weapons.

By the time a U-2 photographed the deployment of MBRMs on October 14, the first ones were combat ready (Plokhy Reference Plokhy2022, 129). Arming them with nuclear warheads stored nearby was a matter of a few days. A number of the tactical nuclear weapons systems were also operational. When the crisis broke, the Soviets had 40,000 men in Cuba, while the CIA continued to assume that there were no more than a couple of thousand.

Kennedy assembled a team of advisors that came to be known as the Executive Committee of the National Security Council (ExComm). Their meeting on October 16 set the agenda for the following two weeks by deliberating on a number of possible reactions. ExComm quickly concluded that the United States had to act forcefully to maintain its reputation among allies and enemies alike and that doing nothing was out of the question. The actions that were seriously debated ranged from a naval blockade to more or less limited airstrikes to a comprehensive airstrike followed by a full-scale invasion. Stand-alone diplomatic solutions were readily dismissed, not least because the Soviets had systematically deceived Kennedy about their intentions and because such solutions would give the Soviets time to complete their military buildup. Politicians’ hope that the offensive weapons could be reliably identified and destroyed “surgically” such that a counterattack on American cities would be impossible was quickly dashed by the limits of prevailing technomilitary capabilities. The Joint Chiefs of Staff were therefore favoring an invasion after comprehensive bombardments as soon as possible.

A minority of advisors around UN Ambassador Adlai Stevenson were nevertheless pressing for a solution that would allow Khrushchev time to rethink the Soviet position and allow for negotiations. Kennedy heeded their advice and settled on declaring a “quarantine” on military shipments on October 22 by televised address (May and Zelikow Reference May and Zelikow2002, 183–89).Footnote ⁶ In an accompanying personal letter to Khrushchev, Kennedy emphasized that the United States would not tolerate offensive nuclear capabilities in Cuba, that the quarantine was the minimum response he could take, and that he was aiming to avoid general nuclear war (May and Zelikow Reference May and Zelikow2002, 189–90). Legally, the quarantine was a limited blockade, which required a declaration of war, hence the unusual term. Even though Khrushchev denounced it as “piracy,” he clearly understood the warning signs and pulled back a number of ships approaching the blockade line. Construction efforts at the missile sites in Cuba showed no signs of slowing down, however. Khrushchev’s response to Kennedy’s letter gave no indication that he might, in fact, be willing to back down – the intended consequence of the quarantine.

At the same time, the United States prepared for a full-scale invasion. Gear and troops were relocated to the southeastern United States. During the week of October 22, progress in troop redeployment was such that a comprehensive aerial bombardment could begin just a week later, followed by a full-scale invasion in yet another week. This preparation time effectively set the clock for a peaceful resolution of the conflict, which for Kennedy could only mean the removal of all offensive weapons from Cuba. Should Khrushchev not budge in response to the blockade, invasion loomed large as the default option.

By Friday morning, in the face of the fact that construction at the missile sites was continuing unabated, ExComm members began to feel gloomy (Ellsberg Reference Ellsberg2017, 200–201). Yet, during the day, the United States received signs via the United Nations, as well as through unofficial channels, that the Soviets might be amenable to a diplomatic solution in which they would withdraw their missiles and bombers in exchange for the United States’ commitment not to invade Cuba. In the course of the evening, this proposal became a reality when a long telegram from Khrushchev to Kennedy arrived that formally made this proposal. On Saturday morning, however, the Soviets publicly announced a second proposal via Radio Moscow in which they added a United States withdrawal of its MRBMs from Turkey to the deal. Kennedy’s advisors were mostly opposed to this swap, even though US leaders had long agreed among themselves that these missiles were technically outmoded (Kennedy Reference Kennedy1969, 83). They feared the United States would lose credibility with its NATO allies by withdrawing these missiles. Yet Kennedy understood clearly that invading Cuba with such a deal on the table would be almost impossible to explain, fearing a reputation loss on a much wider scale. In the end, both parties agreed on the second Soviet proposal with the caveat that the United States’ withdrawal of missiles from Turkey was kept a closely guarded secret that only a handful of Kennedy’s advisors knew about.

In the end, the crisis was resolved by Khrushchev backing down, yielding to superior American force. But what would have happened if he had insisted on keeping his missiles in Cuba? After all, the missiles were now operational, and he had staked out his strategic bet against opposition in the Politburo.Footnote ⁷ Could Kennedy, as a last resort to avoid war, have opted to do nothing, accepting the missiles as a nuisance the United States could live with? After all, even though their presence clearly increased Soviet capabilities, the overall balance of power was, as Robert McNamara declared, barely affected by their presence. Before I can explore the reasons why this is unlikely, it is important to understand how close to nuclear war the crisis actually came and what might have happened in case of an invasion.

