The Exculpating Myth of Accidental War

The risk of war by miscalculation is a popular scholarly subject and a frequent concern of policymakers. Theories of war by miscalculation assume either ignorance by decisionmakers, bad faith by policy participants (usually the military), or a mechanistic series of exchanges that obviate the need for policymakers to display judgment and responsibility. Yet historians struggle to produce evidence that political leaders were either ignorant of the risks they were choosing or emasculated from any ability to affect the course of events. Theories of accidental war are, as Geoffrey Blainey has said, “a description masquerading as an explanation.”^[1] Instead of explaining historical events, the myth of accidental war provides an intricate puzzle for intellectuals, a justification for moralists to fear military subversion of political decisionmakers, and an evasion of culpability for those decisionmakers.

The debate is not esoteric—it has urgent policy applications. U.S. policy regarding Russia’s February 2022 full-scale invasion of Ukraine has been heavily influenced by U.S. President Joe Biden and his administration’s concerns about “a third world war.”^[2] Russia violated the sovereignty and the internationally recognized territory of Ukraine, despite its commitment to respect Ukraine’s security in the 1995 Budapest Memorandum. Failing to achieve its political objectives on the battlefield, Russian strategy shifted to targeting civilian dwellings, destroying infrastructure, and committing mass atrocities. All of these actions violate Russia’s international commitments and overtly challenge the Biden administration’s leadership of the rules-based international order, as well as its pledge to “stand against human rights abuses wherever they occur.”^[3] 

Although rational calculations of power ought to dictate Russian aversion to expanding the war from a trajectory of losing to Ukraine to one of more rapidly and decisively losing should the United States become a direct combatant, the Biden team’s policy is predicated on concerns that uncontrollable escalation could result from U.S. assistance to Ukraine. To avoid an accidental war with Russia, the Biden administration has defended the international order by timidly progressing from declining before and in the early stages of the invasion to provide weapons to Ukraine because they would quickly end up in Russian hands if Russia won as expected, to refusing to provide “offensive weapons,” to providing “offensive weapons” but not those that could reach Russian territory, to providing weapons that could reach Russian territory but making Ukraine promise not to use them at those ranges, to allowing Ukraine to target Russian forces in Crimea, to providing tanks and training in the United States.^[4] Out of concern the United States could unavoidably be drawn into direct combat against Russia, the strongest power in the international order is conveying the limits of its interest to a violent transgressor losing a war against a lesser power.

The idea that war happens by accident is prejudicing U.S. policy toward self-deterrence.

The idea of accidental war also influences U.S. policy toward China, the “pacing challenge” in its great power rivalry, according to the Biden administration’s October 2022 National Security Strategy.^[5] The United States publicizes Chinese infractions, like the construction of island military bases in the South China Sea, but does not enforce compliance. And China is playing that concern to its advantage, undertaking demonstrations of its willingness to violate U.S. sovereignty, initiating dangerous military incidents, explaining provocations as its military acting without political knowledge, and refusing to answer military hotline calls during crises. This shows that the errant notion that war happens unaccountably, that political leaders are powerless to prevent escalation, that militaries will act independently of political control and blunder states into conflict has immediate policy relevance. The idea that war happens by accident is prejudicing U.S. policy toward self-deterrence; national security experts should instead be more concerned about deliberate choices for war as they think about deterrence and escalation control. 

What Is Meant by Accidental War?

It is very difficult to know with precision the combination of political commitment to objectives, military power, the ability to marshal economic resources, and the social fortitude that combatants possess until they fight. Blainey persuasively has argued that war is how relative power is determined in the international order.^[6] As Carl von Clausewitz explained and Russia’s invasion of Ukraine demonstrates, war is a gambling game. If accidental war means nothing more than leaders misjudging their relative power, then every war is a mistake. This is because, as Blainey has concluded, “war is usually the outcome of a diplomatic crisis which cannot be solved because both sides have conflicting estimates of their bargaining power.”^[7] 

Strikingly, across nearly eighty years of the nuclear age . . . there have been few accidents, and none of them led to an automatic escalation or failure by political and military leaders to make judgments dampening the prospects of war.

To be intellectually useful, a robust definition of accidental war matters. At its most basic, accidental war could be mischance—a confused signal, the technical failure of a weapon, or the accidental launch of a missile. Alexander George has termed this kind of scenario “inadvertent war.”^[8] It is striking that across nearly eighty years of the nuclear age, incorporating U.S. and Soviet arsenals in the tens of thousands and less developed countries designing indigenous arsenals and delivery systems, there have been few accidents, and none of them led to an automatic escalation or failure by political and military leaders to make judgments dampening the prospects of war.

