Dan Reiter, Paul Poast
A pillar of American grand strategy since 1945 has been the deployment of forces — sometimes smaller and sometimes larger — abroad. A key logic underpinning smaller deployments is that they serve as tripwires: Attacking them is assumed to inevitably trigger broader intervention, deterring aggression. We question this logic. Not only are small tripwire deployments unlikely to prevent an attacker from capturing its objective and establishing a strong defensive position, tripwire-force fatalities may be insufficient to provoke broader intervention. To deter, forward deployments must be sufficiently substantial to shift the local balance of power. Our claim is examined in three 20th-century deterrence attempts: the successful 1949 American attempt to deter a North Korean attack on South Korea; the unsuccessful 1950 American attempt to deter a North Korean attack on South Korea; and the unsuccessful 1914 British attempt to deter a German attack on Belgium.
Basing U.S. troops close to the front lines of an area where war is likely to break out has been a cornerstone of American grand strategy since World War II.1 From the demilitarized zone between North and South Korea to the plains of West Germany during the Cold War, U.S. troops have been placed directly in the path of expected assaults. The goal of forward-deploying these troops is simple: to deter aggression. Mere verbal promises to protect an ally can be dismissed as “cheap talk.”2 Making an effective threat to protect an ally requires non-verbal measures. After all, actions speak louder than words.
The United States has sometimes sought to achieve deterrence on the cheap through the deployment of smaller “tripwire” forces. While tripwire forces are too small to shift significantly the local balance of power, the idea is that such small deployments can still boost deterrence by increasing the likelihood of American intervention. According to this approach, the deaths of these troops in the early stages of fighting would compel a larger military response because, if America’s leaders were to do nothing in response to those deaths, they would be punished at home and the country would suffer damage to its international reputation. Thus, tripwire-force deployments boost deterrence by increasing credibility.
However, we argue that the deterrence benefits of tripwire-force deployments are exaggerated. These forces will not boost deterrence for two related reasons. First, tripwire deployments do not, in reality, significantly increase the credibility of the deterrent threat. The deaths of troops in combat will not necessarily create a strong public motive to intervene in order to seek revenge or protect a country’s reputation, even given the prospects of broader casualties. Second, even if the deaths of such troops could create these motives to intervene, potential attackers could strike anyway in order to achieve a fait accompli and create a strong defensive position.
In contrast to tripwire forces, the deployment of a larger contingent of troops can actually boost deterrence success. If forward-deployed forces are sufficient to shift the local balance of power, potential attackers will be less confident that they can succeed and less confident that they can achieve a fait accompli before reinforcements arrive. As a result, potential attackers will be less likely to actually launch an attack.
After presenting these general arguments below, we apply them to three empirical cases. We demonstrate that North Korea elected not to attack South Korea in 1949 because of the substantial deployment of U.S. troops to South Korean territory. On the other hand, the deployment of a small American tripwire force to South Korean territory in 1950 failed to deter a North Korean attack. Finally, we argue that, had Britain deployed a sufficient body of troops to Belgium before the July 1914 crisis, Germany would have been deterred from attacking Belgium and escalating the Austrian-Serbian crisis, permitting World War I to be avoided.
Understanding the logic and effectiveness of both tripwire and more substantial deployments is highly relevant to contemporary foreign policy debates. For example, during his presidency, Donald Trump consistently complained about the costs of deploying U.S. troops to allied territories and threatened to reduce American troop deployments abroad.3 Or, consider NATO’s Baltics intervention plan, discussed below, which some consider to be essentially a tripwire plan. In 2018, the United States deployed a small number of marines to Taiwan ostensibly to protect diplomatic personnel there. But observers noted that these troops could serve only as a tripwire in the event of a Chinese invasion.4 Strong faith that the deployment of a small tripwire force significantly bolsters deterrence could push President Joe Biden, who is eager to cut defense spending to both address domestic needs and decrease the deficit, to reduce U.S. troop deployments abroad, especially in South Korea, under the assumption that a smaller force would be sufficient to trigger America’s involvement in a conflict there and deter aggression.5
The U.S. government should stop assuming that undersized tripwire-force deployments have outsized deterrence effects. If a state is going to commit troops to deter aggression, it should do so with the understanding that the troops being deployed ought to be able to shift the local balance of power on their own. When it comes to forward troop deployments, more is better.
Tripwires and the Hope of Deterrence
In order to prevent attacks on other states, some states make deterrent threats against potential attackers. A core tenet of deterrence theory is that such threats are more likely to succeed if the potential attacker believes the defender has credibility — the willingness to execute the threat — and the capability — the ability to carry it out.6 Throughout the development of deterrence theory, scholars and policymakers have paid considerable attention to the credibility part of the deterrence formula, but less attention to the capability element.
Within the U.S. alliance network, tripwire forces have been seen as necessary for boosting deterrence, because alliance agreements alone are not viewed as sufficient to deter an attack. One reason is that alliance treaties frequently contain loopholes and conditions that enable states to find a legal justification not to intervene if an ally is at war, or to offer only very limited assistance.7 Indeed, the United States deliberately included a number of loopholes in its post-1945 treaties, such as the North Atlantic Treaty, including restricting the geographic purview of some treaties, giving signatories the ability to decide what level of intervention would be appropriate, and requiring intervention only if Congress formally declared war, among others. Politicians might be able talk their way out of the obligations of an alliance treaty, as France did when it abandoned Czechoslovakia in 1938, but, it was argued, abandoning Americans who had been killed in battle by hostile forces would be much more difficult.
Within the U.S. alliance network, tripwire forces have been seen as necessary for boosting deterrence, because alliance agreements alone are not viewed as sufficient to deter an attack.