Losing control

The horrifying fact about the Cuban Missile Crisis is its acute potential to descend into a general thermonuclear war, which, according to American war planners would, in the first months alone, have cost hundreds of millions of lives among the citizens of socialist countries in Europe and Asia as well as tens of millions of lives in the United States. These figures include neither deaths in Western Europe nor long-term global deaths from radiation and climate effects (Freedman and Michaels Reference Freedman and Michaels2019, 264). What is shocking about these numbers is that the political goals associated with risking war were comparatively small. For the Soviets, it was a temporarily improved strategic position until its own intercontinental missile force was ready, which they knew would happen in just a few years. For the United States, it was risking thermonuclear war just to scale back what was a significant but, relative to the United States’ considerable nuclear superiority, still a minor temporary strategic gain for the Soviets.Footnote ⁸ As McNamara said, “I don’t think there is a military problem there [with Soviet missiles in Cuba] …. This is a domestic political problem” (May and Zelikow Reference May and Zelikow2002, 70). In other words, the likely gains from strategic risk-taking and the probable costs of war stand in absolutely no rational relation to each other. This is so in spite of the fact that both leaders acted in the belief that they were rational and were doing everything to avoid such a war.Footnote ⁹ The decision-making leading to and during the crisis eschewed that much-cited bedrock of post-Enlightenment military strategy, Clausewitz’s doctrine that war is politics by other means – namely, that war can only be justified by the pursuit of goals that are commensurate with its likely costs (2000).Footnote ¹⁰

The first question must then be why the potential costs could become so monstrous. The answer lies in the then-prevalent American nuclear strategy. The Dwight D. Eisenhower administration bet on “massive retaliation” – that is, an all-out nuclear war – as its main strategy (Freedman and Michaels Reference Freedman and Michaels2019, esp. chaps. 7 and 17; Kaplan Reference Kaplan2020, chap. 1). The reason was that countering Soviet superiority in conventional weapons in Europe by rearming conventionally was deemed too expensive.Footnote ¹¹ The threat of using nuclear weapons was therefore supposed to deter both conventional and nuclear attacks. At the same time, the United States worked frantically to stay ahead of the Soviets in producing and deploying nuclear weapons. The nuclear war plan valid in the fall of 1962 was the Strategic Integrated Operational Plan 1963 (SIOP-63 for short). Since it has not yet been released, scholars have pieced together its provisions based on otherwise documented deviations from SIOP-62, which is much better known. While the latter followed the Eisenhower logic of an all-out nuclear war, designating the simultaneous use of over 4,000 nuclear weapons, the former was adjusted, due to Kennedy’s and McNamara’s urging, to be more discriminating in choosing targets while allowing for more steps of escalation, something they called “flexible response.” Since there was considerable resistance by the military to Kennedy’s and McNamara’s new ideas, the end effect was that there were still 3,500 weapons in the plan with a total destructive power of 6,300 megatons of TNT (Norris and Kristensen Reference Plokhy2022). That is to say, each of these weapons was roughly more than 100 times more powerful than the bombs used against Hiroshima and Nagasaki.

Moreover, it has to be considered that strategies are institutions. To have any effect, war plans existing at the level of the highest command need to be minutely divided into a complex system of subplans down to the rules of engagement for the elementary units of the system, along with a set of contingencies anticipating various situations and associated authorizations for the use of particular weapons. In addition, however, as the theory of institutions tells us, all of these plans have to reside in their individual carriers. Particular strategy-supporting subjects have to be formed in mind and body. Sociologically, strategies exist in drilled-in practices of soldiers on all levels, their near-automated ways of seeing the world, their jokes, their stories about self, enemy, weapons, war and peace, and so on. And they exist in the education of their senses and their feelings. None of this can be changed quickly from above. Institutiosis works usually to the detriment of reforms. Sociologically speaking, there is therefore little reason to assume that in execution SIOP-63 would dramatically deviate from SIOP-62 with its long pedigree of Eisenhower-era routines and military cultures, many with deep roots in World War II. Part of these cultures was a distrust of civilian leadership and a profound sense that the military knew best, including how and when to use nuclear weapons.Footnote ¹² De facto, then, American nuclear war plans thus had very little flexibility for a gradual escalation built into them as was desired by Kennedy and McNamara. SIOP-62 was an all-or-nothing proposition.

The second question must then be what could have triggered the unleashing of SIOP as it was then institutionalized? In the Caribbean war theater, unlike in central Europe, the United States had such superiority in conventional weapons that it was unlikely to feel challenged to go nuclear in a conventional conflict. Of course, quite the reverse is true for the Soviet Union and Cuba. Therefore, the most likely scenario for triggering a quickly escalating war would have been the first use of nuclear weapons, tactical or strategic, by Soviet forces or the perception of an imminent strategic nuclear attack by the United States on the Soviet Union. It does not matter in this context whether these first uses were accidental, only locally but not centrally authorized. Both leaders assumed they had control over their forces, attributing any action of the other side to leadership intention.