But although scholars engaging in the debate use the terms accidental and inadvertent war interchangeably, genuine inadvertence is mostly not what is meant in the debate about accidental war; rather, mostly what is meant is willful subversion of political intent.^[9] George distinguished inadvertent from accidental war, defining the latter as “one that starts because of unauthorized activities by individuals below the leadership level in the chain of command.”^[10] Meanwhile, Marc Trachtenberg has defined accidental war as “one in which the political process— a process that would normally lead to a peaceful settlement of the dispute at hand—is overwhelmed by forces welling up from within the military sphere . . . If the political leaders thought the risks were relatively limited, but the military structure in place created a much greater level of risk in time of crisis, then one could conceivably have what, from the point of view of the political leadership, would be an ‘accidental war.’”^[11] Scott Sagan also has narrowed the aperture of unauthorized activities to the military, defining wars as accidental only when “there would have to be some activity or incident inside the military machine, without which war would not have occurred.”^[12] 

The idea of accidental war has been of particular interest to political scientists rather than historians, which seems a significant detail—the experts on concepts are intrigued, the experts of the case studies find no corroborating examples. No less a source than the distinguished military historian Michael Howard has rejected the idea, writing, “if history shows any record of ‘accidental’ wars, I have yet to find them.”^[13] 

An exhaustive study by Evan Luard of wars dating from 1400 CE to 1985 did not turn up a single case of accidental inception.^[14] Sagan, a political scientist and proponent of the accidental war idea, has admitted as much, acknowledging, “in the major works by historians on the causes of war, however, the whole idea of accidental war is either conspicuous by its absence or explicitly dismissed as conceptually confused and historically irrelevant.”^[15]

Case Studies

While historians have disavowed accidental war, political scientists have put forward several cases that merit consideration: Egypt initiating war against Israel in 1967, the Seven Years’ War, World War I, Japan’s 1931 invasion of Manchuria, and U.S. debates over nuclear strategy. The 1967 War and the Seven Years’ War easily fail the test. World War I, the canonical example of sleepwalking into war, also fails, and in ways revelatory of the decisive political actors involved attempting to shed responsibility for the disastrous outcomes they set in motion. 

What is especially interesting in the World War I case is the emergence of significant civil-military friction. That friction also manifests in U.S. debates over military strategy, becoming both the basis for concern among American political scientists about accidental war and a means for derogating the expertise of military participants in the policy process. The only case that clearly presents as a war that political leaders did not explicitly choose is the invasion of Manchuria, and that one occurred not as an accident, but as part of a broader military takeover of the Japanese government, so its applicability is limited for countries with militaries subordinated to civilian control.

The 1967 War

Egyptian leader Abdel Nasser mobilized and deployed forces to the Sinai, expelled the United Nations (UN) forces, and closed the Straits of Tiran despite Israeli declarations that it would go to war to keep the straits open. The case for Egypt inadvertently precipitating the war leans heavily on Nasser’s statements that he did not believe Israel’s threats, although his actions give contradictory evidence, and on supposed psychological avoidance of an unpleasant reality.

But Egyptian moves in 1967 nicely conform to Blainey’s (and Clausewitz’s) explanation that wars occur because states believe they can attain their political goals by military means. There is little need to resort to a psychological evaluation of Nasser beyond the evidence that he expected war with Israel over closure of the straits, and he chose to take that action. Whether or not he expected war, he acknowledged it as a possibility resulting from his choices. To suggest that he could not accept the loss of prestige of reversing course is simply to say, as Blainey does, that war was preferable to other outcomes.

The Seven Years’ War

The war between France and Great Britain over control of North America, known among Americans as the French and Indian War and in Europe as the Seven Years’ War, decided British predominance (and, incidentally, reduced American settlers’ need for British military protection). The governors of both Virginia and Massachusetts sent alarmed but inaccurate information to the British government in London, then embellished in recounting, that prompted the government to empower governors to defend its forts, which the French subsequently attacked. Sagan has presented the start of the war as an example of conflict started by false reports that “inadvertently magnified its insignificant original cause into a wider conflict.”^[16]

In Sagan’s view of accidental war, “professional military organizations are seen as tools of national security, but tools that take on a life of their own and are very difficult [for] statesmen to control when implementing the state’s security policy in an international crisis.”^[17] None of the evidence Sagan marshalls for the Seven Years’ War, however, was propagated by military authorities, so it does not meet his own definition of accidental war. It was the civilian governors of Virginia and Massachusetts and the head of the Board of Trade that were whipping up concerns. And the French did, in fact, attack British forts, the proximite cause of the war. 