Credibility was the central question of Thomas Schelling’s famous 1966 book, Arms and Influence. How could defenders like the United States credibly signal their willingness to go to war on behalf of their allies? Schelling suggested deploying tripwire forces to the territory of the potential target. A tripwire force is defined as a body of troops that is not large enough to shift the local balance of forces in order to stop or significantly slow down an attack. Schelling’s main claim is that the presence of such a force makes a threat to intervene more credible, because the early and unavoidable deaths of friendly troops in combat would force defender intervention. As Schelling famously wrote of the American tripwire force sent to Berlin,
The garrison in Berlin is as fine a collection of soldiers as has ever been assembled, but excruciatingly small. What can 7,000 American troops do, or 12,000 Allied troops? Bluntly, they can die. They can die heroically, dramatically, and in a manner that guarantees that the action cannot stop there. They represent the pride, the honor, and the reputation of the United States government and its armed forces; and they can apparently hold the entire Red Army at bay. Precisely because there is no graceful way out if we wished our troops to yield ground, and because West Berlin is too small an area in which to ignore small encroachments, West Berlin and its military forces constitute one of the most impregnable military outposts of modern times. The Soviets have not dared to cross that frontier.8
In making this claim, Schelling conflated pride, honor, and prestige to explain why setting off a tripwire would trigger an American response.9 America’s reputation would be damaged if it did not intervene following the deaths of these forces, because it would signal a willingness to allow an enemy to overrun its allies — and even its own troops. Schelling also alludes to the possibility of a response driven by emotion — that the deaths of U.S. troops would trigger a public demand for revenge and an insistence on the deployment of American troops.
Subsequent scholarly work fleshed out, both theoretically and empirically, the credibility element posited by Schelling. Some scholars explored the possibility of making a credible nuclear deterrent threat, i.e., whether the promise of nuclear retaliation for an attack on an ally would be perceived as credible.10 A critical component of this debate was whether long-range nuclear delivery was adequate for creating a credible deterrent threat, or if it was instead necessary for the United States to deploy tactical nuclear weapons in Europe.11 The deployment of such weapons would both serve as a classic tripwire, in that their destruction would compel an escalatory response and create a “use-it-or-lose-it” logic — their quick destruction would mean that the weapons would be kept on a hair trigger. Some scholars maintained that a general nuclear deterrent still required conventional tripwire forces. As Glenn Snyder observed in 1961,
the point of having a trip-wire is not to show that an attack has taken place — which of course would be evident as soon as the Soviet forces crossed the border — but to implicate extra-Continental nuclear powers by means of something stronger than a treaty contract.12
Others used game theoretic and experimental work to explore whether a threat’s credibility could be enhanced by building “audience costs,” whereby a leader is punished by an audience, typically domestic, that is displeased with the leader’s failure to fulfill a stated foreign policy promise or threat.13 Relatedly, scholars developed the concepts of “tied hands” and “sunk costs.”14 A government ties its own hands by raising the costs of failing to follow through on a threat. For example, it might create domestic “audience costs,” such as incurring domestic political consequences (i.e., being voted out of office) for failing to abide by a deterrent commitment. A government sinks costs by taking costly actions to demonstrate how seriously it takes the threat it is making — such as taking pricey mobilizing or arming measures.15
Still other research has connected the desire for revenge with a willingness to support the use of force. Indeed, one survey found that, in a hypothetical scenario, a desire for revenge was a key reason that subjects were willing to support a nuclear attack against Iranian civilians following a deadly Iranian attack on U.S. forces.16 Robert Harkavy points out that the desire for revenge goes beyond a need for “tit-for-tat” retaliation, as the latter is “devoid of much deep emotion.”17 Instead, there is a need for retribution by righting a wrong. Harkavy was considering the perceptions of societies that had been defeated during major war, but the logic of revenge has been found to apply in other contexts, including when a country has not been directly attacked.18
Flaws in Tripwire Logic
While the tripwire logic has been around for a long time and has compelling theoretical attributes, it has two interrelated flaws. First, potential attackers might not see the motives to avenge a tripwire force as increasing the motive to intervene. Second, the temptation of a fait accompli might encourage aggression even if reputation and revenge create motives for a defender to intervene.
Revenge
Regarding the first critique, it is incorrect to assume that all individuals (in or out of government) place a high value on reputation, honor, and revenge, especially in relation to other priorities and values.19 Indeed, a large segment of the population might have a distaste for the use of force and, hence, would punish a leader for even making a threat (but also reward that leader for backing away from the threat).20 Critics of the audience-cost proposition have observed that publics and governments may be willing to accept the reputational costs of failing to execute a deterrence threat if failing to execute undermines other important values, such as a desire to avoid entering difficult to win, costly wars over dubious stakes.21 Publics may put a higher value on consistency in policy, especially if consistency means staying out of conflict.22 Emphasis on reputation and revenge is not strong across all populations. Studies have found that the willingness to demand revenge and defend the national reputation and honor varies interpersonally and across subnational cultures.23
Further, leaders have the ability to shape the degree to which publics perceive harm to national honor for backing down from a threat. According to some surveys, presidential rhetoric can significantly reduce the audience costs that would otherwise be imposed for backing down.24 This is also the case for revenge. In her careful study of the connection between revenge and support for war, Rachel Stein notes that
it is not the case that a highly vengeful public will seize upon the slightest provocation to demand war from a reluctant leader. Rather, broad public endorsement of revenge is a latent source of support for war that leaders can activate by deploying strategically crafted rhetoric.25
Thus, motivated leaders can avoid war even after tripwire troops have been killed.
A leader’s rhetoric aside, the human and financial costs of intervention counterbalances the public desire to intervene for the sake of reputation or revenge. Desiring revenge, or to salvage one’s reputation, does not cause individuals to ignore the costs of war. There is an extensive body of scholarship on the sensitivity of democratic publics toward casualties, and specifically how this sensitivity affects the public willingness to support the use of force. Research on protracted wars has demonstrated persuasively that support for war declines as casualties mount.26
There is less work on how casualties affect the willingness of the American public to support intervening with force. Anecdotally, there are episodes of casualties failing to inspire demands for broader intervention. Casualties suffered by American peacekeeping troops in Beirut in 1983 and Mogadishu in 1993, for example, did not spark a public demand for broader intervention, and indeed, in both cases, American forces were eventually withdrawn without triggering a public backlash.27 Similarly, although several American vessels were attacked by German submarines in the Atlantic Ocean in 1941 before America’s entry into World War II, none of those attacks were sufficient to motivate the country’s entry into the war or retaliation against Germany. As late as October 1941, two American destroyers were attacked, killing 126 Americans, and still there was no public outrage and President Franklin Roosevelt did not order any retaliatory action. Meanwhile, naval recruitment dropped 15 percent, suggesting that concern over becoming a casualty at sea in clashes with German submarines exceeded demands for revenge.28
A leader’s rhetoric aside, the human and financial costs of intervention counterbalances the public desire to intervene for the sake of reputation or revenge.