How do noncentrally authorized uses come about? The whole military apparatus from top command to the smallest combat unit consists of institutions linked by what I call institutiosis. A supreme commander is only as supreme as such a dense fabric of institutional arrangements allows. Importantly, these institutional arrangements are, by design, reproduced differently locally and at the center – that is, their reproduction is not fully centrally controlled. Carriers of infrastructural institutions do, after all, have their own understandings, which are never fully congruent and sometimes even quite divergent from those at the center.Footnote ¹³ Moreover, they are reproduced differently in changing circumstances. In the military, in particular, there is a spectrum of anticipated circumstances ranging from peacetime to wartime operations, with various crisis levels in between. Along this spectrum, the cast of institutional carriers changes; changing contexts produce different opportunities for validation, which push different understandings into the foreground, and the relative weight of validations maintaining these understandings shift, for example, because local commanders begin to matter more than distant ones, not least because the communication infrastructure is no longer the same. Consequently, the subjectivities of actors shift as well.

A combat unit in peacetime, for example, is maintained by barracks routines among familiar insiders, occasional more or less realistic war games, bureaucratic communication across reliable channels, and so on. The modality of validation that prevails to maintain the understanding underwriting these routines is recognition, with an emphasis on the official chain of command. A combat unit in wartime is maintained through interactions with the enemy as well as with insiders. Motivating understandings shift from promotion to survival, from following regulations to getting the job done as it is locally understood. Orienting understandings are filled with real rather than simulated fear and anger, and corroboration by more or less successful action competes in import with recognition by superiors; the authority of distant commanders is regularly superseded by that of proximate superiors and peers because relative communication frequencies shift in favor of the latter. At the height of crises or in war, even the civilian command and the military top brass, both far removed from the actual war theater, will find themselves reconstituted institutionally in different ways because they begin to interact with a different cast of carriers. It is not a far stretch to say that every member of ExComm acquired a subjectivity subtly different from their usual bureaucratic selves during the crisis.

These are the conditions giving rise to what is mystifyingly known as the “fog of war.” But sociologically speaking, it’s just the predictable transformation from routine forms of institutional reproduction to radically uncertain, open-ended forms. It is rather the fog of bureaucratically constituted and rehearsed strategy that prevents planners from comprehending the realities of actual war on the ground. Troops deployed in the field in a situation of intense crisis see themselves on the cusp of war. Sometimes militaries designate their own system of stages between the routine barracks condition and war. The United States was working with a five-step system of defense conditions (DEFCONs, for short). From Wednesday, October 24, American troops were put in the penultimate condition of threat severity, DEFCON-2, one step away from general nuclear war. From October 26, Soviet troops in Cuba saw an American invasion of the island, and thus hot war, as imminent.

To illustrate how institutional dynamics can lead to an escalation potentially ending in what the Americans called at the time “general war,” I will briefly discuss actual occurrences on “Black Saturday,” October 27, followed by an institutional projection of what might have happened during an American invasion. Black Saturday began with the shooting down of a U-2 over Cuba by a Soviet antiaircraft missile battery. The order to shoot was given in spite of the order from the supreme commander of Soviet forces in Cuba not to act without his explicit confirmation. Since he fell ill, however, and could not be reached, his deputies acted on their own in what they saw as good faith. By downing the U-2, the soldiers followed a long-rehearsed air-defense logic that reconnaissance aircraft are typically followed by bombers. The panic and exhaustion after a sleepless night in expectation of the invasion will also have contributed to their decision-making. Not surprisingly, many of Kennedy’s advisors saw this attack as an escalating gesture invalidating Khrushchev’s previously signaled readiness to negotiate about the missiles. Even though Kennedy saw how control was slipping from his own hands, there was still the assumption that Soviet action followed orders directly emanating from the Kremlin (Plokhy Reference Plokhy2022, 240–45). While the military advisors especially were urging Kennedy to bomb at least this one SAM site, Kennedy wisely decided to let it go.

A little later, another U-2 accidentally veered off into Soviet airspace on its way back to Alaska from a prescheduled air-sampling mission to the North Pole to track Soviet nuclear tests. It was intercepted by Soviet fighters, which prompted US fighters to shelter the U-2 as soon as it was out of Soviet airspace. Both US fighters were armed with nuclear air-to-air missiles. Luckily, all of the pilots stayed calm, and nothing happened (Dobbs Reference Dobbs2008, 254–72).Footnote ¹⁴ This action, too, could have served as an indicator to the Soviets that the United States was intent on escalating the situation by sending a reconnaissance plane in preparation for an attack with bombers.Footnote ¹⁵ Kennedy famously commented on this incident, “There is always some sonofabitch who doesn’t get the word.” It would have been more accurate to say that there is always an institutional arrangement (here, prescheduled air sampling) that will produce actions that may provoke totally unexpected reactions in altered circumstances. And it is always difficult to manage all the many arrangements centrally, even within organizations fiercely determined and deeply accustomed to following central orders.