Sagan’s evaluation is that the purported accident of the Seven Years’ War was that neither side genuinely intended to establish control over the Ohio Valley; yet that circumscribed motive excludes the broader stakes that the British and French governments were contesting in the mid-eighteenth century, the primacy of Europe in both sides’ calculations, and the specific acts of war committed by French military forces consistent with French government policy.

World War I

World War I is the case that scholars love to litigate because every state that fought it lost. Even the putative winners were diminished: Germany was crippled by reparations and territorial loss, France was humbled, Russia underwent a revolution, Belgium was laid waste, and the United Kingdom (UK) was nearly bankrupted and beholden to the United States. Even the United States, the moral victor, had its international ambitions hobbled by European condescension and domestic refusal. From the ghastly outcome, it is not clear why any country would have chosen it. Yet leaders of each state that committed to war “insisted on pursuing aggressive policies that they knew risked a localized war at best.”^[18] They all believed that choosing war in the near term was preferable to fighting it later.^[19] And Germany pointedly chose war rather than conciliation during the July crisis of 1914.^[20]

Military perfidy is central to the Guns of August theory of World War I, which posits that rigid military mobilization schedules of several countries as well as the temporal demands and violation of Belgium and Luxembourg’s neutrality of the Schlieffen Plan unacknowledged by the military to unsuspecting political leaders were the causes of war. Thomas Schelling uses World War I as his case for accidental war in his book Arms and Influence.^[21] One influential factor is the charge that Germany’s military leadership undercut a peaceful settlement that their political leadership was negotiating. 

German historians have debunked the argument of automatic escalation and the out-of-control military explanation [in World War I].

But German historians have debunked the argument of automatic escalation and the out-of-control military explanation.^[22] [RD3]As Hal Brands has put it to summarize their work, “the taproot of the conflict was ambition and risk-taking in Berlin.”^[23] Attempts to constrain responsibility solely to General Helmuth von Moltke the Younger’s communications to military counterparts in Austria overstates the role of the military in both countries. While Germany can be fairly judged to have been a military dictatorship by the end of World War I, it was not so at the war’s start.^[24] Germany’s civilian leadership, above all its kaiser and chancellor, weighed the risks of war and had outsized political ambitions that they were unable to achieve by nonmilitary means. Advocacy for war by von Moltke to his Austrian counterpart was not determinative, even if it had been in time to affect government decisions that had already been taken.^[25]

Sagan’s case for World War I , with Chancellor Theobald von Bethmann-Hollweg losing control of policymaking, accords well with Blainey’s and Trachtenberg’s critiques of accidental war. Trachtenberg rightly has pointed out that the fundamental assumption is false of theories of war by miscalculation, “if only we can bring ourselves to calculate more accurately, we could greatly reduce the risk of war.”^[26] It is an unreasonable expectation that anyone can assess risk and the perceived risk of adversaries with precision or even fidelity. As he said of the German government at the inception of World War I, “control and discipline were less than perfect, but this is the way states actually operate, and this kind of thing should be taken as the norm.”^[27] Theories that rely on perfect information and the lack of friction are failed theories. 

Japan’s Invasion of Manchuria

Bitter policy disputes between civilian and military leaders in Japan were already resulting in assassinations in the 1920s; by 1930, the military was the dominant political force in the country, and its assent was necessary for any government. In 1931, the Japanese military fomented an incident in Mukden as a pretext for invading all of Manchuria, ignoring orders from the Japanese imperial government. 

But the military’s actions were consistent with the widely accepted view within the Japanese government that control of all Manchuria was essential to the prosperity of Japan. Nor was insubordination by overseas Japanese military commanders uncommon, as Niall Ferguson has noted.^[28] Blainey has pointed out that the Japanese government “did not recall nor demote the officer activists,” instead defying the League of Nations’ demand that they retreat, thereby legitimating the military’s decisions and suggesting either that the invasion was in line with the government’s objectives or that civilian leaders were dependent on the military.^[29] 

The initiative taken by rebellious officers in Manchuria may have been premature, but it was consistent with the consensus in the Japanese government that control of Manchuria was essential to Japanese security and prosperity.