A few surveys have looked for tripwire effects by asking participants whether they would be more likely to support intervention if friendly forces suffered serious casualties. The results have been mixed. Some surveys have explored this question in the context of audience-cost theory, asking participants whether, if a leader were to make a threat and then back down, that leader would earn higher levels of disapproval if friendly forces had been killed. Michael Tomz’s research presented participants with scenarios in which a country attacked a neighbor and the U.S. president threatened to push out the invaders but then did not take that action. He found that subject disapproval of the president was higher when they were told that 20 Americans had been killed in the initial fighting.29 Similarly, Graeme Davies and Robert Johns found that, when British subjects read about a scenario in which a terrorist group captured British soldiers and the prime minister threatened to take action but did not, their disapproval was higher when they were told of the deaths of four British combat pilots.30 Only one survey experiment has more directly explored for tripwire effects in public opinion by asking respondents to consider scenarios where troops were forward deployed. In their series of surveys on an array of different conflict scenarios, Paul Musgrave and Steven Ward found tripwire effects to be contextually restricted, substantively limited, or non-existent.31
What about international audience costs? Might the prospect of damaging their countries’ international reputation push governments to rescue embattled friendly troops? Some of the reasons why the prospect of domestic audience costs fails to deter are also reasons why the prospects of international audience costs fail to deter. If aggressors believe that a defender’s public is highly sensitive to casualties and wishes to avoid them, then the aggressor may come to believe that the defender will not approve broader intervention even at the risk of damaging its international reputation.32 Aggressors may also see hesitant defender governments as motivated to use treaty loopholes and framing to reduce the international audience costs of deciding against broader intervention.
The Fait Accompli
Regarding the second critique, even if aggressors sometimes recognize that killing an ally’s troops will motivate that ally to intervene to protect its reputation or enact revenge, the prospect of a fait accompli may tempt them to attack anyway.33 A fait accompli is the rapid conquest of territory and defeat of tripwire forces before reinforcements can arrive. Current scholarship proposes that attackers are tempted to launch a fait accompli out of a belief that a defender will accept the aggression, i.e., that the defender will accept a small loss of territory rather than invoke a broader war.34 In some sense, a fait accompli becomes what Schelling called a “salami tactic” if it enables the attacker to gain some territory without inducing a response by the defender.35 We propose an additional motivation: An aggressor might be tempted to launch a fait accompli if it believes that achieving the fait accompli will strengthen its defensive position, thereby further discouraging outside intervention by shifting the balance of power. A successful fait accompli might improve the attacker’s defensive position by providing geographic advantages were it to, say, capture an island or a mountain range. This would permit the attacker to construct stronger defensive positions or use the economic resources of the captured territory to boost its military power.36 Strengthening the defensive position through any of these means raises the defender’s perceived cost of intervening, thereby dissuading it from taking action. Because democracies are very unwilling to intervene in wars with low chances of victory,37 they might become discouraged and accept the conquest, albeit begrudgingly, if faced with the prospect of a long, difficult road to victory. Certainly, autocratic aggressors, such as Japan in 1941 and Argentina in 1982, hoped that their faits accomplis would discourage broader intervention from democratic defenders. Of course, the ability of a fait accompli to shift the military balance varies. The Soviet capture of West Berlin, for example, might not have provided communist forces either local geographical advantages or yielded economic resources that could have been converted to communist military power.
In short, there are good reasons to doubt that tripwire forces significantly boost deterrence. That said, believers in a tripwire effect inevitably point to the most well-known, apparently successful example of tripwire deterrence: the deployment of American and other NATO troops to West Berlin during the Cold War. During the Cold War, Western powers deployed as many as 12,000 troops in the city. This was not nearly enough to defend a city of 900,000, and yet peace obtained. Schelling and others inferred that, because the deaths of those troops in combat would have invoked immediate American intervention, the Soviets decided not to attack.
In the decades since Schelling wrote Arms and Influence, archives have been opened that cast doubt on his interpretation. They indicated that Stalin did not intend to invade West Berlin in 1948, nor did Khrushchev in the 1950s or early 1960s, meaning that tripwire forces did not deter an attack because there was no planned attack to deter.38 Internal discussions within the East German and Soviet governments revolved around sustaining the East German economy and state, stemming the flow of migrants to the West, persuading the West to recognize East Germany, and the possibility of closing access to West Berlin. An actual invasion, including a fait accompli effort against West Berlin, was not discussed, even as a contingency. Further, his public threats and blusterous rhetoric notwithstanding, Khrushchev preferred to avoid actual confrontation with the West over Berlin. In June 1959, Khrushchev remarked privately, “[We] don’t think it’s worth it now to push the west to the wall, so that we will not give the impression that we are seeking the recognition of the GDR [German Democratic Republic]. The Americans don’t want to recognize the GDR.”39
By 1961, as the refugee flow created an accelerating internal crisis for the East German government, there was not even passing mention of invading West Berlin as a solution. Rather, the more belligerent option — which was eventually chosen — was to close the border between East and West Germany. Internal Warsaw Pact discussions of military matters in the summer of 1961 surrounding this decision focused on what the Western reaction would be. The debate was over whether communist forces could deter NATO, not over whether they would invade West Berlin in light of American tripwire forces.40
Flawed assumptions about tripwire deterrence underlie current NATO defense planning. Consider NATO’s strategy toward the Baltic states.41 It presently deploys a battalion in each of the three Baltic states, each of which is “backstopped” by additional forces from Germany, the Czech Republic, and Norway, along with a brigade-sized U.S. force, which is stationed in Poland.42 Some view these forces not as directed toward stopping a Russian invasion of a Baltic NATO member, but instead as a tripwire to provoke American involvement.43 Michael O’Hanlon and Christopher Skaluba recently pointed out that these forces, dubbed NATO’s “enhanced forward deployment,” have not been “a truly integrated combat force; nor has [NATO] deployed many helicopters and air defense systems.”44 An Atlantic Council report described NATO’s current strategy as relying on a “tripwire” and “rapid reinforcement,” warning that this approach does not adequately address Russia’s “time, space, and mass advantages,” affording Russia the ability to achieve a successful fait accompli.45 In other words, NATO’s forward-deployed forces are simply not adequate in size, appropriately placed, or properly equipped to counter a Russian invasion. A 2016 RAND Corporation war game found that Russian forces could reach the outskirts of Riga and Tallinn in approximately 60 hours.46 Given the rapidity of the NATO force’s defeat in the war game, the ability of Russian forces to achieve a fair accompli would leave NATO decision-makers with few feasible options. Would they be willing to initiate an all-out ground war, with the possibility of nuclear escalation? One must therefore seriously question the ability of these forces to “trip the alarm” for a sizeable counter-strike by NATO and the United States.