In the early evening of the same day, B-59, a Soviet diesel submarine, came close to readying a nuclear torpedo for attack on an American aircraft carrier group.Footnote ¹⁶ In his directives to the navy enforcing the limited blockade, McNamara gave strict orders to abstain from any use of live ammunition against a Soviet vessel. At the same time, the United States wondered how to enforce the blockade with respect to submerged submarines that could not be reached by radio communication. The United States thus devised an ad hoc plan for the use of practice depth charges, weaker versions of actual underwater bombs designed to sink submarines, to “signal” to the boat that it was discovered and to force it up. The United States had transmitted its signaling protocol to the Soviets just a few days before, but it is highly unlikely that this protocol ever reached the submarine commanders, who, as long as they were submerged, were virtually incommunicado. When B-59 was rocked by American practice depth charges, it had been submerged for days due to its relentless pursuit by US antisubmarine forces. Its batteries were nearly depleted, and temperatures on board were insufferable. Crew members fainted.Footnote ¹⁷ What was considered a signal by the Americans was considered by the Soviets to be an attack, not least because their boat was actually damaged by the American action. Soviet submarine commanders were under strict orders never to let themselves be forced to the surface. Yet B-59 had to because its battery was empty and because of malfunctioning equipment. According to the men aboard B-59, once up, the boat was greeted with machine-gun fire from a US antisubmarine plane. This led the exhausted captain to order an emergency dive and the order to ready its nuclear torpedo. His motive: the honor of the Soviet navy. Luckily, the chief of staff of the Soviet submarine group sailing together to Cuba was assigned to travel on B-59. He was able to persuade the captain to desist from using the torpedo. Had it been successfully fired at one of the destroyers surrounding B-59, the consequences could have been grave indeed. Said McNamara thirty years later, “No one should believe that a U.S. force could have been attacked by tactical nuclear warheads without responding with nuclear warheads” (Ellsberg Reference Ellsberg2017, 210).

What is noteworthy about this particular episode is that the ability (as distinct from the formal authority) to fire nuclear weapons needs to be delegated at one point to the soldiers operating their delivery systems, whether that be a rocket launcher, a bomber, or a submarine. For technological and organizational reasons, that delegation often has to take place well before commanders may formally authorize their use. In the case of the Soviet submarines in the Caribbean, this happened essentially when they left their home ports. This is precisely because communication is in danger of becoming interrupted, or it is, as in the case of submarines, technologically impossible for long stretches of time. For bombers, this was true as they were leaving their bases. Since communications frequently break down, launching soldiers have to make up their own minds about how to proceed. Often this is done in what they perceive to be the spirit of their orders, adjusted to radically altered circumstances.Footnote ¹⁸ With nuclear weapons, the local enablement of firing is also necessary to maintain a second-strike capability, which is essentially an opportunity for revenge in the event that home bases and command centers have been destroyed in a first strike. As soon as the enablement to fire is delegated to the launching soldiers, local decisions strongly influenced by local circumstances can override central ones. The military in real conflict is, in many ways, no longer the strategically anticipated military.

And again, just a few hours later, an American nuclear missile crew in Okinawa, Japan, appears to have received a “launch” order that luckily met a skeptical local commander who questioned its validity due to the inconsistency between the order and the current defense condition. He demanded reaffirmation of the order by his superiors, only to find them subsequently rescinded (Sherwin Reference Sherwin2020; Tovish Reference Tovish2015).Footnote ¹⁹ This example, along with the dissuasion of B-59’s commander from readying the nuclear torpedo, shows that local decisions may not only destabilize the strategically anticipated institutional fabric, but they can also stabilize it if frontline troops stay oriented towards a broader strategic framework.

By far the greatest danger, however, would have occurred within the context of an attempted American invasion of Cuba – the default plan strongly favored by ExComm should Khrushchev, prompted by the blockade, not back down. The abysmal failure of the United States’ previous attempt to invade the island with the help of 1,200 CIA-trained Cuban refugees in 1961 prompted the US military to prepare a comprehensive plan for invading the island with over 90,000 soldiers.Footnote ²⁰ After a comprehensive six-day bombing campaign specified in operations plan (OPLAN, for short) 312, adjusted for the intelligence gathered about the position of Soviet and Cuban forces during the U-2 and other reconnaissance overflights, amphibious army and marine divisions, as well as several airborne divisions, were supposed to land at various points concentrated around Havana. Importantly, all of the Soviets’ tactical nuclear weapons on the island, as well as all MRBM units, were ready starting September 27. Even more importantly, while the United States knew about the strategic weapons and knew that many of them were ready to fire, it knew next to nothing about the roughly one hundred tactical nuclear weapons deployed.