I share Sagan’s judgment that the Japanese military’s unauthorized 1931 attack in Mukden “is perhaps the most clear-cut historical example” of accidental war and a cautionary tale about the rise of militarism in Japan.^[30] The risk is, however, unextrapolateable to societies in which the military is under civilian control. The invasion of Manchuria was essentially a military coup against the Japanese government, and it threatened to be extended to the Japanese home islands too if the military’s demands were not met. This is most accurately seen in the context of the 1932 military coup in Japan than in isolation as an accidental war. I am less confident than Sagan that “war would have been avoided” without the military’s unauthorized actions in China, though, given the sweep of Japanese ambitions and Japan’s other, authorized military occupations.^[31] The initiative taken by rebellious officers in Manchuria may have been premature, but it was consistent with the consensus in the Japanese government that control of Manchuria was essential to Japanese security and prosperity.

Seven Days in May

Explanations of wars that hinge on military leaders subverting political directions or lacking the elegance of understanding that civilians possess became particularly influential in the aftermath of World War II, a time that coincides with the dawn of the nuclear age. Concerns about accidental war were particularly acute after the Soviet Union crossed the nuclear threshold. While fears of surprise attacks were not new, nuclear weapons changed decision calculations by offering the prospect of surprise attacks on a society-destroying scale.^[32] Great power relations were particularly fraught as the United States and Soviet Union tested each other’s interests and determined the practice of deterrence in the nuclear age. 

Three important transitions took place at the time related to the risks of military supersedence of political authority, and these concerns gave widespread influence to the idea of accidental war starting around 1960. The first transition was the introduction of nuclear weapons, which could produce devastation on a scale previously unimaginable by a single decision and use. Then president Dwight D. Eisenhower worried about accidental war in the nuclear age; Trachtenberg has attributed this to Eisenhower’s concern about nonchalant attitudes about nuclear weapons, but my guess is that the idea more likely stemmed from his rich experience of things going wrong in combat and genuine intellectual wrestling with the meaning of weapons of such destructive magnitude.^[33] Even before the advent of permissive action links and tightened nuclear release procedures, accidental wars were both desperately feared but not in evidence. 

The second transition was the advent of civilian strategists, predominantly academics with economics training. Civilians brought a ratio-centric perspective to the work of defense planning. Schelling, the intellectual architect of so much deterrence theory, worried early on about accidental war. He later professed that he was mystified at the unwillingness of war game participants in the 1960s to cross the nuclear threshold even when it meant losing a conventional war.^[34] The “competition in risk-taking” he identified has been voluntarily constrained by policymakers because of the magnitude of the consequences in the nuclear age.^[35] This development suggests that the more significant the consequences, the more cautious policymakers tend to be, therefore making accidental war less likely.

And the third transition was the coming of age of politicians who had little or no military experience of World War II, a generational change that politically advantaged derogating the superior credentials of the military leaders who had planned and fought the war. Most American policymakers who had fought World War II, particularly those who had been senior military officers, considered nuclear weapons merely an extension of conventional power for planning and operations purposes. This culture clash with the military is best summed up in Alain Enthoven rebutting a senior military planner by saying “General, I’ve fought just as many nuclear wars as you have.”^[36] Nor was Enthoven an unrepresentative case. Herman Kahn also argued that the nuclear age was wholly novel, saying, “the unrealized and unexperienced, but historically plausible, problems of World Wars III and IV are more valuable than the experienced problems of World Wars I and II.”^[37]

Although the American military has routinely passed both those tests, the theory of accidental war casts a shadow of doubt about their reliability.

There are two significant tests of civilian control of the military in the American system: the first is whether elected leaders can ignore military recommendations, and the second is whether the military will carry out policies that elected leaders choose contrary to military counsel. Although the American military has routinely passed both those tests, the theory of accidental war casts a shadow of doubt about their reliability. Both Eisenhower and Kennedy rejected military advice during the 1958 and 1961 Berlin crises, and Kennedy did so again during the 1962 Cuban Missile Crisis, yet none of those instances have served to tamp down concerns about military processes and leaders plunging the nation into wars that political leaders sought to avoid.

During the 1960 presidential campaign, Eisenhower authorized the sharing of intelligence photographs of Soviet nuclear forces collected by U-2 overflights. The reaction of William Kaufmann from the Kennedy campaign was disbelief that they were designing and securing their forces so sloppily. Despite the recklessness of especially early nuclear practice, there have been no accidental nuclear wars or even accidental conventional wars between major powers in the nuclear age. Far from being a fragile balance of terror, the nuclear age has ushered in greater caution by the strongest powers and even middle power diads when engaged in conventional wars.