Deployments Beyond Tripwires
Tripwire-force deployments are insufficient to bolster deterrence because deterrence is about more than signaling. One must instead deploy a force that is sufficiently capable — one that can itself fulfill a deterrent threat. Troop deployments must be able to shift the local balance of power so as to alter the potential attacker’s likelihood of succeeding, or its likelihood of succeeding quickly and/or at a low cost.47
There are three related ways in which forward-deployed troops can help fulfill a deterrent threat. First, they can give local forces a better chance of defeating the attacker’s forces through some combination of increasing troop quantity, improving military technology, and crafting an appropriate military strategy.48 Even if the deployed forces were relatively small in number (but still larger than a simple tripwire force), they could be equipped with force-multiplying weaponry, such as advanced armor.49 Foreign-deployed troops can also shift the local balance of power if equipped with the optimal military strategy. This was John Mearsheimer’s central policy conclusion in his classic 1983 book, Conventional Deterrence: that pre-deployed American troops armed with the right military strategy, and perhaps the right technology, could deter the Soviets from invading West Germany.50 One element of military strategy is integrating with coalition partners. Alliance structures that plan for the effective, combined use of allied forces can serve as an especially powerful deterrent.51 A sufficiently sized troop deployment could also defeat aggression from non-state actors.52 Certainly, the number of deployed troops needed to shift the balance of power varies by context. Under some circumstances, a smaller contingent of troops could be adequate if troops in the region could be rapidly deployed before an aggressor could achieve a fait accompli. Offshore naval power could also play a role, permitting a smaller deployed force to accomplish its deterrence mission.
Tripwire-force deployments are insufficient to bolster deterrence because deterrence is about more than signaling. One must instead deploy a force that is sufficiently capable — one that can itself fulfill a deterrent threat.
Second, even if deployed troops cannot fully blunt an attack, they can slow the attacking forces down. The prospect of a slowed advance can boost deterrence because attackers often seek rapid victory. Slowing an advance can buy time for the defender to deploy larger numbers of troops to the allied territory before it is conquered. If an attacker can be convinced that quick victory is unlikely, deterrence is more likely to hold.53 Deployed troops that are focused on buying time can play a key role in enhancing the fighting capacity of the allies if they are substantial enough to allow the allies to execute a war plan, including permitting the timely mobilization of reinforcements to aid embattled, pre-deployed forces. A small tripwire force is also unlikely to buy much time, something that is not commonly the purpose of such a deployment. Again, the purpose of a tripwire force is to, in the words of Schelling, “die heroically,” not hold the line until help arrives.
Third, deployed troops can preserve peace by making attempts at faits accomplis less attractive. As discussed, the lure of a fait accompli is that, if an attacker can rapidly conquer a territorial objective without engaging in military combat, it may be able to establish a strong defensive position, which would dissuade a counterattack or broader intervention. A substantial forward troop deployment can undermine a fait accompli strategy by reducing the likelihood that the attacker will capture the desired territory and strengthen its defensive position before a response can occur.
In summary, substantial deployments that shift the local balance of power are significantly more likely to deter aggression than tripwire forces are. Some might propose that more substantial troop deployments could boost credibility as well as capability, because such deployments might constitute a sunk cost signal of the defender’s willingness to intervene. It is possible that larger troop deployments convey such a sunk cost signal, though we do not yet have historical, quantitative, or survey research to demonstrate such an effect. Regardless, a key, underappreciated cause of deterrence failure is not an aggressor’s assessment of credibility, but rather its assessment of capability. As we demonstrate in our case studies below, aggressors attempt faits accomplis because they believe they can overwhelm defending forces and capture their territorial objective rapidly, emphasizing the importance capability plays in helping deterrence work.
That said, there are three limits to our argument that large contingents of deployed troops bolster deterrence through improving the capability of the allied state. First, an attacker that is less focused on winning quickly might not be deterred by an increased likelihood of delay. Non-democracies, for example, might be less motivated to win quickly as compared with democracies, because non-democratic leaders are less fearful of the domestic political consequences of a lengthy war.54 North Vietnam was not deterred by the presence of U.S. troops in South Vietnam, in part, because it was willing to wage a long and costly war.55 It is also conceivable that the defender might not have the motivation or ability to make use of a delay to provide more support to the ally. In short, the importance of delay may vary by context.
Second, deployed troops only help if they shift the attacker’s calculus of victory. That is, the mere presence of such troops is not sufficient. The troops need to be the right size, have the appropriate armament, be deployed to the right location, and have the right strategy to change the attacker’s calculations.56 Consider the German attack on the West in 1940, and Belgium in particular. British troops were not deployed to Belgium in peacetime. If they had been, they likely would not have been sent to the area where Germany actually attacked — the Ardennes — because Allied intelligence did not expect the main German attack to take place there, and hence British peacetime deployments might not have deterred a German attack.57
Third, if the ally is sufficiently well armed, then the peacetime deployment of the defender’s troops would not be necessary to strengthen an already robust deterrent. For example, Taiwan currently has sufficient conventional forces to repel a Chinese amphibious invasion, meaning that American force deployments there would likely add little to an already strong capability.58 Similarly, the Israeli Defense Forces are currently sufficiently strong to defeat any conventional ground attack from neighboring countries, making the peacetime deployment of American troops there unnecessary from a capabilities perspective. One policy implication of this is that a defender may be able to maintain a robust deterrent for an ally, not by deploying forces, but by offering substantial material aid to the ally, as the United States does for Taiwan and Israel.