There are two scenarios in which a nuclear altercation between the Soviet Union and the United States could have begun in the context of an invasion of Cuba. First, the relentless five-day bombardment of Cuba that was supposed to proceed with the invasion might have led local commanders to “save” their MRBMs by firing them with or without the Kremlin’s orders. Just imagine that only one or two of them might have come through, destroying significant parts of any American city within reach. Such an attack would have produced immediate casualties in the tens or even hundreds of thousands. Is it doubtful that the United States would have reacted with anything less than the complete nuclear annihilation of Cuba, killing a significant part of its 6 million inhabitants and of the more than 40,000 Soviet troops? How would the Soviets have reacted to such a counterattack? There is a possibility that at this point they may have “cut their losses” in order not to endanger the Soviet Union itself. After all, they knew about their nuclear inferiority. Of course, the sheer number of casualties in the United States might have also prompted Kennedy, or at least some of his commanding officers who thought that nuclear war with the Soviets was inevitable anyway, to attack the Soviet Union directly. What if any of the US bomber pilots constantly circling the Soviet Union and temporarily out of touch would have concluded that World War III had begun, prompting them to pursue their designated targets?

Second, even if the United States had succeeded in taking out all of the Soviet ICBMs, MRBMs, and bombers in the initial days of the air raids, it is extremely unlikely that it would have also destroyed all of the tactical nuclear weapons on the island because they did not know where they were located. There is little doubt that these tactical nuclear weapons would have been fired on invading US troops, leading once more to high casualty scenarios, even if not reaching the numbers of Soviet MRBMs hitting major American cities. Once more, an American nuclear attack on the island would have been likely, as more recently declassified documents show.

Even if, in the course of bombardment and invasion, SIOP-63 had never been implemented in spite of the American practice of rehearsing all-out war, and the conflict could have been essentially limited to the Cuban war theater, the means chosen to achieve the putative political ends remain totally disproportionate to the possible costs. Overarching longer-term goals are lost sight of from within a new everyday struggle that develops its own logic; the forest recedes behind the trees, and strategic rationality flips into the kind of tactical rationalities that appear irrational from a strategic level. The nuclear weapons meant to protect a nation from harm thus can quickly become its nemesis.

Electoral limits on policy options

Throughout the entire missile crisis, Kennedy showed a deep concern for the ways in which his actions (or nonactions) were perceived by the American public (May and Zelikow Reference May and Zelikow2002, passim). How intense the domestic audience orientation in devising action plans really was is revealed by a short exchange between Robert and Jack Kennedy on the eve of the blockade:

So far, the conversation elucidates the putative lessons learnt from “Munich.”Footnote ²¹ But then Robert shifts the conversation to the domestic arena:

In modern mass-mediated democracies, successful performances conveying a message – not facts or issues – win electoral votes (Lempert and Silverstein Reference Lempert and Silverstein2012; Postman Reference Postman1985, chap. 9). Accordingly, political incumbents and their offices are mainly instituted through performances and the public’s reactions to them. Kennedy trained to be good at delivering poignant messages.Footnote ²² He was quite self-conscious of his (and his wife’s) youthful good looks, and he used them to his fullest advantage (Eschner Reference Eschner2017; Soddu Reference Soddu2012). Kennedy was unusually young in every office to which he was elected. He learned to compensate for a dire lack of experience with powerful performances characterized by a charming quick-wittedness, through which he communicated an energetic actor image that instilled hope in people. His quick rise to the presidency corroborated his self-understanding as a political performer. Looking ahead to the 1964 elections, Kennedy was keenly aware of his narrow win in 1960. In 1964, he would have to run on the messages he could spin from his record.

By the second half of 1962, Kennedy had an acute public relations problem, however. His first year in office was marred by consecutive failures: the botched invasion of Cuba in April, a summit with Khrushchev that made him look weak in June, and the building of the Berlin Wall in August. Domestically, he had not been able to score either. He perceived the rising Civil Rights movement and Southern governors’ attempts to crush it as a distraction from foreign policy. He feared a confrontation with the segregationist Southerners in his own party while coming increasingly under fire from the Democrats’ liberal wing for his lack of support for their civil rights agenda. With midterm elections ahead, Kennedy needed to turn around his presidency in the foreign policy arena.