The Kennedy administration was obsessed with concerns about accidental war; it was a major driver of the decisions to restrict nuclear launch authority to the president, to install permissive action links on weapons, and to embrace the strategy of flexible response.^[38] These three transitions—the advent of nuclear weapons, the emergence of civilian strategists, and distrust of military judgments—were all in evidence during a Soviet-precipitated crisis over Berlin early in Kennedy’s tenure.

At his first meeting with Kennedy in June 1961, Soviet Premier Nikita Khrushchev re-issued the ultimatum he had given Eisenhower in 1958: within six months, Western powers needed to withdraw their forces from Berlin and make it a demilitarized city or the Soviets would allow East Germany control of access to it (repudiating the Four Power Pact from World War II). The Eisenhower administration, operating with the confidence of a war-winning U.S. president after six years in office, refused, in a coordinated response with the UK and France. It commenced military planning for preservation of access to Berlin under the North Atlantic Treaty Organization’s (NATO) Supreme Allied Commander Europe (SACEUR), General Lauris Norstad.

The Berlin planning encompassed a variety of circumstances in which NATO forces would prepare “quiet military measures” visible to the Soviets signaling Western commitment and probe Soviet intentions, but quickly escalate from conventional conflict to general nuclear war.^[39] Eisenhower said publicly: 

We are certainly not going to fight a ground war in Europe. What good would it do to send a few more thousands or indeed even a few divisions of troops to Europe? I do not see why we would think that we—with something of a half a million troops, Soviet and some German in East Germany, with 175 Soviet divisions in that neighborhood—why in the world would we dream of fighting a ground war? . . . We want to keep adequate forces and we want to keep as strong in our hearts and our heads as we do in our military, and then carry forward our policies, our firmness in supporting our rights, carrying out our responsibilities in the world, keeping our friends together.^[40] 

Eisenhower’s view that a conventional war was sanguinously pointless—both an unnecessary expense of lives and a sure path to escalating to nuclear use—was broadly shared in the U.S. military and was the strongly preferred policy among allies whose countries would be the conventional battlegrounds.

Norstad supported that view for both strategic and allied reasons during both the Eisenhower and Kennedy administrations. His debates with Kennedy and the secretaries of state and defense defined the doctrine that became flexible response. Norstad’s position never varied: 

I have always accepted the possibility of limited military action in Europe, but normal prudence compels us to realize that, although limited aims may involve only limited effort in other areas of the world, in this critical, highly sensitive region, the assumption that the more likely contingency is one short of nuclear or massive non-nuclear attack does not provide an acceptable basis for planning. A number of our important NATO allies firmly believe this; it is my own judgment that it is so.^[41]

Civilian leadership in the Kennedy administration disagreed; their concern about accidental war led them to believe that war in Europe needed to be ratcheted up slowly and conventionally, giving repeated opportunities for Soviet recalculation. Kennedy described his thinking to Norstad, saying, “it seems evident to me that our nuclear deterrent will not be credible to the Soviets unless they are convinced of NATO’s readiness to become engaged on a lesser level of violence and are thereby made to realize the great risks of escalation to nuclear war.”^[42]

According to then secretary of defense Robert McNamara, Norstad responded, “‘Graduation’ and ‘escalation’ suggest a serial progression in which we move easily and by prepared steps from one stage to another of a development within our own control. This seemed to him unrealistic; he believed that in normal war escalation is apt to be explosive . . . The one central point of concern is this: do we intend to use our nuclear power if necessary?”^[43]

Norstad tried to finesse the differences, saying, “I have tried to develop a rather specific line of action which would meet your requirements and at the same time have a chance of being accepted by our Allies.”^[44] His proposal was a draft planning document containing the following guidance: 

We must prepare a broad catalog of plans to meet the situation at ascending levels of political and military involvement, but we must do so with full awareness that the choice of timing and action rests at least as much with the Soviets as with us . . . The use of armed force, even in limited quantity, risks the danger of explosive expansion to higher levels of conflict, including the highest level. This risk is accepted, and therefore we stand ready to use all forces and weapons available . . . This action might be general war, initiated by the West. If, however, there is reason to believe that the Soviets doubt the seriousness of the West, or if the Western public is not yet prepared for such large-scale military action, NATO forces will launch one or more limited offensive operations . . . The selected, limited-objective offensive operations discussed above are short-term actions which will have served their purpose soon after initiation, probably within hours. By initiating any operation in this category, the commitment of the West will become complete. Whether this step will lead to general war will be determined by the Soviet reaction, which may be prompt and total.^[45] 