History Rewritten? Three Critical Cases of Forward Troop Deployments
We test our theory that tripwire troop deployments will not significantly bolster deterrence, but that more substantial troop deployments will, by examining how troop deployments affected three attempts at deterrence in the 20th century: America’s attempt in 1949 to deter a North Korean attack on South Korea; America’s attempt in 1950 to deter a North Korean attack on South Korea; and Britain’s actions (and largely inactions) to deter a German attack on Belgium in 1914. First, North Korea did not attack South Korea in 1949, in part, because an adequately sized American troop deployment to South Korea shifted the local balance of power. Second, North Korea did attack South Korea in 1950, in part, because American troop deployments to South Korea had been reduced to a tripwire force that was insufficient to shift the local balance of power. Third, British troop deployments to Belgium in 1914 could have deterred a German attack if they had been sufficient to shift the local balance of power. However, a small tripwire force that did not shift the local balance of power would likely not have been sufficient to deter Germany. We chose these cases because they offer variation in the number of troops the defending power deployed (i.e., our independent variable), there are primary documents that provide insight into key decisions made in these cases, and the cases themselves are highly significant in world history and U.S. foreign policy.
Korea, 1949 & 1950
After World War II, the former Japanese colony of Korea was divided into two independent countries: communist North Korea and non-communist South Korea. The leaders of both countries, Kim Il-Sung of North Korea and Syngman Rhee of South Korea, envisioned invading the other Korea and unifying the peninsula under a single government. Kim knew he required Soviet assistance and support to carry out a successful campaign. South Korea became a member of the United Nations in December 1948, placing it under the purview of the U.N. Charter’s Chapter 1 commitment to remove threats and suppress acts of aggression.
In March 1949, Kim first contacted Soviet leader Joseph Stalin about possible Soviet support for an invasion. Stalin “did not object in principle” to Kim’s plan to attack South Korea.59 However, Stalin declined to approve and support such an attack.60 He reached this decision after an extended discussion with Kim about the local balance of power on the Korean Peninsula. The presence of the American forces stationed in South Korea — first deployed there in September 1945 — was of great importance to Stalin. When Kim asked permission to invade the South, Stalin’s first question was to ask how many American troops were stationed there. Kim informed Stalin that as many as 20,000 American forces were there, a very sizable contingent given that the entire South Korean army at that point numbered only 60,000, in Kim’s estimation.61 Stalin expressed concern with the local balance of forces and identified the American troop deployments in the South as one of the key reasons why he was declining to support a North Korean attack, recognizing that American troops would “interfere in case of hostilities.”62 In 1949, then, deterrence held on the Korean Peninsula.
Although Stalin declined to support a North Korean attack on the South in March 1949, Moscow continued to monitor the American troop presence in South Korea. In a secret May 1949 report to Moscow, Soviet intelligence reported that the American and South Korean governments were in the process of negotiating reductions in the American troop presence.63 In September 1949, Moscow asked Pyongyang directly if there were still American troops in South Korea, and how much assistance those troops could offer in the event of a North Korean invasion. North Korea estimated there were between 500 and 2,400 U.S. troops left in South Korea and that, in the event of war, “American instructors will take immediate part in organizing military actions.”64 Internally, American leaders recognized that this reduction in troops would change the U.S. troop presence in South Korea to a “trip-wire.”65
In early 1950, Kim again pressed Stalin to support an invasion. He understood that Stalin demanded rapid victory. In a visit to Moscow in early January, Kim stressed that his army could capture the South Korean capital of Seoul, located near the border, in a matter of days.66 Stalin began to relax his opposition in late January, though he maintained that an invasion must be planned such that “there would not be too great a risk.”67 Stalin saw political advantages to a Korean war: It would help Moscow maintain control of the communist world and prevent China from becoming a challenger for communist leadership.
Kim returned to Moscow in April, and, over the course of this visit, Stalin gave his approval for an invasion. Kim reiterated that the campaign would be over within days and Stalin was persuaded.68 He understood that now a North Korean fait accompli was possible, offering victory before America could send substantial reinforcements. As Stalin declared to Kim that month, “The war should be quick and speedy. Southerners and Americans should not have time to come to their senses. They won’t have time to put up a strong resistance and to mobilize international support.”69 Note Stalin’s mention of “strong” resistance, recognizing that North Korea would confront the weak American tripwire forces — a few hundred troops — that were then deployed to South Korea.
Stalin expressed concern with the local balance of forces and identified the American troop deployments in the South as one of the key reasons why he was declining to support a North Korean attack, recognizing that American troops would “interfere in case of hostilities.”
Kim sought to reassure Stalin that his plan for a fait accompli would avoid the risk of broader American intervention. Kim said that, once the attack had begun, there would be a mass uprising within South Korea against the Rhee government in support of the communist liberation: “Americans won’t have time to prepare and by the time they come to their senses, all the Korean people will be enthusiastically supporting the new government.”70 That is, the fait accompli would help shift the local balance of power against the United States because, after a North Korean victory, the South Korean population would rally to accept and support communist rule.
After Stalin approved the attack in April 1950, he remained focused on the urgency of North Korea achieving a rapid fait accompli. The June 15, 1950 operational plan called for an initial attack on the Ongjin Peninsula on the west coast of South Korea, just over the 38th parallel border.71 This would be followed by the capture of the capital of Seoul, also near the border, and the destruction of the South Korean military. Last, other major cities would be captured and the entirety of the country would be liberated. Stalin continued to stress the need for utmost secrecy, perhaps fearing that, if America sensed an impending invasion, it might send more troops to South Korea. Initial reports of U.S. casualties after the June 25 invasion, specifically of American bombers being shot down, did not disrupt Stalin’s plan of pressing for rapid victory to avoid broader American involvement. Upon hearing these reports, Stalin directed the Soviet ambassador to North Korea to convey to Kim that “the sooner South Korea is liberated the less chance there is for [American] intervention.”72
In summary, it appears Stalin became willing to approve the North Korean invasion of South Korea when he was confident the North could achieve a fait accompli. The substantial reduction of U.S. forces in South Korea, from a consequential detachment to a tripwire force, helped provide this confidence. The evidence does not indicate that the reduction in U.S. forces in South Korea was necessary and sufficient to cause a North Korean invasion, but it does indicate that the communists desired a rapid fait accompli victory; that they realized that the size of the U.S. troop deployments to South Korea affected the likelihood of that victory; and that, as of fall 1949, those troops had been reduced.