As soon as information arrived in August testifying to increasing numbers of Soviet men and weapons arriving in Cuba, Republican members of Congress, along with parts of the press, began to criticize Kennedy for letting it happen. Starting in September, Senator Keating (R, NY) claimed that he knew for a fact that the Soviets were building offensive capabilities in Cuba – a fact that the Kennedy administration denied until the blockade announcement (Sherwin Reference Sherwin2020, 204–6). Once rumors emerged from Cuba through the refugee networks that Keating’s allegation was probably right, leading politicians and parts of the press criticized the administration for “doing nothing.” Sensing a weak spot, the Republicans made Cuba a major campaign issue. Kennedy involved himself in campaigning to project an image of firmness and reasoned responsiveness to the unfolding situation in Cuba.Footnote ²³ Republican critics got Congress to preauthorize the administration to use force against Cuba in the hope of prompting Kennedy into action (May and Zelikow Reference May and Zelikow2002, 20). As the crisis intensified, so did public pressures, making Kennedy increasingly nervous. While a few peace activists urged negotiations to avoid nuclear war, the expectation articulated in major news media and by politicians was that no Soviet bases in Cuba could be tolerated. As soon as negotiated missile swap scenarios entered the discourse, these too were seen mostly critically. No wonder, then, that the “do nothing” option was never seriously considered in the ExComm meetings, not even as a last resort if all else failed, and that almost everybody spoke out against missile swaps. And only after Kennedy’s blockade announcement, five days before the eventual resolution of the crisis, did Republicans withdraw Cuba as a campaign issue.

Following the logic of institutiosis, the question must then be why the Republican strategy of taking Kennedy to task on Cuba could possibly succeed. In answering this question, two institutional arrangements especially matter: the understandings governing the practices of campaigning and the concerns and preferences of voters as they were understood by campaigning politicians and the media. World War II gave rise to the norm that in times of crisis, foreign policy should not be subjected to the antagonism of partisan politics. Indeed, by all objective measures, there was a broad foreign policy consensus for the first quarter-century after the war.Footnote ²⁴ However, this consensus was not the result of a fact-based and reasoned debate leading to agreement on values and aims, nor of a polite self-constraint by the opposition to rally behind the president in foreign affairs. Instead, it was the effect of a peculiar form of partisan politics in which both parties thought that they had to outcompete each other in representing American exceptionalism and anticommunism.

Prompted by the souring relations with the Soviet Union as World War II ended and the massive strike waves that followed, domestic anticommunism became a campaign issue in 1946. Four years later, the global “communist threat” was added (Craig and Logevall Reference Craig and Logevall2009; Logevall Reference Logevall2012; Mann Reference Mann2001). A series of international events and their prevailing interpretation provided corroboration and recognition for the plausibility of this shift. Mao’s 1949 victory in China led the Republican Party, aided by significant parts of the pressFootnote ²⁵ and the China lobby, to accuse President Harry S. Truman of having “lost China.” The 1949 testing of a Soviet nuclear device led the Republicans to accuse Truman of having gambled away the American lead in nuclear weapons by lacking vigilance against Soviet spies. And then in 1950, the continuous border skirmishes between North and South Korea blew up into the invasion of the South by the North. Given the initial success of northern troops, it looked as if Truman might “lose Korea” as well.

The interpretation of these events resonated with the older strata of instituted anticommunist understandings, both discourses and feelings, that had come to be cultivated by business leaders, politicians, and the media after the October Revolution. These sentiments motivated passing laws and forming organizations dedicated to the anticommunist struggle. While red scares came in waves, they could always build on an available institutional infrastructure that was previously established (Ceplair Reference Ceplair2011). In the afterwar years, anticommunism blended with “American exceptionalism,” a strong set of discursive and emotive understandings that the United States is not only unique but superior to other countries, who are called upon to model themselves on it (Lipset Reference Lipset1997; Tyrell Reference Tyrell2022). The astounding two-front victory of the United States in World War II and the United States’ successful initial guidance of both former adversaries to become prosperous democracies provided new corroboration for exceptionalism. This has remained the root of the understanding that the United States can actually convert adversarial dictatorships into democratic allies through military defeat and subsequent economic aid and administrative guidance, which became prominent again in the Iraq War and the War in Afghanistan.

Deeply institutionalized as they were, anticommunism and exceptionalism were mobilized in every electoral cycle. Kennedy himself had avidly campaigned by blaming Eisenhower and Richard Nixon for having “lost Cuba.” He did so in spite of having previously argued that social conditions on the island (to which the United States had contributed significantly) were to blame for the revolution (Schlesinger [Reference Schlesinger1965] 2002, chap. 7). He also eagerly jumped onto the “missile gap” scare, which asserted Soviet superiority in the development, production, and deployment of nuclear missiles. Following the governmental Gaither Report, the missile gap scare was continuously fed by self-serving intelligence reports based on fear-driven speculations extrapolating from shallow circumstantial evidence.Footnote ²⁶ Indeed, the interest and ideological investment in the missile gap myth was so deep that the null results of the U-2 overflights were systematically ignored. Only the first spy satellite images providing comprehensive area coverage settled the issue in the second half of 1960. Kennedy, the versed campaigner, chastised Eisenhower and Nixon for letting a missile gap emerge. He continued to do so after he had been debriefed by the Department of Defense and informed that there was, in fact, no missile gap (Preble Reference Preble2003, 109).