SACEUR’s instructions to military planners at Supreme Headquarters Allied Powers Europe were never issued because, despite several meetings with the president, the gap was unbridgeable between Norstad’s thinking and the civilian leadership of the Kennedy administration. National security advisor McGeorge Bundy wrote to the president saying of Norstad “at root he is a nuclear war man, and all his preferences move accordingly.”^[46] The president’s military advisor, retired general Maxwell Taylor, advised him that “he is convinced that general nuclear war will be the inevitable and rapid outcome of any military action to reopen access to Berlin.”^[47] Then secretary of state Dean Rusk argued for a delay in responding to any provocation; Norstad countered, “it will be necessary to act, not wait—to engage ourselves in a fight—and not to let the interruption of our rights stand unchallenged. You have to react immediately, in order to see if the other man means war. Where will your allies be if you wait?”^[48]

Norstad was accurately reflecting allied concerns that the Kennedy administration was backing away from the U.S. commitment to extended deterrence. Whereas the Eisenhower administration considered the European question to be ensuring Germany’s adherence to the West, the civilian leadership of the Kennedy administration feared an accidental war and calculated that NATO allies had no other choice than to support U.S. strategy.^[49] The State Department optimistically concluded in 1963, “this US discussion has undoubtedly generated some allied questioning but, on balance, we believe that it has had a constructive educational effect.”^[50] France, and to a lesser extent Germany, would prove that wrong.^[51] 

After the Berlin Wall was constructed in 1961, the administration undertook a series of wargames at Camp David to test graduated escalation; both sides in the game sought accommodation and modification of their political objectives rather than provocation.^[52] War by accident never occurred, even when the stakes were theoretical.

Yet the president and his civilian team continued to fear a reprise of the German high command’s supposed inflexibility in 1914, creating plans that dragged an unknowing or powerless civilian leadership into catastrophe. Yet the civilians, not the military, were the prime movers on decisions; it validated their desire to assert civilian expertise in defense planning over that of their more experienced military colleagues. 

Although Norstad put into practice the civilian direction on NATO strategy, McNamara darkly hinted to senior government officials in 1963 that Norstad might start a war.^[53] Trachtenberg has concluded of McNamara that “his assumption here was that one could not trust the military leadership to do the right thing . . . the real problem had to do with the military authorities, with the way they thought, the way they operated, the sorts of things they were likely to do if they were given the authority to do them.”^[54] Sagan’s argument about the risk of accidental war burgeon from the same basis. Both serve to undercut legitimate military perspectives about the dynamic of conflict.

Saving the World

Two other incidents prominent in discussions of accidental war occurred in the 1980s. The first alarm over accidental war occurred in 1983, amid heightened tensions over NATO deployments of intermediate-range nuclear forces and the Soviet Union’s shooting down of a South Korean airliner. Erroneously alerted by Soviet early warning systems that the United States had launched an intercontinental ballistic missile salvo out of the blue, the duty officer, a Soviet lieutenant colonel named Stanislav Petrov of the Air Defense Forces, disbelieved the information absent corroboration and declined to launch a retaliatory strike.^[55] Colonel Petrov was able to understand the stakes and assess information in the context of patterns of behavior (the unlikelihood of an unprovoked U.S. nuclear attack), concluding that restraint was the proper judgment. This outcome demonstrates that, even at the tactical and operational level, civilian and military actors have the ability to control tensions rather than letting events prompt automatic escalation or having military officers operate on judgments untrustworthy to civilian leaders.