One alternative perspective is that the key factor pushing Stalin to greenlight the invasion was not the shifting local balance of power caused by a reduction in U.S. troops, but rather Stalin’s conclusion by January 1950 that America was unwilling to intervene, i.e., that Stalin had doubts about America’s credibility, rather than America’s capability. This view posits that Stalin may have been especially encouraged by Secretary of State Dean Acheson’s January 1950 speech that appeared to place South Korea outside of the American defense perimeter, as Acheson declared that the perimeter “runs along the Aleutians to Japan and then goes to the Ryukyus.”73
Two comments are in order. First, the effects of the 1950 Acheson speech on communist decision-making ought not be exaggerated. Other parts of the speech cast doubt on the idea that America would stand aside in the event of an invasion of South Korea, as Acheson specifically indicated that the United Nations would resist aggression anywhere in the Pacific region. Indeed, the media coverage of the speech implied that South Korea was inside the U.S. defense perimeter.74 Importantly, as noted, in the first half of 1950, after Acheson’s speech, Stalin continued to express concern about the possibility of broader American intervention, emphasizing the need for secrecy and the importance of rapid victory in the South. This was not so much a dismissal of America as an irrelevant paper tiger but rather indicates the importance of accomplishing a rapid fait accompli before broader intervention could occur. As one team of historians summarized it, by May 1950,
Stalin would have concluded from press reports and intelligence that, though the Americans might want to aid … South Korea, it would take many months to mass and get that aid to the western Pacific. The timing was on Kim’s side if he moved quickly and decisively.75
Second, new documentary evidence reveals that, in internal discussions, Soviet and Chinese officials did not take Acheson’s speech as a genuine statement of American disinterest in defending South Korea. Rather, they described Acheson’s speech as “slander,” an attempt to “deceive directly public opinion,” and a “smokescreen.”76 Based on the recent revelation of this and other internal documents, historian Kim Donggil has concluded, “The argument that Acheson’s speech about South Korea resting outside the U.S. defensive perimeter was the deciding factor in Stalin’s decision does not accord with the facts.”77 Another historian concurred, arguing that “recently released Soviet documents [demonstrate] that Acheson’s address had little if any impact on Communist deliberations.”78
World War I
In 1914, the German war strategy depended on rapidly conquering Paris. Germany’s two major adversaries in 1914 were France and Russia, which were allied with each other. Germany recognized that war against one would mean war against both, and war against two major powers simultaneously would be a risky endeavor. Germany’s solution to this problem was the Schlieffen Plan, initially devised in 1905 by then-Chief of the German General Staff Marshal Alfred von Schlieffen.79 Russia’s underdeveloped railroad network meant that the mobilization of its military would take weeks. If Germany could conquer France quickly, German forces could then rapidly redeploy to the east to confront the Russian colossus just as it was becoming ready for war. Moreover, the rapid German conquest of France and Belgium would dissuade Britain from entering a war against Germany. And control of the French coasts would facilitate the use of German submarines and air power against Britain. French and Belgian financial resources could be used to bolster German naval power, eventually permitting a successful German naval blockade of Britain, and thereby compelling Britain to accept German continental hegemony.80 This is the military attraction of the fait accompli. Once achieved, it shifts the balance of power in the attacker’s favor, dissuading the defender from continuing the war.
Critically, the rapid conquest of France required the invasion of neutral Belgium, as the Franco-German frontier was relatively mountainous, making any attack there unattractive.81 The importance of conquering Belgium and France rapidly cannot be overstated. Kaiser Wilhelm recognized this at least as early as 1905.82 When Britain seemed, in early August 1914, to suggest it would remain neutral if Germany did not attack Belgium, Germany rejected the possibility because staying out of Belgium would fatally slow the pace of its offensive. Chief of the German General Staff Helmuth von Moltke commented on August 4,
An attack from German territory [directly to France] would have cost the German army 3 months and would have ensured Russia such a head start that we could no longer have reckoned on a success on both fronts. We had to go via Belgium with all our might on Paris for a quick reckoning with France. This was the only way to victory.83
Britain did not pre-deploy any troops, not even tripwire forces, to Belgian territory in 1914. If it had pre-deployed troops in sufficient numbers to shift the local balance of power, a possibility made politically difficult by Belgium’s official neutrality,84 Germany would have been deterred and World War I might well have been avoided. German forces would have had to face British forces from the outset of hostilities, as opposed to several days or weeks after the war’s outbreak. Schlieffen assumed that, absent the pre-deployment of troops, British forces would not be able to land soon after the outbreak of war:
If in a Franco-German war the English plan to land a force of 100,000 men or more in Antwerp, they can hardly do so in the first days of mobilization. No matter how well they prepare the assembly of their three army corps, their army organization and defence system present so many difficulties that their sudden appearance within the great Belgian fortress is almost inconceivable.85
Moltke agreed.86 Schlieffen was confident that British forces landing after the start of hostilities would not slow down the German push through Belgium into France.87 Schlieffen also scoffed at the prospect of a blockade by the Royal Navy, confident that a short ground war would end the conflict.88
Schlieffen recognized that the prewar deployment of British troops to Belgium would spoil the chances of a rapid conquest, imperiling the entire plan. He wrote in 1912,
A successful march through Belgium on both sides of the Meuse is therefore the prerequisite of a victory. It will succeed beyond doubt, if it is only the Belgian army which tries to obstruct it. But it will be very difficult if the English army, and perhaps even part of the French, is present.89
Maj. Gen. Wilhelm von Hahnke commented in 1911, “Belgian neutrality must be broken by one side or the other. Whoever gets there first, occupies Brussels and imposes a war-levy of some thousand millions, has the upper hand.”90 Moltke was of the same opinion: “The enterprise is only possible if the attack is made at once, before the areas between the forts are fortified.”91 The German military and political leadership, including the Kaiser, Chancellor Theobald von Bethmann-Hollweg, Adm. Alfred von Tirpitz, and Moltke, all agreed with Schlieffen that the key questions were not only whether Britain would intervene, but also when British troops might arrive.92
The focus on seizing Belgium quickly did not shift when Moltke succeeded Schlieffen in planning Germany’s military strategy. However, Moltke did alter another area of Germany’s strategy, which heightened the need for quick success. Moltke decided to respect Dutch neutrality, making it crucial to seize the Belgian fortress city of Liรจge at the very outset of the conflict, a task much more easily accomplished if a substantial peacetime garrison were not present.93 In the words of one historian, “the quick seizure of Liรจge … was … regarded as an indispensable precondition for the successful execution of the great wheeling movement into northeastern France.”94 The city of Liรจge was surrounded by a system of 12 forts, meaning the German advance would have to take these forts quickly to succeed.