Talking tough on Cuba and communism were thus basic declarations of allegiance to a common credo through which politicians established their bona fides. This was also true for high-ranking bureaucrats in the administration. It was therefore an imperative of political survival for Kennedy to get the Soviet missiles out of Cuba – one way or another. The very fact that he had not intervened much earlier, delaying a response until the missiles were operational, would have ended up as a huge political liability had Khrushchev not backed down in response to the blockade. Kennedy’s watchful waiting approach to the crisis had already cost him credibility with military leaders, and more belligerent bureaucrats. The institutional dynamics within which Kennedy was caught, the logics of creating credibility in the ExComm meetings at the time, strongly pressured him toward an invasion.Footnote ²⁷ This does not mean that it would have been entirely inconceivable that Kennedy would have delayed an invasion to try a negotiated solution after all. To make such a counterfactual plausible, it is again important to look for an institutional anchoring. Indeed, a few of his advisors might have been quite willing to support such a move either for reasons of conviction (Stevenson, George W. Ball, Arthur M. Schlesinger, perhaps Llewellyn Thompson) or for personal loyalty toward Kennedy (Robert Kennedy, Salinger). Yet the Kennedy brothers intensely disliked Stevenson (see also below), and this coalition would have had a hard time coming together, especially since Schlesinger and Ball were Stevenson protégés. In sum, then, the institutional pressures on Kennedy – those self-inflicted through his ambitions and those exercised by the majority of his political and of his military advisors – pointed in the direction of an invasion beginning October 29, with all the dangers outlined in the last section.

How big these pressures were (subjectively perceived) and how willing Kennedy was to satisfy them can also be gleaned from the secrecy surrounding the missile trade. Not only was it not discussed publicly after the deal was brokered, but it remained a secret even to most of his advisors. Kennedy could not publicly own the missile trade even though it was not much of a concession at all, given that everyone agreed that the liquid fuel Jupiter system deployed in Turkey was obsolete. The diplomatic problems it caused with a NATO ally, Turkey, were quite resolvable. Kennedy even denied the deal in postcrisis debriefings with Eisenhower, Truman, and J. Edgar Hoover (Sherwin Reference Sherwin2020, 460–61).Footnote ²⁸ There is further evidence in the fact that after the crisis Kennedy colluded with two journalists, one of them a close friend, to publish a longer piece in which Kennedy was praised for his cool sure-handedness in staring down Khrushchev, while Stevenson was denounced as advocating for a second “Munich” precisely by proposing a missile trade (Aslop and Bartlett Reference Aslop and Bartlett1961). Kennedy was shown the article by his friend before publication, and he apparently did not object. One of Kennedy’s advisors, Schlesinger, was so outraged about the article’s misrepresentation that he asked Kennedy twice to make a public statement on behalf of Stevenson to correct the record.Footnote ²⁹ Kennedy chose to remain silent in public and lied to Stevenson in a personal letter that he had not known about the article beforehand.

In a later interview, Kennedy chalked up the inner dynamic of the crisis moving toward a military confrontation to hardliners both in Moscow and in the United States (Sherwin Reference Sherwin2020, 10). From an institutional perspective, this analysis falls short. The power of assuming a hardliner stance in the United States lay in the fact that it reflected deeply instituted public understandings. Ironically, under the historical circumstances outlined, electoral accountability led to a limitation of policy options that were contrary to the most fundamental interest of the electorate: its own continued existence.

Conclusions: A few tentative lessons

The central scandal of the Cuban Missile Crisis lies in the fact that various institutional and, therefore, in principle alterable social arrangements had driven Kennedy and Khrushchev, against better intentions, to the brink of nuclear war. Institutional arrangements led them, in effect, to wager life on Earth for relatively minor strategic and political gains. This should have been then, and still is, a clarion call for a politics of institutional reform. I therefore want to return to the three types of knowledge I have identified as necessary prerequisites for politics (see above). The institutional dynamics of the Cuban Missile Crisis provide a simple form of dystopian knowledge about the relative ease with which direct confrontation between nuclear powers can lead to nuclear war. If these dangers were better known, renewed fear of nuclear war might help to mobilize a new peace movement. Sociological analysis that turns away from debates about the quality of individual political actors identifies two interinstitutional dynamics in the Cuban Missile Crisis that significantly increased the likelihood of war. These are potential targets for reform.

First, the command-and-control measures instituted in military strategies that promised political leaders full control over what was happening were showing significant systemic stress.Footnote ³⁰ The main reason for this stress is that any military under crisis or even war conditions is a different assemblage of institutions than in peacetime because its ongoing institutionalization is put on a radically different footing. This produces quasi-organically an increasing level of autonomy for field forces vis-à-vis the center. To label these common features of institutiosis as accidents is to minimize predictable institutional effects.