The second example of a narrow escape from accidental war was a November 1983 NATO nuclear exercise codenamed Able Archer. Designed to rehearse NATO nuclear alert and release procedures, the exercise simulated a crisis between NATO and the Warsaw Pact. The Soviet leadership, operating on “a potentially lethal combination of Reaganite rhetoric and Soviet paranoia,” feared the exercise was cover for a nuclear attack (as a KGB defector later described it).^[56] The Soviets put their nuclear forces on high alert; U.S. forces noticed the change but did not mirror the Soviet action, and the so-called crisis passed.^[57]

Hysteria about Able Archer being the riskiest confrontation since the 1962 Cuban Missile Crisis is not borne out by the facts.^[58] NATO predictably conducted this exercise at the same time every year, so the Soviet leadership had context to anticipate its occurrence. While some NATO heads of state participated, president Ronald Reagan did not (and explicitly to prevent Soviet conjecture, according to his national security advisor, Robert McFarlane).^[59] Central Intelligence Agency (CIA) analysts at the time concluded that Soviet leaders did not genuinely fear a U.S. attack (although former CIA director Robert Gates takes that as analysts failing to understand Soviet thinking).^[60] Sourcing for the crisis comes largely from debriefing a Soviet defector, who had an interest in heightening the drama of his revelations and his role in preventing escalation. The Soviet military leader of the operations directorate at the time, Marshal Sergey Akhromeyev, claimed not to recall Able Archer.^[61]

The Petrov incident and Able Archer, while often used as cautionary tales about the fragility of the U.S.-Soviet nuclear peace, actually demonstrate the robustness of what Eisenhower called “‘the stability of the stalemate.

The Petrov incident and Able Archer, while often used as cautionary tales about the fragility of the U.S.-Soviet nuclear peace, actually demonstrate the robustness of what Eisenhower called “the stability of the stalemate.”^[62] Far from demonstrating the risk of accidental war, the incidents indicate the lack of mechanistic escalation and the wide role for human judgment—and therefore culpability. As Blainey has summarized, “Perhaps many wars were like traffic accidents, the result of risky driving by nations rather than the result of a wish to crash into a rival.”^[63]

War by miscalculation is once again in vogue as worries about support to Ukraine intensify as Russia’s defeat becomes manifest. And once again, the risk is conjured as a rebuttal to a hard-line policy.^[64] Biden and his national security advisor frequently raise the spectre of needing to avoid World War Three as their justification for rejecting weapons in high quantities or of increasing Kyiv’s capabilities.^[65] This shows the detrimental real world significance of a mistaken concept about the causes of war.

Better strategies would mean taking less counsel of unfounded fears of accidental war and attempting instead to make clear judgments about interests, risk tolerance to protect and advance those interests, strategies for attaining them, and ways of effectively communicating with adversaries. To take a Ukraine example, rather than conveying to U.S. adversaries all they are not willing to do and stingily parsing out assistance to Ukraine out of fear that Russia might escalate to nuclear use against Kyiv, U.S. officials ought to understand that they have the ability to affect that choice and act accordingly. They could publicly state that if the U.S. intelligence community perceives any preparations for nuclear use, Washington would provide Ukraine the targeting information and weapons to preempt the Russian attack, or that if the United States were to fail to preempt it, U.S. officials would send NATO military teams to Ukraine to help manage the effects—so there would be NATO military forces stationed in Ukraine, and they would hunt down and either drag to the Hague or kill everyone involved in the decision and execution of the order. 

The Enduring Popularity of Accidental War

If the idea of accidental or inadvertent war has been so definitively repudiated by the historical cases, why, then, does it continue to command such intense interest? 

For starters, this is because such conjecture is fun. Political scientists play with concepts, teasing out patterns from bodies of evidence. Theoretial accidental wars are plentiful, even if history provides no actual examples. Moreover, the notion of war has grown scarier. The possibility of a single mistake or error of judgment decimating whole cities or countries, even destroying the world, intensifies the consequences of single points of failure.

The idea of accidental war . . . reifies civilian control of military planning . . . and allows leaders to evade culpability.

The idea of accidental war also reifies civilian control of military planning. In almost every example put forward as a case of accidental war by either political scientists or officials in positions of authority during wars, the military is blamed as a usurper of political directives against war or negligence. It is the excuse that political leaders of the time offered, especially in Germany during World War I. Soviet leader Joseph Stalin refused to participate in nuclear release exercises because he feared the military would trick him into an actual launch.^[66] To return to the previous Sagan quote, “for a conflict to be considered an accidental war, there would have to be some activity or incident inside the military machine, without which war would not have occurred.”^[67] Civilian mistakes leading to war are thus airbrushed out of the picture. 