The Belgian military was, by itself, unable to defend the Liรจge fortresses, both because of insufficient numbers (Belgian troops were dispersed throughout the country) and because Belgian troops themselves were inadequately trained and equipped.95 The Belgian chief of staff in 1913 estimated that 60,000 troops would be needed to defend Liรจge, but, when German troops attacked in August 1914, there were only 25,000 Belgian defenders present.96 The Germans contemptuously dismissed Belgian troops as “chocolate soldiers.”97 Moltke declared, “We can count on the somewhat inefficient Belgian forces being quickly scattered.”98 As war broke out and German forces advanced on Liรจge, the Belgian leadership hoped for the rapid arrival of Anglo-French forces to bolster its defenses there.99 French forces were not sent, and British forces arrived too late to help. Defended only by Belgian troops, Liรจge fell to Germany on August 17, only two days later than the German war plan had envisioned.100
British defense planners recognized that the Belgians by themselves might not be able to hold on to the Meuse forts, including at Liรจge, and that, if the forts fell to Germany, British troops could not arrive in time to help save Belgium.
Strikingly, British military planners agreed with Schlieffen. As early as 1906, they recognized that Germany would gain military advantage by going through Belgium to invade France and that the speed of substantial British troop deployments would be essential to thwarting German plans. The British regretted that the projected transport schedule of British troops would not allow for those troops to arrive in time. British officers understood that “Time is of the greatest importance.”101 They understood that the Germans were focused on making “every possible effort to press on and to try to beat us” in reaching Belgian territory.102 Significantly reducing the time by which British troops could arrive in Belgium to confront the German invaders would provide “immense gain.”103 British defense planners recognized that the Belgians by themselves might not be able to hold on to the Meuse forts, including at Liรจge, and that, if the forts fell to Germany, British troops could not arrive in time to help save Belgium.104 A high-level 1911 British report reiterated that Germany might launch a surprise attack on Liรจge and that Belgian troops would be inadequate to defend it alone.105 The British also recognized the low quality of the Belgian troops.106 As late as May 1914, the British military noted that the critical Belgian fortresses, “if defended only by its peacetime [Belgian] garrison, would fall an easy prey to a German coup de main, executed at the outset of hostilities.”107
One piece of evidence that substantial peacetime British troop deployments — meaning larger than a token tripwire force — could have shifted the local balance of power decisively enough to alter German plans in summer 1914 was discussion among German military planners of expected increases in Belgian forces. Belgium began to consider improving its military following the 1911 Agadir military standoff between French and German forces and, in 1913, passed a bill calling for the substantial expansion of the army, from 180,000 to 250,000 troops. While German military planners understood that an improvement in Belgian defenses would undermine the viability of the Schlieffen Plan, they reassured themselves that implementation would be slow. Indeed, though Russia’s growing power over time has traditionally been seen as an incentive for Germany to launch a preventive war, growth in Belgian power over time created similar incentives. This was not because Belgium threatened Germany, but because additional Belgian forces render impossible the quick advances the Schlieffen Plan required.108 Germany’s focus on Belgian military power supports the claim that substantial British troop deployments would have eliminated Germany’s incentives to launch a preventive war in the West, because the local balance of power would have been immediately shifted against Germany, rather than more slowly through the process of Belgian military reforms.
In short, the weakness of Belgian forces, coupled with the absence of adequate British forces in the summer of 1914, fatally weakened prospects for deterring Germany from attacking Belgium. Even if British troop deployments had not been large enough to completely stop the German advance, as long as they were sufficient to slow down German troops this would have disrupted the tight schedule necessary for implementing the Schlieffen Plan. Germany’s decision to allow Austria to attack Serbia was predicated on the assumption that, if the conflict escalated, Britain would not disrupt the rapid conquest of Belgium. If Germany had not been confident in British military irrelevance, it would have restrained Austria, avoiding a broader war.109
A counterargument to our claim is that a small British tripwire force, rather than a large contingent, could have deterred Germany, given that German war plans required Britain remaining neutral. If British troops deployed in Belgium were killed in the first days of the war, British intervention would have been inevitable, even if those troops had not been able to slow or stop the German march through Belgium. But we maintain, based on the views of the German military with regard to British neutrality, that deployment of a smaller tripwire British contingent insufficient to shift the local balance of power would not have deterred Germany.
To understand how German officials viewed British neutrality, one should focus on the views of German military leaders rather than German civilian leaders.110 German military leaders were not concerned about Britain maintaining neutrality. The German military recognized that the operational necessity of invading Belgium made eventual British intervention inevitable. Moreover, the German military leaders would not have been deterred even if they fully believed that Britain was going to intervene, as long as that intervention was not immediate.111 Indeed, in prewar planning discussions, German military leadership appeared unconcerned with Britain remaining neutral. They assumed that British ground forces would eventually arrive, but too late to prevent the German conquest of Belgium and France.112 Historian Annika Mombauer put it starkly, “That Britain would not remain neutral was a mere hiccough for the [German] General Staff.”113 Some German generals even welcomed British intervention, craving the opportunity to defeat British forces in battle, likely assuming that British intervention would be too little, too late to save Belgium or France.114
Moltke’s invasion plan of Belgium actually presumed the eventual presence of British troops on Belgian soil.115 More broadly, he assumed that British intervention in a European war would be inevitable “because England considers Germany stronger than France, is afraid of German hegemony and wants to preserve the balance in Europe.”116 To this end, Moltke wanted to respect Dutch neutrality, in order to protect the German flank in anticipation of British intervention following a German invasion of Belgium.117 On August 4, Moltke declared that seizing Belgium was sufficiently important that it must be done, even if this meant British intervention.118 Moltke also argued that Germany ought to ignore any British commitment to stand aside if Germany respected Belgian neutrality, because such a commitment would not be credible given the threat to the European balance of power if Germany defeated France.119
Thus, British troop deployments before July 1914 (or earlier) would have been unlikely to persuade the German military to avoid war if their only function was to increase the likelihood of British intervention. However, pre-deployed British forces might have persuaded the German military to avert war, by shifting the local balance of forces against Germany and slowing the German advance. In short, the peacetime dispatch to Belgium of a moderate British troop contingent might have prevented what became, at the time, the deadliest war in the history of humanity.