Moreover, such autonomy is not just a regrettable side effect but is necessary for successful institutional reproduction in wartime militaries. Hypercentralized institutional assemblages are easily destroyed. Delegation of control over spatially distributed weapons is the only way to maintain agency once communication fails. Nevertheless, central control remains highly desirable to coordinate the efforts of various subordinate units. At times, judgment is better placed at higher levels, and at other times at lower levels, and it cannot be known in advance which one is right. This too requires judgment.

Nuclear weapons make the command-and-control problem particularly acute because a “single shot” can kill very large numbers of human beings. The revenge feelings triggered by a successful nuclear attack (even of a tactical weapon killing only several thousand) will create psychological and political pressures to respond in kind, thus opening up the potential for uncontrolled escalation. For this reason, the political level needs a particularly tight command-and-control structure. However, command-and-control problems also become more acute with decreasing available response times to enemy moves. Missile attacks require responses in minutes. This, however, favors advance delegation.

Through the rituals surrounding the “nuclear football,” American presidents are given the illusion that they are fully in control at all times.Footnote ³¹ This is a dangerous myth. In thinking through military actions, leaders need to be aware that they will lose control due to unavoidable institutional dynamics. This loss of control can, to a certain extent, be reduced by organizational and technological improvements. Accordingly, the situation today is no longer that of the Cuban Missile Crisis. That said, new technologies and organizational forms always introduce new problems. Perfect control of instituted systems under contingent conditions is an illusion. It would require not only nonmanipulable, continuous communication and perfect local feedback but also human beings operating like automatons. The uses of AI move in this direction. However, this places the system at the mercy of algorithms derived from past experiences that are bound to work more or less well in open futures. Once more a new version of the problem emerges. To speak of command-and-control solutions as “fail-safe,” as Americans did during the Cuban Missile Crisis, is again dangerous myth-making. The conclusion is obvious: nuclear weapons are far too dangerous to be managed by any institutional technological system designed by human beings. The only remedy is to abolish them. Doing so in one fell swoop is politically unrealistic. Yet the logic of institutions makes clear that fewer weapons are better than many and that tactical nuclear weapons should probably be abolished first, not only because they are the least controllable, but also because their threshold for use is lower, while their potential for triggering quick escalation is still very significant.

The Cuban Missile Crisis shows how the interinstitutional dynamics between foreign policy and domestic politics can limit policy choices. Foe image-fueled forms of nationalism that rely on internal mobilization for electoral gain in opposition to an external enemy can quickly become self-defeating. The reason is that they can entrap political leaders in a form of posturing that may be helpful for their reelection while being totally detrimental to everyone’s long-term interest. Even if the removal of the missiles through international negotiations had failed in the Cuban case, the United States would have lived much better with Soviet nuclear missiles in Cuba than with the effects of an even limited nuclear war in the Caribbean. One may wish to press the argument even further: how many casualties was the United States willing to incur in an invasion of Cuba to rectify a quite manageable reduction of its still formidable strategic advantage? The US military estimated that (on the basis of comparable operations in the past) the casualty figure would have been several tens of thousands of soldiers. Just imagine! The fundamental problem is that the political rationale for an invasion could not even be posed in these rational terms because those presenting it would have appeared, given the prevalent nationalist sentiments, like appeasers demanding another Munich. Kennedy’s abysmal treatment of Stevenson after the crisis is a stark reminder of how easily advocates of peace are silenced and denounced in a context dominated by foe image-fueled nationalism.

Kennedy the campaigner, like nearly all other major politicians in the United States, fanned exceptionalist and anticommunist sentiments by talking tough on communism. Kennedy the president first felt driven to execute the CIA’s plan to invade Cuba in April 1961 because he feared the political backlash from canceling it (i.e., not only because the CIA assured him that it would work; Schlesinger [Reference Schlesinger1965] 2002, chap. 7). During the Missile Crisis, Kennedy managed to slow down the march into a military confrontation not least because there was a small group of advisors around Stevenson who urged him to prioritize negotiations. However, the interinstitutional dynamics at play in the situation were such that it would have been hard for him to resist an invasion had the blockade failed to yield the desired results.

The cost for the United States of not attending to counterproductive dynamics between foreign policy and cultivating a foe-image fueled form of nationalism in electoral politics has been immense. It was a central dynamic of the American wars in Korea (Cummings Reference Cummings2010), Vietnam (Logevall Reference Logevall2012), Iraq (Draper Reference Draper2020), and Afghanistan (Whitlock Reference Whitlock2021). Presidents of both parties and of very different temperaments were willing to begin and extend all of these wars beyond any reasonable limit, either to mobilize national support or because they felt compelled to fight for an unambiguous victory for fear of electoral punishment. The sociological upshot: Political communities would want to cultivate a form of communal spirit that rewards attention to the long-term collective interests of the community, which, sadly, is rarely the case.