The concept also allows leaders to evade culpability. If war were simply a car crash, that would allow policymakers to avoid responsibility for their assessments of their own and their adversaries’ relative power, their policy choices on the basis of those assessments, and the consequences of those choices—in effect encouraging negligence in the performance of their duties. In this way, British prime minister David Lloyd George could say “the nations slithered over the brink into the boiling cauldron of war without any trace of apprehension or dismay,” and Germany’s chancellor could say, “things are out of control [es ist die Direktion verloren],” treating their government’s choices as an act of God or the functioning of a natural force like gravity rather than political choices that proved disastrous.^[68] Nor are politicians the only governmental actors eschewing responsibility for wars’ outcomes, as Germany’s military claimed to have been winning the war but to have been betrayed by politicians in what came to be known as the “stab-in-the-back” myth or Dolchstoss im Rücken.^[69]

The concept of accidental war also avoids the question of ideology. War as an accident also has the benefit of not requiring engagement with the academically uncomfortable subject of regime type. Historians prefer unique explanations, and political scientists prefer systemic explanations. Ideology and religion are cultural factors and therefore more difficult to grasp. As novelist Orhan Pamuk has cautioned, “even the most intelligent thinker will, if he talks too long about cultures and civilizations, begin to spout nonsense.”^[70] Yet it is essential to understanding government decisions. Brands’ important challenge to the idea of accidental war noted, “World War I was part of the overarching struggle of the 20th century—the contest between liberalism and its enemies.”^[71] War as an accident circumvents messier and more contentious explanations.

Conclusion

Amos Tversky and Daniel Kahneman’s prospect theory gives some support to concern about war by miscalculation. Their idea, which explains lucky streaks and losing streaks is that people are willing to run greater risks to salvage a failing situation than to improve an already advantageous one.^[72] Applied to international relations, states on the losing side of wars or the disadvantageous side of crises ought to become desperate and take increasing risks. But that is not what appears to have happened at least among the great powers—not in the Berlin crises, or in the Cuban Missile Crisis, or during the 1956 Hungarian uprising. In those cases, billed as the most fraught superpower confrontations of the nuclear age, both the United States and Soviet Union accurately gauged their relative strength and reached political settlements or conceded to avoid war.

Moreover, manipulating risk is a strategy, not an accident. For example, former secretary of state Henry Kissinger’s elevation of nuclear alert levels during the 1973 Arab-Israeli war—something shockingly undertaken without presidential approval—was using the risk of nuclear war to prevent an adversary’s support to a fighting proxy. If that incident had produced war, it would taxonomically have qualified as a failed strategy, not an accidental war. As Trachtenberg has concluded, “the basic point here is that states, by and large, do know what they are doing; to a certain extent, they burn their bridges in the course of a crisis, but they do this with important political objectives in mind.”^[73] 

States can and do choose to de-escalate; there is no automatic ratchet to upward escalation. For example, president Barack Obama declined to honor the red line drawn against Syrian use of chemical weapons in 2012.^[74] Even the most reckless of recent vintage American presidents, Donald Trump, was able to make reasonable risk calculations in crises, by backing away at great reputational cost from striking Iran to retaliate for Iranian attacks on Saudi Arabia and the United Arab Emirates.^[75]

Fear of accidental war apparently causes policymakers to assess whether threat indicators are consistent with prior behavior, providing context that steadies the hand that otherwise might genuinely cause that rare phenomenon of an accidental war.

Advancing the fiction of accidental war does, however, have one potential policy advantage. Fear of stumbling into war can have advantageous policy implications by inducing caution and clarity in crises, as it apparently did on Kennedy.^[76] Especially among nuclear powers, the obviously destructive consequences seem to instill caution. As Trachtenberg has argued, policymakers look for patterns; accidents are routinely met with efforts at confirmation of intentions; fear of accidental war apparently causes policymakers to assess whether threat indicators are consistent with prior behavior, providing context that steadies the hand that otherwise might genuinely cause that rare phenomenon of an accidental war.^[77]

Kenneth Boulding once made the parallel claim that “threat systems are the basis of politics as exchange systems are the basis of economics.”^[78] For all of the intellectual elegance of theorizing about the causes of war, Blainey’s simple evaluation seems most analytically sound: “if two nations are deep in disagreement on a vital issue, and if both expect that they will easily win a war, then war is highly likely. If neither nation is confident of victory, or if they expect victory to come only after long fighting, then war is unlikely.”^[79] Trachtenberg would add that war is also unlikely if the powers engaged genuinely want to avoid it.^[80] And that has been the story of war and peace, especially in the nuclear age.

About the Author

Kori Schake leads the Foreign and Defense Policy Studies team at the American Enterprise Institute. She is the author of Safe Passage: The Transition from British to American Hegemony and has taught at Stanford University and the U.S. Military Academy. 

---

References