Implications for U.S. Foreign Policy
As a policy tool, forward troop deployments are misunderstood in two ways. First, small tripwire deployments do not necessarily have powerful credibility effects, as is often thought. That is because it is not certain that the prospect of a small number of American troop fatalities would automatically engender broader American military intervention. Second, the capability effects of larger troop deployments are underappreciated. If forward troop deployments can substantially shift the local balance of power, even if only by slowing down an attacker’s advance, they can complicate an attacker’s plans and reduce the likelihood of aggression.
The limited utility of tripwire troop deployments should push decision-makers to reject tripwire deployments and instead consider two other options. First, if decision-makers are unwilling to bear the costs of more substantial troop deployments in order to deter attacks on allies, then they should consider reducing foreign commitments to conserve resources. Indeed, advocates of strategic restraint have directly suggested that the United States should eschew tripwire deployments as costly and ineffective.120
Second, if decision-makers are motivated to deter aggression abroad, they should recognize that successful deterrence requires more substantial foreign troop deployments.
This would have important implications for a variety of potential global hotspots. Consider the Russian threat to the Baltic states and Poland, all NATO members. The general concern among NATO commanders is that Russia could overrun tripwire forces and accomplish a quick fait accompli land grab. It could then consolidate its seizure through propaganda, a ceasefire agreement recognizing its territorial capture, and a veiled nuclear threat to deter a NATO counterattack.121 More substantial troop deployments to shift the local balance of power would reduce this risk. At the 2016 Warsaw Summit, NATO agreed to expand its troop commitment to these states, building four multilateral battle groups, each of which contains about 1,200 troops. However, as noted above, some viewed these contingents as inadequate to shift the balance of power, arguing that their aim was to achieve “deterrence by trip wire.”122 Even with these troop commitments, Russia might be tempted to launch a fait accompli, capturing territory before NATO could amass a more sizable force.123 There is also concern about a possible Russian fait accompli attack against the Norwegian Arctic island of Svalbard. Deploying a substantial numbers of troops there, rather than a small tripwire force, could discourage such an attack.124
[I]f decision-makers are unwilling to bear the costs of more substantial troop deployments in order to deter attacks on allies, then they should consider reducing foreign commitments to conserve resources.
Deploying a more significant number of troops could also help deter Chinese aggression. Andrew Krepinevich has proposed an integrated defense of the Japanese-Filipino-Taiwanese archipelago, a plan that includes the pre-deployment of U.S. troops to the region to alter Chinese military calculations of the local balance of power.125 Specifically, pre-deployment and peacetime training exercises involving Japan and the United States on the Senkaku/Diaoyou Islands in the East China Sea could discourage China from launching a fait accompli attack there.126
Our argument also sheds light on U.S. troop deployments to South Korea. A key question is whether current troop deployments to South Korea — about 28,000 troops — really change the local balance of power. Some have argued that the only significant contribution to deterrence the current American troop deployment can make is through tripwire effects, making a broader intervention more likely.127 Our argument suggests that U.S. troops deployed to South Korea will boost deterrence but only if they appreciably change the local balance of power. If such a force is deemed by defense planners to be insufficiently sized or structured to change the balance of power, then increasing or altering U.S. forces deployed to South Korea would provide deterrence benefits.128
Our argument also applies to U.S.-Japanese defense planning. Some propose that the deployment of American troops to Japanese territory, such as Okinawa, can deter an attack through a tripwire effect.129 However, the mere presence of such troops does not necessarily deter an attacker. Deployments must be adequate to shift the local balance of power.
Last, our argument is germane to debates about peacekeeping. Some advocates of peacekeeping have made tripwire arguments that the presence of peacekeepers deters former insurgents from breaking the peace because the deaths of any peacekeeping troops in combat would ensure more substantial international intervention.130 However, the early 1990s demonstrated that peacekeepers did not generate this credibility effect: Overpowered peacekeepers were captured in Bosnia and others were massacred in Rwanda, waiting in vain for rescue reinforcements that did not arrive. Peacekeepers act as a deterrent if they can shift the attacker’s calculus. This may require deployment in sufficient numbers or with the right equipment, such as armor.
We recognize that the pre-deployment of troops capable of shifting the local balance of power has clear limitations. Beyond the need to tailor any deployment to the local combat environment, increased troop deployments may raise financial costs that the public is unwilling to bear. Public opinion polls indicate Americans are hesitant about the prospect of raising defense spending.131 In a fiscal environment in which the COVID-19 pandemic has placed pressure on, and prompted a reconsideration of, the U.S. defense budget,132 increasing resources for foreign troop deployments may not be feasible. Moreover, increased troop deployments could create a moral hazard problem. Specifically, an ally that perceives greater protection from the United States could, in turn, feel more at liberty to engage in provocative or even aggressive action, knowing that it can rely on America for protection if the situation escalates. Some have proposed that moral hazard risks can be reduced by including language in alliance treaties that requires American intervention only if the ally was attacked first. However, others have observed that such “defensive wars” language can be vague, meaning that, even if an ally engages in provocative action, it can still argue that it was attacked first, triggering the alliance commitment.133
The general policy implications of our paper are that policymakers should beware of assuming that small troop deployments can have outsized deterrence effects, and they should recognize that larger and more considered deployments may be necessary to provide real deterrence. Toward the end of the Obama administration, observers argued for just this: the need to move past Cold War faith in tripwire effects and instead devote more resources to the Army to boost troop deployments abroad: “Put simply, forward deployed soldiers and marines are more than just trip-wires and hostages. Allies do not have faith in American commitments because American troops might die; they have faith because American troops can kill and win.”134
No comments:
Post a Comment